Quantitative MRI Biomarkers for Early Muscle Involvement in Late Onset Pompe Disease

Early treatment in late-onset Pompe disease (LOPD) increasingly depends on detecting subclinical muscle involvement. Quantitative muscle MRI (qMRI) has emerged as a promising tool in this context. We conducted a multicenter prospective study in LOPD patients (≥ 10-years-old), stratified into early-stage (Walton 0-1, FVC ≥ 80%) and symptomatic (Walton 2-6). Thigh MRIs were acquired at 3 T using T1-weighted and STIR sequences. Fat fraction (FF) and water T2 (wT2) were calculated in 11 different regions using 6-point Dixon and 17-echo multi-echo spin-echo acquisitions, respectively. Functional, respiratory, and patient-reported outcomes were assessed. The adductor magnus (AM)/rectus femoris (RF) FF ratio was evaluated for its discriminative power. wT2 was analyzed as a binary score per muscle (0 or 1), depending on whether its value exceeded the control mean plus two SD, and summed across 11 muscles to compute an individual wT2 involvement score (WIS). Thirty-three LOPD patients (16 early-stage; 17 symptomatic) and 34 controls were recruited. Thigh FF strongly correlated with clinical scales (ρ up to 0.86). ROC analysis identified AM as the best discriminator (AUC: 0.96, cut off ≥ 7.93%), with similar performance for the AM/RF ratio (AUC: 0.94). wT2 showed weaker correlation with clinical scores compared to FF; symptomatic patients had higher WIS compared to early-stage and controls. Thigh FF is a robust biomarker of disease severity. AM FF and the age-independent AM/RF FF are sensitive to early structural changes. A WIS > 3 may reflect symptomatic disease. If validated longitudinally, these qMRI parameters could help guide optimal timing of therapy initiation.