Editor’s Note
	Two Women

The plethora of “P” values and Kaplan-Meier curves permeating the oncology research literature sometimes makes it easy to forget that the building blocks of clinical trials are people — doctors, nurses and most importantly, patients. In this issue, Kathy Miller presents two women who participated in Phase III randomized trials that had a fundamental impact on our understanding of breast cancer treatment.

The first was a 62-year-old woman who enrolled in the ECOG-1193 trial. This classic trial, which was led by Kathy’s colleague and mentor, George Sledge, addressed the critical question of combination versus sequential chemotherapy in women with metastatic disease. For years this study has been presented and mentioned at oncology meetings, but the definitive paper was only recently published in the February 15, 2003 issue of the Journal of Clinical Oncology. While the paper concludes that sequential single-agent chemotherapy results in the same overall survival as combination chemotherapy, Kathy’s case reveals the challenges in incorporating data from a patient’s individual course to a clinical trial.

This woman was randomized to the single-agent arm, and there was essentially no response to the first agent (doxorubicin) and a modest response to the crossover (paclitaxel), which also caused significant toxicity. However, after the primary randomization, major longstanding complete and near-complete responses were induced with anastrozole, and then capecitabine. The patient eventually died from an unrelated cerebrovascular event while experiencing excellent tumor control from capecitabine. Dr Miller noted that while this woman’s extended survival contributed to the single-agent randomization arm of ECOG-1193, it was the post-trial therapy that seemed to have the greatest effect in prolonging her life.

One can argue that large numbers of patients accrued to a study will obviate outlying clinical events such as these, but any tumor board meeting will provide more than adequate testimony to the heterogeneity of breast cancer, particularly in the metastatic setting. Kathy’s second case, presented at the 2002 San Antonio Breast Cancer Symposium “Meet the Professor” session, demonstrates another critical point about interpreting clinical research.

This 49-year-old woman was enrolled in a historic study — the first major randomized breast cancer clinical trial evaluating an antiangiogenic agent. Many attendees at the San Antonio Breast Cancer Symposium were disappointed that this trial failed to demonstrate its primary endpoint — a time to progression advantage for the combination of capecitabine and bevacizumab compared to capecitabine alone. In the interview for this program, Dr Miller presents the provocative course of her patient with chest wall recurrence to highlight the complexities of interpreting data from clinical trials in patients with metastatic disease.

To Dr Miller’s eyes, this woman’s tumor had a rapid and extremely impressive objective complete response to single-agent capecitabine (see photos below), and the symptoms from her aggressive tumor also completely abated. However, the external review board — evaluating the photos and clinical notes — called this “stable disease.” Kathy concedes that based on the very conservative trial guidelines for external review, this was a correct interpretation, but this case vividly portrays the complexity of determining the antitumor effect of therapies in clinical trials.

In an era of “evidence-based” medicine, clinicians should consider that the foundation for clinical research is the individual patient and that complex biopsychosocial variables make clinical research a less exact science than laboratory investigation. Ultimately, patients and physicians in daily practice routinely confront a panoply of imperfect trial data that must be judiciously evaluated in the context of each patient’s needs and values.

—Neil Love, MD

Home · Search