CA 15-3 elevation in U.S. women without breast cancer.

Abstract

e13609 Background: Cancer antigen 15-3 (CA 15-3) is an epitope of Mucin 1 transmembrane glycoprotein, used to monitor response to therapy in metastatic breast cancer. Elevated CA 15-3 levels have been noted in other malignant and non-malignant conditions. This is the first population-based study examining CA 15-3 elevation in US women without breast cancer. Methods: Women without breast cancer were identified from the National Health and Nutrition Examination Survey 2001-2002 with CA 15-3 measurements via Meso Scale Discovery immunoassay. CA 15-3 levels were dichotomized per previously reported cutoffs of 20, 25, 30, and 40 U/ml, and also logarithmically transformed. Logistic and linear regression models were used to examine associations between CA 15-3 and demographic characteristics, gynecologic/reproductive history, inflammatory markers, non-malignant chronic conditions and non-breast malignancies. Results: Of 2432 women ages 20-85 (mean 48) years old, eighty (3.3%) had CA 15-3 ≥30 U/ml. Based on cutoff ≥30 U/ml, univariate logistic regression showed ages 40-59, uterine fibroids history, and salpingectomy history were associated with normal CA 15-3; Mexican American ethnicity and increasing C-reactive protein were associated with elevated CA 15-3. Multivariate logistic regression confirmed that only ages 40-49 (OR = 0.162, p = 0.0013) and uterine fibroids history (OR = 0.216, p = 0.0373) remained significantly associated. Similar results were found using other cutoffs. Univariate linear regression showed ages 40-59, salpingectomy history, and longer oral contraceptive pill (OCP) use correlated with lower CA 15-3; Black ethnicity, increasing ferritin, endometriosis history, and lung cancer history correlated with higher CA 15-3. Multivariate linear regression revealed significant associations: ages 50-59 (17.1% lower CA 15-3 vs ages < 40, p = 0.028); race other than White, Black, Hispanic, Mexican (32.0% lower CA 15-3 vs White, p = 0.039); longer OCP use (14.0% lower CA 15-3 for 2-5 yrs’ OCP use, p = 0.023; 21.0% lower for > 10 yrs’ OCP use, p = 0.006; vs OCP use < 2 yrs); endometriosis history (22.9% higher CA 15-3 vs no history, p = 0.051). CA 15-3 levels were not significantly associated with chronic liver conditions, coronary artery disease, or metabolic conditions (elevated BMI, diabetes, hyperlipidemia). Conclusions: Perimenopausal women, uterine fibroids history, and longer OCP use were associated with lower CA 15-3 levels, while non-White ethnicity and history of endometriosis were associated with higher CA 15-3. These findings help to elucidate interpretation of CA 15-3 in clinical practice.