Cavernomas

Definition

Well circumscribed benign vascular hamartoma.

Consisting of irregular thick and thin walled sinusoidal vascular channels located within the brain

But lacking intervening neural parenchyma, large feeding arteries, or large draining veins

Aka

Cavernous malformations

Cavernous haemangiomas

Cavernous angiomas

Numbers

Most frequently present in the 40-50s

Small preponderance in women (58%)

incidence of CCMs varies between 0.17 and 0.56 per 100 000 population per year

10% of CNS vascular malformation

10% of paediatric spontaneous cerebral haemorrhage

0.6% of incidental MRI findings

10–20% of CCMs are associated with a developmental venous anomaly

an extreme variation of the normal venous anatomy

Prevalence between 0.16 percent and 0.9 percent,

Familial form has a prevalence of 0.01 percent to 0.03 percent.

Localisation

Supratentorial	60%
Brainstem	30% mainly pons
Basal ganglia	10%
Spinal	Intramedullary, IDEM, extradural, osseous, extradural spinal canal, neuronal

Pathophysiology (Snellings et al 2021)

Features	Sporadic	Hereditary
Fq	50%	20-50%
Genes	MAP3K3 mutation	KRIT1>malcavernin/PDCD10
Multiplicity of lesion	more likely (80-90%)	Less likely 10-20%
Haemorrhagic risk	Higher	Lower
Association with DVAs	Lower	Higher
Radiation exposure	Yes	No

Congenital

A subset of these CCMs acquire a somatic gain of function mutation in PIK3CA (phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha) which fuels lesion growth.

Type	Gene	Chr	Possible function(s)	Notes
CCM1	KRIT1 (k-rev interaction trapped protein 1)	7q11.2-21	Angiogenesis, Inhibit endothelial cells, Apoptosis, Migration	Have other cutaneous vascular lesions Hispanic-American patients of Mexican descent, the proportion of familial cases may be as high as 50% due to a shared mutation in the KRIT1
CCM2	MGC4607 Malcavernin	7p13	Stabilize endothelial cell junctions Maintain endothelial integrity	More likely to be asymptomatic have lower number of cavernomas
CCM3	PDCD10 (programmed cell death 10)	3q26.1	Stimulate cell proliferation regulate angiogenesis Vasculogenesis Regulate apoptotic pathway	More aggressive presentation (greater number of lesions, Scoliosis,, Meningioma, and more frequent haemorrhage and presentation in childhood)
Unspecified gene	ㅤ	3q26.3-27.2	ㅤ	ㅤ

Autosomal Dominant

but with variable of penetrance

Proteins encoded by these CCM genes

are implicated in junction formation between endothelial cells in vessels.
Loss-of-function mutations may result in abnormal blood vessels with gaps between these endothelial subunits

Presentation

Retinal cavernous malformation

Skin venous malformation

Screening

Screen if

adults with CCM(s) + FHx of CCM,
adults with multiple CCM + no FHx.

Reasons

Adult with a single CCM but no family history: 0% to 1%
Adult with multiple CCMs but no family history: 57%
Adult with at least one CCM and a family history: 78% to 94%

See Wilson screening criteria

De Novo formation

Radiation exposure

With latent period of 9 yrs
Immature brain more susceptible
With local gene mutation
Vessel wall changes due to radiation

Radiation induced Sporadic CCM	Non-radiation induced sporadic CCMs
pathologically similar	pathologically similar
multiple lesions	solitary lesion.
Younger age at time of radiation treatment	ㅤ
higher doses of ionizing radiation	ㅤ

Pathology

Macroscopic

Size: 1-5cm

Circumscribed and lobulated with a reddish-purple raspberry appearance

Surrounding brain tissue often contains haemosiderin.

Microscopic

Cavernous malformations show sinusoidal congested vessels arranged back-to-back with little or no interposed brain tissue

Multiple in 50% cases

Can haemorrhage, calcify, thrombose

Pathology "Mulberry lesion": cerebral haemorrhoids

Radiology: popcorn lesions

Light MS: Stains for vWB factor

Blood vessels devoid of muscular and elastic tissue that are lined with endothelial cells that do not have intervening tight junctions.

Electron MS:

Abnormal gapping of tight junction between endothelial cells → leakage of blood

Rare but can form in spinal cord. Associated with spinal radiotherapy

Associated with venous angiomas → do not surgical remove them THEY BLEED

Caverns are filled with blood in various stages of thrombus formation → organisation → dissolution

Haemosiderin deposition and gliosis are found in abutting parenchyma.

