Recent novelties in research and management of cerebrospinal cavernous malformations Authors: Laurèl Rauschenbach, Philipp Dammann, Ulrich Sure Published: 2024 ABSTRACT In recent years, knowledge about cerebrospinal cavernomas has grown considerably, leading to the development of initial guidelines and treatment recommendations. However, due to the rarity and heterogeneity of the disease, the level of evidence remains limited, leaving many questions unanswered and subject to ongoing debate. Therefore, an up-to-date review of this field's latest developments and controversies is reasonable. Introduction Cavernous malformations (CMs) of the central nervous system (CNS), also known as cavernomas, cavernous angiomas, or cavernous hemangiomas, are a distinct type of neurovascular lesion characterized by clusters of abnormally dilated capillaries with thin walls that lack intervening brain parenchyma [8]. Sporadic CMs are often associated with developmental venous anomalies (DVAs), which are benign vascular malformations marked by an atypical venous arrangement in the brain [86]. CMs are predisposed to hemorrhagic events and may rarely exhibit neoplastic-like behavior, leading to their inclusion in the World Health Organization (WHO) classification of CNS tumors [52]. CMs may occur sporadically or as part of a familial syndrome. Sporadic CMs typically present as solitary lesions but may occasionally appear as multiple lesions, especially after previous whole-brain irradiation in infancy or when associated with large DVAs. Familial CMs are inherited in an autosomal dominant pattern and are characterized by multiple lesions resulting from germline loss-of-function mutations in the CCM1 (KRIT1), CCM2 (MGC4607), and CCM3 (PDCD10) genes [80]. When genetic testing is unavailable, the presence of both CMs and DVAs may help differentiate the sporadic from familial forms. In individuals who have not previously undergone irradiation, the presence of a DVA alongside one CM typically suggests a sporadic origin, whereas multiple CMs without a DVA are more indicative of a familial form. Clinically, patients with a solitary lesion, an associated DVA, and no family history of affected relatives are very likely to have a sporadic disease, even in the absence of genetic testing [62]. Figure 1 attempts to illustrate the distinguishing characteristics of sporadic and multiple CMs. DVAs are found incidentally in about 3% of the population [56]. While they typically do not carry a significant risk of hemorrhage, these lesions are believed to be a contributing factor in the development of CMs in their vicinity, which may occur in up to 7% of patients over their lifetime [10]. Current guidelines recommend genetic testing in patients with multiple CMs when there is no associated DVA, history of brain irradiation, or a positive family history [2].Fig. 1Comparison of the clinical and imaging characteristics of solitary and multiple lesions. Shown are two exemplary, axially-layered MRI-SWI sequences: one from a patient with a solitary lesion and associated DVA (left) and another from a patient with multiple lesions without an associated DVA (right). While multiple lesions are often attributable to previous irradiation or a hereditary origin and are less frequently associated with a DVA, the opposite is true for solitary lesions. This allows for an initial assessment of the underlying etiology of the condition, though definitive conclusions cannot be drawn, as exceptions exist on both sides Comparison of the clinical and imaging characteristics of solitary and multiple lesions. Shown are two exemplary, axially-layered MRI-SWI sequences: one from a patient with a solitary lesion and associated DVA (left) and another from a patient with multiple lesions without an associated DVA (right). While multiple lesions are often attributable to previous irradiation or a hereditary origin and are less frequently associated with a DVA, the opposite is true for solitary lesions. This allows for an initial assessment of the underlying etiology of the condition, though definitive conclusions cannot be drawn, as exceptions exist on both sides The clinical presentation of CMs varies widely, ranging from asymptomatic cases to those with seizures or neurological deficits caused by a symptomatic hemorrhage or mass effect. This variability, combined with the significant impact of lesion localization has led to substantial efforts to differentiate which patients require immediate treatment and which may benefit from a conservative, watch-and-wait approach. Surgical resection remains the primary treatment for symptomatic or high-risk CMs, aiming to completely remove the lesion, minimize neurological deficits, and prevent future hemorrhages [46]. Advances in imaging, microsurgical techniques, and intraoperative monitoring have significantly improved surgical outcomes, reducing the risk of iatrogenic morbidity. For patients not eligible for surgery, non-surgical options, such as antiepileptic drugs, offer alternative treatments. Furthermore, ongoing research on stereotactic radiosurgery and pharmacological interventions seeks to identify therapies that may reduce hemorrhage risk or modify the disease course. This narrative review is based on long-term clinical experience from a high-volume referral center for sporadic and familial CMs. It aims to provide an updated synthesis of current strategies and treatment indications for CNS-CMs, drawing on recent literature and the insights gained from our clinical practice. To address the considerable heterogeneity of the disease, we will examine CMs in different anatomical regions of the CNS, including supratentorial and cerebellar, brainstem, and spinal cord lesions. By integrating recent advances and current best practices, we hope to guide clinicians toward state-of-the-art treatments and facilitate patient-centered decision-making based on the latest guidelines, ultimately improving outcomes and quality of life for those affected by this complex condition. General considerations in CNS-CM surgery The treatment options for CNS-CMs range from conservative approaches to surgical interventions, depending on the characteristics of the lesions and the clinical profile of the patients. Although debated, surgical resection remains the gold standard and the only curative option for symptomatic or high-risk CMs with a history of hemorrhage. However, the decision to proceed with surgery often requires careful consideration of the natural history of the disease versus the potential risks of surgery, since the risk of postoperative complications may be equal to or greater than the risk of leaving the lesion untreated [58]. Various factors, both objective and subjective, influence the appropriateness of surgery, including anatomic features (e.g., lesion size and localization), clinical presentation (e.g., symptoms, severity of neurological deficits, number and frequency of bleeding events, and presence of familial cavernomatosis), and patient-specific factors (e.g., age, life expectancy, comorbidities, and personal fears related to surgery, unexpected bleeding events, or sudden unexpected death in epilepsy cases). Randomized controlled trials comparing surgical and conservative treatments are scarce [69], and most available evidence comes from observational studies, which are often limited by selection bias, single-center recruitment, and short follow-up periods [63, 82]. Despite these limitations, recent efforts to conduct higher-quality research studies have been encouraging, though definitive results are still pending [1, 6]. In the meantime, clinical decisions can be guided by existing resources such as the 2013 guidelines from the International League Against Epilepsy [70] and the 2017 guidelines from the Angioma Alliance [2], as well as expert opinions derived from consensus processes [16, 81]. The following general surgical considerations are important in CM surgery:Symptomatology: CMs may cause symptoms through their mass effect or more commonly as a result of hemorrhage, which can exacerbate pre-existing symptoms. Hemorrhagic events should be clearly defined as the acute or subacute onset of neurological symptoms (such as headache, seizure, impaired consciousness, or focal neurological deficits) that can be linked to the anatomical location of a CM and confirmed by imaging [4]. Asymptomatic growth, hemorrhage, or the detection of perifocal hemosiderin deposits alone do not meet the criteria for a hemorrhagic event. This definition helps explain why brainstem and intramedullary CMs are often perceived as more aggressive. Their perceived aggressiveness is likely due to their location in highly eloquent CNS structures, where even small hemorrhages can lead to significant neurological deficits. In rare cases, CMs can cause cluster hemorrhages, characterized by recurrent bleeding episodes occurring in rapid succession, leading to rapid neurological deterioration [16].Risk of hemorrhage: Several studies have evaluated the natural history of CMs and identified risk factors for bleeding. Patients who have experienced a prior hemorrhage are at significantly higher risk for future hemorrhages than those in whom the CM was discovered incidentally [5, 41]. This supports the rationale for a conservative, watch-and-wait approach in asymptomatic patients and favors treatment initiation after an initial or recurrent bleeding event. Our own studies suggest that obese patients and those receiving female hormone treatment may also have an increased risk of experiencing hemorrhages [14, 92].Surgical advances: Recent advances in microsurgical techniques, including neuronavigation [89], imaging [11], and electrophysiological monitoring [51], have enhanced the precision and safety of CM removal. Surgeons now have access to advanced preoperative imaging techniques such as diffusion-tensor MRI with tractography [3], navigated transcranial magnetic stimulation [60], and task-based functional MRI [90]. Depending on the lesion location this can be crucial for preoperative planning, as white matter tracts may be displaced by the CM but can still pass through the adjacent hemosiderin ring, which may be partially resected [12]. For cases involving epilepsy, detailed preoperative epileptological diagnostics, such as video EEG monitoring, are useful [75]. Intraoperative tools such as microsurgical techniques [20, 47], electrophysiological monitoring (cortical and subcortical stimulation) [51, 66], frameless image-guided neuronavigation [89], neuro-navigated ultrasonography [25], and, in certain cases, awake surgery [61], further contribute to minimal surgical morbidity.Extent of resection: Achieving complete lesion removal is crucial for preventing recurrent hemorrhages and ongoing seizures [45]. Debate exists regarding the optimal surgical approach, whether transsulcal [38] or transparenchymal [47], but most agree that a circumferential, intracapsular approach with en bloc resection