DKA CLINICAL OBSTETRICS AND GYNECOLOGY

Volume 61, Number 4, 733–742

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.

The Placenta Accreta

Spectrum:
Downloaded from [Link] by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3dW0s7N5CLiYMlH4wgaY34Fr69rtLH3oWi4+h6VxXUqo= on 10/30/2018

Epidemiology
and Risk Factors
DANIELA A. CARUSI, MD, MSc
Harvard Medical School and Brigham and Women’s Hospital,
Boston, Massachusetts

Abstract: The placenta accreta spectrum has become Once treated as a highly rare condition,
an important contributor to severe maternal morbid- this diagnosis has received increasing at-
ity. The true incidence is difficult to ascertain, but
likely falls near 1/1000 deliveries. This number seems tention over the past decade, and has been
to have increased along with the rate of risk factors. linked to the rising cesarean delivery
These include placenta previa, previous cesarean rate.2–4 With 1 in 3 deliveries in the United
section, use of assisted reproductive technologies, States performed via cesarean section,
uterine surgeries, and advanced maternal age. With concern for abnormal placentation has
increased uterine conservation, previous retained pla-
centa or placenta accreta have become significant risk underscored a need to rethink labor and
factors. Understanding placenta accreta spectrum risk delivery management. However, other fac-
factors facilitates patient identification and safe deliv- tors warrant increasing attention as well,
ery planning. Patients considering elective uterine including advancing maternal age, fertility
procedures or delayed childbirth should consider the treatments, and gynecologic surgery. De-
impact on peripartum morbidity.
Key words: accrete, morbidly adherent placenta, risk fining the full clinical range of this con-
factors dition may impact preconception patient
management and reproductive decision
making. In addition, as efforts to reduce
hemorrhagic morbidity have shifted to
Introduction predelivery risk stratification, understand-
Placenta accreta is one of the most impor- ing the full range of accreta risk factors
tant causes of severe maternal morbidity should improve proactive management of
and mortality in the developed world.1,2 this severe condition.5

Correspondence: Daniela A. Carusi, MD, MSc, Depart-

ment of Obstetrics and Gynecology, Brigham and Wom-
en’s Hospital, Boston, MA. E-mail: dcarusi@bwh. Definition and Terminology
[Link] Descriptions of likely placenta accreta
The author declares that there is nothing to disclose. extend as far back as the 16th century,

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 61 / NUMBER 4 / DECEMBER 2018

[Link] | 733
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
734 Carusi

with the first large case series published in

1937.6 This focused on a clinical defini-
tion: abnormal adherence of the placenta
to the uterine wall. Complete or partial
absence of the decidua basalis, with cho-
rionic villi directly attached to the myo-
metrium, was identified as the pathologic
correlate. Although deep penetration or
perforation of the uterine wall was iden-
tified as “increta” or “percreta,” respec-
tively, this was not encountered in their
cohort of patients with predominantly
unscarred uteri. The condition can be FIGURE 1. The spectrum of clinical and
further subtyped as either “complete” or pathologic findings historically encompassed
“total”; “partial,” with adherence limited by the term “accreta.”
to ≥ 1 placental cotyledons; or “focal,”
with isolated areas of adherence within a
cotyledon.6,7 In light of this concern, some have re-
The medical literature continued to use quired pathologic confirmation to im-
the term “accreta” to describe both clinical prove specificity,13,14 though this risks
cases of abnormal placental adherence and omitting cases in which pathology is
pathologic findings of villi abutting the missed (as might happen with focal accre-
myometrium. The term has also been used tas) or unavailable (such as when no
to encompass all degrees of villous invasion hysterectomy is performed). Further-
(including “increta” and “percreta”) with a more, in some cases a placenta may meet
single term. More recently, efforts have strict pathologic criteria with no clinical
been made to provide more comprehensive morbidity, or even no recognized placen-
definitions. “Morbidly adherent placenta” tal adherence.15,16 Others have restricted
has been used to describe the clinical the clinical definition to cases of placenta
findings, particularly as pathologic correla- previa,17 though this will overlook many
tion is not always available (such as when cases of nonprevia accreta.
the uterus is conserved). Nevertheless, Figure 1 illustrates the range of clin-
“morbidity” has not been strictly defined, ical and pathologic meanings that have
and clinical criteria vary from study to been encompassed by the term “accreta.”
study. The definition has ranged from Debate will likely continue as to whether
curettage for retained portions of adherence without morbidity, or patho-
placenta,8 to “difficult” removal of the logic without clinical findings should be
placenta,8,9 to one which requires a placen- included as “accreta.” These cases
tal bed hemorrhage.3,10,11 A detailed clin- should not be entirely excluded; how-
ical grading system has been proposed for ever, as they likely represent a histologic
prospective research, though this has not and clinical continuum. The more recent
been adopted for reporting purposes, and term Placenta Accreta Spectrum (PAS)
likely cannot be applied on a large scale acknowledges this range.18 Until this
or with retrospective data collection.12 spectrum is better understood and the
The lack of specificity that comes with terminology becomes standardized, re-
some published definitions may lead ports of incidence and mortality need to
to overestimates of accreta incidence be considered in terms of their inclusion
and underestimation of morbidity and criteria. This review will use PAS to
mortality rates. broadly describe all entities illustrated

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Placenta Accreta Epidemiology and Risk Factors 735

TABLE 1. Studies Reporting Incidence of Placenta Accreta Spectrum Disorders, Summarized

According to Case Definitions
Reported Years
Criteria Used Incidence (s) Reported Notes
Reporting from International
Academic Centers and Teaching
Hospitals
All deliveries, diagnosis requires 1/192217-1/337219 1977-1996 1/2510 when either previa or
previa hysterectomy required14
All deliveries, clinical and 1/59913 2004-2009 —
pathologic diagnosis
All deliveries, clinical diagnosis 1/73120 2008-2011 —
only
All deliveries, clinical or 1/53311 1982-2002 1/111 when diagnosis included
pathologic diagnosis retained products of conception8
Singleton cesareans only 1/25021 1988-2011 Clinical diagnosis
Primary unlabored cesareans 1/3339 1999-2002 Clinical or pathologic diagnosis
only
National and state-level reporting
All deliveries, National discharge 1/69422-1/113623 2005-2010 Ireland; Canada
data
All deliveries, with diagnosis 1/216224-1/588225 2009-2012 Nordic countries (clinical criteria
confirmation used); Great Britain (clinical or
histologic)
Deliveries in primiparas only, 1/43526 2003-2012 New South Wales, Australia. Used
discharge data code for “morbidly adherent
placenta”

Adapted from Carusi.27 Adaptations are themselves works protected by copyright. So in order to publish this adaptation,
authorization must be obtained both from the owner of the copyright in the original work and from the owner of copyright in the
translation or adaptation.

in Figure 1, though specific definitions delivery.9 The high incidences reported

will be specified where relevant. in these studies (1/250 to 1/333) reflect
overselection of patients with placenta
previa or previous uterine surgery. This
Incidence shows the importance of reporting the
Table 1 illustrates the range of PAS patient population when citing a specific
incidences that have been reported based PAS incidence.
on various definitions and study popula- The above studies also reflect patients
tions used. Earlier studies reported lower managed in predominantly academic or
incidences of 1/2000 to 3000, but also teaching hospitals. This will overestimate
restricted the diagnosis to patients who the total incidence, as patients with sus-
had placenta previa.14,17,19 With more pected PAS are likely to be referred to
liberal clinical or histologic diagnoses, these tertiary practices. Larger population
the incidences have ranged from 1/500 to databases are needed to determine the
1/700,11,13,20 though were reported as high overall PAS incidence, though coding
as 1/111 when all retained products of for “accreta” or “morbidly adherent pla-
conception requiring curettage were in- centa” is not well standardized. Such
cluded in the definition.8 Others restricted reporting has been possible in Ireland
the patient pool to only those having a and Canada, with reported rates of
cesarean21 or an unlabored cesarean 1/694 and 1/1136, respectively.22,23 Both

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
736 Carusi

Great Britain and the Nordic Countries the national Canadian study reported a
used centralized reporting systems that case fatality rate of 1/819.22 Some have
incorporated specific clinical or patho- quoted a mortality rate as high as 7%2,4;
logic criteria, and reported lower inciden- however, it is important to note that this
ces of 1/2162 to 1/5882 with these more number comes from survey data (prone to
restrictive criteria.24,25 Recently the state reporting bias), and reflects placenta per-
of New South Wales in Australia reported creta only.29
a high incidence of 1/435 when using a
code for “morbidly adherent placenta.” It
is not possible to know whether this Risk Factors
reflects liberal use of the diagnostic code PAS has largely been attributed to abnor-
or a truly high rate of the condition, but mal or deficient areas of decidua, allow-
illustrates the difficulty in comparing ing chorionic villi to adhere to the
rates from one study to another. National underlying myometrium.2 Such damage
reporting in the United States has been may result from previous procedures or
restricted by lack of a specific “accreta” inflammation.6,15,30 Correspondingly,
code in the International Classification of most of the risk factors studied involve
Diseases (ICD-9 CM) coding system, but some degree of uterine trauma or scar-
may become available with the updated ring. The role of the hormonal environ-
ICD-10 system. ment or the trophoblast itself has been less
This variation in case definitions and well addressed, but offers ongoing oppor-
patient sampling makes it difficult to tunities for discovery. Table 2 lists risk
confirm a rising PAS incidence. Three factors that have been mentioned in the
studies have compared these numbers literature, categorized by strength of asso-
over time using consistent definitions. A ciation with PAS.
single Italian teaching hospital reported a
rise from 1/833 to 1/322 when comparing PLACENTA PREVIA
deliveries in the late 1970s to those in 2006 Placenta previa is the most important risk
to 2008.3 Similarly, Ireland’s reporting of factor for PAS, and at one time was
national discharge data showed a 34% rise considered necessary for the
from 1/1266 to 1/943 from 2005 to 2010.23 diagnosis.17,19 Insufficient decidua in the
A true increase is supported by changes in lower uterine segment and cervix may
demographics and risk factors, though predispose to placental adherence with
absolute rates remain difficult to confirm. low implantations. When restricting anal-
yses to patients with placenta previa, PAS
is reported at rates of 1/16 to 1/9
Mortality deliveries.10,14,17,19 When comparing
To report the mortality rate, the total PAS rates in patients with and without
number of PAS cases in a population must previa, odds ratios as high as 50 have been
be known. This is not available in the reported.24,25,31 However, it is well ac-
United States, for reasons described above, cepted that abnormal placentation can
but has been reported with other national occur in women with vaginal births or
databases. Using the Australian Maternal cesarean deliveries in the absence of pla-
Outcomes Surveillance System, a fatality centa previa, though the absolute rates of
rate of 1/143 cases was reported between the condition are far lower in these
2010 and 2012, though this did not include cases.6,15,32,33 Given the strong associa-
centers with <50 births per year.28 The tion between these 2 diagnoses, plans for
Nordic Collaborative study reported no PAS management should be available
deaths over 205 accreta cases,24 whereas with any previa delivery.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Placenta Accreta Epidemiology and Risk Factors 737

TABLE 2. Risk Factors for Placenta risk factors for PAS. One study found a
Accreta Based on Quality lower risk with interrupted than continu-
of Evidence ous hysterotomy suturing,38 whereas an-
Evidence from controlled studies other found an inverse relationship
Placenta previa between PAS development and interpreg-
Previous cesarean section(s) (particularly with a nancy interval after cesarean delivery.39
placenta previa)
Assisted reproductive technology/in vitro
Performing a cesarean on an unlabored
fertilization versus labored uterus may also lead to a
Previous history of placenta accreta (histologic) relative increase in risk, though confound-
Previous history of retained placenta ing by indication should be considered
Previous uterine artery embolization for with this association.40
postpartum hemorrhage
Previous operative uterine procedures
(myomectomy, operative hysteroscopy) MATERNAL AGE
Inconsistent evidence from controlled studies Advanced maternal age, usually defined
Maternal age ≥ 35 as 35 years or greater, has been implicated
Previous dilation and curettage of the uterus in PAS development.11,21,23,31 This rela-
Maternal smoking
Anecdotal evidence from case series and reports
tionship may be confounded by higher
Uterine synechiae or Asherman syndrome parity and previa risk,41 as well as a
Previous endometrial ablation higher probability of previous uterine
Previous uterine fibroid embolization procedures or fertility treatments, but also
Congenital uterine anomalies may represent an altered hormonal or
Previous uterine irradiation
Factors proposed with no supporting evidence
implantation environment. As with cesar-
Use of an intrauterine device ean section, this demographic has been
Adenomyosis changing significantly over time, with an
Current fibroids increasing percentage of US deliveries
occurring in women over 35 years of
Adapted from Carusi.27 Adaptations are themselves works
protected by copyright. So in order to publish this adapta- age.37 Its potential role in abnormal
tion, authorization must be obtained both from the owner of placentation should be considered when
the copyright in the original work and from the owner of
copyright in the translation or adaptation. advising women on reproductive plan-
ning.
PREVIOUS CESAREAN DELIVERY PREVIOUS UTERINE SURGERY
Given that cesarean deliveries are a com- Studies have produced inconsistent results
mon source of scarring in the myometrium regarding PAS development after pre-
and endometrium, this common procedure vious uterine surgery. This may be due
has been associated with the development of to grouping procedures with heteroge-
PAS in subsequent pregnancies.24,32 This is nous risk profiles,23,31 or lack of an
particularly likely when the patient develops appropriate control group with no pre-
a subsequent previa, with the placenta vious uterine procedures.42 Uterine cur-
implanted over the lower segment ettage is likely the most commonly
scar.10,14,34 Multiple studies have shown a encountered uterine procedure in a pa-
linear increase in PAS risk correlating with tient’s surgical history, but its role as an
the number of previous cesareans, both with independent PAS risk factor has been
and without placenta previa.9–11,14,19,24,35,36 unclear. Some have found the relation-
With United States cesarean delivery rates ship to hold only when evaluating multi-
at 32%, this risk factor has become an ple previous procedures,31 whereas others
important focus for risk reduction efforts.37 have found that the relationship is con-
Some studies have focused on modifi- founded by other patient factors.8,10 One
able aspects of cesarean management as study linked accreta development to a

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
738 Carusi

history of recurrent miscarriages, but de- previous history of endometrial ablation.

tails on surgical treatment were not Alternatively, a recent systematic review
given.43 found a 2% rate of abnormally adherent
A recent large study used a statewide placenta in controlled studies reporting
database to link multiple types of gyneco- pregnancies after ablation.44 Although a
logic procedures with later PAS develop- much higher rate (23%) was found among
ment. After restricting the analysis to case reports and case series of pregnancy
women with no previous delivery and after ablation, this likely reflects a large
controlling for age and maternal demo- degree of reporting bias.
graphics, they found a dose-dependent
relationship between the number of pre- PAST OBSTETRIC HISTORY
vious uterine procedures and develop- While previous cesarean section is likely
ment of abnormally invasive placenta. the most commonly encountered risk
These procedures included laparoscopic factor from a previous pregnancy, a pre-
uterine procedures, hysteroscopy, and vious history of accreta or adherent pla-
uterine curettage.26 Figure 2 shows an centa will confer the highest absolute risk.
example of placental adherence to a Previous PAS is a novel risk factor, as
previous hysteroscopic myomectomy scar cases of PAS historically ended in hyster-
in a nulliparous woman. Figure 2A shows ectomy. However, uterine conservation is
the focally abnormal area sonographi- increasingly described in management of
cally, whereas Figure 2B shows the deliv- this condition. For patients who have
ered placenta with adherent myometrium experienced a subsequent pregnancy,
still attached. PAS rates of 13% and 28% have been
With global damage of the endome- reported.21,45 Similarly, patients treated
trium, endometrial ablation would be for early cesarean scar implantations—
expected to show an association with believed to be early precursors of previa
PAS when subsequent pregnancies devel- and PAS46–48—have reported subsequent
op. However, the above study found no PAS rates of 10% to 28%, though the total
cases of PAS among 318 patients with a number reported is small.49,50

FIGURE 2. A placenta accreta densely adherent to uterine myometrium in a previous hys-

teroscopic myomectomy scar. The arrows show the adherent myometrium sonographically (A)
and on gross pathology (B).

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Placenta Accreta Epidemiology and Risk Factors 739

Others have linked PAS to a previous ART results from maternal factors rather
history of retained placenta with manual than the procedure itself. Such risk factors
extraction or retained products of include advanced maternal age and ute-
conception.6,25 Both conditions may in- rine factor infertility arising from pre-
dicate previous adherent placentation, or vious procedures or cavity distortion.
damaged decidua from the procedures Placenta previa may be an intermediate
required to treat them. Previous use of step between ART and PAS, as both
arterial embolization for postpartum ART and multiple gestations have been
hemorrhage has been linked to PAS in linked to low implantations.56,59,60 How-
multiple studies.51,52 Because this proce- ever, a 2015 study showed an independent
dure uses absorbable embolic particles, relationship between transfer of cryopre-
arterial occlusion is considered tempo- served embryos and PAS even after con-
rary, and future pregnancies are generally trolling for maternal factors and previa.57
permitted. It remains unclear whether the Development of a thin endometrium dur-
subsequent PAS risk is related to the ing uterine preparation has been linked to
temporary occlusion of blood flow, or if both previa and PAS development, and
the association is confounded by the cause provides a biologically plausible explan-
or surgical treatment of the initial post- ation for this association.57,61 Further
partum hemorrhage. study of this mechanism may provide an
As previously noted, histologic find- opportunity to reduce PAS risk during the
ings of accreta may be reported with or ART process.
without clinical suspicion of the condi-
tion. Case control studies have linked past OTHER UTERINE ABNORMALITIES
histologic findings of villi abutting myo- Any condition predisposing to an abnor-
metrium to later development of mor- mal decidual-myometrial interface may
bidly adherent placenta.53,54 In a theoretically contribute to PAS risk. This
subsequent cohort study, 36% of patients has been best shown with Asherman
with a previous reported histologic accre- syndrome, with which fibrotic scars ob-
ta had an adherent placenta noted at literate the normal endometrium and
subsequent delivery. However, major potentially obstruct the uterine cavity.
hemorrhage or hysterectomy occurred One case series showed a 9% rate of
only in those who experienced morbidity adherent placenta following treatment of
in the previous pregnancy.55 Asherman syndrome,62 whereas a larger
series found a 2% rate of accreta and 4%
ASSISTED REPRODUCTIVE rate of adherent placenta.63 Pregnancy
TECHNOLOGY (ART) AND outcome is likely related to the extent of
ENDOMETRIAL THICKNESS damage and success of the repair. Adeno-
ART, which includes in vitro fertilization myosis and current fibroids may distort
and intracytoplasmic sperm injection, was the normal endometrium, but have not
first linked to PAS development in 2011.13 been well studied with regard to adherent
Multiple subsequent studies have con- placentation.64 Cigarette smoking has
firmed this association, and a meta-anal- been linked to thinner endometrium with
ysis showed an overall odd ratio of 2.67.56 ART cycles,65 but has not been clearly
This outcome has been further linked to linked to PAS risk. Although 2 studies
specific ART practices, including transfer have shown a positive association be-
of > 1 embryo and use of cryopreserved tween smoking and PAS, a third did not
embryos.57,58 Given that known PAS risk confirm an association.10,23
factors also overlap with infertility, it is Case reports and series have linked PAS
possible that the PAS development with to the presence of Mullerian anomalies7,66,67

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
740 Carusi

or previous uterine fibroid embolization.68,69 maternal morbidity. Obstet Gynecol. 2014;123:

An additional report described PAS and 804–810.
uterine rupture after previous whole body 2. Silver RM. Abnormal placentation: placenta
previa, vasa previa, and placenta accreta. Obstet
irradiation.70 Either of these procedures Gynecol. 2015;126:654–668.
may theoretically damage the endometrium 3. Morlando M, Sarno L, Napolitano R, et al. Pla-
through ischemia or direct damage, respec- centa accreta: incidence and risk factors in an area
tively. Finally, while the use of an intra- with a particularly high rate of cesarean section.
uterine device has been posited as a source Acta Obstet Gynecol Scand. 2013;92:457–460.
4. Belfort MA. Placenta accreta. Am J Obstet
of endometrial damage and subsequent PAS Gynecol. 2010;203:430–439.
development, no research has linked these 5. Dilla AJ, Waters JH, Yazer MH. Clinical vali-
entities.64 At the current time, contraception dation of risk stratification criteria for peripartum
choices should be made without regard to hemorrhage. Obstet Gynecol. 2013;122:120–126.
later PAS risk. 6. Irving F, Hertig A. A study of placenta accreta.
Surg Gynecol Obstet. 1937;64:178–200.
7. Fox H. Placenta accreta, 1945-1969. Obstet Gynecol
Surv. 1972;27:475–490.
Conclusions 8. Gielchinsky Y, Rojansky N, Fasouliotis SJ, et al.
Understanding of accreta pathophysiology Placenta accreta—summary of 10 years: a survey
and epidemiology remains limited, but is of 310 cases. Placenta. 2002;23:210–214.
9. Silver RM, Landon MB, Rouse DJ, et al. Mater-
rapidly growing. As definitions become nal morbidity associated with multiple repeat
standardized, diagnosis and reporting may cesarean deliveries. Obstet Gynecol. 2006;107:
become more consistent. This is essential 1226–1232.
both for incidence reporting and under- 10. Usta IM, Hobeika EM, Musa AA, et al. Placenta
standing of risk factors. Although most previa-accreta: risk factors and complications.
Am J Obstet Gynecol. 2005;193:1045–1049.
work to date has focused on placenta 11. Wu S, Kocherginsky MHibbard JU. Abnormal
previa and uterine scarring, the roles of placentation: twenty-year analysis. Am J Obstet
genetics, local hormonal factors, and the Gynecol. 2005;192:1458–1461.
trophoblast itself have yet to be tapped. 12. Collins SL, Stevenson GN, Al-Khan A, et al.
Understanding the role of known risk Three-dimensional power doppler ultrasonography
for diagnosing abnormally invasive placenta and
factors in this morbid condition may quantifying the risk. Obstet Gynecol. 2015;126:
impact medical decision making. The 645–653.
risks and benefits of performing certain 13. Esh-Broder E, Ariel I, Abas-Bashir N, et al. Pla-
uterine procedures—including myomec- centa accreta is associated with IVF pregnancies:
tomies, curettages, and elective cesarean a retrospective chart review. BJOG. 2011;118:
1084–1089.
deliveries—should take future delivery 14. Miller DA, Chollet JA, Goodwin TM. Clinical
outcomes into account. Furthermore, as risk factors for placenta previa-placenta accreta.
ART has expanded options for delayed Am J Obstet Gynecol. 1997;177:210–214.
childbearing, maternal morbidity related 15. Jacques SM, Qureshi F, Trent VS, et al. Placenta
to abnormal placentation should be con- accreta: mild cases diagnosed by placental exami-
nation. Int J Gynecol Pathol. 1996;15:28–33.
sidered as a potential consequence. Fi- 16. Khong TY, Werger AC. Myometrial fibers in the
nally, appreciation of risk factors in the placental basal plate can confirm but do not
peripartum period may allow better rec- necessarily indicate clinical placenta accreta. Am
ognition and direct management of PAS, J Clin Pathol. 2001;116:703–708.
with the essential goal of minimizing 17. Zaki ZM, Bahar AM, Ali ME, et al. Risk factors
and morbidity in patients with placenta previa
perinatal morbidity. accreta compared to placenta previa non-accreta.
Acta Obstet Gynecol Scand. 1998;77:391–394.
18. Jauniaux E, Ayres-de-Campos D, Diagnosis
References FPA, et al. FIGO consensus guidelines on pla-
1. Grobman WA, Bailit JL, Rice MM, et al. Fre- centa accreta spectrum disorders: introduction.
quency of and factors associated with severe Int J Gynaecol Obstet. 2018;140:261–264.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Placenta Accreta Epidemiology and Risk Factors 741

19. Clark SL, Koonings PP, Phelan JP. Placenta 34. Marshall NE, Fu R, Guise JM. Impact of multiple
previa/accreta and prior cesarean section. Obstet cesarean deliveries on maternal morbidity: a sys-
Gynecol. 1985;66:89–92. tematic review. Am J Obstet Gynecol. 2011;205:
20. Bailit JL, Grobman WA, Rice MM, et al. Mor- 262.e261–262.e268.
bidly adherent placenta treatments and outcomes. 35. Daltveit AK, Tollanes MC, Pihlstrom H, et al.
Obstet Gynecol. 2015;125:683–689. Cesarean delivery and subsequent pregnancies.
21. Eshkoli T, Weintraub AY, Sergienko R, et al. Obstet Gynecol. 2008;111:1327–1334.
Placenta accreta: risk factors, perinatal outcomes, 36. Grobman WA, Gersnoviez R, Landon MB, et al.
and consequences for subsequent births. Am J Pregnancy outcomes for women with placenta
Obstet Gynecol. 2013;208:219.e211–219.e217. previa in relation to the number of prior cesarean
22. Mehrabadi A, Hutcheon JA, Liu S, et al. Con- deliveries. Obstet Gynecol. 2007;110:1249–1255.
tribution of placenta accreta to the incidence of 37. Hamilton BEPD, Martin JA, Osterman MMHS,
postpartum hemorrhage and severe postpartum et al. Births: preliminary data for 2014. Natl Vital
hemorrhage. Obstet Gynecol. 2015;125:814–821. Stat Rep. 2015;64:1–19.
23. Upson K, Silver RM, Greene R, et al. Placenta 38. Sumigama S, Sugiyama C, Kotani T, et al.
accreta and maternal morbidity in the Republic of Uterine sutures at prior caesarean section and
Ireland, 2005-2010. J Matern Fetal Neonatal placenta accreta in subsequent pregnancy: a case-
Med. 2014;27:24–29. control study. BJOG. 2014;121:866–874; discus-
24. Thurn L, Lindqvist PG, Jakobsson M, et al. Ab- sion 875.
normally invasive placenta-prevalence, risk factors 39. Wax JR, Seiler A, Horowitz S, et al. Interpreg-
and antenatal suspicion: results from a large pop- nancy interval as a risk factor for placenta
ulation-based pregnancy cohort study in the Nordic accreta. Conn Med. 2000;64:659–661.
countries. BJOG. 2016;123:1348–1355. 40. Kamara M, Henderson JJ, Doherty DA, et al.
25. Fitzpatrick KE, Sellers S, Spark P, et al. Incidence The risk of placenta accreta following primary
and risk factors for placenta accreta/increta/per- elective caesarean delivery: a case-control study.
creta in the UK: a national case-control study. BJOG. 2013;120:879–886.
PLoS One. 2012;7:e52893. 41. Gilliam M, Rosenberg D, Davis F. The likelihood
26. Baldwin HJ, Patterson JA, Nippita TA, et al. Ante- of placenta previa with greater number of cesar-
cedents of abnormally invasive placenta in primipar- ean deliveries and higher parity. Obstet Gynecol.
ous women: risk associated with gynecologic 2002;99:976–980.
procedures. Obstet Gynecol. 2018;131:227–233. 42. Gyamfi-Bannerman C, Gilbert S, Landon MB,
27. Carusi D. Placenta accreta: Epidemiology and et al. Risk of uterine rupture and placenta accreta
risk factors. In: Silver RM, ed. Placenta Accreta with prior uterine surgery outside of the lower
Syndrome. Boca Raton: CRC Press; 2017:1–12. segment. Obstet Gynecol. 2012;120:1332–1337.
28. Farquhar CM, Li Z, Lensen S, et al. Incidence, 43. Yang J, Wang Y, Wang XY, et al. Adverse
risk factors and perinatal outcomes for placenta pregnancy outcomes of patients with history of
accreta in Australia and New Zealand: a case- first-trimester recurrent spontaneous abortion.
control study. BMJ Open. 2017;7:e017713. Biomed Res Int. 2017;2017:4359424.
29. O’Brien JM, Barton JR, Donaldson ES. The 44. Kohn JR, Shamshirsaz AA, Popek E, et al.
management of placenta percreta: conservative Pregnancy after endometrial ablation: a system-
and operative strategies. Am J Obstet Gynecol. atic review. BJOG. 2018;125:43–53.
1996;175:1632–1638. 45. Sentilhes L, Kayem G, Ambroselli C, et al.
30. Jauniaux E, Jurkovic D. Placenta accreta: patho- Fertility and pregnancy outcomes following con-
genesis of a 20th century iatrogenic uterine dis- servative treatment for placenta accreta. Hum
ease. Placenta. 2012;33:244–251. Reprod. 2010;25:2803–2810.
31. Hung TH, Shau WY, Hsieh CC, et al. Risk factors 46. Michaels AY, Washburn EE, Pocius KD, et al.
for placenta accreta. Obstet Gynecol. 1999;93: Outcome of cesarean scar pregnancies diagnosed
545–550. sonographically in the first trimester. J Ultra-
32. Colmorn LB, Krebs L, Klungsoyr K, et al. Mode sound Med. 2015;34:595–599.
of first delivery and severe maternal complica- 47. Timor-Tritsch IE, Monteagudo A, Cali G, et al.
tions in the subsequent pregnancy. Acta Obstet Cesarean scar pregnancy is a precursor of mor-
Gynecol Scand. 2017;96:1053–1062. bidly adherent placenta. Ultrasound Obstet Gyne-
33. Greenbaum S, Wainstock T, Dukler D, et al. col. 2014;44:346–353.
Underlying mechanisms of retained placenta: 48. Zosmer N, Fuller J, Shaikh H, et al. Natural
evidence from a population based cohort study. history of early first-trimester pregnancies im-
Eur J Obstet Gynecol Reprod Biol. 2017;216: planted in Cesarean scars. Ultrasound Obstet
12–17. Gynecol. 2015;46:367–375.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
742 Carusi

49. Seow KM, Hwang JL, Tsai YL, et al. Subsequent 60. Smithers PR, Halliday J, Hale L, et al. High
pregnancy outcome after conservative treatment frequency of cesarean section, antepartum hem-
of a previous cesarean scar pregnancy. Acta orrhage, placenta previa, and preterm delivery in
Obstet Gynecol Scand. 2004;83:1167–1172. in-vitro fertilization twin pregnancies. Fertil Ster-
50. Washburn EE, Pocius K, Carusi D. Outcomes of il. 2003;80:666–668.
nonsurgical versus surgical treatment of cesarean 61. Rombauts L, Motteram C, Berkowitz E, et al.
scar pregnancies in the first trimester. Arch Risk of placenta praevia is linked to endometrial
Gynecol Obstet. 2017;296:533–541. thickness in a retrospective cohort study of 4537
51. Poggi SH, Yaeger A, Wahdan Y, et al. Outcome of singleton assisted reproduction technology births.
pregnancies after pelvic artery embolization for Hum Reprod. 2014;29:2787–2793.
postpartum hemorrhage: retrospective cohort study. 62. Friedman A, DeFazio J, DeCherney A. Severe
Am J Obstet Gynecol. 2015;213:576.e571–576.e575. obstetric complications after aggressive treatment
52. Soro MP, Denys A, de Rham M, et al. Short & of Asherman syndrome. Obstet Gynecol. 1986;67:
long term adverse outcomes after arterial embo- 864–867.
lisation for the treatment of postpartum haemor- 63. Chen L, Zhang H, Wang Q, et al. Reproductive
rhage: a systematic review. Eur Radiol. 2017;27: outcomes in patients with intrauterine adhesions
749–762. following hysteroscopic adhesiolysis: experience
53. Linn RL, Miller ES, Lim G, et al. Adherent basal from the largest women’s hospital in China.
plate myometrial fibers in the delivered placenta J Minim Invasive Gynecol. 2017;24:299–304.
as a risk factor for development of subsequent 64. Jauniaux E, Chantraine F, Silver RM, et al.
placenta accreta. Placenta. 2015;36:1419–1424. FIGO consensus guidelines on placenta accreta
54. Miller ES, Linn RL, Ernst LM. Does the presence spectrum disorders: Epidemiology. Int J Gynaecol
of placental basal plate myometrial fibres increase Obstet. 2018;140:265–273.
the risk of subsequent morbidly adherent placenta: 65. Weigert M, Hofstetter G, Kaipl D, et al. The
a case-control study. BJOG. 2016;123:2140–2145. effect of smoking on oocyte quality and hormonal
55. Roeca C, Little SE, Carusi DA. Pathologically parameters of patients undergoing in vitro fertil-
diagnosed placenta accreta and hemorrhagic ization-embryo transfer. J Assist Reprod Genet.
morbidity in a subsequent pregnancy. Obstet 1999;16:287–293.
Gynecol. 2017;129:321–326. 66. Henriet E, Roman H, Zanati J, et al. Pregnant
56. Karami M, Jenabi E, Fereidooni B. The associ- noncommunicating rudimentary uterine horn
ation of placenta previa and assisted reproductive with placenta percreta. JSLS. 2008;12:101–103.
techniques: a meta-analysis. J Matern Fetal Neo- 67. Oral B, Guney M, Ozsoy M, et al. Placenta
natal Med. 2018;31:1940–1947. accreta associated with a ruptured pregnant rudi-
57. Kaser DJ, Melamed A, Bormann CL, et al. mentary uterine horn. Case report and review of
Cryopreserved embryo transfer is an independent the literature. Arch Gynecol Obstet. 2001;265:
risk factor for placenta accreta. Fertil Steril. 100–102.
2015;103:1176.e1172–1184.e1172. 68. El-Miligy M, Gordon A, Houston G. Focal my-
58. Takeshima K, Jwa SC, Saito H, et al. Impact of ometrial defect and partial placenta accreta in a
single embryo transfer policy on perinatal out- pregnancy following bilateral uterine artery embo-
comes in fresh and frozen cycles-analysis of the lization. J Vasc Interv Radiol. 2007;18:789–791.
Japanese Assisted Reproduction Technology 69. Takahashi H, Hayashi S, Matsuoka K, et al.
registry between 2007 and 2012. Fertil Steril. Placenta accreta following uterine artery emboli-
2016;105:337.e333–346.e333. zation. Taiwan J Obstet Gynecol. 2010;49:
59. Qin J, Liu X, Sheng X, et al. Assisted reproductive 197–198.
technology and the risk of pregnancy-related 70. Norwitz ER, Stern HM, Grier H, et al. Placenta
complications and adverse pregnancy outcomes percreta and uterine rupture associated with prior
in singleton pregnancies: a meta-analysis of co- whole body radiation therapy. Obstet Gynecol.
hort studies. Fertil Steril. 2016;105:73.e71–85.e76. 2001;98:929–931.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA CLINICAL OBSTETRICS AND GYNECOLOGY
Volume 61, Number 4, 743–754
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.

Pathophysiology of
Placenta Accreta
Downloaded from [Link] by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3dW0s7N5CLiYMlH4wgaY34Fr69rtLH3oWi4+h6VxXUqo= on 10/30/2018

Spectrum Disorders:
A Review of Current
Findings
ERIC JAUNIAUX, MD, PhD, FRCOG,*
and GRAHAM J. BURTON, MD, DSC†
*UCL Institute for Women’s Health, University College London
(UCL), London, UK; and †Department of Physiology,
Development and Neuroscience, The Centre for Trophoblast
Research, University of Cambridge, Cambridge, UK

Abstract: Current findings continue to support the

concept of a biologically defective decidua rather than
Introduction
a primarily abnormally invasive trophoblast. Prior To understand the pathophysiology of
cesarean sections increase the risk of placenta previa accreta placentation it is essential to look
and both adherent and invasive placenta accreta, into the history of its epidemiology. The
suggesting that the endometrial/decidual defect fol- first case reports of placenta accreta (PA)
lowing the iatrogenic creation of a uterine myome-
trium scar has an adverse effect on early implantation. were published in the literature in the
Preferential attachment of the blastocyst to scar tissue 1920s, and the first series in 1937 by
facilitates abnormally deep invasion of trophoblastic the obstetrician Frederick C. Irving and
cells and interactions with the radial and arcuate the pathologist Arthur T. Hertig from the
arteries. Subsequent high velocity maternal arterial Boston Lying-In Hospital.1 In 1927,
inflow into the placenta creates large lacunae, destroy-
ing the normal cotyledonary arrangement of the villi. Dr Forster,2 a scholar in gynecology at
Key words: placenta accreta, placenta increta, placen- the Pathology Department of the Montreal
ta percreta, ultrasound diagnosis, histopathology General Hospital, Montreal, Canada de-
scribed a case of PA with invasive villi
Correspondence: Eric Jauniaux, MD, PhD, FRCOG, following a prior cesarean delivery (CD).
Academic Department of Obstetrics and Gynaecology, This was the only case in 8000 deliveries
Institute for Women’s Health, University College
London, 86-96 Chenies Mews, London, UK. during a 6-year period at the Montreal
E-mail: [Link]@[Link] maternity hospital, indicating the rarity
The authors declare that they have nothing to disclose. of the condition at that time. Modern

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 61 / NUMBER 4 / DECEMBER 2018

[Link] | 743
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
744 Jauniaux and Burton

cesarean sections were introduced in obstet- epidemiologic trends and calculate health
ric practice at the end of the 19th century, care provision. It is essential to confirm the
but were seldom performed until the depth of villous invasion of the uterine
1920s.3 Not surprisingly, only one of the myometrium in order to improve clinical
20 cases of PA personally treated by Irving management. Histopathologic examination
and Hertig1 occurred after a previous CD. also permits immunohistochemical (IHC)
Similarly, in their review of 86 case reports and molecular biological studies of the
up to 1935, only one was found after a prior underlying mechanisms leading to abnor-
CD. They concluded that the predisposing mal adherence or invasive placentation. We
factors of PA at the time were a previous present here a review the recent findings on
manual delivery and/or “vigorous” uterine the pathophysiology of PA. To facilitate the
curettage. They calculated the prevalence of discussion, we use the term placenta accreta
accreta placentation in their population to spectrum (PAS) to include both adherent
be around 1 in 1956 deliveries. The paper and invasive placental disorders.
by Irving and Hertig was published in a
journal that no longer exists, and thus is not BACKGROUND EPIDEMIOLOGY
recorded in PubMed. There are no other By 1966, when Luke et al,4 published their
publications on placenta accreta between series of PAS, nine of their 21 cases had a
1927 and 1944. previous history of CD. All recent epide-
Major technical advances in surgical miologic studies have shown a clear asso-
and anesthetic techniques and the advent ciation between the CD rate and the
of antibiotics and blood transfusions in incidence of accreta placentation in sub-
the 1940s substantially reduced the mor- sequent pregnancies.5–8 There has been an
bidity and mortality of CD.3 As a con- exponential increase in global rates of
sequence, increasingly women not only CD, which was particularly marked over
survive the procedure but also are able to the past 25 years with rates increasing
have one or more subsequent pregnan- from <7% in 1990 to over 19% in 2014.9
cies. Before the advent of ultrasound and Currently, the highest regional CD rates
magnetic resonance imaging (MRI), the are found in Latin America (40% to 50%)
diagnosis of PA was exclusively clinical and the lowest in sub-Saharan Africa (3%
and made at birth. In the 1920s, PA was to 6%). In Europe, the CD rates vary from
diagnosed “when the placenta fails to be 22% in the North to 31% in the South,
delivered, following the birth of the whereas in Northern America the CD rate
child.”2 Irving and Hertig defined PA as is now 32.3%. Countries such as Egypt,
the abnormal adherence either in whole or Turkey, Brazil have national CD rates
in part of “the afterbirth” to the under- over 50% and rising.10 The incidence of
lying uterine wall.1 Their study became a PAS increases with the number of prior
“classic” but none of their cases had CDs,5–8 and a previous cesarean section
villous tissue invading the myometrium. has been the main cause of PAS for the
This fact has led to a lot of confusion for last two decades.8 Changes in the preva-
obstetricians struggling with the delivery lence of PAS secondary to changes in CD
of a “sticky” placenta. Indeed, the clinical rates are often delayed by 1 or 2 decades 8
definition of a superficially adherent PA is depending on local birth rates. Thus,
very similar to that of placental retention countries with both a high birth rate and
and has not changed since the 1930s. a very high CD rate and rising, such as
Like for all other life-threatening path- Egypt, will soon have the highest mater-
ologies in obstetrics and gynecology, nal morbidity and mortality due to PAS.
histopathologic examination remains the For an unknown reason, placenta ac-
gold diagnostic standard to evaluate creta including its invasive forms increta

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Placenta Accreta Pathophysiology 745

and percreta, has increasingly been re- blastocyst for the scar area.8 Women with
ferred to as “morbidly adherent” placenta a prior history of CD, presenting with a
or MAP which is a 19th century terminol- low-lying placenta or placenta previa, now
ogy used to describe placental retention represent the group with the highest risk of
after delivery. Many modern authors still PAS.13 This epidemiologic association also
use the old clinical criteria to define an indicates that a previous uterine scar can
accreta placenta, that is, difficult manual, have an impact on both implantation and
piecemeal removal of the placenta, ab- placentation.
sence of spontaneous placental separation
20 to 30 minutes after birth despite active ACCRETA PLACENTATION
management including bimanual massage Irving and Hertig1 were the first to suggest
of the uterus, use of oxytocin and con- that the pathologic basis of PAS is the
trolled traction of the umbilical cord, complete or partial absence of the decidua
retained placental fragment requiring cur- basalis. Since then, several concepts have
ettage after vaginal birth and heavy been proposed to explain why and how PA
bleeding from the placentation site after occurs.10,15 The oldest concept is based on a
removal of the placenta during CD.11,12 theoretical primary defect of trophoblast
Since the 1920s, the primary method of biology leading to excessive attachment or
management has been emergency hyster- invasion of the myometrium. The current
ectomy, with the diagnosis being con- hypothesis is that of a secondary defect of the
firmed by histopathologic examination endometrial-myometrial interface leading to
in most case reports and series. Although a failure of normal decidualization in the
the vast majority of PAS cases diagnosed area of the uterine scar allowing abnormally
prenatally or at delivery are still managed deep placentation. An abnormal vasculariza-
by hysterectomy, more than half of the tion resulting from the scaring process after
authors do not provide detailed data on surgery, with secondary localized hypoxia
the macroscopic clinical description at leading to both defective decidualization and
birth and/or histopathologic confirmation excessive trophoblastic invasion, has also
of the diagnosis of placenta accreta and been suggested.16
even more very rarely on the differential
diagnosis between adherent and invasive Development of the Uterine Scar
accreta.13,14 Thus, many cohort studies on Major surgical procedures such as cesar-
the diagnosis and management of PAS ean sections or myomectomies that cut
include a mixed bag of cases with various through the entire uterine wall will leave a
degrees of invasion of the myometrium by scar through all the smooth muscular
chorionic villi. The MAP terminology layers of the myometrium. Unlike the
excludes the more invasive forms of accreta epithelial layers of the endometrium and
placentation, and adds to the confusion in uterine peritoneum that heal by regener-
the clinical differential diagnosis at delivery ation and recolonization of scar area, the
between PAS and placental retention.14 myometrium, does not heal by regenerat-
This lack of precise diagnosis can explain ing muscle fibers, but by forming “for-
the wide variation in incidence and preva- eign” substances including collagen.15
lence of PAS reported in the international Myofiber disarray, tissue edema, inflam-
literature in the last 2 decades.8 mation, and elastosis have all been ob-
The single other most important risk served in uterine wound healing after
factor, reported in around 50% of all cases surgery. The resulting scar tissue is less
of PAS disorders, is placenta previa. The risk elastic and more prone to injury/rupture
of previa also increases with the number in subsequent pregnancies than the intact
of prior CDs, suggesting a tropism of the muscle. Different surgical techniques such

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
746 Jauniaux and Burton

as single-layer versus double-layer closure placenta percreta is likely to be similar

of the myometrium, locked versus un- to that of a tubal rupture in an ectopic
locked single-layer closure8 or the suture placentation. The strong epidemiologic
material used for the closure may have an association between placenta previa and
influence on the healing process and the PAS findings suggest that the decidual
risks of uterine rupture, but the evidence defect following the artificial creation of a
regarding the risk of PAS in subsequent scar in the uterine myometrium has an
pregnancies remains limited. adverse effect on both early implantation
Compared with women with a primary by creating conditions for preferential
CD, women who undergo a repeat cesar- attachment of the blastocyst to scar tissue
ean are more than twice as likely to have and facilitating abnormally deep placen-
PAS disorders.8 Multiple CD (MCD) tal invasion.
scars are often associated with a clear loss Leukocyte recruitment to the endome-
of myometrium or a defect (Fig. 1A) with trium is observed during the normal
a direct communication between the en- secretory phase, and has been reported
dometrial cavity and the visceral serosa.15 to be increased following a CD.10,15 The
Residual myometrial thickness is greater uterine artery resistance is increased, and
and scar defect length, but not depth and the volume of uterine blood flow as a
width, is shorter following double-layer fraction of maternal cardiac output is
compared with single-layer closure. Large decreased in women with a previous CD
cesarean scar defects (CSD) may lead to compared with women with a previous
scar dehiscence with advancing gestation vaginal birth.17 Overall, these data sug-
(Fig. 1B) and could even explain rare gest a possible relationship between a
reports of placenta percreta leading to poorly vascularized uterine scar area and
uterine rupture in the first half of preg- an increase in the resistance to blood flow
nancy. Although this is an extremely rare in the uterine circulation. A large scar
complication of placentation, the mecha- area resulting from MCD and scar dehis-
nism of uterine rupture because of a cence is likely to have an impact on

FIGURE 1. A, Transvaginal ultrasound view of the uterus in a nonpregnant woman with a

history of 2 prior CDs. Note the scar defects through the uterine wall at the junction between the
cervix and lower uterine (U) segment (arrow). B, Transvaginal ultrasound view of pregnant
uterus at 7 weeks gestation after 3 prior CDs showing the lower end of the placenta (arrow)
protruding inside a large cesarean scar defect (CSD). The pregnancy was uncomplicated de-
livered at 36 weeks with the edge of the placenta visible under the uterine serosa but with no
clinical evidence of placenta percreta.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Placenta Accreta Pathophysiology 747

endometrial reepithelialization. Decidual- that migrate through the decidual stroma

ization may be sparse or absent in the and down the lumens of the spiral arteries
overlying area and the absence of struc- respectively. The interstitial EVT invade
tured myometrium underneath is more the uterine wall as far as the inner third of
likely to lead to invasive (increta or the uterine myometrium, also called the
percreta) PAS. junctional-zone (JZ), where they fuse to
PAS is not exclusively a consequence of form multinucleated trophoblast giant
cesarean scar and any form of damage, even cells (MNGCs).10,18 In the weeks follow-
small, to the integrity of the uterine lining ing implantation, EVT cells are found both
following curettage, myomectomy, or endo- within and around the spiral arteries in the
metrial resection has been associated with central area of the placenta. EVT grad-
PAS in subsequent pregnancies.8,10 Uterine ually migrate laterally, reaching the pe-
anomalies, adenomyosis and submucous riphery of the placenta around mid-
fibroids have also been associated with gestation. Endovascular EVT cells in
PAS in primigest women.8 Endometritis the central area, destined to become the
can lead to endometrial fibrosis and poor definitive placenta, act as plugs blocking
decidualization and thus to development of the spiral arteries. These plugs prevent a
PAS. This can explain why before the continuous flow of maternal blood from
advent of antibiotics a prior manual remov- entering the placenta during most of the
al of the placenta with or without a uterine first trimester.18 This phenomenon cre-
curettage was the main factor associated ates an environment of physiological
with PAS in subsequent pregnancies.1 The hypoxia inside the gestational sac, which
trauma to the myometrium and the surface is essential for normal fetal-placental
of the endometrium is often limited in an development and which modulates the
uncomplicated curettage procedure, and formation of the membranes of definitive
should not be associated with the absence placenta.
of reepithelialization of the scar area and Both endovascular and interstitial EVT
changes in the surrounding uterine circula- invasion are associated with the physio-
tion compared with the larger and deeper logical conversion of the terminal part of
scars resulting from MCD. If the myome- the uterine circulation, extending as far
trium scar is small, the placenta may simply the basal part of the spiral arteries at the
grow over it, which can explain why more level of the JZ or the inner third of the
than 90% of cases of placenta previa in myometrium.10 Around 30 to 50 spiral
women with one prior CD are not accreta.10 arteries are transformed during the first
Within this context, if PAS develops it is trimester. In normal pregnancies, the
more likely to be superficial (adherent) transformation of spiral arteries into ute-
and focal. roplacental arteries is described as com-
plete around midgestation. There is a
Characteristics of the Accreta Trophoblast gradient in the infiltration of the EVT
There are 2 types of trophoblast: the along the spiral artery, and even in a
villous trophoblast which covers the pla- normal pregnancy not all spiral arteries
cental villi and which is made up of are completely transformed.18 In humans,
cytotrophoblast cells and the syncytiotro- it is obvious that the decidua does not act
phoblast, and the extravillous trophoblast as a barrier but rather as a matrix that
(EVT) which arises from the distal tips of allows the EVT cells to colonize the JZ in
the anchoring villi that normally make a regulated manner. Trophoblast inva-
contact with the decidua basalis. The sion is notably more aggressive and more
EVT differentiate primarily into interstitial penetrative at sites of ectopic implanta-
and endovascular cells subpopulations tion, for example in the Fallopian tube, in

