SUPPLEMENT ARTICLE

Risk of Herpes Zoster in Adults Immunized with Varicella Vaccine

Sophie Hambleton,1,a Sharon P. Steinberg,1 Philip S. LaRussa,1 Eugene D. Shapiro,2 and Anne A. Gershon1
1

Department of Pediatrics, Columbia University College of Physicians and Surgeons, New York, New York; 2Departments of Pediatrics and of Epidemiology and Public Health, Yale University School of Medicine, New Haven, Connecticut

A program of routine varicella vaccination of children 1218 months of age, begun in the United States in 1995, has been very successful in reducing the incidence of varicella. Varicella-zoster virus (VZV), in both wild-type and live attenuated forms, is notable for its ability to produce latent infection of sensory neurons from which it can later reactivate to cause herpes zoster (HZ). Therefore, the effects of vaccination on this secondary VZV-related disease are important to consider; in practice, however, such studies are complicated by the typically long delay between acquisition of the virus and its reactivation. Studies of immunocompromised children have shown that vaccination is relatively protective against HZ in this highly vulnerable group. We now present long-term follow-up data on a group of individuals who received varicella vaccine as healthy young adults 1026 years ago and who have been followed prospectively by means of active surveillance. Among some 2000 person-years of follow-up, 2 cases of HZ have occurred, for a rate of 1.00 case/1000 personyears. Overall, the incidence of HZ in this cohort, therefore, is similar to published data for the US population in the prevaccine era. In 1995, live attenuated varicella vaccine was licensed for routine use with healthy varicella-susceptible children and adults in the United States. More than 57 million doses of the vaccine have been distributed since. The vaccine has been highly effective in preventing varicella, as judged by substantial reductions in varicellaassociated morbidity and mortality [14]. Although a signicant minority of vaccinees develop a modied form of varicella after exposure to varicella-zoster virus (VZV), their illness usually is extremely mild [58]. The effects of widespread use of varicella vaccine on the incidence of herpes zoster (HZ), the disease caused by reactivation of VZV, are still being determined. Initially, there were fears that the incidence of HZ might increase among vaccinees. The live attenuated varicella vaccine retains the ability to establish latent infection and to cause HZ that may be clinically indistinguishable from wild-type disease [9]. In a passive postlicensure surveillance study, typing of clinical specimens by polymerase chain reaction (PCR) analysis revealed that, of 56 suspected cases of HZ among vaccinees, 22 were caused by vaccine-strain VZV and 10 by wild-type VZV (the remaining specimens were inadequate for analysis, negative, or untypeable) [9, 10]. Currently available data do not enable an assessment of the rate at which either type of HZ occurs among immunocompetent individuals. In unvaccinated populations, the main risk factors for reactivation of VZV are advanced age and immunosuppression. Results of several clinical studies of VZV-naive immunocompromised children have sug-

Downloaded from http://jid.oxfordjournals.org/ by guest on December 4, 2012

Potential conicts of interest: A.A.G. lectures and consults on varicella-zoster virus vaccines for Merck and GlaxoSmithKline when invited, and receives research support from Merck. Additionally, Additionally, P.S.L., S.P.S., and A.A.G. are in a contractual relationship with Merck through the Varicella Zoster Virus Identication Program. S.H. and E.D.S. report no potential conicts. Financial support: National Institutes of Health grants AI 01703 and RR022477 (to E.D.S.) and AI 27187 and AI 24021 (to A.A.G.). The work described did not receive specic funding. Supplement sponsorship is detailed in the Acknowledgments. a Present afliation: Childrens Bone Marrow Transplant Unit, Newcastle General Hospital and University of Newcastle, Newcastle upon Tyne, United Kingdom. Reprints or correspondence: Dr. Sophie Hambleton, Childrens BMT Unit, Newcastle General Hospital and University of Newcastle, Newcastle upon Tyne NE4 6BE, United Kingdom (shambleton@doctors.org.uk). The Journal of Infectious Diseases 2008; 197:S1969 2008 by the Infectious Diseases Society of America. All rights reserved. 0022-1899/2008/19705S2-0030$15.00 DOI: 10.1086/522131

S196 JID 2008:197 (Suppl 2) Hambleton et al.

