Shares of drug developer Genta Inc. dropped sharply Tuesday after the Food and Drug Administration questioned whether its leukemia treatment candidate would offer any substantial improvement over chemotherapy.

In a summary prepared for an advisory panel that meets Wednesday, the FDA staff said 10 percent more patients with chronic lymphocytic leukemia (CLL) responded to chemotherapy when Genasense was added, but the difference "is of questionable clinical significance." Adding infusions of Genasense also "appears to add to the toxicity of the treatment," the FDA staff said. Genta's figures showed more patients experienced nausea, fever, fatigue, headaches and other problems.

The response rate, or complete response to the treatment, was 17 percent, compared with a 7 percent response rate for chemotherapy alone. That falls short of a 20 percent rate of remission goal. Also, the median survival rate for patients in the 223-person study on a combination of Genasense and chemotherapy was 33.8 months. The group solely on chemotherapy had a median survival rate of 32.9 months.

THIS is a common theme in analyzing clinical data. Although the drug might be synergistic with chemotherapy in raising the survival rate by a year, overall, when the patient numbers are very large there is no difference. BTW, refractive CLL is a nasty disease and not as treatable as primary ALL and NHL.