REVIEWARTICLE - VASCULAR

Comparative studies of the diagnosis and treatmentof cerebral cavernous malformations in adults:systematic review

Michiel Poorthuis & Neshika Samarasekera &

Krystle Kontoh & Ian Stuart & Buddug Cope &

Neil Kitchen & Rustam Al-Shahi Salman

Received: 1 August 2012 /Accepted: 10 January 2013 /Published online: 31 January 2013# Springer-Verlag Wien 2013

AbstractBackground Cerebral cavernous malformation (CCM) man-agement decisions are usually made after CCM diagnosis issuspected or definitively diagnosed on axial imaging byindirectly comparing a surgeon’s estimate of operative mor-bidity and mortality against published estimates of CCMuntreated clinical course.Methods We used comprehensive electronic strategies tosearch OVID Medline and EMBASE for original studies

published before 2011 of ≥20 adults with CCM that (a)evaluated diagnostic test accuracy, or (b) compared treat-ment with microsurgery or stereotactic radiosurgery againstconservative management in a concurrent or historical con-trol group and reported clinical outcome(s). We used theGrading of Recommendations Assessment, Developmentand Evaluation (GRADE) Working Group’s approach toidentify level 1 or level 2 studies according to the OxfordCentre for Evidence-Based Medicine’s 2011 criteria.Results We found one eligible diagnostic test accuracystudy of 72 patients with brain masses accompanied byvasogenic edema and substantial amounts of blood, whichfound that hyperintense perilesional signal on T1-weightedmagnetic resonance imaging could differentiate CCM fromother causes with excellent specificity (98 %) and reason-able sensitivity (62 %). We found five potentially eligibleobservational studies of adults with a CCM that had alreadybled, but none met level 2 criteria for a “dramatic” effect(the conventionally calculated probability of the two groupsof observations coming from the same population should beless than 0.01 and a rate ratio greater than 10). We found 11potentially eligible observational studies of adults withCCM and epilepsy, but nine studies did not demonstratedramatic effects and the remaining two studies showeddramatic effects, but they were at high risk of bias.Conclusions To address the absence of level 1 or 2 evidenceto support CCM treatment decisions, there is a need forlarge studies of CCM treatment with a concurrent controlgroup, ideally with randomized treatment allocation.

Keywords Cerebral cavernous malformations .

Microsurgery . Stereotactic radiosurgery . Prognosis

Electronic supplementary material The online version of this article(doi:10.1007/s00701-013-1621-4) contains supplementary material,which is available to authorized users.

M. PoorthuisDepartment of Neurology, Rudolf Magnus Institute ofNeuroscience, University Medical Centre Utrecht, PO Box 85500,3500 GA Utrecht, The Netherlands

N. Samarasekera : R. Al-Shahi Salman (*)Division of Clinical Neurosciences, Western General Hospital,University of Edinburgh, Bramwell Dott Building,Edinburgh EH4 2XU, UKe-mail: rustam.al-shahi@ed.ac.uk

K. Kontoh : B. CopeGenetic Alliance UK, Unit 4D, Leroy House, 436 Essex Rd,London N1 3QP, UK

I. StuartCavernoma Alliance UK, Suite 4, Somerleigh Gate,Somerleigh Road,Dorchester DT1 1TL, UK

N. KitchenVictor Horsley Department of Neurosurgery, National Hospitalfor Neurology and Neurosurgery, Queen Square,London WC1N 3BG, UK

Acta Neurochir (2013) 155:643–649DOI 10.1007/s00701-013-1621-4

Introduction

Cerebral cavernous malformations (CCMs) have a typicalappearance on brain magnetic resonance imaging (MRI)(see Fig. 1) [39], and in brain MRI studies they affect onein approximately 600 neurologically asymptomatic people[33]. Probably due to the increasing availability and use ofbrain MRI, annual CCM detection rates appear to be risingfrom approximately one per 500,000 in the USA in 1965–1992 to three per 500,000 in Scotland in 1999–2000 [4, 11].When neurological symptoms lead to a CCM diagnosisin adults, one-quarter are due to intracranial hemor-rhage, another quarter are due to a focal neurologicaldeficit without radiographic evidence of recent hemor-rhage (FND) [2], and the remainder present with epi-leptic seizures [4, 26].

