Clinical spectrum and outcome of crescentic glomerulonephritis in children in developing countries

Pediatr Nephrol (2008) 23:389 394 DOI 10.1007/s00467-007-0647-3 ORIGINAL ARTICLE Clinical spectrum and outcome of crescentic glomerulonephritis in children in developing countries Deepak Dewan & Sanjeev Gulati & Raj K. Sharma & Narayan Prasad & Manoj Jain & Amit Gupta & Alok Kumar Received: 10 March 2007 / Revised: 22 August 2007 / Accepted: 3 September 2007 / Published online: 11 December 2007 # IPNA 2007 Abstract Crescentic glomerulonephritis (CsGN) is an uncommon entity in children. This prospective study was conducted to evaluate the aetiology, clinical spectrum and outcome in children with crescentic glomerulonephritis. The single-centre prospective study comprised of 22 children with biopsy proven CsGN who had been referred to our institute over the period January 2000 to December 2005. These patients were subjected to detailed clinical and biochemical examinations. The diagnosis of underlying renal disease was based on various criteria, including the clinical picture, serology and histopathology. The patients received intravenous methyl prednisolone, oral steroid treatment, and oral cyclophosphamide with or without plasmapheresis. All patients received supportive care, including control of hypertension and oedema and supportive management of renal insufficiency. During this 5-year period, CsGN accounted for 5.1% of all biopsies done in children. The mean age was 12.27 years (range 4 years to 18 years). There were eight girls and 14 boys. The mean duration of symptoms prior to referral was 2.47 months (range 5 days to 21 months). Aetiology was immune complex in 19 cases, anti-glomerular basement membrane (anti-gbm) antibody disease in two cases and pauciimmune (Wegener s granulomatosis) in one case. The D. Dewan : S. Gulati (*) : R. K. Sharma : N. Prasad : A. Gupta : A. Kumar Department of Nephrology, Sanjay Gandhi Post-Graduate Institute of Medical Sciences, Rae Bareily Road, Lucknow 226014, India e-mail: sgulatipedneph@yahoo.com M. Jain Department of Pathology, Sanjay Gandhi Post-Graduate Institute of Medical Sciences, Lucknow, India percentage of crescents ranged from 50% to 100% (mean 70.6%). Twenty-one out of 22 (95.5%) children in our series had hypertension at presentation that required treatment with antihypertensive medications. The serum creatinine level at presentation ranged from 1.5 mg/dl to 11.4 mg/dl (mean 5.5 mg/dl). Of the 22 children, two were lost to follow-up, while the mean follow-up period of the rest of the 20 children was 8.13 months (range 1 month to 43 months). At the last follow-up of the 22 children, ten had stage 5 chronic kidney disease (CKD) and three had stage 4 CKD, while seven children had a calculated glomerular filtration rate (GFR) of >60 ml/min per 1.73 m 2 body surface area. Persistent proteinuria was seen on follow-up in the majority [13/20 (65%)] of patients. The outcome of CsGN in children continues to be poor, in our experience, due to delayed referral and delayed diagnosis. This was correlated histologically by the presence of fibrocellular crescents in the majority of our patient. Thus CsGN should be treated as a renal emergency. A greater awareness of this disease needs to be created amongst the referring paediatricians in developing countries to facilitate early diagnosis and prompt treatment. Keywords Crescentic glomerulonephritis. Chronic kidney disease. Proteinuria Introduction Crescentic glomerulonephritis (CsGN) is an uncommon entity in children [1]. Histopathologically, it is characterized by crescents in 50% or more of the glomeruli and clinically by sudden and progressive decline in renal function. It can accompany most forms of primary glomerulonephritis but is also associated with various systemic diseases. An early

