J Am Soc Nephrol 10: 1900 –1907, 1999

Focal Segmental Glomerulosclerosis: Prognostic Implications

of the Cellular Lesion
MELVIN M. SCHWARTZ,* JONI EVANS,‡ RAY BAIN,‡ and STEPHEN M. KORBET†
*Department of Pathology and †Section of Nephrology, Department of Medicine, Rush Medical College,
Chicago, Illinois; and ‡The Biostatistics Center, George Washington University, Rockville, Maryland.

Abstract. The cellular lesion (CELL), seen in some patients were mainly black (94%), severely nephrotic (67%, .10 g/d),
with primary focal segmental glomerulosclerosis (FSGS), and had a poor response to treatment (23% remission). In
comprises proliferation, hypertrophy, and pathologic changes nephrotic patients, initial serum creatinine, interstitial expan-
in the cells overlying the glomerular scar. The prognosis of the sion $20%, and CELL independently predicted ESRD. How-
cellular lesion was retrospectively studied in 100 patients with ever, the rates of remission in treated nephrotic patients with
FSGS (43 had FSGS-CELL and 57 had FSGS without the FSGS-CELL and FSGS-CS were the same (9 of 17, 53%
cellular lesion (classic segmental scar [CS]). Patients with the versus 17 of 39, 52%), and patients in both groups who
FSGS-CELL lesion were more often black and severely pro- achieved a remission had a 5-yr survival of 100%. Steroid
teinuric and developed more end-stage renal disease (ESRD). treatment was the only variable that predicted remission. Pa-
Nephrotic patients with FSGS-CELL (n 5 39) were more tients with the FSGS-CELL have an increased prevalence of
proteinuric at presentation than patients with FSGS-CS (n 5 ESRD, but the improved prognosis associated with remission
36). ESRD developed more frequently in patients with the is so significant that a therapeutic trial is warranted in all
FSGS-CELL (17 of 39, 44% versus 5 of 36, 14%, P 5 0.005), nephrotic FSGS patients, regardless of the presence of the
and patients with extensive FSGS-CELL ($ 20% glomeruli) cellular lesion.

Primary focal segmental glomerulosclerosis (FSGS) is a glo- believe that the lesion represents an early stage in the evolution
merular disease that causes proteinuria and the nephrotic syn- of glomerular scarring in primary FSGS. However, the clinical
drome in children and adults (1,2). Although patients with significance of the cellular lesion has been controversial. De-
FSGS have a guarded prognosis and often develop renal in- twiler et al. (7) and Valeri et al. (8) suggest that it is associated
sufficiency, some enter remission after treatment (2). The with a poor response to therapy and rapid progression to
classic pathologic description of FSGS was limited to segmen- end-stage renal disease (ESRD); however, our initial experi-
tal glomerular scars (1), but a variety of glomerular lesions, ence in 81 adult patients (2,9) failed to support these conclu-
currently included under the diagnosis of primary FSGS (3,4), sions.
have been used to stratify patients for prognosis and response To further define the prognosis of the cellular lesion, we
to therapy. Glomerular visceral epithelial cell pathology, in- obtained additional follow-up in our original patients (2,9),
cluding proliferation, hypertrophy, vacuolization, protein re- added patients to the study, and included an analysis of the role
sorption droplets, nuclear changes, and separation from the that the proportion of glomeruli with the cellular lesion plays in
glomerular basement membrane, is a frequently recognized determining outcome. We present a retrospective analysis of
histologic feature of idiopathic FSGS (reviewed in reference clinical and pathologic features in 100 patients to assess
(5). This pathologic lesion and changes in the underlying whether the cellular lesion independently predicts progression
glomerular architecture are diagnosed as the cellular lesion of to ESRD in adult patients with primary FSGS.
primary FSGS (5) or collapsing glomerulopathy (6). Although
the nomenclature differs, the glomerular pathology of the cel-
lular lesion and collapsing glomerulopathy is identical, and we Materials and Methods
Study Design and Patient Selection
This study is a retrospective clinicopathologic analysis of all adult
Received December 24, 1998. Accepted March 27, 1999. patients ($15 yr old at presentation) with the diagnosis of primary
This work was presented in part at the 30th annual meeting of the American FSGS who were seen by the Renal Pathology Service and followed by
Society of Nephrology, San Antonio, TX, and has been published in abstract
the Section of Nephrology, Department of Medicine at Rush-Presby-
form (J Am Soc Nephrol 8: 98A, 1997).
Correspondence to Dr. Melvin M. Schwartz, Department of Pathology, Rush-
terian St. Luke’s Medical Center, Chicago, Illinois, between 1975 and
Presbyterian St. Luke’s Medical Center, 1753 West Congress Parkway, Chi- 1996. The diagnosis of primary FSGS was established if there was no
cago, IL 60612. Phone: 312-942-5262; Fax: 312-942-5254; E-mail: pathologic evidence for a primary glomerular disease or systemic
[email protected] disease process that might produce secondary segmental glomerular
1046-6673/1009-1900 sclerosis. The medical records were reviewed, and patients with
Journal of the American Society of Nephrology evidence of systemic disease, other diseases associated with glomeru-
Copyright © 1999 by the American Society of Nephrology lopathy, or a history of morbid obesity, reflux, nephrectomy, a solitary
J Am Soc Nephrol 10: 1900 –1907, 1999 Focal Segmental Glomerulosclerosis 1901

