Transfusion-Associated TT Virus Infection and Its Relationship to Liver Disease

Transcription

1 Transfusion-Associated TT Virus Infection and Its Relationship to Liver Disease AKIHIRO MATSUMOTO, 1 ANTHONY E. T. YEO, 1 J. WAI KUO SHIH, 1 EIJI TANAKA, 2 KENDO KIYOSAWA, 2 AND HARVEY J. ALTER 1 TT virus (TTV) has been proposed as the causative agent of non-a to E hepatitis. We studied the association between TTV viremia and biochemical evidence of hepatitis in blood donors and prospectively-followed patients. TTV was found in 7.5% of 402 donors and in 11.0% of 347 patients before transfusion. The rate of new TTV infections was 4.7% in 127 nontransfused, and 26.4% in 182 transfused patients (P F.0001). The risk of infection increased with the number of units transfused (P F.0001). The rate of new TTV infections in 13 patients with non-a to E hepatitis (23.2%) was almost identical to the rate in 124 patients who were transfused, but did not develop hepatitis (21.8%). Of 45 patients with acute hepatitis C, 40.0% were simultaneously infected with TTV. TTV did not worsen the biochemical severity (mean ALT: 537 in TTV ; 550 in TTV ) or persistence of hepatitis C. In non-a to E cases, the mean ALT was 182 in those TTV positive and 302 in TTV negatives. No consistent relationship between alanine transaminase level and TTV DNA level was observed in 4 patients with long-term, sequential samples. Of 21 viremic subjects, 67% cleared TTV within 5 years (38% in 1 year); 33% were viremic throughout follow-up extending to 22 years. We conclude that TTV is a very common, often persistent infection that is transmitted by transfusion and by undefined nosocomial routes. We found no association between TTV and non-a to E hepatitis and no effect of TTV on the severity or duration of coexistent hepatitis C. TTV may not be a primary hepatitis virus. (HEPATOLOGY 1999;30: ) TT virus (TTV) was discovered in 1997 by representational difference analysis of serial serum specimens from Japanese patients with transfusion-associated non-a to G hepatitis. 1 Because of a temporal association between TTV viremia and elevations in serum alanine transaminase (ALT) level in 3 of 5 patients studied, TTV was proposed as the causative agent of non-a to G hepatitis. 1 TTV was initially described as a single-stranded DNA virus of approximately 3,800 base pairs (bp) and considered to be a member of the parvovirus family. 2 Abbreviations: TTV, TT virus; ALT, alanine aminotransferase; HCV, hepatitis C virus; PCR, polymerase chain reaction; CI, confidence interval; HGV, hepatitis G virus; RDA, representational difference analysis; GBV-C, GB virus, Type C. From the 1 Department of Transfusion Medicine, National Institutes of Health, Bethesda, MD; the 2 Second Department of Internal Medicine, Shinshu University School of Medicine, Matsumoto, Japan. Received January 8, 1999; accepted April 9, Address reprint requests to: Dr. Harvey J. Alter, Department of Transfusion Medicine, Building 10, 1C-711, Clinical Center, National Institute of Health, Bethesda, MD halter@dtm.cc.nih.gov; fax: This is a U.S. Government work. There are no restrictions on its use /99/ $0.00/0 283 A subsequent analysis suggested that it was a single-stranded circular DNA virus in the circovirus family. 3 These genomic characterizations have been questioned and, at present, the agent has not been definitively classified. Seroepidemiological studies have shown TTV to have global distribution. 2,4-6 Although the potential association of TTV with cryptogenic hepatitis is intriguing, the pathological and clinical significance of this virus remains to be established. To assess more fully the etiological role of TTV in the causation of hepatitis, we determined the frequency of acute TTV infections and their relationship to hepatitis in a cohort of prospectivelyfollowed transfusion recipients who did and did not develop hepatitis and in nontransfused controls. Prevalence of the agent was assessed in blood donors and in patients at the time of hospital admission. MATERIALS AND METHODS Selection of Patients in Various Cohorts. Serum samples were obtained from donors in the NIH donor base who resided in the greater Washington region and from donors to the Chesapeake Region of the American Red Cross. Recipients were enrolled in prospective studies of transfusion-associated hepatitis at the National Institute of Health from October 1972 to December The methods of enrollment, frequency, and duration of patient follow-up and criteria for the diagnosis of hepatitis have been described previously. 