Use of a Live Attenuated Varicella Vaccine to Boost Varicella-Specific Immune Responses in Seropositive People 55 Years of Age and Older: Duration of Booster Effect
- Myron J. Levin,
- Deb Barber,
- Eileen Goldblatt,
- Michelle Jones,
- Bonnie LaFleur,
- Christina Chan,
- Diane Stinson,
- Gary O. Zerbe and
- Anthony R. Hayward
- Departments of Pediatrics, Pediatric Infectious Diseases, of Medicine, of Pediatrics, Allergy and Immunology, and of Preventive Medicine and Biometrics, University of Colorado School of Medicine, Denver, Colorado; Merck Research Laboratories, West Point, Pennsylvania
- Reprints or correspondence: Dr. Myron J. Levin, University of Colorado School of Medicine, Dept. of Pediatrics, Pediatric Infectious Diseases, 4200 East Ninth Ave., Box C-227, Denver, CO 80262.
Abstract
Varicella-zoster virus (VZV)-specific T cell immunity was measured in 130 persons ≥55 years of age 6 years after they received a live attenuated VZV vaccine. Circulating T cells, which proliferated in vitro in response to VZV antigen, were enumerated (VZV responder cell frequency assay). Six years after the booster vaccination, the VZV—responding cell frequency (1/61,000 circulating cells) was still significantly (P < .05) improved over the baseline measurements (1/70,000) and appears to have diminished the expected decline in frequency as these vaccinees aged (to 1/86,000). Ten herpes-zoster-like clinical events were recorded. Although the frequency of these events, ∼ 1/100 patient-years, is within the expected range of such events for this age cohort, the number of lesions was small, there was very little pain, and there was no postherpetic neuralgia. These results support the development of a vaccine to prevent or attenuate herpes zoster.
Herpes zoster (HZ) is caused by reactivation of varicella-zoster virus (VZV) that has been latent in sensory ganglia since primary VZV infection, usually in childhood. Since the frequency and severity of HZ increases with aging [1–5], and since VZV-specific T cell—mediated immunity (CMI) decreases with age, it was hypothesized that active immunization of older persons would boost their CMI against VZV and thereby decrease the frequency or severity (or both) of HZ in vaccinees [6].
Beginning in 1988, 202 persons ≥55 years of age were vaccinated with various doses of a live attenuated VZV vaccine. Vaccination was well tolerated, with no serious vaccine-related adverse events reported. Humoral and cell-mediated immune responses were measured after vaccination. Two years after vaccination, the vaccine was immunogenic, but no dose-response relationship was demonstrated [7]. However, 4 years after vaccination, there was some decline in VZV-specific immunity, and the magnitude of the decline was now a function of the immunizing dose [8]. Herein, we report the immunologic results for 130 of these vaccinees, who were followed for 6 years.
Methods
Population
The subjects were vaccinated ≥6 years before the current study with a live attenuated varicella vaccine. They were 55–87 years old at the time of vaccination, and all had a history of varicella but not HZ. Blood samples for VZV-specific testing were obtained yearly. Ten patients who had possible episodes of HZ during the study period were deleted from analysis subsequent to their clinical event. The distribution of vaccinees by age and dose group during each year of follow-up indicated that the nature of the cohort remained constant over time.
Vaccine
The OKA/Merck strain of varicella vaccine was supplied by Merck Research Laboratories. Vaccinees received either 3000, 6000, or 12,000 pfu, or 3000 pfu followed by the same dose 3 months later. A small subset of 10 individuals (55–59 years old) received only 1000 pfu. The vaccine contained 2–4 μg of VZV antigen/3000 pfu.
Frequency of peripheral blood mononuclear cells (PBMC) that proliferate in response to VZV antigen
The VZV antigen was prepared as previously described [7]. The limiting-dilution method that was used to determine the responder cell frequency (RCF) used 24 replicate cultures containing 100,000, 50,000, 25,000, or 12,500 ficoll-hypaque—separated PBMC per well. These were cultured with cell-free VZV antigen for 10 days and then pulsed for 8 h with 0.25 μCi of [3H]thymidine per well. Parallel control cultures were identical except that they were stimulated with a diluted control antigen prepared from uninfected cells. Responder wells were defined as those with greater than the mean counts per minute +3 SD of the 24 replicate parallel control cultures. RCF was interpolated (in a plot of the log of the percentage of nonresponder wells against the cell number per well) as the point at which 37% of VZV antigen—stimulated wells were nonresponders [9]. The readout is the number of PBMC required to detect a single VZV-responding lymphocyte. That is, a falling RCF indicates an increase in the prevalence of specific memory cells. RCF has good intra- and interassay variability. RCF is superior to results from tests based simply on quantification of the extent of proliferation and is more suitable than simple lymphocyte proliferation assays for long-term studies.
