Herpes Zoster Infection After Liver Transplantation: ACase-Control Study

Josh Levitsky,1Andre Kalil,2 Jane L. Meza,3 Glenn E. Hurst,4 and Alison Freifeld2

Prior case series have suggested that herpes zoster (HZ) afterorthotopic liver transplantation (OLT) may lead to seriouscomplications due to visceral involvement. We sought todetermine the incidence, risk factors, and long term out-comes of HZ after OLT. Clinical data from September 1993to April 2004 were collected on all cases of HZ after OLT,and at the same post-OLT time points in age, gender, andtransplant-year–matched HZ-negative controls. Risk factorsfor HZ infection and long-term outcomes were comparedbetween cases and controls. A total of 29 patients developedHZ at a median of 4.9 years (range .5-12.9) after OLT. AllHZ infections except 1 were localized to a single dermatome.Only 8 (28%) were hospitalized and 16 (55%) were treatedwith oral antivirals alone. No patients developed visceralinvolvement or died of HZ infection. No risk factors for HZinfection were identified on multivariate analysis. Of thelong-term outcomes, the estimated 10-year survival waslower (P � .05) for cases than controls. The lower survival inHZ cases was not directly attributable to HZ infection. Inconclusion, this study is the largest series on HZ after OLT.HZ is neither a common nor a serious infection after OLTand can be managed with antiviral therapy with a low likeli-hood of visceral dissemination. (Liver Transpl 2005;11:320–325.)

Varicella-zoster virus (VZV) is a contagious herpesvirus that leads to over 4 million infections, 11,000

hospitalizations, and 100 deaths in the United States everyyear,1 although the incidence is decreasing due to vaccina-tion. A total of 95% of primary varicella infection (chick-

enpox) occurs before adulthood, while reactivation ofVZV, or herpes zoster (HZ), usually occurs later in life andis characterized by a unilateral dermatomal eruption.Reactivations of VZV leading to HZ are more frequent inindividuals with impaired cellular immunity, such as theelderly, human immunodeficiency virus–positivepatients, and other immunosuppressed patients.2–4 Severesystemic complications, such as pneumonia, encephalitis,cerebellar ataxia, and hepatitis may be more likely in thesepopulations as well.

Accordingly, pediatric and adult orthotopic livertransplant (OLT) recipients may be at risk for the devel-opment of HZ and its associated complications. Muchof the published data on HZ infection after OLT aresmall case reports and series.5–9 One study from 1988reported a 7% incidence of HZ infection in 121 adultLT recipients.10 A recent study reported that HZ aftersolid organ transplantation was not uncommon (overallincidence 8.6%, liver 5.7%) in the 1st year after trans-plantation and resulted in significant morbidity.11

However, the statistical analysis performed combinedliver with other solid organ transplants (heart, lung, andrenal) and was not case-controlled. The latest studyreported a larger incidence (12%) of HZ infection inover 200 patients, potentially related to immunosup-pressive therapy.12

Given the need for further data on HZ infectionspecific to OLT, we have performed a case-controlledanalysis investigating the incidence, risk factors, andoutcomes of HZ infection in this population.

Patients and Methods

The University of Nebraska Medical Center electronic organtransplant database and hospital medical records weresearched for all OLT recipients (adult and pediatric) whodeveloped HZ infection from September 1, 1993, to April 1,2004. All patients were followed at our center in this timeperiod. Our electronic database contains complete records onclinical events occurring in these patients both at our medicalcenter and other institutions. Matched controls without HZinfection were selected from an age / gender-ordered masterlist of all liver transplant patients. The 1st control who metour matching criteria (age within 3 years, gender, and trans-plant-year within 3 years) was selected for analysis. All con-trols were analyzed at the same time points after OLT as thecorresponding cases. To ensure that the controls had no priorHZ infection after OLT, the electronic database was searched

Abbreviations: HZ, herpes zoster; OLT, orthotopic liver trans-plantation; VZV, varicella-zoster virus.

From the 1Department of Internal Medicine / Gastroenterology,2Department of Internal Medicine / Infectious Disease, 3Preventive andSocietal Medicine, Biostatistics Section, and 4School of Medicine, Uni-versity of Nebraska Medical Center, University of Nebraska, Omaha,Nebraska.

