Secondary Logo

Journal Logo


Candidate HIV/AIDS vaccines

lessons learned from the World's first phase III efficacy trials

Francis, Donald P; Heyward, William L; Popovic, Vladimir; Orozco-Cronin, Patti; Orelind, Karin; Gee, Carolyn; Hirsch, Adrian; Ippolito, Tina; Luck, Aimee; Longhi, Michael; Gulati, Vineeta; Winslow, Nathan; Gurwith, Marc; Sinangil, Faruk; Berman, Phillip W

Author Information
  • Free


In 2003 the results of the first two efficacy trials of candidate HIV-1 vaccines will be available. The first, the AIDSVAX B/B trial, was started in 1998 in the USA, Canada, Puerto Rico and the Netherlands. The trial will conclude in late 2002 with initial efficacy results available in the first quarter of 2003. The second, the AIDSVAX B/E trial, was started in Bangkok, Thailand in 1999 and will similarly have initial efficacy results available in late 2003.

Each study is unique. The North American/European trial will determine the efficacy of AIDSVAX B/B, a bivalent subtype B gp120 vaccine using MN and GNE8 antigens, in preventing sexual transmission of HIV-1 in geographical areas where subtype B HIV is most prevalent. For this trial, 5108 men who have sex with men (MSM) and 309 high-risk women have volunteered. The Thailand trial will determine the efficacy of AIDSVAX B/E, a bivalent gp120 vaccine containing the MN and A244 envelope antigens from subtype B and subtype E viruses respectively, against blood-borne infection (Fig. 1). This trial involves 2545 injecting drug users (IDU), mostly heroin addicts, in Bangkok where both subtype B and subtype E HIV are prevalent [1].

Fig. 1.
Fig. 1.:
Vial of AIDSVAX B/E vaccine.

The road from initial pre-clinical and clinical results to phase III efficacy trials has been long and difficult. The purpose of this paper is to review the lessons that have been learned through the experience of conducting the first phase III HIV vaccine efficacy trials.

The VaxGen HIV-1 preventive vaccine—AIDSVAX

The rationale for these recombinant gp120 envelope vaccines has been described in detail elsewhere [2]. Briefly, the design and manufacture of gp120 by recombinant DNA technology is similar to that of the licensed hepatitis B vaccine and the candidate herpesvirus type II vaccine. The evaluation of the safety and potential efficacy of these candidate HIV vaccines has, by necessity, also followed a development course similar to that of hepatitis B vaccine where the only susceptible animal model, for both viruses, is the chimpanzee. Pre-clinical studies of AIDSVAX showed impressive chimpanzee protection with minimal toxicity [3,4]. From chimpanzees, evaluation of the vaccine advanced to human studies. In these studies, essentially all recipients developed a robust immune response as judged by both binding and functional antibodies as well as skin-test positivity [1,5]. Evaluation of the safety of AIDSVAX has shown that both formulations have an acceptable safety profile with only mild injection site pain and tenderness being the most common adverse effect.

AIDSVAX B/B and B/E vaccines are composed of rgp120/HIV-1 recombinant envelope glycoproteins produced by expression of the fusion proteins in genetically engineered Chinese hamster ovary cell lines. Based on complementary characteristics, two subtype B antigens, MN and GNE8, were selected for the B/B vaccine intended to cover the majority of subtype B HIV strains present in North America and Europe. The MN antigen is derived from a T-cell tropic or syncytium-inducing B subtype of HIV-1 and requires the CXCR4 receptor to bind to CD4 cells. The GNE8 antigen is macrophage tropic, non-syncytia inducing and requires the CCR5 receptor to bind to CD4 cells. Based on analysis of amino acid sequence, MN and GNE8 differ by approximately 15% and thus should cover around 80% of the subtype B viruses prevalent in North America and Europe.

AIDSVAX B/E consists of the same MN antigen along with the A244 antigen which is derived from the CM244 subtype E strain of HIV. This latter virus is a primary macrophage-tropic, non-syncytia inducing and requires the CCR5 chemokine receptor to bind to CD4. Being from different HIV-1 subtypes, the amino acid sequences of these two antigens differ by 30% and should cover both subtypes B and E viruses which are prevalent in Thailand. The bivalent vaccines contain 300 μg of each antigen with 600 μg of alum adjuvant and are administered by intramuscular injection.

AIDSVAX B/B and AIDSVAX B/E are currently being tested for protective efficacy and safety in randomized, placebo-controlled, double-blinded phase III trials in North America and Europe (VAX004), and in Thailand (VAX003). A total of seven immunizations with either bivalent vaccine are administered at months 0, 1, 6, 12, 18, 24, and 30. In VAX004, 5417 volunteers in North America and Europe at high risk of contracting HIV-1 via sexual transmission will be immunized with AIDSVAX B/B or placebo (2 : 1 ratio). And in VAX003, 2545 volunteers in Thailand at high risk of contracting HIV-1 via injecting drug use will be immunized with AIDSVAX B/E or placebo (1 : 1 ratio).