kidney, or intravenous drug abuse were excluded. Infection with HIV g/d, and a partial remission was defined as a urine protein between
was excluded by clinical criteria and laboratory evaluation. 0.31 and 2.5 g/d.

Definition of FSGS and Pathology Studies Treatment

FSGS is diagnosed when the glomerular lesion involves only a Treatment was limited to patients with FSGS and the nephrotic
portion of some of the glomeruli with others remaining relatively syndrome because historically untreated non-nephrotic patients have
uninvolved. The involved glomeruli have a scarred segment with had an excellent prognosis (10). The decision to treat with steroids, or
obliteration of the glomerular architecture and capillary lumens with immunosuppressive agents, was based solely on physician discretion.
or without adhesions to Bowman’s capsule. This describes the classic However, we generally recommend a trial of steroid therapy in ne-
lesion of FSGS (CS), and patients whose biopsies contained only phrotic patients with primary FSGS in whom renal function is rea-
classic segmental scars are referred to as FSGS-CS. The pathology of sonably well preserved (serum creatinine #3 mg/dl) and in whom it
the cellular lesion of FSGS (CELL) (5,9) is defined by hypercellular- is not otherwise contraindicated. Although no specific therapeutic
ity in Bowman’s space overlying a scarred or collapsed glomerulus, regimen was used, over the past 10 yr we have proposed a more
and the cellular lesion is either segmental or global in its intraglo- aggressive approach to the treatment of nephrotic adults with primary
merular distribution (see Tables 1 and 6). The proliferating/infiltrating FSGS (reviewed in reference (11). This comprises an initial course of
cells are hypertrophic with vesicular nuclei, a prominent nucleolus, prednisone given at a dose of 1 mg/kg per d (up to 80 mg) for 3 to 4
copious frequently basophilic cytoplasm with blebs, vacuoles, peri- mo. In patients demonstrating a complete or partial remission (see
odic acid-Schiff-positive resorption droplets, and focal separation above), the dose of prednisone is then tapered slowly over 2 to 3 mo.
from the underlying glomerular basement membrane. Patients whose Patients unresponsive to the initial therapy are tapered off prednisone
biopsies contained cellular lesions are referred to as FSGS-CELL. more rapidly, over 4 to 6 wk. To minimize potential toxicity in the
Global sclerosis means the entire glomerular tuft is scarred, and the elderly, we use alternate day prednisone (1 to 2 mg/kg up to 120 mg)
capillaries are obliterated without an associated cellular lesion. In for 4 to 5 mo.
Tables 1 and 6, it is apparent that cellular lesions and classic segmen-
tal scars may coexist in renal biopsies. For example, biopsies showing Statistical Analyses
the cellular lesion may also contain glomeruli with classic segmental
Comparison of the clinical and laboratory characteristics among
scars and/or global sclerosis.
patients with cellular lesions and classic segmental scars used the
On the basis of these clinical and morphologic criteria, a total of
Fisher exact test for categorical data (12) and the Wilcoxon rank-sum
100 adult patients with primary FSGS was identified and divided into
test for continuous data (13). Continuous data are reported as mean 6
two groups: patients in whom the glomeruli demonstrated only classic
SD. For the analysis of length of time from biopsy to ESRD, product-
segmental scars (FSGS-CS) and those having the cellular lesion in at
limit life-table distributions were compared with the log-rank test
least one glomerulus with or without other glomeruli with classic
(14). A proportional hazards regression model was used to evaluate
segmental scars or global sclerosis (FSGS-CELL).
potential clinical and laboratory predictors of progression to ESRD
Renal biopsy tissue was divided and processed for light, fluores-
(14). Potential clinical and histologic predictors of remission used
cence, and electron microscopy. In each case, glass slides stained with
logistic regression analysis (15).
hematoxylin and eosin, Masson’s trichrome, periodic acid-Schiff, and
methenamine silver-periodic acid-Schiff (Jones stain) were reviewed.
The average number of glomeruli examined by light microscopy per Results
biopsy was 23 6 17 (median 21, range 5 to 122). Histologic evalu- All Patients
ation was performed by M.M.S., who was blinded to the clinical data Pathology Studies. Histologic evaluation of all biopsies
pertinent to each case. In addition to establishing the diagnosis of demonstrated the cellular lesion (CELL) and/or classic seg-
FSGS, the following features were recorded: the number of glomeruli
mental scars (CS) in 11 6 9% (median 9%) and global scle-
with classic segmental scars; the number of glomeruli with global
scars; and the number of glomeruli with the cellular lesion, subdivided
rosis in 20 6 23% (median 12%) of glomeruli per biopsy. The
into segmental and global cellular lesions. Expansion of the cortical cortical interstitial volume was 20 6 13% (median 19%). The
interstitium was quantified on periodic acid-Schiff-stained sections by cellular lesion (Figure 1) was seen in 43 (43%) biopsies
point counting using an ocular grid. (FSGS-CELL) and classic segmental scars (FSGS-CS) (Figure
2) in 57 (57%) biopsies from patients with primary FSGS
(Table 1). Biopsies from patients with FSGS-CS had more
Clinical Studies and Laboratory Examination glomeruli with classic segmental scars than biopsies from
Demographic as well as clinical and laboratory information at the patients with FSGS-CELL. Although the overall proportion of
time of renal biopsy and at follow-up was obtained on each patient. diseased glomeruli per biopsy (total of CS, global sclerosis,
Presentation was defined as the time when proteinuria was first and CELL) was similar in the two groups (43 6 13% in
detected. A normal serum creatinine was defined as #1.3 mg/dl; renal FSGS-CELL versus 39 6 24% in FSGS-CS, P 5 NS), there
insufficiency, a creatinine .1.3 mg/dl; and ESRD was defined as a
appeared to be more diseased glomeruli with potentially pro-
serum creatinine $5 mg/dl, or the need for renal replacement therapy.
Patients were categorized based on the level of proteinuria at the time
gressive lesions (total of CS and CELL) in biopsies from
of biopsy. Non-nephrotic proteinuria was defined as ,3.0 g/d, ne- patients with FSGS-CELL. The fractional cortical interstitial
phrotic-range proteinuria as $3.0 g/d and massive proteinuria, .10 volume is approximately 5% in normal human kidneys (16),
g/d. Patients with a diastolic BP .90 mmHg or a systolic BP .140 and although it was increased in biopsies from patients with
mmHg were considered hypertensive. In patients with nephrotic pro- FSGS-CELL and FSGS-CS, it was not significantly different
teinuria, a complete remission was defined as a urine protein of #0.3 between the two groups.
1902 Journal of the American Society of Nephrology J Am Soc Nephrol 10: 1900 –1907, 1999

