Main

Hepatitis viruses A through E have traditionally been associated with acute and chronic liver disease. Several lines of evidence point to the existence of additional infectious hepatitis agents(1). A small percentage (0.4%) of posttransfusion hepatitis and 18 to 20% of community-acquired hepatitis are unexplained and remain classified as non-A, non-B, non-C (NANBNC) hepatitis(2–4).

Two new viruses belonging to the Flaviviridae family, hepatitis G virus (HGV) and GB virus C (GBV-C), were discovered recently(5–7). Phylogenetic analysis indicated that HGV and GBV-C are variants of the same viral species and distantly related to hepatitis C virus(5–9).

Earlier reports raised the possibility that this new virus might play an etiologic role in NANBNC hepatitis(5–7,10,11–15). However, despite ample evidence demonstrating persistent GBV-C infection, this virus has not been linked to disease in humans or other primates, and neither seems to influence long-term morbidity and mortality nor cause postliver transplant disease in adults(16–22). It may, however, be associated with increased serum gamma glutamyl transferase (GGT) levels(23).

Most previous studies of GBV-C have been performed in adults. Studies to date in children are limited but important because viral infections such as GBV-C may produce identifiable clinical signs and symptoms not observed in adults. Immunocompromised children are known to be susceptible to GBV-C infection(24). GBV-C and hepatitis C co-infections were not causative agents for postliver transplant hepatitis in children in the United Kingdom(25).

The present study was designed to determine the prevalence of HGV/GBV-C infection in immunocompromised children with liver transplants, assess the relationship between GBV-C viremia and postliver transplant hepatitis, and examine the effect of GBV-C anti-E2 antibodies, which are believed to provide protection against acquisition of GBV-C viral infection(26–29). To our knowledge the role of E2 antibodies has not been studied previously in children.

PATIENTS AND METHODS

The study population consisted of children who underwent liver transplantation at the University of California, San Francisco (UCSF) or California Pacific Medical Center (CPMC) in San Francisco. Sera were collected before and after liver transplantation between April 1996 and February 1998 from 34 out of a total of 268 children (175 at CPMC and 93 at UCSF). Only 41 patients had paired sera available before and after transplant. Seven children who had evidence of hepatitis secondary to other known infectious etiologies (i.e. hepatitis C or hepatitis B) were excluded from the study. The remaining 34 patients formed the study cohort. Patient charts were reviewed for clinical data, including age, gender, ethnicity, and year of liver transplantation, indications for transplantation, and risk factors for parenteral acquisition of viral disease, such as number of blood transfusions, i.v. drug abuse, tattoos, and dialysis. Ethnicity was classified as African-American, Asian, Caucasian, Hispanic, and Other. All patients and/or their parents gave written informed consent before liver transplantation. The Committee on Human Research approved this study.

The study population included 25 girls (73%) and 9 (27%) boys. Median age at the time of transplantation was 1.25 y (range 4 months to 14.5 y). Thirteen children (38%) were Caucasian, four (12%) were Hispanic, four (12%) were Asian, two (6%) were African-American, and 11 (32%) were classified as Other, a group that included those from a multiethnic background. Indications for liver transplant included: biliary atresia(20), intrahepatic bile duct paucity(3), autoimmune hepatitis(3), cryptogenic cirrhosis(3), metabolic liver disease(3), hepatic tumor(1), and fulminant hepatic failure(1).

Pretransplant sera averaged 39.6 d before liver transplant (SD 66.4 d, range 1-300 d), and postliver transplant sera averaged 504.3 d after transplantation (SD 271.2, range 151-1 110 d). All sera were stored at -70°C. GBV-C RNA was amplified by reverse-transcription, polymerase chain reaction (60°C 30 min for 1 cycle, 94°C 40 sec, 63°C 1 min for 35 cycles, 97°C 5 min, 12°C 5 min for 1 cycle). The final step, which denatured the products and cooled them rapidly to 12°C, allowed the GBV-C probe to hybridize to the amplicon. This amplicon/probe complex was detected on the Abbott Laboratories LCx® analyzer (Abbott Park, IL) via a microparticle enzyme immunoassay (MEIA as previously reported(27,30)).

The immunoassay for detection of human antibodies elicited against GBV-C was performed at Abbott Laboratories, using a glycosylated form of GBV-C E2 protein that was purified and used as an antigenic target for detection of human anti-GBV-C(26). Serum samples were assayed by an indirect immunoassay, which used the E2 protein on a solid phase to capture antibodies from human plasma, followed by the addition of an enzyme-conjugated, antihuman antibody for color development(31).