Evaluation of potential cases of HZ
Vaccinees were instructed to inform the investigators whenever they developed pain syndromes or skin lesions that they suspected represented HZ. Patients with suspected HZ were examined; if possible, the lesions were cultured for VZV, and scrapings were tested by VZV-specific direct immunofluorescence. Swabs of lesions were tested for VZV DNA by the polymerase chain reaction if reported after 1994. Acute- and convalescent-phase VZV-specific immunologic tests were done at the initial visit and 4–6 weeks later. Pain was subjectively reported as mild, moderate, or severe. Patients were followed until all symptoms disappeared.
Statistics
The data were plotted using an average RCF status (± 1 SE of the mean) for all subjects at each time point. To establish a presumptive curve representing the expected RCF of our vaccinees if they had not been vaccinated (for curve A in figure 1), we performed a regression analysis of RCF values of our vaccinees at baseline regressed on their age at baseline: RCF = −311.38 + 9.80 (age) − 0.06 (age)2. A series of analyses-of-variance models were used to test which of the descriptive variables were cross-sectionally related to RCF by month on study. Further, repeated measures analysis-of-variance techniques were used to compare covariables at 4 and 6 years. The specific covariables studied were sex, age, dose group, and RCF (coded as responder [> 1 responding cell/100,000 PBMC] vs. nonresponder [< 1 responding cell/100,000 PBMC]) at baseline. A nonlinear stochastic parameter model was fit in order to assess the duration of immunity in the study population [10, 11].
Figure 1
Mean (±SE) of VZV responder cell frequency (RCF) in peripheral blood mononuclear cells (PBMC) from persons ≥55 years old who received live attenuated varicella vaccine, compared with expected RCF of unvaccinated cohort of same age. Each point represents no. of PBMC that must be stimulated to detect 1 VZV-specific responder T cell. Curve A represents expected RCF if subjects had not been vaccinated and was calculated from distribution of RCF values of vaccinees at baseline, using formula: RCF = −311.38 + 9.80 (age) − 0.06 (age)2. Curve B represents mean baseline RCF of all vaccinees. Curve C represents RCF of vaccinees at times after vaccination. Frequency at time 0 was significantly less (P < .05) than at all other time points. No. of observations at various time points: 0 months, 188; 3 months, 190; 12 months, 186; 24 months, 177; 36 months, 166; 48 months, 160; 60 months, 153; and 72 months, 124.
Results
Samples obtained from 130 vaccinees 6 years after vaccination were analyzed. To avoid clinical events that might have produced additional immunologic boosting, we excluded patients who reported that they might have had HZ. The loss of patients from all causes averaged 7% per year. The age-group and dose-group distribution of this remaining cohort is similar to that of the original 202 vaccinees.
The number of T cells in the peripheral blood that proliferated in response to VZV antigen (RCF) is shown for the 6-year period following vaccination (figure 1, curve C). The mean RCF was 1/70,000 circulating cells prior to vaccination, and increased to 1/42,000 circulating cells 5 months after vaccination. This booster effect diminished thereafter, such that at 6 years it was minimal. The half-life of the booster effect based on a model of this curve is 56 months.
There are two ways in which the effect of the vaccination might be under-estimated. The first follows from a technical limitation of the RCF assay, which cannot reliably test > 100,000 cells per well. Thus, all frequencies of responding cells of < 1/100,000 are listed as < 1/100,000, even though a boost might have occurred above this range (e.g., from 1/300,000 to 1/110,000). The second way the effect could be underestimated is to consider that the “booster effect” is not merely the difference between pre- and postvaccination RCF values (difference between curves B and C in figure 1). This definition ignores the predictable loss of responding cells that occurs with aging [12] and which would have been apparent had there been a control group. Thus, we used the original baseline (prevaccination RCF data from a large number of persons between ages 60 and 80) to calculate age-specific RCF values in this age range and constructed a theoretical curve of RCF values that would represent our subjects had they aged without being vaccinated (figure 1, curve A). This representation, which may give a more accurate description of the immunologic effect achieved by vaccination, indicates a continuing enhancement of RCF in vaccinees (difference between curves A and C in figure 1) 6 years after vaccination.
Cross-sectional analysis of variance results indicated that none of the descriptive variables (gender, age, or RCF at the time of vaccination or dose) were significantly related to RCF status 6 years later, and a low R2 value suggested that none of these variables is likely to prove useful in predicting RCF status. The year-specific analysis of variance showed that sex and age at time of vaccination did not influence RCF 4 or 6 years after vaccination, while RCF at baseline and dose each influenced RCF 4 years after vaccination (.0007 and .0481, respectively), although their effect was not apparent at year 6.
At the time of the baseline assessment 6 years ago, 39% of the vaccinees had no detectable VZV-responding cells in our assay. Following vaccination, the percentage of nonresponding vaccinees fell to 10%–12% during the first 3 years after vaccination, was 20.7% at year 4, and rose to 37.7% for the present analysis (figure 2). This was true whether nonresponder status was defined by < 1/80,000 PBMC or < 1/100,000 PBMC. Using a cross-sectional logistic regression, the only significant predictor of nonresponse (defined as RCF < 1/100,000 PBMC) at 6 years was RCF at 5 years. Age was not a predictor of nonresponse in our patient population.
