Cerebral arteriovenous malformations (AVMs) occur with a prevalence of 18 per 100,000 adults,1 and they have an overall annual hemorrhage risk of approximately 2%–3%. Stereotactic radiosurgery (SRS) is a favorable alternative to resection for the management of cerebral AVM, particularly when the perceived risk of resection outweighs the potential benefits owing to AVM location and angioarchitecture and patient comorbidities and for patients unwilling to undergo surgery. SRS can be used as an upfront treatment or as a salvage therapy after incomplete resection or embolization. However, AVM remains in 20%–50% of patients 3–5 years after SRS.2 For these patients, salvage treatment options include resection, endovascular embolization, and repeat SRS, administered alone or in combination. In the event of failed AVM obliteration after initial SRS, complete AVM obliteration has been reported in 55%–70% of patients who underwent repeat SRS2–6 and patients who have undergone repeat radiosurgery experience hemorrhage and radiation-induced changes (RICs) at rates of 7.6% and 7.4%, respectively.7
The purpose of this study was to determine the radiological and clinical outcomes of patients managed with repeat SRS for cerebral AVM after incomplete response to prior SRS. We also identified favorable prognostic factors for obliteration and factors associated with hemorrhage and adverse radiation effects after repeat SRS. Such information may help clinicians better define the role of repeat SRS for AVM patients.
Methods
Patient Population and Inclusion Criteria
This retrospective single-center study included patients managed with repeat SRS for residual AVM after initial single-session SRS. It represents an appreciable update of our previously published analysis, with the addition of data from new patients and previously analyzed patients who received extended follow-up.4 Patients managed with volume-staged or dose-staged SRS, and those without at least 1 clinical and radiological follow-up examination after the second SRS procedure were excluded.
The collected data included patient demographic characteristics, pertinent medical history, and AVM and treatment characteristics at the first and second SRS procedures. Clinical follow-up data after the second SRS procedure included the presence of AVM hemorrhage, seizure outcome, and AVM occlusion, as well as data related to overall survival, salvage therapy needed after repeat SRS, and radiation toxicity. Data about the occurrence and treatment of RICs, radionecrosis, radiation-induced neoplasia, and delayed cyst formation were also collected.
SRS Technique and Radiosurgical Parameters
Details of the Gamma Knife radiosurgery technique were previously described.4 In brief, the Leksell model U Gamma Knife was used until 2001, model C (Elekta Instruments, Inc.) was used from 2001 to 2006, and Gamma Knife Perfexion has been used since 2006. KULA planning software was used from 1989 to 1994, and Gamma Plan software from 1994 to the present.
The patients underwent placement of the Leksell stereotactic frame under local anesthesia and monitored sedation. Children younger than 18 years underwent the procedure under general anesthesia. Stereotactic, thin-slice (1-mm-thick), T1-weighted axial contrast-enhanced and T2-weighted brain MRI, as well as biplane cerebral angiography, was then performed and data were transferred to the Gamma Knife planning system. The chosen margin dose included the entire nidus, which was defined as the arteriovenous shunt. The median (interquartile range [IQR]) margin doses at the initial and repeat SRS procedures were both 20 (5) Gy. Repeat radiosurgery was typically performed 3 to 4 years after prior SRS and administered to patients with radiological presence of persistent nidus on brain MRI or head CT and digital subtraction angiography (DSA).
Clinical and Radiological Follow-Up
The patients underwent clinical and radiological evaluations at 6-month intervals for the first 2–3 years after radiosurgery and yearly thereafter. DSA was requested when nidus obliteration was on follow-up brain MRI. AVM obliteration was defined on the basis of failure to visualize the nidus and early venous drainage on cerebral DSA. In a series of 19,826 consecutive cerebral DSA procedures, the reported incidence rate of all neurological complications was 2.63% and the incidence of all strokes resulting in permanent disability was 0.14%.8 Therefore, patients may refuse to undergo cerebral DSA at follow-up.
