RV144: New discoveries from Old Trials.

There have been a series of papers describing the "Antibody Profile" of participants in the partially successful RV144 Trial that I had described earlier. The first paper looking at the antibody type is a paper by Nicole Yates that appeared in Science Trans Med (2014).

Very little is known about the protective role of vaccine-induced IgG3 antibodies for HIV-1. Previously, it was found that a best known anti-HIV-1 antibody, 2F5, is of IgG3 origin. In the paper by Yates et al., they compared antibody profiles of RV144 trial and the VAX003 trial, which was a modification of the RV144 vaccine.

Background:

RV144 consists of (cladeB/E) ALVAC prime plus 2 protein boosts rgp120 boost

VAX003: consists of (cladeB/E) gp120 protein immunogen.

Results:

*It was found that IgG3 response rates were significantly higher in RV144 compared to VAX003 at peak immunogenicity time point.

**V1-V2 IgG3 antibodies correlated with decreased risk of infection in RV144.

***Most significant finding is that ALVAC prime induced IgG3 preferentially as the IgG3 response was strongest to 92TH023 envelope sequence which is present in ALVAC prime but not in MN and A244 sequences present in protein boost of both RV144 and VAX003.

Minor Points:

In VAX003, two protein boosts elevated Env IgG3 response and declined after 4 protein boosts, suggesting that continued protein/alum boosting may help push the immune response toward lower IgG3 and higher IgG4. Therefore, additional protein boosts can skew IgG3 subclass distribution, even though higher levels of neutralizing antibodies might be induced.

Vaccine-induced IgG3 and IgG responses for either Env or V1-V2 were not significantly predictive of one another suggesting that IgG3 response cannot be inferred from IgG measurement.

However, RV144-induced Env V1-V2 IgG3 response half-life is 1.6-3.7 fold shorter than IgG.

What might be the mechanism of action?

There was a significant association of Env IgG3 and V1-V2 antibodies with antibody-dependent cellular cytotoxicity (ADCC) in RV144, but not VAX003.

Caveat:

Another explanation is that while RV144 focussed on healthy populations, VAX003 used a injection drug user participant group.

Questions (from the paper):

1. Is IgG3 response a marker of HIV-1 risk or does it decrease HIV-1 risk? One has to perform mucosal anti-viral assays or virion capture.

2. The IgG3 response in RV144 and VAX003 might be due to differences in epitope recognition on HIV-1 visions of HIV-1-infected cells.

3. RV144 showed a 60% vaccine effect at 12 months which reduced to 31.2% by 48 months. Is this related to IgG3 response?

Implications:

Other vaccine trials should test for IgG3 response as a correlate of infection risk.

My own comments about the study:

In fact, IgG3 response is not unique to HIV-1. The antibody response is also seen in influenza virus infection. For example, Frasca et al., 2013 showed that there was a robust IgG3 response in healthy individuals vaccinated against H1N1 and there was a 40-50% IgG3 response and the antibody levels were positively correlated with TNFa and IL-6. In fact, activation-induced cytidine deaminase (AID) mRNA significantly correlated with IgG3 and also with IgG1. AID mRNA is a predictive biomarker of optimal in vivo vaccine response.  Indeed, an increase in AID expression  provide important indications for predicting the robustness of antibody response to H1N1 vaccine in both HIV-1 infected patients and healthy controls (Cagigi et al., 2013).

Thus, it might seem that AID mRNA measurement might allow one to understand IgG3 dynamics in the context of HIV-1 vaccine development.