Reproducing Protection

Satellite symposium at the Sydney conference gathers researchers and regulators to discuss the benefits and risks of replicating viral vectors for AIDS vaccines

By Simon Noble

Fingers are crossed that the most promising of the vaccine candidates currently in clinical testing will show some benefit to vaccinated volunteers who later become HIV infected. However, since virtually all of those candidates focus on eliciting cell-mediated immunity, the smart money seems to be on the outcome that, at best, they will have low-to-moderate efficacy, leading to a lower viral set-point that ameliorates disease progression and potentially lowers the likelihood of transmission. If that's likely the best that can be expected from the leading candidates, that means there's still a need for better vectors that induce more robust immune responses. Since most of the obvious options for replication-deficient viral vectors are under active development, some researchers are turning to a more classical approach with a novel twist—replicating viral vectors.

But replicating viral vectors bring with them a new set of risks associated with their ability to propagate. So the regulatory authorities and the general public will have to be convinced that the risk inherent in such a vaccine approach is outweighed by the benefit. These risk-benefit analyses will depend greatly on local conditions; an acceptable risk in a very high incidence environment like Lesotho might well be different from the risk acceptable in Australia, for instance.

To get the conversation started, IAVI held a satellite symposium at the Sydney Conference entitled 'Accelerating the Development of Replicating Viral Vectors for AIDS Vaccines.' The intention was to bring together scientists and regulators to discuss and better understand the issues around the development and evaluation of replicating vectors, and to identify risks that should be addressed during development.

Bang for your buck

Ian Gust of the University of Melbourne introduced the satellite in his role as chair, pointing out that vaccine development is now "much more complex than in the days of Jenner and Pasteur, when you could revolutionize medicine with a vaccine tested in a single study with an n=1." Since many of the licensed vaccines today are intended to prevent relatively rare but severe complications of increasingly unfamiliar diseases, safety has become the prevailing concern of regulatory authorities and, since most vaccinations are given to children, parents. Gust illustrated this overriding consideration of safety by recounting two examples; the recently-approved human papillomavirus vaccine that had to be tested in clinical trials involving more than 70,000 volunteers at a cost of over US$200 million before licensure, and the promising live-attenuated hepatitis A vaccine in the early 1980s that was effective in nonhuman primates (NHPs) but occasionally led to increased transaminases, indicating moderate liver inflammation. The licensee, Smith Kline Beecham, decided not to further develop this vaccine candidate, fearing that some recipients might suffer jaundice, and consequently the hepatitis A vaccine now licensed is a conventional chemically-inactivated vaccine.

Gust pointed out that this contemporary emphasis on inactivated vaccines should not mask the fact that many of the most effective vaccines currently in use—the oral polio vaccine, and vaccines against measles, mumps, rubella, and chickenpox—are live-attenuated viruses. In the case of AIDS vaccines, the best protective efficacy seen so far in NHPs challenged with SIV has been elicited by a live-attenuated vaccine approach. Gust finished off by saying that, given the paucity of good immunogenicity data coming from current vaccine candidates in clinical trial and the relatively robust immunogenicity data described in the live-attenuated SIV model in rhesus macaques, IAVI believes it's time to reconsider the use of replication-competent viral vectors.

Wayne Koff of IAVI then gave a comprehensive overview of the state of the field, emphasizing that almost all of the current candidates in the pipeline are focused on cell-mediated immunity and that none of them elicits the neutralizing antibodies that will likely be required for true sterilizing immunity, nor do any induce effective antiviral activity at the mucosae, including the gut-associated lymphoid tissue (GALT) that is such a critical site in HIV pathogenesis. And, importantly, none of the candidates elicit a combination of responses that might fight HIV on a number of fronts.

Since many researchers accept that the best hope with the current Ad5 candidates is low- to moderate-efficacy, Koff said that "many are beginning to accept that we need to think about novel vectors with a bit more 'oomph'." Although generally risk averse, "the field is evolving, and it is beginning to think outside the box" and there is some limited work on potentially more promising approaches like replicating vectors that might mimic the efficacy of live-attenuated SIV. Koff reiterated the dramatic efficacy that live attenuated vaccines have had against major epidemics of smallpox and polio, once common scourges. HIV is already responsible for more than 25 million deaths and causing 4 to 5 million new infections each year.

