Cranial Cavernous Malformations: Natural History and Treatment.

Abstract

HomeStrokeVol. 49, No. 4Cranial Cavernous Malformations Free AccessReview ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessReview ArticlePDF/EPUBCranial Cavernous MalformationsNatural History and Treatment Christopher J. Stapleton, MD and Fred G. BarkerII, MD Christopher J. StapletonChristopher J. Stapleton From the Department of Neurosurgery, Massachusetts General Hospital, Harvard Medical School, Boston. Search for more papers by this author and Fred G. BarkerIIFred G. BarkerII From the Department of Neurosurgery, Massachusetts General Hospital, Harvard Medical School, Boston. Search for more papers by this author Originally published13 Mar 2018https://doi.org/10.1161/STROKEAHA.117.017074Stroke. 2018;49:1029–1035Other version(s) of this articleYou are viewing the most recent version of this article. Previous versions: January 1, 2018: Previous Version 1 Cavernous malformations (CMs), also known as cavernous angiomas or cavernomas, are low-flow vascular malformations of the brain and spinal cord that consist of clusters of dilated sinusoidal channels lined with endothelial cells that do not exhibit intervening tight junctions. The involved blood vessels lack muscular and elastic layers, and blood at various stages of thrombosis and organization often fills the thin-walled vascular caverns. CMs are grossly distinct from adjacent brain and have a lobulated appearance sometimes likened to a mulberry, with a characteristic dark red or purple color. Hemosiderin and gliosis often surround CMs of the brain and spinal cord; no neural tissue is present inside the lesion. Except for developmental venous anomalies (DVAs) present in ≈33% of CMs,1 no abnormal vascularity is typically seen on digital subtraction angiography, leading CMs to be termed cryptic or occult vascular malformations (Figure 1). Multiple small hemorrhage events lead to a pathognomonic popcorn appearance on magnetic resonance (MR) imaging because of hemosiderin staining from blood products of various ages. Calcifications may be present on computed tomographic imaging (Figure 2).Download figureDownload PowerPointFigure 1. A 48-year-old woman presented with headaches. A, Magnetic resonance imaging of the brain with T2-weighted imaging demonstrated a 1.1-cm lesion in the cerebellar vermis with mixed signal intensity. B–D, Susceptibility-weighted imaging demonstrated hemosiderin within the lesion’s core and a developmental venous anomaly associated with the cavernous malformation (arrowheads).Download figureDownload PowerPointFigure 2. A 28-year-old woman presented with partial seizures without secondary generalization. A, Noncontrast computed tomography of the head demonstrated a 1.4-cm right posterior cingulate hyperdensity, with evidence of mild hemorrhage and calcification. Magnetic resonance imaging of the brain demonstrated the cavernous malformation to have a core of mixed signal intensity, heterogeneous enhancement, and a rim of hemosiderin consistent with a cavernous malformation, as evidenced on (B) susceptibility-weighted imaging, (C) T2-weighted imaging, (D) and T1-weighted imaging with gadolinium.CMs are the second most common type of central nervous system vascular lesion, comprising 10% to 15% of all neurovascular malformations with a prevalence ranging from 0.4% to 0.8%.2,3 Depending on the location and extent of hemorrhage, CMs can present incidentally or with seizure, headache, or focal neurological symptoms. Symptoms can be attributed to hemorrhage inside or outside the lesion or associated mass effect. Sporadic CMs tend to present with a solitary lesion, whereas familial CMs tend to present with multiple lesions throughout the central nervous system, with a strong family history of intracranial hemorrhage or seizures. The management of CMs includes considerations of observation, surgery, and radiosurgery, depending on the clinical presentation and anatomic location. Herein, we review current data on the biology, natural history, and treatment of CMs, with an emphasis on management considerations relative to lesion natural history and anatomic location.Genetics and PathobiologyCMs are present in both sporadic and familial forms, with the latter