Nomogram for predicting invasion in patients with a preoperative diagnosis of ductal carcinoma in situ of the breast

Abstract

Abstract Background The aim of this study was to identify risk factors for invasive breast cancer in patients diagnosed with ductal carcinoma in situ (DCIS) on a preoperative biopsy. These factors were used to develop a nomogram for predicting the risk of invasion in the preoperative setting. Methods This was a retrospective analysis of patients who underwent surgical treatment for DCIS diagnosed before surgery between 1997 and 2009. Multivariable analysis was used to identify clinical, radiological and histopathological factors that may predict upstaging. A nomogram was developed to predict the probability of invasion using multiple logistic regression analysis. This nomogram was subsequently validated using another cohort of patients with a preoperative diagnosis of DCIS between 2010 and 2012. Results Upstaging to invasive cancer occurred in 123 (24.9 per cent) of 493 women treated between 1997 and 2009. A larger DCIS lesion (at least 15 mm), lack of hormone receptor expression, intermediate or high nuclear grade, diagnosis on core biopsy compared with vacuum-assisted biopsy, and non-cribriform subtype of DCIS were significantly associated with upstaging. A nomogram developed using these factors demonstrated good predictive performance (area under the receiver operating characteristic (ROC) curve (AUC) 0·823, 95 per cent confidence interval 0·787 to 0·860). The nomogram showed similar predictive performance in the validation data set, based on another 149 women (AUC 0·700, 0·613 to 0·786). Conclusion Upstaging to invasive cancer in women with a preoperative diagnosis of DCIS is common. A nomogram based on the five most significant factors related to upstaging accurately predicted invasive cancer. This nomogram may be useful when deciding whether to pursue axillary staging with sentinel lymph node biopsy in patients with DCIS.