Herpes zoster (HZ) infection is a frequent and serious complication of organ transplantation that has not been examined in the current era of immunosuppression.
All solid organ transplants performed between 1994 and 1999 (n = 869) at our center were analyzed to determine the incidence, complications and risk factors for developing HZ.
The overall incidence of HZ was 8.6% (liver 5.7%, renal 7.4%, lung 15.1% and heart 16.8%). The median time of onset was 9.0 months. We observed high rates of cutaneous scarring (18.7%) and post-herpetic neuralgia (42.7%). Independent organ-specific risk factors included: female gender and mycophenolate mofetil therapy (liver), and antiviral treatment other than prolonged cytomegalovirus (CMV) prophylaxis (renal and heart). For all organs combined, induction therapy and antiviral treatment other than prolonged CMV prophylaxis were independent predictors for the development of HZ.
Herpes zoster is common and results in significant morbidity for solid organ transplant recipients. Risk factors include induction therapy and antiviral drug therapy other than CMV prophylaxis. The latter variable identifies a subpopulation that is likely at increased risk of latent herpesvirus reactivation. The high first-year post-transplant incidence rate suggests immunization pretransplant, even in varicella zoster virus immunoglobulin seropositive individuals, may be preventative.
Varicella zoster virus (VZV) is the most infectious of the human herpesviruses and infection is almost universal (95%) by adulthood in North American and European populations. Reactivation of latent VZV [herpes zoster (HZ) or shingles] is a painful, cutaneous eruption, dermatomal in distribution, associated with a risk of dissemination. Previously described risk factors for HZ include age, malignancy, HIV infection, organ transplantation, immunodeficiency and treatment with immunosuppressive medications (1).
Herpes zoster leads to significant morbidity and its most frequent complication (2) is post-herpetic neuralgia (PHN), which is associated with persistent pain for 30 or more days after acute infection, lasting years in some cases (3). Unfortunately, immunocompromized patients tend to have the most severe complications of reactivation with a greater tendency for a prolonged course of disease (4). Severely immunocompromized patients with HZ have a risk of dissemination of up to 40%, which can result in mortality rates, despite antiviral therapy, of between 4 and 34% (5–7).
Previously reported incidences vary widely (8–17) and, to our knowledge, no population-based studies have been performed that define the incidence of HZ in adult multiorgan transplant cohorts. However the following incidences have been suggested in this population: renal 3–10%, liver 5–10%, lung 8–12% and hearts 20–25% (7). Possible reasons for this variation include center effect, the immunosuppressive drug (ISD) regimen chosen, antiviral prophylaxis protocols and variable methodologies used for follow up.
Materials and Methods
We performed a retrospective analysis of all solid organ transplants (renal, liver, heart and lung) performed between 1994 and 1999 at the University of Alberta Hospital, Edmonton, Canada. We studied the incidence, complications (PHN and cutaneous scarring) and risk factors for HZ development in the current era of immunosuppressive therapy and routine antiviral prophylaxis. The medical chart and phone interviews were also used to identify any other complications of HZ infection (hospitalization, hepatitis, pneumonia, etc.).
Herpes zoster was defined as the development of a cutaneous vesicular eruption, dermatomal in distribution. Post-herpetic neuralgia was defined as pain persisting in the affected area for more than 30 days after the development of the rash. Cutaneous scarring was defined as skin disfigurement (scars, hypopigmentation) in the distribution of the rash. Diagnosis of HZ was made clinically with or without confirmatory laboratory diagnosis. Patients were followed up for development of HZ, loss of their graft and/or death, or until the study end-date (October 1, 2001).
