Study of the morphologic variants of focal segmental glomerulosclerosis : a Brazilian report

1. Doutor em Ciências; médico da Divisão de Anatomia Patológica do Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo (DAP-HC-FMUSP). 2. Doutora em Ciências; professora da FMUSP. Introduction: Focal segmental glomerulosclerosis (FSGS) is the most frequent primary glomerulopathy in Brazil and its incidence is increasing worldwide. Pathogenesis is related to podocyte injury, which may be due to several factors including viruses, drugs, genetics and immunological factors. In 2004, the Columbia classification of FSGS identified five histological variants of the disease: collapsing (COL), usual (NOS), tip lesion (TIP), perihilar (PHI) and cellular variant (CEL). The objective of this study was to classify the FSGS biopsies in these morphological variants. Methods: One hundred thirty-one cases of renal biopsies with primary FSGS diagnosis, which had been performed at a Brazilian reference center from 1996 to 2006, were classified according to the Columbia criteria. Results: FSGS cases were distributed as follows: 38.2% NOS variant, 36.6% COL, 14.5% TIP, 6.9% PHI and 3.8% CEL. Conclusion: COL variant of FSGS seems to be more prevalent in Brazil in comparison with other centers worldwide, which may be related to environmental and socioeconomic factors.. abstract key words Focal segmental glomerulosclerosis

Introduction: Focal segmental glomerulosclerosis (FSGS) is the most frequent primary glomerulopathy in Brazil and its incidence is increasing worldwide.Pathogenesis is related to podocyte injury, which may be due to several factors including viruses, drugs, genetics and immunological factors.In 2004, the Columbia classification of FSGS identified five histological variants of the disease: collapsing (COL), usual (NOS), tip lesion (TIP), perihilar (PHI) and cellular variant (CEL).The objective of this study was to classify the FSGS biopsies in these morphological variants.Methods: One hundred thirty-one cases of renal biopsies with primary FSGS diagnosis, which had been performed at a Brazilian reference center from 1996 to 2006, were classified according to the Columbia criteria.Results: FSGS cases were distributed as follows: 38.2% NOS variant, 36.6%COL, 14.5% TIP, 6.9% PHI and 3.8% CEL.Conclusion: COL variant of FSGS seems to be more prevalent in Brazil in comparison with other centers worldwide, which may be related to environmental and socioeconomic factors..

Introduction
Focal segmental glomerulosclerosis (FSGS) is a clinical and pathological renal syndrome causing proteinuria, usually in the nephrotic range, and occurring predominantly in young individuals.According to several authors, the worldwide incidence of FSGS is on the rise annually, including in Brazilian statistics (1,8,9) .Data from the Paulista registry of glomerulopathies shows that FSGS is the most common primary glomerulopathy in 11 centers of São Paulo, accounting for 29.7% of cases, followed by nephropathy (20.7%) and immunoglobulin A (IgA) nephropathy (17.8%) (12) .From an etiological standpoint, FSGS may be associated with a variety of conditions, including viral infections (human immunodeficiency virus [HIV], simian virus 40 [SV40] , parvovirus B19), several drugs (heroin, pamidronate, interferon), adaptive to hyperfiltration (as in hypertension, obesity, and sickle-cell anemia), and genetic disorders (familiar forms), and may be idiopathic (11) .Podocyte injury is the core pathophysiological event of FSGS and some studies have found anomalous expression of podocyte proteins in FSGS (2,10,16) .
The Columbia classification for primary FSGS, first published by D'Agati et al., in 2004, proposed five mutually exclusive morphological variants of the disease (histologic criteria shown in Table and Figure).Studies of these variants demonstrated correlation with distinct clinical characteristics and prognostic and therapeutic implications (4,18) .Recognition of these variants through a combination of morphologic and molecular features might bring about a better understanding of the pathogenesis of FSGS and promote identification of new therapeutic targets.The present study aimed to classify cases of primary FSGS diagnosed at the Divisão de Anatomia Patológica do Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo (DAP-HC-FMUSP) into the collapsing (COL), cellular (CEL), usual (NOS), tip lesion (TIP), and perihilar (PHI) morphologic variants.

Methods of case selection
We obtained 525 consecutive biopsies with FSGS diagnosis from the archives of the DAP-HC-FMUSP from 1996 to 2006, of which we selected 131 cases based in the following exclusion criteria: • secondary FSGS due to known etiology (glomerular sclerosis scarring, chronic pyelonephritis, hypertension, reflux nephropathy); • age below 18 and over 45 years-to guarantee, as good as possible, that we were working on primary FSGS, once we did not have access to the patient charts; • less than five glomeruli in a sample to minimize the possibility of include focal segmental glomerulosclerosis secondary to diabetic glomerulopathy, chronic glomerulonephritis etc.; • recurrence in kidney allografts; • biopsies from HIV seropositive patients; • advanced stage of histologic chronicity (more than 90% of sclerotic glomeruli); • biopsies without or with unsatisfactory immunofluorescence; • biopsies with positive immunofluorescence for IgA and/or IgG in glomeruli.
Slides and blocks of each case were separated and reviewed.

