Changes and Problematic Areas in Interpretation of

the AJCC Cancer Staging Manual, 6th Edition, for

Breast Cancer
James L. Connolly, MD

● Context.—Tumor stage is an important prognostic factor nodes, for example, isolated tumor cells, micrometastases,
and guides therapy for patients with breast cancer. The metastases, and the combination of locations. Another dif-
American Joint Committee on Cancer (AJCC) periodically ficult area is assignment of a correct size. The use of clin-
updates the staging standards. This article describes chang- ical and imaging studies for optimal pathologic staging is
es and problematic areas in interpretation of the AJCC discussed. Finally, the proper use of the TNM descriptors
Cancer Staging Manual, 6th edition, for breast cancer and is described.
provides practical advice. Conclusions.—The various practical problems that can
Objectives.—This article reviews the variety of practical arise during the assessment of important prognostic and
problems that can arise during assessment of the patholog- predictive features included in the College of American Pa-
ic stage and other prognostic/predictive factors included thologists Checklist for Evaluation of Resected Breast Can-
in the College of American Pathologists Checklist for Eval- cers are discussed, and specific recommendations are giv-
uation of Resected Breast Cancers. Potential practical dif- en.
ficulties that can arise include the classification of lymph (Arch Pathol Lab Med. 2006;130:287–291)

S ince publication in 2002, there have been changes or

clarifications primarily in what is considered the clas-
sification of pN0(i) and in the definition of isolated tumor
morphologic technique, including hematoxylin-eosin and
IHC), and no ITC cluster greater than 0.2 mm.1

cells (ITCs).1 In the AJCC Cancer Staging Manual, 6th edi- CLASSIFICATION OF ISOLATED TUMOR CELLS
tion, published in 2002, pN0(i⫺) was classified as no re- The classification of ITCs seems straightforward, that is,
gional lymph node metastasis histologically, negative im- ITCs with no cluster larger than 0.2 mm (Figure 1). Other
munohistochemistry (IHC). The pN0(i⫹) classification as than this statement, no guidelines have been elaborated.
of publication in 2002 was no regional lymph node me- The following examples are unofficial and represent how
tastasis histologically, positive IHC, and no IHC cluster I personally would interpret different situations.
greater than 0.2 mm. The classification of pN1mi(i⫹) as A not uncommon occurrence is finding more than 1
of publication in 2002 was micrometastasis detected only group of cells that individually would qualify as ITCs
by IHC.2 (Figure 2). This example could be classified as a lymph
In late 2003, changes or clarifications to these classifi- node with 2 groups of ITCs, pN0(i⫹), or if one includes
cations were made for regional lymph nodes (pN), and the distance between the 2 groups in the measurement, it
the definitions of pN0(i⫹) and pN0(i⫺) were updated to would be classified as pN1a. I believe the intent would be
indicate that i⫾ refers to the presence or absence of ITCs to classify this as a lymph node with 2 groups of ITCs,
detected by any morphologic technique, including hema- pN0(i⫹) rather than pN1a.
toxylin-eosin staining and IHC.1 The reporting element Another common occurrence is diffuse involvement by
pN1mi(i⫹) was eliminated, and pN0(i⫺) became no re- ITCs, as is often seen with infiltrating lobular carcinoma
gional lymph node metastasis histologically, negative mor- (Figure 3). While these are truly ITCs, I believe the intent
phologic findings for ITCs (any morphologic technique, in- would be to classify this as node positive (pN1a), based
cluding hematoxylin-eosin and IHC). on the number of tumor cells.
pN0(i⫹) now reads no regional lymph node metastasis PATHOLOGIC STAGING
histologically, positive morphologic findings for ITCs (any
It is important for pathologists to be aware of the ele-
ments of clinical staging because some of these can be
Accepted for publication October 26, 2005. used by pathologists to fill voids in pathologic staging.
From the Department of Pathology, Beth Israel Deaconess Medical Pathologic staging includes all data used for clinical stag-
Center, Boston, Mass. ing plus data from surgical exploration and resection.2
The author has no relevant financial interest in the products or com-
panies described in this article.
