Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jan 1.
Published in final edited form as: Cancer Prev Res (Phila). 2021 Apr 13;14(7):753–762. doi: 10.1158/1940-6207.CAPR-20-0581

The influence of vitamin D on mammographic density: results from CALGB 70806 (Alliance) a randomized clinical trial

Marie E Wood 1, Heshan Liu 2, Elizabeth Storrick 2, David Zahrieh 2, H Carisa Le-Petross 3, Sin-Ho Jung 4, Patricia Zekan 5, M Margaret Kemeny 6, Jayne R Charlamb 7, Lili X Wang 8, Gary W Unzeitig 9, Candace S Johnson 11, Judy E Garber 10, James R Marshall 11, Isabelle Bedrosian 3
PMCID: PMC8449513  NIHMSID: NIHMS1695621  PMID: 33849913

Abstract

Current therapies for breast cancer prevention only prevent estrogen receptor positive (Er+) disease and toxicity limits use of these agents. Vitamin D is a potential prevention therapy for both Er+ and Er− disease and is safe with few side effects. This study evaluates the effect of one-year of vitamin D supplementation on mammographic density (MD), a biomarker of breast cancer risk in a multicenter randomized controlled trial. Premenopausal women with ≥ 25% MD and no history of cancer, were randomly assigned to 2000IU of vitamin D or placebo orally daily for 1-year. Change in percent MD was evaluated using Cumulus software after all participants completed treatment. Three hundred women enrolled between 1/2011 and 12/2013 with a mean age of 43 and diverse ethnicity (14% Hispanic, 12% African American [AA]). Supplementation significantly increased vitamin D levels compared to placebo (14.5 ng/mL vs −1.6 ng/mL; p<0.0001) with all participants on the Vitamin D arm achieving vitamin D sufficiency at 12 months. Vitamin D was safe and well tolerated. After adjustment for baseline MD, the mean between-arm difference (vitamin D vs placebo) at 1 year was −0.75 [−.26, 1.76 p=0.56]. A greater effect was seen for women with ≥50% MD and AA women, although neither reached significance. This randomized controlled trial demonstrated significant improvement in vitamin D levels with 2000 IU for one year, with 100% of supplemented women achieving sufficiency. However, a null effect was seen regarding change in MD for premenopausal women (the primary outcome of the study).

Keywords: Biomarkers, Breast Cancer Prevention, Chemoprevention, Mammographic Density, Vitamin D

INTRODUCTION

Breast cancer is the second leading cause of cancer related deaths in women [1] and prevention is key to impacting this mortality. Several agents (aromatase inhibitors and selective estrogen receptor modulators) have been shown to prevent breast cancer and are FDA approved for use in prevention. [2–4] However, these agents only reduce rates of estrogen receptor positive (ER+) breast cancer [2] and only tamoxifen is available for premenopausal women. Additionally, these agents have side effects and toxicities that limit uptake [5–8]. With limited choices for premenopausal women, side effects that limit uptake, and the absence of options for estrogen receptor-negative (ER−) cancers, investigation of additional options for chemoprevention are warranted.

As a potential preventive therapy, Vitamin D is safe, has few side effects, and is supported by both biologic and epidemiologic evidence. [9–11] [12–14] Vitamin D has anticarcinogenic properties, demonstrated in both in vitro and in vivo models. [15–20] Epidemiological data reveals a reduced risk of breast cancer for women with higher vitamin D levels. [21–28] In a pooled analysis study, a 58% lower risk for women with normal levels (>38 ng/ml) was seen, compared to women with deficient levels. [23] Similarly, Abbas et al. demonstrate a significantly lower risk in women with vitamin D levels greater than 60 nmol/L (OR 0.45 CI 0.29–0.70). [29] The evidence supporting the inverse relationship between vitamin D and breast cancer risk is stronger among premenopausal women. [26, 30–32] Despite these data, there has not been confirmation of a prevention effect in prospective trials. A trend toward fewer breast cancers was seen when postmenopausal women were given 1100 international units (IU) for four years [33], while no effect was seen for 400 IU daily [34]. In a more recent study of postmenopausal women supplemented with 2000 IU/day for four years, fewer breast cancers were seen in the treatment group. [35] Possible reasons for these negative findings include dosing (400IU) [33, 34] and population tested (postmenopausal women) [35]. Given the strong biologic and epidemiologic data, further study is warranted with higher doses and more targeted populations.

The use of cancer risk biomarkers can circumvent the large sample size and long follow-up needed to demonstrate a preventive effect. Biomarkers that are measurable, modifiable with treatment, with modification is associated with change in cancer risk are the most useful intermediate endpoints. [36] MD has been confirmed as an independent risk factor for development of breast cancer [37–42] and is the most well accepted intermediate biomarker. Women with MD of at least 75% are four to six times more likely than women with MD of ≤10% to develop breast cancer [43]. Observational data suggests an inverse association between MD and vitamin D. [44–47] Additionally, MD appears to be a risk factor for both estrogen receptor positive and negative breast cancer. [48, 49]

Given the strong biological data supporting a possible preventive effect of vitamin D, CALGB 70806 was undertaken to determine the effect of vitamin D supplementation on breast cancer biomarkers in premenopausal women. Here we report on the primary objective of the study: to evaluate the effect of vitamin D supplementation on mammographic density.