7 Study protocols and approvals were reviewed by the appropriate institutional review boards, and all subjects gave written informed consent. Five cohorts were examined for TTV seroprevalence and for the incidence of new infections. They included 402 volunteer blood donors; 140 patients who had surgery but did not require transfusion nor develop hepatitis; 142 patients who were transfused and did not develop hepatitis; 13 patients who developed transfusionassociated non-a to E hepatitis; and 52 patients who developed transfusion-associated hepatitis C. TTV DNA was measured in the presurgery sample and the 6- to 8-week postsurgery/transfusion sample in all patients. In 4 patients with newly-acquired, transfusion-associated infection, TTV DNA was quantitated in serial samples to assess the temporal relationship between TTV DNA level and ALT level. All samples had been stored at 70 C before testing for TTV DNA. Detection of TTV DNA and HCV RNA by Polymerase Chain Reaction. Nucleic acids were extracted from serum samples (100 µl) with proteinase K and sodium dodecyl sulfate. 8 The extracted DNA was dissolved in 15 µl of 12% dimethyl sulfoxide in distilled, deionized water. Perkin-Elmer AmpliTaq Gold DNA polymerase (Roche Molecular Systems, Inc., Branchburg, NJ) with digoxigenin labeling mix (200 µmol/l of datp, dctp, and dgtp each; 190 µmol/l dttp, and 10 µmol/l digoxigenin-11-dutp) were used in the polymerase chain reaction (PCR) procedure. Seminested consensus primers NG059 (5 -ACAGACAGAGGAGAAGGCAACAT-3 ) and NG063 (5 -CTGGCATTTTACCATTTCCAAAG-3 ) were used for the first round PCR. 2 This entailed incubation at 95 C for 10 minutes, followed by 5 cycles at 94 C for 45 seconds; 55 C for 60 seconds;

2 284 MATSUMOTO ET AL. HEPATOLOGY July 1999 and 72 C for 60 seconds; then 55 cycles at 94 C for 60 seconds; 55 C, 60 seconds; finally an additional 7 minutes at 72 C. An amplified product of 286 bp was obtained. The second round PCR was performed with NG061 (5 -GGCAACATGTTATGGATAGACTG-3 ) 2 and NG063 for 25 cycles. This involved incubation at 95 C for 10 minutes; then 5 cycles at 94 C for 45 seconds; 55 C for 60 seconds; and 72 C for 60 seconds; then 20 cycles at 94 C for 60 seconds; 55 C for 60 seconds; and, finally an additional 7 minutes at 72 C. A 271-bp product was obtained. The amplified product was then placed on a 2% agarose gel and electrophoresed. Gels were stained with ethidium bromide and observed under ultraviolet light. Procedures for HCV PCR were performed as described elsewhere. 9 Quantification of TTV DNA and HCV RNA. TTV DNA was quantified using a positive serum external standard. The serially diluted standard contained TTV DNA concentrations, as determined by a limiting dilution method, ranging from 1 to 100,000 equivalents per 100 µl. Detection was performed by the Enzymun-Test DNA detection system on an automated ES300 machine (Boehringer Mannheim GmbH, Indianapolis, IN). First round PCR amplified products were diluted 1:10 with denaturation solution (50 mmol/l NaOH). A total of 100 µl of a diluted sample was mixed with 400 µl of hybridization solution. This consisted of phosphate buffered saline, ph 6.5, containing the 5 end biotinylated capture probe (TTV-CP: 5 -TATGGATAGACTGGCTAAGTAAA- 3 ) at a concentration of 150 ng/ml. The subsequent hybridization for 120 minutes at 37 C was performed in a streptavidin-coated sample tube. The tube was then washed and TTV-specific DNA detected by binding the biotin-labeled nucleic acids to the streptavidin solid phase and by adding antidigoxigenin peroxidase conjugate in Tris-HCV buffer (ph 7.5). The mixture was incubated for 30 minutes at 37 C. After a final wash, an enzyme substrate containing 1.9 mmol/l 2,2 azino-di (3-ethylbenzthiazoline sulfonate), diammonium salt (ABTS) in 100 mmol/l phosphate citrate buffer (ph4.4)-3.3 mmol/l H 2 O 2 in sodium perbotate was added. A color reaction was allowed to develop for 30 minutes, and the optical density (OD) was measured at Å405. The copy number of each sample was determined in relation to the standard curve and expressed as the number of equivalents per ml. HCV RNA was quantified by a commercial assay (Monitor, Roche Molecular