Outcomes of Screening Mammography by Frequency, Breast Density, and Postmenopausal Hormone Therapy

Importance Controversy exists about the frequency women should undergo screening mammography and whether screening interval should vary according to risk factors beyond age.

Objective To compare the benefits and harms of screening mammography frequencies according to age, breast density, and postmenopausal hormone therapy (HT) use.

Design Prospective cohort.

Setting Data collected January 1994 to December 2008 from mammography facilities in community practice that participate in the Breast Cancer Surveillance Consortium (BCSC) mammography registries.

Participants Data were collected prospectively on 11 474 women with breast cancer and 922 624 without breast cancer who underwent mammography at facilities that participate in the BCSC.

Main Outcomes and Measures We used logistic regression to calculate the odds of advanced stage (IIb, III, or IV) and large tumors (>20 mm in diameter) and 10-year cumulative probability of a false-positive mammography result by screening frequency, age, breast density, and HT use. The main predictor was screening mammography interval.

Results Mammography biennially vs annually for women aged 50 to 74 years does not increase risk of tumors with advanced stage or large size regardless of women's breast density or HT use. Among women aged 40 to 49 years with extremely dense breasts, biennial mammography vs annual is associated with increased risk of advanced-stage cancer (odds ratio [OR], 1.89; 95% CI, 1.06-3.39) and large tumors (OR, 2.39; 95% CI, 1.37-4.18). Cumulative probability of a false-positive mammography result was high among women undergoing annual mammography with extremely dense breasts who were either aged 40 to 49 years (65.5%) or used estrogen plus progestogen (65.8%) and was lower among women aged 50 to 74 years who underwent biennial or triennial mammography with scattered fibroglandular densities (30.7% and 21.9%, respectively) or fatty breasts (17.4% and 12.1%, respectively).

Conclusions and Relevance Women aged 50 to 74 years, even those with high breast density or HT use, who undergo biennial screening mammography have similar risk of advanced-stage disease and lower cumulative risk of false-positive results than those who undergo annual mammography. When deciding whether to undergo mammography, women aged 40 to 49 years who have extremely dense breasts should be informed that annual mammography may minimize their risk of advanced-stage disease but the cumulative risk of false-positive results is high.

In 2009, the US Preventive Services Task Force issued guidelines that biennial mammography, rather than the previously recommended mammography every 1 to 2 years, be performed for women aged 50 to 74 years.¹ Because of insufficient evidence, the updated guidelines did not consider the influence of breast cancer risk factors beyond age.² Women with risk factors that increase the chance of advanced-stage breast cancer at diagnosis may benefit from frequent screening to increase the chance of identifying tumors at an early stage. For example, high breast density is associated with larger tumor size,^3-6 positive lymph nodes,^4,7,8 and advanced-stage disease.⁹ Postmenopausal estrogen plus progestogen use for 5 years or more increases the likelihood of development and diagnosis of breast cancer at an advanced stage,^10,11 and risk of advanced-stage disease is increased further in women with dense breasts who use postmenopausal hormone therapy (HT).⁹

Few studies have reported on whether risk factors combined with screening mammography frequency influence outcomes. A decision analysis evaluated the benefits and harms of risk-based screening and found that biennial screening of women aged 40 to 49 years with high breast density and with either a first-degree relative with breast cancer or history of breast biopsy had similar benefit-harm ratios as biennial screening of average-risk women in their fifties.¹² Furthermore, women aged 50 to 69 years with low breast density could be screened less often than biennially without decreased benefit.¹²

Our study aimed to extend the literature by reporting whether the benefits (detection of early-stage disease) and harms (false-positive mammography result or biopsy recommendation) differ among women undergoing screening mammography in community practice by screening frequency according to age, breast density, and postmenopausal HT use.

Data are from the Breast Cancer Surveillance Consortium (BCSC) mammography registries (http://breastscreening.cancer.gov), which are comparable to the US population.^13,14 Registries collected data from community radiology facilities including patient characteristics and clinical information. Radiologists' assessments and recommendations were based on the American College of Radiology's Breast Imaging Reporting and Data System (BI-RADS).¹⁵ Breast cancer diagnoses and tumor characteristics were obtained by linking BCSC data to pathology databases, regional Surveillance, Epidemiology, and End Results (SEER) programs, and state tumor registries, with completeness of reporting estimated at greater than 94.3%.¹⁶ Data were pooled at a central Statistical Coordinating Center. Registries and the Coordinating Center received institutional review board approval for active or passive consenting processes or a waiver of consent to enroll participants, link data, and perform analysis. All procedures were Health Insurance Portability and Accountability Act compliant, and registries and the coordinating center received a federal Certificate of Confidentiality and other protections for the identities of women, physicians, and facilities.