‘‘Imaging findings within 4 months of diagnosis in the
Reprints: James L. Connolly, MD, Department of Pathology, Beth Is- absence of disease progression or through completion of
rael Deaconess Medical Center, 330 Brookline Ave, Boston, MA 02215 surgery(ies), whichever is longer’’ are considered elements
(e-mail: [email protected]). of staging.2 Such imaging findings would include the size
Arch Pathol Lab Med—Vol 130, March 2006 Interpretation of AJCC 6th Edition—Connolly 287
Figure 1. Diagrammatic representation of isolated tumor cells (ITCs). Isolated tumor cells are defined as no regional lymph node metastasis histo-
logically, positive morphologic findings for ITCs (any morphologic technique, including hematoxylin-eosin and immunohistochemistry), and no ITC
cluster greater than 0.2 mm.1 The node shown would be classified as pN0 (i⫹) because there is a single focus of tumor that measures 0.1 mm.
Figure 2. Diagrammatic representation of a lymph node with 2 groups of tumor cells, each measuring 0.1 mm. There are no written guidelines
for classification. I would recommend classifying this example as a lymph node with 2 groups of isolated tumor cells. Others would measure the
greatest distance and classify this as pN1a.
Figure 3. Diagrammatic representation of a lymph node with numerous isolated tumor cells. This pattern is commonly seen with infiltrating
lobular carcinoma. I would classify this example as a positive node (pN1a).
Figure 4. Tumor size. This diagram shows one reason that adding the size of tumors from 2 specimens is not appropriate for staging.
Figure 5. Tumor size. How far apart do tumor nodules have to be to be considered as separate? Again, there are no written guidelines. I would
recommend that if they are clearly separate, the largest be measured for pT size. If they are close, I would refer to imaging studies if possible.
Figure 6. Tumor size. How far apart do tumor nodules have to be to be considered as separate? If imaging or other examinations indicate 1
lesion, measure the greatest dimension. If imaging or other examinations indicate multiple lesions, measure the largest.
288 Arch Pathol Lab Med—Vol 130, March 2006 Interpretation of AJCC 6th Edition—Connolly
of the primary tumor, chest wall invasion, and the pres- ologic) is classified according to the size of the tumor mass
ence of regional or distant metastases.2 or invasive component.2
The pathologic tumor size (pT) can be assigned with
microscopic, but not macroscopic, tumor at margins. With MICROINVASION OF BREAST CARCINOMA
macroscopic tumor at margins the classification is pTX. Microinvasion is the extension of cancer cells beyond the
If a tumor is grossly cut across and there is significant basement membrane into the adjacent tissues with no fo-
tumor in a subsequent excision, an accurate pT classifi- cus more than 0.1 cm in greatest dimension. When there
cation cannot be assigned. It is very important to know are multiple foci of microinvasion, the size of only the
that it is not permissible to add sizes together. I like to use largest focus is used to classify the microinvasion.2 When
the example of an orange cut in half. An 8-cm orange cut there are multiple foci of microinvasive carcinoma, it is
in half gives two 8-cm-diameter halves. The entire orange not appropriate to add them together. The presence of
is 8 cm and not 16 cm (Figure 4). With 2 specimens, one multiple foci of microinvasion should be noted and/or
could use the largest measurement and report the tumor quantified, as it is with multiple larger invasive carcino-
as ‘‘at least pT2.’’ A more accurate estimate may be based mas.
on imaging studies.
Pathologic stage groupings may include any combina- MULTIPLE SIMULTANEOUS IPSILATERAL PRIMARY
tion of pathologic and clinical classifications. For example, CARCINOMAS
Multiple simultaneous ipsilateral primary carcinomas
pT pN pM are defined as infiltrating, macroscopically measurable
pT pN cM carcinomas. Use the largest primary carcinoma to desig-
cT cN pM. nate T classification. It is not appropriate to assign a sep-
The measurement used for classifying the primary tu- arate T classification for the smaller tumor(s); however,
mor (T) is the one judged to be most accurate for that one should record that this is a case of multiple simulta-
particular case (ie, physical examination or imaging, such neous ipsilateral primary carcinomas.2
as mammography or ultrasound).2 A major unanswered question is how far apart tumors
have to be to be considered separate (Figure 5). If the foci
PATHOLOGIC TUMOR SIZE (pT) are clearly separate, measure the largest for pT. This area
requires a lot of judgment. When they microscopically ap-
The pathologic tumor size (pT) for the T classification pear very close, my recommendation would be if imaging
is a measurement of the invasive component only. For ex- or other modalities indicate 1 lesion, measure the greatest
ample, if there is a 4.0-cm intraductal component and a dimension. If on the other hand imaging or other modalities
0.3-cm invasive component, the tumor is classified pT1a. indicate multiple lesions, measure the largest (Figure 6).
