A Nomogram to Predict Noninflammatory Skin Involvement of Invasive Breast Cancer

Abstract

Purpose: The aim of this study was to develop and assess a nomogram to predict noninflammatory skin involvement of invasive breast cancer.

Methods: We developed a prediction model based on SEER database, a training dataset of 89202 patients from January 2010 to December 2016. Multivariable logistic regression analysis was applied to build a predicting model incorporating the feature selected in the least absolute shrinkage and selection operator regression model. Discrimination, calibration, and clinical usefulness of the predicting model were assessed using the C-index, calibration plot, and decision curve analysis. Internal validation was assessed using the bootstrapping validation.

Results: Predictors contained in the prediction nomogram included use of age, race, grade, tumor size, stage-N, ER status, PR status, and Her-2 status. The model shows good discrimination with a C-index of 0.857 (95% confidence interval: 0.807-0.907) and good calibration. High C-index value of 0.847 could still be reached in the internal validation.

Conclusion: This study constructed a novel nomogram with accuracy to help clinicians access the risk of noninflammatory skin involvement by tumor. The assessment of clinicopathologic factors can predict the individual probability of skin involvement and provide assistance to the clinical decision-making.