He then presented a summary of published NHP studies outlining the effectiveness of live attenuated SIV (SIVΔΔnef, Δ3, or Δ5G) protection against the homologous pathogenic SIV challenge, showing that 94% or more of animals in these experiments had been protected (>3 log suppression of viral load at setpoint); in contrast, all other vaccine strategies protected only 7% of NHPs by the same criterion. Encouragingly, recent data indicates that SIVmac239Δnef can lower viral setpoint by 2 logs after heterologous pathogenic SIV E660 challenge (David Watkins, unpublished data; see Groundhog Day, IAVI Report 11(2)).

In closing, Koff said IAVI thinks that replicating vectors that elicit persistent and/or mucosal immunity may be required for control of subsequent HIV infection, and that accelerating their development will require a partnership among vaccine developers, regulatory agency scientists, and developing country scientists, public health officials, institutional review boards, and communities where the clinical trials will take place.

Replicating promise

Chris Parks of IAVI gave an overview of the current pipeline of replicating viral vectors. Listing the potential vaccine strategies in order of increasing immunogenicity/efficacy—nucleic acids, protein/peptide subunits, virus-like particles, inactivated virus, nonpropagating viral vectors, replicating viral vectors, 'Jennerian' vaccines (i.e. related nonpathogenic animal viruses), live attenuated virus—Parks pointed out that, contrarily, this corresponds with a decrease in safety.

But historically the most successful vaccines have come from either live-attenuated vaccines or Jennerian vaccines. Since live-attenuated HIV vaccines are likely not feasible because of safety concerns and there are no candidate Jennerian vaccines, live viral vectors are the next best option in terms of immunogenicity.

Parks then described some of the qualities sought in replication-competent viral vectors: replication and spread of the vector should occur such that it generates abundant and sustained antigen expression, but the replication should be sufficiently restricted so that disease does not result; the antigen expression should induce durable immune responses (antibodies, polyfunctional T cells with a balanced Th1/Th2 response) and immunologic memory; and ideally, the vector would be sufficiently immunogenic to be efficacious without recall to complex heterologous prime-boost regimens. Additionally, it may well be desirable that a replicating vector has cell and tissue tropism that leads to induction of mucosal immunity, and that any pre-existing immunity issues are minimal.

Parks, formerly of Wyeth Vaccines, recounted the story of vesicular stomatitis virus (VSV), one of the leading replicating viral vectors currently being advanced to the clinic by the manufacturer. Prototype VSV vectors developed in John Rose's lab at Yale University showed promising results—NHPs vaccinated with the vectors were protected from disease after challenge with pathogenic simian-human immunodeficiency virus (SHIV), and no adverse events were seen. Wyeth in collaboration with the National Institute of Allergy and Infectious Diseases (NIAID) at the NIH had planned to go into clinical trials but first cautiously conducted a pilot neurovirulence test in NHPs; research from as early as the 1930s had indicated that VSV could be neurovirulent in neonate mice. The results from the pilot NHP study indicated that the prototype vector did cause inflammation and tissue necrosis when inoculated intrathalamically. This result subsequently initiated years of further research that resulted in development of a highly attenuated vector that was far less neurovirulent and retained its immunogenicity. Wyeth and NIAID are now advancing this candidate into clinical trials.

The VSV tale, Parks noted, exemplifies the need to balance safety and efficacy. But he also asked whether it should be necessary to test neurovirulence by direct inoculation into the brain of a vector that will be administered intramuscularly, and said that these questions around safety testing standards need to be discussed. Equally vexing is the issue of the public's evolving view of acceptable risk with regard to vaccination, and Parks contrasted the relative risks associated with some vaccines over the past 60 years or so; smallpox vaccination was associated with notable complications, ranging from erythematous urticaria to death, but was still widely accepted by the general public, whereas the current measles, mumps, and rubella (MMR) combination vaccine has drawn suspicion due to well-publicized but unsupported claims that the vaccine is associated with autism and gastrointestinal disease.

This decreasing tolerance for risk has consequences. Pharmaceutical companies now must demonstrate that a vaccine is nearly 'risk-free' and this is extremely complex, lengthy, and costly. It also increases the probability of vaccine candidate failure during development, and developers are mindful that rare adverse events post-licensure will be costly to both economics and reputation. Parks ended by saying that development of risk-free live vaccines is probably unachievable, which makes replicating vectors less appealing to vaccine developers.

