Pregnancy is an outcome microbicide trials want to avoid for a number of reasons —unfortunately, unexpectedly high rates of pregnancy have been observed in several, though not all of these studies — especially in the studies involving women with multiple sexual partners (which are also being conducted in West Africa where contraceptive use is less frequent). Researchers at last month's Microbicides 2006 conference in Cape Town discussed how pregnancy is affecting ongoing microbicide studies.
In the Cellulose Sulfate (CS) #2 trial, for example, there have been 50 pregnancies per 100 person years (py) in Lagos, and 21 pregnancies per 100 py in Port Harcourt for 32 per 100 py overall. “This is much higher than we expected,” said Dr Vera Halpern of Family Health International.
Similarly, in Savvy Nigeria, there have been 40 pregnancies per 100 py. In Savvy Ghana, one site reported 50 pregnancies per 100 py, and the other reported 76 pregnancies per 100 py, with 64/100 py overall.
The frequency of clinic visits, and the sensitivity of some pregnancy kits may partly be responsible for some of these pregnancies. According to Dr Elof Johannson, of the Population Council, the Carraguard study has had “a lower rate of pregnancies than seen in some studies [below 10%]. Our quarterly visit helps because if you have a very sensitive pregnancy test and you test women very often then you are likely to pick up many early pregnancies which may be spontaneously aborted.” Or may in fact, not be true pregnancies at all.
Indeed, a study presented in a conference poster from the University of Pittsburgh addressed the phenomenon of false or “chemical” pregnancies in more detail. A significant proportion of positive pregnancy tests are, in fact, due to chemical pregnancies - a brief rise in human chorionic gonadotropin (hCG) that has little clinical meaning. The study found that different methods of pregnancy diagnosis could have different positive predictive values for clinically relevant pregnancies by comparing three techniques: monthly urine pregnancy tests alone; monthly pregnancy tests followed by a confirmatory pregnancy test at a one-week interval should the initial test be positive; and pregnancy testing following a missed menses only.
According to a presentation by Dr Elizabeth Raymond of Family Health International, one reason to avoid pregnancies in microbicide studies is the safety of the women who become pregnant in the trials and their babies. “Most of the candidate vaginal microbicides are investigational products. Little or no human data exist regarding their safety in pregnancy.” Even though many of the early microbicides are not absorbed or are poorly absorbed, embryos that are conceived while the mother is using the product “may be exposed by direct passage of the product up through the cervical canal to the uterus and of course, sperm are exposed during intercourse, so therefore trials must establish safeguards to protect pregnant women and their babies from any harmful effects that might occur from these exposures.”
The discovery that a microbicide induces abortions or causes birth defects would likely doom its development.
Also, said Dr Raymond, “these high pregnancy rates have implications for the conduct and interpretation of data in the HIV microbicide trials.” Because of the possible safety concerns, most trials have required that women who become pregnant discontinue using the product immediately. Thus, a high number of pregnancies leads to discontinuation and weakens the study. “The current phase III microbicide trials may be underpowered because of the effect of pregnancy on participant retention,” the authors of the University of Pittsburgh study concluded.
Even if the pregnant patient does not discontinue the microbicide, pregnancy can skew study results. According to a presentation by Dr Wes Rountree, “pregnancy in the Savvy Ghana trial was associated with sexual behaviour changes.” The study found that women who discovered they were pregnant reduced their sexual risk taking with fewer acts of intercourse, fewer unprotected sex acts and fewer partners. “These changes in sexual behavior can affect a woman’s risk of HIV acquisition,” said Dr Rountree “which in turn leads to statistical issues including loss of study power and difficulties in the interpretation of study results.”
Indeed, drop-outs or changes in behaviour due to the high rate of pregnancy in the Savvy Ghana trial could be one of the explanations for the low rate of HIV incidence observed in both the placebo and microbicide arm of that study. The study was closed because it was inadequately powered to detect a significant difference between trial arms at such a low rate of HIV incidence.
However, some women may still have sex or may have little choice about whether they have sex while pregnant. These women may still benefit from a microbicide that prevents HIV transmission and being pregnant, could have an even greater desire to protect themselves and their foetuses. Is it right to routinely exclude them from these studies?
Decisions about how to prevent pregnancies in trials and the identification and care of pregnant trial participants have ethical implications.
Nevertheless, said Dr Halpern “we’re trying to do everything possible to bring these numbers down.” In both CS studies, these efforts include providing free effective contraceptives and increased counselling. However, according to Dr Lut Van Damme, who gave a report on CS trial #1, no woman will be turned away if she decides to refuse contraceptives.
Finally, in the MDP301 study, which has seen only 19 pregnancies per 100 patient years, investigators are exploring the barriers to the use of adequate contraception (which could range from religious objections, misconceptions about the safety of contraception, or even her partner's disapproval). They are also considering putting off the confirmatory pregnancy test by one or two weeks to be certain that the woman has a true pregnancy. But instead of excluding women who become pregnant, the team is considering looking at the safety in pregnancy issue earlier than planned and allowing women who become pregnant to continue using the gel in the trial.
Raymond E et al. Issues related to pregnancies in microbicide effectiveness trials. Microbicides 2006 Conference, Cape Town, OB15, 2006.
Tolley E. Ethical issues related to pregnancy during microbicide trials. Microbicides 2006 Conference, Cape Town, Invited talk, 2006.
Rountree W et al. Change in sexual behavior post-pregnancy in the Savvy Ghana trial. Microbicides 2006 Conference, Cape Town, OB17, 2006.
Schreiber CA. The impact of pregnancy on microbicide clinical trials. Microbicides 2006 Conference, Cape Town, PB67, 2006.