Revisiting Screening in Women With a Family History of Breast Cancer

Abstract

Women with a family history (FH) of breast cancer and without known genetic susceptibility represent a unique population whose lifetime probability of developing breast cancer varies widely depending on familial factors, breast density, and the risk assessment tool used. Recently updated guidelines from the American College of Radiology recommend supplemental annual screening with contrast-enhanced MRI or contrast-enhanced mammography for women with an FH who are high risk (≥20% lifetime risk) or have dense breasts. To date, most screening studies addressing outcomes in women with FH have largely included those also with confirmed or suspected gene mutations, in whom the lifetime risk is highest, with limited data for women at average to intermediate risk who are not known to be genetically susceptible and may not benefit as much from the same screening approaches. Further research focusing specifically on women with FH as the only breast cancer risk factor is warranted to refine risk assessment and optimize a multimodality personalized screening approach.

Introduction

Family history (FH) is a well-established risk factor for developing breast cancer (1–4). Breast cancer in a first-degree relative has been reported in up to 20% of women in the U.S. and amounted to over 9 million women in a 2015 national survey (2,5–7). Women with first-degree FH are estimated to be 2.1 times more likely to develop breast cancer and to have a higher five-year cumulative incidence of breast cancer (19.9 vs 13.0/1000) than those without, based on retrospective study data compiled from 1935 to 2016 (3,4). However, as in clinical practice, most analyses assessing incidence and risk of familial breast cancer are based on patient-provided FH data, with or without knowledge of potential genetic susceptibility or other contributory risk factors.

Although genetic mutations account for less than 10% of all breast cancers, the presence of a mutation significantly increases a carrier’s lifetime probability of developing breast cancer, warranting earlier and more intensive screening (6,8–10). Deleterious mutations of the BRCA oncogenes (eg, BRCA1 and BRCA2) are, based on current knowledge, the most common mutations associated with hereditary breast cancer and may increase a woman's lifetime probability of developing breast cancer to 70% (9,10). In contrast, the lifetime probability for a woman at average risk in the U.S. is approximately 13% (11). Women who are found to be noncarriers of an identified, family specific BRCA1/2 mutation may follow general screening guidelines in the absence of other risk factors because no increased breast cancer risk was observed for these women in a large multinational study (relative risk, 0.39) (8). Furthermore, less than one-quarter of familial cases in the U.S. have been attributed to BRCA mutations, suggesting the presence of other low-penetrance genes as well as nongenetic factors, such as environmental or hormonal exposures (6,12,13).

For women with an FH of breast cancer and no known genetic mutation, the lifetime probability of developing breast cancer is variable, depending on multiple familial factors (relationship, number, and age at diagnosis of family members affected) and the specific risk assessment model used (3,4). Moreover, interpretation of the screening literature for these women is challenging due to the marked heterogeneity of studies regarding the availability of genetic information (ie, BRCA mutation status), definitions of risk categories, and inclusion of other risk factors such as dense breast tissue and personal history of breast disease.

The focus of this article is to review the current screening evidence for women who have FH as their only risk factor for breast cancer.

Risk Stratification

Recently updated guidelines from the American College of Radiology (ACR) recommend that all women have a formal breast cancer risk assessment by the age of 25 years to allow for early identification of those who may be at higher-than-average risk, especially women of Ashkenazi Jewish descent, Black women, and women belonging to other racial and ethnic minority groups. Several clinical risk stratification models have been developed to predict lifetime risk in women with FH to inform screening recommendations and identify potential gene mutation carriers, including the Tyrer-Cuzick, Gail, Claus, BRCAPRO, Breast Cancer Surveillance Consortium (BCSC), and Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm models (14). However, these complex models are multifactorial, with each weighing individual risk parameters to varying degrees, which often leads to different calculations of lifetime breast cancer risk depending on the model used (15). While all of the clinical risk stratification models include FH, the extent to which FH is incorporated into individual risk models varies. For example, the Gail model only includes first-degree family members, while the Tyrer-Cuzick model includes information about several degrees of relatives in addition to their individual risk factors (15). The Claus model incorporates more extensive FH, including age of breast cancer onset of first- and second-degree relatives (15). These variables such as age, number, and degree of affected relatives have implications for risk. For women with FH and no known genetic mutation, lifetime risk ranges from high (≥20%) to average risk (<15%) (16). Early data show potential for imaging-based artificial intelligence to improve breast cancer risk stratification using information from mammography images, although more research is warranted (17,18).

