harold varmus: human cancer is a genetic disease in an unusual way. some altered genes are inherited but that's an uncommon event. much more commonly our genes undergo changes in only a few cells during the course of our lifetime. and certain constellations of changes predispose a cell to becoming a runaway-- a maverick cell that we know is a cancer cell. this is the way that we look at people's dna. dna is much too small to actually see, even under a high powered microscope. but, we can use biochemical reactions to amplify the dna. successive mutations to the hereditary material of certain cells produce oncogenes-- "on" switches that accelerate cell growth. tumor suppressor genes, "off" switches that restrict growth, may also mutate or... become lost from the hereditary makeup of a cell. when this happens, a cell can make billions of copies of its abnormal self. the excess tissue forms a mass-- a tumor. some tumors are benign... they don't invade nearby tissue or spread to other parts of the body. but a malignant tumor is cancer. its cells can invade

harold varmus: it's clear that the aids virus, when it naturally infects the body, does not induce acient immune response to eliminate the virus on its own. that's a daunting prospect for someone who wants to develop a vaccine. secondly, we have a very poor understanding of what components of the virus might be the most potent in inducing a response. i do think it will be a long time before we have an adequate vaccine. alexandra levine worked with jonas salk on what was to be his last major project: looking for a vaccine for h.i.v./aids. if the virus could be altered in some way so that it was seen as foreign by the defense system, then perhaps a patient who was already infected could be vaccinated with this altered h.i.v., forcing his immune system to see it and try to destroy it. that was dr. salk's original thought. he was driven to solve this in the same sense that he was driven to solve the polio issue. those ideas are still current. i do believe that we'll have better ways to deliver the so-called antigens, the immune stimulus, rather than use killed virus, which presents some