Automated Imaging Differentiation in Parkinsonism (AID-P) outperformed both the Magnetic Resonance Parkinsonism Index and neurofilament light chain protein in differentiating PD from atypical parkinsonism.
Derek Archer, PhD
Results from a study comparing different promising biomarkers demonstrated that the Automated Imaging Differentiation in Parkinsonism (AID-P), an advanced diffusion MRI and machine learning developed biomarker, provided the best overall differentiation of Parkinson disease (PD) versus multiple system atrophy (MSA) and supranuclear palsy (PSP) in comparison with the Magnetic Resonance Parkinsonism Index (MRPI) and plasma-based neurofilament light chain protein (NfL).
For PD differentiation versus MSA and PSP, the AID-P had an area under the curve (AUC) of 0.900 (95% CI, 0.830–0.971; P
<.05), which were significantly better than NfL (AUC, 0.747; 95% CI, 0.626–0.869) and MRPI (AUC, 0.689; 95% CI, 0.537–0.842). Additionally, AID-P (AUC, 0.889; 95% CI, 0.765–1.00; P
<.05) and MRPI (AUC, 0.824; 95% CI, 0.664–0.985) measures were greater than NfL (AUC, 0.537; 95% CI, 0.326–0.747) for MSA versus PSP.
To evaluate the ability of the AID-P, MRPI, and NfL to differentiate different forms of parkinsonism, 2 disease specific models—PD vs. MSA/PSP, and MSA vs. PSP—were used. The AID-P (AUC, 0.864; 95% CI, 0.720–1.00), MRPI (AUC, 0.886; 95% CI, 0.786–0.987), and NfL (AUC, 0.761; 95% CI, 0.629–0.894) measures all performed with similar accuracies (all P
Derek B. Archer, PhD, research instructor, Vanderbilt Memory & Alzheimer’s Center, University of Vanderbilt, and colleagues combined measures to determine if any unique combination provided enhanced accuracy, but found no combination performed better than the AID-P alone in differentiating parkinsonisms. While all 3 measures were determined to significantly fluctuate depending on disease severity, all multimodal models performed similarly to the unimodal models (all P
“This is the first study to directly compare these 3 biomarkers, and the findings in the current sample demonstrate that the AID-P, using noninvasive diffusion MRI, provides diagnostic utility in parkinsonism,” Archer and colleagues concluded.
The researchers also noted that AID-P had the highest association with the Movement Disorders Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS; AID-P Radj2
, 26.58%; NfL Radj2
, 15.12%; MRPI Radj2
“When comparing these unimodal models to predict parkinsonian state with models that combined biomarkers, we found that overall, there were no differences between the AID-P and any multimodal model in differentiating parkinsonisms,” Archer and colleagues noted. “In other words, combining NfL and MRPI measures did not improve diagnostic accuracy compared with AID-P.”
The study included 39 participants with PD, 17 with MSA, and 16 with PSP. All patients with MSA had the parkinsonian subtype. Archer and colleagues aimed to compare the utility of the AID-P, MRPI, and NfL to differentiate disease state and explain disease severity in parkinsonism. MRPI was calculated in an automated manner by incorporating the pons (P), midbrain (M), middle cerebellar peduncle (MCP), and superior cerebellar peduncle (SCP) areas into a single metric (P/M)(MCP/SCP).
By inputting the subject space diffusion magnetic resonance imaging (MRI) maps into the AID-P algorithm, researchers were able to obtain the diagnostic probabilities of PD versus MSA/PSP and MSA versus PSP. Outputted probabilities for these analyses were used as inputs into the receiver operating characteristic (ROC) analyses to determine AUC.
To determine if different multimodal combinations of variables were superior than unimodal metrics in differentiating PD versus MSA/PSP and MSA versus PSP, researchers tested 4 different combinations: (1) AID-P + NfL + MRPI, (2) AIDP + NfL, (3) AID-P + MRPI, and (4) NfL + MRPI.
Between both groups, there were no differences in age (P
= .07), sex distribution (P
= .06) or disease duration (P
= .21). MSA and PSP groups had a higher levodopa equivalent daily dose than the PD group (P
Archer DB, Mitchell T, Burciu RG, et al. Magnetic Resonance Imaging and neurofilament light in the differentiation of Parkinsonism. Published online May 1, 2020. doi: 10.1002/mds.28060.