Table 1. Pathology studies: all patientsa

Characteristic FSGS-CS FSGS-CELL P Value

n 57 43
Total glomeruli 19 6 13 29 6 21 ,0.001
median 17 24
CS (%) 15 6 8 466 ,0.001
median 14 0
Global sclerosis (%) 24 6 24 16 6 20 NS
median 15 10
Segmental CELL (%) 0 13 6 12
median 9
Global CELL (%) 0 9 6 11
median 6
Total CELL (%) 0 22 6 18
median 16
Interstitium volume (%) 20 6 12 21 6 13 NS
Figure 1. The cellular lesion. Collapse of all of the glomerular median 17 20
capillaries and diffuse proliferation and hypertrophy of the visceral
epithelial cells. Methenamine silver-periodic acid-Schiff (Jones stain), a
Results are given as mean 6 SD. FSGS, focal segmental
3185. glomerulosclerosis; CS, classic segmental scars; CELL, the cellular
lesion.

Table 2. Characteristics at biopsy: all patientsa

FSGS-
Characteristic FSGS-CS P Value
CELL

n 57 43
Black 28 (49%) 30 (70%) 0.04
Male 34 (60%) 24 (56%) NS
Age (yr) 40 6 15 40 6 18 NS
median 40 39
Hypertension 34 (60%) 23 (53%) NS
Creatinine (mg/dl) 1.7 6 0.8 2.4 6 2.0 NS
median 1.5 1.7
Renal insufficiency 31 (54%) 28 (65%) NS
Proteinuria (g/24 h) 4.8 6 3.6 12.2 6 9.8 ,0.001
median 3.7 8.2
Nephrotic 36 (63%) 39 (91%) 0.002
Figure 2. Classic segmental scar. Glomerulus with a parahilar seg- Proteinuria .10 g/24 h 6 (11%) 19 (44%) 0.001
mental scar (asterisk) and an early adhesion (arrow) without changes Albumin (g/dl) 3.4 6 1.0 2.7 6 0.8 0.001
in the overlying visceral epithelial cells. Methenamine silver-periodic median 3.8 2.5
acid-Schiff (Jones stain), 3185. Presentation to biopsy (mo) 22 6 48 6 6 12 NS
median 3.9 1.5
Clinical Characteristics at Biopsy. Clinical features at a
Results are given as mean 6 SD.
the time of biopsy for all patients are presented in Table 2.
Patients with FSGS-CELL were more likely to be black and to
have higher levels of proteinuria and lower serum albumin
levels than patients with FSGS-CS. Virtually all patients with biopsy was not significantly different between the two groups
the cellular lesion had nephrotic-range proteinuria (91%) com- (Table 3). The proportion of patients with FSGS-CELL pro-
pared with only 63% of patients with FSGS-CS. There was a gressing to ESRD was more than twice that of patients with
trend for patients with FSGS-CELL lesion to have a higher FSGS-CS. At 5 yr, the renal survival was 45% (Figure 3) for
serum creatinine, greater prevalence of renal insufficiency, and patients with FSGS-CELL (median survival 36 mo) compared
shorter interval from clinical presentation to biopsy, but these to 83% for patients with FSGS-CS (median survival 158 mo,
tendencies were not significantly different. P , 0.001).
Outcome. The overall follow-up was longer for patients The univariate analysis of all 100 study patients (Table 4)
with FSGS-CS, but the duration of follow-up from the time of identified males, renal insufficiency, and level of proteinuria at
J Am Soc Nephrol 10: 1900 –1907, 1999 Focal Segmental Glomerulosclerosis 1903

Table 3. Outcome: all patientsa P , 0.01) in whom the 5-yr survival was 47% compared to
76% for patients with FSGS-CS.
Characteristic FSGS-CS FSGS-CELL P Value
Steroid therapy was provided in 56% (42 of 75) of nephrotic
n 57 43 patients. They received high-dose prednisone therapy (60
Presentation to follow-up 75 6 85 39 6 37 0.04 mg/d) for 2.34 6 1.49 mo, mean 6 SD (median 2.00), and the
(mo) total therapeutic course including the steroid taper was 8.07 6
median 44 26 10.01 mo, mean 6 SD (median 5.50 mo). Although a higher
Biopsy to follow-up (mo) 53 6 68 32 6 37 NS proportion of patients with FSGS-CELL received treatment,
median 35 18 the difference was not significantly different. In addition, high-
ESRD 10 (18%) 17 (40%) 0.02 dose prednisone therapy (1.99 6 1.20 mo [median 2.00] versus
2.60 6 1.65 [median 1.90]) and total therapeutic course
a
Results are given as mean 6 SD. (10.38 6 14.89 mo [median 5.60] versus 6.43 6 3.65 [median
5.30]) were not significantly different in patients with
FSGS-CS and FSGS-CELL. Only three patients with
the time of biopsy as clinical features predictive of progression FSGS-CS and four patients with FSGS-CELL received cyclo-
to ESRD. The cellular lesion in any number of glomeruli phosphamide as initial therapy in addition to steroids, and this
($1%) and interstitial expansion involving more than 20% of number was too small to analyze. More than 50% of treated
the biopsy were the histologic features that predicted ESRD. patients entered a remission, and there was no difference in