Supported by the 2004 American Association for the Study of LiverDiseases (AASLD) / Schering Advanced Hepatology Fellowship Program(July 1, 2004 to July 1, 2005) (to J.L.).

The abstract for this article was presented at the American Associationfor the Study of Liver Diseases Annual Meeting, Boston, MA, November2004.

Address reprint requests to Alison Freifeld, MD, Associate Professor ofMedicine, Department of Internal Medicine, Division of Infectious Dis-ease, The University of Nebraska Medical Center, Omaha, NE 68198-5400. Telephone: 402-559-8650; FAX: 402-559-5581; E-mail:afreifeld@unmc.edu

Copyright © 2005 by the American Association for the Study ofLiver Diseases

Published online in Wiley InterScience (www.interscience.wiley.com).DOI 10.1002/lt.20356

320 Liver Transplantation, Vol 11, No 3 (March), 2005: pp 320–325

for the identical terms as above and the hospital medicalrecords were reviewed for any indication of HZ infection.Controls were excluded if there was evidence of a rash sugges-tive of HZ at any time post-OLT.

Our immunosuppressive regimen included calcineurin-inhibitors (cyclosporine or tacrolimus) and corticosteroids.Reduction or elimination of corticosteroids resulting inmonotherapy with calcineurin-inhibitors was attempted. Inthe middle-late 1990s, a gradual increase in the use of tacroli-mus over cyclosporine occurred at our center. Additionalimmunosuppressive agents, such as mycophenolate and aza-thioprine, were used sparingly and only in select cases duringthe study time period. Immediate posttransplant antiviralprophylaxis varied significantly throughout the study timeperiod, but generally consisted of oral acyclovir in cytomega-lovirus-positive recipients and intravenous ganciclovir incytomegalovirus-negative recipients. VZV serologies were notperformed routinely in our patients prior to transplantation.Rejection episodes were treated with intravenous corticoste-roids and, if refractory, antithymocyte globulin.

The clinical course and outcomes of HZ infection in caseswere obtained by reviewing the transplant database and hos-pital medical records. The diagnosis of HZ infection wasmade if either the lesion scraping was positive for VZV directfluorescent antibody or culture or all of the following 3 criteriawere met: 1) classic appearance of HZ lesions; 2) dermatomaldistribution characteristic of HZ; and 3) decision to treat withantiviral therapy. When available, the following data werecollected: time from onset of rash to resolution, dermatomaldistribution, appearance of rash, diagnostic confirmation(e.g., direct fluorescent antibody, culture, PCR), type andlength of antiviral treatment, length of hospitalization, vis-ceral involvement (e.g., pneumonia, hepatitis, encephalitis),other complications during hospitalization (e.g., coexistentinfections, graft dysfunction, renal failure), and complicationsafter hospitalization (e.g., postherpetic neuralgia, cutaneousscarring, recurrent HZ infection, number of acute rejectionsand non-HZ infections in the year following HZ infection,and bilirubin and creatinine 1 year after HZ infection).

Figure 1 shows the specific time points post-OLT for

which data were collected on each patient enrolled. The “HZtime point” is the post-OLT time point when HZ infectionwas 1st diagnosed in cases and the corresponding time pointpost-OLT in matched controls. The variables collected werecompared between cases and controls to determine potentialrisk factors for and outcomes of HZ after OLT. Survival wasdefined as the time from HZ infection to death from anycause. Survival for patients who were alive was censored at thetime of the patient’s last contact date.

Statistical Analysis

Differences in risk factors and long-term outcomes were com-pared between cases and controls using the Wilcoxon signed-rank test for continuous data and McNemar’s test for categor-ical data. A P value of �.05 was considered statisticallysignificant. Multivariate analysis was conducted using gener-alized estimating equations to account for the correlatednature of the data. The continuous and categorical variablesthat were significant at the .20 level in univariate analysis wereentered into the model and a forward selection procedure wasapplied. Continuous variables were analyzed on the log scale.Overall survival was compared between cases and controlsusing the log-rank test.