MATERIALS and METHODS

Trial Design

This double-blinded, placebo-controlled study was conducted through the Alliance for Clinical Trials in Oncology (formerly the Cancer and Leukemia Group B [CALGB]), a National Cancer Institute (NCI)-sponsored cooperative group and registered with clinicaltrials.gov (ClinicalTrials.gov Identifier: NCT01224678) From January 2011 to December 2013, women were recruited from participating institutions. Institutional review board (IRB) approval was obtained at each institution. This study was conducted in accordance with International Ethical Guidelines for Biomedical Research Involving Human Subjects (CIOMS. After providing written informed consent, and local determination of vitamin D levels for stratification, eligible women were randomized (1:1) to receive either 2000 IU of vitamin D or placebo orally every day for 1 year. Study medication was started within 14 days of randomization. Women underwent digital mammography within 6 weeks of randomization (baseline) and ± 2 weeks of the cessation of study medication (1 year). Serum was also drawn at baseline and 1-year post randomization and sent for centralized assay at Roswell Park Cancer Institute for determination of change in Vitamin D levels (ng/mL) over time. Women were seen and examined by a provider at baseline and at 1 year to provide a symptom review, physical examination, and laboratory evaluation. Additionally, women were evaluated every 3 months to provide a symptom review.

Participants

Women were eligible if they were premenopausal (defined as having regular menstrual cycles and ≤55 years old), and had a breast density of at least 25% (scattered fibroglandular density or greater). Women were excluded if they had a history of breast cancer (including DCIS), breast implants or breast reduction surgery, serum calcium level > 10.5 mg/dL, history of hyperparathyroid disease requiring intervention in the past 5 years, history of kidney stones, two or more bone fractures in the past 5 years, diagnosis of osteoporosis requiring treatment, use of vitamin D supplements > 400 IU/ day, hormone replacement therapy, tamoxifen or raloxifene, or concurrent participation in a breast cancer chemoprevention trial. Women using topical estrogens or hormonal contraceptives were eligible if they had been using these agents more than 4 months prior to enrollment. Those on vitamin D therapy at a dose of 400 IU/day or less were also eligible if they had been on that dose for ≥6 months.

Outcome evaluation

Digital mammograms were obtained at baseline and one-year post-randomization from each participant. Imaging on the same machine for each participant was not required as it was felt to be to difficult to require and track this given the multi-institutional nature of the study. Processed digital images were sent and stored at the Imaging and Radiation Oncology Core (IROC) for assessment of MD at the end of the study. Mammograms were processed for measurement with the user-assisted threshold method, Cumulus by a single radiologist (HL), using standard techniques. [50] The reader set the appropriate gray-scale thresholds to define the breast and dense area (cm2), after which the percent dense area was calculated. MD for each woman represented the average percent MD obtained from both breasts and from two different views, cranial caudal (CC) and mediolateral oblique (MLO). The radiologist was blinded to treatment arm and timepoint. MD was determined at the end of the study, after all women completed treatment. A formal reliability assessment was not performed, however any density readings that appeared incongruous were re-interpreted.

Randomization and sample size,

Dynamic allocation based on the methods of Pocock and Simon was used for stratified randomization. [51] Randomization was stratified according to baseline vitamin D levels (insufficient (< 30 ng/mL) vs. sufficient (≥ 30 ng/mL)) and performance of random periareolar fine needle (RPFNA) (institutions that performed the procedure vs. those that did not).

The primary outcome variable was MD, measured by the semi-automated Cumulus method, at both baseline and 1-year post-baseline. [52] We hypothesized that vitamin D supplementation would reduce MD at 1 year in premenopausal women compared with placebo. With 285 eligible women randomized with MD measured at both baseline and one-year post-baseline, a two-sample t-test with a two-sided 0.05 significance level had 92% power to detect a standardized mean difference (vitamin D vs placebo) of −0.4 between the two arms. Chow CK et al. [53] reported a standard deviation of 6.6% in MD change between baseline and one year. Hence, the target effect size was 2.64%. Assuming a 5% dropout rate, 300 women were planned to be randomized.

Statistical Methods

The study was monitored twice annually by the Alliance Data and Safety Monitoring Board. Data quality was ensured by review of data by the Alliance Statistics and Data Center (SDC) and by the study chairperson following Alliance policies. Statistical analyses were conducted using SAS® version 9.4 by the Alliance SDC.

The primary analysis was based on the complete case analysis sample (N=202), defined as the set of eligible women who had an MD measurement at both baseline and 1-year; women who withdrew from the study prior to beginning the study intervention or who were deemed non-compliant were excluded (see Figure 1 for consort diagram). Women were analyzed in the arm to which they were randomized. For the primary analysis, analysis of covariance (ANCOVA) adjusting for baseline MD was used to compare MD at 1-year between the arms. We would conclude that the 1-year MD was statistically different between the arms if the corresponding two-sided P-value was <0.05. Additionally, we report the estimated mean between-arm difference at one year and the corresponding 95% confidence interval (CI) as well as the standardized effect size defined as the estimated mean difference divided by the root mean square error from the ANCOVA model. To assess the robustness of the intervention effect, the primary analysis was repeated adjusting for the following baseline covariates: body mass index, the stratification factor vitamin D sufficiency (sufficient; insufficient), and age.

Figure 1.

Figure 1.

Open in a new tab

Randomization of women enrolled in the study (CALGB 70806 [Alliance]): CONSORT Diagram

* Cancel refers to those patients who withdrew from the study prior to beginning study treatment.

** To be evaluable for the primary endpoint, women had to have mammographic density measured at both baseline and 1-year post baseline.

In a hypothesis-generating exploration, the primary analysis was repeated within the following subgroups: baseline MD (</≥ 50%) and race (AA;Caucasian). Because these analyses were performed on a subset of women, randomization may have been compromised; therefore, the analyses adjusted for the aforementioned baseline covariates. However, due to the small sample size within the AA subset (n = 19), the covariate adjustment was limited to BMI.

RESULTS

Study sample

Three hundred premenopausal women were randomized, between January 2011 and December 2013, from institutions across the United States: 150 women to vitamin D and 150 women to placebo. Two women assigned to vitamin D were deemed ineligible, 8 women withdrew prior to beginning therapy, and 5 women were deemed non-compliant. Of the remaining 285 women, 140 were assigned to vitamin D and 145 to placebo. Sixty-seven percent (202/300) of randomized women had MD data available at both baseline and 1 year (94 vitamin D; 108 placebo). See Consort diagram (Figure 1) for details.

