Abstracts

Rationale: Based on decades of research, traditional neurocognitive models of memory maintain that episodic memory is mediated by the hippocampal region and semantic memory processes (including naming) are supported by lateral temporal neocortex. Naming and verbal memory frequently decline following left (language dominant) anteromedial temporal lobe resection (AMTLR), with verbal memory decline presumed to be a direct consequence of hippocampal removal, and naming decline attributed to resection of lateral temporal cortex. Potentially at odds with traditional memory models, however, are recent findings that suggest a relation between visual object naming and the integrity of the hippocampus. Visual naming has been the gold standard in the assessment of naming; however, recent studies suggest that auditory description naming has greater sensitivity in documenting naming deficits, and cortical stimulation studies have demonstrated a neuroanatomical distinction between auditory and visual naming in lateral temporal cortex. Recent work also suggests that in patients with medial TLE, auditory naming cortex tends to be located slightly posterior to the anterior temporal cortex removed with AMTLR. We speculated that unlike visual naming, the hippocampal region is not involved in auditory naming, and therefore, hypothesized that left AMTLR would not result in auditory naming decline, despite decline in visual naming and verbal memory. Methods: Subjects were 45 pharmacologically intractable medial TLE patients (25 left, 20 right) who underwent ipsilateral AMTLR. Auditory naming, visual naming and verbal memory were tested before surgery and approximately one year postoperatively. Within-group differences between pre and postoperative naming and memory scores were assessed via paired sample T tests. Results: Consistent with our hypothesis, left AMTLR patients exhibited significant decline on four of five of visual naming measures and both short and long delay verbal memory measures, yet showed no decline on any measures of auditory naming. As expected, right AMTLR resulted in no decline on naming or verbal memory measures. Conclusions: Results carry potentially significant clinical implications; rather than discussing risks of global word finding decline, patients considering left AMTLR can be counseled regarding the discrepancy in risks between visual and auditory naming decline, the latter of which might better reflect everyday naming function. The coupling of verbal memory and visual naming in their dependency on the hippocampal region, together with the dissociation between visual and auditory naming, suggests that the neuroanatomical overlap between episodic and semantic memory is incomplete and possibly modality specific.