The advent of MRI has increased the sensitivity of detecting CCMs. Standard T1 and T2 images are highly sensitive and specific for CCMs [5], with susceptibility-weighted imaging (SWI) being the most sensitive modality [31]. CCMs are known to have a characteristic “popcorn” appearance on MRI, owing to the deposition of blood products in varying states of decay in the adjacent brain parenchyma [6]. MRI appearance of CCMs can be [1] hyperintense on T1 and T2 images without active enhancement with gadolinium, [2] central mixed-signal intensity on T1 and T2 surrounded by a hypointense ring on T2 images—these are the characteristic “popcorn” lesions (Fig. 6.2a–c), [3] central isointense to hypointense core on T1 and T2 with a hypointense ring on T2 images (Fig. 6.2d), or [4] punctate hypointense lesions on gradient echo (GRE) that may represent nascent familial CCMs. Subacute hemorrhage or degraded blood product make it difficult to differentiate these lesions from tumors. Besides the lack of uptake with gadolinium T-1 MRI, fluoro-deoxy glucose positron emission tomography (FDG-PET) demonstrates a hypometabolic focus (Fig. 6.3). There are no radiological differences between familial and sporadic CCMs, except that familial CCMs tend to occur in groups, whereas sporadic CCMs are solitary and may occasionally occur adjacent to a developmental venous anomaly [32]. Additionally, there is no angiographic difference between familial or sporadic CCMs, both lacking afferent and efferent vessels [33]. Although contrast enhancement was thought to indicate the presence of other vascular malformations in the past, we now know that this is not always true. CCMs may enhance if time-delayed double doses of gadolinium are used. One retrospective review of 32 patients found that at least 20 % of CCMs showed enhancement, which correlated neither with the size of the CCM nor with the presence of developmental venous anomalies. Therefore, radiographic diagnosis should be based primarily upon morphology and signal intensity rather than enhancement [34].

There are no neurons within the CCM itself [25], and seizures caused by CCMs arise from the neighboring cortex. Over time, CCMs leak minute amounts of blood that degrade into hemosiderin, a toxic substance that leads to neuronal death and synaptic reorganization. In addition, extracellular iron appears to trigger reactive gliosis, likely via free radicals, and inhibit glutamate reabsorption. Collectively, these changes produce a hyperexcitable state. Unlike tumor-induced seizures, mass effect from the CCM does not appear to influence epileptogenesis [15].

Neurons adjacent to CCMs were extracted during surgery and proved to have normal membrane properties, including a normal resting potential [15]. However, more than half of those neurons emitted spontaneous depolarizing discharges that were likely synaptically driven. These neurons also generated large-amplitude postsynaptic potentials resembling paroxysmal depolarization shifts, although these potentials were graded rather than being all or none. These phenomena were more common in neurons sampled from the vicinity of CCMs than those adjacent to neoplasms. Another study compared the intraoperative electrocorticographic (ECoG) discharge patterns of CCMs with those of neurodevelopmental lesions, such as neoplasms and cortical dysplasias, in patients with pharmacoresistant epilepsy [7]. The authors found that coincident mesial temporal discharge bursts were more common in CCMs than in neurodevelopmental lesions. Continuous spiking was seen in both groups correlating with longer disease duration. Interestingly, the absence of coincident bursts in the CCM group was associated with increased density of microglia, with no relationship between the degree of iron deposition or gliosis and the frequency of ECoG epileptiform discharges.

Seizures are the most common symptomatic presentation of supratentorial CCMs [35], and CCMs located in the mesial temporal regions are more likely to result in seizures than in other brain regions. Temporal CCMs, however, respond better to surgery than extratemporal CCMs [36]. Seizure semiology in CCMs varies depending on the location of the CCM [27]. Of note, neither prior intracerebral hemorrhage (ICH) nor focal neurological deficits seem to increase the risk of seizures in patients with CCMs, which stands in contrast to arteriovenous malformations (AVMs) [12].

Incidentally identified CCMs rarely cause clinical symptoms. On the other hand, CCMs that cause a seizure will have a 94 % chance of causing a second seizure [37], which is far more than the chances of having a second seizure due to AVMs (58 %). This difference is believed to be due to the hemosiderin ring surrounding a CCM [12] .

As with other epilepsies, the first-line treatment for seizures caused by CCMs is pharmacotherapy. There are no preferred antiepileptic drugs (AEDs), although some avoid valproic acid as it may cause thrombocytopenia [4]. An AED should be considered after a first seizure as it reduces the time to 2-year seizure freedom when compared to placebo [37]. It is not clear if dual AED therapy offers any additional effect beyond monotherapy [25]. One emerging form of pharmacotherapy is the use of a multiple kinase inhibitor: sorafenib. Though developed for use in renal and hepatocellular carcinoma, it has been used with success to shrink a hepatic cavernous hemangioma, whose size decreased from 1492 to 665 mL after 78 days [38]. Another case report cited similar results with bevacizumab [39].