Table 1.

Subject characteristics and clinical events, by follow-up period.

Age at vaccination, mean (median), years 28.2 (26.1) 28.4 (26.9) 28.1 (26.1) 27.3 (25.6) 24.2 (24.3) 28.0 (26.0)

Follow-up period, a years 04.9 59.9 1014.9 1519.9 20 Total

Subjects, no. 363 129 64 28 6 363

Personyears 1242.1 459.3 213.6 68.5 12.2 1995.7

Vaccine doses, mean (median), no. 1.8 (2) 1.9 (2) 1.8 (2) 1.8 (2) 1.5 (1.5) 2.0 (2)

Seroconversion b rate, % 82.1 89.9 89.1 82.1 100 82.4

HHE Subjects, no. 87 16 5 0 0 108 Rate/year 0.070 0.035 0.023 0 0 0.054

BV Subjects, no. 29 10 0 0 0 39 Rate/year 0.023 0.022 0 0 0 0.02

Person-years after BV, no. (%) 30.8 (2.5) 29.6 (6.4) 14.2 (6.6) 1.7 (2.5) 0 76.3 (3.8)

NOTE. BV, breakthrough varicella; HHE, household exposure to varicella. Follow-up periods are grouped in accordance with years since vaccination. Seroconversion was dened as a uorescent antibody to membrane antigen titer of 1:2 at 48 weeks after receipt of the nal dose of vaccine. Rates are for all study subjects that contributed person-years to the relevant follow-up period.
b a

gested a protective effect against subsequent HZ (as well as varicella) with varicella vaccination [11, 12]. In our laboratory, immunization of healthy young VZV-seronegative adults against varicella began in 1979, in prelicensure clinical trials. Many of these vaccinees were health care workers. We have prospectively followed a cohort of 363 individuals for up to 25 years, by means of active surveillance. We report the incidence of HZ among these vaccinees, most of whom are now middle-aged adults who normally would be considered to be at increasing risk for HZ. METHODS The subjects participated in a variety of prelicensure clinical trials of varicella vaccines between 1979 and 1995. The characteristics of this group and details of their vaccination regimens have been described in a previous report [13]. All studies received institutional review board approval through Columbia University (New York). In brief, subjects were healthy adults with no clinical history of varicella and with serology results negative for VZV (as determined by uorescent antibody to membrane antigen [FAMA] assay); many subjects were health care workers. After informed consent was obtained, subjects were immunized in accordance with a variety of trial protocols. The majority of subjects received 2 doses (range, 13 doses) of research-lot vaccines that differed in potency, from 1500 to 3000 pfu/dose (Merck vaccines) and up to 10,000 pfu/dose (SmithKlineBeecham vaccines). Serologic responses were assessed by FAMA assay between 4 and 8 weeks after vaccination and at various intervals thereafter, in accordance with trial protocols. Samples of serum were obtained subsequently for testing by FAMA assay at annual intervals, when possible. Subjects were encouraged to contact study personnel and their own physicians at the time of any exposure to or possible symptoms of disease due to VZV, regardless of severity. When this occurred, a clinical assessment, samples of serum for testing by

FAMA assay, and vesicular uid for diagnostic studies (including PCR analysis) were obtained. To elicit details of any previously unreported exposure to VZV and/or any episodes of vesicular rash, subjects were contacted annually in the form of a postal questionnaire or, if they failed to respond, by telephone or letter. In the current analysis, follow-up was timed from the date of receipt of the rst dose of varicella vaccine (i.e., rst exposure to vaccine-strain VZV) to the date of last follow-up contact. Subjects with !1 year of follow-up were excluded from the analysis. HZ was diagnosed by report and/or by virus isolation and PCR analysis, as described elsewhere [9]. Person-years after HZ were excluded from the analysis. All data were entered into Microsoft Excel for analysis. RESULTS
Study Sample and Follow-up Duration

