The studies involve mostly gay white men. Resistance, however, may be more prevalent in other groups, such as drug users and their sex partners, researchers said.
About 40,000 new HIV infections occur yearly in the United States. In recent years, powerful drug cocktails have subdued the virus to undetectable levels in many patients. But studies have found the virus persists or comes roaring back in 10 percent to 50 percent.
A researcher involved in one of the studies says the report should "raise a red flag" and says there should be renewed emphasis on safe sex.
The complicated drug regimen has proved difficult to adhere to, and many patients who missed doses or quit taking their medicines developed drug-resistant infections that are now being passed along to others.
HIV is still so new that scientists disagree even about how to define resistance. And since both studies used laboratory tests, no one really knows how the definitions will translate into patient care. Giving high doses of a drug may be enough to overwhelm a virus' resistance, Pomerantz said.
In one study, researchers at the University of California at San Diego defined resistance as a 10-fold increase in HIV's ability to withstand a drug when compared with a laboratory strain.
That study, led by Dr. Susan J. Little, tested 141 patients -- in San Diego, Los Angeles, Dallas, Denver and Boston -- and found that three (2 percent) had HIV with at least 10-fold greater resistance to one or more drugs.
An additional 36 patients (26 percent) had HIV that was 2.5 to 10 times more resistant.
In the other study, researchers at Rockefeller University in New York defined resistance as a threefold increase in HIV's ability to withstand a drug. That study, led by Dr. Daniel Boden of the Aaron Diamond AIDS Research Center, tested 80 subjects in New York and Los Angeles.
Of 67 in whom resistance could be tested, three (4.5 percent) had HIV that was highly resistant -- fivefold resistant -- to multiple drugs. The subjects were among 18 (26.8 percent) with HIV that was at least threefold resistant to at least one drug.
"I think one has to step back and look at the data and say, yes, there is a red flag here, that drug resistant viruses are being transmitted in significant numbers and this may have implications as far as treatment," says Dr. Martin Markowitz of the Aaron Diamond AIDS Research Center.
Testing every newly infected patient for drug resistance would be impractical because the tests cost several thousand dollars and are difficult to interpret, said Dr. Roger J. Pomerantz, an expert not involved in either study.
But if a patient takes a drug cocktail faithfully and it isn't working, testing should be considered to see how the combination of medicines might be reformulate, he said.
"Resistance is slowly increasing," Pomerantz said. "If you were looking at this five years ago, you would see zero."