These vaccines that are currently in phase III evaluation are derived from a long and arduous research and development path that started with the determination of the molecular structure of HIV by researchers at Genentech, Inc. in 1984 [6] (Fig. 2). Since that time, the antigens have been tested, rejected, restructured, and retested. Initial failed attempts to protect chimpanzees [7] resulted in the redesign of the HIV antigen selected for the vaccine altogether. Eventually, after showing that antibodies to gp120 inhibit binding of gp120 to CD4 cells [8], determination of the disulfide and carbohydrate structure of gp120 [9], and showing that monovalent gp120 makes neutralizing antibodies in humans [10], the effort proceeded. Multiple phase I/II clinical trials in humans were conducted to arrive at the optimal dosing and administration schedule. This development endeavor will have taken almost 20 years and has cost over $300 million dollars to complete. Although long and difficult, this process is not unlike that of many other vaccines that have been brought to licensure [11].

Fig. 2.
Fig. 2.:
Timeline for development of the phase III HIV vaccine trials.

Social value of vaccines

Although vaccines are well understood to be the key to controlling epidemic infectious diseases like AIDS, they do not receive the same social, political, or financial support as do therapeutic agents. This contrast is exemplified by a variety of measurements. Consider the immense international activism over the past decade in the USA and more recently in developing countries protesting the lack and cost of anti-HIV therapeutic agents. Multiple activist groups and protest actions have resulted in the establishment of large public funds to help develop these antivirals and, later, to purchase them. The fact that over 20 anti-HIV therapeutic drugs have completed phase III studies and have been licensed attests to the success that has followed this high social valuation. Contrast this result to the lack of social activism and funds established for an HIV/AIDS vaccine. The scientific challenge of developing vaccines is certainly no greater than that posed by developing antiviral agents. Indeed, prior to the HIV/AIDS epidemic, the search for antiviral agents for other viruses seemed overwhelming. In contrast, multiple effective viral vaccines had been developed and were widely used.

It has been different for AIDS. Since the beginning of the epidemic, a large number of antiviral therapeutic agents have entered phase III testing, and many have been subsequently licensed. In contrast, only one preventive HIV candidate vaccine, AIDSVAX, has thus far entered phase III testing. Why? We propose that this effect reflects the lack of social value currently given to vaccines in general and to HIV vaccines in particular—a difficult lesson for those attempting to develop them. Industry will follow social value. If society values a vaccine and there is a potential return on investment, industry can better accept the risk of the development costs. Like therapeutic drugs, hundreds of millions of dollars are required to bring a candidate vaccine through phase III testing. Without societal recognition of the impact of vaccines on disease prevention and the potential for profit incentives for industry, it has been very difficult to convince industry to invest the resources necessary to bring candidate vaccines forward.

Lesson: To stimulate industry to invest in preventive vaccines, society must recognize the value that they deserve. Health care and public health officials must more effectively address this societal need and policy makers must establish funds and mechanisms to retain and stimulate industry involvement in vaccine development.

Taking risk: government, activists and industry

In the early 1990s the United States government, through the National Institutes of Health (NIH), worked with industry and shared some of the expenses of HIV vaccine development. In 1994, two envelope-based recombinant vaccines had completed phase II testing and met the government's criteria to advance to phase III trials. But at the time, a constellation of forces influenced the NIH to alter the phase III advancement criteria and a decision was made not to sponsor the phase III trials. The decision was based on political, financial, and scientific grounds. Although both vaccines had protected chimpanzees from large intravenous challenges of HIV-1, there was concern that the vaccines may not protect humans. The NIH formed a scientific committee that reviewed all of the data and recommended advancement to phase III. In response, some vocal AIDS activists objected. Their concern was that government support of a large vaccine study would divert funds away from their priority of therapeutic drug research and development. In addition, some NIH-supported scientists were concerned that diversion of funds to support vaccine clinical trials would limit the money available for their basic research. The opposition of these two constituencies, together with the opinion of some that a better candidate vaccine, a canarypox vector ‘prime’ followed by a gp120 ‘boost', was ‘just around the corner', led the NIH with input from yet another committee to ultimately decide to not support the phase III trials [12]. This decision was a setback to AIDS vaccine development and left industry with the perception that the government was not a reliable development partner.

Lesson: If the government and its constituents truly desire to stimulate vaccine development, then they must form reliable partnerships that can withstand and confront the controversy and risks, both financial and political, which are required for the development of an HIV vaccine.

