Cytomegalovirus (CMV*), a DNA virus belonging to the herpes family, is an ubiquitous agent; 50-80% of people develop CMV antibodies at some time during their lives (1). The potential severity of CMV infection in immunosuppressed patients is well established (2), and an association of CMV with human immunodeficiency virus disease progression and increased risk of death has been demonstrated (3, 4). In addition to producing protean clinical manifestations including fever, hepatitis, gastrointestinal disease, and pneumonitis in the transplant setting, CMV infection seems to have a direct immunosuppressive effect and is a risk factor for superinfection with opportunistic pathogens, i.e., Gramnegative bacteria, Pneumocystis carinii, and fungi (5, 6). Studies in transplant recipients have shown that CMV infection is associated with leukopenia, a decrease in the T-helper/T-suppressor cell ratio, modulation of non-HLA-restricted cells that kill virus-infected cells including natural killer cells and macrophages, and suppression of cell-mediated immunity (7). It has been proposed that CMV infection potentiates allograft rejection by enhancing the expression of class I and class II histocompatibility antigens and by activating nuclear factor-κB, a transcription factor involved in stimulating a broad range of genes, including those that have significant roles in inflammatory responses (1, 8). In the clinical setting, however, a direct cause and effect relationship between CMV infection and rejection has been difficult to demonstrate because acute rejection often precedes CMV infection. On the other hand, the proposed increased risk of rejection after CMV infection may be related to the reduction in basal immunosuppression that historically had been a routine part of the treatment approach for CMV infection before the availability of ganciclovir.
Hepatitis C virus (HCV) is one of the leading indications for liver transplantation (LT), and although recurrent histological hepatitis is extremely frequent, only a subset of patients seem to develop severe allograft damage (9, 10). Several interrelated host and viral factors have been proposed to explain the variable outcomes, including HCV viral load, genotype and quasispecies, rejection, and the level of immunosuppression (11-13). We sought to investigate how CMV viremia influences the histopathological course of HCV recurrence after LT.
MATERIAL AND METHODS
Patients. The study group was comprised of 43 consecutive patients who underwent LT with cyclosporine-based immunosuppression at our center; all patients had end-stage liver disease secondary to chronic hepatitis C with or without alcoholism and had at least 6 months of histologic follow-up. HCV was diagnosed by ELISA 2.0 (Abbott Laboratories, Chicago, IL) and/or polymerase chain reaction (PCR) as described previously (14). Donor/recipient CMV serostatus was either D-/R- (n=18), D-/R+ (n=9), D+/R+ (n=13), or D+/R- (n=3). CMV prophylaxis consisted of acyclovir at 200 mg p.o. t.i.d for the first 120 days after LT (in 3 patients) or ganciclovir at 10 mg/kg/day i.v. every 12 hr for 7 days followed by acyclovir at 800 mg p.o. q.i.d. for 120 days (in 40 patients).
Virological methods. HCV viremia was assessed by reverse transcription-PCR using conserved 5′-noncoding region primers and a 32P-labeled internal oligonucleotide probe as described previously (14). The reverse transcription-PCR assay sensitivity is less than 100 copies HCV RNA per milliliter of patient serum. HCV genotype was determined by restriction fragment length polymorphism of the 5′-noncoding region as described by Davidson et al. (15), and HCV genotypes were classified according to the nomenclature of Simmonds et al. (16).
CMV infection was surveyed prospectively by protocol and defined by viral isolation from buffy coat or urine using conventional cell culture methods. Mean number (±SEM) of CMV cultures per patient was 11.9±1.56. CMV isolates were frozen at -70°C for further study. DNA was extracted for analysis using a PCR-compatible DNA lysis buffer. Primers for amplification and sequencing were 17-to 21-base oligonucleotides that were selected from the published envelope glycoprotein B (gB) sequences of CMV strain AD169 (17). Primers gB1319 and gB1604 amplified a region of high peptide variability in the gB gene that encodes a portion of gp55 (bases 1319-1604) and yielded a 293- to 296-base pair target, depending on the gB group (18). To identify the genotype to which a strain belonged, restriction enzyme digestion was done on amplified products as previously described (18). The gB amplimer was digested with restriction enzymes HinfI and RsaI, and the digested DNA was analyzed on an 8% polyacrylamide gel. Results of this analysis were used to classify the clinical CMV isolates according to one of four gB types (1 through 4) as previously described (19).
