Prognostic impact of education level of patients with advanced non-small cell lung cancer enrolled in clinical trials

Background: Socioeconomic status can potentially affect prognosis of cancer patients. Our aim was to describe potential differences in demographic and clinical characteristics, treatment, and survival by education level in patients with advanced non-small cell lung cancer (NSCLC) enrolled in clinical trials of first-line treatment. Methods: Individual data of Italian patients with advanced NSCLC (stage IV, or IIIB with supraclavicular nodes or malignant pleural effusion), ECOG performance status (PS) 0-2, enrolled in four phase III randomized trials conducted between 1996 and 2005 were pooled. Information about education was available for 1680 of 1709 patients (98.3%). Patients were divided in two groups according to education level: high (patients with at least high school diploma) or low (those with less than high school diploma). Survival analyses were stratified by treatment arm within trial. Results: There were 312 (19%) and 1368 (81%) patients with high and low education, respectively. Education level was significantly different among birth cohorts, with a time-trend toward higher education level. Patients with high education were significantly younger (median age 65 vs. 70), were less frequently unfit at diagnosis (ECOG PS2 5% vs. 16%), and their tumor type was more frequently adenocarcinoma (47% vs. 37%). Number of treatment cycles received was not significantly different between education groups. Median survival was 9.4 and 7.6. months in high and low education, respectively (p= 0.012). At multivariable analysis, female sex, better PS and high education level (Hazard Ratio 0.85, 95%CI 0.73-0.99, p= 0.03) were independently associated with longer survival. Conclusions: In Italian patients enrolled in four randomized trials of first-line chemotherapy for advanced NSCLC, high education was significantly more frequent among younger patients, and was associated with lower proportion of PS2 patients. Education level did not significantly affect number of chemotherapy cycles received. Overall survival was longer in patients with high education, after adjustment for PS and other prognostic factors. The exact underlying mechanisms of the independent prognostic role of education level are substantially unknown, but lead-time bias (anticipation in diagnosis and time to inclusion in the trial), differences in adherence to care outside the trial procedures, differences in comorbidities and life-style factors may all contribute.