This is an extremely important point. The size of the in-
vasive tumor should be verified microscopically. This rec- SKIN OF BREAST
ommendation does not imply that the size measured on Inflammatory Carcinoma
the slide replaces the gross measurement. The microscopic
Inflammatory carcinoma is a clinicopathologic entity
examination verifies that what is seen grossly is in fact
characterized by diffuse erythema and edema (peau
invasive carcinoma. If the microscopic examination reveals
d’orange) of the breast, often without an underlying pal-
that the gross ‘‘mass’’ is mostly ductal carcinoma in situ
pable mass. These clinical findings should involve the ma-
(DCIS) or a biopsy site reaction, measurement on the
jority of the skin of the breast.2 It is important to remem-
slide(s) may be a more accurate at establishing the true
ber that inflammatory carcinoma is primarily a clinical di-
invasive size.
agnosis. Involvement of the dermal lymphatics alone does
In patients who have undergone multiple core biopsies,
not indicate inflammatory carcinoma in the absence of
measuring only the residual lesion or the largest size on
clinical findings.
the core biopsy may result in significant underclassifica-
tion of the T component and thus understaging of the tu- pT4
mor. In such cases, original tumor size should be recon-
pT4 is a tumor of any size with direct extension to chest
structed on the basis of a combination of imaging and all
wall or skin, only as described as follows.
histologic findings.2
The imaging study that correlates best with the invasive ● T4a is defined as extension to chest wall, not including
size is usually felt to be magnetic resonance imaging fol- pectoralis muscle.
lowed by ultrasound,3 which also is true in the treated ● T4b is defined as edema (including peau d’orange) or
breast.4 ulceration of the skin of the breast, or satellite skin nod-
ules (grossly, not only microscopically seen) confined to
CARCINOMA IN SITU the same breast. Since diffuse peau d’orange would be
Carcinoma in situ, with no evidence of an invasive com- inflammatory carcinoma (T4d), I would assume that
ponent, is classified as Tis, with a subclassification indi- T4b would include limited skin edema.
cating type. Ductal carcinoma in situ is classified Tis Importantly, other skin changes, such as dimpling of the
(DCIS). Lobular carcinoma in situ is classified Tis (LCIS). skin, nipple retraction, or any other skin change (except
Cases with both DCIS and LCIS are classified Tis (DCIS).2 those described under T4b and T4d) may occur in T1, T2,
Paget disease of the nipple without an associated tumor or T3 without changing the classification.2
mass (clinical) or invasive carcinoma (pathologic) is clas-
sified Tis (Paget). REGIONAL LYMPH NODES (N)
Paget disease with a demonstrable mass (clinical) any- The new nodal classification has many categories that
where within that breast or an invasive component (path- are easy to understand, but some may either cause con-
Arch Pathol Lab Med—Vol 130, March 2006 Interpretation of AJCC 6th Edition—Connolly 289
fusion or be difficult for the pathologist to apply. The ma- vasive lobular and mucinous carcinomas. The Nottingham
jor classification change is that the number of involved combined histologic grade (Elston-Ellis modification of the
nodes dramatically changes the N stage. Scarff-Bloom-Richardson grading system5) is recommend-
The overall classification is based on axillary lymph ed.
