Abstracts

Rationale: Individuals with bilateral damage to medial temporal lobe structures show marked memory deficits that are amplified on tasks requiring associative binding operations (e.g., associative recognition memory tasks that necessitate the creation of links between individual pieces of information) relative to tasks that do not (e.g., single item recognition memory). Individuals with unilateral medial temporal damage or dysfunction, such as patients who underwent a temporal lobe excision (TLE) for the treatment of their epilepsy, show memory deficits that are specific to the material preferentially processed by the damaged hemisphere (e.g., left hemisphere for verbal material). This material-specific deficit was shown on item memory tasks and it is unclear whether it also applies to tasks requiring associative binding operations. Very few studies have investigated this question and their findings are mixed. Our goal was to investigate the performance of patients with unilateral TLE on measures that rely or not on associative binding operations. Methods: Twelve patients with language dominant TLE, 12 patients with non-dominant TLE and 12 age-matched controls participated. We used a word-pair recognition paradigm composed of two tasks: 1) a pair recognition task in which participants had to identify pairs containing two studied words irrespective of their pairings, and 2) an associative identification recognition task in which participants were required to discriminate between studied and novel pairings. From performance on these tasks, we derived measures that rely on associative binding operations (i.e., associative identification, associative reinstatement, recollection) and measures that do not (e.g., item memory and familiarity). Results: Our results showed that despite the verbal nature of our tasks, all measures that rely on associative binding operations were impaired in patients with unilateral TLE, irrespective of the side of damage. Familiarity, however, was impaired solely following language dominant resection, which is in keeping with the material-specificity principle. Conclusions: From a neurobiological perspective, our results suggest that, in the context of the current study, measures that rely on associative binding operations are subserved by bilateral MTL regions (likely the hippocampus) while a measure that does not require such operations (familiarity) is supported by unilateral MTL regions (likely the perirhinal cortex).