Related Experiment Video
Updated: May 6, 2026
06:19
Constructing and Visualizing Models using Mime-based Machine-learning Framework
Published on: July 22, 2025
3.1K
Biomarker-based predictive models for prognosis in amyotrophic lateral sclerosis
Xiaowei William Su1, Zachary Simmons2, Ryan Michael Mitchell3
1George M. Leader Family Laboratory, Department of Neurosurgery, The Pennsylvania State University College of Medicine, Hershey.
JAMA Neurology
|October 23, 2013
Summary
Predicting amyotrophic lateral sclerosis (ALS) prognosis is challenging. This study identified plasma and cerebrospinal fluid (CSF) biomarkers using multivariable models to accurately forecast disease duration in ALS patients.
Area of Science:
- Neuroscience
- Biomarker Discovery
- Clinical Prognostics
Background:
- Amyotrophic lateral sclerosis (ALS) presents significant prognostic uncertainty due to wide survival variations.
- Accurate prognosis prediction is crucial for effective clinical management and clinical trial outcomes assessment.
Purpose of the Study:
- To identify predictive biomarkers in plasma and cerebrospinal fluid (CSF) for amyotrophic lateral sclerosis (ALS) prognosis.
- To develop multivariable models for forecasting disease duration in ALS patients.
Main Methods:
- Retrospective analysis of plasma (n=29) and CSF (n=33) biomarkers from ALS patients.
- Measurement of 35 biomarkers using multiplex and immunoassay techniques.
- Statistical modeling to identify biomarker panels for predicting total disease duration.
Main Results:
- Multivariable models incorporating plasma and CSF biomarkers achieved high goodness-of-fit (R² up to 0.962) and predictive accuracy (Cohen κ up to 0.930).
- Key predictive biomarkers included inflammatory markers, interleukins, granulocyte colony-stimulating factor, and l-ferritin.
- Models demonstrated moderate to good agreement between predicted and actual prognostic categories.
Conclusions:
- This study demonstrates a novel multivariable modeling strategy for predicting ALS prognosis.
- The findings support further biomarker discovery in larger cohorts with longitudinal follow-up for improved ALS patient care.