comprising 30% to 50% of cases. Most patients with familial CMs harbor >1 lesion, whereas only 12% to 20% of patients with sporadic CMs have multiple lesions.4 Familial forms of CMs are inherited in an autosomal dominant manner with variable penetrance and are caused by loss-of-function mutations in 1 of 3 genes: KRIT1 (CCM1), malcavernin (CCM2), and PDC10 (CCM3).5 Within these 3 CCM genes, many disease-causing mutations have been identified, underscoring the genetic heterogeneity of the disease process. The clinical, radiographic, and pathological characteristics of the disease do not seem to differ among the aberrant forms of the 3 CCM genes. It has been established that the 3 cerebral cavernous malformation proteins form a complex that interacts with other proteins critical for microvascular integrity, and a mutation in any 1 of the CCM genes may result in the formation of a CM.5,6Familial CMs, thus, result from inherited mutations in any 1 of the 3 implicated CCM genes, whereas sporadic CMs most likely arise from a germline mutation in a single individual or a somatic mutation in a single cell, as may occur after radiation therapy. Cutsforth-Gregory et al reported on 32 radiation-induced CMs managed at the Mayo Clinic between 1989 and 1999. The median latency from radiation treatment to CM diagnosis was 12 years, and lesions were always associated with the radiation port. Shorter latency periods were seen in older patients and patients who received higher radiation doses. Compared with nonradiation-induced CMs, radiation-induced CMs were often diagnosed at a younger age (31.1 versus 42.4 years) but were less likely to be symptomatic at the time of diagnosis (46.9% versus 65.8%). Patients with radiation-induced CMs were more likely to harbor multiple lesions, and these CMs carried a 4.2% risk of symptomatic hemorrhage per patient-year, compared with 2.3% for nonradiation-induced CMs.7 In addition to CMs resulting from radiation effects, there are case reports of CMs arising de novo from a thrombosed DVA,8 as well as from seeding of a biopsy tract.9Clinical PresentationMen and women are affected roughly equally, and patients most frequently come to clinical attention in the third and fourth decades of life. In a recent meta-analysis of 6 studies of 7 cohorts with data from 1620 individual patients, Horne et al reported hemorrhage (35.6%) as the most common clinical presentation, followed by incidental discovery (28.5%), seizure (20.8%), and focal neurological deficit (15.2%). Of the 1620 CMs analyzed, 50% were located in the cerebral hemispheres, 35% in the brain stem, 8% in the basal ganglia or thalamus, and 6% in the cerebellum. The anatomic location of the CM directly affects a patient’s chief complaint. In the analysis by Horne et al,10 86% of patients with seizures on presentation had lobar lesions, whereas 62% and 55% of patients with hemorrhage and focal neurological deficits, respectively, had brain stem CMs.Natural HistoryCM hemorrhage rates vary by study and may be influenced by study design (retrospective versus prospective) and study population (natural history versus surgical series; Table 1). Gross and Du reviewed the natural history of CMs from 12 studies between 1985 and 2015, with a specific focus on lesion natural history and hemorrhage rates. In total, 1610 patients were included in this analysis, of which, 26% across all anatomic locations presented with hemorrhage. In the review, 3 studies assumed that CMs were present from birth, and the annual hemorrhage rates across these particular studies ranged from 0.3% to 2.3% per patient-year. Seven studies defined the observation period as from initial presentation and diagnosis (not birth) to follow-up and reported hemorrhage rates ranging from 0.5% to 10% per patient-year. A pooled analysis of all included studies revealed an annual hemorrhage rate of 2.5% per patient-year across all anatomic locations.11 In the meta-analysis by Horne et al, the 5-year risk of an intracranial hemorrhage from a CM was reported as 15.8% across all anatomic locations. For patients who initially presented without an overt intracranial hemorrhage and a non-brain stem CM, the 5-year risk of hemorrhage was 3.8%, whereas the 5-year risk of a recurrent