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
748 Jauniaux and Burton

the absence of decidua. As EVT cells TABLE 1. Histopathologic and

differentiate, they progressively display a Immunostaining Changes
more migratory phenotype, changing Observed in PAS15 According to
their integrin repertoire from predomi- Anatomic Level
nantly collagen IV receptors to fibronec- Villous trophoblast
tin and then laminin receptors.10,15 An Lower syncytiotrophoblast immunostaining
array of factors operates upstream of for MicroRNA-34a, TGF-β, E-CAD, EGF
c-(erbB-2), VEGFR-2, and RTK Tie-2
these pathways to stimulate trophoblast Higher syncytiotrophoblast immunostaining
invasion and includes; cytokines and for EGFR and TIMP-1
growth factors, such as epidermal growth EVT
factor, vascular endothelial growth factor Increased in the size, numbers and depth of
(VEGF), interleukin-1ß, tumor necrosis myometrial invasion of EVTs
Reduced formation of MNGCs
factor-α and the hyperglycosylated form Higher EVT immunostaining for VEGF and
of hCG; hormones, such as triiodothyr- phosphotyrosine
onine, leptin and gonadotropin-releasing Lower EVT immunostaining for sFLT-1
hormone-1; and low (1%) oxygen Uteroplacental vasculature
concentrations.15,18 Equally important in Decreased proportion of remodeled spiral
arteries
the regulation of placentation are the Greater degree of remodeling in radial/arcuate
inhibitors of trophoblast invasion. The arteries in increta and percreta
precise regulation of trophoblast invasion
will therefore depend on the balance of E-CAD indicates E-cadherin; EGFR, epidermal growth factor
receptor; EVT, extravillous trophoblast; MNGCs, multi-
local concentrations of many factors, and nucleated trophoblast giant cells; RTK, receptor tyrosine
also the composition of the extracellular kinase; RTPCR, reverse transcription polymerase chain reac-
tion; sFLT-1, soluble fms-like tyrosine kinase; TGF-β, trans-
matrix. forming growth factor beta; TIMP-1, tissue inhibitors of matrix
We have recently reviewed the main metalloproteinase; VEGF, Vascular endothelial growth factor;
VEGFR, vascular endothelial growth factor receptor.
findings from histopathologic studies in
PAS,15,18 which are displayed in Table 1. develops as a result of abnormal expression
There are wide variations in study design, of growth-related, angiogenesis-related,
accreta definition, number of cases and invasion-related factors in the different
studied, type of tissue investigated and
the extent of quantification of morpho-
logic changes. In brief, the villous tissue
shows no morphologic changes in PAS
compared to nonaccreta placentas, even
in the invasive areas (Fig. 2). The syncy-
tiotrophoblast in PAS villous tissue shows
reduced immunostaining for MicroRNA-
34a, E-cadherin (E-CAD), epidermal
growth factor EGF c-(erbB-2), trans-
forming growth factor beta (TGF-β),
vascular endothelial growth factor recep-
tor 2 (VEGFR-2) and endothelial cell
receptor tyrosine kinase (RTK) Tie-2.
By contrast, there is increased syncytio- FIGURE 2. Microscopic view of the placental
trophoblast labelling for epidermal bed from a hysterectomy specimen at 34
growth factor receptor (EGFR) and weeks in a pregnancy complicated by placenta
TIMP-1 tissue inhibitors of matrix metal- previa increta (H&E×5) showing the dis-
loproteinase (TIMP-1) in PAS. These data ruption of the decidua by placental villi (ar-
row) invading the myometrium.
suggest that abnormal villous adherence

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Placenta Accreta Pathophysiology 749

trophoblast populations. Lower expression invasiveness or proliferation and that the

of MicroRNA-34a may indicate a dysre- absence of the JZ is of more importance in
gulation of the trophoblastic cellular inva- the pathogenesis. The comparison of ul-
sive capacity. However, the villous trasound features of uterine cesarean scar
syncytiotrophoblast has no invasive ca- with histological findings has shown that
pacity and there is no evidence that these large and deep myometrial defects are
biological changes have any impact on often associated with absence of reepithe-
overall fetoplacental development. Thus, lialization of the scar area.10,15 These
these data are difficult to interpret and findings also emphasize the role of the
could be secondary to a focal environ- subdecidual myometrium of the JZ in
mental change in uteroplacental blood modulating placentation. The absence of
flow, in particular in the oxygen concen- decidua in first-trimester cases of PAS
tration changes within the intervillous negates previous suggestions that decidua
space in the invasive areas. is normal at the beginning of gestation
EVT cells in PAS are increased in size and atrophies as pregnancy proceeds.15
and number, as well as in the depth of
myometrial invasion. However, they form Characteristics of Uteroplacental
fewer MNGCs, indicating that they have Vasculature in PAS
not undergone their normal terminal dif- In invasive PAS, EVTs can be found
ferentiation. Immunohistochemistry has beyond the JZ and chorionic villi inside
shown increased vascular endothelial myometrial vascular spaces.15,18 This
growth factor (VEGF) and phosphotyr- leads to an absence of the normal plane
osine in EVT cells from PAS cases com- of cleavage and prevents placental sepa-
pared to normal controls.16 EVT cells lose ration after delivery. Major hemorrhage
their invasive phenotype through syncy- occurs when the condition has not been
tial-type fusion into MNGCs, and the diagnosed prenatally and manual placen-
secretion of VEGF by MNGCs is likely tal delivery is attempted. Invasion of
to be one of the signals initiating and larger vessels in the outer myometrium
coordinating vascularization in the decid- as far as the uterine serosa in PAS is most
ua and placenta during implantation. certainly determined by access rather than
Lower immunostaining for soluble fms- a trophoblastic malfunction. Deeper than
like tyrosine kinase (sFLT-1), which is a normal EVT invasion through the entire
potent antiangiogenic growth factor, has depth of the myometrium transforms the
been observed in the EVT cells of women arterial vasculature beyond the JZ.18,19 In
presenting with PA. These findings sug- cases of preexisting scar defects, EVT cells
gest that VEGF and sFLT-1 play pivotal can infiltrate directly the tissue around the
roles in the process of pathologic pro- radial and even the arcuate arteries lead-
gramming of EVTs toward increased ing to their excessive dilatation.18 This is
motility and invasiveness in PAS. the most prominent feature of invasive
Placenta increta and percreta are not PAS prenatally on ultrasound, and mac-
due to further invasion of EVT in the roscopically on the uterine surface at
uterine wall. They are likely to arise delivery. It is possible these uteroplacental
secondary to the dehiscence of a scar, vascular changes in the accreta area result
leading to the presence of anchoring villi from both neovascularization and/or in-
deep within the uterine wall, and thus creased recruitment of deep uterine ves-
giving EVTs greater access to the deep sels by EVT and chorionic villi beyond
myometrium and beyond. Overall, these the JZ.
results suggest that accreta placentation Although the numbers of interstitial
does not arise through excessive EVT EVT cells are increased in PAS, several

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
750 Jauniaux and Burton

authors have found that spiral artery TABLE 2. Ultrasound Signs According to
remodeling is focally reduced. The defi- Gray-Scale Versus Color-
ciency is seen more in PAS cases without Doppler Imaging Using the
local decidua,15 and remodeling is some- Standardized Descriptions
times completely absent in the accreta Proposed Recently by the
area.19 Even the most invasive forms of European Working Group on
Abnormally Invasive Placenta
PAS pregnancies are not associated with a (EW-AIP) and the AIP
higher incidence of placental-related dis- International Expert Group
orders such preeclampsia and/or fetal
Grey-scale imaging
growth restriction. These disorders are due Thinning of the uterine myometrium
to a reduction in trophoblast invasion and Bladder wall interruption between the uterine
failure of conversion of the spiral arteries serosa and bladder lumen
most of which retain their vasoreactivity Loss of the subplacental clear zone
beyond 22 weeks of gestation. Preeclampsia Intraplacental lacunae
Extrauterine placental bulge
is essentially a disorder affecting primipar- Focal placental exophytic mass
ous women with ethnic variations whereas Color-Doppler imaging (CDI)
PAS is a disorder of multiparous with no Uterovesical hypervascularity
influence of ethnicity. One can hypothesize Hypervascularity of the placental bed
that in the absence of a decidua, the normal Bridging vessels beyond the uterine serosa
Lacunae feeder vessels
release of proteases and cytokines from
activated maternal immune cells is missing, Modified from Collins et al.20 Adaptations are themselves
impairing arterial remodeling. This phe- works protected by copyright. So, in order to publish this
adaptation, authorization must be obtained both from the
nomenon is limited to the scar area in owner of the copyright in the original work and from the
PAS and thus unlikely to lead to a systemic owner of copyright in the translation or adaptation.
disorder such as preeclampsia.
(EW-AIP) and the Ad-hoc International AIP
ULTRASOUND-PATHOLOGIC Expert Group20 proposed standardized
CORRELATIONS descriptions for reporting the ultrasound
Ultrasound imaging and in particular signs used for the prenatal diagnosis of
color-Doppler imaging (CDI) have en- PAS (Table 2). To facilitate the etiopatho-
abled the investigation in vivo of the logic analysis of the ultrasound changes
development of placental circulations in associated with PAS, we have separated
normal and abnormal pregnancies from them according to their uterine or placen-
the first weeks after implantation. There tal origin:
are now > 1000 cases reports and case
series describing the prenatal diagnosis of Anomalies of the Uterine Wall
PAS at different gestational ages, provid- After a full-term pregnancy, the myome-
ing a unique insight into the development trium remains thinner and more elastic in
of accreta placental tissue and its inter- subsequent pregnancies. If the pregnancy
action with the uterine wall. However, has resulted in a CD, the anterior portion
ultrasound studies of accreta placentation of the lower uterine segment will be trans-
have been as heterogeneous as the histopa- formed by the scarring process(es) into an
thologic studies, in particular regarding the area where the residual myometrial tissue
terminology used to describe the different is mixed with fibrinous tissue. As a result,
ultrasound signs associated with PAS and the scar area is always thinner containing
the lack of detailed clinical description of less muscular tissue and thus more prone
the uterus at birth and histopathologic to dehiscence, in particular during the
correlations.14 In 2016, The European Work- third trimester when the lower segment
ing Group on Abnormally Invasive Placenta stretches to accommodate the growing

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Placenta Accreta Pathophysiology 751

thinning of the uterine myometrial to

<1 mm and the loss of the subplacental
clear zone in the myometrium under the
placental bed, corresponding to the nor-
mal decidua and JZ, have been the most
common findings in case reports and
cohort studies.14 In isolation, these signs
are not specific of PAS (Fig. 3) and may
be affected by the placental position, the
pressure of the ultrasound probe, filling of
the bladder and the amount of scar tissue
in the myometrium.
FIGURE 3. Transvaginal ultrasound view of The hypervascularity of the placental
the lower segment of the uterus in a non- bed was described more recently when
pregnant woman with a history of 3 prior CDs CDI was used more routinely to evaluate
presenting with a major placenta previa cov- women at high risk of PAS. In women with
ering partially the cervix. Note the uterine the highest risk of PAS, those with prior
myometrial of <1 mm in thickness and the multiple CDs and/or presenting with a low-
loss of the subplacental clear zone between the
placental bed and the bladder wall (arrow).
lying or placenta previa, hypervascularity
There was no evidence of PAS at birth. B between the myometrium and the utero-
indicates bladder; P, placenta. vesical posterior wall of the bladder has
been reported in over 80% of cases com-
plicated by PAS.13,14 CDI is not essential to
fetus. This phenomenon will be influenced confirm the diagnosis of PAS in expert
by prior multiple pregnancies and the hands, but may assist in the screening of
number of prior CDs, but also by poor women at higher risk.13 As uterine vascu-
scarring process resulting in permanent larity and vascular dilatation increases with
scar defect with limited endometrial advancing parity, increased vascularity
reepithelialization.15,18 Not surprisingly, under the placental bed is not always
the ultrasound signs defined as the pathognomonic of PAS (Figs. 4, 5).

FIGURE 4. Transabdominal ultrasound longitudinal views of the placental bed at 22 weeks in a

pregnancy after 1 prior CD presenting with a major placenta previa covering the cervix showing:
A, areas of absent myometrium and irregular subplacental clear zone and numerous lacunae
(asterisks) giving the placenta a “moth eaten” appearance on grey-scale imaging; B, increased
and anarchic hypervascularity of the placental bed on CDI. An extended zone of placenta increta
was confirmed at birth. B indicates bladder; F, fetus; P, placenta.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
752 Jauniaux and Burton

FIGURE 5. Transabdominal ultrasound longitudinal views of the placental bed at 28 weeks in a

pregnancy after 2 prior CDs presenting with a major placenta previa covering partially the cervix
showing: A, area suggesting a disruption of the uterine wall between the placental bed and the bladder
(B) with absent myometrium (arrow) and intraplacental lakes on grey-scale imaging. B, CDI indicates
a normal uteroplacental circulation. Note that under the pressure of the ultrasound probes most of the
lakes have disappeared. No PAS was found at birth. B indicates bladder; F, fetus; P, placenta.

Interruption of the bladder wall with myometrium.18 Not surprisingly, in > 50%
loss of the hyperechoic line between the of invasive cases of PAS feeder vessels can
uterine serosa and bladder lumen has be seen entering the lacunae.14 Lacunae
been rarely described on ultrasound.14 must be differentiated on ultrasound from
This sign may result from the villous placental lakes (Fig. 5) which are echolucent
invasion into the muscle of the posterior areas in the center of a cotyledon, under the
wall of the bladder in placenta percreta or chorionic plate or in the marginal areas and
may represent an ultrasound artefact are part of the normal anatomic develop-
arising from the massive hypervascularity ment of the definitive placenta.18
of the placental bed.18 Similarly, the Placental bulge distorting the extra-
presence of bridging vessels across the uterine organs and focal exophytic mass
myometrium and beyond the uterine se- of placental tissue extending beyond the
rosa into the bladder before disappearing serosa have been rarely described in PAS
14
has been reported inconsistently in cases and should only been seen in cases of
of PAS 14 and may be the consequence of placenta percreta.18
an ultrasound artefact.18

Anomalies of the Placenta Conclusions

The presence of intraplacental lacunae caus- The comparison of in-vivo ultrasound
ing large and irregular sonolucent areas features with histopathologic findings is
within the placental mass and giving it a essential in order to better understand the
“moth eaten” appearance (Fig. 4) is the phenomenon of accreta placentation. As
most common ultrasound sign described in we can only witness the consequences of
PAS.14 This anomaly results from the dis- an abnormally deep EVT migration and
tortion of the normal placental cotyledo- villous attachment below the JZ at deliv-
nary anatomy by the unrestricted entry ery, what happens during the initial phase
from the beginning of the second trimester of placentation in PAS several months
of high velocity (turbulent) flow from previously remains a mystery. Both ultra-
the deep arterial vasculature of the sound features and histopathologic data

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Placenta Accreta Pathophysiology 753

support the concept that the morphologic  Correlations of the ultrasound imaging
changes observed in the EVT in PAS are from early in pregnancy with clinical
environmental, and the consequence of an and histopathologic examinations at
unusual and prolonged interaction with birth are pivotal to better understand
the highly vascularized deep myome- the natural evolution of PAS and are
trium, which these cells would normally essential to improve the diagnosis and
not reach in normal placentation. Overall management of this increasingly com-
these findings support the concept of a mon major obstetric complication.
primary deciduomyometrium defect in  Standardized ultrasound and clinical
PAS, exposing the uterine myometrium terminology is essential to allow direct
below the JZ to the migrating EVT. The correlations between antenatal and de-
loss of the normal plane of placental livery findings in PAS and provide more
cleavage from the uterine wall and the accurate population epidemiology data.
excessive vascular remodeling of the
radial and arcuate arteries can explain
the prenatal ultrasound findings and Acronyms
the clinical consequence of accreta pla-
centation at delivery, in particular in its  Placenta accreta: general term used to
invasive forms. Neovascularization or describe the different levels or grades of
dilation of the myometrial vasculature the accrete placenta spectrum.
induced by the trophoblast increases the  Placenta creta: histologic description of
risk of hemorrhage when manual remov- placental villi adherent to the myome-
al of an undiagnosed placenta accreta is trium without interposing decidua.
attempted.  Placenta increta: clinical description of
placenta villi invading the myometrium
down to the uterine serosa.
Teaching Points  Placenta percreta: clinical description
of placenta villi invading the entire
 The recent increase in placenta accreta uterine wall and beyond.
spectrum disorders is directly linked to  Morbidly adherent placenta: 19th Cen-
the increase incidence in uterine scar tury clinical terminology used to de-
from previous cesarean section. scribe placenta retention because of
 There is mounting evidence that abnor- various causes including placenta creta.
mal villous adherence and/or invasion
of the uterine wall is due to a defect
of the junctional zone between the
superficial myometrium and the References
endometrium. 1. Irving C, Hertig AT. A study of placenta accreta.
 Macroscopic and microscopic exami- Surgery, Gynecol Obstet. 1937;64:178–200.
nations of hysterectomy specimens or 2. Forster DS. A case of placenta accreta. Can Med
Assoc J. 1927;17:204–207.
myometrial samples from an abnormal 3. West MJ, Irvine LM, Jauniaux E. Caesarean
placentation site remain the gold stand- section: From antiquity to the 21st century. In:
ard of reference to confirm the diag- Jauniaux E, Grobman W, eds. A Textbook of
nosis of PAS. Caesarean Section. Oxford: Oxford University Press;
 Trophoblast biological changes ob- 2016:9–24.
4. Luke RK, Sharpe JW, Greene RR. Placenta
served with immunohistochemistry in accreta: the adherent or invasive placenta. Am J
PAS are mainly secondary to the devel- Obstet Gynecol. 1966;95:660–668.
opment in a different environment with 5. Silver RM, Landon MB, Rouse DJ, et al. National
no interaction with decidual tissue. institute of child health and human development

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
754 Jauniaux and Burton

maternal-fetal medicine units network. Maternal 13. Jauniaux E, Bhide A. Prenatal ultrasound diag-
morbidity associated with multiple repeat cesarean nosis and outcome of placenta previa accreta after
deliveries. Obstet Gynecol. 2006;107:1226–1232. caesarean delivery: a systematic review and meta-
6. Morlando M, Sarno L, Napolitano R, et al. Pla- analysis. Am J Obstet Gynecol. 2017;217:27–736.
centa accreta: incidence and risk factors in an area 14. Jauniaux E, Collins SL, Jurkovic D, et al. Accreta
with a particularly high rate of cesarean section. placentation: a systematic review of prenatal
Acta Obstet Gynecol Scand. 2013;92:457–460. ultrasound imaging and grading of villous inva-
7. Thurn L, Lindqvist PG, Jakobsson M, et al. siveness. Am J Obstet Gynecol. 2016;215:712–21.
Abnormally invasive placenta-prevalence, risk fac- 15. Jauniaux E, Bhide A, Burton GJ. Pathophysiol-
tors and antenatal suspicion: results from a large ogy of accreta. In: Silver R, ed. Placenta accreta
population-based pregnancy cohort study in the syndrome. Portland: CRC Press; 2017:13–28.
Nordic countries. BJOG. 2016;123:1348–1355. 16. Wehrum MJ, Buhimschi IA, Salafia C, et al.
8. Jauniaux E, Chantraine F, Silver RM, et al. for the Accreta complicating complete placenta previa is
FIGO placenta accreta diagnosis and management characterized by reduced systemic levels of
expert consensus panel. FIGO consensus guide- vascular endothelial growth factor and by epithe-
lines on placenta accreta spectrum disorders: lial-to-mesenchymal transition of the invasive
Epidemiology. Int J Gynecol Obstet. 2018;140: trophoblast. Am J Obstet Gynecol. 2011;204:
265–273. 411.e1–411.e11.
9. Betran AP, Ye J, Moller A-B, Zhang J, et al. The 17. Flo K, Widnes C, Vårtun Å, et al. Blood flow to
increasing trend in caesarean section rates: global, the scarred gravid uterus at 22-24 weeks of
regional and national estimates: 1990-2014. PloS gestation. BJOG. 2014;121:210–215.
One. 2016;11:e014843. 18. Jauniaux E, Collins SL, Burton GJ. Placenta
10. Jauniaux E, Jurkovic D. Placenta accreta: patho- accreta spectrum: pathophysiology and evi-
genesis of a 20th century iatrogenic uterine dis- dence-based anatomy for prenatal ultrasound
ease. Placenta. 2012;33:244–251. imaging. Am J Obstet Gynecol. 2018;218:75–87.
11. Gielchinsky Y, Rojansky N, Fasouliotis SJ, et al. 19. Khong TY, Robertson WB. Placenta creta and
Placenta accreta-summary of 10 years: A survey placenta praevia creta. Placenta. 1987;8:399–409.
of 310 cases. Placenta. 2002;23:210–214. 20. Collins SL, Ashcroft A, Braun T, et al. Proposal
12. Bailit JL, Grobman W, Rice MM, et al. Morbidly for standardized ultrasound descriptions of ab-
adherent placental treatments and outcomes. normally invasive placenta (AIP). Ultrasound
Obstet Gynecol. 2015;125:683–689. Obstet Gynecol. 2016;47:271–275.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA CLINICAL OBSTETRICS AND GYNECOLOGY
Volume 61, Number 4, 755–765
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.

Imaging of Placenta
Accreta Spectrum
Downloaded from [Link] by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3dW0s7N5CLiYMlH4wgaY34Fr69rtLH3oWi4+h6VxXUqo= on 10/30/2018

ELIZA M. BERKLEY, MD, and ALFRED ABUHAMAD, MD

Eastern Virginia Medical School, Norfolk, Virginia

Abstract: Placenta accreta spectrum (PAS) refers to an commonly used to describe this pathologic
abnormally invasive implantation of the placenta into entity. PAS includes accretas, incretas, and
the uterine myometrium. The resultant risk is that of
severe maternal hemorrhage and significant maternal percretas. Unfortunately, severe and life-
morbidity and even mortality. The 2 strongest risk threatening hemorrhage with significant
factors for the development of PAS are a history of a maternal morbidity and even mortality still
prior cesarean section and a placenta previa in the persist today.1,3 Because of these concerns,
current pregnancy. Clinically, most patients are asymp- the American College of Obstetricians and
tomatic but some will present with vaginal bleeding and
abdominal cramping. The goal of this article is to Gynecologists and the Society for Mater-
discuss the common clinical presentation and risk nal Fetal Medicine recently developed a
factors for placenta accreta spectrum, and to review in standardized system describing levels of
detail the ultrasound features/markers of PAS in each maternity care in an effort to ensure that
trimester. pregnant women diagnosed with PAS are
Key words: accreta, previa, ultrasound, Doppler,
lacunae cared for at appropriate “centers of excel-
lence.” The goal is to significantly reduce
maternal morbidity and mortality associ-
ated with PAS.4
Introduction Pathogenesis of PAS is not clear, with
Placenta accreta is defined as the abnormal several existing theories. Abnormal vascu-
implantation of the placenta into the ute- larization resulting from the scarring proc-
rine myometrium. The increasing depth of ess after uterine surgery with secondary
trophoblast invasion defines the pathologic localized hypoxia leading to defective de-
entity as a placenta accreta, increta, or cidualization and excessive trophoblast in-
percreta.1 The first cases were described vasion is the most generally accepted
in the 1930s and reported massive maternal pathophysiology.5–7 Before the 1930s, al-
hemorrhage, severe morbidity, and death.2 most all pregnant women delivered vagi-
Today the terms morbidly adherent pla- nally and a cesarean section was a rare
centa or placenta accreta spectrum (PAS, occurrence. For many reasons, the cesar-
which we will use in this chapter) are ean section rate has increased, reaching a
Correspondence: Eliza M. Berkley, MD, Eastern
peak in 2009 of ∼33%.8 Similarly, there has
Virginia Medical School, Norfolk, VA. E-mail: been a concomitant rise in PAS over the
berkleem@[Link] last half century. The overall incidence is
The authors declare that they have nothing to disclose. around 3 per 1000 deliveries.9,10 In turn,

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 61 / NUMBER 4 / DECEMBER 2018

[Link] | 755
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
756 Berkley and Abuhamad

the previously rare PAS, is now a fairly and include damage to surrounding
common complication of pregnancy. organs, postoperative bleeding and con-
sumptive coagulopathy, transfusion-
related complications, acute respiratory
Clinical Presentation distress syndrome, thromboembolism, in-
Most patients with PAS are asymptomatic. fectious morbidities, multisystem organ
In contrast, some patients will present with failure, and maternal death.15 There is
vaginal bleeding and cramping. These find- also a significant risk for genitourinary
ings are largely related to a concomitant complications including a 15% risk of
placenta previa, which a strong risk factor cystotomy and a 2% risk of ureteral
for PAS. Cesarean sections are also known injury.11 Accurate prenatal diagnosis of
risk factors. A maternal fetal medicine placenta accreta is essential to minimize
network study in 2006 showed that the these risks.
presence of previous cesarean section, es- To improve health outcomes for pa-
pecially multiple cesareans, significantly tients with PAS, extensive delivery plan-
increased the risk of PAS, especially in ning needs to be coordinated in a center
association with the presence of a placenta with appropriate resources, which should
previa. For instance, a patient who has a include access to appropriate subspecial-
placenta previa and 3 previous cesarean ists and a blood bank which can manage
sections has a 40% chance of placenta massive transfusion protocols. The first
accreta. If this same patient does not have step to achieve better outcomes involves
a placenta previa, despite 3 previous cesar- educating obstetricians about the major
ean sections, her risk for a placenta accreta risk factors for PAS, and when these risks
decreases to <1%.11 As such, it is of utmost are identified, protocols must be in place
importance to perform an ultrasound as- to accurately evaluate the patients. Ultra-
sessment for placenta previa in patients sound is integral to prenatal diagnosis.
with previous cesarean delivery. Other risk When PAS cannot be excluded on ultra-
factors for PAS include advancing mater- sound, a referral to an ultrasound center
nal age, multiparity, in vitro fertilization, with extensive experience in the diagnosis
previous uterine surgery, previous uterine and management of PAS is recom-
irradiation, endometrial ablation, Asher- mended. In addition, continued education
man syndrome, uterine leiomyomata, ute- on ultrasound markers of PAS should be
rine anomalies, hypertensive disorders of provided to all imaging centers caring for
pregnancy, and smoking.9 Again ultra- pregnant women.
sound is helpful in assessing for PAS in In the following sections, the authors
patient with many of these risks. will review the significance of each PAS-
The optimal management of PAS is sonographic marker and will present an
still debated in regards to timing approach to the diagnosis of PAS in each
of delivery and surgical approach. How- trimester.
ever, there is general agreement that the Ultrasound is the most commonly used
best outcomes occur when delivery is modality for the diagnosis of PAS. There
performed in a center of excellence with are a vast number of studies reporting on
a multidisciplinary team approach.12 Ma- sonographic markers of PAS. It is impor-
ternal blood loss and morbidity can be tant to evaluate the overall predictive
reduced with a planned cesarean hyster- accuracy of ultrasound in making the
ectomy before the onset of labor and diagnosis of PAS and understand each
by not attempting to remove the placenta markers contribution to its detection.
after delivery of the neonate.13,14 Compli- Small sample size and retrospective de-
cations are higher in undiagnosed cases signs confound the results of early studies.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Imaging of Placenta Accreta Spectrum 757

Newer studies have reported the sensitiv-

ity, specificity, positive, and negative pre-
dictive value for each marker, and some
have proposed scoring systems.
Before we describe the sonographic
features of PAS, it is important to opti-
mize one’s sonographic approach. The
use of transvaginal ultrasound is key
unless there is an absolute contraindica-
tion. The transvaginal approach provides
higher resolution and therefore enhances
visualization. First and most importantly,
it confirms the presence of placenta previa FIGURE 1. Sagittal view of the uterus in a
when suspected on a transabdominal low uterine segment implantation of the GS
ultrasound examination. Second, the in a pregnancy with 3 previous cesarean
transvaginal approach allows for the sections. See the location of the GS in the
evaluation of the cervix for placental lower uterine segment posterior to the B. Note
invasion and the posterior bladder wall the presence of multiple vascular lacunaes
and internal aspects of the bladder. In (arrows) within the placenta, representing
addition, the application of color Doppler another sonographic sign of placenta accreta
in low velocity allows assessment of the in early gestation. This pregnancy resulted in
a placenta percreta. B indicates bladder; GS,
extent of placental and lower segment gestational sac.
vascularization. Table 1 reviews the steps
for optimizing the sonographic approach
to evaluate for the presence of PAS. lower third of the uterus between 8 and
Ultrasound features of PAS may be 10 weeks or primarily occupying the
present as early as the first trimester. The lower uterine segment from 10 weeks
3 main markers include a gestational sac onward.16 It is important, however, to
(GS) implanted in the lower uterine seg- consider ultrasound findings in their clinical
ment (Fig. 1), a GS that is embedded in
a cesarean section scar (Fig. 2), and the
presence of multiple vascular spaces
called lacunae within the placental bed
(Fig. 1). Lower uterine segment implan-
tation is defined as a GS implanted in the

TABLE 1. Optimization of the Ultrasound

Examination in Placenta
Accreta
Use a combined transvaginal and transabdominal
approach
Adjust focal zone(s) to the region of interest
On transabdominal approach, magnify placenta FIGURE 2. Sagittal view of the lower uterine
and scan it in its entirety segment and the Cx in a cesarean section scar
On transvaginal approach, reduce sector width, implantation of the GS in a pregnancy with 1
but ensure posterior bladder wall is in view previous cesarean section. See the location of
Add color Doppler in low-velocity scales and low the GS (arrow) embedded in the cesarean
filters section scar. Cx indicates cervix; GS, gesta-
Save images and/or movie clips of placenta tional sac.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
758 Berkley and Abuhamad

context, and to differentiate between

pregnancies in the lower uterine segment
due to expulsion and pregnancy loss from
those at significant risk for PAS. During
a pregnancy loss, the sac moves when one
applies pressure to the anterior surface
of the uterus. In addition, the sac will not
show circumferential blood flow with the
application of color Doppler.17 In a retro-
spective study, Comstock evaluated early
ultrasound findings before 10 weeks ges-
tation in 7 cases of placenta accreta. One
GS was located near the uterine fundus, FIGURE 3. Color Doppler of a sagittal view
whereas 6 were low-lying and near the of the lower uterine segment and the Cx in a
cervix in patients with a previous cesarean cesarean section scar implantation of the GS
section. In these 6 cases, 2 had first in a pregnancy with 1 previous cesarean sec-
trimester losses in which dilation and tion (same as in Fig. 2). Note the presence
curettage was performed. They both of vascularity on color Doppler surrounding
required hysterectomies due to excessive the GS (arrow). Cx indicates cervix; GS,
blood loss. Of the remaining 4 cases gestational sac.
that delivered near term, 3 were cesarean
hysterectomies, whereas 1 underwent in the lower uterine segment, near the
a vaginal delivery but had persistent cesarean section scar.15 A true scar preg-
vaginal bleeding necessitating a uterine nancy should be implanted within the
artery embolization.18 myometrium, surrounded on all sides by
It is important to note, that not all low- myometrium, and be separate from the
lying GSs lead to PAS. Normal pregnancies endometrium (Fig. 2). The application
have been reported. In these instances, the of low-velocity color Doppler will show
GS should be contiguous with the endome- the surrounding vascularity of placental
trial cavity, a normal thick anterior myo- tissue (Fig. 3). If untreated, significant
metrium should be seen superior to the GS, abnormalities develop in the placenta and
and a continuous white line representing the will lead to a placenta accreta, increta,
bladder-uterine wall interface should be or percreta. Although it may be useful to
seen on ultrasound.17,18 distinguish between true scar pregnancies
There is a significant risk for PAS in and pregnancies that implant in the lower
patients with a previous cesarean section uterine segment adjacent to a scar, both
whose pregnancies implant in the lower carry considerable risk for PAS and exces-
uterine segment. In these cases, the sono- sive hemorrhage. Thus, treatment of both
graphic features include a GS implanted in or conditions is similar.
near the cesarean section scar, a thin appear- Recognition of these ultrasound
ing anterior myometrium, and an irregular markers is important as early diagnosis
placental-myometrial and bladder-uterine may improve the patient’s prognosis and
wall interface17 (Fig. 2). The term cesarean decrease the need for hysterectomy.19 In
scar pregnancy should be differentiated from general, there are 3 treatment modalities:
a low uterine segment implantation. The surgical treatment (abdominal or trans-
former implies a GS that is embedded within vaginal), direct injection of methotrexate
the cesarean section scar. or potassium chloride into the GS, or
Many studies combine cesarean scar expectant management. A study evaluat-
pregnancies with pregnancies that implant ing the management and outcomes of 18

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Imaging of Placenta Accreta Spectrum 759

scar pregnancies diagnosed in the first TABLE 2. Second and Third Trimester
trimester, showed the benefit of early Sonographic Markers of
diagnosis and intervention. Placental Accreta
Balloon catheter placement into the Multiple vascular lacunae26
lower uterine segment or cervix under Loss of the normal hypoechoic retroplacental
transvaginal ultrasound guidance may be zone27
Abnormality of the uterine serosa-bladder
considered as a prophylactic method or interface28
adjuvant treatment in the management of Thinning of the retroplacental myometrium29
cesarean scar or cervical pregnancies. Bal- Bulging of the lower uterine segment30
loon tamponade has been used to decrease Increased placental vascularity on color
bleeding from cesarean scar or cervical Doppler30
pregnancies. It is easy to perform, well
tolerated by patients, decreases complica-
tions from severe bleeding and decreases sonographic findings of PAS in the sec-
the need for surgical intervention.20,21 A ond and third trimester. Lacunae are
study of 18 such cases, reported that typically located deep within the placenta,
balloon catheter placement was successful they have an irregular border, and have
in achieving tamponade in all but one. In 8 turbulent flow with high velocity and low
cases it was placed immediately after me- impendance on color Doppler26 (Fig. 5).
thotrexate injection. In another 8 subjects it The pathogenesis of these findings is
was placed following suction aspiration. In probably related to placental tissue alter-
addition, ultrasound follow-up showed re- ations resulting from long-term exposure
gression of vascularization.20 to pulsatile blood flow.31,32 Multiple vas-
The third marker of placenta accreta in cular lacunae within the placenta are
the first trimester is the presence of an- correlated with a high sensitivity and
echoic areas within the placenta with or low false-positive rate for PAS.26 Some
without documented blood flow on color studies report lacunae to have the highest
Doppler16,22–25 (Fig. 1). Commonly
called placenta lacunae, in 1988 Kerr de
Mendonca described them as a marker
for PAS. Since then, multiple case reports
have supported this first trimester finding
as being associated with PAS. One study
of 10 cases of PAS with first trimester
ultrasound confirmed the presence of
these anechoic areas in 8.16 If the preg-
nancy continues, the lacunae become
more prominent and are well recognized
markers for PAS in the second and third
trimesters of pregnancy (see below).
In the second and third trimester,
multiple ultrasound findings have been FIGURE 4. Sagittal view of the lower uterine
associated with PAS. Table 2 lists the segment in gray scale ultrasound in a patient
common sonographic markers. The easi- with placenta accreta. Note the presence of
est to identify on ultrasound examination multiple anechoic spaces within the placenta,
are placenta lacunae. corresponding to placental lacunaes (arrows).
Multiple vascular lacunae within the Some studies have correlated the presence of
placenta, or a Swiss-cheese appearance ≥ 4 placental lacunaes with a high prevalence
of placenta accreta (see text for details).
(Fig. 4) is one of the most important

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
760 Berkley and Abuhamad

FIGURE 5. Sagittal view of the lower uterine segment and Cx in gray scale (A) and color
Doppler (B) ultrasound in a patient with placenta accreta. Note the presence of multiple an-
echoic spaces within the placenta in A, corresponding to placental lacunaes (arrows). Color
Doppler in B shows vascular flow within the lacunaes. Cx indicates cervix.

sensitivity and positive predictive value other markers. The strength of this marker
among sonographic markers for PAS in is in its negative predictive value of 96% to
the second and third trimester. However, 100%. The presence of a hypoechoic retro-
their positive predictive value varies by placental clear space extending the length
study and ranges from 73% to 100% and of the placenta essentially excludes the
their negative predictive value is between diagnosis of PAS23,33 (Fig. 7).
88% and 100%.26 All studies note that the Many studies report that an irregular
greater the number of lacunae, the higher uterine serosa-posterior bladder wall interface
the risk for PAS. Some studies report a
100% rate of placenta accreta when ≥ 4
lacunae are seen.23,33 It is important to
remember that lacunae may be present in
normal placentas, especially in the ab-
sence of previa. Moreover, PAS has been
reported in the absence of multiple vas-
cular lacunae.34
Loss of the normal hypoechoic (clear)
retroplacental zone between the placenta
and the uterus is another second and third
trimester marker for PAS27,35 (Fig. 6). This
PAS marker has a reported detection rate
of 93%, a sensitivity of 52%, and a specific-
ity of 57%. Unfortunately studies report a
false-positive rate of ≥ 21%.27,34–36 In ad- FIGURE 6. Transabdominal ultrasound in
dition, the loss of the clear space between gray scale in a pregnancy with an anterior
the placenta and uterus is angle dependent placental accreta. Note the absence of the
and can be seen in normal anterior placen- hypoechoic zone (arrows) between the pla-
tas. Therefore, this marker should not be centa and the uterine wall. Compare with a
normal placenta in Figure 7.
used alone, but rather in combination with

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Imaging of Placenta Accreta Spectrum 761

have also been described as varicosities

along the bladder wall or bulging of the
placenta into the posterior bladder wall.
This may be seen as early as the first
trimester but is more commonly noted in
later gestation.23 Initially, this marker was
believed to be highly specific, near 100%,
but not very sensitive.14,30 It was helpful in
cases of placenta percreta but less diag-
nostic when there was no bladder invasion.
However, when one uses a transvaginal
approach and applies color Doppler to the
uterine-bladder interface, the sensitivity
FIGURE 7. Transabdominal ultrasound in improves from 70% to 90% and the spe-
gray scale in a pregnancy with an anterior- cificity increases to ∼99%. The positive
fundal normal placenta. Note the normal and negative predictive values are 96% and
appearance of the retroplacental hypoechoic 92%, respectively.33 In fact, a recent meta-
zone (arrows) between the placenta and the
analysis reported that irregularity of
uterine wall. Compare with a placenta accreta
in Figure 6. this interface was the most specific mar-
ker for invasive placentation (99.75 CI,
99.5%-99.9%).28
is a PAS marker in the second and third Another sonographic marker for PAS
trimesters of pregnancy. The normal inter- is a retroplacental myometrial thickness
face between the uterine serosa and poste- of <1 mm.28,29 Thinning of the myome-
rior bladder wall should appear smooth trium in the upper uterine segment and
and without increased vascularity on fundus is always a concern. However,
sagittal imaging (Fig. 8). In contrast, in determination of myometrial thickness
PAS the sonographer may see an inter- in the lower uterine segment is difficult
ruption of the line, irregularity of the to assess and a cutoff for irregularity is
line, or increased vascularity on color controversial. First, in a normal preg-
Doppler.14,30 Abnormalities of the uterine nancy the lower uterine segment myome-
serosa-posterior bladder wall interface trium thins with labor. In addition, the
lower uterine segment myometrium is
thinner in patients with previous cesar-
eans. One study reports a median lower
uterine segment myometrial thickness of
2.4 mm in the third trimester.37 The range
of sensitivity and specificity for this
marker is reported as 22% to 100% and
100% to 72%, respectively.28,29 Given the
variation, more studies are necessary to
standardize the sonographic approach
and optimal gestational age of assessment
FIGURE 8. Transvaginal ultrasound of a for this marker.
sagittal view of the lower uterine segment in Color Doppler may be used as an
gray scale in a normal pregnancy. Note the adjunct to 2-dimensional ultrasound in
thin, uniform shape of the posterior B wall- the diagnosis of placenta accreta. On
lower uterine segment interface (arrows). B color Doppler the normal subplacental
indicates bladder; Cx, cervix.
venous complex should appear nonpulsatile

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
762 Berkley and Abuhamad

FIGURE 9. Sagittal view of the lower uterine segment and Cx in gray scale (A) and color Doppler
(B) ultrasound in a patient with placenta previa and accreta. Note the presence of multiple anechoic
spaces within the placenta in A, corresponding to placental lacunaes (arrows). Color Doppler in B
shows extensive vascular flow within the lacunaes. Cx indicates cervix; P, placenta.

with low-velocity venous blood flow wave- The application of 3-dimensional ultra-
forms. In contrast, markedly dilated pe- sound in vascular mode has shown promise
ripheral subplacental vascular channels as it allows for a semiquantitative
with pulsatile venous-type flow is sugges- assessment of placental vasculature.42,43
tive of PAS. These vascular channels are Drawbacks exist, however, for its use in
often located directly over the cervix screening for PAS as it requires significant
(Fig. 9). The presence of bridging vessels operator expertise for volume acquisition
linking the placenta and bladder with high and manipulation. Prospective studies are
diastolic arterial blood flow also suggests needed to standardize this technique and
PAS.30,38 Two small-scale studies reported assess its value.
the sensitivity of color Doppler imaging for
the diagnosis of invasive placenta as between DIAGNOSTIC ACCURACY OF
86% and 100% and the specificity between ULTRASOUND
94% and 92%.39,40 Prenatal diagnosis of PAS is imperative to
Other proposed markers for PAS in- optimize maternal outcomes by allowing for
clude a placental bulge or focal exophytic delivery at a center skilled in the maternal
mass. Deviation of the uterine serosa can care of these patients. The primary imaging
cause the placenta to bulge and impinge modality available to pregnant patients is
on the bladder. Disruption of the uterine ultrasound. Therefore, to achieve this goal
serosa with the appearance of the placen- of improved outcomes, women with risk
ta as an exophytic mass is concerning for factors such as placenta previa and previous
placenta accreta and percreta, although cesarean section(s) should receive a detailed
performance characteristics as a marker evaluation by a provider skilled in ultra-
for PAS are unknown.30,41 sound diagnosis of PAS.
Three-dimensional ultrasound has also There is considerable debate, however,
been evaluated in the prediction of PAS. on the overall diagnostic accuracy of

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Imaging of Placenta Accreta Spectrum 763

ultrasound. Early studies concluded that to standardize technique. For example,

overall, gray scale ultrasonography was does one evaluate the uterine serosa and
an excellent tool for the prenatal diagno- bladder interface with transabdominal or
sis of PAS. Sensitivities ranged from 77% transvaginal ultrasound and is the blad-
to 87%, specificities were as high as 96% der empty or full? A transvaginal ap-
to 98%, the positive predictive value proach with a partially full bladder is
ranged from 65% to 93%, and the negative best and can yield a sensitivity of 70%
predictive value was 98%.44 A meta- and specificity of 99%.33 It is important
analysis in 2013 reported a pooled sensi- to define and standardize the approach
tivity of 83%. In addition, another systemic for each sonographic marker of PAS to
review and meta-analysis of 23 studies decrease variability and improve diagnos-
assessed the prenatal accuracy of 4 sono- tic accuracy.
graphic markers. The accuracy of placenta In an effort to improve communica-
lacunae, loss of the hypoechoic retropla- tion, make evaluation of PAS more con-
cental zone, abnormalities of the uterine sistent with a systematic approach, and
serosa and bladder wall interface, and color allow true assessment of the diagnostic
Doppler abnormalities were as follows: performance of each sonographic marker
0.89, 0.88, 0.93, and 0.95 respectively. This or combination of markers, the European
study reported an average sensitivity of Working Group on Abnormally Invasive
91% and average specificity of 97%.45 In Placenta has proposed a standard form
contrast, more recent studies, are unable and protocol. Their goal is to standardize
to replicate such high positive predictive risk into low, intermediate or high in
values. Perhaps the earlier studies over- efforts to reduce maternal morbidity and
estimate the diagnostic accuracy of perinatal complications.47 The authors
ultrasound. Limitations certainly include agree that this is an important first step.
patient selection bias and sizeable interob-
server variation. Other limitations include MAGNETIC RESONANCE IMAGING
small sample sizes, retrospective study (MRI)
designs with wide variability in the defi- MRI is another modality used in the
nition and inclusion criteria leading to evaluation of PAS. In the majority of cases,
inconsistency in performance and skewed ultrasound remains the preferred imaging
sensitivity. To date, the largest prospective method. MRI is less accessible, more ex-
study of sonographic diagnosis of PAS pensive, and has fewer providers skilled in
patients notes a drop in sensitivity to the prenatal diagnosis of PAS. Similar to
54%, a specificity of 88%, and positive some of the initial ultrasound studies, MRI
and negative predictive values of 82% and studies are small and prone to selection
65%.46 It appears that when investigators bias. Like ultrasound, multiple markers are
are blinded to clinical history and multiple used in MRI to evaluate for PAS. The most
experienced providers read the same ultra- common MRI markers include dark pla-
sound images, the accuracy of diagnosing cental bands on t2-weighted imaging, pla-
PAS declines. cental or uterine bulge, disruption of the
Two areas that need further study are area between the uterus and placenta, and
comparing accuracy for antenatal diag- abnormal placental vascularity.45 The sen-
nosis in centers of PAS imaging excellence sitivity of MRI is 80% to 85% overall with
versus those inexperienced in diagnosing a specificity of 65% to 100%.27 MRI may
PAS, and assessing varying ultrasound be complimentary to ultrasound when the
marker’s accuracy with the addition of placenta is posterior or lateral, or when
color Doppler and standardized techni- there is a suspicion of organ involvement
ques. Recent studies underscore the need such as in placenta percreta. In most cases