390 Pediatr Nephrol (2008) 23:389 394 diagnosis and prompt therapy is the key to the management of this entity. Over the years, there have been significant advancements in supportive, as well as specific, therapy. There is a paucity of reports from developing countries [2 9] and there are hardly any series from other parts of the world also. This study was conducted to evaluate the aetiology, clinical spectrum, treatment and outcome of crescentic glomerulonephritis in children. Methods The study group was composed of all children with CsGN referred to our institute over the period January 2000 to December 2005 who were examined prospectively. Crescentic glomerulonephritis was defined as the presence of large epithelial crescents filling the Bowman s spaces in 50% or more glomeruli on biopsy. These patients were subjected to detailed clinical and biochemical examinations. Clinical evaluation included age, gender, duration of symptoms at presentation, presence or absence of hypertension, history of decreased urine output, duration of follow-up time, need for dialysis, treatment given and outcome in relation to proteinuria and renal function. Biochemical parameters studied included serum creatinine level at entry, peak serum creatinine, serum creatinine at last follow-up, serum protein and albumin levels, urine routine and microscopic examination and 24 h urinary protein level. They were also screened for anti-nuclear antibodies (ANAs), anti-neutrophilic cytoplasmic antibodies (ANCAs), anti-double-stranded DNA antibodies; anti-glomerular basement membrane (anti- GBM) antibodies and serum complement levels (C3, C4). All of the patients underwent kidney biopsy. These were evaluated by an expert renal histopathologist who was blind to the clinical outcome. Histopathology examination was done by light and immunofluorescence microscopy. Light microscopic study was performed on renal tissue samples fixed in 10% formalin solution, embedded in paraffin wax and cut into thin sections by routine methods. Sections were stained with haematoxylin and eosin, periodic acid Schiff stain, silver methenamine and Mason s trichrome stain. Features evaluated on histopathology included total number of glomeruli seen, number of glomeruli with crescents, type of crescents, i.e. cellular or fibrocellular, partial or circumferential, presence and degree of sclerosis, and tubulointerstitial involvement. Crescents were defined as cellular, fibrocellular or fibrous, using the following definitions that were modified from the World Health Organization classification of renal disease of Churg et al. [10]. Cellular crescents Presence of cells filling part or all of the Bowman s space. Minimum requirement for this definition was the presence of at least two layers of cells in addition to the visceral and parietal epithelial, with involvement of at least 10% of the glomerular circumference. Fibrocellular crescents A lesion similar to cellular crescents but with the addition of a variable amount of fibrillar material. It was assessed with the silver methenamine and the trichrome stains. Fibrous crescents A lesion within the Bowman s space composed predominantly of fibrous tissue (with functional features of collagen). This lesion, which may represent scarred remains of cellular and fibrocellular crescents, was also assessed with the trichrome stain. Specimens for immunofluorescence studies were snap frozen and stained by a direct method using fluoresceinated antisera monospecific C3, C4, C1q, IgA, IgG and IgM. Electron microscopy was performed only in a few cases, and, hence, the results are not mentioned in the text. Data on the renal replacement therapy and specific immunosuppressive treatment were also analysed. The diagnosis of underlying renal disease was based on various criteria, including the clinical picture, serology and histopathology. Hypertension was defined as a systolic or diastolic blood pressure greater than the 95th percentile for age, based on the Task Force recommendations [11]. Proteinuria was evaluated by dipstick and/or a spot urine protein creatinine ratio. Nephrotic syndrome was defined as proteinuria ( 2+ by qualitative dipstick) or a spot urine protein creatinine ratio of >2, serum albumin <2.5 gm/dl with or without oedema [3]. Haematuria was defined as the presence by dipstick positivity for blood or presence of 5 red blood cells per high-power field on microscopic examination. Serum C3 level was considered low if