Bone Marrow Transplantation, (1998) 21, 1131–1135  1998 Stockton Press All rights reserved 0268–3369/98 $12.00 http://www.stockton-press.co.uk/bmt Incidence and influence of GB C and C virus in patients undergoing bone marrow transplantation

H Akiyama1, N Nakamura1, S Tanikawa1, H Sakamaki1, Y Onozawa1, T Shibayama2, S Tanaka2, F Tsuda3, H Okamoto4, Y Miyakawa5 and M Mayumi4

1Hematology Division and 2Liver Unit, Tokyo Metropolitan Komagome Hospital, Tokyo; 3Department of Medical Sciences, Toshiba General Hospital, Tokyo; 4Immunology Division, Jichi Medical School, Tochigi-Ken; and 5Miyakawa Memorial Research Foundation, Tokyo, Japan

Summary: bone marrow donors. Since most patients have underlying diseases impairing their and receive mye- Markers of GB virus C (GBV-C) and virus loablative therapies before BMT, infection with hepatitis (HCV) were sought in 80 patients before and after they tends to persist and may induce severe hepatitis. underwent BMT in a metropolitan hospital in Tokyo Furthermore, the infection may be modulated by the adop- between 1990 and 1996. RNA of GBV-C was detected tive immunity transferred by a bone marrow donor who has in 14 (18%) patients before BMT. Of the 55 patients been infected or who has overcome the infection. who had been transfused, 14 (25%) possessed GBV-C Recently, a putative hepatitis virus designated GB virus RNA at a frequency significantly higher than in the 25 C (GBV-C) or hepatitis G virus (HGV) has been disco- untransfused patients who were all negative (P Ͻ 0.01). vered.2,3 For the sake of convenience these are collectively HCV RNA was detected in three of the 55 (5%) trans- referred to here as GBV-C. This is a positive, single- fused patients, but in none of the 25 untransfused stranded RNA virus of approximately 9400 bases and patients. Sera at 3 months after BMT were available for classified in the family. GBV-C prevails 57 patients. GBV-C RNA persisted in all 10 patients worldwide, as estimated by the detection of its RNA in 1– who were infected before BMT, while it was detected in 5% of apparently healthy blood donors.3–7 GBV-C is trans- five of the remaining 47 (11%) patients who were not. mited by transfusion and intravenous drug abuse.5,8–11 Co- However, persistent and/or ongoing GBV-C infection infection with GBV-C and HBV or HCV is frequent in had no appreciable influence on patient morbidity or patients with chronic liver disease and in intravenous drug mortality. Two of the 57 patients were positive for HCV users.3,8 GBV-C infection persists in patients with compro- RNA before BMT and this persisted after BMT in both. mised immunity, such as those on maintenance hemodia- HCV RNA became positive in eight of the remaining 55 lysis.5 The ability of GBV-C to induce hepatitis has, how- (15%) patients who were negative before BMT. Of the ever, as yet, not been established. 14 patients who received transfusions screened by the Evidence of infection with GBV-C, HCV and HBV was first-generation test at BMT, seven (50%) became posi- sought in patients with malignant hematological diseases or tive for HCV RNA, a rate significantly higher than the aplastic anemia before and after they underwent BMT, and one of 41 (2%) patients who received transfusions the results were correlated with the number of transfusions screened by the second-generation test (P Ͻ 0.001). they received before and at the time of transplantation. These results indicate that BMT patients are at increased risk of GBV-C infection transmitted by trans- fusions received before and at the time of BMT, and Patients and methods that the risk of HCV infection has decreased after the implementation of the second-generation anti-HCV test. Study population Keywords: GBV-C; HCV; BMT From July 1990 to August 1996, a total of 155 patients underwent BMT in the Hematology Division of Tokyo Metropolitan Komagome Hospital. Only the patients who Patients undergoing BMT are at increased risk of infection underwent allogeneic BMT and whose pre-BMT sera are with blood-borne hepatitis viruses such as virus available were included in this retrospective study. One (HBV) and (HCV), which poses serious hundred and twenty-four patients received allogeneic or 1 problems. These viruses are mainly introduced either by syngeneic BMT and in 80 patients, pre-BMT sera were the transfusions they received before and at BMT or by the available (Table 1). They included 44 males and 36 females with the mean (Ϯs.d.) age of 32 (Ϯ9) years. During BMT, complete blood counts were performed Correspondence: Dr H Akiyama, Hematology Division, Tokyo Metropoli- tan Komagome Hospital, 3–18–22, Honkomagome, Bunkyo-ku, Tokyo, three times a week and blood chemistries obtained two to 113, Japan three times a week. Patients were evaluated retrospectively Received 12 September 1997; accepted 29 December 1997 for veno-occlusive disease (VOD) by the criteria of GBV-C and BMT H Akiyama et al 1132 Table 1 Patient characteristics samples of 10-fold dilution of extracted RNAs were pre- pared, and the sample of the highest dilution positive for Characteristics GBV-C RNA was determined. The result was converted to represent the titer (10N) of the virus per ml of the test Age average (range) 32 (17–53) years serum. Sex M/F 44/36 Diagnosis CML 34 Markers of HCV and HBV AML 13 ALL 15 Until January 1992, blood transfused to patients was SAA 8 screened for antibody to HCV (anti-HCV) by the first-gen- MDS 8 Lymphoma 2 eration enzyme-linked immunosorbent assay (ELISA) Donor using Ortho ELISA I (Ortho Diagnostic Systems, Tokyo, related 67 Japan) and thereafter by the second-generation ELISA unrelated 11 (Ortho ELISA II; Ortho Diagnostic Systems). syngeneic 2 Sera from patients were serially diluted two-fold, and Preparative regimen BU/CY 53 anti-HCV was assayed by hemagglutination using Abbott Ara-C/CY/TBI 19 HCV PHA 2nd Generation (Dainabot, Tokyo, Japan). Sera CY/TLI 8 inducing hemagglutination at dilutions Ͼ25 were con- GVHD prophylaxis sidered to be positive for anti-HCV. RNA of HCV was CsA/MTX 76 ␮ Others 4 determined in RNAs extracted from 100 l of the test serum by a commercial kit (Amplicor HCV detection kit; SAA = severe aplastic anemia; MDS = myelodysplastic syndrome; TLI = Japan Roche, Tokyo, Japan). total lymphoid irradiation. Hepatitis B surface antigen (HBsAg) and the correspond- ing antibody (anti-HBs) were determined by serial two-fold dilutions of the test serum using commercial kits (MyCell; McDonald et al,12 GVHD, graft rejection, interstitial pneu- Institute of Immunology Co. Ltd, Tokyo, Japan), and hem- monia, hemorrhagic cystitis and varicella zoster virus infec- agglutination at dilutions у22 was considered reactive. tion. In 57, sera 3 months after BMT were also available, while in 23 patients, follow-up sera were not available even Statistical analysis though all survived for more than 3 months after BMT. Frequencies between groups were compared with the ␹2 test or Fisher’s exact test using StatView (Abacus Concepts, Bone marrow transplantation Berkeley, CA, USA). Group means were compared using Preparative therapy was given mainly according to the pri- the Student’s t-test. mary diseases and source of bone marrow. Patients with AML and CML were treated with busulfan 16 mg/kg and cyclophosphamide 120 mg/kg (BU/CY). Total lymphoid Results irradiation (TLI) was added in patients who received mar- row from unrelated donors. Patients with severe aplastic GBV-C, HCV and HBV markers before BMT anemia were treated with cyclophosphamide 200 mg/kg Table 2 gives prevalence rates of viral markers in the 80 and TLI 700 cGy (CY/TLI) and those with ALL were patients before BMT. GBV-C RNA was detected in 14 treated with Ara-C 8 g/m2 and cyclophosphamide 120 mg/kg, followed by total body irradiation (TBI), 1200 cGy (18%) patients, HCV RNA in three (4%) and HBsAg in none (P Ͻ 0.001). No patients were vaccinated against (Ara-C/CY/TBI). Other diseases were treated with BU/CY HBV. All three patients with HCV RNA were positive for or Ara-C/CY/TBI. All patients except for those undergoing syngeneic BMT received cyclosporine and short-term methotrexate as prophylaxis for GVHD. Table 2 Markers of GB virus C, hepatitis C virus and infections in patients before BMT

