DO CLINICAL TRIALS WORK?

Linda Joyce Beck, pro miniature golfer

 

Leavenworth

 

You know how clinical trials work.  Let’s say that a research team develops a drug that cures (or more likely inhibits) lung cancer in mice or chickens or monkeys.  To check if it works in humans they first perform a Phase I trial to determine maximum safe dose.  Then they get permission and funding for a Phase II trial, perhaps involving a few dozen to a hundred or so people.  In Phase II they seek to determine whether the drug does humans any good, while continuing to monitor for side effects, etc.  If the drug still looks promising after Phase II, Phase III may be attempted, involving several hundred to several thousand subjects.  If once again the drug appears to be effective they leap with joy, order their BMWs, and apply to the FDA for permission to market.  (Most likely they actually wait for approval before buying the BMW.)  Most drugs never even make it to Phase III, and fewer still are approved for use.  Drug development is a costly, tedious, and frustrating process. 

The “gold standard” for Phase III trials is the “double blind” version.  The trial cohort is divided in two; half receive the drug and the other half get a placebo*.  Neither the patient nor the doctor administering the drug knows which is who.  Only the computer back at headquarters knows that patient A is getting the real stuff and patient B is getting a sugar pill.  When the trial has run its course, that same computer assists sage statisticians to determine (with impecable mathematical rigor) whether the drug arm of the study experienced better outcomes than the placebo arm.  If it did, head for the FDA office nearest you…

Now, however, an article appearing in the NYTimes (thank you, Dick) questions the validity of even “gold-standard” trials.  The example cited is Avastin, which targets angiogenesis in tumors.  (No, I did not use that big word just to impress you.)  Angiogenesis is the process whereby fast-multiplying tumor cells build a blood supply; without sufficient blood they will croak, like any other cell.  Avastin was used to combat a nasty species of brain tumor.  The drug had helped significantly in an earlier, much smaller Phase II study, but the double-blind Phase 3 trial showed no significant difference in survival between the two arms.  In other words, from a much larger selection of patients, very few benefitted from Avastin.  The odd thing was, though, that some benefitted a lot.  Which brings us, I think, to the point:  maybe clinical trials of the traditional sort don’t work mainly because cancers are so heterogeneous.  Maybe it doesn’t make sense to speak of “pancreatic cancer”, for instance, as if it were a single disease that can be confronted by a single remedy.  If one were to examine the genetic damage that lead to cancer in the pancreas, one might find that several different kinds of genetic errors were involved.  Thus a treatment that worked on tumor A might not work on tumor B, even if both were located in the pancreas.  Get my drift?  We come once again to the necessity of matching the treatment to the individual genetic (or epigenetic?) mistake.   More laboratory analysis, more testing – and more money.  

There is a lot more stuff in the original Times article, but I’ll let you read it for yourselves.  It is lunchtime and I’m hungry.  http://nyti.ms/12vDZvw

*Commonly Phase III trials are run with the “standard of care” treatment used in what I have called the placebo arm.  That way you can see if the new drug, which almost certainly is expensive, will do a significantly better job than the usual treatment regimen.