Hepatitis B virus-associated membranous nephropathy.
Glomerulonephritis as an extra hepatic manifestation of chronic HBV infection has now been well documented [1,2,3,4,5]. HBV-associated nephropathy has been described in areas of both high and low endemicity . In Africa HBV-associated nephropathy has been reported from the southern, central and northern regions [7,8,9,10,11]. In the southern African continent the prevalence of HBV-associated nephropathy appears to be higher than the rest of the continent . In KwaZulu/Natal, South Africa, the prevalence of hepatitis B surface antigenaemia (HbsAg) in urban, rural and institutionalised children was reported to be 6.3%, 18.5% and 35.4% and the HBV exposure rates, as shown by the presence of any marker of HBV infection, 19.5%, 65.1% and 70.1% respectively amongst black children . Prior experience of nephrotic syndrome (NS) and its association with HBV in black children, already published in a series of reports, showed HBV-associated nephropathy to be the commonest form of nephrotic syndrome among black patients in KwaZulu/Natal; membranous nephropathy (MN) being the commonest histological type reported [7,14]. The only other large series of HBV-associated nephropathy in southern Africa was from Cape Town of a large cohort of children, mainly of mixed ancestory (coloured), with a small number of black children . There have been no other large studies of this condition amongst black children in Africa. We therefore undertook a series of studies to delineate the spectrum of this disease in black children with regard to the following: clinical presentation, laboratory findings, natural history, biosocial background, genetics (using HLA Class I and II antigens) as well as the impact of treatment and prevention by immunisation. We commenced these studies by reviewing our 20-year experience of 636 children with NS in Durban, South Africa for the period 1976- - 1995. Three hundred and six (48.2%) were blacks, 307 (48.2%) Indians and 23 (3.6%) were a mixed group (coloured); 91 (14.3%) could not be categorised and were excluded from the analysis. In black children, membranous nephropathy accounted for 43% of all cases of NS; 86.2% of these 306 children were associated with hepatitis B virus antigens . This contrasts with the 2% - 5% prevalence of idiopathic membranous nephropathy reported in western countries . We then proceeded to document the clinical features of this disease in black children. One hundred and thirty-three children with NS positive for HBV carriage were studied. In 70 patients the histological type was membranous; 46 of these 70 patients were followed up for a mean of 3.4 years (range 1-11). Spontaneous elimination of both HBsAg and HBeAg occurred in 10 (21.7%) of the 46 patients; 16 (34.8%) cleared HBeAg alone. Co-existing liver disease occurred in 18 (25.7%); hypocomplementaemia (low C3 and C4) in 22 (47.8%) and 5 (10.9%) of these 46 children respectively. Sixty-five (92.9%) of the 70 patients had normal renal function; 1(1.4%) impaired renal function; 3 (4.3%) chronic renal insufficiency and 1(1.4%) end stage renal disease at last hospital visit. Twelve (17.1%) of the 70 patients were in remission; all having cleared HBeAg. HBVMN was clinically indistinguishable from 24 children with idiopathic MN although biochemical characteristics were different. There were 23 patients with histological lesions other than MN. Forty patients with clinical, biochemical and serological findings similar to those with HBVMN and the other histological types, were unbiopsied. This report delineates the natural history of HBV infection in black South African children with NS, the majority of whom have MN. Disease remission in HBVMN parallels elimination of HBV antigens, particularly HBeAg. Comparison of HBVMN with idiopathic MN revealed clinically indistinguishable characteristics but unexplained biochemical differences . Little is understood of the biosocial context in which HBV-associated nephropathy (particularly MN) develops. In the next two studies we evaluated HBV status and proteinuria in family members and household contacts of index children with HBVMN to test the hypothesis that HBV carriage and asymptomatic proteinuria are closely linked and may be causally associated. In the first of these two studies, thirty-one black children with biopsy-proven HBVMN were the index cases. One hundred and fifty-two family members and 43 black household contacts were the subjects of this study. We assessed HBV carrier status by testing for HBV antigens and antibodies using enzyme-linked immunosorbent assays (ELISA) and for HBV DNA by using slot-blot hybridisation and nested polymerase chain reaction. Sequencing of the precore HBV region of HBV was done in a subset of both index cases and subjects. Proteinuria was assessed by measuring the urinary protein: creatinine ratio. Seventy-two (37%) of the 197 family members and household contacts were HBV carriers, and 53 (27%) had a protein: creatinine ratio greater than the physiological limit (protein: creatinine ratio <0.2). Abnormal proteinuria was defined by a protein: creatinine ratio 0.2. Continuous data was compared using analysis of variance. Categorical data were compared using Chi-square test or Fisher’s exact test where appropriate. A probability of <0.05 was considered significant. The