The latter trial, which compared two formulations of intravenous immunoglobulin -- the standard formulation and one prepared from donors known to have high levels of HIV titers (HIVIG) -- in women who were also receiving ZDV, showed that treatment reduced the overall vertical transmission rate to just 5%. The subjects recruited for this trial represented a more heterogeneous population than the one studied in ACTG 076. Most notably, a significant proportion had CD4 counts below 200 cells/mm3 and/or had been treated with ZDV before they became pregnant -- whereas all of the participants in ACTG 076 were antiretroviral-naïve and had CD4 counts in excess of 200 cells/mm3.
The addition of HIVIG to a regimen of ZDV monotherapy had no demonstrable impact on HIV transmission rates, and the preliminary findings indicate that HIVIG also failed to reduce plasma HIV RNA levels. Although the data from ACTG 076 did not establish a clear correlation between HIV RNA levels and likelihood of vertical transmission, the one effective agent in the ACTG 185 regimen -- ZDV -- did have a measurable effect on viral load. This finding suggests that the ability to reduce viral burden should be regarded as a prerequisite for testing any other proposed interventions.
By now, resistance of ZDV is well documented, and clinicians should expect to see viral resistance develop eventually in all patients treated with ZDV, especially when the drug is used as monotherapy. However, the data from ACTG 076 reveal that the induction of genotypic resistance to ZDV was rare in this patient population, all of whom were ZDV-naïve at the time they entered the study. As importantly, the genotypic resistance that did develop in these patients was not associated with a measurable increase in the risk of vertical transmission of HIV infection.
These findings are expected to hold for the women enrolled in ACTG 185, based on the low transmission rate. Prior treatment with ZDV has led to the development of genotypic resistance to the drug in these women, but this resistance is infrequent and is unlikely to have an impact on transmission rates. Clinicians should therefore continue to make ZDV a component of antiretroviral regimens prescribed to pregnant women -- in the hope of preventing vertical transmission of HIV (see the PULL OUT AND SAVE feature in this issue, "U.S. Public Health Service Recommendations for the Use of Antiretroviral Drugs During Pregnancy").
At present we have only minimal data on the efficacy and safety of antiretroviral agents other than ZDV in pregnant women and children (see the two related articles in this issue, "Antiretroviral Therapy in Pregnant Women" and "Antiretroviral Therapy in Children"). This is an unfortunate situation, because common sense and medical ethics both tell us that seropositive women who become pregnant should receive optimal antiretroviral therapy despite their pregnancy.
This paucity of information reflects medicine's traditional proscription against including pregnant women in clinical studies, a holdover from the days when our understanding of fetal development, and our capacity to monitor that development, were considerably more limited than they are today. It fails to reflect the need to maintain optimal maternal health -- which is essential to optimal development of the fetus. It also fails to acknowledge the urgent need to treat a potentially fatal infection in these women.
This situation puts clinicians in a terrible bind: Ethics dictate that we exclude pregnant women from clinical trials, so we have no data on how these women respond to antiretroviral therapy. But ethics also dictate that we provide our pregnant seropositive patients with the best possible medical care, so we find ourselves in the position of prescribing antiretroviral regimens for them in the absence of data from clinical trials.
The paradoxical result of this combination of long-standing tradition, contemporary caution, and ethical obligation is that appropriate antiretroviral therapy is being withheld from some seropositive women, even while others are being treated with drugs for which we have, as yet, no adequate pharmacokinetic data. Fortunately for all concerned, studies are currently being conducted that will help clinicians make decisions about the care of HIV-infected women. We now have data on the pharmacokinetics of ZDV and 3TC in pregnant women, and we will soon have similar data on other agents, among them the protease inhibitors and Glaxo Wellcome's promising new nucleoside analog, abacavir.
