Related Experiment Video
Updated: Jun 26, 2026
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
Proposal of a Risk Stratification Algorithm for Extracranial Malignant Rhabdoid Tumors
Juno Eberl1, Angelika Eggert1,2, Monika Scheer1,3
1Department of Pediatric Hematology and Oncology, Charité-Universitätsmedizin Berlin, Freie Universität Berlin and Humboldt-Universität Zu Berlin, Berlin, Germany.
Background:
Malignant rhabdoid tumors (MRT) are highly aggressive tumors, primarily described in children, but also affecting adults. The objective was to identify independent prognostic factors to predict individual patients' risk and to establish an age-spanning risk stratification system for prospective trials.
Patients And Methods:
Data from the Surveillance, Epidemiology, and End Results Program (SEER 17), accruing patients with extracranial MRT as first malignancy 2000-2019 yielded 324 evaluable cases. The initial selection of probable predictive factors was conducted through univariate analysis, employing the Kaplan-Meier estimator. The Cox proportional-hazard regression model was employed to assess the independent risk of preselected factors. Besides a nomogram, a risk-scoring system was developed based on the sum of hazard ratios to create a 4-risk group and a 3-risk group stratification, respectively.
Results:
3-, 5-, and 10-year overall survival rates were 35.1%, 31.1%, and 27.9% respectively, with corresponding disease-specific survival rates of 38.7%, 34.8%, and 32.8%. Significant prognostic factors identified in the univariate analysis were age, size, site and disease stage, and consequently included in the Cox regression analysis. All factors remained significant. A nomogram was constructed to assess the patient's individual survival probability. Additionally, a risk score system was developed in which individual risk scores were determined by adding up the respective hazard ratios. Patients were stratified into four risk groups, with 3-year survival rates of 77.4%, 51.0%, 26.5%, and 9.0%, respectively. A categorization into three risk groups yielded survival rates of 61.9%, 26.3%, and 8.4%, respectively.
Conclusion:
This analysis identified subsets of eMRT patients who exhibited highly divergent survival probabilities. The derived nomogram may facilitate the estimation of individual survival outcomes. The proposed scoring system may be used to assign patients to risk-tailored therapy regimens. It is our hope that our approach will improve the survival of this aggressive disease.