NDT Advance Access published online on April 18, 2007
Nephrology Dialysis Transplantation, doi:10.1093/ndt/gfm188
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Focal segmental glomerulosclerosis is not a sufficient predictor of renal outcome in patients with membranous nephropathy
1Department of Nephrology and 2Department of Pathology, Radboud University Nijmegen Medical Centre, The Netherlands
Correspondence and offprint requests to: S. F. Heeringa, MD, Department of Nephrology 464, Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands. Email: sf.heeringa{at}gmail.com
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Background. The course of idiopathic membranous nephropathy (iMN) is variable in untreated patients. Accurate prediction of renal outcome would allow optimal treatment decisions. We demonstrated that urinary ß2-microglobulin (ß2M) predicted prognosis in iMN with high sensitivity and specificity. It has been suggested that focal segmental glomerulosclerosis (FSGS) is a discriminative parameter with independent prognostic value.
Methods. We selected patients with iMN biopsied between 1988 and 2002. Biopsies were analysed for the presence of FSGS, interstitial fibrosis and vascular lesions. Serum creatinine, creatinine clearance, proteinuria and blood pressure were recorded at baseline. Outcome variables included remission of proteinuria, renal death (RD) defined as serum creatinine >135 µmol/l or increase of serum creatinine of >50%, or end-stage renal disease (ESRD). In a subgroup of patients, urinary ß2-microglobulin (ß2M) was measured.
Results. We included 53 patients (33M, 20F). Mean age was 51 years, serum creatinine 99 µmol/l, and proteinuria 7.0 g/10 mmol creatinine. FSGS was present in 22 patients. These patients were characterized by a higher serum creatinine at time of biopsy (P = 0.035), more severe interstitial fibrosis (P = 0.001) and higher stage of membranous nephropathy (P = 0.001). During follow-up 24 patients developed RD, almost equally distributed between patients with and without FSGS. Renal survival was numerically, but not significantly, lower in patients with FSGS. In Cox proportional hazard analysis, only serum creatinine at the time of biopsy was an independent predictor of RD or ESRD (P < 0.001). In patients with known urinary ß2M, there was no significant correlation with FSGS score (P = 0.174).
Conclusion. FSGS is not an accurate prognostic marker in iMN. Histological scoring of FSGS is inferior to measurement of urinary proteins in predicting renal outcome in iMN.
Keywords: focal segmental glomerulosclerosis (FSGS); idiopathic membranous nephropathy (iMN); prognosis
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Idiopathic membranous nephropathy (iMN) is one of the most common causes of the nephrotic syndrome in adults [1]. The course of iMN in untreated patients is variable, as approximately 34–62% develop renal insufficiency and 14–56% will have a spontaneous remission [2]. The identification of parameters that predict prognosis with sufficient accuracy would allow individualized treatment and avoid unnecessary immunosuppressive therapy. Well-known risk factors for deterioration of renal function such as proteinuria, age and sex lack sensitivity or specificity [3–5]. In previous studies, we have found that urinary ß2-microglobulin (ß2M) accurately predicts prognosis in patients with iMN [6–8]. Recent studies have provided evidence that morphological parameters—specifically focal segmental glomerulosclerosis (FSGS)—may be used as prognostic parameter [9–12]. In the present study, we have evaluated FSGS as prognostic parameter in patients with iMN. In addition, we performed a meta-analysis of studies analysing the prognostic significance of FSGS in iMN.
|
Subjects and methods |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Patient selection
We included patients with a diagnosis of membranous nephropathy, biopsied between 1988 and 2002 in any of five nephrology units in the Nijmegen region. Only patients with iMN were considered in this study.
Clinical data
From selected patients, their age, sex, serum creatinine, serum cholesterol, serum albumin, proteinuria (expressed per 10 mmol creatinine), creatinine clearance (ECC) calculated according to the Cockcroft Gault formula [13], systolic and diastolic blood pressure and medication had been recorded at the time of biopsy (Table 1), and were frequently measured during follow-up. The mean arterial pressure was calculated as the diastolic blood pressure plus a third of the pulse pressure.
|
In patients who were seen at the University Hospital, additional standardized measurements were done as described to determine urinary excretion of ß2-microglobulin [8].