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
764 Berkley and Abuhamad

however, ultrasound imaging is sufficient 8. Osterman M, Martin JA. Trends in low-risk

and provides enough diagnostic informa- cesarean delivery 1990-2013. Natl Vital Stat Rep.
tion to allow planning for the optimal 2014;63:1–15.
9. Belfort MA. Placenta accreta. Am J Obstet Gynecol.
surgical approach and management of 2010;203:430–439.
PAS.48 10. Hull AD, Resnik R. Placenta accrete and post-
partum hemorrhage. Clin Obstet Gynecol. 2010;53:
228–236.
Conclusions 11. Silver RM, Landon MB, Rouse DJ, et al. Mater-
nal morbidity associated with multiple repeat
The incidence of PAS is on the rise as is cesarean deliveries. Obstet Gynecol. 2006;107:
the associated morbidity and mortality. 1226–1232.
Knowing the risk factors for PAS, opti- 12. Silver RM, Fox KA, Barton JR, et al. Center of
mizing its prenatal diagnosis, and provid- excellence for placenta accreta. Am J Obstet
Gynecol. 2015:561–568.
ing a multidisciplinary approach to care
13. Eller AG, Bennett MA, Sharshiner M, et al.
are critical elements to improving out- Maternal morbidity in cases of placenta accreta
comes. Undoubtedly, ultrasound is the managed by a multidisciplinary care team com-
primary and preferred modality for ante- pared with standard obstetric care. Obstet Gyne-
natal diagnosis of PAS. Efforts to im- col. 2011;117:331–337.
14. Warshak CR, Eskander R, Hull AD, et al.
prove the diagnostic accuracy of
Accuracy of ultrasonography and magnetic reso-
ultrasound will require standardization nance imaging in the diagnosis of placenta accre-
of the definition of markers and of the ta. Obstet Gynecol. 2006;108:573–581.
sonographic approach to diagnosis. A 15. O’Brien JM, Barton JR, Donaldson ES. The
multisociety task force has been as- management of placenta percreta: conservative
sembled to standardize PAS markers and operative strategies. Am J Obstet Gynecol.
1996;175:1632–1638.
and provide the optimal approach for 16. Ballas J, Pretorius D, Hull AD, et al. Identifying
prenatal diagnosis. sonographic markers for placenta accrete in the first
trimester. Am J Obstet Gynecol. 2012;31:1835–1841.
17. Comstock CH, Bronsteen RA. The antenatal
diagnosis of placenta accreta. BJOG. 2014;121:
References 171–182.
1. Usta IM, Hobeika EM, Abu Musa AA, et al. 18. Comstock CH, Lee W, Vettraino IM, et al. The
Placenta previa-accreta: risk factors and compli- early sonographic appearance of placenta accreta.
cations. Am J Obstet Gynecol. 2005;193(suppl): J Ultrasound Med. 2003;22:19–23.
1045–1049. 19. Jurkovic D, Hillaby K, Woelfer B, et al. First-
2. Hertig IC. A study of placenta accreta. J Surg trimester diagnosis and management of pregnan-
Gynecol Obstet. 1937;64:178–200. cies implanted into the lower uterine segment
3. Shellhaas CS, Gilbert S, Landon MB, et al. The cesarean section scar. Ultrasound Obstet Gynecol.
frequency and complication rates of hysterectomy 2003;21:220–227.
accompanying cesarean delivery. Obstet Gynecol. 20. Timor-Tritsch IE, Cali G, Monteagudo A, et al.
2009;114 (pt 1):224–229. Foley balloon catheter to prevent or manage
4. Obstetric care consensus no. 2: levels of maternal bleeding during treatment for cervical and cesar-
care. Obstet Gynecol. 2015;125:502–515. ean scar pregnancy. Ultrasound Obstet Gynecol.
5. Wehrum MJ, Buhimschi IA, Salafia C, et al. 2015;46:118–123.
Accreta complicating complete placenta previa is 21. Jiang T, Liu G, Huang L, et al. Methotrexate
characterized by reduced systemic levels of vas- therapy followed by suction curettage followed by
cular endothelial growth factor and by epithelial- foley tamponade for cesarean scar pregnancy. Eur
to-mesenchymal transition of the invasion troph- J Obstet Gynecol. 2011;156:209–211.
oblast. Am J Obstet Gynecol. 2011;204:e1–e411. 22. Chen CH, Wang PH, Lin JY, et al. Uterine rupture
6. Tantbirojn P, Crum CP, Parast MM. Pathophysi- secondary to placenta percreta in a near-term
ology of placenta accrete: the roll of deciduas pregnant woman with a history of hysterotomy.
and extravillous trophoblast. Placenta. 2008;29: J Obstet Gynaecol Res. 2011;37:71–74.
639–645. 23. Wong HS, Cheung YK, Zuccollo J, et al. Evaluation
7. Strickland S, Richards WG. Invasion of the of sonographic diagnostic criteria for placenta accreta.
trophoblast. Cell. 1992;71:355–357. J Clin Ultrasound. 2008;36:551–559.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Imaging of Placenta Accreta Spectrum 765

24. Yang JI, Kim HY, Kim HS, et al. Diagnosis in 37. Rac MWF, Dasche JS, Wells E, et al. Ultrasound
the first trimester of placenta accreta with pre- predictors of placental invasion: the placenta
vious cesarean section. Ultrasound Obstet Gyne- accreta index. Am J Obstet Gynecol. 2015;212:
col. 2009;34:116–118. 343.e1–343.e7.
25. Shih JC, Cheng WF, Shyu MK, et al. Power 38. Chou MM, Ho ES, Lee YH. Prenatal diagnosis of
Doppler evidence of placenta accreta appearing in placenta previa accrete by transabdominal color
the first trimester. Ultrasound Obstet Gynecol. Doppler ultrasound. Ultrasound Obstet Gynecol.
2002;19:623–625. 2000;15:28–35.
26. Comstock CH, Love JJ, Bronsteen RA, et al. 39. Lerner JP, Deane S, Timor-Tritsch IE. Charcte-
Sonographic detection of placenta accreta in the rization of placenta accrete using transvaginal
second and third trimesters of pregnancy. Am J sonography and color Doppler imaging. Ultra-
Obstet Gynecol. 2004;190:1135–1140. sound Obstet Gynecol. 1995;5:198–201.
27. Gielchinsky Y, Mankuta D, Rojansky N, et al. 40. Levine D, Hulka CA, Ludmir J, et al. Placenta
Perinatal outcome of pregnancies complicated by accreta: evaluation with color doppler US, power
placenta accreta. Obstet Gynecol. 2004;104:527–530. Doppler US, and MR imaging. Radiology. 1997;
28. Wong HS, Cheung YK, Strand L, et al. Specific 205:773–776.
sonographic features of placenta accrete: tissue 41. Kirkinen P, Helin-Martikainin H, Vanninen R,
interface disruption on gray-scale imaging and et al. Placenta accrete: imaging by gray-scale and
evidence of vessels crossing interface-disruption contrast-enhanced color Doppler sonography and
sites on Doppler imaging. Ultrasound Obstet magnetic resonance imaging. J Clin Ultrasound.
Gynecol. 2007;29:239–241. 1998;26:90–94.
29. Twickler DM, Lucas MJ, Balis AB, et al. Color 42. Collins S, Gordon S, Adbula A, et al. Three-
flow mapping for myometrial invasion in women dimensional power Doppler ultrasonography for
with a prior cesarean delivery. J Matern Fetal diagnosing abnormally invasive placenta and
Neonatal Med. 2000;9:330–335. quantifying the risk. Obstet Gynecol. 2015;126:
30. Comstock CH. Antenatal diagnosis of placenta 645–665.
accreta: a review. Ultrasound Obstet Gynecol. 2005; 43. Shih JC, Palacios Jaraquemada JM, Su YN, et al.
26:89–96. Role of three-dimensional power Doppler in the
31. Hull AD, Salerno CC, Saenz CC, et al. Three- antenatal diagnosis of placenta accreta: comparison
dimensional ultrasonography and diagnosis of with gray-scale and color Doppler techniques.
placenta percreta with bladder involvement. Ultrasound Obstet Gynecol. 2009;33:193–203.
J Ultrasound Med. 1999;18:853–856. 44. American College of Obstetricians and Gynecol-
32. Baughman WC, Corteville JE, Shah RR. Placen- ogists (ACOG). Placenta accreta (committee
ta accreta: spectrum of US and MR imaging opinion; no. 529). Obstet Gynecol. 2012;120:
findings. Radiographics. 2008;28:1905–1916. 207–211.
33. Cali G, Giambanco L, Pucchio G, et al. Morbidly 45. D’Antonio F, Iacovella C, Bhide A. Prenatal
adherent placenta:evaluation of ultrasound diag- identification of invasive placentation using
nostic criteria and differentiation of placenta ultrasound: systematic review and meta-analysis.
accreta from percreta. Ultrasound Obstet Gynecol. Ultrasound Obstet Gynecol. 2013;42:509–517.
2013;41:406–412. 46. Bowman ZS, Eller AG, Kennedy AM, et al.
34. Finberg HJ, Williams JW. Placenta accreta: Interobserver variability of sonography for pre-
prospective sonographic diagnosis in patients diction of placenta accreta. J Clin Ultrasound.
with placenta previa and prior cesarean section. 2014;33:2153–2158.
J Ultrasound Med. 1992;11:333–343. 47. Collins SL, Ashcroft A, Braun T, et al. On behalf of
35. Hudon L, Belfort MA, Broome DR. Diagnosis the European working group on abnormally invasive
and management of placenta percreta: a review. placenta. Proposal for standardized descriptors of
Obstet Gynecol Surv. 1998;53:509–517. abnormally invasive placenta (EW-AIP). Ultrasound
36. Royal College of Obstetricians and Gynaecologists. Obstet Gynecol. 2016;47:271–275.
Placenta Praevia, Placenta Praevia Accreta and 48. Rezk MA, Shawky M. Grey-scale and colour Dop-
Vasa Praevia: Diagnosis and Management. Lon- pler ultrasound versus magnetic resonance imaging
don, England: Royal College of Obstetricians and for the prenatal diagnosis of placenta accreta.
Gynaecologists; 2011:26; Green-top guideline 27. J Matern Fetal Neonatal Med. 2014;29:e1–e4.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA CLINICAL OBSTETRICS AND GYNECOLOGY
Volume 61, Number 4, 766–773
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.

Antenatal
Management of
Downloaded from [Link] by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3dW0s7N5CLiYMlH4wgaY34Fr69rtLH3oWi4+h6VxXUqo= on 10/30/2018

Placenta Accreta
CAROLYN HAUNSCHILD, MD,
AMANDA YEATON-MASSEY, MD,
and DEIRDRE J. LYELL, MD
Lucile Packard Children’s Hospital at Stanford, Palo Alto,
California

Abstract: Predelivery diagnosis of placenta accreta, surgical team, an intensive care unit (ICU),
increta, and percreta (from here referred to as placenta and transfusion services.3
accreta, unless otherwise noted) has increasingly
created opportunities to optimize antenatal manage- A thorough review of the patient’s
ment. Despite the increased frequency of placenta prior history, including prior spontaneous
accreta today, occurring in as many as 1 in 533 to 1 in preterm birth and other medical or surgi-
272 deliveries, high-quality data are lacking for many cal comorbidities, should be conducted,
aspects of antenatal management. This chapter will and conditions managed as indicated. An
discuss antenatal management of, and risks faced by,
women with suspected placenta accreta, a condition extensive discussion should be held with
that most frequently requires a potentially morbid the patient and her partner/family (as
cesarean hysterectomy. appropriate) to discuss the severity of
Key words: accreta, antenatal management, delivery the diagnosis and plan of care.
timing, cesarean hysterectomy Pelvic examinations, sexual inter-
course, and rigorous activity should gen-
The incidence of placenta accreta is increas- erally be avoided. Bed rest has not been
ing and might be as high as 1 in 5331 to 1 in found to decrease pregnancy complica-
2722 deliveries. Broadly, once placenta ac- tions and may actually increase the risk of
creta is suspected by adequate imaging venous thromboembolism, decondition-
studies, women with suspected placenta ing, and psychosocial suffering.4–6 Ane-
accreta should be evaluated by a Center of mia should be assessed for and corrected,
Excellence as early as possible to establish a and regardless, oral iron supplementation
plan for care, which may include coordi- should be given to maximize iron stores.7
nated care for those who live far from the
hospital. A Center of Excellence, described
below, consists of at least a multidisciplinary Delivery Location/
Multidisciplinary Care
Correspondence: Deirdre J. Lyell, MD, 300 Pasteur Drive, Women with suspected placenta accreta
MC 5317 Palo Alto, CA. E-mail: dlyell@[Link] should be delivered in a center capable of
The authors declare that they have nothing to disclose. providing care at the level of a “Center of

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 61 / NUMBER 4 / DECEMBER 2018

766 | [Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Antenatal Management of Placenta Accreta 767

Excellence” with multidisciplinary care, delivery at 34 weeks was associated with

consistent experience with accreta deliv- 50% reductions in estimated blood loss,
eries, intensive care facilities, and signifi- transfused packed red blood cells, and
cant transfusion services, as such care has ICU admission.10
been associated with less maternal hem- In our institution, before delivery a
orrhage and better outcomes.3 HIPPA-compliant notification email is
Multidisciplinary care is central to safe sent out to the multidisciplinary care
delivery of women with placenta accreta. team, which also includes obstetric nurs-
It is suggested that such care should ing, placental pathology, and leadership
consist of an experienced maternal-fetal of the main operating room.
medicine physician or obstetrician, pelvic As imaging studies sometimes raise
surgeon (gynecologic oncology or urogy- concern for accreta and may not defini-
necology), anesthesiologist (obstetric or tively rule it out, some women by neces-
cardiac anesthesia), urologist, trauma or sity will be delivered in a specialized
general surgeon, imaging experts (ultra- location despite being found at delivery
sound), interventional radiologist, and not to have accreta.
neonatologist.3 The facilities should in-
clude 24-hour availability of intensive
care specialists, neonatal intensive care
services appropriate for the gestational Antepartum Admission
age of the neonate, and blood services that Data guiding planned antepartum hospi-
include capabilities for massive transfusion, tal admission are limited and there are a
cell saver and perfusionists, experience and wide variety of practice patterns through-
access to alternative blood products, and out the United States. Antepartum ad-
guidance of transfusion medicine special- mission may be beneficial for delivery
ists or blood bank pathologists.3 planning including consultation with
Compared with a standard hospital, multidisciplinary team members, admin-
care by a multidisciplinary surgical team, istration of antenatal steroids, and opti-
defined by 24-hour in-house obstetrician mization of maternal comorbidities. At
gynecologists, anesthesiologists, fully our institution we admit women with
stocked blood banks, immediate avail- placenta accreta to the antepartum service
ability of a gynecologic oncologist, and if they experience bleeding or contrac-
interventional radiology, was associated tions, or at least a day before delivery to
with a 5-fold reduction in composite early facilitate delivery planning. The timing of
maternal morbidity [odds ratio (OR), planned antepartum admission is influ-
0.22; 95% confidence interval (CI), 0.07- enced by the patient’s proximity to the
0.70], less transfusion > 4 units of packed hospital, compliance and reliability with
red blood cells (43% vs. 61%, P = 0.031), prenatal care, and concerns for preterm
and less reoperation within 7 days for labor and expected morbidity of surgery.
bleeding (3% vs. 36%, P < 0.001).8 Fur- Women with severe placenta percreta
ther, a multidisciplinary approach for with previa are often electively admitted
women with planned cesarean hysterec- between 32 and 34 weeks gestation for
tomy at 34 to 35 weeks due to morbidly proximity to the operating room should
adherent placenta, was associated with they need. Relocating the patient to tem-
less median estimated blood loss (2.1 vs. porary housing near the hospital early in
3.0 L, P = 0.025) and need for emergency pregnancy should be considered for wom-
surgery (23% vs. 64%, P = 0.001)9 Others en who live far from the hospital, or for
have shown that implementation for a whom rapid transport to the hospital is
multidisciplinary approach with planned not possible.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
768 Haunschild et al

Given the increased risk of venous throm- with placenta previa and sonographic
boembolism with antepartum admission,11 evidence of placenta accreta utilizing a
mechanical deep venous thrombosis pro- decision analysis approach. The authors
phylaxis should be utilized, and ambulation estimated maternal and neonatal risks
encouraged when possible. Intravenous ac- and benefits and examined 9 different
cess should be considered for all women strategies. The strategy with the highest
hospitalized due to symptoms such as con- quality-adjusted life years and the pre-
tractions or bleeding. ferred strategy in most situations was
delivery at 34 weeks after administration
of betamethasone (without confirmation
Timing of Delivery of fetal lung maturity).
The optimal timing of delivery for women Although the risk of hemorrhage during
with suspected placenta accreta must bal- pregnancy or delivery is thought to increase
ance the risk of catastrophic maternal hem- with gestational age among women with
orrhage inherent to ongoing pregnancy and placenta previa and accreta,12 there are
its maternal/fetal sequelae with potential some data to suggest that a subset of
complications of prematurity. Goals of de- women without contractions, bleeding, or
livery timing further include avoidance of an preterm premature rupture of membranes
unscheduled, emergent delivery, and delivery (PPROMs) may be expectantly managed
in a location without ideal resources. until 36 weeks.14
In 2011 the Eunice Kennedy Shriver Without any randomized controlled
National Institute of Child Health and trials or well-controlled observational
Human Development (NICHD) and So- studies to identify best practices for deliv-
ciety for Maternal-Fetal Medicine ery timing,7 recommendations must be
(SMFM) issued a guideline that the opti- extrapolated from imperfect data, balanc-
mal timing of delivery for women with ing assumptions about degree of cata-
suspected placenta accreta with placenta strophic maternal risk with the intended
previa is 34 0/7 and 35 6/7 weeks.12 This benefit of improved neonatal outcome.
guideline was based on consensus opinion
after review of limited available data,
weighing the maternal, fetal, and neo-
natal risks and benefits of ongoing preg- Scheduled Cesarean
nancy versus delivery. In issuing the Hysterectomy
guideline, the authors acknowledged the Cesarean hysterectomy for placenta ac-
importance of individualized clinical creta is morbid enough that early sched-
management that includes specific mater- uled delivery is recommended to avoid the
nal and fetal risks, comorbidities, assess- additional morbidities associated with
ment of available resources of the practice potential emergent cesarean hysterec-
setting, and patient preferences. Retro- tomy, or delivery at an ill equipped
spective studies upon which the delivery hospital. Peripartum hysterectomy com-
timing recommendation is based may not pared with nonobstetric hysterectomy has
account for many other determinants of higher rates of bladder injury (9% vs. 1%),
patient outcome, including the type of ureteral injury (0.7% vs. 0.1%), reopera-
accreta, delivery circumstances, resource tion (4% vs. 0.5%), transfusion (46% vs.
availability, or patient factors which may 4%), and wound complications (10% vs.
alter morbidity, and as such, care should 3%).15 Women undergoing peripartum
be individualized. hysterectomy for the indication placenta
Robinson and Grobman13 examined accreta compared with uterine atony
various delivery strategies for women experienced 2- to 3-fold higher risks of

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Antenatal Management of Placenta Accreta 769

organ injury to the bladder and ureter,15 earlier gestational age when complicated
with cystotomy seen up to 50% of women by PPROM (30.4 vs. 34.4 wk, P = 0.003)
with placenta percreta.16 and antenatal contractions (33.3 vs.
Women with placenta accreta under- 34.2 wk, P = 0.25). History of ≥ 2 cesar-
going unscheduled, urgent cesarean ean deliveries was also a risk for urgent
hysterectomy may experience more mor- delivery (OR, 4.9; 95% CI, 1.5-16.0;
bidity than scheduled, nonurgent cesar- adjusted OR, 11.4; 95% CI, 1.8-71.1;
ean hysterectomy. In a recent study, P = 0.01).17 Traditional predictors of pre-
women with placenta accreta undergoing term delivery including short cervical
urgent delivery compared with planned length, history of prior spontaneous pre-
delivery had higher composite maternal term delivery, and multiple gestation are
morbidity and mortality (57% vs. 37%, also thought to increase the risk of urgent
P = 0.02).17 In another study, although delivery.20
results did not achieve significance, wom- Risk stratification may help identify
en undergoing unplanned cesarean hys- women at significant risk for early deliv-
terectomy experienced more blood loss ery in patients with antenatally suspected
(3.0 vs. 2.1 L), massive transfusion (43% placenta accreta.20 A recent retrospective
vs. 32%), postoperative complications study stratified women into low and high
(early morbidity, 57% vs. 37%), and risk for preterm delivery. High risk was
ICU admission (31% vs. 23%) than wom- defined as having any one of the following
en who undergo scheduled cesarean risks: bleeding or preterm labor before 34
hysterectomy.18 In addition, emergent weeks, PPROM, or a known risk factor
surgery is associated with higher rates of for preterm delivery (prior spontaneous
postpartum depression and posttraumatic preterm birth, any cervical length meas-
stress disorder.19 urement <2.5 cm, or multiple gestations).
Among “low-risk” women, 3% under-
went unscheduled delivery before 36
weeks, and none underwent emergent
Risk Factors for Unscheduled delivery. In contrast, 19% of “high-risk”
Cesarean Hysterectomy women underwent unscheduled delivery
Several identified risks for early or un- before 36 weeks and 4% were emergent
scheduled cesarean hysterectomy can in- (P = 0.05).
form decisions with regard to timing of
delivery. In a retrospective study of wom-
en with antenatally suspected placenta
accreta, risk factors for unscheduled de- Maternal Hemorrhage and its
livery included vaginal bleeding (63%), Sequelae
uterine contractions (32%), and preterm Hemorrhage is the most common cause of
prelabor rupture of membranes (5%). morbidity and mortality associated with
Each episode of vaginal bleeding during placenta accreta. The risk of severe post-
pregnancy was associated with a 3.8-fold partum hemorrhage is greatest with placen-
increased risk of unscheduled delivery, ta accreta even after controlling for other
and more so in the setting of contractions risk factors.21 Women with placenta accreta
and PPROM.14 Delivery occurred at on average lose between 2 and 3 L of blood
mean gestational ages of 32.9 weeks in at delivery22–24 representing ~40% to 60%
women with antenatal bleeding, com- of a women’s total blood volume. Such
pared with 35.8 weeks in women without massive blood loss necessitates transfusion
episodes of bleeding (P < 0.001). In this in 90% to 95% of women undergoing
same study, pregnancies also ended at an cesarean hysterectomy.25,26 Among women

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
770 Haunschild et al

receiving a blood transfusion, 5% to 40% with placenta accreta of 12.5% by 36

require a massive transfusion defined as weeks, 20.1% by 38 weeks, and 24.4%.
≥ 10 units of packed red cells.10,25,26 In a Most authors report an increased risk of
case series by Stotler et al26 of 66 women bleeding with advancing gestational age
undergoing delivery for placenta accreta, among women with placenta accreta or
11% required ≥ 20 units of packed red cells. percreta.10,12–14,25,33 Warshak and col-
Hemorrhage can lead to severe maternal leagues report a 35% rate of emergent
morbidity including hypoxic encephalop- delivery before their scheduled delivery.
athy, acute tubular necrosis, respiratory This was most frequently for bleeding
distress, and hemorrhagic shock. Data are (22%).33 Bowman et al14 reported a 31%
conflicting with regard to whether the rate of delivery for bleeding before a
degree of placenta accreta (ie, accreta vs. scheduled 36 week delivery among 77
increta vs. percreta) is associated with the women, with 12% delivering early due to
amount of blood product required. Pani- labor. The Society for Perinatal Obstetri-
grahi et al27 reviewed transfusion require- cians conducted a survey among their
ments for a total of 136 patients with members and found that for patients
pathologically confirmed placenta accreta > 35 weeks gestation 93% were delivered
delivered at a single institution and did not for bleeding and 4 of 8 maternal deaths
find a difference in transfusion requirement occurred after 36 weeks.25
based on degree of placental invasion.

Transfusion Medicine Support

Timing of Hemorrhage Predelivery planning anticipating the
The ideal delivery for pregnancies com- need for transfusion (and possibly mas-
plicated by placenta accreta is a planned sive transfusion) is imperative. Delivery
cesarean with hysterectomy as needed planning should involve coordination
that both minimizes maternal hemorrha- with transfusion medicine and occur at a
gic morbidity and neonatal morbidity center with a well stocked blood bank
related to prematurity. The challenge where large volume blood loss can be
however is identifying who will experience safely treated. All women should be as-
antepartum bleeding and at what gesta- sessed for antibodies to allow for appro-
tional age bleeding will occur. Data are priate cross matching of blood products
lacking to provide definitive recommen- before delivery. Predelivery planning thus
dations with regard to when to deliver in prevents isoimmunization that may occur
order to avoid hemorrhage and its seque- with uncrossmatched blood and also pre-
lae. Several groups have described an vents overutilization of type O negative
association between shorter cervical products.
length (< 3 cm) with risk of antepartum
bleeding and earlier delivery among wom-
en with placenta previa; these data may be Fetal/Neonatal Morbidity
extrapolated to women with placenta Studies have been mixed with regard to
accreta.28–32 However, even among wom- whether placenta accreta leads to fetal
en with a cervical length > 3 cm and complications. Some have suggested no
placenta previa, 28% required urgent increase in stillbirth or intrauterine growth
delivery for antepartum hemorrhage.28 restriction,34 whereas others have found
Irrespective of cervical length, the risk that morbidly adherent placenta independ-
of bleeding increases with gestational age. ently increased fetal and neonatal risks,
A study by Rac et al22 reported rates of including low-birth weight, APGAR scores
urgent delivery for bleeding in women <7, total perinatal mortality (6.7%)35 and

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Antenatal Management of Placenta Accreta 771

small for gestational age.36 The rate of before scheduled preterm birth, and is
stillbirth was increased 4.7-fold (95% CI, now recommended for women at risk for
2.4-9.1) in a population-based study of preterm birth as late as 36 5/7 weeks.
morbidly adherent placenta that relied Gyamfi-Bannerman et al42 identified that
upon ICD-10 discharge codes for the newborns of women at risk for late
diagnosis.37 Given these potential risks, preterm birth who received ACS (beta-
we initiate antepartum fetal monitoring methasone) experienced significantly less
and fetal growth assessment ultrasounds respiratory morbidity, (11.6% vs. 14.4%;
among women with suspected placenta RR, 0.8; 95% CI, 0.66-0.97; P = 0.02).
accreta. Respiratory morbidity was defined by a
Most neonatal morbidity among preg- composite outcome of the need for respi-
nancies complicated by placenta accreta ratory support within the first 72 hours of
results from iatrogenic prematurity, and life and any use of continuous positive
morbidity can be quite significant. Given airway pressure or high-flow nasal can-
the risks for neonatal morbidities among nula for at least 2 consecutive hours,
preterm and late preterm newborns, de- supplemental oxygen with a fraction of
livery should ideally occur in a facility oxygen of at least 30% or at least 4
with immediate access to neonatal inten- consecutive hours, extracorporeal mem-
sive care services. brane oxygenation, or mechanical venti-
Respiratory morbidity is the most fre- lation, as well as stillbirth or neonatal
quently encountered neonatal morbidity death. Neonatal hypoglycemia was more
at 34 weeks, as surfactant-producing type frequent among neonates whose mothers
II pneumocytes begin to dominate start- received betamethasone (glucose <40 mg/
ing between 34 and 36 weeks.38 The risk dL, 24.0% betamethasone vs. 14.9% pla-
for respiratory distress/hyaline membrane cebo; RR, 1.61; 95% CI, 1.38-1.88).
disease decreases significantly with each It is also important to note that women
week of gestational age past 34 weeks, were excluded from this study if they had
with a 22-fold reduction by 35 weeks and pregestational diabetes, had previously
a 9-fold reduction by 36 weeks.39 Planned received ACS, were expected to deliver
cesarean delivery itself increases the risk within 12 hours, or had nonreassuring
for neonatal respiratory morbidity, inde- fetal status. The Society for Maternal-
pendent of prematurity.40 Fetal Medicine has recommended that the
Other neonatal morbidities are more same criteria be used when considering
frequent in the late preterm compared whether to administer ACS, unless as part
with term deliveries. In a systemic review of research or a quality improvement
of 22 studies of nearly 30 million infants, project. Women undergoing unscheduled
late preterm (34 wk 0/7 days’ gestation to delivery at any gestational age are pre-
36 wk 6/7 days’ gestation) compared with sumably doing so for urgent reasons, to
term births were associated with not only minimize maternal and neonatal morbid-
more respiratory distress syndrome (RR, ity. Delivery in urgent situations should
17.3) and need for mechanical ventilation not be delayed for ACS administration.
or intubation (RR, 4.9), but also more As predelivery diagnosis of placenta
intraventricular hemorrhage (RR, 4.9), accreta improves, so does the opportunity
necrotizing entercolitis (RR, 7.5), and to minimize maternal morbidity through
neonatal death (RR, 5.9).41 appropriate antenatal care, which is sig-
Antenatal corticosteroid (ACS) admin- nificantly underscored by scheduled deliv-
istration has been shown to reduce ery in a well-equipped facility, with a
neonatal respiratory morbidity, should highly experienced multidisciplinary team,
be administered ideally at least 48 hours neonatal services, and at an optimal time.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
772 Haunschild et al

References 16. Woldu S, Ordonez M, Devine P, et al. Urologic

considerations of placenta accreta: a contempo-
1. Wu S, Kocherginsky M, Hibbard J. Abnormal rary tertiary care institutional experience. Urol
placentation: twenty-year analysis. Am J Obstet Int. 2014;93:74–79.
Gynecol. 2005;192:1458–1461.
17. Shamshirsaz A, Fox K, Erfani H, et al. Outcomes
2. Mogos MF, Salemi JL, Ashley M, et al. Recent
of planned compared with urgent deliveries using
trends in placenta accreta in the United States and
a multidisciplinary team approach for morbidly
its impact on maternal-fetal morbidity and health-
adherent placenta. Obstet Gynecol. 2018;131:
care-associated costs, 1998–2011. J Matern Fetal
234–241.
Neonatal Med. 2016;29:1077–1082.
18. Eller A, Porter T, Soisson P, et al. Optimal
3. Silver RM, Fox K, Barton J, et al. Center of
management strategies for placenta accreta.
excellence for placenta accreta. Am J Obstet
BJOG. 2009;116:648–654.
Gynecol. 2015;212:561–568.
19. de la Cruz C, Coulter M, O’Rourke K, et al.
4. Kovacevich GJ, Gaich S, Lavin JP, et al. The
Posttraumatic stress disorder following emer-
prevalence of thromboembolic events among
gency peripartum hysterectomy. Arch Gynecol
women treated with extended bed rest prescribed
Obstet. 2016;294:681–688.
as part of the treatment for premature labor or
preterm premature rupture of the membranes. Am 20. Perlman N, Little S, Thomas A, et al. Patient
J Obstet Gynecol. 2000;182:1089–1092. selection for later delivery timing with suspected
5. May KA. Impact of prescribed activity restriction previa-accreta. Acta Obstet Gynecol Scand. 2017;96:
during pregnancy on women and families. Health 1021–1028.
Care Women Int. 2001;22:29–47. 21. Ekin A, Gezer C, Solmaz U, et al. Predictors of
6. Maloni JA, Brezinski-Tomasi JE, Johnson LA. severity in primary postpartum hemorrhage. Arch
Antepartum bed rest: effect upon the family. Gynecol Obstet. 2015;292:1247–1254.
J Obstet Gynecol Neonatal Nurs. 2001;30:165–173. 22. Rac MW, Wells CE, Twickler DM, et al. Placenta
7. Belfort M. Placenta accreta. Am J Obstet Gyne- accreta and vaginal bleeding according to gesta-
col. 2010;203:430–439. tional age at delivery. Obstet Gynecol. 2015;125:
8. Eller A, Bennett M, Sharshiner M, et al. Maternal 808–813.
morbidity in cases of placenta accreta managed 23. Bailit J, Grobman W, Rice M, et al. Morbidly
by a multidisciplinary care team compared with adherent placenta treatments and outcomes.
standard obstetric care. Obstet Gynecol. 2011;117: Obstet Gynecol. 2015;125:683–689.
331–337. 24. Wright J, Pri-Paz S, Herzog T, et al. Predictors of
9. Shamshirsaz A, Fox K, Salmanian B, et al. massive blood loss in women with placenta accreta.
Maternal morbidity in patients with morbidly Am J Obstet Gynecol. 2011;205:38.e1–38.e6.
adherent placenta treated with and without a 25. O’Brien J, Barton J, Donaldson E. The management
standardized multidisciplinary approach. Am J of placenta percreta: conservative and operative
Obstet Gynecol. 2015;212:218.e1–218.e9. strategies. Am J Obstet Gynecol. 1996;175:1632–1638.
10. Al-Khan A, Gupta V, Illsley N, et al. Maternal 26. Stotler B, Padmanabhan A, Devine P, et al.
and fetal outcomes in placenta accreta after Transfusion requirements in obstetric patients
institution of team-managed care. Reprod Sci. with placenta accreta. Transfusion. 2011;51:
2014;21:761–771. 2627–2633.
11. Sultan AA, West J, Tata LJ, et al. Risk of first 27. Panigrahi A, Yeaton-Massey A, Bakhtary S, et al.
venous thromboembolism in pregnant women in A standardized approach for transfusion medi-
hospital: population based cohort study from cine support in patients with morbidly adherent
England. BMJ. 2013;347:1–11. placenta. Anesth Analg. 2017;125:603–608.
12. Spong CY, Mercer BM, D’alton M, et al. Timing 28. Stafford I, Dashe J, Shivvers S, et al. Ultrasono-
of indicated late-preterm and early-term birth. graphic cervical length and risk of hemorrhage in
Obstet Gynecol. 2011;118:323–333. pregnancies with placenta previa. Obstet Gynecol.
13. Robinson B, Grobman W. Effectiveness of timing 2010;116:595–600.
strategies for delivery of individuals with placenta 29. Ghi T, Contro E, Martina T, et al. Cervical length
previa and accreta. Obstet Gynecol. 2010;116: and risk of antepartum bleeding in women
835–842. with complete placenta previa. Ultrasound Obstet
14. Bowman Z, Manuck T, Eller A, et al. Risk factors Gynecol. 2009;33:209–212.
for unscheduled delivery in patients with placenta 30. Sekiguchi A, Nakai A, Okuda N, et al. Consec-
accreta. Am J Obstet Gynecol. 2014;210:241. utive cervical length measurements as a predictor
e1–241.e6. of preterm cesarean section in complete placenta
15. Wright J, Devine P, Shah M, et al. Morbidity and previa. J Clin Ultrasound. 2015;43:17–22.
mortality of peripartum hysterectomy. Obstet 31. Zaitoun M, El Behery M, Abd El Hameed A,
Gynecol. 2010;115:1187–1193. et al. Does cervical length and the lower placental

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Antenatal Management of Placenta Accreta 773

edge thickness measurement correlates with clin- 37. Upson K, Silver R, Greene R, et al. Placenta
ical outcome in cases of complete placenta previa? accreta and maternal morbidity in the Republic of
Arch Gynecol Obstet. 2011;284:867–873. Ireland, 2005-2010. J Matern Fetal Neonatal
32. Fukushima K, Fujiwara A, Anami A, et al. Med. 2014;27:24–29.
Cervical length predicts placental adherence and 38. Mahoney A, Jain L. Respiratory disorders in
massive hemorrhage in placenta previa. J Obstet moderately preterm, late preterm, and early term
Gynaecol Res. 2012;38:192–197. infants. Clin Perinatol. 2013;40:665–678.
33. Warshak CR, Ramos GA, Eskander R, et al. 39. Hibbard J, Wilkins I, Sun L, et al. Respiratory
Effect of predelivery diagnosis in 99 consecutive morbidity in late preterm births. JAMA. 2010;
cases of placenta accreta. Obstet Gynecol. 2010; 304:419–425.
115:65–69. 40. Ryan C, Hughes P. Neonatal respiratory morbid-
34. Usta IM, Hobeika EM, Musa AA, et al. Placenta ity and mode of delivery at term: influence of
previa-accreta: risk factors and complications. timing of elective caesarean section. Br J Obstet
Am J Obstet Gynecol. 2005;193:1045–1049. Gynaecol. 1995;102:843–844.
35. Vinograd A, Wainstock T, Mazor M, et al. 41. Teune M, Bakhuizen S, Gyamfi Bannerman C,
Placenta accreta is an independent risk factor et al. A systematic review of severe morbidity in
for late pre-term birth and perinatal mortality. infants born late preterm. Am J Obstet Gynecol.
J Matern Fetal Neonatal Med. 2015;28: 2011;205:374.e1–374.e9.
1381–1387. 42. Gyamfi-Bannerman C, Thom E, Blackwell S.
36. Gielchinsky Y, Mankuta D, Rojansky N, et al. Antenatal betamethasone for women at risk for
Perinatal outcome of pregnancies complicated by late preterm delivery. N Engl J Med. 2016;374:
placenta accreta. Obstet Gynecol. 2004;104:527–530. 1311–1320.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA CLINICAL OBSTETRICS AND GYNECOLOGY
Volume 61, Number 4, 774–782
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.

Surgical Management
of Placenta Accreta
Downloaded from [Link] by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3dW0s7N5CLiYMlH4wgaY34Fr69rtLH3oWi4+h6VxXUqo= on 10/30/2018

Spectrum
BRETT D. EINERSON, MD, MPH,*†
and D. WARE BRANCH, MD*†
*Department of Obstetrics and Gynecology, Division of
Maternal-Fetal Medicine, University of Utah Health, Salt Lake City,
Utah; and †Intermountain Health Care, Salt Lake City, Utah

Abstract: This is a discussion of the standard surgical be the absence of the normal decidual
treatment of placenta accreta spectrum disorders basalis, usually due to uterine scarring after
including preoperative considerations, diagnostic
imaging, surgical steps for cesarean hysterectomy, surgical trauma, and the resulting abnor-
and postoperative management. mal invasion of the trophoblast into my-
Key words: placenta accreta spectrum, cesarean hys- ometrial tissues when the placenta implants
terectomy, ureteral stents at the site of previous scarring. The inci-
dence has increased over the last several
decades in association with increasing rates
of cesarean delivery. For practical pur-
Introduction and Scope of the poses, the likelihood of PAS depends upon
the number of previous cesarean deliveries
Problem a patient has undergone, and hence is
Those readers turning to this review are directly linked to parity. PAS was infre-
doubtlessly aware of the epidemic of pla- quent in the 1960s and 1970s, but recent
centa accreta spectrum (PAS), largely a data indicate a rate of ∼1 in 550 to 1 in 750
result of our legacy of the more liberal use deliveries.1 PAS is now the most common
of cesarean delivery over the last 2 decades. reason for hysterectomy associated with
From a histopathologic perspective, PAS both cesarean delivery2 and peripartum
includes the absence of the normal inter- hysterectomy.3
vening decidua and invasion of the placen- The challenge of managing PAS comes
ta into the myometrium (accreta and with delivery—serious hemorrhage may
increta) or through the myometrium (per- result from failure of the usual hemostatic
creta). The major etiology of PAS seems to mechanisms associated with separation of
Correspondence: Brett D. Einerson, MD, MPH, 30 N
the normally implanted placenta and/or
1900 E, Suite 2B200, Salt Lake City, UT. hypervascularity of adjacent tissues in-
E-mail: [Link]@[Link] volved in the process. Our experience
The authors declare that they have nothing to disclose. suggests that the potential for maternal

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 61 / NUMBER 4 / DECEMBER 2018

774 | [Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Surgical Management of Placenta Accreta 775

bleeding correlates with the (1) degree to operative management. The criteria re-
which the placenta has invaded into and quirements of a placenta accreta referral
“thinned” the underlying myometrium, center have previously been described by
(2) the area of abnormal placental attach- an expert consensus panel.6
ment involved, (3) the degree of hypervascu-
larity associated with adjacent tissues, and PREOPERATIVE IMAGING FOR
(4) the presence or absence of invasion into SURGICAL PLANNING
extrauterine tissues, No universally accepted, highly reliable
PAS is clearly associated with surgical radiographic staging system exists for
morbidity. Over half of women undergoing PAS, so the optimal choice of preopera-
planned cesarean hysterectomy require tive imaging is dependent on the expertise
transfusion of blood products and a third and preferences of local radiologists and
have incidental cystotomy. Ureteral injury surgeons. Although both ultrasound (US)
and the need for reoperation are less and magnetic resonance imaging (MRI)
frequent complications. As we will empha- have good diagnostic capability for
size in this review, the best maternal and PAS,7,8 each imaging modality is imper-
fetal outcomes are dependent upon appro- fect, and cases of inaccurate diagnosis—
priate antepartum diagnosis and care by an and associated surgical surprises—are
expert multidisciplinary team. fairly common. Cases of missed-diagnosis
or under-diagnosis may result in inad-
equate surgical preparation, putting pa-
Preoperative Considerations tients at risk of unnecessary morbidity.
In many ways the preoperative care of Cases of over-diagnosis may result in
patients with PAS mimics that for patients harm from unnecessary procedures and
with cancer, ideally involving a timely and hysterectomies. Multidisciplinary preop-
accurate diagnosis, appropriate referral to erative consensus conferences to review
a specialty center, thoughtful surgical prep- imaging and establish a surgical plan are
aration, and a multidisciplinary collabora- advised and may improve outcomes.
tive approach. Preoperative considerations Many centers use MRI in the preoper-
including the management of blood prod- ative planning for PAS surgery, but the
ucts and the use of preoperative arterial specifically defined value of the informa-
balloons or embolization to reduce hemor- tion MRI provides above and beyond a
rhage are discussed elsewhere in this issue. high-quality US is unclear. MRI, like US,
may be misleading in a surprisingly high
MULTIDISCIPLINARY APPROACH proportion of cases.9 Given the relatively
Although the incidence of PAS is increas- common occurrence of misdiagnosis with
ing, individual obstetricians and low- both US and MRI, the unclear benefit of
volume or mid-volume obstetric hospitals MRI as an adjunct to US, and the much
may only rarely be faced with a case of higher cost of MRI, we recommend using
PAS. Patient outcomes are improved US as the primary imaging modality for
both when the diagnosis is made before diagnosis and surgical planning for PAS.
delivery,4 and when the surgical team is Use of MRI should be limited to cases in
made up of multidisciplinary experts with which there is a specific indication and
experience in PAS treatment.5 As such, it likelihood of clear benefit.
is critically important that patients at risk One aspect of preoperative planning
for PAS (eg, ≥ 3 previous cesareans and that is particularly important is determin-
placenta previa) or with radiographically ing the extent of bladder involvement
suspected PAS be referred to PAS referral near the invading or adherent placental
centers for evaluation, counseling, and tissue. Knowledge of bladder involvement

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
776 Einerson and Branch

are superior to traditional surgical manage-

ment of PAS with regard to maternal mor-
bidity is unknown. In our experience, early
PAS surgery before the third trimester is
not technically easier or less morbid than if
performed in the mid third trimester.
Heavy bleeding in patients with sus-
pected PAS is a strong indication for
prompt delivery regardless of gestational
age. Whether transfer of care to a referral
center can be accomplished will depend on
the rate of hemorrhage, the multidiscipli-
nary capability of the admitting hospital,
and the timeliness of available transport.
These decisions should be made in concert
with accreta specialists at centers of
FIGURE 1. Endovaginal imaging of placenta excellence. Use of a checklist endorsed by
increta with Doppler. The placenta is seen Society for Maternal-Fetal Medicine may
bulging anteriorly toward the Bl. Bl indicates be helpful (Fig. 2).
bladder. To avoid these difficult and potentially
dangerous circumstances, we recommend
patients with suspected PAS who live
preoperatively affords the opportunity to > 45 minutes away from a facility fully
have urologic surgical specialists involved capable of managing all aspects of their
or available for the case. Some radiolog- care consider moving close to the facility at
ists maintain that endovaginal US pro- 30 to 32 weeks gestation, and earlier if
vides the clearest and most reliable vaginal spotting, sonographically short cer-
window for determining the extent of vix, or persistent preterm contractions are
bladder involvement before surgery re- present. Some centers use inpatient obser-
gardless of the patient’s body habitus vation between 30 and 34 weeks gestation.
(Fig. 1). In the absence of bleeding or labor we
perform planned cesarean hysterectomy
TIMING OF DELIVERY for PAS between 34 and 36 week. Ante-
With notable exceptions, most cases of natal corticosteroids should be adminis-
PAS in the United States are delivered tered before preterm delivery if possible to
between 32 and 37 weeks gestation. reduce complications of prematurity.
A goal of achieving at least 34 weeks
gestation is preferred from the standpoint
of neonatal prematurity, but expectant
management of PAS into the late third Surgical Approach to Cesarean
trimester puts the mother at risk of Hysterectomy
labor and hemorrhage. Delivery before Surgical cases for PAS are among the most
neonatal viability may be considered for difficult and life-threatening procedures
patients who desire early termination of performed. Unfortunately, few data are
pregnancy and in patients with severe available to directly inform the optimal
ongoing hemorrhage in the second tri- surgical approach to treatment of PAS;
mester. Whether dilation and evacuation thus the approach is often individualized to
in the second trimester or early hysterec- the patient, surgical team, and resources
tomy in the second or early third trimester available at the institution.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Surgical Management of Placenta Accreta 777

FIGURE 2. An example of a checklist for management of unexpected placenta accreta spectrum.

Planned cesarean hysterectomy is the other surgical approaches that conserve

preferred treatment for PAS in most of the uterus.
the United States. Here we describe the Cesarean hysterectomy should be per-
principles of our approach. Elsewhere in formed in an operating room with ready
this issue the reader can find a discussion access to all of the resources and staff
of alternative and experimental surgical necessary for the unique needs of patients
strategies including in situ conservative with PAS. In many hospitals, this means
management, delayed hysterectomy, and that PAS surgery should be performed in

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
778 Einerson and Branch

FIGURE 2. (Continued).

the main operating room rather than the Subjectively, stents allow for easier iden-
labor ward where there are competing tification and avoidance of the often
demands for the attention and resources atypically displaced ureters throughout
available. Cases of PAS often involve the the case.
care of dozens of health care personnel. Evidence does not favor a particular
The surgical suite should be large enough patient positioning or type of abdominal
to accommodate an anesthesia team, neo- incision. Placement of the patient in either
natal resuscitation staff, surgical specialists, supine or lithotomy positions is reason-
operative and surgical equipment, and able. The latter affords immediate and
several nursing and technical staff. Prox- direct access to the cervico-vaginal field
imity to the neonatal intensive care unit and easy manipulation of the bladder
and blood bank should be considered. catheter. A generous vertical midline
Ureteral stents can be placed before incision or Cherney incision allows for
laparotomy, and may be associated with adequate surgical exposure for exteriori-
reduced risk of ureteral injury and re- zation of the uterus and visualization of
duced overall operative complications.10 the pelvic operative field.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Surgical Management of Placenta Accreta 779

FIGURE 3. Evidence of placenta accreta at laparotomy. Signs of accreta may include bulging
and neovascularization of the lower uterine segment (A), or frank dehiscence of the placenta
through the wall of uterus at the previous cesarean scar site (B).

