It is estimated that more than 1 million cases of herpes zoster (HZ) occur in the United States each year.1 Most cases of HZ occur in adults older than 60 years.2 Although more than 90% of adults have serologic evidence of varicella zoster infection and are therefore at risk of developing HZ, HIV-infected patients have been repeatedly found to have higher incidence rates than the general population.3–8 Even after the institution of antiretroviral therapy (ART), the high rate of HZ persisted in persons living with HIV (PLWH).3,9
Furthermore, studies suggested that PLWH were particularly susceptible to complicated HZ particularly recurrence, multidermatomal involvement, and systemic disease. Studies suggested that early in the combination ART era, ART did not correct the disparity between PLWH and the general population.3 Some studies suggest that ART initiation and the resultant immune reconstitution syndrome might even increase a patient's risk of developing HZ.
As advances in HIV therapy allow patients to live longer and the HIV-infected population ages, it would be plausible to think that the burden of HZ would increase in the HIV-infected population. Moreover, a new varicella vaccine to protect against zoster was approved by Food and Drug Administration in 2006 and could greatly impact the epidemiology of this disease. In this study, we reevaluated the incidence, risk factors, and incidence of complicated HZ in our urban cohort of HIV patients.
Population and Data Collection
The Johns Hopkins University AIDS Service provides both primary and subspecialty care for PLWH in the Maryland region. The clinic maintains an observational database on patients. Trained monitors use structured forms to extract extensive information including demographics, clinical and laboratory data, pharmaceutics, and deaths. Institutional electronic sources supplement the record abstraction.10 Maintenance of the database and use of its contents for analysis is approved by the Institutional Review Board of the Johns Hopkins University School of Medicine. All patients receiving primary HIV care enrolled in this clinic were eligible for inclusion into this study.
For this study, we identified all cases of clinically diagnosed HZ diagnosed between January 1, 2002 and December 31, 2009 in this cohort. Medical records were systematically reviewed for the clinical characteristics, treatment, and sequelae of all new cases of HZ.
We defined HZ as a rash that a health care provider verified as HZ. HZ was considered a first episode if there was no evidence in the medical record of a previous occurrence. Complicated HZ was defined as the occurrence of 1 or more of the following: disseminated zoster (zoster involving 3 or more dermatomes); ocular, visceral, or neurologic complication consistent with HZ but not attributable to an independent pathologic process; or a recurrence of HZ within 180 days of presentation. Consistent with prior literature postherpetic neuralgia (PHN) was defined as pain lasting more than 120 days from the initial onset of the herpes rash.11
HIV risk factors were defined as past or present exposure to injection drug use, men who have sex with men, or high-risk heterosexual contact. Substance use was categorized as current if the use was recorded in the medical record within 1 year of the HZ episode.
Comorbidities were assessed through the patient's medical record. Depression was defined as a diagnosis of depression previously made by a medical or psychiatric provider. Hepatitis C was defined as chronic hepatitis C infection, and hepatitis B as chronic, active hepatitis B infection.
The data we collected were entered and managed using a REDCap (Research Electronic Data Capture) electronic data capture tool hosted by the Johns Hopkins Biostatistics Center.13
For normally distributed variables, means and SDs were calculated. For nonparametric data, medians and interquartile ranges are reported. Normally distributed, continuous variables were compared using the Student t test, categorical variables with the Fisher exact test, and nonparametric data with the Wilcoxon rank-sum test.
Incidence rates were calculated using annual totals of person-time in the cohort. We compared HZ incidence rates over time using Poisson regression. Risk factors were identified using a nested case–control analysis. Three controls were randomly selected for each case from the patients in the cohort based on year of enrollment, baseline CD4, and duration of follow-up. Conditional multivariate logistic regression was used to assess the risk factors for zoster.
To determine predictors of complicated HZ, we also used logistic regression methods. Individual risk factors were first identified using univariate analyses, with a P value cutoff for significance of 0.2. The variables that were identified in univariate analysis were then combined in a multivariable logistic regression model. Potential interactions between variables were evaluated by adding interaction terms to the multivariate models. The final model was selected using a modified backward stepwise selection method for nonsignificant P values and then choosing the model with the lowest Akaike information criterion. All analyses were performed using STATA 11.0 (Stata, Corp, College Station, TX).
