The study, HIVNET 012, released last summer, compared the efficacy of two different antiretroviral regimens at preventing the transmission of HIV from mothers to their children. Half the participants received a very short course of AZT, given to the mother at the onset of labor (600 mg) and throughout delivery (300 mg every three hours), and then post-partum to the infant for the first week of life (4 mg/kg twice daily). The remaining women received a single 200 mg dose of the non-nucleoside reverse transcriptase inhibitor (NNRTI) nevirapine (Viramune) at the onset of labor, and their infants also received a single dose (2 mg/kg) within the first 72 hours of life. The two pills cost a total of $4 at current prices in the US.
The rationale for trying nevirapine was more than purely its affordability. Compared with AZT (or 3TC, another drug sometimes used to prevent perinatal transmission), nevirapine is considerably more potent, able suppress HIV replication much more effectively and quickly. It is rapidly absorbed when taken orally, entering the bloodstream and beginning to work almost immediately. It has a long half-life (61 to 66 hours in pregnant women and 45 to 54 hours in babies), meaning that significant amounts of the drug persist and continue to work many hours after the drug is taken. This is particularly important for perinatal transmission because keeping the maternal viral load low for the first days of life may well reduce the risk of transmission from breastfeeding, most of which seems to occur soon after birth. It has also been well established that nevirapine can cross the placental barrier.
The published data analyzes the first 496 babies (of a total of 616) to reach 14 to 16 weeks, by which time 25.1% of the infants whose mothers took AZT were HIV-positive, versus only 13.1% in the nevirapine arm. Thus, nevirapine reduced the risk of infection by 47%. The p-value was 0.0006, meaning that the correlation is highly statistically significant. The difference between the arms could not be attributed to other factors known to influence perinatal transmission, such as maternal health (baseline characteristics were very similar for the two groups, with mothers having an averages of 426 and 461 CD4 cells, and viral loads of 27,902 and 25,198, for the AZT and nevirapine arms, respectively), duration of labor (8.0 and 9.3 hours, respectively), frequency of Cesarean section (slightly above 10% in each group), and breastfeeding (more than 95% in each group breastfed). Likewise, including in the analysis those babies who had died before they could receive the entire treatment regimen or before they could be tested did not significantly alter the difference between the two treatment groups. Happily, few mothers or infants in either group had serious adverse events deemed attributable to the therapies, although it is still too early to know much about the long-term impact of giving children antiretroviral drugs within the first few days of life.
The news that such an easy and cheap treatment is efficacious is obviously exciting. There are, however, a few notes of concern. First and foremost, this is only one trial, and more data are needed to prove conclusively that nevirapine should replace a short-course of AZT as the regimen of choice in the developing world. Indeed, an earlier, much smaller trial (HIVNET 006) by the same group of researchers did not find as promising results, although the sample size was so small as to make comparisons pointless (see Musoke, et al., AIDS, 13:4:479-86). The comparatively short (14 to 16 weeks) follow-up in HIVNET 012 gives one pause when comparing these results to other recent trials in the developing world, which generally followed babies for longer periods to ensure that the benefits of the antiretroviral therapies were not completely offset by continued exposure to HIV via breastfeeding. A much larger study, the SAINT study (South African Intra-Partum Trial), is currently in the process of enrolling mothers and will compare the efficacy of nevirapine and AZT. Preliminary data should be available at this summer's World AIDS Conference in Durban, South Africa.
Some commentators have worried that this was not a placebo-controlled trial, meaning that it is only possible to judge how effective nevirapine was in comparison to a short course of AZT, rather than the local standard of care, which is currently nothing for much of the world. The ethics of placebo-controlled studies of perinatal transmission has been a subject of fierce debate in the medical community (see, for example, the opposing arguments laid out by Drs. Troyen Brennan and Robert Levine, New England Journal of Medicine, 341:527-34, and a set of summary recommendations published in The Lancet, 353:832-5); however, for understanding the import of this study, the absence of a placebo arm is not so crucial. Given that it is highly unlikely that transmission rates in the AZT arm would be higher than those among untreated individuals, it seems fairly safe to conclude that nevirapine would be at least as effective when compared to no therapy. Rates of transmission among women not taking antiretroviral therapies are generally estimated to be between 25-35%. The fact that there was a 25% transmission rate in the AZT arm of HIVNET 012 suggests that such a short course of AZT is not very effective.
One of the main benefits of the short-course nevirapine regimen is that it is administered at the onset of labor. Since prenatal care is often unavailable in much of the undeveloped world, a short-course nevirapine regimen might be implemented more easily than an AZT regimen that requires prenatal monitoring. Also, the short course may help minimize the risk that resistance to nevirapine, which arises rapidly when it is given as a monotherapy, would compromise the success of future course. Still, the regimen may not be simple enough for widespread use in many developing countries, as many women only venture to hospitals or clinics in the event of a complication. Since the maternal dose must be taken promptly at the onset of labor, delaying a visit until after labor has begun may abrogate some of the beneficial effects of treatment with nevirapine.
Another interesting scenario that requires further study is mass use of nevirapine for all pregnancies, regardless of the mother's HIV status, in certain targeted countries, which the researchers suggest may be analogous to providing iron supplementation. Since the cost and safety of the regimen is so preferable, and since HIV infection is so common in some countries, it may make sense to administer the regimen to all pregnant women. This gets around the formidable problem posed by the fact that in most developing countries widespread screening is rarely available (or necessarily advisable, given the persecution people with HIV are still subjected to in many places), meaning that most HIV-positive women are unaware of their serostatus. Mass treatment may also help with the widespread stigma still associated with HIV in many countries. If everyone takes the pills, no one is stigmatized.
