In a recent study published in npj Parkinson’s disease, researchers evaluated the differences in the progression of Parkinson’s disease (PD) between real-world and research populations.
Study: Disease progression strikingly differs in research and real-world Parkinson’s populations. Image Credit: Kotcha K/Shutterstock.com
Background
Disease-modifying therapies have not been successful for PD despite improved biological understanding and significant investment.
Therapies promising in preclinical and early trials have failed at later stages. There are several challenges to developing disease-modifying therapies for PD; notably, molecular processes driving pathogenesis are poorly understood.
Moreover, different patterns of PD progression and diverse onset ages suggest heterogeneity. Recent studies have uncovered at least six loci associated with genetic heterogeneity in PD progression to Lewy body dementia.
Characterization of PD progression can aid in the design of clinical trials and identify sub-populations. Most studies on PD progression have focused on research data, which might not reflect the population in general.
About the study
In the present study, researchers characterized the progression of PD between research and real-world populations. The Fox Insight longitudinal health and Harvard Biomarkers studies (HBS) were research population datasets.
Real-world data (RWD) sources included the Optum Claims–electronic health records (EHRs) and the Mass General Brigham (MGB) electronic medical records (EMRs).
PD progression was evaluated using common clinical rating scales, including the mini-mental-state examination (MMSE), Montreal cognitive assessment (MoCA), unified PD rating scale (UPDRS), and Hoehn and Yahr (H&Y).
These measures were compared between HBS and MGB by stratifying patients based on the time from initial diagnosis.
Clinical rating scales were actively collected and stored as structured data within the HBS; in MGB, they were passively collected and extracted from clinical notes.
Fox Insight and HBS cohorts had explicit information on PD diagnosis dates. Real-world cohorts required a two-year quiescence period to enroll PD cases.
Further, in RWD sources, individuals were excluded if they developed conditions leading to clinical parkinsonism symptoms during the quiescence period.
Cognitive decline was defined as a diagnosis of mild cognitive impairment, referral to neurology specialists for cognitive decline, or prescriptions of medicines typically used to treat cognitive decline during PD. A linear regression model was used to estimate the average progression of H&Y, UPDRS, and MMSE scores.
Findings
HBS and Fox Insight cohorts included 935 and 36,660 participants, while Optum and MGB cohorts comprised 157,475 and 22,949 individuals, respectively.
Research populations (60.4 years) had a lower age at PD diagnosis than the real-world populations (72.2 years). Follow-up duration was shorter in the Fox Insight study as it is a relatively new resource.
The average follow-up period was 6.1 years in HBS compared to 3.2 years in real-world populations. In general, clinical scores showed smaller changes over time in HBS compared to MGB. Optum lacked clinical rating scores.
MMSE scores were significantly higher throughout the first five years in HBS than in MGB. Notably, the median UPDRS total score doubled by the sixth year in MGB.
Next, the researchers assessed survival against notable clinical events, such as levodopa initiation, H&Y stage 3, discharge to long-term care, depression, fractures, falls, and dyskinesias.
They found that research populations, particularly the Fox Insight cohort, had significantly earlier start of medications. In contrast, PD medications were started later in MGB subjects, albeit levodopa initiation overlapped with the research populations.
Further, real-world populations reached an H&Y score ≥ 3 or met the definition for cognitive decline much sooner after diagnosis than research populations.
Further, levodopa treatment was initiated in over 80% of patients by H&Y stage 2.5. The Optum cohort showed a substantially earlier and higher incidence of falls, fractures, and depression than the MGB cohort.
Conclusions
The study analyzed PD progression across real-world and research populations and illustrated systematic differences and directional biases between them.
Research populations were diagnosed and started on levodopa and other medicines much earlier, with slower changes in clinical scales.
By contrast, PD diagnosis and treatment initiation were much later in real-world populations; they also had accelerated changes in clinical scales. Overall, the findings offer detailed insights into PD progression in different populations.
Study uncovers a previously unknown genetic link to autism spectrum disorder