Histological grading in primary membranous nephropathy is essential for clinical management and predicts outcome of patients

Aims Diagnosis of primary membranous nephropathy (PMN) is mainly based on immunofluorescence/immunohistochemistry findings. However, assessment of specific features on optical microscopy can help to estimate the severity of the disease, guide treatment and predict the response. The aim of this study was to identify, classify and grade the precise histological findings in PMN to predict renal function outcome and guide treatment. Methods and results Histological parameters, including focal segmental sclerosis (FSGS), tubular atrophy (TA), interstitial fibrosis (IF) and vascular hyalinosis (VH), were re‐evaluated in 752 patients with PMN. Their predictive value was estimated separately, and also in a combination score (FSTIV) graded from 0 to 4. Finally, the impact of histology was assessed in the response to immunosuppressive treatment. Mean age of patients was 53.3 (15–85) years and most presented with nephrotic syndrome. FSGS was present in 32% and VH in 51% of the patients, while TA and IF were graded as stage ≥1 in 52% and 51.4%, respectively. The follow‐up period was 122.3 (112–376) months. FSGS, TA and IF and VH were associated with impaired renal function at diagnosis (P = 0.02, P < 0.0001, P = 0.001 and P = 0.02, respectively) and at the end of follow‐up (P = 0.004, P < 0.0001, P < 0.0001 and P = 0.04, respectively). In multiple regression and binary logistic analysis, the presence of FSGS and degree of TA were the most significant parameters predicting renal function outcome, defined either by eGFR (end), FSGS (r = 0.6, P < 0.0001) and TA (r = 0.6, P < 0.0001), or by the endpoint of >50% eGFR reduction, FSGS (P = 0.001) and TA (P = 0.02). Also, patients presented with FSGS, IF, VH and/or with FSTIV > 1 could benefit from immunosuppression, regardless of clinical presentation. Conclusions The presence and degree of four histological indices, FSGS, VH, TA and IF, assessed separately or in combination, and FSTIV score not only predict renal function outcome after long‐term follow‐up, but can also help in the choice of appropriate treatment. Decisions concerning immunosuppressive treatment can be guided by pathology regardless of clinical findings.

estimate the severity of the disease, guide treatment and predict the response. The aim of this study was to identify, classify and grade the precise histological findings in PMN to predict renal function outcome and guide treatment. Methods and results: Histological parameters, including focal segmental sclerosis (FSGS), tubular atrophy (TA), interstitial fibrosis (IF) and vascular hyalinosis (VH), were re-evaluated in 752 patients with PMN. Their predictive value was estimated separately, and also in a combination score (FSTIV) graded from 0 to 4. Finally, the impact of histology was assessed in the response to immunosuppressive treatment. Mean age of patients was 53.3  years and most presented with nephrotic syndrome. FSGS was present in 32% and VH in 51% of the patients, while TA and IF were graded as stage ≥1 in 52% and 51.4%, respectively. The follow-up period was 122.3 (112-376) months. FSGS, TA and IF and VH were associated with impaired renal function at diagnosis (P = 0.02, P < 0.0001, P = 0.001 and P = 0.02, respectively) and at the end of follow-up (P = 0.004, P < 0.0001, P < 0.0001 and P = 0.04, respectively). In multiple regression and binary logistic analysis, the presence of FSGS and degree of TA were the most significant parameters predicting renal function outcome, defined either by eGFR (end), FSGS (r = 0.6,

Introduction
Primary membranous nephropathy (PMN) remains the most common cause of nephrotic syndrome in white adults, representing 20-30% of cases in 50-60-year-old patients, rising to 40% in ages greater than 60 years. [1][2][3] Diagnosis of MN is based on optical microscopy and immunohistochemistry and/or immunofluoresence findings, showing thickening of glomerular basement membrane with or without spikes and granular deposition of immunoglobulin (Ig)G and C3 along the glomerular capillary walls. Electron microscopy demonstrates the subepithelial deposits and helps in disease classification. 2,[4][5][6][7] Recent discovery of circulating pathogenic antibodies, the M-type phospholipase A2 receptor 1 (PLA2R) and the thrombospondin type 1 domain-containing 7A (THSD7A), supported the theory of immunological pathogenesis of the disease and helped to change the term 'idiopathic' MN to 'primary' MN (PMN). [8][9][10][11][12][13] Although the description of PLA2R and THSD7A antibodies has regenerated the clinicians' wish to substitute a serum or urinary biomarker for renal biopsy, we are far from coming to this point, as renal biopsy gives much more information than a histopathologically unsupported diagnosis. Therefore, renal biopsy remains vital for diagnosis of PMN and assessment of disease severity. Classification of the disease on a scale of 0-IV, based on electron microscopy findings, was an attempt to estimate disease severity and categorise patients according to renal lesions, although it has not always been proved to be of clinical significance. 14 Also, electron microscopy is not broadly available within pathology centres. Therefore, there is a need to evaluate optical microscopy and define specific features which can be easily assessed in clinical everyday practice, and have the capacity to estimate severity of the disease and guide treatment.
The present study included patients from the Greek Registry of Primary Membranous Nephropathy, part of the Greek Network of Glomerular Diseases. All patients had undergone a thorough investigation to exclude secondary cases; they had a long-term follow-up, reaching more than 10 years, and their histological reports were re-evaluated in an attempt to define parameters which may determine disease progression. Furthermore, we defined specific histological findings which, in combination with clinical symptoms at presentation, may identify the profile of patients who are likely to benefit from immunosuppressive treatment.

