Long-term survival and toxicity in small cell lung cancer. Expanded Southwest Oncology Group experience.

The Southwest Oncology Group formed a database of 2,501 patients consecutively enrolled in small cell lung cancer (SCLC) trials since 1976. This report summarizes an analysis of this database to determine predictors of 2- and 5-year survival in limited stage disease (LD) and 1- and 2-year survival in extensive stage disease (ED). In addition, we analyzed the frequency of late recurrences, toxicity, and quality of life issues in the long-term survivors. For consecutive studies, greater than or equal to 2-year survival in LD were 15 percent, 21 percent, 28 percent, and 43 percent; 5-year survivals were 5 percent, 9 percent, 8 percent, and 20 percent. In ED, greater than or equal to 1-year survivals were 27 percent, 23 percent, 21 percent, and 42 percent; greater than or equal to 2-year survivals were 6 percent, 5 percent, 3 percent, and 19 percent. Using the logistic regression multivariate model, independent favorable predictors of 5-year survival for patients accrued to our older LD trials were normal lactate dehydrogenase (LDH) values and good performance status. Therapy as employed in these trials was not an independent factor. However, if patients enrolled on more recent trials were included, 2-year predictors could be assessed. Therapy with concurrent chemoradiotherapy and female gender then became additional independent favorable predictors. In ED, a single metastatic site and a normal LHH value were favorable predictors of survival beyond 1 year. The retrospective review of 63 patients with LD who survived at least 5 years found 33 asymptomatic patients with no recurrent disease; 6 with recurrent SCLC, 3 of whom died; 7 who died of non-cancer-related causes or unknown causes; 3 who died of secondary primary lung cancer; and 14 alive with persistent central nervous system symptoms and signs, possibly due to prophylactic brain radiation as given in the first 3 trials. No increased incidence of this syndrome has yet been observed in subsequent trials. For ED patients, 25 of 51 survivors greater than or equal to 2 years subsequently died of recurrent SCLC. The majority of the long-term survivors with ED (35 of 51) had either a single metastatic site or metastases limited to opposite side of the chest or regional nodes. Our multivariate models support the conclusion that aggressive combined modality, concurrent induction therapy, along with favorable prognostic variables, independently contribute to the improved long-term survival we have observed in LD. Longer follow-up is required to confirm that this improvement has occurred with less late toxicity.