Baseline characteristics

For the 300 women randomized, baseline characteristics were balanced across the two arms (Table 1). The median [range] age was 43.0 years [22.0, 59.0], with diverse ethnicity (14% Hispanic, 12% AA). Sixty-four percent of the women had insufficient (<30 ng/mL) vitamin D levels at baseline. For the women in the complete case analysis sample (94 vitamin D; 108 placebo), baseline characteristics were also balanced across the arms with characteristics similar to all randomized women (Table 1).

Table 1.

Baseline characteristics of women enrolled in the study (CALGB 70806 [Alliance]).

All Women Randomized Complete Case†
Placebo
(N=150)		 Vitamin D
(N=150)		 Total
(N=300)		 Placebo
(N=108)		 Vitamin D
(N=94)		 Total
(N=202)
Age
 N 150 150 300 108 94 202
 Mean (SD) 43.0 (6.0) 42.2 (6.5) 42.6 (6.3) 43.1 (6.0) 42.0 (6.0) 42.6 (6.0)
 Median 44.0 43.0 43.0 44.0 43.0 44.0
 Q1, Q3 40.0, 47.0 40.0, 46.0 40.0, 47.0 39.5, 47.0 40.0, 46.0 40.0, 47.0
 Range (24.0–55.0) (22.0–59.0) (22.0–59.0) (25.0–55.0) (24.0–52.0) (24.0–55.0)
Race
 White 124 (82.7%) 114 (76.0%) 238 (79.3%) 89 (82.4%) 76 (80.9%) 165 (81.7%)
 Black or African American 14 (9.3%) 21 (14.0%) 35 (11.7%) 10 (9.3%) 9 (9.6%) 19 (9.4%)
 Asian 9 (6.0%) 5 (3.3%) 14 (4.7%) 6 (5.6%) 4 (4.3%) 10 (5.0%)
 Native Hawaiian or Pacific Islander 0 (0.0%) 2 (1.3%) 2 (0.7%) 0 (0.0%) 1 (1.1%) 1 (0.5%)
 American Indian or Alaska Native 0 (0.0%) 1 (0.7%) 1 (0.3%) 0 (0.0%) 1 (1.1%) 1 (0.5%)
 Not Reported 2 (1.3%) 3 (2.0%) 5 (1.7%) 2 (1.9%) 1 (1.1%) 3 (1.5%)
 More than one race 1 (0.7%) 4 (2.7%) 5 (1.7%) 1 (0.9%) 2 (2.1%) 3 (1.5%)
Ethnicity
 Hispanic 19 (12.7%) 23 (15.3%) 42 (14.0%) 12 (11.1%) 13 (13.8%) 25 (12.4%)
 Non-Hispanic 129 (86.0%) 124 (82.7%) 253 (84.3%) 95 (88.0%) 79 (84.0%) 174 (86.1%)
 Not reported 1 (0.7%) 2 (1.3%) 3 (1.0%) 1 (0.9%) 2 (2.1%) 3 (1.5%)
 Unknown 1 (0.7%) 1 (0.7%) 2 (0.7%) 0 (0.0%) 0 (0.0%) 0 (0.0%)
Region
 Midwest 19 (12.7%) 23 (15.3%) 42 (14.0%) 17 (15.7%) 17 (18.1%) 34 (16.8%)
 Northeast 67 (44.7%) 52 (34.7%) 119 (39.7%) 49 (45.4%) 41 (43.6%) 90 (44.6%)
 South 52 (34.7%) 62 (41.3%) 114 (38.0%) 36 (33.3%) 29 (30.9%) 65 (32.2%)
 West 12 (8.0%) 13 (8.7%) 25 (8.3%) 6 (5.6%) 7 (7.4%) 13 (6.4%)
BMI
 N 150 150 300 108 94 202
 Mean (SD) 26.7 (5.3) 27.4 (6.1) 27.0 (5.7) 26.8 (5.4) 26.5 (5.8) 26.7 (5.6)
 Median 25.7 26.0 25.8 25.7 25.4 25.6
 Q1, Q3 22.8, 29.6 22.9, 30.1 22.8, 29.8 22.8, 29.7 22.5, 29.0 22.8, 29.4
 Range (19.1–53.6) (15.1–46.3) (15.1–53.6) (19.3–53.6) (16.0–46.2) (16.0–53.6)
BMI Categorization ***
 1: Underweight 0 (0.0%) 3 (2.0%) 3 (1.0%) 0 (0.0%) 2 (2.1%) 2 (1.0%)
 2: Normal 69 (46.0%) 59 (39.3%) 128 (42.7%) 49 (45.4%) 43 (45.7%) 92 (45.5%)
 3: Overweight 46 (30.7%) 50 (33.3%) 96 (32.0%) 34 (31.5%) 28 (29.8%) 62 (30.7%)
 4: Obese Class I 25 (16.7%) 20 (13.3%) 45 (15.0%) 17 (15.7%) 12 (12.8%) 29 (14.4%)
 5: Obese Class II 9 (6.0%) 9 (6.0%) 18 (6.0%) 7 (6.5%) 5 (5.3%) 12 (5.9%)
 6: Obese Class III 1 (0.7%) 9 (6.0%) 10 (3.3%) 1 (0.9%) 4 (4.3%) 5 (2.5%)
Vitamin D sufficiency **
 Insufficient (<30ng/mL or <75nmol/L) 95 (63.3%) 97 (64.7%) 192 (64.0%) 67 (62.0%) 59 (62.8%) 126 (62.4%)
 Sufficient (>=30ng/mL or >=75nmol/L) 55 (36.7%) 53 (35.3%) 108 (36.0%) 41 (38.0%) 35 (37.2%) 76 (37.6%)
Vitamin D ng/mL **
 N 150 150 300 108 94 202
 Mean (SD) 27.4 (11.9) 25.8 (11.5) 26.6 (11.7) 28.1 (12.2) 26.4 (9.7) 27.3 (11.1)
 Median 27.0 24.5 26.0 27.5 26.0 27.0
 Q1, Q3 18.1, 32.1 17.0, 32.0 18.0, 32.0 19.3, 32.7 19.8, 31.7 19.5, 32.1
 Range (10.0–72.0) (4.0–65.0) (4.0–72.0) (10.0–72.0) (11.3–60.0) (10.0–72.0)
BI-RADS Density *
 Missing 38 (25.3%) 54 (36.0%) 92 (30.6%) 0 (0.0%) 0 (0.0%) 0 (0.0%)
 2 19 (12.7%) 10 (6.7%) 29 (9.7%) 19 (17.6%) 10 (10.6%) 29 (14.4%)
 3 80 (53.3%) 75 (50.0%) 155 (51.7%) 76 (70.4%) 73 (77.7%) 149 (73.8%)
 4 13 (8.7%) 11 (7.3%) 24 (8.0%) 13 (20.0%) 11 (11.7%) 24 (11.9%)
Open in a new tab
†

