NIH-funded study finds benefits far outweigh risk
Listen to this podcast (MP3 - 16.5 MB).
Barrett Whitener: The antiretroviral drugs used to keep HIV at bay are an unqualified success at preventing the spread of the virus from mother to child. The drugs are not only essential for maintaining the health of a pregnant woman with HIV, they have also nearly eliminated the transmission of HIV to her baby. Among U.S. women, the likelihood of a mother passing the virus on to her child is now less than 1%.
But there have always been concerns about the effects those drugs could potentially have on the developing fetus. Adding to those concerns is the fact that women treated for HIV are now healthier than before, so are more likely to become pregnant.
Now a new study reaffirms the safety of these lifesaving drugs during pregnancy. The researchers reported no increased risk of birth defects for most anti-HIV drugs. Although one drug, Atazanavir, was found to increase the risk, that risk was very slight. Atazanavir is marketed under the name Reyataz. Overall, the researchers concluded that the benefits of antiretroviral drugs during pregnancy far outweigh the risks.
From the National Institutes of Health, I’m Barrett Whitener. This is “Research Developments,” a podcast from the NIH’s Eunice Kennedy Shriver National Institute of Child Health and Human Development, the NICHD.
Joining me today is Paige Williams, lead author of the new study, which was funded by the NICHD and appeared in the Journal of the American Medical Association, or JAMA, Pediatrics. Dr. Williams is senior lecturer on biostatistics at the Center for Biostatistics and AIDS Research at Harvard University’s School of Public Health.
Thank you for joining us today, Dr. Williams.
Paige Williams: Oh, thanks so much for inviting me to speak with you.
Mr. Whitener: Earlier studies have also looked at this possible connection between the mother’s taking anti-HIV drugs and potential consequences for the fetus. What questions still needed answering?
Dr. Williams: Well, first of all, as you mentioned, the benefits of combination therapy are really dramatic, and so much so that now the WHO [World Health Organization] recommends that all pregnant women with HIV start combination treatment during pregnancy and then remain on it for life. So combination treatment usually includes at least three different antiretroviral drugs, which work in different ways to keep levels of the virus very low in the body.
And because more and more women are receiving treatment with these multiple antiretroviral drugs, we felt that there were some questions that still needed answering.
Despite the clear benefits of these treatments in preventing transmission, as you noted, we also want to make sure that these drugs are safe both for the mother and the baby. Past studies have looked at lots of different infant outcomes, including how well the infants grow, how well they learn, hear, and develop. But we felt that one of the most critical outcomes is congenital defects, or birth defects. And this is because of the major impact that these would have on the health and the quality of life, both for the infant and their family.
So the reasons that we felt it was important to conduct our study is, first of all, a lot higher percentage of women now receive combination ARVs [antiretrovirals] early in their pregnancy, during the first trimester when organ systems are first developing and probably most vulnerable. So this is considered a critical window of exposure.
Secondly, several new drugs have been approved within the last 10 years, and the prior studies weren’t able to previously evaluate these adequately because there just wasn’t a high enough percentage of women who had been using those drugs. Most of the past studies only included babies born before 2007, but some of the new drugs approved around 2000 are now widely used.
There’s also the possibility that two or more drugs might act together synergistically to create a higher risk than if either one of these drugs were used without the other. So we felt it was important to look at combinations of drugs.
So overall, because the patterns of use of these drugs have changed over time, we thought it was important to take a new look at the risk of birth defects. And so our study, which is called the SMARTT study—and that’s SMARTT with two Ts; it stands for Surveillance Monitoring of ART [Anti-Retroviral Therapy] Toxicities—it’s being conducted by the Pediatric HIV/AIDS Cohort Study Network, and it’s funded by NICHD, along with many other NIH co-funders.
We have the largest current cohort of HIV-infected women and their babies or children in the U.S., with over 3,000 mother-child pairs. Some of these children are now up to 18 years old and have been followed from birth. The main purpose of our SMARTT study is to monitor the safety of these ARV drugs that mothers take during pregnancy, and this study of birth defects is just one of our many studies.
