Inhibition processes are dissociable and lateralized in human prefrontal cortex

The prefrontal cortex (PFC) is known to make fundamental contributions to executive functions. However, the precise nature of these contributions is incompletely understood. We focused on a specific executive function, inhibition, the ability to suppress a pre-potent response. Functional imaging and animal studies have studied inhibition. However, there are only few lesion studies, typically reporting discrepant findings. For the first time, we conducted cognitive and neuroimaging investigations on patients with focal unilateral PFC lesions across two widely used inhibitory tasks requiring a verbal response: The Hayling Part 2 and Stroop Colour-Word Tests. We systematically explored the relationship between inhibition, fluid intelligence and lesion location using voxel-based lesion symptom mapping (VLSM). We found that PFC patients were significantly impaired compared with healthy comparison group (HC) on both suppression measures of the Hayling and on the Stroop, even when performance on a fluid intelligence test was covaried. No significant relationship was found between patients' performance on each Hayling suppression measure and the Stroop, once fluid intelligence was partialled out, suggesting that the two tests may involve different kinds of inhibition. After accounting for fluid intelligence, we found a significant interaction between tests, Hayling or Stroop, and site, left or right, of PFC damage. This finding suggesting lateralized functional organization was complemented and extended by our VLSM results. We found that performance on both Hayling suppression measures significantly relied on the integrity of a similar and relatively circumscribed region within the right lateral PFC, in the right lateral superior and middle frontal gyri. In stark contrast, performance on the Stroop relies on the integrity of left lateral superior and middle frontal gyri. Thus, lesion location, right or left PFC, is critical in producing impairments on two inhibitory tasks loading similarly on verbal control. This suggests that the two suppression measures of the Hayling and the Stroop are likely to assess dissociable components of executive functions, related to anatomically defined and lateralized PFC circuits. Our findings also suggest that inhibition may actually comprise qualitatively different forms with different neural substrates. This has clinical implications for the diagnosis and treatment of disinhibition impairments, a common behavioural problem caused by PFC lesions. Our results highlight the need to assess inhibition using a variety of tasks and to develop different types of treatments.