Vaccine Clinical Trials

Similar to clinical development of drug products, there are four phases of clinical trials in vaccine development. Phase I trials are referred to early studies with human subjects. The purpose of phase I trials is to explore the safety and immunogenicity of multiple dose levels of the vaccine under investigation. Phase I trials are usually of a small scale. Phase II trials are to assess the safety, immunogenicity, early efficacy of selected doses of the vaccine, and generate hypotheses for later testing. Phase III trials, which are usually large in scale, are to confirm the efficacy of the vaccine in the target population and/or proving consistency of manufacturing processes. Phase IV trials are usually conducted for collecting additional information regarding long-term safety, immunogenicity, or efficacy of the vaccine to fulfill with regulatory requirement and/or marketing objectives after regulatory approval of the vaccine.

In this section, we provide some design considerations, including special statistical considerations, when conducting vaccine clinical trials. Also included are the classification of types of immunogenicity vaccine trials and statistical methods for various study endpoints.

7.8.1 Basic Design and Statistical Considerations

In its guidance, the FDA discusses the design and statistical considerations for clinical studies to demonstrate the safety, immunogenicity, and efficacy of vaccines (FDA, 1997). These design and statistical considerations are briefly described below.

Safety Studies As indicated in the FDA guidance, studies of comparative safety should be randomized and controlled. Follow-up safety should be actively monitored and pro-spectively planned with baseline and specific post-vaccination time of assessment. It is suggested that subjects should be actively monitored on designated post-vaccination days for up to one week for most killed and recombinant vaccines or for 14 or more days for most live vaccines. Follow-up should be continue through at least 30 days for live or killed vaccines. Thus, as an example, for a particular killed vaccine, subjects should be monitored at 6-12 hours, and days 1, 2, 3, 7, and 30 post-vaccination.

Immunogenicity As the objective of comparative immunogenicity studies is to rule out the important difference between the responses to the study vaccine and a control. Such studies should have sufficient power to rule out clinically meaningful differences in geometric mean titers (GMTs) and/or seroconversion rates. In addition, the study design should take into account the intrinsic variability in assays and subjects. A clinically meaningful difference for each response should be defined prospectively in the clinical protocol.

Efficacy Clinical trials for demonstration of vaccine efficacy should be randomized and controlled. Endpoints used to evaluate efficacy could range from disease incidence to a well-established surrogate marker with activity in correlate of protection. A correlate of protection in vaccine efficacy is generally a laboratory parameter that has been shown from adequate and well-controlled trials to be associated with protection from clinical disease. As indicated in the FDA guidance, an immunological correlate of protection is most useful if clear qualitative and quantitative relationships can be determined, e.g., a certain type and level of antibody correlate with protection.

Statistical Considerations In its guidance, the FDA emphasizes the importance of randomization in vaccine clinical trials. Stratified randomization may be recommended if warranted by the inclusion criteria or known disease risk factors. For nonrandomized trials, the FDA requires the validity and reliability of the evaluation of the study vaccine be provided.

For statistical approaches for data analysis, the FDA indicates that both hypotheses testing and the confidence interval approach are appropriate. The one-sided test for superiority is not an issue for regulatory approval. However, it is suggested that a difference-detection trial be designed for demonstration of superiority of the study vaccine as compared to a control.

For study endpoints, the FDA requires that the evaluation of the study vaccine be performed to rule out a prespecified, clinically meaningful difference between the study vaccine and the control and that should be clearly stated in the study protocol. It is also suggested that the assessment of immune response, common adverse reactions, less common adverse reactions, and rare adverse events should be carefully carried out according to specific statistical approaches as described in the guidance.

For sample size, the FDA requires that sample size calculation for each study endpoint (immunogenicity, safety, and efficacy if applicable) should be performed and the largest sample size should be selected as the one for overall trial enrollment. The FDA indicates that sample size calculation may be performed based on confidence interval rather than on power analysis if it is the planned analysis.

Ellenberg (2001) also discussed some statistical considerations in evaluating the safety of combination vaccines. In their review article, Chan et al. (2003) further posted some statistical issues when analyzing data from vaccine clinical trials. These statistical issues include intention-to-treat population versus evaluable subset, missing/coarse data and lost-to-follow-up, analysis with stratification variables, and interim analysis. In addition, Chan et al. (2003) suggested that some special considerations, including surrogate markers of activity, the assessment of immune responses, specificity in endpoint definitions of clinical efficacy based on exposure, confidence in safety, and special health economic considerations of the public health impact of vaccination programs be taken into account when planning a vaccine clinical trial.

7.8.2 Types of Vaccine Immunogenicity Trials

Vaccine immunogenicity trials are used to study the immune response to vaccination, which are usually measured by serum antibody concentration or T-cell responses. In early phases of vaccine development, immunogenicity trials are commonly conducted to assess whether the vaccine can induce immunity before entering large efficacy trials. In addition, immuno-genicity is often used to determine whether an immune marker to the vaccine can be used as a surrogate or correlate of disease protection. Chan et al. (2003) classified vaccine immunogenicity trials into the following categories of (1) superiority immunogenicity study, (2) dose-response immunogenicity study, (3) consistency lots study, (4) bridging study, (5) combination or multivalent vaccine study, and (6) immunological persistence study, which are briefly outlined below.

