Joint model with latent disease age: Overcoming the need for reference time

Area of Science:

• Biostatistics
• Neuroscience
• Computational Biology

Background:

• Neurodegenerative disease progression exhibits significant heterogeneity, complicating therapeutic development.
• Existing progression models struggle with defining a precise reference time, especially when disease onset precedes symptom manifestation.
• Joint models combining longitudinal and survival data are effective but require a well-defined reference time.

Purpose of the Study:

• To propose a novel joint non-linear mixed-effect model incorporating a latent disease age.
• To overcome the limitation of ill-defined reference time in neurodegenerative disease progression modeling.
• To improve the understanding and modeling of heterogeneity in neurodegenerative diseases like Amyotrophic Lateral Sclerosis (ALS).

Main Methods:

• Developed a joint non-linear mixed-effect model with a latent disease age as the longitudinal sub-model.
• Integrated a survival sub-model estimating a Weibull distribution based on the latent disease age.
• Validated the model using simulated data and benchmarked it against a state-of-the-art joint model using ALS patient data.

Main Results:

• The proposed model demonstrated superior performance compared to the state-of-the-art joint model in predicting ALS progression.
• Achieved significantly better results for absolute bias on the ALS Functional Rating Scale Revised score (p < 1.4e-17).
• Showed improved mean-cumulative AUC for right-censored death events (p < 1.7e-03), indicating better survival prediction.

Conclusions:

• The developed joint non-linear mixed-effect model with latent disease age is well-suited for neurodegenerative diseases with unreliable reference times.
• The model effectively captures disease heterogeneity and improves prediction accuracy for both disease progression and survival.
• This approach offers a valuable tool for analyzing complex longitudinal and survival data in clinical research for neurodegenerative conditions.