Seizures are commonly observed following aneurysmal SAH; the incidence varies between 6 and 24 % of all nontraumatic SAH [14, 20, 21]. Between 1 and 28 % of patients with spontaneous SAH develop “early” seizures (i.e., seizures within the first 2 weeks of the onset of hemorrhage). Approximately 1–35 % develop “late” seizures (i.e., seizures occur after 2 weeks of the onset of hemorrhage) [15, 22]. Factors possibly underlying the variation in the incidence of seizures include heterogeneous patient populations, practice differences in pharmacologic seizure prophylaxis, center-specific treatment guidelines, and differences in continuous electroencephalography (cEEG) monitoring and its reporting practices [21, 23]. Population-based data from patients treated in the 1950s and 1960s estimated the 1- and 5-year incidence of epilepsy after SAH at 18 and 25 %, respectively [24]. A 9-year follow-up of patients in the International Subarachnoid Aneurysm Trial (ISAT) showed 10.9 % risk of seizure [25]. In a review of Swedish epilepsy registry data, Adelow et al. found that hospitalization for SAH was associated with an odds ratio of 5.1 (95 % confidence interval: 1.1–23.0) for a subsequent unprovoked seizure [26]. In an older retrospective series, late seizures were reported in 14 % of patients, with a mean latency of 8.4 months [27]. In a systematic review of 25 studies involving 7002 patients, the prevalence of early postoperative seizures was 2.3 %, and late postoperative seizures was 5.5 %, with average time to late seizures was 7.45 months [28].

Epidemiological data regarding seizures in nonaneurysmal SAH is scant. In a series of 12 patients who presented with cortical SAH, seven patients (58 %) had seizures, four were partial, and three generalized [29]. In a series of 34 patients who presented with cortical SAH, three patients (9 %) presented with generalized tonic–clonic seizures [30]. While these are small series, they suggest that the incidence of seizures in cortical nonaneurysmal SAH may be higher than typical aneurysmal SAH, or even than those with perimesencephalic hemorrhage. This high risk of seizure in superficial cortical hemorrhage may be related to the location of blood in contact with the cerebral cortex.

Noncontrast brain computed tomography (CT) is the most common tool used to evaluate acute SAH. The hemorrhage is classified into three distinct patterns according to the blood accumulation within the cranial cavity: diffuse, perimesencephalic, and cortical convexity types. The initial distribution of blood based on the initial imaging remains valid for the first few days after the hemorrhage; substantial changes may occur thereafter. In the first pattern, blood layers in the suprasellar central cisterns, with diffuse peripheral extension (Fig. 4.2a). Blood distribution in this pattern is usually associated with a ruptured cerebral aneurysm located at the circle of Willis branching points. Blood layering in the Sylvian fissure usually suggests aneurysm in the middle cerebral artery distribution (Fig. 4.3a), whereas blood in the interhemispheric fissure with extension to the ventricular system may suggest an aneurysm of the anterior communicating artery (Fig. 4.3b). Blood accumulation in the posterior head region occurs with a rupture of a basilar apex aneurysm (Fig. 4.3c). The second pattern of blood layering in the perimesencephalic and interpeduncular basal cisterns (Fig. 4.2b) is considered benign and usually idiopathic (normal angiogram) and may account for approximately 10 % of all SAH [35]. In the third pattern, blood products localize to a few sulci along the cerebral convexities (Fig. 4.2c), without extension to the basal cistern or ventricles. This latter pattern is less commonly recognized and has only recently been described as a distinct category of SAH [29, 36–38]. This pattern has an estimated incidence of 7 % of all patients with spontaneous SAH [39]. Patients with perimesencephalic SAH have less risk of complications including rebleeding, vasospasm, and hydrocephalus [40]. The accumulation of blood in SAH, based on noncontrast head CT, is graded according to the thickness of the blood layering in the subarachnoid space and whether leakage to the ventricular system had occurred. The Fisher grading scale (Table 4.3) perhaps is the most reliable predictor of vasospasm [41], although it does localize blood accumulation in the brain as highlighted above. Fisher scale remains subjective but with a high inter-reliability value (κ of 0.90) [42].

The risk of seizures varies with the patterns of blood distribution found on the initial head CT. The overall risk of a seizure with aneurysmal SAH is likely higher than those whose bleed is in the perimesencephalic cistern. In a review of 83 patients with spontaneous SAH with negative initial angiography, 49 of those had blood accumulation in the perimesencephalic cistern only [40]. The authors reported that none of the latter group had late epilepsy, but one patient among the diffuse SAH did develop late epilepsy [40]. In a review of 24 patients with superficial cortical convexity hemorrhage, 5 (21 %) presented with seizures; all of whom were older than 60 years of age [43]. When SAH is classified by the Fisher grading system, the risk of epilepsy is higher in those with grade III and IV. In 217 patients with SAH, 17 (7.8 %) developed seizures at onset, all of whom had Fisher grade III and IV. Late epilepsy in the same cohort occurred in 15 patients (6.9 %), nine of whom with grade III and IV on the initial head CT [22].

