The breast cancer HDCT/transplant debacle:

September 1, 2007

In the late 1980s, based on limited studies, high-dose chemotherapy with autologous bone marrow transplantation (HDCT/ABMT) emerged as a prominent procedure in the treatment of metastatic and early-stage breast cancer. Subsequent randomized clinical trials showed that HDCT/ABMT had no benefit, compared with standard chemotherapy; however, in the interim, some 30,000 women had undergone unnecessary treatment.

In the late 1980s, based on limited studies, high-dose chemotherapy with autologous bone marrow transplantation (HDCT/ABMT) emerged as a prominent procedure in the treatment of metastatic and early-stage breast cancer. Subsequent randomized clinical trials showed that HDCT/ABMT had no benefit, compared with standard chemotherapy; however, in the interim, some 30,000 women had undergone unnecessary treatment.

Cancer Care & Economics (CC&E) recently spoke with Richard A. Rettig, PhD, one of four prominent researchers who examined the reasons for this lapse in clinical judgment. The findings were published in the book False Hope: Bone Marrow Transplantation for Breast Cancer (Oxford University Press, 2007). Dr. Rettig is an adjunct senior social scientist for the RAND Corporation, Arlington, Virginia. His three coauthors are Peter Jacobson, JD, MPH; Cynthia M. Farquhar, MD, MPH; and Wade Aubry, MD.

CC&E: What factors propelled the acceptance of high-dose chemotherapy/autologous bone marrow transplantation (HDCT/ABMT) as treatment for metastatic breast cancer?

DR. RETTIG: Widespread clinical utilization of HDCT/ABMT was driven by several factors. First off, the women typically were in their middle years, married with children, and their personal stories evoked great sympathy.

Health insurers initially denied coverage because the procedure was deemed experimental, without solid evidence of effectiveness. But when the mainstream press reported stories of desperate women with cancer being deprived of life-saving, cutting edge technology for financial reasons, there was a surge of public demand. This resulted in litigation, commercial exploitation, and government mandates of insurance coverage.

It's important to note that the oncology community legitimized this treatment long before thorough examination was complete. We found that decisions reached at the front-end of the diffusion process drove subsequent developments for a long period of time.

CC&E: What role did clinical trials play in ending the use of this treatment?

DR. RETTIG: Clinical evaluation by randomized controlled trials was a very slow process. One problem lay in the fact that the procedure was widely available as the trials were being conducted, stifling accrual of patients. When you have oncologists saying to patients, "this is your best hope," why enter a clinical trial when the treatment is already available?

By the end of the 1990s, four trials reported no benefit to the experimental procedure, compared with standard treatment. Two trials that reported positive benefit, when audited, were found to be fraudulent. So in the end, trial results eventually stopped diffusion of the procedure.

CC&E: Was there an inherent flaw in the structure of the first trials that led to the wrong conclusions?

DR. RETTIG: We found that adherence to the gold standard of randomized trials wasn't as strictly observed as it should have been. In our judgment, there was a "jumping the gun" phenomenon on the basis of promising phase II studies, which were single-site, small population, and seldom had adequate controls. Moving too quickly on the basis of rather weak phase II studies was one of the problems that complicated the acquisition of definitive, randomized phase III trial data.

CC&E: Did a general lack of oversight play a role in this failed process?

DR. RETTIG: Yes, and there still is a lack of uniform oversight. We need a more systematic way to manage the evaluation of unregulated new medical procedures or treatments, especially when the interventions are so costly.

There is an institutional deficit regarding our evaluation process. We recommend creating a public/private partnership that includes the NIH, clinical researchers, health insurers, and patients. The new entity would oversee the transition of investigational treatments from hypothesis-generating phase II trials to publicly funded, hypothesis-testing randomized controlled phase III trials.

CC&E: How do we achieve the kind of evidence-based medicine needed to prevent situations like this?

DR. RETTIG: The movement towards evidence-based medicine—with appropriate allowance for physician discretion—is really a decade old in terms of its prominence. The commitment to evidence-based medicine constitutes a conceptual basis for putting medical care on a firmer scientific foundation. This would include, in addition to randomized clinical trials, technology assessments that evaluate the literature for effectiveness of procedures, carefully audited randomized trials, and clinical practice guidelines. These are all critical components to establishing true evidence-based medicine.

I would stress again that it is necessary to strengthen and protect the integrity of the evaluation process on a continuing basis, which involves a firm commitment to randomized controlled clinical trials. I don't think a better system will emerge anytime soon, although there are some interesting discussions going ahead on changing the underlying design of various trials.

CC&E: Does the story of HDCT/ABMT in breast cancer have any particular relevance in today's healthcare debate?

DR. RETTIG: I think it demonstrates the importance of evaluation of new medical interventions to determine what's effective and what's not. There are a variety of new agents flowing from the laboratory into the healthcare system, and they need to be rigorously evaluated. And obviously, postmarketing surveillance certainly is an important element in that process.

I think the story of HDCT/ABMT also points out the underlying tension between making new cancer therapies available for patients who have run out of options and the need to thoroughly validate those therapies before making them widely available.

Finding common ground within the tension that exists between physicians and insurers is also crucial. There is reason to believe that insurers may be more receptive today than they were in the past to seeking that common ground. That's a healthy step forward in the evaluation and validation process. Hopefully, the lessons learned from the HDCT/ABMT experiment will provide some historical guidance as we move forward with new and exciting discoveries.