provides the best results in supratentorial CM surgery. Resection is typically most effective several weeks after a hemorrhage, when the lesion has clearer delineation from surrounding brain tissue and perifocal edema has resolved, allowing for safer and more effective excision [26]. For lesions in deep or eloquent areas, central debulking followed by piecemeal resection is often necessary and feasible, as CMs lack arterial input. Meticulous care must be taken to avoid leaving residual CM tissue, as complete resection is crucial for surgical success.Preserving DVAs: Preservation of DVAs during surgery is important, as these vessels provide normal and usually critical venous drainage for the affected individual. Disruption of a DVA increases the risk of venous thrombosis and infarction, potentially leading to secondary hemorrhage. Some authors propose removing the distal branches of the DVA while preserving the main trunk to prevent CM recurrence, but this approach remains under-researched [88]. Available guidelines and surveys, however, recommend preserving DVAs to avoid secondary complications [2, 16, 81].Epilepsy: In cases of CM-related epilepsy (CRE), the epileptogenic focus can extend beyond the CM itself. In patients with medically refractory seizures, resection should aim to include the hemosiderin-enriched gliotic tissue (if this area is not eloquent) surrounding the CM, as this tissue may also be a source of seizure activity [71]. Removing hemosiderin-enriched tissue is less significant when CMs are resected early in the course of epilepsy, which is generally recommended [28].Case load: For deep or eloquently located CMs, surgery should be performed in high-volume centers with neurovascular experience. Evidence suggests that hospitals with higher case volumes are associated with better neurological outcomes [72], a trend observed in other neurovascular conditions [21]. The rarity of the disease and the complexity of decision-making highlight the importance of treatment in specialized centers with multidisciplinary teams.Quality of life: Patients with CMs may experience a significant decline in quality of life following diagnosis, symptomatic hemorrhage, or surgical intervention [17, 39, 65, 77]. This factor is frequently overlooked when the focus is primarily on focal neurological deficits. However, quality of life considerations should play a central role in counseling affected individuals. These factors should also be taken into account during the final treatment decision-making process and in measuring outcomes throughout therapy. Symptomatology: CMs may cause symptoms through their mass effect or more commonly as a result of hemorrhage, which can exacerbate pre-existing symptoms. Hemorrhagic events should be clearly defined as the acute or subacute onset of neurological symptoms (such as headache, seizure, impaired consciousness, or focal neurological deficits) that can be linked to the anatomical location of a CM and confirmed by imaging [4]. Asymptomatic growth, hemorrhage, or the detection of perifocal hemosiderin deposits alone do not meet the criteria for a hemorrhagic event. This definition helps explain why brainstem and intramedullary CMs are often perceived as more aggressive. Their perceived aggressiveness is likely due to their location in highly eloquent CNS structures, where even small hemorrhages can lead to significant neurological deficits. In rare cases, CMs can cause cluster hemorrhages, characterized by recurrent bleeding episodes occurring in rapid succession, leading to rapid neurological deterioration [16]. Symptomatology: CMs may cause symptoms through their mass effect or more commonly as a result of hemorrhage, which can exacerbate pre-existing symptoms. Hemorrhagic events should be clearly defined as the acute or subacute onset of neurological symptoms (such as headache, seizure, impaired consciousness, or focal neurological deficits) that can be linked to the anatomical location of a CM and confirmed by imaging [4]. Asymptomatic growth, hemorrhage, or the detection of perifocal hemosiderin deposits alone do not meet the criteria for a hemorrhagic event. This definition helps explain why brainstem and intramedullary CMs are often perceived as more aggressive. Their perceived aggressiveness is likely due to their location in highly eloquent CNS structures, where even small hemorrhages can lead to significant neurological deficits. In rare cases, CMs can cause cluster hemorrhages, characterized by recurrent bleeding episodes occurring in rapid succession, leading to rapid neurological deterioration [16]. Risk of hemorrhage: Several studies have evaluated the natural history of CMs and identified risk factors for bleeding. Patients who have experienced a prior hemorrhage are at significantly higher risk for future hemorrhages than those in whom the CM was discovered incidentally [5, 41]. This supports the rationale for a conservative, watch-and-wait approach in asymptomatic patients and favors treatment initiation after an initial or recurrent bleeding event. Our own studies suggest that obese patients and those receiving female hormone treatment may also have an increased risk of experiencing hemorrhages [14, 92]. Risk of hemorrhage: Several studies have evaluated the natural history of CMs and identified risk factors for bleeding. Patients who have experienced a prior hemorrhage are at significantly higher risk for future hemorrhages than those in whom the CM was discovered incidentally [5, 41]. This supports the rationale for a conservative, watch-and-wait approach in asymptomatic patients and favors treatment initiation after an initial or recurrent bleeding event. Our own studies suggest that obese patients and those receiving female hormone treatment may also have an increased risk of experiencing hemorrhages [14, 92]. Surgical advances: Recent advances in microsurgical techniques, including neuronavigation [89], imaging [11], and electrophysiological monitoring [51], have enhanced the precision and safety of CM removal. Surgeons now have access to advanced preoperative imaging techniques such as diffusion-tensor MRI with tractography [3], navigated transcranial magnetic stimulation [60], and task-based functional MRI [90]. Depending on the lesion location this can be crucial for preoperative planning, as white matter tracts may be displaced by the CM but can still pass through the adjacent hemosiderin ring, which may be partially resected [12]. For cases involving epilepsy, detailed preoperative epileptological diagnostics, such as video EEG monitoring, are useful [75]. Intraoperative tools such as microsurgical techniques [20, 47], electrophysiological monitoring (cortical and subcortical stimulation) [51, 66], frameless image-guided neuronavigation [89], neuro-navigated ultrasonography [25], and, in certain cases, awake surgery [61], further contribute to minimal surgical morbidity. Surgical advances: Recent advances in microsurgical techniques, including neuronavigation [89], imaging [11], and electrophysiological monitoring [51], have enhanced the precision and safety of CM removal. Surgeons now have access to advanced preoperative imaging techniques such as diffusion-tensor MRI with tractography [3], navigated transcranial magnetic stimulation [60], and task-based functional MRI [90]. Depending on the lesion location this can be crucial for preoperative planning, as white matter tracts may be displaced by the CM but can still pass through the adjacent hemosiderin ring, which may be partially resected [12]. For cases involving epilepsy, detailed preoperative epileptological diagnostics, such as video EEG monitoring, are useful [75]. Intraoperative tools such as microsurgical techniques [20, 47], electrophysiological monitoring (cortical and subcortical stimulation) [51, 66], frameless image-guided neuronavigation [89], neuro-navigated ultrasonography [25], and, in certain cases, awake surgery [61], further contribute to minimal surgical morbidity. Extent of resection: Achieving complete lesion removal is crucial for preventing recurrent hemorrhages and ongoing seizures [45]. Debate exists regarding the optimal surgical approach, whether transsulcal [38] or transparenchymal [47], but most agree that a circumferential, intracapsular approach with en bloc resection provides the best results in supratentorial CM surgery. Resection is typically most effective several weeks after a hemorrhage, when the lesion has clearer delineation from surrounding brain tissue and perifocal edema has resolved, allowing for safer and more effective excision [26]. For lesions in deep or eloquent areas, central debulking followed by piecemeal resection is often necessary and feasible, as CMs lack arterial input. Meticulous care must be taken to avoid leaving residual CM tissue, as complete resection is crucial for surgical success. Extent of resection: Achieving complete lesion removal is crucial for preventing recurrent hemorrhages and ongoing seizures [45]. Debate exists regarding the optimal surgical approach, whether transsulcal [38] or transparenchymal [47], but most agree that a circumferential, intracapsular approach with en bloc resection provides the best results in supratentorial CM surgery. Resection is typically most effective several weeks after a hemorrhage, when the lesion has clearer delineation from surrounding brain tissue and perifocal edema has resolved, allowing for safer and more effective excision [26]. For lesions in deep or eloquent areas, central debulking followed by piecemeal resection is often necessary and feasible, as CMs lack arterial input. Meticulous care must be taken to avoid leaving residual CM tissue, as complete resection is crucial for surgical success. Preserving DVAs: Preservation of DVAs during surgery is important, as these vessels provide normal and usually critical venous drainage for the affected individual. Disruption of a DVA increases the risk of venous thrombosis and infarction, potentially leading to secondary hemorrhage. Some authors propose removing the distal branches of the DVA while preserving the main trunk to prevent CM recurrence, but this approach remains under-researched [88]. Available guidelines and surveys, however, recommend preserving DVAs to avoid secondary complications [2, 16, 81]. Preserving DVAs: Preservation of DVAs during surgery is important, as these vessels provide normal and usually critical venous drainage for the affected individual. Disruption of a DVA increases the risk of venous thrombosis and infarction, potentially leading to secondary hemorrhage. Some authors propose removing the distal branches of the DVA while preserving the main trunk to prevent CM recurrence, but this approach remains under-researched [88]. Available guidelines and surveys, however, recommend preserving DVAs to avoid secondary complications [2, 16, 81]. Epilepsy: In cases of CM-related epilepsy (CRE), the epileptogenic focus can extend beyond the CM itself. In