Strain ratio elastography for predicting malignant breast lesion: a new grading classification

Roberto A. Sandoval-Villanueva, Centro de Imagenologia Integral IMAX Tampico; Instituto de Estudios Superiores de Tamaulipas, Universidad Anahuac; Tampico, Tamaulipas, México;
Karla A. Lopez-Rodríguez, Centro de Imagenologia Integral IMAX Tampico; Instituto de Estudios Superiores de Tamaulipas, Universidad Anahuac; Tampico, Tamaulipas, México;
David F. Perez-Montemayor, Centro de Imagenologia Integral IMAX Tampico; Instituto de Estudios Superiores de Tamaulipas, Universidad Anahuac; Tampico, Tamaulipas, México;
Karla G. Garcia-Alvarez, Laboratorio de Patologia, Hospital Beneficiencia Española de Tampico, Tampico, Tamaulipas, Mexico;

Introduction: Strain ratio elastography (SRE) lacks a standard classification to establish a cutoff value for predicting malignant breast lesions. This study proposes a new three-grade classification based on SRE cutoff values using the LOGIQ E9 GE Healthcare system and evaluates its diagnostic performance in predicting breast lesion malignancy. Material and Methods: This retrospective cross-sectional study included women with Breast Imaging Reporting and Data System (BI-RADS) breast lesions categories 4 or 5. All lesions were evaluated with conventional ultrasound (US) and strain elastography (SE) with a GE Healthcare LOGIQ E9 system. The semiquantitative three-grade classification cutoff values defined arbitrarily based on SRE were grade 1 (0-1.9, soft pattern), grade 2 (2.0-3.9, intermediate pattern), and grade 3 (4.0-6, hard pattern). Histopathologic diagnosis was the gold standard. Receiver operating characteristic (ROC) analysis was used to compare SRE cutoffs ≥ 2 and ≥ 4. Results: Of 173 breast lesions, 73 (42.0%) were benign, and 100 (58.0%) were malignant. Benign lesions were primarily grade 1 (42.0%) and 2 (33.0%), while 71% of malignant lesions were grade 3 (p < 0.001). The median SRE cutoff value was higher in malignant (5.2) than in benign lesions (2.4) (p < 0.001). A cutoff ≥ 2 showed high sensitivity (93.0%) and low specificity (42.4%) for predicting malignant breast lesions, whereas a value ≥ 4 showed a sensitivity of 71.0%, a specificity of 75.3%, and an accuracy of 72.8%. Conclusion: The proposed SRE three-grade classification demonstrated that a grade 3 hard pattern (cutoff ≥ 4) as defined by the GE Healthcare LOGIQ E9 system achieved good diagnostic performance for predicting malignant BI-RADS 4 or 5 breast lesions; it is useful for characterizing breast lesions.

Keywords: Ultrasound elastography. Strain elastography. Breast cancer. Breast Imaging Reporting and Data System.