UCSF

Lung neuroendocrine tumors (NETs) have few known predictors of survival. We investigated associations of sociodemographic, clinicopathologic, and treatment factors with overall survival (OS) and lung cancer-specific survival (LCSS) for incident lung NET cases (typical or atypical histology) in the California Cancer Registry (CCR) from 1992 to 2019. OS was estimated with the Kaplan-Meier method and compared by sociodemographic and disease factors univariately with the log-rank test. We used sequential Cox proportional hazards regression for multivariable OS analysis. LCSS was estimated using Fine-Gray competing risks regression. There were 6038 lung NET diagnoses (5569 typical, 469 atypical carcinoid); most were women (70%) and non-Hispanic White (73%). In our multivariable model, sociodemographic factors were independently associated with OS, with better survival for women (hazard ratio (HR) 0.62, 95% confidence interval (CI) 0.57-0.68, P < 0.001), married (HR 0.76, 95% CI 0.70-0.84, P < 0.001), and residents of high socioeconomic status (SES) neighborhoods (HRQ5vsQ1 0.73, 95% CI 0.62-0.85, P < 0.001). Compared to cases with private insurance, OS was worse for cases with Medicare (HR 1.24, 95% CI 1.10-1.40, P < 0.001) or Medicaid/other public insurance (HR 1.45, 95% CI 1.24-1.68, P < 0.001). In our univariate model, non-Hispanic Black Californians had worse OS than other racial/ethnic groups, but differences attenuated after adjusting for stage at diagnosis. In our LCSS models, we found similar associations between sex and marital status on survival, but no differences in outcomes by SES or insurance. By race/ethnicity, American Indian cases had worse LCSS. In summary, beyond disease-related and treatment variables, sociodemographic factors were independently associated with survival in lung NETs.