Systems, Branchberg, NJ) according to the directions of the manufacturer. Serological and Chemical Assays. Antibodies to HCV were measured by a second-generation enzyme immunoassay according to the manufacturer s instructions (Abbott HCV EIA 2.0, Abbott Laboratories, North Chicago, IL.). Antibody to hepatitis E virus was measured by the method of Tsarev et al. 10 ALT was measured by a 3-point kinetic assay with an automated photometric analyzer (model 917, Hitachi-Boehringer Mannheim, Indianapolis) and values expressed in international units (normal range: 5-41 U/L). Statistical Analysis. Student s t-test and ANOVA were used to analyze all continuous variables. Duncan s multiple range test was used to differentiate cohorts as needed. The 2 test with Yates correction was used in the analysis of categorical data. Where the number of subjects was less than 5, Fisher s exact test was used. The Mantel-Haenszel 2 test was used for trend analysis. Statistical analyses were performed using SPSS (version 7.5, SPSS Inc., Chicago, IL). RESULTS Prevalence of TTV in Volunteer Blood Donors. The demographics of the 402 volunteer blood donors studied were: 211 male/191 female; average age: 43.8 years (95% CI: ); race: 352 white, 31 black, 9 Asian, 10 other. The overall prevalence of TTV in this donor population was 7.5% (30 of 402). The prevalence of TTV DNA according to race was 5.4% in 352 whites, 16.1% in 31 blacks, 55.6% in 9 Asians, and 10.0% in 10 persons of others races (Table 1). The TTV TABLE 1. TTV DNA Frequency in Volunteer Donors According to Race Race n TTV DNA Total (7.5%) White (5.4%) Black 31 5 (16.1%) Asian 9 5 (55.6%)* Others 10 1 (10.0%) *vs. White: P.00009, vs. Black: P.029, vs. Others: P.057 (Fisher s exact test). White vs. Black: P.035 (Fisher s exact test). prevalence in Asians was significantly higher than that in other races (vs. white, P.0001; vs. black, P.029). A significant difference was also noted between whites and blacks (P.035). Age and sex were not associated with TTV prevalence. Prevalence of TTV in Patients Before Surgery/Transfusion. The prevalence of TTV in a sample obtained from patients before surgery or transfusion was 11.0% (38 of 347). This high prevalence in patients was not significantly different from the 7.5% prevalence observed in blood donors (P.13, Table 2). The background prevalence in patients showed no relationship to their subsequent requirement for blood transfusion or whether the patient developed hepatitis during prospective follow-up. Newly-Acquired TTV Infections in Relation to Blood Transfusion. Excluding the 38 patients who were TTV positive before surgery, TTV DNA was detected following surgery in 48 of 182 (26.4%) patients who were transfused compared with 6 of 127 patients (4.7%) who were not transfused (P.00001, Table 3). This strong association with blood transfusion was further shown when TTV incidence was stratified according to transfusion volume (Fig. 1). There was a stepwise increase in TTV infections as transfusion volume increased from 0 unit to greater than 13 units (P.00001). This stepwise trend was less apparent in patients who developed HCV infection than in those who did not develop hepatitis because the strong co-association of HCV and TTV made it more likely that a patient receiving an HCV positive blood unit would also be exposed to TTV, even if the total number of blood units received were small. Incidence of TTV Infections in Patients Not Transfused. Six of 127 patients (4.7%) who underwent open-heart surgery but were not transfused acquired TTV infection in the perioperative period (Table 3). This raised the question of whether these represented nosocomial TTV infections or activation of a latent infection. Because no antibody assay was available to assess prior TTV exposure, we addressed the possibility of reactivation by comparing the TTV DNA level pre- and postsurgery in those who had pre-existing TTV viremia (Fig. 2). The mean TTV DNA levels in 38 patients before surgery was 3.3 log equivalents/ml (95% CI: ). Although some patients had peri-operative increases in TTV DNA level, an equal number sustained a fall in level and the TABLE 2. Prevalence of TTV DNA in Volunteer Donors and Patients Group n TTV DNA Volunteer donors (7.5%) a Patients* (11.0%) b NOTE. a vs. b: No statistically significant difference (P.13). *Sampled before surgery/transfusion.