For the purposes of this study, the definition of hepatitis was based on histologic assessment. At UCSF, protocol liver biopsies were performed weekly in the early posttransplantation period and annually thereafter. Additional liver biopsies were performed as clinically indicated (i.e. abnormal liver function tests). At CPMC, biopsies were obtained by clinical indication and stained with hematoxylin and eosin. All biopsies were reviewed, by two experienced hepatopathologists (LDF, RGK), using the same histologic scoring index for hepatitis on liver biopsy, which included portal/periportal activity, lobular activity, fibrosis, and cytopathic swelling. Each category was given a score of zero to four (0 absent, 1 minimal, 2 mild, 3 moderate, 4 severe)(32). The maximum histologic score was 16.

For statistical analyses, appropriate values were expressed as mean, median, range, and percentage. To analyze the data, the predictor variable was GBV-C RNA status and the outcome variable was hepatitis (yes/no). For the relationship between the presence or absence of GBV-C RNA and the presence or absence of hepatitis, we used the Fisher's Exact test. For the relationship between GBV-C RNA and histologic score, we used the Wilcoxon test. To evaluate the risk factors for acquiring the virus, we used a conversion rate.

RESULTS

Four of 34 children (12%) were GBV-C RNA positive pretransplant (3 had biliary atresia; 1, fulminant hepatic failure). After transplantation, 14 of 34 (42%) children were GBV-C RNA positive, including the three GBV-C RNA positive pretransplants and 11 newly acquired infections (7 biliary atresia, 1 hepatic tumor, 1 autoimmune hepatitis, 1 cryptogenic cirrhosis, and 1 fulminant hepatic failure). The GBV-C conversion rate was 33% (10 of 30), (CI 17.2-55.7%). There was no correlation between the child's sex, race, or indication for liver transplantation and GBV-C RNA conversion. The only parenteral risk factor for acquiring GBV-C was receipt of blood products. Three patients had a history of blood transfusion before liver transplantation. During and postliver transplant, patients who seroconverted received blood products from a mean of 68 donors (range 4 to 93, SD 34), compared with 43 donors (range 4 to 48, SD 29) for those patients who did not seroconvert. Donor sera were not tested for GBV-C RNA or E2 antibodies. None of the patients had tattoos, none of the patient's parents admitted to a history of drug abuse, and none of the patients required dialysis. Since the history of vertical transmission, breast-feeding, or other nonparenteral factors were not available in most of the patients' charts, these could not be analyzed.

Review of the liver biopsies indicated 23 children (67%) had evidence of hepatitis in the transplanted liver (Table 1). When comparing the grade of hepatitis(32), the average score for children with GBV-C viremia was 2.40 out of a maximum score of 16, while the average score for children without GBV-C infection was 0.63 out of 16. There was no statistical difference in the incidence or severity of hepatitis between the children who were GBV-C RNA positive compared with the GBV-C RNA negative group. (Table 1)

Table 1 Posttransplant hepatitis
Full size table

Overall, 26% (9 of 34) of the children were E2 antibody positive before transplant, and 23.5% (8 of 34) were E2 antibody positive after transplant.

Pretransplant E2 antibody status had no effect on the rate of viral (GBV-C RNA) acquisition. In the pretransplant patient population characterized as E2 antibody positive and GBV-C RNA negative, 33% (3 of 9) acquired GBV-C RNA positivity posttransplant. These data are not substantially different from the pretransplant patient population characterized as E2 antibody negative and GBV-C RNA negative, in which 33% (7 of 21) of the patients acquired GBV-C RNA positivity posttransplant (Table 2). The posttransplant E2 antibody positive and GBV-C RNA positive patients were of particular interest. Two of these patients were antibody positive, RNA negative pretransplant, and two were antibody and RNA negative pretransplant.