frequency of abnormal proteinuria was not significantly different in those with [22 (30.5%) of 72] or without [33 (32%) of 104] HBV carriage. This lack of association remained when carriers were classified into those who were HBsAg positive only and those with active viral replication (HBsAg and/or HBeAg and/or HBV DNA; p = 0.01). Family members were more predisposed to HBV carriage than household contacts, but abnormal proteinuria was present with equal frequency (p = 0.48). Age had a significant impact on proteinuria, with children less than five years being more likely to have abnormal proteinuria (p = 0.008). The prevalence of abnormal proteinuria in family members and household contacts of the index cases was more than in community-based controls. The 10 index HBVMN cases and 14 family members and household contacts that were tested all had HBV of genotype A. The results suggest that the family members and household contacts of children with HBVMN are at very high risk of HBV carriage; they also have asymptomatic proteinuria at a significantly higher rate than community-based controls. The HBV carrier status was not associated with proteinuria. This lack of association was a finding supported by peak prevalences of proteinuria in those under five years but no corresponding peak of HBV carriage. Proteinuria may indicate glomerular basement membrane dysfunction. Environmental and social factors may underpin development of these two disorders, but are insufficient to account for the index cases of HBVMN. The emergence of children with HBVMN from such households additionally depends on unidentified and possibly genetic factors . In the second study of the biosocial background in which the HBV carrier-state with MN develops, we used the same subjects. One hundred and twenty-three unrelated individuals from the communities of the index cases, negative for HBV, served as controls. In this study, proteinuria was assessed using sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and protein: creatinine ratios. Patterns of proteinuria on SDS-PAGE were classified as glomerular, tubular or mixed; IgG and haptoglobulin were suggestive of MN. Seventy-two (36.9%) of the 195 family members and household contacts were HBV carriers; 21 (29.2%) of these carriers had evidence of proteinuria using SDS-PAGE. Twenty-eight (41.2%) of the sixty-eight members of the study group who were HBV negative and 26.8% of the controls also showed proteinuria on SDS-PAGE. This lack of association between HBV carriage and proteinuria remained when controlled for gender and family relationship. Also, HBV was not protective against the development of proteinuria. Age was associated directly with a glomerular pattern of proteinuria (p = 0.007). Those having a pattern of proteinuria suggestive of MN were more likely to have an abnormal protein: creatinine ratio (p = 0.001). Ten (59%) subjects with a membranous pattern of proteinuria and 19 (47.5%) with a non-membranous pattern of proteinuria had microscopic haematuria. Such a pattern of proteinuria was not significantly different between subjects and community based controls (8.7% vs. 6.5%, p = 0.5). Environmental exposures in these subjects may be responsible for the proteinuria, which probably reflects underlying glomerular basement membrane damage. Discordance between the HBV carrier-state and patterns of proteinuria in the study group suggest that interaction between specifically vulnerable individuals and HBV group suggest HBV and MN may not be causally related or that it reflects exceptional interaction between specifically vulnerable individuals and HBV . From the above two studies we inferred that the pathogenetic mechanisms by which individuals with chronic HBV infection develop MN are probably dependent on interactions between viral, host and environmental factors; some evidence suggests a genetic predisposition. We therefore undertook another two studies to explore HLA associations in black children with HBVMN. In the first of these two studies, thirty black children, age range 2 to 16 years, with biospy-proven HBVMN, were the subjects of this study. HLA A, B and C antigens were determined using a two-stage lymphocytotoxic test. HLA DRB1* and DQB1* typing was done using sequence-specific primers. HLA class I and II antigen frequencies of the study subjects were compared to controls that were randomly chosen healthy blood donors from the same population. HLA DQB1*0603 was increased in patients with HBVMN compared to controls (chi-square 13.65, RR 4.3). DRB1*07 and DQB1*02 were increased in frequency in the study subjects but failed to reach statistical significance. There was no significant difference in the frequencies of class I antigens in the study group compared to controls. This study is the first report of HLA associations in black patients with HBVMN in whom Class I and II antigens were determined using molecular methodology. It shows a high frequency of DQB1*0603 in black children with HBVMN compared to controls suggesting a possible genetic predisposition to the development of HBVMN . Following our findings of an HLA Class II association in black children with HBVMN, we proceeded to determine if HLA DQB1*0603 predisposes to HBV carriage and development of abnormal proteinuria in the second study. We studied 70 family members of 14 children with