3 HEPATOLOGY Vol. 30, No. 1, 1999 MATSUMOTO ET AL. 285 TABLE 3. Relationship of Blood Transfusion to the Rate of Acute TTV Infections in Patients Undergoing Open Heart Surgery Group n New TTV Infections* Nontransfused (4.7%) Transfused (26.4%) NOTE. Nontransfused vs. transfused: P *Sampled at 6 to 8 weeks after surgery. Excluded 13 patients TTV DNA before surgery. Excluded 25 patients TTV DNA before surgery. mean level after surgery (3.3 log equivalents/ml [95% CI: ]) was identical to that before surgery. This suggested that neither transfusion nor other peri-operative events uniformly activated TTV replication, and by inference, that TTV was acquired by nosocomial routes in the 4.7% of patients who became TTV DNA positive in the absence of transfusion. Incidence of TTV Infections in Transfused Patients in Relationship to Hepatitis Outcome. Excluding 25 transfused patients with pre-existing TTV DNA, newly-acquired TTV infections were observed during follow-up in 48 of 182 (26.4%) patients (Table 4). The rate of acute TTV infections showed no relationship to hepatitis outcome. Among 13 transfused patients classified as non-a to E hepatitis, 3 (23.2%) became TTV DNA positive following transfusion; this rate was almost identical to that in 124 transfusion recipients who did not develop hepatitis (21.8%). A significantly higher rate of incident TTV infections (40.0%) was seen in patients with transfusion-associated HCV (Table 4). The higher incidence of TTV infection in patients with HCV correlated with transfusion volume. Patients with HCV received significantly (P.004) more blood (mean: 16.9 units; CI: ) compared with patients without hepatitis (mean: 9.2 units; CI: ) or patients with non-a to E hepatitis (mean 11 units; CI: ). When corrected for transfusion volume, there was no significant association between TTV infection and whether or not the recipient developed hepatitis. Relationship of TTV Infection to ALT Level. In the non-a to E hepatitis cohort, the mean peak ALT levels for the 3 subjects FIG. 2. Concentration of TTV before and after surgery in 38 subjects. In 75% (30 of 38) of subjects, TTV concentrations before and after surgery were within 2 log equivalent/ml. The mean TTV concentration before surgery was 3.3 equivalent/ml (95% CI: ) and after surgery was 3.3 equivalent/ml (95% CI: ). with TTV infection was 182 IU/L (95% CI: ) and for the 10 subjects without TTV infection was 302 IU/L (95% CI: ) (Table 5). Though the trend was for a lower ALT level in those infected with TTV, the difference was not significant. In the HCV hepatitis cohort (Table 5), the mean peak ALT level for 18 subjects with TTV infection was 537 IU/L (95% CI: ) compared with 550 IU/L (95% CI: ) in the 27 subjects without TTV infection (P.05). Although there was no association between TTV and ALT level, patients with HCV had significantly higher mean ALT levels (545 IU/L) than patients with hepatitis non-a to E (274 IU/L) (P.05). Thus, significantly higher levels of ALT were observed in patients with HCV than in patients with hepatitis unrelated to HCV, but the coexistence of TTV infection did not influence peak ALT levels in either of these hepatitis populations (Table 5). Relationship of TTV Status to Persistence of HCV Infection. Chronic HCV (duration greater than 1 year) developed in 16 of 18 (88.9%) patients who were coinfected with HCV and TTV compared with 21 of 27 (77.8%) patients who had transfusion-associated HCV in the absence of TTV infection (Table 5). This difference in the HCV chronicity rate was not significant (P.45). Clinical Course of TTV Infection. The clinical course of TTV viremia in an HCV and TTV co-infected patient and in 3 patients with non-a to E hepatitis is detailed in Fig. 3 and 4, TABLE 4. Incidence of TTV Infection in Transfused Patients in Relationship to Hepatitis Outcome Hepatitis Outcome n* New TTV Infections Mean No. Transfusions (95% CI) No hepatitis (21.8%) a 9.2 ( ) d Non-A to E hepatitis 13 3 (23.2%) b 11.0 ( ) e HCV hepatitis (40.0%) c 16.9 ( ) f Total (26.3%) 11.3 ( ) FIG. 1. Relationship between the frequency of TTV infection and number of units of blood transfused. There was a statistical significant association between the number of blood units transfused and the prevalence of TTV (P ). NOTE. a vs. b: Not statistically significant (P 1.00), a vs. c: P.030. d vs. e vs. f: P.0041 (One-way ANOVA). d vs. f: P.05 (Duncan s test). *Excludes 25 patients who tested TTV DNA positive before transfusion.

