NCI EARLY RELEASE OF CLINICAL DATA WILL BE RESTRICTED TO PRODUCTS
Executive Summary
NCI EARLY RELEASE OF CLINICAL DATA WILL BE RESTRICTED TO PRODUCTS readily available for clinical use, according to recently released National Cancer Institute final guidelines. Approved by NCI's National Cancer Advisory Board (NCAB), the guidelines for clinical announcements state that "if the device or drug is new, arrangements must be made with FDA or other appropriate federal agencies to make such readily available to the physician and his patient." A draft of the guidelines was discussed at an NCAB meeting last May. Clinical announcements are best reserved for "common disorders which are serious," said John Durant, VP-Health Affairs at the University of Birmingham. Durant, an NCAB member, headed a subcommittee charged with development of the guidelines. To date, NCI has made two clinical announcements prior to the publication of the studies. The first, in May 1988, focused on the positive effects of adjuvant chemotherapy or hormone therapy in stage one, node-negative breast cancer patients, and was released nine months before publication of the relevant research in the New England Journal of Medicine. The announcement, then called a "clinical alert," caused "a firestorm of controversy," Durant said, in part because physicians did not have sufficient access to the primary data. In contrast, the second clinical announcement (called an 'update') on the combination of levamisole and 5-fluorouracil for Duke's Stage C colon cancer, was released in the fall of 1989 and resulted in "substantial acceptance" by the physician community, Durant said. In May 1989, FDA granted levamisole (Janssen's Ergamisole) Treatment IND status as an adjuvant to 5-FU for the treatment of Duke's Stage C colon cancer. Clinical announcements will "principally" be based on "Phase III therapeutic trials carried out by the investigators supported through NCI's Clinical Trials Cooperative Groups," according to the guidelines. They will usually encompass at least two trials. The "ideal stimulus" for an announcement is a "significant improvement in overall survival"; but, in "selected situations," NCI will consider publicizing "a carefully documented, meaningful increase in improvement in quality of life, complete response rate, disease-free survival, or other surrogates," the guidelines state. At a Nov. 14 forum for science writers, NCI Director Samuel Broder discussed the rationale for clinical announcements, commenting that physicians "have a right to have clear, practical information that allows them to practice their art effectively." To that end, the guidelines require that "a full report of the data should be available . . . concurrent with the distribution" of the announcement. This could take the form of "a conventional publication or it may be electronically available," Broder said. In the case of unpublished data, a clinical announcement "should not be made until the paper is accepted" by a peer-reviewed journal, according to the guidelines.