Ethics approval for this study was obtained. All medical charts were reviewed by one of two investigators. Data was collected using a standardized form for information on recipient demographics, cause of organ failure, date of onset of HZ, immunosuppressive regimens, induction and rejection therapy, antiviral drug exposure, complications incurred and treatment received. Follow-up phone calls were made by investigators to all living transplant recipients to confirm the data collected, in particular to confirm a history of HZ infection and/or complications. A patient was considered a case if chart documentation existed and/or the patient recalled having had an episode of HZ that fulfilled clinical criteria as judged by the investigator. No data on virologic confirmation was available. Only one renal transplant recipient recalled an episode of HZ that was not documented in the chart.
Immunosuppressive regimens in our center
The standard immunosuppressive regimen for renal transplant recipients between 1994 and late 1995 consisted of azathioprine (AZA; 1–2 mg/kg/day) and neoral cyclosporine (CsA). Target 12-h trough levels for CsA were 350 μg/L for the first month and then tapered to a maintenance level of 100–125 μg/L after the first year. Average oral steroid dose at time of transplant is 0.5 mg/kg/day, which is tapered toward an approximate maintenance dose of 0.05 mg/kg/day. In 1995, with the availability of mycophenolate mofetil (MMF), our program converted to use of MMF instead of AZA, at a dose of 1 g twice daily. In 1998, our center converted from CsA to tacrolimus therapy (target trough levels of 12–15 μg/L for the first 3 months post-transplant and then maintenance levels of 5–7 μg/L) as part of the standard regimen. Induction therapy consisting of polyclonal antibodies was given for delayed graft function or to highly sensitized (high-panel reactive antibodies) patients.
All heart transplant recipients received induction with antithymocyte globulin followed by maintenance immunosuppression with AZA (2 mg/kg); cyclosporine and tacrolimus doses and target trough levels were similar to those for renal transplant recipients. Prednisone was initiated at 0.5 mg/kg post-transplant and was then tapered off slowly until completely discontinued by 1 year. In 1999, MMF replaced the use of AZA and was dosed at 1–1.5 g twice daily.
Lung transplant recipients all received induction therapy except for those with hepatitis C infection. Basic immunosuppression was similar to heart transplant recipients except that prednisone was maintained at 0.05 mg/kg/day at 1 year.
Liver transplant recipients had similar cyclosporine and tacrolimus doses and trough target levels as renal transplant recipients. Azathioprine doses consisted of 2 mg/kg initially, which was reduced to 1.5 mg/kg/day on day 15 and then to 1.0 mg/kg by day 30. Each patient received 500 mg of intravenous solumedrol preoperatively, followed by prednisone on day 1 dosed at 100 mg twice daily, which was tapered to 10 mg daily by 6 weeks and then subsequently slowly tapered off over the next several months.
Antiviral drug exposure
Standard guidelines for the prophylaxis and treatment of viral infections after transplantation were in place at our center during the study period. Prior to April 1998, cytomegalovirus (CMV) management in nonlung transplant recipients who were CMV recipient positive or CMV mismatched primarily involved a pre-emptive strategy using prospective laboratory buffy coat surveillance (pp65 antigenemia assay) and intravenous ganciclovir therapy (5 mg/kg q12 h for 2 weeks) when trigger points for CMV disease were observed. For a period of time between 1994 and 1998 some programs used a high-dose oral acyclovir prophylaxis protocol (800 mg qid for 12 weeks) as outlined by Balfour et al. (28).
As of April 1998, oral ganciclovir prophylaxis was used for CMV management in all nonlung transplant recipients. All CMV and Epstein–Barr virus mismatched (D+/R) recipients received oral ganciclovir post-transplant for 14 weeks post-transplant (1000 mg tid). All other donor-recipient subgroups continued to be managed with pre-emptive therapy. Patients were also routinely given IV ganciclovir prophylaxis (2.5 mg/kg/day) for CMV if they received monoclonal or polyclonal antibody therapy for the duration of treatment of rejection episodes. Between 1994 and 1999, all lung transplant recipients other than D–/R– patients received IV ganciclovir prophylaxis for at least 2 months post-transplant.