The annual risk of a first-ever symptomatic ICH fromCCM has been consistently estimated to be 0.4–0.6 % (seeFig. 2). After a first ICH, the annual risk of a second ICH ishigher, but reported estimates vary widely between 3.8 and22.9 % (see Fig. 2). Several studies have identified that therisk of a second ICH decreases over time [3, 7, 16, 51], butstudies have not been consistent about whether female sexor brainstem CCM location raise the risk of ICH, and thereare no prognostic models.

Microsurgery and stereotactic radiosurgery are used totreat CCM in standard clinical practice in many parts of theworld. When deciding whether or not to treat a CCM,doctors usually compare the expected risks and benefits ofthese treatments reported in case series against publishedestimates of untreated clinical course in other populations

(sometimes extrapolating the published risks reported overonly a few years to their patient’s lifetime). However, thisindirect comparison is not ideal because of the uncertaintiesabout CCM untreated clinical course, and the likely varia-tion in treatment outcome according to characteristics of thepatient, CCM, and treatment modality.

Therefore, we sought to determine whether wecould develop CCM management guidelines, followingthe Grading of Recommendat ions Assessment ,Development and Evaluation (GRADE) WorkingGroup’s approach (www.gradeworkinggroup.org) [5].The GRADE approach to formulating recommendationsfor clinical guidelines involves a systematic and explicitapproach to making judgments, with reference tostrict criteria about the methodological quality of theavailable studies (in which comparisons between groupsusually involve assessments of differences in frequencydistributions).

Methods

Identification of questions

We formulated our specific management questions by con-sensus, using the PICO principle, which involves dissectingeach clinical question into its constituent parts and restruc-turing it into four components: Patients [i.e., adults withCCM], Intervention/Indicator, Comparator/Control, and theOutcome of interest. We identified important outcomes indiscussion with the two patient support groups involvedwith the creation of these guidelines. We judged the relativeimportance of the outcomes, and prioritized death and newor worsened clinically symptomatic FNDs (whether or notnew ICH had been confirmed by imaging or pathologicalexamination).

Literature search

We used electronic strategies (see Online Resource 1) tosearch for journal articles about CCMs published prior toJanuary 1, 2011 and indexed in OVID Medline and Embase.Pairs of the group of three reviewers (NS, MP, and RA-SS)reviewed the titles and abstracts of eligible articles andexcluded articles if: they were reviews and did not reportoriginal data, if they did not address our specific manage-ment questions, if they were exclusive to children withCCMs, if they reported fewer than 20 adults with CCMs,if we were unable to extract relevant data from the article, orif they were not designed to address our specific PICOquestions (see Online Resource 2). If there were disagree-ments or uncertainties amongst the pair of reviewers, theywere arbitrated by a third reviewer.

Fig. 1 CCM on brain magnetic resonance imaging. Cross-section-al T2-weighted magnetic resonance imaging, demonstrating a sol-itary cerebral cavernous malformation (arrows) in the temporallobe of the brain

644 Acta Neurochir (2013) 155:643–649

Grading the quality of the evidence

Pairs of the group of three reviewers (NS, MP, and RA-SS)independently graded the quality of evidence available ineach of the remaining articles pertinent to diagnostic andtherapeutic questions using the Centre for Evidence-BasedMedicine (CEBM) Levels of Evidence published in 2011(http://www.cebm.net/index.aspx?o=5653). For studies ofdiagnostic test accuracy, we sought studies graded as level1 (systematic review of cross sectional studies with a con-sistently applied reference standard and blinding) or level 2(individual cross-sectional studies with consistently appliedreference standard and blinding). For studies of the benefitsand harms of treatments, we sought studies graded as level 1(systematic review of randomized trials or n-of-1 trials) orlevel 2 (randomized trial or observational study with dra-matic effect). We used a published definition of a dramaticeffect in an observational treatment study, “a sufficientlyextreme difference between the outcome ranges for treatedand untreated patients might be defined by two rules: (a) thatthe conventionally calculated probability of the two groupsof observations coming from the same population should beless than 0.01 and (b) that the estimate of the treatmenteffect (rate ratio) should be large… We suggest that rateratios beyond 10 are highly likely to reflect real treatmenteffects, even if confounding factors associated with thetreatment may have contributed to the size of the observedassociations.” [20]. If there appeared to be a dramatic effectin an observational treatment study we also judged whether

