Objectives This study aims to create a model that combines ultrasonography (US), magnetic resonance imaging (MRI) examination, and clinicopathological features to predict the axillary pathological complete response (pCR) of patients with breast cancer (BC) who receive neoadjuvant therapy (NAT). Methods This retrospective study included 600 patients with node-positive breast cancer who were eligible for enrollment (clinical stage cT1–4 and cN1-3) and received neoadjuvant therapy from January 2011 to January 2024. Before biopsy and neoadjuvant therapy, these patients underwent ultrasound (US) and MRI imaging of breast lesions and axillary lymph nodes (ALNs), and clinicopathological features were recorded before and after NAT. All imaging evaluations were independently performed by two experienced breast radiologists (with 10 years of experience), and discrepancies were resolved by consensus. Independent risk factors for predicting ALN status after NAT were identified by univariate and multivariate analyses. These independent risk factors were used for nomogram construction. Results Univariate logistic regression analysis revealed that the maximum diameter of the breast lesions on MRI after NAT (p 0.001), MRI ADC-value after NAT (p 0.001), maximum and minimum diameter of the ALN on US after NAT (p 0.001), the Ki67 level (p 0.001), tumor grade 3 (p = 0.017), primary ALN stage cN 2 (p = 0.022), efficacy evaluation of the neoadjuvant therapy, pT stage, MP classification, HR, HER2, and the presence of the Hilum of the lymph gland were significantly associated with ALN pCR after NAT (p 0.05). In the multivariate logistic regression analysis, ypT2 (p 0.001), ypT3 (p = 0.007), HER2 (p 0.001), response PR (p = 0.007), efficacy evaluation (SD/PD) (p = 0.010), and the presence of the Hilum of the lymph gland on US after NAT(p 0.001) were considered independent predictors of ALN pCR after NAT. The area under the curve (AUC) of the nomogram was 0.934(95% CI: 0.913-0.960) in the training set and 0.908 (95% CI: 0.867-0.950) in the validation set, with a sensitivity of 82.0% and a specificity of 89.1% in the training set. Conclusion Our noninvasive model based on US, MRI, and clinicopathological features can help accurately identify patients with ALN pCR after NAT and prevent unnecessary axillary lymph node dissection (ALND).
Cheng et al. (Fri,) studied this question.