patients with medically refractory seizures, resection should aim to include the hemosiderin-enriched gliotic tissue (if this area is not eloquent) surrounding the CM, as this tissue may also be a source of seizure activity [71]. Removing hemosiderin-enriched tissue is less significant when CMs are resected early in the course of epilepsy, which is generally recommended [28]. Epilepsy: In cases of CM-related epilepsy (CRE), the epileptogenic focus can extend beyond the CM itself. In patients with medically refractory seizures, resection should aim to include the hemosiderin-enriched gliotic tissue (if this area is not eloquent) surrounding the CM, as this tissue may also be a source of seizure activity [71]. Removing hemosiderin-enriched tissue is less significant when CMs are resected early in the course of epilepsy, which is generally recommended [28]. Case load: For deep or eloquently located CMs, surgery should be performed in high-volume centers with neurovascular experience. Evidence suggests that hospitals with higher case volumes are associated with better neurological outcomes [72], a trend observed in other neurovascular conditions [21]. The rarity of the disease and the complexity of decision-making highlight the importance of treatment in specialized centers with multidisciplinary teams. Case load: For deep or eloquently located CMs, surgery should be performed in high-volume centers with neurovascular experience. Evidence suggests that hospitals with higher case volumes are associated with better neurological outcomes [72], a trend observed in other neurovascular conditions [21]. The rarity of the disease and the complexity of decision-making highlight the importance of treatment in specialized centers with multidisciplinary teams. Quality of life: Patients with CMs may experience a significant decline in quality of life following diagnosis, symptomatic hemorrhage, or surgical intervention [17, 39, 65, 77]. This factor is frequently overlooked when the focus is primarily on focal neurological deficits. However, quality of life considerations should play a central role in counseling affected individuals. These factors should also be taken into account during the final treatment decision-making process and in measuring outcomes throughout therapy. Quality of life: Patients with CMs may experience a significant decline in quality of life following diagnosis, symptomatic hemorrhage, or surgical intervention [17, 39, 65, 77]. This factor is frequently overlooked when the focus is primarily on focal neurological deficits. However, quality of life considerations should play a central role in counseling affected individuals. These factors should also be taken into account during the final treatment decision-making process and in measuring outcomes throughout therapy. For patients who are not suitable candidates for surgery—such as those with asymptomatic CMs, no history of hemorrhage, or lesions in high-risk locations—conservative management is recommended [2]. Notably, no formal guidelines exist on the necessity or frequency of imaging for conservatively managed patients, emphasizing the need for individualized care based on clinical judgment. We currently recommend regular follow-up MRI examinations for asymptomatic lesions every 3–5 years. However, immediate imaging is essential if new symptoms occur. Radiosurgery is considered a second-line treatment in sporadic cases, though its direct impact on bleeding risk remains controversial [23, 64]. In familial cases, radiosurgery is generally discouraged, as it may lead to the development of new lesions [2]. Antiepileptic drugs may be used in cases of CM-associated epilepsy as part of the conservative management plan. Endovascular treatment has no role in CM therapy. Surgical treatment of sporadic supratentorial and cerebellar malformations For superficial and non-eloquent CMs, surgical risks are generally low, as these lesions tend to be accessible with minimal risk of morbidity. Several clinical conditions justify surgical resection. There is no definitive consensus among experts regarding the resection of asymptomatic lesions in either children or adults [81]. However, potential benefits include preventing future hemorrhage, alleviating the psychological burden of living with a potentially hemorrhagic lesion, and reducing the need for long-term follow-up. Moreover, resection may help facilitate lifestyle or career decisions by eliminating the lesion as a future risk [2]. In our clinical practice, we typically refrain from excising asymptomatic lesions due to the relatively low risk of an initial symptomatic hemorrhage. Surgery, however, should be considered in symptomatic patients to relieve the mass effect, improve neurological function, and prevent recurrent hemorrhages [2]. The indication for surgery is very strong in patients with CRE when seizures are medically resistant, and there is compelling evidence linking the solitary CM to the epileptic focus [2, 70]. Exceptions may be made for patients who refuse anticonvulsive therapy, are noncompliant with their medications, or suffer from a CM with a high rate of hemorrhages [70]. The timing of surgical intervention for CRE remains a subject of debate. Increasing evidence suggests a correlation between the duration of epilepsy and postoperative outcomes, with longer durations of epilepsy being linked to less favorable postoperative seizure control [15, 19, 27, 73]. This has triggered discussions on whether