77P - A prognostic nomogram for the prediction of neuroblastoma

Background

Neuroblastoma (NBL) is the most common and aggressive cancer in infants, and there are no robust predictive nomograms for NBL. In this study, a database from the Therapeutically Applicable Research to Generate Effective Treatments was applied to develop and validate a prognostic nomogram for the prediction of individual 5-year overall survival (OS) probability in patients with NBL.

Methods

A total of 729 eligible NBL patients with their clinicopathological factors and biological characteristics from the database were assigned to either the training cohort (n = 511) or the validation cohort (n = 218). Independent predictors were identified by fitting the Cox model with lasso penalty and then were assembled into a nomogram to predict survival. The model was developed for predicting individual 5-year OS probability and was then internally and externally validated, calibrated and compared in each cohort.

Results

Four independent prognostic factors (International NBL Staging System stage, ploidy, histology and Children’s Oncology Group risk group) were discriminated and achieved favourable prediction efficacy following the lasso model. The prognostic nomogram incorporated those factors that performed well in the training and validation cohorts with OS (HR = 0.36, 95% CI: 0.27-0.48, P < 0.0 001; HR = 0.54, 95% CI: 0.34-0.86, P = 0.0 086; respectively) and recurrence-free survival (HR = 0.52, 95% CI: 0.40-0.67, P < 0.0 001; HR = 0.66, 95% CI: 0.44-0.99, P = 0.044; respectively).

Conclusions

The robust prognostic nomogram with four independent prognostic factors can accurately and diversely predict OS and recurrence-free survival in NBL patients.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.