it was <60 mg/dl, while low C4 level was <15 mg/dl. Anti-GBM antibody titres were considered positive if they were >15 IU/ l, while double-stranded DNA (dsdna) was considered to be positive if it was >30 IU/l. A positive ANCA was defined as >5 IU/l for anti-myeloperoxidase (anti-mpo) or antiproteinase 3 (anti-pr3) antibodies. These patients were treated with methyl prednisolone (MP) and/or orally administered steroids (OS). Methyl prednisolone was given at a dose of 30 mg/kg daily for 3 days, while steroids were given orally in doses of 1 mg/ kg per day and cyclophosphamide orally (OE) at 2 mg/kg per day. Plasmapheresis (PP) was done for two patients with pauci-immune crescentic glomerulonephritis. Fourteen sessions of plasma exchange of 40 ml/kg body weight were given. All patients received supportive care, including control of hypertension and oedema and supportive management of renal insufficiency. Four patients developed side effects from the immunosuppressive therapy in the

Pediatr Nephrol (2008) 23:389 394 391 form of infection requiring hospitalization and treatment with antibiotics, and they all recovered. Children were divided into various stages of chronic kidney disease based on glomerular filtration rates (GFRs) as per the Kidney Disease Outcomes Quality Initiative (K/DOQI) guidelines [12]. Patients were divided into two groups depending upon the percentage of glomeruli with crescents. Group I Patients with crescents in 50% to 79% of glomeruli. Group II Patients with crescents in 80% or more glomeruli. Comparison was made between group I and group II in relation to various clinical, laboratory, treatment and outcome parameters. Statistical methods Data were examined by univariate and multivariate analysis. A two-tailed Student s t-test was used for the unpaired data. Differences among proportions were evaluated by chisquare test. A P value less than 0.05 was considered significant. Renal survival was evaluated by Kaplan Meier survival analysis. Results This single-centre study comprised of 22 consecutive children with biopsy proven crescentic glomerulonephritis (14 boys, eight girls). Examination of 430 renal biopsies from children referred to our institute during the period January 2000 and December 2005 revealed crescentic glomerulonephritis in 22 cases (5.1%). Patients characteristics Epidemiological, clinical and laboratory features of the patients are shown in Table 1, while Table 2 illustrates the histopathological features. The mean age was 12.27 years (range 4 years to 18 years). The mean duration of symptoms prior to referral was 2.47 months (range 5 days to 21 months). The majority of the children (86.4%) presented with decreased urine output and oedema. Twenty-one of the 22 (95.5%) children in our series had hypertension at presentation that required treatment with antihypertensive medications. Levels of serum creatinine at presentation ranged from 1.5 mg/dl to 11.4 mg/dl (mean 5.5 mg/dl). The patients had a variety of underlying conditions or diseases. Aetiology was immune complex in 19 cases, anti-gbm antibody disease in two cases and pauci-immune (Wegener s granulomatosis) in one case. Among the immune complex group, seven patients had post-infectious glomerulonephritis (PIGN), three had mesangiocapillary glomerulonephritis (MCGN), three had IgA nephropathy and two had lupus nephritis (LN). The remaining four cases had other causes (Table 1). The percentage of crescents ranged from 50% to 100% (mean 70.6%). Eight patients had crescents involving 80% or more glomeruli. The majority (14/22, 64%) of the children had fibrocellular crescents. A large number of children (10/ 22, 45%) had sclerosis at presentation. All 22 children had interstitial fibrosis on histopathological examination (Table 2). Of the 22 children, 20 were treated with methyl prednisolone, oral steroid therapy; oral cyclophosphamide therapy. Plasmapheresis was done for children with anti-gbm disease and Wegener s granulomatosis. One patient received mycophenolate mofetil (MMF). The majority of children (15/22, 68%) required dialytic support (haemodialysis or acute peritoneal dialysis) at some time of