How to predict para-aortic node involvement in advanced cervical cancer? Development of a predictive score. A FRANCOGYN study

International audienceINTRODUCTION: Node involvement is one of the main prognostic factors for cervical cancer. Para-aortic lymph node (PALN) assessment is crucial for treating advanced cervical cancer, to define irradiation fields. Objective of this study was to develop a score predicting para-aortic lymph node involvement in patients with advanced cervical cancer. PATIENTS AND METHOD: We performed a multicenter, retrospective, study on 9 French centers from 2000 to 2015, including patients with advanced squamous cell cervix carcinoma who had PALN status assessed by imaging and/or by surgery. Factors associated with a risk of PALN involvement were determined by univariate and multivariate analysis using a logistic regression model. A score was then developed and validated. RESULTS: A total of 1446 patients treated for cervical cancer were included. Of these, 498 had an advanced squamous cell cervical cancer. Ninety-one patients (18.3%) had positive PALN. After univariate and multivariate analysis, tumor size on pelvic MRI, initial SCC, and suspected pelvic node involvement on PET-CT were included in our score. This model allowed the population to be divided into 3 risk groups. Area under the ROC curve of the score was 0.81 (95%CI = 0.72-0.90). In the low-risk group, 9% (28/287) had PALN involvement, whereas in the high-risk group, 43% (22/51) had PALN involvement. CONCLUSION: We developed a simple score predicting PALN involvement in advanced cervical cancers. Three risk groups can be defined, and patients considered to be at low risk may avoid para-aortic staging as well as extensive field irradiation