Renal death was defined as a serum creatinine concentration of >135 µmol/l or an increase of serum creatinine of more than 50% over time. Patients who met these criteria were advised to start with immunosuppressive therapy. As described before, start of immunosuppressive therapy, because of severe nephrotic syndrome, was also considered as renal death (RD) [8].
End-stage renal disease (ESRD) was defined as an ECC <10 ml/min. A complete remission was defined as proteinuria below 0.2 g/10 mmol creatinine and a partial remission as proteinuria between 0.2 and 2.0 g/10 mmol creatinine, without deterioration of renal function.
Follow-up started at the time of biopsy and continued until December 2003 or ended at the time of death or development of ESRD.
Pathology
All biopsy specimens were processed for light microscopy using standardized techniques. For the evaluation and quantitation of FSGS lesions we have used the biopsy slides stained with methenamine silver in which FSGS is best recognized. Biopsies with less than five glomeruli were excluded from the analysis.
The presence of FSGS was evaluated by a qualitative and a quantitative analysis. In the qualitative analysis, all available slides were screened for the presence of FSGS and tubulo-interstitial fibrosis. The risk of overlooking FSGS was thereby considered negligible. In the quantitative analysis, glomerular slides were selected at random and evaluated in detail for the extent of FSGS.
Focal glomerulosclerosis was defined as a focal lesion with mesangial matrix expansion leading to collapse of the glomerular capillary loops. The extent of sclerosis was scored by counting the number of glomerular quadrants with sclerosis. FSGS was expressed as a percentage by dividing the total score of sclerosis by four times the number of evaluated glomeruli. Globally sclerosed glomeruli were excluded in this analysis, but were included in the calculation of total sclerosis. Total sclerosis was defined as the sum of focal sclerosed glomerular quadrants and globally sclerosed glomeruli x 4 divided by the total number of glomeruli x 4.
The interstitium was evaluated for the presence and extent of tubulo-interstitial fibrosis, intimal fibrosis and foam cells. Tubulo-interstitial fibrosis was scored semiquantitatively on a scale of 0–3.
The glomerular deposits were evaluated by electron-microscopy (EM) and the stage of membranous nephropathy was taken from the corresponding pathology report.
Meta-analysis
We performed a thorough search of the MEDLINE, EMBASE and Cochrane Controlled Trial Register databases. The following keywords were used: ‘membranous nephropathy’, ‘membranous glomerulopathy’, ‘FSGS’, ‘outcome’, ‘remission’. The search was limited to studies published in English. The following types of studies were included: randomized controlled trials, non-randomized prospective trials and retrospective studies. Case reports were not included. Only studies analysing the outcome in patients with iMN according to the presence or absence of FSGS were considered. The primary outcome was remission of proteinuria.
Statistical analysis
Statistical analysis was performed using the SPSS version 11.0 for Windows. For differences between groups, Chi-square test, t-test and the Mann–Whitney U test were applied, if appropriate. Correlations were calculated using Spearman's correlation test.
The Kaplan–Meier univariate analysis was used to select variables associated with progression to RD. Significance was determined by the log–rank test. In order to identify independent predictive parameters, Cox proportional hazard was performed in forward stepwise fashion on variables selected by univariate analysis. For this purpose, non-parametric variables were transformed using log 10 or square root transformation. For all analyses, P < 0.05 was considered significant.
