Two new studies have undercut scientists' hopes that people infected with H.I.V. could avoid a lifelong regimen of taking a drug "cocktail" of 20 to 30 pills a day, a regimen that can cost $15,000 a year and cause a variety of side effects.
The studies, being published today in the New England Journal of Medicine, found that efforts to reduce the number of medications failed, allowing the resurgence of the virus.
But scientists say the disappointing results should not discourage badly needed efforts to find better strategies for controlling H.I.V., the virus that causes AIDS.
Patients who are taking the AIDS cocktail of antiviral drugs have been told that they may require such treatment indefinitely, perhaps for life. The two new studies were designed to test another idea, on which many scientists and patients had pinned their hopes: that the drug cocktail might reduce H.I.V. infection to such a low level that some of the drugs could be stopped, and the virus kept in check by the body's immune system and a "maintenance therapy" consisting of fewer drugs. That way, patients might be able to escape some of the adverse effects of medication.
A similar approach has been successfully used in treating tuberculosis and leukemia. The disease is initially hit hard with heavy doses of medication, a treatment known as induction therapy, and then kept at bay with a milder drug regimen. But for H.I.V., induction followed by maintenance therapy did not work.
In both studies, patients initially took three drugs, achieved suppression of the virus, and then, after three or six months, stopped one or two of the drugs. Virus levels rebounded quickly in a significant number of patients, and the studies had to be halted so that the drug cocktail could be reinstated.
"These are truly important and informative studies," said Dr. Jerome Groopman, an AIDS expert at the Harvard Medical School, who was not involved in the research. "They're sobering, because they demonstrate that the virus is an aggressive foe, and once the most intensive pressures are relieved, even partially, it has a tremendous capacity to spring back."
The director of one of the studies, Dr. Diane V. Havlir, an associate professor of medicine at the University of California at San Diego, said that although the results were disappointing, they were not unexpected. Researchers have seen viral levels rebound sharply in some patients who have quit the drug cocktail or tried taking a "drug holiday."
Dr. Havlir said patients in her study had been monitored for signs of relapse, which sometimes occurred within weeks of switching treatments. When that happened, they were urged to resume the triple combination.
Preliminary data on a small number of those patients, Dr. Havlir said, suggest that the virus was resuppressed when the more powerful treatment was resumed.
Dr. Havlir and her colleagues said their findings underscored warnings already given to patients: that combination therapy is a long-term proposition, and breaking the regimen may allow the virus to rebound and evolve resistant strains that defy treatment. Even so, the experiment had to be done, Dr. Havlir said, because doctors and patients alike want to know whether there is any end in sight to the cocktail. She said she and her colleagues had no trouble finding patients eager to enter the study.
"Despite how these drugs have completely changed the face of the H.I.V. epidemic in the United States, we're realizing that there are limitations," she said, citing side effects and a strict, difficult schedule of doses.
One pill in the cocktail must be taken precisely every eight hours, at least one hour before or two hours after a meal. Side effects can include vomiting and diarrhea, and certain drugs, the protease inhibitors, are causing high levels of cholesterol and triglycerides in some patients, increasing their risk of heart attack. Some patients also develop "protease paunches" -- a shift of body fat to the abdomen -- or "buffalo humps" -- fat deposits on the back of the neck that can grow so large and disfiguring that patients have them surgically removed.
"Side effects are catching up with large numbers of patients," said Dr. Groopman, who estimated that 20 percent to 25 percent of patients on protease inhibitors for two or three years had "significant fat redistribution and, importantly, sky high triglycerides and high cholesterol."
He has seen patients with cholesterol levels driven over 700, he said.
If people stay on the drugs for decades, Dr. Groopman asked, "will it happen to everyone?"
Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases, said that in addition to the physical side effects, patients were finding it hard to tolerate the endless disruption of their lives.
"Their whole day is dominated by the schedule of taking drugs," Dr. Fauci said. "After six months to a year they say, 'God, I can't take this any more.' "
Dr. Jay Levy, a professor of medicine at the University of California at San Francisco, said the studies revealed the need for new treatments, particularly vaccines and drugs that would boost the immune system and help it fight the infection. The current antiviral drugs fall short, Dr. Levy said, because they cannot reach viruses hiding within the patient's cells that can re-emerge when the drugs are stopped.
In Dr. Havlir's study, patients took three antiviral drugs for six months: the protease inhibitor indinavir, and two reverse transcriptase inhibitors, lamivudine and zidovudine, or AZT. Then, one-third of the patients continued the three drugs, one-third took indinavir alone, and one-third took AZT and lamivudine.
Among patients who continued taking the drug cocktail, only 4 percent had loss of viral suppression. But in each of the other two groups, within two months or less, 23 percent of the patients relapsed. Those who had taken AZT before the study began were especially vulnerable to relapse.
The second study reported similar results. Directed by Dr. Gilles Pialoux of Rothschild Hospital in Paris, patients initially took the same three drugs as in Dr. Havlir's study, but for only three months instead of six. Then they stayed with that regimen or switched to a combination of either AZT and lamivudine, or AZT and indinavir.
In the French study, 8.6 percent of the patients who continued taking the drug cocktail lost viral suppression. In the other two groups, 31 percent and 22 percent of patients relapsed.
Both groups of scientists said that despite the failure of maintenance therapy in their studies, researchers should keep trying to find alternatives to the current regimen, perhaps involving longer periods on the full cocktail or different drug combinations in the second phase.
Dr. Fauci said his research team was planning such a study, to begin within the next few weeks. Fifty patients taking the drug cocktail whose virus levels have been undetectable for a year or more will stop all three medications, to see whether their immune systems have recovered enough to control the infection. Patients will be tested every one or two weeks to measure their viral levels and counts of CD4 cells, he said, and if virus levels come up, the patients will be treated immediately.
"We don't have overwhelming confidence that we'll succeed," he said, "but we have to do the study. We're all appreciating that it's not feasible to say that patients are going to be on these regimens for 30 years."