Decision time and response accuracy in a conditional motor learning task are impaired independently in unilateral temporal lobe-resected patients.

Learning and memory of declarative knowledge and relational information are dependent on the integrity of medial temporal lobe (MTL). Numerous studies suggest that left lobectomy impairs verbal memory while right lobectomy impairs non-verbal memory. In order to instrumentally quantify material-specific memory impairment after temporal lobe excision, we compared, using a computerized conditional motor associative learning task, patients with surgically treated drug-resistant temporal lobe epilepsy to age-matched controls. We enrolled seven epileptic patients with left (LTR), seven with right (RTR) temporal lobe resection and fourteen controls. During the task, abstract visual stimuli had to be associated, by trial and error, with a spatially oriented joystick motor response. Response and decision time were analyzed. Statistical analysis disclosed that the learning curve slopes of both RTR and LTR patients were significantly shallower compared to controls, LTR patients needed a number of test trials significantly increased compared to RTR patients and controls, the average probability of success in the test trials was significantly lower in LTR patients compared to RTR patients and controls, and RTR patients' decision times were significantly longer than LTR patients and controls. The results suggest that RTR patients, using the preserved verbalization strategy, achieved higher learning scores than LTR patients, which were forced to use a visuo-spatial representation of the stimuli-response association. Accordingly, RTR patients were significantly slower, compared to LTR patients and controls, indicating that processes involving recall were partially impaired, and non-canonical networks for executing a non-verbal task could be in action.