node dissection, with or without sentinel lymph node dis-
section. If the classification is based solely on sentinel TNM DESCRIPTORS
lymph node dissection without subsequent axillary lymph The TNM descriptors are for identification of special
node dissection, it is designated (sn) for sentinel node, for cases of TNM or pTNM classifications; the ‘‘m’’ suffix and
example, pN0(i㛮) (sn).2 Intramammary lymph nodes are ‘‘y,’’ ‘‘r,’’ and ‘‘a’’ prefixes are used. Although they do not
classified as axillary lymph nodes. Cancerous nodules in affect the stage grouping, they indicate cases needing sep-
the axillary fat without evidence of residual lymph node arate analysis.
tissue are classified as positive axillary lymph nodes (N). The ‘‘m’’ suffix indicates the presence of multiple pri-
pN1mi is the classification for micrometastasis, which is mary tumors in a single site and is recorded in parenthe-
a metastasis greater than 0.2 mm with none greater than ses, for example, pT(m)NM.
2.0 mm.2 Therefore, if there is 1 metastasis greater than 2 The ‘‘y’’ prefix indicates those cases in which classifi-
mm and 3 micrometastases, it would be counted as 4 pos- cation is performed during or following initial multimo-
itive nodes. On the other hand, if there is 1 metastasis dality therapy (ie, neoadjuvant chemotherapy, radiation
greater than 2 mm and 3 nodes with ITCs, I would rec- therapy, or both chemotherapy and radiation therapy).
ommend counting this case as 1 positive node. The cTNM or pTNM category is identified by a ‘‘y’’ prefix.
pN1 is the category for metastasis in 1 to 3 axillary The ycTNM or ypTNM categorizes the extent of tumor
lymph nodes and/or in internal mammary nodes with actually present at the time of that examination. The ‘‘y’’
microscopic disease detected by sentinel lymph node dis- categorization is not an estimate of tumor prior to multi-
section, but not clinically apparent. This raises the ques- modality therapy (ie, before initiation of neoadjuvant ther-
tion as to what clinically apparent stands for. Clinically apy).
apparent is defined as detected by imaging studies (ex- The ypTNM may be misleading. The most important
cluding lymphoscintigraphy) or by clinical examination or predictor of prognosis depends on the response of the tu-
grossly visible pathologically. Conversely, not clinically mor to treatment. Patients with advanced disease with a
apparent is defined as not detected by imaging studies complete pathologic response have an excellent progno-
(excluding lymphoscintigraphy) or by clinical examination sis.6,7 Patients with advanced disease with only a partial
or grossly visible pathologically.2 response have a very poor prognosis, even when they ob-
pN1a is metastasis in 1 to 3 axillary lymph nodes. pN1b tain a favorable ypTN classification.7
is metastasis in internal mammary nodes with microscop- The ‘‘r’’ prefix indicates a recurrent tumor when staged
ic disease detected by sentinel lymph node dissection, but after a documented disease-free interval and is identified
not clinically apparent and with negative axillary nodes. by the ‘‘r’’ prefix, for example, rTNM.
I predict this classification will not be commonly used The ‘‘a’’ prefix designates the stage determined at au-
since it is rare to biopsy internal mammary lymph nodes. topsy, for example, aTNM.
pN1c is metastasis in 1 to 3 axillary lymph nodes and in
internal mammary lymph nodes with microscopic disease RESIDUAL TUMOR (R)
detected by sentinel lymph node dissection, but not clin- Tumor remaining in a patient after therapy with cura-
ically apparent. Again, this classification will be rarely tive intent (eg, surgical resection for cure) is categorized
used. If associated with greater than 3 positive axillary by a system known as R classification, as follows.
lymph nodes, the internal mammary nodes are classified
RX Presence of residual tumor cannot be assessed,
as pN3b to reflect increased tumor burden.
R0 No residual tumor,
pN2a now stands for metastasis in 4 to 9 axillary lymph
R1 Microscopic residual tumor,
nodes (at least 1 tumor deposit greater than 2.0 mm).
