Modeling genetic testing (GT) attitudes in a socio-demographically diverse cohort of patients (pts) with cancer: The importance of race/ethnicity and medical mistrust (MM).

Abstract

10562 Background: Germline GT and tumor genomics are relevant to the care of nearly every cancer pt. These tests inform treatment planning by NGS-based tumor genomic profiling (TGP) and hereditary risk assessment. Socioeconomic and race-based disparities in GT are well-known, but the impact of growing MM, fueled by political polarization and social unrest due to overt and institutional racism, has been less studied. In previous research of African American (AA) cancer pts, we identified a consistent association of high levels of MM with negative attitudes toward GT. To inform development of an intervention to support informed decisions about GT in cancer pts, we further characterized the relationship of MM to GT attitudes. Methods: Pts from two Fox Chase Cancer Center (FCCC) sites were recruited (6/2022-12/2022) to complete a cross-sectional survey (n=105, Burholme campus; n=95, Temple University Hospital). A stratified purposive sampling method was employed. Data on pt demographics and disease factors were collected. MM was queried with two validated measures: LaVeist MM Index (MMI: range 0-10, measures healthcare/institution MM) and Thompson Group Based MM Scale (GBMMS: range 1-5, measures race/ethnicity specific MM; suspicion, discrimination, and lack of support subscales). Attitudes toward GT (GTpros=6 items/GTcons=11 items) were measured w/a validated scale (Morren: range 0-10) plus 4 items developed by our team. The study was approved by the FCCC IRB 22-8003. Results: Median age of pts was 62 yrs (range 25-91); 62% were female, 47% AA, 15% Hispanic, 42% married, 47% ≤high school grad, 52% had household income ≤$50K/yr. Mean GBMMS was 2.11 (SD 0.80) and mean MMI was 4.63 (SD 1.54). Mean of GTpros was 8.77 (SD 1.68); GTcons 4.16 (SD 2.32). MM was associated with race/ethnicity by GBMMS (p<0.001) and MMI (p<0.04), as were GBMMS subscales (suspicion, discrimination, lack of support, all p<0.001). Race/ethnicity was not associated w/GTpros (p=0.59) but was w/GTcons (p<0.001). In univariate models both GTpros and GTcons were associated with GBMMS (p=0.005 and p<0.001, respectively) and MMI (p=0.002 and p<0.001, respectively). In a multivariable model examining GTcons, GBMMS score remained significant (p<0.001) even when controlling for race (p<0.01) and an interaction btw race*mistrust. A similar model showed MMI score also remained significant (p=0.04) when controlling for race and interaction effects. Only the GBMMS suspicion subscale remained significant (p<0.02) in a model with both race and an interaction term. Conclusions: MM is markedly more common among racial/ethnic minority cancer pts, but the relationship btw MM and GT attitudes is only partially rooted in race/ethnicity. Oncologists considering germline GT or tumor genomics for pts, and efforts to support informed decision-making, must consider the pervasiveness of MM.