Natural disease progression model with transition probabilities to describes the continuum of Alzheimer’s disease
Noel Patson1,2*, Marwa E.Elhefnawy 2,3*, Samer Mouksassi 4, Goonaseelan (Colin) Pillai 2,5, Emmanuel Chigutsa 6, Ivelina Gueorguieva 7
1 School of Global and Public Health, Kamuzu University of Health Sciences, Blantyre, Malawi. 2 Applied Pharmacometrics Training – Africa Program, c/o Pharmacometrics Africa NPC, Cape Town, South Africa. 3 School of Pharmaceutical Sciences, University Sains Malaysia 4 Integrated Drug Development, Certara, Princeton, NJ, United States. 5 Division of Clinical Pharmacology, University of Cape Town, South Africa 6 Eli Lilly and Company, Indianapolis, Indiana, United States 7 Eli Lilly and Company, Bracknell, United Kingdom. *This project was done as part of the Applied Pharmacometrics Training Fellowship, a capacity strengthening program organised by Pharmacometrics Africa NPC and Certara Email: npatson@kuhes.ac.mw
Introduction and objectives
Understanding the natural Alzheimer`s Disease (AD) progression is important for developing disease modifying therapies [1]. Limited knowledge on natural disease progression and biomarkers has affected the potential impact of pharmacokinetic (PK) and pharmacodynamic (PD) modelling to inform clinical drug development of AD therapies. This work used multistate Markov models to describe natural disease progression, accounting for factors associated with the AD progression.
Methods
We used longitudinal Alzheimer’s Disease Neuroimaging Initiative (ADNI) data[2] collected from 2004 to 2022, which has clinical (cognitive/functional scores), imaging (amyloid plaque assessment by quantifying cortical standardized uptake value ratios, SUVRs), genetic, and biochemical biomarkers for the early detection and tracking of AD. Amyloid plaque was measured by β-amyloid (Aβ) radiotracer in positron emission tomography (PET) and our study focused on patients who had baseline and at least one post-baseline amyloid PET observation, measured by [18F] florbetapir. Using a multistate Markov model, we characterized the Alzheimer’s natural disease progression based on progressive transition from one stage to another among three disease stages, available in ADNI: cognitively normal (CN), mild cognitive impairment (MCI) and Alzheimer disease (AD). The developed model, adjusted for gender and time-varying amyloid plaque, to estimate the transition probabilities across the CN, MCI and AD disease stages. Hazard ratios (HRs) were used to characterize the risk of transition between two progressive stages over time.
Results
One thousand one hundred and thirty-three patients had at least a single amyloid plaque value out of the 2426 patients in the ADNI dataset. The baseline characteristics (e.g. age, gender, race, baseline amyloid plaque) for patients in ADNI substantially differed by disease staging at baseline. For example, patients who were CN and MCI were younger compared to those at AD stage. Our final multistate markov model adjusted for gender and time-varying amyloid plaque as important predictors of natural disease progression. The highest and the lowest mean time spent by a patient in AD and MCI states was 355.5 (SE=134.6) and 97.9 (SE=7.7) months respectively, reflecting a quicker transition of patients from MCI to AD compared to CN to MCI disease states. Patients with a higher amyloid plaque were at a higher risk of transitioning from CN or MCI to AD. For instance, a unit increase (in SUVr) in amyloid plaque was associated with seven-fold increased risk of transitioning from CN to MCI (HR= 7.4; 95% CI: 3.5, 15.5). Gender was not associated with the Alzheimer’s disease progression. Our model consistently predicted the transition probabilities, expected to double at intervals from 1 to 3 years and 3 to 10 years of follow-up. We observed a considerable, 55 back-transitions from MCI to CN that highlighted data quality limitation especially ascertainment of the AD severity: This necessitates improved neuropsychological examination for optimal AD severity classification.
Conclusion
Increased amyloid plaque predicts worsening of AD progression. Since patients progress relatively quickly from MCI to AD, there should be focus on developing anti-amyloid agents targeting early in the AD continuum.
References:
[1] Aisen, P.S., Cummings, J., Jack, C.R. et al. On the path to 2025: understanding the Alzheimer’s disease continuum. Alz Res Therapy 9, 60 (2017). https://doi.org/10.1186/s13195-017-0283-5
[2] Mueller SG, Weiner MW, Thal LJ, et al. Ways toward an early diagnosis in Alzheimer's disease: the Alzheimer's disease neuroimaging initiative (ADNI). Alzheimers Dement. 2005; 1: 55-66