Results

Baseline Data

A total of 29 (26 adults, 3 children) patients developedHZ at a median 4.9 (0.5-12.9) years after OLT. Base-line characteristics were as follows: median age 59(6-71), 66% female, 83% Caucasian, 7% prior livertransplant. The causes of liver disease prior to OLTwere cholestatic liver disease (i.e., primary biliary cir-rhosis primary sclerosing cholangitis, secondary biliarycirrhosis, intrahepatic cholestasis) 48%, cryptogeniccirrhosis 17%, and other 35%. Matched HZ-negativecontrols had the following baseline characteristics:

Figure 1. Timeline and variables for the pre- and post-HZ intervals / time points examined for cases and controls. Thevariables listed were collected within the intervals and at each time point. *Only infections requiring hospitalizations;**Median levels 1 month prior to HZ.

321HZ Infection After Liver Transplantation

median age 59 (6-73), 66% female, 93% Caucasian,14% prior liver transplant. The causes of liver diseaseprior to OLT were cholestatic liver disease 52%, cryp-togenic cirrhosis 24%, and other 24%. There was noevidence of any differences in the baseline characteris-tics between cases and controls: Caucasian race (83 vs.93%; P � .18), retransplantation (7 vs. 14%; P � .41),cause of liver disease (cholestatic 48 vs. 52%; crypto-genic 17 vs. 24%; other 35 vs. 24%; P � .84).

Incidence of HZ Infection

Of the 29 patients who developed HZ during the studytime interval (September 1, 1993, to April 1, 2004), 18were transplanted between September 9, 1987, andAugust 31, 1993, while 11 were transplanted betweenSeptember 1, 1993, and June 26, 2001. Clinical informa-tion prior to August 31, 1993, was recorded by paper chartand not electronically in our transplant database. Unlikeour electronic database after September 1, 1993, the papercharts contain very little information on clinical events,such as HZ infections, occurring in other institutions,making incidence of HZ infection prior to August 31,1993, impossible to calculate. Therefore, to accurately cal-culate the incidence of HZ infection in our study, we onlyincluded those patients (11) who were transplanted in thestudy time period (September 1, 1993, to April 1, 2004)during which the electronic database was being utilized. Asa result, the overall incidence specifically in the study timeperiod was 1.2% (11 patients with HZ infection / 942OLT recipients in this time period).

Clinical Course of HZ Infection

All HZ infections except 1 were localized to a singledermatome. One patient had cutaneous (not visceral)dissemination involving multiple dermatomes. Skinlesions were seen in multiple distributions, includingface / mouth, neck, trunk, back, and extremities. Threepatients had eye involvement that did not result invisual disturbance or loss. Herpes zoster infection wasconfirmed by direct fluorescent antibody in 2 cases inwhich the diagnosis was uncertain. During treatment,10 patients had sufficient documentation of the timeinterval from the onset of HZ infection to lesion reso-lution (crusting): median 7.5 (range 5-30) days. Four(14%) were treated with intravenous acyclovir alone, 9(31%) intravenous followed by oral acyclovir, 11 (38%)oral acyclovir alone, and 5 (17%) with other oral anti-virals. All patients with facial involvement (4) or cuta-neous dissemination (1) were treated with intravenousacyclovir.

No patients developed visceral involvement, such ashepatitis, pneumonitis, or encephalitis. Of the 8 (28%)

patients requiring hospitalization, 7 were hospitalizedin order to receive treatment with intravenous acyclo-vir. None of these 7 patients developed any other med-ical complications, e.g., other infections, graft dysfunc-tion, or renal failure. The patient with cutaneousdissemination stayed in the hospital for 4 weeks second-ary to cyclosporine-induced renal failure, unrelated toHZ infection. On long-term follow-up, 4 (14%) hadpostherpetic neuralgia and 3 (10%) had recurrent HZ.The median interval from initial infection to recurrentHZ in those 3 patients was 180 days (range 120-515).Deaths occurred late (median 2.7 years; range 0.8-8.1)after HZ infection in cases and none were directlyattributable to HZ infection (Table 1).

