Dose-Dependent Neutralizing-Antibody Responses to Vaccinia

Understanding the immune responses to vaccinia and other smallpox vaccines will be important in the development of strategies to vaccinate large numbers of persons by use of new formulations of vaccines and to develop less reactogenic smallpox vaccines. We recently completed 2 dose-response studies of vaccinia (Dryvax; Wyeth) in persons 18–32 years of age who had never previously been vaccinated [1, 2]. In a pilot study, we evaluated vaccinia concentrations of 10^7.8, 10^6.5, and 10^5.0 pfu/mL, given by bifurcated needle (measured vaccinia dose of 10^5.0–10^1.9 on a needle), for vaccinia infection as indicated by vesicle formation and humoral and cellular immune responses [1]. Dose-response effects were evident for cytotoxicity and interferon-γ ELISPOT responses, with the highest concentration of vaccinia inducing greater cell-mediated immune responses. In a larger trial to assess equivalence of major (as defined by Centers for Disease Control and Prevention [3]) cutaneous reaction rates after vaccinia given at 10^8.0 (undiluted), 10^7.5 (1:5 dilution), or 10^7.1 (1:10 dilution) pfu/mL, significant differences in reactogenicity of vaccine included local inflammation (greatest in the undiluted-vaccinia group) and satellite-lesion formation (more frequent in the diluted-vaccinia groups) [2]. We speculated that cell-mediated immune responses, which were greater in the undiluted-vaccine group, were responsible for controlling local viral replication and for preventing satellite-lesion formation

To further our understanding of humoral immune responses after vaccination with smallpox vaccine at different concentrations and to more completely understand possible biologic effects of various doses, we assessed functional neutralizing antibody responses to vaccinia. In the present investigation, we tested, both before vaccination and 28 and 56 days after vaccination, all available serum samples obtained from vaccinated subjects (N=674). Our purpose was to assess the magnitude of neutralizing-antibody responses to different doses of vaccine and to correlate clinical responses with neutralizing-antibody response

Subjects, materials, and methodsHealthy adults 18–32 years of age were selected for absence of a vaccinia scar, no history of vaccinia vaccination, and absence of anti-HIV antibody. Exclusion criteria included vaccination contraindications (i.e., pregnancy, immunosuppression, and eczema) noted in the package insert. In addition, protocol-specified exclusions included history of either vaccination with a live attenuated virus ⩽60 days before the study or receipt of blood products or immunoglobulin ⩽6 months before the study; household contact, sexual contact, or occupational exposure to someone with ⩾1 of the exclusion criteria listed in the package insert; or contact with infants <12 months of age

The study was a randomized, single-blinded trial conducted at the National Institute of Allergy and Infectious Diseases’ Vaccine Treatment and Evaluation Units in St. Louis, Baltimore, and Rochester and at the Respiratory Pathogens Research Unit in Houston, after approvals had been given by the local Institutional Review Boards. Informed consent was obtained from all subjects in this clinical trial, and human-studies guidelines of the US Department of Health and Human Services were followed. All subjects were enrolled during November 2001. A total of 680 subjects were randomly assigned to 1 of the 3 following vaccine-dilution groups: undiluted vaccinia, 1:5 dilution of vaccinia, and 1:10 dilution of vaccinia. The subjects were vaccinated by 15 skin punctures by bifurcated needles, as described elsewhere [2]. Clinical results have been reported elsewhere [2]

In all subjects from whom serum samples were obtained, neutralizing antibody titers were determined by use of 60%-plaque-reduction endpoints modified from an assay described elsewhere [4]. In brief, serum dilutions were incubated with 20–40 pfu of vaccinia for 15 h and subsequently were plated onto tissue-culture wells. The interpolated titer resulting in a 60% reduction in plaque formation was taken as the endpoint. Values were measured for dilutions of 1:4–1:5000. A titer of <1:20 was considered to be seronegative, on the basis of repeated measures of standard antibody-negative serum samples. The 15-h incubation assay resulted in antibody titers ∼400-fold higher than those measured in the 1-h incubation assay [4]. In previous studies, in all subjects with a major cutaneous reaction, the 15-h incubation method was both 100% sensitive and 100% specific at detecting antibody increases after vaccination; all subjects without a major cutaneous reaction did not show antibody increases [1, 4]

