A laboratory error acknowledged this week casts doubt upon earlier findings of Harvard scientists that a combination of three anti-AIDS drug can stop the growth of the HIV virus.

Medical student Yung-Kang Chow, working with a team of researchers in the lab of Dr. Martin S. Hirsch at the Massachusetts General Hospital, announced in February that a combination of AZT, ddl and one of two experimental drugs had been found to block the spread of the HIV virus to other cells in the test tube before the virus could develop sufficient resistance to the drugs.

The much-publicized results seemed to bear out Chow's theory that a combination of anti-HIV drugs would cause the virus to mutate to a non-reproductive form, and 400 human volunteers were selected for studies of the so-called "triple drug therapy." But other scientific teams were unable to replicate the results, prompting the Harvard team to repeat the experiment, The New York Times reported.

The original report had stated that no resistence was found after 10 generations of test-tube testing. But the Harvard researchers found in the additional tests that the virus became resistent to the drug combination after 20 to 30 generations.

"There is a point at which even the wily HIV virus can change itself (mutate) in an attempt to outwit even three drugs," the team said in a statement released yesterday.

Hirsch said Wednesday that the team will send a letter to the scientific journal Nature, which published the original results, informing them of the new findings. But he also said that they still believe in their original theory concerning the effectiveness of drug combinations.

"An unfortunate error, which we regret, occurred in our laboratory," the group said in the statement. "The error, albeit regrettable, does not negate the rationale for triple drug therapy directed against the same viral target."

Hirsch also said that the federally-funded human trials of the drug combination, currently being conducted across the country, will continue despite the new findings.

"Laboratories elsewhere, as late as [Thursday], have confirmed that the three drug attack works better in the test tube than several other two and three drug regimens," he said. "That gives us cautious optimism that the treatment will offer an advance."