Downloaded from http://jid.oxfordjournals.org/ by guest on December 4, 2012

A total of 363 healthy adults (median age, 26 years; range, 17.1 56.7 years) were immunized against varicella and were followed up for at least 12 months; the mean duration of follow-up was 5.5 years (median, 4 years; range, 126 years). At last followup, subjects ranged in age from 21 to 62.7 years, with a mean of 33.7 years and a median of 32.4 years. Table 1 shows the distribution of person-years in our data set, by time since vaccination. The total number of evaluable person-years was 1996, of which 754 and 294 person-years were 15 and 110 years, respectively, after subject vaccination. Table 1 shows age at vaccination, number of vaccine doses received, rate of seroconversion, and annualized rates of household exposure to varicella and of breakthrough varicella, at various intervals, for study subjects who remained in follow-up. Among study subjects who dropped out within the rst 5 years, the seroconversion rate was 78%, whereas 90% of those in longer-term follow-up had seroconversion immediately after vaccination. Rates of house-

Risk of HZ in Adults JID 2008:197 (Suppl 2) S197

hold exposure to varicella and of breakthrough disease also were somewhat higher during this early period of follow-up.
HZ

Only 2 cases of HZ occurred, for an overall rate of 2 cases/ 1996 person-years (1.00 case/1000 person-years; 95% condence interval [CI], 0.123.62 cases/1000 person-years, by Fishers exact test). Case 1. Case 1 has been reported in detail elsewhere [14] and was found to be due to wild-type VZV (on the basis of typing of virus from vesicular uid, by PCR analysis). The study subject had shown initial seroconversion after 2 doses of varicella vaccine, but antibody titers waned by 20 months after vaccination. Subsequently, she showed a secondary increase in antibody titers after denite exposure to wild-type VZV. Despite the lack of accompanying symptoms of breakthrough varicella, the acquisition of wild-type VZV at that time was assumed, which resulted in reactivation that caused classic thoracic dermatomal HZ 6 months later. This subject was 33 years old at the time of the development of HZ, which occurred 3 years after she had received the varicella vaccine. Case 2. Case 2 was reported as classic HZ affecting the left shoulder, when the study subject was 53 years of age. She had received 2 doses of varicella vaccine, administered to the left deltoid, 11 years previously and did not have a history of vaccine-associated rash. As in case 1, this individual had shown a secondary increase in VZV antibody titers after exposure to varicella 4 years after vaccination (7 years prior to the development of HZ). However, this exposure had been accompanied by the development of 2 vesicles on the subjects lower limb that was thought to represent breakthrough varicella. Unfortunately, no clinical samples were obtained from this subject for PCR analysis for VZV. Thus, we do not know whether vaccine-strain or wild-type VZV was responsible for her clinically diagnosed episode of HZ. DISCUSSION HZ is not easy to study, whether at the epidemiological or pathophysiological level. However, the advent of universal varicella vaccination in the United States has brought a new focus to our efforts to understand the reactivation of VZV. The extent to which vaccinated individuals harbor latent infection with vaccine-strain VZV and the contribution that this makes to protective immunity against varicella are unknown. In the current study, we report the long-term incidence of HZ among subjects who had received varicella vaccine as healthy young adults in several prelicensure clinical trials. The observed HZ incidence, at 1.00 case/1000 person-years, is somewhat lower than previously observed rates of HZ among unvaccinated populations, which ranged from 2.15 to 4.05 cases/1000 person-years [1517]. This nding, therefore, is
S198 JID 2008:197 (Suppl 2) Hambleton et al.

