Hepatitis G virus infection before and after liver transplantation. Liver Transplantation Database

Hepatitis G Virus Infection Before and After Liver Transplantation Jay H. Hoofnagle,* Manuel Lombardero,† Yuling Wei,† James Everhart,* Russell Wiesner,‡ Rowen Zetterman,§ Andersen J. Yun,6 Limei Yang,6 and Jungsuh P. Kim6 for the Liver Transplantation Database

SEE EDITORIAL ON PAGE 677 (PESSOA & WRIGHT)

The hepatitis G virus is a newly discovered flavivirus that has been linked to acute and chronic hepatitis of unknown cause. We determined the prevalence of hepatitis G virus infection in 179 selected patients undergoing liver transplantation at three centers participating in the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Liver Transplantation Database. Pretransplantation and posttransplantation specimens were tested for hepatitis G virus RNA by polymerase chain reaction. Before transplantation, 9 of 38 (24%) patients with fulminant hepatic failure, 9 of 62 (15%) with cryptogenic cirrhosis, 3 of 35 (9%) with cholestatic liver disease, and 5 of 44 (11%) with chronic hepatitis C were positive for hepatitis G virus RNA (P 5 .27). Patients with and without viral RNA were similar in clinical features,

T

he genome of a new flavivirus has been identified by two groups of investigators in the serum of patients with acute and chronic liver disease of unknown cause.1,2 This viral agent has been called the hepatitis G virus (HGV) by Linnen et al1 from Genelabs Technologies, Inc and the GB-C virus by Simons et al2-5 from Abbott Diagnostics because of its similarity to the so-called GB-A and GB-B viruses that they had identified in marmoset monkeys with viral hepatitis. These two

From the *Division of Digestive Diseases and Nutrition, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD; the †Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA; the ‡Mayo Clinic and Foundation, Rochester, MN; the §University of Nebraska Medical Center, Omaha, NE; and 6Genelabs Technologies, Inc., Redwood City, CA. Address reprint requests to Jay Hoofnagle, M.D., Bldg 31, Rm. 9A23, National Institutes of Health, Bethesda, MD 20892. Copyright r 1997 by the American Association for the Study of Liver Diseases 1074-3022/97/0306-0004$3.00/0

578

liver test abnormalities, and survival after transplantation. Posttransplantation serum specimens were tested from 73 patients; 9 of 11 (82%) who were positive for viral RNA before transplantation remained positive, but 35 of 62 (56%) patients who were initially negative became positive after transplantation, a rate consistent with that predicted from the number of blood products administered. Only 5% of de novo HGV infections could be attributed to preexisting hepatitis G virus RNA in the donor. Comparison of patients with and without hepatitis G virus infection showed no difference in incidence of hepatitis after transplantation. Thus, hepatitis G virus infection was present in 15% of patients before and appeared de novo in half of patients after liver transplantation. Although hepatitis G virus infection was not associated with poor outcome, the frequency of this infection after transplantation calls for further long-term evaluation. Copyright r 1997 by the American Association for the Study of Liver Diseases

agents share 95% to 97% amino acid homology and are thus likely to be two isolates of the same virus. These agents were initially linked to cases of hepatitis of unknown cause, making the HGV a candidate agent for non-A, non-B, non-C, non-D, non-E (non A-E) hepatitis. However, these agents are detected in 1% to 2% of the general population and 10% to 25% of patients with other known forms of liver disease, particularly chronic hepatitis B and C.6-12 It remains unclear whether the HGV is a cause of acute or chronic liver disease and whether it accounts for some cases of non–A-E hepatitis. Cryptogenic or non A-E hepatitis is a common cause of fulminant hepatic failure and represents a common diagnosis in patients with end-stage liver disease.6-10 We evaluated serum samples from patients who were prospectively entered into a liver transplantation database, selecting all patients with fulminant hepatic failure and cryptogenic cirrhosis and a randomly selected cohort of patients with end-stage chronic hepatitis C and cholestatic liver disease.

