Non-receipt of targeted and immunotherapies in diverse patients with metastatic cancers.

Abstract

93 Background: The rapid evolution of molecularly targeted kinase inhibitors (TKIs) and immune checkpoint inhibitors (ICIs) in the treatment of metastatic cancer, coupled with high drug costs, raises potential for unwarranted variation in patient management. We evaluated correlates of non-receipt of TKIs and ICIs in a diverse, population-based sample of patients with metastatic non-small cell lung cancer (NSCLC), kidney and bladder (GU) cancer, and melanoma. Methods: We identified 2240 patients diagnosed in 2019 with metastatic NSCLC, GU cancer, or melanoma and reported to the Georgia or Los Angeles SEER registries. Using SEER registry data, we linked all patients to a managing oncology physician. We collected data from the managing physician regarding receipt of TKIs and ICIs. We used bivariate analyses and multivariable logistic regression models to assess associations between non-receipt of TKIs and ICIs and patient-level factors. Results: We present results for a preliminary sample (N=894). About half (55%) of the patients had NSCLC, 31% had GU cancer, and 14% had melanoma. The mean age was 62 (SD 12); 59% were male; 65% were White, 16% were Black, 11% were Hispanic, and 7% were Asian; 52% had Medicare, 29% had private insurance, 9% had Medicaid, and 4% had no insurance; 89% resided in urban areas, 11% resided in rural areas. 22% lived in a census tract area with >20% of the population living below the poverty level. The Table shows that in bivariate analyses, non-receipt of TKIs was associated with NSCLC, older age, male sex, non-white race, and non-private or no insurance (all p<0.05). After adjustment, non-receipt of TKIs was associated with NSCLC, older age, male sex, and race. Asian and Hispanic patients had lower rates of non-receipt, and Black patients had higher rates. In bivariate analyses, non-receipt of ICIs was associated with NSCLC and GU cancer (p=0.004); this association remained significant after adjustment. Conclusions: Non-receipt of TKIs was associated with important non-clinical factors including race and sex. Non-receipt of ICIs was associated only with cancer type. These findings suggest that policy-relevant factors such as out-of-pocket costs for oral cancer therapies and differential access to molecular sequencing may contribute to unwarranted variation in receipt of TKIs. Bivariate analyses of non-receipt. Did not Receive TKI Did not Receive ICI % P % P Cancer type NSCLC GU Melanoma 786572 0.002 313015 0.004 Age <50 50-64 65-74 75-84 85+ 5471757986 <0.001 2625302940 0.385 Sex Male Female 7670 0.048 2631 0.085 Race & Ethnicity White Black Hispanic Asian Other 7579655860 0.002 2329312927 0.981 Marital Status Married Unmarried 7373 0.645 2630 0.158 Insurance Medicare Medicaid Private Other government None 7075637080 <0.001 3130215031 0.101 Geography Urban Rural 7375 0.784 2826 0.567 % population below poverty level 0-5% 5-<10% 10-<20% 20-100% 69717774 0.314 29273222 0.138