Complete case is defined as all randomized women who had MD measured at both baseline and 1-year post baseline by the semi-automated Cumulus method.

*

Based on central interpretation.

**

Obtained from the local sites.

***

The BMI classifications were: 1: Underweight= BMI< 18.5; 2: Normal: 18.5 <= BMI < 25; 3: Overweight: 25 < BMI <= 30; 4: Obese Class 1: 30 < BMI <= 35; 5: Obese Class 2: 35 < BMI <=40; 6: Obese Class 3: BMI >40.

Change in Vitamin D Levels

Forty-five percent (136) women had baseline and 12-month vitamin D levels measured centrally at the end of the study. Women receiving vitamin D supplementation had an average increase of 14.5 ng/mL (13.1) by 12 months compared with −1.6 ng/mL (12.4) in women receiving placebo (P < 0.0001).

The proportion of women in the vitamin D arm with sufficient baseline vitamin D levels was 69.8% (44/63) rising to 100% at month 12, an increase of 30.2% [95% CI: 18.9%, 41.5%]. The proportion of women in the placebo arm with sufficient vitamin D levels were 54 / 73 = 74.0% at baseline and 53 / 73 = 72.6% at month 12; a decrease of −1.4% [95% CI: −12.5%, 9.7%].

In these 136 women, there was a clinically significant difference in baseline vitamin D levels determined locally for purposes of stratification as opposed to centrally. The mean baseline vitamin D levels determined locally and centrally were 27.7 ng/mL (11.6) and 37.4 ng/mL (12.8), respectively. While only 36% of women had sufficient vitamin D levels determined locally, 70% of women had sufficient levels when vitamin D was determined centrally. We did examine correlation between baseline vitamin D levels obtained via both sources and they were moderately and positively correlated (Pearson correlation r = 0.68; n = 136) such that there was a general tendency for baseline vitamin D levels measured locally and centrally to increase together.

Safety Analyses

Adverse events (AEs) were submitted at baseline, and at 3, 6, 9, and 12 months as well as 30 days post treatment. In this study, 274 women (134 Vitamin D; 140 placebo) received at least one dose of study drug and had an AE assessment. On the vitamin D arm, 32 women (23.9%) reported at least one AE, while 39 (27.9%) women on the placebo arm did. Nine (6.7%) women on the vitamin D arm reported one or more grade 3+ AEs, while eleven (7.9%) women on the placebo arm reported at least one grade 3+ AE; one of these was deemed to be related to study drug (one woman on the vitamin D arm reported grade 3 constipation, which was deemed to be possibly related to study drug). There was one death on the study, in the placebo group, and determined not to be related to study participation.

Change in Mammographic Breast Density

Based on the 202 women in the complete case analysis sample, the mean (SD) baseline MD based on the semi-automated Cumulus method was 42.4 (18.1) and 38.9 (19.4) in the vitamin D and placebo arms, respectively. The mean (SD) MD at 1-year was 40.5 (19.0) and 37.9 (20.6) in the vitamin D and placebo arms, respectively. Figure 2 shows the baseline MD and 1-year MD values in each arm with the regression lines from the analysis of covariance; the vertical distance between the two regression lines represents the estimated between-arm difference in MD at 1 year. After adjustment for the baseline MD value, the mean between-arm difference (vitamin D vs placebo) [95% CI] at 1-year was −0.75 [−3.26, 1.76]. BMI was significantly and inversely associated with percent density as expected; 42% of women with ≥50% density had normal or low BMI, while only 18% of women who were overweight or obese had ≥50% density. BMI did not significantly change over time nor was there a difference in change by arm.

Figure 2.

Figure 2.

Open in a new tab

Change in Mammographic breast density after one year

Baseline and one-year percent mammographic density in each arm showing the regression lines from the analysis of covariance. Squares show mean values for the two arms. The estimated difference between the two arms from analysis of covariance is the vertical distance between the two regression lines. The estimated between-arm difference (vitamin D vs placebo) [95% CI] at one-year was −0.75 [−3.26, 1.76] (P = 0.56).

Although the estimated MD reduction was 0.75 more on average in the vitamin D arm than in the control arm, the result did not achieve statistical significance (P = 0.56). The corresponding standardized effect size was −0.08. After adjusting for the baseline MD value as well as the other baseline covariates, the mean between-arm difference at 1-year was −0.94 [−3.40, 1.53] (P = 0.46).

Within the subset of women with a baseline MD ≥ 50% the vitamin D effect [95% CI] was −2.82 [−7.48, 1.83] (P = 0.23), while within the subset of women with a baseline MD < 50%, the vitamin D effect was −0.04 [−2.91, 2.83] (P = 0.98). Among the Caucasian women, the vitamin D effect [95% CI] was −1.02 [−3.88, 1.84] (P = 0.49), while within the AA women the vitamin D effect was −2.60 [−7.53, 2.34] (P = 0.30).