By multivariate analysis (Table 5), male gender and renal response to therapy between patients with FSGS-CELL and
insufficiency at biopsy remained significant clinical predictors FSGS-CS (53% and 52%). However, in the subset of patients
of progressive renal disease. The cellular lesion predicted with FSGS-CELL $20% (94% were black and 67% had .10
outcome, but only when combined with histologic evidence of g/24 h of proteinuria) only 23% of those treated entered a
chronic renal disease. Patients with the cellular lesion and remission (data not shown). Only a few untreated nephrotic
$20% interstitial fibrosis were at high risk (relative risk, 27.7) patients experienced spontaneous remissions in each group.
for progression compared to patients with ,20% interstitial The course for patients with both FSGS-CELL and FSGS-CS
fibrosis (relative risk, 2.0). who achieved a remission was excellent with 5-yr renal sur-
vivals of 100% (data not shown). This outcome was compara-
Nephrotic Patients ble to the 5-yr renal survival of 92% observed in the 25
Pathology Studies. Biopsies from nephrotic patients with non-nephrotic patients (four with the FSGS-CELL and 20 with
FSGS-CS had more glomeruli with classic segmental scars FSGS-CS). The prognosis for nephrotic patients who did not
than biopsies from nephrotic patients with FSGS-CELL (Table respond to therapy was grim (Figure 5), and this was especially
6), but the percentage of glomeruli with global sclerosis was true for patients with FSGS-CELL who had a 5-yr renal sur-
similar. Although the proportions of diseased glomeruli per vival of only 16% (versus 69% for patients with FSGS-CS).
biopsy (total of CS, global sclerosis, and CELL) were similar The renal survival was not significantly different between
in the two groups (44 6 24% in FSGS-CELL versus 36 6 24% nonresponsive patients with $20% and ,20% FSGS-CELL
in FSGS-CS, P 5 NS), the proportion of potentially progres- (median survival 21 versus 12 mo, P 5 NS). Univariate anal-
sive lesions (CS and CELL) appeared to be greater in biopsies ysis in nephrotic patients (Table 4) identified renal insuffi-
from patients with the cellular lesion. Eighteen (46%) biopsies ciency and the level of protein excretion at biopsy as clinical
with the cellular lesion had cellular lesions in more than 20% variables associated with an increased risk of progression to
of the glomeruli (39 6 16%, median 33%): The cortical ESRD. Race was not predictive of ESRD. Among the morpho-
interstitial volume was not significantly different between pa- logic features evaluated, global glomerular sclerosis, the cel-
tients with ,20% and $20% cellular lesions. lular lesion, and interstitial fibrosis $20% significantly in-
Clinical Characteristics at Biopsy. Nephrotic patients creased the relative risk of ESRD. In contrast to these findings,
with FSGS-CELL had significantly more proteinuria and hy- treatment and/or remission of proteinuria significantly de-
poalbuminemia than patients with FSGS-CS (Table 7). Even creased the risk of progression to ESRD.
though the serum creatinine levels at biopsy were not signifi- Multivariate analysis of nephrotic patients (Table 5) dem-
cantly different, the proportion of patients with renal insuffi- onstrated creatinine .1.3 mg/dl as the only independent clin-
ciency tended to be greater in patients with cellular lesions. ical predictor of progression to ESRD. Interstitial fibrosis
Among nephrotic patients, black race was no longer predom- $20% and the cellular lesion were the only histologic features
inant in patients with FSGS-CELL. independently predictive of ESRD in nephrotic patients. A
Outcome. The duration of clinical follow-up tended to be significant decrease in the risk of ESRD was associated with
longer in nephrotic patients with FSGS-CS but was not signif- steroid treatment. The risk of ESRD was substantially less in
icantly different from patients with FSGS-CELL (Table 8). patients achieving a remission (relative risk 0.03, P 5 0.001),
Nonetheless, the prevalence of ESRD in patients with FSGS- but this could only be demonstrated when the serum creatinine
CELL was more than three times greater than in patients with was omitted from the multivariate model.
FSGS-CS. The overall renal survival (Figure 4) was worse for Multivariate logistic regression analysis (data not shown) of
nephrotic patients with FSGS-CELL (median survival 36 mo, clinical and histologic factors that might predict a remission in
1904 Journal of the American Society of Nephrology J Am Soc Nephrol 10: 1900 –1907, 1999