Mr. Whitener: Of course, we don’t want to alarm any of our listeners, and from reading your paper, I saw that you and your colleagues described the birth defects you did find as generally of milder severity. But in general, among those that you did find, can you describe them for us?
Dr. Williams: Yes, certainly. So we used a previously developed system for classifying birth defects called the Metropolitan Atlanta Congenital Defects Program, or MACDP. And this system has been used by several of the other studies, particularly by the Antiretroviral Pregnancy Registry, which is an international registry for birth defects. So this registry has specific criteria for defining major birth defects within different organ systems, and some of the types of defects include cardiovascular, musculoskeletal, central nervous system, genital, skin-related, and other categories.
So in our analysis, we treated all major birth defects as an outcome. But some major defects are more severe than others, and this is a limitation of this system. For example, cardiac defects, such as a defect in the wall of the heart, can have pretty severe consequences, with higher risk of death of the baby and often multiple surgeries required. And although rare, these types of defects are really of major concern.
In contrast, certain defects such as skin defects or polydactyly, which is the presence of an extra finger or toe, they may not have the same adverse effect on a child’s life. So these types of defects might be considered as a congenital anomaly of milder severity. So in our analysis to address this limitation, we evaluated each category of defect separately, but this still may not separate the most severe defects from those which are milder. And as you noted previously, the particular drug where we did see an increase in risk, Atazanavir, tended to be associated more with the milder severity defects, like polydactyly and skin defects.
Mr. Whitener: You mentioned earlier about using the so-called SMARTT study approach. Can you tell us a little bit more about how you conducted the study and specifically what you were looking for?
Dr. Williams: Sure. So first we assembled a team of experts in our study to review the potential birth defects that had been reported, and these were defects reported up through June 2012. So at that time, we had over 2,500 mothers and children enrolled. And our team of experts included OB-GYN experts, pediatricians, and infectious disease specialists, most of whom had been treating mothers and their children with HIV for 20 to 30 years.
Using the criteria from the MACDP, which I mentioned earlier, we classified each defect as major, minor, or not a defect. Some of the children had multiple defects, especially children with chromosomal defects like Down’s syndrome. But during our review, the team reviewed all defects for every child, and we were blinded to what ARVs the mothers took during pregnancy so that we couldn’t be biased by any preconceived ideas about risk for specific ARV drugs. So we didn’t know what the mother had taken when we were reviewing the birth defects.
This was a very extensive and thorough process, and it took several years to complete. But once we completed our reviews, we then used the data to evaluate the rate of birth defects among those exposed and unexposed during the first trimester—so early in pregnancy, the first 3 months—to each ARV drug. And we also evaluated certain combinations of ARV drugs which are commonly used.
And we adjusted for the factors that we thought could be confounders. This adjustment for confounders is really important in this type of observational study, because it’s possible that mothers who received a specific drug or combination of drugs are different from those who don’t receive that drug.
Some of the characteristics that we considered as possible confounders were the mother’s age and race, the BMI of the mother before pregnancy, and use of substances like alcohol, tobacco, or other drugs during pregnancy. We also took into account the mother’s health during pregnancy.
It turned out that most of the factors that we looked at were not actually confounders, so they were not included in our adjusted models. But calendar year and the mother’s health in terms of her immune system early in pregnancy, based on her CD4 count, were included in the adjusted models.
Mr. Whitener: So those did make a difference?
Dr. Williams:Yes, those two things did make a difference.
Mr. Whitener: Let’s talk about what you found. Overall, the rate of birth defects was generally low, wasn’t it?
Dr. Williams: Yes, it was generally low. The overall rate was 6.8%, and although that’s relatively low, it is higher than what most of the previous studies have found. We speculated that this could be partly due to the fact that we’re looking very carefully for birth defects, so we had this specific requirement in our study that every child had to be evaluated for birth defects. This might have resulted in higher reporting of birth defects or potential birth defects.
Another possibility is that there have been increasing rates in the general population reported in the literature. So this might be reflecting sort of a general underlying pattern.
We did observe a significant increase over time in the rate of defects in our study, although the rate then dropped slightly in children born after 2010. So I think this issue still warrants further investigation.
Mr. Whitener: What are some possibilities that you’re considering for that increase and then fall-off after 2010?