C L I N I C A L D A T A , T R E A T M E N T A N D O U T C O M E
The clinical data from 752 patients with PMN from 22 centres diagnosed during the period 1995-2015 were obtained from the Registry of Primary Membranous Nephropathy, Greek Network of Glomerular Diseases. Diagnosis was based on optical microscopy and immunohistochemistry and/or immunofluorescence. Electron microscopy was performed on a limited number of patients. Extensive research had excluded secondary cases of MN, such as tumours, systemic diseases, viral infections and previous medication. Renal function and degree of proteinuria at time of diagnosis and at the end of follow-up, the presence of nephrotic syndrome (NS) and renal function impairment at diagnosis, immunosuppressive treatment and outcome were recorded for all patients.
Patients were treated according to Kidney Disease: Improving Global Outcomes (KDIGO) guidelines, immunosuppressive treatment was added either after a 6-month trial with RAASi in cases with partial remission or no response, or earlier, in patients with rapidly progressive disease and/or life-threating complications.
All patients had at least 12 months of follow-up, unless they reached end-stage renal disease (ESRD) or died. End of follow-up was considered the last visit to outpatient clinic, the day started on haemodialysis or death.
The study was approved by the Institutional Review Board of Hippokration General Hospital, Thessaloniki, Greece, Approval number 10/18, approval date 22 March 2016.
Renal biopsy reports were re-evaluated, and characteristics estimated included the presence or absence of focal segmental sclerosis (FSGS), vascular hyalinosis (VH) and immunoglobulin (Ig)G and C3 staining on immunofluoresence and/or immunohistochemistry. The severity of tubular atrophy (TA) and interstitial fibrosis (IF) was rated on a scale of 0, 1 or 2 based on the percentage of affected tubules (<25, 25-50, >50%) or the extension of IF (absent-mild, moderate, severe), respectively.
Combining the four histological parameters, including FSGS (0-1), TA (0-1, 2), IF (0-1, 2) and VH (0-1), we set a FSTIV score, graded in scale from 0 to 4 (when none of the above indices were evident, to one, two, three or four parameters evident, respectively). For practical reasons, each of the histology indices were considered as dichotomous (0 or 1 for present or absent, respectively), TA and IF graded as stage 0 were considered as absent and stages 1 or 2 were considered as present.
Patients who presented with eGFR < 60 ml/min/ 1.73 m 2 , which remained at these levels for at least 3 months, were considered as having chronic kidney disease (CKD). End of follow-up was considered the last visit to outpatient clinic or, for those who reached ESRD or died, the time when they started on the dialysis method or day of death.
Outcome of renal function was assessed by (i) eGFR at the end of follow-up; eGFR (end); (ii) response to treatment, defined as complete, partial or no response; and (iii) endpoint of a >50% reduction in eGFR and/or ESRD. The rate of renal function alteration was estimated by the equation [eGFR (end)-eGFR (start)]/follow-up in years and defined as slope-eGFR.
eGFR was estimated based on the Modification of Diet in Renal Disease study (MDRD) equation. Response to treatment was defined, according to KDIGO, as complete when proteinuria reduced to <0.3 g/24 h with normal levels of serum albumin and normal renal function, partial as the reduction of proteinuria to >50% from baseline, but still remains between 0.3 and 3.5 g/24 h and stable renal function, and no response, when proteinuria levels remained ≥50% from baseline and/or ≥3.5 g/24 h and/or declining renal function. 15

S T A T I S T I C A L A N A L Y S I S
Statistical analysis was performed using SPSS version 23.0 software (SPSS Inc., Chicago, IL, USA) for Windows. P < 0.05 were considered statistically significant. Results from normally distributed variables were expressed as mean AE standard deviation (SD), and Student's t-test or analysis of variance (ANOVA) test were performed to compare differences between groups. Non-normal variables were expressed as medians and interquartile range (IQR), and differences between groups were compared by Mann-Whitney U or Kruskall-Wallis tests. The Kaplan-Meier test and odds ratio (OR) were performed to estimate renal survival and response to treatment in the presence of different histological findings. Multiple regression analysis was performed to assess independent factors associated by renal function outcome, defined by e-GFR (end), slope-eGFR and response to treatment, and binary logistic regression was used when outcome was defined by the 50% reduction in eGFR.

