Association Between Fine Particulate Matter Exposure During Breast Cancer Treatment And Mortality Among Older Women in the United States

Background: Exposure to high fine particulate matter (PM2.5) level among breast cancer survivors has been associated with increased mortality. While biologically plausible, potential interaction between high PM2.5 level and cancer treatment has not been examined.

Objectives: To assess the association between elevated PM2.5 during cancer treatment and mortality among older women with breast cancer (BC).

Methods: Using SEER-Medicare linked with a high resolution PM2.5 data by residential zip code (2007-2016), we assembled a cohort of women aged ≥66 with new stage I-IV BC. We assessed PM2.5 exposure during treatment as < 8 µg/m3 (low), 8-12 µg/m3 (moderate) and >12 µg/m3 (high—above US air quality standard), and types of cancer treatments (surgery [SX], chemotherapy [CTX], radiation therapy [RTX]) during a 9-month landmark period to construct treatment subcohorts (SX, SX + RTX, SX + CTX, SX + RTX + CTX, and RTX and/or CTX). We fit multivariable Cox proportional hazard models to assess association between PM2.5 exposure and 5-year mortality, adjusting for individual-level characteristics (demographics, tumor characteristics, comorbidities), neighborhood SES, SEER state, and pre-treatment PM2.5 exposure. We used multiple imputation to account for missing variables.

Results: Among 65,871 women (mean age 76 years, 83% white, 56% stage I, 32% II, 9% III, 3% IV), 41%, 32%, 13%, 8.4%, and 5.1% received SX + RTX, SX only, SX + Chemo + RTX, SX + Chemo, and CTX and/or RTX. The mean PM2.5 level was 9.0 µg/m3 [34%, 54% and 12% were exposed to low, moderate, and high, respectively]. We observed 11,736 all-cause and 4,628 BC deaths during follow-up. Compared to those exposed to low PM2.5, the risk of mortality was HR 1.02 (95%CI: 0.999-1.05) for moderate and HR 1.10 (95%CI: 1.05-1.14) for high PM2.5. Among SX+RTX subgroup, PM2.5 was associated with higher all-cause mortality for moderate [HR 1.24 (95%CI:1.18-1.31)] and high PM2.5 [HR 1.38 (95%CI: 1.27-1.50)]. Among SX+RTX+CTX subgroup, high PM2.5 was associated with higher mortality [HR 1.33 (95%CI: 1.18-1.51)]. Among the SX+RTX subgroup, BC-specific mortality was elevated for both moderate [HR 1.50 (95%CI:1.36-1.66)] and high PM2.5 [HR 1.81 (95%CI: 1.55-2.12)].

Conclusions: Even below EPA regulatory PM2.5 threshold (12 µg/m3), we observed increased risk of death in women with BC who were exposed to moderate to high PM2.5 level during treatment period. The risk was primarily observed among women who received RTX. As both chest wall RTX and elevated PM2.5 can have acute and chronic effects on the respiratory and cardiovascular system, the increased mortality risk observed in our study is biologically plausible. Further studies assessing potential mediators and PM2.5-treatment interaction are needed.