Local Relapse After Breast-Conserving Therapy for Ductal Carcinoma In Situ

• DOI: 10.1097/ppo.0000000000000025
• Publication Date: 2014-01-20T11:07:03Z
• Authors: Caroline Sweldens; Stephanie Peeters; Erik van Limbergen; Hilde Janssen; Annouschka Laenen; Sujata Patil; Kimberly J. Van Zee; Caroline Weltens
• Keywords: Nomogram; Concordance; Breast-conserving surgery

Purpose Adjuvant treatments after breast-conserving surgery (BCS) for ductal carcinoma in situ to prevent local relapse are considered standard of care. However, patient selection increased morbidity without proven survival benefit remains a challenge. To predict the risk ipsilateral breast tumor (IBTR) BCS, Memorial Sloan-Kettering Cancer Center (MSKCC) developed nomogram. The aim this study was develop our own prediction model IBTR and provide an external validation MSKCC Methods From 1973 2010, 467 patients were treated with BCS at University Hospital Leuven. Clinicopathologic treatment parameters all used create multivariable model. predictive value evaluated using concordance index (C-index) probability estimate (CPE). Multiple imputation account missing data allow be tested on patients. Results Median follow-up 7.2 years, 48 women who IBTR. Omission adjuvant endocrine therapy, younger age, positive or close surgical margins significantly associated bootstrap-corrected C-index 10-year by 0.63 CPE 0.61. probabilities predicted nomogram 0.66 0.61, respectively. Conclusions Despite small number events, need multiple imputation, few radiation, performance somewhat better than This shows that is externally valid. allows users integrate information from 10 different variables more precise stratification use conventional single hazard ratios.