earlier surgical intervention should be favored in patients with CRE, as prolonged use of antiepileptic drugs and the constant awareness of a potentially hemorrhagic lesion can significantly impact a patient’s quality of life [65]. In contrast, deep-seated lesions such as those in the basal ganglia, thalamus, or other eloquent brain regions present a greater challenge. Surgery for these CMs carries a higher risk of morbidity due to their proximity to vital structures [87]. As a result, surgical intervention should be limited to patients who are severely symptomatic or who have experienced previous hemorrhage(s). For asymptomatic patients, surgery is generally not recommended. There is also considerable debate regarding the appropriate timing of surgery in these cases, with some experts advocating for resection after the first hemorrhagic event [2], while others recommend delaying surgery until after a second hemorrhage to reduce the risk of iatrogenic injury [81]. Scoring systems have been developed to objectively assess the appropriateness of surgery for deep-seated or eloquently located CMs. These scoring systems, combined with clinical guidelines from the Angioma Alliance, can serve as valuable tools for preoperative planning [29]. Surgical treatment of sporadic brainstem malformations Brainstem cavernous malformations (BSCMs) are associated with a high risk of symptomatic hemorrhage and iatrogenic morbidity. The brainstem contains densely packed, highly eloquent nuclei and fiber tracts within a small area, making these lesions particularly challenging to access surgically [54]. For decades, BSCMs were considered a surgical no-man's-land [49]. However, advances in imaging, surgical approaches, and microsurgical techniques have fundamentally changed this view. Today, with careful consideration of surgical indications, these procedures can be performed with favorable outcomes [45]. Surgical intervention in BSCM patients should be reserved for symptomatic lesions. Guidelines do not provide a definitive recommendation on the number of hemorrhagic events that should occur before considering surgery. While some experts consider surgery after the first disabling hemorrhagic event controversial, resection is generally recommended after a second hemorrhagic event [2]. This approach is supported by surveys of experienced neurosurgeons, with the majority favoring resection after a second symptomatic hemorrhage [81]. A comprehensive consensus on BSCMs provides detailed guidance, covering a range of clinical scenarios and influencing factors [16]. The consensus defines hemorrhagic events, management of associated DVAs, handling of postoperative remnants, safe entry zones, and strategies for specific anatomic locations. It also addresses the complexity of BSCMs by distinguishing between lesions that are easily accessible and those that are more difficult to access. For easily accessible brainstem lesions with moderate to severe deficits, resection after the first hemorrhagic event is typically favored. In contrast, for difficult-to-access lesions or those associated with mild deficits, surgery is often postponed until after the second hemorrhagic event. In addition to consensus guidelines, several scoring systems have been proposed to predict surgical outcomes for BSCMs [18, 31, 85]. These tools can complement the Angioma Alliance guidelines and Delphi Consensus Projects, helping clinicians refine their surgical decision-making process. Most experts recommend surgery 2 to 8 weeks after the most recent hemorrhage, once a dissection membrane has formed, post-hemorrhagic swelling has subsided, and gliosis has not yet developed [45]. Surgical treatment of sporadic intramedullary malformations Compared to cerebral CMs, there is relatively limited research on spinal malformations. As a result, the Angioma Alliance guidelines do not provide specific recommendations for spinal CMs beyond suggesting that they be managed similarly to BSCMs due to the high eloquence of the spinal cord [2]. Reviews generally support a conservative approach for asymptomatic spinal CMs, with surgery reserved for patients after a first hemorrhagic event [7, 9, 30]. This aggressive treatment recommendation in symptomatic spinal CMs is justified because of recurrent bleeding and an impaired neurological recovery with each subsequent hemorrhagic event [67]. Surgical decision-making is particularly challenging for difficult-to-reach lesions, such as ventrally located non-exophytic CMs or those in the thoracic or lumbar spinal cord. In these cases, surgery may be postponed until after a second hemorrhagic event, particularly if the patient only has mild neurological deficits [36, 68]. While no expert consensus specifically addresses spinal CMs, there appears to be alignment with BSCM guidelines, with most studies recommending surgery within three months of a hemorrhage [7, 9, 36, 68]. One of the few prospective studies in this field advises caution against early intervention, suggesting that surgery should be delayed until a dissection membrane has formed, similar to BSCMs [43]. Surgical treatment of CNS-CM multiplicity Multiple CMs may develop sporadically or as part of familial cavernomatosis. In patients with familial CMs, germline mutations in the CCM1, CCM2, or CCM3 genes result in the continual formation of new lesions, making curative surgery impossible. Thus, surgical treatment in patients with multiple CMs should be limited to symptomatic lesions, following the