Chronic encapsulated intracerebral hematomas are a very rare complication of radiosurgery for cerebral AVM. The incidence in our cohort of treated patients was 1,8 % of patients with brain AVM treated with SRS over a period of 17 years. Other series attest to the rarity of this complication with reported incidences ranging from 0,6 to 4% 1,15–17. However, the long latency period of CEIHs and the cessation of imaging controls once nidus obliteration has been documented may have led to the underestimation of the true prevalence 18. Longer periods of imaging follow-up, at least years post radiosurgery have been suggested 19.
Histologically CEIHs are made of a thickened hematoma capsule with abundant microvasculature that can easily bleed when removed surgically. The hematoma itself is serous and is usually easily aspirated. The gross appearance is similar to chronic subdural hematoma. CEIHs may develop near vascular lesions such as AVMs, cavernous angiomas and venous angiomas. It is thought that CEIHs develop secondary to hemorrhagic episodes of the initial angiomatous lesion with its eventual “self-destruction” or thrombosis 4,16,20,21. In the case of post radiosurgery obliterated AVMs, it is thought that radiation-induced inflammation triggers neo-angiogenesis of fragile new vessels, breakdown of the blood-brain barrier, fluid exudation in the nearby brain, edema and potential cyst formation. Dense vascularization has been found in the capsule of CEIHs and it is thought that bleeding of these fragile vessels results in expansion of the capsule and further bleeding, a mechanism similar to chronic subdural hematoma 22–25. Neovascularization and hematoma expansion appear to be mediated by VEGF (Vascular Permeability Factor), a potent vascular endothelial cell mitogen that promotes neovascularization and vascular permeability 26 associated also to chronic subdural hematoma pathophysiology. Further studies are needed to elucidate the mechanisms of CEIHs post AVM radiosurgery.
On MR imaging, a common finding in all cases of CEIHs was extensive perilesional edema. Most cases demonstrated as low intensity or heterogeneous lesions on T2 weighted imaging with or without a hypodense rim. On contrast enhanced T1 weighted images there existed usually nodular or multinodular enhancement (Table 1). CEIHs were associated with cyst in 62,1% of cases pointing to a possible common pathophysiologic mechanism 3. The latency time from radiosurgery to CEIHs diagnosis was 7,7 years sd 3,7 years. Symptoms, the most common being from headache (44,8%), hemiparesis (41,3%), nausea/vomiting (13,7%) were mostly related to the mass effect of the gradually growing CEIH and the surrounding edema.
Several risk factors have been explored for the development of CEIHs including age, sex, basal ganglia AVM location, irradiated nidus volume, the marginal or total dose, early RICs, repeat radiosurgery, nidus obliteration, pre-radiosurgery embolization, pre-radiosurgery surgery and prior hemorrhage with inconsistent results 15,16,18,21,27. In the present review, the distributions of age, sex, location, marginal dose, nidus obliteration and pre-radiosurgery embolization did not differ from distributions seen in cohorts of AVMs treated by radiosurgery. However, the incidence of radiation induced changes in the years post radiosurgery was unusually high (32,4%). There was also a high percentage (18,9%) of cases which had received repeat radiosurgery. Further studies are needed to ascertain the risk factors and mechanisms of CEIHs that develop post SRS for AVM 14.
CEIHs often caused progressive neurological deficits due to mass effect. The most efficient treatment was complete excision that led to clinical and/or radiological improvement in cases. Partial treatment was less efficient and had to be complemented by complete excision in cases. Conservative management consisting of follow-up or steroid administration was unsuccessful in most cases and had to be complemented by total excision of the hematoma and the capsule to achieve good clinical outcome.
This study is susceptible to a number of biases inherent to any retrospective study and review like the small number of cases, selection bias and publication bias. The time CEIHs were detected was mostly based on the timing of symptom development and asymptomatic CEIHs may have been underreported. Larger studies are needed to further elucidate the pathophysiology, incidence and risk factors related to the development of CEIHs post cerebral AVM radiosurgery.