The value of repeat biopsy in the management of lupus nephritis: an international multicentre study in a large cohort of patients

Transcription

1 Nephrol Dial Transplant (2013) 28: doi: /ndt/gft272 Advance Access publication 24 August 2013 Original Articles The value of repeat biopsy in the management of lupus nephritis: an international multicentre study in a large cohort of patients ABSTRACT Fabio Pagni 1, Stefania Galimberti 2, Paolo Goffredo 3, Maria Basciu 4, Sara Malachina 1, Daniela Pilla 1, Eleonora Galbiati 5 and Franco Ferrario 6 Correspondence and offprint requests to: Franco Ferrario; frferrario@gmail.com Background. The International Society of Nephrology/Renal Pathology Society (ISN/RPS) classification represents the gold standard for the histological evaluation of Systemic Lupus Erythematosus (SLE) nephritis. A repeat biopsy (RB) might be an important tool to provide information on long-term renal outcomes and optimal therapy. Aims of this study were to evaluate the use of the ISN/RPS classification and the role of RB in routine clinical practice. Methods. A total number of 142 patients with SLE nephritis and with adequate reference and RB samples were included in this multicentre retrospective study. A meticulous histological examination was centrally performed on first and RB and compared with clinical variables and follow-up data. Results. Morphological features of the ISN/RPS classification: at first and RB, significant differences were observed between segmental classes (III, IV-S) and Class IV-G in mesangial proliferation, wire loops and tuft necrosis. Clinical features and ISN/RPS classification: the correlation between serum 1 Department of Pathology, University Milano Bicocca, San Gerardo Hospital, Monza, Italy, 2 Department of Health Sciences, Center of Biostatistics for Clinical Epidemiology, University Milano Bicocca, Monza, Italy, 3 Faculty of Medicine, University Milano Bicocca, Monza, Italy, 4 Aggregate School of Pathology, University Statale Milano/University Milano Bicocca, Italy, 5 Nephrology Unit, San Gerardo Hospital, University Milano Bicocca, Monza, Italy and 6 Nephropathology Centre, Department of Pathology, University Milano Bicocca, San Gerardo Hospital, Monza, Italy Keywords: repeat biopsy, SLE nephritis creatinine, proteinuria, blood pressure levels and histological classes at first and RB demonstrated more severe renal disease in Class IV-G, both at first and RB. Agreement between ISN/ RPS classification at first and RB: 40.8% of patients changed the histological class. Fifty per cent of Class II (mild mesangial form) were reclassified as Class IV-G at RB, whereas 18.9% of Class IV-G were reclassified as Class II. The transition among segmental (III/IV-S) and mesangial forms (II/IV-G) was extremely rare. The comparison between the clinical parameters at the final follow-up and the ISN/RPS classification confirmed that the trend of serum creatinine and proteinuria between the different classes was better described at the RB (higher in Class IV-G) than on the first biopsy. Conclusions. The histopathological data suggest that morphological differences between segmental and global forms do exist, possibly due to different pathogenetic mechanisms. An RB strategy could provide additional information on long-term renal outcomes. A strategy of protocol biopsies could be useful in perspective future trials to better understand the therapeutic response and the natural history of this disease. The Author Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. 3014

2 INTRODUCTION The International Society of Nephrology/Renal Pathology Society (ISN/RPS) classification represents the gold standard for the histological evaluation of Systemic Lupus Erythematosus glomerulonephritis (SLE GN) [1 3]. In this classification, diagnostic terminology has been clearly defined, making histological diagnosis more reproducible; as a result, differences among renal pathologists are expected to decrease. ISN/RPS Class III identifies focal segmental GN with necrotizing lesions; Class IV is divided into global (IV-G) and segmental (IV-S) forms. Whether morphological differences are an expression of different pathogenetic mechanisms with different clinical features is still debated and its understanding may provide beneficial information to predict long-term renal outcomes and to identify optimal therapy to prevent end-stage renal failure [4 6]. A repeat biopsy (RB) might be an important tool to study clinical and pathological aspects of SLE GN. Nowadays, there is a scarcity of data on RB in SLE patients, even if current guidelines strongly recommend it [7 9]. Histological features at RB have been reported to correlate with improved