Striking a better bargain with HIV: new interventions needed to save infants and to spare mothers.

During roughly every minute you spend reading this article, a mother somewhere in the world will pass on to her infant HIV, the virus that causes AIDS. Damming this current of HIV transmission would be rather simple from a medical standpoint. But it requires either a daunting financial cost or a Faustian biological bargain in which a mother acting in her infant's best interests may reduce her own later chances of avoiding full-blown AIDS. Researchers are searching for a better deal, but with every passing minute, another newborn gets the virus.

In prosperous countries, a woman with H1V can begin a lifelong course of drugs that not only increases her life span but also cuts her chance of passing the virus on to her baby during pregnancy, birth, and breast-feeding from 25 to 40 percent to less than 2 percent. But in much of the world, the cost of continuous treatment is prohibitive, and physicians must make calculated trade-offs to get the most out of a limited drug supply.

That task gives rise to the Faustian bargain. Two well-timed doses of a single drug, nevirapine--one dose going to the woman as she begins labor and the other going to the infant within the first 3 days of life-can substantially cut a newborn's risk. The treatment drops the baby's chance of acquiring HIV to less than half what it would be without any medication to mother or child.

For that benefit, however, a mother takes on a dire risk. She may develop resistance to anti-HIV drugs that persists for months or years. If nevirapine or related drugs are later used to treat these mothers as their HIV infections progress toward AIDS, the medications may not help.

What's more, drug resistance in the mother may be detrimental to her later babies because nevirapine might no longer reduce the risk of transmission.

Recent studies have evaluated these benefits and risks. Now, health officials are looking for inexpensive treatments that protect both mother and child.

DEFENSIVE DELIVERY Around the world, testing women for HIV is now a basic part of prenatal care. But in developing countries, many women don't receive any medical attention during pregnancy or don't have access to HIV drugs. Worldwide, less than 10 percent of women with HIV receive any treatment to prevent HIV's spread to their infants, estimates James McIntyre of the University of Witwatersrand in Johannesburg, South Africa.

The inexpensive drug nevirapine can rapidly cut the concentration of active HIV particles in the mother and, if given promptly to the newborn, may prevent any virus that slips in from taking hold.

A landmark study called HIVNET 012, which has been under way in Uganda since 1997, has determined that, for preventing mother-to-child transmission, nevirapine is superior to the drug zidovudine, which was previously given to mothers late in pregnancy and to infants ill their first week of life. Nevirapine cut transmission risk during the first 18 months of life to 16 percent, compared to 26 percent transmission with zidovudine.

The trial didn't test advanced drugs, which can reduce infant infection rates far more. Nevirapine, which manufacturer Boehringer Ingelheim markets as Viramune, is cheaper than the more effective drugs and is sometimes provided free of charge.

Since 2000, the Ingelheim, Germany-based company has donated some 400,000 doses of its drug to prevent infections in newborns in dozens of developing countries. Elsewhere, generic doses of it can sell for about 25 cents each, and several international AIDS organizations subsidize the drug for treating pregnant women.

Public perception of single-dose nevirapine treatment took a nosedive after media reports last December exposed flaws in the Uganda study, in particular, incomplete record keeping surrounding possible side effects, such as liver toxicity and skin reactions. The study is funded by the U.S. government and includes researchers at the National Institutes of Health, based in Bethesda, Md.

In the wake of the revelations, a number of people, including key African political figures, discouraged the use of nevirapine in pregnant women. Concerns about HIVNET 012 have had "a chilling effect on the use of this drug," says pediatrician Mark W. Kline of Baylor College of Medicine in Houston.

Turning away from the treatment is wrongheaded, says McIntyre, and most other HIV researchers agree.

Whatever the study's flaws, McIntyre says, its fundamental conclusion is still correct. Single-dose nevirapine is "a safe intervention and effective up to a certain point,' he says. "For the 90 percent of women who have access to nothing, I still do see a role for single-dose nevirapine."

Following the controversy, a nine-member committee of the Institute of Medicine in Washington, D.C., reviewed HIVNET 012. In an April 7 report, the panel reaffirmed the trial's findings. Nevirapine is justifiably the cornerstone of efforts to reduce mother-to-child HIV transmission in much of the world, says Kline, who served on the panel.