All statistical calculations for the meta-analysis were performed using the computer software Review Manager (RevMan) version 4.2.9 provided by the Cochrane Collaboration. A pooled estimate of the odds ratios with 95% confidence intervals was calculated using the random-effects model using the DerSimonian–Laird method. A P-value <0.05 was considered significant. Heterogeneity was evaluated by using the 
2 test. Given the low power of this test, a P-value <0.1 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Clinical and pathological findings
We identified 61 eligible patients. In eight patients, the biopsy contained an insufficient number of glomeruli. Thus, we have studied 53 patients (33 M, 20 F) with a mean (±SD) age of 51 ± 15 years. Mean follow-up was 62 ± 43 months. Mean creatinine was 99 µmol/l and mean proteinuria 7 g/10 mmol. Baseline characteristics of all patients are given in Table 1.
FSGS was present in 22 patients (41.5%). The clinical and pathological findings at baseline of patients without FSGS lesions (n = 31, group I) and patients with FSGS lesions (n = 22, group II) are listed in Table 2. Patients with FSGS had a significantly higher serum creatinine and were numerically older. FSGS was significantly associated with a higher MN stage and more severe tubulo-interstitial lesions (Table 2). Vascular changes were also more marked in patients with FSGS, however, the differences were not statistically significant.
|
For the quantitative analysis of FSGS, we evaluated an average of 336 glomerular profiles per biopsy in group I (min 62, max 830) and 342 glomerular profiles per biopsy in group II (min 63, max 811). A mean of 2.3% of glomerular quadrants was sclerosed in group II (min 0.11, max 13.3%). There were significantly more obsolescent glomeruli in group II (8.1%) than in group I (4.9%), P = 0.034. Serum creatinine at time of biopsy, ECC and serum cholesterol were significantly correlated with the percentage of FSGS (serum creatinine r = 0.339; P = 0.013; ECC r = –0.333; P = 0.031; serum cholesterol r = –0.308; P = 0.039).
Renal outcome
At the time of biopsy, serum creatinine was >135 µmol/l in four patients. Another 20 patients reached the defined end point RD during follow-up. The reason for RD was a serum creatinine >135 µmol/l in 22 patients, a rise of >50% of serum creatinine in one patient and the start of immunosuppressive therapy with normal serum creatinine in one patient.
The outcome was not different between the groups, with development of RD in 41.9% of patients without and 50% of patients with FSGS (Table 3). Renal survival is depicted in Figure 1. Only three patients died, two with FSGS lesions (ns). During follow-up, slightly more patients without FSGS had a complete or partial remission of proteinuria (67.7% vs 50% in group II). Also at the end of follow-up a higher proportion of 20 patients without FSGS (67.7% vs 45.5%) had a complete or partial remission. These differences were not statistically significant. A complete remission at the end of follow-up was attained in 11/31 patients without FSGS and in 4/22 patients with FSGS (ns).
|
|
20 patients received immunosuppressive treatment during follow-up; 13 patients without FSGS, seven patients with FSGS (ns).
Risk factors for RD
To determine which variables were associated with the development of RD, univariate analysis was performed for all of the clinical variables (Table 4). The most discriminate variable was serum creatinine at time of biopsy (P < 0.001, log-rank test). Other variables associated with progression to RD were ECC (P < 0.001), diastolic blood pressure (P < 0.001), mean arterial pressure (P = 0.012), proteinuria (P = 0.004) and serum albumin (P = 0.035). Notably, FSGS was not associated with progression to RD (P = 0.46).
|
Cox proportional hazard analysis was performed using variables selected by univariate analysis. Serum creatinine, at time of biopsy, was the strongest independent predictive factor for progression to RD (P < 0.001). Diastolic blood pressure also had independent significant prognostic value for progression to RD (P = 0.050).
Our analysis could have been biased because at the time of biopsy several patients already had decreased renal function. A subgroup analysis was, therefore, performed limited to patients with a serum creatinine at the time of biopsy of <106 µmol/l. This subgroup consisted of 39 patients, 25 without and 14 with FSGS. During follow-up, 11 patients developed RD, 7/25 without FSGS and 4/14 with FSGS (ns). In univariate analysis, serum albumin (P = 0.0061), diastolic blood pressure (P = 0.007) and mean arterial pressure (P = 0.023) were significantly related to RD. Also in this subgroup, FSGS was not related to RD. Multivariate analysis revealed that serum albumin (P = 0.005) and diastolic blood pressure (P = 0.008) both had significant prognostic value for progression to RD.
In 21 patients, urinary ß2-microglobulin excretion was determined. FSGS was present in eight of 21 patients. Although urinary ß2-microglobulin (uß2M) was higher in patients with FSGS (uß2M: 3865.3 vs 1041 ng/min), the difference was not statistically significant. In a previous study, we have demonstrated that urinary ß2-microglobulin >500 ng/min predicted outcome with high sensitivity and specificity. In the present dataset, we observed development of RD in five out of seven patients with uß2M >500 ng/min and in one of 14 patients with uß2M <500 ng/min (P < 0.01). In Figure 2, urinary ß2-microglobulin excretion is plotted against the percentage of FSGS. There was no strong correlation between these variables (P = 0.17).
|
Meta-analysis of studies analysing the prognostic significance of FSGS in iMN
A total of seven studies, including the present study, fulfilled the inclusion criteria [9–12,14,15] All studies were retrospective by design. Detailed data of one study were unavailable [14]. An overview of all seven studies is given in Table 5. The study by Troyanov et al. did not report the absolute number of patients attaining a remission [15]. Only the remission rate of treated and untreated patients with and without FSGS were reported. However, given that FSGS was present in 25% of patients and the percentage of treated and untreated patients with and without FSGS was not reported to be different, we were able to extrapolate the number of patients with and without FSGS attaining a remission.
|
The six papers included in the review contained a total of 695 patients with membranous nephropathy (235 with FSGS and 460 without FSGS). The results show a significantly higher remission rate for patients with iMN alone (odds ratio, 2.5; 95% CI, 1.3–4.8) compared to patients with iMN and FSGS lesions (Figure 3). However these results must be read with caution as there was considerable heterogeneity between studies. Serum creatinine, proteinuria and blood pressure were higher in patients with FSGS, especially in studies showing a higher remission rate in patients with iMN alone. Of note, the results of the meta-analysis were not altered, if we assumed that a higher percentage of patients with FSGS in the study by Troyanov had been treated.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Patients with iMN would greatly benefit if their prognosis could be predicted with high accuracy. Several studies have defined risk factors for deterioration of renal function such as proteinuria, age and sex [3,4,16–18]. These factors, however, do not have sufficient predictive value for renal outcome to allow firm decisions about start of immunosuppressive therapy. The level and duration of proteinuria in a model introduced by the Toronto Glomerulonephritis Registry proved a better predictor [5,19]. We recently have validated urinary excretion of ß2-microglobulin and IgG as prognostic markers of iMN [6–8]. Sensitivity and specificity exceeded 85%, and when ß2-microglobulin excretion and serum albumin were combined, specificity of these parameters was 100%.
Several studies suggested that the presence of FSGS in the renal biopsy predicted a worse outcome in patients with iMN [9–12]. In the study of Dumoulin et al., FSGS proved an independent predictor of outcome, apparently outweighing clinical parameters such as creatinine and proteinuria. The latter study prompted the current study.
Evidently, FSGS did not predict outcome in our patients with iMN. Our patient cohort consisted of patients with a nephrotic range proteinuria. During follow-up, 45% of patients progressed, which is in agreement with the natural history in nephrotic patients with iMN. Obviously, there were some differences in patient characteristics related to the presence of FSGS. Patients with FSGS were somewhat older and had a significantly higher baseline serum creatinine and more proteinuria. Univariate analysis failed in identifying FSGS as a predictor of RD. Clinical and laboratory parameters proved better predictors and multivariate analysis revealed serum creatinine and diastolic blood pressure at time of biopsy as independent predictive variables.
The conclusions of our analysis could be biased by the fact that some patients already had renal insufficiency at the time of biopsy. Therefore, we separately analysed data of a subgroup of patients with normal renal function at time of biopsy. Again, we failed to identify FSGS as prognostic parameter.
