The Procrustean Bed
Oncologists sometimes try to recruit patients into Phase I (toxicity) clinical trials. But how effective are the experimental treatments provided in such trials? In a recent 2010 study that pooled data from various phase I chemotherapy trials for sarcoma, the partial response rate was 1.6 percent (2 out of 133 subjects) and the complete response (CR) rate was 0.8 percent (1 out of 133). The median progression-free survival was 2.1 months and the median overall survival was 7.6 months. Meanwhile, 18 percent of patients experienced grade 3 or 4 (i.e., critical or life-threatening) toxicity and 12 percent dropped the trial treatment because of toxicity.
Yet here’s the amazing part. The authors of this study, at the Royal Marsden Hospital, London, concluded: “Phase I clinical trials could be considered a therapeutic option in sarcoma…due to the low risk of toxicity” (Jones RL, Olmos D, Thway K, et al. Clinical benefit of early phase clinical trial participation for advanced sarcoma patients. Cancer Chemother Pharmacol. 2010. Available at PubMed, emphasis added).
Pardon me for being blunt, but what universe do these scientists inhabit? I wonder if they themselves would submit to such toxic drugs for a less than one percent chance of a “cure” (a “cure” that in any case may last a month or so). And—it seems almost too obvious to ask—how do these scientists define a “low risk of toxicity”? Grade 4 toxicity classically includes such things as massive hemorrhages, life-threatening infection, more than ten episodes of vomiting in a 24 hour period, etc. Even grade 3 toxicity includes such things as “painful erythema, edema or ulcers and (patients) cannot eat” (http://www.rtog.org/members/toxicity/tox.html)
Sometimes I get the impression that various authors reach their conclusions first and then force their data to fit a preconceived notion. The Greeks had a term for this, a “Procrustean bed.” This term came from a myth about a highwayman named Procrustes, who physically either cut or stretched the limbs of his victims to fit the predetermined length of his torture bed. This term has stuck for any situation in which people stretch (or minimize) the data to conform to some preconceived notion.