Histological assessment. Liver biopsies were obtained by protocol at 7, 14, and 365 days, as indicated by an elevation from the baseline liver function biochemistries, and after treatment for acute cellular rejection. Acute cellular rejection was diagnosed on the basis of biochemical dysfunction and a liver allograft biopsy showing at least two of the three following features: endophlebitis, damage to interlobular bile ducts by infiltrating mononuclear cells, and portal inflammation. HCV recurrence was diagnosed by an allograft biopsy with portal and lobular infiltration by mononuclear cells, piecemeal necrosis, and/or acidophil bodies in the absence of any other specific causes. Disease severity was determined by the Knodell hepatic activity index (20), applied to each biopsy in a blinded fashion by two pathologists (C.L.C. and A.B.), and a consensus score was assigned. The final allograft biopsy was the most recent biopsy at the time of last follow-up, retransplantation, or death. Graft failure (GF) was defined as severe allograft dysfunction due solely to HCV recurrence and leading to death, retransplantation, or listing for retransplantation.
Statistical analysis. One-way analysis of variance and the Mantel-Haenszel chi-square test were used for statistical comparison of means (±SEM) and proportions between groups, respectively. The Kaplan-Meier product-limit estimate was used for the univariate analysis of time-dependent events (i.e., time to allograft cirrhosis) with comparison between groups performed via the log-rank test. Logistic regression analysis with odds ratio (OR) and the corresponding 95% confidence intervals were used to assess the independent effects of CMV viremia on cirrhosis and GF. A P value of less than 0.05 was considered significant. The JMP (SAS Institute Inc, Cary, NC) statistical package was used.
Mean (±SEM) follow-up to date of death or last clinic visit for the cohort was 804.5±65.9 days (range 180-2133). The excretion of virus in urine or isolation from the throat have not been shown to be predictive of CMV disease (1); consequently, for the purposes of this study, two patient groups were identified with respect to CMV viremia. Group 1 (n=8) was the patients with CMV viremia after LT, and group 2 (n=35) was the patients who did not develop detectable CMV viremia. One patient in the latter group did shed CMV in a single urine culture, but blood cultures remained persistently negative. There were no statistically significant differences between the groups with regard to age, initial immunosuppression, incidence of rejection, CMV prophylaxis, HCV genotypes, or mean follow-up (Table 1). The interval to initial HCV recurrence was comparable in both groups (143.4±94.7 days vs. 220.9±48.2 days, for groups 1 and 2, respectively; P=0.47). The proportion of patients developing histological evidence of recurrence within the first year was similar (7 of 8 [87.5%] vs. 23 of 35 [66%]; P=0.4).
The mean total Knodell score of the final allograft biopsy was significantly greater in group 1 patients. Most of the difference was due to the periportal/bridging necrosis and lobular activity subitem scores (Table 2). The mean interval to the development of CMV viremia was not significantly different between patients who developed cirrhosis as compared with those who did not (70.5±26.58 vs. 69.4±23.7 days). In addition, there was no significant difference with regard to the frequency of CMV sampling between the two groups (P=0.6).
The proportion of patients who developed allograft cirrhosis and/or subsequent GF attributable solely to HCV recurrence was significantly higher in patients with CMV viremia. Half (4 of 8) of the CMV viremic patients developed cirrhosis and 37.5% (3 of 8) developed GF versus an 11% (4 of 35) cirrhosis rate and a 5.7% (2 of 35) GF rate in the CMV-negative patients (P=0.027 for cirrhosis and P=0.034 for GF). Accordingly, the cirrhosis-free actuarial survival was significantly diminished in the CMV viremic patients (P=0.012; Fig. 1). The duration of CMV viremia was variable, ranging from 1 to 58 days, and the duration of viremia did not seem to impact on outcome parameters. Two patients were still on acyclovir prophylaxis when they developed CMV viremia, and three patients with CMV viremia were not treated with antiviral therapy because of spontaneous resolution of infection. Only three patients developed CMV disease; patients 1 and 2 (in Table 3) and one additional patient (who eventually developed mild HCV recurrence) had CMV hepatitis. Symptomatic CMV disease was treated with ganciclovir for 3 to 8 weeks.
