Adverse consequences of the COVID-19 pandemic on breast cancer stage distribution and breast cancer disparities.

Abstract

Background: The COVID-19 surge in March 2020 resulted in a hiatus placed on screening mammography programs in support of shelter-inplace mandates and diversion of medical resources to pandemic management. The COVIDrelated economic recession and ongoing social distancing policies continued to influence screening practices after the hiatus was lifted. We evaluated the effect of the hiatus on breast cancer stage distribution on the diverse patient population of a health care system in New York City, the first pandemic epicenter in the United States. Methods: Breast cancer patients diagnosed January 1, 2019 to December 31, 2020 were analyzed, with comparisons of stage distribution and mammography screen-detection for three intervals: Pre-Hiatus, During Hiatus (March 15, 2020 to June 15, 2020), and Post- Hiatus. Results were stratified by African American (AA), White American (WA), Asian (As) and Hispanic/Latina (Hisp) self-reported racial/ethnic identity. Results: A total of 894 patients were identified;of these, 549 WA, 100 AA, 104 As, and 93 Hisp comprised the final race/ethnicity-stratified study population. Overall, 588 patients were diagnosed Pre-Hiatus, 61 During-Hiatus, and 245 Post-Hiatus. Nearly two-thirds (65.5%) of the Pre- Hiatus cases were screen-detected versus 49.2% During-Hiatus and 54.7% Post-Hiatus (p = 0.002). Frequency of tumors diagnosed < 1 cm declined from 41.9% Pre-Hiatus to 31.7% Post-Hiatus (p = 0.035). WA patients were more likely to have screen-detected disease compared to AA in the Pre-Hiatus period (69.1% vs. 56.1%;p = 0.05) but non-significantly more likely to have screendetected disease compared to As and Hisp patients (66.2% vs. 56.9%;p = 0.08). In the Post- Hiatus period, the frequency of screen-detected disease was highest among WA patients (63.0%) compared to all other racial/ethnic groups (AA;48.1%, As-33.3%, and Hisp-40%;p = 0.007). Similar patterns were observed for frequency of tumors diagnosed ≤1cm Pre-Hiatus (WA-44.3% vs AA-26%, p = 0.02;and vs. As-41.3%, Hisp-48%;p = 0.09), and Post-Hiatus (WA-37.7% vs. AA- 18.2%, As-30.8%, Hisp-23.5%;p = 0.25). Conclusions: The 3-month pandemicrelated mammography screening hiatus resulted in a more advanced stage distribution for New York City breast cancer patients, and worsened preexisting race/ethnicity-associated disparities, especially for AA pts.