Phase 3 randomized double-blind clinical trials (RCTs) are very expensive and often fail. When they do the drug company loses their investment in the new drug. As a result they have a big incentive to design the RCT to maximize the chances of success. One way is to exclude weak patients who are less likely to survive the new drug, for example in the passages quoted below, by excluding patients who have a past history of cancer. But the result is that the RCT does not provide evidence about the efficacy of the new drug in helping one of the groups we would like to help.
(p. D5) In a recent report in JAMA Oncology by researchers at the University of Texas Southwestern Medical Center in Dallas, approximately 25 percent of Americans 65 and older and 11 percent of younger adults who were previously treated for cancer were subsequently found to have one or more new cancers in a different site. Depending on the type of original cancer and the person’s age, the risk of developing a second unrelated cancer ranged from 3.5 percent to 36.9 percent. The study covered 765,843 new cancer diagnoses made between 2009 and 2013 and recorded in a population-based national registry, the Surveillance, Epidemiology and End Results (SEER) program.
. . .
The Texas researchers, led by Caitlin C. Murphy, an epidemiologist, undertook the study of new cancers in cancer survivors in hopes of changing the common practice of excluding former cancer patients from clinical trials when they develop another cancer.
“This exclusion is not evidence-based,” Dr. Murphy said in an interview. “Patients with a prior cancer do not necessarily have a worse prognosis than those without a cancer history. They should be allowed to participate in clinical trials, which may be one of their only treatment options. If they’re excluded, a lot of patients are left out from what may be the best available treatment.”
For the full story see:
(Note: ellipsis added.)
(Note: the online version of the story has the date Dec. 25, 2017 [sic], and has the same title as the print version.)
The academic report mentioned above is: