Socioeconomic disparities in the receipt of radiation for node-positive prostate cancer.
53 Background: Radiation therapy in the setting of node-positive prostate cancer has been controversial, although some recent data suggests a survival benefit to radiation in this setting. We evaluated socioeconomic disparities in the receipt of radiation for node-positive prostate cancer to identify groups that may be less likely to receive this potentially life-saving treatment. Methods: We identified 3,283 patients with N1M0 prostate cancer diagnosed 1982-2011 using the Surveillance, Epidemiology, and End Results database who were treated with radiation or no local therapy. We conducted multivariable logistic regression to determine socioeconomic predictors of not receiving radiation treatment. Results: Several patient and demographic factors were associated with a reduced likelihood of receiving radiation: African American (AA) vs non-AA race (31.7% vs. 37.7%, adjusted odds ratio [AOR] 0.74, p = 0.012); unmarried vs married status (31.9% vs 38.6%, AOR 0.72, p < 0.001); bottom third vs top third in income level (33.7% vs. 39.8%, AOR 0.72, p < 0.001); age over 65 versus < = 65 years (34.6% vs 39.8%, AOR 0.81, p = 0.005); diagnosis before 2000 versus starting in 2000 (31.6% vs 43.5%, AOR 0.56, p < 0.001). In a separate analysis, patients under the age of 65 who had Medicaid or no insurance were less likely than patients with other insurance to receive radiation (43.5% vs 55.9%, OR 0.61, p = 0.041), although on multivariable analysis, no significant association persisted (p = 0.512). Conclusions: African American race, unmarried status, lower income level, older age, and insurance status were all associated with significantly reduced odds of receiving radiation therapy for node-positive prostate cancer compared with no local therapy. Given the accumulating data suggesting that radiation therapy can improve survival in node-positive patients, it is increasingly important to understand the reasons for these treatment disparities so that they can be reduced.