In the present study we examine whether hepatitis C virus (HCV) infection status influences glomerular pathologic findings in renal allografts and its effect on graft outcome. Renal allograft biopsies performed between January 1991 and June 1999 were considered. Exclusion criteria were insufficient sample, unknown HCV serological status at time of biopsy and final diagnosis of acute rejection. Light microscopy and immunofluorescence studies were performed on all biopsies. According to a predefined protocol, electron microscopy was carried out. Of 138 eligible renal allograft biopsies, 42 fulfilled at least one exclusion criterion. Of 96 biopsies selected for the study, 44 (45.8%) were from HCV-positive and 52 from HCV-negative recipients. Renal biopsy was performed 74 ± 55 and 60 ± 39 months after transplantation in HCV-positive and HCV-negative groups, respectively (p = 0.12). Of 44 HCV-positive biopsies, 20 (45.4%) showed membranoproliferative glomerulonephritis (MPGN) (16 type I and 4 type III). Conversely, in HCV-negative biopsies there were only three cases of MPGN (2 type I and 1 type III). De novo membranous GN (MGN) was diagnosed in 8/44 (18.2%) HCV-positive and in 4/52 (7.7%) HCV-negative cases. The prevalence of chronic transplant glomerulopathy was similar in HCV-positive and HCV-negative groups (11.4% and 11.5%, respectively). The prognosis of de novo GN (either MPGN or MGN) was worse in HCV-positive than in HCV-negative recipients (relative risk 4.89; 95% confidence interval, 1.15–20.69; p = 0.03). By multivariate analysis, HCV-positive serology infection was the only independent predictor of graft loss (relative risk 2.64; 95% confidence interval, 1.35–5.17; p = 0.005). In diagnostic renal allograft biopsies the presence of de novo immune-mediated glomerulonephritis, especially type I MPGN, is strongly associated with HCV infection and results in accelerated loss of the graft.
Chronic hepatitis C virus (HCV) infection has been associated with glomerular disease in native and transplanted kidneys (1,2). The high prevalence of HCV infection in renal allograft recipients (3,4) may put this population at special risk for developing immune-mediated glomerular lesions. Recent data provided by Hestin et al. (5) indirectly support the involvement of HCV infection in the glomerular pathology of renal allografts. Thus, they reported that pretransplant HCV-positive serology was a major risk factor of proteinuria and graft loss after renal transplantation.
Some authors have observed that HCV infection predisposes patients to acute glomerular lesions on the allograft. Cosio et al. (6) reported a high incidence (around 50%) of acute transplant glomerulopathy in HCV-positive transplanted patients and Baid et al. (7) described the association of HCV infection with pretransplant anticardiolipin antibodies and de novo renal thrombotic microangiopathy. Nevertheless, de novo membranoproliferative glomerulonephritis (MPGN), with or without mixed cryoglobulinemia (8,9), is the most consistent renal lesion associated with chronic HCV infection in renal allografts. HCV-associated MPGN must be differentiated from chronic transplant glomerulopathy (10), which is a sign of chronic rejection-related glomerular damage. Both induce proteinuria and sometimes may appear similar under light microscopy (double contours) and immunofluorescence examination (weak parietal focal and segmental IgM and C3) of the biopsy (8). In this situation differential diagnosis requires electron microscopy, which shows subendothelial electron-dense deposits in MPGN, whereas in transplant glomerulopathy it shows only thickening and duplication of glomerular basement membranes. On the other hand, some authors found de novo membranous glomerulopathy (MGN) to be the most frequent glomerular lesion associated with HCV infection (11). Others (12) reported that the prevalence of de novo MGN was similar in HCV-infected and noninfected kidney allografts. These discordant results reflect the scarcity of systematic pathologic studies on chronic glomerular lesions in the renal transplant population.
Here we analyze the glomerular lesions in renal allograft biopsies following a systematic procedure that included light microscopy, immunofluorescence and electron microscopy. We compare the renal pathologic findings in HCV-positive and HCV-negative transplant recipients and assess whether HCV infection influenced graft outcome.
Materials and Methods
Biopsies carried out on kidney allografts starting from the third month after transplantation and performed in our Institution between January 1991 to June 1999 were identified. We excluded those biopsies done in the first 3 months after transplantation in order to eliminate diagnostic confounding variables such as acute rejection. Biopsies were clinically indicated when proteinuria was higher that 1 g/d, with persistent microhematuria and serum creatinine deterioration. During this period, our standard procedure for diagnostic renal allograft biopsy included oral informed consent.
