EP750 - Differences in Cumulative Incidence of Early Age of Onset Rectal Cancer in Hispanics Disaggregated by Race and Hispanic Background

Purpose/Background: Given that rectal cancer (RC) accounts for a large portion of early age-of-onset (EO) colorectal cancer (CRC) cases and since Hispanics are a heterogeneous population, we wanted to determine: (1) if the cumulative incidence of EO RC varies among Hispanics of varying race and ethnicity (R&E) or Hispanic background (HBackground) and (2) the potential impact of socioeconomic determinants of health (SDOH) on cumulative incidence of EO RC.

Methods/Interventions: The National Cancer Database was queried for adult RC patients diagnosed between 2004 and 2021. Adjusted incidence rate ratios (AIRR) were generated using generalized linear models with Poisson distribution and log link to assess risk factors associated with EO RC. Using non-Hispanic white (NHW) as our reference group, two models were used to assess the impact of SDOH on AIRR of EO RC versus average age-of-onset (AO) RC in Hispanics disaggregated by R&E and HBackground. Model 1 adjusted for facility type, year of diagnosis, and Charlson Deyo comorbidity score, while Model 2 additionally adjusted for patient insurance status as well as neighborhood income and education.

Results/Outcomes: Overall, 263,711 patients with RC were identified. Of these, 18,530 were Hispanics. The proportion of Hispanic patients with EO RC was 10.1%, while that of Hispanics with AO RC was 6.3% (p <.0001). Compared to NH patients, Hispanics in aggregate had 1.35 (95% CI: 1.31:1.40) times higher EO RC cumulative incidence (p <.0001). In Hispanics disaggregated by R&E, EO RC cumulative incidence was higher in all subgroups compared to NHW (AIRR: Hispanic White 1.39 [95% CI: 1.34: 1.44]; Hispanic Black 1.59 [95% CI: 1.27:1.99], and Hispanic Other 1.57 [95% CI: 1.40:1.75], p<.0001). Similarly, EO RC cumulative incidence in Hispanics disaggregated by HBackground was higher in most subgroups compared to NHW (AIRR: Mexicans 1.51 [95% CI: 1.40:1.63]; South/Central American 1.45 [95% CI: 1.28:1.64]; not-otherwise specified [NOS] Spanish 1.43 [95% CI: 1.37:1.49]; other specified Hispanic origin 1.28 [95% CI: 1.02:1.60], and Spanish surname only 1.38 [95% CI: 1.13:1.68], p< 0.05). However, in some HBackground subgroups, the EO RC cumulative incidence was lower (AIRR: Cubans 0.77 [95% CI: 0.59:0.99], p=0.045) compared to NHW while in some no difference in EO RC cumulative incidence was noted (AIRR: Puerto Ricans 0.96 [95% CI: 0.79:1.17] and Dominicans 1.01 [95% CI: 0.74:1.38]). After controlling for SDOH, disparities in subgroups defined by ethnicity or R&E persisted. However, disparities defined by HBackground became insignificant in most subgroups except Mexican, NOS Spanish and South/Central American patients (p <.0001; p<.0001 and p=0.009, respectively). (Table)

Conclusion/Discussion: Hispanic patients are disproportionately affected by EO RC. Disaggregation by HBackground and R&E demonstrates differential cumulative incidence of EO RC amongst varying subgroups of Hispanic patients. When controlling for SDOH, these differences were mitigated in most HBackground subgroups, except in Mexican and South/Central American patients. Therefore, studies aimed at identifying unique environmental, genetic, and cultural interactions, affecting specific subgroups not currently captured by traditional SDOH are warranted.

Table
Adjusted incidence rate ratios for early age of onset versus average age of onset rectal cancer. AO RC: Average age of onset rectal cancer, defined as age over 50 years at diagnosis; EO RC: Early age of onset rectal cancer, defined as younger than 50 years at diagnosis; F: Frequency; AIRR: Adjusted incidence rate ratios; CI: Confidence interval; Ref: Reference group; NHW: Non-Hispanic White; HW: Hispanic White; HB: Hispanic black; HO: Hispanic other; NOS Spanish: Not-otherwise specified Spanish
1 Adjusted for: facility type, year of diagnosis, and Charlson Deyo comorbidity
2Adjusted for: facility type, year of diagnosis, Charlson Deyo comorbidity, neighborhood income, neighborhood education, and insurance status
3 This is the variable that the National Cancer Database labels “Country of Origin”
*Statistically significant value