Abstracts

Rationale: Brain responsive neurostimulation through 2 years is not associated with memory or naming decline, and may be associated with modest performance improvement that varies as a function of seizure onset/stimulation site1. However, it is unknown whether subjects with either low or high performance levels at baseline have different patterns of longitudinal change. In the present study, we report the pattern of change based upon level of baseline performance to evaluate if similar patterns of interval change are present in subjects with high and low baseline function. [1] Loring D.W., Kapur R., Meador K.J., Morrell M.J. Differential neuropsychological outcomes following targeted responsive neurostimulation for partial-onset epilepsy. Epilepsia 2015; 56: 1836?"1844. Methods: Naming (Boston Naming Test, BNT) and verbal memory (Rey Auditory Verbal Learning Test) data were collected from subjects participating in the open-label arm of a randomized controlled trial of responsive neurostimulation with the RNS System (Mountain View, CA). For both neocortical and mesial temporal lobe (MTL) seizure onset groups, naming and memory scores were characterized as "high" or "low" based upon a median split derived from each group independently, and the longitudinal change through 2 years was evaluated using Generalized Estimating Equation (GEE). Results: 93 subjects with seizure onsets in the MTL and 78 subjects with seizure onsets of neocortical origin contributed to this analysis. There were no significant declines in either high or low baseline groups for either verbal learning or on the BNT. Significant performance improvement in verbal learning scores was present in MTL subjects with baseline scores lower than the median (p=0.01), with no significant change present in MTL subjects with higher learning scores. A similar pattern was present in the neocortical group, although this did not reach statistical significance (p=0.07). For subjects with MTL seizure onset, there was no significant improvement in naming performance in either the low or high baseline BNT groups. In contrast, naming improvements were observed in both low and high baseline BNT groups in subjects with neocortical onsets (both p < 0.0001), with no statistically significant between group difference. Conclusions: These results indicate that the maximum improvement in verbal memory is derived for patients with poorer memory scores at baseline, a pattern that is present in both MTL and neocortical groups. In contrast, naming improvement is observed only in neocortical subjects, but improvement is present independent of baseline performance levels. Importantly, there is no concomitant decline in verbal memory for subjects in the "high" baseline group, which demonstrates that the changes over time are not related to regression to the mean associated with performance improvements in one group negating declines associated with stimulation in higher functioning subjects. These data provide further confirmation that long-term (2-year) brain responsive neurostimulation is not associated with cognitive decline for subjects with either higher or lower performance levels at baseline. Funding: None.