Two cylinders of renal tissue were obtained, divided and processed for light microscopy, immunofluorescence and electron microscopy when needed. Two observers, who were blinded as to the HCV status of the patient, evaluated renal lesions according to the Banff schema (13,14) with some modifications, since our major objective was to diagnose glomerular lesions. Thus, according to a predefined protocol (Figure 1), the sample was considered insufficient to evaluate glomerular abnormalities either when it included less than 3 noncompletely sclerosed glomeruli for light microscopy or no glomeruli in the frozen tissue for immunofluorescence or, when required, there was no glomerulus for electron microscopy. Biopsies included in this study were examined by light microscopy (hematoxylin-eosin, periodic acid-Schiff, periodic acid-methenamine silver) and direct immunofluorescence (IgG, IgM, IgA, C3; Dako, Denmark). Electron microscopy examination was performed when either diffuse or focal glomerular parietal immunoglobulin or complement deposition was present, even in cases showing a weak pattern. So, diagnostic biopsies re-examined following the procedure summarized in Figure 1 were classified as chronic allograft nephropathy, chronic transplant glomerulopathy, de novo MPGN, de novo MGN and recurrent primary renal disease. Biopsies with changes of acute rejection, with insufficient sample and performed in recipients with unknown HCV serological status or with hepatitis B coinfection (positive HBsAg) were excluded from the study.
Clinical information on the patient at the time of the biopsy as well as follow-up data were obtained from our transplant database (FileMaker Pro 3.0, Claris Corp, CA, USA) and clinical histories. According to the results of HCV serologic testing of patients, biopsies were classified as HCV positive or HCV negative. The diagnostic test used was second- or third- (since 1993) generation enzyme immunosorbent assay (EIA) (Abbott Labs, IL, USA), followed since 1993 by confirmatory Abbott EIA supplementary assay and since 1996 by serum RNA-HCV determination (Amplicor-Monitor, Roche, Switzerland), respectively. Clinically relevant variables such as primary renal disease, sex, age, transplant number, peak of panel-reactive antibodies, number of HLA-mismatches, donor age, induction therapy with antilymphocyte antibodies, cold ischemia time, delayed graft function (defined as dialysis requirements in the first week after transplantation without evidence of rejection or other cause of allograft dysfunction), acute rejection, serum creatinine and proteinuria at 1 year after transplantation were considered for this analysis.
Groups were compared using the chi-square test for categorical variables, the t-test for normally distributed data and the Mann–Whitney U-test for non-normally distributed variables. The overall graft survival was studied using Kaplan–Meier survival analysis, and differences between the groups were established using the log-rank test. Cox proportional hazards analysis was used to estimate the prognostic effect of certain variables. Variables associated with graft survival were further studied by Cox's analysis. Since the aim of the survival analysis was to assess the influence of HCV infection and histological lesions on graft survival rather than on patient survival, patients who died with a functioning graft were censored.
Numerical results were expressed as mean ± SD. A two-tailed p-value of less than 0.05 was considered significant. Analyses were carried out using Statistica™ 4.1 (Statsoft Inc., Tulsa, OK, USA) and Statview 4.51 (Abacus Concepts Inc. CA, USA) software for Macintosh.
Patient baseline characteristics
A total of 138 diagnostic kidney allograft biopsies were eligible for this study. After clinical and pathologic review, 42 were excluded. Reasons for exclusion were presence of HBsAg in serum (2), histology of grade I acute rejection (9), insufficient sample (18) and unknown HCV serological status (13). Ninety-six biopsies from 93 patients were available for the study. There were three patients (two HCV positive and one HCV negative) with a diagnostic biopsy from their first and second graft.
Immunosuppression therapy in these patients was based on cyclosporine. Of the 96 allografts, 44 were HCV positive (45.8%) and 52 HCV negative. Clinical and analytical baseline characteristics are summarized in Table 1. For most patients there was no consistent information about pretransplant blood transfusion. In all the cases there was at least one HCV serologic screening between renal transplantation and allograft biopsy and prebiopsy HCV serological status did not change in any patient. As shown in Table 1, there were no differences between HCV-positive and -negative groups in cause of end-stage renal disease, gender, donor and recipient age, number of HLA mismatches, utilization of antilymphocyte antibodies, cold ischemia, delayed graft function and acute rejection. In contrast, there were significant differences between HCV-positive and -negative groups in peak value of panel-reactive antibodies and the average of first/second transplants. At 1 year after kidney transplantation, serum creatinine concentrations (170 ± 77 in HCV positive vs. 167 ± 76 µmol/L in HCV negative) and proteinuria (0.8 ± 0.8 in HCV positive vs. 0.9 ± 1.5 g/d in HCV negative) were similar in both groups.
