Do gender and race influence survival in patients with non-small cell lung cancer brain metastases? An outcomes study utilizing the RTOG RPA class stratification

Abstract

7153 Background: To explore gender, race and their interactions in the setting of NSCLC brain metastases only, a single-institution brain database was analyzed, using the RTOG recursive partitioning analysis (RPA) brain metastases classification. Methods: From 1/82 to 9/04, 831 NSCLC pts with brain metastases were registered. RPA criteria for analysis were: class I- Karnofsky performance status (KPS) ≥ 70, age<65 years, primary tumor controlled, no extracranial metastases; class III- KPS<70; class II- all others. Results: Median follow-up was 5.4 months (m) (range 0–122.9). Median age was 62.4 (range 25–90). Median KPS was 80 (range 20–100). There were 485 males [M] (58.4%) and 346 females [F] (41.6%). 824 pts (99%) were either African-American (AA; n = 142[17%]) or White (W; n = 682[83%]). Pts characteristics were balanced when stratified by RPA class and by treatments. Median survival (MS) in months from time of brain metastasis diagnosis for all pts was 5.8. MS in months by gender [F vs. M] and race [W vs. AA] was: 6.3 vs. 5.5, p = 0.013; 6.0 vs. 5.2, p = 0.08, respectively. By RPA class for gender, MS trends (in months) favored F over M in classes I and II but not III: 17.1 vs. 9.5 (p = 0.11); 6.8 vs. 6.0 (p = 0.09), 2.7 vs. 2.5 (p = 0.42), respectively. By RPA class for gender and race, MS trends (in months) favored AAF over AAM in classes I and II but not III: 30.0 vs. 12.4, p = 0.50; 11.2 vs. 4.6, p = 0.021; 3.2 vs. 3.2, p = 0.64, respectively; and WF over WM in classes I but not II or III: 14.4 vs. 9.5, p = 0.11; 6.6 vs. 6.3, p = 0.38; 2.4 vs. 2.3, p = 0.49, respectively. On multivariable analysis, significant variables were gender (p = 0.041; RR 0.83); RPA class (p < 0.0001; RR 0.28, for I vs. III; p < 0.0001; RR 0.51, for II vs. III). Conclusions: Gender significantly influences NSCLC brain metastasis survival while race trends to significance. MS trends by RTOG RPA class suggest race may interact with genderprimarily in class I but pt numbers limited significance. Further characterization of these factors is warranted. No significant financial relationships to disclose.