Table 3 outlines the specific characteristics of the eight patients who developed cirrhosis. All of these patients had HCV genotype 1: 5 patients had HCV genotype 1a and 3 patients had genotype 1b; 28 of the 35 (80%) patients without allograft cirrhosis had genotype 1a or 1b (P=NS). All of the cirrhotic patients had received CMV prophylactic therapy with intravenous ganciclovir for 1 week followed by oral acyclovir for 120 days. Patient 1 died of GF after developing severe HCV recurrence 944 days after LT. Patients 2, 4, and 5 have been listed for hepatic retransplantation after developing cirrhosis at 395, 380, and 302 days, respectively, after LT. Patient 3 was retransplanted 839 days after LT for HCV-related GF; he died of sepsis and multiorgan failure (although his allograft was functioning well) 3.5 months after his second LT. Patients 6, 7, and 8 have developed well-compensated cirrhosis and have not merited consideration for repeat LT. Of note, only 1 of the 8 patients was treated for acute cellular rejection. Logistic regression analysis identified CMV viremia as a strong independent risk factor for the development of cirrhosis (P=0.02; OR, 7.74; 95% confidence interval, 1.38-47.9) and GF (P=0.025; OR, 9.89; 95% confidence interval, 1.34-91.5) from HCV recurrence. We analyzed the distribution of CMV gB genotypes among patients who developed cirrhosis (Table 3) as compared with those who did not. The CMV genotypes of the latter group (n=4) were as follows: 2, 4, 1 and 3 mix, and 1 and 4 mix.
It has been well documented that recurrence of HCV viremia after LT is almost universal, and histological evidence of allograft hepatitis is present in 44-67% of patients within the first year (21, 22). With follow-up to 3 years, the natural history of HCV recurrence seems to be benign in the majority of recipients, although a subset may develop severe allograft damage (9). We hypothesized that CMV coinfection after LT could be a factor in disease progression of HCV recurrence. Comparison of the pretransplant clinical features and initial immunosuppression of patients with and without CMV viremia revealed no significant differences between the two groups. Although CMV viremia was not associated with an increased incidence of or shorter interval to histological HCV recurrence, it was associated with significantly diminished cirrhosis-free survival long term. Interestingly, rejection, which has previously been implicated as a significant cofactor in the development of severe HCV recurrence (13), had a somewhat lower prevalence in the CMV viremic patients, although the difference was not statistically significant.
Previous studies have linked the development of severe HCV recurrence with the use of the monoclonal antibody OKT3, a potent inducer of lymphocyte apoptosis that blocks the ability of cytotoxic T cells to recognize antigen (23, 24). By analogy, nontransplant patients with either primary hypogammaglobulinemia or infection with human immunodeficiency virus seem to develop an accelerated histological course related to HCV, with more rapid progression to cirrhosis and death (25-27). Collectively, these studies provide hints that effective T cell-mediated immunity may play a role in clearing or neutralizing HCV in both the nontransplant and transplant settings. Similarly, the absence of CMV viremia in allogeneic marrow recipients after transfer of donor-derived cytotoxic T lymphocyte clones lends credence to the hypothesis that a reconstituted cellular immune system is essential in the prevention of CMV disease (28). One could speculate that the immune response directed against CMV (presumably cytolytic T cells) is not specific and that CMV viremia, instead of reflecting a causal role in the development of severe HCV recurrence, is just a marker of another process (i.e., profound immunosuppression early after LT), which causes allograft cirrhosis. Alternatively, CMV infection may induce deficiencies in the immune response, blunting the normal effector mechanisms responsible for HCV clearance and indirectly potentiating HCV persistence and pathogenicity (29). Moreover, CMV infection triggers the production and release of tumor necrosis factor-α (30), a key mediator in the pathogenesis of HCV (31, 32). Additional pathways of immune-mediated allograft injury potentially occurring in CMV-HCV coinfected patients include cross-reactive immunological responses against normal cellular proteins or nonspecific host effector responses. Of interest in this regard, molecular homology and immunologic cross-reactivity have been described between an immediate early antigen of CMV and HLA-DRβ (33), and CMV-infected cells produce a glycoprotein homologous to MHC class I antigens (34, 35).
