Defining the Optimal Ki67 Cutoff Values for Survival Prediction in Neoadjuvant Chemotherapy-Treated Patients with Breast Cancer

Chun Jiang¹Tong Zhu²Yunfei Zong³Ruibin Liu⁴Jiang Du⁵Junze Dai⁴Yuxuan Song⁴Dingye Zhang⁴Xin Wang⁴Zhaohu Shi⁴Yinping Jiang⁶Jiawen Bu¹Baifang Ding⁷Xudong Zhu^6,8*

• ¹Shengjing Hospital of China Medical University, Shenyang, China
• ²Panjin Central Hospitl, Panjin, China
• ³People's Hospital of Liaoning Province, Shenyang, China
• ⁴Liaoning Cancer Hospital and Institute, Shenyang, China
• ⁵China Medical University, Shenyang, China
• ⁶University of Kentucky, Lexington, United States
• ⁷Panjin Central Hospital, Panjin, China
• ⁸Markey Cancer Center, Lexington, United States

Background: The rising incidence of breast cancer underscores the need for precise prognostic assessment following neoadjuvant chemotherapy (NAC). Ki67 is widely utilized for prognostic evaluation. However, its clinical applicability remains debated, particularly regarding the optimal cutoff threshold. This study aims to establish the optimal Ki67 cutoff value and evaluate its prognostic significance in predicting survival outcomes in patients with breast cancer undergoing NAC. Methods: A retrospective analysis was performed on 255 patients with breast cancer who received NAC between 2011 and 2024. The optimal Ki67 cutoff value was determined using maximally selected rank statistics. Kaplan–Meier survival analysis was used to evaluate the impact of Ki67 on disease-free survival (DFS) and overall survival (OS). Prognostic variables were selected via Cox regression analysis combined with LASSO dimensionality reduction. Based on these findings, nomogram models incorporating Ki67 and other clinical parameters were constructed to predict 1-year, 3-year, and 5-year DFS and OS, and the models were subsequently evaluated. Results: As a continuous variable, Ki67 presented an increasing and non-linear association with the risk of DFS. Using 20% as the threshold, survival analysis indicated that patients with a high Ki67 proliferation index (Ki67 > 20%) had significantly shortened DFS and OS compared to those with low Ki67 proliferation index. Cox regression analysis also confirmed that Ki67 was a common independent prognostic predictor for both DFS and OS. The nomogram model integrating Ki67, T stage, N stage, and other clinical parameters exhibited strong predictive performance, with the area under the curve (AUC) exceeding 0.900 at all-time points. Calibration plot further validated the model’s accuracy, with a C-index of 0.894 for DFS and 0.788 for OS. Conclusions: A Ki67 cutoff of 20% serves as a reliable predictor of DFS and OS in patients with breast cancer receiving NAC. The developed nomogram models, incorporating Ki67 and other clinical parameters, provide an accurate and clinically valuable tool for individualized prognostic assessment.