Liver Transplantation and Surgery, Vol 3, No 6 (November), 1997: pp 578-585

Hepatitis G Virus and Liver Transplantation

Materials and Methods Patient Cohorts The NIDDK Liver Transplantation Database is a 7-year prospective study of patients undergoing liver transplantation at three centers: Mayo Clinic Foundation, Rochester, MN; the University of Nebraska Medical Center, Omaha, NE; and the University of California, San Francisco, CA.13 All patients evaluated for liver transplantation at these centers who gave written informed consent were enrolled in this database. Detailed information was obtained at the time of initial evaluation, immediately before transplantation, and at defined points after transplantation. Information obtained included demographic features, medical history, symptoms, signs, biochemical laboratory test results, serology, and radiological and histological results. Clinical information was also obtained regarding the donor. Serum was collected from the recipient immediately before and 4 months, 1 year, and 2 years after transplantation and stored in a serum bank. When available, serum was also obtained from the organ donor. Patients underwent protocol liver biopsies immediately after transplantation and 1 and 3 weeks and 1 year after liver transplantation. Liver histology was read by the participating hepatic pathologists using a standard form for staging and grading of the liver histology. Diagnostic categories on the pathology form included ‘‘definite’’ or ‘‘probable’’ viral hepatitis. All data were entered into a computerized database maintained by the Coordinating Center at the University of Pittsburgh, Graduate School of Public Health, Pittsburgh, PA. Between April 1990 and July 1994, 916 patients undergoing initial liver transplantation were enrolled in the database. For the current study, four categories of patients who underwent liver transplantation were identified for serum testing and analysis: patients with fulminant hepatic failure, cryptogenic cirrhosis, chronic cholestatic liver disease, and chronic hepatitis C. The diagnostic criteria were established in advance of this study and analysis.13 Fulminant hepatic failure was diagnosed if a patient had signs of liver failure, including hepatic encephalopathy with no previous history of liver disease. Of 58 patients with fulminant hepatic failure who underwent liver transplantation, adequate serum samples (at least three vials obtained before transplantation) were available from 38. Cryptogenic cirrhosis was diagnosed if a patient had end-stage liver disease with no known etiology, denied a history of significant alcohol use, and tested negative for hepatitis B surface antigen (HBsAg), antibody to hepatitis C virus (anti-HCV), and autoantibodies (antinuclear antibody). Of 93 patients who underwent liver transplantation for cryptogenic cirrhosis, serum was available for testing from 62. The diagnoses of cholestatic liver diseases were based on usual clinical criteria and included cases of primary and secondary

579

biliary cirrhosis and sclerosing cholangitis. Of 88 patients who fit this definition, had not received blood transfusions, and underwent liver transplantation, 35 were randomly selected for testing (17 with primary biliary cirrhosis, 17 with sclerosing cholangitis, and 1 with secondary biliary cirrhosis). The diagnosis of chronic hepatitis C was based on the finding of end-stage liver disease with a history of chronic hepatitis and the presence of anti-HCV in serum by second-generation enzyme immunoassay (EIA). Of 196 patients with chronic hepatitis C in the database, 44 were randomly selected for testing. Samples chosen for these studies may have included some patients tested for HGV infection and reported separately by the individual transplant centers.14

