Skip Navigation

Association of Human Herpesvirus 6 Reactivation with Severe Cytomegalovirus-Associated Disease in Orthotopic Liver Transplant Recipients

  1. Jeffrey A. DesJardin1,3,a,
  2. Eunhui Cho1,3,
  3. Stacey Supran1,3,
  4. Laurie Gibbons2,4,
  5. Barbara G. Werner2,4, and
  6. David R. Snydman1,3
  1. 1Departments of Medicine and Pathology, New England Medical Center, Boston
  2. 2State Laboratory Institute, Massachusetts Department of Public Health, Boston
  3. 3Tufts University School of Medicine, Boston
  4. 4University of Massachusetts School of Medicine, Boston
  1. Reprints or correspondence: Dr. David R. Snydman, New England Medical Center, 750 Washington St., Box 238, Boston, MA 02111 (dsnydman{at}lifespan.org).

  1. Presented in part: 21st International Congress of Chemotherapy, Birmingham, UK, 4–7 July 1999.

  • ↵a Present affiliation: Western Infectious Diseases Consultants, Wheat Ridge, Colorado.

 
Next Section

Abstract

To explore the possible interaction between human herpesvirus 6 (HHV-6) and cytomegalovirus (CMV) in patients who have undergone organ transplantation, stored serum samples from 139 orthotopic liver transplant recipients were tested for HHV-6 immunoglobulin (Ig) G and IgM antibodies. HHV-6 reactivation occurred in 87 patients (62.6%) and was associated with CMV disease (P = .01), severe CMV-associated disease (P = .01), older age (P = .005), and use of muromonab-CD3 (Orthoclone; Orthobiotech) as treatment for rejection (P = .02). Trends for an association between HHV-6 reactivation and invasive fungal disease (P = .12), bacteremia (P = .10), and graft loss (P = .12) were seen. In a multivariate analysis of risk factors for severe CMV-associated disease, HHV-6 reactivation (relative risk [RR], 3.5; 95% confidence interval [CI], 1.2–10.2; P = .02), CMV donor-positive-recipient-negative match (RR, 5.7; 95% CI, 2.5–13.2; P < .001), and elevated serum creatinine level (P < .0001) were independent predictors. HHV-6 reactivation is associated with severe CMV-associated disease in liver transplant recipients.

Human herpesvirus 6 (HHV-6) is a primarily T cell lymphotropic virus first isolated in 1986 [1]. More than 95% of adults have evidence of infection [2]. HHV-6 is known to be an immunomodulator and plays a role in inducing immunosuppression [3]. HHV-6 has been reported to be an opportunistic pathogen in bone marrow, kidney, and liver transplant recipients [4–6]. HHV-6 reactivation occurs frequently in solid-organ transplant recipients, and occasionally HHV-6 may be the only pathogen recovered from patients with encephalitis or interstitial pneumonitis [7]. The interaction between HHV-6 and symptomatic cytomegalovirus (CMV) infection after solid-organ transplantation has been noted to potentiate the likelihood of CMV disease and, perhaps, fungal and bacterial infection, as well as graft rejection in kidney and liver transplant recipients [5, 6, 8, 9]. To further assess the potential interaction between HHV-6 and CMV in solid-organ transplant recipients, we evaluated serum samples available from a previously conducted prospective trial that showed that prophylactic CMV Ig can reduce severe CMV-associated disease in liver transplant recipients [10].

Previous SectionNext Section

Methods

Patient population. We conducted a retrospective analysis of serial serum samples drawn from 139 liver transplant recipients involved in a prospective, randomized, double-blind, placebo-controlled trial that evaluated the effect of CMV Ig prophylaxis on prevention of CMV disease and its complications in patients undergoing liver transplantation [10]. In the original study, 146 patients were enrolled, but serum samples from only 139 enrollees were available for testing.

Patients underwent liver transplantation from December 1987 through June 1990 at 1 of the 4 participating centers for liver transplantation in Boston: Children's Hospital, Massachusetts General Hospital, New England Deaconess, and New England Medical Center. Clinical care was individualized by the transplantation team at each of the participating institutions. Patients were followed weekly for the first 8 weeks after transplantation, monthly for the next 6 months, and at 1 year after transplantation. Baseline laboratory studies were done at enrollment, weekly for 2 months (when special problems arose), and at the aforementioned scheduled follow-up visits. Clinical outcomes relating to CMV virological status, opportunistic disease, immunosuppressive regimens, and rejection were evaluated by use of case definitions reported elsewhere [10].