Subjects may also have been exposed to antiviral drugs (acyclovir or ganciclovir) for the treatment of CMV or herpes simplex virus (HSV) infection, which occurs following transplantation. Cytomegalovirus disease was treated for a minimum of 2 weeks with IV ganciclovir (5 mg/kg bid). Recommended therapeutic guidelines for HSV disease consisted of IV acyclovir (5 mg/kg q8 h) for severe disease and oral acyclovir (200 mg five times a day or 400 mg tid), valacyclovir (I mg bid) or famciclovir (250 mg tid). The recommended duration of treatment was 7–10 days. Routine HSV prophylaxis was used in the first post-transplant month in heart transplant recipients who received induction therapy (oral acyclovir 200 mg tid or 400 mg bid). More prolonged HSV prophylaxis was also used in patients experiencing frequent recurrent oral or genital HSV infections (oral acyclovir 400 mg bid, valacyclovir 250 mg bid or famciclovir 250 mg bid).
Antiviral drug exposure was subdivided into two categories for analyses. Patients were classified as having received prolonged CMV prophylaxis if they received ≥6 weeks of high-dose oral acyclovir or IV and/or oral ganciclovir. Patients were classified as having other antiviral drug exposure if they received treatment for herpesviruses other than VZV (CMV or HSV) or received HSV prophylaxis (Table 1). Because antiviral therapy was often prescribed in an outpatient setting by a variety of practitioners, specific start and stop dates for antiviral regimens were not always documented.
Table 1. Demographics of the study population including virology, immunosuppression regimen and treatment of rejection episodes
Renal n = 434 (%)
Liver n = 263 (%)
Heart n = 119 (%)
Lung n = 53 (%)
All n = 869 (%)
1Forty-eight renal transplant patients' varicella zoster virus (VZV) status is missing and percentages are reported based on known data.
2Eighty-five liver transplant patients and 14 renal transplant patients were neither on azathioprine (AZA) nor on mycophenolate mofetil (MMF) and percentages are reported based on known data.
VZV IgG serostatus pretransplant1
CMV Mismatch (donor and recipient)
Low-dose antiviral therapy
Antiviral exposure all
Antiviral exposure other than prolonged prophylaxis
Viral demographics, immunosuppressive data and induction/rejection profiles for the study population are summarized in Table 1.
The primary endpoint was the incidence of HZ overall and by organ type. Statistical package for the social sciences (SPSS) version 11.0 for Windows was used to conduct the statistical analysis. Risk factors for developing HZ following transplant were evaluated using chi-square analysis, Kaplan-Meier method, and Cox regression analysis. Potential predictors of HZ examined included: age (age < or ≥50), gender, year of transplant (which reflects not only change in immunosuppressive use but also change in overall transplant management practice, surgical techniques, etc.), cause of renal failure, CMV mismatch (MM) status, immunosuppressive drug regimen (ISD), type of calcineurin inhibitor, rejection episode, type of rejection therapy, induction therapy, type of renal transplant (cadaveric vs. living), prolonged antiviral prophylaxis for CMV, antiviral drug exposure other than prolonged CMV prophylaxis, VZV immunoglobulin (IgG) serostatus, and pretransplant episode of HZ. Variables were screened using Kaplan-Meier survival curves and Cox regression, with the variables of interest as a single main effect. Multivariate models were built in a stepwise hierarchical manner, testing the significance of added terms using the likelihood ratio test, for each type of organ transplant and all transplants combined. Interaction between age and potential risk factors was also tested. Results were deemed significant if a 2-tailed p-value was <0.05.
A total of 869 subjects were included in the cohort (Table 1) and had a median follow-up time of 1159 days [3.2 years; range 9 days to 2816 (7.7 years)].
On average, our center performed 100 to 160 transplants per year during this time period. Subjects under the age of 16 years were excluded from this study, as the VZV IgG serostatus and immunosuppression regimens were different for this population. Subjects with missing medical records (15) or who did not survive the first 2 weeks after transplant (46) were also excluded from the study.