Figure 3. Cumulative renal survival for all patients with focal segmental glomerulosclerosis (FSGS) with classic segmental scars (FSGS-CS)
and the cellular lesion of FSGS (FSGS-CELL). P , 0.001 by the log-rank test. Numbers in the table represent patients at risk at each time point.

Table 4. Univariate analysis of predictors of ESRDa Table 5. Multivariate analysis of ESRD in FSGS
Nephrotic Relative
All Patients Participants and Covariates P Value
Patients Risk
Covariate
Relative P Relative P All Patients
Risk Value Risk Value creatinine .1.3 mg/dl 10.7 0.001
Black race 1.99 0.083 2.09 0.114 cell $1%
Male 2.76 0.029 2.44 0.072 interstitial volume ,20% 2.00 0.008
Age (yr) 1.07 0.264 1.12 0.110 interstitial volume $20% 27.7
Hypertension 1.01 0.985 1.10 0.818 male 3.63 0.023
Creatinine .1.3 mg/dl 7.36 0.001 8.33 0.004 Nephrotic patients only
Proteinuria g/24 h 1.05 0.006 1.05 0.030 creatinine .1.3 mg/dl 12.8 0.004
CS $10% 0.42 0.031 0.37 0.032 CELL $1% 6.69 0.001
Global sclerosis $10% 1.73 0.184 2.78 0.024 interstitial fibrosis $20% 6.57 0.001
CELL $1% 3.91 0.001 4.10 0.006 steroid treatment 0.36 0.032
Interstitial volume $20% 3.39 0.004 4.41 0.001 remissiona 0.03 0.001
Steroid treatment 0.43 0.049 a
In nephrotic patients, remission and creatinine could not be
Remission 0.06 0.006 used in the same model (i.e., nonconvergent). When creatinine was
a
not included in the model, remission was highly significant.
ESRD, end-stage renal disease.