Paying for development: influence of ‘vaccine experts'

Without government funding, VaxGen had to look for alternative ways to raise the millions of dollars needed to complete the re-design of the vaccine and to conduct the phase III trials. In the absence of government or foundation support, VaxGen turned to private investors. Importantly, these investors, many seeing the potential profit, proved to be less adverse to risk. But potential investors, not having vaccine expertise, rely on academic researchers to help guide them in their investment decisions. On the surface, this is reasonable. Potential investors will logically turn to people with vaccine expertise before they invest in a vaccine venture. Unfortunately, the difference between academic research and industry biologic development is considerable and not always appreciated by the research community. The difference between the process required to generate the latest bench-top research experiment and the industry process required to produce a candidate biologic for Food and Drug Administration (FDA) approval are often extreme and unconnected. Specifically, the actual value of the latest bench-top research cannot often be determined for years since it must advance through the phases of product development before becoming a licensed product. Unfortunately, many of the researchers exploring these futuristic advances have no experience in the real process of vaccine development and, thus, do not understand the difference between research and development.

Over the last few years there have been many scientific or lay press reports of the latest discovery or experiment predicting a successful AIDS vaccine in the near future. Frequently, these have involved a new vaccine construct that has partially protected a few monkeys in one laboratory. Lacking vaccine development experience and oblivious to the impact that his or her statements may have on the thousands of volunteers in a phase III trial or to the investors who may be planning to put their money into the trial, the laboratory expert announces that their candidate ‘may be the best possibility’ to make an AIDS vaccine. Such public statements, although not intentional, can have substantial adverse effects to ongoing studies and financing. The comments of leaders in the vaccine field can be multiple. For example: (i) ‘chimpanzees are not a good model’ despite the fact that this primate is the only laboratory animal susceptible to HIV-1 and most likely the original source of HIV-1 infection in the wild; (ii) ‘antibodies induced by gp120 do not neutralize ‘primary isolates’ despite the fact that no validation of this assay as a predictor of protection has been provided. Indeed, sera from chimpanzees that were protected against a ‘primary isolate’ challenge could not neutralize HIV-1 in the primary isolate neutralization assay that was claimed to predict protection [12]; finally (iii) ‘recruitment of 5000 MSM volunteers will deplete the pool of volunteers for other trials', despite the fact that many more MSM are eligible for such studies [13].

Lesson: Industry, governments, and investors must choose advisors with broad expertise and experience and not be misled by fads or the research ‘flavor of the week'. Without solid guidance from a broad-based scientific group, free from conflicts of interest, future investors will be unable to judge the potential value of a product and appropriately support industry to develop an effective HIV vaccine. Making such decisions in the highly visible and politically heated field such as AIDS is especially challenging.

Regulatory review

Despite some adverse pronouncements of a few vaccine scientists, VaxGen was successful in obtaining several million dollars needed to re-engineer its bivalent vaccine and launch the phase III trials. A positive finding during this process was the interactions with the USA, Thai, Canadian and Dutch regulatory authorities. In contrast to the relative lack of vaccine development experience among academic scientists, the regulatory authorities and their advisory committee members were highly experienced and knowledgeable. The reviews by these authorities, most notably the US FDA and the Thai AIDS Technical Subcommittee, were helpful and the advice from these groups added to the quality of the phase III studies.

Lesson: Government regulatory authorities can be commended in the choice of their advisors and judicious review of trial protocols to advance HIV vaccine development. These organizations should maintain continued high standards in the future so that, despite the heat of the political moment, solid science- and ethic-based decisions can be made.

Embracing the unknown: viral variation, correlates of protection

Any vaccine progressing to phase III enters the unknown. One of the critical questions of HIV-1 vaccine development has been viral variation—more exactly, the immunologic importance of the genetic variation of the virus and the need for a match between the vaccine candidate and the viruses circulating in the population being entered into phase III trials. Here VaxGen has taken the conservative approach and has assumed that this match is important. Furthermore, we have assumed that both co-receptor preferences by HIV strains, CXCR4 and CCR5, should be represented. Thus, we redesigned the original monovalent MN subtype B gp120 vaccine and constructed two bivalent formulations, AIDSVAX B/B for the North American/European study and AIDSVAX B/E for the Thailand study [2].

Likewise, we do not know how vaccine protection may vary by exposure route, mucosal (sexual) or parenteral (injecting drug use), or what type or combination of immune response are necessary for vaccine protection. For this reason, we are conducting trials in the risk groups of MSM and IDU to determine if route of exposure makes a difference and evaluating their immune response by a variety of assays and measurements to determine the correlate(s) of protection. It is only through efficacy trial results that immune correlates of protection can be learned.

Lesson: All of the answers to important questions will never be known before advancing candidate vaccines into clinical trials, rather we should use trials to answer questions to advance vaccine development. If the need for a vaccine justifies the financial, personal, and scientific risk, one must move into uncharted waters and let the empiric data provide the information.