R E N A L F U N C T I O N A T P R E S E N T A T I O N A N D
Clinical parameters included age, male sex, presence of hypertension, degree of proteinuria, severity of dyslipidaemia, anaemia and histological findings: the degree of global sclerosis, presence of FSGS and VH and severity of TA and IF, and the FSTIV score had a positive correlation with renal function and some patients with severity of proteinuria at the time of diagnosis (Table 1).
Patients with FSGS in renal biopsy had significantly impaired renal function and severe proteinuria at the time of diagnosis (mean rank = 267.76 and 235.98, P = 0.02, and 260.22 and 288.12, P = 0.05, for eGFR and absolute proteinuria (Uprot) respectively, in FSGS (À) and FSGS (+) patients at the time of diagnosis. Differences in the outcome of renal function and rate of deterioration between FSGS (À) and FSGS (+) patients were also significant; mean rank = 235.98, 200.44, P = 0.004 for eGFR (end), 191.9 and 167.7, P = 0.04, for the slope-eGFR and 252. 10 The presence of vascular hyalinosis was also accompanied by a significantly reduced eGFR at the beginning and end of the study (mean rank = 262.41, 234.11, P = 0.02 and 197.05, 175.03, P = 0.04, respectively), while there were statistically significant differences in Uprot and slope-eGFR. The impact of the four histological parameters in renal function at the beginning and end of the study is expressed in Figure 2 and their influence in response to treatment and the >50% reduction of eGFR in Table 2.
Multiple regression analysis showed that the independent factors correlated with renal function at diagnosis were age (r = À0.5, P < 0.0001), degree of proteinuria (r = 0.43, P < 0.0001) and the severity of TA (r = 0.4, P < 0.0001).
Response to treatment in patients who presented with NS was correlated with all four histological parameters: the presence of FSGS (P = 0.005) and VH (P = 0.006) and the degree of TA (P < 0.0001) and IF (P < 0.0001) (Figure 3). However, slope-eGFR was associated only with the presence of FSGS in renal biopsy, and was estimated to À2.2 AE 4.3 versus À1.2 AE 3.2 ml/min/1.73 m 2 /year, P = 0.02, in FSGS (+) and FSGS (À), respectively.
The ORs of the four histological parameters, performed initially on the whole cohort of patients and subsequently to patients with NS and patients with preserved renal function for the response to treatment and deterioration of renal function, defined as >50% reduction of eGFR and/or ESRD, are shown in Table 2.

T H E I N F L U E N C E O F H I S T O L O G Y I N T H E E F F E C T O F I M M U N O S U P P R E S S I V E T R E A T M E N T
We evaluated the impact of FSGS, TA, IF and VH in the response to immunosuppression. Immunosuppressive treatment had a beneficial effect in patients with FSGS, IF (grades 1-2) and VH on renal biopsy, but not in those without FSGS, IF and VH ( Figure 4). Moreover, the combination FSTIV score also seemed to positively influence effect of immunosuppressive treatment. In order to simplify analysis, we divided patients into those with FSTIV score to ≤1 (n = 342) and with FSTIV score >1 (n = 410), and evaluated the possible beneficial effects from immunosuppressive treatment. As depicted in Figure 5, immunosuppressive treatment had no effect in the former group; besides, there was a significantly better outcome of renal function after immunosuppression in the second cohort of patients with advanced histology, and this effect was constant even after long-term follow-up.