Determination of GBV-C RNA Viral markers Total Transfusion P valuea n = 80 (%) ␮ RNAs were extracted from 100 l of serum with an extrac- (+)(−) tion reagent containing guanidinium isothiocyanate and n = 55 (%) n = 25 (%) phenol (ISOGEN-LS; Nippon Gene Co. Ltd, Tokyo, Japan) and dissolved in 5.3 ␮l of distilled water treated with GBV-C RNA 14 (18) 14 (25) 0 Ͻ0.01 diethylpyrocarbonate (Sigma Chemical, St Louis, MO, HCV RNA 3 (4) 3 (5) 0 USA). Samples were reverse transcribed and the cDNA was Anti-HCV 5 (6) 5 (9) 0 HBsAg 0 0 0 amplified by PCR using primers taken from well-conserved Anti-HBs 6 (8) 5 (9) 1 (4) areas in the 5′ untranslated region of the , by the 9 method described previously. aCompared between the patients with previous transfusions and those with- For semi-quantitative analysis of GBV-C RNA, serial out. GBV-C and BMT H Akiyama et al 1133 anti-HCV; no patients had seronegative HCV infection. GBV-C RNA in patients after BMT Previously transfused patients had a higher frequency of In 57 of 80 patients, sera obtained at 3 months after BMT GBV-C RNA than untransfused patients (14 of 55 or 25% were also available. GBV-C RNA persisted in all 10 vs none of 25 or 0%, P Ͻ 0.01). All three patients with patients who had it before BMT. Five of the remaining 47 HCV RNA had received transfusions; two of these were (11%) patients who were negative for GBV-C RNA before positive for GBV-C RNA. BMT acquired it after BMT (de novo infection) (Table 5). RNAs of GBV-C and HCV in patients with various In those five patients, ALT levels before BMT were underlying diseases are shown in Table 3, with special ref- within normal limits (Ͻ40 IU/l), but they increased in three erence to the number of transfused units (best estimated patients after BMT, one of whom was coinfected with HCV number of exposures to blood donors) they had received. and had the highest ALT level. Titers of GBV-C RNA All 13 patients with AML had received 50 or more units ranged from 101/ml to у104/ml in these patients but there of blood, of whom 10 (77%) had GBV-C RNA and two was no correlation between RNA titer and level of ALT. were positive for HCV RNA. Both also possessed GBV-C Their maximum total bilirubin levels were less than RNA. In contrast, of 21 patients with other diseases who 1.4 mg/dl at 3 months and no patients developed VOD. had received у50 units, only two (10%) were positive for All 57 patients received transfusions during the course GBV-C RNA (P Ͻ 0.001). However, patients with diseases of BMT (total: 931 units). There were no differences in the other than AML had received rather fewer blood units numbers of blood units received between the five patients (mean Ϯ s.d.: 143 Ϯ 84 vs 322 Ϯ 292 units). who contracted de novo GBV-C infection and the 42 Only two of the 21 (10%) patients who had received Ͻ50 patients who did not (16 Ϯ 11 vs 14 Ϯ 11). units possessed GBV-C RNA, and neither were positive for HCV RNA. HCV infection in patients after BMT Effects of GBV-C on BMT patients Of the 57 patients whose sera were available at 3 months after BMT, two possessed HCV RNA before BMT which VOD occurred in none of the 14 patients with GBV-C persisted. HCV RNA became positive in eight of the = RNA, and in six of the 66 (9%) patients without (P 0.32). remaining 55 (15%) patients (Table 6). After BMT, ALT In the 14 patients with GBV-C, the mean of the maximum levels increased in seven to higher levels than before BMT. levels of total bilirubin during the first 3 months was A moderately elevated level of ALT (60 IU/l) in the 0.9 mg/dl (0.4–2.1) (Table 4). Serum alanine amino- remaining one patient (Patient No. 51) decreased to normal transferase (ALT) levels were within three times the normal after BMT. Anti-HCV antibody became detectable in range up to 20 days after BMT in all but two patients and seven, while one other patient became positive for anti- within twice the usual range between day 20 and 3 months HCV antibody without HCV RNA. in all but two patients. There were no differences in the Blood products were screened for HCV by the first- incidence of graft-versus-host disease, graft rejection, inter- generation ELISA before February 1992 when the second- stitial pneumonia, hemorrhagic cystitis and varicella zoster generation ELISA was implemented. Of the 55 patients virus infection between patients with and without GBV-C without pre-transplantation HCV RNA, 14 had received RNA. No differences were observed in recovery of transfusions at BMT screened by the first-generation Ͼ × 9 Ͼ WBC to 1 10 /l, reticulocytes 0.5% and platelets ELISA, and eight (57%) of these became positive for HCV Ͼ × 9 50 10 /l, or in the period during which platelet trans- RNA after BMT. By contrast, none of the 41 transfused fusions were required (data not shown). patients screened by the second-generation ELISA developed HCV RNA at a significantly lower frequency Table 3 Prevalence rates of GBV-C RNA and HCV RNA in patients (P Ͻ 0.001). stratified by disease and number of blood units before BMT

