Bristol-Myers Squibb Co. said on Sunday its experimental drug for rheumatoid arthritis showed highly promising results in two late-stage clinical trials.

The drug, abatacept, could be the first of a new class of treatments known as T-cell co-stimulation modulators, which slow the body’s response to inflammation. The results will be formally presented this week at the annual meeting of the American College of Rheumatology.

Both phase III trials met their main goal, which was to show a 20 percent improvement in the signs and symptoms of the disease, as measured according to a standard known as ACR.

The results come as Bristol-Myers struggles to recover from an accounting scandal, increased competition to some of its products from cheaper generics, and as it prepares to lose patent protection in 2006 for its biggest-selling product, the cholesterol fighter Pravachol.

Rheumatoid arthritis is a hereditary condition in which the immune system attacks joints. It affects 2.3 million people in the United States, and analysts estimate abatacept could generate annual sales ranging from $500 million to more than $1 billion if it is approved.

The first trial, lasting a year, tested abatacept in patients who had failed to respond to methotrexate, a common existing treatment. The trial showed that 73.1 percent of patients taking abatacept plus methotrexate achieved a 20 percent improvement in the signs and symptoms of the disease, compared to 39.7 percent taking methotrexate plus placebo.

The second trial, lasting six months, tested patients who had failed to respond to TNF-inhibitors, which block a protein known as tumor necrosis factor. These include such drugs as Johnson & Johnson’s Remicade, Amgen Inc.‘s Enbrel and Abbott Laboratories Inc.‘s Humira.

The trial showed that 50.4 percent of patients taking abatacept together with another treatment in the same class as methotrexate achieved a 20 percent improvement in the symptoms of the disease, which can cause crippling pain in the joints. Only 19.5 percent of patients taking standard treatment plus a placebo achieved an improvement of 20 percent.

Abatacept, like Remicade, is given by infusion. The other TNF-inhibitors are given by injection.

While abatacept was not tested on a head-to-head basis with any of the TNF-inhibitors, Joel Kremer, director of research at the Center for Rheumatology in Albany and lead investigator on the first trial, said they appeared to have similar levels of efficacy.

“In terms of the initial studies on both the effectiveness of this intervention appears to be similar to the studies with TNF drugs,” he said.

Analysts, who expect abatacept to be launched in the fourth quarter of next year, said the drug was likely to be used at first in patients who did not respond to the TNF-inhibitors.

Kremer also said that was how doctors were likely to begin their prescription patterns. But he said they may soon be more likely to prescribe abatacept after a patient had failed one TNF-inhibitor rather than all three, and he said that abatacept could within a short period of time be prescribed instead of a TNF-inhibitor.

“With any new drug people need to get a level of comfort with its use,” he said. “In the beginning it probably won’t be used instead of a TNF-inhibitor, but I predict that within a short period of time (it) will be.”

Kremer said abatacept had a good safety profile, which, if borne out, could help it win advocates.

TNF-inhibitors carry precautions that include a possible increased risk of lymphoma, a blood cancer, and certain blood-related disorders such as leukopenia, in which the white blood cell count is too low, and thrombocytopenia, when blood platelets are too low.

Bristol-Myers said the most common adverse events seen in its trials were headache, nasopharyngitis, an irritation of the upper throat, and nausea.