Two-year serial whole-brain N-acetyl-L-aspartate in patients with relapsing-remitting multiple sclerosis

Objectives: To test the hypotheses that 1) patients with relapsing-remitting multiple sclerosis (RR-MS) exhibit a quantifiable decline in their whole-brain concentration of the neural marker N-acetyl-L-aspartate (WBNAA), that is 2) more sensitive than clinical changes and 3) may provide a practical outcome measure for proof-of-concept and larger phase III clinical trials. Methods: Nineteen patients (5 men and 14 women) with clinically definite RR-MS, who were 33 ± 5 years old (mean ± SD), had a disease duration of 47 ± 28 months, and had a median Expanded Disability Status Scale (EDSS) score of 1.0 (range 0-5.5), underwent MRI and proton magnetic resonance spectroscopy (¹H-MRS) semiannually for 2 years (5 time points). Eight matched control subjects underwent the protocol annually (3 time points). Their global N-acetyl-L-aspartate 1HMRS signal was converted into absolute amounts by phantom replacement and into WBNAA by dividing with the brain parenchymal volume, VB, from MRI segmentation. Results: The baseline WBNAA of the patients (10.5 ± 1.7 mM) was significantly lower than that of the controls (12.3 ± 1.3 mM; p < 0.002) and declined significantly (5%/year, p < 0.002) vs that for the controls who did not show a decline (0.4%/year, p > 0.7). Likewise, VB values of the patients also declined significantly (0.5%/year, p < 0.0001), whereas those of the controls did not (0.2%/year, p > 0.08). The mean EDSS score of the patients increased insignificantly from 1.0 to 1.5 (range 0-6.0) and did not correlate with VB or WBNAA. Conclusions: WBNAA of patients with RR-MS declined significantly at both the group and individual levels over a 2-year time period common in clinical trials. Because of the small sample sizes required to establish power, WBNAA can be incorporated into future studies.