Incidence and timing of VZV reactivation
The proportion of organ transplant recipients overall who developed HZ following transplant was 8.6% (75/869)(Table 2). The median age was 49.1 (range 16.0–74.0) years, with 44 men and 31 women developing HZ. The median time to onset of HZ was 9.0 (0.6–69.3) months and the mean time to onset was 13.9 (9 days – 5.8 years). The majority of subjects (62.7%) developed zoster during the first post-transplant year (Figure 1), and receiving prolonged CMV prophylaxis did not decrease the risk of developing HZ. We did not observe any cases of disseminated zoster. As patterns of localization were not routinely documented in the chart and subject recall was unreliable, no interpretation related to patterns of distribution could be made. We found an overall prevalence of VZV IgG of 98.4% (Table 1). Varicella zoster virus IgG status was missing for 48 renal transplant recipients of which two developed HZ post-transplant. The calculated incidence density based on a total follow-up time of 1 007 242 days (2757.7 years) was 2720 per 100 000 person-years (18). Interestingly, although no HZ episodes post-transplant were observed in patients who had HZ before transplant, the proportion of patients that developed HZ after transplantation, that had HZ before transplantation, is not significantly different from who did not have HZ before transplantation (0/25 vs. 75/844; p = 0.16). However, this may be related to insufficient power and a low event rate.
Table 2. Cases of herpes zoster treatment and complications
Patients with post-transplant HZ by organ type
n = 32
n = 15
n = 20
n = 8
n = 75
Age ≥50 years
Mean ± SD
43.8 ± 15.8
54.2 ± 9.2
53.6 ± 6.5
49.2 ± 5.7
49.1 ± 12.5
Time to onset of HZ (months)
Mean ± SD
15.5 ± 17.8
12.4 ± 11.8
11.1 ± 7.9
17.1 ± 13.4
13.9 ± 14.0
Treatment with antiviral agents
Cases of post-herpetic neuralgia
Cases of cutaneous scarring
Morbidity and mortality associated with HZ
Significant evidence of morbidity was found in the form of cutaneous scarring (18.7%) and PHN (42.7%). The duration of depigmentation and duration of pain were confirmed by chart review. There was one severe complication in a 47-year-old female renal transplant recipient who developed spinal cord involvement confirmed by magnetic resonance imaging. She suffered significant neurological and psychological morbidity, in the form of gross motor impairment, requiring a walking aid. She was the only patient hospitalized for HZ. No mortality directly attributable to HZ was recorded in the 75 subjects.
Treatment of HZ
Treatment of HZ is summarized in Table 2. All patients were either tapered off steroids by 1 year post-transplant or left on low-dose maintenance steroid therapy. No further steroid therapy was given upon diagnosis of HZ. Information regarding other analgesia prescribed was not available through a retrospective study because there was great variability in subject reporting, who prescribed the analgesics (family physicians or specialists) and which medication was actually used for pain control. No data was available on timing of administration of antiviral drug therapy in relationship to the onset of symptoms.
Effect of prolonged CMV prophylaxis and other antiviral drug exposure
During the period of study, 79 of the 157 CMV mismatched (D+R–) patients (Table 1) received prolonged CMV prophylaxis for 14 weeks. Of the 75 patients who developed HZ, 12 were D+R– but only two received prolonged CMV prophylaxis; 11/12 were transplanted before 1998 when administration of oral ganciclovir prophylaxis to CMV- mismatched patients became standard practice at our institution. Of the patients, 29.5% (n = 255) were exposed to antiviral agents for reasons other than prolonged CMV prophylaxis, which included: 89 (34.9%) HSV prophylaxis, 78 (30.6%) gcv treatment, 51 (20%) acv treatment, 24 (9.4%) gcv and acv treatment, and 13 (5.1%) received HSV prophylaxis as well gcv or acv treatment.