Brain MRI is highly effective for identifying AVM obliteration.9 In patients who refused cerebral DSA, AVM occlusion was defined on the basis of absent contrast enhancement of the nidus on T1-weighted postcontrast MRI and lack of flow voids on T2-weighted brain MRI. If residual AVM was noted on follow-up brain MRI at 3–4 years after prior radiosurgery, the patient was reevaluated and additional treatments were considered.
Adverse Radiation Events
RICs were defined as perinidal hyperintensities on T2-weighted follow-up brain MRI.4 They were classified as asymptomatic, symptomatic if accompanied by global or focal neurological deficits in the absence of bleeding, and permanent if neurological deficits persisted after conservative management or if surgical intervention was needed.
Statistical Analysis
All statistical analyses were performed with R programming in RStudio.10,11 The rates of AVM obliteration, post-SRS hemorrhage, and RIC were calculated with the Kaplan-Meier method. The log-rank test was employed to assess differences in Kaplan-Meier curves. Univariate and multivariate analyses were computed with Cox proportional hazards modeling. The assessed factors that may have affected the rates of AVM obliteration, post-SRS hemorrhage, and RIC included age (≤ 18 vs > 18 years), sex, AVM diameter (≤ 20 vs > 20 mm) and location (deep [thalamus, brainstem, basal ganglia, and corpus callosum] vs superficial), SRS parameters (margin dose, prescription volume, and 12-Gy volume), history of hemorrhage or RIC, and Spetzler-Martin (SM) grade. In this study, p < 0.05 was defined as statistically significant.
Results
Patient and AVM Attributes
A total of 170 patients underwent repeat single-session SRS after initial single-session SRS and were included in the current study. There were 90 (52.9%) females, and the median (IQR) patient ages at the first and second SRS procedures were 28 (21.5) and 32 (22.5) years, respectively. The main presenting signs or symptoms at AVM diagnosis were intracranial hemorrhage (62.3%), headache (26.3%), motor deficits (26.3%), seizures (25.1%), and visual (15%) and speech (11.4%) disturbances. AVM was asymptomatic in 8.4% of patients (Table 1).
Patient and brain AVM characteristics
| Characteristic | Value | |
|---|---|---|
| Entire cohort | ||
| Sex | ||
| Male | 80 (47.1) | |
| Female | 90 (52.9) | |
| Median (IQR) time to 2nd SRS, mos | 44 (23) | |
| AVM location | ||
| Superficial | 103 (60.6) | |
| Deep | 67 (39.4) | |
| Surgical intervention prior to 2nd SRS | ||
| None | 106 (62.4) | |
| Partial resection | 16 (9.4) | |
| VPS/EVD placement | 10 (5.9) | |
| AVM embolization | 45 (26.5) | |
| Characteristic at SRS procedure | 1st SRS | 2nd SRS |
| Median (IQR) age, yrs | 28 (21.5) | 32 (22.5) |
| Symptoms at SRS | ||
| None | 5 (2.9) | 12 (7.1) |
| Hemorrhage | 101 (59.4) | 109 (64.1) |
| Seizures | 43 (25.3) | 42 (24.7) |
| Headache | 30 (17.6) | 46 (27.1) |
| CN palsy | 6 (3.5) | 5 (2.9) |
| Visual disturbances | 28 (16.5) | 27 (15.9) |
| Speech deficits | 12 (7.1) | 19 (11.2) |
| Motor deficits | 41 (24.1) | 44 (25.9) |
| Cerebellar | 5 (2.9) | 3 (1.8) |
| Tinnitus | 0 (0) | 1 (0.6) |
| Bruit | 1 (0.6) | 2 (1.2) |
| SM grade | ||
| I | 13 (7.6) | 20 (11.8) |
| II | 47 (27.6) | 68 (40) |
| III | 71 (41.8) | 76 (44.7) |
| IV | 38 (22.4) | 6 (3.5) |
| V | 1 (0.6) | 0 (0) |
| Median (IQR) AVM diameter, mm | 26.5 (12) | 14 (9.8) |
| Median (IQR) AVM vol, cm3 | 3.6 (3.53) | 1 (1.7) |
| Median (IQR) prescription dose, Gy | 20 (5) | 20 (5) |
| Median (IQR) isodose line, % | 50 (0) | 50 (9) |
CN = cranial nerve; EVD = external ventricular drain, VPS = ventriculoperitoneal shunt.
