The added advantage of automated breast ultrasound to mammographically detected different breast lesions in patients with dense breasts

Breast cancer is the most commonly diagnosed malignancy in women worldwide. Women with dense breast tend to have 15–25% lifetime risk of breast cancer due to decrease of mammographic sensitivity. Automated breast ultrasound (ABUS) is a new promising tool for detection of breast lesions masked by dense glandular tissue at mammography. The sensitivity of digital mammography in detecting breast lesions was 60.7%, specificity 91.6%, PPV 85%, NPV 75%, and accuracy 78%. The sensitivity of ABUS in detecting breast lesions was 92.86%, specificity 77.78%, PPV 76.47%, NPV 93.33%, and accuracy 84.38%. The sensitivity of handheld ultrasound (HHUS) in detecting breast lesions was 89.29%, specificity 88.89%, PPV 86.21%, NPV 91.43%, and accuracy 89.06%. The sensitivity of ABUS in detecting breast lesions was much higher than mammography in dense breast while the digital mammography (DM) had higher specificity. So, implementation of both DM and ABUS to get benefit of DM specificity as well as ABUS sensitivity were highly recommended.

of breast lesions. However, conventional handheld US (HHUS) has several limitations such as operator dependence and the requirement of a considerable amount of radiologist time for whole-breast US [10].
ABUS has several advantages over HHUS, such as higher reproducibility, less operator dependence, and less required physician time for image acquisition [11].
Recent studies have reported that ABUS is promising in US screening for women with dense breasts and can potentially replace handheld second-look US in a preoperative setting [10].

Aim of work
The aim of this study was to assess the ability of ABUS to detect mammographically occult breast lesions at dense breasts, assessing the diagnostic parameters of ABUS compared to digital mammography as well as HHUS in detection of breast lesions in dense breast. The secondary outcome was to prove the effectiveness of using ABUS as a screening tool in dense breasts in BIR-ADS 0 mammography results.

Methods
Prospective study was conducted on 59 patients presented with either palpable breast mass or as a part of early screening starting from January 2017 till July 2018. Their ages ranged from 24 to 81 years (mean age, 41 ± 10 SD years). Three cases were represented with bilateral lesions and two cases had two lesions in the same breast.

Inclusion criteria
Dense breast (ACR C or ACR D) on digital mammography.

Exclusion criteria
Breasts with American College of Radiology (ACR) A (predominantly fatty breast) or ACR B (scattered glandular tissue) detected with digital mammography were excluded.

Limitations of the ABUS in the study
Exclusion of axillary regions from the field of view. The absence of tools to assess vascularity and tissue elasticity. Artifacts that occur during data acquisition remain an issue that can cause false positive results or can obscure actual findings. The area where the most significant artifact usually occurs is in the subareolar region.

Methods
All of the cases (n = 59) were subjected to both digital mammography and automated breast ultrasound, as well as routine handheld ultrasound. They were asked to expose the upper part of the body. No other special preparations were needed.

a) Digital mammography examination protocol design
A craniocaudal (CC) and a medio-lateral oblique (MLO) views were obtained with the patient in a standing position. Breast compression was applied. Images were acquired with a mammography system Senographe Essential, GE Healthcare fullfield digital mammography machine. Senographe Essential has dual anode (rhodium molybdenum) with CsI digital detector.

b) Automated breast examination protocol design
All participants underwent ABUS examination. All ABUS exams were done with an ABUS system (Invenia TM ABUS, Automated Breast Ultrasound System, GE Healthcare, Sunnyvale, CA, USA) with high frequency probe. The transducer length is 15.3 cm, with 6-15 MHz frequency. The gray scale levels were 256 with frame rate 10 frames/second. The examination was performed in the supine position.
A cushion placed under the shoulder that helped to spread out the breast tissue evenly, with the nipple pointing to the ceiling. A hypoallergenic lotion placed evenly on the breast with an additional amount on the area of the nipple.
A disposal membrane was used to aid an acoustic coupling and one of the three levels of compression was applied to spread out the breast evenly with respect to image quality and patient comfort. The ABUS scan was continuous and automated. During the acquisition, women were asked not to move and to breathe smoothly.
Volume acquisitions were obtained in the axial plane starting from the inferior part of the breast with coronal and sagittal reconstruction.
Image data automatically acquired a 15.4 cm × 17.0 cm volume from the skin to the chest wall up to 5 cm deep with 0.2 mm thickness of each slice. For each breast, three volumes were obtained: the central (anteroposterior) volume with the nipple in the center of the footprint, the lateral volume that included the upper outer part of the breast tissue with the nipple located in the inferior-medial corner, and the medial volume that included the inner and inferior part of the breast tissue. A nipple marker was placed in every examination for the accurate co-ordinance. For optimal image quality, a selection between three breast sizes was made. In women with larger breasts additional views were taken to avoid tissue exclusion.

c) Handheld ultrasound images (HHUS)
The gel was applied to the breasts and ultrasound examination was done using radial and antiradial techniques with axilla, with probe frequency 18-5 MHz.