Discussion

This randomized, multi-institutional clinical trial, carried out in an ethnically and geographically diverse sample of premenopausal women, is one of the first to evaluate the effect of vitamin D in women at increased risk due to breast density. Vitamin D therapy was tolerable, resulting in few side effects. Further, supplementation resulted in a greater shift towards vitamin D sufficiency at 12 months with 100% of women assayed at 1 year reaching sufficient vitamin D levels.

Despite the significant increase of vitamin D with supplementation, the study did not achieve its primary objective: a significant reduction in MD for women taking vitamin D. This is now the third randomized clinical trial to examine the effect of vitamin D on MD in premenopausal women, all with null results. [54, 55] These null data are supported by several observational studies. [56–59]

However, unlike the other prospective trials we did observe a larger reduction in MD among women with ≥50% density (−2.82 [−7.48, 1.83]) and AA women (−2.60 [−7.53, 2.34]). While not statistically significant, these exploratory subgroup analyses are important. Women with MD between 50–75% have a relative risk of breast cancer 2.92 (2.49–3.42) times that of women with MD less than 50%, and women with ≥75% density have a relative risk of 4.64 (3.64–5.91). [41] Furthermore, high MD maybe a risk factor for triple negative breast cancer (TNBC) [49, 60] among younger women. [61]

Finding a larger reduction in MD among African American women who took vitamin D may be important. African Americans have been shown to have lower vitamin D levels. [62] In fact, 74% of AA women in this study had insufficient vitamin D levels. Moreover, AA women are known to have greater risk for triple negative breast cancer. [63–65] Several studies (including a systematic review) suggest that low vitamin D levels may be a risk factor for TNBC. [66–69]

Currently, there are no known agents for prevention of TNBC. However, two recent laboratory-based studies suggest that vitamin D may have preventive effects specifically for TNBC. [70, 71] Our subgroup analysis in women with ≥50% MD and in AA women (two populations at greater risk for TNBC) may lend support for vitamin D as an agent for prevention of TNBC.

While this was a randomized placebo-controlled multi-site clinical trial, there are limitations to our study. This trial was conducted at sites across the United States, and logistically we could not require that sites have mammogram images taken from the same machine. Given our null findings this should be considered a limitation of the study. Although supplementation resulted in 100% of women reaching sufficient vitamin D levels, one third of the 94 women on the vitamin D arm who were evaluable for the primary endpoint were not assayed for vitamin D levels. Consequently, the lack of an observed meaningful difference in MD between the arms may have been due to inadequate vitamin D supplementation in a third of the women. Additionally, there was a considerable amount of missing MD data, but based on characteristics of these data we do not feel that the trial results were at risk of significant bias due to incomplete outcome data. Nevertheless, the missing data did lead to a reduction in the usable sample size. Based on a complete case analysis population, the standardized effect size of −0.08 was considerably smaller than hypothesized (−0.40). This result may have been also been due to the premenopausal status of the study population – a group known for lower change in MD over time.[72, 73]

Conclusion:

Although supplementation with vitamin D led to a substantial increase in the vitamin D levels of women, it did not significantly decrease the MD of study participants. Exploratory subset analyses are consistent with the hypothesis that women with elevated MD and/or AA women are more likely to achieve a benefit with vitamin D supplementation. Given that African American women are clearly at elevated risk for TNBC and that increased MD may be a risk factor for TNBC, this is an interesting finding. Findings from this study highlight the need for identifying the right agent for the right population (precision medicine), a concept equally relevant for prevention as it is for treatment.

Acknowledgments:

We would like to thank Drew Seisler for his contributions to data management and statistical programming. The authors are grateful to the patients who consented to participate in this clinical trial and the families who supported them. We wish to acknowledge the accrual of patients to this study by ECOG-ACRIN (supported by CA180820 and CA21115) and SWOG (supported by 1U10CA180888).

Support:

Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under Award Number UG1CA189823 (Alliance for Clinical Trials in Oncology NCORP Grant), U10CA21060, U10CA180867, U10CA180866, UG1CA189817, and UG1CA189858. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Also supported in part by funds from 5R21CA137650–2 (PI M. Wood). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

DATA SHARING STATEMENT: The Alliance for Clinical Trial for Oncology has a data sharing responsibility with NCTN to report the clinical data used in the primary publication within approximately six months of the publication date. At that time, the data will be deposited in a central location, with a data dictionary for researchers to retrieve.

Conflict of interest disclosure statement: The Authors declare that there is no conflict of interest.

Trial Registration: Vitamin D and Breast Cancer Biomarkers NCT01224678 clinicaltrials.gov

Prevention Relevance Statement: Current therapies for breast cancer prevention only prevent estrogen receptor positive (Er+) disease and are underutilized due to toxicity and side effects. Vitamin D is a potential prevention therapy for both Er+ and Er− disease and is safe with few side effects.

The following institutional networks participated in this study:

Bay Area Tumor Institute NCORP, Oakland, CA, Lisa Bailey, UG1CA189817; Dana-Farber / Partners CancerCare LAPS, Boston, MA, Harold Burstein, U10CA180867; Dartmouth College - Norris Cotton Cancer Center LAPS, Lebanon, NH, Konstantin Dragnev, U10CA180854; Doctor’s Hospital of Laredo, Laredo, TX, Gary Unzeitig; Duke University - Duke Cancer Institute LAPS, Durham, NC, Jeffrey Crawford, U10CA180857; Florida Hospital Orlando, Orlando, FL, Carlos Alemany; Heartland Cancer Research NCORP, Decatur, IL, Bryan Faller, UG1CA189830; Hematology Oncology Associates of Central New York-East Syracuse, East Syracuse, NY, Jeffrey Kirshner; Mayo Clinic LAPS, Rochester, MN, Steven Alberts, U10CA180790; MedStar Georgetown University Hospital, Washington, DC, Filipa Lynce; Mount Sinai Hospital, New York, NY, Lewis Silverman; Mount Sinai Medical Center, Miami Beach, FL, Michael Schwartz; New Hampshire Oncology Hematology PA-Hooksett, Hooksett, NH, Douglas Weckstein; Northern Indiana Cancer Research Consortium, South Bend, IN; The Ohio State University Comprehensive Cancer Center LAPS, Columbus, OH, Claire Verschraegen, U10CA180850; Roswell Park Comprehensive Cancer Center LAPS, Buffalo, NY, Ellis Levine, U10CA180866; Southeast Clinical Oncology Research (SCOR) Consortium NCORP, Winston-Salem, NC, James Atkins, UG1CA189858; State University of New York Upstate Medical University, Syracuse, NY, Stephen Graziano; UC San Diego Moores Cancer Center, La Jolla, CA, Lyudmila Bazhenova; UNC Lineberger Comprehensive Cancer Center LAPS, Chapel Hill, NC, Thomas Shea, U10CA180838; University of Chicago Comprehensive Cancer Center LAPS, Chicago, IL, Hedy Kindler, U10CA180836; University of Oklahoma Health Sciences Center LAPS, Oklahoma City, OK, Adam Asch, U10CA180798; University of Texas MD Anderson Cancer Center LAPS, Houston, TX, Kelly Hunt, U10CA180858; University of Vermont and State Agricultural College, Burlington, VT, Steven Ades, and Walter Reed National Military Medical Center, Bethesda, MD, Karen Zeman.