the risk of bias in the study was high, and if it was the articlewas excluded (see Online Resource 2). Three reviewers(NS, MP and RA-SS) agreed on the overall quality of theevidence relevant to each PICO question, which was avail-able after this selection process, and other members of theguidelines panel approved the final grading.

Results

Diagnostic test accuracy of brain imaging to diagnose CCM

Prior to the use of brain MRI in clinical practice in the1980s, CCMs were often classified as angiographically oc-cult (or “cryptic”) vascular malformations if histologicalexamination had not been undertaken to definitively diag-nose CCM [48]. The typical appearance of CCMs on T1-and T2-weighted MRI sequences was described in an un-blinded study of ten patients with pathologically confirmedCCM in 1987 [39]. Awider variety of MRI appearances hasbeen described subsequently [42, 55]. MRI sequences sen-sitive to the magnetic susceptibility artefact of iron (such asgradient echo and susceptibility weighting) have been foundto be more sensitive for the detection of multiple CCMs incomparison to conventional spin echo in studies undertakenin families with inherited CCMs (but without pathologicalconfirmation) [13]. The detection of recent hemorrhagefrom a CCM is dependent on the use of the appropriateimaging modality at the appropriate time to detect acute or

Fig. 2 Risk of symptomatic intracranial hemorrhage during fol-low-up in studies of the untreated clinical course of >20 partic-ipants with cerebral cavernous malformations. Areas of point

estimates are proportional to the sample size of each study. Errorbars represent 95 % confidence intervals (if available or calcula-ble) [1, 3, 7, 16, 18, 19, 21, 27, 29, 31, 32, 36, 37, 40, 51, 55]

Acta Neurochir (2013) 155:643–649 645

subacute blood products. This is relevant to both clinicalpractice and research, which has led to definitions andreporting standards for diagnosing CCM hemorrhage [2],although these criteria have not been validated.

We sought studies testing the accuracy of CCM diagnos-tic criteria applied to a consecutive series using a patholog-ical reference standard and MRI as the index test.

We excluded potentially eligible studies because: theywere restricted to patients meeting CCM MRI diagnosticcriteria thus precluding identification of false negatives [17],they were restricted to lesions consistent with CCMs onhistological examination thus precluding identification offalse positives [42, 48], observers were not blinded to theindex test or reference standard [9, 10, 35, 48], or we wereunable to extract relevant data [9, 10].

Our criteria were met by one study of 72 patients withbrain masses accompanied by vasogenic edema and sub-stantial amounts of blood, which found that a T1 hyperin-tense perilesional signal could differentiate CCM from othercauses with excellent specificity (98 %) and reasonablesensitivity (62 %) [54].

Treatment

We sought studies of CCM treatment involving at least 20adults that examined surgical resection [8] and/or stereotac-tic radiosurgery [45], in which a group of adults receivingtreatment was compared to either another group receiving adifferent treatment or to a conservatively managed(untreated) group of adults. We stratified studies by adults’mode of clinical presentation because the future risks ofhemorrhage from CCMs appear are higher if the CCMinitially comes to medical attention with a hemorrhage.

How should an adult with an incidentally discovered CCMbe managed?

We did not find any eligible studies of adults with inciden-tally discovered CCMs, comparing two different manage-ment strategies for reducing the risk of future ICH or FND,or improving functional outcome.

How should an adult with a CCM that has caused one ICHor FND be managed?

We found five potentially eligible observational studies ofadults with CCMs that had already caused one ICH or FND,comparing two different management strategies for reducingthe risk of future ICH or FND, or improving functionaloutcome [15, 24, 31, 46, 53]. Three studies compared sur-gery with conservative management in a total of 81 adultswith brainstem CCMs [15, 24, 46]; one study comparedgamma knife stereotactic radiosurgery with conservative

management in 41 adults with CCMs in any location [53];and one study compared surgery with gamma knife stereo-tactic radiosurgery with conservative management in 68adults with deep and brainstem CCMs [31]. However, noneof these studies was randomized and none demonstrateddramatic effects according to the CEBM criteria [20].

