Objective: We evaluated the sensitivity of baseline MRI measures to predict future PD conversion, and whether baseline MRI associates with MDS prodromal likelihood ratio (LR) scores at 10-year follow-up (FU).

Background: Identifying baseline MRI markers that predict phenoconversion to Parkinson’s disease (PD) and prodromal progression is critical for early disease detection and intervention trials.

Method: This study retrospectively analyzed imaging data from N=69 fifteen non-manifesting PD mutation carriers (LRRK2, or GBA1), and N=46 healthy non-carriers (Age Mean±SD= 48.46±10.97 years). All participants underwent baseline MRI. Over up to 15 years FU,10 participants converted to PD. Five pre-specified predictors were examined: the functional PD-related pattern (fPDRP), bilateral substantia nigra free-water fraction (SN FWF), normalized gray matter (GM) volume, and bilateral caudate and putamen FWF. Cox regression with LOOCV-based combination analysis addressed conversion prediction. Spearman correlations and linear regression with FDR correction was applied.

Results: No predictor significantly predicted PD conversion after FDR correction. GM volume showed the most consistent trend across all analyses: lower baseline GM volume was associated with higher conversion risk (AUC=0.67; HR=0.64 per SD [0.36–1.15]; p=0.133) and with higher prodromal LR scores at 10 years (Spearman r=−0.20; p=0.087; R²=0.047). Exhaustive combination analysis (25 models, LOOCV) showed no combination improved discrimination beyond GM volume alone. fPDRP showed near-chance conversion prediction (AUC=0.49) and negligible association with LR scores (r=−0.03), despite a numerically higher group mean in converters driven by two outliers. The multivariate Cox model was non-significant overall (LR χ²=2.12; p=0.548).

Conclusion: GM volume showed a trend as a sensitive baseline MRI marker in pre-manifest PD mutation carriers, predicting both future clinical conversion and subclinical prodromal progression. The consistent null findings for fPDRP and SN FWF suggest these measures may not yet be informative at the pre-manifest stage. Results are exploratory due to relatively small sample size (EPV=3.7) and highlight the need for investigating these aims in larger longitudinal cohorts.

« Back to 2026 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/baseline-mri-predictors-of-phenoconversion-and-prodromal-progression-in-non-manifesting-pd-mutation-carriers-a-longitudinal-retrospective-analysis/