Assessment of Nonmass Lesions Detected with Screening Breast US Based on Mammographic Findings

Academic Background

Breast cancer is one of the most common malignant tumors among women worldwide, and early detection and diagnosis are crucial for improving patient survival rates. Breast ultrasound (US), as a non-invasive and radiation-free imaging modality, plays a significant role in the screening and diagnosis of breast diseases. Particularly in women with dense breasts, breast US can effectively detect early breast cancers that are difficult to identify with mammography. However, a type of imaging finding known as “nonmass lesions” (NMLs) is frequently encountered during breast US examinations. NMLs typically appear as discrete areas of altered echotexture compared to surrounding breast tissue but lack the three-dimensional characteristics or distinct margins and shapes of masses. Although NMLs are common in breast US, distinguishing between benign and malignant NMLs remains challenging, especially in screening settings where the malignancy rate of NMLs is low, potentially leading to high false-positive rates and unnecessary biopsies.

This study aims to retrospectively analyze the imaging features of NMLs detected during breast US screening, particularly those associated with malignancy, and to evaluate the diagnostic performance of these features in the context of mammographic findings. The results are expected to provide clinicians with more accurate diagnostic criteria, reduce unnecessary biopsies, and optimize breast screening strategies.

Source of the Paper

This study was conducted by a team of radiology experts from multiple institutions, including Seoul National University Hospital, Seoul National University College of Medicine, Seoul National University Medical Research Center, University of Ulsan College of Medicine, University of California, Irvine, and Samsung Medical Center. The primary authors include Su Min Ha, Woo Jung Choi, Boo Kyung Han, Hak Hee Kim, Woo Kyung Moon, and others. The paper was published in Radiology in 2024, titled Assessment of Nonmass Lesions Detected with Screening Breast US Based on Mammographic Findings.

Research Process

Study Design

This multicenter retrospective study included asymptomatic women who underwent breast US screening at three tertiary hospitals in South Korea between January 2012 and December 2019. The participants were women with dense breasts or risk factors for breast cancer, and NMLs were detected during their US examinations. All patients underwent mammography within three months before or after the US examination and had a final pathological diagnosis. The study excluded lesions reclassified as masses or associated with masses during US, as well as patients lacking pathological diagnosis or with less than two years of follow-up.

Data Collection and Processing

The research team screened eligible cases from the radiology databases of the participating hospitals and retrospectively analyzed the US images. Three experienced radiologists independently evaluated the features of NMLs, including echogenicity, distribution, associated features (such as calcifications, posterior shadowing, and ductal abnormalities), and lesion size. Additionally, mammographic findings, including calcifications, asymmetry, and architectural distortion, were assessed.

Statistical Analysis

Univariable and multivariable logistic regression analyses were used to evaluate the association between imaging features of NMLs and malignancy. Diagnostic performance metrics, including sensitivity, specificity, and positive predictive value (PPV), were calculated for each US feature. Furthermore, inter-reader agreement for US features was assessed using the Fleiss κ statistic.

Main Results

Patient Characteristics

The study included 993 patients with a mean age of 50 years. Among them, 914 (92.0%) NMLs were benign, and 79 (8.0%) were malignant. In the 198 patients with positive mammographic findings, the malignancy rate was 28.8%, while in the 795 patients with negative mammographic findings, the malignancy rate was only 2.8%.

Imaging Feature Analysis

Malignant NMLs were significantly larger than benign NMLs (2.6 cm vs. 1.9 cm). In terms of echogenicity, the proportion of mixed echogenicity was significantly higher in malignant NMLs (33% vs. 14.3%). Additionally, malignant NMLs were more frequently observed in segmental distribution (62% vs. 18.1%) and were associated with calcifications (44% vs. 4.8%), posterior shadowing (24% vs. 7.4%), and ductal abnormalities (33% vs. 22.5%).

Diagnostic Performance

In multivariable analysis, associated calcifications (OR=21.6), posterior shadowing (OR=6.9), segmental distribution (OR=6.2), and mixed echogenicity (OR=5.0) were significantly associated with malignancy. The PPVs for these features were 44%, 22%, 22.9%, and 16.6%, respectively. In patients with positive mammographic findings, the PPVs of these US features were significantly higher than in those with negative mammographic findings.

Inter-reader Agreement

The assessment of inter-reader agreement for US features showed that the lower bound of the 95% confidence interval for the Fleiss κ statistic ranged from 0.63 to 0.81, indicating good to excellent agreement among readers.

Conclusion

The findings of this study suggest that associated calcifications, posterior shadowing, segmental distribution, and mixed echogenicity are suspicious features of malignant NMLs in breast US screening. These features can help clinicians more accurately identify malignancies, particularly in patients with positive mammographic findings. However, given the low malignancy rate of NMLs in patients with negative mammographic findings, clinicians should combine US and mammographic findings to make informed decisions about biopsy, thereby avoiding unnecessary invasive procedures.

Research Highlights

1. Key Findings: This study is the first to systematically evaluate the association between imaging features of NMLs detected during breast US screening and malignancy, identifying associated calcifications, posterior shadowing, segmental distribution, and mixed echogenicity as independent predictors of malignant NMLs.
2. Clinical Significance: The results provide clinicians with more accurate diagnostic criteria, helping to optimize breast screening strategies and reduce unnecessary biopsies.
3. Methodological Innovation: The study employed a multicenter retrospective design, combined with multivariable logistic regression analysis and inter-reader agreement assessment, ensuring the reliability and reproducibility of the findings.

Other Valuable Information

The study also noted that with the upcoming release of the 6th edition of the Breast Imaging Reporting and Data System (BI-RADS), NMLs may be included as a new category, which will further promote the clinical application of NMLs. Additionally, the research team recommended that in patients with negative mammographic findings, follow-up rather than immediate biopsy should be considered when US features such as nonsegmental distribution and isoechoic pattern are present, to reduce unnecessary invasive procedures.

This study provides important clinical evidence for distinguishing between benign and malignant NMLs in breast US screening, offering significant scientific value and clinical application prospects.