How should an adult with a CCM that has causedmore than one ICH or FND be managed?

We did not find any eligible studies of adults with CCMsthat had caused more than one ICH or FND, comparing twodifferent management strategies for reducing the risk offuture ICH or FND, or improving functional outcome.

How should an adult with a CCM that has caused epilepticseizure(s) be managed?

We found 11 potentially eligible observational studies ofadults with CCMs that had caused a first seizure or epilepsy,comparing two different management strategies for reducingthe risk of future seizures or improving functional outcome[6, 14, 23, 25, 28, 34, 41, 43, 44, 49, 52]. Two studiesappeared to show a dramatic effect, but we excluded bothof them because we judged them to be at high risk of bias[34, 44]. One study compared 14 adults with CCMs under-going surgical resection with 16 adults with CCM undergo-ing gamma knife stereotactic radiosurgery, but treatmentselection bias, unknown differences in epilepsy severity,and unblinded outcome assessment may have accountedfor these findings [44]. The other study compared 16 adultswith CCMs undergoing surgical resection to 15 adults whowere managed conservatively, but unknown differences inthe intractability of patients’ epilepsy and antiepileptic drugtherapy, as well as unblinded outcome assessment, mayhave accounted for these findings [34]. Nine further studiesdid not demonstrate dramatic effects: seven compared sur-gery with conservative management in a total of 204 adultswith supratentorial CCMs [6, 14, 25, 28, 41, 43, 49], onestudy compared surgery with radiosurgery in 29 adults withCCMs in any supratentorial location [23], and one studycompared lesionectomy with other types of surgical resec-tion in 56 adults with CCMs in any supratentorial location[52].

How should a CCM be managed in mothersbefore, during, and after birth?

We did not find any studies of mothers with CCMs, com-paring two different management strategies before, duringor after birth for reducing the risk of future ICH or FND, orimproving functional outcome. We excluded a few casereports of different outcomes for mothers with CCMs during

646 Acta Neurochir (2013) 155:643–649

pregnancy, and one observational cohort study of 85 womenthat did not find any ICH or FND during pregnancy[36]. We found one narrative review article stating thatcaesarean section may be indicated and pregnancy maybe contraindicated for some women with CCMs [30],but we could not support this finding in the light ofcurrent evidence.

Should the existence of a CCM influence the prescriptionof antithrombotic or thrombolytic medications?

We did not find any studies comparing two different antith-rombotic or thrombolytic management strategies for adultswith vaso-occlusive disease who also had a CCM. Weexcluded a few case reports of different outcomes for adultswith CCMs who received antithrombotic or thrombolytictherapy [12, 22, 38, 47]. We found one narrative reviewarticle that recommended that anticoagulant treatmentshould be contraindicated for adults with CCM [30], butwe could not support this finding in the light of currentevidence.

Discussion

This systematic review did not find any studies of thediagnosis and treatment of CCMs of level 1 or 2 qualityaccording to the Centre for Evidence-Based Medicine’s2011 criteria, which enabled us to make few specific rec-ommendations when writing guidelines for the clinical in-vestigation and management of adults with CCMs(www.cavernoma.org.uk/opus452.html). Another recentsystematic review of studies of surgery for adults withCCMs and intractable epilepsy confirmed the methodolog-ical problems which we have found with the entire body ofevidence supporting CCM treatment [50]. Some of theobservational studies found encouraging effects associatedwith treatment, but none of these met the published statisti-cal threshold for a dramatic effect in a non-randomizedobservational study (20).