diagnostic predictive values, while serology, creatinine levels and activity indices show a modest relation [10, 11]. Moreover, recent studies on RB in adult and children demonstrated that a second biopsy during a flare or at the end of maintenance phase of therapy could guide physicians to safer practices with improved results [12 15]. Enrolling a large number of patients diagnosed with SLE GN, the current international multicentre study aimed at evaluating the use of ISN/RPS classification and the role of RB in routine practice. Furthermore, this paper provides an overview of the most important findings in this compound field of nephropathology, matching meticulous morphological data and clinical variables. MATERIALS AND METHODS Patients This retrospective multicentre study included data on 142 patients with a histological diagnosis of SLE nephritis between 1969 and 2006, who were subsequently subjected to RB. Patients were eligible if they had SLE, as defined by the European League Against Rheumatism (EULAR [7]), and had a positive biopsy for SLE GN. In addition, adequate biopsy samples containing >10 glomeruli were required for inclusion. Thus, 37 SLE nephrites were excluded due to the absence of enough glomeruli or inconclusive histopathological results. Patients were enrolled at the following centres: Nephropathology Centre San Gerardo Hospital (n = 25), San Carlo Borromeo Hospital (n = 23), Policlinico Ca Granda Hospital (n = 23), Kitano Hospital (n = 22), Sendai Shakaihoken Hospital (n = 15), Clinic Barcelona Hospital (n = 14); University of Antioquia (n = 7), St Vincent s Hospital (n = 7), Manzoni Hospital (n = 4) and San Martino Hospital (n = 2). Proteinuria, serum creatinine and blood pressure levels were registered in 96 patients (67.7%) at the time of the first biopsy, 87 patients (61.3%) at RB and 79 (55.6%) during the follow-up visit. Patients were treated according to single institutional guidelines. Most often, Class III and IV SLE nephritis received induction therapy with corticosteroids and cyclophosphamide. Cyclophosphamide and mycophenolate mofetil were administered in the maintenance phase of therapy [16]. Histopathology All histological diagnoses (reference and RB) were centrally reviewed at the Nephropathology Centre of the San Gerardo Hospital, according to a well-accepted international form already employed in previous international trials on SLE GN [17]. Two renal pathologists (F.F., F.P.), unaware of patients clinical data, evaluated the biopsies according to the ISN/RPS classification. The scoring system of Pollak (Austin modified) for activity and chronicity was also calculated [18]. For every case, the histological slides included haematoxylin and eosin, periodic acid Schiff (PAS), Masson trichrome and periodic acid methenamine silver stains for light microscopy. The following histological parameters were examined in each renal biopsy: (i) glomeruli: mesangial proliferation, endocapillary proliferation, extracapillary proliferation, tuft necrosis, wire loops, glomerular sclerosis; (ii) interstitium: inflammatory infiltrates, fibrosis; (iii) tubules: atrophy, necrosis, intraepithelial infiltrates. The study analysed exclusively morphological data; however, immunofluorescence (IgG, IgA, IgM, C3, C1q, fibrinogen, kappa and lambda light chains) was always performed for central diagnosis confirmation. Electron microscopy results were available in 80 first biopsies and 55 RBs for additional verification. Statistical analysis Descriptive analyses were performed on the histological and clinical data collected at first and RB, and also on followup information. Cohen s weighted kappa was used to calculate the agreement between the ISN/RPS classification at the two biopsies. For the comparisons of selected histological and clinical characteristics in Class IV-G versus IV-S and versus III, the Wilcoxon Mann Whitney test and the Fisher exact test were utilized for continuous and categorical variables, respectively. All tests were two-sided with a significance level of RESULTS Patients characteristics Patients were more commonly female (n = 125, 88%) and the mean age at the reference biopsy was 29.8 years (standard deviation [SD] = ±10.3). The first biopsy mean time after SLE diagnosis was 1.2 years (SD = ±2.8), with a mean interval between the first and the RB of 4.9 years (SD = ±4.9). In 51 out of 142 patients (36%), the RB was performed during a quiescent phase of the disease to evaluate a possible reduction of immunosuppressive therapy. In 19 patients (13.3%), the indication for RB was persistent non-nephrotic proteinuria, in SLE nephritis histopathology and RB