After the initial vaccination, analyses were performed separately for subjects with a major cutaneous reaction and for subjects without a major cutaneous reaction. For each vaccine-dilution group and each time point, serostatus (“positive” was defined as a 60%-plaque-reduction neutralizing-antibody titer ⩾1:20) and geometric mean titers (GMTs) were summarized. Pairwise comparisons of the GMTs were made between vaccine-dilution groups, by use of a linear regression model (the “GLM” procedure in SAS). When it was necessary to do so, the baseline titer level (at day 0) was controlled by being made a covariate in the model. P values were not adjusted for multiple comparisons. The linear regression model was also employed to examine the association between neutralizing-antibody levels and local/systemic reactions. The model controlled for vaccine-dilution group and baseline neutralizing-antibody level. The fitness of regression models was assessed by the F test of model significance, and their strength of association was assessed by use of R². The nature of the association is explained in terms of n-fold increase, in neutralizing-antibody titer, per unit change in the underlying parameter

ResultsThe demographic characteristics of the 665 adults with a major cutaneous reaction after dose 1 of vaccinia are summarized in table 1. There were no significant differences in either sex, race, or age

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 1

Examples of a typical vaccinia pustule with a pronounced inflammatory response (top) a pustule with a satellite lesion (middle) and a pustule with a satellite lesion and relatively less local inflammation (bottom)

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Demographic characteristics of 665 adults with a major cutaneous reaction after dose 1 of vaccinia, by vaccine dosage

Antibody status both before vaccination (day 0) and on days 28 and 56 after vaccination, for all subjects with a major cutaneous reaction and from whom serum samples were obtained, are shown in table 2. Before vaccination, only 24 subjects (3.6%) were seropositive for neutralizing antibody, whereas 640 (96.2%) were seronegative. On day 28 after vaccination, all (100%) of 663 subjects tested manifested neutralizing antibody; on day 56, all but 1 (i.e., >99%) of 661 subjects tested manifested neutralizing antibody

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 2

Antibody status of vaccinated subjects with a major cutaneous reaction, stratified by vaccine dosage, before vaccination and at days 28 and 56 after vaccination

GMTs of neutralizing-antibody titers in serum, for each time point, are shown in table 3. In all of the vaccine-dilution groups, GMTs were <20 before vaccination, and neutralizing-antibody titers increased between day 0 and either day 28 or day 56; however, at day 28 the subjects given either dilution of vaccinia had significantly higher neutralizing-antibody titers than did subjects given undiluted vaccine (P=.008 and P=.003, for the subjects given undiluted vaccine vs. those given either a 1:5 or 1:10 dilution of vaccine, respectively, and these statistically significant differences persisted through day 56 [table 3]). Neutralizing antibodies had begun to wane by day 56, and their titers had fallen to 1008 and 1026 in the 1:5 and 1:10 vaccine-dilution groups, respectively, and to 796 in the undiluted-vaccine group. No significant difference in neutralizing-antibody titer was found when the results for the 1:5 vaccine-dilution group were compared with those for the 1:10 vaccine-dilution group, on either day 28 or day 56

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 3

Geometric mean titer (GMT) of neutralizing antibodies in serum from subjects with a major cutaneous reaction after the first vaccination

Previously, we had assessed the common, generally mild adverse events associated with vaccinia [1, 2]. To evaluate the possible correlation between neutralizing-antibody response and adverse events, we used linear regression models to evaluate neutralizing-antibody titers on days 28 and 56, as well as skin-lesion diameter, erythema, induration, satellite-lesion formation, lymphadenopathy, and fever (table 4). Higher neutralizing-antibody titers at days 28 and 56 were significantly associated with larger skin lesions, erythema, and fever

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 4

Association between after-vaccination events (parameters) and neutralizing-antibody titers in serum, at days 28 and 56 after vaccination

