Hepatitis E Virus-Related Cirrhosis in Kidney-and Kidney–Pancreas-Transplant Recipients

Authors


* Corresponding author: Nassim Kamar, kamar.n@chu-toulouse.fr

Abstract

Hepatitis E virus (HEV) infection was thought to be responsible for acute hepatitis that did not become chronic. However, we have recently reported that HEV infection can evolve to chronic hepatitis, at least in solid-organ transplant patients. We report on two cases of rapidly progressive of HEV-related cirrhosis that occurred in two organ-transplant patients. Case 1: A kidney–pancreas-transplant patient developed acute HEV hepatitis 60 months after transplantation, which evolved to chronicity as defined by persisting elevated liver-enzyme levels and positive serum HEV RNA. At 22 months after the acute phase, she presented with cirrhosis and portal hypertension, that is ascites and esophagus varices. Case 2: A kidney-transplant patient developed acute hepatitis 36 months after transplantation, which persisted and remained unexplained for 38 months. Then, HEV RNA was searched for in their serum and stools, and was found to be positive in both. Retrospective analysis of available stored serum, mainly the serum obtained at the acute phase, confirmed the diagnosis of chronic hepatitis E. In both cases, a liver biopsy showed cirrhosis. We conclude that HEV infection cannot only evolve to chronic hepatitis, but can also be responsible for rapidly progressing cirrhosis in organ-transplant patients.

Introduction

Hepatitis E virus (HEV) is an emerging disease in industrialized countries (1). Cases of autochthonous acute hepatitis E infection have been reported worldwide. Recent studies have reported a high prevalence of anti-HEV IgG antibodies in blood donors in the southwest of France, that is 16.6% (2). HEV infection was thought to be responsible for acute hepatitis that does not become chronic. However, we have recently reported that HEV infection can evolve to chronic hepatitis, at least in solid-organ transplant patients (3). Herein, we report on two cases of rapidly progressive HEV-related cirrhosis that occurred in a kidney–pancreas and a kidney-transplant patient.