Nevertheless, says McIntyre, in light of the controversy, the need to find better alternative treatments has never been more urgent.

CRUSHING RESISTANCE At the crux of the nevirapine dilemma is HIV's skill at developing resistance to antiretroviral drugs. For the same reason that doctors urge patients struggling against bacterial infections to complete any prescribed course of antibiotics, physicians prefer to hit HIV with steady doses of medication rather than with intermittent wallops.

To become resistant to nevirapine, HIV needs only one chance mutation at any of several sites in its genetic code. Substantial quantities of the drug linger in the body for up to 3 weeks after a single dose, and that lag gives the virus plenty of time to evolve.

A study of Thai women last year demonstrated the consequences of nevirapine resistance. Gonzague Jourdain of the Harvard School of Public Health in Boston and his colleagues recruited 1,844 pregnant women infected with HIV. Some of the women received a dose of nevirapine as they entered labor.

The women were observed for up to 4 months after delivery. If their health worsened, they were prescribed a daily antiviral drug regimen that included nevirapine.

Among the women who had not taken nevirapine, the drug regimen benefited two out of three by suppressing viral activity for at least 6 months. Among those who had received nevirapine during delivery, less than half benefited from the later drug regimen, Jourdain and his colleagues reported in the July 15, 2004 New England Journal of Medicine.

Other studies indicate that most genetic mutations that make HIV impervious to nevirapine simultaneously give the virus resistance to related drugs known as non-nucleoside reverse transcriptase inhibitors (NNRTIs). This group includes many of the most widely used anti-HIV drugs.

In the short term, emerging resistance may make little difference to a mother who couldn't afford HIV drugs. But the World Health Organization has an operation under way that aims to deliver, by the end of this year, a steady supply of HIV drugs to 3 million infected people who can't afford to purchase them. Infected mothers with dependent infants might be a natural choice for priority care, but because of nevirapine exposure during childbirth, many of those women may not get the full benefit of the drug. New methods of detecting resistance (see box) are assessing how long nevirapine resistance lasts. Sarah Palmer of the National Cancer Institute in Frederick, Md., and her colleagues took blood samples from 29 pregnant women who were about to receive singledose nevirapine and from the same women up to three times in the 14 months following delivery. They then tested each of the blood samples for nevirapine-resistant HIV.

Standard viral testing indicated that drug resistance arose in the majority of the women and then fell, with no resistance measurable 11 months after the single dose of the drug. But according to a new assay that Pahner's team developed, 88 percent of women who'd shown evidence of drug-resistant HIV on at least one postdelivery test still carried some resistant virus 11 months after delivery.

Those data, which Palmer presented in February in Boston at the Conference on Retroviruses and Opportunistic Infections, suggest that at least 22 percent of women receiving nevirapine during childbirth still harbor drug resistance more than a year later.

What these findings mean in terms of health and sickness remains unclear, Palmer says.

Apart from the danger that nevirapine resistance may pose to mothers, Neil Martinson of Johns Hopkins University in Baltimore is concerned about the vulnerability of subsequent children born to those women. Nevirapine resistance in mothers might render subsequent maternal doses of the drug useless, making transmission of HIV more likely at the end of a later pregnancy, he says.

In a pilot study of 186 women that Martinson and his colleagues recently conducted in Soweto, South Africa, prior use of nevirapine seemed to increase the risk of HIV transfer during a later pregnancy. Martinson says that the resulting transmission rate, 14 percent by 6 weeks after birth, wasn't as high as he feared it might be and that the difference observed between the groups might be due to unrelated factors.

Nevertheless, he and his colleagues say, more data are urgently needed to inform doctors' decisions, as HIV-infected women are increasingly returning to birth wards in places such as Soweto after taking nevirapine at least once before.

NEW APPROACHES To stop resistance from arising in women who take nevirapine for the sake of their children, researchers are investigating several combinations of relatively inexpensive drugs that might inhibit the virus from adapting to any single medication.

For example, Francois Dabis of Victor Segalen University in Bordeaux, France, and his collaborators treated 329 pregnant women in the Ivory Coast from the 32nd week of gestation until 3 days after each had delivered. The researchers administered two daily antiviral drugs, zidovudine and lamivudine. Moreover, they gave each woman a dose of nevirapine at the beginning of labor and each child a dose of that drug on the second day after birth. The newborns further received daily zidovudine for the first week.

