P178 Coupling brain network simulation with pharmacokinetics for Parkinson's disease: towards patient-usable digital twins
William Lytton*127, Donald Doherty17, Adam Newton17, June Jung1, Samuel Neymotin13, Salvador Dura Bernal1, Thomas Wichmann57, Adriana Galvan57, Hong-Yuan Chu47, Yoland Smith57,Husan Abdurakhimov6, Jona Ekström6, Henrik Podéus Derelöv16, Elin Nyman6, Gunnar Cedersund6
1 Downstate Health Science University, Brooklyn NY USA
2 Kings County Hospital, Brooklyn NY USA
3 Nathan Kline Institute, Orangeburg NY USA
4 Georgetown University, Washington DC USA
5 Emory University, Atlanta GA USA
6 Linköping University, Linköping, Sweden
7 Aligning Science Across Parkinson's (ASAP) Collaborative Research Network, Chevy Chase, United States*billl@neurosim.downstate.edu
Introduction
Parkinson’s disease (PD) is characterized by complex motor deficits in multiple sites. Starting with dopamine (DA) depletion in substantia nigra, brain dysfunction subsequently occurs in primary motor cortex (M1), basal ganglia (BG) and other areas. At first, dysfunction is a direct consequence of reduced DA. Then, through the dynamics of compensation and decompensation, these other areas become themselves pathophysiological. We used simulation to explore the focal M1 pathophysiology seen in mouse models. We are now integrating pharmacokinetic (PK) models to consider how therapy (Rx) can normalize dynamics.
Methods
We adapted our NEURON/NetPyNE M1 neuronal network (NN) model to simulate PD, reducing pyramidal-tract layer 5 neuron (PT5B) excitability. We coupled a prior ODE PK model to evaluate DA, NE levels produced by L-DOPA, L-DOPS Rx, respectively, modulating parameters based on local DA,NE levels. Parameter optimizations explored PK outputs into network activity to look at 1. dose-timing; 2. gut absorption (bioavailability, gastric delays); 3. multi-compartment distribution (blood, fat, muscle, brain); 4. blood-brain barrier (BBB) crossing; 5. precursor conversion to DA, NE; 6. drug metabolism and clearance.
Results
We focused on NE since locus coeruleus (LC) degeneration directly affects M1 cells, while DA loss directly affects BG neurons. Our untreated network simulations showed elevated PT5B activity despite reduced PT5B excitability. This paradoxical firing rate increase was associated with enhanced LFP beta-band oscillatory power with beta bursts. NE Rx shifted network activity to 20-35 Hz high-beta activity with reduction in excessive beta power, partly normalizing activity.
Discussion
Our hybrid PK-NN model demonstrated related potential clinical Rx of PD with correction of the pathophysiological changes that produce motor dysfunction, thus starting to link treatment with well-being. Partial normalization of beta oscillations and firing rates with L-DOPS treatment may add to treatment outcomes. We can isolate clinically-modulatable effects including dose-timing, gut pretreatment, precursor transformation, and clearance to shape target-neuron effects. We hope to thereby improve effect/side-effect balance to reduce dyskinesias, wearing-off, and freezing. Future model iterations will extend to digital twin applications to a provide tools to assist patients in personally optimizing their own therapy.
Acknowledgements
This research was funded in part by Aligning Science Across Parkinson’s [ASAP-020572] through the Michael J. Fox Foundation for Parkinson’s Research (MJFF). For the purpose of open access, the author has applied a CC BY public copyright license to all Author Accepted Manuscripts arising from this submission.
Supported in part by STRATIF-AI funded by Horizon Europe agreement 101080875.
References
none