Virologic Assays Testing for HBsAg, antibody to HBsAg (anti-HBs), antibody to hepatitis B core antigen (anti-HBc), and antibody to hepatitis A virus (anti-HAV) (including immunoglobulin M [IgM] anti-HAV) was performed with commercial immunoassays. All patients were tested for anti-HCV by second-generation EIA (EIA-2) either when the serum was collected or from stored serum. Samples selected for this study were also tested for HGV RNA by reverse-transcription–polymerase chain reaction (RT-PCR) using primers from the NS5 region of the genome.1 RNA was extracted from 125 µL of serum by the purescript method with reagents from Gentra Systems (Minneapolis, MN). The pelleted RNA was redissolved in 25 µL of water. A 10-µL volume of RNA was used in the RT-PCR reaction, the equivalent of 50 µL of serum. RT-PCR was done in duplicate on each extracted sample in a 96-well format, using random primers and MMLV reverse transcriptase.1 Positive controls for 470 and NS5 and negative controls were run with each plate of samples. PCR was performed with four primers, two from the 470 region (sense primer ‘‘470-201-211R,’’ 58-CGAATGAGTCAGAGGACGGGGTAT-38, and anti-sense primer ‘‘470-20-1-77F,’’ 58-CTCTTTGTGGTAGTAGCCGAGAGAT-38) and two from the upstream NS5 region (sense primer ‘‘GV57-4512MF,’’ 58-GGACTTCCGGATAGCTGARAAGCT-38 and anti-sense primer ‘‘GV57-4657MR,’’ 58GCRTCCACACAGATGGCGCA-38 [R 5 A or G]). The RT-PCR yielded two PCR products, one of 134 bp (470 region) and one of 145 bp (NS5), that were probed using two different methods. The 470 region product, which was biotinylated (on the 211R primer), was hybridized to an oligonucleotide probe (probe ‘‘470-20-1-152F,’’ 58TCGGTTACTGAGAGCAGCTCAGATGAG-38) that had a chemiluminescent label (TBR) attached for electrochemical luminescence detection (Perkin Elmer Q-PCR 5000). Samples that tested positive or indeterminate for the 470 product were further tested for the NS5 product using a 32P (New England Biolabs, Beverly, MA: polynucleotide kinase and

580

Hoofnagle et al

buffer) end-labeled probe (probe ‘‘GV57-89M,’’ 58CYCGCTGR TTTGGGGTGTACTGGAAGGC-38: [R 5 A or G; Y 5 C or T]).1 All positive and indeterminant samples were reextracted and retested. The limits of detection of the probe-based assays were 10 RNA copies of positive control RNA transcripts per reaction or approximately 200 genome equivalents per milliliter of serum.

Table 1. Comparison of Patients With Fulminant Hepatic Failure With and Without HGV RNA in Serum

Feature

Statistical Analyses Differences in proportions, means, and medians were assessed with x2 tests, t tests, and Mann-Whitney U tests, respectively. Log-rank tests were used if the proportions compared involved censored data. Cox regression models were used to obtain relative risks of patient death and graft failure after posttransplantation testing for HGV RNA, adjusting for the time of conditional survival up to the date blood was drawn for testing.

Results Fulminant Hepatic Failure Of 38 patients undergoing liver transplantation for various etiologies of fulminant hepatic failure, 9 (24%) were reactive for HGV RNA on a pretransplant serum sample (Table 1). HGV RNA was detected in 1 of 2 patients with acute hepatitis A, 3 of 7 with hepatitis B, 1 of 6 with drug-induced hepatitis, 1 of 3 with other forms of acute liver failure, and 3 of 20 with fulminant hepatitis of unknown cause (cryptogenic). Thus the prevalence of HGV RNA among patients with fulminant hepatic failure of known cause (6 of 18, 33%) was twice that found among those with acute liver failure of unknown cause (3 of 20, 15%), although the difference did not reach statistical significance. Moreover, patients with and without HGV RNA in serum were not different in regards to sex, age, duration of disease, liver test results, history of blood transfusion or drug abuse, or survival. Chronic End-Stage Liver Disease Of 141 patients with chronic end-stage liver disease, 17 (12%) were reactive for HGV RNA before transplantation. Patients with HGV RNA in serum included 9 of 62 (15%; 95% confidence interval [CI], 7% to 26%) with cryptogenic cirrhosis, 5 of 44 with chronic hepatitis C (11%; 95% CI, 4% to 25%), and 3 of 35 with chronic cholestatic liver disease (9%; 95% CI, 2% to 23%). These differences among these three groups were not statistically significant (P 5 .68). The rate of HGV RNA positivity among the 38 patients with fulminant hepatic failure (24%; 95% CI, 11% to 40%) was twice as

HGV Positive (n 5 9)

HGV Negative (n 5 29)

