Abstracts

Rationale: Heart rate variability (HRV) is influenced by autonomic function and a decrease in HRV may predict a poor prognosis in cardiac diseases. Although periictal alterations in HRV, including a reduction in RMSSD (root mean square of successive differences) after generalized convulsive seizures (GCS), have been reported in people with epilepsy (PWE), substantial interindividual variability exists, and the mechanisms responsible for these alterations are not clear. The relationship between HRV and postictal respiratory depression has also not been clearly defined despite the latter’s relevance to sudden unexpected death in epilepsy (SUDEP). We recently showed that low interictal central CO₂ chemosensitivity as measured by the hypercapnic ventilatory response (HCVR) was associated with worse respiratory depression after GCS. Because autonomic function is modulated by the respiratory network, we speculated that central CO₂ chemosensitivity would be associated with HRV, and that this relationship would be strongest when cortically mediated influences on breathing are minimal: during sleep and after a GCS. We therefore analyzed the relationship between HCVR and HRV during the interictal and postictal periods.

Methods: Retrospective analysis of Sainju et al. Epilepsia 2019 PWE admitted to the epilepsy monitoring unit (EMU) for video EEG study, and who experienced GCS. EEG was recorded using scalp electrodes with a standard 10-20 placement system. Interictal HCVR was measured using a modified hyperoxic rebreathing technique and an Ultima PFX Respiratory Gas Analyzer. HRV was calculated during 5-minute artifact free recordings from interictal resting wakefulness, REM, and non-REM sleep, and from the postictal period. HRV was expressed by time-domain (SDNN, RMSSD, pNN50), frequency-domain (VLF, LF, HF, LF/HF ratio) and nonlinear measures (SD1, SD2). Univariate analyses were performed using Spearman’s correlation test.

Results: Both a GCS and an interictal HCVR were available for 24 patients; with median age 34.5 [26-45] years, median duration of epilepsy 12 [5-25] years, and median number of current AEDs 2 [2-3]. Three patients had no REM sleep recorded. HCVR slope varied from 0.17 to 5.10 (median 1.89) L/min/mm Hg. Heart rate awake, NREM, REM, and post-ictal was 72.6 [67.2-75.7], 64.5 [56.7-69.3], 64.8 [57.8-72.5], and 112.4 [96.2-123.3] (bpm). RMSSD awake, NREM, REM, and post-ictal was 31.7, [15.7-74.2], 52.6 [21.1-86.7], 41.3 [27.5-78.3], and 24.7 [10.3, 56.9] (ms). Following GCS, HCVR slope was inversely correlated to RMSSD (rho -0.46, p=0.02), SD1 (rho -0.46, p=0.02), and HF (ms²) (rho -0.42, p=0.04). HCVR slope was positively correlated to VLF (%) (rho 0.43, p=0.04). There was no significant correlation between HCVR and interictal HRV measures in wakefulness, REM, or NREM sleep.

Conclusions: In patients with medically refractory epilepsy, low interictal HCVR correlated with high HRV after GCS. This may reflect variable serotonergic modulation of autonomic function. Low central CO₂ chemosensitivity and high parasympathetic activity may increase the risk of severe respiratory depression, bradycardia, and SUDEP after GCS.

Funding: Please list any funding that was received in support of this abstract.: 1R01NS113764, U54TR001356, U01 NS090414.