A cost-effectiveness study by a group of health economists working with the Makerere/Johns Hopkins team suggests that the universal treatment approach is the most efficient way of preventing perinatal transmission. Since nevirapine itself is inexpensive but counseling and testing programs require considerably greater outlays for infrastructure and staffing, giving the antiretroviral to all pregnant women in countries with relatively high seroprevalence (their model, using conservative estimates of the efficacy of the therapy, suggested that it was cost-effective for countries with seroprevalence rates above 10%) would be highly cost effective. Indeed, in their model, providing the drug would be more cost effective than well-known public health strategies such as vaccination against polio.
Still, the decision to give antiretrovirals to all pregnant women is not one that should be taken lightly, so it was good news that the same group found that counseling and testing along with the targeted provision of nevirapine was also quite cost effective. Even before the HIVNET 012 data was released, another recent cost-effectiveness study (published in BMJ, 381:1650-6) had found that even two other, more expensive regimens used to prevent perinatal transmission (the so-called PETRA and CDC-Thai regimens; see Treatment Issues, February 1999) provide very good bang for the buck. Dr. Neil Söderlund and colleagues examined the cost-effectiveness of providing different regimens of antiretroviral therapies in a clinic setting in South Africa and found that were the South African government to offer free AZT in accord with the CDC-Thai regimen, it would actually save money over the long run, because of the impact of decreased costs associated with hospitalizations and lost productivity. Unfortunately, at a meeting of the AIDS Economics Network in Washington, D.C. on July 27, Dr. Söderlund said that the South African government has heretofore been less than enthusiastic about these findings. Still, these two models provide real hope that forward-looking policy planners in the developing world will now have the ammunition needed to convince national governments to fund programs aimed at providing antiretroviral therapy to prevent perinatal transmission.
The cause of ensuring that the developing world will have affordable access to antiretrovirals to prevent perinatal transmission received a major boost when, at the second conference on "Global Strategies for the Prevention of HIV Transmission from Mothers to Infants," the not-for-profit organization Global Strategies for HIV Prevention launched a "Call to Action" aimed at providing both information and free drugs to countries that would otherwise be unable to afford them. Happily, the Elizabeth Glaser Pediatric AIDS Foundation immediately signed onto the Call, ponied up $1 million to get the project off the ground and took out a full page ad in The New York Times on September 8 to promote the campaign. Further, according to a representative of Boehringer-Ingelheim, the pharmaceutical company that holds the international patent rights to nevirapine, various multinational organizations, including UNAIDS and the WHO, and numerous smaller non-governmental organizations have already expressed interest in nevirapine-based prevention programs. However, the process of getting the drug to the countries where it could make the largest impact has been slowed by the fact that nevirapine was not registered in many countries prior to the HIVNET 012 breakthrough, meaning that there is a lot of red-tape to be cut through before the drug can be legally sold. If these first steps are representative of a growing political commitment to end perinatal transmission, then there truly is hope that the advances seen in preventing the infection of infants in the US can be replicated worldwide.
Last February, we described a French study that had found rare and fatal neurological complications in two infants enrolled in a trial examining the efficacy of using AZT and 3TC to prevent mother-to-child transmission (Treatment Issues, February 1999). Concerns about the possibility of a lethal side effect related to the use of antiretrovirals sent a shock wave through the community of researchers working on this issue and prompted a review of data from the numerous National Institutes of Health and Centers for Disease Control and Prevention. In reports presented at various summer conferences, including the 27th AIDS Clinical Trials Groups Meeting in July in Washington, D.C., "Global Strategies for the Prevention of HIV Transmission from Mothers to Infants" in Montreal and at last year's ICAAC in San Francisco, it was announced that of the more than 15,000 infants known to have been exposed in utero to antiretroviral therapies in US studies of perinatal transmission, none seem to have died as a result of drug toxicity.
However, it is clear that the last chapter of this story has not yet been written. The French group is particularly concerned that the antiretrovirals are damaging the body's mitochondrial DNA, resulting in the neurological symptoms seen in the two fatalities, as well as problems with muscle, colon, liver, pancreatic and other organ function. They have found six additional infants who have evidence of such mitochondrial toxicity, although all are still alive. The US cohort data will now be examined for signs of similar non-fatal mitochondrial defects.
Another report presented in both Montreal and San Francisco discussed the pros and cons of Cesarean sections. While it appears clear that elective C-sections -- those done in a planned, rather than emergency, manner -- can reduce the risk of transmission, especially in women who have not taken antiretroviral therapy, the benefits for those women who have taken antiretrovirals may not outweigh the risks the surgery carries, which include fevers and infection. It is clear that antiretroviral therapy is the safest and most effective way to reduce the risk of mother-to-child transmission, but Cesarean sections are a reasonable option for women who are aware of the risks but who want to further reduce the chance of transmission.
Taken together, the recent findings sketch a very hopeful picture of the future of perinatal transmission: cheap and safe treatment made available to all in need because governments have realized that it is cost-effective to do so. With a little luck and some political backbone, most of the 12 million HIV-positive women of childbearing age will also live to see such a day.
Back to the GMHC Treatment Issues Winter 1999-2000 contents page.