Measuring AIDS vaccine efficacy: intermediate versus full-scale trials
The vast majority of preventive AIDS vaccine trials to date have been Phase I studies that enroll small numbers of volunteers and primarily measure the safety and immunogenicity of a vaccine .candidate. Studies with small numbers of volunteers cannot provide any information about whether or not the vaccine candidate prevents HIV infection or disease. This information can only be gathered in large-scale “efficacy” trials. Traditional efficacy trials are called Phase III trials. The three completed and ongoing Phase III efficacy trials of preventive AIDS vaccines all enrolled thousands of people and cost hundreds of millions of dollars.
There are now several AIDS vaccine candidates completing early phases of testing and moving towards large-scale efficacy trials. This is a promising development, but it also raises new challenges since many sponsors feel that it may not be possible to test every AIDS vaccine candidate in a Phase III trial.
One reason for this is that the field has limited financial and human resources for clinical trials. Time is another factor. It can take two or more years to prepare for such a large-scale trial and then up to five years to get an answer about vaccine efficacy.
Given the urgent need for an AIDS vaccine, sponsors are considering testing some candidates in intermediate-size efficacy trials, also known as Phase IIb trials. These studies are smaller and less expensive than Phase III trials and could still provide some preliminary indication of a candidate’s efficacy.
Intermediate-size trials have been used to test many other types of medicines and vaccines, including cancer drugs and, recently, an experimental vaccine against human papillomavirus (HPV) which causes genital warts and cervical cancer. However Phase IIbs have not yet been used to test AIDS vaccine candidates.
This could change in the next few years. The International AIDS Vaccine Initiative, the HIV Vaccine Trials Network and Merck and Co. are all currently considering Phase IIb trials. As plans for these trials advance it will be important for communities and AIDS vaccine advocates to understand the strengths and limitations of this approach to evaluating AIDS vaccine candidates.
Comparing Phase IIb and Phase III trials
Phase IIb and Phase III trials take the same overall approach to measuring vaccine efficacy. Both trials divide volunteers into two groups: volunteers in one group receive the experimental vaccine, and volunteers in the other group receive an inactive substance called a placebo. Neither the trial staff nor the volunteers know who has been assigned to receive the vaccine or the placebo until the study is over. This is called a double-blinded, placebo-controlled study.
All volunteers in both types of trials are regularly tested for HIV and receive condoms and risk reduction counseling which emphasizes that volunteers should not assume that they have received, or are protected by, the experimental vaccine. However some volunteers still become infected with HIV despite these services. It is important to remember that the vaccine cannot cause HIV and that no volunteers in these trials are ever intentionally exposed to HIV.
The number of volunteers and the duration of both Phase IIb and Phase III trials are determined by the rate of HIV infections or “inci¬dence” in the community where the trial is going to take place. The higher the incidence, the fewer volunteers and/or shorter the follow-up period required. A Phase IIb trial would enroll fewer volunteers than a Phase III trial done in the same population. In general, Phase IIb trials are likely to be about half the size of a Phase III trial.
At the end of the study, researchers “unblind” the study, which means that they learn who received the vaccine and who received the placebo. They then look for evidence that the vaccine helped protect against HIV infection, or helped to reduce the severity of disease in people who became HIV-infected. (See Primer to learn more about the different types of AIDS vaccine efficacy.) To do this, researchers compare the number of new HIV infections in the vaccine and placebo groups. They also look at markers of HIV disease including viral load and CD4+ cell counts in volunteers who became infected with HIV. If differences are detected, statistical tests are performed to determine whether they are due to the vaccine or just a coincidence.
The main difference between Phase IIb and Phase III trials lies in the precision of the conclusions that can be drawn from a trial. A Phase III trial can make more accurate estimates of vaccine efficacy than a Phase IIb trial done in the same population. Phase III trials can also detect lower levels of efficacy than Phase IIb trials. This is because accuracy is directly related to the number of people studied in a trial. When there are more volunteers, there are likely to be more people who become infected through blood or sexual exposure. These infections are the key “endpoints” for an AIDS vaccine trial. The more endpoints there are, the more confident sponsors can be that a possible vaccine effect is real and not a coincidence.
Phase IIb trials are not as precise. A Phase IIb trial would only be able to tell if a vaccine candidate was very effective or not effective at all, and could not reliably detect moderate or low levels of efficacy. Instead, a Phase IIb trial might provide “inconclusive” data about a candidate with moderate efficacy, meaning that it wouldn’t be known for certain if it had any beneficial effects.
Why Phase IIb trials?
The risk of conducting a Phase IIb trial is that sponsors may end up with an inconclusive answer. The possible benefit is that sponsors may be able to find out relatively quickly whether or not a particular candidate shows signs of efficacy or has very high efficacy. This is sometimes called a “proof of concept” trial. Vaccine and drug developers sometimes test early versions of promising candidates in “proof of concept” trials before investing in design, testing and manufacturing of a final candidate in a large efficacy trial.
A Phase IIb trial that provided “proof of concept” for a particular candidate might also help suggest correlates of protection for preventive AIDS vaccines. A correlate of protection is an immune response that corresponds to a high degree of vaccine protection. For example, antibody against the hepatitis B virus is the correlate of protection for hepatitis B vaccine. Physicians can measure the level of anti-hepatitis B virus antibody in the blood of a vaccine recipient to confirm that he or she is likely to be protected from hepatitis B virus infection. This way, a vaccine recipient knows whether he or she is protected from hepatitis B virus without being exposed to the virus itself.
One of the major challenges in the AIDS vaccine field is that the correlates of protection are not well understood. At present, trial sponsors analyze the type and quantity of “vaccine-induced” immune responses in Phase I and II trial volunteers but it is not known for certain that the immune responses they are measuring will protect against HIV infection or disease.
Both Phase III and Phase IIb trials could help identify correlates of protection, but Phase IIb trials could potentially accelerate this process by providing rapid estimates of efficacy. Once a candidate shows efficacy, researchers can analyze immune responses to try to learn which immune responses are associated or “correlated” with vaccine protection. These correlates could then be used to help make decisions about whether or not future candidates should be tested in large-scale trials.
Phase IIb trials could also be used to gather information about partially effective vaccines (see Primer). The current generation of AIDS vaccines will be evaluated for their ability to reduce viral load and HIV disease in vaccine recipients who become HIV-infected. Scientists believe that a vaccine-induced reduction in viral load would be beneficial, but do not know how much the viral load would have to drop, or how long it would have to last, to have a health benefit for the volunteer.
Phase IIb trials could gather information on these types of questions, allowing the field to fine-tune its goals for partially effective candidates. These goals could then be used to shape the design of Phase III trials.
If Phase IIb trials move ahead, the AIDS vaccine field will have to do additional education and outreach to explain that some efficacy trials will be designed as information-gathering tools, and will not lead directly to a “license” for widespread use, even if that candidate appears effective. One reason for this is that sponsors may choose to conduct a Phase IIb trial of an earlier version of the candidate while they are developing manufacturing plans for their final product (see Primer, April 2004). In this case, another efficacy trial would be tested once the final product had been completed. Another reason is to gain more precise information, since Phase IIb trials usually provide a general idea of whether a candidate is effective or not.
Phase IIb trials are a new development in AIDS vaccine research and communities, researchers and sponsors will need to work together to find effective ways of explaining the contribution that these studies can make to the field.
All articles written by Emily Bass