nephrotic patients found steroid treatment as the only covariate patients because nephrotic-range proteinuria is an independent
independently predictive of a remission (relative risk 10.6, P 5 predictor of outcome in FSGS (10) and because non-nephrotic
0.001). Factors predictive of decreased relative risk of remis- patients with FSGS are rarely treated. We confirmed that renal
sion were male gender (relative risk 0.23, P 5 0.012) and survival is excellent in nephrotic patients with FSGS who
$20% cellular lesions (relative risk 0.03, P 5 0.002). achieve a remission of proteinuria (reviewed in reference (17).
In addition, the present study further supports a role for therapy
Discussion by showing that remissions associated with steroid treatment
We describe two classes of patients with primary FSGS that are equally likely in nephrotic patients with the cellular lesion
are distinguished by the presence or absence of the cellular and those with only classic segmental scars. Thus, a therapeutic
lesion and differences in clinical features and prognosis. The response favorably affects outcome in primary FSGS patients
levels of protein excretion at presentation and progression to with nephrotic-range proteinuria independent of the presence
ESRD were greater in patients with any amount of glomerular of the cellular lesion.
involvement by the cellular lesion, but we focused on nephrotic The cellular lesion is a descriptive diagnosis that encom-
J Am Soc Nephrol 10: 1900 –1907, 1999 Focal Segmental Glomerulosclerosis 1905

Table 6. Pathology studies: nephrotic patientsa Table 8. Outcome: nephrotic patientsa

Characteristic FSGS-CS FSGS-CELL P Value Characteristic FSGS-CS FSGS-CELL P Value

n 36 39 n 36 39
Total glomeruli 21 6 14 29 6 21 0.03 Presentation to follow-up 65 6 80 39 6 39 NS
median 18 24 (mo)
CS (%) 16 6 9 466 ,0.0001 median 39 26
median 15.5 0 Biopsy to follow-up (mo) 52 6 78 34 6 38 NS
Global sclerosis (%) 19 6 23 16 6 21 NS median 25 18
median 10 4 ESRD 5 (14%) 17 (44%) 0.005
Segmental CELL (%) 0 14 6 12 Spontaneous remission 1 (3%) 2 (5%) NS
median 10 Patients treated 17 (47%) 25 (64%) NS
Global CELL (%) 0 10 6 11 Remission 9 (53%)b 13 (52%) NS
median 6 complete 6 (35%) 8 (32%) NS
Total CELL (%) 0 24 6 18 partial 3 (18%) 5 (20%) NS
median 17 a
Results are given as mean 6 SD.
Interstitial volume (%) 20 6 13 22 6 14 NS b
Absolute number (percentage of treated).
median 17 19
a
Results are given as mean 6 SD.
prognoses. Patients with FSGS-CELL $20% are almost ex-
a clusively black, have more severe proteinuria, and are less
Table 7. Characteristics at biopsy: nephrotic patients
responsive to treatment than either patients with FSGS-CELL
Characteristic FSGS-CS FSGS-CELL P Value ,20% or FSGS-CS, and the increased prevalence of ESRD
observed in patients with FSGS-CELL can be attributed to the
n 36 39 adverse natural history of the patients with FSGS-CELL
Black 23 (64%) 28 (72%) NS $20%. Thus, the cellular lesion, when it is widely distributed
Male 22 (61%) 23 (59%) NS among the glomeruli in a renal biopsy, has adverse implica-
Age (yr) 39 6 15 38 6 17 NS tions for prognosis and therapeutic response. Because sponta-
median 38 36 neous remission of FSGS is unusual (2), it is likely that the
Hypertension 17 (47%) 20 (51%) NS remissions seen in the treated patients are induced by therapy,
Creatinine (mg/dl) 1.7 6 0.8 2.5 6 2.1 NS although this has never been tested, prospectively. The physi-
median 1.4 1.7 cian should not fall into the “slough of despond” when con-
Renal insufficiency 18 (50%) 28 (72%) 0.06 fronted by the cellular lesion in the renal biopsy of a nephrotic
Proteinuria (g/24 h) 6.6 6 3.4 13.2 6 9.7 0.001 patient. Although pessimism is justified when the cellular
median 5.7 8.4 lesion is widespread, the present experience demonstrates that
Proteinuria .10 g/24 h 6 (17%) 19 (49%) 0.003 a majority of patients with more focal cellular lesions (FSGS-
Albumin (g/dl) 3.1 6 1.0 2.6 6 0.7 0.008 CELL #20%) and even some patients with extensive glomer-
median 3.2 2.5 ular involvement by cellular lesions (FSGS-CELL .20%) will
Presentation to biopsy 13 6 28 5 6 10 NS respond to treatment as well as patients with FSGS-CS.
(mo) Clinical observations suggest that glomerular visceral epi-
median 3 1.3 thelial cell injury is the initiating event in the pathogenesis of
a
Results are given as mean 6 SD. primary FSGS. The most convincing evidence supporting the
evolution from the cellular lesion to glomerular scarring is seen
in serial biopsies from renal transplant recipients who develop
passes the proliferative/infiltrative cellular changes seen in recurrent FSGS: Early onset of massive proteinuria coincides
Bowman’s space in some renal biopsies that fall within the with minimal light microscopic changes, while ultrastructural
morphologic and clinical spectrum of FSGS (5), and in our pathology is limited to foot process effacement. With time the
opinion, the cellular lesion is identical to the descriptions and cellular lesion develops, but classic segmental scars are only
illustrations of collapsing glomerulopathy (6). In this study, the seen in renal biopsies taken late in the clinical course (18).
cellular lesion predicted the outcome of ESRD, a result that is Additionally, the cellular lesion is frequently present in renal
in accord with the findings of Detwiler et al. (7) and Valeri et biopsies taken from patients near to the onset of the nephrotic
al. (8) in collapsing glomerulopathy. We extended these ob- syndrome (5,8). In contrast, classic segmental scars dominate
servations by stratifying patients with the cellular lesion ac- the pathology later in the course suggesting that the cellular
cording to the proportion of glomerular involvement and dem- lesion heals with the formation of a classic segmental scar.
onstrated that patients with biopsies showing ,20% and The relationship between the extent of the cellular lesion and
$20% cellular lesions have different clinical features and ESRD demonstrates that involvement of a larger proportion of
1906 Journal of the American Society of Nephrology J Am Soc Nephrol 10: 1900 –1907, 1999