What we do not yet have is comparable data on the safety and efficacy of the protease inhibitors and abacavir in newborn infants. If these drugs are to be used in pregnant women -- to treat HIV infection and/or to interrupt vertical transmission -- it is essential that we know what effect they will have on neonates. Although some progress has been made in this regard, pregnant women and infants remain the groups for which we have the least concrete information. Put bluntly, it is both inappropriate and unfair to selectively delay the delivery of optimal therapy to any patient population, either because they present particular problems or because they have the smallest voice.
Those who treat HIV-infected women have come to the conclusion that all of these women should receive antiretroviral therapy that achieves maximum suppression of viral replication -- regardless of whether they are pregnant. Prudence dictates that any regimen chosen for pregnant seropositive women should include ZDV, because of its proven utility in diminishing mother-to-infant transmission of HIV infection.
Before initiating any form of antiretroviral therapy in pregnant women, clinicians need to inform these patients about the range of therapeutic options now available to seropositive individuals. This discussion must encompass the known benefits -- and risks -- of various therapeutic combinations, and it must also cover what it not known about the safety and efficacy of such combinations.
Women who have never taken antiretroviral agents may elect to delay therapy until after the first trimester (14 weeks) of gestation, to minimize the theoretical risk of teratogenicity. Women who are already on combination therapy when they become pregnant may elect to discontinue therapy during the first trimester. If they do, they should stop all of their drugs simultaneously. When therapy is reinitiated, all drugs should be administered simultaneously. This will minimize the chances that resistance will develop to a particular drug or class of drugs.
We know for certain that ZDV therapy reduces the risk of vertical transmission of HIV infection. What we don't yet know is why. Given the modest reductions in viral load that are achieved with ZDV monotherapy, this alone cannot explain the magnitude of the decrease in transmission seen in ZDV-treated patients. It seems that administering the drug to the mother -- so that it is present in the infant at the time of exposure to the virus -- is essential, if transmission is to be interrupted. Several small studies indicate that transmission rates are significantly higher if ZDV is given only to infants born to seropositive mothers and not to the mothers themselves.
This finding is consistent with the observation that 30% to 50% of HIV-infected infants have detectable levels of virus in their blood within 48 hours of birth. Of the remainder, more than 90% have detectable virus within two weeks of birth. These facts suggest that most infected infants acquire infection around the time of delivery; if infection occurred earlier, it would be possible to detect the virus in virtually all babies at birth. And this, in turn, leads one to hypothesize that it is the presence of ZDV in the neonate's blood and tissues around the time of delivery, when infection is most likely to occur, that accounts for the drug's high rate of success in interrupting transmission of HIV. For ZDV to be present in the fetus's blood at this crucial juncture, the mother must be treated, at least during labor and delivery.
All HIV-infected women should receive optimal antiretroviral therapy. Many of these women are diagnosed during pregnancy precisely because it is pregnancy that brings them into contact with healthcare providers. Consequently, all pregnant women should be counseled about HIV infection and offered testing. Those who do test positive should be strongly encouraged to begin antiretroviral therapy with a regimen that includes ZDV -- for their own sake, and for the sake of their unborn child. This is now the standard of care for pregnant, HIV-infected women, and there is no acceptable excuse for failing to provide such care.
The total number of new cases of pediatric AIDS declined in 1996, the first time such a decline has been reported since the C.D.C. began keeping statistics on the epidemic. This heartening trend was noted both in hard-hit urban centers like New York City, San Francisco, and Miami, and in largely rural states like Wisconsin and North Carolina.
Although we are a long way from our goal -- which is to eliminate vertical transmission entirely -- the modest reductions already achieved have saved hundreds of millions of dollars in healthcare costs. Indeed, it has been noted that the administration of ZDV to pregnant women, an outgrowth of ACTG 076, has resulted in savings to the healthcare system that exceed the total cost of the entire pediatric AIDS Clinical Trials Group program. More importantly, the use of ZDV in these women has saved hundreds, and will save thousands, of young lives -- in all, a very wise investment in our future.
Catherine M. Wilfert, M.D., is Professor Emerita, Division of Infectious Diseases, Department of Pediatrics, Duke University Medical Center, Durham, NC.