Digital Mammography and Digital Breast Tomosynthesis

Mammography remains the standard for breast cancer screening because it is the only modality to have demonstrated decreased mortality through multiple randomized control trials (19). Annual mammographic screening may begin earlier than the age of 40 years in women with an FH of breast cancer but is not recommended before 30 years—this is because breast cancer incidence is low and the potential harm of radiation exposure outweighs the benefit (3,10,20,21). Guidelines from the National Comprehensive Cancer Network recommend that women at high risk (≥20% lifetime risk) initiate screening 7 to 10 years before the age of the youngest family member who was diagnosed (21). In a recent study using BCSC data for 306 147 women, 11% of whom reported a first-degree FH of breast cancer, the five-year cumulative incidence of breast cancer was found to increase as the relative’s diagnosis age decreased (3). The authors therefore recommended initiation of screening 5 to 8 years before the relative’s diagnosis age, if diagnosed at 45 years or younger, when their five-year cumulative incidence matched that of a 50-year-old woman (15.2/ 1000) (3).

Digital mammography (DM) relies on the detection of morphologic changes in the breast and therefore has lower sensitivity in dense breast tissue, which is seen in 46% of women in the U.S. undergoing screening (22). Higher breast density is also an independent risk factor for developing breast cancer (23). Digital breast tomosynthesis (DBT) is a newer mammographic technique that generates quasi-3D slices through the breast. The main benefit of DBT over DM in screening in the U.S. is improved recall rates, with more varied results for improved cancer detection rates (CDR) (24,25). The sensitivity of DM ranges between 31% and 87% in women reporting FH as at least one risk factor and may be lowest in young women who initiate screening earlier and tend to have denser breasts (20,26–28). Moreover, several studies have demonstrated limited benefit of mammography before the age of 40 years in women with genetic mutations who are undergoing annual MRI screening, especially BRCA1/2 mutation carriers who have increased radiation susceptibility (20,29–31). Nevertheless, screening DM has detected cancers that are MRI occult in higher-risk women, mostly ductal carcinoma in situ (DCIS) (27,32–35). Thus, the ACR recommends annual DM/DBT for women at high risk with FH who are noncarriers beginning at the age of 30 years (20).

MRI

Breast MRI is a vascular technique that uses intravenously injected gadolinium-based contrast to depict tumor neovascularity. MRI has proved superior to conventional screening techniques in women who are at average and above-average risk (28,33,36–38).

Current ACR and American Cancer Society screening guidelines recommend annual MRI for women with a strong FH and a lifetime risk calculation of 20% or more (20,39) (Figure 1). For such high-risk women, supplemental screening with MRI yields an incremental CDR of 8 to 24 cancers per 1000 when combined with DM, and 13 to 16 cancers per 1000 when combined with DM plus US (28). Most studies performed to evaluate MRI in this high-risk setting included women with both hereditary and familial risk (Table 1). Only a few studies have specifically focused on women with FH and no known gene mutation. The first study was a meta-analysis of six prospective MRI screening trials (32–34,36,40,41) performed between 1997 and 2011 that included 2226 women with FH (lifetime risk >15%) without a known genetic mutation (42). All women underwent screening mammography and MRI within three months, yielding a pooled combined DM/MRI CDR of 12 cancers per 1000. The addition of MRI significantly improved the sensitivity of screening from 55% (DM alone) and 89% (MRI alone) to 98% (combined MRI and DM; P <.001), although specificity was reduced from 94% (DM alone) and 83% (MRI alone) to 79% (combined MRI and DM; P = .002). The benefit of screening MRI was observed across all age groups. Though breast density was not specifically evaluated in the meta-analysis, one study found that MRI performance did not differ on subgroup analysis of age or breast density, suggesting that MRI remains beneficial in older women with nondense breasts (34). Results from this meta-analysis are limited by heterogeneity of inclusion criteria and risk assessment tools used.

Another prospective study randomized women with FH (and no mutation of BRCA or TP53 genes) to undergo either annual screening with MRI and biennial DM or annual DM alone (26). This study design differed from prior paired studies in which each patient had both MRI and DM at the same time. Women with a personal history of DCIS were not excluded from the study. Compared with DM, MRI detected significantly more cancers (CDR, 14/1000 vs 5/1000; P <.001) but had lower specificity (84% vs 91%; P <.001). Most importantly, the MRI-detected cancers on incidence rounds were smaller, node negative, and of an earlier stage compared with those detected by DM, suggesting potential for decreased breast cancer mortality.