Host immunity to CMV infection is directed at numerous viral proteins and glycoproteins, and it is possible that strain variation affects immune recognition. Previous sequence analyses by our group of the envelope gB gene of CMV, involved in neutralization-related epitopes, has allowed classification into four variant groups (19). It has been shown in related herpes infections that gB is involved in cell entry, and therefore, variations in gB could affect viral tropism (36). We did not detect a difference in the distribution of gB genotypes between those patients who developed severe versus milder HCV recurrence or between patients with CMV disease as compared with asymptomatic infection, although the numbers were relatively small; a larger study is needed before any conclusions can be made. Similarly, the emergence and persistence of distinct genotypes and quasispecies have been implicated as important factors in the pathogenesis of HCV infection. However, two recent large studies of HCV-seropositive LT recipients failed to show an association between viral factors (genotypes and level of viremia) and postLT disease severity (37, 38). The results of the current series were in agreement with these findings; all the patients who developed allograft cirrhosis had HCV genotypes 1a or 1b as did 80% of patients without severe recurrence (P=NS). Recent preliminary data suggest that selection of quasispecies variants of HCV may contribute to the disease process after LT (12).
Several additional aspects of our study deserve special mention. Demonstration of CMV in blood (viremia) has been shown to be highly correlated with the presence of clinically active or impending infection, whereas isolation of the virus from urine or respiratory secretions is frequently a marker of asymptomatic infection and has not been shown to be predictive of CMV disease (1). In our present study, CMV infection was surveyed for prospectively using conventional culture methods. It is possible that the use of more sensitive laboratory tests for CMV antigen or DNA would have yielded an earlier and higher rate of viral detection. Most investigators agree, however, that buffy coat PCR has poor specificity and positive predictive value for CMV disease (39, 40). Studies comparing the results of plasma PCR and conventional culture have revealed a closer correlation (41). In a recent study of 41 LT recipients, conventional cell culture was the better technique for the diagnosis of symptomatic CMV infection (42). Considering the high frequency of sampling in our cohort (11.9 per patient), which increases the sensitivity, as well as the exclusion of patients with short follow-up, we believe that our study does not significantly underestimate the prevalence of CMV viremia, and this is corroborated by the range of CMV infection described in other studies (1, 6).
The significance of the variable duration of CMV viremia after LT remains undefined, and some investigators have hypothesized that spontaneous resolution of CMV viremia might constitute a risk factor for late CMV-related liver dysfunction, prompting a search for DNA in the liver allograft (43). Future investigations from our group will use double-label in situ hybridization methods to determine the cellular loci (hepatocytes, reticuloendothelial cells, ductal cells) and the intensity of CMV DNA and HCV RNA in serial biopsy specimens from dually infected patients. There are several possible, although not mutually exclusive, mechanisms by which CMV infection may lead to more severe HCV recurrence. An improved understanding of the molecular mechanisms of CMV-induced immunosuppression will permit optimal pharmacologic management of the immune response and the development of new strategies for antiviral therapy.
In conclusion, this study demonstrates that after LT for chronic HCV, patients who developed CMV viremia incurred a greater risk of allograft cirrhosis (OR >7) and GF (OR >9) from severe HCV recurrence. Numerous host and viral factors have been implicated to explain the variable course of HCV after LT. Given that the processes leading to severe HCV recurrence are complex and undoubtedly multifactorial, it is all the more compelling that one factor-CMV viremia-conveyed such a high risk. With time, the full impact of early CMV infection on HCV recurrence may become appreciated, perhaps challenging the old adage that CMV infection after LT does not adversely affect graft or patient survival (5). These findings may argue for the relative merits of preemptive CMV treatment of all HCV positive transplant recipients compared with treatment of only those patients receiving more intense immunosuppression or those with symptomless viral excretion. On the other hand, this approach would not eliminate the potential problem of CMV-HCV synergism because both infections are chronic, anti-CMV drugs cannot be given indefinitely, and drug resistance after LT has been reported (44).
Acknowledgments. The authors thank Dr. Jay A. Nelson, Ph.D., for helpful discussions and invaluable insight into viral pathogenesis.
This work was presented in part at the 47th Annual Meeting of the American Association for the Study of Liver Diseases, Chicago, IL, November 1996.
Abbreviations: CMV, cytomegalovirus; GF, graft failure; HCV, hepatitis C virus; LT, liver transplantation; OR, odds ratio; PCR, polymerase chain reaction.
1. Patel R, Snydman DR, Rubin RH, et al. Cytomegalovirus prophylaxis in solid organ transplant recipients. Transplantation 1996; 61: 1279.
2. Van den Berg AP, Van Son WJ, Haagsma EB, et al. Prediction of recurrent cytomegalovirus disease after treatment with ganciclovir in solid-organ transplant recipients. Transplantation 1993; 55: 847.