The implication of our findings for clinical practice is thatwhether and when to treat CCMs remains uncertain in theabsence of observational studies of CCM treatment with acomparison group, at low risk of bias, showing a dramaticeffect. The magnitude of the benefit shown in only a fewcomparative studies of CCM treatment has not been largeenough to discount the need for randomized controlled trials[20]. A growing body of evidence indicates that the risk ofICH from CCMs declines over time [3, 7, 16, 51]. Firstly,this questions whether the declining annual rate of ICH afterstereotactic radiosurgery indicates anything other than thenatural history of CCM haemorrhage. Secondly, this ques-tions whether extrapolation of annual CCM hemorrhage

rates to patients’ lifetimes is justified. Therefore, indirectcomparison of the early risks of CCM treatment from caseseries to extrapolated estimates of the lifetime risk of CCMhemorrhage among patients in different populations may beflawed, questioning the basis for decisions about CCMtreatment in current practice.

The strengths of this systematic review are: the involve-ment of patients and representatives in determining themanagement questions and outcomes that were addressed;a comprehensive electronic literature search strategy; thor-ough independent evaluation of all the relevant literature bytwo reviewers; and reference to explicit methodologicalmarkers of study quality in the CEBM 2011 Levels ofEvidence. However, these criteria rendered us unable toidentify any studies that were of sufficient methodologicalquality to make recommendations for clinical practice.Amongst the many published studies reporting the outcomeof CCM treatment in case series or comparative studies,there are encouraging observations which should not bedisregarded. Despite the imprecision of many of the effectestimates and infrequency of comparison groups in manyobservational studies of CCM, experts may find their resultsinformative. Ultimately, these observations provide hypoth-eses to be confirmed or refuted in future studies with clearlydefined eligibility criteria, larger sample sizes, and experi-mental designs.

Therefore, the implications of our findings for futureresearch are that large, comprehensive observational stud-ies/registries, at low risk of bias, comparing CCM treat-ments against conservative management in a concurrentcontrol group are required. The findings of these controlledobservational studies are most needed for groups of patientsfor whom the superiority of CCM treatment or conservativemanagement is particularly uncertain, or where the effects oftreatment (such as stereotactic radiosurgery) are more un-certain. If these do not show dramatic differences betweenconservative management and CCM treatment, then ran-domized controlled trials—accounting for known prognos-tic factors [3, 7, 16]—would be justified to resolve thischallenging dilemma for neurosurgeons, neurologists, andtheir patients.

Funding

RA-SS has received funding for research from the MedicalResearch Council, the Stroke Association, Chest Heartand Stroke Scotland, the Chief Scientist Office of theScottish Government Health Department, and CavernomaAlliance UK. NS has received funding from the MedicalResearch Council and the Stroke Association. MP, KK,BC, IS, and NK have not declared any competinginterests.

Acta Neurochir (2013) 155:643–649 647

Acknowledgments The creation of these guidelines was jointlyfunded by Cavernoma Alliance UK and a Facilitating Networks grantto Genetic Alliance UK. NS was funded by a Medical ResearchCouncil/Stroke Association clinical research training fellowship. RA-SS was funded by a Medical Research Council senior clinicalfellowship.

Conflicts of interest None.

References

1. Aiba T, Tanaka R, Koike T, Kameyama S, Takeda N, Komata T(1995) Natural history of intracranial cavernous malformations. JNeurosurg 83(1):56–59

2. Al-Shahi Salman R, Berg MJ, Morrison L, Awad IA, AngiomaAlliance Scientific Advisory Board (2008) Hemorrhage from cav-ernous malformations of the brain: definition and reporting stand-ards. Stroke 39(12):3222–3230

3. Al-Shahi Salman R, Hall JM, Horne MA, Moultrie F, JosephsonCB, Bhattacharya JJ, Counsell CE, Murray GD, Papanastassiou V,Ritchie V, Roberts RC, Sellar RJ, Warlow CP (2012) Untreatedclinical course of cerebral cavernous malformations: a prospective,population-based cohort study. Lancet Neurol 11(3):217–224

4. Al-Shahi R, Bhattacharya JJ, Currie DG, Papanastassiou V, RitchieV, Roberts RC, Sellar RJ, Warlow CP, Scottish IntracranialVascular Malformation Study Collaborators (2003) Prospective,population-based detection of intracranial vascular malformationsin adults: the Scottish Intracranial Vascular Malformation Study(SIVMS). Stroke 34(5):1163–1169