REFERENCES

  • 1.Siegel RL, Miller KD, and Jemal A., Cancer statistics, 2016. CA Cancer J Clin, 2016. 66(1): p. 7–30. [DOI] [PubMed] [Google Scholar]
  • 2.Fisher B, et al. , Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst, 1998. 90(18): p. 1371–88. [DOI] [PubMed] [Google Scholar]
  • 3.Goss PE, et al. , Exemestane for breast-cancer prevention in postmenopausal women. N Engl J Med, 2011. 364(25): p. 2381–91. [DOI] [PubMed] [Google Scholar]
  • 4.Cuzick J, et al. , Anastrozole for prevention of breast cancer in high-risk postmenopausal women (IBIS-II): an international, double-blind, randomised placebo-controlled trial. Lancet, 2014. 383(9922): p. 1041–8. [DOI] [PubMed] [Google Scholar]
  • 5.Dent SF, et al. , Aromatase inhibitor therapy: toxicities and management strategies in the treatment of postmenopausal women with hormone-sensitive early breast cancer. Breast Cancer Res Treat, 2011. 126(2): p. 295–310. [DOI] [PubMed] [Google Scholar]
  • 6.Makubate B, et al. , Cohort study of adherence to adjuvant endocrine therapy, breast cancer recurrence and mortality. Br J Cancer, 2013. 108(7): p. 1515–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Waters EA, et al. , Use of tamoxifen and raloxifene for breast cancer chemoprevention in 2010. Breast Cancer Res Treat, 2012. 134(2): p. 875–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Smith SG, et al. , Factors affecting uptake and adherence to breast cancer chemoprevention: a systematic review and meta-analysis. Ann Oncol, 2016. 27(4): p. 575–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Feldman D, et al. , The role of vitamin D in reducing cancer risk and progression. Nat Rev Cancer, 2014. 14(5): p. 342–57. [DOI] [PubMed] [Google Scholar]
  • 10.Krishnan AV, Swami S, and Feldman D., Equivalent anticancer activities of dietary vitamin D and calcitriol in an animal model of breast cancer: importance of mammary CYP27B1 for treatment and prevention. J Steroid Biochem Mol Biol, 2013. 136: p. 289–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Chakraborti CK, Vitamin D as a promising anticancer agent. Indian J Pharmacol, 2011. 43(2): p. 113–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Wu Y, et al. , Association of Vitamin D3 Level with Breast Cancer Risk and Prognosis in African-American and Hispanic Women. Cancers (Basel), 2017. 9(10). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Yin L, et al. , Meta-analysis: serum vitamin D and breast cancer risk. Eur J Cancer, 2010. 46(12): p. 2196–205. [DOI] [PubMed] [Google Scholar]
  • 14.McDonnell SL, et al. , Breast cancer risk markedly lower with serum 25-hydroxyvitamin D concentrations >/=60 vs <20 ng/ml (150 vs 50 nmol/L): Pooled analysis of two randomized trials and a prospective cohort. PLoS One, 2018. 13(6): p. e0199265. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Anzano MA, et al. , 1 alpha,25-Dihydroxy-16-ene-23-yne-26,27-hexafluorocholecalciferol (Ro24–5531), a new deltanoid (vitamin D analogue) for prevention of breast cancer in the rat. Cancer Res, 1994. 54(7): p. 1653–6. [PubMed] [Google Scholar]
  • 16.Jacobson EA, et al. , Effects of dietary fat, calcium, and vitamin D on growth and mammary tumorigenesis induced by 7,12-dimethylbenz(a)anthracene in female Sprague-Dawley rats. Cancer Res, 1989. 49(22): p. 6300–3. [PubMed] [Google Scholar]
  • 17.Marchionatti AM, et al. , Antiproliferative action of menadione and 1,25(OH)2D3 on breast cancer cells. J Steroid Biochem Mol Biol, 2009. 113(3–5): p. 227–32. [DOI] [PubMed] [Google Scholar]
  • 18.Mehta RG, et al. , Prevention of preneoplastic mammary lesion development by a novel vitamin D analogue, 1alpha-hydroxyvitamin D5. J Natl Cancer Inst, 1997. 89(3): p. 212–8. [DOI] [PubMed] [Google Scholar]
  • 19.Nolan E, et al. , Dissociation of vitamin D3 and anti-estrogen mediated growth regulation in MCF-7 breast cancer cells. Mol Cell Biochem, 1998. 188(1–2): p. 13–20. [PubMed] [Google Scholar]
  • 20.Spina CS, et al. , Vitamin D and cancer. Anticancer Res, 2006. 26(4A): p. 2515–24. [PubMed] [Google Scholar]
  • 21.Bertone-Johnson ER, et al. , Plasma 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D and risk of breast cancer. Cancer Epidemiol Biomarkers Prev, 2005. 14(8): p. 1991–7. [DOI] [PubMed] [Google Scholar]
  • 22.Crew KD, et al. , Association between plasma 25-hydroxyvitamin D and breast cancer risk. Cancer Prev Res (Phila), 2009. 2(6): p. 598–604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Garland CF, et al. , Vitamin D and prevention of breast cancer: pooled analysis. J Steroid Biochem Mol Biol, 2007. 103(3–5): p. 708–11. [DOI] [PubMed] [Google Scholar]
  • 24.Gissel T, et al. , Intake of vitamin D and risk of breast cancer--a meta-analysis. J Steroid Biochem Mol Biol, 2008. 111(3–5): p. 195–9. [DOI] [PubMed] [Google Scholar]
  • 25.John EM, et al. , Vitamin D and breast cancer risk: the NHANES I Epidemiologic follow-up study, 1971–1975 to 1992. National Health and Nutrition Examination Survey. Cancer Epidemiol Biomarkers Prev, 1999. 8(5): p. 399–406. [PubMed] [Google Scholar]
  • 26.Lin J, et al. , Intakes of calcium and vitamin D and breast cancer risk in women. Arch Intern Med, 2007. 167(10): p. 1050–9. [DOI] [PubMed] [Google Scholar]
  • 27.Robien K, Cutler GJ, and Lazovich D., Vitamin D intake and breast cancer risk in postmenopausal women: the Iowa Women’s Health Study. Cancer Causes Control, 2007. 18(7): p. 775–82. [DOI] [PubMed] [Google Scholar]
  • 28.Rossi M, et al. , Vitamin D intake and breast cancer risk: a case-control study in Italy. Ann Oncol, 2009. 20(2): p. 374–8. [DOI] [PubMed] [Google Scholar]
  • 29.Abbas S, Chang-Claude J, and Linseisen J., Plasma 25-hydroxyvitamin D and premenopausal breast cancer risk in a German case-control study. Int J Cancer, 2009. 124(1): p. 250–5. [DOI] [PubMed] [Google Scholar]
  • 30.Berube S, et al. , Vitamin D and calcium intakes from food or supplements and mammographic breast density. Cancer Epidemiol Biomarkers Prev, 2005. 14(7): p. 1653–9. [DOI] [PubMed] [Google Scholar]
  • 31.Shin MH, et al. , Intake of dairy products, calcium, and vitamin d and risk of breast cancer. J Natl Cancer Inst, 2002. 94(17): p. 1301–11. [DOI] [PubMed] [Google Scholar]
  • 32.Thomson CA, et al. , Pilot study of dietary influences on mammographic density in pre- and postmenopausal Hispanic and non-Hispanic white women. Menopause, 2007. 14(2): p. 243–50. [DOI] [PubMed] [Google Scholar]
  • 33.Lappe JM, et al. , Vitamin D and calcium supplementation reduces cancer risk: results of a randomized trial. Am J Clin Nutr, 2007. 85(6): p. 1586–91. [DOI] [PubMed] [Google Scholar]
  • 34.Chlebowski RT, et al. , Calcium plus vitamin D supplementation and the risk of breast cancer. J Natl Cancer Inst, 2008. 100(22): p. 1581–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Lappe J, et al. , Effect of Vitamin D and Calcium Supplementation on Cancer Incidence in Older Women: A Randomized Clinical Trial. JAMA, 2017. 317(12): p. 1234–1243. [DOI] [PubMed] [Google Scholar]
