Impaired neurovascular coupling to ictal epileptic activity and spreading depolarization in a patient with subarachnoid hemorrhage: possible link to blood-brain barrier dysfunction

Abstract

Spreading depolarization describes a sustained neuronal and astroglial depolarization with abrupt ion translocation between intraneuronal and extracellular space leading to a cytotoxic edema and silencing of spontaneous activity. Spreading depolarizations occur abundantly in acutely injured human brain and are assumed to facilitate neuronal death through toxic effects, increased metabolic demand, and inverse neurovascular coupling. Inverse coupling describes severe hypoperfusion in response to spreading depolarization. Ictal epileptic events are less frequent than spreading depolarizations in acutely injured human brain but may also contribute to lesion progression through increased metabolic demand. Whether abnormal neurovascular coupling can occur with ictal epileptic events is unknown. Herein we describe a patient with aneurysmal subarachnoid hemorrhage in whom spreading depolarizations and ictal epileptic events were measured using subdural opto-electrodes for direct current electrocorticography and regional cerebral blood flow recordings with laser-Doppler flowmetry. Simultaneously, changes in tissue partial pressure of oxygen were recorded with an intraparenchymal oxygen sensor. Isolated spreading depolarizations and clusters of recurrent spreading depolarizations with persistent depression of spontaneous activity were recorded over several days followed by a status epilepticus. Both spreading depolarizations and ictal epileptic events where accompanied by hyperemic blood flow responses at one optode but mildly hypoemic blood flow responses at another. Of note, quantitative analysis of Gadolinium-diethylene-triamine-pentaacetic acid (DTPA)-enhanced magnetic resonance imaging detected impaired blood-brain barrier integrity in the region where the optode had recorded the mildly hypoemic flow responses. The data suggest that abnormal flow responses to spreading depolarizations and ictal epileptic events, respectively, may be associated with blood-brain barrier dysfunction.

Figure 1

Cluster of SDs over a period of 1 h on day 4 after aSAH. Traces 1–3 show the monopolar recordings of the DC-ECoG at electrodes 3, 5, and 6 in the frequency range below 45 Hz. Four SDs, identified by large, negative slow potential changes (SPCs) propagate from electrode 6 (E6) to electrodes 5 and 3 (E5, E3) (direction of propagation indicated by black arrows). Traces 4–6 show the parallel monopolar recordings of the near DC/AC-ECoG at electrodes 3, 5, and 6 in the frequency range between 0.01 and 45 Hz. Here again, black arrows point into the direction of propagation of the SDs. In the high-frequency, alternating current range of the ECoG (bandwidth 0.5–45 Hz), SD causes silencing of spontaneous activity, that is, spreading depression. This is observed as a rapid reduction in ECoG power. Traces 7–9 display the bandpass filtered ECoG recorded at electrodes 3, 5, and 6. The integral of power of the high frequency signal (traces 13–15) is used for visual enhancement of the amplitude loss and helps to determine the duration of the depression period beginning at the initial decrease and ending at the start of the recovery, as described previously (Dreier et al., 2006). Note that the spreading depression of spontaneous activity propagates in a similar fashion as the SPCs from E6 to E3. Trace 16 displays changes in p_tiO₂, measured with the intraparenchymal oxygen sensor (see Methods section). Note that the first two SDs have essentially hyperoxic responses (green single asterisks, mild hypoxia indicated by green single arrows). The third SD causes a profound hypoxic response (triple arrows) followed by hyperoxia (triple asterisks). Optode 3 (trace 17) shows a hyperemic response to the SD recorded at electrode 3, whereas optode 5 displays a clearly hypoemic response to the third SD. This is coincident with the p_tiO₂ response described above. Note, however, that there is a delay in rCBF decrease after p_tiO₂ decrease. This is likely due to the spatial distance between the Licox probe (located between electrodes 3 and 4) and optode 5 (located near electrode 5).

Figure 2

Propagation of ictal epileptic activity during a recording period of 1 h on day 11 after aSAH. Here, ictal epileptic field potentials (iEFPs) (indicated by black asterisks in traces 1–6) coincide with sizable changes in the high-frequency AC-ECoG (bandwidth 0.5–45 Hz, displayed in traces 7–15). Similar to SD, ictal epileptic activity is associated with negative peaks in the DC/AC-ECoG but note that their amplitudes are much smaller than those of SDs. The amplitude of the spontaneous bandpass filtered high frequency ECoG activity (0.05–45 Hz) as well as the amplitudes of the power and the integral of the power show a sharp increase during ictal epileptic activity (traces 7–15) in contrast to SD where they show a decrease (depression) (compare Fig. 1). Also note that compared to SDs, ictal epileptic activity propagates in the opposite direction, from E3 to E5 to E6 (black arrows). P_tiO₂ (trace 16) shows a steep increase in tissue oxygenation in response to ictal epileptic activity. Traces 17–18 show changes in rCBF measured with optodes neighboring electrodes 3 and 5. Although the rCBF response at optode 3 is hyperemic, optode 5 shows a shallow rCBF decrease in response to the ictal epileptic activity similar to the rCBF response to SD at optode 5 (compare Fig. 1). The two bottom traces demonstrate the parallel suppression of LF-VF (bandpass 0.01–0.2 Hz). Note that this suppression is similar to suppression of LF-VF during SD.

Figure 3

Evaluation of BBB permeability. (A) Cerebral MRI scan on day 15 after aSAH shows BBB dysfunctional regions marked in green. (B) CT image showing the optoelectrode strip overlaid with the BBB findings of A. The tissue surrounding electrode 4 and 5 (circled in red) shows higher BBB permeability. To achieve registration between scans, MRI scan with highest correlation to the CT was rotated according to the skull orientation for maximum fit. Electrodes are positioned close to the skull, resulting in minor distortions between the models and allowing accurate overlay of findings. (C) Frequency distribution of SD and ictal epileptic events over 12 days of ECoG recording after aSAH. Two spreading convulsions that occurred on days 7 and 9 were counted as SDs. The incidence of SD has a peak on day 2 (41 SDs), whereas ictal epileptic events peak on day 8 (10 events). Highest co-occurrence of SD and ictal epileptic events is on days 5 (13 SDs, 1 ictal epileptic event), and 6 (10 SDs, three ictal epileptic events).