Medical intractability in epilepsy is defined as the failure of two AEDs to achieve full seizure control, and must necessitate a surgical evaluation [40]. With CCMs, a prompt surgical evaluation is even more highly recommended than with most other focal epilepsies since longer duration of seizures may be associated with decreased chances of seizure freedom after lesionectomy [25]. As with other focal epilepsies, the adverse effects of AEDs as well as the functional and occupational limitations imposed by uncontrolled seizures will further clarify the need of early surgical evaluation.

A thorough preoperative evaluation is certainly recommended. This should include at least a seizure-protocol brain MRI and continuous video-electroencephalograph (EEG) monitoring with ictal recordings. The intracarotid amobarbital procedure, neuropsychological testing, and other preoperative procedures may be recommended in individual cases. Sometimes, intracranial monitoring may be used in order to identify the epileptic margins of CCMs. Diffusion tensor imaging has been used to help identify white matter tracts passing through the surrounding hemosiderin ring [32]. Functional imaging such as single-photon emission computed tomography (SPECT) or functional magnetic resonance imaging (fMRI) remains a useful adjunct [35].

The extent of optimal surgical resection is unknown due to a lack of adequate data that compare lesionectomies with more extensive resections [35, 41]. Lesionectomies involve the removal of only the CCM itself, whereas tailored resections involve removal of all or part of the hemosiderin rim surrounding the CCM [25]. The borders of such resections can be further refined with ECoG monitoring [42]. Only around 60 % of patients whose lesionectomy is limited to the malformation itself will achieve Engel class I surgical outcome [43], likely due to limited or no resection of hemosiderin-laden tissue [25]. Importantly, sparing subcortical hemosiderin deposition to preserve white matter tracts does not appear to influence the outcome [35].

As with other lesional epilepsies, resection of CCMs from eloquent areas poses greater difficulties. Extraoperative intracranial monitoring or intraoperative awake mapping is often indicated in such situations. In one series of nine patients with dominant hemispheric CCMs of whom six had seizures, five patients attained seizure freedom off AEDs [44]. Multiple CCMs do not contraindicate surgery, and one report mentions a patient who achieved seizure freedom after removal of ten malformations [41]. It is uncertain, however, if asymptomatic CCMs should be removed alongside epileptogenic ones [25].

Gamma Knife radiosurgery has also been used with CCMs since the mid-1980s [45]. The focally applied radiation leads to obliteration of the CCM lumen by virtue of endothelial proliferation. This process can take as long as 3 years to complete. The optimal dose of radiation has not been determined. Furthermore, the risk of bleeding persists until ablation is complete and only approximately half of all patients will achieve seizure freedom, which is less than what is seen with lesionectomy. Stereotactic radiosurgery may still be useful for patients with CCMs near the eloquent cortex [35]. In one study of 49 patients with pharmacoresistant epilepsy due to CCMs, radiosurgery achieved seizure freedom in 26 patients (53 %) [46]. Of note, medial temporal CCMs in that series were associated with a higher risk of failure .

Patients with mesial temporal CCMs are more likely to experience seizure freedom after lesionectomy than those with CCMs in other brain regions [36], but some authors suggest that this difference is short lived [47]. Otherwise, there is no correlation between the location of the CCM and seizure freedom [35]. Of note, neither the size of the lesion nor the presence of perioperative bleeding has an effect on long-term outcome. Also, age at the time of surgery does not seem to affect the outcome, although the data about this are inconsistent. Patients who experience secondary generalization or high seizure frequency (defined in one series of 60 patients as one or more seizures per month over 1 year) have less favorable outcomes after surgery [35].

In one survey, 70 % of 168 patients who underwent lesionectomies or more radical resections of CCMs achieved Engel I surgical outcome after 1 year of follow-up [47]. This rate declined to 65 % 3 years after surgery. Part of this decline in seizure freedom rates may be due to the loss of seizure-free patients to follow up. Only nine patients had Engel IV surgical outcome, of whom four had multifocal slowing and epileptic discharges preoperatively suggesting other seizure foci [47]. The rate of AED discontinuation after surgery is also unclear [25].

Surgical complications pertaining to surgery are rare. In the immediate postoperative period, 17 % of patients experience focal neurological deficits, but this falls to 2.6–8 % during follow-up. There are no reported deaths due to surgery [35]. This low complication rate correlates with patient satisfaction. In one survey, patients were asked if they would have surgery again and 90 % of respondents stated they would [42]. Given all of the above, surgery should be considered in those with supratentorial CCMs and epilepsy [47].

For those with familial CCMs, genetic testing is also recommended as a screening MRI may be negative due to disease latency. Furthermore, those with familial CCMs who have successfully undergone lesionectomy may still require periodic MRI screening as new lesions can form at the rate of 0.2–0.4 CCMs per patient year [11] .