Case Reports

Patient 1: A 55-year-old woman received a first deceased donor dual kidney and pancreas transplantation for diabetes mellitus. Initial immunosuppression was based on induction therapy with antithymocyte globulins, followed by triple sequential immunosuppressive therapy that included tacrolimus (targeting trough levels of 8–12 ng/mL), azathioprine (2 mg/kg/day) and steroids (5 mg/day). She was insulin-free and her serum creatinine level was 93 μmol/L. She did not experience any acute rejection after transplantation. Sixty months posttransplantation, she presented with fatigue without fever. No abnormalities were detected on physical examination. Biological examination revealed acute hepatitis associated with an increase in alanine aminotransferase (ALT) levels from 13 to 156 IU/L (N = 5–55), aspartate aminotransferase (AST) levels from 22 to 115 IU/L (N = 5–40) and gamma-glutamyl transpeptidase (γGT) levels from 8 to 47 IU/L (N = 15–73). Total bilirubin levels also increased from 11 to 16 μmol/L (2 to 21 μmol/L). Serum creatinine level remained stable at 96 μmol/L. There was no significant proteinuria. C-reactive protein was 10 mg/L (N < 3 mg/L). The prothrombin test was 100% (N > 70%). Biliary-tract complications were ruled out by abdominal ultrasonography. Toxin- and drug-related causes of abnormal liver-function tests were ruled out by anamnesis. The patient's alcohol consumption was lower than 10 g/day. Hepatitis B, C and E virus (HBV, HCV, HEV) serologies and HAV IgM were negative (HEV serology was determined by an enzyme immunoassay, Abbott HEV EIA, Rungis, France). HBV DNA and HCV RNA were also negative. There was no evidence of CMV, EBV and herpes simplex virus (HSV) infections. Antinuclear autoantibodies (ANA), anti-liver-kidney microsomal (LKM) antibodies, anti-smooth-muscle cell (SMC) antibodies and antimitochondrial antibodies were not detected. Epstein Barr virus DNA concentration in the blood was unchanged as compared to 4 months before the acute hepatitis episode, for example 4.3 log10 copies/mL. In contrast, serum HEV RNA was found to be positive (HEV RNA was detected by a real-time polymerase chain reaction using Taqman: This amplified a 189-bp product located in the ORF2 region) (4). Phylogenetic analysis revealed that the strain belonged to genotype 3f. A liver biopsy showed moderate hepatitis with a Metavir score of A2F0. Portal inflammation was moderate with lymphoid follicules, rare parcellar necrosis but important lobular necrosis. Detection of CMV, HBV and EBV was negative using immunohistochemistry or in situ hybridization. At acute HEV diagnosis, CD4-positive T cells, CD8-positive T cells and CD19-positive cells counts were, respectively 216/mm3, 367/mm3 and 26/mm3. There were no available CD4, CD8 and CD19 counts before the acute phase. In contrast, 1 year after the acute phase, CD4-positive T cells, CD8-positive T cells and CD19-positive cells counts were, respectively 208/mm3, 291/mm3 and 21/mm3. There was no evident transmission route. This patient lived in an urban region in the southwest of France and had not traveled outside France during the last year. Her treatment had not been significantly modified and we decided to monitor her liver parameters. Azathioprine and steroid doses remained unchanged. Tacrolimus trough levels at diagnosis, 3, 6, 12 and 15 months later were, respectively 10.8, 8.7, 8, 7 and 7 ng/mL. Liver-enzyme abnormalities, as well as the detection of HEV RNA in the serum, persisted (Figure 1A). A second liver biopsy was performed 16 months after diagnosis. The latter revealed signs of moderate hepatitis with a Metavir score of A2F2; it also showed signs of acute cholangitis with epithelial alteration and bile-duct infiltration by inflammatory cells. Six months later, that is 22 months after the acute phase, she was admitted for fatigue, weight loss (−5 kg in 6 months), jaundice and diarrhea. Physical examination revealed the presence of ascites, which was confirmed by ultrasonography. At admission, her biological parameters were as follow: ALT level at 211 IU/L, AST at 148 IU/L, γ-GT level at 169 IU/L, total bilirubin level at 217 μmol/L, conjugated bilirubin level at 119 IU/L and CRP level at 36 mg/L. HEV RNA was still positive in the serum and the stools. Anti-HEV IgG was also detected. In contrast, all other previously quoted viral serologies and viremias were unchanged. An esophagogastroduodenoscopy revealed the presence of esophagus varices. Colonoscopy showed signs of congestive and inflammatory mucosa. Unfortunately, a few days later, she developed a septic shock due to an Escherichia coli-induced colitis followed by a multiorgan failure. She died a few days after admission. A postmortem liver biopsy confirmed the presence of cirrhosis with a Metavir score of A1 F4 (Figure 2).

Figure 1.

Liver-enzyme levels and virological parameters in two patients who developed HEV-related cirrhosis. Abbreviations: AST = aspartate aminotransferase; ALT = alanine aminotransferase; γ-GT = gamma-glutamyl transpeptidase; ND = not done.

Figure 2.

Histological pattern of liver biopsies taken from patient 2 at diagnosis and at last follow-up. (A) Initial liver biopsy shows a low fibrosis score: F1 (Masson's trichrome, ×100). (B) Thirty-eight months later, liver biopsy revealed a cirrhosis: F4 (Masson's trichrome, ×100).

