Refusal of Surgery in Pituitary Adenoma Patients: A Population-Based Analysis

SIMPLE SUMMARY: This study presents an evaluation of the clinical and sociodemographic factors predictive of surgery refusal in pituitary adenoma (PA) patients and the effect of surgical receipt on overall survival outcomes based on patient data from the National Cancer Database. To our knowledge, the impact of pituitary refusal on outcomes has not been investigated by a research team. Here, we identified age > 65, African American race, increased comorbidities, and government insurance or being uninsured as independent predictors of surgery refusal. We also demonstrate a significant decrease in overall survival in PA patients with macroadenoma who refuse surgery compared to those who receive surgery. We hope these findings can help physicians understand the sociodemographic factors that influence surgery refusal in PA patients, as well as the importance of surgery in appropriate patients with macroadenomas. ABSTRACT: We characterized the clinical and sociodemographic factors predictive of surgery refusal in pituitary adenoma (PA) patients. We queried the National Cancer Database (NCDB) to identify adult PA patients treated from 2004–2015 receiving or refusing surgery. Multivariate logistic regression and Cox proportional-hazards analysis identified clinical and/or sociodemographic factors predictive of surgery refusal or mortality, respectively. Of the 34,226 patients identified, 280 (0.8%) refused surgery. On multivariate logistic regression, age > 65 (OR: 2.64; p < 0.001), African American race (OR: 1.70; p < 0.001), Charlson-Deyo Comorbidity (C/D) Index > 2 (OR: 1.52; p = 0.047), and government insurance (OR: 2.03; p < 0.001) or being uninsured (OR: 2.16; p = 0.03) were all significantly associated with surgery refusal. On multivariate cox-proportional hazard analysis, age > 65 (HR: 2.66; p < 0.001), tumor size > 2 cm (HR: 1.30; p < 0.001), C/D index > 1 (HR: 1.53; p < 0.001), having government insurance (HR: 1.66; p < 0.001) or being uninsured (HR: 1.67; p < 0.001), and surgery refusal (HR: 2.28; p < 0.001) were all significant predictors of increased mortality. Macroadenoma patients receiving surgery had a significant increase in overall survival (OS) compared to those who refused surgery (p < 0.001). There are significant sociodemographic factors that influence surgery refusal in PA patients. An individualized approach is warranted that considers functional status, clinical presentations, and patient choice.