3 patients (22.5%) nephrotic syndrome, while the remaining 40 patients (28.2%) experienced a significant worsening of the renal function or an acute renal failure. ISN/RPS classification at first and RB The distribution of the ISN/RPS classes at first and RB, and the transitions from one class to the other are shown in Table 1. Class IV was the most represented both at first (50.7%) and RB (41.5%), with the majority of patients belonging to IV-G (40.8 and 36.5%, respectively). Eighty-four out of the 142 patients (59.2%) did not change class when the RB was evaluated; Classes IV-G and V showed the most stable behaviour, with more than 70% of the patients remaining in the same class. The moderate level of agreement between the ISN/ RPS classification at first and RB is reflected by a weighted Cohen s index of 0.48 (95% CI: ). Among the reclassified patients, the transitions between Classes II and IV-G, and between III and IV-S are of particular interest. Fifty per cent of patients in Class II switched to Class IV-G at RB, while 18.9% in Class IV-G were reclassified as Class II. Three patients in Class III (20%) evolved in Class IV-S, and 35.7% included in Class IV-S were reclassified as Class III. Histopathology at first and RB Tables 2 and 3 show the main histological characteristics of the glomerular tufts on first and RB (Figures 1 and 2). As expected, a moderate or severe mesangial proliferation was observed in almost all Class IV-G patients (Figure 1E), while it was less expressed in Class IV-S (94.8 versus 35.7%, P = and 90.3 versus 57.1%, P = at reference and RB, respectively). Endocapillary and extracapillary proliferation did not differ in Classes III, IV-S and IV-G, even if severe endocapillary proliferation was poorly represented in Class III. Tuft necrosis was more often noted in IV-S patients (Figure 1C) than in IV-G cases (42.8 versus 6.8%, P = and 42.8 versus 1.9%, P = at first and RB, respectively). The presence of wire loops decreased from Classes IV-G to III through IV-S. The comparison of Class IV-G and IV-S at first biopsy was statistically significant (75.9 versus 14.3%, P = ), while at RB the difference was attenuated and of borderline significance, probably due to the small number of patients in IV-S class (63.4 versus 28.5%, P = ). Wire loops were almost absent in Class III. The distribution of segmental (ranging from 20 to 24.1%) and global sclerosis (ranging from 35.7 to 46.6%) was similar in the three classes at first biopsy, while segmental sclerosis was more represented in Class III (70.6%) and global sclerosis in Class IV-G (73%) at RB. Tables 4 and 5 show the histological differences in the interstitial lesions on the first and RB, respectively. Interestingly, on first biopsy, tubulitis was not differently expressed in the three proliferative classes, ranging from 16 to 36%. On RB, no infiltrate or tubulitis was found in Class III and IV-S patients, differently from Class IV-G patients where infiltrates and tubulitis were present in the 33 and 25% of cases, respectively. The interstitial fibrosis, which was similar on the first biopsy in the three classes (ranging from 7.1 to 17.2%), increased in Class IV-G (57.7%) and in IV-S (28.6%) on RB. The mean value of the Pollak Austin chronicity index at first biopsy was significantly higher in Class IV-G (1.81) than in Classes III (1.3) and IV-S (0.13; P = and P = , respectively). At RB, this index was >3 in all classes, with a peak of 3.93 for IV-G (Table 5). On the contrary, the activity index showed a more stable behaviour in the two biopsy occasions. ISN/RPS classification and clinical features at first and RB As shown in Table 6, atfirst biopsy, we observed a higher serum creatinine mean value in Class IV-G (1.4 mg/dl) than in Classes III (0.8 mg/dl) and IV-S (1.0 mg/dl; P = and P = , respectively). The same trend was observed at RB. The mean value of proteinuria at first renal biopsy was not different in Classes III (2.3 g/day) and IV-S (2.3 g/ day), while it increased to 4.1 g/day in Class IV-G; similar Table 1. ISN/RPS classification at first and second biopsy of the 142 SLE patients with nephritis ISN/RPS class At RB Total II III IV-S IV-G V Mixed At first biopsy II 5 (27.7) 1 (5.5) 0 9 (50) 1 (5.5) 2 (11.1) 18 (12.7) III 1 (6.6) 9 (60) 3 (20) 1 (6.6) 0 1 (6.6) 15 (10.6) IV-S 1 (7.1) 5 (35.7) 3 (21.4) 1 (7.1) 0 4 (28.5) 14 (9.8) IV-G 11 (18.9) (70.6) 2 (3.4) 4 (6.9) 58 (40.8) V 0 1 (4.2) (75) 5 (20.8) 24 (16.9) Mixed 0 1 (7.6) 1 (7.6) 0 3 (23) 8 (61.5) 13 (9.1) Total 18 (12.7) 17 (12.0) 7 (4.9) 52 (36.6) 24 (16.9) 24 (16.9) 142 (100) Results are reported as n and (%). The percentages shown in total are calculated on the overall sample of 142 patients, while those in the body of the table describe the percentage distribution of the RB reclassification in each class at first biopsy. F. Pagni et al. 3016

