Clinical factors predictive of overall survival (OS) and the identification of prognostic groups in patients (pts) with unresectable stage III non-small cell lung cancer (NSCLC) treated with chemoradiotherapy on Cancer and Leukemia and Group B trial (CALGB) 39801.

7535 Background: CALGB 39801 was designed to test whether treatment with induction chemotherapy and concurrent chemoradiotherapy (arm B) would improve OS in comparison to identical chemoradiotherapy alone (arm A), and demonstrated no significant benefit in OS for induction therapy. The objective of this analysis was to identify factors predictive of OS, and to use relevant factors to dichotomize pts into prognostic groups. METHODS: Between July 1998 and May 2002, 331 pts were studied and included in a Cox proportional hazard regression analysis investigating previously identified prognostic factors: age (< 70 vs. ≥ 70 years), gender, race/ethnicity, hemoglobin (hgb) (< 13 vs. ≥13), performance status (PS) (0 vs.1), pretreatment weight loss (wt loss) (<5% vs. ≥ 5%), and treatment arm. RESULTS: Cox regression analysis identified weight loss ≥ 5%, age ≥ 70, PS of 1, and hgb < 13 as predictive of worse survival (p<0.05), but not treatment arm (p=0.55). The median survival for pts with 0 (n=66), 1 (n=100), 2 (n=100), or ≥ 3 (n=65) risk factors were 24, 18, 10, and 8 months, respectively (p=0.0001). The pts were dichotomized into "poor prognosis" (PP) defined as ≥2 factors (n=165) and "good prognosis" (GP) defined as ≤ 1 factors (n=166). The hazard ratio (HR) for overall survival for the PP in comparison GP was 1.88 (95% CI, 1.49 to 2.37; p-value < 0.0001); the median survival times (MST) observed were 9 and 18 months, respectively (p<0.0001). The reasons for discontinuing treatment, and the rates of hematologic and non-hematologic adverse events were similar between the two groups. In the PP group the OS was similar between arms A (n=82) and B (n=83) (HR=0.97, 95% CI, 0.70 to 1.4; p=0.34); MST of 8.7 and 9.5 months, respectively. In the GP the OS was similar between arms A (n=79) and B (n=87) (HR=0.86, 95% CI, 0.63 to 1.1; p=0.87); MST of 19.3 and 17.6 months, respectively. CONCLUSIONS: Factors predictive of OS can be used to dichotomize pts into prognostic groups. Induction chemotherapy was not beneficial in either prognostic group. No significant financial relationships to disclose.

Duke Scholars

Author James Emmett Herndon II Biostatistics & Bioinformatics, Division of Biostatistics

Author Michael John Kelley Medicine, Medical Oncology