Association of Cancer with Risk and Mortality of COVID-19: Results from the UK Biobank

Although cancer has been associated with COVID-19 risk and mortality in hospital-based studies, few population-based studies have been reported. Utilizing data from the UK Biobank (UKB), a population-based prospective cohort, we formally tested the association of over 44 different types of cancer with COVID-19 infection and mortality among 7,661 subjects who were tested by June 17, 2020. Compared to non-cancer subjects, cancer subjects (N=1,521) had significantly lower overall risk for COVID-19 infection [odds ratio (OR) and 95% confidence interval (CI): 0.79 (0.68-0.92), P=2.60E-03]. However, a trend of higher risk for COVID-19 mortality was found among 256 COVID-19 positive cancer patients, especially for hematologic cancers such as non-Hodgkin lymphoma [3.82 (1.17-12.01), P=0.02]. In cancer patients, while few demographic, lifestyle, genetic and comorbidity factors predicted risk for COVID-19 infection, older age, male sex, heart disease and hypertension significantly predicted COVID-19 mortality. The lower risk for COVID-19 infection is likely due to extra caution in COVID-19 prevention and more testing among cancer patients, an encouraging finding that demonstrates the feasibility of intervention. These results, if confirmed in future releases of UKB data and other independent populations, may provide guidance for COVID-19 prevention and treatment among cancer patients.