R2 Macroscopic residual tumor.
pN2b is metastasis in clinically apparent (histologically
confirmed) internal mammary lymph nodes in the ab- During the past few years there have been a number of
sence of axillary lymph node metastasis. questions concerning implementation of the College of
pN3a now stands for metastasis in 10 or more axillary American Pathologists cancer protocols. The protocols re-
lymph nodes (at least 1 tumor deposit greater than 2.0 flect the AJCC 6th edition. The following are questions not
mm) or metastasis to the infraclavicular lymph nodes. already addressed in this article.
pN3b is metastasis in clinically apparent ipsilateral inter-
Q. In the Nottingham histologic scoring for invasive
nal mammary lymph nodes in the presence of 1 or more
breast tumor, is it mandatory to report the actual mi-
positive axillary lymph nodes, or in more than 3 axillary
totic count and total Nottingham score?
lymph nodes and in internal mammary lymph nodes with
A. If you report histologic grade by the combined Not-
microscopic disease detected by sentinel lymph node dis-
tingham score, scores for each of the 3 elements (score
section, but not clinically apparent. pN3c is metastasis in
1, 2, or 3) will be required (as well as the total score),
ipsilateral supraclavicular lymph nodes. Metastasis to any
but you will not have to separately list a specific mi-
other lymph node, including cervical or contralateral in-
totic count in addition to the score.
ternal mammary lymph nodes, is classified as distant
(M1). Q. If the Nottingham histologic scoring system for in-
vasive breast tumor is not used, is it mandatory to
HISTOLOGIC GRADE report the actual mitotic count and total Nottingham
All invasive breast carcinomas, with the exception of score?
medullary carcinoma, should be graded. This includes in- A. In contrast, a specific mitotic figure count will be re-
290 Arch Pathol Lab Med—Vol 130, March 2006 Interpretation of AJCC 6th Edition—Connolly
 quired for those who report histologic grade by a sys- men. If there are no changes in the required elements,
tem other than Nottingham. no protocol would be necessary. If there are changes
in the required elements, those elements at least
Q. Are checklists necessary for core biopsies? would have to be reported.
A. No. Checklists are for excision specimens.
I hope these observations have helped clarify issues with
Q. Do you report out items from the first specimen on the current staging system.
the second? Do you note it some special way? References
A. The gross and microscopic diagnoses of an individual 1. Singletary SE, Greene FL, Sobin LH. Classification of isolated tumor cells:
pathology report are primarily concerned with that clarification of the 6th edition of the American Joint Committee on Cancer staging
specimen, but the tumor summary (our wording) or manual. Cancer. 2003;98:2740–2741.
2. Greene FL, Page DL, Fleming ID, et al, eds, for the American Joint Com-
checklist can and should include information ob- mittee on Cancer. AJCC Cancer Staging Manual. 6th ed. New York, NY: Springer-
tained from earlier specimens. For instance, estrogen Verlag; 2002.
receptor, progesterone receptor, and HER-2 evalua- 3. Berg WA, Gutierrez L, NessAiver MS, et al. Diagnostic accuracy of mam-
mography, clinical examination, US, and MR imaging in preoperative assessment
tions may have already been done on the needle bi- of breast cancer. Radiology. 2004;233:830–849.
opsy, or perhaps a sentinel lymph node biopsy was 4. Yeh E, Slanetz P, Kopans DB, et al. Prospective comparison of mammog-
done on an earlier excision. raphy, sonography, and MRI in patients undergoing neoadjuvant chemotherapy
for palpable breast cancer. AJR Am J Roentgenol. 2005;184:868–877.
Q. Do reports for DCIS only need use the breast check- 5. Elston CW, Ellis IO. Pathological prognostic factors in breast cancer, I: the
value of histological grade in breast cancer: experience from a large study with
list? long-term follow-up. Histopathology. 1991;19:403–410.
A. No. 6. Ogston KN, Miller ID, Payne S, et al. A new histological grading system to
assess response of breast cancers to primary chemotherapy: prognostic signifi-
Q. Are College of American Pathologists protocols re- cance and survival. Breast. 2003;12:320–327.
quired for both lumpectomy and a subsequent mas- 7. Kuerer HM, Newman LA, Smith TL, et al. Clinical course of breast cancer
patients with complete pathologic primary tumor and axillary lymph node re-
tectomy? sponse to doxorubicin-based neoadjuvant chemotherapy. J Clin Oncol. 1999;17:
A. It depends on the findings in the mastectomy speci- 460–469.

Arch Pathol Lab Med—Vol 130, March 2006 Interpretation of AJCC 6th Edition—Connolly 291