Spinocerebellar ataxia type 1 (SCA1) is an inherited movement disorder characterized by progressive degeneration of the cerebellum and brainstem¹. It was recently demonstrated that volumetric MRI is more sensitive to disease progression in SCA1 than standardized clinical scores in a 2-year longitudinal study². In addition, MRS was shown to detect highly significant differences in neurochemical profiles between patients with SCA1 and healthy controls³. As such, structural MRI combined with state-of-the-art MRS might provide a robust neuroimaging protocol to non-invasively and objectively monitor cerebral pathology. The aim of this study was to use MRS with volumetric MRI to determine the sensitivity of these two techniques to onset and progression of neurodegeneration in patients with SCA1.16 early-moderate stage patients with SCA1 (genetically confirmed) and 21 healthy control subjects were enrolled in this study. All MR measurements were carried out on a 3 T Siemens scanner using the standard body coil for excitation and 32-channel head coil for reception. Subjects were scanned at baseline and followed up at ~18 and ~36 months. Ataxia severity was assessed at each visit with the Scale for the Rating and Assessment of Ataxia (SARA) score, which yields a composite ataxia score in the range of 0 (no ataxia) – 40 (most severe ataxia)⁴. T₁-weighted MPRAGE images (T_R=2530ms; T_E=3.65ms; FA=7°; slice thickness=1mm; 224 slices; FOV=256×176 mm²; matrix size=256×256) were acquired to position the volume-of-interest (VOI). Proton spectra were acquired from 3 regions: cerebellar vermis, cerebellar hemisphere and pons in all subjects using the semi-LASER sequence⁵ (T_E=28ms, T_R=5 s, 64 averages). MRS spectra were processed in Matlab and quantified with LCModel as previously described6. Only metabolites that were reliably quantified (Cramér-Rao lower bounds ≤ 50% and correlation r>-0.5) from at least half of the spectra from a particular brain region were included in the final analysis. Intracranial, cerebellar, and brainstem substructures (medulla oblongata, pons, midbrain, and the superior cerebellar peduncles) volumes for each scan were obtained using Freesurfer^7,8. Statistical analysis was performed by estimating a linear change (i.e. slope) in metabolite concentrations and in volumes in for each person (scaled by using their number of days from baseline) and then scaling relative to each person’s baseline value) to obtain their the “percent change per year.”. Mean value of these changes and the standard deviation (SD) within group are reported. The effect size² which represents the ratio between mean %change to SD of %change is also reported.Using structural MRI, the pontine volume decreased highly significantly over 3 years in SCA1 compared to control subjects (Figure 1): pons volume was reduced by 1.8±1.2 % (P<2e^-6) per year (Figure 2). The largest effect size (Table 1) was observed for the pons (-1.54 vs. 0.33 in SCA1 vs. controls, respectively) among brain regions, indicating that this region is most sensitive to disease progression over time, consistent with the previous longitudinal MRI study². With MRS, a significant reduction in [tNAA]/[Ins] (i.e. ratio of (N-acetylaspartate + N-acetylaspartylglutamate) to myo-inositol) was observed in SCA1 relative to controls (Figure 1) in all 3 regions studied, and the pons showed the largest sensitivity over 3 years based on an effect size of -0.84. The pontine [tNAA]/[Ins] was found to decrease by 4.0±4.8 % (P<0.02) per year in SCA1 patients (Figure 2). Both MRI and MRS measurements were more sensitive to disease progression than SARA scores (Table 1). Longitudinal changes in pontine volume, pontine [tNAA]/[Ins] ratio and SARA scores for all subjects are shown in Figure 3. Group separation was observed in all three measurements, except for the two pre-symptomatic SCA1 patients (SARA scores of ≤1) represented by the red and green lines in Figure 3. [tNAA]/[Ins] of one of the 2 pre-symptomatic subjects (green line) was clearly among SCA1 patients and below the lowest control [tNAA]/[Ins] across all 3 time points, starting at baseline, while the pons volume of this subject overlapped with control values at baseline. This observation suggests that MRS might be more sensitive to detect early changes in disease onset compared to structural MRI. The other presymptomatic subject (red line) showed no neurochemical or clinical signs of disease, but a pons volume at the low end of the control range. Both volumetric MRI and MRS measurements at 3 T are more sensitive to progressive neurodegenerative changes than clinical assessment. In this study volumetric MRI was most sensitive to disease progression while MRS might be more sensitive to detect early disease.