4 Table 2. Glomerular morphological features overall and by ISN/RPS classes at first renal biopsy INS/RPS class Table 3. Glomerular morphological features overall and by ISN/RPS classes at RB II (n = 18) INS/RPS class II (n = 18) III (n = 15) III (n = 17) IV-S (n = 14) IV-S (n =7) IV-G (n = 58) IV-G (n = 52) V V Mixed (n = 13) Mixed Total (n = 142) Mesangial proliferation 1 (5.6) 4 (26.7) 5 (35.7) 55 (94.8) a 3 (12.5) 4 (30.7) 72 (50.7) (moderate-severe) Endocapillary 0 11 (73.3) 10 (71.4) 34 (58.6) 1 (4.1) 11 (84.6) 67 (47.2) proliferation <50% Endocapillary 0 1 (6.6) 9 (64.2) 19 (32.7) 0 5 (38.4) 34 (23.9) proliferation >50% Enxtracapillary 0 6 (40) 12 (85.7) 31 (53.4) 0 9 (69.2) 58 (40.8) proliferation <50% Enxtracapillary 0 2 (13.3) 0 13 (22.4) 0 2 (15.3) 17 (12) proliferation >50% Tuft necrosis 0 2 (13.3) 6 (42.8) 4 (6.8) a 0 1 (7.7) 13 (9.1) Wire loops 1 (5.6) 0 2 (14.3) 44 (75.9) a 0 3 (23.1) 50 (35.2) Segmental sclerosis 0 3 (20) 3 (21.4) 14 (24.1) 2 (8.2) 7 (53.8) 29 (20.4) Global sclerosis 5 (28) 7 (46.6) 5 (35.7) 21 (36.2) 4 (16.6) 9 (69.2) 51 (35.9) Results are reported as n and (%). a Significant test for the comparison of class IV-G versus IV-S. Total (n = 142) Mesangial proliferation 1 (5.6) 5 (29.4) 4 (57.1) 47 (90.3) a 2 (8.3) 9 (37.5) 68 (47.9) (moderate-severe) Endocapillary 1 (5.6) 12 (70.6) 4 (57.1) 17 (32.6) 2 (8.3) 14 (58.3) 50 (35.2) proliferation <50% Endocapillary 1 (5.6) 2 (11.8) 3 (42.8) 3 (57.7) 0 5 (20.8) 14 (9.9) proliferation >50% Enxtracapillary 0 5 (29.4) 5 (71.4) 19 (36.5) 2 (8.3) 18 (75) 49 (34.5) proliferation <50% Enxtracapillary (14.2) 8 (15.3) 0 4 (16.6) 13 (9.2) proliferation >50% Tuft necrosis 0 3 (17.6) 3 (42.8) 1 (1.9) a 0 1 (4.2) 8 (5.6) Wire loops 0 1 (5.9) 2 (28.5) 33 (63.4) 0 2 (8.3) 38 (26.8) Segmental sclerosis 0 12 (70.6) 2 (28.5) 19 (36.5) 5 (20.8) 11 (45.8) 49 (34.6) Global sclerosis 6 (33.3) 11 (64.7) 3 (42.8) 38 (73) 11 (45.8) 13 (54.1) 82 (57.7) Results are reported as n and (%). a Significant test for the comparison of Class IV-G versus IV-S. findings were noted at RB. Arterial hypertension at first biopsy tended to be more common in Class IV-G than in IV-S (45.7 versus 0%, P = ). The highest prevalence of hypertension in Class IV-G (59.5%) was confirmed at RB. These data indicate more severe renal damage in Class IV-G both at first and RB SLE nephritis histopathology and RB

5 FIGURE 1: (A) Class III. A well-delineated area of tuft necrosis with intense endocapillary hypercellularity due to mesangial proliferation and inflammatory cells (arrows). The other part of the glomerular tuft is quite normal (Trichrome stain 400 ). (B) Class III. Immunohistochemistry revealed localized segmental infiltration of monocytes/macrophages (arrows, CD68 pgm1, 400 ). (C) Class IV-S. Segmental necrotizing lesions in two glomerular tufts (arrows; Trichrome 250 ). (D) Class IV-S. Immunohistochemistry revealed localized segmental infiltration of monocytes/macrophages (arrows, CD68pgm1, 400 ). (E) Class IV-G. Severe global mesangial proliferation with marked endocapillary hypercellularity. Diffuse huge subendothelial deposits with wire loops (Trichrome, 400 ). (F) Class IV-G. Immunohistochemistry revealed diffuse infiltration of monocytes/macrophages (CD68pgm1, 400 ). ISN/RPS classification and follow-up clinical features The average interval between RB and final follow-up was 7.7 years. ISN/RPS classification at first biopsy did not successfully stratify patients (Table 7), as the mean serum creatinine at follow-up was similar in patients classified as IV-S (2.1 mg/ dl) and IV-G (2.3 mg/dl). A more accurate clinical and prognostic indication came from RB values, as IV-G class patients had worst mean serum creatinine (2.8 mg/dl) compared with IV-S (0.8 mg/dl). High mean creatinine (2 mg/dl) and proteinuria levels (2.5 g/day) were observed at follow-up in patients classified as Class II at first biopsy. These values decreased in patients who had a Class II diagnosis at RB (0.9 mg/dl and 0.3 g/day, respectively). DISCUSSION Up to date, the 2003 ISN/RPS classification is considered the gold standard for the histological evaluation of SLE GN [1 3]. This classification introduced several important quantitative and qualitative differences between Class III, IV-S and IV-G lesions. To our knowledge,this multicentre collaboration included one ofthelargestcohortsofpatientswhounderwentrbforsle GN, offering the opportunity to evaluate the ISN/RPS classification in routine practice and to test the predictive value of RB in the management of disease. The study was designed to have the morphological evaluation of the original slides centrally 3018 F. Pagni et al.