We evaluated women aged 40 to 74 years with and without breast cancer (Figure 1). Analyses of tumor characteristics included women who were diagnosed as having incident invasive breast cancer or ductal carcinoma in situ, either screen detected or interval cancer, between 1996 and 2008 and who had at least 2 screening mammography examinations before diagnosis. Women were classified based on the time between the 2 most recent screening examinations as either annual (9-18 months apart), biennial (>18-30 months apart), or triennial (>30-42 months apart) (Figure 2). We restricted analyses to breast cancers diagnosed within a specified follow-up period after a woman's index examination (screening mammography before breast cancer diagnosis): within 1 year for annual, 2 years for biennial, and 3 years for triennial screening, as would be done in a randomized trial (Figure 2). To allow adequate follow-up, we included only index examinations that occurred at least 1 year before the end of complete cancer data collection by a woman's BCSC registry for annual interval, at least 2 years for biennial interval, and at least 3 years for triennial interval.

For the cumulative false-positive probabilities analysis, we included first and subsequent screening mammography examinations from 1994 to 2008 from women without a history of breast cancer and without a breast cancer diagnosis within 1 year after mammography. We censored women at their prior screening examination if their self-reported time since last examination differed from that in the database by more than 6 months, to ensure an accurate count of mammography examinations.

Demographic and breast health history information were obtained on a self-administered questionnaire completed at each mammography examination. We obtained information on history of first-degree relatives (mother, sister, or daughter) with breast cancer and current postmenopausal HT use at the time of mammography. Women aged 50 to 74 years with hysterectomy information were included in analyses by hormone type. Women with a uterus receiving HT were classified as using estrogen plus progestogen (combination HT), whereas women without a uterus receiving HT were classified as using estrogen only, as previously described.¹¹ We used self-reported race/ethnicity to categorize women as non-Hispanic white, non-Hispanic black, Hispanic, Asian/Native Hawaiian/Pacific Islander, Native American/Native Alaskan or other/mixed race. If self-reported race/ethnicity was missing, we used information from cancer registries. Breast density was categorized by radiologists at the time of clinical interpretation using BI-RADS breast density categories: 1 = almost entirely fat; 2 = scattered fibroglandular densities; 3 = heterogeneously dense; 4 = extremely dense.

Breast cancers were classified according to the American Joint Committee on Cancer (AJCC) staging system, sixth edition.¹⁷ We defined advanced-stage disease as AJCC stages IIb, III, or IV and large tumors as greater than 20 mm in diameter. The AJCC sixth edition staging was classified as early- or late-stage disease based on SEER summary stage and other tumor characteristics (see eAppendix).

Mammography examinations were considered screening based on the indication reported by radiologists. To minimize misclassifying diagnostic mammography as screening, we excluded examinations that were unilateral or were preceded by a breast imaging study within 9 months.

A false-positive recall or biopsy recommendation was defined as no invasive carcinoma or ductal carcinoma in situ diagnosis within 1 year after a positive screening examination result or before the next screening examination, whichever occurred first. A screening examination was considered positive for recall if the initial BI-RADS assessment was 0 (needs additional imaging evaluation); 4 (suspicious abnormality); 5 (highly suggestive of malignancy); or 3 (probably benign finding with a recommendation for immediate evaluation). A screening examination result was considered positive for biopsy recommendation if the final BI-RADS assessment after all imaging workup and within 90 days after the screening examination was 4 or 5—or was 0 or 3 with a recommendation for biopsy, fine needle aspiration, or surgical consultation. Examinations were excluded from the biopsy recommendation analysis if the final assessment, 90 days after the screening mammography, was BI-RADS 0, with a recommendation for additional imaging, nonspecified workup, or missing a recommendation.