Risk Factors for HZ

Of the risk factors analyzed on univariate analysis, onlythe number of non-HZ infections in the 12 monthspreceding the HZ time point was significantly higher incases than controls (P � .05) (Table 2). There was alsoa trend toward higher doses of prednisone in the casescompared to the controls (P � .07). On multivariateanalysis, none of the risk factors with P � .2 (i.e., acuterejection, non-HZ infections, number on prednisone,prednisone dose, median creatinine, median bilirubin)in univariate analysis were significantly differentbetween cases and controls.

Outcomes

Figure 2 displays the Kaplan-Meier survival curves forcases and controls. The estimated 10-year survival rate

Table 1. Causes of Death After Herpes Zoster Infection

Patient EtiologyInterval from HZto Death (Years)

1 Sepsis (VRE) 4.42 Sepsis (unknown etiology) 1.93 Sepsis (unknown etiology) 2.34 Sepsis (pseudomas) 5.45 Sepsis (foot gangrene) &

renal failure2.3

6 Recurrent HCC 4.27 PTLD 8.18 Heart failure (ischemic) 1.69 Heart failure (ischemic) 3.2

10 Liver failure (chronicrejection)

.8

11 Unknown 2.7

Abbreviations: VRE, Vancomycin-resistant enterococcus;HCC, hepatocellular carcinoma; PTLD, post-transplantlymphoproliferative disorder.

322 Levitsky et al.

was 73% for the controls (95% confidence interval:52-94%) compared to 54% for the cases (95% confi-dence interval: 33-75%) (P � .05). In both univariateand multivariate analysis, none of the other outcomes(12-month infection and rejection rate, bilirubin, cre-atinine) were significantly different between cases and

controls. No patients were lost to follow-up or wereretransplanted after the HZ time point.

Discussion

In this case-control study of HZ infection after liver trans-plantation, the occurrence of HZ was infrequent and clin-ically mild. All but one patient had unidermatomal infec-tions and only 28% of patients were hospitalized. Themajority were treated as outpatients, often with oral anti-viral medications. Only 1 patient developed cutaneousdissemination and had a prolonged hospitalization due toissues unrelated to HZ infection. Postherpetic neuralgiawas seen in 14% of patients, similar to what is seen inimmunocompetent patients, and lower than in the elderlypopulation.13,14 No patients developed organ involve-ment or died from HZ infection.

In the study time period (September 1993 to April2004), the overall incidence of HZ infection after OLTwas low at 1.2%. It is difficult to directly compare theincidence in our study to those reported in healthypopulation studies, as multiple differences in studydesign, length of follow-up, and incidence calculationsexist among these studies.13,15–18 The largest of thesestudies showed a 0.43% 2-year incidence of HZ infec-tion among healthy patients in a community practice.18

Table 2. Risk Factors and Outcomes Before and After Herpes Zoster Time Point (Univariate Analysis)

Case (n � 29) Control (n � 29)

Risk factors (pre-HZ)Antiviral drug exposure—no. (%)

�3 months post-OLT 22* (79%) 18* (78%)3–12 months post-OLT 6* (21%) 5* (22%)�12 months pre-HZ time point 11 (38%) 9 (31%)

Acute rejection—no. (%)†,‡ 9 (31%) 4 (14%)Non-HZ infections—no. (%)† 6 (21%) 2 (7%)Immunosuppression—no. (%)

FK-506 5 (17%) 8 (28%)Cyclosporine 24 (83%) 21 (72%)Prednisone† 25 (86%) 28 (97%)Azathioprine or Mycophenylate 0 (0%) 0 (0%)

Median FK-506 level (range) 8.8 (4.7–21.5) 10.2 (2.8–15)Median cyclosporine level (range) 239 (109–761) 278 (96–553)Median prednisone dose (range)† 5.0 (0–60) 5.0 (0.5–15)Median creatinine (range)† 1.6 (0.6–6.3) 1.5 (0.6–7.1)Median bilirubin (range)† 0.7 (0.3–5.8) 0.6 (0.2–3)

Outcomes (post-HZ)Acute rejection—no. (%) 3 (10%) 3 (10%)Non-HZ infections—no. (%) 5 (17%) 2 (7%)Median creatinine (range) 1.4 (0.5–8.3) 1.5 (0.5–6.3)Median bilirubin (range) 0.6 (0.3–2.1) 0.6 (0.2–1.8)Estimated 10-year survival† 54% 73%

*Data available on 28 cases and 23 controls.†P value � .2.‡All rejection episodes responded to steroid monotherapy.

