[External validation of nomogram for the prediction of recurrences of pregnancy-associated breast cancer]. / Cancer du sein associé à la grossesse : validation d'un nomogramme prédictif de la récidive.

OBJECTIVE: Diagnosis and therapeutics are challenging in pregnancy-related breast cancer and a multidisciplinary approach is therefore essential. A nomogram predictive of recurrence at 36 months of PABC has recently been developed, the aim was to evaluate it on an independent cohort. PATIENTS AND METHODS: This was a monocentric retrospective study including all patients with PABC between January 2004 and March 2018. A Kaplan-Meier method was used for the analysis of recurrence-free survival. Nomogram performance was analysed on 3 indicators: discrimination by a receiver operator characteristic curve (ROC), calibration and net clinical benefit by a decision curve analysis. RESULTS: Forty-one patients have been included. The survival rate was 83 % at 3 years. The nomogram showed acceptable discrimination capacities on receiver-operator characteristic analysis (c-index: 0.78). Decision-curve analysis showed a positive net-benefit for relapse probability ranges between 5 % and 35 %. CONCLUSION: The performance of this prediction model is acceptable but the net clinical benefit low. In conclusion, prediction of recurrence remains challenging with a continued need for more precise tools.