Values are shown as number (%) unless indicated otherwise.
At the initial SRS procedure, 13 AVMs were classified as SM grade I, 47 as grade II, 71 as grade III, 38 as grade IV, and 1 as grade V. The median (IQR) AVM volume was 3.6 (3.53) cm3, and the median (IQR) AVM diameter was 26.5 (12) mm. At the repeat SRS procedure, 20 AVMs were SM grade I, 68 were grade II, 76 were grade III, and 6 were grade IV. The median (IQR) AVM volume and diameter at repeat SRS were 1 (1.7) cm3 and 14 (9.8) mm, respectively (Table 1, Fig. 1).
Sankey diagram demonstrating the distributions of the treated AVMs in terms of SM grade prior to the first and second radiosurgery procedures and at last follow-up.
In the interval between the first and second radiosurgery procedures, 11 patients underwent endovascular embolization of the AVM, 1 underwent resection, and 2 underwent insertion of a ventriculoperitoneal shunt for hydrocephalus after AVM hemorrhage.
Patient Follow-Up
Ten patients had died of non–AVM-related conditions at a median (IQR) clinical follow-up of 57 (66.5) months, and 1 patient died of brain AVM hemorrhage 8 months after repeat SRS. Intracranial hemorrhage occurred in 14 (8.2%) patients at a median (IQR) interval of 40.5 (68.5) months after the second SRS procedure (Table 2, Fig. 2A). On multivariate analysis, age younger than 18 years (HR 5.4, 95% CI 1.7–17, p = 0.004) and maximum AVM diameter > 20 mm at repeat SRS (HR 6.2, 95% CI 1.8–21, p = 0.004) were associated with increased risk of hemorrhage after the second SRS procedure (see Table 5).
Patient clinical and radiological outcomes
| Characteristic | Value |
|---|---|
| Median (IQR) follow-up, mos | |
| Clinical | 57 (66.5) |
| MRI | 53 (71) |
| DSA | 36.5 (30.3) |
| Radiological | |
| Actuarial obliteration rate (95% CI), % | |
| 3-yr | 37.6 (29.5–44.8) |
| 5-yr | 57.3 (48.3–64.8) |
| Median (IQR) time to obliteration, mos | 36 (28) |
| Confirmation of obliteration | |
| MRI | 27 (15.9) |
| DSA | 83 (48.8) |
| RIC | |
| Median time to RIC (IQR), mos | 8 (6) |
| Crude asymptomatic RIC incidence | 44 (25.9) |
| Crude symptomatic RIC incidence | 15 (8.8) |
| Speech deficit | 2 (1.2) |
| Motor deficit | 9 (5.3) |
| Seizures | 5 (3.6) |
| Hydrocephalus | 1 (0.6) |
| Headache | 1 (0.6) |
| Crude permanent symptomatic RIC incidence | 9 (5.3) |
| Median (IQR) time to cyst formation, mos | 66 (55) |
| Crude cyst formation incidence | 4 (2.4) |
| Post-SRS hemorrhage | |
| Median (IQR) time to post-SRS hemorrhage, mos | 40.5 (68.5) |
| Crude post-SRS hemorrhage incidence | 14 (8.2) |
Values are shown as number (%) unless indicated otherwise.
Kaplan-Meier curves demonstrating the intracranial bleeding rate after repeat SRS (A), the occlusion rate of brain AVM after the second radiosurgery procedure (B), and the rate of RICs after repeat SRS for brain AVM (C).