Vaccine trial cohorts

To successfully evaluate the efficacy of an HIV vaccine, thousands of individuals at risk for HIV infection are required who, despite advice to the contrary, will continue risk-taking behavior, albeit at a potentially lower level. One of the concerns expressed before VaxGen launched the two phase III studies was that the counseling provided during the study would eliminate risk-taking behavior altogether, or lower it to a degree that determination of efficacy was not possible. Here again, experienced counselors and social/behavioral scientists who had attempted to eliminate risk-taking behaviors in previous studies involving MSM and IDU, knew that risk-taking behaviors continue despite the best education and counseling measures available. Indeed, multiple longitudinal studies of both MSM and IDU clearly indicated the continued risk of these individuals in the face of behavioral counseling [14,15].

Another concern expressed before launching these studies was whether a vaccine trial was even feasible in these populations. Some wondered whether at-risk persons would volunteer for vaccine trials and, if they did, would they comply with the rigorous, multi-year schedule of such a trial. The concern here was reasonable given that the volunteers are expected to attend at least 17 clinic visits, receive seven vaccine or placebo injections, have at least 16 blood draws, and undergo multiple intensive education and counseling sessions and interviews over a 3-year trial period. Consider the ramifications of a major effort that was successful in enrolling thousands of volunteers only to have a high lost-to-follow-up rate render the data useless in determining vaccine efficacy. Fortunately, there were data available that gave us reassurance that such a negative outcome was unlikely. Early experience with the hepatitis B vaccine trials among MSM demonstrated that the recruitment of thousands of volunteers was possible and that follow-up rates were excellent [16,17]. More recent HIV-study experience showed that MSM could be recruited into cohorts and that the compliance was excellent. In Thailand, equally promising data had been collected regarding the feasibility of conducting vaccine trials among IDU. Early studies by the World Health Organization (WHO)/UNAIDS, the US Centers for Disease Control and Prevention (CDC), and the Bangkok Metropolitan Administration showed that Thai IDU had a high incidence of infection, that they could be followed over time, and that they would volunteer for a vaccine trial [15,18].

These studies provided solid estimates of the incidence of infection in the face of intensive education and counseling, willingness of high-risk persons to participate in vaccine trials, and the expected lost-to-follow-up rates. With this valuable information, the number of volunteers required for a study could be accurately estimated. Furthermore, the process of conducting these cohort studies served as excellent training and capacity building exercises for clinic staff in recruiting, counseling and following at-risk individuals—skills essential for the success of any subsequent phase III study.

Lesson: Conducting cohort studies well in advance of planned efficacy trials is essential. Those considering investing the large amounts of resources to conduct a multi-year trial need data to ensure that the trial is feasible. The data collected by WHO/UNAIDS, CDC and NIH-funded studies were very reassuring to those of us considering investing time and money into large efficacy trials. In addition, as cohorts are developed, staff expertise and research infrastructure are strengthened among a core group of researchers. This expertise proved to be key in establishing eligibility criteria, developing recruitment materials, clinic staff expertise, and general study conduct. Once a level of expertise was achieved, it was far easier to transfer that expertise as new clinics joined the studies.

Working with foreign governments

Essential for our success in Thailand was the strong national commitment and interest of the local government authorities and academic institutions toward vaccine research. Their willingness to advance the field of HIV vaccine development, especially in the face of wavering United States government commitment, was critical to VaxGen moving forward in Thailand. Initially, WHO and UNAIDS assisted Thailand with the development of a National Plan for HIV Vaccine Development which provided a roadmap and an organized framework through which appropriate candidate vaccines could be moved from research to licensure. To ensure that trials would be conducted with the highest scientific and ethical standards, expert committees were appointed along with systems to conduct ethical and technical reviews of proposed HIV vaccine trials. This approach was bolstered by parallel international reviews. Indeed, the AIDSVAX B/E trial was reviewed by no fewer than eight committees or institutions including the Technical Subcommittee for HIV Vaccine Development of the Thailand National AIDS Committee, the Ethical Review Committee of the Thailand Ministry of Public Health, the Bangkok Metropolitan Administration Ethical Review Committee, Mahidol University Ethical Review Committee, the UNAIDS Vaccine Advisory Committee, the US Centers for Disease Control and Prevention Institutional Review Board, the US Office for the Protection of Research Risks, and the US FDA. Although time consuming, this intensive review of the first international phase III trial was considered necessary by VaxGen to assure the international scientific community that this trial was being properly conducted.

Lesson: Government health authorities can, indeed, be a powerful influence and integral part of vaccine development in support of public health. The courage of those to advance vaccine trials in the face of controversy deserves applause. In addition, the assistance of an independent advisor, here provided by WHO/UNAIDS, helped local authorities establish review procedures of the highest international standards. Although arduous, the intensive international review and approval process has ensured that the Thailand trial is conducted under the highest scientific and ethical standards.

