Gold Standard Clinical Trials - Science or Marketing?

This is the second in a new series on analyzing information. Credit Robert Whitaker’s “Anatomy of an Epidemic” for getting the ball rolling, as there is no way of critiquing his book without evaluating the evidence he relied upon. Only then can we ask ourselves if his conclusions are justifiable.

This is a different proposition entirely than whether Whitaker is right or wrong. You can decide that for yourself.

In my previous piece, I commented on some findings at the very bottom of the evidence food chain (a series of wholly unscientific reader polls I had conducted here), but that actually told us something. In this piece, I will flip it. I will talk about a study at the very top of the food chain, but that tells us nothing. Let’s get started ...

Two years ago, in my capacity as a journalist, I was at the 12th International Congress on Schizophrenia Research in San Diego, just down the hill from where I live. The top brain scientists in the world were presenting and in attendance, including 2000 Nobel Laureate Arvid Carlsson. I recall going to a table with my morning coffee and greeting the researchers there with, “Hi, I’m the only C student at this table.”

Part of the Congress involved poster sessions, where scientists stand in front of blow-ups of their latest findings. This affords an invaluable opportunity for the researchers at the conference to engage their fellow researchers in one-on-one discussion. It also affords a golden opportunity for people like me to walk up to some future Nobel Laureate and ask, “What’s a neuron?”

The first three or four days of posters involved cutting edge brain research. I was like a kid in a candy store ... a wrestler in a chair factory, a hippie in a drum circle. (Did you see that Super Bowl commercial?) The last day of the conference featured posters on treatment. Instantly, I felt I had been sucked back in time a half century.

Some very enthusiastic Johnson&Johnson people informed me of a new study of theirs that shed some very important new light on schizoaffective disorder.

Really? I thought. Schizoaffective is probably the most confusing diagnostic classification in the entire DSM. Everyone agrees that it has something to do with that undelineated middle ground where bipolar and schizophrenia bleed into one another, but what is it exactly? Schizophrenia lite? Bipolar heavy? Some kind of schizophrenia-bipolar mix? Or a stand-alone illness in its own right?

Naturally, I was interested. I looked at the poster. "A Randomized, Double Blind, Placebo-Controlled Study of Flexible Dose Paliperidone ER in the Treatment of Patients with Schizoaffective Disorder," I read.

I smiled politely, engaged in very short chit-chat, then made my way to other posters. In one glance, I knew, the study told me nothing. Here’s the deal:

Randomized, double blind, placebo-controlled studies are the gold standard of treatment research. Perhaps the most celebrated example of a clinical trial is James Lind’s 1747 experiment involving 12 scurvy-infected sailors. He divided his patients into six pairs. Five of the pairs showed no improvement. Of the pair receiving oranges and lemons, one had fully recovered after five days, the other had almost recovered. (Then they ran out of fruit.)

A more refined study would have pitted a much larger group of orange-and-lemon people against a proportionally large group taking a look-alike, taste-alike placebo (say a citrus pill vs a sugar pill). To keep the study honest, one would have made sure the clinicians handing out the pills had no way of knowing which pill was which (the double-blind).

I knew at a glance the J&J people had all their “i”s dotted and “t”s crossed. Studies of this nature are immensely expensive - my best guess is the $20 million-range - designed by some of the smartest people in the world overseeing an immense and highly complex undertaking where an infinity of things can go wrong (and often do).

The catch? Even though the study was conducted scientifically, to rigorous standards, it was not a scientific study. It was a marketing exercise.

Paliperidone is the generic name for J&J’s recently-approved Invega, which is Son of Risperdal. Risperdal represents the first of the new-generation atypical antipsychotics (discounting Clozaril) introduced in the early 1990s and hyped as superior to old-generation antipsychotics.

We already know that as problematic as antipsychotics may be, they are very effective in knocking out psychosis. The first studies involving Thorazine proved that six decades ago. Since then, antipsychotics have been the treatment of choice for schizophrenia and any condition involving psychosis, including bipolar and schizoaffective, accounting for some $20 billion annually in sales.

So what’s new? Nothing much, really, except that no drug company had ever undertaken a trial on a schizoaffective population. What was the point?

Keep in mind that when James Lind was investigating scurvy, the British Royal Navy was not issuing citrus fruit to its sailors. His findings would give them cause to change their practice, though it would take them another fifty years to get around to it.

Contrast this with the Invega study. Is a clinician, reading this study, really going to change his or her practice by putting all his schizoaffective patients on an antipsychotic? He already has them on an antipsychotic.

But could the study possibly be persuasive in getting a clinician to change her practice from prescribing say Geodon or Zyprexa to Invega? Ah, a marketing question.

More to come ...