4 286 MATSUMOTO ET AL. HEPATOLOGY July 1999 TABLE 5. Associations of TTV With Peak ALT Levels in Patients With Hepatitis Group n Peak ALT (IU/L; 95% CI.) HCV Chronicity Posttransfusion non-a to E hepatitis TTV ( ) a TTV ( ) b Posttransfusion HCV hepatitis TTV ( ) c 16 (88.9%) e TTV ( ) d 21 (77.8%) f NOTE. a vs. b vs. c vs. d: P.0041 (One-way ANOVA). a b vs. c d: P.05 (Duncan s test). a vs. b: No statistically significant difference. c vs. d: No statistically significant difference. e vs. f: No statistically significant difference (Fisher s exact test, P.45). respectively. In the HCV and TTV co-infected subject (Fig. 3), ALT and HCV RNA levels were approximately synchronous in their rise and fall. Notably, HCV RNA rose early and sharply before the first appearance of elevated ALT and declined to baseline immediately preceding the fall in ALT. In contrast, TTV DNA remained at high level after the ALT had normalized, declined to nondetectable at week 108 when the ALT was rising and rose again at week 120 and beyond when the ALT had again normalized. Among the non-a to E hepatitis subjects (Fig. 4), in case A, TTV and ALT levels were in parallel for the first 12 to 14 weeks, after which they generally dissociated. In case B, the initial rise in TTV DNA preceded the rise in ALT by 10 weeks; by week 32 TTV DNA had become undetectable but ALT levels remained elevated. In case C, TTV levels remained at a low-level plateau despite progressively increasing ALT levels. Overall, there was no clear relationship between TTV DNA level and ALT level in any of the cases that were carefully observed over time. Persistence of TTV Infection. Serial samples sufficient to assess the persistence of TTV infection were available from 21 patients. Of these, 14 (67%) cleared TTV DNA during the course of follow-up (Fig. 5). Eight patients (38%) became TTV DNA negative within 1 year of the onset of their FIG. 4. TTV Infection in 3 subjects with transfusion-associated hepatitis. Levels of ALT (shaded areas) and TTV DNA ( ) in 3 subjects infected only with TTV are plotted against time. Qualitative results of PCR for TTV DNA are shown (, TTV present;, TTV absent). The horizontal dashed lines indicate the normal ALT level. The vertical dashed lines indicate a point of blood transfusion. In each patient, the relation between the TTV DNA and ALT was inconsistent. In case A, both TTV DNA and ALT level fluctuated but the peaks of TTV DNA and peaks of ALT level were not synchronized. In case B, the peaks of TTV DNA and peaks of ALT level were again not synchronized. Moreover, even though TTV became undetectable after 32 weeks, ALT levels remained elevated. In case C, TTV levels progressively decreased despite increasing ALT levels. FIG. 3. Combined TTV and HCV co-infection in a subject with transfusion-associated hepatitis. The levels of ALT (shaded areas), TTV DNA ( ) and HCV RNA ( ) are shown plotted against the time. Qualitative results of PCR (, virus present;, virus absent) are shown. The horizontal dashed lines indicate the normal ALT level. The vertical dashed lines indicate the point of blood transfusion. HCV RNA and ALT levels were synchronized whereas TTV and ALT levels showed no relationship. infection whereas another 6 patients (29%) became TTV negative in the succeeding 5 years. Seven subjects (33%) remained TTV DNA positive throughout follow-up. Although follow-up was truncated by death in 2 patients and refusal for further sampling in a third, none of the patients who had viremia for 6 years lost the virus in subsequent testing. In 1 patient, viremia was documented for 22 years.