6 FIGURE 2: (A/B) Class III. First and RB of the same patient with the same segmental necrotizing lesion (arrows, PAS, 400 ). (C/D) Class IV-S. First and RB of the same patient with similar segmental tuft necrosis (arrows, Silver stain, 400 ). (E/F) Class IV-G. Global mesangial proliferation, endocapillary hypercellularity and wire loops both at first and RB (H&E, 400 ). (G/H) Regression of a first biopsy IV-G lesion to Class II at RB (H&E and PAS, 400 ). (I/L) Progression of Class II to Class IV-G at RB (Trichrome, 250 ) SLE nephritis histopathology and RB

7 Table 4. Interstitial lesions and activity/chronicity indices overall and by ISN/RPS classes at first renal biopsy INS/RPS class Table 5. Interstitial lesions and activity/chronicity indices overall and by ISN/RPS classes at RB II (n = 18) INS/RPS class II (n = 18) III (n = 15) III (n = 17) IV-S (n =7) IV-G (n = 52) V Mixed Total (n = 142) Interstitial infiltrates 1 (5.9) (32.7) 4 (16.7) 4 (16.7) 26 (18.3) Tubulitis 1 (5.9) (25) 4 (16.7) 4 (16.7) 22 (15.4) Interstitial fibrosis 1 (5.9) 1 (5.9) 2 (28.6) 30 (57.7) 5 (20.8) 5 (20.8) 44 (31) Tubular atrophy 1 (5.9) 0 2 (28.6) 33 (63.5) 7 (29.1) 4 (16.7) 47 (33) Mean activity index 2.6 (3) 3.4 (2.8) 6 (4.2) 3.3 (2.9) 2.5 (3.4) 2 (3.3) 3.3 (3.3) (SD) Mean chronicity index (SD) 3.4 (3.4) 3.1 (2.4) 3.6 (1.1) 3.9 (3) 2.8 (3.2) 4.6 (3.2) 3.6 (2.7) Results are reported as n and (%), unless otherwise specified. IV-S (n = 14) IV-G (n = 58) V Mixed (n = 13) Total (n = 142) Interstitial infiltrates 1 (5.6) 2 (13.3) 3 (21.4) 7 (12.1) 2 (8.3) 4 (16.7) 19 (13.3) Tubulitis 1 (5.6) 4 (26.7) 5 (35.7) 9 (15.5) 2 (8.3) 4 (16.7) 25 (17.6) Interstitial fibrosis 1 (5.6) 2 (13.3) 1 (7.1) 10 (17.2) 1 (4.2) 8 (33) 23 (16.1) Tubular atrophy 1 (5.6) 1 (6.7) 1 (7.1) 12 (20.7) 2 (8.3) 10 (41.7) 27 (19) Mean activity index 3.1 (2.6) 3.7 (3.1) 4.5 (4.5) 5.4 (4.2) 4.4 (3.6) 3 (2.2) 4.5 (3.8) (SD) Mean chronicity index (SD) 0.7 (1.2) 1.3 (2.2) 0.1 (0.3) 1.8 (2.1)* 1.6 (2.5) 3.2 (2) 1.5 (1.8) Results are reported as n and (%), unless otherwise specified. statistically significant test for the comparison of class III vs IV-S. *statistically significant test for the comparison of class IV-G vs IV-S. performed by two experienced renal pathologists in a highly specialized nephropathology centre. Different studies have shown pathological and clinical differences between IV-S and IV-G classes. The vast majority suggested that IV-S and IV-G subclasses do not represent a simple continuum of disease; nevertheless, these findings were based on limited samples of patients and follow-up [19, 20]. Overall, patients with Class IV- G lesions had greater subendothelial deposits on immunofluorescence, and were characterized by wire loops and hyaline deposits on light microscopy. Conversely, Class IV-S showed predominant mesangial deposits and a higher rate of glomerular fibrinoid necrosis of viable glomeruli. Other differences included thepresenceofmembranoproliferativefeaturesonlyinclassiv- Glesions[21]. We observed that focal (III) and diffuse segmental (IV-S) glomerular lesions were characterized by segmental endocapillary proliferation, fibrin deposition, intense inflammatory lesion with karyorrhexis, basement membrane wall destruction and crescents (Figures 1A, C and 2A D), while the presence of severe mesangial proliferation and wire loops was glomerular characteristics strictly related to IV-G patients (Figures 1E and 2E andf). The segmental lesions showing necrosis were similar to the lesions of systemic vasculitis [22], suggesting a possible 3020 F. Pagni et al.

