What Next?

As data analysis proceeds on the STEP trial, some future trials are placed in a temporary holding pattern

By Kristen Jill Kresge

Science is an empirical process. "We had a hypothesis, we tested it, and it didn't work," says Emilio Emini, head of vaccine research and development at Wyeth, referring matter-of-factly to results of the Phase IIb STEP trial, which showed Merck's adenovirus serotype 5 (Ad5) candidate was not effective.

After the sudden failure of MRKAd5, many are wondering what will happen next with the string of other cell-mediated immunity (CMI) candidates that are in various stages of clinical testing or preclinical development—many of which use an adenovirus vector, albeit in slightly different form than Merck's. Some of these candidates contain different immunogens and are also being tested in heterologous prime-boost regimens, unlike Merck's, which involved multiple immunizations with the same vaccine. Other candidates use different serotypes of Ad, including Ad35, Ad26, or a chimeric version of Ad5. "They [different serotypes of Ad] are probably as different as two other viral vectors altogether," says Dan Barouch of Beth Israel Deaconess Medical Center in Boston. These differences may result in better immunogenicity and a higher likelihood of efficacy, and researchers make several arguments as to why they should also get a fair trial. "[The] STEP results proved that this product failed and should not be construed as indicative that all adenoviral vectors or other viral vectors will fail," says Peggy Johnston of the Division of AIDS, part of the National Institute of Allergy and Infectious Diseases (NIAID).

PAVE 100 was the next Phase IIb test-of-concept trial on tap—it was scheduled to begin just weeks after Merck and NIAID announced that immunizations in the STEP trial were stopped. This NIAID-sponsored 8,500-person trial aims to test the safety and efficacy of a heterologous prime-boost combination of DNA and Ad5 vector-based vaccine candidates that were developed at the Vaccine Research Center (VRC), part of NIAID, in collaboration with the HIV Vaccine Trials Network (HVTN), IAVI, and the US Military HIV Research Program (USMHRP). This same regimen was also to be tested in a Phase II trial, known as V002, conducted by IAVI in Rwanda, Kenya, Uganda, and Zambia. The start of both of these trials was immediately delayed.

"There are substantial differences between the Merck product and the VRC product," says Gary Nabel, director of the VRC. One difference is the heterologous prime-boost regimen. In both preclinical and clinical studies, researchers at the VRC report that the DNA vaccine candidate appears to effectively prime the immune system, resulting in an enhanced immune response following the boost vaccination with Ad5. The exact mechanism for this is unknown (see One-two combination, IAVI Report, May-June 2007), but Nabel says this combination "elicits a quantitatively and probably qualitatively different immune response," including more diverse CD4+ T-cell responses and an increased magnitude of CD8+ T-cell responses than when Ad5 is used alone.

Nabel also points to the differing efficacy between MRKAd5 and the VRC's DNA/Ad5 combination when evaluated in the SIV/nonhuman primate (NHP) model. In this model, he says, "The Merck vaccine really doesn't work," whereas the VRC's DNA/Ad5 candidates promoted long-term survival in SIV-infected nonhuman primates and suppressed viral load during the early stages of SIV infection (Science 312, 1530, 2006).

But when the latest data from the STEP trial was released at the HVTN meeting on November 7, researchers began grappling with additional questions (see A STEP back?). For now it is too soon to determine if there is any real link between enhanced risk of HIV infection and receipt of the MRKAd5 vaccine in some individuals. Some groups, including the AIDS Vaccine Advocacy Coalition (AVAC), are now advocating that other efficacy trials should be postponed until "definitive conclusions" can be drawn. But many researchers think it is still imperative to test other candidates that induce different immune responses. "I certainly feel there are ways to go forward safely, but we have to do that together," says Scott Hammer of Columbia University and chair of the PAVE 100 protocol team.