5. Atkins D, Best D, Briss PA, Eccles M, Falck-Ytter Y, Flottorp S,Guyatt GH, Harbour RT, Haugh MC, Henry D, Hill S, Jaeschke R,Leng G, Liberati A, Magrini N, Mason J, Middleton P, MrukowiczJ, O’Connell D, Oxman AD, Phillips B, Schunemann HJ, EdejerTT, Varonen H, Vist GE, Williams JW Jr, Zaza S (2004) Gradingquality of evidence and strength of recommendations. Br Med J328(7454):1490–1494

6. Banfi P, Guaschino E, Delodovici ML, Peron S, Tomei G, Bono G(2006) Cerebral cavernous angiomas and epilepsy: clinical featuresand therapy in 20 cases. Bollettino—Lega Italiana ControL’Epilessia 133–134:125–127

7. Barker FG II, Amin-Hanjani S, Butler WE, Lyons S, Ojemann RG,Chapman PH, Ogilvy CS (2001) Temporal clustering of hemor-rhages from untreated cavernous malformations of the centralnervous system. Neurosurgery 49(1):15–24

8. Bertalanffy H, Benes L, Miyazawa T, Alberti O, Siegel AM, SureU (2002) Cerebral cavernomas in the adult. Review of the litera-ture and analysis of 72 surgically treated patients. Neurosurg Rev25(1–2):1–53

9. Biondi A, Scialfa G (1988) Morphological and blood flow MRfindings in cerebral vascular malformations. J Neuroradiol 15(3):253–265

10. Biondi A, Scotti G, Scialfa G, Landoni L (1986) Magnetic reso-nance imaging of cerebral cavernous angiomas. Acta Radiol Suppl369:82–85

11. Brown RD Jr, Wiebers DO, Torner JC, O’Fallon WM (1996)Incidence and prevalence of intracranial vascular malformationsin Olmsted County, Minnesota, 1965 to 1992. Neurology 46(4):949–952

12. Califf RM (1999) Glycoprotein IIb/IIIa blockade and thrombo-lytics: early lessons from the SPEED and GUSTO IV trials. AmHeart J 138(1 Pt 2):S12–S15

13. Campbell PG, Jabbour P, Yadla S, Awad IA (2010) Emergingclinical imaging techniques for cerebral cavernous malformations:a systematic review. [Review]. Neurosurg Focus 29(3):E6

14. Congia S, Saba S, Cannas A, Mascia MM, Tuveri A, Casula A,Onali E, Pinna F, Mascia G, Mela A (2001) Epileptic seizures andcavernous angiomas. Bollettino—Lega Italiana Contro L’Epilessia113–114:163–166

15. Esposito P, Coulbois S, Kehrli P, Boyer P, Dietemann JL,Rousseaux P, Auque J, Maitrot D (2003) Place of the surgery inthe management of brainstem cavernomas. Results of a multi-centric study. [French]. Neurochirurgie 49(1):5–12

16. Flemming KD, Link MJ, Christianson TJ, Brown RD Jr (2012)Prospective hemorrhage risk of intracerebral cavernous malforma-tions. Neurology 78(9):632–636

17. Frischer JM, Pipp I, Stavrou I, Trattnig S, Hainfellner JA, Knosp E(2008) Cerebral cavernous malformations: congruency of histo-pathological features with the current clinical definition. J NeurolNeurosurg Psychiatry 79(7):783–788

18. Fritschi JA, Reulen HJ, Spetzler RF, Zabramski JM (1994)Cavernous malformations of the brain stem. A review of 139cases. Acta Neurochir (Wien) 130(1–4):35–46

19. Ghannane H, Khalil T, Sakka L, Chazal J (2007) Analysis of aseries of cavernomas of the central nervous system: 39 non oper-ated cases, 39 operated cases, 1 dead. [French]. Neurochirurgie 53(2–3 Pt 2):217–222

20. Glasziou P, Chalmers I, Rawlins M, McCulloch P (2007) When arerandomised trials unnecessary? Picking signal from noise. Br MedJ 334(7589):349–351