8 Table 6. Main clinical features overall and by ISN/RPS classes at first and RB ISN/RPS class at 1 biopsy II (n = 14/18) III (n = 11/15) IV-S (n = 6/14) IV-G (n = 47/58) V (n = 13/24) Mixed (n = 5/13) Total (n = 96/142) Mean creatinine (mg/dl) 0.89 (0.3) 0.84 (0.3) 1.02 (0.2) 1.44 (1.2) 0.82 (0.2) 2.34 (2.8) 1.23 (1.1) Mean proteinuria (g/day) 1.54 (1.8) 2.32 (1.7) 2.27 (1.7) 4.13 (3.7) 2.92 (2.4) 7.81 (10.2) 3.5 (3.9) % Arterial hypertension a ISN/RPS class at RB II (n = 14/18) III (n = 9/17) IV-S (n = 3/7) IV-G (n = 37/52) V (n = 15/24) Mixed (n = 9/24) Total (n = 87/142) Mean creatinine (mg/dl) 0.94 (0.4) 0.89 (0.2) 1.22 (0.4) 1.58 (1.0) 0.91 (0.3) 1.09 (0.5) 1.2 (0.8) Mean proteinuria (g/day) 2.50 (5.3) 1.84 (1.7) 3.42 (3.4) 3.46 (2.7) 3.26 (2.3) 3.66 (2.8) 3.10 (3.1) % Arterial hypertension a a Blood pressure levels >140/90 mmhg in two independent values. Results are reported as mean and (sd), unless otherwise specified Table 7. Main clinical features at last follow-up by ISN/RPS classes as defined at first and RB ISN/RPS class at 1 biopsy SLE nephritis histopathology and RB Interval between 1 biopsy and last follow-up (years) II (n = 11/18) III (n = 10/15) IV-S (n = 6/14) IV-G (n = 38/58) V (n = 9/24) Mixed (n = 5/13) Total (n = 79/142) 11.5 (7.5) 14.3 (6.6) 11.1 (9.4) 13.4 (8.2) 15.0 (10.8) 7.7 (6.6) 12.9 (8.1) Creatinine (mg/dl) 2 (2.7) 0.9 (0.3) 2 (2.9) 2.3 (2.8) 0.9 (0.3) 1.2 (0.4) 1.8 (2.4) Proteinuria (g/day) 2.5 (2.5) 0.5 (0.5) 0.3 (0.2) 1 (1.3) 2.1 (2.1) 2.3 (3.6) 1.3 (1.9) ISN/RPS class at RB II (n = 12/18) III (n = 9/17) IV-S (n = 4/7) IV-G (n = 33/52) V (n = 12/24) Mixed (n = 7/24) Total (n = 77/142) Interval between RB and 10.0 (4.3) 6.2 (5.9) 9.0 (5.3) 8.5 (7.9) 5.4 (6.9) 3.8 (2.1) 7.7 (6.8) last follow-up (years) Creatinine mg/dl (SD) 0.9 (0.3) 0.9 (0.2) 0.8 (0.5) 2.8 (3.1) 1 (0.3) 0.9 (0.3) 1.8 (2.4) Proteinuria g/day (SD) 0.3 (0.2) 0.4 (0.5) 0.4 (0.3) 1.7 (1.9) 2.3 (2.1) 1.9 (3) 1.3 (1.9) Results are reported as mean and (SD). Downloaded from at Universita Milano Bicocca on December 27, 2013

9 role of cellular immunity in the pathological process of these cases. This hypothesis is different from the generally accepted theory of immune complex deposition in SLE GN. The similarity of the segmental proliferative lesions to anti-neutrophil cytoplasmic antibodies (ANCA)-related GN suggests also a possible aetiological role of ANCA in SLE GN. It has been reported that 15 20% of patients with SLE have detectable ANCAs [23, 24]. Moreover, our personal experience with a previous histological series of SLE nephritis cases (data not published) seemed to confirm a different intraglomerular infiltration of monocyte/macrophages between segmental (Classes III and IV-S) and global forms (IV-G, Figure 1B, D snd F). Future studies will probably clarify the role of these inflammatory populations in the pathogenesis of SLE nephritis. The evaluation of the association between ISN/RPS classification and clinical parameters confirmed the importance of the distinction between Classes III, IV-S and IV-G. Indeed, IV-G lesions had worse proteinuria, higher serum creatinine and blood pressure levels than Classes III and IV-S, both at first and RB. Moreover, RB might be an important tool to study clinical and pathological aspects of SLE GN. Histological features at RB have been reported to correlate with improved diagnostic predictive values [10, 11]. At RB, we found that transitions from segmental (Classes III and IV-S) to global (IV-G) or mesangial forms (Class II) were very rare. Moreover, in the mixed class, we did not observe any shift among the segmental (V + III and V + IV-S) and global forms (V + IV-G), or vice versa. On the contrary, we described reciprocal changes between Classes III and IV-S. Probably due to therapy success, 18% of IV-G cases had a strong regression to Class II, whereas 50% of Class II cases had a histological progression to Class