HIV Microbicides and the Risks of Clinical Trials

Contributed by guest blogger: Julia Ding ’12

Once preliminary studies suggest that a drug is safe for human use, clinical trials are conducted in order to further investigate the effects and possible adverse reactions of the drug. The example of HIV microbicides has shown that caution and careful scrutiny is highly important for these trials. HIV microbicides are chemical entities which, when applied before vaginal or rectal intercourse, prevent the transmission of the virus. Of the potential microbicide agents that have been studied, two compounds classified as polyanions were thought to be promising for inhibiting HIV-1 transmission: carrageenan and cellulose sulfate (CS). However, these compounds were deemed in phase III clinical trials to be ineffective as microbicides.

In addition to that discovery, the more surprising and disturbing result of these trials was that the HIV microbicides appeared to actually enhance the rates of HIV infection. Pirrone and colleagues examined the validity of this claim in a study reassessing the in vitro activities of the compounds. Cells were infected with different strains of HIV-1 in the presence of three different polyanions: CS, λ-carrageenan (LC), and destran sulfate (DS). Resulting assays showed that all of these compounds exhibited antiviral activity against both R5 and X4 HIV-1 strains. However, further experiments also discovered that application and removal of polyanion microbicides prior to HIV exposure enhanced and increased the rates of HIV-1 infection. The compounds were added to cell cultures and washed out prior to HIV-1 infection to simulate the natural loss of the compound after vaginal application. In both HIV-susceptible cells and regular human cells, the results indicated an increase in the percentage of cells infected, unrelated to any change in cell viability. The level of enhancement was found to be dependent on the target cell, its co-receptor phenotype, the compound identity and concentration, and the timing of the viral challenge. While the mechanism through which HIV-1 transmission increased in the in vitro experiments is still unclear, these factors suggest that the nature of the host cell also plays a role in polyanion-dependent HIV-1 infection.  This data provides a discouraging outlook on the use of these compounds as effective microbicides, while introducing new questions about its mechanisms of action.

This study provides us with many valuable insights about not only the microbicide technology itself, but also the risks and complications associated with clinical trials. The data suggested a significant increase in HIV-1 infection after the application and removal of the two microbicides. Furthermore, it emphasized the need for intense scrutiny of compounds prior to clinical trials, considering the dangers they may pose on human subjects. While previous studies supported the use of polyanion microbicides as a safe and possibly effective means of preventing HIV-1 transmission in women, the effects of the leakage and loss of the product over time was not taken into consideration, and significantly more women on the drug were found to have contracted HIV than if they had not taken it. The study also provides us with an example of the vital role clinical trials play in the testing of a drug, and how certain adverse effects may be missed through in vitro studies that only become apparent when applied to real world uses.



Julia Ding is a senior at Vassar College, with a major in Science, Technology and Society.


7 thoughts on “HIV Microbicides and the Risks of Clinical Trials”

  1. Thank you for your interest in my blog post, but I’m not quite sure where I misstated data from a CG trial. The study that I reference which was stopped prematurely due to safety concerns was testing a 6% cellulose sulfate gel, not a carrageenan-based compound. Is there a peer-reviewed article that you could refer me to so that I can better understand the results from your study?

  2. Ally, those are some really interesting points you’ve brought up! After reading your comment, I went back to the Van Damme paper, and it turns out that there was only one significant difference between the 6% cellulose sulfate gel and the placebo gel: The active gel had a pH of 7.5 , while the placebo had a pH of 4.4. While the placebo gel was within the typical range of vaginal pH, the CS was not. It’s unfortunate that the paper does not go into detail about what affects this difference may have, but it would be an interesting factor to examine in future studies.

  3. Julia Ding’s post is based on a Virology Journal article which misstates data from the Population Council’s Carraguard (CG) trial. The Population Council is reaching out to authors Pirrone and her colleagues to request a correction as well.

    In fact, our study found that among more than 3000 women using Carraguard, there was no increased infection with HIV. HIV was reduced in the CG arm, although not statistically significantly different from the placebo arm.

    There were no safety-related differences between women using Carraguard and women using the placebo.

    This short piece provides an overview of the Carraguard findings:

  4. Marianne – thanks for your comment, and its great to see Vassar alums on this site! I see you are listed in the acknowledments section of the NEJM paper. That paper reports 25 women using CS became infected with HIV during the course of the trial, compared to 16 using the placebo. My understanding is that the preliminary pre-publication analysis suggested a statistically significant difference leading to termination of the trial, although this difference did not turn out to be significant in later analysis. And the comparison here is to the placebo group, not the general population. So I take it from your comment that the placebo group, receiving counseling and condoms, has a lower rate of infection than the general population. However, adding in CS did not further reduce the infection rate. Am I interpreting it correctly?

  5. While you raise some very interesting points, the sentence that states that “significantly more women on the drug were found to have contracted HIV than if they had not taken it” is not correct. The baseline prevalence for HIV in South Africa runs between 10-45%, depending on the person’s age, location, etc., with incidence rates ranging from 5-15% at the clinical sites used for these prevention trials. Every person in the trial is told to use a male condom with each act of intercourse in addition to their assigned microbicide gel, active or placebo. If the same number of women were followed in daily life and used no product, the number of HIV events would have been much higher over the same period of time. I was personally involved in the CS trial and the only positive, proof-of-concept microbicide gel trial to date, tenofovir vaginal gel. These trials are extremely difficult to carry out but are done so with great care and with the highest ethical regard for the participants.

  6. While the use of polyanion microbicides is interesting and contains potential useful results for future medical research the ethics of this experiment considering that for many cases it actually increased the chance of HIV.

  7. The experiments discovers the application and removal of polyanion microbicides prior to HIV exposure increases the rate of HIV infection; could this be due to a change in pH?
    Microbicides would protect by inactivating HIV or preventing attachment, entry and replicating in target cells. Would it also lead to dissemination from target cells present in semen or already infected cells within the female reproductive tract? The vagina and semen have different pHs, therefore could the microbicides cause any imbalance in vaginal pH leading to an increased rate of infection? When the microbicide is removed, would pH imbalance result in a temporary increased exposure rate?

Comments are closed.