Abstract P6-02-07: Is real-time rescanning by radiologists necessary for all ACR BI-RADS US category 3 to 5 lesions in the diagnostic setting?

Abstract

Abstract Purpose: To assess the value of radiologists' rescanning before final interpretation for American College of Radiology (ACR) Breast Imaging Reporting and Data System ultrasonography (US) category 3–5 lesions. Materials and Methods: Image data on 1070 patients with 1070 category 3–5 breast lesions with a pathological diagnosis scanned between January 2016–June 2017 were included. Both real-time and static images were acquired for each lesion. The diagnostic performance of assessing static images and rescanning by the radiologist were each calculated. The positive predictive values (PPVs) of each category in the two groups were calculated and compared. Results: The sensitivity, specificity, PPV, and negative predictive value for real-time US were 98.9%, 58.2%, 44.8% and 99.4%, and for static images were 98.9%, 57.1%, 44.1% and 99.3%, respectively. Areas under the curves were not significantly different if final assessment was only dichotomized as negative or positive (≥ category 3: P = .566). Significant differences were observed if more detailed classification was performed (category 3–5 without subcategory: 0.969 vs. 0.955, P = .0113; category 3–5 with subcategory 4A–4C: 0.915 vs. 0.855, P < .0001). All PPVs of each category for each assessment were within the reference range provided by the ACR in 2013 except subcategory 4B of static image evaluation, which was also significantly higher than that of real-time assessment (54.8% vs. 40.7%, P = .037),BI-RADS US of real-time and static images evaluation correlate with pathology are listed in Table 1. The overall agreement of the two approaches was moderate (κ = 0.43 if lesions were assigned to category/subcategory 3, 4A–4C, 5). Table1BI-RADS US category/subcategoryPPV provided by ACRReal-time US n (%)Static image US n (%)P3>0 &≤2%4674581.000Benign 464(99.4%)455(99.3%) Malignant 3(0.6%)3(0.7%) 4A>2% &≤10%2883010.257Benign 271(94.1%)276(91.7%) Malignant 17(5.9%)25(8.3%) 4B>10% &≤50%911350.037Benign 54(59.3%)61(45.2%) Malignant 37(40.7%)74(54.8%) 4C>50% &<95%50790.473Benign 5(10.0%)4(5.1%) Malignant 45(90.0%)75(94.9%) 5≥95%174971.000Benign 3(1.7%)1(1.0%) Malignant 171(98.3%)96(99.0%) BI-RADS US Breast Imaging Reporting and Data System for Ultrasonography, PPV positive predictive value, ACR American College of Radiology Conclusion: Both static image evaluation and real-time assessment had similar diagnostic performance for most lesions; however, the diagnostic performance of static image evaluation for subcategory 4B lesions was lower. Real-time rescanning by the interpreter is strongly recommended for lesions of subcategory 4B after assessing static images. YH and JM contributed equally to this study. Citation Format: Hu Y, Mei J, Jiang X, Liu Q, Gong C. Is real-time rescanning by radiologists necessary for all ACR BI-RADS US category 3 to 5 lesions in the diagnostic setting? [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P6-02-07.