Patient 2: A 39-year-old man received a third deceased donor kidney transplantation for chronic glomerulonephritis. Initial immunosuppression was based on induction therapy with antithymocyte globulins, followed by triple sequential immunosuppressive therapy that included tacrolimus (targeting trough levels of 8–12 ng/mL), mycophenolate mofetil (500 mg b.i.d) and steroids (5 mg/day). His serum creatinine level was 148 μmol/L. He did not experience any acute rejection after transplantation. At 32 months after transplantation, a routine biological study revealed acute hepatitis. ALT levels had increased from 15 to 127 IU/L, AST had increased from 13 to 68 IU/L, γ-GT levels from 57 to 119 IU/L and total bilirubin from 6 to 12 μmol/L. He was asymptomatic and no abnormalities were evident on physical examination. C-reactive protein level was 5 mg/L (N < 3 mg/L); the prothrombin index was 100% (N > 70%). Serum creatinine level had also increased from 148 to 207 μmol/L. There was no significant proteinuria. Biliary-tract complications were ruled out by abdominal ultrasonography. Toxin- and drug-related causes of abnormal liver function tests were ruled out by anamnesis. The patient's alcohol consumption was lower than 10 g/day. Hepatitis B and C virus (HBV, HCV) serologies and HAV IgM were negative. HBV DNA and HCV RNA were also negative. There was no evidence of CMV, EBV or HSV infections. ANA, anti-LKM, anti-SMC and antimitochondrial antibodies were not detected. At diagnosis, CD4-positive T cells, CD8-positive T cells and CD19-positive cells counts were, respectively 189/mm3, 461/mm3 and 38/mm3. There were no available CD4, CD8 and CD19 counts before the acute phase. In contrast, 1, 2 and 3 years after the acute phase, CD4-positive T cells were respectively 204/mm3, 93/mm3 and 176/mm3; CD8-positive T cells were respectively 517/mm3, 251/mm3 and 664/mm3; and CD19-positive cells counts were, respectively 25/mm3, 20/mm3 and 31/mm3. His treatment had not been modified. Tacrolimus trough level was always ranging between 8 and 10 ng/mL. MMF (1 g/day) and steroid (5 mg/day) doses were unchanged. Because of persisting elevated liver-enzyme levels, a liver biopsy was performed 1 year after the acute phase. It revealed signs of mild nonspecific hepatitis with a Metavir score of A1F1. At that time, the patient was switched from tacrolimus to cyclosporine A because of new-onset diabetes mellitus. Liver-enzyme elevation persisted but no etiology of the hepatitis was found (Figure 1B). Later, that is 38 months after the acute phase, the patient was referred to our institute for persistent elevated liver-enzyme levels and weight loss (−10 kg in 24 months). No abnormalities were found on physical examination. Biological parameters revealed liver-enzyme elevation (Figure 1B). A whole-body CT-scan, a colonoscopy, as well as an esophagogastroduodenoscopy did not reveal any abnormalities no esophageal varices. All previously quoted viral serologies and viremias were unchanged, except for EBV viremia, which became slightly positive, that is 3 log10 copies/mL. In contrast, IgG anti-HEV, IgM anti-HEV and both serum and stool HEV RNA were found to be positive. Phylogenetic analysis revealed that the strain belonged to genotype 3f. Retrospective analysis showed that HEV RNA was undetectable before the acute hepatitis episode, and was repeatedly detected in available serum (Figure 1B). Liver biopsy revealed cirrhosis associated with parcellar necrosis, cholangitis and moderate inflammation (Metavir score of A2 F4) (Figure 2). No signal was detected by in situ hybridization for EBV-encoded RNA There was no evident transmission route. This patient lived in an urban region in the southwest of France and had not travelled outside France since his last kidney transplantation.

Discussion

HEV infection seems to be a zoonosis transmitted via the fecal–oral route. Even though, its clinical presentation is quite similar to hepatitis A virus infection, it seems to be more severe, leading to a mortality rate of about 1% in the general population (5). However, Peron et al. have recently shown that HEV-infected patients were older and had lower ALT levels compared to those infected by hepatitis A (6). When HEV infection occurs during pregnancy, the mortality rate can reach 30% (7). Fulminant hepatitis has been also previously reported and requires a liver transplant in some cases and has led to death in the other instances (8–10). In organ-transplant patients, acute HEV-induced hepatitis has been reported in both kidney- and bone-marrow-transplant patients (3,11,12). HEV-induced pancreatitis has also been reported (13). Even though two cases of persistent HEV infection have been previously reported (14,15), HEV was, until very recently, considered to be responsible for acute hepatitis that did not become chronic. However, recently, we have shown that HEV infection can evolve to chronicity, at least in solid-organ transplant patients (3). Indeed, we have identified 14 cases of acute hepatitis E infection that have occurred in three liver-, nine kidney- and two-kidney–pancreas-transplant patients. Eight of the 14 patients evolved to chronic hepatitis, which was confirmed by the presence of persisting elevated liver-enzyme levels and positive serum and/or stool HEV RNA at 15 (10–24) months after the acute phase. Furthermore, Gerolami et al. recently described a case of HEV-related cirrhosis in a kidney-transplant patient (16). Herein, we report two other cases of rapidly progressing chronic HEV hepatitis that led to severe cirrhosis.