Table 1. : Baseline characteristics of HCV-positive and HCV-negative patients
HCV+ n = 44
HCV– n = 52
Abbreviations: KT, number of kidney transplant; PRA, peak of panel-reactive antibodies; Antilymphocyte Ab, antibodies used as immunosuppressive induction therapy; CIT, cold ischemia time; DGF, delayed graft function.
Primary renal disease
38 ± 13
40 ± 14
23 ± 35
7 ± 17
HLA mismatch (A + B + DR)
2.8 ± 1.3
3.0 ± 1.0
Antilymphocyte Ab (yes/no)
Donor age (years)
36 ± 15
35 ± 15
23 ± 4
23 ± 7
Acute rejection (yes/no)
Clinical and analytical data at the time of renal allograft biopsy
Diagnostic renal allograft biopsies were performed several years after kidney transplantation (Table 2). Non-nephrotic proteinuria (1–3 g/d) was an indicator for these allograft biopsies in both groups. Between HCV-positive and -negative groups there were clinically relevant differences. Indeed, 16/44 HCV-positive recipients showed nephrotic syndrome (proteinuria > 3 g/d and serum albumin < 35 g/L) and in 7/44 the reason for performing the biopsy was a gradual deterioration in serum creatinine. In contrast, in HCV-negative recipients, only 6/52 had nephrotic syndrome and in 24/52 the indication for biopsy was serum creatinine deterioration. As observed in Table 2, proteinuria was higher and serum albumin lower in the HCV-positive than in the HCV-negative group. Serum creatinine at the time of biopsy was similar in both groups. The presence of persistent microhematuria in the urinary sediment was more frequent in HCV-positive (73%) than in HCV-negative (35%) recipients.
Table 2. : Clinical data of HCV-positive and HCV-negative patients
HCV + n = 44
HCV – n = 52
Months after transplant
74 ± 55
60 ± 39
Reason for biopsy
Serum creatinine (µmol/L)
216 ± 91
236 ± 90
3.2 ± 2.5
2.1 ± 2.3
Serum albumin (g/L)
36.4 ± 5.8
40.5 ± 4.9
Pathological results are summarized in Table 3 and illustrated in Figures 2 and 3. There were some cases of recurrence of the primary renal disease, mainly MPGN and focal and segmental glomerular sclerosis in both HCV-positive and -negative groups. According to the Banff criteria for evaluation of renal allograft biopsies, all cases had changes of chronic allograft nephropathy, HCV-positive and HCV-negative groups having similar and moderate degrees of interstitial fibrosis, vascular hyalinosis and intimal fibrosis (data not shown). In Table 3, samples without glomerular lesions were classified as isolated chronic allograft nephropathy. On the other hand, there were biopsies with glomerular lesions, which were classified as transplant glomerulopathy, de novo MPGN or MGN. Thus, de novo MGN, and especially MPGN with subendothelial electron dense deposits (type I), were mainly found in HCV-infected patients. There were few cases with concomitant subendothelial and subepithelial electron dense deposits (type III MPGN). Among those biopsies classified as MPGN, there were three, one HCV-negative and two positive, superimposed chronic transplant glomerulopathy abnormalities. Nevertheless, the occurrence of chronic transplant glomerulopathy in our study population was similar in HCV-positive and -negative allografts.
Table 3. : Renal pathology in HCV-positive and HCV-negative renal allografts
HCV + n = 44
HCV – n = 52
Abbreviations: CAN, chronic allograft nephropathy; TxGP, chronic transplant glomerulopathy; MPGN, membranoproliferative glomerulonephritis (type I and type III); MGN, membranous glomerulonephritis; PRD, primary renal disease. Recurrences of PRD were type I MPGN (1) and IgA nephropathy (1) in the HCV-positive group, and type I MPGN (2), type II primary crescentic glomerulonephritis (1), focal and segmental glomerular sclerosis (3), MGN (1) and IgA nephropathy (1) in the HCV-negative group.
De novo MPGN
20 (45.4%) 16 (I), 4 (III)
3 (5.8%) 2 (I); 1 (III)
De novo MGN
Hepatitis C virus, renal pathology and graft survival
The duration of follow-up after the biopsy (the follow-up finished when the patient suffered graft loss or, if the graft was functioning, on December 31, 1999) was 25.3 ± 20.2 months in HCV-positive and 18.3 ± 15.7 in HCV-negative patients, respectively (p = 0.07). During the follow-up, 34 (77.3%) HCV-positive and 22 (42.3%) HCV-negative cases developed end-stage renal disease and returned to dialysis (p = 0.001). Four patients from the HCV-negative group died with a functioning allograft (one because of a bladder carcinoma and three because of cardiovascular events), whereas no deaths with a functioning allograft were observed in the HCV-positive group. There were no cases of death or graft loss attributed to liver disease in any of the study groups.
