Sunday, July 25, 2010

Was the Tenofovir Gel Microbicide Trial Ethical?

Was the trial ethical and were the results valid? The two questions go hand in hand.

It's quite hard to decide if there was something unethical or invalidating about the Tenofovir Gel Microbicide trial because there could be important details about the trial that have not yet been published. For a start, the trial assumes that most HIV transmission among the participants is sexual. But did the researchers involved actually establish that HIV transmission was all or even mostly sexual? If they did, they haven’t said so in the paper.

The question is important because, during the course of the trial, researchers would have had the opportunity to find out, for each participant who became infected with HIV, how they became infected. One might say they had a duty to find out how they became infected. Were the partners of all the women who seroconverted HIV positive? I don't believe this data was collected. If it was, it should have been published because those who were infected non-sexually should have been excluded from the results. The gel is supposed to protect against heterosexually transmitted HIV, not, for example, HIV transmitted by unsafe medical procedures.

Participants were given the gel as a prophylactic against HIV. They were given to understand that they would be protected against HIV infection. They wouldn't have been told that they were protected against non-sexual HIV transmission. But they don't appear to have been warned about the possibility of non-sexual transmission. If the researchers didn’t exclude the possibility of non-sexual HIV transmission, and they don’t appear to have done so, the results are of questionable validity.

Nearly 900 people were recruited to take part in a trial and it was known that some, perhaps a lot of participants, would become infected with HIV before the end of the trial. It was not known how many would become infected or which participants. Perhaps non-participants in the area may face an even higher risk of becoming infected. But I don't think that excuses those running the trial for failing to ensure the safety of those taking part, or for failing to establish the cause of each HIV infection.

Participants were "provided with comprehensive HIV prevention services (HIV pre- and post-test counseling, HIV risk reduction counseling, condoms, and STI treatment), reproductive health services...". They were made aware of the risk of being infected with HIV sexually, but not non-sexually; this is not comprehensive. But despite this preparation, which surpasses the level of prevention available to most people in most African countries, HIV transmission rates were extremely high.

Why, given all these precautions, including very high condom use, were transmission rates so high? And why did the researchers not make any attempt to find out how people were becoming infected? Did they not have a duty to find out if their gel could even have had any influence on rates of HIV infection, or what level of influence it could have had?

In fact, it is not accurate to say that "All women were counselled on the risk of HIV and encouraged to use condoms at all times..." as one report said. All women were counselled on risks of *sexual* HIV transmission. Condoms wouldn't have been much use if any of the women were infected non-sexually.

A BBC article inadvertently puts its finger on the problem: "A vaginal gel has significantly cut the rate of women contracting HIV from infected partners...". But the trial has not established if partners were infected or uninfected. So we don't know if the vaginal gel has achieved this, or exactly what it has achieved.

The same article goes on: "Such a gel could be a defence for women whose partners refuse to wear condoms." But reported levels of condom use were very high and they increased during the course of the trial. Condoms should be far more effective than they appear to have been during this trial. Condom manufacturers must be asking if it is safe to use the gel when using condoms or if the gel actually weakens them or causes them to burst. I certainly hope they are asking these questions.

The article cites one of the researchers as claiming: "Women who used the gel more consistently were much less likely to be infected…". But women using the gel more consistently also had less sex. And those having more sex tended to use the gel less consistently. It's hard to know exactly how to interpret the results of this trial. If the results were truly valid and significant, how would we know? The research seems to be incomplete in many ways.

Executive director of UNAIDS, Michel Sidibe, is quoted as saying: "For the first time we have seen results for a woman-initiated and controlled HIV prevention option." Well here's another one: Patient Observed Sterile Treatment (POST). In order to protect themselves from non-sexual HIV transmission, people need to be made aware that HIV is not always transmitted sexually and of the precautions they can take. Concentrating on sexual risk alone leads to the sort of travesty that this gel trial appears to have been. If the Executive director of UNAIDS is not aware of this, he should resign.

Can we say clearly that the results of the trial could not have been due to chance? Technically, the result is statistically significant, other things being equal. But we just don't know enough (or are not being told enough). The paper notes that, "Overall, condoms were reportedly used in 80.3% of sex acts; increasing from 78.5% in the first 6 months to 84.3% in months 18-24". This means that as condom use went up, the efficacy of the gel seemed to go down.

The paper goes on "we observed declining HIV incidence rates in the placebo gel arm. This may have been due to their declining coital frequency and increasing condom use." So, are they saying that over time, the efficacy of the placebo gel increased as the efficacy of the Tenofovir gel decreased? That seems like a strange result. If the trial had gone on for five years, would these two trends have continued? Not only is the trial questionable ethically, it is also of questionable validity.

The trial could be unethical because the researchers would have known in advance, or should have known in advance, that at least some HIV is transmitted non-sexually. Yet they made no effort to protect people from this. Nor did they make any effort to inform people that such a risk exists. And the results are of questionable validity because we don't know how people became infected and, therefore, why some people didn't become infected. There could have been many factors that increased the risk of infection and others that decreased the risk. Tenofovir gel may have had some effect, but we don't know what effect it had or why it had an effect.


No comments: