1 Poorthuis

Stereotactic radiosurgery for cerebral cavernous malformations: a systematic review

and meta-analysis

Authors

Michiel H.F. Poorthuis MD,1 Leon A. Rinkel BSc,2 Simon Lammy MRCS (Ed),3 Rustam Al-

Shahi Salman MA PhD FRCP(Edin).4

Affiliations

1. Department of Neurology and Neurosurgery, Brain Center Rudolf Magnus, University

Medical Center Utrecht, Utrecht, The Netherlands.

2. University of Groningen, University Medical Centre Groningen, Groningen, The

Netherlands.

3. Department of Neurosurgery, Institute of Neurological Sciences, Queen Elizabeth

University Hospital, Glasgow, UK.

4. Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh, UK.

Corresponding Author

Professor Rustam Al-Shahi Salman, Centre for Clinical Brain Sciences, University of

Edinburgh, Chancellor's Building, 49 Little France Crescent, Edinburgh, EH16 4SB, UK.

Rustam.al-shahi@ed.ac.uk.

Email addresses co-authors:

MP: m.h.f.poorthuis@gmail.com, LR: l.a.rinkel@student.rug.nl, SL: s.lammy@nhs.net

1

1

2

3

4

5

6

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

1

2 Poorthuis

Keywords

Cerebral cavernous malformations, epilepsy, intracranial hemorrhage, stereotactic

radiosurgery, Gamma Knife.

Word count

Abstract – 259 words

Manuscript – 2345 words

Number of references: 49

Number of tables: 5

Figures 1

Funding

This study was not industry-sponsored.

MP: Reports no disclosures

LR: Reports no disclosures

SL: Reports no disclosures

RA-SS: Reports no disclosures

Statistical analysis:

All statistical analysis were performed by MP, University Medical Centre Utrecht, and LR,

University Medical Centre Groningen.

Supplementary data:

Appendix 1: Checklists for reporting standards

Appendix 2: Search strategies

2

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

2

3 Poorthuis

Appendix 3: Overview of overlapping studies

Appendix 4: Excluded studies, with the reasons for their exclusion

Appendix 5: Characteristics of each included study

Appendix 6: Summary of methodological characteristics of the included studies

Appendix 7: Risk of bias summary

3

48

49

50

51

52

53

3

4 Poorthuis

ABSTRACT

Background and objective The efficacy of stereotactic radiosurgery (SRS) for the treatment

of cerebral cavernous malformations (CCMs) is uncertain, so we set out to quantify clinical

outcomes after SRS for CCM and compare them to microsurgical excision or conservative

management.

Methods We searched PubMed and Ovid EMBASE from inception until June 1, 2018 for

peer-reviewed publications describing clinical outcomes after SRS for ≥10 people with CCM

in cohorts with or without a comparison group treated with neurosurgical excision or

conservative management. Two reviewers independently extracted data from the included

studies to quantify cohort characteristics and the incidence of the primary outcome (death

attributable to CCM and/or its treatment) and secondary outcomes (non-fatal symptomatic

intracerebral hemorrhage [ICH] and non-hemorrhagic persistent focal neurological deficit

[FND]). We assessed whether comparative studies showed a “dramatic” effect (meaning the

conventionally calculated probability of two differently managed patient groups from the

same population was <0.01 with a rate ratio greater than 10).

Results We included 30 cohort studies involving a total of 1,576 patients undergoing SRS for

CCM. Four non-randomized studies compared SRS to other treatment strategies, but did not

demonstrate dramatic associations. During a median follow-up of 48 (IQR 35-62) months

after SRS, the annual incidences (95%CI) of outcomes were: death 0.18% (0.10-0.31), ICH

2.40% (2.05-2.80), FND 0.71% (0.53-0.96), and the composite of death, ICH or FND 3.63%

(3.17-4.16). Outcomes did not differ by CCM location or type of SRS.

Conclusions After SRS for CCM, the annual incidences of death, ICH and FND are low and

seem comparable to outcomes without SRS. A randomized trial of SRS for CCM is needed.

4

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

4

5 Poorthuis

Registration This systematic review was registered (PROSPERO CRD42016025463 ) .

5

78

5

6 Poorthuis

INTRODUCTION

Cerebral cavernous malformations (CCMs) are intracranial vascular malformations that are

found in 0.15% to 0.44% of the population.1, 2 CCM may present with intracerebral

hemorrhage (ICH), non-hemorrhagic focal neurological deficits (FND),3 or epileptic seizures.

All of these complications may recur and might be prevented by CCM treatment.

In clinical practice, patients with CCM are managed conservatively or treated with

neurosurgical excision or stereotactic radiosurgery (SRS), but the role of the latter remains

controversial. However, randomized trials have not been performed and observational studies

at low risk of bias have not shown dramatic associations between treatment and better

outcome.4 Therefore, the most effective treatment modality for CCM is unknown and this

therapeutic dilemma has been identified as the top priority for further research by a priority

setting partnership.5

A systematic review assessed the outcomes of SRS for brainstem CCM, but it had

limitations.6 The review included five retrospective case series, but did not assess their risk of

bias.7 The authors performed a meta-analysis comparing the risk of ICH before and after SRS,

but such un-controlled before-and-after study designs are unreliable,8 and the reduction in

ICH risk after SRS may simply reflect the known reduction in the risk of CCM re-bleeding

over time.9 The review was restricted to studies of brainstem CCM published before

September 2013 and included only 178 patients, but many other studies report outcomes after

SRS.

6

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

6

7 Poorthuis

Consequently, a recent North American guideline deemed that there was only class II-III level

B-C evidence to support a recommendation that SRS might be considered for solitary CCM

with previous symptomatic ICH located in the CCM lies in eloquent areas that carry an

unacceptably high risk for neurosurgical excision (class IIb, level B), and a recommendation

against SRS for asymptomatic CCM, familial CCM, and CCM that are surgically accessible

(class III, level C).10

Therefore, we set out to perform an updated systematic review of all studies of SRS for CCM

in all locations and assess studies’ risk of bias in order to quantify all important clinical

outcomes after SRS for CCM with precision and compare these outcomes to microsurgical

excision or conservative management in studies with comparison groups.

7

103

104

105

106

107

108

109

110

111

112

113

114

7

8 Poorthuis

MATERIALS AND METHODS

Registration and reporting standards

We conducted this systematic review according to a protocol that we registered (PROSPERO

CRD42016025463 ) , and report the results consistent with the Preferred Reporting Items for

Systematic Reviews and Meta-analyses (PRISMA) (appendix 1).11, 12

Search strategy

We used comprehensive electronic strategies (appendix 2) to search PubMed Medline, Ovid

EMBASE and The Cochrane Library from inception to June 1, 2018 for articles meeting our

eligibility criteria to update our previous systematic review and meta-analysis of all

treatments for CCM.13 We performed backward citation searching using the bibliographies of

included studies and searched for randomized trials in ClinicalTrials.gov.

Eligibility criteria

We included randomized trials or observational cohort studies (with or without a comparison

group), published in any language in peer-reviewed journals, reporting ten or more patients of

any age diagnosed with CCM by MRI or histopathology, who were treated with SRS,

reporting clinical outcomes of interest (at least death, but also ICH and FND) during a

quantified period of follow-up of at least 30 days.

8

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

8

9 Poorthuis

Study selection

Three reviewers (MHFP, LR, and SL) independently used Covidence to screen all titles and

abstracts for eligibility, and assess the full text of potentially eligible studies for final

inclusion. In case of disagreement, MHFP, LR, SL and RA-SS resolved discrepancies in

consensus meetings. If multiple studies of the same cohort were published, we prioritized

inclusion of the study reporting a comparison group, but otherwise we included the study with

the largest sample size (appendix 3).

Data extraction

Two authors (MHFP and LR) independently extracted the characteristics of each cohort

according to the treatment received (SRS [Gamma Knife, linear accelerator, CyberKnife],

neurosurgical excision, or conservative management).

Study characteristics

We extracted aspects of study design such as institution, country, years in which treatment

took place, setting study, number of study centers, study design, design of data-collection,

design of follow-up, consecutiveness, and use of selection criteria.

Risk of bias

We used Cochrane criteria7 to assess the following domains: 1) random sequence generation;

2) allocation concealment; 3) blinding of participants and personnel; 4) blinding of outcome

assessment; 5) incomplete outcome data; 6) selective outcome reporting; and 7) other bias.

9

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

9

10 Poorthuis

Each potential source of bias was independently graded as high, low, or unclear risk of bias

by two reviewers (MHFP and LR).

Patient and CCM characteristics

We recorded the total number of patients, total number of CCMs, presenting symptoms, age,

number of children, number of female patients, number of patients with family history, and

number of patients with a genetic mutation. We recorded CCM location, number of patients

with multiple CCMs, size, and number of CCMs associated with developmental venous

anomaly. We recorded the number of patients that had undergone neurosurgical excision or

SRS at baseline.

CCM treatment

We recorded the type of stereotactic radiosurgery, maximum and margin dose, and number of

patients undergoing repeat radiosurgery.

Outcomes

The primary outcome was death attributed to CCM and/or its treatment. The secondary

outcomes were: non-fatal symptomatic ICH; non-fatal and non-hemorrhagic persistent FND;

a composite of death attributed to CCM and/or its treatment or non-fatal ICH or non-fatal and

non-hemorrhagic persistent FND; epileptic seizures; adverse radiation effects; and functional

outcome. We collected the numbers of primary and secondary outcomes (using authors’

descriptions of ICH, because most studies began before standards for reporting of CCM ICH

were published,3 but we contacted authors for clarification if it was unclear whether ICHs

were symptomatic or accompanied by radiographic evidence of new hemorrhage). We aimed

10

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

10

11 Poorthuis

to collect these outcomes in three timeframes: 1) 30 days after intervention; 2) between 30

days and 1 year after intervention; 3) more than 1 year after intervention. If data were

provided in such a way that it was not possible to separate the outcomes according to these

timeframes, we collected all outcomes during the entire duration of follow-up.

Statistical analyses

We quantified the distribution of cohort-level characteristics with descriptive analyses. For

each comparative observational study, we determined whether it found a “dramatic”

association, defined as the conventionally calculated probability of the two differently

managed patient groups from the same population being <0.01 with a rate ratio >10.14 We

quantified the occurrence of outcomes after SRS either during the total person-years of

follow-up described, or by multiplying the median (or mean if median was not provided)

follow-up duration by the total number of treated patients. We calculated annual incidence

rates and their 95% CI with Poisson distributions using the Poisson command in STATA 15.

We intended to perform the following subgroup analyses: 1) outcomes in cohorts that

included 80% or more patients with brainstem CCM, 2) outcomes in cohorts that described

Gamma Knife SRS, and 3) by mode of CCM presentation. Our analyses are based on the

published aggregate results of individual studies, so individual-level data cannot be made

publicly available. All extracted data from the individual studies and the code used for

performing the meta-analyses can be made available upon reasonable request to the

corresponding author.

11

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

11

12 Poorthuis

Ethical approval

Approval from an institutional or regional review board was not required as all data used in

this study was extracted from publications.

Data availability

The analysis for this study is based on published results from individual studies. Therefore,

individual-level data cannot be made publicly available. All extracted data from the individual

studies and the code used for performing the meta-analyses can be made available upon

reasonable request to the corresponding author.

12

209

210

211

212

213

214

215

216

12

13 Poorthuis

RESULTS

After screening 361 references and excluding ineligible studies and smaller overlapping

reports of included studies (appendices 3 and 4), we included 30 studies of SRS treatment for

CCMs (figure 1 and appendix 4):15-44 three studies compared SRS to neurosurgery, one study

compared SRS to neurosurgery and conservative management,17, 19, 21, 35 and one study

compared two cohorts receiving different types of SRS,24 leaving 25 cohort studies of SRS

alone.15, 16, 18, 20, 22, 23, 25-34, 36-44

Study characteristics

In the 31 cohorts receiving SRS in 30 studies, including a total of 1,576 patients, the median

cohort-level attributes were: sample size 34 patients, duration of follow-up 48 months, age at

treatment 40 years, 48% female, 91% presented with ICH due to CCM, 65% CCM were

infratentorial and 35% were supratentorial (Table 1 and appendix 5). Twenty-nine (97%)

studies were from single centres15-21, 23-44 and one (3%) was multicenter (appendix 6).22 Fifteen

(50%) studies were from Asia,16, 19, 21-26, 28, 32, 34, 35, 38, 41, 43 nine (30%) from Europe,15, 17, 18, 20, 27, 29, 31,

39, 42 four (13%) from North America,30, 33, 36, 37 and two (7%) from South America.40, 44

In general, included studies were at medium to high risk of bias (appendices 6 and 7). None of

the studies was randomized, none concealed treatment allocation, and none used blinding of

allocation or outcome assessment, so they were at high risk of performance bias (appendix 7).

In the five studies (17%) with a comparison group, three studies were at high risk of selection

bias because there were significant differences between groups, but it was unclear if there

13

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

13

14 Poorthuis

were differences in the other two. Three (10%) cohorts identified patients prospectively,15, 30, 33

eight (27%) studies identified consecutive patients,15, 18-20, 26, 39, 41, 44 and four (13%) followed

patients prospectively. Twenty-seven (90%) studies appeared to be at low risk of attrition bias

and two (7%) confirmed full reporting of pre-specified outcomes. All studies were at low risk

of performance and detection bias for the primary outcome, since the outcome death is not

influenced by lack of blinding of participants and personnel.

Outcome after SRS compared other treatment strategies

One non-randomized study compared radiosurgery to neurosurgery and conservative

management21, three compared radiosurgery to neurosurgery,17, 19, 35 and one compared

Gamma Knife and linear accelerator SRS.24 These studies showed no dramatic associations

with primary and secondary outcomes and none was at low risk of bias (table 2).

Outcome after SRS

In 31 separate cohorts reporting the outcome after SRS in 30 studies for a total of 1,576

patients, 1,386 (88%) were treated with Gamma Knife SRS, 159 (10%) were treated with

linear accelerator SRS, and 31 (2%) were treated with CyberKnife SRS. The median margin

dose was 15 Gy and the median maximum dose was 25 Gy. We were unable to quantify

outcomes in our prespecified timeframes, so we calculated the incidence of outcomes per 100

person-years of follow-up over the durations of follow-up described in included studies. We

were unable to reliably quantify the frequency of adverse radiation effects, epileptic seizures,

and functional outcome in these studies because of variation between studies in the methods

14

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

14

15 Poorthuis

and completeness of reporting these outcomes. We were unable to quantify outcomes

according to mode of CCM presentation, because studies did not stratify their reporting of

outcomes by mode of presentation (which was mostly ICH, table 1).

The annual incidence of the primary outcome of death attributable to CCM or SRS was 0.18%

(95% CI 0.10-0.31) (table 3). The annual incidence of the secondary outcomes were:

symptomatic ICH 2.40% (95% CI 2.05-2.80), non-hemorrhagic persistent FND 0.71% (95%

CI 0.53-0.96), and in the 26 cohorts reporting on all components of the composite outcome,

the annual incidence was 3.63% (95% CI 3.17-4.16).

In nine studies that included 80% or more patients with brainstem CCM, 303 patients (298

[98%] of whom had a brainstem CCM) were treated with a median margin dose of 13 Gy and

median maximum dose 26 Gy. Their annual risks of death, ICH, and FND were not different

from the primary and secondary outcomes of the main analysis (table 4).

In twenty studies, 1,386 patients were treated with Gamma Knife SRS with a median margin

dose of 14 Gy and median maximum dose 30 Gy. Their annual risks of death, ICH, and FND

were not different from the primary and secondary outcomes of the main analysis (table 5).

15

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

15

16 Poorthuis

DISCUSSION

In this systematic review of SRS for CCM, we did not find any randomized trials, but

included 30 observational studies involving a total of 1,576 patients (median sample size 34,

age 40 years, 91% presented with ICH, 65% infratentorial CCM, and follow-up 48 months).

These studies were at medium to high risk of bias because of the lack of randomization,

allocation concealment, and blinding, although the risk of attrition bias was generally low.

Five (17%) non-randomized studies compared SRS to another treatment strategy, but none of

these studies was both at low risk of bias as well as finding dramatic associations with

outcome. Nonetheless, our pooled estimates of the outcomes after SRS reflect clinical practice

around the world, in which the annual incidences were: death attributable to CCM or SRS

0.18% (95% CI 0.10-0.31), symptomatic ICH 2.40% (95% CI 2.05-2.80), non-hemorrhagic

persistent FND 0.71% (95% CI 0.53-0.96), and the composite of all these outcomes was

3.63% (95% CI 3.17-4.16). We found no differences in studies focused on brainstem CCM or

using Gamma Knife SRS. We were unable to consistently extract and summarize data on

adverse radiation effects, epileptic seizures, or functional outcome.

The strengths of this review were its prespecified protocol, comprehensive literature searches,

absence of language restrictions, standardized assessment of risk of bias, the precision of the

pooled estimates of outcomes after SRS compared to a previous systematic review,6 and

exploration of two sub-groups. However, this review has some limitations. The included

studies were non-randomized, mostly lacked comparison groups, and were at moderate to

high risk of bias. We obtained aggregate data, and not individual patient data, so we could

only explore whether presentation with ICH and type of SRS were associated with differences

16

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

16

17 Poorthuis

in outcome, but we could not determine time trends in outcome over time. One new cohort of

45 patients and an update of an included cohort were published after the date of our literature

search and were not included,45-47 but are unlikely to have changed our conclusions.

Our pooled estimates of the risks of SRS for CCM over approximately four years after

treatment can be applied to clinical practice, especially for patients with CCM that have

caused ICH. The risks after SRS appear similar to the untreated clinical course of CCM, by

indirect comparison with the overall 5-year risk of ICH found in a patient-level meta-analysis

(15.8%, 95% CI 13.7-17.9).9 However, it is not possible to be confident in the beneficial

effects of SRS for CCM because of the shortage of comparative studies at low risk of bias,

leaving reassurance only from indirect comparisons with untreated clinical course or

neurosurgery.13, 48

Our findings have several implications for future research. Comparative, ideally randomized,

studies at low risk of bias with blinded assessment and standardized definitions of outcome3

are needed to be certain about the overall risks and benefits of SRS for CCM. The outcomes

observed in the cohort studies in this systematic review can inform sample size calculations in

future randomized trials (https://www.nihr.ac.uk/funding-and-support/funding-opportunities/

18186-active-management-vs-conservative-management-for-symptomatic-brain-cavernoma/

9700).

17

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

17

18 Poorthuis

ACKNOWLEDGEMENTS

We thank Bruce Pollock, Roman Liščák, Yang-Hsin Shih, Remedios López-Serrano, Sharon

Jay, and Yoshihisa Kida for providing further information about some of the cohorts included

in this systematic review. We are very grateful to Gabor Nagy, Jozsef Dobai, László Bognar,

Dániel Bereczki, Michael Poon, Murat Akyol, Ebru Mumcu, Shoichiro Sato, Kazutaka

Sonoda, Joan Marti Fabregas and Elena Lebedeva for helping with the translation and data

extraction from publications in Hungarian, Chinese, Turkish, Japanese, Spanish and Russian.

18

329

330

331

332

333

334

335

336

18

19 Poorthuis

Appendix. Author contributions

Name Location Role Contribution

Michiel Poorthuis

University Medical Centre Utrecht, The Netherlands

First author

Design and conceptualized study; Screened titles and abstracts of initial search and its update. Assessed full-text articles and reference lists of included studies; Extracted data from included studies; analyzed the data; drafted the manuscript for intellectual content

Leon Rinkel

University Medical Centre, Groningen, The Netherlands

Author

Screened titles and abstract of search update. Assessed full-text articles and reference lists of included studies; Extracted data from included studies; analyzed the data; drafted the manuscript for intellectual content

Simon Lammy

Queen Elizabeth University Hospital, Glasgow, UK

Author

Design and conceptualized study; Registered this study on PROSPERO. Performed initial literature search and removed duplicates; screened title and abstracts of initial search and assessed full-text articles; revised the manuscript for intellectual content

Rustam Al-Shahi Salman

Centre for Clinical Brain Sciences, University of Edinburgh, UK

Corresponding author

Design and conceptualized study; Interpreted the data; revised the manuscript for intellectual content; study supervision

19

337

338

19

20 Poorthuis

Table 1. Summary of the characteristics of all studies of SRS

Characteristics Median (inter-quartile range)

Cohorts reporting

characteristic

Patients reported

Number of patients treated 34 (16-82) 31 1576Duration of follow-up, months 48 (35-62) 30 1563Mid-year of treatment 2001 (1997-2014) 30 1565Age, years 40 (37.6-41.7) 28 1521Female sex, % 48 (45-57) 27 1504Multiple CCMs, % 7 (0-15) 23 1075Children, % 0 (0-0) 15 337Size, mm 15.5 (14.9-18) 4 342Size, cm3 1.37 (0.6-1.86) 20 807Previously treated with SRS 0 (0-2.4) 20 797Previously treated with surgery 0 (0-10.3) 19 705Re-treated after SRS with SRS 0 (0-1.0) 17 679Treated after SRS with surgery 0 (0-4.4) 16 576CCM associated with DVA, % 7 (0-23) 8 283CCM location Total supratentorial, % 35 (4-73) 28 1430 Lobar, % 18 (0-60) 28 1430 Basal ganglia and thalamus, % 13 (0.1-23) 28 1430 Total infratentorial, % 65 (27-96) 28 1430 Brainstem, % 61 (18-96) 28 1430 Cerebellum, % 0 (0-7.1) 28 1430Clinical presentation without symptoms, %

0 (0-0) 27 1409

Clinical presentation with ICH, % 91 (53-100) 26 1321Clinical presentation with seizures, %

4.8 (0-22) 26 1301

Marginal SRS dose (Gy) 15 (13-16) 30 1463Maximum SRS dose (Gy) 25 (22-30) 20 1218

CCM = cerebral cavernous malformation, SRS = stereotactic radiosurgery, DVA = developmental venous anomaly, ICH = intracerebral hemorrhage, Gy = Gray

20

339

340341342

20

21 Poorthuis

Table 2. Overview of outcomes in comparative cohorts

First author, year of

publication

Outcome Comparison of

treatment modalities

Number of

patients treated

(index / reference

group)

Number of patients

with outcome

(index / reference

group)

Statistical analyses comparing the

outcomes after treatment

1. Frischer et al, 201417 Death GKS vs. NS 38 / 29 0 / 0 Not performed

2. Hsu et al, 200719 Death GKS vs. NS 14 / 15 0 / 0 Not performed

ICH GKS vs. NS 14 / 15 0 / 0 Not performed

FND GKS vs. NS 14 / 15 0 / 0 Not performed

3. Kayali et al, 200421 Death GKS vs. NS vs. CM 13 / 3 / 21 0 / 1 / 0 Not performed

ICH GKS vs. NS vs. CM 13 / 3 / 21 0 / 0 / 0 Not performed

FND GKS vs. NS vs. CM 13 / 3 / 21 0 / 0 / 0 Not performed

4. Kim et al, 200224 Death GKS vs. LINAC 11 / 11 0 / 0 Not performed

ICH GKS vs. LINAC 11 / 11 1 / 0 Not performed

FND GKS vs. LINAC 11 / 11 0 / 2 No significant difference in

occurrence of FND

5. Shih and Pan, 200535 Death GKS vs. NS 30 / 16 0 / 0 Not performed

21

343

344

21

22 Poorthuis

ICH GKS vs. NS 30 / 16 1 / 0 Not performed

FND GKS vs. NS 30 / 16 NA / 0 Not performed

Abbreviations: ICH = symptomatic intracerebral hemorrhage. FND = Non-hemorrhagic persistent focal neurological deficit. GKS = Gamma

Knife Surgery. LINAC = Linear accelerator.

NS = Neurosurgery. CM = Conservative Management. NA= not available.

22

345

346

347

348

22

23 Poorthuis

Table 3. Incidence of outcomes after SRS

Cohorts (%)

Patients Total number of outcome events / total person-years

Median per cohort (inter-quartile range)

Outcome event annual incidence (95% CI) per 100 person-years

Follow-up, PY 30* 1563 - 124 (53-344) -Primary outcomeDeaths attributable to CCM or its treatment

30 1563 13 / 7305 0 (0-0) 0.18 (0.10-0.31)

Secondary outcomesICH 28 1403 160 / 6670 3 (1-4) 2.40 (2.05-2.80)FND 27 1383 44 / 6191 2 (0-3) 0.71 (0.53-0.96)Composite outcome of death, ICH or FND

26 1261 209 / 5754 5 (2-6) 3.63 (3.17-4.16)

CCM = cerebral cavernous malformation, CI = confidence interval, FND = non-fatal non-hemorrhagic persistent focal neurological deficit, ICH = non-fatal symptomatic intracranial hemorrhage with persistent focal neurological deficit.* One study describing SRS, neurosurgical intervention, and conservative management provided the total number of patient-years for all patients combined. Total number of patient-years could not be estimated for the different management options separately.

23

349

350351352

353354355

23

24 Poorthuis

Table 4. Incidence of outcomes after SRS in cohorts with ≥80% brainstem CCMs

Cohorts (%)

Patients Total number of outcome events / total person-years

Median per cohort (inter-quartile range)

Outcome event annual incidence (95% CI) per 100 person-years

Follow-up, PY 9 303 - 123 (70-146) -Primary outcomeDeaths attributable to CCM or its treatment

9 303 2 / 1163 0 (0-0) 0.17 (0.04-0.69)

Secondary outcomesICH 9 303 30 / 1163 3 (2-4) 2.58 (1.80-3.69)FND 9 303 8 / 1163 0 (0-1) 0.69 (0.34-1.38)Composite outcome of death, ICH or FND

9 303 40 / 1163 5 (2-6) 3.44 (2.52-4.69)

CCM = cerebral cavernous malformation, CI = confidence interval, FND = non-fatal non-

hemorrhagic persistent focal neurological deficit, ICH = non-fatal symptomatic intracranial

hemorrhage with persistent focal neurological deficit.

24

356

357

358

359

24

25 Poorthuis

Table 5. Incidence of outcomes after Gamma Knife SRS

Cohorts (%)

Patients Total number of outcome events / total person-years

Median per cohort (inter-quartile range)

Outcome event annual incidence (95% CI) per 100 person-years

Follow-up, PY 20 1386 - 156 (88-421) -Primary outcomeDeaths attributable to CCM or its treatment

20 1386 11 / 6504 0 (0-1) 0.17 (0.09-0.31)

Secondary outcomesICH 18 1226 139 / 5870 3 (1-6) 2.37 (2.01 -2.80)FND 17 1206 31 / 5390 1 (0-3) 0.58 (0.40-0.82)Composite outcome of death, ICH or FND

16 1084 173 / 4953 5 (2-9) 3.49 (3.01-4.05)

CCM = cerebral cavernous malformation, CI = confidence interval, FND = non-fatal non-

hemorrhagic persistent focal neurological deficit, ICH = non-fatal symptomatic intracranial

hemorrhage with persistent focal neurological deficit.

25

360

361

362

363

364

25

26 Poorthuis

REFERENCES

1. Flemming KD, Graff-Radford J, Aakre J, et al. Population-Based Prevalence of Cerebral Cavernous Malformations in Older Adults: Mayo Clinic Study of Aging. JAMA Neurol 2017;74:801-805.2. Morris Z, Whiteley WN, Longstreth WT, Jr., et al. Incidental findings on brain magnetic resonance imaging: systematic review and meta-analysis. BMJ 2009;339:b3016.3. Al-Shahi Salman R, Berg MJ, Morrison L, Awad IA, Angioma Alliance Scientific Advisory B. Hemorrhage from cavernous malformations of the brain: definition and reporting standards. Angioma Alliance Scientific Advisory Board. Stroke 2008;39:3222-3230.4. Poorthuis M, Samarasekera N, Kontoh K, et al. Comparative studies of the diagnosis and treatment of cerebral cavernous malformations in adults: systematic review. Acta Neurochir (Wien) 2013;155:643-649.5. Al-Shahi Salman R, Kitchen N, Thomson J, et al. Top ten research priorities for brain and spine cavernous malformations. Lancet Neurol 2016;15:354-355.6. Lu XY, Sun H, Xu JG, Li QY. Stereotactic radiosurgery of brainstem cavernous malformations: a systematic review and meta-analysis. J Neurosurg 2014;120:982-987.7. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. : The Cochrane Collaboration., 2011.8. Torgerson DJ, Torgerson CJ. The Limitations of Before and After Designs. Designing Randomised Trials in Health, Education and the Social Sciences: An Introduction. London: Palgrave Macmillan UK, 2008: 9-16.9. Horne MA, Flemming KD, Su IC, et al. Clinical course of untreated cerebral cavernous malformations: a meta-analysis of individual patient data. Lancet Neurol 2016;15:166-173.10. Akers A, Al-Shahi Salman R, Awad IA, et al. Synopsis of Guidelines for the Clinical Management of Cerebral Cavernous Malformations: Consensus Recommendations Based on Systematic Literature Review by the Angioma Alliance Scientific Advisory Board Clinical Experts Panel. Neurosurgery 2017;80:665-680.11. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med 2009;6:e1000100.12. Moher D, Liberati A, Tetzlaff J, Altman DG, Prisma Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009;6:e1000097.13. Poorthuis MH, Klijn CJ, Algra A, Rinkel GJ, Al-Shahi Salman R. Treatment of cerebral cavernous malformations: a systematic review and meta-regression analysis. J Neurol Neurosurg Psychiatry 2014;85:1319-1323.14. Glasziou P, Chalmers I, Rawlins M, McCulloch P. When are randomised trials unnecessary? Picking signal from noise. BMJ 2007;334:349-351.15. Aboukais R, Estrade L, Devos P, Blond S, Lejeune JP, Reyns N. Gamma Knife Radiosurgery of Brainstem Cavernous Malformations. Stereotact Funct Neurosurg 2016;94:397-403.16. Chung TT, Lin BJ, Tang CT, Hueng DY, Ju DT. Cyberknife stereotactic radiosurgery for intracranial cavernous malformations: Preliminary results from one medical center. Journal of Medical Sciences 2012;32:017-025.17. Frischer JM, Gatterbauer B, Holzer S, et al. Microsurgery and radiosurgery for brainstem cavernomas: effective and complementary treatment options. World Neurosurg 2014;81:520-528.18. Fuetsch M, El Majdoub F, Hoevels M, Muller RP, Sturm V, Maarouf M. Stereotactic LINAC radiosurgery for the treatment of brainstem cavernomas. Strahlenther Onkol 2012;188:311-316.19. Hsu PW, Chang CN, Tseng CK, et al. Treatment of epileptogenic cavernomas: surgery versus radiosurgery. Cerebrovasc Dis 2007;24:116-120; discussion 121.

26

365

366

367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413

26

27 Poorthuis

20. Jay SM, Chandran H, Blackburn TP. Gamma knife stereotactic radiosurgery for thalamic & brainstem cavernous angiomas. Br J Neurosurg 2012;26:367-370.21. Kayali H, Sait S, Serdar K, Kaan O, Ilker S, Erdener T. Intracranial cavernomas: analysis of 37 cases and literature review. Neurol India 2004;52:439-442.22. Kida Y, Hasegawa T, Iwai Y, et al. Radiosurgery for symptomatic cavernous malformations: A multi-institutional retrospective study in Japan. Surg Neurol Int 2015;6:S249-257.23. Kim BS, Yeon JY, Kim JS, Hong SC, Lee JI. Gamma knife radiosurgery of the symptomatic brain stem cavernous angioma with low marginal dose. Clin Neurol Neurosurg 2014;126:110-114.24. Kim DG, Choe WJ, Paek SH, Chung HT, Kim IH, Han DH. Radiosurgery of intracranial cavernous malformations. Acta Neurochir (Wien) 2002;144:869-878; discussion 878.25. Kim MS, Pyo SY, Jeong YG, Lee SI, Jung YT, Sim JH. Gamma knife surgery for intracranial cavernous hemangioma. J Neurosurg 2005;102 Suppl:102-106.26. Lee CC, Pan DH, Chung WY, et al. Brainstem cavernous malformations: the role of Gamma Knife surgery. J Neurosurg 2012;117 Suppl:164-169.27. Liscak R, Urgosik D, Simonova G, Vymazal J, Semnicka J. Gamma knife radiosurgery of brain cavernomas. Acta Neurochir Suppl 2013;116:107-111.28. Liu HB, Wang Y, Yang S, Gong FL, Xu YY, Wang W. Gamma knife radiosurgery for brainstem cavernous malformations. Clin Neurol Neurosurg 2016;151:55-60.29. López-Serrano R, Martinez NE, Kusak ME, Quiros A, Martinez R. Significant Hemorrhage Rate Reduction after Gamma Knife Radiosurgery in Symptomatic Cavernous Malformations: Long-Term Outcome in 95 Case Series and Literature Review. Stereotact Funct Neurosurg 2017;95:369-378.30. Lunsford LD, Khan AA, Niranjan A, Kano H, Flickinger JC, Kondziolka D. Stereotactic radiosurgery for symptomatic solitary cerebral cavernous malformations considered high risk for resection. J Neurosurg 2010;113:23-29.31. Nagy G, Razak A, Rowe JG, et al. Stereotactic radiosurgery for deep-seated cavernous malformations: a move toward more active, early intervention. Clinical article. J Neurosurg 2010;113:691-699.32. Park SH, Hwang SK. Gamma knife radiosurgery for symptomatic brainstem intra-axial cavernous malformations. World Neurosurg 2013;80:e261-266.33. Pollock BE, Garces YI, Stafford SL, Foote RL, Schomberg PJ, Link MJ. Stereotactic radiosurgery for cavernous malformations. J Neurosurg 2000;93:987-991.34. Sager O, Beyzadeoglu M, Dincoglan F, et al. Evaluation of linear accelerator (LINAC)-based stereotactic radiosurgery (SRS) for cerebral cavernous malformations: a 15-year single-center experience. Ann Saudi Med 2014;34:54-58.35. Shih YH, Pan DH. Management of supratentorial cavernous malformations: craniotomy versus gammaknife radiosurgery. Clin Neurol Neurosurg 2005;107:108-112.36. Stea RA, Schicker L, King GA, Winfield JA. Stereotactic linear radiosurgery for cavernous angiomas. Stereotact Funct Neurosurg 1994;63:255-265.37. Tsien C, Souhami L, Sadikot A, et al. Stereotactic radiosurgery in the management of angiographically occult vascular malformations. Int J Radiat Oncol Biol Phys 2001;50:133-138.38. Üçüncü Kefeli A, Şengöz M, Peker S. Gamma Knife Radiosurgery for Hemorrhagic Brainstem Cavernomas. Turk Neurosurg 2017;29:14-19.39. Fedorcsak I, Nagy G, Dobai JG, Mezey G, Bognar L. [Radiosurgery of Intracerebral Cavernomas--Current Hungarian Practice]. Ideggyogy Sz 2015;68:243-251.40. Garcia-Munoz L, Velasco-Campos F, Lujan-Castilla P, Enriquez-Barrera M, Cervantes-Martinez A, Carrillo-Ruiz J. [Radiosurgery in the treatment of brain cavernomas. Experience with 17 lesions treated in 15 patients]. Neurochirurgie 2007;53:243-250.41. Jia G, Zhang J, Ma Z, Qiu B, Hou Y. [Therapeutic effect of gamma knife on intracranial cavernous angioma]. Zhong Nan Da Xue Xue Bao Yi Xue Ban 2014;39:1320-1325.42. Karabulut Gül Ş, Mayadaǧli A, Özşeker N, et al. Stereotactic radiotherapy in the treatment of cavernoma. Turkiye Klinikleri Journal of Medical Sciences 2014;34:29-32.

27

414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465

27

28 Poorthuis

43. Liu AL, Wang CC, Dai K. [Gamma knife radiosurgery for cavernous malformations]. Zhongguo Yi Xue Ke Xue Yuan Xue Bao 2005;27:18-21.44. Moreno-Jiménez S, Reyes-Moreno I, Aceves GAG, et al. Cavernous malformations treated with linac-based radiosurgery at the INNN: Preliminary results regarding volumetric response. Archivos de Neurociencias 2008;13:8-12.45. Nagy G, Burkitt W, Stokes SS, et al. Contemporary radiosurgery of cerebral cavernous malformations: Part 1. Treatment outcome for critically located hemorrhagic lesions. J Neurosurg 2018:1-9.46. Nagy G, Stokes SS, Eross LG, et al. Contemporary radiosurgery of cerebral cavernous malformations: Part 2. Treatment outcome for hemispheric lesions. J Neurosurg 2018:1-9.47. Park K, Kim JW, Chung HT, Paek SH, Kim DG. Long-Term Outcome of Gamma Knife Radiosurgery for Symptomatic Brainstem Cavernous Malformation. World Neurosurg 2018;116:e1054-e1059.48. Moultrie F, Horne MA, Josephson CB, et al. Outcome after surgical or conservative management of cerebral cavernous malformations. Neurology 2014;83:582-589.

28

466467468469470471472473474475476477478479480481

28