GENE THERAPY RESEARCHERS SHOULD REPORT ALL SERIOUS TOXICITIES, EXPECTED AND UNEXPECTED, RAC SUGGESTS; NIH COMMITTEE MAY CONTACT IRBs ON INFORMED CONSENT

Executive Summary

Investigators involved in gene therapy trials should report serious adverse events, even if the events are not unexpected, committee member Brigid Leventhal, MD, Johns Hopkins, suggested at the National Institutes of Health's Recombinant DNA Advisory Committee meeting Dec. 2. Leventhal, who heads the RAC working group on data management, told the committee that "nobody has reported any unexpected toxicities" in gene therapy protocols to the RAC. However, "there have been a couple of instances of really severe toxicities which were in fact expected." The RAC should "insist on reporting of all severe toxicities whether they're expected or not," she urged. She cited the experience with Lilly's FIAU as one example of the need for caution in this area. At RAC's Dec. 2-3 meeting, the committee discussed three issues which fall under its purview but over which it has little actual regulatory clout: informed consent forms; gene therapy adverse event reporting; and investigator conduct in executing a RAC-endorsed protocol. Committee members alluded to the many gaps in RAC authority. Leventhal noted that the RAC has attempted to restrict "the maximum number of patients...on almost every protocol." However, if an investigator adds on more patients, "we're going to be upset, but we haven't actually said what we're going to do about it." In the specific instance of adverse event reporting, FDA Center for Biologics Evaluation & Research Office of Therapeutics Research & Review Director Janet Woodcock, MD, suggested that the committee contact FDA about reporting changes that the agency is contemplating ("The Pink Sheet" Nov 22, p. 15). FDA is also dealing with how to handle the issue of gene therapy safety: the idea of a national gene therapy patient registry was proposed by FDA's Suzanne Epstein, PhD, at an FDA Vaccines & Related Biologics Advisory Committee meeting in October ("The Pink Sheet" Nov. 1, p. 19). Epstein noted at the meeting that it "is not clear...whether resources will be forthcoming and leadership will appear." RAC members also discussed ways to get changes in consent forms that the committee feels are incomplete or inadequate. The group invited NIH Office for Protection from Research Risks Director Gary Ellis, PhD, to speak about OPRR's role in informed consent. Institutional review boards have the authority to revise forms submitted by investigators. RAC member Doris Zallen, PhD, Virginia Polytechnic Institute, asked if the NH-1 director, to whom RAC is an advisor, has authority over consent forms. Ellis answered that the director may refuse to make an award to an institution; Ellis recommended that the "threshold for that be very, very high." While the committee did not submit the question to a vote, there was general consensus that any informed consent form should state several things: that the sponsoring institution is responsible for patient care costs to treat any injury associated with an experimental treatment; that study sponsors may request an autopsy following the death of a patient on an approved protocol; and that both male and female trial participants should practice an effective method of birth control while enrolled on a protocol. Reimbursement in the case of injury arose as an especially compelling issue in informed consent. Gary Chase, PhD, Johns Hopkins, said that on some informed consent forms, "there is a cave-in to legal advice when people could be facing financial catastrophe." He argued that "gene therapy differs to some extent from other areas not because it's inherently more dangerous, but because to some extent it represents a national strategy...and in national strategies against disease, I think it's horribly unfair that an individual patient who's already suffering a tremendous amount" be subjected to financial insecurity in case of injury. NIH's Ellis said that the reimbursement issue "is bigger and broader than gene therapy, although it is very important to gene therapy. This is the kind of category that requires deliberation by a national commission on ethical issues in biomedical research. A letter from the RAC would be very useful." He added: "OPRR doesn't have the clout." Gene therapy pioneer French Anderson, MD, University of Southern California, said that he would strongly suggest" that the RAC stipulate in writing its standards for informed consent. Such a document "should be given to the [principal investigator] before the protocol is submitted, and if [the standards are] not in the informed consent when [the protocol] [arrives], it gets returned," he said. "I'm an advocate for investigators, but it's not fair for the committee to have to do the education of every principal investigator on something as straightforward as informed consent," Anderson said. Ellis also welcomed the idea of sending out RAC guidance on gene therapy issues in informed consent to IRBs through the OPRR. A locus of the committee's frustration with investigator conduct continues to be an "emergency" single-patient protocol that was approved by NIH without formal RAC review in December 1992 ("The Pink Sheet" Jan. 4, T&G-4). The RAC was considering a proposed multiple-patient follow-on trial in glioblastoma patients. The proposed trial, submitted by Robert Sobol, MD, and Ivor Royston, MD, San Diego Regional Cancer Center, would have treated patients with irradiated tumor cells and genetically altered, IL-2-secreting fibroblasts. The glioblastoma follow-on trial was rejected by the RAC 10 to 5, with one abstention. A trial that would perform the same procedure in colon cancer patients using the patients' own resected tumor was approved by the RAC in a vote of 14-2. In discussion of the glioblastoma trial, Leventhal said she personally was "angry" about the single-patient compassionate use trial: "Now suddenly there's an abstract submitted to a scientific meeting, there are [stories] coming out in the press saying this [therapy] is promising," with no mention that "in the weeks before she got her immunotherapy she got radiotherapy and was apparently able, before the gene therapy started to be weaned off [Merck's] Decadron, which must suggest that she was having a response to something that was going on before it started, not to mention the fact that she got tamoxifen during the gene therapy." The RAC is still questioning what is needed in a protocol to gain endorsement. During the meeting, the committee had discussed a number of thresholds for RAC approval: whether the patient would not be harmed; whether there was some suggestion of efficacy; or whether the trial was well designed. Carmen told the committee that he based his decision to disapprove the glioblastoma protocol on "the fruit from the poisonous tree threshold test." NIH should "never have permitted the single-patient protocol to be performed under its auspices," he maintained. Carmen added that "there is no...reliable scientific data that can be gleaned from the single patient protocol, which was not an experimental research protocol but an heroic measures protocol, and there are no animal data...that are all approbative of the questions that [the RAC is] addressing."