Study design

A successful HIV vaccine is one that either prevents infection or, if not, modifies disease if infection occurs. For the primary endpoint, prevention of infection, one must have an assay that differentiates infected volunteers from uninfected volunteers. For AIDSVAX, this requirement has been straightforward as a standard immunoblot can easily discriminate HIV infection (showing antibodies to multiple proteins) from vaccine-induced antibody (to the envelope alone). However, with the more complex DNA or vector vaccine candidates of the future, which may include antigens from multiple HIV epitopes, distinguishing infection from immune response to the vaccine may present unique challenges.

For the secondary endpoint, prevention or modification of disease, much has changed since the beginning of the AIDSVAX trials. At that time, the US FDA was reluctant to use what were referred to as ‘surrogate tests’ to predict the eventual clinical outcome of HIV-infected persons. But as the years have progressed, increasing data have documented that the concentration of virus in the blood of infected persons correlates well with disease progression. Such assays allow studies to predict clinical outcomes without having to follow volunteers for a decade. However, it must be realized that evaluation of secondary endpoints for efficacy may be significantly delayed from the time of primary endpoint efficacy evaluation due to the time required for long-term follow-up of the HIV-infected individual.

Lesson: Investment into natural history studies provides invaluable data for validating the correlation between laboratory assays and clinical disease progression. We have experienced few problems with the use of ELISA tests to screen for HIV infection during the course of our efficacy trials. However, as HIV vaccine candidates become more complex by the addition of other viral antigens, distinguishing vaccinated volunteers from HIV-infected volunteers may prove increasingly difficult. We have also learned that the primary and secondary endpoints can be viewed as separable trials, and that one has to plan for the fact that the primary endpoint of prevention of infection will be complete before the secondary endpoint data to evaluate prevention of disease will become available. This fact has implications for the timing of unblinding and determination of ‘final’ results. In addition, it is important in the evaluation of candidate HIV vaccines to consider the combined effect of prevention of infection and the public health benefit of ameliorating or delaying the progression of disease in the vaccinated who also become infected.

Volunteer recruitment

At the onset of VaxGen's trials, there was concern that 5000 MSM, 300 at-risk women and 2500 IDU could not be recruited into the trials or would substantially reduce the number of people willing to participate in future vaccine trials, implying that the AIDSVAX trials would preclude the conduct of another trial [13]. Our recruitment experience demonstrated that many people are willing to volunteer for HIV vaccine trials. When asked why they volunteer, trial participants indicated an overwhelming desire to do something to help stop the AIDS epidemic. It is worthwhile to note that the recruitment strategies for the North American/European trials have varied greatly in different geographical settings. Some have relied on person-to-person ‘snowball’ referral, others have relied on the general press to publicize the study, and others have relied on community-specific press to publicize the study. In Thailand, it was more challenging as the IDU community is of limited size and a high prevalence of pre-existing HIV infection excluded as much as 40% of those interested in joining.

Lesson: There were thousands of eligible high-risk persons who wanted to volunteer, even after our trials were fully enrolled. With well-developed and -planned recruitment strategies, large-scale trials can be successfully enrolled.

Volunteer retention

Success of any vaccine trial depends on maximizing the proportion of volunteers who will complete the full 3-year follow-up. In large part the key to this success depends on the talents and dedication of the clinic staff and their ability to establish a meaningful relationship with each volunteer. This close relationship helps to engender the commitment to continue a study that becomes rather arduous over time. In addition, there are other practical strategies that can improve follow-up. Initially to maximize recruitment, the North American study enrolled 60 sites across the USA and Canada. During the course of the study, we found the study population to have a rather high rate of relocation. Over 10% of the volunteers in this study have moved with some moving two or three times. Fortuitously, the wide geographic coverage of our sites has allowed almost all relocated volunteers to be successfully transferred to another study site.

In Thailand the issue has been different. In Bangkok, incarceration of IDU is relatively frequent with approximately 25% reporting incarceration in jail or prison every 6 months. This fact has necessitated establishment of a close working relationship between the health authorities of the Bangkok Metropolitan Administration and the Bangkok Department of Incarceration so that follow-up visits could be accomplished in jail or prison without compromising the confidentiality of volunteers and engendering adversity to the volunteer either from guards or other inmates.

Lesson: Organization flexibility and innovation are important characteristics of conducting a successful trial. If the clinics and staff cannot adjust to the needs of volunteers, then volunteers will be lost to the study and the statistical power to determine the efficacy of the vaccine will be weakened.

