Prognostic factors of patients with human epidermal growth factor receptor 2-positive breast cancer following neoadjuvant therapy: Development and validation of a predictive nomogram

Objective

Human epidermal growth factor receptor 2 (HER2)–positive breast cancer exhibits an aggressive phenotype and poor prognosis. The application of neoadjuvant therapy (NAT) in patients with breast cancer can significantly reduce the risks of disease recurrence and improve survival. By integrating different clinicopathological factors, nomograms are valuable tools for prognosis prediction. This study aimed to assess the prognostic value of clinicopathological factors in patients with HER2-positive breast cancer and construct a nomogram for outcome prediction.

Methods

We retrospectively analyzed the clinicopathological data from 374 patients with breast cancer admitted to the Fourth Hospital of Hebei Medical University between January 2009 and December 2017, who were diagnosed with invasive breast cancer through preoperative core needle biopsy pathology, underwent surgical resection after NAT, and were HER2-positive. Patients were randomly divided into a training and validation set at a ratio of 7:3. Univariate and multivariate survival analyses were performed using Kaplan-Meier and Cox proportional hazards regression models. Results of the multivariate analysis were used to create nomograms predicting 3-, 5-, and 8-year overall survival (OS) rates. Calibration curves were plotted to test concordance between the predicted and actual risks. Harrell C-index and time-dependent receiver operating characteristic (ROC) curves were used to evaluate the discriminability of the nomogram prediction model.

Results

All included patients were women, with a mean age of 50 ± 10.4 years (range: 26–72 years). In the training set, both univariate and multivariate analyses identified residual cancer burden (RCB) class, tumor-infiltrating lymphocytes(TILs), and clinical stage as independent prognostic factors for OS, and these factors were combined to construct a nomogram. The calibration curves demonstrated good concordance between the predicted and actual risks, and the C-index of the nomogram was 0.882 (95 % CI 0.863–0.901). The 3-, 5-, and 8-year areas under the ROC curve (AUCs) were 0.909, 0.893, and 0.918, respectively, indicating good accuracy of the nomogram. The calibration curves also demonstrated good concordance in the validation set, with a C-index of 0.850 (95 % CI 0.804–0.896) and 3-, 5-, and 8-year AUCs of 0.909, 0.815, and 0.834, respectively, which also indicated good accuracy.

Conclusion

The nomogram prediction model accurately predicted the prognostic status of post-NAT patients with breast cancer and was more accurate than clinical stage and RCB class. Therefore, it can serve as a reliable guide for selecting clinical treatment measures for breast cancer.