same criteria used for solitary symptomatic lesions. Asymptomatic lesions should not be treated surgically [16]. Interestingly, studies indicate that patients with multiple CMs do not necessarily have a higher cumulative risk of hemorrhage compared to those with solitary CMs when analyzed over short-term follow-up periods of 5 years [41, 74]. Ongoing research is focused on determining whether patients with CCM3 mutations are at a higher risk for early onset of symptoms and increased incidence of hemorrhage compared to those with CCM1 or CCM2 mutations, though this remains an area of active investigation [22, 76]. Clinical practice – balancing evidence-based medicine and surgical experience The current data and recommendations for managing CM patients have greatly improved the quality of care at our center and are well integrated into clinical practice. However, extensive clinical experience remains essential for accurately assessing surgical risks and providing informed patient counseling. At our center, patients attend a dedicated clinic, where they receive comprehensive information about their condition and, if necessary, undergo additional diagnostic evaluations. In our approach, asymptomatic patients are generally not considered for surgery, as the operative risk typically outweighs the low risk of spontaneous bleeding in stable CMs. Exceptions may be made for superficial, non-eloquent lesions if there is a strong and well-justified patient preference for surgery. Patients with a history of hemorrhage, as a prior bleeding event is the most significant predictor of future bleeds, are prioritized for surgical evaluation. For superficial supratentorial and cerebellar CMs with a history of bleeding, we offer surgical resection, usually after the first bleed. CMs located in eloquent or cortical areas require additional preoperative imaging and intraoperative monitoring, such as diffusion-tensor MRI with tractography, intraoperative electrophysiology, or awake surgery, to ensure safe resection. Sometimes, a second symptomatic hemorrhage causes an enlargement of the lesions that may then allow a less invasive dissection for the removal. For patients with CRE, surgery is indicated if they have drug-resistant epilepsy, as defined by the ILAE, i.e. failure to achieve seizure freedom with two antiepileptic drugs. Exceptions are made for patients with previous hemorrhages and those who prefer not to undergo long-term medication. Preoperative evaluation of the epileptogenic focus is always necessary to confirm a clear link between the lesion and epilepsy. Deep supratentorial CMs, such as those in the thalamus or basal ganglia, are assessed based on surgical accessibility, neurological morbidity, and patient preference, with surgery potentially considered after two symptomatic bleeding episodes. BSCMs are not surgically treated if asymptomatic. However, unlike the Angioma Alliance guidelines, we adopt a more individualized approach, sometimes opting for surgery after a single bleed. For accessible, superficially located lesions and in patients with moderate to severe deficits, surgery is considered after the first bleed. In contrast, for deep-seated lesions in patients with mild symptoms, we typically defer surgery until a second bleed. For spinal CMs, we are usually more aggressive compared to BSCMs, with surgery considered after the first symptomatic bleed in accessible lesions, because a second hemorrhage in these patients often causes a new irreversible functional deficit. Patients with newly occurring or recurrent hemorrhages are typically admitted to the hospital and monitored for several days, depending on their neurological status, either in the intensive care unit, intermediate care unit, or general ward. Once the MRI has been updated and cluster bleeding has been ruled out, these patients are referred for rehabilitation. Patients with an indication for surgery undergo the procedure approximately 2–6 weeks after the latest hemorrhage (especially in eloquent lesions), while those without a surgical indication enter a "watch and wait" phase. Concerns and perspectives Research into cavernous malformations continues to evolve, with an increasing volume of scientific literature contributing valuable insights into the epidemiology, etiology, pathogenesis, natural history, and therapeutic strategies for these lesions. The development of systematic literature reviews has led to the creation of initial treatment guidelines, such as those from the Angioma Alliance in 2017, which have been supplemented by various consensus projects that address unresolved issues in CM management. However, research into CMs faces significant limitations. The heterogeneity of the disease and the rarity of hemorrhagic events present challenges for designing large, randomized controlled trials, particularly for brainstem surgery. Additionally, existing guidelines and consensus projects may not fully translate into clinical practice, as they are often subject to differing interpretations by proponents of either conservative or surgical approaches. For example, the term "symptomatic" is used broadly in studies to describe a wide range of neurological conditions, from mild dysesthesias to severe motor deficits. Similarly, terms like "easily accessible" or "inoperable" are highly subjective and strongly vary among surgeons. Consequently, the management of CMs remains largely individualized, particularly in smaller neurosurgical centers with a low