VaxGen's AidsVax: company snubs vaccine lobby over details of trial methodology

The failure of AidsVax to prevent infection with HIV - in clinical trial results published in 2003 - triggered an intense debate about the controversial product and its manufacturer, VaxGen Inc of Brisbane, California. Mail to this website, maintained by Brian Deer, shows that existing material on a VaxGen-AidsVax index is read by significant numbers. This page seeks to further inform the discussion

TO: Don Francis & Jim Key
VaxGen

FROM: Chris Collins on behalf of AIDS Vaccine Advocacy Coalition
DATE: February 20, 2003
RE: Data Analysis Plan

As we prepare for your impending news on the AIDSVAX trial, it has become very clear to us that we cannot provide informed and accurate comment on the results reported in your statements unless we have access to the plan by which you are analyzing trial data.

We understand your position that the data analysis plan is proprietary or a trade secret. However, full public access to this plan, in our view, would greatly increase understanding of your data, and improve the company's credibility. Most importantly, it would allow the kind of independent scientific response that will be expected by the public, and particularly our own constituency, as we try to understand the true meaning of the results from the world's first AIDS vaccine efficacy trial.

In the absence of access to the full data analysis plan, we will be forced to rely on the "statistical methods" portion of the AIDSVAX 004 Final Protocol (dated May 21, 1998) as the sole basis of our assessment of your results. If there are any substantive differences between the protocol and the formal analysis plan, please reply by identifying them for us so our response can be accurate and fair.

Additionally, our expectation is that VaxGen's announcement will not exceed reporting of the primary and secondary endpoints and additional outcome measure results as detailed in the protocol. We believe it is essential that your announcement report the results on these endpoints, to the extent they are available, without overstating the outcomes or attempting to make claims for AIDSVAX that are not warranted by the primary data analysis determined in advance by the protocol.

Again, we respectfully request that you forward the data analysis plan to us at your earliest convenience, preferably before you announce your results.

Our public comment, and our interaction with the media, will be guided by the level of detail, evidence of balance in your statements, and the credibility of the external review that accompanies announcement of your results. We must say that we have some reservations about limiting direct access to the data to yourselves, CDC, and study investigators, since all of these have been chosen by you and are interested parties in the trial. Since the DSMB was also chosen by you and FDA is not likely to object to the details of your statement prior to filing for licensing, we feel it is our responsibility to base our response solely on hard data criteria established before the trial began before the details have been subjected to independent peer-review scrutiny.