A Web-based Prediction Model for Early Death in Patients With Metastatic Triple-negative Breast Cancer.

Abstract

Background: Triple-negative breast cancer (TNBC) is a subtype of breast cancer characterized by the absence of expression of estrogen receptor, progesterone receptor, or human epidermal growth factor receptor 2. This subtype of breast cancer is known for its high aggressiveness, high metastatic potential, tendency for recurrence, and poor prognosis. Patients with metastatic TNBC (mTNBC) have a poorer prognosis and a higher likelihood of early death (survival time ≤3 months). Therefore, the development of effective individualized survival prediction tools, such as prediction nomograms and web-based survival calculators, is of great importance for predicting the probability of early death in patients with metastatic TNBC.

Methods: Patients diagnosed with mTNBC in the Surveillance, Epidemiology, and End Results database between 2010 and 2015 were included in the model construction. Univariate and multivariate logistic regression analysis was performed to identify risk factors associated with early death in patients with mTNBC and predictive prognostic nomograms were constructed. The accuracy of the nomograms was verified using receiver operating characteristic curves, and GiViTi Calibration belt plots were used to evaluate the model consistency. The clinical applicability of the nomograms was evaluated using decision curve analysis. On the basis of the predictive prognostic nomograms, a network survival rate calculator was developed for individualized survival prediction in patients with mTNBC.

Results: A total of 2230 patients diagnosed with mTNBC were included in the Surveillance, Epidemiology, and End Results database for this study. After strict exclusion criteria, 1428 patients were found to be eligible for the study. All the patients were randomly divided into a training cohort and a validation cohort in a ratio of 7:3. Independent risk factors for mTNBC, including age, tumor size, brain metastasis, liver metastasis, surgery, and chemotherapy, were identified and integrated to construct the prediction nomogram and survival calculator. Results of receiver operating characteristic curves, calibration curves, and decision curve analysis curves from the training and validation cohort confirmed that the developed nomogram and web-based survival calculator in this study could accurately predict the probability of early death in patients with mTNBC.

Conclusions: In this study, we developed a reliable prediction nomogram and web-based survival calculator for predicting the probability of early death in patients with mTNBC. These tools can assist clinical physicians in identifying high-risk patients and developing personalized treatment plans as early as possible.