Amyotrophic lateral sclerosis (ALS) is a devastating disease recognized as a priority by the Veteran Health Administration because of its increased incidence in veterans, thus earning it a service connection status. Current treatment is largely supportive, and there is dire need for the development of therapeutics that slow disease progression. However, better outcome measures and prediction tools are required to improve research efficiency and facilitate the discovery of successful treatments for ALS. This proposal is to develop novel and important research tools that can be used to design more conclusive ALS clinical trials. First, a new ALS disability scale will be developed with improved sensitivity and reliability to be used as an outcome measure for ALS clinical trials. The current clinical outcome scale for ALS trials is the revised ALS Functional Rating Scale (ALSFRS-R). The ALSFRS-R is easy to administer and correlates with patient survival. However, the scale is relatively insensitive to changes in function and is not specific for changes in true disease status. Indeed, changes in patient comfort level due to palliative interventions are recorded as “improvements” in the ALSFRS-R even though the underlying disease is progressing. Better mathematical tools are available for the development of outcome measure scales. Rasch analysis is a modern test-theory technique that produces validated, reliable, responsive, linearly-weighted scales with meaningful overall sum- scores for use in clinical trials. Rasch analysis has been used to improve outcome questionnaires for other neurologic diseases. By combining clinical expertise with the mathematically rigorous Rasch methodology, a new ALS disability scale will be created that will overcome many of the limitations of the ALSFRS-R. Secondly, predictive models will be developed and validated in collaboration with a data analytics expert to allow for stringent patient stratification during clinical trial enrollment, improving trial efficiency and reducing costs. Simulations with the best existing models, which rely on non-parametric, non-linear modeling techniques, show the ability to reduce the number of patients needed for an ALS trial by 20%, and these models significantly outperformed the predictions of expert ALS clinicians. Refinement of these models to utilize only baseline data, incorporate outcome measures with higher sensitivity, and use biomarkers as predictor variables, will improve predictive power and provide a valuable tool for trial design and patient care. Finally, immune factors will be investigated as biomarkers of disease activity and as biologic variables in ALS prediction models. Previous studies indicate that pro-inflammatory markers are associated with faster disease progression in ALS, while anti-inflammatory markers are seen with slower disease course. Markers of T-cell populations are promising prognostic biomarkers based on animal and human studies and based on widesprea...