Disease n Age No. of Best estimated No. of exposures to Discussion (years) patients the blood donors transfused GBV-C was detected in 14 of the 80 (18%) BMT patients (%) у50 Ͻ50 whose sera were available. In addition, RNA of GBV-C became detectable in five of the 47 (11%) patients within n GBV-C HCV n GBV-C HCV 3 months of BMT. Inclusive of the 10 patients who were RNA RNA RNA RNA infected before BMT, 15 of the 57 patients (26%) were positive for GBV-C RNA after BMT. The frequency of CML 34 35 ± 8 11 (32) 1 1 0 10 0 0 ALL 15 31 ± 9 15 (100) 10 0 0 5 1 0 GBV-C infection in the BMT patients was much higher AML 13 28 ± 8 13 (100) 13 10 2 0 than in the general population of Japan, which is reported SAA 8 23 ± 3 8 (100) 4 1 0 4 0 0 as 1.2%.13 Hence, patients undergoing BMT are at MDS 8 34 ± 9 7 (88) 6 0 1 1 1 0 increased risk for GBV-C infection, as was the case for ± NHL 2 30 11 1 (50) 0 1 0 0 HBV and HCV before screening and exclusion of contami- 1 Total 80 32 ± 9 55 (69) 34 12 3 (9%) 21 2 0 nated blood units. These results confirm a recent report (35%) (10%) from the UK documenting the detection of HGV RNA in 20 of 33 (61%) recipients of BMT14 and a report of 42% MDS = myelodysplastic syndrome; NHL = non-. from Spain.15 GBV-C and BMT H Akiyama et al 1134 Table 4 Patients with GBV-C RNA in serum before BMT

Patient No. Age and sex Disease Transfused units Maximum ALTb Maximum total bilirubin

Before BMT At BMT Before Ͻday 20 Ͻ3 months Before Ͻ3 months HCV (IU/l) (mg/dl) RNA

48 28F AML 170 4 19 90 Ͻ40 0.6 0.7 − 52 25F AML 480 46 19 134 Ͻ40 0.9 1.1 + 70 33M ALL 8 24 16 42 46 0.6 1.1 − 75 41M AML 165 7 33 44 51 0.3 0.4 − 77 20F AML 70 7 27 182 Ͻ40 0.8 1.0 − 85 26M SAA 364 20 8 53 39 0.3 0.8 − 93 27M AML 224 7 45 68 Ͻ40 0.2 1.4 − 97 39F AML 970 15 22 85 59 0.6 0.9 + 106 39F CML 340 18 27 85 Ͻ40 0.8 0.7 − 110 21M AML 577 22 12 20 159 0.5 0.7 − 112 45M MDS 27 20 7 12 133 0.5 1.1 − 137 25F AML 100 4 18 73 Ͻ40 0.5 0.4 − 146 19F AML 290 12 19 41 Ͻ40 0.9 0.7 − 155 23M AML 850 100 21 90 61 0.3 2.1 −

aTransfused units, best estimated number of exposures to the blood donors. bALT, alanine aminotransferase (normal: Ͻ40 IU/l).

