Aspirin Metabolites and Mammographic Breast Density in Premenopausal Women

Aspirin shows potential as a chemopreventive agent due to its anti-inflammatory effects, and aspirin use has been associated with a decreased risk of several cancers, including breast cancer (1). High mammographic breast density (MBD), an independent risk factor for breast cancer development, has been associated with an increased inflammatory response (2). To date, only a limited number of studies have investigated the associations of aspirin use with MBD, but findings across studies are inconsistent (3–6). However, no study has investigated the associations of aspirin metabolites with MBD. Therefore, we determined the relationship between aspirin metabolites [salicyluric glucuronide, 2-hydroxyhippurate (salicylurate), salicylate, and 2,6-dihydroxybenzoic acid] and MBD in premenopausal women.

Our study included 705 premenopausal women who had their annual screening mammograms at Washington University School of Medicine in St. Louis, MO. The analytic sample for multivariable analyses included 700 women with complete MBD data. All participants provided written informed consent, and the study was approved by the Washington University School of Medicine Institutional Review Board and conducted in accordance with the Declaration of Helsinki. Study methodology and exclusion criteria have been published previously (7). Untargeted metabolomic profiling was performed at Metabolon Inc., employing techniques similar to those of a previous study (8). Peak area data were normalized using ComBat to reduce batch effects (9). Missing values for any aspirin metabolite were imputed using the 10-nearest neighbor method. We used Volpara 1.5 (Volpara Health) to quantify volumetric percent density (%; VPD), dense volume (cm³; DV), and non-DV (cm³; NDV).

We evaluated the distribution of demographic and clinical characteristics by self-reported use of aspirin in the past 12 months. We performed multivariable linear regression models to assess the least square means of VPD, NDV, and DV by quartiles of aspirin metabolites, adjusting for age, age at menarche, body fat %, race, oral contraceptive use, family history of breast cancer, alcohol consumption, body shape at age 10 (based on the Stunkard pictogram), and parity and age at first birth. Covariates with missing observations were imputed using multivariate imputation. MBD measures were log₁₀ transformed to ensure normality and backtransformed for interpretation. We calculated the P value for trend across the quartiles of each aspirin metabolite by operationalizing the metabolite as an ordinal variable and setting all values within each quartile to the median of that quartile. All statistical analyses were conducted in RStudio (RRID:SCR_000432)/R (RRID:SCR_001905), and associations were considered significant if the P value was <0.05.

Deidentified data are available from the corresponding author upon reasonable request.

Study participants were, on average, 46 years old; 71.8% were non-Hispanic White, and 23.1% were non-Hispanic Black. The average MBD was 10.5% (SD = 7.6), 82.0 cm³ (SD = 45.5), and 1,057.1 cm³ (SD = 717.0) for VPD, DV, and NDV, respectively. Approximately 13% of participants reported taking aspirin in the past 12 months. Aspirin users had higher mean peak area of 2-hydroxyhippurate (salicylurate), salicylate, and salicyluric glucuronide levels than nonusers, but only salicyluric glucuronide levels were increased monotonically by both aspirin dose (1–2 tablets per day: N = 83, Mean = 1,140,663.7, SD = 2,736,137.2, and ≥3 tablets per day: N = 4, Mean = 1,380,476.0, SD = 1,851,058.8) and frequency (1 day/week: N = 47, Mean = 888,129.3, SD = 2,701,288.4; 2–3 days/week: N = 17, Mean = 1,199,897.9, SD = 3,027,482.1; and ≥4 days/week: N = 23, Mean = 1,654,637.0, SD = 2,458,330.8; Table 1). There were no significant monotonic associations between salicyluric glucuronide, 2-hydroxyhippurate (salicylurate), salicylate, and 2,6-dihydroxybenzoic acid and any measure of MBD in the overall analysis (Table 2).

Salicyluric glucuronide, 2-hydroxyhippurate (salicylurate), and salicylate levels were positively associated with having used aspirin in the past 12 months, but only salicyluric glucuronide was monotonically associated with both frequency of use and dose. This demonstrates that salicyluric glucuronide is a highly reproducible measure of aspirin use. However, we observed no significant monotonic associations across the quartiles of each metabolite with MBD in premenopausal women.

Our study is an important new addition to the field. To the best of our knowledge, this is the first study to investigate the associations between aspirin metabolites and MBD. We and others have investigated the associations of self-reported aspirin use and MBD and found no association regardless of duration of use (3, 4, 6). This body of work suggests that aspirin use may not be associated with MBD, although one study found an inverse association between aspirin use (collected from the patients’ medical records) and MBD evaluated as Breast Imaging Reporting & Data System (BI-RADS) categories, with the strongest inverse associations among younger women (5). Our study has several strengths. We evaluated aspirin metabolites, which offer a more objective measure of aspirin use than self-report and minimize some of the biases in previous studies. We adjusted for many potential confounders. Limitations include measuring metabolites at a single time point, which may not capture long-term aspirin use. A small number of participants (∼13%) reported taking aspirin; hence, our results may not be generalizable to populations with more frequent use and our study may not be powered to detect the association between aspirin use and MBD. However, the focus of our study is on aspirin metabolites and MBD, and the aspirin metabolites were measurable in our study population with minimal [approximately 20% missing for salicyluric glucuronide and 2-hydroxyhippurate (salicylurate)] or no missing values (salicylate and 2,6-dihydroxybenzoic acid), suggesting that self-reported aspirin use may not accurately capture the full exposure to salicylates. This may be due to the effect of additional exogenous exposures such as nonaspirin medications, supplements, and diet on these metabolites (10). Hence, the null associations observed do not completely rule out an association between aspirin use and MBD. Younger women are more likely to use other NSAIDs, but our study did not report on the associations of NSAID metabolites with MBD because most of the metabolites were below the level of detection in participants (>50%). Hence, analyses utilizing these metabolites are not likely to be robust. Last, because our study population is limited to premenopausal women, future research exploring this association in postmenopausal women is necessary.

In conclusion, aspirin metabolites [salicyluric glucuronide, 2-hydroxyhippurate (salicylurate), salicylate, and 2,6-dihydroxybenzoic acid] were not monotonically associated with MBD in premenopausal women. Additional studies investigating the associations of additional aspirin and other NSAID metabolites in plasma and breast tissues with MBD are necessary.

No disclosures were reported.

R.K. Singh: Formal analysis, writing–original draft. K.R. Getz: Formal analysis, writing–original draft, writing–review and editing. J.K. Kyeyune: Writing–original draft, writing–review and editing. M.S. Jeon: Formal analysis. C. Luo: Writing–review and editing. J. Luo: Formal analysis, writing–review and editing. A.T. Toriola: Conceptualization, resources, supervision, funding acquisition, writing–original draft, writing–review and editing.

We would like to thank all the women who participated in this study. This work was supported by funding from the NIH/NCI (R01CA246592 to A.T. Toriola), and K.R. Getz was supported by T32CA190194. The content is solely the responsibility of the authors and does not represent the official view of the NIH.