3. Leach CT, Detels R, Hennessey K, et al. A longitudinal study of cytomegalovirus infection in human immunodeficiency virus type 1-seropositive homosexual men: molecular epidemiology and association with disease progression. J Infect Dis 1994; 170: 293.
4. Webster A, Lee CA, Cook DG, et al. Cytomegalovirus infection and progression towards AIDS in haemophiliacs with human immunodeficiency virus infection. Lancet 1989; 2: 63.
5. Stratta RJ, Shaeffer MS, Markin RS, et al. Cytomegalovirus infection and disease after liver transplantation. Dig Dis Sci 1992; 37: 673.
6. Wiesner RH, Marin E, Porayko MK, Steers JL, Krom RAF, Paya CV. Advances in the diagnosis, treatment, and prevention of cytomegalovirus infections after liver transplantation. Gastroenterol Clin North Am 1993; 22: 351.
7. Rook AH. Interactions of cytomegalovirus with the human immune system. Rev Infect Dis 1988; 10 (suppl 3): S460.
8. Kowalik TF, Wing B, Haskill JS, Azizkhan JC, Baldwin AS, Huang ES. Multiple mechanisms are implicated in the regulation of NF-kappa B activity during human cytomegalovirus infection. Immunology 1993; 78: 405.
9. Rosen HR, O'Reilly PM, Shackleton CR, et al. Graft loss following liver transplantation in patients with chronic hepatitis C. Transplantation 1996; 62: 1773.
10. Gane EJ, Portmann BC, Naoumov NV, et al. Long-term outcome of hepatitis C infection after liver transplantation. N Engl J Med 1996; 334: 815.
11. Gane EJ, Naoumov NK, Qian KP, et al. A longitudinal analysis of hepatitis C replication following liver transplantation. Gastroenterology 1996; 110: 167.
12. Gretch DR, Polyak SJ, Wilson JJ, Carithers RL, Perkins JD, Corey L. Tracking hepatitis C virus quasispecies major and minor variants in symptomatic and asymptomatic liver transplant recipients. J Virol 1996; 70: 7622.
13. Sheiner PA, Schwartz ME, Mor E, et al. Severe or multiple rejection episodes are associated with early recurrence of hepatitis C after orthotopic liver transplantation. Hepatology 1995; 21: 30.
14. Gretch DR, Lee W, Corey L. Use of aminotransferase, hepatitis C antibody, and hepatitis C polymerase chain reaction RNA assays to establish the diagnosis of hepatitis C virus infection in a diagnostic virology laboratory. J Clin Microbiol 1992; 30: 2145.
15. Davidson F, Simmonds P, Ferguson JC, et al. Survey of major genotypes and subtypes of hepatitis C using RFLP of sequences amplified from the 5′ noncoding region. J Gen Virol 1995; 76: 1197.
16. Simmonds P, Holmes EC, Cha TA, et al. Classification of hepatitis C into six major genotypes and a series of subtypes by phylogenetic analysis of the NS-5 region. J Gen Virol 1993; 74: 2391.
17. Chou S, Dennison KM. Analysis of interstrain variation in cytomegalovirus glycoprotein B sequences encoding neutralization-related epitopes. J Infect Dis 1991; 163: 1229.
18. Chou S. Molecular epidemiology of envelope glycoprotein H of human cytomegalovirus. J Infect Dis 1992; 166: 604.
19. Fries BC, Chou S, Boeckh M, Torok-Storb B. Frequency distribution of cytomegalovirus envelope glycoprotein genotypes in bone marrow transplant recipients. J Infect Dis 1994; 169: 769.
20. Knodell RG, Ishak KG, Black WC, et al. Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis. Hepatology 1981; 5: 431.
21. DiBisceglie AM. Liver transplantation for hepatitis C: the promise and the challenge. Hepatology 1995; 22: 660.
22. Chazouillerees O, Kim M, Mamish D, et al. Quantitation of hepatitis C virus RNA in liver transplant recipients. Gastroenterology 1994; 106: 994.
23. Rosen HR, Shackleton CR, Higa L, et al. Use of OKT3 is associated with early and severe recurrence of hepatitis C after liver transplantation. Am J Gastroenterol 1997; 92(9): 1453.
