The addition of tomosynthesis to digital mammography has been shown to reduce recall rates and increase cancer-detection rates in the general population.21 Several studies have reported a 40% rate increase or more in invasive cancer detection when compared with digital mammography alone and a simultaneous 15% reduction in the rate of false-positive results.21-23 Two prospective European studies evaluated the efficacy of FFDM in combination with digital breast tomosynthesis for breast cancer screening.22,23 Skaane et al22 reported a 40% increase in the detection rate of invasive cancers with a simultaneous 15% rate reduction in false-positive results in 12,621 screening examinations when using FFDM in combination with digital breast tomosynthesis compared with FFDM alone. In an analysis of 7,292 screening examinations, Ciatto et al23 demonstrated an increased cancer-detection rate from 5.3 to 8.1 cancers per 1,000 women screened, with a simultaneous 17% reduction in recall rate. Two single-site observational studies performed by Rose et al24 and Haas et al25 also demonstrated statistically significant reductions in recall rates of 37% and 30%, respectively, although both groups demonstrated an increase in cancer detection, and neither reached statistical significance — possibly due to small sample sizes.
In a retrospective, multisite study evaluating the use of digital breast tomosynthesis in combination with FFDM among 173,663 patients, Friedewald et al21 demonstrated an increase in the rate of invasive cancer detection from 2.9 to 4.1 per 1,000 cases after adding digital breast tomosynthesis to FFDM. This was a relative increase of 41%.21 Although an increase was also observed in the rate of biopsy among patients screened with FFDM in combination with digital breast tomosynthesis (19.3 vs 18.1 per 1,000 cases for the digital mammography cohort), a concomitant 21% relative increase was seen in positive predictive values for biopsy, thus reflecting the higher yield of malignancy in women undergoing biopsy from the group undergoing FFDM in combination with digital breast tomosynthesis.21 The association with fewer unnecessary tests and biopsies, with a simultaneous increase in cancer-detection rates, would support the potential benefits of tomosynthesis as a tool for screening.21 This would, in theory, be particularly useful as an adjunct to FFDM screening in women with dense breasts by eliminating confounding, overlapping breast tissue, thereby helping to better detect masses and decrease recall rates.
In a study by Lee et al,26 simulation models of breast cancer were utilized. Compared with biennial FFDM alone, the researchers showed that combined biennial FFDM and digital breast tomosynthesis for US women aged 50 to 74 years with dense breasts would avert 1 additional breast cancer– related death per 2,000 women screened as well as 405 false-positive findings on screening examinations per 1,000 women.26 They concluded that adding tomosynthesis to biennial FFDM for women aged 50 to 74 years with dense breasts was likely to improve health outcomes at a reasonable cost.26
Thus, the evidence supports the finding that digital breast tomosynthesis improves detection rates of breast cancer with both increased sensitivity and specificity. However, its potential drawbacks include increased interpretation time and higher doses of radiation. Thus, further studies are needed to demonstrate whether screening with digital breast tomosynthesis in combination with FFDM results in a greater decrease in breast cancer–related mortality rates compared with FFDM alone.
The reduced sensitivity rate of mammography in women with dense breasts prompted the search for a widely available, reproducible, and cost-effective adjunct screening tool. Ultrasonography offers an affordable option for the detection of small masses without additional ionizing radiation or intravenous contrast. Robust support for ultrasonography as a supplementary imaging modality for the screening of dense breasts comes from data collected by Berg et al.27 Their multicenter, randomized trial included 2,809 women with dense breasts who were at increased risk for breast cancer. The authors compared the performance of 2-view mammography alone to 2-view mammography combined with hand-held survey ultrasonography performed by a radiologist.27 Mammography in combination with ultrasonography yielded 11.8 breast cancers per 1,000 women compared with 7.6 breast cancers per 1,000 women screened with mammography alone.27 In a substudy of the data from Berg et al,27 3 annual screening examinations or incident rounds were analyzed from 2,659 eligible women.28 The data revealed an increase in the rate of cancer detection each year when supplemental ultrasonography was utilized. A total of 32 additional cancers were detected by ultrasonography alone, 30 of which were invasive cancers.28 The median size of the invasive tumors was 10 mm, and 96% of these had node-negative disease.28 Overall, the data for the second and third years revealed greater rates of sensitivity for cancer detection when supplemental screening ultrasonography was combined with mammography.28 Thus, data from these incremental studies suggest a possible benefit in continuing annual screening with supplemental ultrasonography in conjunction with mammography.
A retrospective study was performed by Hooley et al29 in Connecticut after implementation of the Connecticut Public Act 09-41, which requires that increased breast density be directly communicated to the patient. Hand-held, whole-breast ultrasonography was performed by technologists in 935 women with dense breasts in both the diagnostic and screening populations.29 An additional 3.2 cancers were detected per 1,000 women screened in the first year of implementation.29 Similarly, Weigert et al30 retrospectively evaluated the utility of ultrasonography for screening among women with dense breasts from 6 Connecticut practices by compiling 8,647 screening ultrasonography sessions during the first year of test availability following the legislation. The addition of hand-held ultrasonography performed by certified technologists resulted in an additional 3.25 cancers detected per 1,000 women with dense breast and normal findings on mammography.30 The data from these studies support that handheld ultrasonography — whether performed by a radiologist or trained technologist — could increase the sensitivity rate for detecting breast cancer in women with dense breasts.
A multicenter study by Tagliafico et al31 evaluated the cancer-detection rate of tomosynthesis and hand-held screening ultrasonography after a negative finding was obtained on 2D diagnostic mammography among 3,231 self-referred women with heterogeneously dense or extremely dense breasts. Adding tomosynthesis detected an additional 13 breast cancers (12 of which were also identified on ultrasonography, whereas 1 was seen on tomosynthesis alone). Hand-held screening ultrasonography performed by a dedicated breast imaging radiologist detected an additional 23 breast cancers.31 These cancers were detected with low false-positive recall and biopsy rates. The false-positive recall rates for tomosynthesis and ultrasonography were 1.7% and 2.0%, respectively; the false-positive biopsy rate was 0.7% for both groups.31 The lower false-positive rate could have contributed to operator experience (ultrasonography was performed by a dedicated breast imaging radiologist) as well as lack of recall for what was likely a benign finding in the clinical practice setting.31
Hesitation still exists in implementing routine, supplemental ultrasonography screening despite the data from the aforementioned studies. Using handheld 2D ultrasonography to detect small masses is labor intensive. Operator variability, shortages of trained personnel, and reductions in radiologist efficiency for image acquisition all contribute to the widespread discouragement for whole-breast surveys.27 To combat some of these challenges, 3D automated whole-breast ultrasonography has been introduced as an alternative modality. Three-dimensional automated breast ultrasonography received premarket approval by the US Food and Drug Administration (FDA) in September 2012.32 It is approved for use in combination with mammography for breast cancer screening in women who are asymptomatic who have normal or benign findings on mammography, have had no prior clinical breast intervention, and have dense breast parenchyma (Figure 3).32 Its automated algorithms give the clinician the ability to obtain reproducible imaging data from volumes of the breast within a short time interval.