HCV serological status at the time of biopsy as a predictor of allograft survival was examined with the log-rank test method and plotted using Kaplan–Meier estimates (Figure 4). The product limit estimated for the median renal survival was 40.4 and 15.9 months in HCV-negative and HCV-positive patients, respectively (p < 0.001). Among diagnostic biopsies performed in HCV-positive, as compared with those of HCV-negative patients, the relative risk of graft loss was 3.06 (95% confidence interval, 1.72–5.44; p < 0.001).
We also analyzed whether renal pathology influenced graft outcome (Figure 5, panel A). We found that graft survival depended on renal pathology (p = 0.008). By the log-rank test method, diagnostics associated with poor graft survival were de novo GN (p = 0.008) and transplant glomerulopathy (p = 0.02). For relative risks and confidence intervals, see Table 4. We further analyzed this issue by stratifying de novo GN and chronic allograft nephropathy groups according to HCV serology (Figure 5, panel B). Diagnosis of de novo GN in HCV-positive patients was associated with a very poor graft outcome, even worse than de novo GN in HCV-negative (p = 0.03 by the log-rank test method).
Table 4. : Results of univariate and multivariate analysis
Univariate analysis RR (95% CI)
Multivariate analysis RR (95% CI)
Abbreviations: RR and CI denote relative risk and confidence interval, respectively. De novo GN means de novo glomerulonephritis (either MPGN or MGN). For other abbreviations, see Table 3. HCV positive, proteinuria at the time of the biopsy, de novo GN and TxGP were predictors of graft survival in univariate analysis. p-values in the table are from multivariate analysis.
De novo GN
Among the other variables, only proteinuria at the time of the biopsy was also associated with a poorer renal allograft survival by univariate analysis (Table 4). Other variables, such as kidney transplant number, panel-reactive antibodies, serum albumin and microhematuria, were not predictors of graft outcome. Multivariate Cox's regression analysis showed that HCV infection was the only independent predictor of graft survival in diagnostic renal allograft biopsies (Table 4).
In this study we examined the association of HCV serological status and the development of de novo GN on renal allografts. Sixty-three percent of diagnostic renal allograft biopsies performed in HCV-positive recipients showed pathologic findings of de novo GN (45% MPGN and 18% MGN). Conversely, in HCV-negative patients, de novo GN, either MPGN or MGN, was seldom encountered.
Several reports suggest that HCV infection is associated with immune-mediated glomerular lesions in renal allografts (8–10) as well as in liver transplant recipients (15–17). However, in those studies, particularly in the renal transplant setting, major concerns are the few cases described, the absence of HCV-negative controls and the lack of follow-up. The present study was designed to overcome these issues. First, irrespective of the anti-HCV status, we re-examined diagnostic biopsies performed in our Institution since 1991. Second, we excluded cases with confounding variables such as coinfection with hepatitis B virus or acute rejection in the biopsy. Third, considering both the limitations of the Banff evaluation for diagnosing glomerular abnormalities and the particularities of glomerular pathology in renal allografts, a systematic procedure was used, which included electron microscopy to differentiate chronic transplant glomerulopathy from de novo MPGN (10). Once a consistent diagnosis was achieved, we evaluated clinical outcome according to HCV serology. We are aware that our study has limitations, such as the fact that some transplanted patients could have HCV RNA in the absence of detectable anti-HCV antibodies (18), although it has been reported that serologic responses to HCV are less impaired in renal transplants than in solid organ recipients (18).
Some baseline characteristics that are different in HCV-positive and -negative groups can be explained by their close relationship with mechanisms of HCV transmission among end-stage renal disease patients. Indeed, previous kidney transplantation is a risk factor for HLA sensitization and, in the past, before HCV screening was routinely performed in the donor, it was a way to transmit HCV to a recipient (18). Other forms of sensitization such as blood transfusion could also be an important source of HCV transmission (18). Well-known predictive variables of graft outcome, such as acute rejection (19), delayed graft function (20), serum creatinine (21) and proteinuria at 1 year after transplantation (22), were similar in HCV-positive and -negative groups, thus corroborating that both groups were comparable considering their baseline characteristics.