HHV-6 antibody determination. Serum samples obtained within 72 h after transplantation, every other week for 2 months, and at 8 months and 1 year after transplantation, had been stored at −20°C; these were evaluated for IgG HHV-6 antibodies using ELISA (Advanced Biotechnologies). Samples were initially screened at a 1:20 dilution. Two samples with the lowest and highest optical density from each patient from the preliminary screening were diluted serially from 1:50 to 1:12,800 and reevaluated for HHV-6 IgG antibody by use of the same ELISA. At the time of infusion of CMV Ig or placebo, a preinfusion serum sample was tested. Dilutions for each paired sample were tested on the same plate [11].

Serum samples obtained within 72 h after transplantation and at 6 and 12 weeks were screened for HHV-6 IgM antibody by means of a commercially available indirect immunofluorescence assay (Advanced Biotechnologies). Samples were diluted beyond the manufacturers' recommendation (1:20 to 1:80) to increase the IgM assay specificity [11]. At the 1:20 dilution, we had noted nonspecific reactivity in some control samples. Reviewers of the immunofluorescence assay were blinded to the CMV antibody result and the source of serum sample. A fluorescence intensity of ⩾2+ was scored as positive [11].

CMV evaluation. CMV antibody was measured in all patients' and liver donors' serum samples by use of complement fixation, latex agglutination, and ELISA [10]. Urine and throat-wash specimens (or throat swabs from young children) for isolation of virus and serum for analysis of CMV antibody were collected before the first infusion, weekly for 2 months, and then monthly for 6 months. Peripheral blood leukocytes were obtained for isolation of virus every other week for the first 2 months and then every month for 6 months, as well as at the time of clinical illness compatible with CMV disease and at the end of 1 year. Tissue specimens obtained by means of biopsy and at autopsy were cultured for virus. Human foreskin fibroblasts and MRC-5 cells were used at the different centers for routine CMV cultures and determination of viral cytopathic effects for up to 6 weeks. When it was clinically indicated, CMV culture with shell vial assay was used at several centers to rapidly diagnose CMV infection in various types of specimens [10].

Definitions. “HHV-6 reactivation” was defined as either a ⩾4-fold increase in IgG antibody or the presence of IgM antibody. “CMV disease” was defined as clinical evidence of organ dysfunction along with biopsy proof of CMV in the affected organ, as documented either by virus isolation or histological evidence of CMV in that organ. “Severe CMV-associated disease” was defined as biopsy-proven CMV disease in ⩾2 organs or the presence of CMV pneumonia or invasive fungal disease in presence of CMV infection [10].

“CMV syndrome” was defined as unexplained fever of ⩾3 days' duration in association with 1 of the following: pneumonitis without other cause, leukopenia (< 4.0 × 109 leukocytes/mm3), or thrombocytopenia (< 100 × 109 platelets/L) on ⩾3 consecutive days after withdrawal of azathioprine or ganciclovir, or atypical lymphocytosis (>5% of peripheral leukocytes). Hepatitis was not included in the definition, because liver dysfunction is common after liver transplantation.

Statistical analysis. The –2 and t tests were used for univariate analysis. Multivariate analysis was done using Cox proportional hazards models controlling for HHV-6 infection, CMV Ig prophylaxis, and other potential confounders. All statistical analyses were done by use of the SAS system for Windows, version 6.12 (SAS Institute) [12].

Previous SectionNext Section

Results

HHV-6 serology and infection. HHV-6 reactivation was observed in 87 (62.6%) of the 139 patients after liver transplantation. HHV-6 reactivation was detected by a ⩾4-fold increase in HHV-6 IgG antibody and presence of HHV-6 IgM antibody in 39 patients; ⩾4-fold increase in HHV-6 IgG antibody without detection of HHV-6 IgM antibody, in 35 patients; and presence of IgM antibody without a ⩾4-fold increase in IgG antibody, in 13 patients. The median time of onset of HHV-6 IgM antibody was 6 weeks (range, 0–12 weeks). All serum samples tested at the time of transplantation from both control and CMV Ig-treated transplant recipients were positive for HHV-6 IgG antibody (median number tested per patient, 8 serum samples; range, 2–10 serum samples).

Patient characteristics. The distribution of patient characteristics between those who did and did not develop HHV-6 reactivation is shown in table 1. There was no significant difference in race, sex, renal function, United Network of Organ Sharing (UNOS) status at time of transplantation, primary liver disease, blood product use, or steroid use. HHV-6 reactivation was more likely to occur among those who received muromonab-CD3 (OKT3) treatment for rejection, in older persons, and among patients who received prednisone, azathioprine, and cyclosporine as initial immunosuppression. Use of CMV Ig did not affect the likelihood of reactivation.

Table 1
Table 1

Distribution of characteristics in orthotopic liver transplant recipients with and without human herpesvirus 6 (HHV-6) reactivation.