We describe the distribution of risk factors among women with and without breast cancer. Among cancer cases, we estimated the proportion with invasive cancer vs ductal carcinoma in situ. Among women with invasive cancer, we estimated distributions of tumor characteristics (stage, size, and lymph node status) at diagnosis by age, screening interval, breast density, and HT use at the study closest to cancer diagnosis. We used logistic regression to estimate odds ratios (ORs) and 95% confidence intervals of adverse (vs more favorable) invasive tumor characteristics associated with screening intervals by breast density, HT use, and age group. Models were adjusted for age in years, BCSC mammography registry, and race/ethnicity. Because breast cancer among women aged 40 to 49 years with fatty breasts is uncommon (<1%), we combined women aged 40 to 49 years with fatty breasts and those with scattered fibroglandular densities.¹⁸ Similarly, breast cancer among women aged 50 to 74 years with extremely dense breasts is uncommon (2%-6%), so we combined women aged 50 to 74 years with heterogeneously and extremely dense breasts.¹⁸

We estimated the probability of a false-positive first mammography result using logistic regression including breast density and screening interval terms in the model and adjusted for BCSC registry. Probability estimates were standardized to the BCSC mammography registry distribution using indirect (marginal) standardization. We modeled the cumulative probability of false-positive results after 10 years of subsequent screening using previously developed methods.¹⁹ Briefly, we fit logistic regression models for false-positive results at each subsequent screening round conditional on screening round number, total number of screening rounds before censoring, screening interval, breast density, and BCSC mammography registry. All estimates were stratified by age (40-49 vs 50-74 years) and by type of HT use for women aged 50 to 74 years. We combined estimates of the false-positive risk at each subsequent screening round according to age at first examination and HT use at each examination to obtain woman-level cumulative false-positive probabilities after 10 years of repeated screening. We report fitted values from this model by breast density, screening interval, age, and HT use.

Analyses of tumor characteristics were performed using SAS version 9.2 statistical software (SAS Institute Inc). Analyses of cumulative false-positive probabilities were performed using R 2.10.1 (R Foundation for Statistical Computing).

We included 11 474 women with breast cancer; the majority were 50 years or older and white and had heterogeneously dense or extremely dense breasts. Percentages of interval cancers increased with increasing screening interval (Table 1).

The proportion of tumors associated with less-favorable prognostic characteristics (stage IIb or higher, size >20 mm, and positive lymph nodes) was higher among women with high breast density (heterogeneously dense or extremely dense) compared with women with low or average breast density (fatty or scattered fibroglandular densities) (Table 2). Within density categories, the proportion with less favorable prognostic tumor characteristics did not vary by screening interval except among women with extremely dense breasts, forwhom a 3-year screening interval was associated with a higher proportion of advanced stage, large tumors, and positive lymph nodes (Table 2).

We calculated ORs comparing the risk of less favorable tumor characteristics by screening interval (Table 3). Compared with annual mammography, women aged 50 to 74 years undergoing biennial mammography were not at increased risk of less favorable tumor characteristics regardless of breast density or HT use. Women undergoing biennial mammography receiving combination HT with heterogeneously dense or extremely dense breasts had a non–statistically significant increased risk of advanced stage (OR, 1.56; 95% CI, 0.88-2.80) and large tumor size (OR, 1.59; 95% CI, 0.97-2.61). No differences were observed in tumor characteristics among women aged 50 to 74 years undergoing triennial vs biennial mammography. In contrast, women aged 40 to 49 years with extremely dense breasts undergoing biennial compared with annual mammography were at increased risk of advanced stage (OR, 1.89; 95% CI, 1.06-3.39) and large tumor size (OR, 2.39; 95% CI, 1.37-4.18).

In a sensitivity analysis, model results were similar with adjustment for family history of breast cancer. We did not include this factor in the main model because missing values reduced our sample size.

We included 922 624 women who underwent 2 099 648 screening examinations; more than half of women with extremely dense breasts were aged 40 to 49 years (Table 4).