5 HEPATOLOGY Vol. 30, No. 1, 1999 MATSUMOTO ET AL. 287 FIG. 5. Long-term follow-up of TTV viremia in 21 subjects. Fourteen of 21 (66.7%) subjects became TTV negative during observation period. One patient was not followed from 4 to 10 years after surgery because of geographical relocation (a). However, when refollowed, TTV was not detected. Eight of 38 (33.3%) subjects became TTV negative within 1 year. Two subjects died from heart disease (b) and lung cancer (d). One patient has refused to be followed up since 4 years after the surgery (c). In 4 subjects, TTV persisted for more than 7 years. DISCUSSION A substantial proportion of cases of transfusion-associated hepatitis, 11 community-acquired hepatitis, 12 fulminant hepatitis, 13 and unexplained chronic hepatitis/cirrhosis 14 cannot be accounted for by infection with established hepatitis viruses A, B, C, D, and E. This has led to the continued search for new hepatitis agents. The discovery of HCV using blind cloning and immunoscreening 15 established the principle that viruses could be identified through molecular technology before they had been physically characterized, grown in culture, or serologically defined. Molecular approaches similar to those employed for HCV led to the subsequent discovery of the hepatitis G virus (HGV) 16 and representational difference analysis (RDA) was used to discover another candidate non-a to E hepatitis agent that was designated GBV-C. 17 It was later shown that HGV and GBV-C were variants of the same agent 16 and that neither was a likely cause of unexplained hepatitis in humans Hence, the search for viral causes of cryptogenic liver disease continued. In 1997, Nishizawa et al. 1 using RDA, found a novel agent in the serum of 3 of 5 patients with transfusion-associated hepatitis. This agent was designated TT virus (TTV) after the initials of the first patient in whom it was discovered. A series of publications suggested that TTV was a parvovirus 2 and that it was the probable causative agent for some cases of cryptogenic acute and chronic hepatitis. 1,2 The current study was designed to test the hypothesis that TTV was causally related to transfusionassociated hepatitis. Using primers and probes similar to those reported by Okamoto et al. 2 we found TTV in high prevalence (7.5%) in United States volunteer blood donors and in 11% of patients at the time of hospital admission. There was a markedly higher prevalence in donors of Asian origin (55.6%), but the number of Asian donors studied was very small. The prevalence of TTV in Japanese donors has been reported to be 11%, and similar rates have now been found throughout the world. It is to be noted that these prevalences are measures of active viremia and not simply cumulative measures of past exposure. Hence, TTV is an exceedingly common, globally distributed agent. 2,4,5 In the United States, the rate of viremia far exceeds that for HCV (0.3%), HBV (0.1%), and HGV (1.5%). It is clear that TTV is parenterally transmitted. In our study, we found a very strong association with blood transfusion when comparing patients who were prospectively followed after open heart surgery; 26% of transfused patients were acutely infected with TTV compared with 5% of patients who were not transfused (P.0001). In addition, there was a strong association (P.0001) between transfusion volume and TTV infection with a step-wise increase in risk reaching 35% in those who received 13 or more units of blood. Unexpectedly, we found that near 5% of patients undergoing open-heart surgery were acutely infected with TTV in the peri-operative period in the absence of blood transfusion. This suggested either nosocomial transmission of TTV by routes that are currently undefined or reactivation of a latent infection. In the absence of an established antibody assay for TTV, we were unable to directly assess prior exposure to this agent and the potential for reactivation of latent infection. In lieu of this limitation, we assessed peri-operative changes in TTV DNA level in patients who were viremic at the time of admission. As shown in Fig. 2, although TTV DNA level increased in some patients following surgery, this seemed to be a random event and the mean TTV DNA level was identical before and after surgery. The relatively stable DNA levels do not suggest increased TTV replication as the result of the multiple events that surround hospitalization and surgery and indirectly argue against reactivation of latent virus in this setting. More likely, the finding of new TTV infections in 5% of nontransfused subjects reflects transmission by nonparenteral or hidden parenteral routes. In this regard, it is of interest that TTV has been found in bile and stool 21 implying that this agent may be disseminated by the fecal-oral route in addition to its parenteral modes of transmission. The prevalence of an infectious agent reflects both its ease of dissemination and its rate of persistence. Among 21 TTV infected patients followed long-term, we found that 38% cleared the virus within 1 year, that 29% had long-term infection, but ultimately cleared the virus within 5 years and, that 33% had persistent infection throughout follow-up that ranged as long as 22 years. Although the rate of persistent TTV infection is not as high as that for HCV (85%), 7 its higher prevalence probably reflects the fact that it is disseminated by nonparenteral routes as well as sharing with HCV a variety of parenteral transmission patterns. The key element of this study was the availability of serial samples from prospectively followed patients who did and did not develop transfusion-associated hepatitis. We found that the rate of newly-acquired TTV infections was virtually identical in patients who did not develop hepatitis (22%) as it was in patients diagnosed as acquiring non-a to E hepatitis (23%). There was a significantly higher rate of TTV infection in those who developed HCV (40%) as compared with the other populations studied. There are 2 likely explanations for this observation. First, patients who developed HCV received significantly more blood than those who did not develop hepatitis (mean 16.9 units vs. 9.2 units) and hence had a greater chance of exposure to the highly prevalent TTV agent.