21. Hasegawa T, McInerney J, Kondziolka D, Lee JY, Flickinger JC,Lunsford LD (2002) Long-term results after stereotactic radiosur-gery for patients with cavernous malformations. Neurosurgery 50(6):1190–1197

22. Henninger N, Ahmad N, Morris JG (2010) Intravenous thrombol-ysis in a patient with known cavernous malformation: a first casereport. Am J Emerg Med 28(1):117.e1–117.e3

23. Hsu PW, Chang CN, Tseng CK, Wei KC, Wang CC, Chuang CC,Huang YC (2007) Treatment of epileptogenic cavernomas: surgeryversus radiosurgery. Cerebrovasc Dis 24(1):116–120

24. Huang AP, Chen JS, Yang CC, Wang KC, Yang SH, Lai DM, TuYK (2010) Brain stem cavernous malformations. J Clin Neurosci17(1):74–79

25. Iakovlev G, Devaux B, Ghossoub M, Beuvon F, Brami F, Roux FX(2005) Cerebral cavernomas, epilepsy and seizures. Natural historyand therapeutic strategy. [French]. Neurochirurgie 51(1):3–14

26. Josephson CB, Leach JP, Duncan R, Roberts RC, Counsell CE, Al-Shahi Salman R, Scottish Audit of Intracranial VascularMalformations (SAIVMs) steering committee and collaborators(2011) Seizure risk from cavernous or arteriovenous malfor-mations: prospective population-based study. Neurology 76(18):1548

27. Kim DS, Park YG, Choi JU, Chung SS, Lee KC (1997) Ananalysis of the natural history of cavernous malformations. SurgNeurol 48(1):9–17

28. Kivelev J, Niemela M, Kivisaari R, Dashti R, Laakso A,Hernesniemi J (2009) Long-term outcome of patients with multi-ple cerebral cavernous malformations. Neurosurgery 65(3):450–455

29. Kondziolka D, Lunsford LD, Kestle JR (1995) The natural historyof cerebral cavernous malformations. J Neurosurg 83(5):820–824

30. Labauge P, Denier C, Bergametti F, Tournier-Lasserve E (2007)Genetics of cavernous angiomas. Lancet Neurol 6(3):237–244

31. Mathiesen T, Edner G, Kihlstrom L (2003) Deep and brainstemcavernomas: a consecutive 8-year series. J Neurosurg 99(1):31–37

32. Moriarity JL, Wetzel M, Clatterbuck RE, Javedan S, Sheppard JM,Hoenig-Rigamonti K, Crone NE, Breiter SN, Lee RR, RigamontiD (1999) The natural history of cavernous malformations: a pro-spective study of 68 patients. Neurosurgery 44(6):1166–1171

33. Morris Z, Whiteley WN, Longstreth WT Jr, Weber F, Lee YC,Tsushima Y, Alphs H, Ladd SC, Warlow C, Wardlaw JM, Al-Shahi

648 Acta Neurochir (2013) 155:643–649

Salman R (2009) Incidental findings on brain magnetic resonanceimaging: systematic review and meta-analysis. Br Med J.doi:10.1136/bmj.b3016

34. Noto S, Fujii M, Akimura T, Imoto H, Nomura S, Kajiwara K,Kato S, Fujisawa H, Suzuki M (2005) Management of patientswith cavernous angiomas presenting epileptic seizures. SurgNeurol 64(6):495–498

35. Pinker K, Stavrou I, Knosp E, Trattnig S (2006) Are cerebralcavernomas truly nonenhancing lesions and thereby distinguish-able from arteriovenous malformations? MRI findings and histo-pathological correlation. Magn Reson Imaging 24(5):631–637

36. Porter PJ, Willinsky RA, Harper W, Wallace MC (1997) Cerebralcavernous malformations: natural history and prognosis after clin-ical deterioration with or without hemorrhage. J Neurosurg 87(2):190–197

37. Porter RW, Detwiler PW, Spetzler RF, Lawton MT, Baskin JJ,Derksen PT, Zabramski JM (1999) Cavernous malformations of thebrainstem: experience with 100 patients. J Neurosurg 90(1):50–58

