Objective: (1) To characterize motor therapeutic use in the Parkinson’s Progression Markers Initiative (PPMI) early, untreated (de novo) cohort, and (2) to develop an evidence-based first-line treatment strategy based on baseline characteristics.
Background: Parkinson’s disease (PD) motor symptoms are largely managed by pharmacological agents which prevent dopamine metabolism (MAO-B inhibitors) or mimic/supplement dopamine (dopamine agonists, levodopa). Treatment selection is often dictated by provider preference and training and patient attributes. We utilized a data-driven approach to select the first-line drug and optimize individual patient outcomes.
Method: 392 de novo PD participants with motor treatment information were identified in the PPMI database, accessed 10/5/19. A subset of participants with complete visit data prior to initiating treatment and 6-month post-treatment data were selected for analysis. Participants initiating multiple first-line therapies were excluded. Outcomes were MDS-UPDRS I-IV and Schwab & England Activities of Daily Living (ADL). An individual’s optimal treatment for each of the 5 outcomes was estimated using a linear marginal model with a SuperLearner library with machine learning models estimating the confounding between baseline covariates and each treatment assignment and outcome. Estimates and confidence intervals of treatment rule advantages were computed, representing the change in expected future outcomes under the estimated rule relative to the current standard of care.
Results: A plurality of participants initiated monotherapy with MAO-B inhibitors (n=1129, 32.9%), followed by levodopa (n=125, 31.9%), and dopamine agonists (n=79, 20.2%); 27 (6.9%) initiated with both a dopamine agonist and MAO-B inhibitor. The advantages for each outcome were reductions of 1.32, 1.45, 1.58, and 0.39 in MDS-UPDRS parts I, II, III (on), and IV respectively and an increase of 1.93 in ADL at 6 months after initiating treatment (Table 1).
Conclusion: The variety in treatment options for PD motor symptoms necessitates a personalized and data-driven approach to treatment selection. Our future work will explore a composite outcome to optimize treatment across multiple symptom domains and expand to a longer follow-up time and treatment alterations.
To cite this abstract in AMA style:
M. Javidnia, J. Jones, A. Ertefaie, C. Venuto. Treatment characterization and decision-making strategy for first-line pharmacological motor symptomatic therapies in Parkinson’s disease [abstract]. Mov Disord. 2020; 35 (suppl 1). https://www.mdsabstracts.org/abstract/treatment-characterization-and-decision-making-strategy-for-first-line-pharmacological-motor-symptomatic-therapies-in-parkinsons-disease/. Accessed September 7, 2024.« Back to MDS Virtual Congress 2020
MDS Abstracts - https://www.mdsabstracts.org/abstract/treatment-characterization-and-decision-making-strategy-for-first-line-pharmacological-motor-symptomatic-therapies-in-parkinsons-disease/