Table 2 Tracking of E2 Ab and RNA Pretransplant and Posttransplant
Full size table

DISCUSSION

The liver transplant model provides a means of assessing the relationship of GBV-C to liver injury. Immunocompromised children with liver transplants would be more prone to virus acquisition and may be more likely to exhibit clinical manifestations of infection. Because many of the liver transplant recipients underwent protocol liver biopsies, we had the opportunity to examine these transplanted organs for evidence of hepatic injury. In our study, the prevalence of GBV-C infection in children recovering from liver transplantation was 42%, similar to that reported in other postliver transplantation populations (24-67%)(16–18,25,33). The high frequency of de novo infection may most aptly be explained by transmission of the virus through blood product transfusions given to the patients (mean of 68 donors per recipient). Based on the estimated GBV-C RNA prevalence in American blood donors (1.4-1.7%)(4,5,13,34) and the mean number of donor exposures per transplant recipient, the expected prevalence of GBV-C viremia posttransplant would be 1-(0.983)68 = 0.68 for patients who acquired GBV-C, and 0.52 for patients who did not convert. In both groups, the risk is high (more than 50%) and associated with the risk of receiving blood products. None of the patients had other parenteral risk factors for acquisition of viral disease such as i.v. drug abuse, tattoos, or dialysis. A lower prevalence rate has been reported in children who received fewer transfusions and were not immunocompromised(35,36).

Four of the children (12%) in our study were GBV-C RNA positive before transplantation and three of the four received blood product transfusions (1-3 donors each) before transplantation. A 12% pretransplant GBV-C RNA positivity rate is consistent with the rather broad pretransplant percentage (8-39%) reported for adults with liver disease(5,11,12). One of the children who was GBV-C RNA positive pretransplant had fulminant hepatic failure of unknown etiology. It has been suggested that GBV-C virus may be related to hepatic failure. Alternatively, the GBV-C RNA may have been an "innocent bystander," secondary to blood transfusion as previously suggested(2–4,11,37–40). No etiology for acquisition of GBV-C RNA positivity was identified in the fourth child, an infant, although we were unable to evaluate the possibility of vertical transmission that had been previously reported(41).

GBV-C infection in immunocompromised children apparently did not contribute significantly to the incidence or severity of postliver transplant hepatitis. Posttransplant biopsies documented that hepatitis was present in similar numbers in GBV-C RNA positive and negative children, which is similar to the results of studies in postliver transplanted adults(16–18,33) and GBV-C and hepatitis C co-infected children(25). Although patients with GBV-C viremia had slightly higher histologic scores on biopsy after transplantation, the difference was not statistically significant. However, immunosuppression may contribute to prolonged GBV-C viremia. Long-term follow up of the children in our study demonstrated that GBV-C RNA could be detected as long as 37 months after transplant, which is longer than the GBV-C viremia described in postliver transplanted adults(16,33), but shorter than that reported for children after neonatal transfusion(42) or infant cardiac surgery(36,43).

Previous studies in postliver transplanted adults indicated that the appearance of E2 antibody correlated with resolution of GBV-C viremia and possible protection against further GBV-C infection(26–28). We noted that none of our patients were E2 Ab positive or GBV-C RNA positive before transplant, which is consistent with previous observations. However, the presence or absence of pretransplant E2 antibody had no effect on the percent of patients who acquired GBV-C RNA positivity posttransplant. Exactly 33% of patients who were E2 Ab negative pretransplant or E2 Ab positive pretransplant acquired GBV-C RNA posttransplant. These results contrast with previous reports in which the presence of E2 antibody affected the rate of GBV-C RNA acquisition(26–28). Furthermore, four children lost their E2 antibodies posttransplantation and one of them acquired GBV-C infection. The ages of the patients with E2 antibodies pretransplant ranged from 4.5 mos to 14.5 y, suggesting that the E2 antibodies present in these children were not transient or due to transplacental transfer of maternal antibodies. Since posttransplant sera were obtained at the average of 504.3 d after transplant, it is highly unlikely that these patients acquired E2 Ab by means of passive transfusion via blood products.

The loss of the E2 antibodies, and the acquisition of GBV-C in the presence of E2 Ab, suggests that the immunity E2 Ab provides in immunocompromised children is neither complete or long standing after liver transplantation. We were unable to examine the differences in immunosuppression regimens to determine whether this influenced the loss of the E2 antibodies. This information may prove useful in the future development of a GBV-C virus vaccine.

In summary, the prevalence of GBV-C RNA positivity after liver transplantation in children was high. We found no correlation between GBV-C viremia and the incidence or severity of postliver transplant hepatitis. In immunocompromised children with liver transplants, E2 antibodies did not appear to afford complete protection against GBV-C, and were not present in some children after transplantation. However, the possibility of transient viremia cannot be ruled out. Further studies of this newly identified virus need to be performed in other groups of children to fully understand its clinical significance.