The opening of the PAVE 100 trial is still postponed and the protocol team will be meeting soon to discuss possible changes to the trial design. "It has to be amended in light of the STEP trial," says Hammer. "We do not have the details of that amendment in place. The regimen won't change but the study design might." Some possible alterations might involve the populations enrolled in the trial or the monitoring of the data collected while the trial is underway to ensure the safety of the volunteers.

"I am confident that the team will come up with a plan that will have equipoise and participant safety and scientific integrity all wrapped into it," says Hammer. This revised trial plan will then be sent to an oversight committee for review and comment. "The go, no go decision is not in our hands," says Hammer. "[It] is in the hands of the Division of AIDS at NIAID. Peggy [Johnston] and Tony's [Tony Fauci, director of NIAID] office [make] the ultimate decision."

"[We are] looking very carefully at the differences between products," says Johnston. "As we decipher the immune responses that are made by this vaccine [MRKAd5] in humans, we can then have a better comparison with other products in the pipeline and make a best assessment whether the differences are sufficient for moving ahead," she adds.

Andreas von Bubnoff contributed reporting to this article.

Another issue raised by the STEP trial is the use of the Phase IIb test-of-concept trial to evaluate the efficacy of AIDS vaccine candidates. The idea of using these smaller, less expensive trials has become fashionable in the field as a way to get a quick read on whether or not a candidate is likely to protect against HIV infection or provide some level of partial protection that could limit disease progression. The STEP trial was the first to use a Phase IIb trial to evaluate an AIDS vaccine candidate—though similar trials have been used for other vaccines—and it successfully showed that this design can yield earlier results with fewer volunteers than a full Phase III trial. "The STEP study trial design was an enormous success," says Steve Self, a biostatistician with the HIV Vaccine Trials Network (HVTN).

Many people praised Merck for deciding to evaluate their Ad5 candidate in a Phase IIb test-of-concept trial and for planning an early analysis by the data safety monitoring board, which gave decisive information on the efficacy of the product even faster. "It enabled us to get an answer as quickly as possible," says Peggy Johnston of the Division of AIDS, part of the National Institute of Allergy and Infectious Diseases (NIAID). "That, in hindsight, proved to be an excellent decision." Andrew McMichael of Oxford University agrees. "Maybe we should do more [such] trials rather than the full blown 10,000-person Phase III trial."

But some argue that smaller trials, an idea known as screening-test-of-concept or STOC trials, could provide preliminary efficacy data for candidates that induce cell-mediated immunity even faster than Phase IIb trials. This novel clinical trial concept has been championed by IAVI as a way to conduct rapid, less costly trials in far fewer volunteers (AIDS 21, 2259, 2007).

These trials would involve 500 to 1,000 volunteers in areas with high HIV incidence, compared to the 3,000 participants in the Phase IIb STEP study or the 8,500 volunteers in the original plans for PAVE 100. "We at IAVI feel that it's important to move quickly and be as efficient as possible in collecting clinical data to guide the field," says Pat Fast of IAVI.

But the STOC trials will also provide more limited information than can be collected from larger Phase IIb studies. The current STOC design would not allow researchers to determine if a candidate protects against HIV infection. It would only allow researchers to detect a difference in viral load in volunteers who do acquire HIV, despite vaccination. Emilio Emini of Wyeth calls the concept of STOC trials a "nice design" but warns that "you can't cut the corners too tightly."

"If we think that there may be differences in acquisition of infection, then that's not the design to do," says Johnston. But many researchers think the best possible hope for cell-mediated immunity candidates that don't elicit neutralizing antibodies is a reduction in viral load in vaccinated individuals if they become HIV infected—especially now, given the results of the STEP trial. Still some are cautious. "We still don't know if the basic assumption is correct," says José Esparza of the Bill & Melinda Gates Foundation. "After the current results, we need to be extra careful with our assumptions."

Ian Gust of the University of Melbourne and a member of IAVI's board of directors, says that both Phase IIb and STOC trials have validity, but he views the use of STOC trials as an attempt to move the field forward as rapidly as possible. "I don't think IAVI should reflect the prevalent or most conservative point of view," he says.—KJK