Nomogram for predicting outcomes in elderly women with mucinous breast cancer: A retrospective study combined with external validation in southwest China

Abstract

Objective

Mucinous breast cancer (MBC) is a kind of breast cancer (BC), which is rare in clinic, mainly for women, because of the low incidence rate, so there is no unified standard treatment protocol. Elderly patients have a poor prognosis due to their combined comorbidities. This study aims to investigate the effect of surgery and chemoradiotherapy on the prognosis of elderly female MBC patients and construct nomograms for predicting the OS and CSS in elderly female MBC patients.

Methods

Data for female MBC patients over 65 years are obtained from the Surveillance, Epidemiology and End Results (SEER) database, patients were divided into two groups: the training set and the validation set. External validation data of the prediction model were provided by Kunming Hospital of Traditional Chinese Medicine. We used Cox regression modeling, which was used to identify independent risk factors affecting patient prognosis. After avoiding confounding bias according to the multifactorial Cox regression model, we used these screened statistically significant results to construct column-line plots. The performance of the model was tested using the consistency index (c-index), the calibration curve, and the area under the operating characteristic curve of the receiver (AUC). Subsequently, we used decision curve analysis (DCA) to examine the potential clinical value of our nomograms.

Results

A total of 8103 elderly MBC female patients were extracted from the database SEER and were assigned to the training and validation set, randomly. A total of 83 patients from Kunming Hospital of Traditional Chinese Medicine were used in the external verification set. After multifactorial Cox regression analysis, we found that age, race, T-stage, M-stage, surgical approach, radiotherapy, and tumor size were independent risk factors for OS in elderly MBC patients. Similarly, independent risk factors of CSS included age, marital status, N stage, M stage, surgical approach, chemotherapy, and tumor size. The C-index for the OS training, validation, and external verification set were 0.731 (95%CI 0.715–0.747), 0.738 (95%CI 0.724–0.752), and 0.809 (95%CI 0.731–0.8874). The C-index of the training set, the validation set, and external verification set for CSS were 0.786 (95%CI 0.747–0.825), 0.776 (95%CI 0.737–0.815), and 0.84 (95%CI0.754–0.926), respectively. The AUC, calibration curves and DCA also showed good accuracy.

Conclusions

In this study, we construct a new nomogram to predict the prognosis of elderly patients with MBC. The nomograms have undergone internal and external validation and have been confirmed to have good clinical applicability. At the same time, we found that for elderly female MBC patients, surgery and radiotherapy significantly benefit their survival, but chemotherapy is not conducive to patient survival.