Before laparotomy, we review imaging this procedure can be found at [Link]

studies to identify the location of the ly/HysterotomyStaple.
placenta on the uterine wall so that the After hysterotomy the neonate is deliv-
placenta can be avoided during hystero- ered and handed to the neonatal resusci-
tomy. Upon entry into the abdomen a tation team. The placenta is left in situ
pelvic survey is performed to visually assess and care is taken to avoid disrupting the
the extent of placental invasion or dehis- placental bed. The umbilical cord is
cence through the previous uterine scar trimmed, tied, and tucked into the uterine
(Fig. 3). Intraoperative US can be used to cavity. At this point oxytocin is adminis-
map the position of the placenta. A classical tered to improve uterine tone. The hys-
hysterotomy is performed in a location well terotomy, traditionally cut or stapled, is
away from the site of the placenta. As most closed quickly with a running locked
cases of PAS are associated with previa, this suture to achieve hemostasis. In our ex-
incision is usually fundal and requires perience, packing the uterine cavity with
the exteriorization of the uterus through the lap sponges before uterine closure offers
abdominal incision before delivery of the no clear advantages and may reduce
neonate. exposure in the surgical field by expand-
Fundal hysterotomy can be the source ing the uterus. A self-retaining retractor
of significant blood loss ( > 500 to 1000 and abdominal packing are used to im-
mL) during the case, and may contribute prove surgical exposure and expand the
to the need for transfusion. As such, we operating field in preparation for hyster-
have adopted the approach described by ectomy.
Belfort et al11 to reduce blood loss during The most important surgical principle
this step using a linear stapling-cutting during hysterectomy for PAS is to isolate
device to make a hysterotomy and secure and secure the uterine circulation early in
hemostatic edges along the length of the the case and, if possible, before addressing
hysterotomy. A video and description of the area of highest concern for massive

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
780 Einerson and Branch

FIGURE 4. Thinned (A) and attenuated (B) uterine wall with underlying placental tissue.

hemorrhage—the lower uterus and adher- damage to the anteriorly located ureter,
ent bladder. which can be retracted medially, and
The hysterectomy proceeds by ligating posterolateral-located internal iliac vein.
and dividing the round ligament, below The utero-ovarian ligaments are iso-
which the avascular paravesical and para- lated, secured, and transected. Ovaries
rectal spaces can be accessed and the are preserved and packed into the
major pelvic vessels and ureters visual- abdomen.
ized. The ureters should be identified in Isolating and ligating the uterine arteries
the retroperitoneal space as they course and other uterine vasculature—which is
into the pelvis. The pelvic ureters can be often atypically and impressively dilated in
seen crossing over the bifurcation of the the setting of PAS—along the lateral sides
common iliac arteries from lateral to of the uterus can be particularly difficult.
medial then down along the pelvic side Securing this vasculature before proceeding
wall anterior and medial to the internal with a full bladder dissection is, in our
iliac arteries. The act of opening the experience, important in reducing the risk
retroperitoneal space below the round of massive hemorrhage, but some degree of
ligament draws the ureter medially, away lateral bladder dissection and ureterolysis is
from the lateral pelvic side wall and often needed to safely isolate the uterine
vessels. Preoperative placement of ureter- arteries. The uterine wall is often extremely
al stents can aid in the continuous identi- thinned along the lateral aspect and every
fication and easy palpation of the ureters attempt should be made to avoid disrupting
throughout the procedure. the underlying placenta (Fig. 4). The ex-
Ligation of the anterior division of the posed uterine arteries can then be clamped,
internal iliac artery (hypogastric artery) can transected, and ligated. Further dissection
be performed at this point to reduce pulse and securing of vessels along the cardinal
pressure and blood flow to the uterus ligaments to a level below the placenta is
through the uterine artery, theoretically then undertaken. It is at this point, after
reducing blood loss. Evidence to support securing the vasculature, that we proceed
this additional step is lacking with regard with bladder dissection (Fig. 5), a process
to blood loss or the need for transfusion, that can be tedious and bloody, and often
and operative time may be longer. If involves inadvertent cystotomy. Whether to
performed, care should be taken to avoid perform supracervical or total hysterectomy

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Surgical Management of Placenta Accreta 781

of cesarean hysterectomy for PAS: pre-

vention of postoperative venous throm-
boembolism (VTE) and postoperative
management of urinary tract injury.

VTE PROPHYLAXIS
For patients with PAS, risk factors for
postoperative VTE include pregnancy
and delivery, major open surgery, pro-
longed surgical time, and the postpartum
state. Other fairly frequent risk factors
include varicose veins, central venous
catheter access, bedrest, and excessive
blood loss with the need for transfusion.
The combination of any several of these
would indicate a Caprini score of ≥ 5,
suggesting a several percent risk of post-
operative VTE in the absence of throm-
FIGURE 5. Bladder dissection. boprophylactic measures, and perhaps a
risk in excess of 6%. Common additional
should be individualized to the patient, sur- risk factors for VTE in the PAS popula-
gical course, and operating surgeon. tion may include history of previous VTE,
When in some cases bleeding from prepregnancy obesity, and the presence of
the placenta is brisk and potentially life- a heritable thrombophilia.
threatening, a strategy of rapid clamping The American College of Chest Physi-
and cutting along the hysterectomy plane cians’ (ACCP) guidelines recommend
and across the uterine corpus just below pharmacologic and mechanical VTE pro-
the placenta may be necessary to secure the phylaxis general and abdominal-pelvic sur-
uterine blood flow and stop hemorrhage by gery patients at high risk for VTE, such as
removing the placenta and uterus from the the PAS patients, who are not at high risk
surgical field. After removal of the uterus for major bleeding complications.12 We use
and placenta the remaining untied pedicles enoxaparin at as dose of 0.5 mg/kg body
and bleeding surfaces can be addressed with weight starting within 6 hours after com-
improved exposure. This technique of rapid pletion of surgery and sequential calf com-
hysterectomy is more likely to cause damage pression stockings. For those patients in
to pelvic structures including the bladder and whom the postoperative risk of bleeding is
ureter, and should probably not be performed clinically concerning, the ACCP guidelines
except in episodes of massive blood loss. recommend mechanical prophylaxis until
After removal of the specimen, the the risk of bleeding diminishes and phar-
unusually vascular tissues of the upper macologic prophylaxis may be initiated.
vagina and cervix often require meticulous We continue pharmacologic prophylaxis
hemostasis. The vaginal cuff or cervical for 4 weeks.
closure proceed in the typical way for a
hysterectomy. POSTOPERATIVE MANAGEMENT OF
UROLOGIC INJURY
Even in experienced hands, incidental
Postoperative Care cystotomy occurs in about one third of
In addition to the usual postoperative PAS patients undergoing cesarean hyster-
care, 2 issues deserve emphasis in cases ectomy. The dome of the bladder is the

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
782 Einerson and Branch

site of most injuries; involvement of the 4. Eller AG, Bennett MA, Sharshiner M, et al.
trigone is rare in our experience. Depend- Maternal morbidity in cases of placenta accreta
ing upon the size of the bladder injury, we managed by a multidisciplinary care team com-
pared with standard obstetric care. Obstet Gyne-
recommend Foley drainage of the bladder col. 2011;117:331–337.
for 7 to 14 days postoperatively. Whether 5. Shamshirsaz AA, Fox KA, Erfani H, et al. Multi-
or not to cystographically examine the disciplinary team learning in the management of
bladder before removal of the catheter is a the morbidly adherent placenta: outcome im-
clinical decision based on the size and provements over time. Am J Obstet Gynecol. 2017;
216:612.e1–612.e5.
complexity of the repaired injury. 6. Silver RM, Fox KA, Barton JR, et al. Center of
Ureteral injury requiring anastomosis excellence for placenta accreta. Am J Obstet Gynecol.
or bladder reimplantation may occur, but 2015;212:561–568.
is fortunately infrequent. A ureteric stent 7. D’Antonio F, Iacovella C, Palacios-Jaraquemada J,
facilitates healing. A Foley catheter et al. Prenatal identification of invasive placentation
using magnetic resonance imaging: systematic re-
should be used to maintain bladder drainage view and meta-analysis. Ultrasound Obstet Gynecol.
for at least 7 to 14 days postoperatively and 2014;44:8–16.
may be removed after cystogram confirma- 8. Meng X, Xie L, Song W. Comparing the diag-
tion of the absence of a leak. The ureteric nostic value of ultrasound and magnetic reso-
stent may be removed 1 to 2 months post- nance imaging for placenta accreta: a systematic
review and meta-analysis. Ultrasound Med Biol.
operatively, and the ureteric surgical site 2013;39:1958–1965.
evaluated via cystography or intravenous 9. Einerson BD, Rodriguez CE, Kennedy AM, et al.
pyelography.13 Thereafter, ureteric and renal Magnetic resonance imaging is often misleading when
function should be assessed at 3 to 6 months used as an adjunct to ultrasound in the management
and again at 12 months. of placenta accreta spectrum disorders. Am J Obstet
Gynecol. 2018;218:618.e1–618.e7.
10. Eller AG, Porter TF, Soisson P, et al. Optimal
management strategies for placenta accreta.
BJOG. 2009;116:648–654.
References 11. Belfort MA, Shamshiraz AA, Fox K. Minimizing
1. Silver RM, Branch DW. Placenta accreta spec- blood loss at cesarean-hysterectomy for placenta
trum. N Engl J Med. 2018;378:1529–1536. previa percreta. Am J Obstet Gynecol. 2017;216:
2. Shellhaas CS, Gilbert S, Landon MB, et al. The 78.e1–78.e2.
frequency and complication rates of hysterectomy 12. Gould MK, Garcia DA, Wren SM, et al. Pre-
accompanying cesarean delivery. Eunice Ken- vention of VTE in nonorthopedic surgical pa-
nedy Shriver National Institutes of Health and tients. Antithrombotic Therapy and Prevention of
Human Development Maternal-Fetal Medicine Thrombosis, 9th ed: American College of Chest
Units Network. Obstet Gynecol. 2009;114: Physicians Evidence-Based Clinical Practice
224–229. Guidelines. Chest. 2012;141(suppl):e227S–e277S.
3. Wright JD, Herzog TJ, Shah M, et al. Region- 13. Sharp HT, Adelman MR. Prevention, recogni-
alization of care for obstetric hemorrhageand its tion, and management of urologic injuries during
effect on maternal mortality. Obstet Gynecol. gynecologic surgery. Obstet Gynecol. 2016;127:
2010;115:1194–1200. 1085–1096.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA CLINICAL OBSTETRICS AND GYNECOLOGY
Volume 61, Number 4, 783–794
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.

Conservative
Management of
Downloaded from [Link] by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3dW0s7N5CLiYMlH4wgaY34Fr69rtLH3oWi4+h6VxXUqo= on 10/30/2018

Placenta Accreta
Spectrum
LOÏC SENTILHES, MD, PhD,* GILLES KAYEM, MD, PhD,†
and ROBERT M. SILVER, MD‡
*Department of Obstetrics and Gynecology, Bordeaux University
Hospital, Bordeaux, France; †Department of Obstetrics and
Gynecology, Trousseau Hospital, AP-HP, Paris, France; and
‡Department of Obstetrics and Gynecology, University of Utah
School of Medicine, Salt Lake City, Utah

Abstract: The purpose of this review was to assist strategies aiming to avoid a peripartum
obstetricians and gynecologists in considering the most hysterectomy and its related morbidity
appropriate conservative treatment option to manage
women with placenta accreta spectrum according to and consequences. The main goals are to
their individual need and local expertise of the heath care decrease severe maternal morbidity re-
team. The issue is challenging, as the quality of evidence lated to the placenta accreta spectrum
with regard to efficacy is poor, and is mainly based on (PAS), especially the amount of blood
retrospective studies with limited sample size. loss, and, consequently, the risk of mas-
Key words: placenta accreta spectrum, morbidly adher-
ent placenta, abnormally invasive placenta, placenta sive transfusion and coagulopathy, as well
percreta, conservative management, leaving placenta as operative injury and its potential con-
in situ sequences such as vesicouterine fistula.
A second goal may be to attempt to preserve
the option of future pregnancies, knowing
that fertility is often inextricably linked with
Introduction societal status and self-esteem. Four types of
Conservative management of placenta conservative management have been de-
accreta is defined as all procedures or scribed: extirpative treatment,1 expectant
management or the leaving placenta
Correspondence: Loïc Sentilhes, MD, PhD, Department
of Obstetrics and Gynecology, Bordeaux University in situ,2 1-step conservative surgery,3 and
Hospital, Place Amélie Raba Léon, Bordeaux, France. the triple-P procedure.4 For each of these
E-mail: loicsentilhes@[Link] procedures, there is only low-quality evi-
L.S.: wrote the first draft of the report. dence available derived from retrospective
The authors declare that they have nothing to disclose. case series.

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 61 / NUMBER 4 / DECEMBER 2018

[Link] | 783
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
784 Sentilhes et al

The Extirpative Approach the placenta in cases of unusual and

The concept of this approach is simple; unexplained difficulties before the occur-
the aim is to avoid leaving retained rence of massive hemorrhage.
placental tissue in the uterine cavity.
Retained placenta is a common cause of
postpartum hemorrhage (PPH), and Leaving Placenta In Situ
complete removal decreases the risk of
bleeding.5,6 Therefore, the procedure con- Without Hysterectomy or
sists of manually removing the placenta to Expectant Management
obtain an “empty” uterus. Unfortunately,
in cases of PAS, this procedure often SHORT-TERM AND MID-TERM
results in massive hemorrhage. Kayem MATERNAL OUTCOME
and colleagues performed a retrospective This approach consists of leaving the
study comparing 2 consecutive epochs. In placenta in situ and waiting for complete
the first one, the extirpative approach was resorption. It was first described mainly in
routinely applied for PAS, whereas in the France2 and initially was termed “con-
second one, the placenta was left in situ. servative treatment of placenta accreta.”
Mean number of red blood cells trans- As other conservative approaches have
fused (3230 ± 2170 vs. 1560 ± 1646 mL; been since described, it is more accurate
P < 0.01), disseminated intravascular co- to use the term “leaving the placenta
agulation [5 (38.5%) vs. 1 (5.0%); P = 02], in situ without hysterectomy” or “expect-
and hysterectomy rates [11 (84.6%) vs. 3 ant management.”13
(15%); P < 0.001] were reduced using the The goals of this approach are to avoid
placenta in situ approach.1 Moreover, the morbidity associated with hysterec-
when a cesarean hysterectomy for suspi- tomy, preserve fertility, and still avoid
cion of PAS has been planned, Eller et al7 hemorrhage. Planned cesarean hysterec-
showed that early maternal morbidity tomy while leaving the placenta in situ is
was increased when placental removal considered the “gold standard” treatment
was attempted, compared with the pla- for PAS6–12 but is associated with high
centa left undisturbed in situ (67% vs. rates of severe maternal morbidity (40%
36%; P = 0.04). Consequently, several au- to 50%12) and loss of fertility. Cesarean
thorities recommend that manual placen- hysterectomy with placenta percreta is
tal removal should be avoided in cases of even more morbid, with reported mortal-
planned cesarean hysterectomy.8–11 The ity rates up to 7%.14 By leaving the
downside of this approach is the potential adherent placenta in situ after the delivery
for unnecessary hysterectomy if the pa- of the child, one can expect a significant
tient does not really have PAS. In con- decrease of blood flow within the uterus
clusion, extirpative approach with a and even the parametrium. This also will
forcible manual removal of the placenta occur within the placenta, and the pla-
should be abandoned unless there is a low centa will progressively and spontane-
probability of PAS.12 Unfortunately, ously detach from the uterus and even
manual removal of the placenta usually adjacent organs by necrosis. It is analo-
happens in most cases with undiagnosed gous to cutting the foot of the ivy that is
placenta accreta. In our opinion, women incrusted into a stone wall, and waiting
with strongly suspected PAS should never for its death before removing it to avoid
have attempted manual removal of the weakening the wall. This approach is
placenta. For women with risk factors for particularly attractive for severe PAS with
PAS or mild suspicion of PAS, caregivers adjacent organ invasion to avoid opera-
should stop attempts to manually remove tive complications and injuries.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Conservative Management of Placenta Accreta 785

In contrast, expectant management has They found 48 case reports describing the
important risks. These include intrauter- outcome of 60 women as well as 2 French
ine infection, placental abscess and even case series including 31 women, yielding
sepsis, as well as unpredictable massive available data on 91 cases. Of the 26
hemorrhage. Moreover, it requires long- women managed conservatively while
term monitoring until complete resorp- leaving the placenta in situ without the
tion of the placenta. use of additional therapies, 22 (85%) had
In practice, the exact position of the a favorable outcome. Expectant manage-
placenta is determined by a preoperative ment failed in 4 (15%) of the patients who
ultrasound. Before initiating cesarean deliv- required hysterectomy due to severe hem-
ery, all materials required for an immediate orrhage or infection.15
conversion to hysterectomy are readily avail- To increase statistical power and satisfac-
able. Laparotomy is made by a midline tory external validity, a French multicenter
vertical cutaneous incision, often enlarged retrospective study was conducted to deter-
above the umbilicus. The uterine approach mine maternal outcome after conservative
uses a midline or “classical” incision at a treatment.2 Of 45 university hospitals in
distance from the placental bed. After deliv- France, 40 (88.9%) agreed to participate in
ery of the child, and only in cases wherein the study, and 25 used conservative treatment
PAS is unlikely, the obstetrician carefully for placenta accreta at least once.
attempts to remove the placenta by con- Placenta accreta was diagnosed according
trolled cord traction (see below); if the to the following clinical and histologic cri-
placenta does not easily separate from the teria: (1) it was partially or totally impossible
uterus, it confirms the diagnosis of PAS. In to manually remove the placenta with no
this case, the cord is tied with suture, cut at discernable cleavage plane between all or
the site of insertion, and the uterine cavity is part of the placenta and uterus; (2) prenatal
closed. Postoperative antibiotic therapy diagnosis of placenta accreta on sonogram,
(amoxicillin and clavulanic acid) is usually confirmed by the failure of gentle attempts to
administered prophylactically for 5 days to remove it during the third stage of labor or at
minimize the risk of infection, although cesarean delivery; (3) evidence of placental
efficacy is uncertain. Adjunctive procedures invasion at the time of surgery; (4) and
[embolization or vessel ligation, temporal histologic confirmation of accreta on hyster-
internal iliac occlusion balloon, methotrex- ectomy specimen. Women treated with an
ate (MTX), and hysteroscopic resection of extirpative approach or a planned cesarean
retained tissues] may be used to attempt to hysterectomy were excluded from this study.
decrease morbidity or to hasten placental Conservative management in case of placen-
resorption. As with antibiotic treatment, ta accreta was defined by the decision of the
none of these interventions is proven to obstetrician to leave the placenta partially or
improve outcomes. totally in situ, with no attempt to remove it
In France, the first conservative treatment forcibly. When placenta accreta was not
took place in 1993; the number of proce- suspected before delivery, it was diagnosed
dures increased steadily, particularly during when it was impossible to detach the placenta
the 2000s.4 First, only very few data about by gentle manipulation, and conservative
maternal outcome after conservative man- treatment was defined as leaving part or all
agement were available. Moreover, they of it in the uterus. The study included 167
were from small case reports and case series cases of placenta accreta with 59% of pla-
from individual tertiary-care institutions.1,12 centas left partially in situ and 41% left
In 2007, Timmermans et al15 reviewed totally in situ. Outcomes are summarized
available case series of placenta accreta in Table 1.2 Success rates were similar to
managed by leaving the placenta in situ. previous reports,1,12 with successful uterine

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
786 Sentilhes et al

TABLE 1. Maternal Morbidity After Conservative Treatment for Placenta Accreta, Modified
From Sentilhes et al2
Placenta Accreta, Including
Characteristics Percreta (n = 167)
Placenta left in situ 167 (100)
Partially 99 (59.3)
Entirely 68 (40.7)
Primary postpartum hemorrhage 86 (51.5)
No additional uterine devascularization procedure 58 (34.7)
Additional uterine devascularization procedure 109 (65.3)
Pelvic arterial embolization* 62 (37.1)
Vessel ligation* 45 (26.9)
Stepwise uterine devascularization 15 (9.0)
Hypogastric artery ligation 23 (13.8)
Stepwise uterine devascularization and hypogastric artery ligation 7 (4.2)
Uterine compression suture* 16 (9.6)
Balloon catheter occlusion 0
Methotrexate administration 21 (12.6)
Primary hysterectomy 18 (10.8)
Cause of primary hysterectomy
Primary postpartum hemorrhage 18/18 (100)
Postpartum prophylactic antibiotic therapy > 5 d 54 (32.3)
Transfusion patients 70 (41.9)
Units of packed red blood cells transfused > 5 25 (15.0)
Transfer to intensive care unit 43 (25.7)
Infection 47 (28.1)
Septic shock 1 (0.6)
Sepsis 7 (4.2)
Vesicouterine fistula 1 (0.6)
Uterine necrosis 2 (1.2)
Deep vein thrombophlebitis or pulmonary embolism 4 (2.4)
Secondary postpartum hemorrhage 18 (10.8)
Delayed hysterectomy 18 (10.8)
Median interval from delivery to delayed hysterectomy (d) 22 (9-45)
Cause of delayed hysterectomy
Secondary postpartum hemorrhage 8/18 (44.4)
Sepsis 2/18 (11.1)
Secondary postpartum hemorrhage and sepsis 3/18 (16.7)
Vesicouterine fistula 1/18 (5.6)
Uterine necrosis and sepsis† 2/18 (11.1)
Arteriovenous malformation 1/18 (5.6)
Maternal request 1/18 (5.6)
Death 1 (0.6)
Success of conservative treatment 131 (78.4)
Severe maternal morbidity 10 (6.0)

Data presented as mean ± SD, or as median with interquartile range in parentheses, or as number of patients with percentages in
parentheses.
Some patients had > 1 type of morbidity.
Success of conservative treatment was defined as uterine preservation.
A primary hysterectomy took place within the first 24 hours, whereas a delayed hysterectomy took place > 24 hours after
delivery.
*The total number of additional uterine devascularization procedures exceeds the number of patients, because some patients had
> 1 such procedure.
†These 2 patients had bilateral supraselective embolization of the uterine arteries due to primary postpartum hemorrhage on the
day of delivery.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Conservative Management of Placenta Accreta 787

preservation (no hysterectomy) in 78.4% of PAS, and the rate of previous cesarean was
cases. Importantly, severe maternal morbid- lower than in other series.2,16 It is noteworthy
ity occurred in only 6% (10/167) (defined as that conservative management was typically
any of the following: sepsis, septic shock, proposed to women who desired further
peritonitis, uterine necrosis, postpartum ute- pregnancies.2,12 This may account for the
rine rupture, fistula, injury to adjacent or- characteristics of the population, wherein
gans, acute pulmonary edema, acute renal only 53% of women had a previous cesarean,
failure, deep vein thrombophlebitis or pul- but 96% had at least 1 risk factor. Moreover,
monary embolism, or maternal death).2 One we used stringent and uniformly applied
maternal death related to multiorgan failure criteria to define PAS (see above) to mini-
occurred in a patient with marrow aplasia, mize this limitation. Finally, histopathologic
nephrotoxicity with acute renal failure, fol- examination confirmed the diagnosis of pla-
lowed by peritonitis with septic shock, after centa accreta in all immediate (18/18) and
injection of MTX in the umbilical cord. all but 1 delayed hysterectomies (17/18).2
Other rare morbidities included vesico- Despite these reassuring elements with regard
vaginal fistula and arteriovenous fistula for- to possible selection bias, we acknowledge
mation. These complications have also been that such bias may exist, and that some
reported by others12,13 and are similar to women may not have had PAS. Indeed, the
those reported after planned cesarean problem of being certain of PAS concerns all
hysterectomy.7,8,10 studies related to placenta accreta/increta/
The placenta spontaneously and com- percreta when no hysterectomy specimen for
pletely resorbed in 75% of cases after a histopathologic confirmation is available.
median of 13.5 weeks (min: 4 wk, max: 60 Consequently, assessment of conservative
wk). Hysteroscopic resection and/or cur- management of PAS is difficult. In a review
ettage were performed to remove any detailing correlations between ultrasound
remaining placenta in 25% at a median and pathologic findings, only 72/1078 cases
of 20 weeks after delivery (min: 2 wk, had histopathologic descriptions.17
max: 45 wk).2 Strengths of this retrospec- Data are even more scarce with regard to
tive study included a large number of conservative management of placenta per-
cases and participating centers, which creta. Pather and colleagues reported 3 cases
increases the study’s external validity. of placenta percreta treated with conserva-
Thus, it is reasonable to anticipate similar tive management and also performed a
results in other University teaching hos- review of available data. They found 57
pitals that may have limited experience in cases of suspected placenta percreta that
conservative treatment of placenta accre- were managed conservatively with the pla-
ta, but where blood banks, pelvic arterial centa left in situ. Hysterectomy was avoided
embolizations, obstetric subspecialties, in 60% of cases and 42% experienced major
obstetric anesthesia, interventional radi- morbidity (including sepsis, coagulopathy,
ology, urology, and gynecologic oncology hemorrhage, pulmonary embolism, fistula,
are readily available. and arteriovenous malformation).18 In a
Limitations include its retrospective design similar review, Clausen et al19 retrieved 36
and the absence of histologic confirmation of cases of placenta percreta managed by leav-
PAS in cases without hysterectomy.16 Ac- ing the placenta in situ. Delayed hysterec-
cordingly, some of these cases may not truly tomy was required in 58% of cases. In the
have been PAS, resulting in possible bias French national study that reported the
and underestimation of maternal morbidity largest series of consecutive cases of placenta
associated with conservative management. percreta with attempt to leave the placenta
Indeed, only about half of women had a in situ (n = 18), prenatal diagnosis by ultra-
placenta previa and prenatal suspicion of sonography or magnetic resonance imaging

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
788 Sentilhes et al

(MRI) was performed in 14 cases and during morbidity.12 Similarly, caregivers should
labor (at the time of the cesarean) in 4 cases.2 be aware that false-positive results, which
Conservative treatment was successful for 10 may occur in up to 28% of cases,7 also
of 18 cases (55.6%) of placenta percreta, and increase maternal morbidity. They may
severe maternal morbidity occurred in 3 of lead caregivers to perform unnecessary
the 18 (16.7%). Of the 8 cases of placenta surgical procedures with their inherent
percreta with bladder involvement, conser- complications. It makes sense to consider
vative treatment was successful in 6 cases the context. Thus, our current practice is to
(75%), and severe maternal morbidity oc- attempt to gently remove the placenta by
curred in 2 (25%).2 Although morbidity was cord traction only in cases when the diag-
considerable, it was favorable in comparison nosis of placenta accreta is uncertain. An
with planned cesarean hysterectomy in example would be a nulliparous woman
women with percretas.14 with a history of curettage in whom ultra-
These results show that leaving the sound revealed intraplacental lacunae in a
placenta in situ is a reasonable option for low-lying anterior placenta and no visible
women with PAS who are properly coun- evidence of placenta accreta during the
seled and motivated, in particular if they cesarean.12
desire future pregnancies. It also is critical
that they agree to close follow-up monitor- MTX ADJUVANT TREATMENT
ing in centers with adequate equipment and Some authors have proposed the use of
resources.2,6,8–11 However, many questions MTX to hasten placental resolution.21 Its
remain unanswered because only scare data efficacy for this indication has never been
are available with regard to the various shown, and only case reports and small
adjunctive treatments and procedures used case series with no control groups have
in a conservative approach with the placenta been reported.15 Accordingly, the Royal
left in situ. College of Obstetricians and Gynaecolo-
gists (RCOG) as well as the International
GENTLE ATTEMPTED REMOVAL Federation of Gynecology and Obstetrics
OF THE PLACENTA (FIGO) do not recommend its routine
There are few data available to answer this use.10,11 The low rate of placental cell
question. The main drawback of attempted division in the third trimester compared
removal of the placenta is that this proce- with early pregnancy raises the question
dure can cause severe bleeding with a risk of of whether MTX has any effect on pla-
maternal hemorrhagic complications and cental resorption. In addition, MTX can
hysterectomy. Its main advantage is to rarely cause serious harm such as neutro-
potentially avoid leaving an in situ placen- penia or medullary aplasia, even with a
ta, if it is really not a case of PAS, as well as single dose in a young patient.12 These
to remove the nonadherent portion of the complications are particularly morbid in
placenta, when the placenta adheres parti- the setting of infection, which is one of the
ally to the myometrium. This can reduce more common complications of conser-
the volume of placenta left in the uterus, vative management.2 Finally, the only
potentially reducing the risk of bleeding and case to our knowledge of maternal death
infection. It is important to emphasize after conservative treatment was secon-
imprecision in the antenatal diagnosis of dary to a cascade of complications (bone
PAS using Doppler ultrasound and/or marrow suppression, sepsis, and renal
MRI. These 2 imaging modalities are good failure) attributed to an intraumbilical
but imperfect for the diagnosis of PAS.12,20 cord administration of MTX.2 For these
The consequences of a false-negative result reasons, we do not advocate the use of
are obvious, that is, increased maternal MTX in cases of conservative treatment.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Conservative Management of Placenta Accreta 789

PROPHYLACTIC SURGICAL practice, we typically observe the patient

OR RADIOLOGIC UTERINE in the hospital for up to 8 days and
DEVASCULARIZATION administer prophylactic antibiotics for
There are also very few data on the use of 5 days. This epoch is the time of highest
these adjuvant techniques. Prophylactic risk for bleeding and infection. Before
devascularization can be achieved by discharge, the woman and her partner
techniques used to treat PPH (emboliza- should be advised about the need for
tion, bilateral uterine artery ligation, close, long-term monitoring. There is still
stepwise uterine devascularization, and a risk for bleeding and infection, and the
bilateral ligation of hypogastric arteries), size and vascularization of the retained
although these uterine-sparing procedures placenta often does not meaningfully
may be relatively less effective in cases of change for several weeks. The following
placenta accreta.22,23 Angstmann et al24 symptoms require emergency medical at-
showed that prophylactic embolization tention: hyperthermia, severe pelvic pain,
before performing cesarean hysterectomy foul smelling vaginal discharge, and bleed-
may reduce the risk of blood loss with ing. She should also be advised about the
accreta. Thus, it is possible that prophy- possibility of abnormal and persistent vag-
lactic devascularization could reduce the inal discharge. There should be a multi-
risk of secondary hemorrhage in the set- disciplinary team available with the skills
ting of conservative treatment.25 It could to manage complications 24 hours a day,
also theoretically accelerate placental res- 7 days a week.12,26 Patients are seen for
olution. In fact, in a retrospective com- outpatient clinic visits weekly for the first
parative study, the median delay for 2 months. If she is asymptomatic, monthly
complete placental resorption was signifi- visits are conducted until complete resorp-
cantly shorter when women underwent an tion of the placenta. The visits include a
embolization (median = 17 wk; q1011.5; clinical examination (bleeding, tempera-
q3023; range, 1 to 38 wk), compared with ture, and pelvic pain), pelvic ultrasound
women who did not undergo emboliza- (size of retained tissue), and laboratory
tion (median = 32 wk; q1018; q3048.8; screen for infection (hemoglobin and leu-
range, 12 to 111 wk) (P = 0.036). Unfortu- kocytes, c-reactive protein, and vaginal
nately, the reason for embolization was sample for bacteriological analysis).2 Of
not clearly reported by the authors.25 In course, the efficacy of most of these meas-
contrast, devascularization may cause ures is uncertain, but, in theory, they may
harm.2,23 In the French multicenter series reduce the risk of serious complications.
of 167 cases of placenta accreta treated We do not routinely use MRI and
conservatively, the only 2 cases of uterine serial beta human chorionic gonadotro-
necrosis occurred in patients (62 total) pin (βhCG) levels for monitoring. Soyer
who underwent arterial embolization.2 et al27 used MRI to follow-up 23 women
Other adverse effects of uterine artery with placenta left in situ for PAS. The
embolization have also been reported.12 median delay for complete placental re-
The risk-benefit ratio of routine devascu- sorption was 21.1 weeks (1 to 111 wk).
larization procedures in conservative They found a significant correlation be-
management of placenta accreta remains tween the degree of vascularity on early
to be determined. phase of dynamic MRI and delay of
complete placental resorption (r = 0.69;
MONITORING OF CONSERVATIVE P < 0.001). They speculate that MRI
MANAGEMENT may help predict delay for complete
Unfortunately, there are no data with placental resorption.27 It is not clear
regard to this important issue. In our whether decreasing levels of βhCG correlate

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
790 Sentilhes et al

with the rate of involution of placental hysteroscopic procedure.30 It seems that

tissue. Khan et al28 and Torrenga et al29 hysteroscopic resection may shorten recov-
described several cases of placenta left ery time without harmful effects in symp-
in situ followed with serial βhCG levels. tomatic women. The role of prophylactic
Serum βhCG levels decreased to minimal hysteroscopy or the timing of it in asymp-
levels in 5 months in the Khan study tomatic women is unknown. The safety
and in 5 to 10 weeks in the Torrenga and feasibility of high-intensity focused
study. In both studies, βhCG levels did not ultrasound in the treatment of placenta
correlate with the volume of remaining accreta after vaginal delivery was rece-
tissue,28,29 raising questions about its useful- ntly tested in 12 women with placenta
ness for the monitoring of these patients. accreta.31 The average period of residual
Thus, measuring serum βhCG on a weekly placental involution was 36.9 days. High-
basis can reassure to some extent, but intensity focused ultrasound treatment did
low levels do not guarantee complete not apparently increase the risk of infection
placental resolution.11 Consequently, pla- or hemorrhage, and no patient required
cental resorption should be documented by hysterectomy.
ultrasound imaging.11
DELAYED INTERVAL
SYSTEMATIC HYSTEROSCOPIC HYSTERECTOMY
RESECTION OF PLACENTAL TISSUE Another possible advantage of leaving the
THAT DOES NOT RESORB placenta in situ is to plan a delayed interval
Again, data with regard to this issue are hysterectomy, after partial involution of the
scarce. As mentioned previously in the placenta and decreased uterine vascularity.
French retrospective study, hysteroscopic This may decrease hemorrhagic morbidity
resection or curettage or both were used and risk of injury to adjacent organs. This
to remove retained placenta in 29 (25.0%) strategy seems most attractive in women with
cases, at a median of 20 weeks (range, 2 to placenta percreta, who are at highest risk for
45 wk) after delivery.2 These results high- blood loss and urinary tract injury. Excellent
light the fact that this procedure is performed outcomes have been reported using this
frequently. Nevertheless, no information approach in percreta cases.32 In contrast, this
with regard to the reason for performing this approach requires 2 surgeries instead of 1,
procedure (due to pain, bleeding, and/or and both may be quite morbid. Moreover,
infection, to hasten placental resorption, on there is a risk of hemorrhage or infection
maternal request or systematically) was avail- prompting the need for emergency hysterec-
able. In a small cohort of 23 women with tomy during the planned interval. Finally,
placenta left in situ for placenta accreta, 12 the optimal timing of planned delayed hys-
underwent hysteroscopy under ultrasound terectomy is uncertain.13 It may only be
guidance due to pain and/or bleeding with possible to truly ascertain whether delayed
retained placental tissue.30 The use of bipolar interval hysterectomy is effective through
energy was limited as much as possible to appropriate randomized clinical trials.
minimize risk of uterine perforation. The
median size of the retained placenta was LONG-TERM MATERNAL OUTCOME
54 mm (13 to 110 mm). No complications AND SUBSEQUENT FERTILITY AND
occurred due to hysteroscopic resection. OBSTETRICAL OUTCOME
Complete removal (11/12) was achieved Few data are available with regard to
after 1, 2, and 3 hysteroscopic procedures subsequent pregnancies in women with
in 5 (41.7%), 2 (16.7%), and 4 (33.3%) conservative management of PAS using
cases, respectively. One delayed hysterec- the placenta in situ approach, but suc-
tomy was performed after “failure” of the cessful pregnancies have been reported.12

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Conservative Management of Placenta Accreta 791

However, these reports are biased toward too small and prone to bias to truly
successful outcomes. Therefore, an attempt compare strategies. In fact, it is possible
was made to contact all women included in that severe maternal morbidity is increased
the French national retrospective study in cases of conservative treatment because
who did not undergo a hysterectomy, to unpredictable infectious complications, ute-
estimate fertility and pregnancy outcomes rine necrosis, and secondary hemorrhage
after successful expectant management.22 associated with conservative treatment can
Follow-up data were available for 96 be dramatic. As with delayed interval
(73.3%) of the 131 women included in the hysterectomy, the relative merits of planned
study. There were 8 women who had severe cesarean hysterectomy and conservative
intrauterine synechiae and were amenor- management will only be elucidated
rheic. Of the 27 women who reported through properly designed clinical trials.
wanting more children, 3 women were Until such trials are completed, it seems
attempting to become pregnant (mean du- reasonable to counsel women about planned
ration, 11.7 mo; range, 7 to 14 mo), and 24 cesarean hysterectomy and conservative
(88.9%) women had had 34 pregnancies (21 management. A major consideration is
third-trimester deliveries, 1 ectopic preg- whether or not future childbearing is desired.
nancy, 2 elective abortions, and 10 early A planned cesarean hysterectomy may be the
pregnancy losses) with a mean time to best option if the patient has no desire for
conception of 17.3 months (range, 2 to 48 more children, is older, and/or multiparous.
mo). All 21 deliveries resulted in healthy Nevertheless, we believe that conservative
babies born after 34 weeks of gestation. management is a reasonable option for
Placenta accreta recurred in 6 of 21 cases patients who are properly counseled and
(28.6%) and was associated with placenta motivated—for example, women who want
previa in 4 cases. PPH occurred in 4 (19%) the option of future pregnancies, who agree
cases; this was associated with PAS in 3 and to close follow-up monitoring, and who are
to uterine atony in 1. These results show in centers with adequate equipment and
that successful expectant management for resources.1,2,6,11,12
placenta accreta can be associated with Moreover, leaving the placenta in situ may
successful subsequent fertility and preg- prove to be the most appropriate choice in
nancy, although there is an increased risk the most severe cases of PAS, in particular in
of recurrent PAS.22 cases of adjacent organ invasion,12 wherein
radical surgery is often associated with severe
PLANNED CESAREAN maternal morbidity.14 Others also favor this
HYSTERECTOMY VERSUS approach, even in the United States, where
CONSERVATIVE TREATMENT conservative management is less common
This is one of the most important unre- than in France. An US survey noted that
solved questions with regard to the manage- 14.9% of providers would attempt to leave
ment of PAS. However, there is only 1 the placenta in situ in a hemodynamically
small retrospective study directly compar- stable patient,33 and 32% had attempted
ing maternal outcomes following planned conservative management for PAS.34
cesarean hysterectomy (n = 16) versus con-
servative management while leaving the
placenta in situ (expectant management) One-Step Conservative Surgery
(n = 10).33 No differences were observed This is an alternative conservative proce-
between groups except for estimated blood dure that has been described by one
loss, which was lower in the conservative author.3 It consists of resecting the in-
treatment group (3625 ± 2154 vs. 900 ± 754 vaded area of the uterus together with the
mL; P < 0.05).33 Of course, this study was placenta and reconstructing the uterus. It

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
792 Sentilhes et al

is performed at the time of cesarean delivery with regard to the technique. It is important
as a “one-step procedure.”3 This strategy to note that the 1-step procedure may be
aims to combine the advantages of leaving less reproducible and generalizable than
the placenta in situ approach (ie, preserving conservative treatment, because it requires
fertility) and of cesarean hysterectomy (no a novel and specific surgical procedure.
persistent high risk of bleeding or infection Successful use of the procedure by other
after the procedure). The main steps of this groups and prospective trials will ultimately
uterine-sparing technique achieved through clarify the merits of 1-step conservative
a median or Pfannenstiel incision are as therapy.
follows: (1) vascular disconnection of newly
formed vessels and the separation of in-
vaded uterine from invaded vesical tissues; The Triple-P Procedure
(2) performance of an upper segmental The team of Chandraharan proposed a
hysterotomy; (3) resection of all invaded novel uterine-sparing procedure for PAS
tissue and the entire placenta in 1 piece; (4) termed “the triple-P procedure.”4 The main
use of surgical procedures for hemostasis; steps of this procedure include the following:
(5) myometrial reconstruction in 2 planes; (i) preoperative placental localization using
and (6) bladder repair if necessary.3 transabdominal ultrasound to identify the
Palacios-Jaraquemada et al3 described superior border of the placenta in order to
outcomes of this 1-step conservative surgery deliver the fetus by an incision above the
in 68 women presenting with placental in- upper border of the placenta; (ii) pelvic
vasion of adjacent organs [invasion of the devascularization involving preoperative
posterior upper bladder (n = 46; group 1), placement of intra-arterial balloon catheters
and of the posterior lower vesical area with inflation after delivery; and (iii) no
(n = 22; group 2)]. Uterine preservation was attempt to remove the placenta with en bloc
achieved in 95.7% (44/46) and 27.3% (6/22) of myometrial excision and uterine repair. It
cases, respectively. The indications for the 18 seems important to ensure that a 2 cm
hysterectomies were segmental circumferen- margin of myometrium is retained in the
tial rupture > 50% (n = 13), coagulopathy lower lip of the uterine incision to facilitate
(n = 2), infection (n = 1), and uncontrolled closure of the myometrial defect.4 Bleeding
hemodynamic instability (n = 2). The follow- from the separated and adherent part of the
ing complications were reported (mostly in placenta is controlled by oversewing the
group 2): lower ureteral injuries (n = 2), defect. If the posterior wall of the bladder
vesical fistula (n = 1), hematoma in the vag- is involved, placental tissue invading the
inal cuff (n = 1), and uterine infection (n = 1). bladder is left in situ to avoid cystotomy.
Among the 50 women with uterine preserva- The authors reported a small series,
tion, follow-up was available in 42. Menses comparing outcomes after (n = 19) and
were recovered between 3 and 16 months. before implementation of the triple-P pro-
Ten women had another uneventful preg- cedure (n = 11). In the past, PAS was treated
nancy and delivery with no recurrence with an elective cesarean delivery, using an
of PAS.3 In another publication, Palacios- incision into the uterine fundus, leaving the
Jaraquemada35 reported 45 pregnancies placenta in part or entirely in the uterus,
following a 1-step procedure for placenta unless PPH occurred and peripartum hys-
accreta. Among these 45 pregnancies, 44 terectomy was required (control group).4
were uneventful and only 1 was complicated Demographic characteristics were compa-
by a recurrence of PAS.35 rable between groups, with a percreta rate of
As we have limited experience with 54.5% and 68.4% (P = 0.35), respectively.
the 1-step conservative surgery, we find it There was no statistical difference for the
difficult to make strong recommendations estimated mean blood loss (2170 vs. 1700

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Conservative Management of Placenta Accreta 793

mL; P = 0.44) and the rate of transfusion placenta accreta. Obstet Gynecol. 2004;104:
(45.5% vs. 47.4%; P = 0.61). However, the 531–536.
rates of PPH (54.5% vs. 15.8%; P = 0.035) 2. Sentilhes L, Ambroselli C, Kayem G, et al.
Maternal outcome after conservative treatment
and hysterectomy (27.3% vs. 0.0%; P = of placenta accreta. Obstet Gynecol. 2010;115:
0.045) were lower in the triple-P group. 526–534.
One major complication (5%) occurred in 3. Palacios-Jaraquemada JM, Pesaresi M, Nassif
a woman treated with triple-P (right com- JC, et al. Anterior placenta percreta: surgical
mon iliac and external iliac artery thrombo- approach, hemostasis and uterine repair. Acta
Obstet Gynecol Scand. 2004;83:738–744.
sis).4 This is a known complication of 4. Teixidor Viñas M, Belli AM, Arulkumaran S, et al.
temporal internal iliac occlusion balloon Prevention of postpartum hemorrhage and hyster-
catheters.12 As with the 1-step procedure, ectomy in patients with morbidly adherent placen-
these data should be considered preliminary, ta: a cohort study comparing outcomes before and
and further studies are needed to assess after introduction of the triple-P procedure. Ultra-
sound Obstet Gynecol. 2015;46:350–355.
relative efficacy. It is noteworthy that the 5. Sentilhes L, Merlot B, Madar H, et al. Postpar-
RCOG and FIGO do not recommend bal- tum haemorrhage: prevention and treatment.
loons for cesarean hysterectomy or conserva- Expert Rev Hematol. 2016;9:1043–1061.
tive treatment10,11 due to untoward effects, 6. Sentilhes L, Vayssière C, Deneux-Taraux C, et al.
although the issue remains controversial. Postpartum hemorrhage: guidelines for clinical
practice from the French College of Gynaecolo-
gists and Obstetricians (CNGOF) in collabora-
tion with the French Society of Anesthesiology
Conclusions and Intensive Care (SFAR). Eur J Obstet Gynecol
Except for extirpative treatment (best Biol Reprod. 2016;198:12–21.
reserved for cases unlikely to be PAS), 7. Eller AG, Porter TF, Soisson P, et al. Optimal
management strategies for placenta accreta.
conservative management for PAS may be BJOG. 2009;116:648–654.
a reasonable alternative option to planned 8. Publications Committee, Society for Maternal-
cesarean hysterectomy in well-selected Fetal Medicine1, Belfort MA. Placenta accreta.
cases. It is important that women treated Am J Obstet Gynecol. 2010;203:430–439.
with expectant management have appro- 9. Committee on Obstetric Practice. Committee
opinion no. 529: placenta accreta. Obstet Gynecol.
priate counseling and close surveillance 2012;120:207–211.
after delivery. The best-studied conservative 10. Royal College of Obstetricians and Gynaecolo-
approach is expectant care after leaving the gists (RCOG). Placenta Praevia, Placenta Praevia
placenta in situ. Although comparable out- Accreta and Vasapraevia: Diagnosis and Manage-
comes to planned cesarean hysterectomy ment. London: RCOG, Greentop; 2011.
11. Sentilhes L, Kayem G, Chandraharan E, et al.
have been reported, the approach is of FIGO Placenta Accreta Diagnosis and Manage-
uncertain efficacy due to bias in case ment Expert Consensus Panel. FIGO consensus
selection and uncertainty with regard to guidelines on placenta accreta spectrum disorders:
the diagnosis of PAS. Prospective trials are conservative management. Int J Gynaecol Obstet.
desperately needed to assess the true risks 2018;140:291–298.
12. Sentilhes L, Goffinet F, Kayem G. Management
and benefits of conservative management of placenta accreta. Acta Obstet Gynecol Scand.
overall, as well as for each approach. The 2013;92:1125–1134.
prospective PACCRETA study has been 13. Fox KA, Shamshirsaz AA, Carusi D, et al.
initiated to answer some of the questions Conservative management of morbidly adherent
raised in this chapter.20 placenta: expert review. Am J Obstet Gynecol.
2015;213:755–760.
14. O’Brien JM, Barton JR, Donaldson ES. The
management of placenta percreta: conservative
and operative strategies. Am J Obstet Gynecol.
References 1996;175:1632–1637.
1. Kayem G, Davy C, Goffinet F, et al. Conserva- 15. Timmermans S, van Hof AC, Duvekot JJ. Con-
tive versus extirpative management in cases of servative management of abnormally invasive

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
794 Sentilhes et al

placentation. Obstet Gynecol Surv. 2007;62: embolization for the conservative treatment of
529–539. placenta accreta. Clin Radiol. 2012;67:1089–1094.
16. Silver RM, Branch DW. Placenta accreta spec- 26. Silver RM, Fox KA, Barton JR, et al. Center of
trum. N Engl J Med. 2018;378:1529–1536. excellence for placenta accreta. Am J Obstet Gynecol.
17. Jauniaux E, Collins SL, Jurkovic D, et al. Accreta 2015;212:561–568.
placentation: a systematic review of prenatal 27. Soyer P, Sirol M, Fargeaudou Y, et al. Placental
ultrasound imaging and grading of villous inva- vascularity and resorption delay after conserva-
siveness. Am J Obstet Gynecol. 2016;215:712–721. tive management of invasive placenta: MR imag-
18. Pather S, Strockyj S, Richards A, et al. Maternal ing evaluation. Eur Radiol. 2013;23:262–271.
outcome after conservative management of pla- 28. Khan M, Sachdeva P, Arora R, et al. Conserva-
centa percreta at caesarean section: a report of tive management of morbidly adherant placenta
three cases and a review of the literature. Aust N Z —a case report and review of literature. Placenta.
J Obstet Gynaecol. 2014;54:84–87. 2013;34:963–966.
19. Clausen C, Lönn L, Langhoff-Roos J. Management 29. Torrenga B, Huirne JA, Bolte AC, et al. Post-
of placenta percreta: a review of published cases. partum monitoring of retained placenta. Two
Acta Obstet Gynecol Scand. 2014;93:138–143. cases of abnormally adherent placenta. Acta
20. Kayem G, Deneux-Tharaux C, Sentilhes L. Obstet Gynecol Scand. 2013;92:472–475.
PACCRETA group. PACCRETA: clinical situa- 30. Legendre G, Zoulovits FJ, Kinn J, et al. Con-
tions at high risk of placenta ACCRETA/percreta: servative management of placenta accreta: hys-
impact of diagnostic methods and management on teroscopic resection of retained tissues. J Minim
maternal morbidity. Acta Obstet Gynecol Scand. Invasive Gynecol. 2014;21:910–913.
2013;92:476–482. 31. Bai Y, Luo X, Li Q, et al. High-intensity focused
21. Mussalli GM, Shah J, Berck DJ, et al. Placenta ultrasound treatment of placenta accreta after
accreta and methotrexate therapy: three case vaginal delivery: a preliminary study. Ultrasound
reports. J Perinatol. 2000;20:331–334. Obstet Gynecol. 2016;47:492–498.
22. Sentilhes L, Trichot C, Resch B, et al. Fertility 32. Lee PS, Bakelaar R, Fitpatrick CB, et al. Medical
and pregnancy outcomes following uterine devas- and surgical treatment of placenta percreta to
cularization for severe postpartum haemorrhage. optimize bladder preservation. Obstet Gynecol.
Hum Reprod. 2008;23:1087–1092. 2008;112:421–424.
23. Sentilhes L, Gromez A, Clavier E, et al. Predic- 33. Jolley JA, Nageotte MP, Wing DA, et al. Man-
tors of failed pelvic arterial embolization for agement of placenta accreta: a survey of mater-
severe postpartum hemorrhage. Obstet Gynecol. nal-fetal medicine practitioners. J Matern Fetal
2009;113:992–999. Neonatal Med. 2012;25:756–760.
24. Angstmann T, Gard G, Harrington T, et al. 34. Esakoff TF, Handler SJ, Granados JM, et al.
Surgical management of placenta accreta: a co- PAMUS: placenta accreta management across
hort series and suggested approach. Am J Obstet the United States. J Matern Fetal Neonatal Med.
Gynecol. 2010;202:38.e1–38.e9. 2012;25:761–765.
25. Bouvier A, Sentilhes L, Thouveny F, et al. 35. Palacios-Jaraquemada JM. Diagnosis and man-
Planned caesarean in the interventional radiology agement of placenta accreta. Best Pract Res Clin
cath lab to enable immediate uterine artery Obstet Gynecol. 2008;22:1133–1148.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA CLINICAL OBSTETRICS AND GYNECOLOGY
Volume 61, Number 4, 795–807
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.

