[External multicentric validation of two scores predicting the risk of relapse in patients with borderline ovarian tumors: The nomogram of Bendifallah and the score of Ouldamer]. / Validation externe de 2 scores prédictifs de récidive de tumeurs mucineuses et séreuses borderline de l'ovaire : le nomogramme de Bendifallah et le score d'Ouldamer.

BACKGROUND: Borderline ovarian tumors are rare and can occur in young women. For these patients, a fertility sparing surgery should be discussed. Two predicting borderline ovarian tumor relapse risk models were developed in 2014 (Nomogram of Bendifallah) and 2017 (Score of Ouldamer). This study aimed to valid in an external population, these two scores using a multi-institutional BOT database. METHODS: In this bicentric and retrospective study, all consecutive patients comprising the variable nomogram documented treated between January 2006 and December 2012 for BOT in centre hospitalier de Poissy-Saint-Germain and hôpital René-Huguenin were included. A ROC model was established for each predicting scores. RESULTS: Sixty-five patients were included in the study. Twelve patients showed a recurrence (19%), three of them experienced an infiltrative cancer (5%). The median time of recurrence was 25 months (range: 8-115). The concordance index for the Nomogram of Bendifallah and the Score of Ouldamer were 0.88 (IC 95% [0.78-0.98]) and 0.87 (IC 95% [0.77-0.96]) respectively. CONCLUSION: This study from an independent population valids the Bendifallah nomogram and Ouldamer score for clinical use in predicting borderline ovarian recurrence.