Objective: To characterize the sensitivity of smartphone and -watch sensor features to levodopa and dopamine agonist (DA-Tx) treatment.

Background: Digital health technology tools (DHTTs) enable remote and frequent monitoring of core motor symptoms in Parkinson’s disease (PD). They may increase sensitivity to disease progression and thus facilitate detection of treatment effects in clinical trials.

Method: Early PD participants (<2y disease duration) of the Phase II PASADENA Part 1 study of prasinenezumab (NCT03100149), a humanized monoclonal antibody targeting aggregated alpha-synuclein, completed Roche PD Mobile Application v2 smartphone ‘active tests’ at-home every morning and carried the phone/smartwatch for passive monitoring during the day. Analyses investigated (A) long- and (B) short-term effects of DA-Tx. (A) analyzed data from 6 fortnights (two-week periods) prior to and after DA-Tx start in n=114/316 who started DA-Tx within the Part 1 (52 weeks). (B) Acute effects of levodopa were analyzed with in-clinic OFF/ON sensor data from n=58 participants; to ensure stable levodopa responses, only data after a 6-month stabilization period were analyzed. Both analyzed 17 pre-selected features with mixed effects models. All results were interpreted with respect to 80% confidence intervals.

Results: A total of 12 out of 17 sensor features showed positive long-term DA-Tx effects: speeded tapping variability, hand-turning speed, spiral drawing celerity, u-turn speed, spontaneous turn speed during daily life, postural and rest tremor power, gesture power in daily life and symbol-digit modalities number correct. However, MFCC2 variance in free speech decreased (more monotonicity) following DA-Rx. Only tremor sensor features differentiated same-day in-clinic OFF/ON levodopa assessments.

Conclusion: The Roche PD Mobile Application detected both long- and short-term effects of DA-Tx. Specifically, sensor features reflecting tremor and bradykinesia in both active testing and passive monitoring during daily life improved upon DA-Rx over long-term. Only tremor sensor features were sensitive to acute effects of levodopa, potentially reflecting the difficulty in washing out levodopa or different pharmacodynamics of levodopa on bradykinesia and tremor systems. These findings support the use of DHTTs to detect symptomatic treatment effects.

« Back to MDS Virtual Congress 2021

MDS Abstracts - https://www.mdsabstracts.org/abstract/effect-of-dopaminergic-treatment-on-digital-sensor-features-in-phase-ii-pasadena-part-1-in-early-parkinsons-disease/