hide
Free keywords:
-
Abstract:
Introduction:
Neuroinflammation after CNS injury is a common feature. Next to its role during recovery processes after CNS injury, a higher risk to develop dementia has been also attributed to neuroinflammation after e.g. traumatic brain injury [1]. In patients with spinal cord injury (SCI), a large epidemiological study reported that patients with SCI are at higher risk of dementia than age- and sex-matched controls [2]. Experimental evidence has shown in the mice model of chronic SCI that trauma was associated with chronic neuroinflammation and also impaired neurogenesis within the dentate gyrus of the hippocampus [3] along with decreased BDNF levels [4, 5]. The magnitude of which correlated with cognitive impairments of spatial navigation, object recognition and memory function [3]. To date, structural and metabolic changes of the hippocampus after human SCI have not been explored yet. This pilot study therefore investigated whether metabolic changes in the hippocampus occur after human SCI and whether these changes are related to memory performance.
Methods:
All participants underwent MRS measurement on a 3T scanner (Philips, Netherlands) with a 8 channel SENSE coil. T1-weighted images (1x1x1mm3) were used to place the spectroscopic voxel (16x10x12mm3) at the right hippocampus and the metabolite cycling (MC) PRESS technique [6] was applied as localization sequence. Each MRS measurement contained 256 signal averages and the data were fitted using LC Model [7]. Total N-Acetyl-Aspartate (tNAA), total creatine (tCr), choline containing compound (tCho), myo-Inositol (mI), and glutamate/glutamine (Glu+Gln=Glx) were quantified (CRLB<25%). All participants were assessed with the visual and verbal working memory test [8] on immediate (time-point 1) and mid-term (recalling of approx. 2 hours after learning, time-point 2) memory function. Statistical analyses were performed using R (R Core Team, 2016, Version 3.4.3). Group differences on metabolites were assessed with the Kruskal-Wallis test and Spearman's rank correlations were used to investigate associations between metabolites and memory function.
Results:
So far we recruited 6 chronic SCI patients (4 men, age [median, (range)]: 58 (40 – 75) years, years since injury: 11.5 (4-31) years) and 10 healthy controls (7 men, age: 47.0 (35 – 68) years). All investigated metabolites in the hippocampus were numerically lower in SCI patients compared to healthy controls. Glx was significantly lower in SCI patients compared to healthy controls (p=0.045) (see Fig. 1). Additionally, Glx levels were negatively associated with visual memory scores for all participants (time-point 1: p=0.036, R2= 0.277; time-point 2: p=0.028, R2=0.30; see Fig. 2AB), but in particular for the SCI patients (time-point 1: p<0.001, R2=1; time-point 2: p=0.049, R2=0.659; see Fig. 2CD).
Supporting Image: ohbm19_hippocampus_v2_glx.png
Conclusions:
This is the first study showing that metabolite levels are lower in the hippocampus after traumatic SCI. Although the number of patients was small, insights to the metabolic profile of hippocampal areas are revealed. In particular, we found altered levels of Glx, which were associated with decreased memory function. N-methyl-d-aspartate subtype glutamate receptors (NMDA) are required for long-term potentiation and long-term depression of hippocampal CA1 synapses, the proposed cellular substrates of learning and memory [9]. Our results indicate that SCI does not only lead to degeneration and demyelination of primarily affected tracts, but it is likely that trans-neuronal complex remodeling of primarily unaffected brain regions occurs which might also affect memory function. This pilot study therefore provides unbiased, quantitative readouts of hippocampal changes after SCI relating to memory function, which might serve as future biomarkers.