caseload, though it is increasingly integrated into an evidence-based framework rather than being guided solely by expert opinion. While surgical treatment has advanced and improved in safety, practical limitations persist. Emerging technologies such as radiosurgery [23, 64] and laser interstitial thermal therapy for CRE cases [53] have been applied in the treatment of CMs, though no clear recommendations exist for their use. At this time, we do not recommend using these treatment options, as the literature partly provides conflicting results, is currently limited to short-term follow-up investigations and both approaches do not finally eliminate the risk of re-hemorrhages. Moreover, radiation-induced CM formation is well documented, even after stereotactic radiosurgery [23, 64]. Finally, studies that focus on CM size as a primary outcome parameter also have limited relevance due to the unpredictable nature of the disease. Recent interest in drug repositioning—using existing medications such as HMG-CoA reductase inhibitors, platelet aggregation inhibitors, or beta-adrenergic receptor antagonists—has gained momentum, although current evidence is mixed. Regardless of the pending results, current developments in pharmacotherapy testify to the success of evidence-based medicine in CM research, exemplified by HMG-CoA reductase inhibitors. Enabled by the decoding of increased RhoA/ROCK activity in CMs [24], in vitro and in vivo experiments were conducted several years ago, which ultimately progressed from retrospective observational studies to prospective clinical trials [13, 35, 55, 83]. The results of these studies are eagerly awaited, particularly the forthcoming data from the Phase I/II randomized, placebo-controlled, double-blinded AT CASH EPOC trial (NCT02603328) on atorvastatin, which exhibits pleiotropic effects, including RhoA/ROCK inhibition. At present, the therapeutic use of beta-adrenergic receptor antagonists like propranolol [42, 48] and aspirin [59, 91], which have shown a reduction in bleeding rates in retrospective trials, has not yet been incorporated into our clinical routine recommendations. However, phase 3 trials on propranolol and aspirin are anticipated and could significantly impact future management. The recognition that CMs share characteristics with neoplastic structures has redirected research toward the pharmacological modulation of CM-associated signaling pathways [78]. Preclinical studies are investigating therapies targeting pathways such as RhoA/ROCK (e.g., NRL-1049 or Fasudil), mTOR (e.g., Rapamycin), superoxide dismutase (e.g., REC-994), VEGF (e.g., Semaxanib or Sorafenib), MAP (e.g., Ponatinib), and TIE2 (e.g., Rebastinib). These studies have already led to the initiation of early clinical trials, including the randomized, double-blind, placebo-controlled SYNCAMORE trial on the use of REC-994 [57]. Gene therapies, though theoretically promising for addressing dysfunctional genes, are not yet viable treatment options [40, 84]. Furthermore, efforts are underway to identify biomarkers that could predict bleeding events [33]. The advantage of highly sensitive and specific biomarkers would be the ability to monitor disease progression in a way that complements MRI imaging. While this field of research is still in its early stages, a significant number of potential biomarkers have already been identified. These now require further investigation in follow-up studies to assess their clinical applicability [34, 50, 79]. So far, biomarkers have not yet been incorporated into clinical practice, so diagnosis and monitoring continue to rely on neurological examination and imaging diagnostics. The increasing use of artificial intelligence (AI) across various medical fields suggests that AI will play a significant role in the diagnosis and treatment of neurovascular diseases in the future. This trend is already evident, with a growing body of literature focused on AI-based biomarker discovery for hemorrhage risk prediction [44], identification of targeted therapeutic options [32], and advancements in surgical planning [37]. However, these areas remain subjects of ongoing research and therefore currently do not play a role in our present clinical care. Conclusion Cerebrospinal CMs present a complex clinical challenge due to their unpredictable natural history, variable clinical presentations, and the potential for significant neurological morbidity. Advances in diagnostic imaging, microsurgical techniques, and intraoperative monitoring have improved the outcomes for patients undergoing surgical resection, which remains the cornerstone of treatment for symptomatic or high-risk CMs. However, the decision to intervene surgically must be carefully weighed against the risks and the management of deep-seated or eloquently located lesions remains a subject of debate. Non-surgical treatment, including antiepileptic pharmacotherapy and watchful waiting, is a viable option for patients with asymptomatic or less accessible lesions. Meanwhile, emerging therapies, such as stereotactic radiosurgery and pharmacological approaches targeting specific molecular pathways, may offer future alternatives, although studies are needed to establish their efficacy and safety. The development of evidence-based guidelines and consensus recommendations has provided clinicians with a framework for managing this rare and heterogeneous condition, but the limited availability of high-quality evidence underscores the need for ongoing research.