hemorrhage was 18.4%. Patients with brain stem CMs were noted to have significantly higher 5-year rates of initial (8%) and recurrent (30.8%) hemorrhage.10Table 1. Natural History SeriesAuthorsNo. of StudiesNo. of PatientsPresentation, %Location, %Hemorrhage RateIncidentalSeizureHemorrhageNeurological DeficitLobarDeepBrain StemCerebellumHorne et al106162028.520.835.615.250835.56.4Non-brain stem: 3.8% (5 y)Brain stem: 18.4% (5 y)Gross and Du1112161022.1*28.8*13.9*29.8*65.3†8.2†19.1†7.4†Congenital: 1.7% per patient-yearFollow-up: 2.7% per patient-year‡Taslimi et al12251295……………………Non-brain stem: 0.3% per patient-yearBrain stem: 2.8% per patient-year*Data derived from 1360 patients.†Data derived from 758 patients.‡Congenital: lesion assumed to be present since birth (time=0). Follow-up: lesion not assumed to be present since birth; time of diagnosis designated as time=0.Taslimi et al12 performed a meta-analysis of 25 natural history studies published between 1976 and 2015 that included 1295 patients with CMs. As reported, CM hemorrhage rates vary across populations and anatomic locations, Taslimi et al sought to pool patients into homogenous groups in an effort to report more specific hemorrhage rates. In their analysis, annual incidence rates of hemorrhage were 0.3% and 2.8% per patient-year for non-brain stem and brain stem CMs, respectively. Similarly, estimated annual rehemorrhage rates per patient-year were quoted at 6.3% and 32.3% for non-brain stem and brain stem CMs, respectively, with a median time to rehemorrhage of 10.5 months. The mortality rate after CM hemorrhage was low at 2.2%. Further, focused study of 323 patients from 5 of the original 25 studies demonstrated that rehemorrhages tended to occur within 2 years of the incident hemorrhage.12 This concept was explained by Barker et al13 in a retrospective review of patients from the Massachusetts General Hospital series, in which 141 patients were studied of whom 63 experienced a rehemorrhage. In this study, the monthly rehemorrhage hazard for untreated CMs was 2% during the first 2.5 years after initial hemorrhage, which then abruptly decreased to <1% per month thereafter. This phenomenon was termed temporal clustering—a concept that greatly affects the interpretation of efficacy rates after radiosurgery for CMs. Quoted risk factors for CM hemorrhage include young age,13 female sex,11 deep location, and prior hemorrhage.10–12,14Information on the overall risk of seizure onset in patients diagnosed with CMs has not been studied to the same degree as hemorrhage. In studies by Del Curling et al15 and Kondziolka et al,16 new-onset seizures were noted at a rate of 1.5% per patient-year, whereas Moriarity et al3 noted seizures to occur at 2.4% per patient-year. It is estimated that epilepsy develops in 35% to 70% of symptomatic CMs, with 40% of these patients having seizures refractory to antiepileptic medications.17,18Management ConsiderationsObservation/Conservative ManagementInterventional treatment in the form of surgery or radiosurgery is not without risk. As such, asymptomatic CMs may be observed over time, depending on the patient’s age and lesion location.19 In fact, CM hemorrhages tend to be small (<1.8 cm3),19 the annual risk of hemorrhage declines after 2.5 years,13 and most patients have good functional outcomes, despite the hemorrhage.12 Moreover, even if a lesion is symptomatic, observation may be the preferred management strategy if the risk of intervention is felt to outweigh the natural history risk of the lesion, such as may be the case with CMs located in deep-seated locations or eloquent cortex. Aside from antiepileptic medications to control seizures, there are no specific medications to treat CMs. Further, no established guidelines exist to guide the frequency of cranial imaging in patients being conservatively managed.SurgerySurgical resection is the gold standard of interventional treatment. Indications for surgery may include multiple hemorrhages, neurological deficit, and progressive seizures, unless the location carries an unacceptably high surgical risk. When a DVA is present alongside the CM, this should be left intact because DVAs provide a route of venous egress for normal central nervous system parenchyma. The