Predictive Nomographic Model for Tumour Recurrence in Children with Wilms' Tumour

This study aims to retrospectively analyse factors affecting recurrence in children with Wilms' tumour (WT) and establish a nomographic model for predicting WT recurrence 12 months after operation.

Children with Wilms' tumour admitted to our hospital from January 2020 to January 2022 were retrospectively analysed. According to the recurrence of children recorded in the medical record system, they were divided into case group and control group. General demographic data and disease characteristics were collected, and multivariate logistics regression analysis was used to evaluate recurrence factor 12 months after operation. A nomogram model was constructed to predict the recurrence factors of WT, and an ROC (receiver operating characteristic) curve was drawn to verify the predictive efficacy of the nomogram model. Calibration and decision curves were constructed.

A total of 119 children with WT were selected and divided into case group (25 children with recurrence) and control group (94 children without recurrence). Multivariate logistic regression analysis showed that clinical stage, histological risk group, Ki-67, tumour volume, capsule rupture and incomplete resection were all risk factors for the recurrence of WT (odds ratio (OR) >1, p < 0.05). The R package rms was used to draw a nomogram for the risk of recurrence. The C-index of clinical stage, histological risk group, Ki-67, tumour volume, capsule rupture and incomplete resection was 0.885 (95% CI: 0.810-0.956). The accuracy of the nomogram model in predicting the risk of recurrence was verified. The ROC curve showed that the Area Under Curve (AUC) value of the model was 0.911 (95% CI: 0.854-0.969; p < 0.05). The calibration and reference curves of the risk prediction model were similar, which proved the high consistency between the predicted risk and the actual risk of the recurrence of WT. The net benefit rate of the prediction model in the threshold range was high.

Clinical stage, histological risk group, Ki-67, tumour volume, capsule rupture and incomplete resection were risk factors for the recurrence of WT. A predictive nomogram model was constructed using these factors to screen children with WT with a high risk of recurrence for early intervention.