their disease, while seven patients never received any form of renal replacement therapy (RRT). Of the 22 children, follow-up data were not available for two patients, while the mean follow-up period for the rest of the 20 children was 8.13 months (range 1 month to 43 month). Of these, one died during dialysis, another died after transplantation, three had persistent renal dysfunction (chronic kidney disease not requiring dialysis), and eight were dialysis dependent [two on continuous ambulatory peritoneal dialysis (CAPD) and six on haemodialysis]. Only seven (35%) had normalization of serum creatinine levels. Among eight patients with crescents in 80% or more glomeruli, four (50%) progressed to end-stage renal disease (ESRD) compared to four (28.57%) patients with crescents in <80% of glomeruli. Thus, at their last follow-up examination, ten children had stage 5 CKD and three had stage 4 CKD, while seven children had a calculated GFR of >60 ml/min per 1.73 m 2. Persistent proteinuria was seen on follow-up in the majority [13/20 (65%)] of patients. On Kaplan Meier survival analysis, the mean survival time was 14.44 months (95% confidence interval of 6.25 to 22.62) (Fig. 1). Discussion Crescentic glomerulonephritis (CsGN) remains an uncommon entity in children. The severity of the condition, the multiple aetiologies and the associated poor prognosis have led to the use of multiple treatment regimens. There have been no published series in the English literature over the past 10 years, although there are small case series [6]. At our centre this entity accounted for 5.1% of kidney biopsies in children done over the past 5 years. Our incidence is similar to that described by Srivastava et al. and much

392 Pediatr Nephrol (2008) 23:389 394 Table 1 Epidemiological, clinical and laboratory features S. No Age (years) Gender Duration of symptoms (months) Aetiology Creatinine at entry (mg%) No. of glomeruli with crescents/ total Followup (months) Treatment given Duration of immunosuppression Creatinine at last follow-up (mg%) Stage of CKD U. protein at last follow-up 1 15 M 4 PSGN 4.6 14/17 2 MP+OS+OE 1.5 3.8 4 3+ 2 7 M 0.5 IgA 6.3 08/10 19 MP+OS 12 4.5 5 4+ 3 8 F 0.667 PIGN 2.94 13/26 2.2 MP+OS+OE 02 0.84 2 25 mg% 4 11 F 1 MCGN 2.7 06/12 4 MP+OS+OE 03 1.1 2 3+ 5 17 F 0.5 MCGN 6.9 15/15 1 OS+OE 0.5 3.53 5 4+ 6 15 M 2 MesPGN 5.4 08/14 4 OS 03 1.33 2 115 mg/day 7 9 M 2.5 MPGN 1.5 10/19 3 MP+OS+OE 02 3.4 4 4+ 8 4 M 0.167 PIGN 7 20/22 43 MP+OS+OE 12 0.73 2 60 mg% 9 16 F 21 LN 2.4 07/13 1 MP+OS+OE 0.5 2.5 2 1000 mg% 10 6 M 1 Anti-GBM 7.3 18/18 1.5 OS+OE+PP 1.0 6.11 5 65 mg% 11 10 M 0.33 PSGN 9.2 13/25 24 MP+OS+OE 12 0.63 2 90 mg% MP+OS+OE followed by MMF 12 14 M 0.5 Anti-GBM 2 05/07 8 & ACEI 07 1.38 2 20 mg% 13 17 F 3 LN 4.8 09/13 9.8 MP+OS+OE 09 10 3 3+ 14 14 M 0.267 PIGN 9 06/11 14 OS 12 14 5 4+ 15 10 M 2 X 6.7 20/22 15 MP+OS+OE 12 16 5 4+ 16 16 F 0.5 PIGN 6.4 22/30 6 MP+OS 05 5.3 5 4+ 17 14 M 1.5 MesPGN 1.8 08/09 13 OS 12 2.9 4 1.625 18 7 F 2 PIGN 2.1 15/15 0 X X X 2 X 19 18 M 2 X 9.59 07/12 3.5 OS 03 7.8 5 4+ 20 16 M 6 X 7.1 08/14 0 MP+OS+OE X X 5 X 21 17 F 1 Wegener s 11.4 21/30 1 MP+OS+OE +PP 0.5 5.6 5 3+ 22 9 M 1 X 3.93 19/31 4 Dexa+OS+OE 03 6.5 5 3+ PSGN Post-streptococcal glomerulonephritis, PIGN post-infectious glomerulonephritis, MCGN mesangiocapillary glomerulonephritis, LN lupus nephritis, MesPGN mesangioproliferative glomerulonephritis, MPGN membranoproliferative glomerulonephritis, GBM glomerular basement membrane, MP methylprednisolone, OS orally steroids, OE orally cyclophosphamide, PP plasmapheresis, MMF mycophenolate mofetil, Dexa Dexamethasone, ACEI ACE inhibitor