The first patient developed acute HEV hepatitis that evolved to chronicity. At 22 months after the acute phase, she presented with cirrhosis and portal hypertension, that is ascites and esophagus varices. The second patient developed persisting hepatitis, which remained unexplained until the last visit, that is 38 months after the acute phase. At his last visit, HEV RNA was looked for in the serum and the stools, and was found to be positive in both. Retrospective analysis of available stored serum, mainly the serum obtained at the acute phase, confirmed the diagnosis of chronic hepatitis E. A liver biopsy revealed cirrhosis. We did not identify any source of HEV contamination. However, we speculate that the virus was transmitted via water or food intake.

Similar to the case report by Gerolami et al. (16) the most striking finding is the rapidity of evolution to cirrhosis, especially in the setting of kidney transplantation. The progression of liver fibrosis may have been accelerated in patient 1 by azathioprine therapy, which is known to induce hepatitis (17). Nevertheless, patient 2 developed HEV-related cirrhosis without receiving azathioprine. In anti-HCV positive/RNA-positive kidney-transplant patients, we have previously shown, by means of sequential liver biopsies performed every 3–4 years after transplantation, that the rate of liver-fibrosis progression was 0.09 ± 0.03 Metavir units/year (18). Only 3 patients out of the 51 developed cirrhosis a long time after transplantation (18). This finding suggests that HEV infection in kidney-transplant patients is more severe than HCV infection. This might be related to a different immune response. Srivastva et al. found that CD4(+) IFN-gamma-secreting cells and natural killer T cells, may be involved in the pathogenesis of hepatitis E (19). Pal et al. have suggested that the greater severity of hepatitis E infection among pregnant women may be related to a Th2 response (20). Interestingly, both patients who developed cirrhosis had a low CD4-positive T cell count at initial presentation. In addition, serum EBV DNA was (patient 1) or became (patient 2) positive during follow-up. These two findings reflect a heavy immunosuppression in both patients who received induction therapy by RATG followed by a triple sequential immunosuppressive therapy. Furthermore, Patient 2 was grafted twice previously. Hence, the cumulative immunosuppression may play an important role in the persistence of the virus and in liver fibrosis progression. Indeed, in case of reduced CD4 positive cells count induced by immunosuppressive therapy, CD8 positive T cells are not able to clear viruses (21). Further immunological investigations are required to understand the severity of HEV infection in organ-transplant patients.

The treatment of chronic HEV infection after kidney transplantation is unknown. Interferon-α therapy is not recommended in this setting because of the high risk of acute rejection (22). Total lymphocyte and CD2, CD3 and CD4 lymphocyte subset counts were found to be significantly lower in solid-organ transplant patients that evolved to chronicity compared to those who were cleared off HEV, suggesting that the T-cell response seems to play a role in HEV clearance (3). Hence, reducing immunosuppressive drugs that target the T cells may be a possible therapeutic option. The monitoring of CD4 positive cells may be used to adapt the immunosuppressive therapy in organ-transplant patients infected by HEV. Very recently, an HEVrecombinant-protein vaccine was found to be effective in preventing hepatitis E infection (23). Hence, we suggest that a vaccination against HEV should be evaluated for patients who live in an endemic area, before undergoing an organ transplantation.

We conclude that HEV infection cannot only evolve to chronic hepatitis but can also be responsible for rapidly progressing cirrhosis.

Acknowledgment

We thank Mrs Martine Dubois for her technical assistance.

Ancillary