Assessment of Breast Tumor Size in Architectural Distortion: Comparing DM, DBT, US, CEM, and MRI

Overview

This study evaluated the accuracy of five imaging modalities—digital mammography (DM), digital breast tomosynthesis (DBT), ultrasound (US), contrast-enhanced mammography (CEM), and magnetic resonance imaging (MRI)—in measuring invasive breast tumor size in cases presenting as architectural distortion (AD). The correlation of imaging measurements with pathological tumor size was assessed, including the impact of including thin spicules in mammographic measurements and the influence of histological subtype.

Background

Architectural distortion (AD) is a subtle breast imaging finding characterized by disruption of normal breast architecture without a definite mass, representing about 6% of screening mammography abnormalities. AD can be caused by both benign and malignant processes, with malignancy more likely in diagnostic mammography settings. Accurate tumor size assessment is critical for staging and treatment planning; however, imaging modalities vary in their measurement accuracy, especially in AD cases where wispy spicules complicate size estimation. This study addresses the gap in comparative data on the accuracy of DM, DBT, US, CEM, and MRI in measuring tumor size in AD presentations.

Data Highlights

From January 2018 to December 2022, 63 AD lesions from 59 patients with biopsy-proven carcinoma were retrospectively analyzed. All lesions underwent DM and US, while subsets underwent DBT (29 lesions), MRI (34 lesions), and CEM (40 lesions). Mammographic measurements were performed twice: one including only the lesion core with thick spicules and another including thin, wispy spicules. Tumor sizes were measured to the nearest millimeter and compared to pathological invasive tumor size (pT-stage) as the reference standard.

Key Findings

• All 63 lesions were visible on DM; 60 lesions (95.2%) had an ultrasound correlate.
• DBT, MRI, and CEM were performed selectively, primarily in patients with dense breasts (BI-RADS C and D).
• Including thin spicules in mammographic measurements increased lesion size but the study aimed to determine if this improved correlation with pathological size.
• Imaging modalities varied in their correlation with pathological tumor size, with MRI and CEM expected to provide better accuracy due to contrast enhancement capabilities.
• The study also evaluated whether histological subtype influenced the accuracy of imaging size measurements.

Clinical Implications

Accurate preoperative tumor size estimation in AD cases is essential to guide surgical planning and avoid incomplete excision or overtreatment. Understanding which imaging modality best correlates with pathological size can optimize clinical staging and treatment decisions. Additionally, clarifying whether to include thin spicules in mammographic measurements may standardize reporting and improve size estimation accuracy.

Conclusion

This comparative analysis highlights the variability in tumor size assessment across imaging modalities in breast cancers presenting as architectural distortion. Incorporating advanced imaging techniques such as MRI and CEM may enhance size estimation accuracy, and careful consideration of mammographic spicule inclusion is warranted to improve correlation with pathological findings.

References

1. American College of Radiology BI-RADS -- Breast Imaging Reporting and Data System
2. Study on AD prevalence and malignancy rates -- Various Authors
3. TNM Staging and Tumor Size Assessment -- AJCC 8th Edition
4. Comparative Imaging Studies in Breast Cancer -- Multiple Sources