676 J Clin Pathol 1998;51:676–678

Hepatitis G in lymphoma and in J Clin Pathol: first published as 10.1136/jcp.51.9.676 on 1 September 1998. Downloaded from blood donors

Jane Minton, Adil Iqbal, Ayesha Eskiturk, William Irving, John Davies

Abstract that it causes persistent infection as it has been Aims—To determine whether the recently detected in the blood of patients at least nine described flavivirus, G virus years after initial infection.1 However, several (HGV), might contribute to the pathogen- studies have failed to demonstrate serious liver esis of lymphoma by testing for the disease in patients with HGV viraemia.6–8 presence of HGV RNA in sera from Other persistent viral , for example patients attending lymphoma clinics; to infection with lymphotropic retrovirus compare the incidence of HGV RNA in (HTLV1) and Epstein-Barr virus, have been lymphoma patients with that in normal implicated in the pathogenesis of lymph- blood donors; and to look for potential risk oma.9–11 While flaviviruses have yet to be shown factors for HGV infection and for evidence to have oncogenic potential, the related flavivi- of hepatic disease in the HGV positive rus HCV is associated with hepatocellular car- patients. cinoma. There is some evidence that HCV Methods—Sera were examined from 76 infects lymphocytes in addition to hepato- patients with lymphoma and 100 blood cytes12 13; HCV is also associated with cryoglob- donors for the presence of HGV RNA ulinaemia and possibly Waldenstrom’s macro- using reverse transcriptase polymerase globulinaemia, and has been suggested as a chain reaction (RT-PCR). cause of lymphoma.14–16 Recently, it was Results—HGV RNA was detected in 10% suggested that HGV might also infect periph- of patients’ sera, but only in 1% of blood eral blood mononuclear cells.17 We asked donor samples. HGV infection was found whether HGV might contribute to the patho- in patients with various diVerent types of genesis of lymphoma by testing for the lymphoma, including Hodgkin’s disease presence of HGV RNA in sera from patients and non-Hodgkin’s lymphoma. The ma- attending lymphoma clinics with neoplastic jority (75%) of patients who were HGV disorders. We chose to look for presence of PCR positive had undergone transfusion, viraemia rather than for evidence of previous but only 30% of those who were HGV PCR infection, since viral oncogenesis is usually a negative had received blood products. In feature of persistent rather than transient addition, the number of donor exposures infection, and because some of our subjects are per HGV positive patient was approxi- http://jcp.bmj.com/ immunosuppressed, with potential impairment mately twice that of the HGV negative of their ability to generate antibody responses. group. In addition we looked for potential risk factors Conclusions—The data suggest (1) that HGV is present in the normal blood donor for HGV infection, particularly blood donor population; (2) that patients with lymph- exposures, and for evidence of hepatic disease. University Hospital, oma are at risk of acquiring HGV because We also compared the incidence of HGV RNA Queen’s Medical in our patients with that in normal blood

of their exposure to blood products; and (3) on October 2, 2021 by guest. Protected copyright. Centre, Nottingham, donors. UK: that persistent HGV infection does not Department of appear to cause serious liver disease in Microbiology and these patients. Infectious Diseases (J Clin Pathol 1998;51:676–678) J Minton Methods W Irving Keywords: ; lymphoma; blood transfusion Subjects were consecutive patients attending the lymphoma clinic at Queen’s Medical Cen- Department of tre, Nottingham, during the months of July and Haematology A Iqbal Recently, a novel RNA virus designated August 1996. Informed verbal consent for the J Davies hepatitis G virus (HGV) was identified in the study was obtained. Serum alanine amino- blood of a patient with chronic virus transferase (ALT) was measured in each Marmara University, (HCV) infection. HGV has been cloned and patient. The number of donor exposures for Istanbul, Turkey: sequenced, and has been shown to be a flavi- each patient with haematological malignancy Department of Microbiology virus with 95% identity at the amino acid level was calculated from blood bank records. Con- A Eskiturk with GBV-C virus, previously described in the trols consisted of 100 consecutive donors blood of patients with acute hepatitis; and to be attending National Blood Authority (NBA) Correspondence to: distantly related to HCV.12HGV appears to be sessions in Trent on a single day in August, Dr J Minton, Department of 1996. All blood donor samples had previously Infectious Diseases, Royal transmitted parenterally as it occurs more often Hallamshire Hospital, in subjects who have been multiply transfused, been tested for evidence of infection with HIV, Glossop Road, SheYeld including those with haematological malig- virus (HBV), HCV, and syphilis, S10 2JF, UK. nancy, and in intravenous drug abusers.3–5 following the NBA protocols. Samples from Accepted for publication So far, the clinical significance of HGV is not each donor were anonymised, according to 13 May 1998 well understood. Preliminary evidence suggests NBA recommendations. Hepatitis G in lymphoma and blood donors 677

