Influence of tumor volume on survival in patients irradiated for non-small-cell lung cancer

Purpose: To investigate the importance of CT-defined total tumor volume (TTV) on overall survival (OS) in patients with unresectable or medically inoperable non-small-cell lung carcinoma (NSCLC). Methods and Materials: Between 1991 and 1998, 150 evaluable patients with Stage I-IIIB NSCLC were treated with three-dimensionally planned conformal radiotherapy and curative intent at Duke University Medical Center. On the treatment-planning CT, the primary tumor and nodal volumes were identified and subsequently combined to form the TTV. The TTV was compared with the stage and outcome with respect to OS, local progression-free survival, and distant failure-free survival using the Kruskall-Wallis analysis of variance and Kaplan-Meier actuarial method. To account for the potentially confounding effects of therapeutic and patient-specific covariates on survival, the Cox proportional hazard regression model was used. Results: The TTVs in patients with Stage I disease (median 19 cm(3)) were smaller than in patients with Stage II (median 80 cm(3)) or Stage III (median 97 cm(3); p <0.001) disease. The Stage II TTVs were not significantly different from those of Stage III (post-hoc test according to Bonferroni). Prolonged OS was independently associated with a small TTV (<80 vs. >80 cm(3) [median]; p = 0.01), young age (<60 vs. >= 60 years; p = 0.03), high Karnofsky performance status (<= 70 vs. >70; p = 0.04), and female gender (p = 0.04). Both stage (p = 0.7) and T stage (p = 0.06) were of less importance for OS than was the TTV, according to multivariate modeling. Increased local progression-free survival (p = 0.001) and distant failure-free survival (p = 0.03) were independently associated with a small TTV (i.e., <80 cm(3)). The results were unchanged if the TTV was analyzed as a continuous variable. Conclusion: A strong independent association between a small CT-defined TTV and prolonged survival in patients with NSCLC selected for curative/definitive RT was found. Future therapeutic studies in NSCLC should consider stratifying/adjusting for differences in TTV to avoid confounding effects on survival from variations in the TTV at baseline. (C) 2002 Elsevier Science Inc.