extent in the primary sensory cortex and premotor area. , , This is in keeping with the anatomy of the corticospinal tract which originates predominantly from these regions of the neocortex.34 These observations are also in accordance with the topographical pathology of ALS in which the precentral gyrus, followed by the postcentral gyrus, is most severely affected.1 Inconsistent abnormalities in NAA between MRS studies of the frontal lobe can at least in part be explained by variable affliction of the frontal lobes in this disease.21,26,29
The accuracy of MRS in aiding the diagnosis of ALS has been assessed in few studies. Chan8 found the sensitivity and specificity to be 79% and 90% respectively with a motor cortex NAA/Cr cut-off of 2.5 (2.12 standard deviations below the control mean). With a 2.5 standard deviation cut-off for NAA/Cho Pohl35 found the sensitivity to increase with the degree of certainty in the diagnosis from 47% to 64% (specificity was not provided).
Abnormal NAA indices have shown correlations with clinical parameters such as strength16,36 and severity of some UMN signs.8,14,15,18,21,22,27,29,36 Diminished frontal lobe NAA/Cr was found to correlate with cognitive dysfunction.26 Correlations have not been demonstrated with spasticity or muscle stretch reflexes.36 Most,17,23,36 but not all,16 note a relationship to the site of onset of symptoms (bulbar versus limb), but not symptoms or disease duration.16'17'22'30'36
Reports of associations with disease severity measures have been favourable. Correlations have been noted with the ALS Severity Scale,36 Norris Scale,17'26 and Jablecki Scale'16 but not with the ALSFRS.17 The ALSFRS bulbar subscores have been correlated with glutamate/glutamine but not NAA spectral intensities from the brainstem.33
The lack of uniformity with clinical correlations is likely in part due to the fact that disability scales usually are influenced by both LMN and UMN dysfunction and some elements are heavily weighted for disability secondary to LMN impairment. Weak or absent correlations with measures of UMN function' such as reflex score and spasticity scales (eg. Ashworth)' is not surprising since these are imprecise and ordinal in nature.
Was this article helpful?