Main Menu · Search ·Current Issue ·Contact ·Archives ·Centennial ·Letters to the Editor ·FAQs
On Becoming a StatisticAssessing a breast-cancer therapyby John F. Lauerman
|
In Europe, in the first half of the nineteenth century, the use of mathematics and statistics in medical and behavioral research approached zealotry. In 1828, Louis Villermé adopted a statistical approach to public hygiene, demonstrating a relationship between the conditions in the poorest neighborhoods of Paris and their residents' health problems. In Italy, Giacomo Tomassini began using statistics to establish the efficacy of therapies, and the French physician Pierre C.A. Louis, considered one of the founders of medical statistics, developed the "numerical method" of patient study, an unprecedentedly precise statistical analysis of therapy. Florence Nightingale said that to the extent statistics revealed truths about humanity, its study was "a religious service."
Until the Enlightenment, healing techniques had largely been based either on cultural traditions or on the observations of individuals who seldom had the opportunity to test their ideas. The power of statistics paved the way to modern medicine. Thousands of individuals have been converted into statistics, allowing us to learn from the accumulated circumstances of their lives and deaths.
In medicine, the use of statistics to disentangle observation, opinion, fact, and desire remains crucial. The hope inspired by the mere suggestion of treatment can be healing, evidenced by the way patients with heart disease frequently respond as well to placebo as to actual drug therapy. Weary of watching patients suffer, physicians themselves often pray for therapies to succeed, root for new interventions, and suffer when a promising candidate fails.
High-dose chemotherapy with bone-marrow transplant (BMT) was such a beacon for many women with breast cancer and their doctors. Researchers had theorized that the highest possible doses of chemotherapy would stand a better chance of removing all the cancer cells in the body, a vital step in the effort to keep cancer from spreading. Patients saw the procedure as a faint glow of hope in the battle against a disease that kills thousands of the 175,000 women diagnosed each year with the condition.
Unlike cancer, hope is contagious, and in the late eighties and early nineties, hope for a breast-cancer cure spread throughout the community of people affected by this terrible disease. In many states, insurance companies were obliged to pay for chemotherapy with BMT, despite their skepticism about its effectiveness. BMT itself made the treatment extremely dangerous. Chemotherapy, which preferentially kills growing, dividing cells, targets cells in the bone marrow that are necessary to replenish the immune system. Cancer researchers reasoned that they could increase anticancer medication and still protect these bystanders from collateral damage by removing them before the procedure and replacing them afterward.
In recent years, somewhere between 5,000 and 6,000 women (there are various estimates) received the treatment annually, at a total cost of more than $500 million. At first, 15 to 20 percent of the women who underwent the harsh treatment died of their "cure." With experience, physicians lowered that rate to 5 percent of BMT patients, yet proof that this expensive, sometimes fatal, therapy did anyone any good was still lacking.
The answer was supposed to come from a series of clinical trials funded by the National Cancer Institute, each one looking at a slightly different application of high-dose chemotherapy with BMT. Unfortunately, the results, reported this spring, were not as conclusive as people had hoped.
Associate professor of medicine Eric Winer, M.D., director of the breast oncology division at Dana Farber Cancer Institute's Gillette Center for Women's Cancers, does note "some encouraging bits of evidence." He points to one 783-patient study, performed at the University of Chicago, in which all patients received high-dose chemotherapy, whether or not they underwent BMT. About two-thirds of the patients in both the experimental and the control group fared well, although some of the gains were offset by toxic side effects. Still, this indicates to Winer that high-dose chemotherapy can work, if it can somehow be made safer.
Of course, making aggressive cancer treatment safer was what BMT was supposed to accomplish. Perhaps if the studies had been able to enroll more patients more quickly, the results might have been more clear and timely.
Unfortunately, although thousands of women underwent the BMT protocol, only a few hundred entered the National Cancer Institute-sanctioned trials, the source of the recently released report. Most, if not all, of the women who received treatment outside the NCI trials were part of clinical studies of some kind, usually smaller studies that lacked great statistical potential because the patients were not randomized. Their outcomes, therefore, were not compared with those of randomly selected patients who did not receive the BMT protocol.
"In my experience," says associate professor of medicine Thomas Spitzer, M.D., director of Massachusetts General Hospital's bone-marrow transplant program, "patients often decline participation in trials because they've already made up their minds one way or another. Some women think the treatment is too toxic or risky, while others insist on receiving it."
The fact is that, before trials begin, we don't know whether a treatment like BMT gives patients a better chance. "If we knew the answer," says Winer, "we wouldn't have been doing the trial." Now those trials are over and we still don't know.
Choosing whether to roll the dice with a randomized study or put your faith in an unproven treatment tests both the nerve and the conscience. When our son, James, tested positive for an antibody suggesting he might be susceptible to Type 1 diabetes, we consulted both friends and specialists about how to proceed. Putting him in a trial program meant that he would either receive an unproven, potential, treatment that could delay the onset of diabetes, or enter the untreated control group.
The decision we had to make for James did not involve life and death--at least, not at the time--yet it brought out the same issues that face women who want the most aggressive treatment possible for their breast cancer. Was the price of knowledge too high for us? Would we consent to losing an opportunity to delay diabetes by several years so that future generations would have an answer? Everyone we knew and trusted advised us to jump out of the randomization process and take a chance on the treatment.
Fortunately, subsequent test results contradicted the first set, and the need to decide subsided. But I wonder what I would have done. How would we ever know whether the treatment worked, if we ourselves were unwilling to take the dark road toward an answer?
When we speak of becoming a statistic, it's usually with a feeling of anonymous suffering and despair. But lost in that phrase are the many brave individuals who risked or sacrificed their lives to acquire the knowledge that we need to treat lethal diseases effectively. Perhaps not all of us consider the study of statistics "a religious service." But recognizing the value of statistics, and the toil that goes into obtaining them, is to recognize the parting of two paths: one a return to the Dark Ages, and the other, toward an enlightenment that will last for generations.