surgical risk is typically low for lesions in the cerebrum and cerebellum and, as such, the presence of a single hemorrhage event, focal neurological deficits, or epilepsy may prompt consideration of resection.During surgery, complete removal is required because lesion remnants may hemorrhage or cause seizures.20,21 If the goal of surgery is a reduction in seizures, many advocate removal of the surrounding hemosiderin ring.22 Resection of the CM is facilitated by entry into and dissection within the gliotic plane surrounding the CM proper. With careful use of microsurgical techniques and instruments, the lesion can be circumferentially dissected and removed without permanent injury to the surrounding parenchyma or without significant bleeding from the CM.23 Unlike arteriovenous malformations, which are high-flow vascular shunts, CMs are low-flow lesions, and entry into the core of the CM does not typically result in a meaningful degree of hemorrhage. This allows for piecemeal removal of the lesion when necessary for CMs in deep-seated locations (eg, brain stem) or other eloquent locations where parenchymal invasion would be detrimental. Other maneuvers for ensuring safe surgical resection include internally debulking the CM and then dissecting the capsule away from surrounding structures. The use of stereotaxy and neuronavigation may allow the neurosurgeon to tailor the craniotomy and cortical entry point to the area of interest to minimize collateral damage.23RadiosurgeryCMs in deep-seated and eloquent locations may have unacceptably high surgical risk profiles. In these instances, stereotactic radiosurgery (SRS) has been proposed and used as a means to stall the natural progression of CMs (Table 2). In an early report from the University of Pittsburgh, Kondziolka et al24 performed Gamma Knife radiosurgery on 47 patients with hemorrhagic CMs (44 of whom had experienced at least 2 hemorrhages) in a critical intraparenchymal location remote from a pial or ependymal surface, features rendering surgical excision high risk. Of the 47 lesions, 24 were located in the pons/midbrain, 3 in the medulla, 9 in thalamus, 3 in the basal ganglia, 4 in the deep parietal lobe, and 4 in the deep temporal lobe. After a mean follow-up of 3.6 years, the authors noted a significant reduction in hemorrhagic events after radiosurgery (mean maximal dose of 32 Gy delivered), from 32% per lesion-year before treatment to 1.1% per lesion-year at 2 years after treatment. With respect to procedural complications, 12 patients (26%) developed new neurological deficits that were associated with referable MR imaging changes after radiosurgery, but only 2 patients (4%) had permanent neurological injury. In a later publication from the same group, Lunsford et al25 retrospectively evaluated the safety of efficacy of SRS in patients with solitary CMs prone to hemorrhage deemed high risk for resection. Between 1988 and 2005, 103 patients underwent SRS, with 98% of patients experiencing at least 2 hemorrhages before treatment. As in their prior publication, the authors again noted a reduction in the annual rate of symptomatic CM hemorrhage: 32.5% before SRS, 10.8% within 2 years of SRS, and 1.1% after 2 years of SRS. In this updated series, SRS treatment paradigms were adjusted to reduce complication rates, including limiting treatment volume and reducing maximal doses in critical anatomic locations. With these measures, 14 patients (13.5%) experienced new neurological deficits because of adverse radiation effects—a decline from 26% in the prior series. In this same publication, Lunsford et al also reviewed the available literature on radiosurgery for CMs. For 542 total patients, pre-SRS annual hemorrhage rates ranged from 2% to 35.5%, whereas hemorrhage rates >2 years post-SRS ranged from 1.6% to 8.8%. Adverse radiation effects ranged from 7% to 59%. One report included in the review was that of Amin-Hanjani et al, which provided an account of experience of Raymond N. Kjellberg with proton beam therapy from the Harvard Cyclotron Laboratory. In this report on 98 patients, the authors noted a decline in hemorrhage rates from 17.3% per lesion-year before treatment to 4.5% per lesion-year after treatment (after a latency period of 2 