Patient Demographics, Clinical Characteristics, and Incidence of Zoster
Between January 1, 2002 and December 31, 2009, there were 4353 patients with 19,752 person-years (PY) of follow-up time. During this period, there were 321 cases of HZ in 262 individuals. One hundred eighty-three cases (57%) were incident cases. The remaining 138 cases were recurrences that occurred during the study period. Incident cases were less likely than recurrent cases to be on ART at the time of the event (odds ratio, 0.42; P = 0.004). The incidence rate over the entire study period was 9.3 events per 1000 PY of follow-up and varied from 4.7 per 1000 PY in 2009 to 13.1 in 2003 (Fig. 1). There was no statistically significant difference between the annual incidence rates and the overall rate for the study period (P > 0.1).
Among incident HZ, the majority of patients were male (62%), African American (75%), and had heterosexual sex as their HIV risk factor (52%) (Table 1). The median age was 39 years (range, 18–68 years). Most patients actively used tobacco (63%), alcohol (46%), or illicit drugs (27%).
The median CD4 count at HZ was 278 cells per cubic millimeter (interquartile range, 153–444 cells /mm3; 18% with CD4 <350 cells/mm3), and the median HIV-1 RNA was 2581 copies per milliliter (range, <400 to > 700,000 copies/mL; 37% undetectable).
Most (75%) patients were on ART at the time of their zoster event. Of the 137 patients on ART, 10 (5.5%) had started ART within the past 90 days. Most (86%) patients had at least 1 comorbid condition. The most common conditions were hepatitis C (53%), depression (42%), and hepatitis B (26%) (Table 1). Notably, none of the patients in this study received the HZ vaccine in our clinic during the study interval.
Thirty cases were seen by an outside care provider at the time of their diagnosis and had unknown dermatomal involvement. Among the remaining 153 patients, 67% had single dermatomal involvement, with the most common location being the thorax (involved in 60% of patients). Nineteen patients (11%) had disseminated HZ (Table 2).
Among 168 cases for whom we had treatment data, 158 (94%) were treated with antiviral therapy for HZ. The most commonly prescribed antiviral was valacyclovir (77%). Thirty-six patients (20%) were hospitalized for their HZ. The hospitalized patients had a mean CD4 count of 233 cells per cubic millimeter, 68% were already on ART at the time of the episode, and 26 patients (74%) were treated with intravenous acyclovir.
In multivariate analysis, factors associated with an initial HZ outbreak included having initiated ART within 90 days of the episode (adjusted odds ratio [AOR], 4.02; 95% confidence interval (CI), 1.31 to 12.41), having a detectable HIV-1 RNA at the time of the episode (AOR, 1.49; 95% CI, 1.00 to 2.24), and having a CD4 count between 350 and 500 cells per cubic millimeter (AOR, 2.02; 95% CI, 1.14 to 3.57) or a CD4 count below 350 cells per cubic millimeter (AOR, 2.46; 95% CI, 1.42 to 4.23) as compared with patients with CD4 count above 500 cells per cubic millimeter (Table 3). Age, sex, race, HIV risk factor, current ART use, and comorbid depression, hepatitis B, or hepatitis C were not associated with a zoster outbreak in univariate or multivariate analyses.
Presenting Symptoms and Complications of Zoster
Not all patients initially presented within the Johns Hopkins system with their episode of HZ. For patients whose available medical records included their initial presentation (n = 133), the majority complained of pain (82.0%), rash (53.4%), and pruritis (33.8%). Over one quarter of patients (28.3%) had complicated zoster. The most common complications were PHN (11.9%), disseminated zoster (10.7%), bacterial superinfection (6.3%), ocular involvement (5.7%) and meningioencephalitis (2.8%). In univariate logistic analysis, having a CD4 count ≤50 cells per cubic millimeter (odds ratio, 2.86; 95% CI, 1.01 to 8.09) was significantly associated with developing complicated zoster (Table 4). However, this finding was no longer statistically significant after adjusting for age and viral load.
In this study, the incidence of HZ among an urban cohort of HIV-infected persons was 9.3 per 1000 PY. This rate has decreased significantly since a previous study conducted earlier in the combination ART era using this same cohort (32 cases/1000 PYs, between 1997 and 2001) and now seems to be approaching that of the general population estimated at approximately 3.5 cases per 1000 PY.3,14 This finding is in contrast to data from 2 other cohorts—the Veterans Health database and Olmsted County, Minnesota—that both showed small but significant increases in HZ over time (2000–2007 and 1996–2001, respectively).15,16
The observed decrease in incidence rate in our clinic might be explained by improvements in addressing the risk factors for HZ specific to PLWH. Consistent with our earlier study and other studies, we found that a lower CD4 count was associated with increased risk of incident HZ.3,4,8 Indeed, immune suppression is consistently a risk factor for HZ outbreak in this population, with a CD4 count below 350 cells per cubic millimeter conferred greater risk than a CD4 count between 350 and 500 cells per cubic millimeter. Given that the median CD4 count of our population has steadily increased from 2000 through 2009 from 298 to 431 cells per cubic millimeter, this finding highlights the importance of restoring immune function in protecting against HZ. The effectiveness of contemporary ART in improving immune function is likely to be the primary reason for the decrease in HZ incidence.