Image analysis
The digital mammography and automated ultrasound data were evaluated by two experienced radiologists in consensus; both observers were unaware of the pathological data of each patient.

Digital mammography images
Assessment of breast composition, mass characterization (shape, margin density), asymmetry, calcification, mass number, location, axillary lymphadenopathy, extension, skin thickening, retraction and architectural distortion, and BIRADS classification was done.

Automated ultrasound images and handheld ultrasound images
Assessment of mass characterization (shape, margin orientation, echopattern, posterior feature, calcification), mass number, location, axillary lymphadenopathy, skin thickening, retraction, and BIRADS classification were done. Additionally, for ABUS, we assessed lesions' character in coronal view.
All breast masses included in this study were interpreted as above described and then the accuracy in reaching the final diagnosis was calculated for digital mammography and automated ultrasound as well as HHUS.
Pathological results were used as the gold standard of reference for the 64 lesions apart from 13 lesions which were proven by HHUS criteria to be benign (9 diagnosed as fibroadenomas (follow up) and 4 as simple cysts).
Samples were obtained with fine needle aspiration cytology (FNAC), cytology of nipple discharge, core biopsy, surgical excision, and/or radical surgery. Analysis of the samples was performed in the pathology department by a group of well-trained expert pathologists.

Statistical analysis
Data were coded and entered using the statistical package SPSS (Statistical Package for the Social Sciences) version 25. Data were summarized using mean, standard deviation, median, minimum and maximum in quantitative data, and using frequency (count) and relative frequency (percentage) for categorical data.
Comparisons between quantitative variables were done using the non-parametric Mann-Whitney test. For comparing categorical data, chi square (χ2) test was performed. Exact test was used instead when the expected frequency is less than 5. Correlations between   quantitative variables were done using Spearman correlation coefficient. Standard diagnostic indices including sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic efficacy were calculated. Testing for agreement between different methods in numerical data was done using the intra class coefficient (ICC) with 95% confidence interval (95% CI). P value less than 0.05 was considered as statistically significant.

Patient's demographics
This prospective study included a total of 59 patients with 64 lesions presented with breast masses (detected by clinical examination or by mammography examination). Their ages ranged from 24 to 81 years (mean age, 41 ± 10 SD years). Thirty-six lesions (56.2%) were diagnosed as benign while 28 (43.8%) lesions were diagnosed as malignant.
Pathological results were used as the gold standard of reference apart from 13 lesions which were proven by HHUS criteria to be benign, 9 diagnosed as fibroadenomas (follow up) and 4 as simple cysts.
Each examination (digital mammography, ABUS, and HHUS) was evaluated regarding the following criteria according to the 5th edition of BIRADS lexicon.
As regards breast density, 44 (68.8%) of breasts examined were ACR C, while 20 (31.2%) were ACR D. No significant correlation in our study between glandular tissue composition breast and malignancy (P value 0.221).
As regards the BIRADS evaluation of different lesions by mammography, 27 lesions were considered BIRADS 0 for further evaluation by other imaging tools, 1 lesion was considered BIRADS II, 19 lesions were considered BIRADS III, 16 lesions were considered BIRADSIV, and 1 lesion was considered BIRADS V (Table 1).

II. HHUS
As regards lesion detectability, it was higher than mammography, it could detect 15 lesions missed by mammography; however, it was lower than ABUS. ABUS could detect duct papilloma interpreted by HHUS   as dilated ducts, and a second lesion was retro areolar lesion, interpreted by HHUS as dilated ducts with increased internal vascularity (Table 2).

III. ABUS
As regards lesion detectability, it was the highest by ABUS; it could detect 17 lesions missed by mammography, and 2 lesions missed by HHUS.