Figure 2. Kaplan-Meier survival curves for both cases andcontrols after the HZ time point. As demonstrated, caseshad a significantly lower long-term survival than controls.

323HZ Infection After Liver Transplantation

Another study showed that 9.9% of elderly patientsreported a previous episode of HZ in their lifetime.17

Given this information, the 10-year HZ incidence of1.2% in our transplant recipients is probably similar tothat of the general population. The reason for this maybe that most of our patients developed HZ long afterOLT (median 4.9 years) when immunosuppression wasgenerally low. At the time of HZ, our patients weregenerally on low doses of prednisone, had moderatelylow levels of calcineurin-inhibitors, and did not havesignificant renal or liver dysfunction.

In contrast to our study, other studies have reportedmore serious complications and higher incidences ofHZ in OLT recipients and in other immunosuppressedindividuals, such as recipients of bone marrow trans-plantation and patients with human immunodeficiencyvirus, cancer, and collagen-vascular diseases.3,19–27

Hepatitis, idiopathic thrombocytopenic purpura, andocular involvement due to VZV after OLT have beenreported, although it is uncertain if these complicationswere due to primary or reactivation VZV.5,7,8 Threestudies evaluating �600 patients reported an overallincidence of 5.7 to 12% of HZ infection afterOLT.10–12 Gourishankar et al.11 reported that themajority had HZ within 1 year after OLT and oftendeveloped postherpetic neuralgia. Azathioprine wasused in over 80% of their patients. Similarly, Herrero etal.12 reported a higher incidence (12%) and shorteronset of HZ after OLT (median 1.9 years) compared toour study and found that the only factor related to HZreactivation was treatment with azathioprine or myco-phenolate. These higher incidences and significantlyearlier presentations of HZ after OLT in comparison toour study are likely secondary to differences in thedegree of immediate and long term post-OLT immu-nosuppression among our centers. The fact that none ofour patients were on azathioprine or mycophenolatelikely played a substantial role in the reduced incidenceand delayed clinical presentation in our study.

This case-control study was performed to identifyany potential risk factors for the development of HZand compare outcomes after HZ. Only nonviral infec-tions and prednisone dose were identified as potentialrisk factors on univariate analysis. However, in multi-variate analysis, none of the factors examined were sig-nificantly different between cases and controls. In par-ticular, there was not evidence of a difference in thenumber of acute rejection episodes and non-HZ infec-tions, prior antiviral therapy, type and level of immu-nosuppression, and renal and liver dysfunction betweencases and controls. For short-term outcomes (�1 yearafter HZ time point), there was also no evidence of adifference in the development of acute rejection,

non-HZ infections, renal dysfunction, and liver dys-function between cases and controls.

An interesting, unexpected finding was poorer survivalin HZ cases compared to controls. Deaths occurred yearsafter HZ infection and were often potentially related toimmunosuppression (e.g., sepsis, malignancy) (Table 1).A larger control group would have more clearly elucidatedthe validity of this association. That being said, an episodeof HZ infection may be a surrogate marker of T-cell dys-function and adverse long-term outcomes. While HZinfection in OLT recipients appears self-limited with fewcomplications and no immediate mortality, it may indi-cate a somewhat worse overall prognosis and should alertthe clinician to reevaluate the degree of long-term immu-nosuppression.

There are several limitations of this study. First, whileexcellent communication generally exists between ourcenter and the patients’ long-term care providers, such thatclinical events like VZV are routinely documented in ourdatabase, there may have been some cases of HZ notreported to our center. Thus, it is possible that the inci-dence of HZ after OLT reported in this study underesti-mates the true incidence. Second, while we were primarilyinterested in recent trends in HZ infection after OLT, wecould not determine the incidence of HZ infection priorto September 1993, when over 50% of our patients weretransplanted. Third, there may have been other risk factorsand outcomes not analyzed in our study. Aside from sur-vival and causes of death, we did not investigate otherlong-term outcomes �1 year after the HZ time point, e.g.,graft survival, other infections, or acute / chronic rejection.Finally, the reasonably large number of variables evaluatedin comparison to the small number of cases may havehindered the detection of statistically significant risk fac-tors in our univariate and multivariate analyses. The lackof differences among cases and controls may have been dueto a low power to detect such differences.