38. Pozzati E, Zucchelli M, Marliani AF, Riccioli LA (2006) Bleedingof a familial cerebral cavernous malformation after prophylacticanticoagulation therapy. Case report. Neurosurg Focus 21(1):e15

39. Rigamonti D, Drayer BP, Johnson PC, Hadley MN, Zabramski J,Spetzler RF (1987) The MRI appearance of cavernous malforma-tions (angiomas). J Neurosurg 67(4):518–524

40. Robinson JR, Awad IA, Little JR (1991) Natural history of thecavernous angioma. J Neurosurg 75(5):709–714

41. Rougier A, Castel JP, Cohadon F (1989) Different modalities for thetreatment of hemispheric cavernomas. [French]. Neurochirurgie 35(2):115–119

42. Schefer S, Valavanis A, Wichmann W (1991) MRT morphologyand classification of cerebral cavernomas. [German]. Radiologe 31(6):283–288

43. Secchi P, Calloni MV, Freschi R, Giorgetti A, Mariani G, MoriniR, Perrone P, Piazza I, Zancaner F (1998) Epilepsy and cerebralvascular malformations: retrospective study of 71 cases. Bollettino—Lega Italiana Contro L’Epilessia 102–103:111–113

44. Shih YH, Pan DH (2005) Management of supratentorial cavernousmalformations: craniotomy versus gamma knife radiosurgery. ClinNeurol Neurosurg 107(2):108–112

45. Steiner L, Karlsson B, Yen CP, Torner JC, Lindquist C,Schlesinger D (2010) Radiosurgery in cavernous malformations:anatomy of a controversy. J Neurosurg 113(1):16–21

46. Tarnaris A, Fernandes RP, Kitchen ND (2008) Does conservativemanagement for brain stem cavernomas have better long-termoutcome? Br J Neurosurg 22(6):748–757

47. The Stroke Prevention in Atrial Fibrillation Investigators (1996)Bleeding during antithrombotic therapy in patients with atrialfibrillation. Arch Intern Med 156(4):409–416

48. Tomlinson FH, Houser OW, Scheithauer BW, Sundt TM Jr,Okazaki H, Parisi JE (1994) Angiographically occult vascularmalformations: a correlative study of features on magnetic reso-nance imaging and histological examination. Neurosurgery 34(5):792–799

49. Vespignani H, Bouilleret V, Taillandier L, Schaff JL, Auque J,Huot JL, Bracard S, Weber M (1994) Cavernous angiomasrevealed by epileptic seizures: 57 cases reports. Epilepsies 6(2–3):97–111

50. von der Brelie C, Schramm J (2011) Cerebral cavernous malfor-mations and intractable epilepsy: the limited usefulness of currentliterature. Acta Neurochir (Wien) 153(2):249–259

51. Wang C-C, Liu A, Zhang J-T, Sun B, Zhao Y-L (2003) Surgicalmanagement of brain-stem cavernous malformations: report of 137cases. Surg Neurol 59(6):444–454

52. Yeon JY, Kim JS, Choi SJ, Seo DW, Hong SB, Hong SC (2009)Supratentorial cavernous angiomas presenting with seizures: sur-gical outcomes in 60 consecutive patients. Seizure 18(1):14–20

53. Yoon PH, Kim DI, Jeon P, Ryu YH, Hwang GJ, Park SJ (1998)Cerebral cavernous malformations: serial magnetic resonance im-aging findings in patients with and without gamma knife surgery.Neurol Med Chir (Tokyo) 38(Suppl):25–261

54. Yun TJ, Na DG, Kwon BJ, Rho HG, Park SH, Suh YL, Chang KH(2008) A T1 hyperintense perilesional signal aids in the differen-tiation of a cavernous angioma from other hemorrhagic masses.Am J Neuroradiol 29(3):494–500

55. Zabramski JM, Wascher TM, Spetzler RF, Johnson B, Golfinos J,Drayer BP, Brown B, Rigamonti D, Brown G (1994) The naturalhistory of familial cavernous malformations: results of an ongoingstudy. J Neurosurg 80(3):422–432

Acta Neurochir (2013) 155:643–649 649