IV-G. Our RB data confirmed that III/IV-S and IV-G classes could be different entities. The changes from the first assigned class and the frequent discrepancy between reference biopsy results and clinical markers at follow-up suggest that RB might be used as an important tool in guiding physicians to safer practices. Many investigators have studied the use of RB in case of SLE nephritis flare or renal function worsening [13, 25]. Since 36% of our patients had quiescent disease at RB, one could conclude that a strategy of RB may be useful in decreasing or stopping the therapy regimens. In our study a Class II diagnosis at RB was a good prognostic indicator, while the same class at the reference biopsy was not related to good clinical follow-up data, probably due to progression to more aggressive forms. On the contrary, a RB diagnosis of IV-G was a negative factor when compared with the final follow-up. Given its retrospective nature, this study had some limitations. First of all, due to the rarity of RB in routine practice, this cohort of 142 patients with SLE and RB was obtained in an extended period of almost 40 years; therefore, some clinical data were inevitably missing. Furthermore, the authors were not able to precisely weight the correlation between different therapeutic regimens and morphological and clinical features of the RB. Thus, also considering controversial data on onetime biopsy strategy, which affirm that segmental and global subclasses could have similar renal outcomes [26], new perspective multicentre clinical trials should be performed to possibly include RB in SLE GN management guidelines and to confirm our results. CONCLUSIONS Morphological differences between ISN/RPS classes, in particular segmental and global forms, were evident and possible expression of different pathogenetic mechanisms. The transition from one ISN/RPS class to another during the natural history of SLE GN is a frequent event, but the transition among segmental (III, IV-S) and mesangial forms (II, IV-G) is very rare. An RB can be useful in monitoring the pathological evolution of the disease. Larger perspective trials, using protocol biopsies, are needed to investigate more deeply the relationship between histology and clinical data in an RB strategy. CONTRIBUTORS Pietro Napodano, MD Nephrology Unit, Hospital S. Carlo Borromeo, Milan, Italy; Gabriella Moroni, MD Croff Department, Hospital Policlinico Ca Granda, Milan, Italy; Eri Muso, MD Department of Medicine, Division of Nephrology and Dialysis, Kyoto University, Japan; Joh Kensuke, MD Division of renal pathology, Sendai Shakaihoken Hospital, Senday City, Japan; Ricardo Cervera, MD Department of Autoimmune diseases, Hospital Clinic Barcelona, Spain; Luis Arias, MD Division of Nephrology, University of Antioquia, Medellin, Colombia; Prue Hill, MD Anatomical Pathology, St Vincent s Hospital Melbourne, Australia; Francesco Locatelli, MD Nephrology Unit, Hospital Manzoni, Lecco, Italy; Giacomo Garibotto, MD Department of Internal Medicine, University of Genoa, Hospital San Martino, Genova, Italy CONFLICT OF INTEREST STATEMENT None declared. (See related article by Tesar. Rare transformation in repeat renal biopsies suggests a different pathogenesis of segmental and global lesions in proliferative lupus nephritis. Nephrol Dial Transplant 2013; 28: ) REFERENCES 1. Weening JJ, D Agati VD, Schwartz MM et al. The classification of glomerulonephritis in systemic lupus erythematosus revisited. J Am Soc Nephrol 2004; 15: Markowitz GS, D Agati VD. Classification of lupus nephritis. Curr Opin Nephrol Hypertens 2009; 18: Markowitz GS, D Agati VD. The ISN/RPS 2003 classification of lupus nephritis: an assessment at 3 years. Kidney Int 2007; 71: Appel GP, Cohen DJ, Pirani CL et al. Long-term follow-up of patients with lupus nephritis: a study based on the classification of the World Health Organization. Am J Med 1987; 83: F. Pagni et al.

