PLX4032 raises the seemingly straightforward question: data or life? The melanoma drug candidate from Roche highlights the sometimes opposing needs of doctors and regulators when it comes to experimental treatments.
"It doesn't make sense to say, ‘We want you for a statistic' instead of giving them a chance at life," says the mother of a 22-year-old son who died of melanoma, of the comparative-data-based clinical trial system of drug approval. Some melanoma specialists familiar with PLX4032 agree, and would trade the data for faster access to the drug, reports the New York Times. A doctor involved in earlier testing of the drug says his use of it will be unaffected by the results of any trial comparing it with a chemotherapy drug, the only other treatment available and therefore the benchmark.
The 22-year-old had been randomly assigned to the control group (chemotherapy) in a trial of the drug, which has shown very good results so far in slowing the spread of a certain tumor type. His cousin, meanwhile, was being given PLX4032 and enjoying those good results. And in this case, because the control treatment is administered by infusion and the candidate by pill, the randomization was transparent to doctors and subjects. The 22-year-old died while his cousin still lives.
PLX4032 makes the case that some drugs under development may be so superior to existing treatments that use of a control group amounts to needless suffering and perhaps unnecessarily early death of those subjects. The Times quotes an FDA official who says that promising candidates for the worst cancers might require individual evaluation rather than comparative trials. For the regulator, that would mean "flexibility and an open public discussion."
But for now, such drugs--as effective as they appear to be--are also very effective at bringing conflict to the doctors who administer them: Help the current patient and forgo trial data collection, or help the many into the future by collecting the data to earn the drug's market approval?
- here's the article