HHV-6 reactivation and CMV infection. There was no relationship between CMV infection or detection of CMV viremia and HHV-6 reactivation (table 2). In addition, there was no relationship between CMV donor and recipient serological status and HHV-6 reactivation. Two possible trends were evident: the CMV-negative donor and recipient pair were less likely to have HHV-6 reactivation, and CMV viremia may have been more likely to have been detected in those with reactivation of HHV-6. However, there was a striking association between CMV disease (P = .01) and HHV-6 reactivation, as well as severe CMV-associated disease and HHV-6 reactivation (P = .01).

Table 2
Table 2

Relationship of human herpesvirus 6 (HHV-6) reactivation to cytomegalovirus (CMV) infection and CMV-associated disease in liver transplant recipients.

HHV-6 reactivation and other outcomes. Table 3 lists the relationship between HHV-6 reactivation and non-CMV-related outcomes. Trends for development of invasive fungal disease, bacteremia, and need for retransplantation were seen, but none were statistically significant.

Table 3
Table 3

Relationship of human herpesvirus 6 reactivation with outcomes other than cytomegalovirus (CMV) in a cohort of liver transplant recipients.

Multivariable models of HHV-6 and CMV outcomes. In a series of analyses to determine whether HHV-6 reactivation was an independent predictor for CMV disease (table 4), 3 factors were significantly associated: transplant from a CMV-seropositive donor to a CMV-seronegative recipient (RR, 4.2; 95% CI, 2.1–8.2; P < .001), ventilator support (UNOS status 4; RR, 2.9; 95% CI, 1.3–6.4; P = .008), and any OKT3 use (RR, 2.7; 95% CI, 1.0–7.1; P = .05). HHV-6 reactivation (RR, 1.9; 95% CI, 0.8–4.4; P = .16) showed a trend, but inclusion of HHV-6 reactivation in this model did not reach statistical significance. Examination of the subset of patients with CMV end-organ disease showed the same trend.

Table 4
Table 4

Multivariate analysis of human herpesvirus 6 (HHV-6) reactivation as risk factor for cytomegalovirus disease in liver transplant recipients.

In addition, we analyzed the effect of HHV-6 reactivation on development of severe CMV-associated disease. Both HHV-6 reactivation (RR, 3.5; 95% CI, 1.2–10.2; P = .02), transplantation from a CMV-seropositive donor to a CMV-seronegative recipient (RR, 5.7; 95% CI, 2.5–13.2; P < .001), and elevated creatinine level at transplantation (RR, 1.7 per mg/dL increase in creatinine; 95% CI, 1.3–2.1; P < .0001) were all significantly associated with severe CMV-associated disease.

In an analysis of the effect of CMV Ig prophylaxis on severe CMV-associated disease, we were able to demonstrate that CMV Ig was still protective (RR, 0.4; 95% CI, 0.2–0.9; P = .03) independent of HHV-6 reactivation and transplantation from CMV-seropositive donor to CMV-seronegative recipient.

Previous SectionNext Section

Discussion

We have examined the association of HHV-6 reactivation with CMV disease, severe CMV-associated disease, and other outcomes in a cohort of prospectively followed orthotopic liver transplant recipients. On univariate analysis, we were able to show a strong association between HHV-6 reactivation and CMV disease as well as severe CMV-associated disease. Although only a trend toward an association between HHV-6 reactivation and CMV disease was found in the multivariable model, we could demonstrate an independent relationship of HHV-6 reactivation and severe CMV-associated disease.

Although the data are based on serological analysis only for HHV-6 reactivation, the rates of reactivation for HHV-6 seen in this group of patients are virtually identical to rates defined by molecular diagnostic techniques [5, 8, 9], and they are actually higher than those seen with use of culture techniques [5]. The advantage of our analysis is the unbiased assessment of HHV-6 reactivation with disease associations, the prospectively collected data, the prospectively defined end points in the study cohort, and the absence of antiviral prophylaxis. Because of the use of serological testing, we could not examine timing of reactivation in relation to CMV events or perform time-dependent analyses. Although HHV-6 antibody may be present in CMV Ig, the multiple time points tested and testing of serum samples at 4 and 8 months after the last infusion allowed us to differentiate passively transfused antibody from active infection.

We could demonstrate only a trend for the association of HHV-6 reactivation with other outcomes, such as invasive fungal infection, bacteremia, or need for retransplantation. Other studies have demonstrated a stronger association [5, 6, 8, 9], and the reason for our discrepant findings could be related to the confounding effects of a high proportion of CMV-seropositive donor to CMV-seronegative recipient mismatches, the use of OKT3 in a large proportion of patients, or other factors, including prophylaxis with CMV Ig.

Of interest is the fact that we were able to independently demonstrate CMV Ig protection from severe CMV-associated disease, even in the presence of HHV-6 reactivation, because this was a primary end point in the original study. The independent effect of CMV Ig was consistent with that seen in renal transplant recipients [11].

It is clear that virus-virus interactions may play an important role in many outcomes associated with organ transplantation [6, 11, 13]. Our studies support the role of HHV-6 in CMV disease and severe CMV-associated disease in orthotopic liver transplant recipients and add to the growing evidence regarding the importance of virus-virus interactions in organ transplant recipients. Future prospective studies in a variety of transplant settings and with specific antiviral agents will be necessary to define the role of each.

Previous SectionNext Section

Acknowledgements

We thank Inga Svensson, for help in serological assays, and Roselia Martinez, for help in manuscript preparation.

Previous SectionNext Section

Footnotes

  • Financial support: National Institutes of Health (AI-07329) and an unrestricted educational grant from MedImmune, Inc.

  • Received December 22, 2000.
  • Revision received May 2, 2001.
Previous Section
 

References

  1. ↵
    1. Salahuddin SZ,
    2. Ablashi DV,
    3. Markham PD,
    4. Josephs SF,
    5. Sturzenegger S,
    6. Kaplan M
    . Isolation of a new virus, HBLV, in patients with lymphoproliferative disorders. Science 1986;234:596-601.
    Abstract/FREE Full Text
  2. ↵
    1. Lopez C,
    2. Pellett P,
    3. Stewart J,
    4. et al
    . Characteristics of human herpesvirus 6. J Infect Dis 1988;157:1271-3.
    FREE Full Text
  3. ↵
    1. Flammand L,
    2. Gossehn J,
    3. Stefanescu I,
    4. Ablashi D,
    5. Menezes J
    . Immunosuppressive effect of human herpes virus 6 on T cell function: suppression of interleukin-2 synthesis and cell proliferation. Blood 1995;85:1263-71.
    Abstract/FREE Full Text
  4. ↵
    1. Kadakia MP,
    2. Rybka WB,
    3. Stewart JA,
    4. et al
    . Human herpesvirus 6: infection and disease following autologous and allogeneic bone marrow transplantation. Blood 1996;87:5341-54.
    Abstract/FREE Full Text
  5. ↵
    1. Rogers J,
    2. Rohal S,
    3. Carrigan DR,
    4. et al
    . Human herpesvirus-6 in liver transplant recipients. Transplantation 2000;69:2566-73.
    CrossRefMedlineWeb of Science
  6. ↵
    1. Kidd IM,
    2. Clark DA,
    3. Sabin CA,
    4. et al
    . Prospective study of human betaherpesviruses after renal transplantation: association of human herpresvirus 7 and cytomegalovirus co-infection with cytomegalovirus disease and increased rejection. Transplantation 2000;69:2400-4.
    CrossRefMedlineWeb of Science
  7. ↵
    1. Singh N,
    2. Paterson DL
    . Encephalitis caused by human herpesvirus-6 in transplant recipients. Transplantation 2000;69:2474-9.
    CrossRefMedlineWeb of Science
  8. ↵
    1. Lautenschlager I,
    2. Linnavuori K,
    3. Hockerstedt K
    . Human herspesvirus-6 antigenemia after liver transplantation. Transplantation 2000;69:2561-6.
    CrossRefMedlineWeb of Science
  9. ↵
    1. Dockrell DH,
    2. Mendez JC,
    3. Jones M,
    4. et al
    . Human herpesvirus 6 seronegativity before transplantation predicts the occurrence of fungal infection in liver transplant recipients. Transplantation 1999;67:399-403.
    CrossRefMedlineWeb of Science
  10. ↵
    1. Snydman DR,
    2. Werner BG,
    3. Dougherty NN,
    4. et al
    . Cytomegalovirus immune globulin prophylaxis in liver transplantation: a randomized, double-blind, placebo controlled trial. Ann Intern Med 1993;119:984-91.
    Abstract/FREE Full Text
  11. ↵
    1. DesJardin JA,
    2. Gibbons L,
    3. Cho E,
    4. et al
    . Human herpesvirus 6 reactivation is associated with cytomegalovirus infection and syndromes in kidney transplant recipients at risk for primary cytomegalovirus infection. J Infect Dis 1998;178:1783-6.
    Abstract/FREE Full Text
  12. ↵
    SAS/STAT user's guide. Version 6. 4th ed. Cary, North Carolina: SAS Institute; 1990.
  13. ↵
    1. Singh N,
    2. Carrigan DR
    . Human herpesvirus-6 in transplantation: an emerging pathogen. Ann Intern Med 1996;124:1065-71.
    Abstract/FREE Full Text
| Table of Contents

This Article

  1. Clin Infect Dis. (2001) 33 (8): 1358-1362. doi: 10.1086/323336
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 54 (5)
  1. Clinical Infectious Diseases
  1. Alert me to new issues

Most