Superiority Immunogenicity Study The objective of superiority immunogenicity studies is to assess the superiority of immunogenicity of the vaccine under study as compared to a placebo control. Superiority immunogenicity studies are often conducted in early phases of vaccine development, which can also be performed to claim superiority of one vaccine as compared to another from a different manufacturer with respect to immune responses.

Dose-Response Immunogenicity Study During the development of a new vaccine, dose-response immunogenicity studies are often conducted to assess the immunologic responses across different dose levels of the vaccine. A well-established dose response can help in determining the minimum effective dose and the safe dose. In addition, dose-response studies are useful in studying the kinetic of potency decay and in determining the release and end-expiry dose level and shelf-life of the vaccine.

Consistency Lots Study Before a vaccine can be licensed, the FDA requires that evidence of analytical consistency among at least five lots of vaccine and clinical consistency in at least three lots of vaccine from the same manufacturing process be provided. Frey et al. (1999) suggested that vaccines from three consistency lots and a control vaccine be used for a typical clinical consistency lots study.

Bridging Study When there are minor changes in manufacturing process, storage conditions, routes of administration, or dosing schedules after regulatory approval, the FDA requires a bridging study be conducted to demonstrate that such changes do not have adverse effects on the vaccine effectiveness. An immunogenicity bridging study is usually designed as a noninferiority trial aimed to exclude a clinically significant difference in the immune response between the modified vaccine and the current vaccine.

Combination or Multivalent Vaccine Study A combination vaccine is defined as a vaccine that consists of two or more live organisms, inactivated organisms, or purified antigens combined either by the manufacturer or mixed immediately before administration (FDA, 1997). A combination vaccine is intended to prevent multiple diseases or to prevent one disease caused by different strains or serotypes of the same organism. For establishment of efficacy of a combination vaccine, the immunogenicity of all vaccine components in the combination or multivalent vaccine should be performed to rule out clinically significant differences in immune response rates and/or GMTs between the combined vaccine and the separate but simultaneously administered antigens. In addition, acceptable levels of immuno-genicity should be demonstrated for each serotype or component.

Immunological Persistence Study In practice, it is recognized that the duration of vaccine-induced immunity will be considerably longer than the time span of the clinical studies. Thus, immunological persistence study is often conducted to collect data regarding immune responses over multiple years. Life table or time-to-event data analysis is then used to assess the cumulative immunological persistence rate for determination of long-term immunological persistence.

7.8.3 Statistical Methods

As indicated in Chan et al. (2003), one of the most critical steps of evaluations of a new vaccine is to assess the protective efficacy of the vaccine against the target disease. An efficacy trial is often conducted to evaluate whether the vaccine can prevent the disease or reduce the incidence of the disease in the target population. For immunogenicity analysis, the immune response rate and the geometric mean titer or concentration of immune response post-vaccination are usually considered to assess vaccine immunogenicity. In what follows, statistical methods for analysis of these endpoints are briefly outlined.

Disease Incidence/Immune Response

Consider a vaccine clinical trial comparing a new vaccine with a control. Subjects who meet the inclusion/exclusion criteria are randomly assigned to receive either the test vaccine (T) or the placebo control (C). Let pT and pC be the true disease incidence rates or immune response rates of the nT vaccines and nC controls randomized in the trial, respectively. Thus, the relative reduction in disease incidence for subjects in the vaccine group as compared to the control groups is given by pc - Pt n =--

1 pC

In most vaccine clinical trials, nhas been widely used and is accepted as a primary measure of vaccine efficacy. Note that a vaccine is considered 100% efficacious (i.e., n = 1) if it prevents the disease completely (i.e., pT = 0). On the other hand, it has no efficacy (i.e., n = 0) if pT = pC. Let xT and xC be the number of observed diseases for treatment and control groups, respectively. It follows that the natural estimators for pT and pC

are given by

By Taylor's expansion and the Central Limit Theorem (CLT), ¡3 is asymptotically distributed as a normal random variable with mean ¡3 and variance given by

For a given confidence level of 1 - a, the 100 x (1 - a) confidence interval of ¡is given by

(fi - Z(a/2)<, ¡3 + Z(a/2)<), where < is obtained according to (7.8.1) by replacing pT andpc by pT andpPc, respectively. This leads to the following 100 x (1 - a) confidence interval for n :

(1 - exp(3 + Z(a/2)<3), 1 - exp(/3 - Z(a/2)<3)).

