Cognitive deterioration and functional compensation in ALS measured with fMRI using an inhibitory task

Abstract

Amyotrophic lateral sclerosis (ALS) is characterized by degeneration of upper and lower motor neurons, resulting in progressive weakness and muscle atrophy. Recent studies suggest that nondemented ALS patients can show selective cognitive impairments, predominantly executive dysfunction, but little is known about the neural basis of these impairments. Oculomotor studies in ALS have described deficits in antisaccade execution, which requires the implementation of a task set that includes inhibition of automatic responses followed by generation of a voluntary action. It has been suggested that the dorsolateral prefrontal cortex (DLPFC) contributes in this process. Thus, we investigated whether deterioration of executive functions in ALS patients, such as the ability to implement flexible behavior during the antisaccade task, is related to DLPFC dysfunction. While undergoing an fMRI scan, 12 ALS patients and 12 age-matched controls performed an antisaccade task with concurrent eye tracking. We hypothesized that DLPFC deficits would appear during the antisaccade preparation stage, when the task set is being established. ALS patients made more antisaccade direction errors and showed significant reductions in DLPFC activation. In contrast, regions, such as supplementary eye fields and frontal eye fields, showed increased activation that was anticorrelated with the number of errors. The ALS group also showed reduced saccadic latencies that correlated with increased activation across the oculomotor saccade system. These findings suggest that ALS results in deficits in the inhibition of automatic responses that are related to impaired DLPFC activation. However, they also suggest that ALS patients undergo functional changes that partially compensate the neurological impairment.

Keywords: amyotrophic lateral sclerosis; antisaccade; cognitive control; fMRI; prefrontal cortex; task set.

Figure 1.

Behavioral paradigm. Representation of stimuli and timing of events for the four trial types. Trials were pseudo-randomly presented and intermixed with periods of fixation on the neutral fixation stimulus that lasted 1.5, 3, and 4.5 s. Arrows indicate the correct saccade directions for the saccade trials and were not actually displayed. Fixation-only trials are not shown in this figure.

Figure 2.

Eye movement behavioral data. A, Sample eye traces depicting correct antisaccade trials and an erroneous antisaccade trial (direction error) followed by a correction. B, Cumulative probabilities of saccade distributions for the two groups (pooled SRT across subjects). Positive Y values indicate correct saccades, whereas negative Y values indicate direction errors; dashed lines indicate antitrials; solid lines indicate protrials; black lines indicate control; red lines indicate ALS patients. Gray shaded region represents the region categorized as “express saccades” (90 ≤ SRT ≤ 160 ms). Asterisks indicate significant shifts in error rates between the control group and the ALS group. C, Mean percentage direction errors (initial saccade away from target on prosaccade trial, toward target on antisaccade trial). D, Mean SRTs on correct trials. E, Mean percentage of express saccades (90–160 ms). F, Mean intrasubject CVSRT. Error bars indicate SEM. †p < 0.05, significance for group × task interactions only. ‡p < 0.01, significance for group × task interactions only. *p < 0.05. **p < 0.01, significance for main effects of task.

Figure 3.

Saccade network. Contrast map of combined correct prosaccade trials and antisaccade trials corrected using false discovery rate at p < 0.01 (t value = 5.0, df = 11). The identified ROIs were cluster-corrected across the population of voxels with p < 0.05 (49 contiguous voxels, as estimated by Brain Voyager's Cluster-level Statistical Threshold Estimator with 1000 iterations).

Figure 4.

ROI activation time course. Average activity time courses in the preparatory network ROIs during pro (solid lines) and anti (dashed lines) for control (black) and ALS patients (red). Gray bar represents the time points used for subsequent analysis.

Figure 5.

ROI procatch and anticatch trials analysis. Group × condition interaction analysis using the mean β weight peak activity during procatch and anticatch trials for control and ALS groups. Note the heightened activity increase in the anticatch condition in all areas in the ALS group. ACC also showed a significant activity increase in the anticatch condition; however, the group × condition interaction did not reach significant levels. Error bars indicate SE. *p < 0.05.

Figure 6.

Correlations of activation in selected ROIs of the saccade network with SRT. Pearson's correlations between mean β weight peak activity during anticatch trials and SRT during antisaccades in all ALS patients. Note how individuals with a larger average activity during the preparatory phase in many areas, including DLPFC, show faster reaction times when executing correct antisaccades.

Figure 7.

Direction errors analysis. Left, Mean β weight peak activity during correct and error trials in the antisaccade task for the ALS group. Note the significant reduced activity in DLPFC during error trials. Right, Correlations between mean β weight peak activities during antisaccade error trials with the percentage of errors. Note how the increased activity in the insula, SEF, and FEF correlated with fewer errors. Error bars indicate SE. *p < 0.05.