A total of 15 subjects did not have a major cutaneous reaction after dose 1, and 14 of these 15 subjects were revaccinated, on or near day 7, with the same vaccine dilution as had been given in dose 1. Preexisting neutralizing antibody was present in 6 of the 15 subjects, suggesting that they had received prior unrecognized vaccination. When they were revaccinated, 7 of the 15 subjects, including 3 of the subjects who initially had been seropositive, developed a skin vesicle characteristic of vaccinia vaccination. Of the 7 subjects without a major cutaneous reaction after 2 vaccinations, 3 had preexisting neutralizing antibody (1 of these 3 had a >4-fold increase in neutralizing-antibody titer, and the other 2 had no increase); and 3 of the other 4 subjects without a major cutaneous reaction developed a >4-fold increase in neutralizing-antibody titer. In total, 4 of the 7 subjects without clinical evidence of a vesicular skin lesion had an increase in neutralizing-antibody titers (GMT, 169 [range, 13–788] on day 28; GMT, 105 [range, 11–722] on day 56)

DiscussionPreviously, we had reported that there is a dose-response effect for cell-mediated immune responses, with cytotoxic T cell responses being greatest in undiluted-vaccinia recipients versus diluted-vaccinia recipients [1]; subjects given undiluted vaccinia had significantly greater local inflammatory response (figure 1A) and significantly fewer satellite lesions than did subjects given diluted vaccinia [2] (figure 1B and 1C). These observations are consistent with the observation reported here—that is, that significantly higher neutralizing-antibody titers are generated in subjects who receive diluted vaccinia. The significant increase in the frequency of satellite-lesion formation suggests increased local viral replication, which would result in more antigen production, which in turn stimulates higher neutralizing-antibody titers. The increased inflammatory response seen in undiluted-vaccinia recipients may reduce local viral replication in some of them—and therefore reduce the occurrence of local satellite lesions. Linear regression modeling (controlling for dosage and the level of preexisting neutralizing antibody) revealed that lesion size and fever were significantly associated with greater neutralizing-antibody responses; lesion size may also correlate with increased viral replication. The magnitude of these effects was a 1.5- and 1.9-fold increase in neutralizing antibody for each centimeter increase in lesion size, at days 28 and 56, respectively, and a 1.4-fold increase in the presence of fever, at both day 28 and day 56

Local inflammatory response, with redness and induration, is believed to be related to cellular immune responses, which were more vigorous in the undiluted-vaccinia group. Children who had either defects in cellular immunity or severe combined immune deficiency and who previously had been vaccinated occasionally developed progressive local ulceration leading to either loss of limb or death [5–13]. Neutralizing antibody is believed to have controlled the dissemination of vaccinia to the skin, the central nervous system, or other organs, but cellular responses (now believed to be CD8 plus cytotoxic T lymphocytes) were necessary to control local viral replication. The observations in this report and in our previous studies are consistent with these clinical observations and with the pathogenesis proposed

Of the 7 subjects without a primary vesicle after having been given 2 doses of vaccine, 4 developed neutralizing antibody; clearly, viral replication may occur in the absence of clinically apparent skin vesicles, but these cases represent <1% of vaccinated naive subjects. A review of photographs of the skin of the subjects without a major cutaneous reaction confirms the absence of vesicle formation. After 15 skin punctures with a bifurcated needle, a 2–3-mm scab was present in many of the subjects by day 7, and viral replication may have been present in these small lesions

In the present study, all subjects with a major reaction developed neutralizing antibody, and elsewhere it has been shown that nearly all subjects tested develop vigorous cell-mediated immune responses [1]. The neutralizing-antibody titers reported in the present study, in conjunction with the vigor of the CTL responses reported elsewhere [1], provide a standard by which to compare responses to other vaccinia vaccines such as tissue culture–produced vaccines and modified vaccinia Ankara, as these other products enter clinical evaluation

• Received March 3, 2003.
• Accepted August 5, 2003.

• © 2004 by the Infectious Diseases Society of America