24. Rosen HR, Martin P. OKT3 and hepatitis C: defining the risks. Transplantation 1997; 63: 171.
25. Bjoro K, Froland SS, Yun Z, Samdal HH, Haaland T. Hepatitis C infection in patients with primary hypogammaglobulinemia after treatment with contaminated immune globulin. N Engl J Med 1994; 331: 1607.
26. Martin P, DiBisceglie AD, Kassiandes C, Lisker-Melmav M, Hoofnagle J. Rapidly progressive non-A, non-B hepatitis in patients with human immunodeficiency virus infection. Gastroenterology 1989; 97: 1559.
27. DiMartino V, Rufat P, Martinot M, et al. Influence of HIV infection on the outcome of chronic hepatitis C. Hepatology 1996; 24: 382A.
28. Walter EA, Greenberg PD, Gilbert MJ, et al. Reconstitution of cellular immunity against cytomegalovirus in recipients of allogeneic bone marrow by transfer of T-cell clones from the donor. N Engl J Med 1995; 333: 1038.
29. Gehrz RC, Liu YNC, Eckhardt J, Klaus A. Relevance of immune responses to pathogenesis of cytomegalovirus-associated diseases. Transplant Proc 1991; 23: 75.
30. Nokta MA, Hassan MI, Loesch K, Pollard RB. Human cytomegalovirus-induced immunosuppression: relationship to tumor necrosis factor-dependent release of arachidonic acid and prostaglandin E2
in human monocytes. J Clin Invest 1996; 97: 2635.
31. Larrea E, Garcia N, Qian C, Civeira MP, Prieto J. Tumor necrosis factor-alpha gene expression and the response to interferon in chronic hepatitis C. Hepatology 1996; 23: 210.
32. Levy G. Hepatitis C: diagnosis and treatment. ILTS postgraduate course: Recurrent liver disease after liver transplantation. Chicago, November 8, 1996.
33. Funinami RS, Nelson JA, Walker L, Oldstone MB. Sequence homology and immunologic cross-reactivity of human cytomegalovirus with HLA-DR beta chain: a means for graft rejection and immunosuppression. J Virol 1988; 62: 100.
34. Rubin RH. The indirect effects of cytomegalovirus infection on the outcome of organ transplantation. JAMA 1989; 261: 3607.
35. Mosnier J-F, Scoazec J-Y, Marcellin P, et al. Expression of cytokine-dependent immune adhesion molecules by hepatocytes and bile duct cells in chronic hepatitis C. Gastroenterology 1994; 107: 1457.
36. Cai WH, Gu B, Person S. Role of glycoprotein B of herpes simplex virus type 1 in viral entry and cell fusion. J Virol 1988; 62: 2596.
37. Zhou S, Terrault NA, Ferrell L, et al. Severity of liver disease in liver transplantation recipients with hepatitis C virus infection: relationship to genotype and level of viremia. Hepatology 1996; 24: 1041.
38. Vargas H, Wang L, Laskus T, et al. Infecting HCV genotypes: a final report on HCV genotype influence on the outcome of liver transplantation for HCV-related end-stage liver disease. Hepatology 1996; 24: 182A.
39. Mutimer D, Matyi-Toth A, Elias E, et al. Quantitation of cytomegalovirus in blood of liver transplant recipients. Liver Transpl Surg 1995; 1: 395.
40. Delgado R, Lumbreras C, Alba C, et al. Low predictive value of polymerase chain reaction for diagnosis of CMV disease in liver transplant recipients. J Clin Microbiol 1992; 30: 1876.
41. Schmidt CA, Oettle H, Peng R, et al. Comparison of polymerase chain reaction from plasma and buffy coat with antigen detection and occurrence of immunoglobulin M for the demonstration of cytomegalovirus infection after liver transplantation. Transplantation 1995; 59: 1133.
42. Patel R, Smith TF, Espy M, et al. A prospective comparison of molecular diagnostic techniques for the early detection of cytomegalovirus in liver transplant recipients. J Infect Dis 1995; 171: 1010.
43. Drouet E, Colimon R, Michelson S, et al. Monitoring levels of human cytomegalovirus DNA in blood after liver transplantation. J Clin Microbiol 1995; 33: 389.
© Williams & Wilkins 1997. All Rights Reserved.
44. Rosen HR, Benner KG, Flora KD, et al. Development of ganciclovir resistance during treatment of primary cytomegalovirus infection following liver transplantation. Transplantation 1997; 63: 476.