Pediatr Nephrol (2008) 23:389 394 393 Table 2 Histopathological features (C cellular, FC fibrocellular, P partial, Cf circumferential, TI tubulo-interstitial, GBM glomerular basement membrane) Patient number Total of glomeruli Percentage of glomeruli with crescents Cellular fibrocellular Percentage of glomeruli with sclerosis TI involvement Partial/ circumferential Immunofluorescence 1. 17 82.35 C-FC 0 Y Cf Inadequate 2. 10 80 FC 0 Y P-Cf 3. 26 50 C 0 Y P-Cf IgG 2 3+, C3 2+ 4. 12 50 FC 0 Y C3 only 5. 15 100 C 0 Y P-Cf 6. 14 57.14 C 28.57 Y P-Cf 7. 19 52.63 C 21.05 Y P-Cf IgM 3+, C3 3+, IgG 2+ 8. 22 90.90 C-FC 9.09 Y P-Cf 9. 13 53.84 C 0 Y IgG 2+, IgA +, C3 3+, C1q + 10. 18 100 C-FC 0 Y Cf 11. 25 52 FC 0 Y P-Cf Coarse granular along GBM IgM & C3 2+, IgG 1+ 12. 7 71.42 C 0 Y Cf 13. 13 69.23 FC 30.76 Y Cf 14. 11 54.54 C-FC 45.45 Y P-Cf 15. 22 90.90 Hyalinised 90.90 Y 16. 30 73.33 C-FC 20 Y P-Cf 17. 9 88.88 C-FC Y 18. 15 100 C Y P-Cf 19. 12 58.33 FC-F 75 Y 20. 14 57.14 FC-F 35.71 Y 21. 30 70 C Y 22. 31 61.29 C-FC 25.80 Y C3 2+, IgG 4+ lower than that reported by Miller et al., who observed crescentic lesions in 56/372 patients (15.05%) [5, 7]. A notable feature in our study was a delayed diagnosis and delayed referral, as observed by the mean duration of symptoms prior to referral of 2.47 months. Hence, the Cumulative Survival 1.2 1.0.8.6.4.2 0 10 Survival Function 20 30 Follow up in months Fig. 1 Kaplan Meier survival analysis showing mean renal survival time (months) during the follow-up period 40 50 Not recovered majority of our patients had severe disease at presentation, as evidenced by oliguria in 19/22 (86.4%), and 15/22 (68%) were dialysis dependent at presentation. Late referral may be related to the fact that our institute is a tertiary care hospital. Such patients are first treated by the primary care physicians; subsequently, they go to a higher centre to be finally referred to our institute. Hence, a substantial time elapses before they reach us. The prevalence of renal failure at presentation has been reported to be 50 70% in previous studies [2 5]. However, the duration of symptoms has not been correlated with the severity in most of the previous studies. In our study this high incidence of advanced renal failure requiring dialysis was also histopathologically correlated by the high frequency of chronicity changes seen on kidney biopsy, as evidenced by the presence of fibrocellular crescents in the majority of the patients [(14/ 22, 64%)] and the presence of glomerular sclerosis in a significant number of children [(10/22, 45%)] at presentation. These findings were almost similar to those in the study by the Southwest Pediatric Nephrology Study Group, where it was observed that only 37% of the crescents were cellular, while the majority (63%) were fibrous/fibrocellular crescents [3].

394 Pediatr Nephrol (2008) 23:389 394 Most series in children suffer from having used different inclusion criteria and from being retrospective and lacking immunofluorescence studies on all patients; therefore, the precision of the diagnoses is somewhat suspect. In our experience immune complex glomerulonephritis was the commonest aetiologic category (86.4%), and the majority of these children had post-infectious glomerulonephritis. Amongst the immune complex group, mesangiocapillary glomerulonephritis continues to be an important cause of CsGN. The aetiology in our study is in concordance with the findings by other authors, wherein immune-complex glomerulonephritis accounted for the majority of the cases [4, 5]. In contrast, Miller et al. observed that most children in their series had well-defined systemic or primary glomerular disease, and only two of the 56 patients with a crescentic lesion had disease that had a definite poststreptococcal aetiology [7]. Similarly, post-streptococcal GN (PSGN) accounted for 12% of CsGN in another multicentre collaborative series [3]. Thus, post-streptococcal glomerulonephritis continues to be the commonest aetiology in developing countries like ours. However, in an earlier series of 22 children with rapidly progressive glomerulonephritis (RPGN) and diffuse crescents in 80% to 100% of glomeruli, nine cases were associated with evidence of systemic disease, seven were considered postinfectious, and six were presumed to be idiopathic, one with anti-gbm antibody [13]. In our study, seven (35%) patients had a complete recovery from renal failure, and persistent proteinuria was seen in a majority [13/20, (65%)] of the patients on followup. The poor outcome in children with PSGN with crescentic transformation has been reported previously [2, 14]. Previous studies also have revealed varying