Figure 4. Cumulative renal survival for all nephrotic patients with FSGS-CS and FSGS-CELL. P , 0.0001 by the log-rank test. Numbers in
the table represent patients at risk at each time point.

Figure 5. Cumulative renal survival for nephrotic patients with FSGS-CS and FSGS-CELL who did not achieve a remission. P , 0.0001 by
the log-rank test. Numbers in the table represent patients at risk at each time point.

glomeruli ($20%) is usually associated with a therapeutically distributions of the cellular lesion represent discrete diseases
unresponsive and progressive form of FSGS. Because patients with different etiologies.
with FSGS-CELL $20% frequently have other glomeruli with The cellular lesion is presumed to arise from glomerular
either global sclerosis or classic segmental scars, the presence visceral epithelial cells, but the markers that characterize the
of the cellular lesion in many other glomeruli indicates the mature epithelial cell are not present in the cellular lesion
potential to develop further scarring and progression to renal (19,20). Therefore, the presumptive origin of the cellular lesion
insufficiency. In contrast, FSGS is less aggressive in patients is based on its morphology, including the location in the
with FSGS-CELL ,20% and FSGS-CS, and irreversible dam- urinary space outside the glomerular basement membrane and
age is usually restricted to diseased glomeruli. These observa- transitions observed between glomerular visceral epithelial
tions suggest that there is a critical proportion of glomerular cells and the cellular lesion (20). Proliferation is the definitive
involvement by the cellular lesion that favors progressive glo- morphologic feature of the cellular lesion, as demonstrated by
merular disease. However, it is also possible that different increased cells in Bowman’s space, the presence of mitotic
J Am Soc Nephrol 10: 1900 –1907, 1999 Focal Segmental Glomerulosclerosis 1907

