CROI: Microbicide failure poses more questions than answers

The recent trial of the candidate microbicide cellulose sulphate (Ushercell) was closed because of 35 seroconversions amongst trial subjects, all of which happened at the three African sites of the trial.

The trial, run by CONRAD, a joint initiative of Eastern Virginia Medical School and the US Agency for International Development (USAID), was stopped on 27th January this year when the trial’s Independent Data Monitoring Committee noticed a higher rate of HIV infections in the women using the microbicide than in those using the placebo.

At the same time another trial using cellulose sulphate run by Family Health International in Nigeria was also stopped as a precaution, though FHI had not seen the same rate of seroconversions.

In an extraordinary presentation at the Fourteenth Conference on Retroviruses and Opportunistic Infections that posed more questions than it answered, the trial’s Principal Investigator, Dr Lut van Damme, said that “all of the seroconversions had been in the same direction” though she also said that until the data had been properly analysed she would not reveal specifically how many seroconversions had been in women using the microbicide and how many in those using the placebo.

The trial was stopped when the probability of the findings reached 0.1, in other words when it became clear that there was a less than one in ten chance that the greater rate of seroconversions in the women receiving the microbicide was a random effect.

The fact that the seroconversions all happened in the African sites (or maybe even in one site), despite the fact that local HIV prevalence appeared to be at least as high in one of the two Indian sites, may indicate that some site-specific factor led to the apparent failure of the microbicide.

CONRAD’s Dr Gustavo Doncel told the conference that extensive preclinical testing had not revealed that the 6% cellulose sulphate (CS) gel was in any way harmful to cells in the vagina, created no inflammation or induce any pro-inflammatory cytokines, and was harmless to naturally occurring bacterial ‘flora’ such as the lactobacilli.

It blocked attachment of HIV to both CD4 receptors and CCR5 co-receptors and had been found to be active against HIV from subtypes A through to G and HIV-2. It was also active against other sexually transmitted infections including herpes, HPV and Chlamydia and had some contraceptive potential.

Safety studies had largely compared the safety of the CS gel with KY jelly, though one had compared it with the spermicide Conceptrol.

A total of 318 women had taken part in the safety studies before the HIV prevention study started and one of the safety studies had lasted as long as six months.

In the HIV prevention trial 2,574 high-risk heterosexual women were to be randomised 1:1 to receive either the microbicide or an inert gel. If CS proved to be 50% effective at stopping HIV infection, this trial would have an 80% power to demonstrate efficacy if 80% of participants completed the trial. If HIV incidence was the expected 4 per 100 trial subjects per year, it would need 66 seroconversions to prove statistically whether the microbicide was superior to the placebo.

The trial was taking place at five sites: at Cotonou in Côte d’Ivoire, Kampala in Uganda and Durban in South Africa, and at Chennai in India and at Bagalkot, a remote rural part of Karnakata state near Bangalore in India.

There was a remarkably high HIV prevalence in the women screened for the trial. By the time of the trial closure, 2733 women had been screened, and 1,400 had been excluded, 1,020 of them (37% of those screened) because they turned out to have HIV.

HIV prevalence was 56% at the Durban site - but also 45% at Bagalkot in India.

By the time of the trial closure 1,333 women had been enrolled though only 326 had completed twelve months. The fact that 3.8% of the women enrolled had already seroconverted by this time may indicate an HIV infection rate considerably in excess of the 4% forecast for the study – even though Dr van Damme said that participants had received safer sex counselling and free condoms at every visit and that condom use had been “intensive”.

Dr Van Damme said that the last follow-up appointment would be on 31st March and there would be closing visits for all participants in April. However she did not expect a full data analysis of the trial to be completed till autumn this year.