The only way to really tell the difference between a safe and effective drug and a placebo is to test the drug in a rigorous, controlled study. This is another reason to doubt Celgene's ( CELG) current plans to seek FDA approval of Revlimid for myelodysplastic syndrome (MDS). Celgene's Revlimid data in MDS are all from single-arm studies, which means there isn't any way for the FDA to determine if the safety problems that have cropped up are tied to the drug or to patient co-morbidities. I'm singling out Celgene here, but the company is certainly not alone. In my opinion, Iressa's recent failure will have long-term effects on how the FDA reviews and approves drugs.