Characteristic haemosiderin depositions are present in the basal lamina and intervening neural tissue is absent, which differentiates them from capillary telangiectasia

Natural history of Cavernomas

General: issues with natural history studies

Describes selected patients who did not receive treatment for some reason
Describes a short untreated follow-up of those who did receive treatment.
Large majority of those studies are hospital-based and have not looked at the population in general.

Horne et al 2016: meta-analysis of hospital + population based study

Higher risk of bleed in brainstem than other locations
Higher risk of presenting with ICH or new focal neurological deficits than other presentation methods
5 yrs risk of ICH if left un tx

ㅤ	Without initial presentation of ICH/FND	With initial presentation of ICH/FND
Non brainstem CCM	3.8%	18.4%
Brainstem CCM	8.0%	30.8%

Modified for memorization

Annual risk	Without initial presentation of ICH/FND	With initial presentation of ICH/FND
Non brainstem CCM	1%	4%
Brainstem CCM	2%	8%

50% of the CCM patients become two years seizure-free within 5 years following the diagnosis of epilepsy

Al-Shahi Salman et al. 2012: Population study & Flemming et al. 2012

Annual risk of first-ever CCM-related haemorrhage

Low (0.4–0.6% per year)

Annual risk of a subsequent haemorrhage

much higher (3.8–23% per year)
This risk decreases over time

Variable annual haemorrhage rate of

6.2% in patients who presented with a bleed
2.2% in patients presenting with non-haemorrhage related symptoms and
0.33% in patients with incidental CCMs

Presentation with haemorrhage, male sex and multiplicity of lesions can increase the haemorrhage risk

Taslimi 2016: Meta-analysis of 25 studies

Hemorrhage definition

Acute or subacute onset of symptoms accompanied by radiologic, pathologic, surgical, or CSF evidence of recent extralesional or intralesional haemorrhage

Annual haemorrhage rates per person:

0.3% non-brainstem
2.8% brainstem

Annual RE-haemorrhage rates per person:

6.3% non-brainstem
32.3% brainstem
median 10.5 months to rehaemorrhage

Classification

Zabramski classification of cavernoma

Lesion Type	MRI Signal Characteristics	Pathologic Characteristics	Adult overall 3 year MRI evidence of haemorrhage (Jeon et al 2013)
Type I Subacute	T1: hyperintense core T2: hyper- or hypointense core with surrounding hypointense rim	Subacute hemorrhage, surrounded by a rim of hemosiderin-stained macrophages and gliotic brain	9.47%
Type II	T1: reticulated mixed-signal core T2: reticulated mixed-signal core with surrounding hypointense rim	Loculated areas of hemorrhage and thrombosis of varying ages, surrounded by gliotic, hemosiderin-stained brain; in large lesions, areas of calcification may be seen	4.74%
Type III Chronic	T1: iso- or hypointense core T2: hypointense with a hypointense rim that magnifies the size of the lesion GE: hypointense with greater magnification than T2	Chronic resolved hemorrhage, with hemosiderin staining within and around the lesion	1.43%
Type IV	T1: poorly seen or not visualized at all T2: poorly seen or not visualized at all GE: punctate hypointense lesions	Multiple lesions in the category were pathologically documented as telangiectasias	ㅤ

Differential diagnosis

Hypointensity on hemosiderin-sensitive sequences Haemorrhagic metastases

Cavernous malformation (type IV) Granuloma

Calcium Infectious/inflammatory nodules

Thrombosed arteriovenous malformation

Microbleed due to cerebral amyloid angiopathy

Calcified primary brain tumor

Microbleed due to hypertension

Trauma

History of radiation to the brain

Foreign body: metal/glass

Cardiac surgery

Presentation

Incidentally (25-40%)

Esp: with increase with MRI

Asymptomatic is it because they

haemorrhage in a non eloquent area OR
they don't haemorrhage at all

Seizure 60%

Due to inflammation, perilesional hemosiderin and gliosis

Most common for frontal and temporal lobe cavernomas

Josephson 2011: 5-year risk of first time seizure in cavernoma patients

6% symptomatic
4% incidental (annualised: 0.9%)
Higher risk of epilepsy in Cavernomas but can be well controlled with antiepileptics
In patients with a single seizure, the risk of a recurrent seizure is as high as 94% over 5 years;

Progressive neurological deficit 50%

Dependent on location of cavernoma

Haemorrhage 20% → only for symptomatic radiologically proven

Definition: a clinical event with acute/subacute onset of symptoms with radiological, pathological, surgical or CSF evidence of either recent extralesional or intralesional haemorrhage

Almost all cavernomas that have a hemosiderin around it has had a small leak

Most haemorrhage outcome is good because it is small

Risk of haemorrhage in cerebral CMs (Akers 2017)