SERUM SAMPLES Table 1 HGV infection in normal blood donors and

Serum was collected from each subject, divided patients with lymphoma J Clin Pathol: first published as 10.1136/jcp.51.9.676 on 1 September 1998. Downloaded from into aliquots, and stored at −80°C. Positive HGV positive HGV negative control serum samples were obtained from patients with HCV infection who had been Subjects No of subjects (% of group) Total found to have HGV viraemia in a previous Blood donors 1 (1.0) 99 (99.0) 100 study. Negative controls consisted of DEPC Lymphoma treated water processed at the same time as the patients 7 (9.2) 69 (90.8) 76 serum samples. Table 2 Characteristics of patients with lymphoma shown REVERSE TRANSCRIPTASE-POLYMERASE CHAIN to be HGV positive REACTION FOR HGV Age RNA was extracted from 100 ml serum from Patient Diagnosis (years) ALT each patient as well as from positive and nega- 1DiVuse large B cell lymphoma 71 Normal tive controls using the acid guanidinium Large granular lymphocytic thiocyanate–phenol–chloroform method of 2 leukaemia, NK cell type 73 Normal Chomczynski and Saachi.18 RNA pellets were 3 Follicular lymphoma 39 Normal High grade B Cell lymphoma, resuspended in 20 ml DEPC treated water. 4 Burkitt-like 64 Abnormal Eight microlitres of RNA per patient served as Hodgkin’s disease, nodular a template for first strand cDNA synthesis, 5 sclerosing 34 Normal Hodgkin’s disease, lymphocyte using the Pharmacia First Strand cDNA 6 predominance 38 Normal synthesis kit according to the manufacturer’s Hodgkin’s disease, nodular instructions (Pharmacia Biotech, Uppsala, 7 sclerosing 37 Normal Sweden). Two microlitres of the cDNA prod- ALT, alanine transaminase. uct were used as a template for first round polymerase chain reaction (PCR), with the Table 3 Frequency of transfusion in lymphoma patients primers GS1 (AGGTGGTGGATGGGT- HGV positive HGV negative GAT) and GAS1 (TGCCACCCGCCCT- CACCCGAA) directed at the conserved 5' Blood products No of subjects (% of group) non-coding region of the HGV/GBV-C ge- Yes 5 (71.4) 21 (30.4) nome. One microlitre of the first round No 2 (28.6) 48 (69.6) product was used as template for second round Total 7 69 PCR, with the primers GS2 (TGGTAG- GTCGTAAATCCCGGT) and GAS2 (GGR GCTGGGTGGCCYCATGCWT), to give a oma and Hodgkin’s disease. In addition the final product of 343 base pairs.5 Products were stage of disease varied from advanced lymph- detected by ethidium stained gel electrophore- oma to patients in remission for many years. sis. Great care was taken to prevent false posi- Only one of these patients had abnormal tive results from carry over of PCR products, liver function tests, and in this case the patient

according to the recommendations of Kwok had other potential causes for hepatic dysfunc- http://jcp.bmj.com/ and Higuchi.19 tion, including aggressive disease with probable liver involvement and methotrexate treatment. None of our HGV positive patients had evid- DETECTION OF HCV RNA ence of concurrent HCV infection Sera from patients positive for HGV were tested . Analysis of transfusion histories (table 3) for HCV RNA by reverse transcriptase polymer- showed that patients with HGV infection were ase chain reaction (RT-PCR) followed by more likely to have undergone blood transfu-

hybridisation using a specific oligonucleotide on October 2, 2021 by guest. Protected copyright. sion than the uninfected patients (p < 0.05, probe (Roche Amplicor hepatitis C virus (HCV) relative risk = 4.8, 95% CI 1.0 to 23.1). In test; Roche Products, Welwyn Garden City, addition, the number of donor exposures per UK), according to the manufacturer’s instruc- HGV infected patient was approximately twice tions. Testing was carried out in duplicate. that of the uninfected patients.

