Abstract
BACKGROUND: The development of a robust and clinically applicable predictive model for pathological complete response (pCR) following neoadjuvant therapy (NAT) in human epidermal growth factor receptor 2 (HER2)-positive breast cancer (BC) is of critical importance. METHODS: In this retrospective study, 393 female patients with stage II-III BC who received NAT followed by surgery between May 2021 and December 2023 were included. Clinicopathological data, apparent diffusion coefficient (ADC) values from breast MRI, and pathological remission after NAT were collected. The change rate of ADC values after two cycles of NAT (ΔADC(0-2)%) was calculated. The efficacy of NAT regimens containing trastuzumab plus pertuzumab (HP) and trastuzumab plus pyrotinib (HPy) was compared. A nomogram predicting pCR was constructed, and its performance was evaluated. The model was internally validated using the bootstrap resampling method. RESULTS: The rate of total pathological complete response (tpCR) in the overall population was 68%. There was no statistically significant difference in tpCR between the HP and HPy groups (P > 0.05). Hormone receptor (HR) negativity, HER2 3+, high Ki-67 index, moderate-highly (M-H) infiltrated tumor-infiltrating lymphocytes (TILs), and ΔADC(0-2)% > 36.2% were independently associated with tpCR (P < 0.05). The nomogram integrating these variables exhibited good discrimination (AUC, 0.75) and calibration ability (P = 0.925), as well as valuable clinical applicability. CONCLUSION: Both HP and HPy combined with chemotherapy can be considered as optional NAT regimens for HER2-positive BC. The nomogram incorporating common clinical indicators provides a basis for clinicians to predict NAT efficacy at an earlier stage.