Evaluation of Inequities in Cancer Treatment Delay or Discontinuation Following SARS-CoV-2 Infection

Abstract

Importance: There is a disproportionately greater burden of COVID-19 among Hispanic and non-Hispanic Black individuals, who also experience poorer cancer outcomes. Understanding individual-level and area-level factors contributing to inequities at the intersection of COVID-19 and cancer is critical.

Objective: To evaluate associations of individual-level and area-level social determinants of health (SDOH) with delayed or discontinued cancer treatment following SARS-CoV-2 infection.

Design, setting, and participants: This retrospective, registry-based cohort study used data from 4768 patients receiving cancer care who had positive test results for SARS-CoV-2 and were enrolled in the American Society for Clinical Oncology COVID-19 Registry. Data were collected from April 1, 2020, to September 26, 2022.

Exposures: Race and ethnicity, sex, age, and area-level SDOH based on zip codes of residence at the time of cancer diagnosis.

Main outcomes and measures: Delayed (≥14 days) or discontinued cancer treatment (any cancer treatment, surgery, pharmacotherapy, or radiotherapy) and time (in days) to restart pharmacotherapy.

Results: A total of 4768 patients (2756 women [57.8%]; 1558 [32.7%] aged ≥70 years at diagnosis) were included in the analysis. There were 630 Hispanic (13.2%), 196 non-Hispanic Asian American or Pacific Islander (4.1%), 568 non-Hispanic Black (11.9%), and 3173 non-Hispanic White individuals (66.5%). Compared with non-Hispanic White individuals, Hispanic and non-Hispanic Black individuals were more likely to experience a delay of at least 14 days or discontinuation of any treatment and drug-based treatment; only estimates for non-Hispanic Black individuals were statistically significant, with correction for multiple comparisons (risk ratios [RRs], 1.35 [95% CI, 1.22-1.49] and 1.37 [95% CI, 1.23-1.52], respectively). Area-level SDOH (eg, geography, proportion of residents without health insurance or with only a high school education, lower median household income) were associated with delayed or discontinued treatment. In multivariable Cox proportinal hazards regression models, estimates suggested that Hispanic (hazard ratio [HR], 0.87 [95% CI, 0.71-1.05]), non-Hispanic Asian American or Pacific Islander (HR, 0.79 [95% CI, 0.46-1.35]), and non-Hispanic Black individuals (HR, 0.81 [95% CI, 0.67-0.97]) experienced longer delays to restarting pharmacotherapy compared with non-Hispanic White individuals.

Conclusions and relevance: The findings of this cohort study suggest that race and ethnicity and area-level SDOH were associated with delayed or discontinued cancer treatment and longer delays to the restart of drug-based therapies following SARS-CoV-2 infection. Such treatment delays could exacerbate persistent cancer survival inequities in the United States.

Conflict of interest statement

Conflict of Interest Disclosures: Dr Llanos reported receiving grant funding from Conquer Cancer, the American Society for Clinical Oncology (ASCO) Foundation, during the conduct of the study. Dr Ganesan reported receiving personal fees from EQRx Inc, Silagene Inc, Merck & Co Inc, and Kayothera; grant funding from Gandeeva Therapeutics and M2Gen; and having a spouse employed by Merck & Co Inc, outside the submitted work. No other disclosures were reported.

Figures

Figure.. Adjusted Survival Curves of Time to Restart Pharmacotherapy Following a Confirmed Positive SARS-CoV-2 Test Result, Stratified by Race and Ethnicity

Includes 857 patients with cancer. A Cox proportional hazards regression analysis was performed to assess differences in the restart rate of the first or only anticancer drug treatment patients were scheduled to receive, by race and ethnicity, with adjustment for age and cancer type. Compared with non-Hispanic White patients, the restart rate was 19% less (hazard ratio [HR], 0.81 [95% CI, 0.67-0.97]; P = .03) for non-Hispanic Black patients, 13% less (HR, 0.87 [95% CI, 0.71-1.05]; P = .15) for Hispanic patients, and 21% less (HR, 0.79 [95% CI, 0.46-1.35]; P = .39) for non-Hispanic Asian American or Pacific Islander patients.