Unlike our earlier study, we also found that the recent ART initiation was strongly associated with incident disease, suggesting a possible role of immune dysregulation during immune reconstitution in producing HZ outbreaks. The association between early immune reconstitution after starting ART and HZ outbreaks has been previously observed in smaller studies, and it may be at least partially mediated through an increase in CD8 cell counts.17,18
Although the incidence rate has decreased in our population, it is still greater than the general population, especially when age is considered, and we continue to observe a significant complication rate (28%) among cases. If one excludes PHN, the rate is still 21%—far more than the 10%–13% rate observed in the general population.16,19,20 Our data are consistent with previous reports of higher complication rates in PLWH; however, the complication rate observed in this study was less than the 53% complication rate previously described in our clinic population.3,7 Importantly, since the earlier analysis, the definition of PHN has become more conservative now only including pain 120 days after the onset of rash, as compared with the previous definition that included pain that persisted 28 days after the disappearance of the rash. This change in definition might also account for the large decrease in the proportion of patients experiencing PHN (from 28% to 12% in this study). We identified a CD4 count ≤50 cells per cubic millimeter as a factor associated with complications on univariate analysis, but no factors were statistically significant on multivariate regression. Previous studies have identified low CD4 count, older age, and particular HIV risk factors to be associated with complications.3,19 Only 50 of our patients experienced complications, which was perhaps not a sufficient number to detect risk factors with only a small effect size. Our results are nevertheless consistent with the literature suggesting a higher complication rate in PLWH. Better access to improved HIV-related care may reduce the gap in complication rates between PLWH and the general population. Age was not a risk factor for either an HZ episode or an HZ complication in this study. Age is a known risk factor for HZ in the general population, but our population is still relatively young.
There are several potential limitations to this study. Our results are derived from patients followed at a single, urban institution with a relatively high proportion of injection drugs users and minority ethnicity and may not generalize to other clinical settings. We were unable to assess the impact of the 2 varicella vaccines. Although young, our population is still old enough to have had very low rates of varicella vaccination in childhood. Both the varicella and zoster vaccines are live-attenuated vaccines and as such could potentially be dangerous in an immunosuppressed population. Data on the use of the vaccine in HIV-infected adults are limited, although suggest that the vaccine can be immunogenic in HIV-infected patients.21 Based on expert opinion, vaccination of HIV-infected persons older than 8 years who are in Centers for Disease Control and Prevention clinical class A or B and have CD4 counts ≥200 cells per cubic millimeter may be considered.22 Although not currently licensed for this use, the zoster vaccine can be given to those who do not have AIDS or clinical manifestations of HIV, and a CD4 >200 cells per cubic millimeter or >15% of total lymphocytes.23 No patient in our clinical cohort had received a dose of zoster vaccine during the study period. Finally, our results rely on the documentation in the medical record, although we have no reason to believe that the clinical documentation of events and complications has decreased during our or the previous study periods. In addition, our methods of abstracting these data have not changed over time.
In summary, this study offers several important findings. First, incidence rates of HZ have decreased significantly since earlier in the combination ART era, moving closer to the general population rate, but still remain elevated. Second, complication rates of HZ remain high in PLWH, despite this group's much younger median age. Despite the high complication rate, and the high incidence rate in PLWH, not a single patient in our study population had been vaccinated against HZ. Finally, and perhaps most importantly, there are several modifiable risk factors for incident HZ, including having a detectable HIV-1 RNA level and a low CD4 count. However, HZ does seem to be associated with immune reconstitution, so that the clinician should be aware of the higher risk of HZ shortly after ART is started. Although age was not a risk factor for HZ in this study, it is a known risk factor for HZ in the general populations, and as the HIV-infected population continues to age on effective ART, we may see a greater burden of HZ.
1. 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;352:2271–2284.
2. Gilden DH, Kleinschmidt-DeMasters BK, LaGuardia JJ, et al.. Neurologic complications of the reactivation of varicella-zoster virus. N Engl J Med. 2000;342:635–645.
