Prognosis of high-risk neuroblastoma (HRNB) remains poor despite multimodal therapies. Better prediction of survival could help to refine patient stratification and better tailor treatments. We established a mechanistic model of metastasis in HRNB relying on two processes: growth and dissemination relying on two patient-specific parameters: the dissemination rate μ and the minimal visible lesion size Svis. This model was calibrated using diagnosis values of primary tumor size, lactate dehydrogenase circulating levels, and the meta-iodobenzylguanidine International Society for Paediatric Oncology European (SIOPEN) score from nuclear imaging, using data from 49 metastatic patients. It was able to describe the data of total tumor mass (lactate dehydrogenase, R2 > 0.99) and number of visible metastases (SIOPEN, R2 = 0.96). A prediction model of overall survival (OS) was then developed using Cox regression. Clinical variables alone were not able to generate a model with sufficient OS prognosis ability ( P = .507). The parameter μ was found to be independent of the clinical variables and positively associated with OS ( P = .0739 in multivariable analysis). Critically, addition of this computational biomarker significantly improved prediction of OS with a concordance index increasing from 0.675 (95% CI, 0.663 to 0.688) to 0.733 (95% CI, 0.722 to 0.744, P < .0001), resulting in significant OS prognosis ability ( P = .0422).

High Risk Neuroblastoma (HRNB) is the second most frequent solid tumor in children. Prognosis remains poor despite multimodal therapies. Mathematical models have been developed to describe metastasis, but their prognosis value has yet to be determined and none exists in neuroblastoma. We established such a model for HRNB relying on two coefficients: α (growth) and μ (dissemination). The model was calibrated using diagnosis values of primary tumor size, lactate dehydrogenase circulating levels (LDH) and the meta-iodo-benzyl-guanidine (mIBG) SIOPEN score from nuclear imaging, using data from 49 metastatic patients treated according to the European HR_NBL1 protocol. The model was able to accurately describe the data for both total tumor mass (LDH, R^2>0.99) and number of visible metastasis (SIOPEN, R^2=0.96). Statistical analysis revealed significant association of LDH with overall survival (OS, p=0.0268). However, clinical variables alone were not able to generate a Cox-based model with sufficient prognosis ability (p=0.507). The parameter μ was found to be independent of the clinical variables and positively significantly associated with OS (p = 0.0175 in multivariate analysis). Critically, addition of this novel computational biomarker to the clinical data drastically improved the performances of predictive algorithms, with a concordance index in cross-validation going from 0.755 to 0.827. The resulting signature had significant prognosis ability of OS (p=0.0353). Mechanistic modeling was able to describe pathophysiological data of metastatic HRNB and outperformed the predictive value of clinical variables. The physiological substrate underlying these results has yet to be explored, and results should be confirmed in a larger cohort.