Comparison of patient characteristics and outcome between a single-institution phase II trial and a cooperative-group phase II trial with identical eligibility in metastatic melanoma

Differences in overall survival and response rates are often noted when promising single-institution phase II treatment regimens are evaluated in a cooperative group setting. One reason for this discrepancy may be the differences in patient characteristics at the time of registration. In the metastatic melanoma literature, the prognostic factors for survival that are most frequently identified are the number of metastatic sites, visceral involvement, performance status, liver involvement, and possibly the disease-free interval and gender. A prognostic factor for response appears to be sites of involvement. A comparison of patient characteristics and outcome was conducted for patients entered in similar phase II melanoma trials at a single institution versus those in a cooperative group. Sixty-four patients at Wayne State University (WSU) were compared with 96 patients who had nearly identical eligibility criteria and were registered in the Southwest Oncology Group (SWOG). All patients were receiving comparable phase II treatments for metastatic melanoma. Southwest Oncology Group patients were significantly older (p < 0.001), had worse performance status (p = 0.03), had more visceral involvement (p = 0.001), and were more likely to have two or more metastatic sites (p = 0.02). No significant differences in gender (p = 0.55), absence or presence of liver involvement (p = 0.12), or disease-free interval were noted. These disparities, despite similar eligibility, may partly explain the observed differences in survival (WSU median = 10 months, SWOG median = 7 months, p = 0.13) and response rates (WSU = 31%, SWOG = 15%: p = 0.02) between the two groups of patients. Investigators should report these important patient characteristics in treatment reports. These differences highlight the difficulty in comparing single-institution and cooperative-group phase III trials, even with comparable patient eligibility. This serves to emphasize the need for well-designed phase III trials when comparing treatment approaches and stratification for the prognostic factors identified.