Table 5 Patients who developed GBV-C RNA in serum after BMT

Patient Age and Disease Transfused unitsa Before BMT 3 months after BMT Anti- No. sex HCVd Before At BMT Anti-HCVc ALTb GBV-C HCV RNA Anti-HCVd ALTb GBV- HCV BMT (IU/l) RNA (IU/l) C RNA RNA

53 27F CML 16 34 I 10 −− − 409 102/ml +− 121 24F SAA 158 11 II 11 −− − Ͻ40 у104/ml −− 131 28M CML 0 16 II 22 −− − 69 102/ml −− 152 41M ALL 85 11 II 24 −− − 90 у104/ml −− 182 40F AML 100 7 II 6 −− − Ͻ40 101/ml −−

aTransfused units; best estimated number of exposures to the blood donors. bALT, alanine aminotransferase (normal: Ͻ40 IU/l). cAnti-HCV, the method of anti-HCV measurement of blood products used at BMT (Ortho ELISA I or II). dAnti-HCV, data of anti-HCV of patients’ sera measured by Abbott HCV PHA second generation.

Table 6 Patients who were infected with HCV after BMT

Patient No. Age and sex Disease Transfused unitsa Before BMT 3 months after BMT

Before At BMT Anti-HCVc ALTb HCV Anti-HCVd ALTb HCV Anti-HCVd BMT (IU/l) RNA (IU/l) RNA

30 40F CML 16 10 I 16 −− 116 ++ 43 26M CML 4 9 I 10 −− 114 ++ 47 24M SAA 46 15 I 23 −− 115 ++ 51 44M ALL 60 7 I 30 −− Ͻ40 ++ 53 27F CML 16 34 I 10 −− 409 +− 60 32F CML 0 34 I 7 −− 806 ++ 62 31M CML 2 14 I 42 −− 1005 ++ 73 27M ALL 46 6 I 41 −− 293 ++