P

Male (n) 4 (44.4%) 14 (48.3%) 1.00 Age (yr) 31.0 6 19.3 37.9 6 18.5 .22 Duration Dz (d) 6.9 6 6.3 10.5 6 10.5 .52 Blood transfusion 2 (22%) 4 (13.8%) 1.00 IV drug abuse 0 (0%) 0 (0%) ALT (U/L) 1,384 6 1,336 1,825 6 2,695 .80 AST (U/L) 1,219 6 1,429 1,603 6 2,437 .58 Bilirubin (mg/dL) 31.5 6 9.0 24.6 6 11.7 .06 Albumin (g/dL) 3.0 6 0.3 2.9 6 0.5 .69 Protime (s) 18.1 6 12.8 16.0 6 12.9 .38 Etiology Hepatitis A 1 (11.1%) 1 (3.4%) Hepatitis B 3 (33.3%) 4 (13.8%) Hepatitis D 0 (0%) 1 (3.4%) Wilson’s .38 disease 1 (11.1%) 0 (0%) Drug-induced 1 (11.1%) 5 (17.2%) Heat stroke 0 (0%) 1 (3.4%) Unknown 3 (33.3%) 17 (58.6%) 1-yr survival 8 (88.9%) 21 (72.4%) .41

6

NOTE. Data represent mean (%) or mean 6 SD. Abbreviations: Dz, disease; IV, intravenous; ALT, alanine transaminase; AST, aspartate transaminase.

high as that among the 141 patients with end-stage liver disease (12%; 95% CI, 7% to 19%), but this difference did not reach statistical significance (P 5 .11), and the rates in patients with fulminant versus chronic liver failure of unknown cause were the same (15%; 3 of 20 v 9 of 62, respectively). Combined analysis of the three patient groups who underwent liver transplantation for end-stage liver disease showed only minor differences in demographic, clinical, biochemical, or serological features between those with and without HGV infection (Table 2). Patients with HGV RNA in serum were somewhat younger than those without HGV RNA (mean age, 44.3 versus 51.0 years; P 5 .08), but they were similar with regard to other clinical features and degree of abnormality of liver biochemical test results. One-year survival was 76.5% among patients who were HGV positive and 85.5% among those who were negative before transplantation (P 5 .47). Histories of blood trans-

Hepatitis G Virus and Liver Transplantation

Table 2. Comparison of Patients With End-Stage Liver Disease With and Without HGV RNA in Serum

Feature

HGV Positive (n 5 17)

HGV Negative (n 5 124)

P

Male (n) Age (yr) Duration Dz (yr) Blood transfusion ALT (U/L) AST (U/L) Bilirubin (mg/dL) Albumin (g/dL) Protime (s) 1-yr survival

7 (41.2%) 44.3 6 14.5 4.6 6 3.8 5 (29.4%) 110 6 98 150 6 110 9.3 6 9.4 2.7 6 0.8 3.9 6 4.2 13 (76.5%)

61 (49.2%) 51.0 6 12.2 5.6 6 6.0 46 (37.1%) 91 6 102 138 6 114 6.5 6 8.4 3.1 6 0.6 3.2 6 4.9 106 (85.5%)

.54 .08 .87 .52 .50 .54 .26 .12 .44 .47

NOTE: Data represent number (%) or mean 6 SD from combined analysis of three cohorts; patients with cryptogenic cirrhosis (n 5 62), cholestatic liver disease (n 5 35), and chronic hepatitis C (n 5 44). Abbreviations: Dz, disease; ALT, alanine transaminase; AST, aspartate transaminase.

fusion and parenteral exposures were no more frequent among HGV RNA–positive than among HGV RNA–negative patients. HGV RNA After Liver Transplantation Serum specimens obtained after liver transplantation were available from 28 patients with fulminant hepatic failure and 45 with cryptogenic cirrhosis (10% within 6 months and 85% at 1 year or later). The results of the 73 patients in the two diagnostic groups were combined for analysis. Among 11 patients who were HGV RNA positive before transplantation, 9 (82%) remained positive after transplantation. The 2 patients who were nonreactive underwent transplantation for fulminant hepatic failure (1 case of hepatitis A and 1 of hepatitis B) and tested negative for HGV RNA on days 6 and 49 after transplantation (both underwent retransplantation shortly thereafter, and 1 of them tested HGV RNA positive after the second transplant). Among the 62 patients who were HGV RNA negative before transplantation, 35 (56%) became HGV RNA positive after transplantation (31 were tested between 1 and 2 years posttransplantation). The rate of de novo HGV infection after transplantation was similar for patients who underwent transplantation for fulminant hepatic failure (11 of