In a prospective study of 559 high-risk women with BRCA mutations or a strong FH (lifetime risk ≥20%) who underwent triple modality screening (DM, US, and MRI), specificities and positive predictive values (PPVs) of MRI were significantly lower in the non–BRCA mutation group compared with the BRCA mutation group (specificity, 87% vs 94%, P <.001; PPV, 15% vs 33%, P = .018) (34). This observation was supported by two more recent retrospective studies that compared screening MRI performance among various groups of women at elevated risk, finding that MRI performance was poorest among women who reported FH as their only risk factor (43,44). Both studies included women with FH (greater and less than 20% lifetime risk), BRCA1/2 gene mutations, personal history of breast cancer, chest wall radiation, or high-risk lesions. MRI positive predictive values (PPV1, 10%; PPV3, 12%–14%) and CDRs (7.6–8/1000) were lowest in the FH groups in both studies, with PPVs below benchmark. The sensitivity of MRI in the FH groups was 73% to 77%. While the retrospective nature of these two studies limits verification of reported FH details, better risk prediction with personalized supplemental screening strategies may be helpful for women with FH as their only risk factor.

Few MRI screening studies to date have included women with FH and less than 20% lifetime risk, with limited subanalyses for women in the lower-risk category (37,40,44,45). In one retrospective study of 1019 women with an FH of breast or ovarian cancer, of which 1.5% were BRCA mutation carriers, screening MRI showed higher yield in women with stronger FH (more than two first-degree relatives and >25% lifetime risk) compared with those with lower risk FH (one first-degree relative and <20% lifetime risk), with PPVs of 13% and 6%, respectively (45). Women with lower-risk FH who have dense breasts and desire supplemental screening should undergo MRI according to the updated ACR guidelines (20). Although the benefit of screening MRI was sustained across all breast densities in a study evaluating 2120 women with average breast cancer risk (incremental CDR, 6.9/1000), more data would be helpful to support MRI screening in women with lower-risk FH and nondense breasts (38).

Abbreviated MRI (AB-MRI) is a shortened MRI protocol typically consisting of at least one pre- and one postcontrast sequence with acquisition times under 10 minutes (46). When used for screening, AB-MRI has demonstrated performance equivalent to full protocol MRIs, with reduced acquisition and interpretation times making it more favorable for this application (47–49). In the ECOG-ACRIN 1141 prospective multisite randomized trial comparing screening AB-MRI and DBT in average-risk women with dense breasts, AB-MRI was associated with a significantly higher invasive CDR, again supporting the use of MRI in women with dense breasts (48). Currently, the main challenges of AB-MRI implementation are limited accessibility, workflow, and cost (50). For women with average-intermediate risk due to FH, AB-MRI is a promising screening option, especially for those with extremely dense breasts in whom DM sensitivity is lowest.

Contrast-Enhanced Mammography

Contrast-enhanced mammography (CEM) is a newer dual-energy mammographic technique that acquires low-energy and high-energy images after intravenous injection of iodinated contrast and generates recombined images to obtain morphologic and functional information. The low-energy images are equivalent to standard DM, and the recombined images depict tumoral enhancement similar to contrast-enhanced breast MRI. A recent CEM meta-analysis of 10 605 patients reported a pooled sensitivity and specificity of 95% and 81%, respectively; however, the 60 studies represented a mix of screening and diagnostic evaluations (51). Thus, current data on screening CEM are limited by heterogeneity of study design and populations, as well as small sample sizes.

A few single-institution prospective and retrospective studies have evaluated screening CEM in women with above-average risk due to a variety of factors, including FH with no genetic mutation, which comprised 22% to 56% of the study participants (Table 2). These studies compared CEM to DM using low energy images as a surrogate for DM,and/or compared CEM to MRI or US. Contrast-enhanced mammography consistently outperformed DM and DM plus US with high sensitivities and specificities reaching 83% to 100% and 76% to 95%, respectively, and CDRs of 13.1 to 15.5 per 1000, similar to MRI. The specificities of CEM and MRI were also found to be comparable (95% and 94%, respectively) in the first prospective pilot screening study of 307 women, of whom more than half had FH (52). While CEM performance in screening is promising, a recent study of 157 intermediate-to-high–risk women demonstrated a significantly lower true positive rate in the 16% of women with FH as risk factor compared with those with non-FH risk factors (0% vs 18%; P = .03) including personal history of breast cancer, high-risk lesion, or chest wall radiation, suggesting that FH may confer a lower risk than these other risk factors (53).