Factors associated with brain AVM occlusion after repeat radiosurgery
| Characteristic | Univariate | Multivariate | ||
|---|---|---|---|---|
| p Value | HR (95% CI) | p Value | HR (95% CI) | |
| Female sex | 0.1 | |||
| Age <18 yrs at SRS | 0.01 | 0.45 (0.24–0.84) | 0.04 | |
| Diameter >20 mm | 0.01 | 0.51 (0.3–0.86) | 0.04 | 0.59 (0.35–0.99) |
| RIC after 1st SRS | 0.07 | |||
| Pre-SRS hemorrhage | 0.27 | |||
| SM eloquence | 0.51 | |||
| SM veins | 0.64 | |||
| Margin dose >19 Gy | 0.01 | 1.8 (1.1–3) | 0.01 | 1.8 (1.1–3) |
| RIC after 2nd SRS | 0.04 | 1.5 (1–2.2) | 0.16 | |
Factors associated with RICs after repeat SRS
| Characteristic | Univariate | Multivariate | ||
|---|---|---|---|---|
| p Value | HR (95% CI) | p Value | HR (95% CI) | |
| Female sex | 0.85 | |||
| Older age at SRS | 0.16 | |||
| Prescription vol >2 cm3 | 0.01 | 1.3 (1.1–1.6) | 0.17 | |
| RIC after 1st SRS | 0.02 | 1.8 (1.1–3) | 0.006 | 5.1 (1.6–16.3) |
| Pre-SRS hemorrhage | 0.04 | 0.54 (0.3–0.99) | 0.2 | |
| SM criteria | ||||
| Eloquence | 0.24 | |||
| Veins | 0.17 | |||
| Deep location | 0.67 | |||
| Margin dose >19 Gy | 0.56 | |||
| 12-Gy vol >2.5 cm3 | 0.05 | 2.7 (1–7.4) | 0.81 | |
Factors associated with intracranial hemorrhage after repeat radiosurgery for brain AVM
| Characteristic | Univariate | Multivariate | ||
|---|---|---|---|---|
| p Value | HR (95% CI) | p Value | HR (95% CI) | |
| Female sex | 0.4 | |||
| Age <18 yrs at SRS | 0.004 | 5.4 (1.7–17) | 0.03 | 4 (1.2–13.4) |
| RIC after 1st SRS | 0.95 | |||
| Pre-SRS hemorrhage | 0.83 | |||
| SM veins | 0.25 | |||
| Deep location | 0.15 | |||
| Diameter >20 mm | 0.004 | 6.2 (1.8–21) | 0.04 | 3.9 (1.1–14.4) |
Radiological Outcome
Follow-up with brain MRI and cerebral DSA was performed at medians (IQRs) of 53 (71) months and 36.5 (30.3) months, respectively. AVM obliteration was achieved in 110 (64.7%) patients at a median (IQR) of 36 (28) months after repeat SRS. AVM obliteration was confirmed with brain MRI and cerebral DSA in 27 (15.9%) and 83 (48.8%) patients, respectively (Fig. 2B). The actuarial 3-, 5-, and 10-year AVM obliteration rates were 37.6%, 57.3%, and 80.9%, respectively (Table 2). Margin dose ≥ 19 Gy was associated with higher obliteration rates (HR 1.8, 95% CI 1.1–3, p = 0.01). Residual AVM diameter greater than 20 mm (HR 0.59, 95% CI 0.35–0.99, p = 0.04) and age less than 18 years were associated with lower obliteration rates (HR 0.5, 95% CI 0.26–0.95, p = 0.04) (Table 3).
Adverse Effects
At a median of 8 months after repeat SRS, asymptomatic, symptomatic, and permanent RICs occurred at rates of 25.9%, 8.8%, and 5.3%, respectively (Fig. 2C). Symptomatic RICs occurred in 15 (8.8%) patients at a median (IQR) of 8 (6) months after repeat SRS and were permanent in 9 (5.3%) of these patients (Table 2). RIC after the first SRS procedure was the only factor that conferred increased risk of RIC after the second SRS procedure (HR 5.1, 95% CI 1.6–16.3, p = 0.006) (Table 4). All patients with symptomatic RICs were managed with a short course (generally approximately 2 weeks) of dexamethasone. Persistent symptoms occurred in 3 patients: 1 was treated with 4 months of oral dexamethasone, 1 with a single dose of bevacizumab, and 1 patient was managed with 6 months of oral dexamethasone and 4 cycles of bevacizumab.
