Cerebral cavernous malformations (CCM) are relatively common vascular malformations in the brain and spinal cord. The familial form of cerebral cavernous malformation (FCCM) is far less common, accounting for only a small minority of cavernous malformations but certainly underdiagnosed. Patients present with a variation of neurologic symptoms, and MR imaging is often the first diagnostic test. Our purpose is to review the MR features of FCCM.
Cerebral cavernous malformations (CCM) or cerebral cavernous angiomas are vascular malformations in the brain and spinal cord. In sporadic cases, up to a third of such cavernous malformations are multiple.
Familial cerebral cavernous malformation syndrome (FCCM) is defined as the presence of multiple CCM (typically five or more) or the occurrence of CCM in at least two members of a family or the presence of a mutation in one of the three genes causing FCCM (Table 1) [1, 2].
|Diagnostic Criteria of FCCM (At Least 1 of the Following Criteria):|
|– the presence of multiple CCM (typically 5 or more)|
|– the occurrence of CCM in at least two members of a family|
|– the presence of a mutation in one of the three genes causing FCCM.|
Due to the dynamic nature of CCM, new lesions may appear at a rate of between 0.2 and 0.4 lesions per patient-year . This explains why older patients present with diffuse CCM, although young patients may present with already diffuse FCCM.
The disease most commonly presents with seizures (38%–55%) and focal neurological deficits (35%–50%) . A (recurrent) cerebral haemorrhage or a nonspecific headache are less frequently encountered symptoms. Exceptionally, a spontaneous paraplegia may occur. However 25–50 percent of individuals with CCM in general remain symptom free throughout their lives . As many persons affected by FCCM remain asymptomatic, an underestimation of its frequency is very probable. Repeated minor bleeding from the lesions may cause degeneration. Resulting dementia and parkinsonism have been reported in patients with FCCM . Autopsy revealed that even 90 percent of persons with a CCM never had symptoms during their lives .
Although cerebral cavernous malformations have been reported in children, the majority of patients present with symptoms between the second and fifth decades.
There is a coexistence of 33 percent between cavernous malformations and developmental venous anomalies [8, 9]. Skin (9%), retinal (5%) and liver lesions have occasionally been reported. Only one of our patients had a cutaneous hemangioma .
CCM are composed of closely clustered and enlarged capillary channels (called caverns) with a single layer of endothelium. The mature vessel wall elements are absent, as is the normal intervening brain parenchyma. They range from a few millimeters to several centimeters and can – due to their dynamic nature – increase (or sometimes decrease) in size and increase in number over time.
The suspicion of FCCM by physical examination, family history, and brain and spinal cord MRI may be confirmed by genetic testing. Three genes are known to cause mutations in FCCM: KRIT-1 (CCM-1), CCM-2, and PDCD-10 (CCM-3) . At least one other unspecified gene (CCM-4) located on the long arm of chromosome three may also cause FCCM . Moreover, an association of parenchymal FCCM and dural-based meningiomas has also been reported in PDCD10-mutated individuals .
FCCM is an autosomal dominant disease. Each child of an individual with FCCM has a 50 percent chance of inheriting the mutation. A much higher incidence of FCCM has been reported in Hispanic-American individuals of Mexican descent [3, 14], in which the proportion of familial cases (in the general population only about 20%) is estimated to be up to 50 percent. These persons seem to be related to a common ancestor with a mutation in the KRIT-1 gene .
In typical cases, MRI shows multiple bilateral and diffuse focal regions of susceptibility induced signal loss (Figure 1) of variable size, well seen on gradient-echo (GRE) sequences, or even better on susceptibility-weighted imaging (SWI). GRE reveal up to triple the number of lesions seen on SE T2-WI, and SWI even reveals an additional tripling [15, 16]. They range in size from a few millimeters to several centimeters in diameter. However, all T2* sensitive sequences result in a degree of blooming, seen as a hazy halo of signal loss around the lesion, and in an overestimation of the actual size of the lesions. Blooming is most obvious around larger lesions (Figure 1).
On MRI, four types of lesions are found (Table 2) . A type 1 lesion with hyperintense core on SE T1-WI and SE T2-WI suggests recent hemorrhage (Figure 2A and 2B). Peripheral oedema is best seen on FLAIR-weighted images (Figure 2C). A reticulated mixed signal core on SE T1-WI and reticulated mixed signal core with surrounding hypointense rim on SE T2-WI is compatible with a type 2 lesion (Figure 3). Hypointense lesion on SE T1-WI and hypointense lesion with hypointense rim SE T2-WI and blooming on GRE-T2* are referred to as type 3 lesions (Figure 4), whereas punctuate hypointense lesions on GRE-T2* or SWI, not seen on SE T1-WI, nor on SE T2-WI, are designated as type 4 lesions (Figure 5).
The identification of type 1 lesions, with a high frequency of recurrent bleeding, and type 4 lesions, which possibly represent new lesions, is important. The lesions may demonstrate changing signal. High signal on SE T1- and T2-WI centrally with a variable degree of perilesional oedema, best seen on fluid attenuated inversion recovery (FLAIR) sequences, suggests a more recent bleeding. The clinical significance of small lesions seen on MRI remains unknown. However, detection of multiple lesions is helpful in distinguishing CCM from FCCM. Moreover, the meticulous evaluation of the number, location, and size of the lesions is important, as there may be an increased risk of hemorrhage with certain analgesic medications (nonsteroidal anti-inflammatory drugs) and the frequently used acetylsalicylic acid. The risk-benefit of medications that increase the frequency of hemorrhage (heparin and coumarin-type drugs) should be weighed thoroughly before use.
