Shear-Wave Elastography for the Differential Diagnosis of Breast Papillary Lesions

Clinicopathological and Imaging Features of Breast Papillary Lesions and Their Association with Pathologic Nipple Discharge

No studies have evaluated whether any clinicopathological or imaging characteristics of breast papillary lesions are associated with pathological nipple discharge (PND). We analyzed 301 surgically confirmed papillary breast lesions diagnosed between January 2012 and June 2022. We evaluated clinical (age of patient, size of lesion, pathologic nipple discharge, palpability, personal/family history of breast cancer or papillary lesion, location, multiplicity, and bilaterality) and imaging characteristics (Breast Imaging Reporting and Data System (BI-RADS), sonographic, and mammographic findings) and compared malignant versus non-malignant lesions and papillary lesions with versus without PND. The malignant group was significantly older than the non-malignant group (p < 0.001). Those in the malignant group were more palpable and larger (p < 0.001). Family history of cancer and peripheral location in the malignant group were more frequent than in the non-malignant group (p = 0.022 and p < 0.001). The malignant group showed higher BI-RADS, irregular shape, complex cystic and solid echo pattern, posterior enhancement on ultrasound (US), fatty breasts, visibility, and mass type on mammography (p < 0.001, 0.003, 0.009, <0.001, <0.001, <0.001, and 0.01, respectively). On multivariate logistic regression analysis, peripheral location, palpability, and age of ≥50 years were factors significantly associated with malignancy (OR: 4.125, 3.556, and 3.390, respectively; p = 0.004, 0.034, and 0.011, respectively). Central location, intraductal nature, hyper/isoechoic pattern, and ductal change were more frequent in the PND group (p = 0.003, p < 0.001, p < 0.001, and p < 0.001, respectively). Ductal change was significantly associated with PND on multivariate analysis (OR, 5.083; p = 0.029). Our findings will help clinicians examine patients with PND and breast papillary lesions more effectively.

Monitoring Breast Cancer Response to Neoadjuvant Chemotherapy Using Ultrasound Strain Elastography

Strain elastography was used to monitor response to neoadjuvant chemotherapy (NAC) in 92 patients with biopsy-proven, locally advanced breast cancer. Strain elastography data were collected before, during, and after NAC. Relative changes in tumor strain ratio (SR) were calculated over time, and responder status was classified according to tumor size changes. Statistical analyses determined the significance of changes in SR over time and between response groups. Machine learning techniques, such as a naïve Bayes classifier, were used to evaluate the performance of the SR as a marker for Miller-Payne pathological endpoints. With pathological complete response (pCR) as an endpoint, a significant difference (P < .01) in the SR was observed between response groups as early as 2 weeks into NAC. Naïve Bayes classifiers predicted pCR with a sensitivity of 84%, specificity of 85%, and area under the curve of 81% at the preoperative scan. This study demonstrates that strain elastography may be predictive of NAC response in locally advanced breast cancer as early as 2 weeks into treatment, with high sensitivity and specificity, granting it the potential to be used for active monitoring of tumor response to chemotherapy.

Comparative Diagnostic Accuracy of Contrast-Enhanced Ultrasound and Shear Wave Elastography in Differentiating Benign and Malignant Lesions: A Network Meta-Analysis

Background: We performed a network meta-analysis to compare the diagnostic accuracy of contrast-enhanced ultrasound (CEUS) and shear wave elastography (SWE) in differentiating benign and malignant lesions in different body sites.

Methods: A computerized literature search of Medline, Embase, SCOPUS, and Web of Science was performed using relevant keywords. Following data extraction, we calculated sensitivity, specificity, positive likelihood ratio (LR), negative LR, and diagnostic odds ratio (DOR) for CEUS, and SWE compared to histopathology as a reference standard. Statistical analyses were conducted by MetaDiSc (version 1.4) and R software (version 3.4.3).

Results: One hundred and fourteen studies (15,926 patients) were pooled in the final analyses. Network meta-analysis showed that CEUS had significantly higher DOR than SWE (DOR = 27.14, 95%CI [2.30, 51.97]) in breast cancer detection. However, there were no significant differences between CEUS and SWE in hepatic (DOR = −6.67, 95%CI [−15.08, 1.74]) and thyroid cancer detection (DOR = 3.79, 95%CI [−3.10, 10.68]). Interestingly, ranking analysis showed that CEUS achieved higher DOR in detecting breast and thyroid cancer, while SWE achieved higher DOR in detecting hepatic cancer. The overall DOR for CEUS in detecting renal cancer was 53.44, 95%CI [29.89, 95.56] with an AUROC of 0.95, while the overall DOR for SWE in detecting prostate cancer was 25.35, 95%CI [7.15, 89.89] with an AUROC of 0.89.

Conclusion: Both diagnostic tests showed relatively high sensitivity and specificity in detecting malignant tumors in different organs. Network meta-analysis showed that CEUS had higher diagnostic accuracy than SWE in detecting breast and thyroid cancer, while SWE had higher accuracy in detecting hepatic cancer. However, the results were not statistically significant in hepatic and thyroid malignancies. Further head-to-head comparisons are needed to confirm the optimal imaging technique to differentiate each cancer type.

A Radiomics Approach With CNN for Shear-Wave Elastography Breast Tumor Classification

This paper proposes a segmentation-free radiomics method to classify malignant and benign breast tumors with shear-wave elastography (SWE) data. The method is targeted to integrate the advantage of both SWE in providing important elastic with morphology information and convolutional neural network (CNN) in automatic feature extraction and accurate classification. Compared to traditional methods, the proposed method is designed to directly extract features from the dataset without the prerequisite of segmentation and manual operation. This can keep the peri-tumor information, which is lost by segmentation-based methods. With the proposed model trained on 540 images (318 of malignant breast tumors and 222 of benign breast tumors, respectively), an accuracy of 95.8%, a sensitivity of 96.2%, and a specificity of 95.7% was obtained for the final test. The superior performances compared to the existing state-of-the-art methods and its automatic nature both demonstrate that the proposed method has a great potential to be applied to clinical computer-aided diagnosis of breast cancer.