Delayed cyst formation occurred at a median (IQR) of 66 (55) months after repeat SRS in 4 (2.4%) patients (Table 2). Delayed cysts were asymptomatic in 3 patients and were managed with observation. One patient developed hydrocephalus and was managed with ventriculoperitoneal shunt insertion.
Additional Interventions After Repeat SRS
Ten patients required a third SRS procedure, which resulted in AVM occlusion in 6 of these patients. Microsurgical resection was employed in 1 patient and resulted in AVM cure.
Radiosurgery-Associated Malignancy
There was 1 previously reported case of presumed radiation-induced meningioma 12 years after SRS.12 The tumor has remained stable and asymptomatic, and it has not required any intervention.
Discussion
AVM Obliteration
Repeat SRS may be considered for patients with residual brain AVM after incomplete AVM obliteration following initial SRS. Repeat SRS reportedly results in residual AVM obliteration rates between 58% and 71%.4,6,13–15 Smaller residual AVM volume,6,14 residual AVM volume < 1.5 cm,4 volume reduction ≥ 50% after the first SRS procedure,14 higher prescription dose,6,13 prescription dose > 20 Gy,4 maximum dose > 40 Gy,4 and male sex6 have been associated with increased residual AVM obliteration rates. A history of intracranial hemorrhage was associated with decreased residual AVM obliteration rates after repeat SRS.6
In our series of 170 brain AVM patients treated with repeat SRS, the AVM obliteration rate was 64.7% and the median time to residual AVM obliteration after the second SRS procedure was 36 months. The actuarial 3-, 5-, and 10-year AVM obliteration rates were 37.6%, 57.3%, and 80.9%, respectively. A margin dose ≥ 19 Gy was associated with higher AVM obliteration rates. Residual AVM diameter > 20 mm and age < 18 years were associated with lower obliteration rates (Table 3). It is our practice to offer retreatment with either SRS or microsurgery to patients with a patent AVM approximately 3–4 years after the first or second SRS procedure. In our study, 60 patients had residual AVM after the second radiosurgery procedure at last follow-up. Of these, 10 patients underwent a third SRS procedure and 1 underwent AVM resection. In addition, 1 patient underwent microsurgical resection for a residual arteriovenous fistula, 3 patients were still in the latency period for AVM occlusion, and 3 declined repeat SRS treatment.
Post-SRS Intracranial Hemorrhage
The main aim of initial and repeat brain AVM radiosurgery is to prevent intracranial hemorrhage. A major limitation of AVM radiosurgery is the latency period from SRS to AVM occlusion, during which intracranial hemorrhage may still occur. The annual incidence of intracranial hemorrhage during the latency period after repeat AVM radiosurgery is between 1.8% and 3.7%,4,6,14,15 which is similar to the annual incidence of intracranial hemorrhage during the latency period after initial SRS.4,6 Kano et al. reported that larger target volume at repeat SRS and higher hemorrhage rate prior to repeat SRS were associated with increased risk of hemorrhage after repeat SRS.14 In our analysis, 1 of the 11 deaths was due to intracranial hemorrhage 8 months after repeat SRS. The annual risk of bleeding within 5 years after repeat radiosurgery for residual AVM was less than 2% per year and comparable to the annual intracranial hemorrhage incidence of untreated brain AVM. AVM diameter > 20 mm and age < 18 years conferred increased risk of intracranial hemorrhage during the latency period after repeat radiosurgery (Table 5).
