Purpose: Various parameters have been reported to be correlated with response to interleukin-2 (IL-2) therapy A multiinstitutional study was performed to assess by multivariate analysis the predictive value of known clinical and biologic melanoma prognostic markers recorded before the onset of IL-2 therapy on the likelihood of objective clinical response. Patients and Methods: Serum C-reactive protein (CRP), IL-6, and lactate dehydrogenase (LDH) levels were measured in 81 metastatic melanoma patients included in different IL-2-based regimens before the starting of IL-2 therapy. Clinically defined prognostic groups, ie, patients with superficial or visceral metastases, were also analyzed for response correlates. Patients were evaluated for response to treatment 4 to 6 weeks after completion of one course of therapy. Results: On univariate analysis, the pretreatment values of CRP (P = .001), IL-6 (P = .007), and LDH (P = .02) and site of metastases (P = .0004) were correlated with clinical response. However, only CRP (P < .007) and clinically defined group (P < .004) were independent predictors on multifactorial analysis. Indeed, when adjusted to CRP, IL-6 fended to improve patient selection, but did not reach statistical significance (P = .07). Furthermore, using multivariate survival analysis based on the Cox proportional hazards model, only CRP was found to be an independent prognostic factor for survival (P < .0001) Conclusion: In this study, patients with high serum levels of CRP and/or visceral organ involvement before therapy were unlikely to respond to IL-2 therapy, Therefore, clinical classification based on the site of metastases and serum CRP determination before the start of IL-2 therapy may help to improve selection of melanoma patients who may benefit from IL-2 and could prevent unnecessary morbidity. (C) 1996 by American Society of Clinical Oncology.
