CCEP N1 reciprocity was weaker and inconsistent across eight SEEG patients

We tested whether cortico-cortical evoked potential (CCEP) N1 response presence is reciprocal in an independent center. The analysis used expert-checked N1 rows and contact coordinates for eight SEEG patients from the STReEF cohort (OpenNeuro ds005448, UMC Utrecht). A prespecified contact-level analysis included 12,181 bidirectional dyads at least 5 mm apart. The fixed-effect adjusted odds ratio (OR) for a reverse response when a forward response was present was 1.75 (95% CI 1.52–2.01; p = 4.0 × 10⁻¹⁵); seven of eight patient estimates were above 1. However, heterogeneity was high (I² = 0.81), one patient showed a significant reverse association (OR 0.30, permutation p = 0.002), and the exact patient-level sign-flip test was not significant (p = 0.23). The prespecified support rule was therefore not met. The unadjusted Mantel–Haenszel OR was 4.98. A secondary ECoG analysis of five patients gave an adjusted OR of 2.39 (95% CI 2.10–2.73). The SEEG estimate was smaller than the OR of 3.42 reported in the earlier ds004774 study (z = −3.11, p = 0.002). These results agree with reciprocity in direction at the dyad level, but the effect was about half the earlier estimate and was not consistent across patients. Absence of an N1 row was treated as no response, and responses were aggregated to contacts.

cortico-cortical evoked potentials · N1 response · response reciprocity · stereo-EEG · intracranial stimulation · OpenNeuro ds005448 · replication