The mild clinical presentation and good outcome ofHZ after OLT do not support the use of adult vaccina-tion to prevent HZ infection before or after OLT. Onereport advocated the use of the VZV vaccine in VZVimmunoglobulin G–negative adults prior to OLT toreduce the risk of primary varicella infection or chick-enpox.5 In our study time period, only 1 adult (of 678adult recipients) had chickenpox after OLT (data notshown). Given the rare occurrence of primary VZVafter OLT and the mild presentation of reactivationVZV in adults, the recommendation to vaccinate trans-plant candidates or recipients to prevent VZV infectionis probably not justified. Furthermore, the VZV vaccineis a live attenuated vaccine with costs and potentialrisks, in particular when given to transplant recipientsor other immunosuppressed adults.28

324 Levitsky et al.

Our data also suggest that the majority of infectionscan be managed as an outpatient with either oral orintravenous antiviral medications, as patients did wellwith either therapy route. Overall, the treatment of HZin a liver transplant recipient should generally parallelthat of a healthy adult with HZ. The decision to treatwith oral or intravenous antiviral therapy should bebased on a case-by-case clinical assessment of the degreeof immunosuppression and the extent, location, andseverity of HZ lesions.

In summary, of the �1,000 patients transplanted atour center in the last 10 years, only a small number ofrecipients developed HZ infection. No risk factors for HZinfection were identified on multivariate analysis. Infec-tions occurred years after OLT, were clinically mild andtreatable, and did not result in any significant immediateor short-term complications. Most HZ infections in OLTrecipients may be treated with oral antiviral therapy,reserving intravenous therapy for more extensive infec-tions. Surprisingly, despite the relatively benign nature ofHZ in post-OLT patients, long-term survival is lower inthis group compared to recipients without HZ. HZ maybe a marker of T-cell dysfunction from long-term immu-nosuppression and potentially worse long-term outcomes.Overall, HZ occurs infrequently long after OLT, is not alife-threatening infection, and can be treated with a lowlikelihood of organ dissemination.

References1. Center for Disease Control. National Center for Infectious Dis-

eases. Infectious Disease Information. Varicella-zoster virusinfection (VZV infection, VZV infection, shingles, zoster).Available at: http://www.cdc.gov/ncidod/diseases/list_varicl.htm. MMWR. Decline in Annual Incidence of Varicella inSelected States, 1990-2001. Accessed December 16, 2004.

2. LaRuss P. Clinical manifestations of varicella. In: Arvin AM, Ger-shon AA. Varicella-zoster virus: virology and clinical management.Cambridge: Cambridge University Press, 2000:206–207.

3. Locksley RM, Flournoy N, Sullivan KM, Meyers JD. Infectionwith varicella-zoster virus after marrow transplantation. J InfectDis 1985;152:1172–1181.

4. McGregor RS, Zitelli BJ, Urbach AH, Malatack JJ, Gartner JC.Varicella in pediatric orthotopic liver transplant recipients. Pedi-atrics 1989;83:256–261.

5. Kusne S, Pappo O, Manez R, Pazin G, Carpenter B, Fung JJ,Starzl TE. Varicella-zoster virus hepatitis and a suggested man-agement plan for prevention of VZV infection in adult livertransplant recipients. Transplantation 1995;60:619–621.

6. Breinig MK, Zitelli B, Starzl TE, Ho M. Epstein-Barr virus,cytomegalovirus, and other viral infections in children after livertransplantation. J Infect Dis 1987;156:273–279.

7. Singh N, Gayowski T, Yu VL. Herpes zoster-associated idio-pathic thrombocytopenic purpura in a liver transplant recipient:a case report and overview. Transpl Int 1995;8:58–60.

8. Papanicolaou GA, Meyers BR, Fuchs WS, Guilory SL, Mendel-son MH, Sheiner P, et al. Infectious ocular complications in

orthotopic liver transplant patients. Clin Infect Dis 1997;24:1172–1177.