The Utilization
of Interventional
Downloaded from [Link] by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3dW0s7N5CLiYMlH4wgaY34Fr69rtLH3oWi4+h6VxXUqo= on 10/30/2018

Radiologic
Procedures in the
Surgical Management
of Placenta Accreta
Syndrome
VINEET K. SHRIVASTAVA, MD,
and MICHAEL P. NAGEOTTE, MD
Miller Children’s & Women’s Hospital/Long Beach,
Long Beach, California

Abstract: The role of Interventional radiologic proce-

dures for the management of suspected placenta Background
accreta spectrum (PAS) has evolved considerably over In 1947, Wilen1 forewarned that increasing
last 3 decades. In this article, the authors describe cesarean deliveries would inevitably lead to
the various techniques of vascular occlusion for the increasing incidence of placenta previa,
management of PAS and provide a brief review of the
literature examining the pros and cons in the use of suggesting a rise in “accreta” would be
these devices. the logical result. Placenta accreta spec-
Key words: placenta accreta spectrum, interventional trum or PAS, comprises placenta accreta
radiology, balloon catheters, embolization, surgical and its variants, increta or percreta which
management represent the failure of the placenta to
separate normally after birth. There has
been a 10-fold rise in the incidence of
morbidly adherent placenta over the last
Correspondence: Vineet K. Shrivastava, MD, 2888
4 decades, largely attributed to an overall
Long Beach Blvd, Ste 400 Long Beach, CA 90806. rising cesarean rate.2 As per ACOG guide-
E-mail: vshrivas@[Link] lines, definitive surgical management by
The authors declare that they have nothing to disclose. cesarean hysterectomy is the recommended

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 61 / NUMBER 4 / DECEMBER 2018

[Link] | 795
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
796 Shrivastava and Nageotte

course of action for PAS.3 However, the and placement of the interventional de-
most critical aspect faced in the surgical vice. Initial reports focused on placement
management of PAS is massive hemor- of BC in the hypogastric arteries; how-
rhage best demonstrated by Miller et al4 in ever, more proximal aspects of the vascu-
which 66% of patients undergoing cesarean lar tree have also been investigated
hysterectomy for either placenta accreta or including placement in the common iliac
one of its variants had blood loss > 2 L. arteries as well as the abdominal aorta.
The complexity encountered at the time of Another area of investigation is the tim-
surgery can be further exacerbated by the ing of deployment of the intervention. For
amount of blood loss, thus increasing the example, some recommend routine use of
risk for surgical complications and making BCs and/or embolization following the
the management of abnormal placentation delivery of the infant. Others advise the
one of modern obstetrics greatest chal- alternative approach of using these inter-
lenges. In an effort to mitigate bleeding ventions only when the surgical circum-
associated with cesarean hysterectomy, in- stances warrant it (ie, when hemorrhage is
terventional radiologic (IR) procedures encountered).5
have been utilized. In general, the approach towards the
In this article, the authors address patient starts in the presurgical period.
potential benefits and examine controver- Implicit in the informed consent conver-
sies surrounding the use of IR procedures sation is the need for accurate and timely
at time of cesarean in patients with antenatal recognition of an abnormally
morbidly adherent placentas. adherent placenta; PAS that are recog-
nized intraoperatively pose a distinctly
different clinical problem. The procedure
Why IR? is typically initiated after administration
The integration of techniques to reduce of regional anesthetic and after transfer to
blood flow and facilitate surgery by IR is the IR suite. In some cases, IR is available
a natural evolution of efforts to decrease in the operating room or labor and
morbidity and mortality. The prevailing delivery suite. Common femoral arterial
theory is that a reduction of perfusion by puncture is performed under local anes-
occluding blood flow (whether tempora- thetic and 5 to 8 French vascular sheaths
rily or for a prolonged period), improves are placed. Using pulsed low-dose fluoro-
visualization of the surgical field by re- scopic guidance by the Seldinger techni-
ducing the degree of hemorrhage. This in que, occlusive balloons catheters are
turn allows for a hysterectomy under inserted bilaterally. Exact placement of
controlled circumstances, decreases the the catheters is dependent on the inten-
need for transfusion of blood and blood tions of the operator in terms of what
products and possibly reduces the occur- aspect of the vascular tree is targeted.
rence of surgical complications. Once both BCs were correctly positioned,
There have been several IR techniques a test volume of dilute water-soluble
described in the literature with regard to contrast material is usually injected to
patients with PAS; these include prophy- inflate the occlusion balloons to the opti-
lactic placement of intra-arterial occlusive mal size, then subsequently deflated until
balloon catheters (OBCs) alone, balloon needed. Once positioning is satisfactory,
catheter (BC) placement plus postdelivery the catheters are securely taped to the
embolization, and postdelivery emboliza- skin. The deployment of OBC is per-
tion alone. In addition to these differing formed in the period following delivery
approaches, there are also distinctions of the infant at the discretion of the
reported with regard to the ideal timing surgeon.6 In addition, the placement of

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
IR Procedures for Management of Placenta Accreta Syndrome 797

arterial sheaths allows for postdelivery groups. The earliest studies centered
embolization, which has been reported around BC placement in the hypogastric
following delivery but before proceeding artery (internal iliac), likely due to the
with hysterectomy or for bleeding that perception that occlusion here will have
persists following hysterectomy. The ben- the most direct benefit with minimal
efit of embolization is a more prolonged extraneous effect. Data are limited on
vascular occlusion to further suppress the use of BC alone for hysterectomy in
pulse pressure and allow for activation PAS; however, there is significant var-
of innate clotting systems without the iance in reported outcomes (Table 1).
potential trauma resulting from BC. The first study noting benefit was pub-
Although the use of all of these techni- lished in 2007 by Tan and colleagues, who
ques shows promise in theory, data are reported improved outcomes in 11 patients
mixed with regard to efficacy. There is some of which underwent cesarean hyster-
significant variation among studies with ectomy following bilateral internal iliac
regard to technique used (ie, OBCs vs. artery balloon occlusion. Following delivery
embolization, etc.), location of the inter- of the infant, the surgeons deployed the
vention in the vascular tree, whether to use balloons and determined if hysterectomy
embolization (and which particle to use for was necessary. A control group consisting
embolization) and timing of when the of 14 similar patients underwent cesarean
intervention is used (routine vs. indicated delivery followed by hysterectomy without
use). Early data were primarily case series; occlusive balloon placement. The authors
over the years retrospective studies with noted significant reduction in intraoperative
comparisons with historical controls have blood loss (2011 vs. 3316 mL; P = 0.042)
been reported. With the exception of one and the volume of blood transfused (1058
small trial in 2015 by Salim and colleagues, vs. 2211 mL; P = 0.005) in patients that
there are no prospective randomized trials received BC and ultimately underwent hys-
evaluating any IR procedures in the man- terectomy. No differences between groups
agement of PAS at the time of hysterec- were noted in length of hospitalization and
tomy. intensive care unit admission.6 Some of the
subjects in the study group did not have
hysterectomy performed; 5 patients had
placental extraction, but 2 of these had
Prophylactic Occlusive significant postpartum hemorrhage that re-
Balloon Catheters quired additional therapies including hys-
terectomy in one case.
INTERNAL ILIACS AND BRANCHES The next study to demonstrate benefit in
There have been many retrospective re- BC was conducted by Ballas and col-
ports and studies on usage of OBC, with leagues. In this study, 59 of 117 subjects
these early reports championing suc- received BCs in the hypogastric artery and
cesses, purporting decreased blood loss had significant reduction in mean esti-
and improved maternal outcome.7–10 As mated blood loss (2165 vs. 2837 mL;
stated previously, the goal of BC is to P = 0.02), decreased subjects experiencing
reduce intraoperative blood loss, decrease > 2500 mL blood loss and fewer patients
the need for transfusion, reduce operative requiring massive transfusions ( ≥ 6 units
time, and enhance ease of surgery. packed RBC units). The outcomes of this
Although case series offered insight into study suggest considerable benefits from
the relative effectiveness of these inter- BC. However, it is noteworthy that nearly
ventions, true efficacy is difficult to estab- every patient receiving BC were antenatally
lish in the absence of a comparison diagnosed with PAS compared with 29% in

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
798 Shrivastava and Nageotte

TABLE 1. Summary of Literature of Patients With BC Placement in Internal

Iliacs ± Embolization for Management of Abnormal Adherent Placentas
(Organized by Year)
Total Patients Balloons/
References (# w/Balloons) Embolization Benefit Comment
15
Salim et al 27 (13) Yes/no No Powered to units of blood
transfused
Only randomized trial
Cali et al12 53 (30) Yes/no Yes Benefit observed in percretas
No benefit in accreta/increta
Ballas et al11 117 (59) Yes/no Yes Majority of controls not
diagnosed antenatally
Only study to deploy balloons
situationally vs. routinely
Angstmann 26 (8) Yes/yes Yes-from Patients in both groups
et al29 embolization received balloons, difference
between groups was
embolization
Tan et al6 25 (11) Yes/yes Yes Notes benefit in blood loss and
need for transfusion.
No difference in mean Hgb
change
Shrivastava 69 (19) Yes/no No No difference noted when
et al14 emergency cases were
excluded from control group
Three cases of thrombosis in
BC group
Bodner et al28 28 (6) Yes/yes No No benefit with addition of
embolization
Levine13 9 (5) Yes/no No

Dark shade indicates negative studies with BCs; white shade, positive studies with BCs; BC, balloon catheters; Hgb, hemoglobin.

the hysterectomy alone group. Further, hysterectomy alone. A strength of this

when only PAS subjects antenatally diag- study as compared with the previous 2, is
nosed were compared with or without that all patients were antenatally diag-
balloons, there were no differences in mean nosed. As observed in Ballas and col-
estimated blood loss (P = 0.19).11 This leagues, the BC group had a reduction
suggests that cases of intraoperatively en- in the mean estimated blood loss (846 vs.
countered placenta accreta or its variants 1156 mL; P = 0.036) and number of units
skew these data in hysterectomy alone transfused (0.47 vs. 1.96 units; P = 0.011)
groups towards greater blood loss. This compared with controls. The author,
fact is intuitive; patients with antenatally however, noted something novel; the pos-
diagnosed PAS are presumably clinically itive effect was more demonstrable with
stable (ie, not bleeding at time of delivery, placenta percreta compared with placenta
not in active labor, etc.) and had improved accreta/increta. This suggests that the
outcomes relative to cases diagnosed dur- benefits of BC are more apparent in cases
ing delivery. at higher risk for bleeding such as percre-
Cali and colleagues, in 2014, showed tas. This study highlights the fact that
benefit in 30 subjects who received pro- varied results with regard to BC and PAS
phylactic BCs in the internal iliacs com- may be due to the heterogeneity of ab-
pared with 23 historical controls who had normal placentation.12

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
IR Procedures for Management of Placenta Accreta Syndrome 799

In contrast to these results, other stud- measured endpoints including the mean
ies found no benefit form BCs in PAS. In number of packed RBC units transfused
1999, Levine and colleagues performed or in calculated blood loss. Further, 2
the first prospective study of 5 patients women receiving BCs had reversible ad-
who received prophylactic BC and com- verse effects (leg pain and weakness of
pared their outcomes against historical unclear etiology and reversible buttock
controls who had an unsuspected diagno- pain). Unfortunately, not all patients
sis of placenta accreta. In this study, no received the same surgical procedure of
significant differences in estimated blood hysterectomy, as conservative manage-
loss (5025 vs. 4653 mL), units of packed ment was an option. In fact, fewer than
red cells transfused (5.5 vs. 4), and hospi- half of the women had hysterectomies
tal stay (7 vs. 5 d) were noted between the performed initially (6/13 in the treatment
groups.13 The inherent flaw in this study, and 7/14 in the control group). Conse-
similar to the one by Ballas et al, was the quently, this study was severely under-
comparison of antenatally diagnosed ver- powered to be able to assess the potential
sus unsuspected placenta accreta. How- benefit of BCs in patients with placenta
ever, in this case there was no difference accreta.15
observed in outcomes.
Shrivastava and colleagues in a larger COMMON ILIACS
study also observed no differences when The previous studies described above dem-
BC’s were used in a study of 69 subjects, onstrated benefit in subjects that had BC
19 of whom received OBCs. In this inves- placed “downstream” in branches of the
tigation there was no demonstrable dis- internal iliac. In some cases, BCs are placed
tinction in mean estimated blood loss in the hypogastric arteries or more prox-
(2700 vs. 3000 mL; P = 0.79), number imally to occlude the internal iliac division.
of units transfused (10 vs. 6.5 units; However, as noted above, data are mixed
P = 0.60), postoperative hospital days with regard to efficacy, potentially due to
(5 vs. 4 d; P = 0.85) and operative time compensatory collateral circulation being
(182 vs. 180 min; P = 0.85). As with pre- established once the occlusion occurs. This
vious studies, the hysterectomy alone prompted attempts to occlude more prox-
groups had a mixture of both antenatally imal vessels, starting in the common iliacs
diagnosed and intraoperatively encoun- (Table 2). In a retrospective study, 13
tered (19%) abnormally adherent placen- patients with BC in the common iliacs
tas. However, when the intraoperatively who were identified presurgically with
diagnosed group was analyzed, there con- PAS were compared with historical
tinued to be no difference in the prespe- controls.16 In this descriptive study, the
cified primary outcomes.14 As with earlier authors demonstrated a significant reduc-
studies, the BCs were placed in the hypo- tion in blood loss in the BC group
gastric arteries (internal iliac arteries). compared with patients with hysterectomy
There has been one publication of a alone (1902 vs. 4445 mL; P = 0.0402).
randomized control trial of precesarean The authors did not report on any other
prophylactic BCs for suspected PAS’s by outcome variables but noted 2 serious
Salim and colleagues. They report on the complications (15.8%) involving arterial
prospective randomization of 27 women thrombotic occlusions requiring intrave-
with suspected placenta accreta to either nous medical therapy. The number of
preoperative prophylactic BCs in the an- subjects in this study was very limited.
terior division of the internal iliacs or to a The use of control data from a previous
control group with no BCs. They report publication is problematic with the intro-
no difference between groups in any duction of considerable bias. For example,

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
800 Shrivastava and Nageotte

TABLE 2. Summary of Literature of Patients With BC Placement in Common Iliacs for

Management of Abnormal Adherent Placentas
Total Patients
References (# w/Balloons) Comparison Group Benefit Comment
17
Ono et al 57 (29) Yes: 3-arm study Yes Management based on epochs in single
I. Hyst alone institution
II. Surgical Observed reduction in operative blood
ligation loss in BC group
III. BCs 5 patients in BC with complications or
questionable efficacy
Al-Hadethi 52 (25) Yes No Notes no benefit in blood loss and
et al18 mean Hgb change
2 thrombotic complication from BC
Common Iliac arterial dissection
and thrombus
Bilateral iliac dissection
Chou et al16 35 (13) Yes: historical Yes Comparison with previous publication
controls Primary outcome was reduction in
operative blood loss
2 thrombotic complications
11 of 13 Balloon patients had
hysterectomy

Dark shade indicates negative studies with BCs; white shade, positive studies with BCs; BC, balloon catheters; Hgb, hemoglobin.

there can be changes in practice and im- the interpretation of efficacy. In addition,
provement in surgical acumen over time. the authors excluded cases, potentially
Common iliac BC placement was also introducing bias. In 2 patients the bal-
performed by Ono and colleagues in 2018, loons reportedly ruptured—with no de-
who reported on 57 patients who under- scribed sequelae, in another patient there
went hysterectomy for presumed PAS’s was balloon migration into the external
over a 29-year period (1985 to 2014). iliac and in at least 2 patients balloons
They compared 3 groups including hys- were “ineffective.” Interestingly, as in
terectomy alone (13 patients), patients Cali and colleagues, the authors in this
who had surgical ligation of the internal study noted that the significant findings
iliacs before hysterectomy (15 patients), were driven primarily by more invasive
and BC in the common iliacs before placentation. When placenta percreta are
hysterectomy (29 patients). Each techni- excluded from the groups, there was no
que was the only method of management benefit from balloons.
during epochs at this single institution. In However, not all studies found benefit
this study, the authors reported that from common iliac placement of BCs. In
operative blood loss was significantly their study, Al-Hadethi and colleagues
reduced in BC group compared with evaluated 25 subjects who had common
hysterectomy alone and surgical ligation iliac BC versus 27 control subjects who
at time of delivery (2027 vs. 3786 mL vs. did not. Subjects were identified retro-
4175 mL, respectively). In addition, there spectively using pathology and radiology
was a concomitant reduction in number databases. Groups were similar with re-
of units transfused.17 However, as with gard to operative blood loss (1992 vs.
Chou and colleagues, these authors com- 2455 mL; P = 0.25) and postoperative he-
pared groups in different epochs over a moglobin (87 vs. 90 g/L; P = 0.40).18
29-year period, potentially confounding Although this study counters the findings

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
IR Procedures for Management of Placenta Accreta Syndrome 801

presented by Chou and colleagues, the ranging 20 to 60 minutes. Further, some

identification of subjects by using data- protocols used intermittent deflation after
bases potentially influences the findings. 15 to 20 minutes. This was performed in
Specifically, not all patients had hysterec- an effort to reduce ischemia-reperfusion
tomies performed in both groups, suggest- injury, which could result in tissue ne-
ing cases without PAS were included in crosis, multiorgan dysfunction and signif-
the analysis—potentially diluting the ben- icant metabolic acidosis.
efits of the balloons. Once again, 2 major The use of aortic OBC has been used in
complications were observed in the 2 distinct situations involving morbidly
study group. One patient had traumatic adherent placenta; either at time of hyster-
removal of arterial sheath resulting ectomy or for attempted uterine preserva-
in common iliac dissection and acute tion. Although avoidance of hysterectomy
thrombus formation, requiring surgical was described occasionally in BC’s placed
thrombectomy. The second patient had in more distal locations, the goal of using
bilateral iliac dissection at the site of aortic balloon occlusion increasingly used
balloon occlusion requiring angioplasty to facilitate uterine preservation. An early
and stenting. study that attempted to demonstrate this
was by Panici and colleagues, in which they
ABDOMINAL AORTA were significantly able to reduce the need
As studies in BCs in the common iliacs for hysterectomy in magnetic resonance
were ongoing, some authors postulated imaging (MRI)-confirmed multifocal ac-
that even more proximal placement of creta or increta subjects who had no ante-
occlusive balloons into the infrarenal partum bleeding and desired future
abdominal aorta would fully prevent the fertility. In this study, manual removal of
establishment of collateral blood flow the placenta was routinely tried in both
into the pelvis, thus reducing extraneous groups. Patients with aortic OBC had
bleeding from diverted blood flow. This significantly fewer hysterectomies 13%
technique was described in a report by (2/15) versus the control group that refused
Paull et al19 in 1995 in the successful OBC 50% (9/18) (P = 0.03). In addition,
management of placenta percreta. the aortic BC patients had less blood
The technique for this procedure is loss (950 vs. 3375 mL; P = < 0.001), less
ostensibly the same as iliac BCs place- packed blood transfusions (0 vs. 4 units;
ment. The abdominal aortic balloon is P = < 0.001) and less postoperative days
placed before a scheduled cesarean deliv- (3 vs. 8.5 d; P = < 0.001). The authors
ery. Following infiltration of local anes- claimed that using aortic BC’s granted a
thesia, the BC is inserted into the “higher degree of pelvic devascularization”
abdominal aorta utilizing the Seldinger and intimated that aortic BC offers a more
puncture technique through the femoral effective and possibly safer fertility sparing
artery approach and placed below the option compared with leaving the placenta
renal arteries. Placement is usually con- in situ.20 The use of MRI for confirmation
firmed by injecting contrast material in for PAS offers credibility in the underlying
the balloon and obtaining imaging to pathologic diagnosis, however, as with
verify its location. Following delivery of most of these retrospective studies the out-
the infant, the balloon is deployed by comes are clouded by a lack of random-
syringe push of a predetermined amount ization. In addition, it is implied that cases
till complete occlusion of blood flow is of severe architectural destruction are not
achieved. The reported occlusion time candidates for placental removal under
of the abdominal aorta varies among aortic balloon occlusion. Lastly, all of these
reports, with authors reporting times patients reached 36 weeks’ gestation and

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
802 Shrivastava and Nageotte

had no episodes of antepartum bleeding, a and days in intensive care unit were lower
cohort of patients that are innately stable in the aortic BC group than in control
and perhaps more amenable to conserva- groups. The average time of balloon
tive therapy. inflation was 23 minutes, which is com-
Chen and colleagues performed a sim- paratively less than previous studies.
ilar study using occlusive aortic balloons The authors speculate that this may
on patients undergoing cesarean delivery have resulted in relatively fewer compli-
with suspected morbidly adherent placen- cations (lower extremity thrombosis in
ta. The authors had a similar group of 2 patients).22 The significant findings of
controls. In this study, patients were in 2 reduced operative hemorrhage and con-
different categories; (1) overt placental servation fertility were further echoed
encroachment whom were immediately in 2 subsequent retrospective studies.23,24
taken for hysterectomy (2) subjects with However, as stated earlier, the inclusion
less apparent clinical invasion following of unstable women in the control groups
attempted placental removal with adjunc- biases the apparent benefits of balloon
tive surgical measures of wedge resec- occlusion in these retrospective compar-
tions, placental bed sutures, etc. The isons.
control group was similarly managed Results of studies using occlusive aortic
except they did not have aortic BC—the BC in the management of PAS are sum-
reason these patients did not have occlu- marized in Table 3. All showed benefit for
sion was not specified. In this study, as the balloons, which also was noted in a
with Panici and colleagues, the use of systematic review.25 In every study there
occlusive aortic BC’s was associated with is an observed reduction in the observed
less intraoperative blood loss (1155 vs. operative blood loss, a reduction in the
2017 mL; P = 0.001), less operative time amount of blood transfused, and the
(118.5 vs. 145 min; P = 0.07) and less possibility of avoiding hysterectomy all
transfused red blood cell products (580 together. The strength of these studies is
vs. 1017 mL; P = 0.016). Unfortunately, the large number of patients enrolled, the
there was no analysis of outcomes within lower complication risk compared with
each of the subgroups. One complication more distal balloons and the selection of
was reported when the occlusive balloon subjects using modern imaging techni-
migrated cephalad and blocked the renal ques. A limitation, as with any study of
artery resulting in acute renal failure.21 conservatively managed patients with
The largest study to evaluate the PAS, is the lack of pathologic assessment
occlusive aortic BC in the management raising questions about the accuracy of
of PAS was performed by Wu and col- the diagnosis. Another major limitation is
leagues and included 268 patients who the lack of randomization in all of the
were antenatally diagnosed with PAS studies; this introduces considerable bias,
using ultrasound or MRI. Two-hundred especially since the sickest patients were
thirty patients were managed with OBC sometimes considered too unstable for the
and compared with 30 control group balloon procedure.
patients (who did not receive BC either
because of lack of availability of operat- COMPARISONS
ing room or because the patient was There is only has one report comparing
unstable). As seen in earlier studies, oper- the placement of aortic versus internal
ating room time (64.1 vs. 92.1 min), esti- iliac BCs. In a study by Wang and
mated blood loss (921 vs. 2790 mL), the colleagues, a prospective nonrandomized
incidence of hysterectomy (0 vs. 3), blood study was performed comparing efficacy
transfusion volume (422 vs. 1580 mL), of aortic occlusion (57 patients) versus

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
IR Procedures for Management of Placenta Accreta Syndrome 803

TABLE 3. Summary of Literature of Patients With Balloon Catheter Placement in the

Abdominal Aorta for Management of Abnormal Adherent Placentas
Total Patients Comparison
References (# w/Balloons) Group Benefit Comment
23
Zeng et al 86 (48) Yes Yes Combined use of Bakri balloon in
strategy to avoid hysterectomy
Maternal outcomes improved in patients
with increta, including a reduction in
hysterectomy
Not as effective in the management of percreta
Xie et al24 71 (30) Yes Yes Controls voluntarily refused interventional
procedure
Noted significant differences in mean
EBL and decrease in Hgb after
surgery between groups
No differences in duration of surgery,
transfusion or need for hysterectomy
Wu et al22 268 (230) Yes Yes Controls were either not stable or no IR
availability
Significantly lower operative time, EBL
and need for hysterectomy reported
Chen and Xie21 43 (20) Yes Yes Significantly less operative blood loss,
less operative time and less units
of blood products transfused
One complication reported with BC migration
occluding renal arteries
Panici et al20 33 (15) Yes Yes Manual placental removal on all patients
Control group was patients who refused BC
2 of 15 BC patients required hysterectomy
vs. 9 of 18 non-BC patients
Reduced blood loss in BC group with
hysterectomy

BC indicates balloon catheters; Hgb, hemoglobin; IR, interventional radiology.

internal iliac occlusion (48 patients). The aortic balloon group. Patients in the
selection into groups was based on a aortic balloon group had less blood loss
change in practice over time. Patients (450.4 vs. 619.2 mL; P < 0.001), required
were identified based on sonographic less blood to be transfused (480.7 vs.
and/or MRI imaging, risk factors, desire 614.6 mL; P < 0.001), and lower fetal
to preserve fertility and no antepartum radiation doses (3.8 vs. 6.6 mGy;
bleeding. Embolization was utilized in P < 0.001). There were no major proce-
patients with continued hemorrhage (gel- dure related complications in either
atin sponge particles) and hysterectomy group. The authors concluded that occlu-
was performed for refractory bleeding. sion of the internal ililacs is not as
Angiograms were performed at the effective as occlusion of the aorta in the
end of procedures to evaluate vascular management of PAS’s.26 The implication
dissection or thrombosis. Postoperative of this study is that the lack of clear
evaluations were performed at 1, 3, and benefit from distal occlusion of the iliac
6 months. Successful cesarean section branches can be attributed to the estab-
with preservation of the uterus was lishment of collateral circulation that can
achieved in all but 2 patients, both in the be avoided by aortic occlusion.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
804 Shrivastava and Nageotte

Shahin and Pang performed a systematic to stem blood loss at time of hysterec-
review and meta-analysis evaluating the tomy, but also used in some cases as part
effectiveness of endovascular interventional of a strategy to preserve the uterus. Some
procedures on the management of hemor- studies attempted to evaluate the effects
rhage associated with PAS. In this review, of arterial embolization following OBCs
69 studies met inclusion criteria, including in preparation for hysterectomy. As with
1395 patients undergoing endovascular OBC investigations, there are conflicting
management for hemorrhage. This analysis data and limited numbers of studies with
evaluated balloon occlusion along all as- appropriate control groups investigating
pects of the vascular tree, internal iliac arterial embolization before planned hys-
occlusion (470 patients; 34%), abdominal terectomy. The first study to evaluate this
aorta (460 patients; 33%), hypogastric ar- issue was by Bodner and colleagues who
tery (181 patients; 13%), and common iliac evaluated 28 consecutive subjects, 6 of
occlusion (21 patients; 5%) and made com- whom received BC plus embolization
parisons with no endovascular intervention. compared with 22 who received hysterec-
The primary endpoint was estimated vol- tomy alone. They found no differences in
ume of blood loss. Compared with no mean estimated blood loss (2.8 vs. 2.6 L;
endovascular intervention, prophylactic en- P = 0.40), volume of replaced blood prod-
dovascular intervention is effective for hem- ucts (6.5 vs. 6.3 units; P = 0.47), or operating
orrhage control during or after deliveries room time (338 vs. 228 min; P = 0.052).28
complicated by abnormal placentation. Im- In contrast, Angstmann and colleagues
portantly, patients with OBCs in the ab- compared a staged embolization hysterec-
dominal aorta had the lowest blood loss tomy protocol on 8 subjects to 18 controls
during delivery and were less likely to have a and observed benefits from embolization.
hysterectomy performed when compared In this study, patients with abnormal pla-
with other locations for BC placement.27 centas had BCs placed in the IR suite on
In addition, the amount of maternal and the day of planned delivery. Following
fetal radiation exposure was lower as was placement, the patient was transferred to
the number of observed complications in the OR where a cesarean delivery was
patients who had aortic OBC (despite the performed. If the patient was stable follow-
fact a larger introducer is required resulting ing delivery of the infant, the patient was
in the need for a vascular closure device). Of then transferred back to the IR suite and
course, the limitations and biases of the embolized before returning to the OR for
individual studies are also pertinent to the hysterectomy. These cases were compared
review. Moreover, there was significant with patients who underwent hysterectomy
heterogeneity across studies in terms of alone, some which were not stable enough
patient selection, intervention used, timing for the embolization procedure. The au-
of intervention, and goal of the intervention thors did not clearly delineate why some
(for example uterine preservation). Taken patients were embolized and others were
together, these problems limit our ability to not. The authors noted embolized subjects
make definitive conclusions about the effec- had less mean blood loss (553 vs. 4517 mL;
tiveness of endovascular intervention. P = 0.0001) and units transfused (0.5 vs. 7.9
units; P = 0.0013).29 However, there was
BALLOON CATHETERS AND/OR considerable bias in the use of controls that
EMBOLIZATION were “not stable” for the embolization
As with OBCs, early reports supported portion of the protocol. The uses of these
the use of embolization in the manage- subjects skew these data towards greater
ment of abnormal placentation. In some blood loss, making interpretation of the
instances, embolization was not only used results difficult.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
IR Procedures for Management of Placenta Accreta Syndrome 805

It is possible that one reason for the lack of occlusion of distal branches downstream
observed benefit from embolization for PAS are targeted and direct, but allows for
is likely for the same reason why benefit is establishment of substantial collateral cir-
not demonstrated in BC placed distally in the culation from the plexus of surrounding
vasculature; collateral circulation. Another vessels diverting blood flow more distally,
possible reason could be technical difficulties. continually delivering blood to the surgical
Embolization in these cases are more chal- field. Angiographic studies of patients
lenging not only due to the extensive collat- undergoing internal iliac artery ligation
eral circulation but also due to the increased identified branches of the lumbar, sacral,
caliber of vasculature as a consequence of rectal, femoral, and even internal thoracic
pregnancy; achieving vascular occlusion with arteries as the origin of collateral circula-
particle matter requires more product and tion, preventing pelvic ischemia.9 In an
greater technical expertise. effort to curtail these effects in distal
OBC, some authors have advocated against
STATE OF THESE DATA routine deployment of balloons following
As alluded to earlier, a lack of data from delivery of the infant. Instead, they advise
randomized prospective trials limits our deployment only at times of “torrential
ability to evaluate the effectiveness of the hemorrhage.5” This technique of delayed
interventional procedures. Available in- inflation was performed only in the Ballas
formation is derived from retrospective et al11 trial which found benefit in OBC’s.
studies with (often biased) comparison Another strategy to avoid the establishment
groups. The difficulty in using these stud- of collaterals is for more proximal place-
ies is inherent in their nature—lack of ment of the balloons in the vascular tree,
randomization, multiple confounding typically in common the common iliacs and
variables, and introduction of bias. For the aorta. As described earlier, this techni-
example, in studies that failed to demon- que allows for a “drier surgical” field but
strate benefit, a valid criticism is that BCs increases the risk for end-organ ischemia.
were placed in subjects with more omi- Preliminary data suggest less operative
nous appearing placentas thus preventing blood loss with more proximal placement.
greater blood loss but obscuring the ac- It should be noted the majority of studies on
tual advantage (selection bias). This could aortic placement were conducted in Asia,
explain the absence of observed benefit in where there may be differences in approach,
Shrivastava and colleagues. In contrast, diagnosis and treatment compared to West-
other studies included controls that were ern medical practices.
deemed too unstable to have IR proce- Another factor that makes it hard to
dures such as the reports by Ballas and assess benefit from interventional proce-
colleagues and Angstmann and col- dures is the remarkable heterogeneity of
leagues. The benefits of IR in these studies abnormal placentation. This was demon-
may have been due to more difficult cases strated by Cali et al12 who noted that the
being included in the control groups. The benefits of OBC’s were limited to placenta
findings noted in these studies, beneficial percretas. The implication of this finding
or not, could be a reflection of the differ- suggests that the performance of a cesar-
ing characteristics of the groups instead of ean hysterectomy is subject to an intrinsic
the studied interventions. amount of blood loss. Typical blood loss
Another plausible explanation for failure with planned cesarean hysterectomy for
to observe benefit from OBC’s is the routine cases may be so low that OBCs do
establishment of collateral circulation. The not lead to meaningful improvement.
vascular tree that supplies the gravid uterus With the exception of Cali et al’s study,
is very extensive and redundant. Selective the others evaluated the effects of OBC’s

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
806 Shrivastava and Nageotte

on all types of abnormally adherent pla- 4. Miller DA, Chollet JA, Goodwin TM. Clinical
centa as a group, with no subanalysis risk factors for placenta previa-placenta accreta.
based on severity. These data highlight Am J Obstet Gynecol. 1997;177:210–214.
5. Hull AD, Resnik R. Placenta accreta and post-
the need for a presurgical classification partum hemorrhage. Clin Obstet Gynecol. 2010;53:
of morbidly adherent placentas so that 228–236.
appropriate comparisons are made, and 6. Tan CH, Tay KH, Sheah K, et al. Perioperative
interventions appropriately assessed. endovascular internal iliac artery occlusion balloon
It also is critical to consider complica- placement in management of placenta accreta. AJR
Am J Roentgenol. 2007;189:1158–1163.
tions of IR. As with efficacy, it is hard to 7. Kidney DD, Nguyen AM, Ahdoot D, et al.
accurately determine risk from retrospec- Prophylactic perioperative hypogastric artery bal-
tive data. Shrivastava and coworkers loon occlusion in abnormal placentation. AJR
noted a 19% rate of complications! In Am J Roentgenol. 2001;176:1521–1524.
the context of that study, and in the 8. Oei SG, Kho SN, Broeke EDMt, et al. Arterial
balloon occlusion of the hypogastric arteries:
absence of clear benefit, it was difficult A life-saving procedure for severe obstetric
to recommend continued use of these hemorrhage. Am J Obstet Gynecol. 2001;185:
devices. The types of complications en- 1255–1256.
countered ranged from subdermal hema- 9. Shih JC, Liu KL, Shyu MK. Temporary balloon
toma formation to vascular dissection occlusion of the common iliac artery: new ap-
proach to bleeding control during cesarean hys-
warranting arterial bypass. Chou and terectomy for placenta percreta. Am J Obstet
colleagues noted a 15.8% rate of compli- Gynecol. 2005;193:1756–1758.
cations from thrombotic occlusions, 10. Carnevale FC, Kondo MM, de Oliveira Sousa
which were managed medically using W Jr, et al. Perioperative temporary occlusion of
balloon occlusion of the common iliac the internal iliac arteries as prophylaxis in cesar-
ean section at risk of hemorrhage in placenta
arteries. Complication rates after transi- accreta. Cardiovasc Intervent Radiol. 2011;34:
ent occlusion of the aorta are lower. This 758–764.
may be due to thicker arterial walls in 11. Ballas J, Hull AD, Saenz C, et al. Preoperative
more proximally located vasculature, intravascular balloon catheters and surgical out-
which may be relatively resistant to in- comes in pregnancies complicated by placenta
accreta: a management paradox. Am J Obstet
timal damage and thrombosis formation. Gynecol. 2012;207:216.e1–216.e5.
The optimal use of endovascular inter- 12. Cali G, Forlani F, Giambanco L, et al. Prophy-
ventional procedures for PAS remains lactic use of intravascular balloon catheters in
unclear. However, recent data on prox- women with placenta accreta, increta and percreta.
imal occlusion are encouraging and war- Eur J Obstet Gynecol Reprod Biol. 2014;179:36–41.
13. Levine AB, Kuhlman K, Bonn J. Placenta accre-
rant further study. Prospective studies ta: comparison of cases managed with and with-
with randomization are needed in presur- out pelvic artery balloon catheters. J Matern Fetal
gically classified patients, to accurately Med. 1999;8:173–176.
assess the efficacy and risk of these thera- 14. Shrivastava V, Nageotte M, Major C, et al. Case-
pies for PAS. control comparison of cesarean hysterectomy
with and without prophylactic placement of intra-
vascular balloon catheters for placenta accreta.
Am J Obstet Gynecol. 2007;197:402.e1–402.e5.
References 15. Salim R, Chulski A, Romano S, et al. Precesarean
1. Wilens I. Placenta accreta found at cesarean section prophylactic balloon catheters for suspected pla-
for placenta previa. Am J Surg. 1942;LVIII:434–437. centa accreta: a randomized controlled trial.
2. Silver RM, Landon MB, Rouse DJ, et al. Mater- Obstet Gynecol. 2015;126:1022–1028.
nal morbidity associated with multiple repeat 16. Chou MM, Kung HF, Hwang JI, et al. Tempo-
cesarean deliveries. Obstet Gynecol. 2006;107: rary prophylactic intravascular balloon occlusion
1226–1232. of the common iliac arteries before cesarean
3. Practice AcoO. ACOG Committee opinion. hysterectomy for controlling operative blood loss
Number 266, January 2002: placenta accreta. in abnormal placentation. Taiwan J Obstet Gyne-
Obstet Gynecol. 2002;99:169–170. col. 2015;54:493–498.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
IR Procedures for Management of Placenta Accreta Syndrome 807

17. Ono Y, Murayama Y, Era S, et al. Study of the reduced maternal morbidity of placenta
utility and problems of common iliac artery increta/percreta. Medicine (Baltimore). 2017;96:
balloon occlusion for placenta previa with accre- e8114.
ta. J Obstet Gynaecol Res. 2018;44:456–462. 24. Xie L, Wang Y, Luo FY, et al. Prophylactic use of
18. Al-Hadethi S, Fernando S, Hughes S, et al. Does an infrarenal abdominal aorta balloon catheter in
temproray bilateral balloon occlusion of the com- pregnancies complicated by placenta accreta.
mon iliac arteries reduce the need for intra-operative J Obstet Gynaecol. 2017;37:557–561.
blood transfusion in cases of placenta accretism? 25. Manzano-Nunez R, Escobar-Vidarte MF, Naranjo
J Med Imaging Radiat Oncol. 2017;61:311–316. MP, et al. Expanding the field of acute care surgery:
19. Paull JD, Smith J, Williams L, et al. Balloon a systematic review of the use of resuscitative
occlusion of the abdominal aorta during caesar- endovascular balloon occlusion of the aorta (RE-
ean hysterectomy for placenta percreta. Anaesth BOA) in cases of morbidly adherent placenta. Eur J
Intensive Care. 1995;23:731–734. Trauma Emerg Surg. 2018;44:519–526.
20. Panici PB, Anceschi M, Borgia ML, et al. Intra- 26. Wang YL, Duan XH, Han XW, et al. Compar-
operative aorta balloon occlusion: fertility pres- ison of temporary abdominal aortic occlusion
ervation in patients with placenta previa accreta/ with internal iliac artery occlusion for patients
increta. J Matern Fetal Neonatal Med. 2012;25: with placenta accreta—a non-randomised pro-
2512–2516. spective study. Vasa. 2017;46:53–57.
21. Chen M, Xie L. Clinical evaluation of balloon 27. Shahin Y, Pang CL. Endovascular interventional
occlusion of the lower abdominal aorta in patients modalities for haemorrhage control in abnormal
with placenta previa and previous cesarean sec- placental implantation deliveries: a systematic
tion: a retrospective study on 43 cases. Int J Surg. review and meta-analysis. Eur Radiol. 2018;28:
2016;34:6–9. 2713–2726.
22. Wu Q, Liu Z, Zhao X, et al. Outcome of 28. Bodner LJ, Nosher JL, Gribbin C, et al. Balloon-
pregnancies after balloon occlusion of the infrare- assisted occlusion of the internal iliac arteries in
nal abdominal aorta during caesarean in 230 patients with placenta accreta/percreta. Cardio-
patients with placenta praevia accreta. Cardiovasc vasc Intervent Radiol. 2006;29:354–361.
Intervent Radiol. 2017;96:1–7. 29. Angstmann T, Gard G, Harrington T, et al.
23. Zeng C, Yang M, Ding Y, et al. Preoperative Surgical management of placenta accreta: a co-
infrarenal abdominal aorta balloon catheter hort series and suggested approach. Am J Obstet
occlusion combined with Bakri tamponade Gynecol. 2010;202:38.e1–38.e9.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA CLINICAL OBSTETRICS AND GYNECOLOGY
Volume 61, Number 4, 808–827
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.

