A predictive model for calculating the likelihood of recurrent uterine fibroids after surgical intervention

Abstract

Aim. To develop a predictive model for calculating the likelihood of recurrent uterine fibroids after myomectomy.Materials and Methods. We consecutively recruited 149 women aged 19 to 45 years who underwent myomectomy. Prediction of uterine fibroids was carried out using multivariate analysis including a classification tree (IBM SPSS Statistics). To construct a classification tree for predicting recurrent uterine fibroids, the following patient-related parameters were used as the input features: age, family history, nulliparity, past medical history of myomectomy, obesity, rapid fibroid growth, multiple fibroids at myomectomy, BCL-2, Ki-67, and VEGF expression.Results. When analyzing the resulting classification tree, we can conclude that the most significant factors defining the recurrence of uterine fibroids are low parity, multiple fibroids at myomectomy, intensive VEGF expression, and weak expression of BCL-2.Conclusion. Our original model allows the identification of the most significant predictors of recurrent uterine fibroids and might be proposed as a useful tool for clinical practice.