figures (5), and the cell cycle marker, Ki-67 (20), but the sclerosis: Pathology, histological variants, and pathogenesis.
degree of proliferation is difficult to reconcile with the limited [Review]. Am J Kidney Dis 22: 874 – 883, 1993
regenerative capacity of mature glomerular visceral epithelial 5. Schwartz MM, Lewis EJ: Focal segmental glomerular sclerosis:
cells (21). The mechanism of this anomalous proliferation may The cellular lesion. Kidney Int 28: 968 –974, 1985
be related to the demonstrated loss of WT1 protein from the 6. Weiss MA, Daquioag E, Margolin EG, Pollak VE: Nephrotic
syndrome, progressive irreversible renal failure, and glomerular
glomerular visceral epithelial cells (20). WT1 is a zinc finger,
“collapse”: A new clinicopathologic entity? Am J Kidney Dis 7:
tumor suppressor factor normally expressed in mature podo-
20 –28, 1986
cytes, and it is critical in normal renal development. Loss of 7. Detwiler RK, Falk RJ, Hogan SL, Jennette JC: Collapsing glo-
WT1 may allow the expression of factors (21,22) that result in merulopathy: A clinically and pathologically distinct variant of
proliferation. The pathogenesis of glomerular scarring follow- focal segmental glomerulosclerosis. Kidney Int 45: 1416 –1424,
ing loss of WT1 is unknown, but it is noteworthy that abnor- 1994
malities in the WT1 gene have been demonstrated in several 8. Valeri A, Barisoni L, Appel GB, Seigle R, D’Agati V: Idiopathic
forms of the congenital nephrotic syndrome (22), suggesting collapsing focal segmental glomerulosclerosis: A clinicopatho-
that loss of WT1 expression leads to progressive glomerular logic study. Kidney Int 50: 1734 –1746, 1996
sclerosis in these inherited diseases. It is not known whether 9. Schwartz MM, Korbet SM, Rydell J, Borok R, Genchi R: Pri-
WT1 expression is lost in limited forms of the cellular lesion mary focal segmental glomerular sclerosis in adults: Prognostic
with a favorable response to treatment and a relatively good value of histologic variants. Am J Kidney Dis 25: 845– 852, 1995
prognosis. However, widespread loss of WT1 expression is 10. Korbet SM, Schwartz MM, Lewis EJ: Primary focal segmental
glomerulosclerosis: Clinical course and response to therapy.
associated with a poor renal prognosis whether the loss is the
Am J Kidney Dis 23: 773–783, 1994
result of a point mutation in the Denys-Drash syndrome (22),
11. Korbet SM, Schwartz MM, Lewis EJ: The prognosis of focal
a viral infection in HIV-associated nephropathy, or unknown segmental glomerular sclerosis of adulthood. Medicine (Balti-
causes in collapsing glomerulopathy (20). more) 65: 304 –311, 1986
The pathologic finding of the cellular lesion in a nephrotic 12. Agresti A: Categorical Data Analysis, New York, John Wiley &
patient with primary FSGS portends an adverse outcome, but Sons, 1990
in our experience these patients are as likely to have a thera- 13. Conover WJ: Practical Nonparametric Statistics, New York,
peutic response as nephrotic patients with FSGS-CS. In con- John Wiley & Sons, 1980
trast, extensive glomerular involvement by the cellular lesion 14. Kalbfleisch JD, Prentice RL: The Statistical Analysis of Failure
portends a poor prognosis that is due to the adverse outcome Time Data, New York, John Wiley & Sons, 1980
seen in patients with the FSGS-CELL .20%. Given the lack of 15. Kleinbaum DG, Kupper LL, Morgenstem H: Epidemiological
response of the widely distributed form of the cellular lesion Research: Principles and Quantitative Methods, Belmont, Cali-
(.20%) to current therapy, we recommend that it be included fornia, Lifetime Learning, 1982
in the stratification of patients with FSGS for therapeutic trials. 16. Venkatachalam MA, Kriz W: Anatomy. In: Heptinstall’s Pathol-
ogy of the Kidney, 5th Ed., edited by Jennette JC, Olson JL,
Despite the negative prognostic connotation of the cellular
Schwartz MM, Silva FG, Philadelphia, Lippincott-Raven, 1998,
lesion, the favorable outcome associated with a remission of
pp 3– 66
proteinuria strongly supports therapeutic intervention in ne- 17. Korbet SM: Primary focal segmental glomerulosclerosis. J Am
phrotic patients with primary FSGS regardless of the presence Soc Nephrol 9: 1333–1340, 1998
of the cellular lesion. 18. Korbet SM, Schwartz MM, Lewis EJ: Recurrent nephrotic syn-
drome in renal allografts. Am J Kidney Dis 11: 270 –276, 1988
Acknowledgment 19. Bariety J, Mandet C, Meyrier A: Podocyte transdifferentiation
We thank Dr. Edmund J. Lewis for his critical review of the into macrophage-like cells in cellular variants of human FSG
manuscript. [Abstract]. J Am Soc Nephrol 9: 532A, 1998
20. Barisoni L, Kriz W, Mundel P, D’Agati V: The dysregulated
References podocyte phenotype: A novel concept in the pathogenesis of
1. Churg J, Habib R, White RH: Pathology of the nephrotic syn- collapsing idiopathic focal segmental glomerulosclerosis and
drome in children: A report for the International Study of Kidney HIV-associated nephropathy. J Am Soc Nephrol 10: 51– 61, 1999
Disease in Children. Lancet 760: 1299 –1302, 1970 21. Combs HL, Shankland SJ, Setzer SV, Hudkins KL, Alpers C:
2. Rydel JJ, Korbet SM, Borok RZ, Schwartz MM: Focal segmental Expression of the cyclin kinase inhibitor p27kip1 in developing
glomerular sclerosis in adults: Presentation, course, and response and mature human kidney. Kidney Int 53: 892– 896, 1998
to treatment. Am J Kidney Dis 25: 534 –542, 1995 22. Yang Y, Jeanpierre C, Dressler GR, Lacoste M, Neaudet P,
3. D’Agati V: The many masks of focal segmental glomeruloscle- Gubler M-C: WT1 and PAX-2 podocyte expression in Denys-
rosis. Kidney Int 46: 1223–1241, 1994 Drash syndrome and isolated diffuse mesangial sclerosis. Am J
4. Schwartz MM, Korbet SM: Primary focal segmental glomerulo- Pathol 154: 181–192, 1999

This article can be accessed in its entirety on the Internet at

http://www.lww.com/JASN along with related UpToDate topics.