Another bias could have been the inclusion of patients with biopsies containing 5–9 glomeruli. Patients with FSGS may not have been identified correctly. In our study, biopsies of 16 patients contained 5–9 glomeruli. In 11 patients, FSGS was not observed. Five of these developed RD. Even if we would assume that all these patients had FSGS, the analysis would not reveal FSGS as an independent prognostic marker.
Unfortunately, we lacked information on urinary ß2-microglobulin excretion in many patients. Urinary ß2-microglobulin is an accurate predictor of outcome in patients with iMN. In a validation study, using predefined threshold values, sensitivity and specificity exceeded 85%. Still, the available data show a weak correlation between percentage of FSGS and urinary ß2-microglobulin excretion (Figure 2). This underlines the weakness of FSGS as a prognostic marker.
Several studies have evaluated FSGS as a predictor of outcome. Our meta-analysis of studies with available data shows that outcome is worse in patients with iMN and FSGS lesions. The meta-analysis does indicate that the presence of FSGS identifies high-risk patients. However, our data suggest that FSGS is not an independent risk factor, in agreement with all other studies but one. In most studies, patients with FSGS had higher serum creatinine, more proteinuria and more hypertension. A multivariate analysis was done in one of the older studies and in two recent studies, which both concluded that the presence of FSGS was not an independent predictor of outcome [9,14,15]. FSGS emerged as an independent prognostic factor by multivariate analysis only in the study of Dumoulin [9]. These authors studied 72 patients with iMN, FSGS being present in 42%. The patients in this study were younger and had somewhat better renal function. The outcome in these patients was rather grim, with only 24% developing persistent remission and 53% progressing to ESRD after ten years. These differences in patient characteristics may explain the different results.
Especially, the timing of renal biopsy in relation to the natural outcome of the disease may be relevant. We propose the following sequence of events in progressive iMN: proteinuria 
tubular injury reflected by increased urinary ß2-microglobulin focal sclerosis renal failure. The data of Dumoulin suggest that this model may be valid: some patients without FSGS at initial biopsy progressed to renal failure. These patients were rebiopsied and in all of them the new biopsy revealed typical FSGS lesions. Thus, the presence of FSGS may be a better predictor of outcome in studies where biopsies are taken in a later phase of the disease.
It will be difficult to attain high specificity and sensitivity with histological parameters. Most studies use only qualitative measures. Furthermore, in routine practice renal biopsies often contain inadequate amounts of tissue. In our study, eight out of 61 biopsies (13%) contained less than five glomeruli. Thus, overall accuracy would never reach 90%, even if specificity was 100%. The abovementioned studies, in fact, noted a specificity of 56–59%, clearly too low to be used as a guide for treatment decisions.
In conclusion, FSGS is not an accurate predictor of renal outcome in patiens with iMN. Renal biopsy results cannot be used to guide decisions on immunosuppressive therapy.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Dr Saskia Heeringa was supported by a student-grant of the Dutch Kidney Foundation (OW07).
Conflict of interest statement. None declared.
