Can breast MRI accurately exclude malignancy in mammographic architectural distortion?

Eur Radiol. 2020 May;30(5):2751-2760. doi: 10.1007/s00330-019-06586-x. Epub 2020 Jan 30.

Abstract

Objectives: To investigate the diagnostic accuracy of problem-solving breast magnetic resonance imaging (MRI) in excluding malignancy in a cohort of patients diagnosed with mammographic architectural distortion (MAD).

Methods: The Institutional Review Board approved the study. Imaging database with 40,245 breast MRIs done between January 2008 and September 2018 was retrospectively reviewed. The study included all exams considered problem-solving MRI for MAD. Two radiologists reviewed the imaging data. Outcome was determined by the pathology results of biopsy/surgical excision or at least 1 year of clinical and radiological follow-up. Predictors for malignancy were examined, and appropriate statistical tests were applied.

Results: One hundred seventy-five patients (median age 53 years) fulfilled the inclusion criteria and formed the study cohort. No cancers were diagnosed in 106 patients with a negative MRI. Out of 69 women with positive MRI findings, 48 (70%) had benign outcome defined either by pathology result or by negative follow-up, and 21 (30%) yielded malignancy. Malignancy was significantly associated with positive MRI (p < 0.001) and older age (p = 0.014). Falsely positive MRIs were frequently found in women with radial scars. The sensitivity, specificity, negative predictive value, positive predictive value, and overall accuracy of breast MRI were 100% (95% CI 84 to 100%), 68% (CI 61 to 76%), 100% (CI 95 to 100%), 30% (CI 26 to 36%), and 73% (95% CI 66-79), respectively.

Conclusion: A negative breast MRI in patients with MAD was reliable in excluding malignancy in this cohort and may have a role as a precision medicine tool for avoiding unnecessary interventions.

Key points: • MRI shows a high negative predictive value in MAD cases. • MRI displays low accuracy in differentiating malignancy from RS. • MRI is a reliable non-invasive method to exclude malignancy in women with mammographic architectural distortion, potentially avoiding unnecessary biopsies and surgeries.

Keywords: Architectural distortion; Breast neoplasm; Mammography; Problem solving; Radial scar.

MeSH terms

  • Biopsy
  • Breast / diagnostic imaging*
  • Breast Neoplasms / diagnosis*
  • Female
  • Humans
  • Magnetic Resonance Imaging / methods*
  • Mammography / methods*
  • Middle Aged
  • Reproducibility of Results
  • Retrospective Studies
  • Unnecessary Procedures