Newsletter #59–May/June 2000
Disappointment and Deceit in High-Dose Chemotherapy Trials
by Jane Sprague Zones
To the dismay of women's health advocates around the world, the National Cancer Institute (NCI) announced in February that fraudulent trial results on the breast cancer treatment commonly referred to as bone-marrow transplant had been reported at the annual meeting of the American Society of Clinical Oncology (ASCO) last spring.
Werner Bezwoda, chair of the hematology and oncology departments at the University of Witwatersrand in Johannesburg, South Africa, admitted that he had misrepresented research indicating benefit for women with advanced breast cancer who were treated with high-dose chemotherapy (HDC). Acknowledging that he had "committed a serious breach of scientific honesty and integrity," Bezwoda resigned from his university positions. (Six weeks later, following a disciplinary hearing, he was officially dismissed by university officials.)
Bezwoda published the first randomized clinical trial data on the use of HDC with autologous bone marrow transplant (ABMT) in 1995, showing that this very arduous and toxic treatment was more effective than conventional chemotherapy in women with metastasized breast cancer. This study contradicted the findings of a meta-analysis conducted that year by ECRI, a respected technology assessment group, which found no increased benefit for HDC/ABMT. (See What is the Price of Hope, Despite Growing Use of HDC, Many Questions Remain, NCI Withholds Trial Results on HDC/ABMT and Preliminary Results from NCI Show HDC Offers Little Benefit, BCA Newsletters 33, 50, 53, and 54 for more on this topic.) It was also noted at the time that women receiving conventional treatment in the Bezwoda study did not fare as well as women receiving equivalent treatment elsewhere.
In April 1999, the NCI and ASCO released the results of four randomized clinical trials on HDC with either ABMT or stem-cell rescue (SCR), a more recent adaptation of the treatment. None of the four studies showed that either treatment prolonged life for women with metastatic or high-risk primary breast cancer when compared to conventional treatment. A fifth trial, headed by Bezwoda, showed significant benefit for women receiving HDC.
Bezwoda's work has been used by proponents for several years to encourage breast cancer patients to undergo this expensive and debilitating treatment outside the purview of randomized research trials. Oncologists and their patients were optimistic enough about HDC's potential that they were willing to forego randomized trials (which would give women a 50 percent chance of being assigned to the control group, in which conventional treatment would be used). Bezwoda's presentation at the ASCO meeting last spring prompted plans to replicate the study in the United States, but incomplete data on the South African group receiving conventional chemotherapy raised enough concern that a team of oncologists visited the University of Witwatersrand in January to review Bezwoda's research. Although some data was withheld from them, what they did see did not jibe with the information presented at the ASCO conference.
The fraudulent data was evidently related to treatment and results for women in the control group. It appeared unlikely that participants in Bezwoda's study were hurt by this misconduct, although the ability to follow up on participants is heavily compromised by lack of documentation. The university conducted an extensive investigation, including an examination of data from the study Bezwoda published in 1995. The results of the university's investigation were posted March 10 on the Web site of the prestigious British medical journal, the Lancet, prior to their publication in the periodical.
Investigators found numerous breaches. Bezwoda had completely bypassed the university's Institutional Review Board while conducting his experiments and was unable to provide any evidence of patients' informed consent. He also lied about treatments that were used, misrepresented the race of his trial participants (most of whom were black women), and possibly fabricated data on nonexistent subjects.
The Lancet released an accompanying commentary discussing controversial measures to oversee the validity of clinical trials, and another examining the current status of HDC research and practice in women with breast cancer. The latter argues against further pursuing this line of treatment.
In an unusual move following the announcement of the fraudulent data, the New England Journal of Medicine released through its Web site an article reporting the results of an NCI-sponsored multicenter randomized trial of HDC/SCR in women with metastatic disease. (The article was posted on March 3, more than a month before its scheduled April 13 publication date.) The study, which was led by Edward Stadtmauer of the University of Pennsylvania Cancer Center, was one of the five reported at last spring's ASCO meeting. The article confirmed the authors' preliminary results, showing that HDC prolonged neither the women's lives nor the length of time the cancer took to progress.
Money Matters
Ten days after the NCI released the news of Bezwoda's fraudulent study, Aetna—at press time the nation's largest insurance company—announced that it would no longer cover HDC/SCR for women with breast cancer outside of clinical trials. Only about a thousand of the estimated 30,000 breast cancer patients who have undergone this experimental treatment in the past decade did so within a randomized trial, greatly slowing researchers' ability to study the treatment's safety and effectiveness.
Financial conflict of interest plays a major role here. According to an October 1999 article in the New York Times, the profit margin for HDC/SCR—which typically costs $100,000 to $200,000—is 15 percent.1 Response Oncology, a corporation of for-profit health centers, began offering the procedure in 1989 and netted $128 million in 1998, largely from transplants performed on women with breast cancer.
Although fraud has rarely been uncovered in breast cancer research, it is common enough in clinical trials that the NCI has established an Office of Scientific Integrity to review such misconduct. To protect the participants, it is extremely important that controls be instituted in research projects, particularly those with great financial implications. Despite the discouraging evidence on the procedure that has been mounting for more than four years, representatives of the NCI and the oncology community continue to hold out promise for variations on HDC. Desperate for therapies for advanced breast cancer, medical researchers are reluctant to abandon highly toxic treatments and may continue to come up with new formulations. BCA emphasizes the need to turn our attention and considerable resources to finding less toxic means to treat breast cancer at every stage. The one beneficial outcome of this saga may be the acknowledgment that experimental treatment must be assumed to be experimental until sound, ample evidence is reviewed and presented. For years BCA has encouraged women with breast cancer who are considering HDC to undertake the treatment only through participation in randomized trials.
We urge insurers to provide coverage to women in clinical trials. If insurance coverage for HDC had been available to women only under these conditions, we would have had these results years ago, and thousands of women would have been spared this harsh and ineffective treatment.
1 Kolata, Gina, and Eichenwald, Kurt, "Health Business Thrives on Unproven Treatment, Leaving Science Behind," New York Times, October 3, 1999.