CCM can be multiple (in up to a third of the patients). However, in FCCM the number of cavernomas is higher, typically five or more . Hundreds of lesions may be identified, normally increasing with the person’s age. However, young patients may already show numerous CCM (Figure 6).
Lesions may be located either supratentorially (75%) or infratentorially (25%)  (Figure 7). Almost half of the infratentorial lesions occur in the brainstem, and these are frequently associated with symptoms related to hemorrhage. Cavernous malformation can lead to death from severe intracranial hemorrhage, particularly when located in the brainstem . MRI is a useful tool to assess evolution of the number of CCM over time (Figure 8). Screening MRI brain examinations of family members is useful to detect infra- and supratentorial lesions in the brain (Figure 6).
Lesions may be detected only occasionally in the spinal cord (about 5%). Secondary superficial siderosis [18, 19] is exceptional in superficially located lesions. The presence of a single (or less than five) CCM – especially in young persons – without a history of FCCM does not exclude this diagnosis at all (Figure 3). MRI evaluation over time shows the appearance of new lesions and of acute (often asymptomatic) hemorrhages (0.7% lesions per patient-year) that change in signal intensity and increase in size over time . In children, hemorrhage with an aggressive presentation may be more likely.
Intravenous gadolinium contrast administration is not required for the diagnosis of (F)CCM, but contrast-enhanced MRI may be useful in identifying more complex vascular malformations which are sometimes associated with CCM. It may also help to distinguish CCM from other types of vascular brain malformations such as capillary telangiectasia, aneurysms, and arteriovenous malformations and are indicated if surgical resection is considered (detection of the venous drainage in order to preserve it). CCM are rarely visualized on angiography because of the small size of the afferent vessels, the presence of thrombosis, and the relatively low flow in a CCM.
Some clinicians advocate MRI of the spinal cord at the time of diagnosis to serve as a baseline for future follow-up. A control MRI of the brain every one to two years (GRE or SWI) is indicated, also in individuals with obvious new neurologic symptoms. Interpretation may be difficult, as new lesions may be asymptomatic. (Symptomatic) siblings should undergo MRI of the brain to determine presence, size, and location of lesions.
Surgical removal of lesions associated with seizures or focal deficits from recurrent hemorrhage or mass effect may be justified, even when a large number of other lesions is present. Follow-up MRI may evaluate postoperative outcome (lesion resection, bleeding, loss of brain parenchyma).
All other causes of cerebral microhaemorrhages are to be considered.
In cerebral amyloid angiopathy (Figure 11), numerous small foci may have a similar imaging appearance, with a predilection, however, for the peripheral subcortical white matter. There may also be a history of prior (larger) lobar haemorrhage or superficial siderosis.
Neurocysticercosis lesions are more uniform and smaller in size; they may calcify by time and are typically more regularly delineated. CT or quantitative susceptibility mapping  on MRI may confirm the calcifications.
Radiation-induced cavernous malformations are radiologically and even pathologically indistinguishable from FCCM . Anamnesis of previous radiation (frequently during childhood) may reveal the diagnosis.
More rare are hemorrhagic metastases (Figure 14) and de novo bleeding in cerebral vasculitis and radiation vasculopathy.
Microsurgical removal of a CCM lesion may be justified if associated with seizures or focal deficits from recurrent hemorrhage or mass effect. To reveal the structural and the functional abnormalities of a seizure, the data from EEG, MRI, and SPECT may be focused and integrated. The spatial relationships may be demonstrated by co-registering images of the abnormalities on the MRI. A recent technique of subtraction known as ictal SPECT co-registered to MRI (SISCOM) may reveal a hyperperfusion focus. The SISCOM focus and its relationship to the brain may serve as a map for subsequent surgical resection. A follow-up MRI is an excellent technique to monitor the postoperative outcome. MRI control after surgery in a patient with CCM traditionally shows complete resection, a minimal loss of brain parenchyma, and a small marginal region of susceptibility artifact (Figure 15).
The familial form of cerebral cavernous malformation is uncommon. However, as this autosomal dominant pathology presents with a variety of neurological symptoms and as, on the other hand, many persons affected by FCCM remain asymptomatic, radiologists should include FCCM in their differential diagnosis. The presence of a single CCM in an individual, even without a history of FCCM, does not exclude this diagnosis, and a control MRI of the brain after one to two years is advised. MRI with gradient-echo sequences (GRE) or even preferably susceptibility-weighted imaging (SWI) of the brain is indicated for the diagnosis and to serve as a baseline examination. The identification of the type of lesions is important to evaluate the risk of bleeding relapse. The knowledge of the presence of FCCM is important, as there may be an increased risk of hemorrhage with certain medications. Control MRI of the brain is indicated over time, certainly with obvious new neurologic symptoms, and after surgical removal. Symptomatic and possibly even asymptomatic siblings may also undergo MRI of the brain to determine presence, size, and location of the lesions.
The authors declare that they have no competing interests.