HBVMN positive for HLA DQB1*0603 selected from the first study. Associations of HLA DQB1*0603 to HBV carriage and abnormal proteinuria were determined using the mean probability ratio (LOD scores). Forty-seven of the 70 (67%) family members were positive for HBV infection. Nineteen (27%) had abnormal range proteinuria. LOD scores in the study subjects with DQB1*0603 who were HBV negative vs. those with DQB1*0603 who were HBV positive was not significant (anti-log sum = 2.0559 and average 0.23). When a similar calculation was done for abnormal proteinuria, there were no significant findings (anti-log sum = 3.8587 and average 0.43). This lack of association between HLA DQB1*0603 with either HBV carriage or abnormal proteinuria in family members suggests that additional factors may play a role in predisposing children to chronic HBV carriage and the development of MN. We therefore conclude that the main effect of HLA DQB1*0603 which distinguishes HBVMN from family members is the degree of proteinuria which is a reflection of the severity of glomerular basement membrane damage in the latter . In the next study we proceed to investigate the efficacy of Interferon alpha 2b (INTRON A ®) in the treatment of HBV-associated nephropathy in black children. Twenty-four black children with biopsy-proven HBV-associated nephropathy were recruited into the study during the period April 1997 to June 1999. Five defaulted treatment and were excluded from the primary analysis. IFN 2b was administered for 16 weeks. Response to treatment was defined as loss of HBeAg, decrease in proteinuria, and prevention of deterioration in renal and liver function. A control group of 20 patients was followed up for the same period. Ten (52.6%) of the treated children responded with clearance of HBeAg by 40 weeks. None cleared HBsAg. All responders showed remission of proteinuria, 90% maintained normal renal function and 1 (10%) showed improvement of renal function. HBV DNA levels decreased in this group. Nine patients did not clear HBeAg; none showed remission of proteinuria, 2 showed deterioration of renal function. Liver enzymes rose during treatment but subsequently declined irrespective of response to therapy. No serious side effects were encountered. Only 5% of controls showed spontaneous clearance of HBeAg, and none had remission of proteinuria. Black children with HBV-associated nephropathy show accelerated clearance of HBeAg with remission of proteinuria following treatment with IFN 2b. IFN 2b was well-tolerated . We then went on to investigate the impact of HBV vaccination in South Africa over 6 years on HBV-associated MN. HBV vaccine has resulted in a decline in the incidence of HBV carriage and hepatocellular carcinoma in South East Asia. Vaccine efficacy in Africa has not been adequately assessed. King Edward VIII Hospital, Durban, South Africa, is the only tertiary referral centre for the province of KwaZulu/Natal for children with renal diseases. HBV vaccine was introduced into the Extended Programme on Immunisation (EPI) in April 1995; vaccine coverage rates between 1995-2001 for children for the first, second and third doses were 85.4%, 78.2% and 62.0% respectively. HBV status was determined using radioimmunoassay (1984 – 1991) or ELISA. MN was confirmed on renal biopsy. The hospital average annual incidence of HBVMN was compared pre and post-vaccination, and according to age groups. Between 1984 and 2001 there were 119 children with HBVMN; the mean age was 7 years (range 1 to 14 years) and 101(85%) were males. The average annual rate ratio (aRR) per 105 child population was 0.25. The aRR of 0.03 for the years 2000-2001, was significantly lower than the aRR of 0.22 during the pre-immunisation period (1984 – 1994) [p = 0.003; RR = 0.12 (95% CI: 0.03 – 0.5)]. The aRR in 2000-2001 for children 0 – 4 years (0.00) and 5 – 10 years (0.09) were significantly lower than in the pre-vaccination years (0.16 and 0.46, p = 0.01 and 0.02 respectively). Thus, HBV vaccine, even at low coverage for the full EPI schedule, reduced the hospital incidence of HBVMN by six years . From this series of studies we concluded that prior to the introduction of the HBV vaccine into the Expanded Programme on Immunisation in Children, HBV-associated nephropathy, particularly MN was the commonest form of NS in black children. Several studies have suggested on the basis of epidemiological, clinical and immunological evidence a causal association between chronic HBV carriage and the development of nephropathy. In our present series of studies we have findings that lend further support to the causal association between HBV carriage and development of nephropathy, particularly MN, in black children. We have shown that genetic and other environmental factors may also play a role in determining the degree of proteinuria. Those children with abnormal range proteinuria less than the nephrotic range show no association with HBV carriage or genetic factors with regard to HLA linkage. The efficacy of interferon treatment in elimination of the HBV and abrogation of proteinuria following clearance of the virus (particularly the HBeAg) as well as the impact of routine HBV immunisation in preventing HBV carriage and subsequent development of nephropathy lends further support to our findings. The impact of viral load has yet to be investigated.