Regulatory affairs

The regulatory experts presenting at the satellite session emphasized many of the same practical aspects highlighted above. Jim Ackland of Global BioSolutions, a regulatory consultancy based in Australia, pointed out that it's not just the potential for reversion to virulence and adverse events that concerns regulators with regard to replicating viral vectors. He pointed out that, as with any novel product, there are both predictable and unpredictable potential risks, and the latter can only be assessed through careful clinical development. Predictable risks can often be assessed during preclinical development, and these include manufacturing and the potential for contamination with adventitious agents, genetic stability of the vector and its antigenic insert, toxicity, virulence, transmission, and recombination with wild type or other viruses that might facilitate reversion to virulence.

Ackland also noted that there are only limited guidelines available for viral vaccines from regulatory authorities and these tend to be generic in nature, and emphasized the need to engage regulatory authorities in the development process and appreciate their challenges and expectations. Good regulation is based on good science—for regulators "it's not publish or perish, it's document or perish," he said—and it's important to be familiar with the existing guidelines and identify precedent.

Keith Peden of the FDA indicated that replicating virus vectors presented authorities with a potential regulatory nightmare since they can evolve rapidly in the human host, which he summed up as "predicting, and regulating, the unpredictable." He reiterated the importance of involving regulatory authorities early and throughout the development process. As to the choice of the vector, he emphasized the importance of virus tropism in the selection process, cautioning that neurotropic or cardiotropic viruses should be avoided. He also counseled to choose prudently an acceptable cell substrate on which to grow the vector. He stated that the use of recombinant DNA virus clones is preferable to generate the vaccine virus, as this eliminates concerns due to adventitious agents in any original virus vector and provides a constant source of the vaccine. Generation of the vaccine virus from a recombinant DNA clone in qualified cells under cGMP would simplify the regulatory path. Peden did illustrate that regulatory positions do change when data are provided by reference to the recent discussions at the FDA regarding the use of continuous cell lines for vaccine production, saying that products manufactured in continuous cell lines (e.g., PER.C6 cells and Madin-Darby canine kidney cells) have been permitted to enter clinical trials.

Gary Grohmann of the Therapeutic Goods Administration, the Australian Government regulatory body, concurred with Peden and said that "it's obvious we are moving beyond the conventional." He stressed the need for public acceptance of replicating viral vectors, and asked if there was the need for a communication forum specifically for this purpose. Gust made the point that regulatory authorities traditionally reflect the feeling of a community.

The regulatory context and challenges in developing countries were discussed by Helen Rees of the University of Witwatersrand, Johannesburg, and she began by pointing out that the risk-benefit analysis of any AIDS vaccine approach will differ by region. She pointed out that some of the inequities in global health research greatly affect attitudes in developing countries, including biomedical regulatory attitudes, such as the "10/90" gap that sums up the fact that only 10% of global investment in health research is committed to solving 90% of the world's health problems, and that there are more African researchers in the US then there are in the whole of Africa. Also, Rees said that "the 'guinea pig' concern continues to overshadow all other considerations" and affects political attitudes toward research and clinical trials conducted in developing countries.

There has been a huge increase in the number of clinical trials being conducted in developing countries—from fewer than 1000 trials in 1990-92 to more than 5000 in 1996-98—but clinical trial review is still an emerging area of expertise and therefore these countries are still largely reliant on the FDA and European Agency for the Evaluation of Medicinal Products (EMEA) as the default standard. Similarly, the capacity to evaluate trial ethics is still emerging, so it must be made clear that trials are being done in the context of sufficient ethical and regulatory oversight. Recently there have been moves to streamline regulation in East, Southern, and West African regions, "but it will be catch up, if a replicating vector is developed then it will have to be considered and ways to evaluate it developed."

In discussion after the presentations, Rees said that the media also needs education so they can communicate difficult concepts in simple language. She also said that many in developing countries look toward the World Health Organization (WHO) as a reputable, independent, and trustworthy body that can facilitate acceptance of new concepts, and strongly recommended that the WHO be engaged in an expert consultation to continue the discussion about the acceptance of replicating viral vectors. Koff agreed that now is the time to begin this dialogue since the VSV vector should be ready to go into clinical trials in 12 to 18 months and other replicating vectors are under active consideration.