Imbokodo trial results highlight challenge of developing an HIV vaccine and urgent need for access to proven prevention tools

At the end of August, disappointing results were released from Imobokodo, a phase IIb clinical trial evaluating the efficacy of an HIV vaccine candidate developed by the Janssen Pharmaceutical Companies of Johnson & Johnson. [1]

Among cisgender women recruited in five African countries, the vaccine regimen didn’t offer significant protection against acquisition of HIV infection. The outcome underscores the difficulty of inducing protective immunity against HIV compared to other pathogens, and highlights the importance of making effective biomedical prevention options – oral or injectable pre-exposure prophylaxis (PrEP) and the dapavirine ring – accessible to those in need, including women in the communities where Imobokodo took place. 

The Imbokodo Trial

The trial enrolled 2,637 cisgender women aged 18–35 years old in Malawi, Mozambique, South Africa, Zambia, and Zimbabwe. The vaccine comprised priming immunisations with adenovirus serotype 26 (Ad26) vectors encoding four different HIV mosaic antigens that include elements from multiple virus clades, followed by boosts combining the Ad26 vector with an HIV gp140 envelope protein (from a clade C virus) in alum adjuvant.

HIV clades are groups of similar virus variants that primarily circulate in particular geographic regions across the globe (e.g. clade B is predominant in the Americas while clades A, C and D are most common on the African continent). A key idea behind the vaccine was to induce immune responses capable of targeting HIV from many clades.

Behavioral risk factors for HIV acquisition were evenly distributed between the vaccine and placebo groups in the trial. Uptake of oral PrEP, which was made available to all participants, was low – only around 3% of participants displayed detectable tenofovir levels, and the proportion in each arm with levels considered likely to be effective was even lower at 0.2% in the placebo group and 0.4% in the vaccine group (this slight difference was not statistically significant).

Trial results

The results were that 63 of 1,109 participants (5.68%) who received a placebo or dummy immunisation acquired HIV over a two-year period, compared with 51 of 1,079 participants (4.73%) given the vaccine. There were 25% fewer HIV infections in the vaccine group compared to placebo, but the difference was not statistically significant – in other words, it may have resulted from chance.

Statisticians calculate a confidence interval (CI) that gives a range of possibilities for where the true result may lie and, in this case, it was anywhere from a 10% increase in the risk of HIV acquisition to a 49% reduction (–10% to +49%). For a result to be considered statistically significant, the lower end of the CI must be above zero.

Some researchers have speculated that the apparent 25% difference between the vaccine and placebo arms might have reached statistical significance if the trial had been larger, but this is not certain.

The hint of a slight reduction in HIV incidence among vaccine recipients cannot be considered formal evidence of vaccine-induced protection from infection. But it does allow researchers to explore whether there might be notable differences between vaccine recipients who acquired HIV and those who did not, such as particular types of vaccine-induced immune responses.

Importantly, the vaccine regimen was found to be safe, with the typical reaction of mild local injection site pain and redness reported as more common among vaccine compared to placebo recipients.

Implications for the HIV vaccine field

The most immediate question for HIV vaccine research was whether another ongoing efficacy trial of the Johnson & Johnson vaccine should continue. Mosaico is studying a similar regimen in 3,600 cisgender men and transgender people who have sex with cisgender men and/or transgender people.

After review by the trial’s Data Safety Monitoring Board (DSMB), it was recommended that Mosaico continue for several reasons: it includes a slightly different protein boost immunisation (consisting of a mosaic rather than clade C gp140 protein), the prevalent HIV clade where the trial is taking place is different from Imbokodo (clade B versus clade C), and the route of exposure to HIV is also likely to be different (rectal versus vaginal).

The potential for Mosaico to produce a better result is unclear. There may be some reason to hope that the larger sample size of the trial could lead to a similar degree of vaccine efficacy achieving statistical significance. But it would be surprising if the Mosaico results showed a level of efficacy sufficient to justify licensure (e.g. >50% reduction in risk of HIV acquisition).