consistent with the hypothesis that immunocompetent, vaccinated individuals are relatively protected against HZ, compared with unvaccinated persons harboring wild-type VZV. However, the CIs associated with the only 2 cases of HZ in our study group are necessarily wide. We did not have a control group in our study, and published data do not inform on the incidence of HZ among the relevant control population, namely, individuals who developed natural varicella as adults. Comprehensive cross-sectional studies have concentrated on age at the time of development of HZ, rather than the time since varicella, and give rates that increase steeply with age, particularly after the sixth decade [15, 16]. It is argued that the relationship between advancing years and the risk of HZ reects an age-related decline in cell-mediated immunity against VZV. However, it seems likely that this relationship would be offset at least partially by later acquisition of primary VZV infection. Hence, the rate of HZ among individuals acquiring wild-type VZV in adulthood would be expected to lie somewhere between that of their agematched peers (most of whom had experienced varicella in childhood) and that of younger people. Although rates of HZ are lowest among children, Donohue et al. [15] estimated a rate of 1.33 cases/1000 person-years among susceptible individuals in the age group !15 years. In the same study, rates of HZ increased progressively with age, from 1.92 cases/1000 person-years (age group, 2534 years) to 3.13 cases/1000 personyears (age group, 4554 years). The extremes for these values were similar to the 95% CIs of the incidence rates that we found in our vaccinated cohort. Thus, we can surmise that, as a group, our vaccinees have not obtained protection from varicella at the expense of an excess risk of HZ, when compared with their age-matched peers. However, the degree to which they may be protected against HZ, when compared with adults who have had wild-type VZV infection, cannot be estimated because of the lack of control subjects in this study. Some other limitations of our study should be emphasized. Follow-up data were partial, although the comparison presented in table 1 suggests that those subjects with longer followup are likely to be representative of the original population. Our data are heavily skewed toward the early years after vaccination (table 1) and, therefore, are less informative with respect to the later risk of HZ. In addition, our study subjects may have experienced higher than usual rates of occupational exposure to VZV, which may be associated with protection against HZ in adults who had varicella [18]. This same occupational exposure of vaccinated adults may be associated with an increase in the incidence of breakthrough varicella and perhaps, subsequently, of HZ. With these caveats in mind, we nonetheless are heartened by the relative rarity of HZ in this aging population of vaccinees. It has been shown previously that the risk of HZ in children

Downloaded from http://jid.oxfordjournals.org/ by guest on December 4, 2012

with leukemia is strongly linked to prior VZV-related rash, whether caused by wild-type or vaccine-strain VZV [12]. This observation is consistent with the hypothesis that latent VZV infection follows the infection of sensory neurons that innervate infected skin. Such a model suggests that HZ caused by vaccinestrain VZV would be associated with a history of vaccine-associated rash, whereas HZ caused by wild-type VZV would be associated with prior breakthrough varicella. The current data do not enable us to address this issue. In future larger studies, the inuence of these relatively common complications on the risk of HZ among vaccinees will be important to examine. In our cohort, which experienced only a moderate rate of breakthrough varicella (2%/person/year), 1 case of HZ occurred after clinically evident breakthrough varicella, whereas the other case appeared to have occurred after subclinical breakthrough varicella. We therefore must consider the possibility that there may be an excess risk of HZ among the minority of individuals who acquire wild-type VZV infection despite vaccination. Furthermore, we must recognize that this risk may be conferred even if the breakthrough varicella is subclinical, as in case 1. To document the changing epidemiology of HZ in the era of varicella vaccination, large prospective studies currently are in progress in sentinel areas of the United States. Interpretation of their results will be complicated by the relative paucity of historical data; furthermore, the data that are available suggest that the incidence of HZ was increasing in all age groups before the introduction of the varicella vaccination program [15]. Careful analysis will be required in order to distinguish between the direct and indirect (herd) effects of mass vaccination, because of the potential impact of reduced virus circulation on the longevity of immunity against HZ [18]. The Shingles Prevention Study (Veterans Affairs San Diego Healthcare System) recently reported impressive trial data demonstrating the efcacy of a high-potency varicella vaccine among older individuals [19]. Such a vaccine shows the potential for a signicant reduction in the incidence of HZ. In the meantime, our current data add to the existing body of evidence suggesting that individuals vaccinated against varicella are not at heightened risk of subsequently developing HZ.