Discussion
Breast cancer is the most common malignancy in women from developed and developing countries. Detection and treatment of breast cancer in its earliest possible stage are the ultimate goal. Thus, the role of radiologists in imaging the breast is vital. At present, Xray mammography is the "gold standard" for screening and early detection of breast cancer [8]. Women with dense breast tissue have a high risk of developing breast cancer in a ratio of 15-25% [9]. ABUS has a promising role in patients with dense breasts in detecting the hidden lesions, as it is a non-operator dependent and it needs less time of interpretation by a radiologist, helping to improve the workflow [10].
Wilczek et al. [12] stated in a study on 1668 asymptomatic women, age 40-74 years, with heterogeneously dense parenchyma (ACR C) or extremely dense breast (ACR D) that the increase in sensitivity of screening for full field digital mammography and 3D ABUS versus FFDM alone was 36.4%. The difference in specificity was − 0.7%.
Giger et al. [13] reported in a study done on 185 asymptomatic women with BI-RADS C or D breast density that the sensitivity was 57.5% for FFDM alone and 74.1% for FFDM with ABUS, yielding a statistically significant increase in sensitivity (P < 0.001) (relative increase = 29%). Overall specificity was 78.1% for FFDM alone and 76.1% for FFDM with ABUS (P = 0.496).
To summarize, this study showed the same results compared to the above four studies that ABUS showed an average of 30% increase in sensitivity in detecting breast malignancy in dense breast compared to digital mammography. As regards specificity, mammography had higher specificity than ABUS in all fore mentioned studies except Wilczek et al. [12], who showed near results of specificity between DM and ABUS but still higher specificity for DM.
Choi et al. [16] evaluated a large population of asymptomatic women who were subdivided into two groups (1866 patients for ABUS and 3700 patients for HHUS) and showed that diagnostic accuracy and specificity were significantly higher for ABUS than HHUS (respectively, diagnostic accuracy 97.7 vs. 96.5% and specificity 97.8 vs. 96.7).
Vourtsis et al. [17] performed a study that included women with breast density category C or D (aged 48.6 ± 10.8 years) were recruited. All participants underwent ABUS and HHUS examination; a subcohort of 1665 women also underwent a mammography. The overall agreement between HHUS and ABUS was 99.8%; kappa = 0.994, P < 0.0001. In this study, the overall agreement between HHUS and ABUS was kappa = 0.694, P < 0.0001, which is lower compared to the above study.
Chen at al [15]. stated that there were significant differences between the malignant and benign masses with respect to retraction phenomenon and hyperechoic rim in the coronal plane of the ABUS. For retraction phenomenon, both the specificity and positive predictive value of a malignant diagnosis reached 100%, and the accuracy and false-positive rate were 96.8% and 0, respectively; for the hyperechoic rim, the specificity, negative predictive value, and accuracy of a benign diagnosis were 92.8%, 95.3%, and 95.9%, respectively.
These results are going with this study that retraction phenomenon has a significant relation with malignant pathology (P value < 0.001) with 100% specficity and 75% sensitivity, while complete hyperechoic rim has significant relation with benign pathology with (P value < 0.001) with 90.5% specificity and 52.8% sensitivity.
Rella et al. [18] stated that the coronal plane also improves the evaluation of lesion margins; benign tumors are often surrounded by a continuous hyperechoic rim, while breast cancers can present a discontinuous hyperechoic rim. In this study, 14 cases showed incomplete (discontinuous) hyperechoic rim, 9 of them were benign (64%) while 5 (36%) were malignant.
Finally, Skane et al. [19] proved that combined mammography and ABUS reading by the same radiologist improved diagnostic performance and resulted in higher observer agreement. Consequently, combined reading mode should be "standard" if ABUS was implemented in screening for women with dense breasts. Prospective studies were necessary before the implementation of ABUS (See figure on previous page.) Fig. 1 A 53-year-old patient presenting with left breast lump. a, b Mammography revealed ACR D breast with UOQ irregular high densityspeculated mass lesion. Mammography, MLO, and CC views of ACRD revealed suspicious left breast lesion (red arrow); however, proper assessment of the lesion could not be done for further imaging (BIRADS IV). c HHUS images. HHUS show irregular speculated hypoechoic mass lesion (BIRADS IV) could be recommended in population-based screening. This study also went finally with same recommendation as mammography still could detect DCIS before IDC development as per our knowledge; further research is also recommended for this point. The potential role of ABUS in the follow-up of benign lesions was supported by its considerable reliability in the recording of lesion location, distance from the nipple, and lesion size, these features suggested potential use in the follow-up of benign lesions as per Chang et al. [20].

Conclusion
The sensitivity of ABUS in detecting breast lesions is much higher than mammography in dense breast, while the DM has higher specificity. We recommend the implementation of both DM and ABUS to get benefit of DM specificity as well as ABUS sensitivity.