Risk of first-time haemorrhage among incidentally discovered CMs is very low (0.08%)
2 factors most consistently associated with increased bleeding risk

Previous haemorrhage: CMs initially presenting with haemorrhage:

Horne 2016

hazard ratio = 5.6 (Horne 2016)
5 yr risk of bleed ICH: 18.4% (Horne 2016)

Salman 2012 lancet

5 yr risk of bleed for 1st ICH: 2.4%
5 yr risk of bleed for recurrent ICH: 29.5%
Salman 2012 summary of all paper

Annual risk of 1st ICH: 0.3-0.6%
Annual risk of recurrent intra-cranial haemorrhage: 3.8-22.9%

Familial cases have higher risk

Location of haemorrhage: brainstem CMs:

hazard ratio = 4.4

Salman 2012 summary of other paper

21-60.2% annual risk

Inconsistent findings reported for risk diﬀerences with female gender, CM size, and CM multiplicity
Annual risk of recurrent haemorrhage declines over time (Horne 2016)

6.2% (1st year) —> 2.0% (5th year)

Higher annual ICH rates are reported in familial CMs (4.3–6%) than in sporadic cases but this could be artefactual
Pregnancy & parturition are not thought to be risk factors for haemorrhage
Risk of haemorrhage from a CM may not be increased by platelet inhibitors or anticoagulation but this is based on uncontrolled studies that likely avoided treatment of patients with recent haemorrhage—> don’t give them anticoagulation if can
No relation of physical activity to haemorrhage from CMs has been identified

Paediatric cavernomas (<25 yrs old)

Do not bleed more often than adults
Al-Holou 2011: n=92 pt with n=110 cavernomas

Haemorrhage rate

In general

1.6% per patient-year
0.9% per cavernoma-year

Symptomatic group

8.0% per patient-year

incidental group

0.2% per patient-year

Previous bleed: higher risk of bleeding
Younger age of bleeding in the familial population

Hydrocephalus

Radiology

19 to 31% are radiologically occult

not sensitive—> miss small lesions
Poor mass effect
15% has calcification

Higher epileptogenicity

MRI:

Gradient echo T2 most sensitive

Popcorn: mixed signal core with low signal rim
There is a difference in MRI features based on presence and absence of bleed and the age of the bleed
Rim of hypointensity can extend and stain the adjacent brain parenchyma (blooming effect) --> making an fake larger lesion

Makes it very much clearly and with greater blooming effect

Angiography:

Does not demonstrate lesion

Hence called angiographically occult lesions, or cryptic vascular malformations

Unless has anomalous venous malformation

Management

Conservative management

Indicated for

Incidental cavernomas
Symptomatic CCMs that are associated with high treatment risks, such as those in eloquent locations (e.g. brainstem)

Repeat MRI 6 monthly - annually for 3 years

Medication

Propranolol use for familial cerebral cavernous malformation

lower risk of intracerebral haemorrhage
Lower risk of new persistent or progressive focal neurological deficit

Evidence

Lanfranconi 2023 Pilot Treat_CCM trial

Propranalol vs standard care N=83
The incidence of symptomatic intracerebral haemorrhage or focal neurological deficit was

propranolol = 1·7 cases per 100 person-years
standard care = 3·9 cases per 100 person-years

Studies comparing Intervention vs conservative

Intervention

No observational studies comparing microsurgical resection with radiosurgery

Overall incidence of death, haemorrhage or new/worsened permanent deficit after resection or SRS, is approximately 6 per 100 person-years.

Microsurgery

Indications

Symptomatic cavernomas in non eloquent areas or brainstem

multiples haemorrhages,
CCMs with mass effect causing symptoms, and
patients with intractable epilepsy with an identified epileptogenic focus

If really need to go to brainstem look for the following factors

Cavernoma should reach the pial or ependymal surface, in order to minimize operative neurological deficits
Use relatively safe entry zones
Use intraoperative electrophysiological

Goal:

complete excision of malformation
If for seizures excision of surrounding haemosiderin brain)

Technique

Surgery planning

Find shortest trajectory to cavernoma
Avoiding ‘blind-spots’ along the surgical corridor.

Done by

Utilising an extended line to the skull that incorporates one point in the centre of the CCM and a second point at where the CCM manifests at the pial surface

Identification of Cavernoma

Yellow discolouration
Neuronavigation

For non eloquent Cavernomas

Circumferential dissection of the glial plane, emptying of the caverns to achieve internal debulking with subsequent facilitation of resection, by rolling it away from the glial plane.
Ideally, an en-bloc resection should be attempted, since this minimizes the risk of incomplete resection.

