Ethnicity and socioeconomic deprivation in early phase clinical trials in a UK tertiary referral centre

Abstract

Background Early phase trials evaluate the safety and efficacy of new treatments. To capture potential differences in responses and toxicities, trial cohorts should represent the heterogeneity of the overall population. We aimed to examine the ethnicity and socioeconomic deprivation of patients (pts) referred to the Experimental Cancer Medicine Team (ECMT) at the Christie NHS Foundation Trust. Methods Retrospective analysis was done on pts referred from Jan 2018 - Dec 2019. Data on ethnicity, postcodes, response as per RECIST, toxicities and overall survival (OS) were collected. Using Office for National Statistics data we compared ethnicity distributions with national proportions. Postcodes were used to generate indices of multiple deprivation 2019 (IoMD); 1 most and 10 least deprived. Analyses were performed using univariable and multivariable models (UVA & MVA). Contribution of the 7 IoMD domains were studied using rank correlation matrices. Results Of 1375 referrals, 247 pts were enrolled onto trials. 1060 referred pts and 231 trial pts had ethnicity recorded. 93% of pts referred to ECMT were White British compared to 78% in England, 85% in the North West and 77% in Greater Manchester (p<0.001). In UVA, pts in IoMD deciles 1-3 had median OS of 14.3m (95% CI 10.6 - NR) compared to 7.7m (95% CI: 5.8-10.4) in deciles 4-10 (HR 2.03 95% CI 1.32-3.13 p=0.001). On adjusting for response in MVA, this association between IoMD decile and OS was not seen (HR 1.33 95% CI 0.81-2.19, p=0.262). Considering IoMD domains in UVA, decile correlates highly with each domain aside from barriers to housing and living environment. MVA, after adjusting for IoMD decile, reveals that higher the living environment decile, poorer the OS (HR 1.14 95% CI 1.04-1.24 p=0.004) but no association was noted for barriers to housing. Similar to IoMD decile, this association weakens after adjusting for response (HR = 1.08 95% CI 0.98-1.17 p = 0.108). No association was noted between IoMD and toxicity. Conclusions Pts from non-white backgrounds were underrepresented in referrals to ECMT. Outreach to these communities is needed to explore barriers to referral and enrolment. Further work is required to study the association between IoMD, its constituent domains such as living environment and OS.