Peripartum
Anesthesia
Downloaded from [Link] by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3dW0s7N5CLiYMlH4wgaY34Fr69rtLH3oWi4+h6VxXUqo= on 10/30/2018

Considerations for
Placenta Accreta
CHRISTINE M. WARRICK, MD,
and MARK D. ROLLINS, MD, PhD
Department of Anesthesiology, University of Utah, Salt Lake City,
Utah

Abstract: Placenta accreta spectrum is becoming more hysterectomy and has potential for mas-
common and is the most frequent indication for peri- sive hemorrhage. Additional peripartum
partum hysterectomy. Management of cesarean delivery
in the setting of a morbidly adherent placenta has sequelae of morbidly adherent placenta
potential for massive hemorrhage, coagulopathies, and include dilutional coagulopathy, transfu-
other morbidities. Anesthetic management of placenta sion reactions, transfusion-related acute
accreta spectrum presents many challenges including lung injury, transfusion associated car-
optimizing surgical conditions, providing a safe and diac overload, electrolyte derangements,
satisfying maternal delivery experience, preparing for
massive hemorrhage and transfusion, preventing coagu- acute kidney injury, and intensive care
lopathies, and optimizing postoperative pain control. unit (ICU) admission. Women with ab-
Balancing these challenging goals requires meticulous normal placentation are also at risk for
preparation with a thorough preoperative evaluation of injury to the bowel, bladder, and ureters;
the parturient and a well-coordinated multidisciplinary as well as thromboembolism.
approach in order to optimize outcomes for the mother
and fetus. PAS requires well-planned coordination
Key words: obstetric anesthesia, placenta accreta, of multiple specialists and presents a vari-
obstetric hemorrhage, massive transfusion, neuraxial, ety of challenges for the anesthesiologist
anesthesia, breastfeeding, preoperative evaluation including preparation for rapid acute hem-
orrhage, massive transfusion, coagulop-
athy, possible airway difficulties, extended
Background operative time, movement to other loca-
Placenta accreta spectrum (PAS) is the tions (eg, interventional radiology) and
most common indication for peripartum optimizing postoperative pain manage-
Correspondence: Mark D. Rollins, MD, PhD, 30
ment. The anesthesiologist needs to bal-
North 1900 East, SOM RM 3A416, Salt Lake City, ance these challenges while ensuring
UT. E-mail: [Link]@[Link] accomplishment of multiple simultaneous
The authors declare that they have nothing to disclose. goals of maternal safety, optimal operating

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 61 / NUMBER 4 / DECEMBER 2018

808 | [Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Anesthesia Considerations for Placenta Accreta 809

conditions, and minimization of risk to the be focused on optimizing all co-morbid

fetus for optimal neonatal outcome. Clear conditions and risk reduction before the
communication between all team members time of delivery.
throughout the detailed planning, intra- Underlying cardiovascular disease is
operative care, and postoperative manage- present in 1% to 4% of pregnancies and is
ment is critical to obtaining the best currently the number one cause of preg-
outcomes. nancy-related death in the United States.1,2
This article details important anes- Careful risk assessment of maternal car-
thetic considerations in managing a pa- diac status should be performed. Pregnant
tient with known PAS. Although the patients with chest pain, syncope, severe
primary focus is patients scheduled for a arrhythmias, high-grade murmurs, or clin-
combined cesarean delivery and hysterec- ically significant shortness of breath should
tomy, information is also provided for undergo appropriate clinical evaluation and
management of unanticipated PAS only work-up. High-risk cardiac diseases include
recognized during delivery. The purpose valvular disease, pulmonary hypertension,
of this article is to allow the obstetrician congenital heart disease, potential aortic
and other team members to have a better disease (eg, Marfan syndrome), and dilated
understanding of the decisions and con- cardiomyopathy. In certain known cardiac
siderations of the anesthesiologist and the conditions (eg, Eisenmenger syndrome, se-
implications of the interventions. vere heart failure) it is ideal if consultation
can occur early enough, irrespective of
presence of PAS, to allow the option of
pregnancy termination rather than risk sig-
Preoperative Assessment nificant maternal morbidity. Cardiac risk
and Preparation scoring systems include the WHO classifi-
Preoperative assessment of the patient cation, the CARPREG scoring system, and
with known or suspected PAS should the ZAHARA Risk Score.3 In general,
focus on components of the medical elements of the cardiac risk assessment are
history and obstetric risk factors that based on presence of prior arrhythmia, prior
may place the patient at increased risk cardiac event, presence of pulmonary hyper-
for hemorrhage or end organ damage. tension, valvular abnormalities, medical
Early consultation and evaluation by the therapy before pregnancy, left ventricular
anesthesiologist allows additional labora- dysfunction, abnormal left heart outflow,
tory or diagnostic studies, timely input cyanotic heart disease, and poor NYHA
from other specialists and referrals if functional class. Changes in circulating
needed, ability to have multidisciplinary volume, heart rate, oxygen demand, and
discussions around unique patient needs, clotting factors exacerbate the underlying
and in some circumstances, time to trans- cardiac condition and increase morbidity
fer care of the patient to another facility if during pregnancy. The most common car-
current resources are inadequate. In addi- diac complications during pregnancy in
tion to a review of major organ systems, women with known heart disease are ar-
obstetric history, and a focused physical rhythmias, thromboembolic events, and
examination, specific elements that sig- heart failure.2,4
nificantly alter planned perioperative In patients with known cardiac disease,
management include prior anesthetic is- the antepartum evaluation should include
sues (eg, known difficult airway, malig- an electrocardiogram, serial echocardiog-
nant hyperthermia), coagulopathy, raphy, and assessment of the effect of
chronic pain, and refusal of blood prod- maternal cardiac status on the fetus.
ucts should be assessed. All efforts should Patients with preexisting cardiac disease

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
810 Warrick and Rollins

may become hemodynamically compro- at greater risk of cardiomyopathy and

mised due to the stress of increased blood thromboembolic events, and have a 5-fold
volume and cardiac output associated increase of in-hospital death.6 Because of
with pregnancy. In addition, the large the increased prevalence of obesity, OSA is
fluid shifts at cesarean delivery combined becoming more common with 15% to 20%
with high risk for massive hemorrhage of obese pregnant women estimated to have
and subsequent hypovolemia place these OSA.6 If OSA is suspected based on screen-
patients at even greater risk. Peripartum ing questionnaires or other risk factors such
management typically includes thrombo- as age, body mass index, presence of hyper-
prophylaxis and optimization of volume tension, or snoring, referral should be made
status. In certain circumstances, intra- to an anesthesiologist and a sleep medicine
operative management may include en- specialist for formal diagnosis and treatment
docarditis prophylaxis, central venous or with CPAP. Difficult intubation and ven-
pulmonary artery monitoring, intraoper- tilation are associated with OSA patients
ative echocardiography, potential need and postoperative pain management should
for ECMO or cardiac bypass, and recov- focus on multimodal analgesia with mini-
ery in a critical care setting. mization of respiratory depression.
Pulmonary adaptations of pregnancy Obesity presents significant challenges
result in increased minute ventilation and to the anesthesiologist and other peripar-
oxygen consumption. The decreased pul- tum care providers. In addition to in-
monary reserve and increased airway creased rates of gestational diabetes and
edema associated with pregnancy make preeclampsia, obesity in pregnancy sig-
respiratory management more challeng- nificantly increases anesthetic risks and
ing. Presence of pulmonary diseases such complications. These include failed epi-
as obstructive sleep apnea (OSA) or re- dural anesthesia, hemodynamic instabil-
active airway disease may guide the deci- ity with neuraxial anesthesia, difficult
sion to begin the case with a general intubation, respiratory depression with
anesthetic and controlled airway rather opioids, aspiration with induction of gen-
than leaving the possibility of converting eral anesthesia, and difficult intravenous
to general anesthesia during a period of access.7 Early preoperative consultation is
potential hemodynamic instability with essential in this population to not only
compromised respiratory function. formulate the optimal perioperative plan,
For asthmatics, spirometry and pulmo- but to prepare them for possible risks,
nary function assessment should be part of complications, and interventions.
routine antepartum care.5 Patients with Women with diabetes should have pre-
poorly controlled asthma despite routine operative and postoperative blood glucose
first-line treatment should be referred to more assessment. Both hyperglycemia and hypo-
intensive medical therapy. It is far better for glycemia are associated with poor surgical
pregnant asthmatics to undergo treatment outcomes. It is prudent to assess blood
and optimization with asthma medications glucose levels for diabetic patients intra-
than undergo surgery and delivery in the operatively. The type of anesthetic used
setting of poorly controlled symptoms and influences glucose levels during surgery.
exacerbations. Pregnant women are at higher General anesthesia is more frequently
risk of influenza infection and its associated associated with hyperglycemia, increased
morbidities and should receive the influenza catecholamines, cortisol, and glucagon
vaccination as part of their care. compared with neuraxial anesthesia.8
Women with either preexisting or gesta- Although the optimal blood glucose level
tional OSA often have associated hyper- for the perioperative period remains uncer-
tensive disorders of pregnancy, diabetes, are tain, most medical governing organizations

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Anesthesia Considerations for Placenta Accreta 811

recommend intraoperative glucose levels section or intrauterine procedures in the

remain <180 mg/dL.8 Treating hyperglyce- setting of postpartum hemorrhage and
mia is equally important in the postoper- retained placenta. However, it is impor-
ative period as glucose levels > 180 mg/dL tant to note that nulliparous women with
are associated with adverse clinical out- no history of cesarean have had PAS.11
comes including surgical site infections, Detailed knowledge of prior abdominal
delayed wound healing and increased operations can facilitate the anesthesiolo-
length of stay.9 gist in estimating time of both incision
Women scheduled for cesarean hysterec- to delivery and total time of a planned
tomy should be evaluated for anemia and cesarean hysterectomy. This will help
have a careful assessment for (history of) determine an appropriate anesthetic plan.
coagulopathies. Laboratory studies evaluat-
ing potential anemia, thrombocytopenia, PHYSICAL EXAMINATION AND
and coagulopathy should be obtained at LABORATORY STUDIES
an early enough time point so that appro- A focused physical examination should
priate hematology consultation and recom- include a baseline blood pressure meas-
mendations can be obtained if a coagulation urement, and an airway, heart, and
disorder is identified. In addition, if signifi- lung examination, consistent with the
cant anemia is noted, early evaluation will ASA Practice Advisory for Preanesthesia
allow time for determination of the under- Evaluation.12 The potential for a difficult
lying cause and potential intervention to airway as assessed by mouth opening,
raise red cell mass (eg, iron or B12 supple- neck mobility, and neck circumference,
mentation, erythropoietin). may guide anesthetic choice in the setting
of an anticipated difficult intubation or
OBSTETRIC HISTORY difficult mask ventilation. When a neu-
A portion of the preoperative evaluation raxial anesthetic is planned, the patient’s
should focus on obstetric history that back should be examined. Sites for possi-
could affect intraoperative hemodynamic ble vascular access should be noted to
stability, coagulopathy, and end organ better plan what lines will be required if
morbidity (eg, preeclampsia with or with- large bore peripheral intravenous access is
out severe features, gestational thrombo- anticipated to be challenging.
cytopenia). Even though many cases of Although not required for standard
PAS are antenatally suspected based on admissions to labor and delivery,12 a com-
imaging, definitive diagnosis is not known plete blood count should be evaluated pre-
until laparotomy. Consequently, it is im- operatively for the presence of anemia. In
portant for the anesthesiologist to obtain addition, a basic metabolic panel, liver
an obstetric history. Placenta accreta is function test, and coagulation tests (PT/
associated with presence of placenta pre- INR, partial thromboplastin time, fibrino-
via and history of previous cesarean with gen level) should also be assessed.
an incidence as high as 67% in women
who have a previa and 5 previous cesar- DISCUSSION OF ANESTHESIA RISKS,
ean deliveries.10 In addition to history of BENEFITS, AND CONSENT
cesarean delivery and presence of placen-
ta previa, multiparity, history of curet- Maternal Considerations
tage, uterine anomalies, and maternal age The anesthesiologist should have a shared
> 35 are risk factors. One should always decision-making approach in discussing
have an index of suspicion for morbid risks and benefits of anesthetic type with
placental adherence in women with pla- the patient to choose an anesthetic with
centa previa and history of cesarean the best fit. Patients wish to understand the

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
812 Warrick and Rollins

risks, benefits, and alternatives for the maternal levels.16 A 2012 Cochrane system-
anesthetic they will receive, and it is essential atic review and meta-analysis noted that for
that this discussion is meaningful and docu- elective cesarean deliveries there was no
mented properly.13 The Joint Commission difference in neonatal outcome between
on Accreditation of Healthcare Organiza- use of general or regional anesthetic
tions states that all elements of the consent techniques.17 This analysis primarily exam-
discussion be documented “in a form, prog- ined acute measures of neonatal wellbeing
ress notes, or elsewhere in the record.”14 including 5-minute APGAR scores and
Even in cases where the plan may be for use need for resuscitation. Beginning in 2003,
of neuraxial anesthesia throughout the case, numerous animal studies noted neuronal
the possibility of general anesthesia should apoptosis in the developing brain after
always be discussed. This discussion should exposure to a wide variety of anesthetic
focus on realistic risks and benefits of agents including volatile anesthetics and
general anesthesia for the fetus and mother, propofol.18 In 2016, the US Food and Drug
including possible aspiration during intuba- Administration (FDA) advisory committee
tion, failed intubation, hemodynamic insta- issued a warning that “repeated or lengthy
bility, prolonged intubation, fetal exposure use of general anesthetic and sedation drugs
to anesthetic medications, invasive monitor- during surgeries or procedures in children
ing procedures, and possible ICU admis- younger than 3 years of age or in pregnant
sion. Neuraxial analgesia carries the risk of women during their third trimester may
unanticipated failure or intolerance leading affect the development of children’s
to conversion to general anesthesia; post- brains.”19 In the FDA discussion, it is noted
dural puncture headache; hemodynamic that in most cases the exposure to general
instability; and less common risks of spinal anesthetic agents or sedating medications in
bleeding, infection, and nerve damage. animal studies noting neuronal apoptosis
In addition, a detailed discussion of was > 3 hours. There are only extremely
likelihood of blood product transfusion limited data about anesthetic exposure in
and risks and benefits of various blood human fetuses. One retrospective study
products should take place and be docu- found no association between the use of
mented. This discussion should occur general anesthesia for cesarean and the
early enough that if the patient has beliefs incidence of learning disabilities at age 5.20
preventing some or all blood product Currently, no general anesthetic agent is
transfusion (eg, Jehovah’s Witness), strat- known to be superior to another, and it is
egies can be optimized to maximize red encouraging to note that two recent well-
cell mass and individual wishes around designed clinical trials found limited expo-
each type of product and cell salvage can sure to general anesthesia early in life was
be understood and documented. It is not associated with long-term neurocogni-
important to discuss each product indi- tive deficits.21
vidually as beliefs vary between individu-
als and in one series, nearly 50% of TIMING AND LOCATION OF
pregnant Jehovah’s Witnesses agreed to DELIVERY/SURGERY
transfusion of some blood products.15 The timing and location for delivery of
PAS by planned cesarean hysterectomy is
Effects of General Anesthesia on the Fetus critically important. ACOG suggests that
With use of general anesthesia, inhaled stable patients with PAS be delivered
halogenated agents readily cross the placen- between 34 0/7 and 35 6/7 weeks’
ta to the fetus. After use of general anes- gestation,22,23 as improved outcomes are
thesia for cesarean delivery, fetal levels noted for patients with planned cesarean
of isoflurane reach approximately 70% of hysterectomy as opposed to unplanned

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Anesthesia Considerations for Placenta Accreta 813

urgent delivery in the setting of antenatal multidisciplinary teams, having all surgi-
bleeding.24 The patient and surgical plan cal instrumentation available, and rou-
should be discussed among a multidisci- tinely handling massive hemorrhage and
plinary team, ideally with initial meeting transfusion.
occurring early in the third trimester of Performing cases in a hybrid operating
pregnancy. All members of the team room with interventional radiology capa-
should be updated about relevant changes bility is increasingly being performed as
to the patient’s medical status as she this minimizes the risk of intra-arterial
progresses in her pregnancy (eg, antepar- balloon catheter displacement upon pa-
tum bleeding). Contingency plans for tient transport from interventional radi-
urgent or emergent cesarean in the setting ology to the operating room. However,
of preterm premature rupture of mem- data supporting routine utilization of
branes, preterm labor, significant hemor- interventional radiology for patients with
rhage or other obstetric and fetal concerns abnormal placentation is lacking.26
should be in place with all team members
informed of such plans.22 BLOOD PRODUCT PREPARATION
Cesarean hysterectomy for patients Type and screen should be completed
with abnormal placentation should be before the date of surgery to evaluate
performed in an institution where physi- and plan for antibodies and difficult
cians trained in maternal fetal medicine, crossmatch. Type and crossmatch should
obstetric anesthesia, neonatology, critical be completed before moving the patient to
care, pelvic surgery, vascular surgery, and the operating room for surgery. Packed
interventional radiology are available. red blood cells (RBCs) and FFP should be
ICU space for postoperative care should available at the time of surgery in the
be available. The facility should have a operating room for planned cesarean
blood bank capable of massive transfu- hysterectomy. Although considerable
sion with readily available fresh frozen variation exists, an example blood prod-
plasma (FFP), cryoprecipitate, and coag- uct preparation might include 6 RBC
ulation factors. Laboratory capabilities units and 4 FFP units.27 It is also impor-
should be considered, given the potential tant to communicate the potential need
for multiple laboratory samples to con- for platelets and cryoprecipitate with the
tinually assess hemoglobin and hema- blood bank as early as possible, as these
tocrit levels as well as frequent blood products may be periodically in short or
gas assessments. On the basis of these limited supply at some institutions.
factors, a tertiary care center or center
of excellence for placenta accreta is ANTEPARTUM ANTICOAGULATION
often preferred, since outcomes are im- CONSIDERATIONS
proved for patients managed at a center The National Partnership for Maternal
of excellence for PAS with a multidisciplinary Safety (NPMS), recently created a safety
team.24,25 bundle that suggests recommendations
In addition to hospital facility, patients for venous thromboembolism (VTE)
should be managed in an area of the prophylaxis.28 If these recommendations
hospital that is most capable of handling are followed, it is likely more pregnant
the case. There should be enough room patients will receive pharmacologic
for a multidisciplinary team and easy prophylaxis in the antepartum period.
access to surgical supplies. Although this For pregnant patients requiring either
may be accomplished on the labor and unfractionated heparin or low–molecular-
delivery unit of some hospitals, the main weight heparin, doses required may be great-
operating room may be better suited for er than typical in nonpregnant patients

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
814 Warrick and Rollins

secondary to changes in plasma volume, for massive transfusion. The state inpa-
increased levels of procoagulant proteins, tient database of New York cited placenta
and lower levels of anticoagulant proteins. accreta as the most common cause for
For each anticoagulation drug and dosing massive transfusion in obstetrics.32 Pa-
regimen there are specific contraindications tients should have at minimum two large
or need for delay to placement of neuraxial bore IVs (16 or 14 G) placed before
blockade. In cases where the anesthetic plan incision to facilitate rapid blood and fluid
for PAS has a neuraxial component, a administration. Those with difficult IV
detailed multidisciplinary discussion includ- access should have a large bore central
ing anesthesiology, obstetrics, and other line or introducer sheath in place for large
specialty physicians (eg, hematology, cardi- rapid volume resuscitation. Although a
ology) is critical in determining an optimal central line may also be useful for vaso-
anticoagulation plan with appropriate risk pressor infusions as a method to tempo-
and benefit considerations in both the case rize hemodynamic instability due to
of scheduled and urgent cesarean hysterec- sympathectomy or volume loss and as
tomy. A 2018 consensus statement by The an estimate of volume status, it is not
Society for Obstetric Anesthesia and Peri- routinely used in all institutions.
natology with regard to thromboprophylaxis Invasive hemodynamic monitoring is
during pregnancy provides decision aids for prudent in women with PAS as it allows
use of neuraxial anesthesia (Fig. 1).29 for continuous evaluation of blood pres-
sure and confirmation of perfusing heart
rate and rhythm in the setting of potential
Intraoperative Management rapid fluid shifts. In addition, it provides
and Considerations quick access for blood gases, blood
Although many of the perioperative anes- count, coagulation tests, and other labo-
thetic considerations are similar to those for ratory studies. In cases with significant
cesarean delivery,30 a significant amount of maternal cardiac pathology, intraopera-
detailed additional preparation is needed tive echocardiography is indispensable in
in order to optimize maternal and fetal optimizing fluid management, titrating
outcomes in cases of PAS. Consequently, ionotropic and vasoactive medications,
an itemized checklist (Table 1) is useful to and determining ventricular function.
ensure all planning and operational details Even in cases of no known maternal
have been accomplished. Before entering cardiac morbidity, rapid availability of
the operating room, it is important to intraoperative echocardiogram is indis-
complete an interim history and physical pensable in helping diagnose causes of
examination, confirm availability of blood unexplained hypotension or hypoxia.33
products, neonatal and other specialty Intraoperative echocardiography provides
teams, consents (cesarean delivery, hyster- real-time assessment of preload, myocar-
ectomy, blood products, and anesthetic dial contractility, cardiac output, valvular
plan), availability of invasive monitoring, function, ventricular wall motion abnor-
rapid infusers, vasopressors, and airway malities, and presence of thromboemboli.
equipment. A detailed extended timeout Consequently, it can markedly assist in
before entering the operating room is ideal diagnosing critical events, guide fluid re-
to ensure everything is in place. placement, and assist in titration of ap-
propriate classes of vasopressor therapy.
VASCULAR ACCESS AND INVASIVE
MONITORING ANESTHETIC TECHNIQUES
Patients with abnormal placentation are Anesthesia for planned cesarean hyster-
at risk for massive hemorrhage and need ectomy may comprise general anesthesia,

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Anesthesia Considerations for Placenta Accreta 815

FIGURE 1. Thromboprophylaxis and neuraxial anesthesia during pregnancy. A flowchart to

assist with neuraxial procedures in an obstetric patient receiving anticoagulation with either
unfractionated heparin (A) or LMWH (B). aPTT indicates activated partial thromboplastin
time; GA, general anesthesia; LMWH, low–molecular-weight heparin; SEH, spinal epidural
hematoma; SQ, subcutaneous; UFH, unfractionated heparin. Figures are created from in-
formation presented in Leffert et al.29

neuraxial anesthesia, or a combination of compelling imaging findings, PAS is not

both (Table 2). Although many different diagnosed when the abdomen is open.
techniques have been successfully used, Choice of technique may often be indi-
no data from prospective trials supports vidualized and determined by patient
an “optimal” anesthetic for planned ce- comorbidities, degree and certainty of
sarean hysterectomy. In patients with placental invasion, specific patient prefer-
PAS and planned cesarean hysterectomy, ences, resource availability, and provider
massive and rapid intraoperative blood familiarity and preference.
loss is common. Reported blood loss Standard preoperative fasting guidelines
often ranges from 2000 to 5000 mL and should be followed for planned cases. Irre-
significant coagulopathies can frequently spective of planned anesthetic technique,
develop. Yet in a few cases, despite aspiration prophylaxis should be performed

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
816 Warrick and Rollins

TABLE 1. Perioperative Checklist for TABLE 1. (Continued)

Management of Morbidly
Intraoperative
Adherent Placenta • If applicable induction of general anesthesia
Antenatal Period with rapid sequence technique
• Multidisciplinary meeting with maternal fetal • Additional IV access, arterial line, central
medicine team and other specialty consultants venous access placed as appropriate
• Baseline blood laboratory studies drawn and • Use of forced air warmer to maintain core
reviewed temperature ≥ 36°C
• Anesthesiology consult focused on • Placement of ureteral stents or arterial balloon
comorbidities, obstetric history, and any lab catheters by other specialties if appropriate
abnormalities • Preincision timeout
• Additional periodic multidisciplinary meetings • Prophylactic antibiotic administration
to formulate optimal delivery and surgical plan • Maternal mean arterial pressure and heart rate
• Contact scheduler and confirm booking of maintained near baseline prior to delivery
appropriate delivery location After Delivery
• Notify IR, blood bank, NICU, and cell salvage • Possible conversion to general anesthesia
technicians if applicable (before potential for hemorrhage)
Day Prior or Day of Surgery • Tranexamic acid administration
• Review staging and location of adherent • Fluid administration and blood products
placenta guided by clinical judgment and laboratory
• Anti-D prophylaxis administered if appropriate studies
• Check fasting guidelines ordered and followed • Periodic laboratory studies (blood gas,
• Confirm availability of specialty teams complete blood count, coagulation/clotting
(eg, IR, gynecologic surgery, cell salvage, studies)
and OR nursing) • Remain in contact with blood bank to
• Repeat examination of airway, heart, lungs, communicate transfusion needs
spine, and IV access sites • Possible transport to IR suite or use of arterial
• Review laboratory studies (complete blood balloons if appropriate
count, coagulation screen, electrolytes, liver • Possible need to redose antibiotics
function tests) Case Completion
• Review any anticoagulation • Consider extubation (awake)
• Confirm blood product availability • Transport to surgical ICU
• Cross match 6 units PRBCs, 4 units FFP, • Pain control orders in place
1 unit platelets, with additional units available • Plan for VTE prophylaxis reviewed
• Confirm consents for cesarean delivery, • Debrief
hysterectomy, blood products, and anesthetic
plan FFP indicates fresh frozen plasma; ICU, intensive care unit;
• Before entering OR, review (preprocedure IR, interventional radiology; MFM, maternal fetal medicine;
NICU, neonatal intensive care unit; OR, operating room;
timeout): PRBCs, packed red blood cells; VTE, venous thromboemb-
• Surgical plan olism.
• Anesthetic plan (general or neuraxial with Table includes information presented in (1) Walker et al31
possible conversion to general anesthesia) and (2) Panigrahi et al.27
• Plan for invasive monitoring (arterial line,
central venous access, intraoperative cardiac for all patients due to the possible need to
echo)
• Blood products present in OR
convert to general anesthesia and increased
• Neonatal isolette nearby and neonatal risk of aspiration in pregnant women. Some
resuscitation team available anesthesia providers prefer the use of H2
• Rapid fluid/blood product infuser available blockers, which increase gastric pH within an
Intraoperative hour in pregnant women. Combined antacids
• Aspiration prophylaxis and H2 blockers are more effective in increas-
• Baseline vitals obtained, standard monitors ing gastric pH than antacids alone or no
placed
• Appropriate fluids, vasopressors, and airway pharmacologic intervention. Current ASA
equipment ready guidelines for Obstetric Anesthesia state,
• If applicable, spinal or epidural catheter “…before surgical procedures (eg, cesarean
insertion with test dose for epidural delivery or postpartum tubal ligation),

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
TABLE 2. Specific Considerations Regarding Anesthetic Options for Patients With Placenta Accreta Spectrum
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.

Spinal with Conversion Epidural With CSE With Conversion

Type of Anesthetic to GA Epidural Conversion to GA CSE to GA GA
Onset Rapid Slow Slow Rapid Rapid Rapid
Duration Unlimited Titratable block Unlimited Titratable block Unlimited Unlimited
Maternal Can meet neonate, then Can meet neonate, Can meet neonate, then Can meet neonate, Can meet neonate, then Mother under GA
satisfaction under GA anxiety and under GA anxiety and under GA
discomfort with time discomfort with time
Airway Secured during Controlled by patient Secured during Controlled by patient Secured during Secured
conversion to GA conversion to GA conversion to GA
during possible HD during possible HD during possible HD
instability instability instability

Anesthesia Considerations for Placenta Accreta

Neonatal anesthetic Minimal Minimal Minimal Minimal Minimal Limited
exposure
Neuraxial Intrathecal opioids Epidural opioids or Epidural opioids or Intrathecal/epidural Intrathecal/epidural None
postoperative continuous continuous opioids or continuous opioids or continuous
pain control postoperative postoperative epidural postoperative epidural postoperative epidural
epidural infusion infusion infusion infusion
Failed/incomplete < 1% 5%-10% 5%-10% ≤ 5% ≤ 5% NA
neuraxial
blockade
Effect on uterine ≥ 1 MAC inhaled None ≥ 1 MAC inhaled None ≥ 1 MAC inhaled ≥ 1 MAC inhaled
tone anesthetic decreases anesthetic decreases anesthetic decreases anesthetic decreases
uterine tone uterine tone uterine tone uterine tone
Hemodynamic effect Sympathectomy, possible Sympathectomy, Sympathectomy, possible Sympathectomy, possible Sympathectomy, possible No sympathectomy.
HD instability during possible HD HD instability up to HD instability HD instability up to Considered more HD
first 60 to 90 min after instability throughout 60-90 min following throughout case 60-90 min following stable. Inhaled
spinal case last epidural bolus last epidural bolus anesthetics/propofol
[Link]

cause decreased
myocardial depression
and decreased SVR
Risk of PDPH < 1% Approximately 1%, but Approximately 1%, but Approximately 1%, but Approximately 1%, but None
50% if known dural 50% if known dural 50% if known epidural 50% if known epidural
puncture puncture needle dural puncture needle dural puncture

CSE indicates combined spinal epidural; GA, general anesthesia; HD, hemodynamic; MAC, minimum alveolar concentration; NA, not available; PDPH, postdural puncture headache;
SVR, systemic vascular resistance.

817
 DKA
818 Warrick and Rollins

consider the timely administration of the rate of failed intubation for cesarean
nonparticulate antacids, H2 receptor antag- delivery under general anesthesia to be
onists, and/or metoclopramide for aspiration 1:533.36 Although considered safe and ap-
prophylaxis.”12 propriate, the relative risk of mortality
Antibiotics for surgical wound infection using general anesthesia for cesarean deliv-
prophylaxis should be administered ery is 1.7 times that of neuraxial anesthesia,
before incision. Although there are no with the majority associated with intubation
specific recommendations with regard failure or induction difficulties.37 The overall
to antibiotic prophylaxis for combined rate of anesthesia complications leading to
cesarean hysterectomy, a recent ACOG maternal mortality is incredibly low. Recent
Practice Bulletin provides current dosing data from the United States Pregnancy
recommendations for planned hysterec- Mortality Surveillance System found anes-
tomy (which are the same recommen- thesia complications represented only 0.2% of
dations for cesarean delivery).34 A first- pregnancy-related mortality between 2011
generation cephalosporin is the first-line and 2013, or ~1 death per 2.9 million live
choice (2 g IV; 3 g IV if ≥ 120 kg). If births.38
a significant penicillin or cephalosporin Choice of induction agent is often based
allergy is of concern, a combination of on the hemodynamic stability of the mother.
metronidazole or clindamycin plus genta- Propofol is a rapid onset hypnotic but
micin or aztreonam is recommended. Re- decreases both cardiac output and mean
peated dosing of the cephalosporin should arterial pressure, so is less desirable in
occur at 4 hours to maintain appropriate hemodynamically unstable patients. Etomi-
plasma levels. In addition, antibiotic ad- date also provides quick onset because of its
ministration should be repeated anytime high lipid solubility and has minimal effects
estimated blood loss exceeds 1500 mL. on the cardiovascular system, which may
be desirable. However, it can increase
General Anesthesia the possibility of seizures in patients at risk,
The option of general anesthesia provides a and induction doses of etomidate can result
rapid, reliable, and titratable anesthetic with- in decreased neonatal cortisol production
out concern for additional hemodynamic (though the clinical significance is unknown).
instability secondary to sympathectomy Ketamine functions as an amnestic, hypnotic,
created by neuraxial anesthetic techni- and analgesic. It acts as a sympathomimetic,
ques. However, general anesthetics increasing arterial pressure, heart rate, and
decrease peripheral vascular resistance cardiac output. Because of these character-
and are direct myocardial depressants. istics, ketamine, similar to etomidate, is an
With general anesthesia for a planned appropriate choice for a patient with hemo-
case, the airway is secured before inci- dynamic compromise. Standard induction
sion, which is ideal for patients with doses (eg, 1.5 mg/kg) do not result in neonatal
presumed difficult mask ventilation or in- compromise. However, significantly larger
tubation. Optimal positioning for securing doses can cause uterine hypertonia, which
the airway should occur before laryngo- may result in decreased uterine perfusion.
scopy, with backup instrumentation and Following induction, halogenated in-
airway equipment immediately available. haled anesthetics are frequently used for
Availability of video laryngoscopy is maintenance before delivery, but later
highly recommended, especially in obese reduced or eliminated to minimize uterine
patients.35 Following preoxygenation, rap- atony. Other agents that do not affect
id sequence induction with cricoid pressure uterine tone (eg, propofol, opioids, ben-
is used to minimize aspiration risk. A recent zodiazepines, and muscle relaxants) are
multi-institutional database review found then often administered.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Anesthesia Considerations for Placenta Accreta 819

General anesthesia may be preferable manipulation of the peritoneum. Neuraxial

for patient comfort in the setting of large techniques carry the additional risks of
bore IV access, arterial line placement, failure or intolerance and subsequent con-
central venous line placement, and mas- version to general anesthesia. The use of
sive transfusion with use of rapid infusion neuraxial anesthesia creates a sympathec-
devices. However, this technique prevents tomy, which may make maintenance of
the mother from viewing the neonate appropriate blood pressure and perfusion
immediately after birth and exposes the more challenging with acute severe hemor-
fetus to general anesthetics (see the Effects rhage. In addition, if significant coagulop-
of General Anesthesia on the Fetus sec- athy develops during the procedure, the
tion). In addition, anesthetic exposure patient is at increased risk of epidural
time may be further increased by ureteral hematoma and removal of the catheter
stent placement and/or intra-arterial bal- may need to be delayed.
loon placement, should either of those
interventions be selected. Combined Neuraxial and General
Anesthesia
Neuraxial Anesthesia The combination of a neuraxial technique
Neuraxial anesthesia, in the form of spinal, followed by conversion to general anesthesia
epidural, or combined spinal epidural, after delivery offers some of the benefits of
offers the benefits of limiting fetal exposure both techniques. Anesthetic exposure to the
to anesthesia, avoiding risks of airway fetus is minimized and the mother is able to
interventions in the parturient (unless the experience the delivery. In the case of a spinal
surgical block becomes inadequate), pre- anesthetic, intrathecal opioids may be ad-
venting anesthetic contribution to uterine ministered for postoperative analgesia, and
atony, and allowing the mother to much of the sympathectomy may be abated
experience the delivery. Use of spinals or by the time of conversion to general anes-
epidurals also allow the administration of thesia following delivery (depending on dos-
neuraxial opioids to assist in postoperative ing). With the use of an epidural or CSE, the
analgesia. Presence of an epidural catheter neuraxial catheter can be used to prolong
can be used for postoperative analgesia and block duration if there is an extended period
to improve patient compliance in supine of preincision preparation (stents or intra-
positioning after closure of the groin sites arterial balloon placement). Dosing can often
from the intra-arterial balloons. Although be timed to decrease the impact of the
neuraxial has many benefits, it is important remaining sympathectomy at the time of
to point out the use of a “single shot” spinal planned conversion to general anesthesia
as the sole anesthetic is suboptimal for this following delivery. However, with all the
type of case and will likely lead to con- techniques of planned conversion to general
version to general anesthesia due to its anesthesia there is always risk of unantici-
limited duration. A catheter-based neurax- pated difficult airway and intubation under
ial technique (epidural or CSE) allows for suboptimal circumstances owing to hemor-
prolongation of blockade when needed, rhage and hemodynamic instability. These
and the ability to utilize neuraxial anesthe- planned conversion techniques would be a
sia for urgent postoperative procedures if it poor choice for the patient with a suspected
is left in situ. However, patients may difficult airway.
become anxious or restless with this tech-
nique due to prolonged surgical procedure VASOPRESSORS
time, necessitating the use of anxiolytics or For many years, ephedrine was the stand-
sedative medications. The patient is also ard treatment for spinal-induced hypoten-
subject to nausea and discomfort during sion in the setting of cesarean delivery.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
820 Warrick and Rollins

However, about 10 years ago it was deter- has occurred vasopressor selection should
mined that ephedrine crosses the placenta be determined primarily by the current
to a greater extent than phenylephrine and status of the patient and underlying cause
is associated with fetal acidosis.39 Phenyl- of hypotension.
ephrine and ephedrine are both safe op-
tions in pregnancy in reasonable doses, but FLUIDS AND TRANSFUSION
phenylephrine is now the preferred pressor Intravenous fluids and blood products
before delivery and is typically adminis- should be administered warm to avoid
tered as an infusion following initiation of hypothermia induced coagulopathy. Hem-
neuraxial anesthesia.40 Vasopressor agents orrhage during PAS cases is frequently
and dosing are frequently titrated to keep acute and rapid. Use of massive transfusion
both maternal blood pressure and heart protocols helps to ensure blood products
rate parameters near baseline in an effort to can be administered rapidly and in appro-
maintain cardiac output and uteroplacen- priate ratios. The optimal blood transfusion
tal perfusion. Glycopyrrolate may be used ratio of packed RBCs to FFP, in obstetric
to increase maternal heart rate if needed. hemorrhage is unknown. However, data
Use of norepinephrine to treat spinal-in- from trauma cases supports avoidance of
duced hypotension during cesarean deliv- dilutional coagulopathy by administering
ery is currently being studied, as the FFP and platelets early in the resuscitation
combined alpha and beta activity may process in a [Link] ratio of RBCs: FFP:
prevent the reflex bradycardia often platelets (a [Link] ratio implies use of the
seen with phenylephrine administration. more common platelet thrombophreresis
Studies comparing phenylephrine versus unit rather than individual platelet
norepinephrine infusions suggest that use “packs”). Although more recent trauma
of norepinephrine in pregnancy is reason- studies did not demonstrate a mortality
able and may result in a relatively higher difference in using a [Link] versus a [Link]
heart rate and cardiac output compared transfusion ratio, the majority of academic
with phenylephrine alone. Nonetheless, the centers with MTPs use a 1:1 ratio of RBC’s
clinical significance is uncertain since neo- and plasma.43 The ACOG committee opin-
natal outcomes are similar with the use of ion for PAS recommends transfusion of
both drugs, and further study is needed blood products during hemorrhage in a 1 to
before norepinephrine is considered a first- 1 ratio.22
line agent of choice.41 Epinephrine use As coagulation status and red cell mass
during pregnancy is typically reserved for can change rapidly, frequent laboratory test-
acute resuscitation settings (eg, anaphy- ing is likely beneficial until hemorrhage is
laxis, cardiac arrest). There is concern for controlled. Point of care testing can signifi-
reduced uterine blood flow after adminis- cantly reduce the time delay between sam-
tration of epinephrine because of the sig- pling and results, which may be beneficial in
nificant uterine vasoconstriction that may guiding transfusion management. Although
occur, theoretically increasing fetal hypo- use of viscoelastic testing (TEG and RO-
xia. However, in certain circumstances, the TEM) may ultimately prove to improve
best fetal resuscitation is maternal resusci- outcomes, there is currently insufficient evi-
tation, and this may require use of epi- dence to support that either standard coag-
nephrine. Vasopressin is a second-line ulation testing or viscoelastic testing is better
vasopressor option for management of in management of obstetric hemorrhage
hypotension in the septic pregnant when result availability times are similar.43
patient.42 However, in theory, vasopressin Cryoprecipitate should be readily available
may activate uterine V1a receptors and to treat disseminated intravascular coagula-
stimulate uterine contractions. Once delivery tion and low fibrinogen levels.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Anesthesia Considerations for Placenta Accreta 821

The recent results of a large multi- cells bacteria, and other procoagulants.46
institutional trial note the benefit of Although there are two reports of hypo-
tranexamic acid (TXA) in reducing mortal- tension related to the use of cell salvage in
ity from hemorrhage in women experienc- the peripartum period, most cases resulted
ing postpartum bleeding when 1 g is in good outcomes and the technique is
administered within 3 hours of delivery.44 considered appropriate for use in massive
Most importantly, the results helped pro- obstetric hemorrhage.47 While not evidence
vide evidence of safety with no increase in based, many institutions advocate for using
rates of thromboembolic events including wall suction to remove all amniotic fluid
pulmonary embolism, myocardial infarc- and irrigate the field with additional crys-
tion, or stroke. Consequently, administra- talloid, which is then removed before
tion of TXA is likely beneficial following initiation of cell salvage suction. In patients
delivery in cesarean hysterectomy cases but who are Rh−, anti-D immunoglobulin
proof is lacking. should be administered in coordination
Although fibrinogen levels are elevated in with Kleihaur Betke testing to prevent
pregnancy and fibrinogen levels serve as a alloimmunization as variable amounts of
predictor of severe postpartum hemorrhage, fetal RBCs are transfused to the mother
there is currently no evidence to assure that with use of cell salvage.
maintenance of fibrinogen at ≥ 200 mg/dL
and preemptive use of fibrinogen concen- UNANTICIPATED PAS
trate is beneficial in management of obstet- In the case of unanticipated PAS,
ric hemorrhage.45 Conversely, levels the anesthesiologist should focus on
< 200 mg/dL are associated with an in- maintaining hemodynamic stability, man-
creased risk of hemorrhagic complications aging massive hemorrhage and keeping
and some authorities advise keeping levels open communication with the obstetric
above this threshold. surgical team. Massive transfusion proto-
Unlike use of TXA, the “off-label” use cols should be activated and open com-
of recombinant factor VIIa (rFVIIa) for munication with the blood bank initiated.
treatment of postpartum hemorrhage has The anesthesiologist managing a patient
been associated with an increased risk of with unanticipated PAS should call for
thromboembolic events. Current expert additional anesthesia staff to assist in
opinion and hemorrhage protocols in placing invasive monitoring, additional
obstetrics recommend rFVIIa be reserved peripheral IV access, and/or central access
for use as a final option when all other for volume resuscitation, as well as mobi-
treatments have failed to rectify ongoing lization and ongoing use of rapid trans-
severe hemorrhage and coagulopathy. fusion devices. Vasopressors and
In addition to transfusion of allogenic uterotonics should be made readily avail-
blood products, cell salvage has been used able. The patient should be maintained
successfully in numerous cases of obstetric euthermic with the use of fluid warmers
hemorrhage despite the theoretical concern and forced air warming systems. Addi-
of amniotic fluid embolism and is appro- tional personnel are often needed to facil-
priate for use in cesarean hysterectomy itate blood sampling for frequent
cases. Although the exact component of laboratory and point of care testing, and
amniotic fluid responsible for clinical retrieval of blood products. Blood prod-
events associated with amniotic fluid em- ucts should be administered based on the
bolism remain unknown, the use of cell clinical situation as opposed to waiting for
salvage with a leukocyte reduction filters laboratory results to guide management.
has been demonstrated to remove tissue Conversion to general anesthesia with a
factor, alpha fetoprotein, fetal squamous controlled airway should be considered

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
822 Warrick and Rollins

early if there is concern for hemodynamic the time of spinal placement can be used
instability. Open communication with the for postoperative pain management, as
obstetric team can facilitate a discussion it provides ~18 to 24 hours of analgesic
about other surgical management options benefit. Common side effects of intra-
(including intrauterine balloon tampo- thecal morphine include pruritus and
nade, compression sutures, or hysterec- nausea. The potential for delayed respi-
tomy), calling for additional assistance ratory depression with administration of
from general or gynecologic surgical col- neuraxial morphine warrants respiratory
leagues, or embolization via intervention- monitoring. However, rates of clinically
al radiology. If the patient remains significant respiratory depression are
hemodynamically stable following discov- <0.1% with current intrathecal dosing
ery of PAS, the team may discuss trans- ( ≤ 250 mcg). Higher intrathecal doses
ferring the patient to a facility or area of (150 to 250 mcg) prolong analgesia, but
the hospital with improved capability for increase rates of unwanted side effects
handling PAS. compared with lower intrathecal doses
(50 to 100 mcg).49 If an epidural was
placed, 1.5 to 3 mg of epidurally admin-
Postoperative Management istered morphine provides analgesic ben-
Management of PAS with cesarean hyster- efit, but also has side effects of pruritus
ectomy can carry significant morbidity. and nausea that are reduced with lower
Ideally the need for an ICU admission dosing (1.5 mg).50 Women with OSA or
following surgery should be determined morbid obesity are at increased risk for
with a bed reserved before starting the postoperative respiratory depression and
case, but in some circumstances, unantici- continuous pulse oximetry monitoring
pated intraoperative complications may should be utilized during the first day of
warrant unanticipated postoperative inten- recovery.6
sive care. In a retrospective single institu- Continuous epidural infusions of dilute
tion study, need for ICU admission of local anesthetic with opioid can used for
patients with placenta accreta was postoperative analgesia. They normally
~40%.48 Irrespective of postoperative loca- provide excellent analgesia, but carry
tion, close nursing supervision, continuous disadvantages of decreased mobility (po-
monitoring and serial examination, and tentially increasing VTE risk) and poten-
laboratory studies are likely beneficial. tially complicating use of pharmacologic
VTE prophylaxis.51 This technique may
PAIN MANAGEMENT be most useful in patients with a history of
The relatively more extensive incision and chronic pain.
surgical procedure make postoperative Although systemic opioids are often
pain management more challenging fol- used in the form of patient-controlled
lowing a cesarean hysterectomy compared analgesic and have historically been the
with an uncomplicated cesarean delivery. primary treatment for postoperative pain,
Adequate pain control is best achieved use of opioid sparing adjuncts are beneficial
using a multimodal approach. Options to to reduce opioid requirements, minimize
consider include neuraxial analgesia, re- unwanted side effects of opioid therapy
gional blocks, systemic and neuraxial (sedation, pruritus, nausea, prolonged
opioids, and nonopioid systemic analgesics bowel dysfunction) and potentially
including acetaminophen and nonsteroidal decrease rates of progression to chronic
anti-inflammatory drugs (NSAIDs). opioid dependence. Systemic opioids
If a spinal component is a planned part should be provided only on an as-needed
of the anesthetic, intrathecal morphine at basis for breakthrough pain, with patients

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Anesthesia Considerations for Placenta Accreta 823

receiving scheduled doses of nonopioid of analgesic efficacy are inconsistent, and

analgesics. Once the patient can take oral there is a high level of fetal transfer.51
medications, oral opioids are preferred in Gabapentin is not commonly used for ma-
most patients as there is no evidence that ternal postoperative pain because of increased
intravenous opioids provide superior anal- sedation, visual disturbances, and concern
gesic benefit. with regard to breast milk transfer. Ketamine
NSAIDs such as ketorolac and ibuprofen may improve postoperative pain but it has
provide significant opioid sparing pain relief significant undesired side effects for a new
but should be used with caution if there is mother including hallucinations, dizziness,
concern for renal impairment or potential disturbed dreams, and lightheadedness. The
for ongoing bleeding given the platelet degree and significance of breast milk transfer
dysfunction associated with these medica- has not been well studied.
tions. Ketorolac decreases postcesarean Wound infiltration with local anesthetics
pain scores when administered in scheduled and use of transversus abdominis plane
doses (eg, 30 mg IV every 6 h). The patient (TAP) blocks may also be used to improve
should be transitioned to scheduled oral postoperative analgesia. The efficacy of
ibuprofen (eg, 600 mg ibuprofen every 6 h) both these options in the setting of neurax-
for continuation of the NSAID benefits ial morphine and nonopioid systemic
once advanced to oral intake.49 medications remains uncertain. A meta-
Acetaminophen provides postoperative analysis examining postoperative epidural
analgesia and reduces the need for opioids. use compared with local anesthetic wound
A study examining scheduled acetaminophen infiltration noted comparable pain scores in
dosing with as-needed opioids versus as- the first 48 hours following abdominal
needed acetaminophen/opioid combination surgery.53 TAP blocks are proven most
tablets found that scheduled acetaminophen effective in patients who did not receive
resulted in less total opioid consumption either neuraxial morphine or placement of
following cesarean delivery.52 In 2009, the an epidural catheter for postoperative pain
FDA decreased the recommended maximum control. In addition, TAP blocks can be
daily dose of acetaminophen from 4000 to used for patients with poorly controlled
3250 mg. A scheduled dose of 1 g oral postoperative pain despite other treat-
acetaminophen every 8 hours is reasonable ments. TAP blocks involve injection of a
following cesarean delivery once the patient is long acting local anesthetic between inter-
advanced to oral intake. There is currently no nal oblique and transverse abdominis
evidence that intravenous preparations are muscle layers using ultrasound guidance.
superior to oral acetaminophen. Avoiding There can be analgesic sparing in the mid-
opioid/acetaminophen combination medica- line depending on placement location and
tion reduces unnecessary opioid use and helps technique. In some instances, placement of
to avoid exceeding recommended maximum a catheter in the muscle plane for continu-
doses of acetaminophen.49 ous local anesthetic administration (eg,
Other agents such as gabapentin and 0.5% ropivacaine subfascially at 5 mL/h)
ketamine are not typically considered may be optimal as duration from a single
first-line analgesic agents and are often injection is often <12 hours.54
reserved for patients with a history of
chronic pain. Gabapentin may provide BREASTFEEDING
some pain reduction and improved ma- There are numerous documented benefits
ternal satisfaction with a single 600 mg to breastfeeding and it is considered safe
oral preoperative dose or ongoing treat- to breastfeed following general and/
ment with 300 mg every 8 hours for more or neuraxial anesthesia (Fig. 2). All
severe postoperative pain, but the findings anesthetics and analgesics enter the breast

