Validation of CTS5 Model in Large-scale Breast Cancer Population and Combination of CTS5 and Ki-67 Status to Develop a Novel Nomogram for Prognosis Prediction.

BACKGROUND: More than half of patients with early-stage estrogen receptor-positive (ER+) breast cancer relapse after completing 5 years of adjuvant endocrine therapy, so it is important to determine which patients are candidates for extended endocrine therapy. The clinical treatment score after 5 years (CTS5) is a prognostic tool developed based on postmenopausal ER+ breast cancer to assess the risk of late distant recurrence (LDR) after 5 years of adjuvant endocrine therapy for breast cancer. We aimed to externally validate the prognostic value of CTS5 in premenopausal and postmenopausal patients and combined with Ki-67 to develop a new model to improve the ability of prognosis prediction. METHODS: We included a total of 516 patients with early-stage ER+ breast cancer who had received 5 years of adjuvant endocrine therapy and were recurrence-free for 5 years after surgery. According to menopausal status, we divided the study population into 2 groups: premenopausal and postmenopausal women. The CTS5 of each patient was calculated using a previously published formula, and the patients were divided into low, intermediate, and high CTS5 risk groups according to their CTS5 values. Based on the results of the univariate analysis ( P <0.01), a multivariate COX proportional hazards regression analysis was conducted to establish a nomogram with significant variables ( P <0.05). The discriminative power and accuracy of the nomograms were assessed using the concordance index (C-index), calibration curve, and area under the time-dependent receiver operating characteristic curve. Discrimination and calibration were evaluated by bootstrapping 1000 times. Finally, we utilized decision curve analysis to assess the performance of our novel predictive model in comparison to the CTS5 scoring system with regard to their respective benefits and advantages. RESULTS: The median follow-up time was 7 years (6 to 9 years). The 516 women were categorized by CTS5 as follows: 246(47.7%) low risk, 179(34.7%) intermediate risk, and 91(17.6%) high risk. Using the CTS5 score as a continuous variable, patients' risk score was significantly positively associated with recurrence risk in both premenopausal and postmenopausal subgroups. For HER2- premenopausal patients and HER2+ postmenopausal patients, the CTS5 score was positively correlated with LDR risk. Patients with a Ki-67≥20% had a higher risk of LDR regardless of menopausal status. Using the CTS5 score as a categorical variable, the high-risk group of HER2- premenopausal patients had a higher risk of LDR. However, the CTS5 model could not distinguish the risk of LDR in different risk groups for HER2+ postmenopausal patients. In the high-risk group, patients with Ki-67≥20% had a higher risk of LDR, regardless of menopausal status. We developed a new nomogram model by combining the CTS5 model with Ki-67 levels. The C-indexes premenopausal and postmenopausal cohorts were 0.731 and 0.713, respectively. The nomogram model was well calibrated, and the time-dependent ROC curves indicated good specificity and sensitivity. Furthermore, decision curve analysis demonstrated that the new model had a wider and practical range of threshold probabilities, resulting in an increased net benefit compared with the CTS5 model. CONCLUSIONS: Our study demonstrated that the CTS5 model can effectively predict the risk of LDR in early-stage ER+ breast cancer patients in both premenopausal and postmenopausal patients. Extended endocrine therapy is recommended for patients with Ki-67≥20% in the CTS5 high-risk group, as well as premenopausal patients with HER2-. Compared with CTS5, the new nomogram model has better identification and calibration capabilities, and further research is required to validate its efficacy in large-scale, multicenter, and prospective studies.

Survival nomogram for patients with de novo metastatic breast cancer based on the SEER database and an external validation cohort.

Background: On average, 5-10% of patients are diagnosed with metastatic breast cancer (MBC) at the initial diagnosis. This study aimed to develop a nomogram to predict the overall survival (OS) in these patients. Methods: The nomogram was based on a retrospective study of 9435 patients with de novo MBC from the Surveillance, Epidemiology, and End Results (SEER) database. The predictive accuracy and discriminative ability of the nomogram were determined using the concordance index (C-index), area under the time-dependent receiver operating characteristic curve (AUC), and calibration curve. Decision curve analysis (DCA) was employed to evaluate the benefits and advantages of our new predicting model over the 8th edition of the American Joint Committee on Cancer (AJCC) Tumor Node Metastasis (TNM) staging system. The results were validated in a retrospective study of 103 patients with de novo MBC from January 2013 to June 2022 at an institution in northwest China. Results: Multivariate analysis of the primary cohort revealed that independent factors for survival were age at diagnosis, pathological type, histological grade, T stage, N stage, molecular subtype, bone metastasis, brain metastasis, liver metastasis, lung metastasis, surgery, chemotherapy, and radiotherapy. The nomogram achieved a C-index of 0.688 (95% confidence interval [CI], 0.682-0.694) in the training cohort and 0.875 (95% CI, 0.816-0.934) in the validation cohort. The AUC of the nomograms indicated good specificity and sensitivity in the training and validation cohorts, respectively. Calibration curves showed favorable consistency between the predicted and actual survival probabilities. Additionally, the DCA curve produced higher net gains than by the AJCC-TNM staging system. Finally, risk stratification can accurately identify groups of patients with de novo MBC at different risk levels. Conclusions: The nomogram showed favorable predictive and discriminative abilities for OS in patients with de novo MBC. Other populations from different countries or prospective studies are needed to further validate the nomogram.