References
1. Seward JF, Watson BM, Peterson CL, et al. Varicella disease after introduction of varicella vaccine in the United States, 19952000. JAMA 2002; 287:60611. 2. Va zquez M, LaRussa PS, Gershon AA, et al. Effectiveness over time of varicella vaccine. JAMA 2004; 291:8515. 3. Nguyen HQ, Jumaan AO, Seward JF. Decline in mortality due to varicella after implementation of varicella vaccination in the United States. N Engl J Med 2005; 352:4508. 4. Zhou F, Harpaz R, Jumaan AO, Winston CA, Shefer A. Impact of varicella vaccination on health care utilization. JAMA 2005; 294: 797802. 5. Gershon AA, Steinberg SP, Gelb L. Live attenuated varicella vaccine use in immunocompromised children and adults. NIAID Collaborative Varicella Vaccine Study Group. Pediatrics 1986; 78:75762. 6. Gershon AA, Steinberg S, Gelb L. Live attenuated varicella vaccine: efcacy for children with leukemia in remission. NIAID Collaborative Varicella Vaccine Study Group. JAMA 1984; 252:35562. 7. Seward JF, Zhang JX, Maupin TJ, Mascola L, Jumaan AO. Contagiousness of varicella in vaccinated cases: a household contact study. JAMA 2004; 292:7048. 8. Va zquez M, LaRussa P, Gershon A, Steinberg S, Freudigman K, Shapiro E. The effectiveness of the varicella vaccine in clinical practice. N Engl J Med 2001; 344:95560. 9. LaRussa P, Steinberg S, Shapiro E, Va zquez M, Gershon A. Viral strain identication in varicella vaccinees with disseminated rashes. Pediatr Infect Dis J 2000; 19:10379. 10. Sharrar RG, LaRussa P, Galea S, et al. The postmarketing safety prole of varicella vaccine. Vaccine 2000; 19:91623. 11. Broyer M, Tete MT, Guest G, Gagnadoux MF, Rouzioux C. Varicella and zoster in children after kidney transplantation: long-term results of vaccination. Pediatrics 1997; 99:359. 12. Hardy IB, Gershon A, Steinberg S, LaRussa P. The incidence of zoster after immunization with live attenuated varicella vaccine: a study in children with leukemia. N Engl J Med 1991; 325:154550. 13. Ampofo K, Saiman L, LaRussa P, Steinberg S, Annunziato P, Gershon A. Persistence of immunity to live attenuated varicella vaccine in healthy adults. Clin Infect Dis 2002; 34:7749. 14. Hammerschlag MR, Gershon AA, Steinberg SP, Clarke L, Gelb LD. Herpes zoster in an adult recipient of live attenuated varicella vaccine. J Infect Dis 1989; 160:5357. 15. Donahue JG, Choo PW, Manson JE, Platt R. The incidence of herpes zoster. Arch Intern Med 1995; 155:16059. 16. Insinga RP, Itzler RF, Pellissier JM, Saddier P, Nikas AA. The incidence of herpes zoster in a United States administrative database. J Gen Intern Med 2005; 20:74853. 17. Jumaan AO, Yu O, Jackson LA, Bohlke K, Galil K, Seward JF. Incidence of herpes zoster, before and after varicella-vaccinationassociated decreases in the incidence of varicella, 19922002. J Infect Dis 2005; 191: 20027. 18. Thomas S, Wheeler J, Hall AJ. Contacts with varicella or with children and protection against herpes zoster in adults: a case-control study. Lancet 2002; 360:67882. 19. Oxman MN, Levin MJ, Johnson GR, et al. A vaccine to prevent herpes zoster and postherpetic neuralgia in older adults. N Engl J Med 2005; 352:227184.

Downloaded from http://jid.oxfordjournals.org/ by guest on December 4, 2012

Acknowledgments
Supplement sponsorship. This article was published as part of a supplement entitled Varicella Vaccine in the United States: A Decade of Prevention and the Way Forward, sponsored by the Research Foundation for Microbial Diseases of Osaka University, GlaxoSmithKline Biologicals, the Sabin Vaccine Institute, the Centers for Disease Control and Prevention, and the March of Dimes.

Note added in proof. Since this article was accepted for publication, additional person-years of follow-up have accrued, producing a revised incidence of herpes zoster of 0.91 cases/1000 person-years (95% condence interval, 15.3300.6 cases/1000 person-years).
Risk of HZ in Adults JID 2008:197 (Suppl 2) S199