581

21, 52%) and for cryptogenic cirrhosis (24 of 41, 59%; P 5 .79). The high frequency of de novo infection led to analyses of the possible source of HGV transmission. Two major possibilities were investigated, the liver donor and blood transfusions. Serum was available from 20 of the 35 liver donors of these recipients with de novo HGV infection, and only 1 (5%) was positive for HGV RNA. Thus, de novo HGV infection occurred in half of the patients after transplantation but could be attributed to HGV infection in the organ donor in only 5% of cases. All patients received blood, plasma, and platelet transfusions during transplantation and in the perioperative period. Comparison of the 35 patients who became HGV RNA positive after transplantation with the 27 who remained negative showed no significant differences in the total amount of blood products received. The average exposure for patients with HGV was 20 units of packed red cells, 31 units of fresh-frozen plasma, and 33 units of platelets (total, 84 units) and for the uninfected patients was 16 units of packed red cells, 31 units of fresh-frozen plasma, and 26 units of platelets (total, 73 units). The total number of blood product exposures ranged from 18 to 349, with a median of 55 units in both groups. Although testing for HGV RNA on the individual units could not be done, the probability of exposure to an HGV RNA–positive unit could be estimated based on the assumption that 1.5% of the blood donor population was HGV RNA positive and that the average recipient received 55 units: probability 5 [1 2 (0.985)55]. These assumptions yielded an expected infection rate of 56.5%, which was identical to the observed rate (35 of 62 patients, 56.5%). To further assess the possibility that the high rate of HGV infection was caused by blood product exposure in the perioperative period, patients were categorized into four groups by the level of blood product exposure and thus by their probability of receiving an HGV RNA–positive unit (Table 3). Across all levels of exposure, the observed and expected rates of HGV infection were consistent, suggesting that the frequency of HGV infection could be explained by the exposure to blood products. The occurrence of hepatitis after liver transplantation was assessed in the patients who underwent transplantation for cryptogenic liver disease (15 with fulminant hepatic failure and 45 with cryptogenic cirrhosis) with and without HGV RNA in serum. In this analysis, the 6 patients with fulmi-

582

Hoofnagle et al

Table 3. Observed Versus Expected Rates of De Novo Infection With HGV After Liver Transplantation Probability of Exposure

No. of Patients

Average No. Blood Units Exposure

Expected Rate of HGV Exposure

Actual Rate of De Novo HGV Infection

24% to 40% 40% to 60% 60% to 80% 80% to 100%

7 28 15 12

24.5 45.5 80.5 186.8

30.8% 49.5% 69.8% 89.7%

28.6% 60.7% 66.7% 50%

NOTE. x2 5 2.852 (3 df), P 5 .42 for goodness of fit, indicating that there is no evidence of inconsistency with the hypothesis of infection caused by perioperative blood.

nant hepatic failure of known cause were excluded. Based on data obtained at 4 months, 1 year, and 2 years after transplantation, there were no differences in the incidence or degree of aminotransferase elevation between the 38 patients with and the 22 patients without HGV RNA in serum (Table 4). The diagnosis of ‘‘probable’’ or ‘‘definite’’ viral

Table 4. Comparison of Patients With Cryptogenic Acute or Chronic Liver Disease With and Without HGV RNA After Transplantation

Feature Month 4 Median ALT (U/L) ALT abnormal Median AST (U/L) AST abnormal Hepatitis by biopsy* 1 yr Median ALT (U/L) ALT abnormal Median AST (U/L) AST abnormal Hepatitis by biopsy* 2 yr Median ALT (U/L) ALT abnormal Median AST (U/L) AST abnormal Hepatitis by biopsy*