rates of complete recovery in a range of 14 50% of children [3 5]. However, all these studies were conducted more than 10 years ago, and there was a significant heterogeneity in the treatment protocols. El-Husseini et al. reported a series composed of 128 patients with CsGN, followed up for 10 years (mean 34±28 months); at the last follow-up, 44 (34.4%) had normal kidney function, 29 (22.7%) had renal impairment, 33 (25.8%) had chronic renal failure, and 22 (17.2%) had died, highlighting the poor prognosis of CsGN [15]. However, their study was not of children, and the mean age of their subjects was 22.7±14 years. Over the years there has been a broad consensus over the use of pulse methyl prednisolone therapy in CsGN and also significant advancements in the supportive therapy. Our study involved the use of a common therapeutic protocol. Our results are better than those reported previously in another study from our country, with one-third of our patients showing recovery of renal functions [5]. The majority of the children had evidence of chronic kidney disease. Thus, in our study, immune complex glomerulonephritis was the commonest aetiology of crescentic glomerulonephritis in children. However, the outcome of crescentic GN in children continues to be poor, due to delayed referral and delayed diagnosis. This was correlated histologically by the presence of fibrocellular crescents in the majority of our patients. Thus, crescentic GN should be treated as a renal emergency. A greater awareness of this disease needs to be created amongst the referring paediatricians to facilitate early diagnosis and prompt treatment. References 1. Bidani AK (1992) Idiopathic rapidly progressive glomerulonephritis and Goodpasture s syndrome. In: Edelmann CM Jr (ed) Pediatric kidney disease, 2nd edn. Little, Brown & Company, New York, pp 1223 1245 2. Roy S 3rd, Murphy WM, Arant BS Jr (1981) Poststreptococcal crescentic glomerulonephritis in children: comparison of quintuple therapy versus supportive care. J Pediatr 98:403 410 3. Southwest Pediatric Nephrology Study Group (1985) A clinicopathologic study of crescentic glomerulonephritis in children. A report of the Southwest Pediatric Nephrology Study Group. Kidney Int 27:450 458 4. Jardim HM, Leake J, Risdon RA, Barratt TM, Dillon MJ (1992) Crescentic glomerulonephritis in children. Pediatr Nephrol 6:231 235 5. Srivastava RN, Moudgil A, Bagga A, Vasudev AS, Bhuyan UN, Sundraem KR (1992) Crescentic glomerulonephritis in children: a review of 43 cases. Am J Nephrol 12:155 161 6. Dilima MG, Adhikari M, Coovadia HM (1981) Rapidly progressive glomerulonephritis in black children. A report of 4 cases. S Afr Med J 60:829 832 7. Miller MN, Baumal R, Poucell S, Steele BT (1984) Incidence and prognostic importance of glomerular crescents in renal disease of childhood. Am J Nephrol 4:244 247 8. Arieff AI, Pinggera WF (1972) Rapidly progressive glomerulonephritis treated with anticoagulants. Arch Inter Med 129:77 84 9. Cunningham RJ 3rd, Gilfoil M, Cavallo T, Brouhard BH, Travis LB, Berger M, Petrusick T (1980) Rapidly progressive glomerulonephritis in children: a report of 13 cases and a review of literature. Pediatr Res 14:128 132 10. Churg J, Sobin LH (1981) Renal disease: classification and atlas of glomerular disease. In: World Health Organization (WHO) (ed) Monograph, New York, Igaku-Shoin Medical Publishers, Tokyo, pp 3 9 11. National Heart, Lung and Blood Institute (1987) Report of the second task force on blood pressure control in children. Pediatrics 79:1 25 12. The K/DOQI Clinical Practice Guidelines on Chronic Kidney Disease Work Group (2002) K/DOQI clinical practice guidelines for chronic kidney disease. Evaluation, classification and stratification. Am J Kidney Dis 39:S1 S266 13. Habib R (1974) The major syndromes. In: Royer P, Habib R, Mathieu H (eds) Pediatric nephrology. Saunders, Philadelphia, pp 246 262 14. Vijayakumar M (2002) Acute and crescentic glomerulonephritis. Indian J Pediatr 69:1071 1075 15. El-Husseini AA, Moustafa FE, Fouda MA, Sobh MA (2002) Crescentic glomerulonephritis in Egypt: clinical and histopathological risk factors. J Nephrol 15:281 289