Predicting Lymph Node Involvement in Borderline Ovarian Tumors with a Quantitative Model and Nomogram: A Retrospective Cohort Study

Abstract

Purpose: This study aimed to establish a predictive model for lymph node involvement (LNI) in patients with borderline ovarian tumor (BOT) using clinicopathological factors.

Patients and methods: We collected clinical data from consecutive patients who underwent lymphadenectomy for BOT between 2001 and 2018 and analyzed their clinicopathological features. Multivariate logistic regression was used to identify all independent risk factors associated with LNI; these were then incorporated into the prediction model.

Results: In total, we included 248 patients with BOT who were undergoing lymphadenectomy. These were divided into a training cohort (n=174) and a validation cohort (n=74). When considering histopathological data, 16 and 5 patients were identified to have LNI in the training and validation cohorts, respectively. Overall, 13.5% (21/156) patients with serous BOT had LNI while 0% (0/92) patients with non-serous BOT had LNI. We identified several predictors of LNI: the largest tumor being ≥ 12.2cm in diameter, the presence of lesions on the ovarian surface, and the presence of pelvic or abdominal lesions. We created a prediction model and nomogram that incorporated these three risk factors for serous BOT. The model achieved good discriminatory abilities of 0.951 and 0.848 when predicting LNI in the training and validation cohorts, respectively. The LNI-predicting nomogram had an area under curve (AUC) of 0.951 and generated well-fitted calibration curves.

Conclusion: Non-serous BOT may not require lymphadenectomy as part of surgical staging. The individual risk of LNI in patients with serous BOT can be accurately estimated using our prediction model and nomogram. The use of LNI criteria provides a practical way to support the clinician in making an optimal decision relating to surgical scope for patients with BOT.

Keywords: borderline ovarian tumor; lymph node involvement; lymphadenectomy; nomogram; prediction model.

Figure 1

Diagram showing how participants were recruited.

Figure 2

Receiver operator curve (ROC) for the largest tumor diameter cutoff values in the training cohort. The area under the ROC was 0.723 (95% CI: 0.603–0.843).

Figure 3

Receiver operator curve (ROC) of the model was used to estimate the probabilities of lymph node involvement in both the training and validation cohorts. (A) the training cohort; the area under the ROC was 0.951 (95% CI: 0.911–0.991). (B) the validation cohort; the area under the ROC was 0.848 (95% CI: 0.590–1.000).

Figure 4

Nomogram for estimating lymph node involvement (LNI) risk and its predictive performance. (A) The nomogram was formulated based on the proportional conversion of each regression coefficient in multivariate logistic regression to a 0- to 100-point scale by using the rms package in the R environment (version 3.3). (B) Calibration plots showed good agreement with regards to the presence of LNI when compared between the risk estimation provided by the nomogram and the histopathological confirmation of surgical specimens in the training cohort. (C) Calibration plots showed good agreement with regards to the presence of LNI when compared between the risk estimation by the nomogram and the histopathological confirmation of surgical specimens in the validation cohort.