Modeling Amyotrophic Lateral Sclerosis Progression: Logic in the Logit

Background Amyotrophic lateral sclerosis functional rating scale-revised (ALSFRS-R) has emerged as a clinical prognostic marker for clinical and research purposes in amyotrophic lateral sclerosis (ALS). However, tools for predicting disease progression are still underdeveloped. The aim of this study was to mathematically model ALS progression to provide a reliable and personalized approach to the prognosis for ALS patients. Also, it aimed to provide a reliable prediction tool for the current and newly diagnosed patients. Methods Twenty patients from the South-East England Amyotrophic Lateral Sclerosis register (SEALS) database were included in the analysis. A non-linear logistic regression model was used to describe disease progression from baseline health to the theoretical maximum disease. The reliability of predicted variables and correlation between model parameters were assessed separately for each subject. Results The logistic regression model best described the disease progression in patients with a high progression rate. Most notably, the model fitted better when a patient has progressed enough to approximately the midpoint of the functional rating scale. The model failed to characterize the disease course in patients defined as slow progressors. Furthermore, the linear relationship between the rate of progression and time since onset at ALFRS-R score of 24 was evident in 65% of patients. Conclusion These results indicate that the rate of disease progression and time when ALSFRS-R declines to half the maximum score are correlated with functional outcomes. Nonetheless, the logistic model failed to describe disease course in patients with slow progression rates. Different rates of progression can be attributed to the genetic heterogeneity of ALS. Thus, clinicians and patients can benefit from adding a gene factor to the equation. With the outlined limitations, the model can provide a good prognostic tool.