Numerous studies have identified the risks for early and late seizures following SAH (Table 4.4). In an earlier retrospective analysis by Ohman et al., risk factors associated with seizures in SAH included history of hypertension, presence of infarction on imaging, and the duration of coma [44]. A Korean study reported risk factors for early-onset seizures after aneurysmal SAH that included younger age (age less than 40), poor clinical and radiological grades (based on Hunt–Hess and Fisher grading systems), acute hydrocephalus, and rebleeding prior to surgical occlusion of the aneurysm [20]. Two studies evaluated the risk factors for late seizures following SAH [20, 45]. Both of these studies showed that the amount of blood within the cranial cavity, based on the initial head scan, is a consistent predictor of late seizures. In a post hoc analysis of 2143 patients enrolled in the ISAT, a Fisher grade greater than 1 on the initial head CT showed a trend in predicting posthemorrhagic epilepsy (hazard ratio 1.34, 95 % CI 0.62–2.87). Another study reported a Fisher grade of 3 or greater as a risk factor for seizures [22]. Other independent risk factors for late seizures include cortical infarction and hematoma formation (intracerebral or subdural hematoma) [45]. Seizures at onset in aneurysmal SAH were believed to be an important risk factor for delayed epilepsy [14, 46, 47]. In contrast, other work reported seizures at onset are not a predictive factor for late epilepsy [48–50].

Seizure following SAH may occur by either direct injury caused by blood or blood products in the subarachnoid space or due to vasospasm. Early clinical seizures are thought to be related to large amounts of blood in the subarachnoid space and damage to the motor cortex or the insula, the combined effects of a space-occupying lesion with mass effect, focal ischemia, and blood products catalyzing the release of large amounts of glutamate and the generation of oxygen-free radicals [50, 51]. In addition to the direct pathologic effects, seizures may also worsen the extent of injury from the inciting neurologic injury by increasing metabolic, excitotoxic, and oxidative stress on at-risk brain, leading to irreversible injury. Vasospasm is a common complication that usually occurs during the acute phase of SAH and may lead to focal cerebral ischemia. Angiographic vasospasm follows a biphasic pattern, with early arterial spasm within minutes of hemorrhage and delayed phase vasospasm, which is often associated with delayed cerebral ischemia [52, 53]. Early spreading depolarization has also been proposed as a risk factor for post-SAH seizures [54]. Spreading depolarization is a wave of massive ion translocation between the intracellular and extracellular space, near-complete sustained depolarization of neurons, glial depolarization, neuronal swelling, distortion of dendritic spines, and an abrupt, large, negative change of the slow electrical potential [55]. By contrast, epileptiform activity is characterized by a milder sustained depolarization with less pronounced disturbance of ion homeostasis up to the so-called ceiling level [56]. Whereas sustained depolarization remains below the inactivation threshold for the generation of action potentials during epileptiform activity, this threshold is exceeded during spreading depolarization. Therefore, during seizures, neurons typically fire synchronous action potentials at a high frequency, while silencing (spreading depression) of spontaneous activity is observed during spreading depolarization [57]. Late seizures may occur in the setting of superficial siderosis, blood in subarachnoid space resulting hemosiderin deposition in the subpial layer of the brain [58]. Besides late seizures, superficial siderosis is usually characterized by the presence of progressive ataxia and hearing loss.

Nonconvulsive seizures are associated with elevations in glycerol, presumably because of cell membrane breakdown [59, 60]. One theory regarding cause is that seizures in acutely injured brain have marginally maintaining homeostasis that leads to excessive metabolic demand and increased blood flow thereby compromising at-risk brain tissue leading to additional brain injury with cell death. An alternative hypothesis is that electrographic seizures or periodic epileptiform patterns (Fig. 4.4) are surrogate markers for more severely injured brain. However, most authors agree that prolonged seizures may damage the brain and are associated with secondary injury [61].

The treatment for ruptured cerebral aneurysms has changed over the past two decades. The ISAT showed significantly better outcomes associated with endovascular coiling compared to surgical clipping performed through craniotomy [62]. The same observations were confirmed in the Barrow Ruptured Aneurysm Trial (BRAT) [63]. In addition to minimal invasive surgery to secure a ruptured aneurysm, critical care and seizure detection technology have improved over the last decade [31], allowing for early treatment and prevention. Along with these changes, the recommendations for seizure treatment with primary and secondary prevention have changed.

Long-term follow-up of patients enrolled in the ISAT and other examples suggested a significantly lower risk of seizures in patients treated with coil embolization when evaluated at 2- and 14-year follow-up, compared to those treated with surgical clipping [25, 64]. Currently, no randomized clinical trials have investigated the benefit and outcomes of seizure prophylaxis following SAH. The administration of prophylactic antiepileptic drugs (AEDs) has in the past been part of a standard protocol for patients undergoing neurosurgical procedures such as craniotomy. A survey conducted in 2002 by the American Association of Neurological Surgeons showed that 24 % of neurosurgeons routinely prescribed AEDs for 3 months after SAH regardless of whether seizures occurred at presentation, in hospital, or not [16, 48, 49, 65]. In contrast, only 4 % of German neurosurgeons routinely prescribe AEDs for patients with aneurysmal SAH [66]. More recently, however, routine use of AEDs for SAH patients has come under question [23] because of lower seizure rates associated with the minimal invasive technology to obliterate cerebral aneurysms [15, 67]. The Stroke Council of America guidelines recommend against routine perioperative use of AEDs in aneurysmal SAH because of the paucity of evidence of benefit [12]. However, the guidelines suggest that AED prophylaxis may be considered in the posthemorrhagic period and longer term for those with risk factors for seizure recurrence. These include a prior seizure, parenchymal infarction or hematoma formation, and middle cerebral artery aneurysm [12]. Marigold et al. attempted to evaluate the utility of AEDs in primary and secondary prevention using the Cochrane database [68]. They concluded that there is insufficient evidence to support routine use of AED for primary and secondary seizure prevention after SAH. No specific recommendation exists to stratify the risk of seizures or late epilepsy by the location of hemorrhage in the brain on the initial head CT. For example, patients with perimesencephalic hemorrhage are usually at the lowest risk of initial seizure or late epilepsy; whereas superficial cortical hemorrhage typically has a higher risk of associated seizures. The recommendations for the latter two groups usually follow any individual with aneurysmal SAH.