Nomogram to predict live birth rate after fertility-sparing surgery for borderline ovarian tumours

Can a nomogram be used to predict the individual probability of live birth (LB) in women with borderline ovarian tumours (BOTs) receiving primary fertility-sparing surgery? A nomogram built according to the woman's age, histological subtype (serous versus mucinous), type of ovarian surgical treatment and FIGO stage can accurately predict the probability of LB in women with BOT. Current prediction models determine the probability of pregnancy after medically assisted reproduction (MAR) and form the basis of patient counselling to guide the decision as to whether to consider in vitro fertilization but do not take into account prediction of the LB rate. This was a retrospective multi-centre study including 187 women with fertility-sparing surgery for BOT diagnosed between January 1980 and December 2013. A multivariate logistic regression analysis of selected factors and a nomogram to predict the subsequent LB rate was constructed. A bootstrapping technique was used for internal validation. Fifty-one women had LB (27.3%). Taking into account multiple pregnancies, the overall LB rate was 40.1% (75/187). Federation International of Gynaecology and Obstetric (FIGO) stage, age at diagnosis, histological subtype and surgery type were included in the nomogram. The predictive model had an AUC of 0.742 (95% CI, 0.644-0.825) and 0.72 (95% CI, 0.621-0.805) before and after the 200 repetitions of bootstrap sample corrections, respectively, and showed a good calibration. The retrospective nature of the study cannot exclude all biases. Our nomogram is based on simple criteria, but did not take into account the evaluation of ovarian reserve. It demonstrates a fair relevance, but requires external validation before routine use. Clinicians are increasingly interested in such tools to support the patient in making an informed decision about treatment options. This nomogram contributes to the decision-making by defining simple risk factors of poor LB probability that can help identify good candidates for MAR. No external funding was used for this study. There are no conflicts of interest to declare. N/A.