Fast ripples reflect increased excitability that primes epileptiform spikes

Abstract The neuronal circuit disturbances that drive inter-ictal and ictal epileptiform discharges remain elusive. Using a combination of extra-operative macro-electrode and micro-electrode inter-ictal recordings in six pre-surgical patients during non-rapid eye movement sleep, we found that, exclusively in the seizure onset zone, fast ripples (200–600 Hz), but not ripples (80–200 Hz), frequently occur <300 ms before an inter-ictal intra-cranial EEG spike with a probability exceeding chance (bootstrapping, P < 1e−5). Such fast ripple events are associated with higher spectral power (P < 1e−10) and correlated with more vigorous neuronal firing than solitary fast ripple (generalized linear mixed-effects model, P < 1e−9). During the intra-cranial EEG spike that follows a fast ripple, action potential firing is lower than during an intra-cranial EEG spike alone (generalized linear mixed-effects model, P < 0.05), reflecting an inhibitory restraint of intra-cranial EEG spike initiation. In contrast, ripples do not appear to prime epileptiform spikes. We next investigated the clinical significance of pre-spike fast ripple in a separate cohort of 23 patients implanted with stereo EEG electrodes, who underwent resections. In non-rapid eye movement sleep recordings, sites containing a high proportion of fast ripple preceding intra-cranial EEG spikes correlate with brain areas where seizures begin more than solitary fast ripple (P < 1e−5). Despite this correlation, removal of these sites does not guarantee seizure freedom. These results are consistent with the hypothesis that fast ripple preceding EEG spikes reflect an increase in local excitability that primes EEG spike discharges preferentially in the seizure onset zone and that epileptogenic brain regions are necessary, but not sufficient, for initiating inter-ictal epileptiform discharges.