aTransfused units, best estimated number of exposures to the blood donors. bALT, alanine aminotransferase (normal: Ͻ40 IU/l). cAnti-HCV, the method of anti-HCV measurement of blood products used at BMT (Ortho ELISA I or II). dAnti-HCV, data of anti-HCV of patients’ sera measured by Abbott HCV PHA second generation. GBV-C and BMT H Akiyama et al 1135 None of the blood units or bone marrow donors were ongoing GBV-C infection on the morbidity or mortality in tested for GBV-C RNA. However, there is evidence to indi- the patients who underwent BMT. cate of GBV-C to patients by transfusions they received before and during BMT. Before BMT, GBV-C References RNA was detected significantly more frequently in patients who had received transfusions than in those who had not 1 Shuhart MC, McDonald GB. Gastrointestinal and hepatic (25 vs 0%, P Ͻ 0.01). Furthermore, GBV-C RNA was complications. In: Forman SJ, Blume KG, Thomas ED (eds). Bone Marrow Transplantation. Blackwell Scientific: Massa- detected more frequently in patients with AML who chusetts, 1994, pp 454–481. received more units of blood before BMT than in patients 2 Leary TP, Muerhoff AS, Simons JN et al. Sequence and gen- with the other diseases (77 vs 10%, P Ͻ 0.001). omic organization of GBV-C: a novel member of the Flavivir- The incidence of VOD was not raised in patients with idae associated with non-A-E hepatitis. J Med Virol GBV-C, and neither was the incidence of severe liver dam- 1996; 48: 60–67. age during the first 3 months after BMT. Maximum levels 3 Linnen J, Wages J, Zhang-Keck ZY et al. Molecular cloning of ALT were within three times the normal range and the and disease association of hepatitis G virus: a transfusion- maximum levels of total bilirubin during the 3 months were transmissible agent. Science 1996; 271: 505–508. 4 Dawson GJ, Schlauder GG, Pilot-Matias TJ et al. Prevalence less than 2.1 mg/dl in patients with GBV-C, suggesting an studies of GB virus-C infection using reverse transcriptase- insignificant association between GBV-C and acute liver polymerase chain reaction. J Med Virol 1996; 50: 97–103. toxicity, which is also consistent with previous reports.14,15 5 Masuko K, Mitsui T, Iwano K et al. Infection with hepatitis At BMT, 657 blood units (plus 47 bone marrow GB virus C in patients on maintenance hemodialysis. New donations) were given to the 47 patients negative for GBV- Engl J Med 1996; 334: 1485–1490. C RNA. Assuming 1.2% of blood donors were infected 6 Brown KE, Wong S, Buu M et al. High prevalence of GB with GBV-C, at least eight contaminated blood units virus C/hepatitis G virus in healthy persons in Ho Chi Minh were transfused. City, Vietnam. J Infect Dis 1997; 175: 450–453. 7 Wang Y, Chen HS, Fan MH et al. Infection with GB virus C ALT levels after BMT increased in three of the five and hepatitis C virus in hemodialysis patients and blood (60%) patients who acquired GBV-C RNA. One of the donors in Beijing. J Med Virol 1997; 52: 26–30. three patients with GBV-C RNA, however, became positive 8 Aikawa T, Sugai Y, Okamoto H. Hepatitis G infection in drug for HCV RNA which could have been responsible for the abusers with chronic hepatitis C (letter). New Engl J Med elevated ALT. ALT levels were within three times the nor- 1996; 334: 195–196. mal range at most in these patients. In view of a variety 9 Shimizu M, Osada K, Okamoto H. Transfusion-transmitted of factors potentially inducing impaired liver function in hepatitis G virus following open heart surgery (letter). Trans- fusion 1996; 36: 937. recipients of BMT, elevated ALT levels in the two patients 10 Wang JT, Tsai FC, Lee CZ et al. A prospective study of trans- infected with GBV-C cannot readily be ascribed only to fusion-transmitted GB virus C infection: similar frequency but 1 GBV-C. different clinical presentation compared with hepatitis C virus. On the other hand, HCV RNA is a more reliable marker Blood 1996; 88: 1881–1886. of HCV infection than is anti-HCV in BMT patients, since 11 Alter HJ, Nakatsuji Y, Melpolder J et al. The incidence of impaired immune responses can prevent seroconversion. transfusion-associated hepatitis G virus infection and its HCV RNA became detectable in eight patients after BMT relation to liver disease. New Engl J Med 1997; 336: 747–754. and one of these was negative for anti-HCV. HCV RNA 12 McDonald GB, Sharma P, Matthews DE et al. Venocclusive disease of the liver after bone marrow transplantation: diag- was detected in the recipients of blood units screened by nosis, incidence, and predisposing factors. Hepatology 1984; the first-generation anti-HCV ELISA significantly more fre- 4: 116–122. quently than in those screened by the second-generation 13 Yoshikawa A, Fukuda S, Itoh K et al. Infection with hepatitis anti-HCV ELISA (57 vs 0%, P Ͻ 0.001). Thus, HCV infec- G virus and its strain variant, the GB agent (GBV-C), among tion in recipients of BMT appears to have decreased as the blood donors in Japan. Transfusion 1997; 37: 657–663. sensitivity of the tests increases. 14 Skidmore SJ, Collingham KE, Harrison P et al. High preva- These results indicate that patients with BMT are at lence of hepatitis G virus in bone marrow transplant recipients increased risk for GBV-C infection transmitted to them by and patients treated for acute leukemia. Blood 1997; 89: 3853–3856. transfusions they receive before and at the time of BMT, 15 Rodriguez-Inigo E, Tomas JF, de Soria VGG et al. Hepatitis and that the risk of HCV infection has decreased after the C and G virus infection and liver dysfunction after allogeneic implementation of the second-generation anti-HCV test. All bone marrow transplantation: results from a prospective study. in all, there were no appreciable effects of persistent and/or Blood 1997; 90: 1326–1331.