Individualized model for predicting pathological complete response to neoadjuvant chemotherapy in patients with breast cancer: A multicenter study

Abstract

BackgroundPathological complete response (pCR) is considered a surrogate for favorable survival in breast cancer (BC) patients treated with neoadjuvant chemotherapy (NACT), which is the goal of NACT. This study aimed to develop and validate a nomogram for predicting the pCR probability of BC patients after NACT based on the clinicopathological features.MethodsA retrospective analysis of 527 BC patients treated with NACT between January 2018 and December 2021 from two institutions was conducted. Univariate and multivariate logistic regression analyses were performed to select the most useful predictors from the training cohort (n = 225), and then a nomogram model was developed. The performance of the nomogram was evaluated with respect to its discrimination, calibration, and clinical usefulness. Internal validation and external validation were performed in an independent validation cohort of 96 and 205 consecutive BC patients, respectively.ResultsAmong the 18 clinicopathological features, five variables were selected to develop the prediction model, including age, American Joint Committee on Cancer (AJCC) T stage, Ki67 index before NACT, human epidermal growth factor receptor 2 (HER2), and hormone receptor (HR) status. The model showed good discrimination with an area under the receiver operating characteristic curve (AUC) of 0.825 (95% CI, 0.772 to 0.878) in the training cohort, and 0.755 (95% CI, 0.658 to 0.851) and 0.79 (95% CI, 0.724 to 0.856) in the internal and external validation cohorts, respectively. The calibration curve presented good agreement between prediction by nomogram and actual observation, and decision curve analysis (DCA) indicated that the nomogram had good net benefits in clinical scenarios.ConclusionThis study constructed a validated nomogram based on age, AJCC T stage, Ki67 index before NACT, HER2, and HR status, which could be non-invasively applied to personalize the prediction of pCR in BC patients treated with NACT.

A nomogram based on platelet-to-lymphocyte ratio for predicting pathological complete response of breast cancer after neoadjuvant chemotherapy

“…The nomogram is a simple and effective tool for predicting the outcome [20]. The nomogram prediction model established based on the results of regression analysis can predict the pathological response of tumor patients after NAC [21][22][23][24], and the prediction ability of these models is pretty good. Multivariate regression analysis in our study showed that PLR, PLT, WBC and tumor grade were independent predictive factors of pCR.…”

Machine learning prediction of pathological complete response and overall survival of breast cancer patients in the Montefiore Health System in the Bronx

“…This is in keeping with several other studies. Qian et al found lower T scores and smaller tumor size correlated with higher pCR rates [51]. Goorts et al reported lower T stages had signi cantly higher pCR [53].…”

“…Tumor biology consistently emerges as a factor linked to pCR in multiple studies [50,51]. Previously reported racial disparities in survival may be due to additional factors which are potentially inter-related and would therefore be di cult to isolate from one another, such as socioeconomic differences, differences in insurance, and differences in treatment.…”

Development and Assessment of a Novel Core Biopsy-Based Prediction Model for Pathological Complete Response to Neoadjuvant Chemotherapy in Women with Breast Cancer

“…In addition, based on the imaging, Li et al [ 56 ] built an imaging model with diffusion-weighted MRI in neoadjuvant immunotherapy, with an AUC of 0.73 for predicting pCR. One recent study developed a prediction model with an AUC of 0.825 based on age, AJCC T stage, Ki67, HER2, and hormone receptor status [ 57 ]. However, this study had a relatively small sample size (n = 527).…”

“…In contrast to prior studies [ 24 , 55 , 57 ], the ER, PR, Ki67, and p53 thresholds were established based on ROC analysis rather than simply categorizing them as positive or negative. In order to facilitate clinical application, we created a user-friendly nomogram based on our Cox model.…”