Hurley, Dan Neurology Today: 6 July 2017 - Volume 17 - Issue 13 - p 35doi: 10.1097/01.NT.0000521718.33696.b8
VANCOUVER—The first clinical trial to employ a smartphone app to continuously measure symptoms associated with Parkinson's disease (PD) has found the technology to be well-accepted by patients during 24 weeks of use and highly correlated with traditional clinical measures, according to a poster presented here on June 6 at the International Congress of Parkinson's Disease and Movement Disorder?s.
Indeed, the smartphone measurements proved sensitive enough to detect upper-limb tremors that were not reported or present during standard physician-administered assessments.
“If you compare healthy control data to the patients rated as having no tremor, the data from the smartphone is clearly picking up a tremor,” said Michael Lindemann, PhD, from Roche's Innovation Center in Basel, Switzerland, on behalf of the research team.
“It's not only consistent with how physicians have measured the disease so far, it actually augments the picture, offering a unique window into patients' day-to-day life in an unobtrusive fashion,” Dr. Lindemann said.
Incorporated into a phase 1 clinical trial of a drug under development by Roche and Prothena Biosciences, the study involved 44 patients with early-stage PD. The research also included 35 healthy controls recruited in a separate study using an identical approach. Patients were asked to carry out six tests of symptoms every morning, including sustained phonation, gait, balance, dexterity, rest tremor, and postural tremor. In addition, they were asked to carry the smartphone with them throughout the day so that the device's accelerometer and gyroscope could passively monitor their movements. Results were compared to standard clinical measures, including the Movement Disorder Society-Unified Parkinson's Disease Rating Scale (MDS-UPDRS) obtained by clinicians.
Adherence to the protocol was surprisingly good throughout the duration of the trial. Patients performed the six active tests on average at least three times per week, although slowly dropping from over 75 percent adherence at the outset of the trial to about 50 percent by the end. Likewise, the patients carried the smartphone with them to permit passive monitoring about six hours per day at the outset of the trial, dropping to about four hours per day by the end. Even so, that was enough to generate a total of 24,104 hours of passive monitoring during the course of the study.
The smartphones' measurements of the patients' performance of the daily tests were generally consistent with the physicians' ratings on the MDS-UPDRS for these symptoms: rest tremor (p=0.04); postural tremor (p=0.010); dexterity (p=0.04); balance (p<0.001); and gait (p<0.001).
“Moreover, active test features for healthy controls were often significantly different from PD patients whose motor symptoms were scored as ‘0’ (that is, absent), suggesting that active test data collected in the two weeks before and after a clinical visit may augment the clinical picture obtained at site visits,” the poster stated.
Using machine learning to detect the patients' walking, standing, sitting and other movements during passive monitoring, the study found significantly lower gait activity (36 percent less) than healthy controls, reflecting less mobility. It also detected 17 percent fewer standing up/sitting down transitions per hour compared to healthy controls, and 38 percent fewer turns per hour while walking than healthy controls.
Commenting on the study, E. Ray Dorsey, MD, MBA, professor of neurology and director of the Center for Health and Technology at the University of Rochester Medical Center in New York, said: “This is the first time a smartphone has been incorporated into a clinical trial in Parkinson's disease. This is actually a big deal. Rather than just conducting assessments twice a year in the clinic, this study shows you can obtain continuous, objective assessments in the real world. It could be especially useful in drug trials, to see if somebody is showing an improvement over the course of a study. That would be powerful information.”
This report was one of several reports from the Neurology Today Conference Reporter coverage of the International Congress of Parkinson's Disease and Movement Disorders. Look for more coverage from the meeting here: http://bit.ly/NTCR-MDS-NT.
In an upcoming issue of Neurology Today, we will provide broader look at the challenges and opportunities of using wearable technologies in neurology clinical trials.
•. International Congress of Parkinson's Disease and Movement Disorders Abstract 541. Lipsmeier F, Fernandez Garcia I, Wolf D, et al. Successful passive monitoring of early-stage Parkinson's disease patient mobility in Phase I RG7935/PRX002 clinical trial with smartphone sensors http://http://www.mdsabstracts.org/abstract/successful-passive-monitoring-of-early-stage-parkinsons-disease-patient-mobility-in-phase-i-rg7935prx002-clinical-trial-with-smartphone-sensors/.