Spinal cord repair in MS: does mitochondrial metabolism play a role?

Abstract

Objective: To investigate the mechanisms of spinal cord repair and their relative contribution to clinical recovery in patients with multiple sclerosis (MS) after a cervical cord relapse, using spinal cord (1)H-magnetic resonance spectroscopy (MRS) and volumetric imaging.

Methods: Fourteen patients with MS and 13 controls underwent spinal cord imaging at baseline and at 1, 3, and 6 months. N-acetyl-aspartate (NAA) concentration, which reflects axonal count and metabolism in mitochondria, and the cord cross-sectional area, which indicates axonal count, were measured in the affected cervical region. Mixed effect linear regression models investigated the temporal evolution of these measures and their association with clinical changes. Ordinal logistic regressions identified predictors of recovery.

Results: Patients who recovered showed a sustained increase in NAA after 1 month. In the whole patient group, a greater increase of NAA after 1 month was associated with greater recovery. Patients showed a significant decline in cord area during follow-up, which did not correlate with clinical changes. A worse recovery was predicted by a longer disease duration at study entry.

Conclusions: The partial recovery of N-acetyl-aspartate levels after the acute event, which is concurrent with a decline in cord cross-sectional area, may be driven by increased axonal mitochondrial metabolism. This possible repair mechanism is associated with clinical recovery, and is less efficient in patients with longer disease duration. These insights into the mechanisms of spinal cord repair highlight the need to extend spinal cord magnetic resonance spectroscopy to other spinal cord disorders, and explore therapies that enhance recovery by modulating mitochondrial activity.

Figure 1 Spinal cord N-acetyl-aspartate concentration at each time point Spectroscopic volume of interest between C1 and C3 on coronal images at baseline (A), 1 month (B), 3 months (C), and 6 months (D) in a patient who recovered. Corresponding spectra obtained by LCModel analyses at baseline (E), 1 month (F), 3 months (G), and 6 months (H). The N-acetyl-aspartate (NAA) was 4.1 mM (%SD 12) at baseline (E), 3.6 mM (%SD 11) at 1 month (F), 3.7 mM (%SD 11) at 3 months (G), and 4.0 mM (%SD 11) at 6 months (H). The other peaks near NAA (such as the resonance at 2.1–2.5 ppm) seem to change concordantly, but changes in these peaks were not investigated because they are not reliably quantified in the spinal cord using our methodology.

Figure 2 Longitudinal changes in N-acetyl-aspartate levels Mean N-acetyl-aspartate (NAA) (millimolar) by month with SE in controls, patients who recovered, and patients who did not recover. Changes in NAA seem to be nonlinear. In particular, in patients who recovered during the follow-up period, there was a significant and sustained increase of NAA from baseline to 6 months and from 1 month to 6 months.

Figure 3 Correlations between clinical recovery and either N-acetyl-aspartate changes or disease duration (A) Plot of the raw Expanded Disability Status Scale (EDSS) changes between 6 months and baseline vs the N-acetyl-aspartate (NAA) changes (millimolar) between 6 months and 1 month (R² = 0.57). This graph shows that the increase in NAA after 1 month is associated with greater extent of recovery (i.e., greater EDSS reduction). (B) Plot of the raw EDSS changes between 6 months and baseline vs the disease duration (years) (R² = 0.39). This graph shows that a longer disease duration at baseline is associated with lower extent of recovery (i.e., smaller EDSS reduction). Fitted lines are shown.

Figure 4 Longitudinal changes in spinal cord cross-sectional area Mean spinal cord cross-sectional area (square millimeters) by month with SE in controls, patients who recovered, and patients who did not recover. Both patients who recovered and those who did not showed a significantly greater decline of spinal cord area over time compared with controls, which remained significant when models were repeated with data collected after 1 month and adjusted for baseline cord swelling.