HGV Positive (n 5 38)

HGV Negative (n 5 22)

n

Value

n

Value

P

31 31 37 37 17

29 26% 26 16% 24%

18 18 22 22 8

28 44% 30.5 32% 13%

.55 .22 .72 .20 .64

34 34 37 37 31

34.5 41% 31 32% 10%

15 15 19 19 14

27 27% 25 26% 7%

.33 .36 .38 .76 1.00

28 28 31 31 12

30.5 29% 31 35% 17%

16 16 18 18 6

16.5 19% 21.5 17% 17%

.06 .72 .08 .20 1.00

Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase. *Definite or probable viral hepatitis diagnosis made by pathologist on pathology form. Biopsy at 2 years includes any biopsy performed between years 2 and 5.

hepatitis was made on liver biopsy on at least one occasion in 21 patients (13 with HGV RNA and 8 without) over the 3-year period after transplantation. The frequency of a biopsy-demonstrated diagnosis of hepatitis was no more common among patients with and without HGV RNA in serum (Fig. 1). Comparison of the two groups for patient and graft survival in the usual sense was not possible because the analysis was limited to those patients who were tested for HGV RNA after transplantation, and the samples used were from blood drawn an average of 442 days (HGV RNA negative) and 450 days (HGV RNA positive) after transplantation. However, with adjustment for the exact time of the blood sampling, the two groups had similar death and graft failure rates after blood was drawn (data not shown).

Discussion The hepatitis G virus is a newly described flavivirus that has been shown to cause acute and chronic infection in humans.1-5 Although it was originally described as a hepatitis virus, it is still uncertain whether the virus infects hepatocytes and whether infection causes liver injury.10 The frequency of HGV infection is high among patients with liver disease, but most liver injury in these patients appears to be caused by other factors, such as hepatitis B or C.9 The reason HGV is frequent among patients with chronic hepatitis is that the modes of transmission of these viruses are similar. Not only HBV and HCV, but also HGV, appears to be transmissible by blood transfusion, parenteral drug abuse, and accidental needle stick exposures.9 The role of sexual or close personal contact in spreading HGV is unknown. In this study, we analyzed the frequency and clinical associations of HGV infection among pa-

Hepatitis G Virus and Liver Transplantation

583

Figure 1. Proportion of patients with cryptogenic liver disease who underwent liver transplantation and remained without a biopsy-documented diagnosis of either ‘‘probable’’ or ‘‘definite’’ viral hepatitis after transplantation. The 38 patients who tested HGV RNA positive (solid line) are compared with the 22 patients who tested HGV RNA negative (dotted line) after liver transplantation. The HGV RNA–positive patients include 9 who were positive before transplantation (reinfection) and 35 who became positive after transplantation (de novo infection).

tients undergoing liver transplantation as part of a prospective database analysis of the outcomes of transplantation. HGV infection was found before transplantation in 24% of patients with fulminant hepatic failure, 15% with cryptogenic cirrhosis, 11% with chronic hepatitis C, and 9% with endstage cholestatic liver disease. These rates were not statistically significantly different, but the numbers of patients tested were small and the overall rate of HGV RNA positivity was twice as high among patients with fulminant liver disease (24%) as among those with chronic end-stage liver disease (12%). Although HGV infection was common among patients with fulminant hepatic failure, it was not necessarily the cause of the liver failure. Indeed, HGV infection was present in similar percentages of patients with acute and chronic hepatitis of known and unknown cause. Thus the liver injury in these patients could not be definitively attributed to HGV infection. Instead, HGV infection was more likely caused by the frequency of parenteral and nonparenteral exposures to this virus among these groups of patients. Similar findings have been reported by other groups of investigators.14-16 The frequency of de novo infection with HGV among patients undergoing liver transplantation

who tested negative for this viral RNA before transplantation was striking in this study. In only 1 case of de novo infection could the donor be shown to be the source of the HGV RNA in serum. The most likely source of de novo HGV infection was the blood and blood product transfusions that were administered to patients undergoing liver transplantation. Another possibility is that the high rates of HGV RNA detected in patients after liver transplantation represented reactivation of latent disease that was actually present before transplantation. However, the observed rates of de novo infection in this study were consistent with the rates one would expect given the level of exposure to blood products (median of 55 units) and the current rates of HGV RNA detection among volunteer blood donors in the United States (1.4% to 1.8%).1,9,10 Similar findings in liver transplant recipients have been reported by other groups of investigators.14,17 Despite its frequency, HGV infection after liver transplantation was not associated with a poor outcome or an increase in the rates of histological hepatitis, liver injury, graft failure, or death. Patients with and without HGV RNA on pretransplantation blood samples had similar posttransplantation survivals. Among patients who had HGV RNA detectable in serum after transplantation, the aver-

584

Hoofnagle et al

age alanine transaminase and aspartate transaminase elevations and the rates of occurrence of biopsy-documented hepatitis were similar to those of patients who did not develop HGV infection. These findings are important because of the implications of HGV infection after transplantation. If acute hepatitis or adverse outcomes were more frequent among patients acquiring HGV infection after transplantation, routine screening of blood and organ donors for HGV would be appropriate. Unfortunately, at present the only reliable assays for this viral infection are molecular biological techniques such as RT-PCR, methods that are not easily adapted to routine blood bank use. Development of serological assays for anti-HGV and advances in PCR methodology and other techniques for detecting low levels of virus in serum may ultimately make screening of blood for HGV practical. In the meantime, it is important to better document the clinical implications of this viral infection, and liver transplantation provides an important cohort of patients in whom to evaluate the long-term consequences of HGV infection. These findings indicate that HGV infection is common among patients undergoing liver transplantation for either acute or chronic liver failure and occurs de novo in a high percentage of patients after transplantation. Nevertheless, in this study there was no evidence that HGV infection was associated with more severe disease before transplantation or with the occurrence of hepatitis, liver disease, or decreased survival after transplantation.

Acknowledgment The Liver Transplantation Database is supported by contracts N01-DK-0-2251, N01-DK-0-2252, N01-DK-02253, and N01-DK-0-2254 from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). Members who contributed to this study included Katherine M. Detre, M.D., Dr.P.H. (Principal Investigator), A. Jake Demetris, M.D. (Coinvestigator), Steven Belle, Ph.D. (Coinvestigator), Yuling Wei, M.S. (Project Coordinator), Manuel Lombardero, M.S. (Statistician), Eric Seaberg, M.P.H. (Statistician), Sharon Lawlor, M.B.A. (Data Manager/Analyst), Heather Eng, B.A. (Data Analyst), Shannon Fitzgerald, M.S. (Graduate Assistant), Jacqueline Haber, B.S. (Data Manager Assistant), and Gerald L. Swanson, B.S. (Systems Analyst), from the Department of Epidemiology, University of Pittsburgh Graduate School of Public Health, Pittsburgh, PA; Russell Wiesner, M.D. (Principal Investigator), Ruud Krom, M.D. (Coinvestigator); Michael K. Porayko, M.D. (Coinvestigator), and Lori Schwerman, R.N. (Research Nurse Coordinator), from the Departments of Medicine and

Surgery, Mayo Clinic Foundation, Rochester, MN; Rowen K. Zetterman, M.D. (Principal Investigator), Byers Shaw, Jr., M.D. (Coinvestigator), and Karen Taylor, R.N., B.S.N. (Research Nurse Coordinator), from the Departments of Medicine and Surgery, the University of Nebraska Medical Center, Omaha, NE; Nancy Ascher, M.D., Ph.D. (Principal Investigator), John Lake, M.D. (Principal Investigator), and Cherie Bremer-Kamp, R.N. (Research Nurse Coordinator), from the Departments of Medicine and Surgery, University of California, San Francisco, CA; and James Everhart, M.D., M.P.H. (Program Director), and Jay H. Hoofnagle, M.D. (Division Director), from the Division of Digestive Diseases and Nutrition, NIDDK, National Institutes of Health, Bethesda, MD.