|
References |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionAcknowledgementsReferences |
|---|
- Haas M, Meehan SM, Karrison TG, Spargo BH. (1997) Changing etiologies of unexplained adult nephrotic syndrome: A comparison of renal biopsy findings from 1976–1979 and 1995–1997. Am J Kidney Dis 30:621–631.[Web of Science][Medline]
- du Buf-Vereijken PWG, Branten AJW, Wetzels JFM. (2005) Idiopathic membranous nephropathy: outline and rationale of a treatment strategy. Am J Kidney Dis 46:1012–1029.[CrossRef][Web of Science][Medline]
- Reichert LJ, Koene RA, Wetzels JF. (1998) Prognostic factors in idiopathic membranous nephropathy. Am J Kidney Dis 31:1–11.[Web of Science][Medline]
- Honkanen E, Tornroth T, Gronhagen-Riska C, Sankila R. (1994) Long-term survival in idiopathic membranous glomerulonephritis: can the course be clinically predicted? Clin Nephrol 41:127–134.[Web of Science][Medline]
- Pei Y, Cattran DC, Greenwood C. (1992) Predicting chronic renal insufficiency in idiopathic membranous glomerulonephritis. Kidney Int 42:960–966.[Web of Science][Medline]
- Reichert LJ, Koene RA, Wetzels JF. (1995) Urinary excretion of ß2-microglobulin predicts renal outcome in patients with idiopathic membranous nephropathy. J Am Soc Nephrol 6:1666–1669.[Abstract]
- Reichert LJ, Koene RA, Wetzels JF. (1997) Urinary IgG excretion as a prognostic factor in idiopathic membranous nephropathy. Clin Nephrol 48:79–84.[Web of Science][Medline]
- Branten AJW, du Buf Vereijken PWG, Klasen IS, et al. (2005) Urinary excretion of ß2-microglobulin and IgG predict prognosis in idiopathic membranous nephropathy: a validation study. J Am Soc Nephrol 16:169–174.
[Abstract/Free Full Text] - Dumoulin A, Hill GS, Montseny JJ, Meyrier A. (2003) Clinical and morphological prognostic factors in membranous nephropathy: significance of focal segmental glomerulosclerosis. Am J Kidney Dis 41:38–48.[CrossRef][Web of Science][Medline]
- Wakai S and Magil AB. (1992) Focal glomerulosclerosis in idiopathic membranous glomerulonephritis. Kidney Int 41:428–434.[Web of Science][Medline]
- Lee HS and Koh HI. (1993) Nature of progressive glomerulosclerosis in human membranous nephropathy. Clin Nephrol 39:7–16.[Web of Science][Medline]
- Van Damme B, Tardanico R, Vanrenterghem Y, Desmet V. (1990) Adhesions, focal sclerosis, protein crescents and capsular lesions in membranous nephropathy. J Pathol 161:47–56.[CrossRef][Web of Science][Medline]
- Cockcroft DW and Gault MH. (1976) Prediction of creatinine clearance from serum creatinine. Nephron 16:31–41.[Web of Science][Medline]
- Shiiki H, Saito T, Nishitani Y, et al. (2004) Prognosis and risk factors for idiopathic membranous nephropathy with nephrotic syndrome in Japan. Kidney Int 65:1400–1407.[CrossRef][Web of Science][Medline]
- Troyanov S, Roasio L, Pandes M, Herzenberg AM, Cattran DC. (2006) Renal pathology in idiopathic membranous nephropathy: a new perspective. Kidney Int 69:1641–1648.[CrossRef][Web of Science][Medline]
- Honkanen E, Tornroth T, Gronhagen-Riska C. (1992) Natural history, clinical course and morphological evolution of membranous nephropathy. Nephrol Dial Transplant 7:35–41.
[Abstract/Free Full Text] - Noel LH, Zanetti M, Droz D, Barbanel C. (1979) Long-term prognosis of idiopathic membranous glomerulonephritis. Study of 116 untreated patients. Am J Med 66:82–90.[CrossRef][Web of Science][Medline]
- Davison AM, Cameron JS, Kerr DN, Ogg CS, Wilkinson RW. (1984) The natural history of renal function in untreated idiopathic membranous glomerulonephritis in adults. Clin Nephrol 22:61–67.[Web of Science][Medline]
- Cattran DC, Pei Y, Greenwood CMT, Ponticelli C, Passerini P, Honkanen E. (1997) Validation of a predictive model of idiopathic membranous nephropathy: its clinical and research implications. Kidney Int 51:901–907.[Web of Science][Medline]
Accepted in revised form: 9. 3.07
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Gupta, A. Sharma, P. J. Mahanta, T. G. Jacob, S. K. Agarwal, T. S. Roy, and A. K. Dinda Focal segmental glomerulosclerosis in idiopathic membranous glomerulonephritis: a clinico-pathological and stereological study Nephrol. Dial. Transplant., October 6, 2009; (2009) gfp521v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||