Supplemental screening US performed after CEM reduced specificity in two studies by detecting false positives and no additional cancers, indicating that US may not be valuable after a negative CEM exam, though further research is warranted (54,55). Diagnostic US performed after a positive CEM exam, however, is clinically useful because enhancing lesions with sonographic correlates are more likely to be malignant and can undergo US-guided biopsies (53,56).

For women who are at average-intermediate risk because of FH and also have dense breasts, annual screening CEM or US may be considered as an alternative to MRI, as per the updated ACR guidelines (20). The ACR-sponsored Contrast-Enhanced Mammographic Screening Trial is an ongoing multisite prospective randomized study that will compare screening DBT to CEM in average-intermediate risk women with dense breasts, which may validate the use of CEM as a screening tool for lower-risk women with FH.

Contrast-enhanced mammography may also be performed to screen women with a strong FH (lifetime risk ≥20%) who have contraindications to MRI, such as claustrophobia, metallic implants, or gadolinium-based contrast allergies (20) (Figure 2). Studies have shown that most high-risk patients prefer the experience of CEM over MRI, citing faster acquisition, less noise, greater comfort, and significantly lower rates of anxiety (57,58). Additionally, for high-risk women with a strong FH who are undergoing supplemental screening MRI, CEM may be performed in lieu of annual DM/DBT and staggered with MRI to provide vascular-based screening every six months.

US

Supplemental screening US is a valuable adjunct to mammography, with the largest benefit observed among women who have dense breasts. The ACRIN 6666 trial showed that of 2659 women with elevated breast cancer risk and dense breasts, supplemental screening US detected 32 additional mammographically occult cancers, yielding an incremental CDR of 3.7 per 1000 women, of which the majority were invasive (94%), small (median, 10 mm), and node negative (96%) (59). This benefit has been reproduced in more recent studies comparing screening US with DBT in women with dense breasts (60,61). In the prospective ASTOUND-2 trial that included 5300 women with dense breasts, screening US yielded a significantly higher incremental CDR than DBT (4.9 vs 2.8 cancers per 1000 women; P = .015) compared with standard DM (61). In a larger retrospective study of 13 968 women, screening US yielded an incremental CDR of 2.4 per 1000 women that were occult on DBT and significantly smaller, more invasive, and of a lower stage (invasive disease) than DBT-detected malignancies (60). Women with FH of breast or ovarian cancer accounted for nearly half (55/113, 49%) of patients diagnosed with cancer in this study overall. Nevertheless, all of these studies included women with dense breasts with or without additional risk factors, and subgroup analyses for women with FH were not performed.

For women undergoing high-risk screening MRI, supplemental screening US offers no additional benefit. Several prospective studies evaluating triple modality screening (DM, US, MRI) in high-risk women with FH and genetic mutations showed that US increases false positive findings and reduces specificity without improving sensitivity (Table 3). Of the 186 cancers diagnosed in 2827 women in these seven studies, no cancer was detected on US alone. In one subgroup analysis, screening US demonstrated significantly lower specificity and PPV in women without BRCA mutations compared with BRCA mutation carriers (specificity, 96% vs 99%, P = .02; PPV, 16% vs 60%, P = .01, respectively), suggesting that women without genetic susceptibility may not benefit as much from supplemental US (34). Additionally, a few CEM studies have found that the use of screening US after negative CEM increases the false positive rate without improving cancer detection and may therefore be unnecessary, although further research with larger patient populations is warranted (54,55).

Thus, women with FH may benefit from supplemental screening US if they are at high risk and unable to undergo MRI or CEM, or if they have dense breasts and prefer US instead of contrast-based imaging (Figure 3).

Conclusion

An FH of breast cancer without known genetic susceptibility confers an increased risk that varies depending on the pattern of family members affected and may justify earlier and/or more frequent screening. Current evidence supports annual DM/DBT and MRI for high-risk women with strong FH. Women with lower-risk FH, and particularly those with dense breasts, should consider supplemental screening with MRI, CEM, or US, noting that US has limited benefit after MRI or CEM. Interestingly, a few studies reported inferior performance of screening MRI and CEM in women with FH as sole risk factor compared with women with gene mutations and non-FH risk factors, suggesting that women with only FH may benefit less from these supplemental screening modalities. Further prospective research evaluating screening outcomes in women with FH as their only breast cancer risk factor, particularly those with average-intermediate risk or nondense breasts, is warranted to optimize screening strategies and inform recommendations for these women.

Funding

None declared.

Conflict of Interest Statement

V. M. has research support from Pfizer unrelated to this study.

References