Among renal allograft recipients, the prevalence of pretransplantation anti-HCV is 11% to 49% [reviewed in (4)]. In our study, HCV prevalence was 46%, whereas in our overall renal transplant population it is 23.7% (8), indicating that diagnostic allograft biopsy was more frequently performed in HCV-positive patients and suggesting that HCV infection promoted renal allograft pathology. Likewise, in the HCV-positive group there were some clinical data suggesting the presence of glomerular lesions in the graft, such as nephrotic syndrome and microhematuria. This suspicion was confirmed by pathologic examination, since MPGN was found in 45.4% of the diagnostic allograft biopsies in anti-HCV-positive patients. We have previously demonstrated (8) that lower titers of type II IgMκ cryoglobulins containing enriched HCV RNA participate in de novo type I MPGN in HCV-infected allografts. In the present study, we also found several cases of de novo MGN (18.2% in HCV-positive vs. 7.7% in HCV-negative patients) among diagnostic renal allograft biopsies. The association between HCV infection and de novo MGN, although more controversial, was previously reported by Morales et al. (11). In fact, other authors found a similar prevalence of MGN in HCV-negative and HCV-positive patients (12). Although studies in native kidneys suggested that immune complexes containing HCV-core proteins were involved in the development of MGN, no clear pathogenic mechanism has been described among renal allografts (23).
Several concurrent factors may account for the development of de novo immune-mediated glomerulonephritis in HCV-infected allografts. After renal transplantation, HCV-infected patients showed an increase of the viral titer, indicating HCV replication (4). Immunosuppression and HCV itself (24,25) may also modify the lymphocyte response and antibody formation against HCV. Consequently, in combination, modulation of the lymphocyte activity and increased viral load may produce an antigen–antibody imbalance, which associated with susceptibility of the renal allograft, predisposes to the long-term development of de novo GN. The hypothesis that kidney grafts are particularly susceptible to immune-mediated GN is supported by our previous observation that de novo MPGN appear despite the fact that cryocrites are low (8), and this may explain both the absence of extrarenal manifestations of HCV-related cryoglobulins and their attenuated histological expression on the allograft.
Association of chronic transplant glomerulopathy with HCV-infection is controversial. Cosio et al. (26) reported a high prevalence of transplant glomerulopathy among HCV-positive patients. Hestin et al. (5) studied renal histology, without electron microscopy, from 26 allografts with proteinuria and found that chronic transplant glomerulopathy was the most frequent glomerular lesion in both HCV-positive and HCV-negative patients. In our approach, which included electron microscopy examination, we found that chronic transplant glomerulopathy prevalence was not different between HCV-positive and HCV-negative groups. Furthermore, among HCV-positive patients, the de novo immune-mediated glomerular lesion was the most prevalent glomerular abnormality, indicating that diagnosis of chronic transplant glomerulopathy may be overestimated when electron microscopy is not routinely performed.
Besides the pathologic abnormalities associated with chronic HCV infection, we assessed their effect on graft survival. There is still controversy about the impact of HCV on patient and graft survival, although most authors agree that, irrespective of the antibody status, long-term patient survival depends on the severity of the associated liver disease (27). Curry et al. (24) recently reported that some HCV patients with active infection had expansion of CD5 + B lymphocytes which was associated with production of rheumatoid factor and protection against the development of progressive liver disease. This observation suggests that HCV-infected patients with autoimmunity, like those with de novo MPGN, probably have less progression of the hepatic disease and, interestingly, it may explain the absence of deaths due to chronic liver disease among our HCV-positive group.
Pathologic diagnoses associated with poor graft survival were de novo GN (either MGN or MPGN) and chronic transplant glomerulopathy. In agreement, Suri et al. (28) recently reported that the presence of transplant glomerulopathy on a biopsy specimen is followed by accelerated graft loss. Although, HCV-positive, proteinuria at the time of the biopsy, transplant glomerulopathy and de novo GN were predictors of graft survival in the univariate analysis, the only independent predictor of graft survival in the multivariate analysis was HCV infection. We can suggest a possible explanation for this finding, since we observed that de novo GN in HCV-infected patients was associated with extremely poor graft survival, even worse than in HCV-negative recipients. This different prognosis of de novo GN according HCV infection may simply be a reflection of the different physiopathology of glomerular lesions in HCV-positive (usually the nephritogenic component of cryoglobulins, IgMκ) and HCV-negative allografts.
In summary, we have shown that in diagnostic renal allograft biopsies the presence of de novo immune-mediated glomerulonephritis, especially type I MPGN, is strongly associated with HCV infection and results in accelerated loss of the graft. This makes it necessary to define pretransplant risk markers of de novo GN among HCV-infected patients, to implement an active policy of renal biopsies in this population, and to develop reasonable therapeutic strategies aimed to minimize the detrimental effects on graft survival of HCV-associated GN.
These results were presented orally at The Joint American Transplant Meeting, Transplant 2000, Chicago, May 13–17, 2000.