Clinically sufficient vitamin D levels at breast cancer diagnosis and survival outcomes in a prospective cohort of 3,995 patients after a median follow-up of 10 years.

Abstract

10510 Background: There have been suggestive findings for better cancer survival with vitamin D supplementation in the recent VITAL trial. The findings are consistent with meta-analyses based on earlier randomized trials testing daily supplement vitamin D intake. As there is no ongoing or planned randomized trial of vitamin D supplementation in sight for women after breast cancer diagnosis, we evaluated relationships between serum levels of vitamin D and breast cancer outcomes in a large prospective cohort of breast cancer survivors. Methods: We measured 25-hydroxyvitamin D (25OHD) levels in serum samples collected at the time of diagnosis from 3,995 women with incident breast cancer enrolled in the Pathways Study, a large prospective cohort established in 2006 at Kaiser Permanente Northern California with active follow-up (FU). Potential determinants of 25OHD levels, including a polygenic score, were examined. Vitamin D levels were categorized based on clinical cutoffs as deficient ( < 20 ng/ml), insufficient (20 to < 30 ng/ml), or sufficient (≥30 ng/ml). These levels were then evaluated in relation to overall survival (OS), breast cancer-specific survival (BCSS), recurrence-free survival (RFS), and invasive disease-free survival (IDFS). Cox proportional hazards models were adjusted for non-clinical, clinical, and treatment factors and were further stratified by stage, estrogen receptor (ER) status, and body mass index (BMI). Results: Vitamin D supplement use, lower BMI, and self-reported white race were the strongest determinants of higher 25OHD levels. The polygenic score was significantly associated with 25OHD levels but explained only 0.3% of the variance. The median FU was 9.6 years (range: 0.3-13). Compared to those with deficient vitamin D levels, patients with sufficient levels had significantly better survival outcomes, which remained after controlling for various covariates (OS: HR [95% CI] = 0.73 [0.58-0.91]; BCSS: HR = 0.78 [0.56-1.09]; RFS: HR = 0.79 [0.65-0.97]; IDFS: HR = 0.82 [0.68-0.99]). Associations were similar by ER status, but stronger among patients with more advanced stage disease and those with under-weight or normal BMI. Black patients had the lowest 25OHD levels, which contributed to their poorer survival compared to white patients. Adding 25OHD levels to the Cox model of OS lowered the HR associated with Black vs. white race from 2.03 (1.57-2.62) to 1.79 (1.37-2.32). Conclusions: Sufficient vitamin D levels at the time of diagnosis were associated with improved breast cancer prognosis. Consistent with results from randomized trials, our findings from a large observational cohort of breast cancer survivors with long FU provide the strongest evidence to date for maintaining sufficient vitamin D levels in breast cancer patients, including among Black women and those with more advanced stage disease.