Preoperative risk assessment model for identification of lymph node metastasis in early cervical cancer.
5600 Background: The aim of this study was to develop a preoperative risk prediction model for lymph node metastasis in patients with early cervical cancer. Methods: The medical records of 504 patients with early cervical cancer who underwent hysterectomy and pelvic/paraaortic lymphadenectomy between 2007and 2012 in our center were retrospectively reviewed. According to the order of surgery performed, data between 2007 and 2010 were allocated to a model development cohort (n=314), and data between 2011 and 2012 were allocated to an external validation cohort (n=190). By using preoperative clinicopathologic data, magnetic resonance imaging (MRI) data, and positron emission/computed tomography (PET/CT) data, a multivariate logistic model was created. Based on this model, predictive nomogram was developed and externally validated. Results: Age, tumor size measured by MRI, and lymph node metastasis on PET/CT were found to be independent risk factors for nodal metastasis. Developed nomogram incorporating these three predictors showed good discrimination and calibration, with a bootstrap-adjusted concordance index of 0.772. Also, the validation set showed good discrimination with a bootstrap-adjusted concordance index of 0.783. Conclusions: We have developed a robust model to predict lymph node metastasis in patients with early cervical cancer. This new tool may be useful to clinicians and patients when deciding lymphadenectomy and maybe useful in designing clinical trials. [Table: see text]