Object naming after epilepsy surgery in the dominant left temporal lobe: risk factors, time course and long-term outcome

Background: Deterioration in naming function is a common sequelae after epilepsy surgery in the language-dominant temporal lobe but information on recovery and long-term outcome is scarce. We, therefore, assessed short-term and long-term outcome of object naming in patients undergoing surgery in the temporal lobe and determined factors affecting deterioration and recovery of naming function.

Method: Object naming (Boston naming test) before surgery, at early follow-up (FU, 6-12 months) and late FU (≥2 years) was assessed in people with epilepsy (PWE) undergoing resections in the language-dominant left and non-dominant right temporal lobe.

Results: Sixty-six patients with left temporal lobe epilepsy (LTLE) and 87 control patients with right temporal lobe epilepsy (RLTE) were included. At early FU, 28 patients with LTLE (42%) and three patients with RTLE (3%) showed a significant naming decline. In patients with LTLE, risk for deterioration increased with lower verbal memory before surgery, older age at seizure onset and was particularly high with posterior temporal resections (≥40 mm from the temporal pole) and seizure onset >16 years. Of the patients with LTLE with early naming decline, 11 patients (39%) recovered fully in their naming abilities at late FU, averaging almost 10 years. Recovery was associated with the degree of postoperative naming decline at early FU. PWE with a decline of less than 10 items (<20%) had a good prognosis of recovery at late FU. Postoperative seizure control had no significant effect on recovery.

Conclusions: In our cohort, less than 50% of PWE showed significantly deteriorated naming function after resection of the dominant temporal lobe. If a decline occurred, it appeared to recover to a certain degree and remained as a permanent deficit in 26% of the patients. Long-term outcome of visual object naming can be predicted by the degree of early postoperative decline.