Monitoring social harms

When initially proposing the advancement of AIDSVAX into phase III trials, there were those who expressed strong concerns regarding potential harms that such a study could cause [19]. The two major concerns were: (i) vaccine trial participants, expecting protection, may increase their risk-taking behaviors; and (ii) volunteers could experience discrimination or other social harms on revealing their participation in a trial or testing false positive when tested for HIV outside of the trial. Regarding the potential for increased risk behavior, measures were taken in the screening and recruitment of volunteers in the phase III studies to ensure that volunteers understood the concepts of blinding, randomization, and the experimental nature of the trial, and that they should not assume any protection from AIDSVAX. For other social issues, the services of experienced clinic staff were made available to help mitigate any potential harm that may occur during their trial participation. In addition, systems were established to quantify these potential issues so that an independent Data and Safety Monitoring Board (DSMB) could monitor the potential for social harms throughout the study. Throughout both studies, we have been reassured that trial participation has not resulted in a significant number of social harms such as discrimination, loss of work or insurance, or travel restrictions. In fact, social harms have been very uncommon with the majority being due to reactions by friends or family members to whom volunteers have revealed their participation in the trial. In these few instances, clinic staff have assisted to calm reactions if needed. Regarding risk-taking behaviors, volunteers in the VaxGen trials are regularly questioned and counseled regarding their risk behaviors. During the study, self-reported risk behavior has decreased in both AIDSVAX studies, most prominently in Thailand where pre-trial education and counseling had not been as extensive as among MSM in North America and Europe. However, it must be remembered that when a safe and effective preventive HIV vaccine becomes available, intensive risk reduction education and counseling must be continued and delivered in conjunction with vaccine delivery programs.

Another potential social issue, the problem of false-positive ELISA assays, is not a major problem with a gp120 vaccine. Because of the near absence of gp120 protein in the virus concentrates used for ELISA tests, few AIDSVAX recipients, despite the presence of high titer anti-gp120, test ‘false’ positive. If the ELISA is positive, it is very easy to differentiate vaccine-induced antibody (envelope reaction only) from HIV infection by Western blot testing.

Lesson: Careful monitoring is essential in HIV vaccine trials. Building in regular assessments of potential medical and social harms in the design of the study has enabled us to address issues quickly and to ensure that the study is not causing the social harms that were predicted.

Monitoring vaccine safety

Although extensive pre-clinical animal toxicity studies and clinical phase I and phase II studies showed no evidence of major clinical adverse side-effects of AIDSVAX, the number of doses to be given during the phase III trials would be much greater and the possibility of an as yet undefined adverse effect was real. To document the potential medical harm attributed to vaccine, all adverse events, regardless of their relationship to the vaccine are recorded. In addition, close monitoring of local reactions at the injection site is standard procedure. All events are collected and analyzed for the biannual DSMB meetings. The DSMB, having the vaccine–placebo randomization code, compares the incidence of medical events in the vaccine group to that in the placebo group. The DSMB is asked specifically whether any serious event is more prevalent in the vaccine recipients. In addition, the potential of ‘immune enhancement’ is closely monitored, i.e., the hypothetical possibility that the vaccine increases susceptibility to infection or accelerates disease progression once infection has occurred. To monitor these possibilities, the DSMB is provided with rates of infection, RNA viral load measurements and CD4 cell counts for those who are HIV-infected (vaccinees and placebo recipients) and asked to ascertain whether there is any potential harm due to vaccine. In all reviews, the DSMB has reassured us that there is no evidence for these adversities.

Lesson: In studies with large trial populations, such as our trials with 7900 volunteers followed over 3 years, the number of adverse events is sizeable and the effort required to track and report them is immense. For both trials as of 1 October 2002 1111 serious adverse events have been reported, including 134 deaths. Fortunately, the DSMB reviews have assured us that the distribution of medical events between vaccine and placebo recipients does not suggest any vaccine-related safety problems. Nevertheless, the effort required to document the details of each case is large and needs to be planned for in future studies.

The effort required and added value through broad collaborations

These phase III efficacy trials have been immense endeavors in terms of the amount of commitment, effort, and data collected. For the two trials combined, almost 900 staff have been employed to screen over 12 000 potential trial volunteers in order to enroll the 7900 volunteers. Over 135 000 clinic visits have been made by the volunteers where almost 56 000 injections have been administered, 127 000 blood draws made, and 1.2 million case report forms filled out which if stacked up would be 330 feet high and exceed the height of the Statue of Liberty (Table 1). This enormous effort could not have been achieved without the dedication of the volunteers and the clinic staff who have been equally committed to the search for an effective HIV vaccine. Of note has been the added value of our collaborations with the US CDC, the US National Community Advisory Board, and the Thai institutions including the Bangkok Vaccine Evaluation Group, Mahidol University, and the Bangkok Metropolitan Administration. Through these collaborations, VaxGen has tremendously benefited from the broad addition of clinical and epidemiologic expertise, and social–behavioral and laboratory research experience. Our final results will far exceed just knowing whether gp120 prevents HIV infection. They will yield very valuable data on effectiveness of recruitment and retention strategies for trial volunteers, risk behavior among MSM and IDU, HIV-1 genetic variation in North America and Thailand, evolving resistance of HIV to antiretroviral therapy, and many other data that can advance HIV vaccine development.