Using conventional methods, the investigators monitored the blood of the infants and mothers for drug-resistant HIV for 6 weeks after delivery.

Fewer than 10 percent of the women appeared to develop drug resistant virus during the study, and most of those who did had virus resistant to lamivudine rather than nevirapine, the researchers found. Less than 5 percent of the infants acquired HIV, Dabis reported at the Boston conference in February. He concludes that a short course of combination therapy before delivery leading up to a one-time-only dose of nevirapine on delivery day reduces the risk of maternal resistance to that drug.

In a separate study, Palmer and her colleagues compared rates of nevirapine resistance in 10 women who had received single-dose nevirapine only and 22 similar women who had also received either 4 or 7 days of Combivir, a drug that contains zidovudine and lamivudine, following delivery.

Using the newer, more sensitive assay, the researchers found that resistance was three times as frequent with nevirapine alone as with both Combivir and nevirapine. Palmer described those findings on June 8 at the 14th International HW Drug Resistance Workshop in Quebec City.

Another approach under consideration is withholding the mom's half of the single-dose nevirapine treatment, thus preventing virus within her from being exposed to the drug and developing into a resistant form. At least two teams of researchers are exploring this option.

One trial was conducted by Robert Shapiro of Beth Israel Deaconess Medical Center in Boston and Ibou Thior of the Botswana-Harvard School of Public Health AIDS Initiative Partnership in Gaborone, along with their colleagues working in Botswana. They demonstrated that in women receiving zidovudine during pregnancy, a single dose of nevirapine to the child soon after birth seems as effective at preventing infection as is a dose to the mother during labor and to the newborn. However, the study didn't estimate the risk of such women developing zidovudine-resistant infections, which could limit their options for future treatment.

In a different study, McIntyre and his colleagues treated infants whose mothers hadn't received nevirapine with either a single dose of nevirapine or 6 weeks of daily zidovudine. In both approaches, less than 20 percent of the infants became infected during their first 3 months of life. The study didn't follow the infants beyond that age and didn't assess zidovudine resistance.

While the researchers haven't directly compared either of these regimens with a regimen that includes giving mothers nevirapine, they suggest that, in certain situations, exclusively treating infants to prevent infection may be a valid alternative to giving drugs to mothers.

In light of the recent studies, the World Health Organization is considering whether to more strongly encourage health workers in developing countries to use, where affordable, a combination of nevirapine and one or more other drugs, rather than nevirapine alone.

Permanent, multiple-drug therapy should continue to top the list for people who have access to it, says Dabis. But numerous simpler alternatives are needed because the availability of specific drugs varies greatly from region to region. For many women, he says, single-dose nevirapine is still the most effective practical treatment for preventing infections in newborns.

The several recent studies on mother-to-infant HIV transmission have marched in parallel through phases of design, data collection, and analysis, Dabis says. "We will have to live with this data for quite a while," he says, because many of the scientists are near the end of their current round of research, and completing the next round of studies will take several years.

Although many questions remain, Dabis says, making recommendations about medical treatment of HIV-infected pregnant women is something "we cannot defer."

RELATED ARTICLE: Recognizing Resistance.

New techniques are more sensitive

To get a better handle on the prevalence of drug-resistant virus after nevirapine use, scientists have developed blood tests that increase the sensitivity with which they can identify resistance. Jeffrey Johnson of the Centers for Disease Control and Prevention in Atlanta, along with colleagues in the United States and Johannesburg, South Africa, have designed one new method, which employs a technique known as real-time polymerase chain reaction. They use it to detect even small numbers of HIV particles that carry either of the two most common nevirapine-resistance mutations.

Using blood samples from 50 South African women, the researchers compared their test with a conventional one. According to the conventional test, infections in 10 of the women acquired the most common form of nevirapine resistance after exposure to the drug. The new test, however, revealed drug resistance in at least 16 others, Johnson's team reports in the July 1 Journal of Infectious Diseases.

They also found that nine of the women, rather than the five suggested by an older test, had acquired drug resistance via the second-most-common nevirapine-resistance mutation.

Overall, nearly two-thirds of the women in the study developed at least one of these two forms of resistance. There is "a considerable amount of resistance" that conventional tests don't generally detect, Johnson says.--B.H.