Extremely Low Disease Incidence In many cases, the disease incidence rate is extremely low. In this case, a much larger scale of study is required to demonstrate vaccine efficacy as described in the preceding subsection. For sufficiently large sample sizes and small incidence rates, the numbers of cases in the vaccine groups and the control groups may be approximated by independent. Poisson distribution with rate parameters XT (~ nTpT) and XC (~ nCpC), respectively. As a result, the number of cases in the vaccine group given the total number of cases (denoted by S) is distributed as a binomial random variable with parameter 0, i.e., b(S, 0), where

R + u 1 - n + u where u = nC/nT, because 0is a decreasing function in n, testing the hypotheses that

Hq: n < no vs. Ha: n > nQ is equivalent to testing the following hypotheses:

Let xT and xC be the number of the observed diseases for the treatment and control, respectively. A natural estimator for 6 is given by 6 = xT /(xT + xC). The test statistic is given by

Under the null hypothesis, T is asymptotically distributed as a standard normal random variable. Hence, we reject the null hypothesis at a level of significance if T > Z(a).

Relative Vaccine Efficacy In vaccine trials, when the control is a licensed vaccine (an active control), the relative efficacy n can be evaluated through the relative risk (i.e., R = PT/PC) based on the relationship n = 1— R. If the absolute efficacy of the control (i.e., nC) has been established, one can estimate the absolute efficacy of the test vaccine by nT = 1— R(1 — nC).

For a comparative vaccine trial, it is often designed as a noninferiority trial by testing the following hypotheses:

H0: R > R0 vs. Ha: R < R0, where R0 > 1 is a prespecified noninferiority margin or a threshold for relative risk. In practice, the hypotheses regarding relative risk are most often performed based on log-scale. In other words, instead of testing the above hypotheses, we usually consider the following hypotheses:

Ho:log(Ä) > log(R<) vs. Ha:log(Ä) < log(R<).

This becomes the two-sample problem for relative risk.

Composite Efficacy Measure As indicated by Chang et al. (1994), in addition to the prevention of the disease infection, a test vaccine may also reduce the severity of the target disease as well. As a result, it is suggested that a composite efficacy measure be considered to account for both incidence and severity of the disease when evaluating the efficacy of the test vaccine. Chang et al. (1994) proposed the so-called burden-of-illness composite efficacy measure.

Suppose nT subjects were assigned to receive treatment while nC subjects were assigned to receive control (placebo). Let xT and xC be the number of cases observed in the treatment and the control group, respectively. Without loss of generality, we assume the first xT subjects in the treatment group and xC subjects in the control group experienced the events. Let s,j, i = T, C; j = 1,..., xi be the severity score associated with the jth case in the ith treatment group. For a fixed i = T or C, it is assumed that sj are independent and identically distributed random variables with mean ^ and variance c2. Let pi be the true event rate of the ith treatment group. The hypotheses of interest is given by

The test statistic is given by

Under the null hypothesis, Chang et al., 1994 showed that T is asymptotically distributed as a standard normal random variable. Hence, we would reject the null hypothesis if \T\ > Z(a/2).

It should be noted that safety considerations of vaccines are different from that of pharmaceutical products. The methods and measurements chosen for assessment of safety of a vaccine depend on a number of factors, such as the type of vaccine and its specific mechanism for eliciting immune responses. As vaccination can cause allergic or anaphylactic reactions due to the induction of the immune system, it is suggested that all safety variables encountered in a vaccine clinical trial and the analytical approach should be specified in the study protocol. As vaccines will typically be administered to millions of otherwise healthy individuals, it is strongly recommended that rare but serious adverse events be carefully evaluated.

In addition to the statistical methods for analysis of vaccine efficacy and immunogenic-ity endpoints described above, Durham et al. (1998) considered a nonparametric survival method to estimate the long-term efficacy of a cholera vaccine in the presence of warning protection. For evaluation of long-term vaccine efficacy, as indicated by Chan et al. (2003), the analysis of time-to-event may be useful for determining whether breakthrough rates among vaccinees change over time. It, however, should be noted that sample size calculation may be different depending on the study objectives, the hypotheses of interest, and the corresponding appropriate statistical tests.

Clinical development for vaccine has recently received much attention both from regulatory agencies such as the US. FDA and the pharmaceutical industry. For example, Ellenberg and Dixon (1994) discussed some important statistical issues of vaccine trials (related to HIV vaccine trials). O'Neill (1988b) and Chan and Bohidar (1998) gave asymptotic and exact formulas for sample size and power calculations for vaccine efficacy studies, respectively. Chan et al. (2003) provided a comprehensive review of vaccine clinical trials and statistical issues that are commonly encountered in vaccine clinical trials.

How To Bolster Your Immune System

How To Bolster Your Immune System

All Natural Immune Boosters Proven To Fight Infection, Disease And More. Discover A Natural, Safe Effective Way To Boost Your Immune System Using Ingredients From Your Kitchen Cupboard. The only common sense, no holds barred guide to hit the market today no gimmicks, no pills, just old fashioned common sense remedies to cure colds, influenza, viral infections and more.

Get My Free Audio Book


Responses

  • Bellisima
    How to design vaccine immunogenicity bridging trials?
    2 months ago

Post a comment