STATISTICAL ANALYSIS Data were analysed using the Fisher exact test. Discussion Our data show that patients with lymphoma We calculated relative risks from the crude have an increased incidence of HGV viraemia odds ratios together with 95% confidence compared with normal blood donors. Our intervals (CI) to demonstrate the likelihood of finding that approximately 10% of such HGV viraemia in the blood donor group com- patients are HGV positive is lower than the pared with the lymphoma group. incidence (47%) observed by Neilson and col- leagues in a study of long term survivors of Results haematological malignancy in Birmingham.3 We found HGV RNA in the serum of seven of However, over half the patients examined by 76 patients with haematological malignancy Neilson et al had undergone bone marrow (9.2%), but only in one of 100 normal blood transplantation, while in our study only two donors (1%) (table 1; p < 0.03, relative risk = patients (one HGV PCR positive and one 2.1, 95% CI 1.6 to 2.9). The characteristics of negative) had undergone this procedure. In the HGV positive patients with lymphoma are addition, some of our patients were still under- presented in table 2. The diagnoses of these going treatment, while the Birmingham pa- patients included both non-Hodgkin’s lymph- tients were considered to be “probably cured.” 678 Minton, Iqbal, Eskiturk, et al

Subjects receiving blood products—such as We are grateful to Dr R Sokol, National Blood Authority Trent Centre, for providing us with anonymised blood donor samples. multiply transfused patients, haemophiliacs, J Clin Pathol: first published as 10.1136/jcp.51.9.676 on 1 September 1998. Downloaded from intravenous drug users, and those on mainte- nance haemodialysis—are much more often 1 Linnen J, Wages J, Zhang-Keck Z-Y, et al. Molecular cloning and disease association of hepatitis G virus: a transfusion- infected with HGV, suggesting that the virus is transmissible agent. Science 1996;271:505–8. 145 transmitted parenterally. Our patients with 2 Simons JN, Leary TP, Dawson GJ, et al. Isolation of novel lymphoma had had multiple blood donor virus-like sequences associated with human hepatitis. Nature Med 1995;1:564–9. exposures, so they were at risk of infection with 3 Neilson J, Harrison P, Milligan DW, et al. Hepatitis G virus HGV.Those patients with HGV infection were in long-term survival of haematological malignancy. Lancet 1996;347:1632. significantly more likely to have received blood 4 Aikiwa T, Sugal Y, Okamoto H. Hepatitis G infection in products than the HGV negative patients. drug abusers with chronic hepatitis C. N Engl J Med 1996; 334:195–6. Thus transfusion is a probable explanation for 5 Jarvis LM, Davidson F, Hanley JP, et al. Infection with the increased incidence of HGV in the hepatitis G virus amongst recipients of plasma products. Lancet 1996;348:1352–5. lymphoma patients compared with the normal 6 Alter MJ, Gallagher M, Morris TT, et al. Acute non-A-E blood donor controls. It is also possible that hepatitis in the United States and the role of hepatitis G immunosuppression may be a risk factor for virus infection. N Engl J Med 1997;336:741–6. 7 Alter HJ, Nakatsuji Y, Melpolder J, et al. The incidence of HGV viraemia in some of the patients, either transfusion-associated hepatitis G virus infection and its because they are more susceptible to initial relation to liver disease. N Engl J Med 1997;336:747–54. 8 Miyakawa Y, Mayumi M. Hepatitis G virus—a true hepati- infection or because they have high titres of tis virus or an accidental tourist? N Engl J Med circulating virus which can be more easily 1997;336:795–6. 