A B

C D
Supplementary Figure 1: Units with significant increases in peak gaussian smoothed firing rate during high-frequency oscillation (HFO) events exhibit higher HFO event rates.Histograms of the number of units showing significant (f.d.r.corrected p<0.001) increases in peak firing rate in red and number of units with insignificant increases in peak firing rate (blue) as a function of the log10 transformed number of HFO events recorded by the proximal macroelectrode contact for ripples on oscillations (A, RonO), fast ripples on oscillations (B, fRonO), ripples on spikes (C, RonS), fast ripples on spikes (D, fRonS).Note that some units showed insignificant increases in HFO related firing (insignificant) despite relatively high HFO counts.For fRonO 57% of the multi-units were insignifact, whereas 68% of the single units were insignificant.Abbreviations: RonO=ripple on oscillation; fRonO=fast ripple on oscillation; RonS = ripple on spike; fRonS = fast ripple on spike.Supplementary Figure 2: Solitary and pre-epileptiform spike fast ripple on oscillation (fRonO) related increases in peak firing above baseline are greater in microelectrode recordings sorted as multi-unit activity (right) as oppossed to those sorted as single-unit activity (left).In addition the increase in firing rate corresponding to the after-going epileptiform spike are present in the multi-unit activity (right) but absent in the single-unit activity (left).Abbreviations: sec= second; macro= macroelectrode; AP= action potential; fRonO= fast ripple on oscillation Note that a relatively small, but distinct, subpopulation of events occurred within <10 msec of the spike.(C) Plot of the pre-spike fRonO power as a function of latency before the spike.Note that relatively larger powers (see A) were seen across a broad range of latencies.However, the power distribution for fRonO events occurring <10 ms before spikes exhibits relatively higher powers than the distribution of fRonO events with a latency >10 ms.Abbreviations: fRonO=fast ripple on oscillation; ms=millisecond.(left) and fRonS (right) event-unit trials at each measured peri-event latency.Note that most frequently RonS/ fRonS event-unit trials fire maximally at the time of RonS/fRonS onset or within 20 msec after the onset.However, probability of maximal firing is also higher than baseline at 10 msec before RonS/fRonS onset, and marginally higher up to 60 msec before RonS/fRonS onset.Abbreviations: Sec=second; RonS: ripple on spike; fRonS: fast ripple on spike.Histogram of the latency in milliseconds (ms) between the fRonO and the after-going fRonS.Note that a distinct subpopulation of fRonS events followed fRonO events by <10 ms.(C) Plot of the fRonS following fRonO power as a function of the latency between the fRonO and the aftergoing fRonS.Note that relatively decreased powers (see A) were seen across a broad range of latencies.However, the power distribution of fRonS events following a fRonO by <10 ms exhibits relatively higher powers as compared to the fRonO power distribution with latencies >10 ms.The difference between these two power distributions suggests that fRonS following fRonO by <10 ms may be generated by a distinct mechanism.Abbreviations: fRonS=fast ripple on spike; ms=millisecond.after Gaussian smoothing, minus the mean baseline unit firing rate measured 750 msec before the fRonO for each peri-fRonO trial (hfodiff, d.f.=224,266).The random effect terms were patient id, macroelectrode contact ID of the Behnke-Fried electrode, and the unit id to determine that the effects were observed across all units.The fixed effect of the model was whether, in the trial, the fRonO preceded a sharply contoured epileptiform spike, with or without a high-frequency oscilation, within 300 msec ("1" preSpike), or was a solitary fRonO ("0" pre-Spike).Abbreviations CI: confidence interval.Supplementary Table 3: Results of generalized linear mixed effects model for the outcome of the maximum unit firing rate during the fast ripple on oscillation (fRonO) detected in the macroelectrode, (action potentials/sec) after Gaussian smoothing, minus the mean baseline unit firing rate measured 750 msec before the fRonO for each peri-fRonO trial (hfodiff, d.f.=224,266).The random effect terms were patient id, macroelectrode contact ID of the Behnke-Fried electrode, and the unit id to determine that the effects were observed across all units.
Wilkinson-Roger's Notation: hfodiff ~ 1 + preSPIKE + sumu + sumu:preSPIKE +(1|patient) + (1|electrode) + (1|unit) Single units trended toward a lower hfodiff with fRonO compared with multi-units.However, preSpike status resulted in a significantly lower hfodiff for single units compared with multi-units.4: Results of generalized linear mixed effects model for the outcome of the maximum unit firing rate during the fast ripple on oscillation (fRonO) detected in the macroelectrode, (action potentials/sec) after Gaussian smoothing, minus the mean baseline unit firing rate measured 750 msec before the fRonO for each peri-fRonO trial (hfodiff, d.f.=224,266).The random effect terms were patient id, macroelectrode contact ID of the Behnke-Fried electrode, and the unit id to determine that the effects were observed across all units.
The fixed effects of the model were 1) the log10(power) of the trial's fRonO measured in arbitrary units; and 2) whether, in the trial, the fRonO preceded a sharply contoured epileptiform spike, with or without a high-frequency oscilation, within 300 msec ("1" preSpike), or occurred as a solitary fRonO ("0" preSpike).":" refers to interaction term.Abbreviations CI: confidence interval.
Wilkinson-Roger's Notation: hfodiff ~ 1 + power + preSPIKE + power:preSPIKE + (1|patient) + ( 1|electrode) + (1|unit) The increase in the preSpike fRonO associated peak firing rate, with respect to baseline, (hfodiff) as compared to the hfodiff of solitary fRonO is due to the power:preSpike interaction.Additionally, the pre-spike fRonO had higher power than the solitary fRonO (Figure 2F).Despite this, the preSpike status alone correlated with a decreased hfodiff.preSpike status alone should positively correlate with hfodiff if the solitary fRonO group contained more null trials (hfodiff<=0).These null trials would only influence the power:preSpike interaction if they corresponded to fRonO events with a uniform power distribution.Since artifactual fRonO are more likely to be lower power events, from noise, it is less likely that artifact contamination accounts for the smaller hfodiff in the solitary group.d.f.=224,266).The random effect terms were patient id, macroelectrode contact ID of the Behnke-Fried electrode, and the unit id to determine that the effects were observed across all units.The fixed effect of the model was whether, in the trial, the fRonO preceded a sharply contoured epileptiform spike, with or without a high-frequency oscilation, within 300 msec ("1" preSpike) or occurred as a solitary fRonO ("0" preSpike).Abbreviations CI: confidence interval.hfodiff, d.f.=2,603,800).The random effect terms were patient id, macroelectrode contact ID of the Behnke-Fried electrode, and the unit id to determine that the effects were observed across all units.The fixed effect of the model was whether, in the trial, the RonO preceded a sharply contoured epileptiform spike, with or without a high-frequency oscilation, within 300 msec ("1" preSpike), or occurred as a solitary RonO ("0" pre-Spike).Abbreviations CI: confidence interval.hfodiff, d.f.=2,603,845).The random effect terms were patient id, macroelectrode contact ID of the Behnke-Fried electrode, and the unit id to determine that the effects were observed across all units.The fixed effects of the model were 1) the log10(power) of the trial's RonO measured in arbitrary units, and 2) whether, in the trial, the RonO preceded a sharply contoured epileptiform spike, with or without a high-frequency oscilation, within 300 msec (preSpike "1") or occurred as a solitary RonO (preSpike "0") .":" refers to interaction term.Abbreviations CI: confidence interval.
A trialwise comparison of the latency of the maximum Gaussian smoothed unit firing rate relative to ripple (RonS) or fast ripple on spike (fRonS) onset.Shown is a histogram of the number of RonS An examination of fast ripple on spike (fRonS) following fast ripple on oscillation (fRonO) power with increased temporal resolution.(A) Normalized histogram of fRonS event power in macroelectrode recordings.fRonS power was reduced when it followed a fRonO (t-test, p<1e-5, Cohen's d=0.21) (B)

Table 2 :
Results of generalized linear mixed effects model for the outcome of the maximum unit firing rate during the fast ripple on oscillation (fRonO) detected in the macroelectrode, (action potentials/sec)

Table 5 :
Results of generalized linear mixed effects model for the outcome of the mean baseline unit firing rate measured 750 msec before the fast ripple on oscillation (fRonO) for each peri-fRonO trial (bl,

Table 6 :
Results of generalized linear mixed effects model for the outcome of the maximum unit firing rate during the ripple on oscillation (RonO) detected in the macroelectrode, (action potentials/sec) after Gaussian smoothing, minus the mean baseline unit firing rate measured 750 msec before the RonO for each peri-RonO trial (

Table 7 :
Results of generalized linear mixed effects model for the outcome of the maximum unit firing rate during the ripple on oscillation (RonO) detected in the macroelectrode, (action potentials/sec) after Gaussian smoothing, minus the mean baseline unit firing rate measured 750 msec before the RonO for each peri-RonO trial (