Discussion
Diagnosis of PMN and exclusion of secondary cases are widely based on immunohistochemistry and/or immunofluorescent findings, and classification currently depends on electron microscopy. However, electron microscopy is not widely available in every centre, and current classification has not been proved to correlate with disease outcome. 14,16 Therefore, in the present study we intended to search for a simplified grading in PMN, including histological parameters, which can be easily identified and evaluated in a routine base optical microscopy. Histology in MN does not include proliferative lesions. Mesangial proliferation is unusual and may act as an indicator of secondary forms of disease, and crescent formation and endothelial proliferation are extremely rare. Besides, tubulointerstitial lesions are widely accepted as the most important findings for disease outcome. [17][18][19] We therefore decided to integrate only four parameters in the assessment of renal biopsies, namely: (i) the presence of FSGS, representing glomerular damage, (ii) degree of TA, (iii) severity of IF, to assess tubulointerstitial damage and (iv)  presence of VH, to estimate the vasculature implication. Our objective was to review the importance of ordinary histological findings into a routine pathology assessment of PMN, and to provide parameters with prognostic significance and indices to guide treatment approach. Furthermore, we combined the above four     histological parameters in a novel, concise score (the FSTIV score), and assessed its significance in predicting response to immunosuppressive treatment and renal function outcome in long-term follow-up. FSGS and vascular hyalinosis were present in onethird and half of cases, respectively, while moderate to severe (grades 1 and 2) tubular atrophy and interstitial fibrosis were seen in more than 50% of the patients. Clinical presentation of our patients covered all aspects of disease, from mild proteinuria and normal renal function to severe renal impairment with or without nephrotic syndrome, hypertension and haematuria. Clinical parameters, such as age, male sex, hypertension and degree of proteinuria and histological findings, including obsolescent glomeruli, presence of focal sclerosis and vascular hyalinosis, severity of tubular atrophy and interstitial fibrosis, correlated with severity of renal function impairment at the time of renal biopsy. These findings are in agreement with previous results, which showed that clinical parameters with predictive significance were age, male sex, hypertension and renal insufficiency. 20 There has been a great deal of debate concerning the accuracy and efficacy of the levels of proteinuria and eGFR in determining treatment strategies, but although the story goes back to the early 1980s there is still no accurate answer, even after the use of biomarkers such as beta-2-microglobulin or N-acetylbeta-glucosaminidase, or the description of anti-PLA2R antibodies. [20][21][22][23][24] Therefore, histology should be thoroughly analysed in order to estimate the severity of disease and make decisions concerning treatment choices. In the present study we found that, in the whole cohort of patients, all four parameters examined, including FSGS, TA, IF and VH, had a significant negative correlation with eGFR levels at presentation and also with renal function outcome; however, the most important of them, as proved by the multiple regression analysis, being the presence of FSGS and severity of TA and IF.
In the present study, we estimated eGFR at the end of follow-up and the ≥ 50% reduction of eGFR in order to evaluate outcome of the disease, and we also estimated slope-eGFR as an indicator of renal function deterioration rate. We also used KDIGO definition to assess the response to treatment approach. KDIGO is the most important and widely accepted clinical practice guideline document for the assessment and treatment of primary and secondary glomerular diseases. KDIGO guidelines for PMN are mainly based upon two main clinical characteristics, the presence of NS and the presence of CKD. All treatment strategies and decisions, and the decision to start on immunosuppressive treatment, are made upon the degree of proteinuria and the preservation of renal function. 15 Therefore, additionally to the whole cohort of patients, we examined the importance of histology in the presence of NS and preserved renal function.
Regarding patients who presented with NS; FSGS, TA and IF had a significant role in final outcome. However, the presence of FSGS was the only finding associated with the rate of declining renal function. Among patients with preserved renal function, FSGS, TA and IF were important in both outcome, and rate of deterioration.
The evaluation of FSGS as a predictor factor of renal function outcome has gained much interest but results remain conflicting, with some investigators supporting a pathogenic role of FSGS in declining renal function, providing FSGS and the severity of tubulointerstitial fibrosis as the only independent factors of disease outcome, [25][26][27][28] while others doubt that there is any correlation of FSGS with renal function outcome. 29,30 Based on the results from our study, FSGS and VH are important findings in the presence of NS, and additionally the presence of FSGS predicts a worse outcome in patients with preserved renal function at diagnosis. Apparently, neither FSGS nor VH should be regarded as findings secondary to co-factors, such as advanced age and hypertension, but they seem to act independently as histological parameters determining progression of renal function.
The simplified FSTIV score, resulting from the combination of the most common histological findings in PMN, also had a significant impact on the outcome and response to treatment. A similar score, the MEST score, based on the presence of mesangial (M), endocapillary (E), hypercellularity, glomerular sclerosis (S) and tubular atrophy (T), has been previously applied to estimate the severity of IgA nephropathy. 31,32 FSTIV score is based on optical microscopy findings, can easily be described by a pathologist, and can assist the clinician to estimate disease severity, predict renal function outcome and decide upon treatment approach.
An interesting finding of our study was the influence of histology on the effect of immunosuppressive treatment. Regardless of clinical presentation, patients with FSGS, VH and IF (degrees 1 and 2) in renal biopsy had a significantly better outcome when they received immunosuppressive treatment combined with those who did not. Furthermore, patients with an FSTIV score >1 seemed to benefit from immunosuppression, instead of those with less severe histology, and this beneficial effect was stable after longterm follow-up. These findings, in agreement with previous results which showed that even patients with severe tubulointerstitial injury may gain remission after treatment, 33 support the administration of immunosuppressive treatment in PMN, especially in the presence of the above histological features.
In conclusion, after evaluation of certain histological findings in PMN, we found that FSGS, TA, IF and VH are significant parameters predicting renal function outcome and response to treatment and should be considered as imperative factors in the management of PMN. The combination of these four findings revealed that the FSTIV score, a novel, simplified and comprehensive score, can not only help in grading PMN, but can also be used as guidance for treatment strategies.