Regression analyses of prognostic factors in metastatic malignant melanoma.

Patients entered into phase II trials in metastatic malignant melanoma should be carefully selected in order to ensure that they live long enough to permit a meaningful evaluation of the efficacy of a given drug. In this selection emphasis has been put on performance status. However, also for patients with a good performance status, survival is often short. The purpose of this study has been to identify supplementary prognostic factors as these could be of help in the design of phase II trials. From 1978-1986, 177 consecutive patients were given various chemotherapy regimens for metastatic malignant melanoma in the Norwegian Radium Hospital. About 92% had a performance status of ECOG 0-2. Median survival was 4.0 months (0-30 months). Multivariate survival analysis selected lactate dehydrogenase (LDH) greater than 450 U/l, presence of brain metastases, leukocyte count greater than 10 x 10(9)/l, and erythrocyte sedimentation rate (ESR) greater than 15 mm/h as significant prognostic factors indicating short survival with low probability of surviving 3 months. Patients with normal values of LDH, leukocyte count, and ESR had a median survival of 11.5 months with 94% surviving 3 months. We conclude that this information could have an impact on the design of phase II trials.