Results
After application of the exclusion criteria, a total of 131 eligible cases remained and were assessed according to the histologic criteria proposed by the FSGS classification, which revealed 50 cases of NOS; 48, COL; five, CEL; 19, TIP; and nine, PHI (Table and Figure).The most prevalent discrepancy between the authors regarding the Columbia classification of FSGS was related to the CEL and COL variants, since there were cases previously classified as COL that were reclassified as NOS or CEL variant.

Discussion of results
Increases in the number of patients with chronic kidney disease (CKD) have led to growing interest in a better understanding of the renal conditions involved in CKD, including FSGS.Improved knowledge of pathogenic mechanisms would allow development of individualized and more efficient therapeutic actions.The recognition of histologic variants can be the first step in this direction in the case of primary FSGS, this has translated into recent definition of the five variant forms of the disease, namely, usual (NOS), collapsing (COL), cellular (CEL), perihilar (PHI), and tip lesion (TIP) variants.It is now known that certain histologic features in FSGS are associated with poor clinical progression, including rapid deterioration of renal function and marked proteinuria (3,14,18) .The present series included 131 cases of primary FSGS diagnosed between 1996 and 2006, which were classified according to the Columbia classification criteria (4) as 50 NOS (38.2%),48 COL (36.6%), 19 TIP (14.5%), nine PHI (6.9%), and five CEL (3.8%) (Table ).Thomas et al. (7) published a series of 197 cases recorded at the University of North Carolina between 1982 and 2001 and reported that 42% of cases were of the NOS variant, 17% were TIP, 13% were COL, 26% were PHI, and 3% were CEL (18) (Table ).However, we think the authors were less strict in selecting cases of primary FSGS than us.Degeens et al. (5) observed 32% of NOS, 37% of TIP, 26% of PHI, 5% of COL and no CEL variant in an adult Dutch population (5) .Our data revealed a higher prevalence of the COL variant, diverging from the above cited case series.These data raise the vision of a possible environmental or infectious factor.Given that it is well known the bigger incidence in black race and close resemblance to HIV-associated nephropathy, a search for possible viral etiology and socioeconomic influences appears justified.Merlet-Bénichou et al. has found reduced numbers of glomeruli and larger glomerular volumes in newborn rats exposed to toxins in utero and in infants born with small placentas and low birth weights; these observations suggest that socioeconomic differences, such as poor nutrition or exposure to toxins during gestation, might affect the developing kidney and result in a reduced glomerular population with larger glomerular volume (6,13) .
In the present study, lesions were most easily classified into the TIP, NOS, or COL variants using the Columbia histological criteria; there was substantial overlap of criteria for the NOS and PHI variants, as well as with COL and CEL variants.Some authors question the existence of a separate CEL variant, and claim it is merely a form of the COL variant (17) .Others agree that both variants are very difficult to distinguish histologically and common pathophysiological pathways affecting cell cycle regulatory proteins have been established (3,14,15) .In our sample, there was also overlap of histologic findings for both variants.
In cases classified as COL, we found the coexistence of marked tubulointerstitial injury, sometimes with microcystic tubular dilatation, however similar findings were also present in other variants, particularly in advanced chronic disease, although these changes were more marked in COL variant.Therefore, these findings should not be considered specific to COL variant FSGS.
The distinction between the NOS and PHI variants also seemed unclear when using the current criteria.We believe the criterion of > 50% perihilar involvement varies according to the level of histological section.According to the proposed criteria, the diagnosis of PHI variant FSGS should only be considered when over 50% of sclerotic glomeruli show perihilar lesions, which may vary from one histologic section to another of a single biopsy.Several cases classified as NOS exhibited perihilar sclerosis and hyalinosis in additional sections which highlights the importance of having a representative sample in order to distinguish between the NOS and PHI variants.Stokes et al. did not include cases of the PHI variant in their series, as they believed this variant often refers to secondary FSGS, however, they found evidence of perihilar lesions in other variants mostly in NOS (17) .Maybe they found the same difficulty in the PHI versus NOS distinction.
In our sample, the TIP variant was particularly difficult to diagnose when the apical lesion was represented marginally, with no apparent relation to the origin of the proximal convoluted tubule, and, in some cases, coexisted with morphologic aspects similar to those of COL FSGS.These lesions sometimes presenting confluence of hypertrophic podocytes, parietal cells, and tubular epithelial cells, deformed the glomerular tuft, giving it an aspect alike to a collapsing injury.A search for classical lesions in contact with the origin of the tubule in additional histologic sections and the absence of podocyte proliferation can assist in differential diagnosis (7) .and 3.8% CEL, according to Columbia classification, which points to a bigger prevalence of the COL variant than in other international centers.Additional studies to better explain the socioeconomic role on these prevalences might be of value on this field.