Neurophysiological investigations of drug resistant epilepsy patients treated with vagus nerve stimulation to differentiate responders from non‐responders

Abstract

Background and purpose: In patients treated with vagus nerve stimulation (VNS) for drug resistant epilepsy (DRE), up to a third of patients will eventually not respond to the therapy. As VNS therapy requires surgery for device implantation, prediction of response prior to surgery is desirable. It is hypothesized that neurophysiological investigations related to the mechanisms of action of VNS may help to differentiate VNS responders from non-responders prior to the initiation of therapy. Methods: In a prospective series of DRE patients, polysomnography, heart rate variability (HRV) and cognitive event related potentials were recorded. Polysomnography and HRV were repeated after 1 year of treatment with VNS. Polysomnography, HRV and cognitive event related potentials were compared between VNS responders (≥50% reduction in seizure frequency) and non-responders. Results: Fifteen out of 30 patients became VNS responders after 1 year of VNS treatment. Prior to treatment with VNS, the amount of deep sleep (NREM 3), the HRV high frequency (HF) power and the P3b amplitude were significantly different in responders compared to non-responders (P = 0.007; P = 0.001; P = 0.03). Conclusion: Three neurophysiological parameters, NREM 3, HRV HF and P3b amplitude, were found to be significantly different in DRE patients who became responders to VNS treatment prior to initiation of their treatment with VNS. These non-invasive recordings may be used as characteristics for response in future studies and help avoid unsuccessful implantations. Mechanistically these findings may be related to changes in brain regions involved in the so-called vagal afferent network.

Biomarkers of seizure response to vagus nerve stimulation: A scoping review

“…They observed an increase in P3 amplitude only in VNS responders 29 . The same group reported similar findings in the study published by Wostyn et al 36 Hödl et al later studied the P3b component, a subcomponent of the P3, and reported that responders had lower P3b amplitudes at baseline during VNS ON conditions compared to VNS OFF conditions 31 …”

“…The second class of biomarkers pertained to patterns of electrophysiological brain activity, 22,23,[32][33][34][35][36][37][24][25][26][27][28][29][30][31] which often yielded inconsistent results. All these studies investigate EEG or MEG as a measure of electrical brain activity.…”

“…29 The same group reported similar findings in the study published by Wostyn et al 36 Hödl et al later studied the P3b component, a subcomponent of the P3, and reported that responders had lower P3b amplitudes at baseline during VNS ON conditions compared to VNS OFF conditions. 31 Two studies investigated the potential of the pdBSI, a measure of EEG symmetry, as an objective measure of VNS efficacy; however, whereas the first reported higher pdBSI values across several frequency bands as being associated with worse VNS response, 33 a subsequent validation study refuted these findings and found no significant association between pdBSI and VNS responsiveness. 34 Hödl et al used polysomnography to study the association between sleep waves and VNS responsiveness, and noted statistically significant differences in preoperative deep sleep (NREM 3) patterns in responders compared to nonresponders.…”

“…Six studies investigated preoperative HRV, cytokine levels, and tryptophan metabolite levels as systemic biomarkers of VNS efficacy. 31,[40][41][42][43][44] Using ECG, Liu et al reported that VNS nonresponders had lower preoperative HRV compared to responders. 41 Measures of HRV included the (1) time-domain measures RMSSD (the root mean square of the successive differences between heartbeats) and pNN50 (the proportion of successive heartbeats differing by >50 milliseconds divided by the total number of beat-tobeat intervals), which reflect cardiac parasympathetic control of heart rate; (2) frequency domain measure HF (high-frequency power for the frequency range 0.15-0.40 Hz), which is mediated by respiration and is a marker of vagal activity; and (3) nonlinear domain measure SD1 (the standard deviation of instantaneous beat-to-beat interval variability), which reflects short-term HRV caused by respiration.…”

“…42 Finally, Hödl et al reported lower HF power in responders compared to nonresponders, both at baseline before VNS and 1 year after VNS therapy. 31 In a study of inflammatory biomarkers of VNS response, Aalbers et al compared preoperative and postoperative plasma cytokine levels between patients randomly assigned to high-output or low-output VNS stimulation. Whereas they did not find significant differences between groups, they found that lower baseline plasma levels of interleukin-6 (IL-6), a pro-and anti-inflammatory cytokine that mediates fever and the acute phase reaction, was associated with greater reduction in seizure frequency in both groups.…”