Related Pages Learning About Clinical Trials What clinical trials are, how they work, and other educational materials for patients and professionals.

Key Words

Clinical trial, phase 1, cytotoxic, targeted therapy, side effects, safety. (Definitions of many terms related to cancer can be found in the Cancer.gov Dictionary.)

Summary

The risk of a cancer patient dying from an experimental treatment while participating in a phase 1 clinical trial decreased about tenfold between 1991 and 2002, according to a study led by researchers from Harvard Medical School. During the same period, response rates to treatment (as measured by tumor shrinkage) also decreased, though by a much smaller amount. Researchers suggest that both these trends may be explained in part by the emergence of new targeted cancer drugs that are less toxic and that often work by stopping tumor growth rather than by shrinking tumors.

Source

Journal of the American Medical Association, November 3, 2004.

Background

Phase 1 clinical trials are the earliest studies of new treatments in people. The goals of a phase 1 trial are usually to find out whether an experimental treatment is safe in people and to establish the best dose to test in larger studies.

Participation in phase 1 trials can involve significant risks and uncertain benefits. Because the experimental treatment has not been previously tested in people, little is usually known about its possible adverse effects. What’s more, patients who enroll in phase 1 trials of new cancer treatments typically have advanced disease that is no longer responding to conventional treatment. Past estimates have suggested that only about five percent of participants benefit from the treatment they receive in a phase 1 trial.

However, the conduct of phase 1 trials has changed in recent years, raising the question of whether they have become safer or perhaps more beneficial for participants. The current study was an effort to look into that question.

The Study

Researchers reviewed data from 213 phase 1 trials involving 6,474 cancer patients. The trials were selected from those submitted for presentation at the annual meeting of a major cancer society between 1991 and 2002. Each trial tested a single experimental drug – either a conventional drug (called cytotoxic), which kills both normal and cancer cells, or a so-called targeted drug, which aims more directly at cancer cells, leaving most normal tissue alone. Targeted drugs tend to cause fewer side effects (called “adverse” or “toxic” reactions).

The investigators looked at the “toxic death rate” - that is, how many patients died from the adverse effects of the experimental drug. They also looked at how many patients responded to treatment, using the standard definition of response: tumor shrinkage of more than 50 percent. The study team was led by Thomas G. Roberts, Jr., M.D., of Harvard Medical School in Cambridge, Mass.

Results

A total of 231 patients (2.1 percent) died while they were enrolled in a phase 1 trial included in the study. Of these, 35 deaths (0.54 percent of all the patients involved in the 213 trials) were caused by the experimental drug.

The toxic death rate declined from 1.1 percent during the first four years covered by the study (1991 to 1994) to 0.06 percent during the last four years (1999 to 2002). Thus, the chance that a patient would die from the experimental treatment was 10 times as great during the first four years studied as during the last four years. Patients treated with targeted drugs were one-fourth as likely to die from the treatment as patients treated with cytotoxic drugs.

A total of 243 patients (3.8 percent) responded to treatment according to the traditional definition of “response” - that is, their tumors shrank by more than half. Overall, the response rate declined from 6.2 percent during the first four years covered by the study to 2.5 percent during the last four years. Thus, patients were less than half as likely to respond to treatment (as measured by tumor shrinkage) during the last four years studied (1999 to 2002) as during the first four years (1991 to 1994).

However, note the researchers, this does not necessarily mean the later trials were less beneficial to participants. Tumor shrinkage may not accurately reflect the beneficial effect of targeted drugs. Many targeted drugs stop the tumor from growing rather than cause the tumor to shrink. For this reason, the researchers suggest that the time it takes a tumor to start growing again (“time to progression”) might be a better measurement of the benefit of targeted drugs.

The authors suggest three possible reasons for the decline in treatment-related deaths over the study period.

• Almost half (47 percent) of the trials included in the study involved targeted drugs, which tend to have fewer adverse effects than conventional cytotoxic drugs.
• Supportive care for cancer patients improved over the study period. For example, new and better drugs were introduced to combat severe nausea and infections, common side effects of conventional cancer treatment.
• Oversight of cancer clinical trials became more stringent, resulting in more careful scrutiny of patients prior to enrollment. Thus, patients with a higher risk of dying from experimental treatment were less likely to be enrolled in a phase 1 clinical trial.

Limitations

The researchers looked at only those phase 1 trials that tested a single experimental drug. They did not look at trials in which an experimental drug is added to standard therapy, a common practice in cancer clinical trials. They also excluded trials in which patients received radiation therapy in addition to an experimental drug. Including these trials in the analysis might have found a higher rate of toxic death.

In addition, the trial results selected for inclusion in the review had all been published in medical journals. This may have resulted in “publication bias,” because studies with unfavorable outcomes are less likely to be published, say the authors of an accompanying editorial, Eric X. Chen, M.D., Ph.D., and Ian F. Tannock, M.D., Ph.D., of the University of Toronto in Canada.

Comments

The finding that phase 1 trials are safer should not be interpreted to mean that patients who take part in these trials are more likely than in the past to be cured or to live longer - the study did not address those issues.

Nonetheless, says James Doroshow, M.D., director of the Division of Cancer Treatment and Diagnosis at the National Cancer Institute, “this study demonstrates that the risk of participating in early clinical trials has decreased substantially. The fact that traditional response criteria [tumor shrinkage] were used means that the potential clinical benefit of the therapies studied - in terms of quality of life benefit or prolonged disease stabilization - was probably underreported.”

Back to Top