Left extratemporal coverage was associated with lower recall in RAM SEEG patients, while temporal coverage was not

In 254 patients with stereo-electroencephalography (SEEG) from the Restoring Active Memory (RAM) study, lower verbal free-recall rate was associated with a greater fraction of left extratemporal bipolar pairs, but not with a greater fraction of left temporal pairs. The preregistered analysis used the same patients and covariates as the prior overall laterality analysis and classified pairs by their MNI midpoint. Left extratemporal fraction, controlling for left temporal fraction, had partial Spearman rho = -0.188 (95% CI -0.318 to -0.056; Holm-adjusted p = 0.0075). Left temporal fraction, controlling for left extratemporal fraction, had rho = -0.067 (95% CI -0.196 to +0.069; Holm-adjusted p = 0.532). The difference between standardized rank-regression coefficients was not significant (+0.124, 95% CI -0.095 to +0.339; Holm-adjusted p = 0.533); therefore, the analysis does not show that the extratemporal association is larger than the temporal association. Left medial temporal coverage also did not predict recall, and four alternative temporal-box limits gave the same pattern. These between-patient associations use MNI boxes as proxies for lobe and suspected seizure onset; the data contain no seizure-onset-zone, language-dominance, lesion, or neuropsychology labels. The findings support controlling for left extratemporal coverage in between-patient RAM recall-rate analyses, but do not establish a causal effect.

verbal memory · free recall · stereo-electroencephalography · SEEG · epilepsy · RAM study · temporal lobe · extratemporal coverage · MNI · partial Spearman correlation