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
824 Warrick and Rollins

FIGURE 2. Breastfeeding following general anesthesia. This graphic depicts the specific breast-
feeding safety of various medications that may be administered for a general anesthetic. IV indicates
intravenous; NMBAs, neuromuscular blocking agents. Created from information presented in
Wanderer and Rathmell.55

milk to some degree and are transferred to studies have measured volatile anesthetic gas
the neonate. However, drug levels in levels in breast milk following administration
breast milk follow a passive diffusion of volatile based general anesthesia. How-
model and correlate with maternal plas- ever, it is known that inhalational agents are
ma concentrations. As a result, as mater- rapidly excreted and have poor oral bioavail-
nal plasma drug concentrations decrease, ability. There are also no studies evaluating
so do breast milk drug concentrations. In neuromuscular blocking agent transfer to
addition, it is important to understand breast milk. It is unlikely they cross the
that the dose the infant receives is based blood-milk duct membranes secondary to
on bioavailability, and drugs that are their large size and low lipid solubility. They
administered intravenously have a differ- also have a poor oral bioavailability. Con-
ent bioavailability when consumed orally. sequently, although trace amounts of some
Notably, early after birth the volume of anesthetic agents are present in breast milk, it
colostrum consumed by the infant is far is safe to breastfeed following general anes-
less than the volume of breast milk con- thesia and is typically the practice of mothers
sumed later on in life. who had received an urgent cesarean delivery
Induction agents can be used safely in under general anesthesia.
breastfeeding mothers.56,57 Propofol has low Similarly, mothers can typically breastfeed
oral bioavailability and is rapidly metabo- while taking medications for postoperative
lized in infants. Etomidate has rapid clear- pain. Transfer of local anesthetics from
ance and is undetectable in breast milk neuraxial analgesia is minimal. Acetamino-
4 hours after maternal administration. There phen and ibuprofen can both be safely used
are currently no human studies evaluating in breastfeeding mothers without concern
the transfer of ketamine in breast milk. No for the neonate as the relative infant dose

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Anesthesia Considerations for Placenta Accreta 825

is typically <3% of the mother’s. It is safe United States: is placenta accreta an increasingly
for mothers to breastfeed with standard important contributor? Am J Obstet Gynecol. 2015;
prescribed doses of opioids for treatment of 213:384.e1–384.e11.
2. Elkayam U, Goland S, Pieper PG, et al. High-risk
postoperative pain following cesarean deliv- cardiac disease in pregnancy: part I. J Am Coll
ery. However, caution and physician consul- Cardiol. 2016;68:396–410.
tation is recommended for mothers who 3. Hebson C, Saraf A, Book WM. Risk assessment
require larger doses of opioids. The 2 excep- and management of the mother with cardiovas-
tions to this are meperidine and codeine. cular disease. Clin Perinatol. 2016;43:1–22.
4. Ntiloudi D, Giannakoulas G, Parcharidou D, et al.
Meperidine carries an increased risk of neo- Adult congenital heart disease: a paradigm of epide-
natal respiratory sedation if administered miological change. Int J Cardiol. 2016;218:269–274.
to breastfeeding mothers, and there have 5. Dombrowski MP, Schatz M. Bulletins-obstetrics
been cases of neonatal cyanosis, bradycardia, ACoP: ACOG practice bulletin: clinical manage-
and apnea after its administration.56 Codeine ment guidelines for obstetrician-gynecologists
number 90, February 2008: asthma in pregnancy.
is best avoided in breastfeeding mothers, Obstet Gynecol. 2008;111:457–464.
as there is significant transfer to the breast 6. Dominguez JE, Street L, Louis J. Management of
milk and infants with specific genetic poly- obstructive sleep apnea in pregnancy. Obstet
morphisms are at increased risk of significant Gynecol Clin North Am. 2018;45:233–247.
respiratory depression.57 7. Gaiser R. Anesthetic considerations in the obese
parturient. Clin Obstet Gynecol. 2016;59:193–203.
Information with regard to specific med- 8. Duggan EW, Carlson K, Umpierrez GE. Perio-
ication safety during breastfeeding and safe perative hyperglycemia management: an update.
alternatives is available for patients and Anesthesiology. 2017;126:547–560.
health care providers at the National Institute 9. Thompson BM, Stearns JD, Apsey HA, et al.
of Health LactMed database (found at: Perioperative management of patients with dia-
betes and hyperglycemia undergoing elective sur-
[Link] gery. Curr Diab Rep. 2016;16:2.
htm. Accessed August 24, 2018). 10. Silver RM, Landon MB, Rouse DJ, et al. Maternal
morbidity associated with multiple repeat cesarean
deliveries. Obstet Gynecol. 2006;107:1226–1232.
Summary 11. Bailit JL, Grobman WA, Rice MM, et al. Mor-
Scheduled and urgent management of PAS bidly adherent placenta treatments and outcomes.
represent challenging surgical cases that Obstet Gynecol. 2015;125:683–689.
require significant multidisciplinary plan- 12. American Society of Anesthesiologists Commit-
tee on Standards and Practice Parameters. Prac-
ning and frequent close communication. tice guidelines for obstetric anesthesia: an updated
Optimal management requires perioperative report by the American Society of Anesthesiolo-
care providers to have a detailed under- gists Task Force on obstetric anesthesia and the
standing of the pathophysiology of abnor- Society for Obstetric Anesthesia and Perinatol-
mal placentation, the physiological changes ogy. Anesthesiology. 2016;124:270–300.
13. Tierney S, Perlas A. Informed consent for
of pregnancy, invasive monitoring, trans- regional anesthesia. Curr Opin Anaesthesiol. 2018;
fusion medicine, critical care, and effective 31:614–621.
postoperative pain control. Although mater- 14. O’Leary CE. Informed consent for anesthesia: has
nal hemorrhage remains a significant con- the time come for a separate written consent
tributor to maternal mortality, appropriate document? ASA Newsletter. 2006;70:11–12.
15. Gyamfi C, Berkowitz RL. Responses by pregnant
peripartum management of these challenging Jehovah’s Witnesses on health care proxies. Ob-
cases can help to reduce maternal morbidity stet Gynecol. 2004;104:541–544.
and optimize fetal outcomes. 16. Dwyer R, Fee JP, Moore J. Uptake of
halothane and isoflurane by mother and baby
during caesarean section. Br J Anaesth. 1995;74:
379–383.
References 17. Afolabi BB, Lesi FE. Regional versus general
1. Creanga AA, Bateman BT, Butwick AJ, et al. anaesthesia for caesarean section. Cochrane Data-
Morbidity associated with cesarean delivery in the base Syst Rev. 2012;10:CD004350.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
826 Warrick and Rollins

18. De Tina A, Palanisamy A. General anesthesia 32. Mhyre JM, Shilkrut A, Kuklina EV, et al. Mas-
during the third trimester: any link to neuro- sive blood transfusion during hospitalization for
cognitive outcomes? Anesthesiol Clin. 2017;35: delivery in New York State, 1998-2007. Obstet
69–80. Gynecol. 2013;122:1288–1294.
19. American Society of Anesthesiologists. ASA Re- 33. Barber RL, Fletcher SN. A review of echocardiog-
sponse to the FDA Med Watch Warning. 2016. raphy in anaesthetic and peri-operative practice. Part
Available at: [Link]/advocacy/fda-and- 1: impact and utility. Anaesthesia. 2014;69:764–776.
washington-alerts/washington-alerts/2016/12/asa- 34. Committee on practice bulletins—gynecology.
response-to-the-fda-med-watch?month=12&cate ACOG practice bulletin No. 195: prevention of
gory=Washington%20Alert. Accessed September infection after gynecologic procedures. Obstet
25, 2018. Gynecol. 2018;131:e172–e189.
20. Sprung J, Flick RP, Wilder RT, et al. Anesthesia 35. Aziz MF, Kim D, Mako J, et al. A retrospective
for cesarean delivery and learning disabilities in a study of the performance of video laryngoscopy in
population-based birth cohort. Anesthesiology. an obstetric unit. Anesth Analg. 2012;115:904–906.
2009;111:302–310. 36. D’Angelo R, Smiley RM, Riley ET, et al. Serious
21. Chinn GA, Sasaki Russell JM, Sall JW. Is a short complications related to obstetric anesthesia: the
anesthetic exposure in children safe? Time will serious complication repository project of the
tell: a focused commentary of the GAS and Society for Obstetric Anesthesia and Perinatol-
PANDA trials. Ann Transl Med. 2016;4:408. ogy. Anesthesiology. 2014;120:1505–1512.
22. Committee on obstetric practice. Committee 37. Hawkins JL, Chang J, Palmer SK, et al. Anesthe-
opinion no. 529: placenta accreta. Obstet Gynecol. sia-related maternal mortality in the United States:
2012;120:207–211. 1979-2002. Obstet Gynecol. 2011;117:69–74.
23. Safety ACoP, Quality I: ACOG committee opinion 38. Creanga AA, Syverson C, Seed K, et al. Preg-
No. 526: standardization of practice to improve nancy-related mortality in the United States,
outcomes. Obstet Gynecol. 2012;119:1081–1082. 2011-2013. Obstet Gynecol. 2017;130:366–373.
24. Silver RM, Fox KA, Barton JR, et al. Center of 39. Ngan Kee WD, Khaw KS, Tan PE, et al. Placental
excellence for placenta accreta. Am J Obstet transfer and fetal metabolic effects of phenylephrine
Gynecol. 2015;212:561–568. and ephedrine during spinal anesthesia for cesarean
25. Eller AG, Bennett MA, Sharshiner M, et al. delivery. Anesthesiology. 2009;111:506–512.
Maternal morbidity in cases of placenta accreta 40. Ngan Kee WD. The use of vasopressors during
managed by a multidisciplinary care team com- spinal anaesthesia for caesarean section. Curr
pared with standard obstetric care. Obstet Gyne- Opin Anaesthesiol. 2017;30:319–325.
col. 2011;117:331–337. 41. Vallejo MC, Attaallah AF, Elzamzamy OM,
26. Kodali BS. Bloodless trilogy? Anesthesia, obstet- et al. An open-label randomized controlled clin-
rics and interventional radiology for cesarean ical trial for comparison of continuous phenyl-
delivery. Int J Obstet Anesth. 2010;19:131–132. ephrine versus norepinephrine infusion in
27. Panigrahi AK, Yeaton-Massey A, Bakhtary S, prevention of spinal hypotension during cesarean
et al. A standardized approach for transfusion delivery. Int J Obstet Anesth. 2017;29:18–25.
medicine support in patients with morbidly ad- 42. Pacheco LD, Saade GR, Hankins G. Severe sepsis
herent placenta. Anesth Analg. 2017;125:603–608. during pregnancy. Clin Obstet Gynecol. 2014;57:
28. D’Alton ME, Friedman AM, Smiley RM, et al. 827–834.
National Partnership for Maternal Safety: con- 43. O’Brien KL, Shainker SA, Lockhart EL. Trans-
sensus bundle on venous thromboembolism. fusion management of obstetric hemorrhage.
Anesth Analg. 2016;123:942–949. Transfus Med Rev. 2018;32:249–255.
29. Leffert L, Butwick A, Carvalho B, et al. The 44. Collaborators WT. Effect of early tranexamic
Society for Obstetric Anesthesia and Perinatology acid administration on mortality, hysterectomy,
Consensus Statement on the anesthetic manage- and other morbidities in women with post-partum
ment of pregnant and postpartum women receiv- haemorrhage (WOMAN): an international, rand-
ing thromboprophylaxis or higher dose omised, double-blind, placebo-controlled trial.
anticoagulants. Anesth Analg. 2018;126:928–944. Lancet. 2017;389:2105–2116.
30. Rollins M, Lucero J. Overview of anesthetic 45. Collins PW, Cannings-John R, Bruynseels D,
considerations for cesarean delivery. Br Med Bull. et al. Viscoelastometric-guided early fibrinogen
2012;101:105–125. concentrate replacement during postpartum hae-
31. Walker MG, Pollard L, Talati C, et al. Obstetric morrhage: OBS2, a double-blind randomized
and Anaesthesia Checklists for the management controlled trial. Br J Anaesth. 2017;119:411–421.
of morbidly adherent placenta. J Obstet Gynaecol 46. Grainger H, Catling S. Intraoperative cell salvage
Can. 2016;38:1015e1023. in obstetrics. J Perioper Pract. 2011;21:264–270.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Anesthesia Considerations for Placenta Accreta 827

47. Goucher H, Wong CA, Patel SK, et al. Cell salvage 53. Ventham NT, Hughes M, O’Neill S, et al. Sys-
in obstetrics. Anesth Analg. 2015;121:465–468. tematic review and meta-analysis of continuous
48. Esakoff TF, Sparks TN, Kaimal AJ, et al. Diag- local anaesthetic wound infiltration versus
nosis and morbidity of placenta accreta. Ultra- epidural analgesia for postoperative pain follow-
sound Obstet Gynecol. 2011;37:324–327. ing abdominal surgery. Br J Surg. 2013;100:
49. Sutton CD, Carvalho B. Optimal pain manage- 1280–1289.
ment after cesarean delivery. Anesthesiol Clin. 54. Bollag L, Richebe P, Ortner C, et al. Transversus
2017;35:107–124. abdominis plane catheters for postcesarean deliv-
50. Singh SI, Rehou S, Marmai KL, et al. The efficacy ery analgesia: a series of five cases. Int J Obstet
of 2 doses of epidural morphine for postcesarean Anesth. 2012;21:176–180.
delivery analgesia: a randomized noninferiority 55. Wanderer JP, Rathmell JP. Anesthesia & breast-
trial. Anesth Analg. 2013;117:677–685. feeding: more often than not, they are compatible.
51. Carvalho B, Butwick AJ. Postcesarean delivery Anesthesiology. 2017;127:A15.
analgesia. Best Pract Res Clin Anaesthesiol. 2017;31: 56. Cobb B, Liu R, Valentine E, et al. Breastfeeding
69–79. after anesthesia: a review for anesthesia providers
52. Valentine AR, Carvalho B, Lazo TA, et al. regarding the transfer of medications into breast
Scheduled acetaminophen with as-needed opioids milk. Transl Perioper Pain Med. 2015;1:1–7.
compared to as-needed acetaminophen plus 57. Dalal PG, Bosak J, Berlin C. Safety of the breast-
opioids for post-cesarean pain management. Int feeding infant after maternal anesthesia. Paediatr
J Obstet Anesth. 2015;24:210–216. Anaesth. 2014;24:359–371.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA CLINICAL OBSTETRICS AND GYNECOLOGY
Volume 61, Number 4, 828–840
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.

Blood Products in the

Management of
Downloaded from [Link] by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3dW0s7N5CLiYMlH4wgaY34Fr69rtLH3oWi4+h6VxXUqo= on 10/30/2018

Abnormal
Placentation
LUKE A. GATTA, MD,* EVELYN L. LOCKHART, MD,†
and ANDRA H. JAMES, MD, MPH*
*Department of Obstetrics and Gynecology, Duke University,
Durham, North Carolina; and †Department of Pathology,
University of New Mexico, Albuquerque, New Mexico

Abstract: A critical tool in the successful management increta, and percreta) and placenta previa is
of patients with abnormal placentation is an estab- blood loss. With the uterine arteries at term
lished massive transfusion protocol designed to rap-
idly deliver blood products in obstetrical and surgical receiving 12% of cardiac output,1 obstetrical
hemorrhage. Spurred by trauma research and an hemorrhage can quickly deteriorate into
understanding of consumptive coagulopathy, the past disseminated intravascular coagulation, mul-
2 decades have seen a shift in volume resuscitation tiorgan failure, and is the critical and root
from an empiric, crystalloid-based method to bal- cause of maternal mortality.2 There is a
anced, targeted transfusion therapy. The present
article reviews patient blood management in abnor- multifactorial cause to the bleeding: abnor-
mal placentation, beginning with optimizing the mal placentation increases the potential for
patient’s status in the antenatal period to the labo- obstetric bleeding (originating from the
ratory assessment and transfusion strategy for blood disruption of spiral arteries within the post-
products at the time of hemorrhage. partum uterus), the anatomic complexity
Key words: placenta accreta, placenta previa, trans-
fusion, blood products, postpartum hemorrhage, increases the risk for surgical bleeding (due
patient blood management to incisions, ruptured vessels, or lacerations),
and taken together, both increase the risk for
coagulopathic bleeding (due to the consump-
Background tion of platelets and clotting factors).3 The
patient with PAS can routinely experience
The immediate and most important ma-
a 3 to 5 L blood loss,4 accompanied by
ternal complication of placenta accreta
an impressive need for transfusion. One
spectrum (PAS, typically placenta accreta,
retrospective study of 66 patients with known
Correspondence: Andra H. James, MD, MPH, DUMC placenta accreta found that 95% were trans-
3967, Durham, NC. E-mail: [Link]@[Link] fused, with 39% requiring > 10 units of
The authors declare that they have nothing to disclose. packed RBC units.5 Another study of women

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 61 / NUMBER 4 / DECEMBER 2018

828 | [Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Blood Products 829

with placenta previa noted a 10-fold risk of available for those who need them.11 PBM
massive hemorrhage.6 The potential for includes measures to avoid or minimize
transfusion warrants multidisciplinary man- transfusion such as anemia management,
agement for cases of known or suspected cell salvage, and the maintenance of hemo-
abnormal placentation, and optimal care stasis.
begins before delivery. Improved outcomes
have been demonstrated in tertiary care ANEMIA MANAGEMENT
centers with a multidisciplinary team with With or without abnormal placentation,
experience in abnormal placentation.7,8 iron deficiency anemia is common in
Among the most important criteria for these pregnancy, and is associated with adverse
centers is a comprehensive blood bank, with neonatal and pregnancy outcomes.12 Par-
an established massive transfusion protocol ticularly in abnormal placentation, where
to efficiently deliver blood products.9 In blood loss is to be expected, delivery plan-
addition to the restoration of blood volume, ning should ensure an ideal preoperative
transfusion medicine optimizes the medical hemoglobin (Hb) level. Managing anemia
management of massive hemorrhage and preoperatively has been shown to reduce
correction of coagulopathy. transfusion requirements, and treatment of
Data on the transfusion of blood prod- anemia is one of the pillars of PBM.13
ucts specifically in abnormal placentation Effective management requires advanced
is limited. Thus, recommendations are screening for anemia, which is routinely
largely driven by studies on postpartum included in prenatal care.14,15 In pregnancy,
hemorrhage (PPH) and traumatic bleed- anemia has been defined as an Hb < 11.0 g/
ing. Although an international consensus dL (33% hematocrit) in the first trimester,
panel recommended caution applying ob- <10.5 g/dL (32% hematocrit) in the second
servational data from trauma research to trimester, and <11.0 g/dL (33%) in the third
obstetrical population, advances in trans- trimester.16 Women with iron deficiency
fusion management continue to be guided anemia should receive 100 to 200 mg oral
by the latter.10 Empiric and allogenic elemental iron daily, which usually requires
blood transfusion is no longer viewed supplementation above a standard prenatal
as the default therapy in blood loss, and vitamin. Cost-effectiveness studies support
has become one among multiple treat- the development of preoperative anemia
ments that should be used. This article clinics for high-risk patients, such as those
reviews blood products and the trans- undergoing major orthopedic surgery,17 and
fusion management strategies of massive academic centers have developed anemia
hemorrhage as it is applied to patients clinics to consult and manage patients with
with abnormal placentation, and dis- anemia in anticipation of blood loss.13,18 A
cusses guideline recommendations from mainstay of the anemia clinic includes the
the antenatal to the postoperative period. outpatient administration of IV iron for
patients who are unable to tolerate oral
iron, unable to absorb it appropriately, or
have greater than expected needs. There are
Antenatal Patient Blood strong data to support the administration of
Management (PBM) iron in obstetric patients, with several
Delivery planning includes PBM, an evi- randomized controlled trials of oral verses
dence-based, multidisciplinary approach IV iron showing benefit.19–25 The ACOG
identifying and optimizing patients at high Practice Bulletin on Anemia in Pregnancy
risk for transfusion. The goal of PBM is recommends IV iron in patients unable to
to reduce the need for allogenic blood, tolerate oral iron or with severe deficiency,
while ensuring that blood components are although a threshold Hb level for the use of

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
830 Gatta et al

IV iron has not been established.14 Both many centers use a massive transfusion
oral and IV iron are effective in raising Hb protocol to provide blood components dur-
levels, although a 2 g/dL increase is more ing ongoing bleeding, these protocols are
likely to be achieved with IV iron. IV iron not designed to address blood preparation
may be started as early as in the second before hemorrhage. Although antenatal ul-
trimester, with a repeat Hb/hematocrit test- trasound provides early diagnosis of abnor-
ing completed in 4 weeks to assess response. mal placentation and aids delivery planning,
There are minimal risks associated with it does not provide guidance in predicting
oral iron supplementation, with nausea and blood loss or transfusion requirements.5
constipation the most common adverse Interestingly, in the retrospective study ref-
effects.26 The most serious risk of IV iron erenced, spanning 14 years and including
therapy is anaphylaxis, and the risk profile 66 patients with histopathology-confirmed
differs by formulation. A recent retrospective placenta accreta, blood transfusion require-
study of Medicare patients who received iron ments did not differ between the pathology
infusion found that the risk of anaphylaxis subtypes (percreta, increta, or accreta).
was 68 per 100,000 for iron dextran versus 24 Imaging did, however, have bearing on the
per 100,000 for nondextran products (iron transfusion demands in women with pla-
sucrose, gluconate, and ferumoxytol). centa previa and not PAS. In a retrospective
cohort study of 210 women undergoing a
PREOPERATIVE AUTOLOGOUS primary cesarean delivery for placenta pre-
DONATION via (defined within 2 cm of the cervical os),
Preoperative autologous donation, collection, women with an anterior placenta had an
and storage of a patient’s blood before adjusted odds ratio of 3.13 [95% confidence
scheduled surgery is no longer recommended interval (CI), 1.18-8.36] for blood trans-
in obstetrics. Autologous donation was fusion compared with women with a poste-
touted as a way to eliminate the risk of rior placenta, although the number of units
transmitting infectious disease, particularly of blood was not reported.31 Delivery plan-
HIV and HCV, through the reduction of ning for patients with abnormal placenta-
allogeneic blood transfusion.27 The risk of tion must include advanced communication
infectious transmission has steeply declined, with the blood bank, particularly for pa-
currently 1/1.47 million units for HIV and tients with rare blood types or antibodies to
1/1.15 million for HCV.28 Noninfectious specific blood group antigens.
transfusion risks are not mitigated by autol-
ogous donation, including transfusion-
associated lung injury (TRALI) and trans- PBM, Intraoperative
fusion-associated circulatory overload (TA- The goal of the management of massive
CO).29 Following donation, a patient’s Hb hemorrhage due to abnormal placentation
may not return to baseline before delivery, is the control of obstetrical, surgical, and
leading to a decrease in maternal iron stores coagulopathic bleeding in order to maintain
and exacerbating anemia.28 Therefore, pre- perfusion and prevent ischemia. Early co-
operative autologous donation is uncom- agulation testing, serial laboratory monitor-
mon and no longer recommended in ing and, in some instances, point of care
obstetrics except in rare circumstances.30 technologies can be used to assess maternal
coagulation profile and to guide ongoing
BLOOD PREPARATION correction of coagulopathy.
Predicting the need for blood products is
case specific, and no consensus exists for LABORATORY MONITORING
predelivery transfusion planning in patients In patients with known abnormal placen-
with abnormal placentation. Although tation, a recent complete blood count and

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Blood Products 831

an updated type and screen should be data can elucidate an optimum fibrinogen
maintained per individual blood bank level for intervention.
requirements. At the time of hemorrhage,
baseline coagulation laboratory studies THE ADVENT OF VISCOELASTIC
should be obtained, and include Hb/ COAGULATION TESTING
hematocrit, platelets, electrolytes, pro- Research from massive bleeding in trauma
thrombin time (PT), partial thromboplas- suggests that the hemorrhage itself contrib-
tin time (PTT), fibrinogen, and, if utes to coagulopathy.40 In this paradigm,
clinically indicated, an arterial shock bleeding is not only a consequence of coagul-
panel.32 Reassessment should be repeated opathy, but a main contributor. Thus, coag-
frequently, every 45 to 60 minutes, until ulation parameters are expected to change
the hemorrhage is controlled, as it will over the course of hemorrhage, rendering it
assist in medical decision-making and necessary to have accurate and timely data.
targeted therapy.10,33 Traditional laboratory testing is associated
Hypofibrinogenemia is an important with delayed turnaround times,33 ranging
predictor of severe PPH. Charbit et al34 from 30 to 60 minutes after the blood sample
assessed coagulation laboratory values at is received by the laboratory. Recognizing
the time of second-line uterotonic admin- that current data are necessary as coagulop-
istration in 128 patients experiencing PPH. athy evolves, clinicians at the University of
The only laboratory value that predicted Washington developed an emergency hem-
progression to severe PPH, as defined by orrhage panel (EHP) with the goal of a
(1) transfusion of ≥ 4 packed RBC units, turnaround time of <20 minutes. Through
(2) Hb decrease of <4 g/d, (3) procedural expedited transport, prioritizing critical sam-
intervention, or (4) death, was a fibrinogen ples, and shortened centrifuge time, the EHP
level <200 mg/dL. With viscoelastometric turnaround time was 14 ± 3 minutes,33 com-
testing, this can translate to a Fibtem < 15 parable with point of care viscoelastic testing.
mm.35 In light of these data, rapid identi- Point of care viscoelastic coagulation testing
fication of hypofibrinogenemia is recom- (PCVT) such as thromboelastography or
mended for patients with PPH. Although rotational thromboelastometry, have gained
hypofibrinogenemia is known to be indi- popularity for their utility in assessing con-
cative of severe PPH, the role of targeted sumptive coagulopathy in ongoing blood loss
replacement remains unclear. Guidelines and in guiding transfusion efforts.41 These
vary, although most recommend mainte- tests use a small volume of whole blood to
nance above 200 mg/dL.36–38 In a recent assess hemostatic function from clot forma-
multicenter, blinded, randomized con- tion to propagation and, finally, to clot lysis.
trolled trial, fibrinogen concentrate versus Various assays for rotational thromboelas-
placebo was administered to 55 women tometry analysis can rapidly isolate defects in
with ongoing bleeding and Fibtem < 15 hemostatic and thrombolytic pathways, with
mm. The adjusted incidence rate ratio for a turnaround time <20 minutes in most
the number of allogenic units transfused in cases.42 With demonstrated utility in the
the fibrinogen group, compared with the trauma literature43,44 it has been suggested
placebo group, was 0.72 (95% CI, 0.3-1.7; that PCVT guide targeted blood product
P = 0.45), leading the authors to conclude resuscitation in PPH. In one tertiary care
that replacing fibrinogen did not improve hospital, a retrospective cohort study was
outcomes in PPH.39 It is important to note conducted to assess the clinical outcomes
that the aforementioned study excluded (blood transfused, rate of intensive care unit
patients with abnormal placentation, and admission) in patients managed with PCVT
it would seem prudent to replace fibrino- versus traditional empiric resuscitation.42
gen to maintain > 200 mg/dL until further Among 86 women with severe PPH, 28 were

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
832 Gatta et al

managed by PCVT, and 58 were in the non- reporting. The benefit of a massive trans-
PCVT group. There was no statistically fusion protocol is that it can be used at any
significant difference between PCVT and time, such as the situation where abnormal
non-PCVT with regard to the volume of placentation is diagnosed at the time of
crystalloid, colloid, albumin, and cryopreci- delivery. Massive hemorrhage protocols
pitate administered. However, the patients should have a clear criteria for activation,
in the PCVT group received significantly followed by a subsequent designation of
fewer transfusions of RBCs (< 0.0001), fresh professional roles to acquire appropriate
frozen plasma (FFP < 0.0001), and platelets blood and respond to transfusion needs.
(< 0.0001), and the estimated blood loss was Activation should begin with the early
significantly lower in the PCVT group [me- recognition of excessive blood loss, tradi-
dian, 2000 mL (1600 to 2500) vs. 3000 mL tionally considered to be > 500 mL for
(2000 to 4000); P = 0.001]. The rate of vaginal delivery or > 1000 mL for cesarean.
puerperal hysterectomy was also significantly Historically, this is appreciated by the
lower in the PCVT group (25%, 7/28 vs. delivering clinician, and is limited by a
53.5%, 31/58; P = 0.013) as were postopera- near-universal propensity to underestimate
tive intensive care unit admissions (3.5% vs. blood loss.45,46 In light of this, in 2014 a
43.1%). The results of the study suggest that multidisciplinary team under the reVITAL-
morbidity can be significantly reduced with ize Initiative, sought to standardize
timely assessment of the patient’s hemostatic language used, describing PPH as “cumu-
state. It is of the opinion of the authors that lative blood loss of ≥ 1000 ml or blood loss
traditional, serial laboratory monitoring re- accompanied by sign or symptoms of
mains the mainstay due to widespread avail- hypovolemia within 24 hours following
ability and familiarity in practice, but PCVT the birth process, including intrapartum
is an important adjunct to understanding loss.”47 This new definition deemphasizes
coagulation. Although institutional and clini- subjective estimation of volume and appre-
cian preferences will vary with respect to ciates aberrations in maternal vital signs
laboratory testing, there appears to be a and clinical assessment of hemodynamic
potential role for PCVT in timely assessment status. After activation of a massive trans-
of hemostasis. Regardless of the testing used, fusion protocol, a well-trained response
both PCVT or traditional (and expedited) team should have access to a hemorrhage
coagulation testing are valuable in the med- cart and emergency released blood
ical management of actively bleeding pa- products.48 The hemorrhage cart includes
tients. a tray of instruments as well as uterotonics
(available refrigerated), and a checklist is
MASSIVE TRANSFUSION PROTOCOL available through the California Maternal
The ACOG Committee Opinion on Quality Care Collaborative.49 Emergency
Placenta Accreta recommends that institu- released blood products should be type
tionally established massive transfusion specific for the patient, or universally com-
protocols be followed.4 Although no single patible (group O, Rhesus negative RBCs,
algorithm for massive transfusion protocol or AB plasma) if blood type is not available
is appropriate for each facility, consensus or pending. An institution-specific massive
guidelines have been collected by the Na- transfusion protocol provides some defense
tional Partnership for Maternal Safety.32 against the development of coagulopathy.
The latter, which includes representatives
from a number of organizations, developed BLOOD ADMINISTRATION
a consensus bundle of guidelines for PPH As previously mentioned, the recommenda-
organized into 4 components: readiness, tions with regard to PPH and transfusion
recognition and prevention, response, and have been influenced by trauma studies.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Blood Products 833

Historically, the goal of hypovolemic shock to coagulopathy,56 as normal coagulation is

resuscitation efforts have been to support suppressed with decreasing temperature. Co-
perfusion and urine output through the agulation reactions are slowed by ~5% with
maintenance of blood pressure, in order to each degree Celsius drop in temperature,57
prevent or reverse the metabolic derange- and in a multicenter prospective cohort study
ments associated with ischemia from acute of adults with blunt injury with hemorrhagic
blood loss. In the early 2000s, it was shock, there was a dose-response relationship
recognized that shock management should with a temperature <34°C resulting in the
prevent the development of intravascular greatest mortality.58 In the aforementioned
coagulopathy, and there was a trend away study, hypothermia was found to be an
from crystalloid-led resuscitation.50,51 Em- independent risk factor for mortality in
piric administration of crystalloid is no patients requiring massive transfusion.
longer the mainstay of therapy. Observatio-
nal studies of blood replacement in trauma CELL SALVAGE
suggested a 1:1 ratio of RBCs to plasma was Cell salvage is a strategy used to decrease
associated with improved hemostasis and the need for allogenic blood transfusion.
survival.52,53 In a randomized controlled Historically, cell salvage has been avoided
trial assessing major bleeding in trauma in obstetric patients because of the theo-
patients, the early administration of plasma, retical risk of amniotic fluid embolism or
platelets, and RBCs in a [Link] compared alloimmunization secondary to the expo-
with [Link] found that more patients in the sure to fetal RBC antigens. However, with
treatment group achieved hemostasis, and advances in cell salvage technology, risks in
fewer died of exsanguination within the first obstetric patients appear to be comparable
24 hours, although it is important to note with the general population.59 In obstetric
that overall mortality was not statistically patients, there are 2 modifications: (1) use of a
significant between the 2 groups at 24 hours separate suction source to waste blood and
and 30 days.53 The optimal ratio in an amniotic fluid collected before placental de-
obstetric population has not yet been deter- livery; (2) addition of a leukocyte depletion
mined, but a massive transfusion protocol filter to the circuit to reduce levels of con-
should rapidly dispense blood in a defined taminants before transfusion of cell-salvaged
ratio designed to avoid dilutional coagulop- blood, including squamous fetal cells and
athy that may result from overtransfusion amniotic fluid–derived tissue factor (both of
of one product.54 Obstetric massive trans- which are implicated in amniotic fluid embo-
fusion protocols that have incorporated lism). In a study comparing maternal venous
higher RBC to plasma ratios between 1:1 blood from postwash, postfiltration cell-
and 2:1 have been successful in controlling salvaged blood, there was no difference in
hemorrhage and reducing blood product the level of contaminants.60 Of note, the cell
utilization.38,55 The National Partnership salvage technology is unable to distinguish
for Maternal Safety consensus bundle for maternal from fetal erythrocytes, and there-
obstetric hemorrhage supports an RBC to fore fetal antigens remain. For Rh negative
plasma ratio between 1:1 and 2:1, as well mothers, Rh immune globulin is required
as the administration of a dose of apheresis postoperatively after a Keihauer-Betke test is
platelets (or a 6 to 8 “pack” of pooled used to quantify maternal exposure to fetal
platelets) for approximately every 6 to 8 RBCs. In a study of 20-patients undergoing
units of RBCs.32 cesarean hysterectomy, 15 patients received
It is imperative to note that optimizing autologous blood after cell salvage, with a
resuscitation requires attention to the pa- mean volume of 1476 ± 247 mL. Of the 15
tient’s thermal and acid-base status. Hypo- patients receiving autologous blood, 13 did
thermia and acidosis are known contributors not require allogenic transfusion.61

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
834 Gatta et al

TABLE 1. Blood Products Used in Postpartum Hemorrhage

Target Expected
Components Preparation Volume Laboratory Benefit Other
Red Blood Erythrocytes Isolated from 55%-65% Hemoglobin > 8 g/ Raise Must establish
Cells plasma by hematocrit dL, between 7.0 hemoglobin serologic
centrifuga- in 300-400 and 10 g/dL by 1 g/dL compatibility or
tion, mL during active (hematocrit use universal O−
sedimenta- bleeding by 3%)
tion, or by
apheresis; 42 d
shelf life
Plasma Albumin, Separation of 200-250 mL; Target PTT and PT Both FFP and The National
coagulation whole blood unless <1.5 × normal in PF24 have Partnership for
factors, or by apheresis- active bleeding roughly equal Maternal Safety
fibrinolytic apheresis; derived: hemostatic consensus
proteins, once 400-600 potency in bundle
immunoglobu- collected, mL; labile hemorrhage, recommends a
lins, frozen −18°C coagulation although RBC to plasma
anticoagulant within 8 h and factors vary levels of FV, ratio between 1:1
proteins stored up to a by FVIII, and and 2:1
year until processing protein C are
thawed before and storage slightly
use (FFP). In conditions reduced in
contrast, PF24
plasma frozen
within 24 h
after
phlebotomy
(PF24) is
frozen within
24 h of
collection
Cryoprecipitate Fibrinogen, factor Thawing FFP, > 80 IU factor Principle use is in Wide variation in
VIII, factor and VIII, > 150 hypofibrinogene- response: one
XIII, VWF recovering mg mia guidelines unit can
insoluble fibrinogen vary, generally increase
precipitate; in ~5-20 mL recommend fibrinogen by
transfused of plasma fibrinogen > 125- up to 10 mg/
individually 200 mg/dL dL
or pooled
Platelets Platelets, Prepared by 5.5×1010 > 50,000/µL Increase 5000- While platelets do not
suspended in apheresis or platelets if platelets in active 10,000/µL bear RhD
plasma from whole derived from bleeding platelets if antigen, trace
blood; whole blood; derived from RBCs in this
suspended in 3.0×1011 whole blood; product
plasma with platelets if 30,000- necessitate Rh
coagulation derived from 50,000/µL if negative platelets
factors apheresis derived from (or RhIg may be
apheresis given to prevent
alloimmunization)

FFP indicates fresh frozen plasma; PT, prothrombin time; PTT, partial thromboplastin time.

Blood Products Association of Blood Banks32,36,62–64

Depending on the collection system used, (Table 1).
a single donation of blood contains
450 to 500 mL of whole blood stored in ADVERSE REACTIONS
an anticoagulant-preservative solution To prevent adverse reactions, there are 3
and has a minimum hematocrit of 38%. main requirements before transfusing
These donations are processed in an aseptic RBCs in premenopausal female recipients:
manner and separated into various compo- first, the ABO group must be matched,
nents for use, guided by the American second, the Rh(D) group must be matched,

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Blood Products 835

and finally, the recipient sample must be NEW DEVELOPMENTS IN BLOOD

screened for antibodies that may cross react COMPONENT PATHOGEN
with antigens from the donor sample. REDUCTION
The most commonly reported transfu- Two systems for blood component patho-
sion reaction is a febrile, nonhemolytic gen reduction have been approved for use
transfusion reaction (FNHTR), caused by within the US. Although the US blood
activation of the cytokine cascade within supply is safer than ever before, some
the recipient, with an incidence of 0.3% to bacteria, viruses, prions, and parasites
6% in the transfusion of RBCs.65 While can still be transmitted. The Intercept
benign, the diagnostic criteria for FNHTR System (Cerus, Concord, CA) uses a
(an increase in 1°C, accompanied by chills, psoralen which binds pathogen nucleic
rigors, and malaise) suggest a more threat- acid and is activated by UVA light,
ening reaction. A platelet transfusion car- preventing pathogen replication. Cur-
ries a higher risk of FNHTR due to the rently the system is approved in the US
higher concentration of leukocyte antigens for both plasma and platelets, with on-
in the preparation.66 going development for use in RBCs.
In contrast, the hemolytic transfusion Octaplas (Octapharma, Hoboken, NJ) is
reaction is a potentially fatal complication plasma product that is pathogen reduced.
due to the immune-mediated destruction of Plasma undergoes filtration and solvent-de-
donor products by preexisting recipient anti- tergent reagent treatment to inactivate lipid-
bodies. Improvements in blood procurement enveloped viruses, and affinity column filtra-
and banking strategies have reduced the tion to reduce prion proteins.71
incidence, although hemolytic transfusion
reactions may still occur in compatible blood
due to non-ABO antibodies. Hemostatic Agents
Two severe and potentially lethal reac- There are almost no specific data on the
tions to transfusion, particularly massive portion of transfused accreta or previa
transfusion, include TACO and TRALI67 patients who have received hemostatic
which have similar presentations. TACO agents such as tranexamic acid (TXA),
results from the transfusion of blood prod- fibrinogen concentrate, or recombinant
ucts with subsequent volume overload, re- factor VIIa (rVIIa) but guidance is based
sulting in cardiogenic pulmonary edema on information from studies on PPH.
leading to respiratory distress. The diagnosis
is made in a patient receiving transfusion ANTIFIBRINOLYTIC THERAPY
developing hypoxemia (SpO2 < 90%) with Although it has long been understood
new onset bilateral infiltrates on chest radio- that severe hemorrhage is associated with
graph, and clinically evident left atrial activation of the fibrinolytic pathway,72
hypertension.68 TRALI presents similarly before this decade, antifibrinolytic ther-
to TACO, with hypoxemia and dyspnea, apy has been infrequently used in the US.
and is due to increased vascular permeability TXA, a lysine analog that prevents plas-
secondary to donor leukocyte antibodies min-mediated fibrin degradation, has
leading to noncardiogenic pulmonary been shown to decrease bleeding compli-
edema.69 TRALI and TACO may be dis- cations in trauma and cardiac surgery.73
tinguished by the cardiogenic origin, with the In 2017, the results from the World
former having an elevated left atrial pressure. Maternal Antifibrinolytic (WOMAN) tri-
Treatment in massive hemorrhage requires al found that early use of TXA was
aggressive respiratory support, usually with associated with decreased bleeding-re-
invasive ventilation or extracorporeal mem- lated mortality.74 In this multinational,
brane oxygenation.70 double-blind and intention to treat trial,

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
836 Gatta et al

women experiencing PPH were random- combination of FVIIa with tissue factor
ized to receive either 1 g IV TXA or initiates the clotting cascade, thus a sig-
placebo. Death due to bleeding was sig- nificant concern exists for the develop-
nificantly reduced in the treatment group ment of thrombosis. The largest reported
(155, or 1.5%) compared with the placebo series of off-label use of rFVIIa in obstet-
group (191, or 1.9%); RR, 0.81; 95% CI, rics is from the Australian and New
0.65-1.00; P = 0.045. Thromboembolic Zealand Haemostasis Registry.81 The in-
events did not differ significantly in the vestigators recorded all off-label use of
TXA versus the placebo group. In their rFVIIa for treatment of acute PPH in
conclusions, the authors suggested that 105 cases; accreta accounted for 16% of
TXA should be given empirically soon the patients. The majority (78%) received
after the onset of excessive bleeding. a single dose (median dose = 92 mcg/kg)
Partly because of this trial, TXA has been with a positive response (defined subjec-
included in massive transfusion protocols. tively by the clinician) in 76%. In this
More than 30 studies evaluating the subanalysis of the registry, at least
efficacy of TXA have been reported, with 2 patients developed thromboembolism.
no significant increase in thromboembolic When the entire registry was assessed (not
complications.75–78 The recommended limited to obstetrical patients),82 rFVIIa
dose is 1 g IV within 3 hours of delivery, was used for critical bleeding in 3322
and a single dose should suffice in most patients. In this observational study, a
cases, with the half-life lasting 7 to 8 hours subjective cessation of bleeding was re-
although another dose may be given with- ported in 74%. Approximately 11% of the
in 30 minutes of the first dose if bleeding patients experienced a thromboembolic
does not slow.78 TXA has also been used event. Because of the recognized throm-
for the prevention of PPH at the time of boembolic risk, the European Society of
cesarean delivery. A meta-analysis of Anesthesia recommends rFVIIa be con-
randomized controlled trials showed a sidered only as a last resort for PPH.83
lower incidence of PPH or severe PPH Clinicians considering the use of rFVIIa
among women who received TXA com- in uncontrolled PPH should optimize the
pared with placebo.79 There is currently a patient’s hemostatic potential, with cor-
large (n = 11,000) NIH-funded, random- rection of hypofibrinogenemia, clotting
ized placebo-controlled trial to assess factor deficiency, and thrombocytopenia
whether TXA lowers the risk of PPH in before administration. Failure to opti-
women undergoing a cesarean delivery mize fibrinogen can result in a lack of
(NCT03364491). response to rFVIIa. In addition, the pa-
tient ideally should have a pH of ≥ 7.2, as
RECOMBINANT FACTOR VIIA rFVIIa activity significantly decreases in
In obstetrics, rFVIIa has been used in the acidosis.84
management of severe PPH unresponsive
to blood component therapy. rFVIIa was FIBRINOGEN CONCENTRATE
approved by the US Food and Drug As mentioned previously, the identifica-
Administration for the management of tion of fibrinogen <200 mg/dL as an un-
severe bleeding in patients with hemo- equivocal biomarker for progression to
philia and inhibitors to FVIII. Since then, severe PPH spurred research on the opti-
it has been used to help manage severe mal method to correct hypofibrinogene-
bleeding in a variety of clinical scenarios. mia. Particularly in Europe, there has been
In a survey of off-label use in rFVIIa, an effort to use fibrinogen concentrate
> 70% was used in cardiac surgery, 7% in rather than cryoprecipitate to replace
trauma, and <1% in obstetrics.80 The fibrinogen. Fibrinogen concentrate, like

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Blood Products 837

other plasma-derived factor products, is the important role for adjuvant pharmaco-
purified and carries an infinitesimally therapy in blood management for patients
small risk of viral transmission. Multiple with abnormal placentation.
case reports and a small series of its
use in the management of obstetrical
hemorrhage have been published, with References
doses ranging from 2 to 4 g and given in 1. Thaler I, Manor D, Itskovitz J, et al. Changes in
combination with plasma or cryopre- uterine blood flow during human pregnancy. Am
cipitate.85–87 A randomized controlled trial J Obstet Gynecol. 1990;162:121–125.
conducted at 4 hospitals in Denmark 2. Mehrabadi A, Hutcheon JA, Liu S, et al. Con-
tribution of placenta accreta to the incidence of
compared the preemptive administration postpartum hemorrhage and severe postpartum
of fibrinogen concentrate (2 g IV) versus hemorrhage. Obstet Gynecol. 2015;125:814–821.
saline in patients with normal fibrinogen 3. James AH, Lockhart EL. Blood management for
levels who had an estimated blood loss of patients with placenta accreta. In: SIlver RM, ed.
> 500 mL after vaginal delivery or > 1000 Boca Raton, FL: Taylor & Francis Group; 2017:153.
4. ACOG. Committee opinion no. 529: placenta
mL after cesarean delivery. There was no accreta. Obstet Gynecol. 2012;120:207–211.
difference in the rate of transfusion be- 5. Stotler B, Padmanabhan A, Devine P, et al. Trans-
tween the 2 groups, nor were there any fusion requirements in obstetric patients with
thromboembolic events in either group. To placenta accreta. Transfusion. 2011;51:2627–2633.
assess the efficacy and risks, there is an 6. Crane JM, Van den Hof MC, Dodds L, et al.
Maternal complications with placenta previa. Am
active multicenter, randomized, double- J Perinatol. 2000;17:101–105.
blind, placebo-controlled trial, the fibrino- 7. Wright JD, Herzog TJ, Shah M, et al. Region-
gen in hemorrhage of DELivery (FIDEL) alization of care for obstetric hemorrhage and its
trial, testing fibrinogen concentrate versus effect on maternal mortality. Obstet Gynecol.
placebo in women with PPH requiring 2010;115:1194–1200.
8. Eller AG, Bennett MA, Sharshiner M, et al.
second-line uterotonics.88 Maternal morbidity in cases of placenta accreta
managed by a multidisciplinary care team com-
pared with standard obstetric care. Obstet Gyne-
Summary col. 2011;117(pt 1):331–337.
Patients with abnormal placentation are at 9. Silver RM, Fox KA, Barton JR, et al. Center of
excellence for placenta accreta. Am J Obstet
risk for massive hemorrhage, requiring Gynecol. 2015;212:561–568.
large-volume transfusion. Delivery planning 10. Abdul-Kadir R, McLintock C, Ducloy AS, et al.
includes optimizing the patient’s hemato- Evaluation and management of postpartum hem-
logic status, and must utilize centers of orrhage: consensus from an international expert
excellence with well-equipped blood banks, panel. Transfusion. 2014;54:1756–1768.
11. Neb H, Zacharowski K, Meybohm P. Strategies to
trained transfusion services, with an estab- reduce blood product utilization in obstetric prac-
lished massive transfusion protocol to deliv- tice. Curr Opin Anaesthesiol. 2017;30:294–299.
er appropriate blood products. Guided by 12. Sun D, McLeod A, Gandhi S, et al. Anemia in
trauma research, the management of coa- pregnancy: a pragmatic approach. Obstet Gynecol
gulopathy has shifted away from empiric Surv. 2017;72:730–737.
13. Guinn NR, Guercio JR, Hopkins TJ, et al. How
resuscitation towards targeted therapy do we develop and implement a preoperative
based on expedited serial laboratory testing anemia clinic designed to improve perioperative
and PCVT analysis. The early identification outcomes and reduce cost? Transfusion. 2016;56:
of coagulopathy provides an opportunity to 297–303.
prevent ongoing bleeding and the need for 14. ACOG. ACOG Practice Bulletin No. 95: anemia
in pregnancy. Obstet Gynecol. 2008;112:201–207.
excessive allogenic transfusion. The intro- 15. McDonagh M, Cantor A, Bougatsos C, et al.
duction of TXA, now supported by a large Routine Iron Supplementation and Screening for
randomized controlled trial, has paved the Iron Deficiency Anemia in Pregnant Women:
way for future clinical trials which highlight A Systematic Review to Update the US Preventive