  • 36.Henry NL and Hayes DF, Cancer biomarkers. Mol Oncol, 2012. 6(2): p. 140–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Boyd NF, et al. , Breast tissue composition and susceptibility to breast cancer. J Natl Cancer Inst, 2010. 102(16): p. 1224–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Shepherd JA, et al. , Volume of mammographic density and risk of breast cancer. Cancer Epidemiol Biomarkers Prev, 2011. 20(7): p. 1473–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Pinto Pereira SM, et al. , Localized fibroglandular tissue as a predictor of future tumor location within the breast. Cancer Epidemiol Biomarkers Prev, 2011. 20(8): p. 1718–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Pettersson A, et al. , Mammographic density phenotypes and risk of breast cancer: a meta-analysis. J Natl Cancer Inst, 2014. 106(5). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.McCormack VA and dos Santos Silva I., Breast density and parenchymal patterns as markers of breast cancer risk: a meta-analysis. Cancer Epidemiol Biomarkers Prev, 2006. 15(6): p. 1159–69. [DOI] [PubMed] [Google Scholar]
  • 42.Boyd NF, et al. , Mammographic density and the risk and detection of breast cancer. N Engl J Med, 2007. 356(3): p. 227–36. [DOI] [PubMed] [Google Scholar]
  • 43.Boyd NF, Mammographic density and risk of breast cancer. Am Soc Clin Oncol Educ Book, 2013. [DOI] [PubMed]
  • 44.Chai W, Maskarinec G, and Cooney RV, Serum 25-hydroxyvitamin D levels and mammographic density among premenopausal women in a multiethnic population. Eur J Clin Nutr, 2010. 64(6): p. 652–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Brisson J, et al. , Synchronized seasonal variations of mammographic breast density and plasma 25-hydroxyvitamin d. Cancer Epidemiol Biomarkers Prev, 2007. 16(5): p. 929–33. [DOI] [PubMed] [Google Scholar]
  • 46.Lope V, et al. , Serum 25-hydroxyvitamin D and mammographic density in premenopausal Spanish women. J Steroid Biochem Mol Biol, 2019. 189: p. 101–107. [DOI] [PubMed] [Google Scholar]
  • 47.Straub L, et al. , Mammographic Density and Vitamin D Levels - A Cross-sectional Study. Geburtshilfe Frauenheilkd, 2017. 77(3): p. 257–267. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Antoni S, et al. , Is mammographic density differentially associated with breast cancer according to receptor status? A meta-analysis. Breast Cancer Res Treat, 2013. 137(2): p. 337–47. [DOI] [PubMed] [Google Scholar]
  • 49.Ma H, et al. , Is there a difference in the association between percent mammographic density and subtypes of breast cancer? Luminal A and triple-negative breast cancer. Cancer Epidemiol Biomarkers Prev, 2009. 18(2): p. 479–85. [DOI] [PubMed] [Google Scholar]
  • 50.Byng JW, et al. , Analysis of mammographic density and breast cancer risk from digitized mammograms. Radiographics, 1998. 18(6): p. 1587–98. [DOI] [PubMed] [Google Scholar]
  • 51.Pocock SJ and Simon R., Sequential treatment assignment with balancing for prognostic factors in the controlled clinical trial. Biometrics, 1975. 31(1): p. 103–15. [PubMed] [Google Scholar]
  • 52.Boyd NF, L.G.A., Byng JW, et al. , Mammographic densities and breast cancer risk. Cancer Epidemiol Biomarkers Prev, 1998. 10(3–4):113–26. [DOI] [PubMed] [Google Scholar]
  • 53.Chow CK, et al. , Effect of tamoxifen on mammographic density. Cancer Epidemiol Biomarkers Prev, 2000. 9(9): p. 917–21. [PubMed] [Google Scholar]
  • 54.Brisson J, et al. , A Randomized Double-Blind Placebo-Controlled Trial of the Effect of Vitamin D3 Supplementation on Breast Density in Premenopausal Women. Cancer Epidemiol Biomarkers Prev, 2017. 26(8): p. 1233–1241. [DOI] [PubMed] [Google Scholar]
  • 55.Crew KD, et al. , Randomized Double-Blind Placebo-Controlled Biomarker Modulation Study of Vitamin D Supplementation in Premenopausal Women at High Risk for Breast Cancer (SWOG S0812). Cancer Prev Res (Phila), 2019. 12(7): p. 481–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Duma N, et al. , Assessing vitamin D and mammographic breast density in Alaskan women. Clin Pract, 2020. 10(4): p. 1253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Heo DS, et al. , The association between 25-hydroxyvitamin D and mammographic density in healthy pre- and postmenopausal women regardless of the menstrual cycle phase: a cross-sectional study. Nutr Cancer, 2014. 66(1): p. 97–103. [DOI] [PubMed] [Google Scholar]
  • 58.Bertrand KA, et al. , Premenopausal plasma 25-hydroxyvitamin D, mammographic density, and risk of breast cancer. Breast Cancer Res Treat, 2015. 149(2): p. 479–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Amadou A, et al. , Serum 25-Hydroxyvitamin D3 and Mammography Density among Mexican Women. PLoS One, 2016. 11(8): p. e0161686. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Razzaghi H, et al. , Association between mammographic density and basal-like and luminal A breast cancer subtypes. Breast Cancer Res, 2013. 15(5): p. R76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Bertrand KA, et al. , Mammographic density and risk of breast cancer by age and tumor characteristics. Breast Cancer Res, 2013. 15(6): p. R104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Holick MF, Vitamin D deficiency. N Engl J Med, 2007. 357(3): p. 266–81. [DOI] [PubMed] [Google Scholar]
  • 63.Dietze EC, et al. , Triple-negative breast cancer in African-American women: disparities versus biology. Nat Rev Cancer, 2015. 15(4): p. 248–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Newman LA and Kaljee LM, Health Disparities and Triple-Negative Breast Cancer in African American Women: A Review. JAMA Surg, 2017. 152(5): p. 485–493. [DOI] [PubMed] [Google Scholar]
  • 65.Newman LA, et al. , Association Between Benign Breast Disease in African American and White American Women and Subsequent Triple-Negative Breast Cancer. JAMA Oncol, 2017. 3(8): p. 1102–1106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Lope V, et al. , Serum 25-hydroxyvitamin D and breast cancer risk by pathological subtype (MCC-Spain). J Steroid Biochem Mol Biol, 2018. 182: p. 4–13. [DOI] [PubMed] [Google Scholar]
  • 67.Tommie JL, Pinney SM, and Nommsen-Rivers LA, Serum Vitamin D Status and Breast Cancer Risk by Receptor Status: A Systematic Review. Nutr Cancer, 2018. 70(5): p. 804–820. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Peppone LJ, et al. , The association between breast cancer prognostic indicators and serum 25-OH vitamin D levels. Ann Surg Oncol, 2012. 19(8): p. 2590–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.de Sousa Almeida-Filho B, et al. , Vitamin D deficiency is associated with poor breast cancer prognostic features in postmenopausal women. J Steroid Biochem Mol Biol, 2017. 174: p. 284–289. [DOI] [PubMed] [Google Scholar]
  • 70.Shan NL, et al. , Vitamin D compounds inhibit cancer stem-like cells and induce differentiation in triple negative breast cancer. J Steroid Biochem Mol Biol, 2017. 173: p. 122–129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 71.LaPorta E and Welsh J., Modeling vitamin D actions in triple negative/basal-like breast cancer. J Steroid Biochem Mol Biol, 2014. 144 Pt A: p. 65–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72.Azam S, et al. , Determinants of Mammographic Density Change. JNCI Cancer Spectr, 2019. 3(1): p. pkz004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Maskarinec G, et al. , A longitudinal investigation of mammographic density: the multiethnic cohort. Cancer Epidemiol Biomarkers Prev, 2006. 15(4): p. 732–9. [DOI] [PubMed] [Google Scholar]

RESOURCES