For eloquent CCMs

requires an entirely different approach.
An ‘inside-out’ dissection is preferable, which is done by entering and emptying the CCM to make it collapse on itself.
This is followed by careful dissection of the plan between capsule and eloquent brain tissue to minimize surrounding brain injury

If the CCM is associated with a DVA, the latter should be preserved during resection as it represents part of the normal venous system.

Only small branches of the DVA can be cauterized but sacrifice of the main trunk(s) may lead to venous infarction with potentially disastrous consequences

Outcome

Poor prognostic factors after surgery

Brainstem CCM
Increasing age
Male sex

Kivelev et al. 2011

Outcomes according to a grading scale
Location

1 point: Supratentorial
2 points: Infratentorial basal ganglia, or spinal

Pre-existing neurological deficit (1 point if present).
A good functional outcome (Glasgow outcome score 5: normal activities, minor neurological deficits) was achieved in

Grade 1: 87%
Grade 2: 79%
Grade 3: 46%

Seizure freedom in CCM patients with medically refractory epilepsy can be allegedly achieved approximately 88% (seizure freedom for 80% of patients with chronic epilepsy due to CCMs and 91% CCM patients with sporadic seizures-von der Brelie et al., 2013).

Depending on the definition of seizure and what defines seizure freedom

Moultrie 2014

Non randomized group study

Surgery is worse than conservative management for both the risk of first haemorrhage or Focal neurological deficit and oxford handicap score
Surgical risks were lower if

patients presented with haemorrhage or
in patients with CCMs outside the brainstem

Mortality was comparable;

Surgery group: one patient (4.0%) in the surgical group died after a seizure
Conservative group: four patients (3.7%) experienced CCM-related deaths

Surgery brings forward all the risk of CCM if completely excised when compared to conservative

With further follow-up, the difference between conservative and surgery might diminish because of new neurologic events in the conservatively managed group, but this may not turn out to be the case because the risk of recurrent CCM haemorrhage seems to decline over time and CCM sometimes regrow after surgery and bleed again.

Risk of postoperative complications (composite of death within 30 days, symptomatic haemorrhage, new or worsened focal neurological deficit) was estimated at

7.7% for studies looking at resection of predominantly supratentorial CCMs
50% for resection of brainstem CCMs

SRS

Aim of SRS: prevention of haemorrhage by inducing thrombosis-obliteration in the cavernoma.

Indicated

Highly eloquently or surgically inaccessibly areas, such as basal ganglia and brainstem (where the cavernoma does not present at the surface).
Small <3cm
Seizure control

Radiosurgery not advised in familial CCM

Concern of proliferation of further CMs as RT in general can cause CCM

Controversial as outcome seem similar to placebo

Poorthuis 2019

After SRS for CCM, the annual incidences of death, ICH, and FND are <5% and seem comparable to outcomes without SRS.
As the risk of recurrent intracranial haemorrhage declines over time (Al-Shahi Salman et al., 2012; Flemming et al., 2012) it is questionable if the reduced haemorrhage risk after SRS that some studies have described can be attributed to this treatment
As CCMs are angiographically occult lesions, the effectiveness of SRS cannot be assessed by imaging as in AVMs, but relies solely on the basis of occurrence of haemorrhage during follow-up period

CARE Pilot RCT

(n33) medical management + surgery (Resection or SRS) VS
(n34) medical management alone

Primary clinical outcome

Symptomatic intracranial haemorrhage or new persistent or progressive non-haemorrhagic focal neurological deficit due to cerebral cavernous malformation or surgery during at least 6 months of follow-up.
Occurred in

6% medical management and surgery
6% medical management alone

Dose

12-15Gy
Given at least 3 months after last bleed
Avoid DVA

Outcome

Morbidity

Post-treatment haemorrhages

Present in 7% patients after SRS

Adverse radiation effect.

6.7% data from Sheffield

Seizure free

CCM patients with medical refractory epilepsy reported that about 40–50% of patients were seizure- free at follow-up
epileptogenic cavernomata demonstrated 87% of patients were without seizures after surgery compared to 64% after SRS (Hsu et al., 2007)

Pregnancy and CCM

Provide prepregnancy counselling regarding genetic risk in familial form

If the patient has a seizure disorder,

consider the AED with the lowest teratogenic potential
Advise folate supplementation if the patient is taking seizure medication during pregnancy MRI without contrast, if needed, during pregnancy or postpartum while breastfeeding

Vaginal delivery is an acceptable mode of delivery unless recent haemorrhage or neurologic deficits preclude such