6 288 MATSUMOTO ET AL. HEPATOLOGY July 1999 Second, the donor who transmitted HCV would have an increased likelihood of also transmitting TTV because of the co-association of these agents through shared parenteral exposures. Although TTV and HCV were frequently associated, there was no evidence that TTV increased the biochemical severity of coexistent HCV or that it increased the likelihood that HCV infection would become chronic. Similarly, in patients classified as non-a to E hepatitis, there was no evidence that TTV increased the biochemical severity of the hepatitis. Further, in patients who were not transfused, but nonetheless acquired peri-operative TTV infection, none developed hepatitis. Finally, when serial ALT levels were compared with serial TTV DNA levels there was poor temporal correlation and a general dissociation between these 2 values. In summary, we found no evidence that TTV infection was more frequent in patients with hepatitis than in identically followed patients without hepatitis, no evidence that TTV was the cause of non-a to E hepatitis or that it worsened the severity or persistence of coexistent HCV, no evidence that acute TTV infection in the absence of transfusion resulted in hepatitis and no consistent correlation between TTV DNA level and ALT level. The data from this study therefore provide no support to the concept that TTV is a primary hepatitis virus. One cannot exclude the possibility that TTV could cause hepatitis as a secondary manifestation of its pathogenic spectrum, such as would be the case for Epstein Barr Virus or cytomegalovirus, but there is little to support this from our data or that in the literature. Also, one cannot exclude the possibility that a particular strain of TTV might be more hepatotropic and more likely to cause liver injury than more common strains, as has been proposed. 2,4 A recent study that utilized novel primers has suggested that TTV is far more prevalent than initially reported, with rates in Japan that exceed 90%. 22 This higher prevalence, if confirmed, would not alter the conclusions of our study, namely that TTV is a common, often persistent infection that does not appear to be specifically associated with post-transfusion hepatitis and that does not worsen the course of coexistent HCV. Indeed, the higher prevalence would further strengthen these conclusions. It is probable that TTV like HGV/GBV-C, is an innocent bystander in cases of cryptogenic hepatitis that are a result of an as yet unrecognized hepatitis agent or to nonviral factors. In one respect, we have become victims of the sophistication of our technology. In probability, RDA and related molecular techniques will continue to uncover new infectious agents. Because these agents will have been sought in the blood of persons with particular diseases, the first presumption on discovery will be that they are the cause of the disease under investigation. When the agent is very prevalent in the general population, as is the case for HGV and TTV, such a presumption of causality is inherently flawed. Very carefully selected controls, large numbers of patients, sequential samples before and after infection, coded panels and documentation of intra-hepatic replication are required before a newly discovered agent might warrant classification as a hepatitis virus. TTV has not reached that level of scientific validity and, based on the data in this study, it is unlikely that it will. Acknowledgment: We thank Professor Keiji Mitamura of Showa University, Tokyo, Japan, for providing control serum samples. REFERENCES 1. Nishizawa T, Okamoto H, Konishi K, Yoshizawa H, Miyakawa Y, Mayumi M. A novel DNA virus (TTV) associated with elevated transaminase levels in post-transfusion hepatitis of unknown etiology. Biochem Biophys Res Commun 1997;241: Okamoto H, Nishizawa T, Kato N, Ukita M, Ikeda H, Iizuka H, Miyakawa Y, et al. Molecular cloning and characterization