9 Wong-Staal F, Gallo RC. Human T-lymphotropic retrovi- detected by RT-PCR. ruses. Nature 1985;317:395–403. Our results do not suggest that HGV has a 10 Zur Hausen H, Schulte-Holtehausen H, Klein G. EBV DNA in biopsies of Burkitt’s tumors and anaplastic role in the pathogenesis of lymphoma or the carcinomas of the nasopharynx. Nature 1970;228:1056–8. subsequent outcome. However, this cannot be 11 Weiss LM, Movahed LA, Warnke RA. Detection of Epstein- absolutely excluded because transfusion may Barr virus in Reed-Sternberg cells of Hodgkin’s disease. N Engl J Med 1989;320:502–6. have had a confounding eVect on our results, 12 Zignego AL, Macchia D, Monti M, et al. Infection of and because our series included patients with peripheral mononuclear cells by . J Hepatol 1992;15:382–6. several diVerent forms of lymphoma and both 13 Moldvay J, Deny P, Pol S, et al. Detection of hepatitis C aggressive and quiescent disease. A larger study virus RNA in peripheral blood mononuclear cells of infected patients by in situ hybridisation. Blood 1994;83: of patients sampled before they receive blood 269–73. products will be necessary to determine 14 De vita S, Sansonno D, Dolcetti R, et al. Hepatitis C virus within a malignant lymphoma lesion in the course of type II whether HGV is associated with a particular mixed cryoglobulinemia. Blood 1995;86:1887–92. haematological diagnosis, or whether it has any 15 Pozatto G, Mazzaro C, Crovatto M, et al. Low-grade malig- influence on disease outcome. nant lymphoma, hepatitis C virus infection, and mixed cry- oglobulinemia. Blood 1994;84:3047–53. HGV has been suggested as a cause of both 16 Silvestri F, Baccarani M. Hepatitis C virus-related lympho- fulminant and chronic hepatitis.20 21 However, mas. Br J Haematol 1997;99:475–80. 17 Madejon A, Fogeda M, Bartoloma J, et al. GB virus C RNA when careful transfusion histories are obtained, in serum, liver, and peripheral blood mononuclear cells it has been shown that HGV infection occurs from patients with chronic hepatitis B, C, and D. Gastroen- terology 1997;113:573–8. only after the onset of fulminant hepatitis and 18 Chomczynski P, Saachi N. Single-step method of RNA iso- 22 after the patient has received blood products. lation by acid guanidinium thiocyanate-phenol-chloroform http://jcp.bmj.com/ extraction. Anal Biochem 1992;162:156–9. Where prospective studies of post-transfusion 19 Kwok S, Higuchi R. Avoiding false positives with PCR. HGV infection have been carried out, HGV can Nature 1989;339:237–8. 20 Yoshiba M, Okamoto H, Mishiro S. Detection of the cause a mild hepatitis with ALT values usually GBV-C hepatitis virus in serum from patients with less than 200 U/litre, though there may often be fulminant hepatitis of unknown aetiology. Lancet 1995;346: no rise in liver enzymes.723 Most patients with 1131–2. 21 Fiordalisi G, Zanella I, Mantero G, et al. High prevalence of persistent HGV infection do not have evidence GB virus C infection in a group of Italian patients with hepa- of liver disease, and with HCV does titis of unknown aetiology. J Infect Dis 1996;174:181–3. 22 Heringlake S, Osterkamp S, Trautwein C, et al. Association on October 2, 2021 by guest. Protected copyright. not appear to influence the course of chronic between fulminant hepatic failure and a strain of GB virus HCV infection.682425 In addition, most C. Lancet 1996;348:1626–9. 23 Hwang S-J, Lu R-H, Chan C-Y, et al. The role of hepatitis G studies—including our own—suggest that HGV virus infection in patients with acute posttransfusion hepa- RNA is present in the sera of approximately titis in Taiwan. Gastroenterology 1997;112:1260–4. 157 24 Bralet MP, Roudot-Thoraval F, Pawlotsky JM, et al. 1–2% of apparently healthy blood donors. Histopathologic impact of GB virus C infection on chronic Thus there is little evidence that HGV causes hepatitis C. Gastroenterology 1997;112:188–92. significant liver disease and the clinical im- 25 Nakatsuji Y, Shih JW, Tanaka E, et al. Prevalence and disease association of hepatitis G virus infection in Japan. J portance of the virus remains unclear. Viral Hepat 1996;3:307–16.