3. Gebo KA, Kalyani R, Moore RD, et al.. The incidence of, risk factors for, and sequelae of herpes zoster
patients in the highly active antiretroviral therapy era. J Acquir Immune Defic Syndr. 2005;40:169–174.
4. Veenstra J, Krol A, van Praag RM, et al.. Herpes zoster
, immunological deterioration and disease progression in HIV
-1 infection. AIDS. 1995;9:1153–1158.
5. Weller TH. Varicella and herpes zoster
: a perspective and overview. J Infect Dis. 1992;166(suppl 1):S1–S6.
6. Buchbinder SP, Katz MH, Hessol NA, et al.. Herpes zoster
and human immunodeficiency virus infection. J Infect Dis. 1992;166:1153–1156.
7. Glesby MJ, Moore RD, Chaisson RE. Clinical spectrum of herpes zoster
in adults infected with human immunodeficiency virus. Clin Infect Dis. 1995;21:370–375.
8. Glesby MJ, Hoover DR, Tan T, et al.. Herpes zoster
in women with and at risk for HIV
: data from the women's interagency HIV
study. J Acquir Immune Defic Syndr. 2004;37:1604–1609.
9. Song JY, Lee JS, Jung HW, et al.. Herpes zoster
-infected patients in the highly active antiretroviral therapy era: Korean HIV
cohort study. J Acquir Immune Defic Syndr. 2010;53:417–418.
10. Moore RD. Understanding the clinical and economic outcomes of HIV
therapy: the Johns Hopkins HIV
clinical practice cohort. J Acquir Immune Defic Syndr Hum Retrovirol. 1998;17(suppl 1):S38–S41.
11. Dworkin RH, Portenoy RK. Pain and its persistence in herpes zoster
. Pain. 1996;67:241–251.
12. Schneider E, Whitmore S, Glynn KM, et al.. Revised surveillance case definitions for HIV
infection among adults, adolescents, and children aged <18 months and for HIV
infection and AIDS among children aged 18 months to <13 years—United States, 2008. MMWR Recomm Rep. 2008;57(RR-10):1–12.
13. Harris PA, Taylor R, Thielke R, et al.. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377–381.
14. Hope-Simpson RE. The nature of herpes zoster
: a long-term study and a new hypothesis. Proc R Soc Med. 1965;58:9–20.
15. Rimland D, Moanna A. Increasing incidence of herpes zoster
among veterans. Clin Infect Dis. 2010;50:1000–1005.
16. Yawn BP, Saddier P, Wollan PC, et al.. A population-based study of the incidence and complication rates of herpes zoster
before zoster vaccine introduction. Mayo Clin Proc. 2007;82:1341–1349.
17. Domingo P, Torres OH, Ris J, et al.. Herpes zoster
as an immune reconstitution disease after initiation of combination antiretroviral therapy in patients with human immunodeficiency virus type-1 infection. Am J Med. 2001;110:605–609.
18. Martinez E, Gatell J, Moran Y, et al.. High incidence of herpes zoster
in patients with AIDS soon after therapy with protease inhibitors. Clin Infect Dis. 1998;27:1510–1513.
19. Galil K, Choo PW, Donahue JG, et al.. The sequelae of herpes zoster
. Arch Intern Med. 1997;157:1209–1213.
20. Ragozzino MW, Melton LJ III, Kurland LT, et al.. Population-based study of herpes zoster
and its sequelae. Medicine (Baltimore). 1982;61:310–316.
21. Benson C, Hua L, Andersen J, et al.. ZOSTAVAX is generally safe and immunogenic in HIV
+ adults virologically suppressed on ART
: results of a phase 2, randomized, double-blind, placebo-controlled trial. Presented at: Conference for Retroviruses and Opportunistic Infections; March 5–8, 2012; Seattle, WA. 2012:19.
22. Marin M, Guris D, Chaves SS, et al.; Advisory Committee on Immunization Practices, Centers for Disease Control and Prevention (CDC). Prevention of varicella: recommendations of the advisory committee on immunization practices (ACIP). MMWR Recomm Rep. 2007;56(RR-4):1–40.
23. Harpaz R, Ortega-Sanchez IR, Seward JF; Advisory Committee on Immunization Practices (ACIP) Centers for Disease Control and Prevention (CDC). Prevention of herpes zoster
: recommendations of the advisory committee on immunization practices (ACIP). MMWR Recomm Rep. 2008;57(RR-5):1–30; quiz CE 2–4.