Predictors of developing HZ following renal transplant
Factors associated with an increased risk of developing post-transplant HZ are shown in Table 3. Univariate analysis identified three factors to be associated with developing HZ in solid organ transplant recipients: age ≥50 years, induction therapy and antiviral drug exposure other than prolonged CMV prophylaxis during the post-transplant course (Table 3). Female gender and receiving MMF therapy were risk factors for developing HZ in liver transplant recipients. Despite a trend of increased antiviral exposure in heart transplant recipients, these individuals were at reduced risk of developing HZ if they experienced any rejection episode. Detailed analyses of this result did not provide additional information and this finding remains unexplained. Overall, in multivariate analysis, both induction therapy and antiviral drug exposure other than prolonged CMV prophylaxis conveyed a twofold increased risk for developing HZ for all solid organ transplant recipients (Table 3).
Table 3. Hazard ratio and 95% confidence interval1 of herpes zoster by each organ and all organs together
Renal (n = 434)
Liver (n = 263)
Heart (n = 119)
Lung (n = 53)
All (n = 869)
1HR and 95% CI are only reported if p < 0.1
2Antiviral exposure other than prolonged prophylaxis consisted of treatment with gcv and/or treatment with acv and/or use of low-dose antiviral prophylaxis.
HR = hazard ratio
Age ≥50 years
MMF vs. AZA
Tacrolimus vs. CsA
Treatment with gcv
Low dose antiviral therapy
Antiviral exposure other
than prolonged prophylaxis2
MMF vs. AZA
Antiviral exposure other
than prolonged prophylaxis2
When comparing the HR for developing H2 between renal transplant recipients and heart and lung transplant recipients, renal transplant recipients have a two-fold increased risk (Figure 2). Liver transplant recipients appear to be at less risk than renal recipients, although this result did not reach statistical significance.
Our study's incidence rate of HZ of 8.6% in a multiorgan transplant population is higher than previously reported (7). Thus HZ is not an uncommon complication of solid organ transplantation. In fact, we may be underestimating the risk because of our dependence on subject recall and the retrospective nature of the study. Misclassification bias of the diagnosis of HZ and the sequelae of HZ is another potential limitation of this study, as is the potential confounding effect of treatment or exposure to other antiviral agents.
Previous reports describe the onset of HZ following solid organ transplantation to be between 100 days and 8 months (7). We observe a delayed presentation with only 62.7% of the subjects developing zoster by 12 months after transplant. Our mean time to onset (13.9 months) is later than previously reported and may suggest that factors related to organ transplantation in the present era delay the onset of VZV reactivation; however, there was a vast range of time of onset (9 days – 5.8 years). Our median time of onset of 9 months (range 0.6–69.3 months) is still slightly prolonged but is generally more consistent with prior reports. In bone marrow transplantation, an even earlier onset of between 2 and 10 months (19–21) has been reported; however, a recent study suggests that the delayed mean time of onset of 16 months is the result of prolonged exposure to antiviral therapy (17). Routine antiviral prophylaxis for prevention of CMV disease has been shown to reduce the mortality from zoster (22) but does not appear to affect the incidence of HZ (16). Of the patients that developed HZ in our study, almost 50% of subjects were exposed to some antiviral therapy either before or after developing HZ and 20% received prolonged CMV prophylaxis (Table 3). Suppression of VZV by antiviral drug use may have delayed both the reactivation of VZV and the reconstitution of the immune response to VZV.
We also found that 1.6% of recipients were VZV-seronegative before transplant. This demonstrates that even in the adult transplant population, pretransplant screening and immunization of all susceptible individuals should be considered to avoid the development of primary VZV infection.