When screening women aged 50 to 74 years with scattered fibroglandular densities not receiving HT, the cumulative probability of a woman receiving at least 1 false-positive mammography result after 10 years was 49.8% with annual, 30.7% with biennial, and 21.9% with triennial screening (Table 5). Estimates were similar among estrogen-only users. Among women aged 50 to 74 years undergoing annual mammography, estimates were highest among women with extremely dense (65.8%) or heterogeneously dense breasts (68.1%) receiving combination HT. Estimates were lowest for women aged 50 to 74 years with fatty breasts (30.3% with annual, 17.4% with biennial, and 12.1% with triennial mammography not receiving HT), and even low among HT users. The cumulative probability of at least 1 false-positive mammography result after 10 years was highest among women aged 40 to 49 years undergoing annual screening with heterogeneously dense (68.9%) or extremely dense breasts (65.5%). Estimates of the cumulative probability of a woman receiving at least 1 false-positive biopsy recommendation after 10 years had a similar pattern to that of false-positive mammography results: (1) risk decreased as screening interval increased, (2) risk was lowest among women with fatty breasts, and (3) risk was highest among combination HT users with dense breasts (Table 5).

We found that biennial screening mammography for most women aged 40 to 49 and 50 to 74 years, even among those with high breast density or receiving combination HT, results in a similar risk of presenting with advanced-stage disease as annual screening mammography. Notably, most women who undergo annual mammography are at high risk of false-positive mammography results and biopsy recommendations without added benefit from more frequent screening. However, a small proportion of women aged 40 to 49 years with extremely dense breasts are more likely to present with advanced-stage disease if they undergo biennial vs annual screening mammography. This benefit is counterbalanced by a higher risk of cumulative false-positive mammography results with annual screening.

Our results are consistent with those of randomized controlled trials, a population-based screening program, a community-based study, and statistical models that report annual mammography has minimal if any additional benefit over biennial mammography for women aged 50 to 74 years.^12,20-24 For women aged 40 to 49 years, less data are available on effectiveness by screening interval. Of 6 statistical models that incorporate US population-based breast cancer incidence and mortality information from the SEER program and US population-based mammography outcomes from the BCSC, 4 showed no additional deaths averted by annual vs biennial screening for women aged 40 to 49 years.²³ A recent BCSC community-based study found no statistically significant absolute difference in the overall proportion of advanced-stage cancer with biennial compared with annual screening.²⁴ We add to the literature by showing that for women aged 40 to 49 years with extremely dense breasts, who have increased risk of advanced-stage disease⁹ and missed breast cancers by mammography,^25,26 annual screening has added benefit to detect breast cancer at an earlier stage than biennial screening. As others have shown,^12,23,27 we found the added benefit of annual screening is offset by increased risk of false-positive mammography results and breast biopsy recommendations. The 12% to 15% of women aged 40 to 49 years with extremely dense breasts,¹⁸ whose risk of breast cancer is similar to average-risk women aged 50 to 59 years,^12,27,28 will need to decide if the added benefit is outweighed by the additional harms of annual screening including doubling the number of mammograms and increased risk of false-positive mammography results and breast biopsy recommendations. For the majority of women aged 40 to 49 years without extremely dense breasts, biennial mammography is associated with a similar risk of advanced-stage disease as annual screening, and the cumulative risk of false-positive screening results and biopsy recommendations is lower.

Although postmenopausal combination HT use for 5 years or more increases the likelihood of developing breast cancer that is diagnosed at an advanced stage,^9-11 and this risk is increased even more in women with dense breasts,⁹ the risk of advanced disease did not differ significantly in HT users with dense breasts undergoing biennial vs annual mammography. Perhaps frequent screening does not decrease the risk of advanced-stage disease in women with dense breasts receiving combination HT because increased breast density obscures identification of tumors and/or tumors grow rapidly in a short period.^29-31 Alternatively, we may have had insufficient statistical power to observe a benefit from annual mammography. Our results need confirmation to determine whether biennial mammography increases the risk of advanced disease in women with extremely dense breasts receiving combination HT.