Figure 2
Presence of ≥1 VZV responder cell/100,000 peripheral blood mononuclear cells (responder) from persons ≥55 years old who received live attenuated varicella vaccine at different times after vaccination. No. of observations at various time points: 0 months, 188; 3 months, 190; 12 months, 186; 24 months, 177; 36 months, 166; 48 months, 160; 60 months, 153; and 72 months, 124.
A passive reporting system consisting of annual questioning and reminders to vaccinees resulted in 10 events being reported to us as potential episodes of HZ (table 1). Some events actually may not have been the result of VZV reactivation. The frequency of these events was 1/100 patient-years, similar to that predicted from natural history studies. In three instances, VZV was isolated or detected by polymerase chain reaction. In some others, a comparison of acute and convalescent RCF assays suggested a recent VZV infection. In 4 of the culture-negative cases, the results of IFA tests of lesion scrapings were negative. In all reported events, the number of lesions was small and the pain was usually mild and of short duration. There was no obvious relationship to age, interval since vaccination, dose, or last RCF in the year prior to suspected HZ. Some patients had a suspected HZ event when their last RCF determination showed a high frequency of VZV-responding cells. A single VZV isolate was available for typing, and it was found to be wild-type VZV.
Table 1
Possible herpes zoster in persons ≥55 years old who were immunized with a live attenuated varicella vaccine.
Conclusions
At least one measure of VZV-specific CMI can be enhanced in older persons by the administration of a live attenuated VZV vaccine. This “booster effect” was observed over a 6-year follow-up period. The effect diminished over time, with the half-life of the effect being 56 months. However, considering the ongoing loss of VZV-specific CMI with aging, the vaccine-induced effect was still significant 6 years after vaccination. The magnitude of the booster effect at 6 years was not predicted by the age or gender of the vaccinee or by the dose administered. The baseline immunologic state of the vaccinee predicted response at 4 but not 6 years after vaccination. Furthermore, vaccination decreased the number of individuals who had no detectable VZV-specific CMI by the assay we used. This percentage, which was 39% prior to vaccination, declined to 10%–12% for the first 3 years after vaccination but then approached the baseline proportion for the 6-year analysis.
The use of the RCF assay to measure VZV-specific CMI is an arbitrary choice. We suspect that the RCF assay is a surrogate marker for the Th1 type responses that we hypothesize will limit reactivation of VZV [13]. The vaccine used is known to enhance VZV-specific class I cytotoxic T lymphocyte activity in older persons [14]. In fact, the value of RCF as a predictor of clinical protection is unknown. Our follow-up system to detect HZ in vaccinees represents ∼1050 postvaccination patient-years. This is a small number, and interpretation is complicated by the difficulty in determining which clinical syndromes being reported were truly VZV related. However, given these reservations, it would appear that the incidence of HZ in this age group was not reduced by vaccination and that the episodes reported as HZ were atypically mild and not followed by post-herpetic neuralgia. Some of the breakthrough cases of HZ occurred in vaccinees who had VZV-specific responding cells in the year prior to presumptive HZ, which raises questions about the value of RCF as a marker for protection, although the RCF in the period just prior to reactivation remains unknown.
The strategy of active immunization of older people to prevent HZ can be adequately tested only in a very large, placebo-controlled trial. The immunogen used for this trial should reflect the observations made in this report, namely, that the booster effect with the current vaccine has a half-life of < 5 years and that a large proportion of vaccinees lose a detectable booster effect after 6 years. This latter observation may be especially important since it is possible that individuals who develop HZ come from a cohort at high risk as defined by the absence of detectable CMI.
Footnotes
-
Informed consent was obtained from the patients in accordance with human experimentation guidelines of the US Department of Health and Human Services and those of the University of Colorado Combined Investigative Review Board and were followed in the conduct of the clinical research.
-
Financial support: Merck Research Laboratories, West Point, Pennsylvania.
- © 1998 by the Infectious Diseases Society of America
References
This Article
- AbstractFree
- » Full Text (HTML)Free
- Full Text (PDF)Free
Services
Citing Articles
Google Scholar
PubMed
Related Content
Share
- Email this article
-
CiteULike
Connotea
Delicious
Facebook
Google+
Mendeley
Twitter
Published on behalf of
IDSA Members: For your free access to this journal, log in via the IDSA members area.
The Journal
Impact Factor: 6.288
5-Yr impact factor: 6.103
For The Media
For Authors
- Services for authors
- Self-archiving policy
- IDSA Conflict of Interest policy
- Instructions for Authors
- ICMJE Form
- Submit Now!
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy
Corporate Services
Alerting Services
Related IDSA Publications
