Vicky Chi-Dan Chang

Purpose: Although mechanistic evidence suggests that some per- and polyfluoroalkyl substances (PFASs) may be carcinogenic, epidemiologic findings for PFAS and breast cancer remain inconsistent, partly because of differing study designs and limited evaluation by hormone receptor status. Methods: We conducted a case-cohort study within the Sister Study that included 1,609 breast cancer cases and 2,269 subcohort participants. We oversampled estrogen receptor (ER)–negative cases (324 ER-negative, 1,120 ER-positive) to improve statistical power. Seventeen PFASs were quantified in prediagnostic plasma using liquid chromatography-high-resolution mass spectrometry. Sampling-weighted Cox models were used to estimate associations with overall breast cancer incidence and by hormone receptor status, adjusting for confounders and coexposures. We conducted a meta-analysis of seven studies with prediagnostic blood PFAS measurements. Results: We observed associations between certain PFAS and ER-negative tumors. Positive associations were observed for perfluoro-n-octanoic acid (PFOA; hazard ratio [HR] _{Q4 v Q1} , 1.67 [95% CI, 1.10 to 2.53], P trend = .09), perfluorooctane-1-sulfonic acid (PFOS; HR _{Q4 v Q1} , 1.93 [95% CI, 1.15 to 3.22], P trend = .01), and potentially n-methylperfluoro-1-octanesulfonamidoacetic acid (N-MeFOSAA; HR _{Q4 v Q1} , 1.42 [95% CI, 0.99 to 2.04], P trend = .12). By contrast, perfluoro-n-tridecanoic acid was inversely associated with ER-negative tumors (PFTrDA; HR _{Q4 v Q1} , 0.63 [95% CI, 0.41 to 0.97], P trend = .05). Perfluorooctane sulfonamide (PFOSA), which was infrequently detected (6.9%), was positively associated with ER+ tumors (HR _{detectable v nondetectable} , 1.36 [95% CI, 1.03 to 1.81]). The meta-analysis further supported associations for both PFOA (relative risk [RR], 1.30 [95% CI, 1.05 to 1.62]) and PFOS with ER-negative breast cancer (RR, 1.20 [95% CI, 0.91 to 1.58]). Conclusion: Findings from this large prospective study, together with meta-analyses, suggest that PFOA and PFOS may increase ER-negative breast cancer risk. Given the widespread exposure to PFAS and limited established risk factors for ER-negative tumors, these results have important public health implications.

Objective Atrazine, an herbicide widely used in US agriculture, has been associated with kidney cancer and non-malignant kidney disease. However, the potential mechanisms underlying these associations remain unclear. We evaluated atrazine use and biomarkers of kidney function and nephrotoxicity among male farmers in the Biomarkers of Exposure and Effect in Agriculture Study. Methods Our investigation included three groups of farmers defined based on atrazine use: (1) both recent (last 3 months) and past use (continuous users; n=83); (2) high lifetime use (≥178.5 lifetime days) but no recent use (former high users; n=88) and (3) never/low lifetime use (Results Compared with farmers with never/low atrazine use, continuous users had statistically significantly lower eGFRcr-cys (−9.4%; 95% CI –16.1% to –2.2%) and higher serum creatinine (9.4%; 95% CI: 1.6% to 17.9%) and cystatin C (10.8%; 95% CI 2.0% to 20.4%); no associations with these markers were observed for the former high users. We also observed higher uric acid for both former high and continuous users and lower urea nitrogen for former high users, although these associations were not statistically significant. Urine albumin-creatinine ratio and KIM-1 levels did not differ across groups. Conclusions Our findings add to the evidence that continued atrazine use is associated with diminished kidney function.

Background: Occupational exposure to endotoxin has been associated with reduced lung cancer risk. The mechanisms underlying this association are unclear, though immunological alterations likely play a role. Farmers who perform certain tasks (eg, raising hogs) can be highly exposed to endotoxin. We, therefore, leveraged measurements of circulating immune markers from a prior investigation among male farmers in the Biomarkers of Exposure and Effect in Agriculture study to evaluate associations with newly developed estimates of endotoxin exposure. Methods: Our investigation included 122 non-smoking farmers from Iowa, oversampling those raising hogs. Serum levels of 60 markers were measured using multiplex bead-based assays and ELISA. Based on an algorithm linking measurement-based task intensity estimates with self-reported task frequency, we estimated cumulative endotoxin exposure in the 30 days up to sample collection. We used multivariable linear regression to estimate geometric mean ratios of immune markers across exposure quartiles. Results: Higher endotoxin exposure in the last 30 days was associated with increased levels of fibroblast growth factor-2, macrophage inflammatory protein-3 alpha/CCL20 and sIL-4R (P _trend ≤0.02) and decreased levels of macrophage-derived chemokine/CCL22 (P _trend =0.02). We also identified novel associations with several additional markers; those with the highest (vs lowest) exposure to endotoxin had decreased levels of TARC/CCL17, sCD27 and IL-1B. Conclusions: Several circulating immune markers were associated with endotoxin levels in an exposure–response manner. Our findings are consistent with prior work on hog farming and suggest possible biological mechanisms through which endotoxin may confer a reduced risk of lung cancer.