9. Haagsma EB, Klompmaker IJ, Grond J, Bijleveld CMA, TheTH, Schirm J, et al. Herpes virus infections after orthotopic livertransplantation. Transpl Proc 1987;19:4054–4056.

10. Singh N, Dummer JS, Kusne S, Breinig MK, Armstrong JA,Makowka L, et al. Infections with cytomegalovirus and otherherpesvirus in 121 liver transplant recipients: transmission bydonated organ and the effect of OKT3 antibodies. J Infect Dis1988;158:124–131.

11. Gourishankar S, McDermid JC, Jhangri GS, Preiksaitis JK. Her-pes zoster infection following solid organ transplantation: inci-dence, risk factors, and outcomes in the current immunosuppres-sive era. Am J Transplant 2004;4:108–115.

12. Herrero JI, Quiroga J, Sangro B, Pardo F, Rotellar F, Alvarez-Cienfuegos JA, Prieto J. Herpes zoster after liver transplantation:incidence, risk factors, and complications. Liver Transpl 2004;10:1140–1143.

13. Ragozzino MW, Melton LJ 3rd, Kurland LT, Chu CP, PerryHO. Population-based study of herpes zoster and its sequelae.Medicine (Baltimore) 1982;61:310–316.

14. Choo PW, Galil K, Donahue JG, Walker AM, Spiegelman D,Platt R. Risk factors for postherpetic neuralgia. Arch Intern Med1997;157:1217–1224.

15. Thomas SL, Hall AJ. What does epidemiology tell us about riskfactors for herpes zoster. Lancet Infect Dis 2004;4:26–33.

16. Richards P. Shingles in one family practice. Arch Fam Med1996;5:42–46.

17. Schmader K, George LK, Burchett BM, Pieper CF, HamiltonJD. Racial differences in the occurrence of herpes zoster. J InfectDis 1995;171:701–704.

18. Donahue JG, Choo PW, Manoson JE, Platt R. The incidence ofherpes zoster. Arch Intern Med 1995;155:1605–1609.

19. Buchbinder SP, Katz MH, Hessol NA, Liu JY, O’Malley PM,Underwood R, Holmberg SD. Herpes zoster and human immu-nodeficiency virus infection. J Infect Dis 1992;166:1153–1156.

20. Melbye M, Grossman RJ, Goedert JJ, Eyster ME, Biggar RJ.Risk of AIDS after herpes zoster. Lancet 1987;1:728–730.

21. Glesby MJ, Moore RD, Chaisson RE. Clinical spectrum of her-pes zoster in adults infected with human immunodeficiencyvirus. Clin Infect Dis 1995;21:370–375.

22. Engels EA, Rosenberg PS, Biggar RJ. Zoster incidence in humanimmunodeficiency virus-infected hemophiliacs and homosexualmen, 1984–1987. J Infect Dis 1992;166:1153–1156.

23. Morgan D, Mahe C, Malamba S, Okongo M, Mahanja B, Whit-worth J. Herpes zoster and HIV-1 infection in a rural Ugandancohort. AIDS 2001;15:223–229.

24. Schimpff S, Serpick A, Stoler B, Rumack B, Mellin H, JosephJM, Block J. Varicella-zoster infection in patients with cancer.Ann Intern Med 1972;76:241–254.

25. Sokal JE, Firat D. Varicella-zoster infection in Hodgkin’s dis-ease. Am J Med 1965;39:452.

26. Anotnelli MAS, Moreland LW, Brick JE. Herpes zoster inpatients with rheumatoid arthritis treated with weekly low-dosemethotrexate. Am J Med 1991;90:295–298.

27. Nagasawa K, Yamauchi Y, Tada Y, Jusaba T, Niho Y, YoshikawaH. High incidence of herpes zoster in patients with systemiclupus erythematosus: an immunological analysis. Ann RheumDis 1990;49:630–633.

28. Levitsky J, Te HS, Faust TW, Cohen SM. Varicella infectionfollowing varicella vaccination in a liver transplant recipient.Am J Transplant 2002;2:880–882.

325HZ Infection After Liver Transplantation