years). Radiation-induced complications resulted in a 16% incidence of permanent neurological deficits and a 3% mortality rate.26 In each of the reported series, the efficacy of radiosurgery was seen after at least a 2-year latency period. This observation parallels the published data on radiosurgery for arteriovenous malformations, in which lesion obliteration rates increase over time. In a multicenter retrospective analysis of 509 unruptured arteriovenous malformations, Ding et al27 reported actuarial obliteration rates of 44.1% at 3 years, 70.3% at 5 years, 76.8% at 7 years, and 85.2% at 10 years, with the steepest rate of occlusion occurring between 2 and 5 years after treatment.Table 2. Radiosurgical SeriesAuthorsNo. of CentersNo. of PatientsLocationPreradiosurgery Hemorrhage Rate, %Postradiosurgery Hemorrhage RateFollow-Up, y (Mean)ComplicationsKondziolka et al24147Critical intraparenchymal lesion remote from pial or ependymal surface56.5*0–2 y: 8.8%3.6Transient: 26%2–6 y: 1.1%Permanent: 4%Lunsford et al251103Lobar: 9.7%32.5*0–2 y: 10.8%5.713.5%Deep: 26.2%>2 y: 1.1%Brain stem: 64.1%Amin-Hanjani et al26195Lobar: 24.5%17.4*†0–2 y: 22.4%5.5Transient: 16.8%Deep: 18.4%>2 y: 4.5%Permanent: 7.3%Brain stem: 57.1%*Date of first hemorrhage designated as time=0.†Data derived from 73 patients with at least 1 confirmed clinical hemorrhage.The natural history of CMs suggests that after an incident hemorrhage, patients will experience a dramatic decline in hemorrhage risk after a period of 2.5 years without treatment.13 Given that an interval of 2 to 3 years is typically used as the time after which the efficacy of SRS is observed, many have called the use of SRS for CMs into question. Current evidence, therefore, favors expectant management of CMs deemed inappropriate for surgical resection, given this notion of temporal clustering and the high rates of radiation-induced injury observed in radiosurgical series.Stereotactic Laser AblationGiven the morbidity associated with both the natural history and surgical and radiosurgical treatment of CMs in deep-seated, eloquent, or surgically inaccessible locations, some authors have experimented with modern, minimally invasive methods of treatment. McCracken et al were the first to report experience with MR thermography-guided stereotactic laser ablation in a safety and feasibility series of 5 patients with drug-resistant epilepsy. In all patients, imaging during treatment and immediately post-procedure confirmed the desired extent of thermal ablation, and delayed postprocedure imaging (6–21 months) showed a reduction in lesion size with surrounding liquefactive necrosis (indicating extended lesionotomy). No adverse or unexpected neurological sequelae were noted, and 4 patients (80%) achieved freedom from disabling seizures (Engel class 1).28AnatomyCerebrumThe management paradigm for CMs varies depending on the anatomic location and the patient’s presenting symptoms. Incidental and asymptomatic lesions may be observed over time, whereas CMs with repeated hemorrhages or those that cause seizures or focal neurological deficits may be considered for surgical resection to reduce neurological sequelae. Amin-Hanjani et al published an early report on the risks of surgical management of CMs. In total, 94 patients with 97 CMs were reviewed, with 63 lesions (65%) located in the cerebrum, 14 (14.4%) in the brain stem, and 8 (8.2%) in the cerebellum. Of the 63 lesions in the cerebrum, 21 (55.3%) presented with seizures, whereas 8 (80%) and 13 (86.7%) presented with neurological deficits and headache, respectively. An excellent or good neurological outcome was observed in 96.8% of patients with lobar CMs, and 97% of patients presenting with seizures were seizure free after surgery. For cerebral lesions, 13 patients (20.6%) experienced complications, only 3 (4.8%) of which were permanent.29Seizures are the most common symptom associated with cerebral CMs. The pathogenesis of CM epileptogenesis may be secondary to slow lysis of red blood cells sequestered inside the sinusoids of the CM. Over time, pigments diffuse into the surrounding brain, inducing a gliotic reaction that develops into a seizure focus. Englot et al performed a systematic review of the published literature on