To evaluate multimodal MRI biomarkers—iron content, neuromelanin, and perfusion—across three longitudinal timepoints in a randomized trial of mesenchymal stem cell (MSC) infusion

Sensitive imaging biomarkers that reflect disease biology and clinical outcomes are critically needed in PD. Intravenous MSC infusions have shown promise for improving motor symptoms, but their effects on the brain remain unclear. Quantitative MRI offers mechanistic insight: quantitative susceptibility mapping (QSM) measures iron deposition, neuromelanin-sensitive MRI detects cell loss, and arterial spin labeling (ASL) quantifies perfusion as a marker of neural activity.

Forty-nine participants were randomized 1:1:1 to three placebo infusions, mixed (placebo + MSC), or MSC-only (three infusions). MRI was performed at baseline (TP01), interim (TP02), and final (TP03). Imaging metrics were extracted from atlas-defined ROIs (substantia nigra, putamen, locus coeruleus, cerebral crus) and analyzed relative to MDS-UPDRS total scores. Primary analyses assessed within-subject change (Δ imaging, ΔUPDRS) across arms and timepoints.

ASL putamen perfusion increased progressively across groups (Placebo: +3.4%, Mixed: +7.3%, MSC-only: +15.2%), paralleling clinical stabilization or improvement (ΔMDS-UPDRS: Placebo +9.6, Mixed +5.3, MSC-only +1.8). Correlations between perfusion change and ΔUPDRS were strongest in the MSC group (r = +0.22, n.s. after FDR). ΔLC neuromelanin contrast ratio showed weaker trends (MSC r = –0.26, n.s.), while ΔSN QSM and ΔCrus asymmetry indices were less robust but directionally consistent. Dose–response ordering was preserved across arms, and mixed-effects models confirmed significant arm × time interactions for putamen perfusion (p = 0.032, FDR-adjusted), with suggestive effects for LC neuromelanin (p ≈ 0.07).