Prior studies report that postmenopausal combination HT increases the risk of abnormal mammography.^10,26 We found that for women aged 50 to 74 years with average or high breast density, receiving combination HT increases the cumulative probability of receiving at least 1 false-positive mammography result after 10 years, and the magnitude of the risk is similar to that of women aged 40 to 49 years with high breast density. Women receiving combination HT should be informed of the increased risk of false-positive mammography results compared with women their age not receiving HT and that stopping HT can reduce this risk.^32,33

Women with fatty breasts are at low risk of breast cancer, regardless of age, menopausal status, family history of breast cancer, history of breast biopsy, and HT use.^9,34 Moreover, women with fatty breasts are at reduced risk of advanced-stage disease.⁹ Our results show that women with fatty breast density have the lowest cumulative probability of false-positive mammography results or biopsy recommendations after 10 years of screening. This low probability is probably because radiologists can easily discern whether a lesion is suggestive of malignancy in women with fatty breasts because the characteristics of the lesion are not obscured by normal fibroglandular tissue, so fewer women are recalled for diagnostic evaluation. Taken together, our results suggest there is no added benefit of screening women with fatty breasts annually and false-positive results are low compared with women with high breast density. One study supporting our findings found that mammography every 3 to 4 years was cost-effective for women aged 50 to 79 years with fatty breasts and no other risk factors.¹²

In observational studies, women at high breast cancer risk may undergo more frequent screening than low-risk women, which could spuriously inflate advanced disease rates among frequent screening. To minimize this potential bias, we evaluated the proportion of cases with advanced disease. Although more than 10 000 breast cancers were identified among women undergoing screening mammography, we had limited statistical power to examine subgroups of women with fatty breasts and those undergoing triennial mammography. Misclassification of BI-RADS density because of modest interrater agreement between radiologists^35-37 could result in under or overestimation of associations with breast cancer outcomes by density category. We evaluated numerous comparisons; some may be significant by chance. Thus, it is important to consider the magnitude of differences and confidence interval widths. Also, most study examinations were film screen. The sensitivity of digital mammography is higher in women with extremely dense breasts,³⁸ which could result in a smaller difference in tumor outcomes between annual, and biennial screening than we observed. We did not adjust for body mass index because data were missing in 50% of women, mostly because facilities do not collect this information.

In conclusion, women aged 50 to 74 years, regardless of breast density or HT use, can undergo biennial rather than annual mammography because biennial screening does not increase the risk of presenting with advanced disease but does substantially reduce the cumulative risk of a false-positive mammography result and biopsy recommendation. Women aged 40 to 49 years with extremely dense breasts who choose to undergo mammography should consider annual screening to decrease the risk of advanced-stage disease but should be informed that annual screening leads to a high cumulative probability of a false-positive mammography result because of the additional screening examinations.

Correspondence: Karla Kerlikowske, MD, General Internal Medicine Section, San Francisco Veterans Affairs Medical Center, 4150 Clement St, Mailing Code 111A1, San Francisco, CA 94121 (Karla.Kerlikowske@ucsf.edu).

Accepted for Publication: January 15, 2013.

Published Online: March 18, 2013. doi:10.1001/jamainternmed.2013.307

Author Contributions: Dr Kerlikowske had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Kerlikowske, Zhu, Miglioretti, and O’Meara. Acquisition of data: Kerlikowske, Zhu, Geller, and Miglioretti. Analysis and interpretation of data: Kerlikowske, Zhu, Hubbard, Geller, Dittus, Braithwaite, Wernli, Miglioretti, and O’Meara. Drafting of the manuscript: Kerlikowske, Hubbard, Geller, and O’Meara. Critical revision of the manuscript for important intellectual content: Kerlikowske, Zhu, Hubbard, Geller, Dittus, Braithwaite, Wernli, Miglioretti, and O’Meara. Statistical analysis: Zhu, Hubbard, and Miglioretti. Obtained funding: Kerlikowske, Geller, and Miglioretti. Administrative, technical, and material support: Kerlikowske and Dittus. Study supervision: Miglioretti.

Funding/Support: This work was supported by the National Cancer Institute–funded Breast Cancer Surveillance Consortium cooperative agreement (grants U01CA63740, U01CA86076, U01CA86082, U01CA63736, U01CA70013, U01CA69976, U01CA63731, U01CA70040) and the National Cancer Institute–funded grants RC2CA148577, R03CA150007, and P01 CA107584. The collection of cancer data used in this study was supported in part by several state public health departments and cancer registries throughout the United States For a full description of these sources, please see http://breastscreening.cancer.gov/work/acknowledgement.html.

Disclaimer: The design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript do not represent those of the National Cancer Institute, and this organization had no role in the final decision to submit the manuscript for publication.

Additional Contributions: We thank the participating women, mammography facilities, and radiologists for the data they have provided for this study.