seizure freedom after surgical resection of supratentotrial CMs. A total of 1226 patients were included across 31 studies, with 361 patients presenting with medically refractory epilepsy. After CM resection, 915 patients (75%) were seizure free. Prognosticators of favorable postsurgical outcomes included small CM size (<1.5 cm), absence of multiple CMs, medically controlled seizures, and the lack of seizure with secondary generalization. Of note, extended resection of the hemosiderin-stained tissue did not predict seizure freedom. Despite the observed efficacy, however, the authors note that medical management with antiepileptic medications is still recommended for patients with single or small lesions, with medically controlled epilepsy or the absence of secondary generalization.30Although surgical resection of symptomatic lobar lesions may be considered standard treatment, surgery for CMs in deep or eloquent supratentorial locations remains controversial. Even though hemorrhage from untreated lesions may result in progressive neurological deterioration, surgical resection of lesions in these locations carries serious risks. Chang et al reported the experience at the University of California, San Francisco, in which 79 patients underwent surgical resection of CMs in eloquent and deep brain regions, including 27 lesions (34.1%) in the basal ganglia, 23 (29.1%) in sensorimotor cortex, 3 (3.8%) in language cortex, 6 (7.6%) in the thalamus, 10 (12.7%) in visual cortex, and 10 (12.7%) in the corpus callosum. Found in 53 patients (67%), the most common presentation was hemorrhage with resulting focal neurological deficits. Postoperative MR imaging confirmed complete CM resection in 76 patients (96.2%). Good outcomes were achieved in 77 patients (97.4%), with 64 patients (81%) showing an improvement in their neurological status. Six patients (7.6%) had transient neurological injuries, and 1 patient (1.3%) was made permanently worse.31 For patients with hemorrhagic CMs and fixed neurological deficits with lesions in these critical and eloquent locations, the rationale for consideration of surgical resection is that surgery is less likely to add neurological morbidity (than if the patients were neurologically intact) and may reduce the risk of further neurological deterioration from persistent hemorrhage. If only a single hemorrhage event has occurred, surgical resection may be considered if the patient has a fixed neurological deficit and there is convincing pial or ependymal extension of the CM or an obvious hematoma cavity that provides a surgical corridor. Otherwise, conservative management is advised.32CerebellumThe management of cerebellar CMs is similar to that of cerebral CMs. In a series by de Oliveira et al, an acute or sudden presentation was noted in all patients, consisting of headache and components of a cerebellar syndrome: ataxia, dysmetria, nystagmus, speech abnormalities, or vomiting. Of the 10 patients highlighted, 9 (90%) had Zabramski type II hemorrhages, and all 10 showed imaging evidence of acute hemorrhage.33 At the time of last follow-up, 8 patients (80%) were in excellent condition, whereas 2 patients (20%) were in good condition but had neurological deficits similar to their preoperative status.34Brain StemEven more controversial than the resection of CMs in deep and eloquent locations is the resection of brain stem lesions, although considerations for surgical resection remain similar. In essence, if repeated hemorrhages lead to progressive neurological morbidity, surgical resection may be entertained (Table 3). This is especially true for lesions that abut a pial surface or for which a safe surgical corridor exists. For patients who recover or stabilize from a single hemorrhage event (or who are asymptomatic), observation is recommended.Table 3. Brain Stem Surgical SeriesAuthorsNo. of CentersNo. of PatientsLocation, %Preoperative Hemorrhage, %Complications, %MidbrainPontomesencephalicPonsPontomedullaryMedullaPorter et al35110026153910169735Abla et al36130011.215.443.111.918.596.953Garcia et al37110420.1…60.1…18.39928In 1999, Porter et al detailed the Barrow Neurological Institute experience with 100 patients with 103 brain stem CMs managed between 1984 and 1997. By anatomic