PREDICTORS OF SURVIVAL FOLLOWING RELAPSE OR PROGRESSION OF SMALL-CELLLUNG-CANCER - SOUTHWEST-ONCOLOGY-GROUP STUDY 8605 REPORT AND ANALYSISOF RECURRENT DISEASE DATA-BASE

Background. Survival after relapse or progression of small cell lung c ancer (SCLC) is poor. The Southwest Oncology Group (SWOG) initiated a study of modulation of cyclophosphamide (Cy) resistance in this popula tion. At study closure, the value of testing new regimens in previousl y treated patients was being debated nationally: Is there an independe nt impact of treatment over favorable prognostic factors? Thus, the au thors analyzed the SWOG recurrent SCLC data base. Methods. A 12-hour i nfusion of cytosine arabinoside (ara-C) was used as a potential repair inhibitor of Cy-induced DNA damage in patients with relapsed SCLC. A data base of successive SWOG studies in recurrent SCLC was formed. The independent contribution to survival of prognostic factors, type of p rior chemotherapy (CT), time from diagnosis, and type of CT on relapse were then assessed.Results. There were 3 partial responses observed i n 67 patients, with substantial myelotoxicity. The median survival was 2.5 months, and 16% lived beyond 6 months. The multivariate analysis of the recurrent SCLC data base found that a normal lactate dehydrogen ase (LDH) and second-line treatment with etoposide plus cisplatin (EP) , if not initially treated with either alternating or complex multidru g regimens, were the only independent predictors of improved survival. The 2-year survival of this subset was 20%. Conclusions. The Cy/ara-C program cannot be recommended for patients with recurrent SCLC. Howev er, EP independently contributed to improved survival in patients with out complex prior CT and a normal LDH. This finding supports future tr ials of new approaches in certain subsets of SCLC patients with limite d prior treatment.