Prophylactic Cranial Irradiation (PCI) Improves Overall Survival (OS) in Patients with Limited Stage Small Cell Lung Cancer (SCLC) Without Lymph Node Involvement

Abstract

Roughly one in three patients present with limited stage at the time of diagnosis in SCLC, although the disease has a high propensity to metastasize to distant sites. Combined modality therapy has become the standard of care, and incorporation of PCI has been shown to improve OS in several studies and meta-analyses. Data are limited in patients who received PCI and who did not have evidence of lymph node involvement (N0). We retrospectively investigated the importance of prognostic factors, including receipt of PCI, in patients with limited stage N0 SCLC who were registered in the Surveillance, Epidemiology, and End Results (SEER) database. From the SEER database, we identified a total of 887 patients with limited stage SCLC without lymph node involvement, who were diagnosed from 1988-1997 (the years for which PCI was coded in the SEER registries). Kaplan-Meier analyses, log-rank tests, and multivariate Cox proportional hazards models were used to examine the impact of prognostic factors, including receipt of PCI, on OS. Median age at diagnosis was 69 years and median OS was 15 months. The majority of the patients were white (n=736, 83.0%) and male (n=495, 55.8%). Primary site of disease distribution consisted of upper lung (n=434, 48.9%), lower lung (n=236, 26.6%), main bronchus (n=76, 8.6%), middle lung (n=50, 5.6%), overlapping sites (n=14, 1.6%), and not otherwise specified (n=77, 8.7%). A minority of patients (n=66, 7.4%) underwent PCI as part of their initial therapy. Most patients had T2 disease (n=257, 29.0%). With multivariate analysis accounting for age at diagnosis, race, sex, primary site of disease, year of diagnosis, and T stage, patients who were coded as having received PCI had improved OS (HR=0.62, p<0.001) with median survival of 37 months compared to 14 months for those who did not. Patient race, sex, or site of primary disease had no significant prognostic impact on survival. For the 10 year period in our study, those who were diagnosed from 1993-1997 had better OS (HR=0.86, p<0.05) than those diagnosed from 1988-1992. Finally, patients with T1 disease (HR=0.72, p<0.01) had improved OS compared with patients with larger and/or invasive tumors. In patients with limited stage SCLC who were coded as having N0 disease at the time of initial staging, PCI is associated with improved OS in our hypothesis-generating study. Limitations in these analyses include inability to account for selection biases, under-reporting receipt of PCI, and lack of data on chemotherapy use. These results are consistent with previous work that supports the use of PCI in patients with limited stage SCLC and may help to inform future discussions regarding prevention of metastases to the brain.