Socioeconomic factors and outcomes among patients with recurrent/metastatic head and neck cancer receiving immunotherapy.

Abstract

153 Background: Immune checkpoint inhibitors (ICI) are now a therapeutic standard for recurrent/metastatic head and neck squamous cell cancer (R/M HNSCC). There is limited data on the impact of demographics and socioeconomic (SES) factors on outcomes in this population, and we sought to evaluate these in our single institution cohort. Methods: R/M HNSCC patients (pts) receiving ICI were retrospectively reviewed from an institutional database. SES factors included income, distance to center (dist), marital status, race, and insurance. Median household income by residence zip code was obtained from the US Census Bureau. Time to ICI initiation (TTI) was time from initial visit recommending ICI and first ICI dose. Opiate use was calculated using morphine equivalents prior to ICI initiation and either at best response or end of ICI if no response. Associations between SES factors with overall survival (OS) and TTI were assessed using Cox proportional hazards regression. Binary outcomes were assessed using logistic regression and included ER visits/unplanned hospitalizations (UH) and increase in opioid use. Analyses were adjusted for disease characteristics, smoking status, ECOG, and demographics. Results: Between 1/2012-12/2019, 152 pts received ICI; 124 (82%) were male, with median age of 64 years (range 23-90), and 103 (68%) were partnered/married. The most common races were 114 white (75%), 14 Asian (9%) and 6 Hispanic, any race (4%). Out of 149 (98%) insured pts, 27 (18%) were Medicaid and 69 (46%) Medicare. Median dist was 39 miles (Q1 21, Q3 100), and median income was $80,586 (Q1 $61,202, Q3 $103,059). The most common primary sites were oropharynx (36%), oral cavity (22%), and nasopharynx (7%); 29 (19%) had an ECOG ≥2. While on or within 100 days of ICI, 69 (45%) had ER visits, and 57 (38%) had UH. Increased dist was associated with improved OS (4th vs 1st quartile, p = 0.0002; HR 0.33; 95% CI [0.18,0.59]); we observed no other SES association with OS. Increased opioid use was associated with Medicaid/no insurance (p = 0.05; OR 2.89; 95% CI [1.02,8.77]). No SES association with TTI was found, although there was a nonsignificant trend of higher TTI with increasing dist. We saw no correlation with ER/UH and any SES variables. Conclusions: Among R/M HNSCC pts receiving ICI, insurance had an impact on opiate usage, suggesting more advance disease/higher burden of symptoms and indicating need for augmentation of supportive care in this group. Higher dist was associated with improved OS, even accounting for performance status, which may reflect increased resources in this group. Further studies should examine pt factors that may contribute to disparities in the setting of novel therapies for R/M HNSCC pts.