of a novel DNA virus (TTV) associated with posttransfusion hepatitis of unknown etiology. Hepatology Res 1998;10: Takahashi K, Hoshino H, Ohta Y, Yoshida N, Mishiro S. Very high prevalence of TT virus (TTV) infection in general population of Japan revealed by a new set of PCR primers. Hepatol Res 1998;12: Simmonds P, Davidson F, Lycett C, Prescott LE, MacDonald DM, Ellender J, Yap PL, et al. Detection of a novel DNA virus (TTV) in blood donors and blood products. Lancet 1998;352: Naoumov NV, Petrova EP, Thomas MG, Williams R. Presence of a newly described human DNA virus (TTV) in patients with liver disease. Lancet 1998;352: Charlton M, Adjei P, Poterucha J, Zein N, Moore B, Therneau T, Krom R, et al. TT-virus infection in North American blood donors, subjects with fulminant hepatic failure, and cryptogenic cirrhosis. HEPATOLOGY 1998; 28: Alter HJ, Purcell RH, Shih JW, Melpolder JC, Houghton M, Choo QL, Kuo G. Detection of antibody to hepatitis C virus in prospectively followed transfusion recipients with acute and chronic non-a, non-b hepatitis. N Engl J Med 1989;321: Okamoto H, Yotsumoto S, Akahane Y, Yamanaka T, Miyazaki Y, Sugai Y, Tsuda F, et al. Hepatitis B viruses with precore region defects prevail in persistently infected hosts along with seroconversion to the antibody against e antigen. J Virology 1990;64: Alter HJ, Sanchez-Pescador R, Urdea MS, Wilber JC, Lagier RJ, Di Bisceglie AM, Shih JW, et al. Evaluation of branched DNA signal amplification for the detection of hepatitis C virus RNA. J Viral Hepat 1995;2: Tsarev SA, Tsareva TS, Emerson SU, Kapikian AZ, Ticehurst J, London W, Purcell RH. ELISA for antibody to hepatitis E virus (HEV) based on complete open reading frame-2 protein expressed in insect Cells: Identification of HEV infection in primates. J Infect Dis 1993;168: Alter HJ. Transfusion transmitted hepatitis C and non-a, non-b, non-c. Vox Sang 1994;67(Suppl 3):19S-24S. 12. Rochling FA, Jones WF, Chau K, DuCharme L, Mimms LT, Moore B, Scheffel J, et al. Acute sporadic non-a, non-b, non-c, non-d, non-e hepatitis. HEPATOLOGY 1997;25: Ferraz ML, Silva AE, Macdonald GA, Tsarev SA, Di Biscelgie AM, Lucey MR. Fulminant hepatitis in patients undergoing liver transplantation: evidence for a non-a, non-b, non-c, non-d, and non-e syndrome. Liver Transpl Surg 1996;2: Kodali VP, Gordon SC, Silverman AL, McCray DG. Cryptogenic liver disease in the United States: further evidence for non-a, non-b, and non-c hepatitis. Am J Gastroenterol 1994;89: Kuo G, Choo QL, Alter HJ, Gitnick GL, Redeker AG, Purcell RH, Miyamura T, et al. An assay for circulating antibodies to a major etiologic virus of human non-a, non-b hepatitis. Science 1989;21: Linnen J, Wages J, Yong Z, Keck Z, Fry KE, Krawczynski KZ, Alter HJ, et al. Molecular Cloning and disease association of hepatitis G virus: a transfusion- transmissible agent. Science 1996;271: Simons JN, Leary TP, Dawson GJ, Pilot-Matias TJ, Muerhoff AS, Schlauder GG, Desai SM, et al. Isolation of novel virus-like sequences associated with human hepatitis. Nature Med 1995;1: Alter HJ, Nakatsuji Y, Melpolder J, Wages J, Wesley R, Shih JWK, Kim JP. The incidence of transfusion-associated hepatitis G virus infection and its relation to liver disease. N Eng J Med 1997;336: Tanaka E, Alter HJ, Nakatsuji Y, Shih JWK, Kim JP, Matsumoto A, Kobayashi M, et al. Effect of hepatitis G virus infection on chronic hepatitis C. Ann Intern Med 1996;125: Alter HJ. G-pers creepers, where d you get those papers? A reassessment of the literature on the hepatitis G virus. Transfusion 1997;37: Okamoto H, Akahane Y, Ukita M, Fukuda M, Tsuda F, Miyakawa Y, Mayumi M. Fecal excretion of a non-enveloped DNA virus (TTV) associated with posttransfusion non-a to G hepatitis. J Med Virol 1998;56: Takahashi K, Ohta Y, Mishiro S. Partial 2.4kb sequences of TT virus (TTV) genome from 8 Japanese isolates: diagnostic and phylogenetic implications. Hepato Res 1998;12: