Sep 20 2009
“Gonzalez Regimen” for Cancer of the Pancreas: Even Worse than We Thought (Part II: Loose Ends)
Last week I discussed the dismal results of the “Gonzalez Trial” for cancer of the pancreas,* as reported in an article recently posted on the website of the Journal of Clinical Oncology. I promised that this week I’d discuss “troubling information, both stated and unstated [in the report],” and also some ethical issues. More has come to light in the past few days, including Nicholas Gonzalez’s own voluminous, angry response to the JCO article. I’ll comment upon that below, but first a brief review.
The trial was begun in 1999 under the auspices of Columbia University, after Rep. Dan Burton had pressured NCI Director Richard Klausner to fund it. It was originally conceived as a randomized, controlled trial comparing the “Gonzalez Regimen” to standard chemotherapy for cancer of the pancreas. In the first year, however, only 2 subjects had been accrued, purportedly because those seeking Gonzalez’s treatment were not willing to risk random assignment to the chemotherapy arm. In 2000, the protocol was changed to a “prospective, cohort study” to allow potential subjects to choose which treatment they would follow. Gonzalez himself was to provide the ‘enzyme’ treatments.
After that there was little public information about the trial for several years, other than a few determination letters from the Office of Human Research Protections and a frightening account of the experience of one subject treated by Gonzalez. By 2006 or so, those of us who pay attention to creeping pseudomedicine in the academy were wondering what had become of it. About a year ago we found out: the trial had been quietly “terminated” in 2005 after it met “pre-determined stopping criteria.” As explained here, that meant that the Gonzalez group had not fared well.
Four years after the trial’s ‘termination,’ the report was finally published: The Gonzalez cohort had not only fared much worse than the cohort that received chemotherapy, but it had fared worse than a comparable group of historical controls. Here, again, is the survival graph from the JCO paper:

The Gonzalez group had also fared much worse in ‘quality of life’ scores, which included a measure of pain.
Now let’s read between the lines. Forgive me for taking shortcuts; I’m a little pressed for time. Continue Reading »