Dr. Williams: Well, we actually had two different cohorts in our study. One was the static cohort, and this included children who had generally been rolled over from a prior study. And it’s possible that those children in the static cohort had mothers less likely or more likely to participate in our study than those who were followed in the dynamic cohort, which is a cohort that was followed in our study from birth. So that was one possibility of potential selection bias—that the mothers may have been more or less willing to participate if their child had a specific problem than those whose children did not have any problem. So that’s something that, in our further investigations, we’ll have to look more carefully to see whether the rate continues to drop or holds steady or increases.
But I think the important finding is that we did not see a significant increase in the overall rate of defects for those who were exposed to ARVs in the first trimester compared to those who started treatment later. And I should point out that almost 100 percent of these women received some treatment during pregnancy. There was only 1 to 2%who received no treatment at all. So our comparisons are really looking at those who started early in pregnancy versus those who started later. And this is very reassuring. I think it provides good support for the current guidelines of treating women with HIV during pregnancy.
Mr. Whitener: You mentioned that one of the main questions you had was whether the rates of birth defects changed as some of the newer anti-HIV drugs were used. What did you find in that regard?
Dr. Williams: Yes, so in terms of the newer ARV drugs, we found some associations for children exposed to two different drugs--Atazanavir, which goes under the name Reyataz, and Ritonavir, which goes under the name Norvir, and is often used in combination with drugs in a class called protease inhibitors. Ritonavir is used at a lower dosage to boost the levels of the other drugs that are used in combination with it.
So those two drugs were linked to higher rates of birth defects. And even when we adjusted for calendar year and the mother’s immune health, they remained linked to prior rates of birth defects. So although the overall risk of birth defects was low, the relative increase that we observed with Atazanavir, which had about twice the risk, is notable.
When we looked at specific types of defects, as I mentioned earlier, the associations for this particular drug were strongest for musculoskeletal and skin defects, and many of these defects were considered a milder type of defect. Nevertheless, this suggests that some additional attention be given to look at the safety of Atazanavir exposure early in pregnancy.
Mr. Whitener: Well, this is, as you say, all extremely reassuring. Given what you found in the study, what do you think the next phase of this research may be?
Dr. Williams: Well, we’re certainly continuing to monitor birth defects in our SMARTT study, and we will need to update our analyses at some point to take into account the additional years of information that we’ve collected since 2012, which is when we cut off the information for our published study.
However, there’s also lots of other safety outcomes that need to be addressed. Two of our other studies evaluating language skills in infants also found an increased risk of language delays with the same drug, Atazanavir. So that is really highlighting the possible safety signal with that drug, and so we’re carefully looking at that drug in other ways.
Another important study has looked at Tenofovir exposure. Tenofovir is marketed under the name Viread. And this drug has been linked to problems with bone development. And we did a study that was a sub-study of SMARTT to see if it was linked—exposure during pregnancy was linked to lower bone mineral density in newborns. And that study was led by George Siberry, who’s also at NICHD. Findings of that study of lower bone mineral density in newborns suggests further evaluation in older children should also be done.
Mr. Whitener: So for expectant mothers or women who are thinking about becoming pregnant who have HIV, it sounds like overall, the news is very good at this point.
Dr. Williams: Yes, I think so. I found our results were very reassuring. We really didn’t see any individual drugs in two of the drug classes with any association with higher risk of birth defects. And I think even the finding of Atazanavir exposure really needs to be confirmed in further studies. So overall, I think the low risk and lack of associations with the individual drugs that we looked at is very reassuring.
Mr. Whitener: Okay, thanks so much for speaking with us today.
Dr. Williams: Thank you.
Mr. Whitener: I’ve been speaking with Dr. Paige Williams, lead author of the study, “Congenital Anomalies and In-Utero Antiretroviral Exposure in Human Immunodeficiency Virus-Exposed Uninfected Infants.” The study was published in JAMA Pediatrics.
About the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD): The NICHD sponsors research on development, before and after birth; maternal, child, and family health; reproductive biology and population issues; and medical rehabilitation. For more information, visit the Institute's website at http://www.nichd.nih.gov/.