Herpes zoster can be a serious disease and heightened awareness of this condition is necessary to facilitate early diagnosis and treatment. We report a high rate of cutaneous scarring (18.7%) and PHN (42.7%) in our population. This is significantly higher than previously reported in the transplant population (7,23) and we observed one serious complication of spinal cord involvement with significant disability. The rate of PHN is of particular concern given that 96.1% patients received antiviral treatment for HZ and were on maintenance steroid therapy as part of the standard immunosuppressive regimen post-transplant, although likely at low doses. Although there was a lack of consistent information relating time of onset of treatment to complication rates, consideration of early diagnosis and treatment of HZ in this population can prevent or minimize the incidence of PHN. There was no mortality in our study attributable to HZ. However there were 77 subjects for whom information was not collected with respect to the cause of death and history of HZ.
Age and induction therapy have been shown to be risk factors for the development of HZ for all transplants and we found this in an unadjusted analysis. However, in the multivariate analysis, in the presence of induction therapy, age did not factor out as an independent predictor of developing HZ and this was not owing to confounding or effect modification. This observation contrasts with the previous reports in healthy and elderly populations, but confirms the results of BMT studies (17). The mean age of our population was 46.8 years with 43.8% of our study subjects being aged older than 50 years. In the absence of a disproportionate number of subjects aged older than 50 years, an underestimate of the effect of age may occur. With less rigid criteria for recipient age evolving in transplantation, the age effect may become a more prominent observation in the future. However, the degree of immunosuppression (e.g. induction therapy) associated with newer and more potent immunosuppressive agents may simply outweigh the impact of age on the incidence of HZ following organ transplantation.
Previous reports have shown that exposure to antiviral prophylaxis for CMV protects against VZV infection in BMT recipients (17). Our study did not demonstrate this association in multivariate analysis, however, few patients in our study received ‘true prolonged CMV prophylaxis’ (18%). Most of the antiviral drug exposures, other than prolonged CMV prophylaxis, consisted of receipt of low-dose acyclovir for the prevention or treatment of HSV and/or intravenous ganciclovir treatment for CMV disease. This latter variable was associated with an increased risk for the development of HZ and more likely identifies a subset of patients with increased global susceptibility to the reactivation of latent herpesviruses infections.
A direct association exists between the intensity of immunologic suppression and the frequency and severity of reactivated HZ (24). Induction therapy clearly posed a risk for developing HZ for all organ types, with heart and lung transplant recipients having the greatest risk, which correlates with routine induction therapy. We did not show specific maintenance immunosuppressive agents to be independent risk factors in multivariate analysis; however, in unadjusted analysis, liver transplant recipients receiving MMF demonstrated an association with a higher risk of developing HZ. Mycophenolate mofetil is considered a more immunopotent agent than AZA; however, there is no evidence in the literature at present from randomized control trials that MMF specifically places organ transplant recipients at a greater risk of developing diseases as a result of other herpesviruses, such as CMV or HSV, or post-transplant lymphoproliferative disorder.
Given that the ultimate goal would be to completely prevent the occurrence of HZ in this population, currently there is no proven prevention method and therapies of acute HZ and PHN are only partially effective. An interesting although nonsignificant finding of our study is that subjects that developed zoster pretransplant experienced no recurrence following transplantation. A natural hypothesis would be that a pretransplant immunity boost may protect the recipient from infection post-transplant. The development of the varicella vaccine has resulted in controversy with respect to its effect on the natural history of HZ and its role in boosting immunity of adults at risk (25). Varicella zoster virus-cell-mediated immune responses increase in healthy elderly subjects immunized with vaccine and significant benefit is observed when elderly subjects already harboring the latent VZV receive what amounts to a booster vaccination (26). Seropositive BMT recipients immunized with the varicella vaccine experienced greater protection against developing HZ and this correlated with reconstitution of CD4 T-cell immunity against VZV (27). Fehr et al. support vaccination in potential recipients that are VZV IgG-negative or who have very low VZV antibody titers (6), and the high incidence of disease in the early post-transplant period makes solid organ transplant recipients an ideal population to evaluate prospectively the utility of pretransplant administration of the varicella vaccine as a means of reducing the incidence and morbidity resulting from HZ.