Table 1
Table 1:
Factors involved in the completion of the two phase III HIV vaccine efficacy trials.

Lesson: Although phase III efficacy trials with HIV vaccine candidates are enormous endeavors, they can and should be done as quickly as possible to provide valuable information and insights into approaches to reach the goal of developing an effective HIV vaccine. This effort can be tremendously enhanced through broad collaborations between industry, governments, and academic institutions.

What will be learned?

In the year 2003, it is anticipated that both of these phase III trials will be unblinded, and the initial results made public. For the first time, placebo-controlled data from large-scale HIV vaccine trials will become available to the scientific community. Regardless of levels of overall AIDSVAX efficacy observed, these placebo-controlled trials will greatly increase knowledge concerning the degree of protection provided by different levels of antibody that bind to the various domains in the envelope protein for each of the subtypes included in AIDSVAX. Envelope protein sequencing of intercurrent infections from placebo and AIDSVAX recipients will allow a much better understanding of the interaction of HIV subtype variability and potential protection by vaccine-induced antibodies. Additionally, important information, such as secular and geographic trends in changes in envelope protein sequences of intercurrent HIV infections, the impact of changes in levels of risk behavior on acquisition of infection, and the potential impact of an HIV vaccine on the progression of HIV infection will also become available.

Lesson: Vaccine efficacy trials are a means, not necessarily an end, to developing an effective vaccine. Much can be learned from the results of the VaxGen phase III trials in 2003. It is unfortunate that this type of data might have been available much earlier [13], and that no additional phase III HIV vaccine results will be available again until 2007 at the earliest.

Looking ahead

Now as we approach the conclusion of both efficacy trials, we have new and tough challenges ahead. As Henry Kissinger said, ‘Each success only buys an admission ticket to a more difficult problem'. If efficacy of gp120 is demonstrated in our trials, the scientific and public health community, investors, and indeed the world will be exuberant and expect an HIV vaccine to be widely available the next day at an affordable price. Although there is nothing in the world more desirable, unfortunately this is not possible. The licensure process, including establishment of scale-up manufacturing facilities, will still take at least another 2 years to complete. In addition, recommendations and guidelines for vaccine use must be developed, systems for vaccine administration that include continued intense education and counseling on HIV prevention must be established, and third-party funding mechanisms must be established to assist developing countries with vaccine purchase and distribution. Even with a safe, effective, and available HIV vaccine, years of hard work lay ahead to bring HIV prevention by vaccine to a reality.

But what if efficacy is not demonstrated? We must not overlook how much has been learned through these trials to advance HIV vaccine development. We must not let a ‘no efficacy’ result be interpreted as a ‘failed trial'. This conclusion in itself would be a severe impediment to future trials. The current AIDSVAX phase III trials will bring forth valuable information for future HIV vaccine development.

Lesson: Regardless of phase III trial results, much has been and will be learned from these trials. Moreover, if results show efficacy, years of work lie ahead to manufacture the vaccine in large scale, establish guidelines and systems for vaccine use and availability, and create funding mechanisms to provide it to the world at an affordable cost.


The world's first HIV vaccine efficacy trials will conclude in early 2003. We, as the industry partner in this multi-partner collaboration, have learned many valuable lessons that can be applied in the future to support the successful development of an HIV vaccine, whether it is gp120 or a future product yet to be tested (Table 2). Vaccines are developed and manufactured by industry. However, this task cannot be accomplished by industry alone. To achieve the goal of providing a safe and effective HIV vaccine for those in need, industry, investors/funders, clinical and research partners—a collaboration of both public and private institutions are needed. The government of the USA, together with governments of other industrialized countries, needs to work with existing international institutions and partners (e.g., the International AIDS Vaccine Initiative, WHO/UNAIDS, World Bank, the Global Alliance for Vaccines Initiative, and private foundations) to ensure that people in developing countries who are the most severely stricken by the AIDS epidemic are provided a safe, effective and affordable HIV vaccine.

Table 2
Table 2:
Problems and solutions for conducting HIV vaccine efficacy trials.


VaxGen thanks the thousands of volunteers in VAX003 and VAX004 for their courage and dedication which had made these phase III trials possible; The VAX003 Bangkok Vaccine Evaluation Group and the 61 study sites of VAX004 for their leadership, commitment, and work on these trials; the US National Institutes of Health and the Centers for Disease Control and Prevention; Thailand MOPH – US CDC Collaboration; Genentech, Inc.; the VaxGen Data and Safety Monitoring Board; and E. Li, M. Chernow, N. Virani-Ketter, G. Alonzo, J. Jermano, J. Curd, T. Gregory, and J. Obijeski for their efforts in the AIDSVAX trials.