10 5. Askenazi D, Myones B, Kamdar A et al. Outcomes of children with proliferative lupus nephritis: the role of protocol renal biopsy. Pediatr Nephrol 2007; 22: Chen YE, Korbet SM, Katz RS et al. Value of a complete or partial remission in severe lupus nephritis. Clin J Am Soc Nephrol 2008; 3: Bertsias G, Ioannidis JP, Boletis J et al. Report of a task force of the EULAR Standing Committee for International Clinical Studies Including Therapeutics. Ann Rheum Dis 2008; 67: Hahn BH, McMahon MA, Wilkinson A et al. American College of Rheumatology Guidelines for screening, treatment, and management of lupus nephritis. Arthritis Care Res 2012; 64: Gordon C, Jayne D, Pusey C et al. European consensus statement on the terminology used in the management of lupus glomerulonephritis. Lupus 2009; 18: Hill GS, Delahousse M, Nochy D, Bariéty J. Class IV-S versus class IV-G lupus nephritis: clinical and morphologic differences suggesting different pathogenesis. Kidney Int 2005; 68: Mittal B, Hurwitz S, Rennke H, Singh AK. New subcategories of class IV lupus nephritis: are there clinical, histologic and outcome differences? Am J kidney Diseases 44: Alsuwaida A, Husain S, Alghonaim M et al. Strategy for second kidney biopsy in patients with lupus nephritis. Nephrol Dial Transplant 2012; 27: Daleboudt GM, Bajema IM, Goemaere NN, van Laar JM, Bruijn JA, Berger SP. The clinical relevance of a repeat biopsy in lupus nephritis flares. Nephrol Dial Transplant 2009; 24: Lu J, Tam LS, Lai FM et al. Repeat renal biopsy in lupus nephritis: a change in histological pattern is common. Am J Nephrol 2011; 34: Sidiropoulos PI, Kritikos HD, Boumpas DT et al. Lupus nephritis flares. Lupus 2005; 14: Gurevitz SL, Snyder JA, Wessel EK, Frey J, Williamson BA. Systemic lupus erythematosus: review of the disease and treatment options. Consult Pharm 2013; 28: Furness PN, Taub N. Interobserver reproducibility and application of the ISN/RPS classification of lupus nephritis: a UKwide study. Am J Surg Pathol 2006; 30: Schwartz MM, Bernstein J, Hill GS, Holley K, Phillips EA. Predictive value of renal pathology in diffuse proliferative lupus glomerulonephritis. Lupus Nephritis Collaborative Study Group. Kidney Int 1989; 36: Yu F, Tan Y, Wu LH, Zhu SN, Liu G, Zhao MH. Class IV-G and IV-S lupus nephritis in Chinese patients: a large cohort study from a single center. Lupus 2009; 18: Yu F, Tan Y, Liu G, Wang SX, Zou WZ, Zhao MH. Clinicopathological characteristics and outcomes of patients with crescentic lupus nephritis. Kidney Int 2009; 76: Najafi CC, Korbet SM, Lewis EJ, Schwartz MM, Reichlin M, Evans J. Significance of histologic patterns of glomerular injury upon long-term prognosis in severe lupus glomerulonephritis. Kidney Int 2001; 59: Ferrario F, Vanzati A, Pagni F. Pathology of ANCA-associated vasculitis. Clin Exp Nephrol 2012; [Epub ahead of print] 23. Behara VY, Whittier WL, Korbet SM, Schwartz MM, Martens M, Lewis EJ. Pathogenetic features of severe segmental lupus nephritis. Nephrol Dial Transplant 2010; 25: Chin HJ, Ahn C, Lim CS et al. Clinical implications of antineutrophil cytoplasmic antibody test in lupus nephritis. Am J Nephrol 2000; 20: Sprangers B, Monahan M, Appel GB. Diagnosis and treatment of lupus nephritis flares-an update. Nat Rev Nephrol 2012; 8: Haring CM, Rietveld A, van den Brand JA, Berden JH. Segmental and global subclasses of class IV lupus nephritis have similar renal outcomes. J Am Soc Nephrol 2012; 23: Received for publication: ; Accepted in revised form: SLE nephritis histopathology and RB