Perhaps the most optimistic possibility is that Mosaico, combined with prior results from Imbokodo and RV144 (an older vaccine trial in Thailand where a reduction in HIV incidence of around 30% crept above the threshold for statistical significance), could add to a body of evidence that current experimental vaccine approaches can offer a low level of protective efficacy. [2]

Work could then continue to try to figure out if there is any way of improving efficacy to reach the threshold considered necessary for licensure, or if it represents the ceiling for what these types of vaccine can achieve.

The broad common thread between all experimental HIV vaccines tested to date is that, while there are many nuances to each regimen, they induce immune responses against the virus that are known as non-neutralising. In the case of the Johnson & Johnson vaccine, these include T cell and antibody responses intended to mediate clearance of virus-infected cells. Induction of these responses was associated with protection against persistent infection in the SIV/macaque animal model. [3]

If researchers can’t figure out a way to enhance the efficacy of vaccines that induce non-neutralising responses, the primary option that remains is solving the extremely difficult challenge of inducing broadly neutralising antibodies (bNAbs) with HIV vaccines.

Neutralising antibodies are considered a critical component of the protective response created by many licensed vaccines (including recently developed COVID-19 vaccines). Stimulating the production of bNAbs against HIV is far more difficult than for most other pathogens, for a variety of reasons, including:

• The virus’s outer envelope protein is cloaked in glycan (sugar) molecules that are difficult for antibodies to penetrate or attach to.
• The mutation rate of the envelope protein is extremely high, with very few parts of the protein offering stable, conserved targets for antibodies.

Despite these challenges, rare bNAbs with strong anti-HIV activity have been identified and characterised using new techniques that allow researchers to fish the B cells that produce them out of blood samples taken from people with HIV. The bNAbs are typically not able to benefit the individuals they were sampled from, likely due to being present at low levels in the face of high amounts of virus.

But the bNAbs can be grown to high levels in the lab and are being tested as both preventive and therapeutic interventions, delivered via infusion or subcutaneous injection. They also provide a guide for the type of bNAb response an HIV vaccine would need to induce to be successful.

In recent years, HIV vaccine constructs that aim to coax B cells down the first steps toward the long-term goal of bNAb production have begun to enter clinical trials. Results presented at the beginning of 2021 indicate that it’s possible to bolster the number of B cells that may represent a starting point for bNAb production (these results, while promising, were unfortunately widely misunderstood in the media). [4]

Researchers are now employing Moderna’s mRNA technology to test whether delivering specially engineered proteins into the body can enhance the numbers of B cells that may have the potential to produce bNAbs. The launch of these trials has generated considerable news coverage. But even if successful, additional vaccines will need to be designed to take the B cells further toward the goal of bNAb generation. An article in the UK’s Independent newspaper does a good job of conveying that this work is at an early stage. [5]

In addition to bNAbs, some researchers are exploring whether there might be novel types of immune responses – not yet induced by any vaccine candidates – that could offer some protection against HIV. The main hope is a strategy that employs a weakened form of cytomegalovirus (CMV) as a vaccine vector to deliver HIV components into the body. The vaccine has shown promise in the SIV/macaque model, consistently protecting around 50% of immunised animals. Protection is associated with induction of an unusual type of CD8 T cell response. Phase I clinical testing in humans is now underway. [6]

Implications for access to biomedical prevention

In the absence of an imminent HIV vaccine, the Imbokodo results underscore the urgent need to make existing biomedical HIV prevention interventions accessible to those who most stand to benefit from them. The cisgender African women who participated in the trial represent a critically important population, as is demonstrated by the high rate of HIV infections that occurred. Many biomedical HIV prevention trials have recruited large cohorts of dedicated volunteers from communities of African women, but the payoff in terms of accessible interventions has so far been minimal.

This inequity must be addressed by scaling up availability of oral PrEP, and – when approved – the vaginal dapavirine ring and long-acting injectable PrEP. These points are emphasised in AVAC’s statement on the Imbokodo results and in a follow up interview with their regional stakeholder engagement advisor Nandisile Luthuli and executive director Mitchell Warren (conducted by Tim Murphy for the Body.com). [7, 8]

Additional resources