References 1. Linnen J, Wages J Jr, Zhang-Keck ZY, Fry KE, Krawczynski KZ, Alter HJ, et al. Molecular cloning and disease association of hepatitis G virus: A transfusion-transmissible agent. Science 1996;271:505-508. 2. Simons JN, Leary TP, Dawson GJ, Pilot-Matias TJ, Muerhoff AS, Schlauder GG, et al. Isolation of novel virus-like sequences associated with human hepatitis. Nat Med 1995;1:564-569. 3. Leary TP, Muerhoff AF, Simons JN, Pilot-Matias TJ, Erker JC, Chalmers ML, et al. Sequence and genomic organization of GBV-C: A novel member of the Flaviviridae associated with human non-A-E hepatitis. J Med Biol 1996;48:60-67. 4. Simons JN, Pilot-Matias TJ, Leary TP, Dawson GJ, Desai SM, Schlauder GG, et al. Identification of two flavivirus-like genomes in the GB hepatitis agent. Proc Natl Acad Sci USA 1995;92:3401-3405. 5. Schaluder GG, Dawson GJ, Simons JN, Pilot-Matias TJ, Gutierrez RA, Heynen CA, et al. Molecular and serologic analysis in the transmission of the GB hepatitis agents. J Med Virol 1995;46:81-90. 6. Masuko K, Mitsui T, Iwano K, Yamazaki C, Okuda K, Meguro T, et al. Infection with hepatitis GB virus C in patients on maintenance hemodialysis. N Engl J Med 1996;334:1485-1490. 7. Yoshiba M, Okamoto H, Mishiro S. Detection of the GBV-C hepatitis virus genome in serum from patients with fulminant hepatitis of unknown aetiology. Lancet 1995;346:1131-1132. 8. Aikawa T, Sugai Y, Okamoto H. Hepatitis G infection in drug abusers with chronic hepatitis C. N Engl J Med 1996;344:195-199. 9. Alter MJ, Gallagher M, Morris TT, Moyer LA, Meeks EL, Krawczynski K, et al. Acute non–A-E hepatitis in the United States and the role of hepatitis G virus infection. N Engl J Med 1997;336:741-746. 10. Alter HJ, Nakatsuji Y, Melpolder J, Wages J, Wesley R, Shih JW-K, et al. The incidence of transfusion associated hepatitis G virus infection and its relation to liver disease. N Engl J Med 1997;336:747-754. 11. Schmidt B, Korn K, Fickenstein B. Molecular evidence

Hepatitis G Virus and Liver Transplantation

for transmission of hepatitis G virus by blood transfusion. Lancet 1996;908:347-349. 12. Nubling CM, Lower J. GB-C genomes in a high-risk group, in plasma pools, and in intravenous immunoglobulin. Lancet 1996;347:68-69. 13. Wei YL, Detre KM, Everhart JE. The NIDDK Liver Transplantation Database. Liver Transplant Surg 1997; 3:10-22. 14. Berenguer M, Terrault NA, Piatak M, Yun A, Kim JP, Lau JYN, et al. Hepatitis G virus infection in patients with hepatitis C virus infection undergoing liver transplantation. Gastroenterology 1996;111:1569-1575.

585

15. Munoz S, Alter H, Liang TJ, Nakatsuji Y, Shih J, Reddy R, et al. Hepatitis G virus is present in serum of patients with fulminant hepatitis of unknown etiology [abstract]. Hepatology 1996;24:293A. 16. Jeffers LJ, Piatak M, Bernstein DE, Reddy KR, Lifson JD, Yun A, et al. Hepatitis G virus infection in patients with acute and chronic liver disease of unknown etiology [abstract]. Hepatology 1995;22:182A. 17. Berg T, Naumann U, Fukumoto T, Bechstein WO, Neuhaus P, Lobeck H, et al. GB virus C infection in patients with chronic hepatitis B and C before and after liver transplantation. Transplantation 1996;62:711-714.