Abstracts

Rationale: A few studies have described the long-term course of cognitive changes after anteromedial temporal lobectomy (ATL) for refractory epilepsy. How pre-surgical risk factors, post-surgical seizures and anti-epilepsy drugs (AED) combine to influence acute post-surgical changes and later progression has not been described.Methods: We analyzed scores on a standard cognitive test battery before, and at 2 and 5 years after ATL within a 7-center, prospective, observational study. Seizure activity and AED use were queried quarterly. Multivariate, individual growth curve analysis tested the independent association of pre-surgical cognitive risk factors, AED use and seizure control with early (pre-ATL to 2 years) and late (2 years to 5 years) cognitive change. Results: 221/318 subjects (70%) had at least one post-surgical evaluation. There were no IQ changes overall when effects of post-surgical AEDs and seizures were controlled. Reliable improvement occurred as often as decline (15% vs. 12%). Monotherapy was associated with higher scores on IQ and other time-dependent tests across all sessions. Reducing to monotherapy increased FSIQ by 4.7 points at 2 years (11% with reliable (>=11 point) improvement) vs. 0.5 points in non-reducers (4% with reliable improvement) (p < .01). Similar trends were seen at 5 years and on other time-dependent tests. Effects of seizure control were rare. Proportions of dominant subjects who declined vs. improved by >=1 SD differed between those with normal MRI or bilateral MTS (54% vs. 15%) and those with only ipsilateral MTS (30% vs. 45%)(p <. 01). Late verbal memory decline >= 1SD occurred in 25% of subjects, associated only with contralateral MTS (p < .01), independent of side of surgery or ipsilateral MTS. Reliable naming decline was more common in dominant vs. non-dominant cases (42% vs. 17%)(p=.0001), regardless of MRI findings. Non-verbal memory did not change in any group. Memory and naming were unrelated to AEDs or seizure control.Conclusions: Improvements after ATL in non-memory, cognitive functions have been attributed to test practice effects, elimination of ‘nociferous’ cortex or reducing AEDs. In this study, IQ decline or improvement beyond that explained by practice was uncommon (12-15%), but improvement was more likely with reducing to monotherapy. AED reduction, not seizure control, appears to be the primary source of improvement in non-memory functions after ATL. AED reduction primarily affects processing speed, as effects were greatest on time-dependent tests. The common occurrence of early verbal memory change more so in dominant cases with normal MRI or bilateral MTS is consistent with prior studies suggesting that decline reflects loss of ‘functionally adequate’ medial temporal structures. The novel observation that late verbal memory changes are associated with contralateral MTS in both dominant and non-dominant cases suggests that late memory changes are influenced by the ability of residual structures to mediate function (i.e.,‘functional reserve’). This information may be useful for determining short- and long-term cognitive prognosis following ATL.