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
838 Gatta et al

Services Task Force Recommendation. Rockville 29. Covin RB, Ambruso DR, England KM, et al.
(MD): US Preventive Services Task Force Evi- Hypotension and acute pulmonary insufficiency
dence Syntheses, formerly Systematic Evidence following transfusion of autologous red blood
Reviews; 2015. cells during surgery: a case report and review of
16. Pavord S, Myers B, Robinson S, et al. UK the literature. Transfus Med. 2004;14:375–383.
guidelines on the management of iron deficiency 30. Liumbruno GM, Liumbruno C, Rafanelli D.
in pregnancy. Br J Haematol. 2012;156:588–600. Autologous blood in obstetrics: where are we
17. Rosencher N, Kerkkamp HE, Macheras G, et al. going now? Blood Transfus. 2012;10:125–147.
Orthopedic Surgery Transfusion Hemoglobin 31. Young BC, Nadel A, Kaimal A. Does previa
European Overview (OSTHEO) study: blood location matter? Surgical morbidity associated
management in elective knee and hip arthroplasty with location of a placenta previa. J Perinatol.
in Europe. Transfusion. 2003;43:459–469. 2014;34:264–267.
18. Kapadia BH, Banerjee S, Issa K, et al. Preoper- 32. Main EK, Goffman D, Scavone BM, et al. Na-
ative blood management strategies for total knee tional partnership for maternal safety: consensus
arthroplasty. J Knee Surg. 2013;26:373–377. bundle on obstetric hemorrhage. Obstet Gynecol.
19. al-Momen AK, al-Meshari A, al-Nuaim L, et al. 2015;126:155–162.
Intravenous iron sucrose complex in the treatment 33. Chandler WL, Ferrell C, Trimble S, et al. Devel-
of iron deficiency anemia during pregnancy. Eur J opment of a rapid emergency hemorrhage panel.
Obstet Gynecol Reprod Biol. 1996;69:121–124. Transfusion. 2010;50:2547–2552.
20. Bayoumeu F, Subiran-Buisset C, Baka NE, et al. 34. Charbit B, Mandelbrot L, Samain E, et al. The
Iron therapy in iron deficiency anemia in preg- decrease of fibrinogen is an early predictor of the
nancy: intravenous route versus oral route. Am J severity of postpartum hemorrhage. J Thromb
Obstet Gynecol. 2002;186:518–522. Haemost. 2007;5:266–273.
21. Al RA, Unlubilgin E, Kandemir O, et al. Intra- 35. Collins PW, Lilley G, Bruynseels D, et al. Fibrin-
venous versus oral iron for treatment of anemia in based clot formation as an early and rapid
pregnancy: a randomized trial. Obstet Gynecol. biomarker for progression of postpartum hemor-
2005;106:1335–1340. rhage: a prospective study. Blood. 2014;124:
22. Shafi D, Purandare SV, Sathe AV. Iron deficiency 1727–1736.
anemia in pregnancy: intravenous versus oral 36. Collins P, Abdul-Kadir R, Thachil. Subcommittees
route. J Obstet Gynaecol India. 2012;62:317–321. on Women’ s Health Issues in T, Haemostasis, on
23. Froessler B, Cocchiaro C, Saadat-Gilani K, et al. disseminated intravascular c. management of coa-
Intravenous iron sucrose versus oral iron ferrous gulopathy associated with postpartum hemorrhage:
sulfate for antenatal and postpartum iron defi- guidance from the SSC of the ISTH. J Thromb
ciency anemia: a randomized trial. J Matern Fetal Haemost. 2016;14:205–210.
Neonatal Med. 2013;26:654–659. 37. Mavrides E, Allard S, Chandraharan E, et al.
24. Kochhar PK, Kaundal A, Ghosh P. Intravenous Prevention and management of postpartum hae-
iron sucrose versus oral iron in treatment of iron morrhage: green-top guideline no. 52. BJOG.
deficiency anemia in pregnancy: a randomized 2017;124:e106–e149.
clinical trial. J Obstet Gynaecol Res. 2013;39: 38. Tanaka H, Matsunaga S, Yamashita T, et al.
504–510. A systematic review of massive transfusion protocol
25. Shi Q, Li J, Yao Q, et al. Intravenous iron sucrose: in obstetrics. Taiwan J Obstet Gynecol. 2017;56:
an alternative for oral iron in pregnancy with 715–718.
iron deficiency anemia. Ann Hematol. 2015;94: 39. Collins PW, Cannings-John R, Bruynseels D,
523–525. et al. Viscoelastometric-guided early fibrinogen
26. Friedrisch JR, Friedrisch BK. Prophylactic iron concentrate replacement during postpartum hae-
supplementation in pregnancy: a controversial issue. morrhage: OBS2, a double-blind randomized
Biochem Insights. 2017;10:1178626417737738. controlled trial. Br J Anaesth. 2017;119:411–421.
27. Forgie MA, Wells PS, Laupacis A, et al. Preop- 40. Cohen MJ, Christie SA. Coagulopathy of trauma.
erative autologous donation decreases allogeneic Crit Care Clin. 2017;33:101–118.
transfusion but increases exposure to all red blood 41. Afshari A, Wikkelso A, Brok J, et al. Thrombe-
cell transfusion: results of a meta-analysis. Inter- lastography (TEG) or thromboelastometry (RO-
national Study of Perioperative Transfusion (IS- TEM) to monitor haemotherapy versus usual care
POT) Investigators. Arch Intern Med. 1998;158: in patients with massive transfusion. Cochrane
610–616. Database Syst Rev. 2011:CD007871.
28. Vassallo R, Goldman M, Germain M, et al. 42. Snegovskikh D, Souza D, Walton Z, et al. Point-
Preoperative autologous blood donation: waning of-care viscoelastic testing improves the outcome
indications in an era of improved blood safety. of pregnancies complicated by severe postpartum
Transfus Med Rev. 2015;29:268–275. hemorrhage. J Clin Anesth. 2018;44:50–56.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Blood Products 839

43. Veigas PV, Callum J, Rizoli S, et al. A systematic 56. Maegele M, Schochl H, Cohen MJ. An update on the
review on the rotational thrombelastometry coagulopathy of trauma. Shock. 2014;41(suppl 1):
(ROTEM(R)) values for the diagnosis of coagul- 21–25.
opathy, prediction and guidance of blood 57. Wolberg AS, Meng ZH, Monroe DM III, et al.
transfusion and prediction of mortality in trauma A systematic evaluation of the effect of temper-
patients. Scand J Trauma Resusc Emerg Med. ature on coagulation enzyme activity and platelet
2016;24:114. function. J Trauma. 2004;56:1221–1228.
44. Wikkelso A, Wetterslev J, Moller AM, et al. 58. Reynolds BR, Forsythe RM, Harbrecht BG, et al.
Thromboelastography (TEG) or thromboelas- Hypothermia in massive transfusion. J Trauma
tometry (ROTEM) to monitor haemostatic treat- Acute Care Surg. 2012;73:486–491.
ment versus usual care in adults or children with 59. Goucher H, Wong CA, Patel SK, et al. Cell salvage
bleeding. Cochrane Database Syst Rev. 2016: in obstetrics. Anesth Analg. 2015;121:465–468.
CD007871. 60. Waters JH, Biscotti C, Potter PS, et al. Amniotic
45. Patel A, Goudar SS, Geller SE, et al. Drape fluid removal during cell salvage in the cesarean
estimation vs. visual assessment for estimating post- section patient. Anesthesiology. 2000;92:1531–1536.
partum hemorrhage. Int J Gynaecol Obstet. 2006;93: 61. Elagamy A, Abdelaziz A, Ellaithy M. The use of
220–224. cell salvage in women undergoing cesarean hys-
46. Toledo P, Eosakul ST, Goetz K, et al. Decay in terectomy for abnormal placentation. Int J Obstet
blood loss estimation skills after web-based di- Anesth. 2013;22:289–293.
dactic training. Simul Healthc. 2012;7:18–21. 62. Callum JL, Karkouti K, Lin Y. Cryoprecipitate:
47. Menard MK, Main EK, Currigan SM. Executive the current state of knowledge. Transfus Med Rev.
summary of the reVITALize initiative: standard- 2009;23:177–188.
izing obstetric data definitions. Obstet Gynecol. 63. American Association of Blood Banks ARC,
2014;124:150–153. America’s Blood Centers, Armed Services Blood
48. Main EK, Goffman D, Scavone BM, et al. Na- Program. Circular of information for the use of
tional Partnership for Maternal Safety: consensus human blood and blood component, Washington,
bundle on obstetric hemorrhage. J Obstet Gynecol DC. 2013. Available at: [Link]/downloads/
Neonatal Nur. 2015;44:462–470. BiologicsBloodVaccines/GuidanceCompliance
49. Casper LLR. OB Hemorrhage: Carts, Kits, and RegulatoryInformation/Guidances/Blood/UCM36
Trays California: CMQCC OB Hemorrhage [Link]. Accessed June 1, 2018.
Task Force. 2009. Available at: [Link]/ 64. Bonnet MP, Deneux-Tharaux C, Dupont C, et al.
resources-tool-kits/toolkits/ob-hemorrhage-toolkit. Transfusion practices in postpartum hemorrhage:
Accessed June 1, 2018. a population-based study. Acta Obstet Gynecol
50. Holcomb JB. Damage control resuscitation. Scand. 2013;92:404–413.
J Trauma. 2007;62(suppl 6):S36–S37. 65. Kennedy LD, Case LD, Hurd DD, et al. A
51. Holcomb JB, Jenkins D, Rhee P, et al. Damage prospective, randomized, double-blind controlled
control resuscitation: directly addressing the early trial of acetaminophen and diphenhydramine
coagulopathy of trauma. J Trauma. 2007;62: pretransfusion medication versus placebo for the
307–310. prevention of transfusion reactions. Transfusion.
52. del Junco DJ, Holcomb JB, Fox EE, et al. 2008;48:2285–2291.
Resuscitate early with plasma and platelets or 66. Heddle NM. Pathophysiology of febrile nonhe-
balance blood products gradually: findings from molytic transfusion reactions. Curr Opin Hema-
the PROMMTT study. . J Trauma Acute Care tol. 1999;6:420–426.
Surg. 2013;75(suppl 1):S24–S30. 67. Osterman JL, Arora S. Blood product transfu-
53. Holcomb JB, Tilley BC, Baraniuk S, et al. Trans- sions and reactions. Hematol Oncol Clin North
fusion of plasma, platelets, and red blood cells in a Am. 2017;31:1159–1170.
[Link] vs a [Link] ratio and mortality in patients with 68. Rana R, Fernandez-Perez ER, Khan SA, et al.
severe trauma: the PROPPR randomized clinical Transfusion-related acute lung injury and pulmo-
trial. JAMA. 2015;313:471–482. nary edema in critically ill patients: a retrospective
54. Bolliger D, Gorlinger K, Tanaka KA. Patho- study. Transfusion. 2006;46:1478–1483.
physiology and treatment of coagulopathy in 69. Curtis BR, McFarland JG. Mechanisms of trans-
massive hemorrhage and hemodilution. Anes- fusion-related acute lung injury (TRALI): anti-
thesiology. 2010;113:1205–1219. leukocyte antibodies. Crit Care Med. 2006;34
55. Shields LE, Wiesner S, Fulton J, et al. Compre- (suppl):S118–S123.
hensive maternal hemorrhage protocols reduce 70. Andreu G, Boudjedir K, Muller JY, et al. Analysis
the use of blood products and improve patient of transfusion-related acute lung injury and possible
safety. Am J Obstet Gynecol. 2015;212:272–280. transfusion-related acute lung injury reported to the

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
840 Gatta et al

french hemovigilance network from 2007 to 2013. 80. Andersen ND, Bhattacharya SD, Williams JB,
Transfus Med Rev. 2018;32:16–27. et al. Intraoperative use of low-dose recombinant
71. Vassallo R, Bachowski G, Benjamin RJ, et al. activated factor VII during thoracic aortic
A compendium of transfusion practice guidelines, operations. Ann Thorac Sur. 2012;93:1921–1928;
Washington, DC. American Red Cross. 2013. discussion 8–9.
Available at: [Link]/sites/arc/ 81. Phillips LE, McLintock C, Pollock W, et al.
files/59802_compendium_brochure_v_6_10_9_13. Recombinant activated factor VII in obstetric
pdf. Accessed June 1, 2018. hemorrhage: experiences from the Australian
72. Kruithof EK, Tran-Thang C, Gudinchet A, et al. and New Zealand Haemostasis Registry. Anesth
Fibrinolysis in pregnancy: a study of plasminogen Analg. 2009;109:1908–1915.
activator inhibitors. Blood. 1987;69:460–466. 82. Zatta A, McQuilten Z, Kandane-Rathnayake R,
73. Shakur H, Roberts I, Bautista R, et al. Collabo- et al. The Australian and New Zealand Haemo-
rators C-t. Effects of tranexamic acid on death, stasis Registry: ten years of data on off-licence use
vascular occlusive events, and blood transfusion of recombinant activated factor VII. Blood Trans-
in trauma patients with significant haemorrhage fus. 2015;13:86–99.
(CRASH-2): a randomised, placebo-controlled 83. Kozek-Langenecker SA, Afshari A, Albaladejo P,
trial. Lancet. 2010;376:23–32. et al. Management of severe perioperative bleed-
74. Collaborators WT. Effect of early tranexamic ing: guidelines from the European Society of
acid administration on mortality, hysterectomy, Anaesthesiology. Eur J Anaesthesiol. 2013;30:
and other morbidities in women with post-partum 270–382.
haemorrhage (WOMAN): an international, rand- 84. Meng ZH, Wolberg AS, Monroe DM III, et al.
omised, double-blind, placebo-controlled trial. The effect of temperature and pH on the activity
Lancet. 2017;389:2105–2116. of factor VIIa: implications for the efficacy of
75. Novikova N, Hofmeyr GJ. Tranexamic acid for high-dose factor VIIa in hypothermic and acidotic
preventing postpartum haemorrhage. Cochrane patients. J Trauma. 2003;55:886–891.
Database Syst Rev. 2010):CD007872. 85. Fenger-Eriksen C, Lindberg-Larsen M, Christen-
76. Ducloy-Bouthors AS, Jude B, Duhamel A, et al. sen AQ, et al. Fibrinogen concentrate substitution
High-dose tranexamic acid reduces blood loss in therapy in patients with massive haemorrhage
postpartum haemorrhage. Crit Care. 2011;15: and low plasma fibrinogen concentrations. Br J
R117. Anaesth. 2008;101:769–773.
77. Xu J, Gao W, Ju Y. Tranexamic acid for the 86. Bell SF, Rayment R, Collins PW, et al. The use of
prevention of postpartum hemorrhage after ce- fibrinogen concentrate to correct hypofibrinoge-
sarean section: a double-blind randomization naemia rapidly during obstetric haemorrhage. Int
trial. Arch Gynecol Obstet. 2013;287:463–468. J Obstet Anesth. 2010;19:218–223.
78. Pacheco LD, Hankins GDV, Saad AF, et al. 87. Glover NJ, Collis RE, Collins P. Fibrinogen
Tranexamic acid for the management of obstet- concentrate use during major obstetric haemor-
ric hemorrhage. Obstet Gynecol. 2017;130: rhage. Anaesthesia. 2010;65:1229–1230.
765–769. 88. Ducloy-Bouthors AS, Mignon A, Huissoud C,
79. Simonazzi G, Bisulli M, Saccone G, et al. et al. Fibrinogen concentrate as a treatment for
Tranexamic acid for preventing postpartum postpartum haemorrhage-induced coagulopathy:
blood loss after cesarean delivery: a systematic A study protocol for a randomised multicentre
review and meta-analysis of randomized con- controlled trial. The fibrinogen in haemorrhage of
trolled trials. Acta Obstet Gynecol Scand. DELivery (FIDEL) trial. Anaesth Crit Care Pain
2016;95:28–37. Med. 2016;35:293–298.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA CLINICAL OBSTETRICS AND GYNECOLOGY
Volume 61, Number 4, 841–850
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.

The Role of Centers

of Excellence With
Downloaded from [Link] by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3dW0s7N5CLiYMlH4wgaY34Fr69rtLH3oWi4+h6VxXUqo= on 10/30/2018

Multidisciplinary
Teams in the
Management of
Abnormal Invasive
Placenta
ALIREZA A. SHAMSHIRSAZ, MD,
KARIN A. FOX, MD, MEd, HADI ERFANI, MD, MPH,
and MICHAEL A. BELFORT, MD, PhD
Department of Obstetrics and Gynecology, Division of Maternal
Fetal Medicine, Baylor College of Medicine and Texas Children’s
Hospital, Houston, Texas

Abstract: Abnormal invasive placenta (AIP) causes is accomplished in centers with multidisciplinary expertise
significant maternal and perinatal morbidity and and experience in the care of AIP. This article highlights
mortality. With the increasing incidence of cesarean the desired features for developing and managing a multi-
delivery, this condition is dramatically more common in disciplinary team dedicated to the treatment of AIP in
the last 20 years. Advances in grayscale and Doppler center of excellence.
ultrasound have facilitated prenatal diagnosis of abnormal Key words: abnormal invasive placenta, center of
placentation to allow the development of multidisciplinary excellence, multidisciplinary team
management plans. Outcomes are improved when delivery
Abnormal invasive placentation (AIP),
Correspondence: Alireza A. Shamshirsaz, MD, Department which in this article will be defined as
of Obstetrics and Gynecology, Division of Maternal Fetal any degree of morbid placental invasion,
Medicine, Baylor College of Medicine Texas Children’s is associated with significant risk of mor-
Fetal Centre, Texas Children’s Hospital, Suite F1020,
Houston, TX. E-mails: alirezashamshirsaz@[Link]; tality, and in survivors, a considerable
shamshir@[Link] degree of morbidity which may include
The authors declare that they have nothing to disclose. coagulopathy and the adverse effects of

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 61 / NUMBER 4 / DECEMBER 2018

[Link] | 841
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
842 Shamshirsaz et al

massive transfusion, ureteral and/or other Patients with AIP present an even greater
organ injury, infection, and in some cases challenge to the system and its providers than
the need for reoperation. Even in the best non–AIP-related obstetric hemorrhage by
of circumstances, these cases present a virtue of the increased complexity of the
challenge to the obstetrician-gynecologist. surgery required, and utilization of available
The rate of invasive placentation appears to resources (both human and material). Care-
be rising, almost assuredly as a consequence ful perioperative planning and preparation
of the rising rate of cesarean delivery.1 minimize the morbidity and mortality in
Accordingly, all obstetric providers need AIP. Indeed, the presence of a dedicated,
to have a high index of suspicion for this multidisciplinary team has been associated
condition, and should be aware of risk with improved outcomes.8–10 This team
factors for invasive placentation. Given should include representation from all service
the substantial morbidity and increasing lines that may be involved during the ante-
incidence of AIP, it is paramount that we natal, intrapartum, and postnatal periods.
evaluate and implement strategies to im- The key to success is coordinated teamwork
prove outcomes. Although the optimum between providers who are experienced in
management of AIP is not supported by treating AIP. Ideally, multidisciplinary pre-
any level I trials, in the United States most operative consultation will occur along with
patients with a known diagnosis undergo a the use of a checklist to ensure comprehen-
planned cesarean hysterectomy. sive management (Table 1). Team training,
The incidence of peripartum hysterectomy continuous quality improvement, and on-
in the United States is estimated to be 0.8 per going simulation can further enhance the
1000 deliveries, and when confined to cesar- effectiveness of the team.11 Optimally, surgi-
ean deliveries the incidence increases to 1 per cal management will be planned, but the
200 cesarean deliveries.2,3 The most common ability to rapidly assemble essential team
indications for peripartum hysterectomy are members at any time allows for continuous,
invasive placentation (64.2%) and atony- high-quality care for patients with AIP.
related postpartum hemorrhage (26.9%).4,5 Descriptions of the roles of essential
Peripartum hysterectomy, particularly in team members are outlined below (Table 2).
cases of AIP, is often technically challenging
and is associated with a higher risk of
complications than abdominal hysterectomy
performed for a less acute indication.6 The Imaging Experts
particular challenges associated with peripar- Accurate antenatal diagnosis of AIP is a
tum cesarean hysterectomy include the crucial first step in planning and prepara-
following: (i) a more technically difficult tion for delivery. Expertise and experience
dissection of often edematous or highly in pelvic imaging is essential. Several
vascular tissue, (ii) abnormal neovasculariza- studies confirm that there is a decreased
tion and obliterated tissue planes (particularly incidence of hemorrhage and other ma-
at the uterine-bladder interface), (iii) need for ternal complications in those cases where
partial bladder resection, and (iv) a hemody- AIP is antenatally suspected rather than
namically unstable patient who is undergoing incidentally encountered at the time of
large-volume transfusion of blood products. delivery.12,13 The primary modality for
Given the increased risks and technical chal- the initial prenatal diagnosis of AIP is 2D
lenges, maternal perioperative morbidity and ultrasound used at the time of routine
mortality from major obstetric hemorrhage, anatomy or growth scanning. The use of
of any etiology, are lower in women deliver- 3-dimensional imaging and color Doppler
ing in high-volume hospitals compared with has been shown to improve the diagnostic
those delivering in low-volume hospitals.7 capabilities of ultrasound. However, a

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Multidisciplinary Management of AIP 843

TABLE 1. Abnormal Invasive Placenta Safety Checklist

Antepartum and BPR pre-surgical checklist
Consults Consents Laboratory Nursing
MFM Cesarean section Hemoglobin ___ Patient allergies:__/banded __
GYN oncologist Hysterectomy Hematocrit ___ Admission navigator complete
Urology Cystoscopy with bilateral stent Prenatal lab Newborn identification to OR
placement
Anesthesiology Epidural/ spinal/ general Type and cross within EFM in BPR
72 hours
Intensive care Arterial line Other: VTE prophylaxis: SCDs on
Interventional Central line
radiology
Blood Bank Medications
Notified: ___ (date) Pre-op anesthesia
4 units RBCs and FFP Antibiotic to OR
to OR

Pre-brief completed

OR pre- and intra-operative checklist

Anesthesiology Risk of blood loss (> 1000 mL) Surgical specialties
Antibiotic Blood in OR (4 units RBC; 4 units FFP; 4 and 4 in blood MFM and GYN-Oncology
prophylaxis bank)
Methergine in OR Cell saver Cesarean section/ hysterectomy
Hemabate in OR Urology
Misoprostil in OR Nursing Instruments/ equipment and supplies
Normothermia VTE prophylaxis: SCDs on Interventional radiology
measures
EFM during/ after epidural/ spinal On stand-by
Instruments, equipment, supplies
Other equipment/ supplies requested by
surgeon: _____

MFM indicates maternal-fetal medicine; OR, operating room; RBC, red blood cells.

pro forma for the standardized reporting but rather is optional and encouraged for
of ultrasound findings in such cases de- centers with experience in color Doppler
veloped by the European Working Group imaging.14 The field in which one subspeci-
for Abnormally Invasive Placenta (EW- alizes [eg, maternal-fetal medicine (MFM)
AIP), color Doppler imaging is not required, or radiology] is likely less important than the
individual evaluator’s knowledge and expe-
rience with the antenatal diagnosis of AIP.
TABLE 2. Multidisciplinary Team for The sensitivity and specificity for sec-
the Management of Abnormal ond-trimester and third-trimester ultra-
Invasive Placenta sonography for the identification of AIP
Core Services Ancillary Services has been reported to be as high as 80% to
97%.15–17 However, these data may over-
Obstetrics/maternal-fetal Vascular surgery
medicine
estimate the diagnostic accuracy of ultra-
Gynecologic oncology/ Trauma surgery sound, because they are derived from
surgery a preselected, referred population of
Urology Operating room patients, and are frequently interpreted
staff by a single expert with a priori knowledge
Anesthesiology Psychiatry/
psychology
of the suspected abnormality. In large
Imaging Social worker population-based studies, which include
Blood bank/transfusion patients screened in both academic cen-
medicine ters and in community clinics, antenatal
Neonatalogy diagnosis has been reported to occur in
Interventional radiology
Critical care medicine
~50% of cases.18,19 The diagnosis of AIP
can be increased by specific referral of

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
844 Shamshirsaz et al

those patients with known risk factors for the peculiarities of AIP (as mentioned above)
this condition. It is thus recommended that there is great potential for massive hemor-
patients with a history of any prior uterine rhage. The type of hemorrhage seen in these
surgery (and especially multiple cesarean cases is unlike that seen outside of obstetrics,
sections), placenta previa, endometrial abla- and even the most experienced trauma,
tion or uterine embolization, and first- vascular and cancer surgeons may have
trimester or second-trimester bleeding in difficulty in managing massive hemorrhage
the presence of other AIP risk factors be from deep in the lateral pelvic sidewall from
referred for specialist evaluation and imag- a bleeding percreta if they are called in after
ing. Use of a checklist directed at recognizing the fact in a patient in-extremis and have
risk factors for AIP may alert sonographers limited experience with AIP. The exsangui-
to focus on a closer inspection of the placenta. nations that occurs via the abnormal vascu-
Referral is also recommended any time a lature of a percreta, always fed by an
screening ultrasound suggests an abnormality extensive collateral circulation that includes
(abnormal placental appearance, abnormal the external iliac branches as well as direct
uterine shape, increased vascularity of the branches from the aorta (inferior mesenteric,
myometrial wall, current or prior cesarean ovarian, and sacral arteries) and cervico-
scar pregnancy, ectopic pregnancy).11 vaginal branches, occurs so rapidly that
efforts to keep up with coagulation factors,
red cell transfusion, electrolyte balance and
Obstetrician/MFM Specialists volume status are often overwhelmed. Com-
Once the diagnosis is suspected, patients pounding this is the completely abnormal
should be evaluated by a provider(s) pelvic architecture which is often deformed
experienced in AIP management and by overlying and bulging or invasive placen-
receive care from such providers through- tal tissue with distorted retroperitoneal struc-
out the antenatal period. tures (including the ureters, major blood
Early admission to the antepartum unit vessels and nerves). All of this is further
3 to 5 days before a scheduled cesarean complicated by limited access within the
hysterectomy is strongly advised for mul- narrowest portion of the bony pelvis. The
tidisciplinary consultation and antenatal specific credentials of the surgeon (eg,
corticosteroid administration which is co- gynecologic oncologist, general obstetrician-
ordinated by the obstetrician/MFM team gynecologist, MFM specialist, etc.) are prob-
members. When we analyzed data from our ably less important than the cumulative and
AIP patients delivered between January ongoing experience and the surgical skill and
2013 and May 2017, ~50% required urgent acumen gained from the consistent exposure
cesarean hysterectomy before their planned from managing cases of AIP.
delivery date due to contractions, bleeding, We have found that the establishment
or other maternal or fetal indications. This of a dedicated multidisciplinary team is a
finding is consistent with other cohorts20,21 necessary but not entirely sufficient approach
and highlights the importance of counseling to optimizing outcomes for patients with
patients about signs and symptoms of labor AIP. The team must not only possess multi-
and of being close to a hospital with disciplinary skills, but also needs an appro-
appropriate facilities as delivery nears. priate volume of patients to maintain smooth
and consistent high-level performance.22 We
have demonstrated that patient outcomes
Expert Surgeons improve over time with increasing experience
Appropriate surgical expertise and ongoing within a stable, multidisciplinary team per-
experience with AIP cases is very important forming at least 2-3 cases per month. We
for the safe management of AIP. Because of propose that even small, collective changes in

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Multidisciplinary Management of AIP 845

team dynamics and system process will lead underwent delivery with combined spinal-
to continuous improvement in clinical out- epidural anesthesia with conversion to gen-
comes. Surgical expertise and multidiscipli- eral endotracheal anesthesia for completion
nary teamwork will also inspire innovation of the hysterectomy.
and lead to ongoing development of im- Preoperative team discussion with the
proved surgical technique.23,24 anesthesiologist allows time to identify
airway or vascular access concerns, to
review the anticipated level of surgical
Anesthesiologists difficulty, and to identify whether the
The unique physiological changes of preg- patient is a candidate for regional anes-
nancy require specialized knowledge and thesia for cystoscopy, stent placement,
training in obstetric anesthesiology, partic- and cesarean delivery. We believe that
ularly when delivery entails complex surgery the approach should be individualized
and massive blood loss. Anesthesiologists after input from key team members (in-
experienced in both massive hemorrhage cluding the patient), with safety as the
and obstetrics are key players in the care of overarching and guiding principle.
women with AIP. All patients with AIP
should have a preoperative consultation and
evaluation by an anesthesiologist with expe- Urologists
rience in such cases. We recommend place- Abnormally invasive placenta into the blad-
ment of at least 2 large-bore intravenous der and pelvic side walls may obscure visual-
catheters as well as access for hemodynamic ization of the ureters, especially when the
monitoring (eg, arterial line in all and central placenta is low lying. The possibility of
venous line in some cases). A rapid blood intraoperative cystotomy and resection of
warmer/infusion and cell-saver assist the part of the bladder, and the potential for
anesthesiology team with blood product ureteral injury during hysterectomy, is con-
and fluid resuscitation. Pneumatic compres- siderably higher in cases of AIP than in less
sion devices and adequate warming equip- complex hysterectomies.3 Patients with AIP
ment to maintain maternal core temperature frequently require urological intervention
at a euthermic level are also essential. to prevent or repair injury to the urinary
The optimal type of anesthesia for AIP tract.25 In cases of bladder invasion we
for delivery is unclear, and there are pros prefer to perform an intentional cystotomy
and cons to starting with general endo- to allow delineation of the involved bladder
tracheal anesthesia versus starting with wall, and then to resect that portion of
regional analgesia and converting to gen- deeply invaded bladder tissue rather than
eral anesthesia intraoperatively after deliv- to engage in extensive dissection at the
ery of the baby and assessment of the bladder/placenta interface. We believe that
degree of AIP. General endotracheal anes- this avoids the sudden and massive hemor-
thesia may be preferred in cases compli- rhage that can occur when the placenta is
cated by antepartum bleeding due to the inadvertently entered during bladder dissec-
potential for ongoing massive hemorrhage tion. We performed intentional cystotomy in
and prolonged surgery. When massive ~15% of our cases. Our data suggest that
blood loss is expected, a complete sympa- persistent vesicovaginal fistula is very rare
thectomy that can occur with spinal anal- [1.4% (2/142)] with this approach.
gesia may impair the patient’s sympathetic Although not universally recommended,
response to sudden onset hypovolemia, and there are mounting data to suggest lower risk
thereby limit her ability to vasoconstrict of ureteral injury after placement of ureteral
and increase her systemic vascular resist- stents.10,26,27 We found ureteral stents espe-
ance. In our series, close to 40% of patients cially helpful in cases when bleeding that

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
846 Shamshirsaz et al

obscures the anatomy occurs deep in the part of our team, and they direct the
pelvis since the ureter can be readily palpated preparation of essential blood products,
and clamps can be applied with less of a risk suggest alternatives if type-specific prod-
of ureteric involvement. ucts are in short supply, and assist in
monitoring trends in coagulation profiles,
whether using standard laboratory tests
or thromboelastography. In intraopera-
Blood Bank/Transfusion tive emergencies, the transfusion medicine
Medicine Services specialist may need to take over the
Our Hematology-Pathology colleagues decision-making about blood product
are critical to the team and run a state- transfusion and make recommendations
of-the art, well-stocked blood bank that about electrolyte management. The need
functions at full capacity 24 hours a day, for an adequate blood bank that can
7 days a week. The most severe maternal support massive transfusion if required
complications we see in these cases result at time of delivery cannot be overstated.
almost exclusively from massive hemor-
rhage. The median estimated blood loss in
the setting of AIP ranges from 1500 to Interventional Radiologists
8000 mL. Many women require multiple The capacity to perform safe and timely
units of blood and other blood products. therapeutic interventional radiologic pro-
The reported median number of units of cedures is required. There is debate about
blood required ranges between 5 and 6 the routine use of pelvic artery balloon
units.8–10 In our recent series, 15% of all occlusion. Advocates believe that using
patients with AIP (19/130) developed balloon catheters to occlude the uterine or
coagulopathy, and 16% (21/130) required internal iliac arteries substantially de-
activation of our massive transfusion creases blood loss in cases of AIP, espe-
protocol. Replacement of red cells and cially percreta.28,29 Typically balloons are
coagulation products is not the only life- placed but not inflated before initiating
saving activity required. Electrolyte dis- the surgery. They are then inflated after
turbances, particularly severe ionized delivery of the neonate so as to not
hypocalcemia from rapid infusion of large compromise fetal blood supply. However,
volumes of citrate that accompany mas- the procedure has considerable risks such
sive transfusion, must be expeditiously as arterial damage, occlusion of major
and continuously addressed. In our series, vessels with tissue infarction, and
10% of our patients (13/130) experienced infection.29–31 The opposing opinion is
critically low (panic level) serum total based on the fact that by occluding the
calcium ( ≤ 7 mg/dL) and ionized calcium internal iliac arteries blood is simply
(< 0.7 mmol/L) during transfusion that diverted to the considerable collateral
necessitated immediate treatment. circulation to the uterus and placenta
The facility and blood bank should (via the external iliac artery) which results
have a well-established massive transfu- in a marked engorgement of deeper and
sion protocol. There should be plentiful less accessible vessels in the pelvis and
type and cross matched packed red blood increased risk of blood loss. There may
cells, fresh frozen plasma, cryoprecipitate, be a role for common iliac balloon
platelets, cell-saver technology and alter- occlusion32 or even balloon occlusion of
nate blood products, such as lyophilized the infrarenal aorta above the inferior
fibrinogen concentrates (ie, Riastap) and mesenteric artery.
factor concentrates (Kcentra). Transfu- Although prophylactic use of balloons
sion medicine specialists are an essential is controversial, radiographic embolization

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Multidisciplinary Management of AIP 847

of selected large vessels in some cases may realistic expectations about neonatal length of
be lifesaving. Ongoing hemorrhage after stay, neonatal hospital course, the ability to
hysterectomy, may be diffuse due to con- breastfeed or hold the neonate, and the effects
sumptive coagulopathy and may not be of prematurity. In our series 83% (107/130) of
amenable to simply identifying and “tying neonates had neonatal ICU admission with a
off” a bleeder. In such cases, radiographic median length of 18 days.
embolization can be used. The ability to
embolize in the operating room, ideally
using a hybrid operating room or alter- Critical Care Medicine
natively a portable fluoroscopy C-arm and
other equipment allows this to be per-
Specialists
Many patients with AIP will require ICU
formed in a safe environment. Under no
admission in the immediate postpartum
circumstances should a hemodynamically
period. In our cohort, ~50% were admitted
unstable bleeding patient be removed from
to our obstetric ICU. In addition, 28% (37/
an operating room and taken to an inter-
130) of our patients required postoperative
ventional radiology suite for embolization
ventilator support for ≥ 4 hours mostly
in the hopes of treating ongoing massive
because of concerns about the pulmonary
hemorrhage.
effects of extensive fluid resuscitation. This
may be prolonged in cases complicated by
transfusion-related acute lung injury or
Neonatologists pulmonary edema. Some women require
In general, delivery between 34 and 35 weeks
vasopressor support and invasive hemody-
of gestation is advised to balance the risk
namic monitoring necessitating close sur-
of maternal bleeding against neonatal
veillance and expert nursing care. Others
prematurity1,33; however, earlier delivery is
may need specialized care that can only be
warranted in the setting of active bleeding
provided in an intensive care setting. The
and in women with preterm premature
intensivist caring for the AIP patient should
rupture of membranes or increasing uterine
have experience in postsurgical care and the
activity.20 Conversely, some patients who
recognition and management of ongoing
remain very stable throughout pregnancy,
intra-abdominal hemorrhage. The surgical
with no bleeding and no preterm contrac-
team should have open access to the unit, as
tions may be able to safely deliver at 36
some women will have complications that
weeks’ gestation.21 The average gestational
require reoperation. Early detection of
age of delivery in women with AIP is
these complications is vital to avoid acute,
typically between 34 and 36 weeks of gesta-
irreversible decompensation.
tion, usually following medically indicated
preterm birth.8–10 AIP itself may not have a
direct effect on neonatal outcomes but cer-
tainly leads to the need for indicated preterm Ancillary Services
birth.34 In our series, ~50% of our patients (Intraoperatively)
had an unscheduled or emergent delivery Utilizing surgical technologists and circulat-
earlier than the planned 34 to 35 weeks due ing nurses familiar with invasive placentation
to bleeding or uterine contractions.35 We is extremely helpful, as with any specialized
recommend antenatal corticosteroid admin- surgical procedure. Planned cesarean hyster-
istration to improve neonatal outcomes ectomy for AIP is often best performed in a
before 37 weeks. Antenatal consultation with “main” operating suite with equipment that
a neonatologist and a tour of the neonatal may not be readily available on a labor and
intensive care unit (ICU) provides families delivery unit (eg, large self-retaining retrac-
the opportunity to ask questions and to have tors, cystoscopy equipment, ureteral stents

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
848 Shamshirsaz et al

and point-of-care testing devices allowing attain and retain the skills required to
rapid determination of hematocrit, electro- serve as a senior surgeon leading a
lytes, coagulation parameters, blood gases, team. We estimate that ~5 years and
and lactate). In such cases, labor and delivery close to 100 cases are necessary before
nursing personnel should also circulate to this goal can be optimally realized.
address issues specific to obstetrics and neo- (2) Stability: Having the same surgeons and
natal resuscitation. If an institution prefers to surgical team leads to an improved
perform these procedures on the labor and understanding among all team members.
delivery unit, then arrangements should be This also encourages ongoing, sequen-
made to have the appropriate equipment and tial, systems and process improvement
personnel available, which may include gy- that can lead to improved outcomes.
necologic oncologists, vascular surgeons, (3) Resources and purpose: An “AIP team”
trauma surgeons, or general surgeons. is an expensive proposition and rarely is
an economically positive undertaking
for the surgical team, the anesthesiology
Ancillary Services group, or the hospital. Regional Centers
(Preoperatively/ of Excellence are needed to allow con-
tinued high-level functioning and the
Postoperatively) development of such a team should
Other important groups of people involved
be considered as an essential quality
in the multidisciplinary management of
and safety endeavor and a community
AIP are social workers and psychiatrists/
service.
psychologists. The psychological and social
(4) Trust and communication: Establishing
impact of placenta accreta has yet to be
an AIP referral center can cause disquiet
fully elucidated; however, studies suggest
in systems in which people may pride
that traumatic birth experiences increase a
themselves on their surgical competency
woman’s risk for postpartum depression or
and dedication to their individual pa-
posttraumatic stress disorder.36 Psychiatrists
tients. We suggest that clear communi-
and social workers are available to provide
cation between referral center physicians
immediate and ongoing support to women
and referring providers, along with com-
and couples facing stressful and high-risk
mitment to the highest levels of safety
pregnancy
and service for all patients, will be well
In conclusion, we believe that the follow-
received and will represent a win-win
ing principles are important when developing
proposition for all parties.22,37
and managing a multidisciplinary team dedi-
cated to the treatment of AIP:
(1) Experience: Given the complex nature
of these cases it is important to select
team members based on exceptional References
surgical skill and dedication to the
1. Belfort MA. Publications committee, Society for
mission. Although the fundamental Maternal-Fetal Medicine. Placenta accreta. Am J
approach and specialized surgical skills Obstet Gynecol. 2010;203:430–439.
(ie, ureterolysis, isolation and ligation 2. Bateman BT, Mhyre JM, Callaghan WM, et al.
and/or sealing of vascular bundles and Peripartum hysterectomy in the United States:
abnormal neovascularization) remain nationwide 14-year experience. Am J Obstet
Gynecol. 2012;206:63.e1–63.e8.
constant, the presentation, location 3. Shellhaas CS, Gilbert S, Landon MB, et al. The
and depth of invasion varies in each frequency and complication rates of hysterectomy
case, and a significant amount of time accompanying cesarean delivery. Obstet Gynecol.
and additional training is required to 2009;114:224–229.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
Multidisciplinary Management of AIP 849

4. Flood KM, Said S, Geary M, et al. Changing increta, and percreta in the UK: a population-
trends in peripartum hysterectomy over the last based descriptive study. BJOG. 2014;121:62–70;
4 decades. Am J Obstet Gynecol. 2009;200: discussion 1.
632.e1–632.e6. 20. Bowman ZS, Manuck TA, Eller AG, et al. Risk
5. Glaze S, Ekwalanga P, Roberts G, et al. Peripartum factors for unscheduled delivery in patients with
hysterectomy: 1999 to 2006. Obstet Gynecol. 2008; placenta accreta. Am J Obstet Gynecol. 2014;210:
111:732–738. 241.e1–241.e6.
6. Wright JD, Devine P, Shah M, et al. Morbidity 21. Rac MW, Wells CE, Twickler DM, et al. Placenta
and mortality of peripartum hysterectomy. Obstet accreta and vaginal bleeding according to gesta-
Gynecol. 2010;115:1187–1193. tional age at delivery. Obstet Gynecol. 2015;125:
7. Wright JD, Herzog TJ, Shah M, et al. Region- 808–813.
alization of care for obstetric hemorrhage and its 22. Shamshirsaz AA, Fox KA, Erfani H, et al. Multi-
effect on maternal mortality. Obstet Gynecol. disciplinary team learning in the management of
2010;115:1194–1200. the morbidly adherent placenta: outcome im-
8. Eller AG, Bennett MA, Sharshiner M, et al. provements over time. Am J Obstet Gynecol.
Maternal morbidity in cases of placenta accreta 2017;216:612.e1–612.e5.
managed by a multidisciplinary care team com- 23. Belfort MA, Shamshirsaz AA, Fox K. Minimiz-
pared with standard obstetric care. Obstet Gyne- ing blood loss at cesarean-hysterectomy for pla-
col. 2011;117:331–337. centa previa percreta. Am J Obstet Gynecol. 2017;
9. Al-Khan A, Gupta V, Illsley NP, et al. Maternal 216:78.e71–78.e72.
and fetal outcomes in placenta accreta after 24. Belfort MA, Shamshirsaz AA, Fox KA. A tech-
institution of team-managed care. Reprod Sci. nique to positively identify the vaginal fornices
2014;21:761–771. during complicated postpartum hysterectomy.
10. Shamshirsaz AA, Fox KA, Salmanian B, et al. Am J Obstet Gynecol. 2017;217:222.e1–222.e3.
Maternal morbidity in patients with morbidly 25. Norris BL, Everaerts W, Posma E, et al. The
adherent placenta treated with and without a urologist’s role in multidisciplinary management
standardized multidisciplinary approach. Am J of placenta percreta. BJU Int. 2016;117:961–965.
Obstet Gynecol. 2015;212:218.e1–218.e9. 26. Tam Tam KB, Dozier J, Martin JN Jr. Ap-
11. Silver RM, Fox KA, Barton JR, et al. Center of proaches to reduce urinary tract injury during
excellence for placenta accreta. Am J Obstet management of placenta accreta, increta, and
Gynecol. 2015;212:561–568. percreta: a systematic review. J Matern Fetal
12. Eller AG, Porter TF, Soisson P, et al. Optimal Neonatal Med. 2012;25:329–334.
management strategies for placenta accreta. 27. Merritt AJ, Crosbie EJ, Charova J, et al. Prophy-
BJOG. 2009;116:648–654. lactic pre-operative bilateral ureteric catheters for
13. Warshak CR, Ramos GA, Eskander R, et al. major gynaecological surgery. Arch Gynecol Ob-
Effect of predelivery diagnosis in 99 consecutive stet. 2013;288:1061–1066.
cases of placenta accreta. Obstet Gynecol. 2010;115: 28. Dubois J, Garel L, Grignon A, et al. Placenta
65–69. percreta: balloon occlusion and embolization of
14. Alfirevic Z, Tang AW, Collins SL, et al. Ad-hoc the internal iliac arteries to reduce intraoperative
International AIP Expert Group. Pro forma for blood losses. Am J Obstet Gynecol. 1997;176:
ultrasound reporting in suspected abnormally 723–726.
invasive placenta (AIP): an international consen- 29. Ballas J, Hull AD, Saenz C, et al. Preoperative
sus. Ultrasound Obstet Gynecol. 2016;47:276–278. intravascular balloon catheters and surgical out-
15. Berkley EM, Abuhamad AZ. Prenatal diagnosis comes in pregnancies complicated by placenta
of placenta accreta. Is sonography all we need? accreta: a management paradox. Am J Obstet
J Ultrasound Med. 2013;32:1345–1350. Gynecol. 2012;207:216.e1–216.e5.
16. Comstock CH, Bronsteen RA. The antenatal 30. Greenberg JI, Suliman A, Iranpur P, et al. Pro-
diagnosis of placenta accreta. BJOG. 2014;121: phylactic balloon occlusion of the internal
171–181. iliac arteries to treat abnormal placentation: a
17. Jauniaux E, Bhide A. Prenatal ultrasound diag- cautionary case. Am J Obstet Gynecol. 2007;197:
nosis and outcome of placenta previa accreta after 470.e1–470.e4.
cesarean delivery: a systematic review and meta- 31. Bishop S, Butler K, Monaghan S, et al. Multiple
analysis. Am J Obstet Gynecol. 2017;217:27–36. complications following the use of prophylactic
18. Bailit JL, Grobman WA, Rice MM, et al. Mor- internal iliac artery balloon catheterization in a
bidly adherent placenta treatments and outcomes. patient with placenta percreta. Int J Obstet Anesth.
Obstet Gynecol. 2015;125:683–689. 2011;20:70–73.
19. Fitzpatrick KE, Sellers S, Spark P, et al. The 32. Clausen C, Stensballe J, Albrechtsen CK, et al.
management and outcomes of placenta accreta, Balloon occlusion of the internal iliac arteries in

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.
 DKA
850 Shamshirsaz et al

the multidisciplinary management of placenta using a multidisciplinary team approach for

percreta. Acta Obstet Gynecol Scand. 2013;92: morbidly adherent placenta. Obstet Gynecol. 2018;
386–391. 131:234–241.
33. Robinson BK, Grobman WA. Effectiveness of 36. Halperin O, Sarid O, Cwikel J. The influence of
timing strategies for delivery of individuals with childbirth experiences on women’s postpartum
placenta previa and accreta. Obstet Gynecol. traumatic stress symptoms: a comparison between
2010;116:835–842. Israeli Jewish and Arab women. Midwifery.
34. Balayla J, Bondarenko HD. Placenta accreta and 2015;31:625–632.
the risk of adverse maternal and neonatal out- 37. Fox KA, Shamshirsaz AA, Belfort MA. Center of
comes. J Perinat Med. 2013;41:141–149. excellence for placenta accreta. In: Silver RM, ed.
35. Shamshirsaz AA, Fox KA, Erfani H, et al. Out- Placenta Accreta Syndrome. Boca Raton, Florida:
comes of planned compared with urgent deliveries CRC Press; 2017:113–122.

[Link]
Copyright r 2018 Wolters Kluwer Health, Inc. All rights reserved.