location, 39 CMs (37.9%) were in the pons, 16 (15.5%) in the medulla, 16 (15.5%) in midbrain, 15 (14.6%) at the pontomesencephalic junction, 10 (9.7%) at the pontomedullary junction, 2 (1%) in the midbrain-hypothalamus/thalamus region, and 5 in >2 brain stem levels. A retrospective annual hemorrhage rate of 5% per lesion-year was estimated. In total, 86 patients (83.5%) underwent microsurgical resection, with all patients found to have an associated DVA. After surgery, 73 patients (87%) were the same or better, 8 patients (10%) were worse, and 3 (4%) died. Lesions were chosen for resection if they satisfied one of the following criteria: (1) abutted the pial surface or was exophytic, (2) produced repeated hemorrhages causing progressive neurological deficits, (3) had acute hemorrhage extending outside the lesion capsule, or (4) had significant mass effect because of large intralesional hemorrhage.35 Abla et al subsequently reported on the more recent experience from the Barrow Neurological Institute, in which 260 patients were treated between 1985 and 2009. In this series, 252 patients presented with hemorrhage. A total of 137 patients (53%) developed new or worsened neurological symptoms after surgery, but these symptoms were only permanent in 93 patients (36%). Perioperative complications were seen in 74 patients (28%). Rehemorrhages were seen in 18 patients (6.9%), and 12 (4.6%) patients required reoperation for residual or recurrent CMs. Postoperatively, the estimated annual hemorrhage rate was 2% per patient.36Garcia et al recently published the University of California, San Francisco, experience on the surgical treatment of brain stem CMs in 104 patients between 1997 and 2012. A retrospective annual hemorrhage rate of 4.5% per lesion was estimated. Most patients (72%) presented with cranial neuropathies, whereas headache (42.2%) and motor deficits (36.5%) were also common presentations. In this series, 15 patients (14.4%) experienced worsened cranial nerve or motor dysfunction postoperatively, and 4 patients (3.9%) required reoperation for hemorrhage within the resection cavity. The authors developed a proposed grading system for surgical risk, with the following factors and point assignments: size (≤2 cm, 0 points; >2 cm, 1 point), crossing axial midpoint (no, 0 points; yes, 1 point), DVA (no, 0 points; yes, 1 point), age (≤40 years, 0 points; >40 years, 1 point), and hemorrhage acuity (0–3 weeks [acute], 0 points; 3–8 weeks [subacute], 1 point; >8 weeks [chronic], 2 points). Patients with grades of 0 or I had universally favorable functional outcomes, whereas those with grades of VI or VII had universally unfavorable functional outcomes. The rate of good functional outcome for the remaining grades was as follows: II, 94.7%; III, 86.2%; IV, 76.5%; and V, 68.4%.37 In well-selected patient populations by experienced neurovascular or skull base surgeons, acceptable outcomes after surgery for brain stem CMs may be achieved.ConclusionsCMs are low-flow vascular malformations with a distinctive appearance on MR imaging. Although CMs may result in seizures, focal neurological deficits, or headache because of lesion hemorrhage, ≈40% of cases are incidentally discovered. Natural history studies suggest that the annual rate of lesion hemorrhage is 0.3% and 2.8% per patient-year for non-brain stem and brain stem CMs, respectively, and that hemorrhage events tend to cluster within a 2.5-year period after the incident hemorrhage. Surgical resection is considered for cerebral and cerebellar CMs that hemorrhage, cause neurological deficits, and are superficial in location. Lesions in deep-seated and eloquent cortex are frequently observed, although surgery may be undertaken if patients experience numerous hemorrhage events with neurological injury. Patients with lobar CMs and seizures are managed initially with antiepileptic medications. However, for progressive seizure and neurological disability, surgical resection is advised. The management of brain stem CMs is controversial. Although most lesions are observed, several centers have advocated for surgical resection in well-selected patient populations, whereas others advocate for SRS for these high-risk lesions.DisclosuresNone.FootnotesCorrespondence to Fred