1.Francis DP, Gregory T, McElrath MJ, Belshe RB, Gorse GJ, Migasena S, et al. Advancing AIDSVAX® to Phase III. Safety, immunogenicity, and plans for phase III.AIDS Res Hum Retroviruses 1998, 149 (suppl 3):S325–S331.
2.Berman PW. Development of bivalent rgp120 vaccines to prevent HIV type 1 infection.AIDS Res Hum Retroviruses 1998, S277–S289.
3.Berman PW, Gregory TJ, Riddle L, Nakamura GR, Champe MA, Porter JP, et al. Protection of chimpanzees from infection by HIV-1 after vaccination with recombinant glycoprotein gp120 but not gp160.Nature 1990, 345:622–625.
4.Berman PW, Murthy KK, Wrin T, Vennari JC, Cobb EK, Eastman DJ, et al. Protection of MN-rgp120-immunized chimpanzees from heterologous infection with a primary isolate of human immunodeficiency virus type I.J Inf Dis 1996, 173:52–59.
5.Hladik F, Bender S, Akridge RE, Hu YX, Galloway C, Francis D, et al. Recombinant HIV-1 glycoprotein 120 induces distinct types of delayed hypersensitivity in persons with or without pre-existing immunologic memory.J Immunol 2001, 166: 3580–3588.
6.Muesing MA, Smith DH, Cabradilla CD, Benton CV, Lasky LA, Capon DJ, et al. Nucleic acid structure and expression of the human AIDS/lymphadenopathy retrovirus.Nature 1985, 313:450–458.
7.Berman PW, Groopman JE, Gregory T, Clapham PR, Weiss RA, Ferriani R, et al. Human immunodeficiency virus type 1 challenge of chimpanzees immunized with recombinant envelope glycoprotein gp120.Proc Natl Acad Sci USA 1988, 85:5200–5204.
8.Lasky L, Nakamura G, Smith DH, Fennie C, Shimasaki C, Patzer E, et al. Delineation of a region of the human immodeficiency virus type 1 gp120 glycoprotein critical for interaction with the CD4 receptor.Cell 1987, 50:975–985.
9.Leonard CK, Spellman MW, Riddle L, Harris RJ, Thomas JN, Gregory TJ, et al. Assignment of intrachain disulfide bonds and characterization of potential glycosylation sites of the type 1 recombinant human immunodeficiency virus envelope glycoprotein (gp120) expressed in Chinese hamster ovary cells.J Biol Chem 1990, 265:10373–10382.
10.Schwartz DH, Gorse G, Clements ML, Belshe R, Izu A, Duliege AM, et al. Induction of HIV-1 neutralizing and syncytium-inhibiting antibodies in uninfected recipients of HIV-1 IIIB rgp120 subunit vaccine.Lancet 1993, 342:69–73.
11.Heyward WL, MacQueen KM, Goldenthal KL. HIV vaccine development and evaluation: realistic expectations.AIDS Res Hum Retroviruses 1998, 14 (suppl 3):S205–S210.
12.Thomas P. Big Shot. Passion, Politics and the Struggle for an AIDS Vaccine. New York: PublicAffairs; 2001.
13.King RT. FDA allows large-scale trial of AIDS vaccine. New York: Wall Street Journal, June 3, 1998:1.
14.Buchbinder SP, Douglas JM, McKirnan DJ, Judson FN, Katz MH, MacQueen KM. Feasibility of human immunodeficiency virus vaccine trials in homosexual men in the United States; risk behavior, seroincidence, and willingness to participate.J Infect Dis 1996, 174:954–961.
15.Vanichseni S, Kitayaporn D, Mastro TD, Mock PA, Raktham S, Des Jarlais DC, et al. Continued high HIV-1 incidence in a vaccine trial preparatory cohort of injection drug users in Bangkok, Thailand.AIDS 2001, 15:397–405.
16.Szmuness W, Stevens CE, Harley EJ, Zang EA, Oleszko WR, William DC, et al. Hepatitis B vaccine: demonstration of efficacy in a controlled clinical trial in a high risk population in the United States.N Engl J Med 1980, 303:833–841.
17.Francis DP, Hadler SC, Thompson SE, Maynard JE, Ostrow DG, Altman N, et al. Prevention of hepatitis B with vaccine: report from the Center for Disease Control multi-center efficacy trial among homosexual men.Ann Intern Med 1982, 97:362–366.
18.MacQueen KM, Vanichseni S, Kitayaporn D, Lin LS, Buavirat A, Naiwatanakul T, et al. Willingness of injecting drug users to participate in an HIV vaccine efficacy trial in Bangkok, Thailand.J Acquir Immun Defic Syndr 1999, 21:243–251.
19.Chesney MA, Chambers D, Kahn JO. Risk behavior for HIV infection in participants in preventive HIV vaccine trials: a cautionary note. J Acquir Immun Defic Syndr 1997, 16: 206–271.

HIV; candidate vaccine; clinical trial; phase III

© 2003 Lippincott Williams & Wilkins, Inc.