Adverse Radiation Events
RICs usually occur 6 to 18 months after radiosurgery for brain AVM.16 In a systematic review that evaluated repeat radiosurgery for brain AVM, the mean rate of symptomatic RIC was 7.4%.7 Factors associated with increased risk of adverse radiation events after repeat SRS include prior embolization,14 higher SM AVM grade,14 and 12-Gy volume at repeat SRS.13
In our study, asymptomatic, symptomatic, and permanent RICs occurred at rates of 25.9%, 8.8%, and 5.3%, respectively, at a median of 8 months after repeat radiosurgery. RIC after the first radiosurgery procedure was the only factor associated with an increased risk of RIC after repeat SRS. At our institution, a short course of oral corticosteroids is the typical initial treatment option for symptomatic RIC. Patients with persistent symptoms and patients who decline surgery after cessation of corticosteroids are managed with a more protracted course of corticosteroids. Patients who do not respond to corticosteroids and those who develop corticosteroid-related adverse effects are treated with bevacizumab. In a prior study, delayed post-SRS cysts occurred with an incidence rate of 3% at a mean latency period of 78 months, and approximately one-third of these patients were symptomatic and required surgical intervention.17 In our cohort, delayed cysts occurred in 2.4% of patients at a median of 66 months after second SRS. One patient developed cyst-associated hydrocephalus and required insertion of a cystoperitoneal shunt.
Radiosurgery-Associated Malignancy
The risk of SRS-associated malignancy is low.18,19 In a multicenter study of 14,168 patients, the overall incidence of radiosurgery-associated malignancy was 6.8 per 100,000 patient-years, which is similar to the incidence rates of the general populations of the United States and some European countries.19 In our cohort, 1 patient developed a radiosurgery-associated asymptomatic meningioma 12 years after repeat radiosurgery.12
Limitations
Our study was limited by its retrospective and single-center design. In addition, we reported the outcomes of patients treated at different periods with different Gamma Knife and neuroimaging technologies. Therefore, due to advancements in the Gamma Knife and radiological technologies used, in addition to our center’s increasing experience, it is likely that the patients treated at the end of the study period may have benefited the most from repeat AVM radiosurgery.
Conclusions
Repeat SRS results in the obliteration of residual AVM in the majority of patients and confers a low risk of complications. Smaller AVM nidus, age ≥ 18 years, and prescription dose portended increased probability of AVM obliteration after repeat SRS. The incidence rates of intracranial hemorrhage and symptomatic RIC after repeat SRS were comparable to those after initial SRS.
Disclosures
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.
Author Contributions
Conception and design: Sheehan, Pikis. Acquisition of data: Pikis, Mantziaris, Ramanathan, Xu. Analysis and interpretation of data: Pikis, Mantziaris. Drafting the article: Pikis. Critically revising the article: Sheehan, Xu. Reviewed submitted version of manuscript: all authors. Statistical analysis: Mantziaris. Study supervision: Sheehan.
References
- 1↑
Al-Shahi R, Warlow C. A systematic review of the frequency and prognosis of arteriovenous malformations of the brain in adults. Brain. 2001;124(Pt 10):1900–1926.
- 2↑
Schlienger M, Nataf F, Lefkopoulos D, et al. Repeat linear accelerator radiosurgery for cerebral arteriovenous malformations. Int J Radiat Oncol Biol Phys. 2003;56(2):529–536.
- 3
Kano H, Iqbal FO, Sheehan J, et al. Stereotactic radiosurgery for chordoma: a report from the North American Gamma Knife Consortium. Neurosurgery. 2011;68(2):379–389.
- 4↑
Yen CP, Jain S, Haq IU, et al. Repeat γ knife surgery for incompletely obliterated cerebral arteriovenous malformations. Neurosurgery. 2010;67(1):55–64.
- 5
Liscák R, Vladyka V, Simonová G, et al. Arteriovenous malformations after Leksell gamma knife radiosurgery: rate of obliteration and complications. Neurosurgery. 2007;60(6):1005–1016.
- 6↑
Stahl JM, Chi YY, Friedman WA. Repeat radiosurgery for intracranial arteriovenous malformations. Neurosurgery. 2012;70(1):150–154.
- 7↑
Awad AJ, Walcott BP, Stapleton CJ, Ding D, Leed CC, Loeffler JS. Repeat radiosurgery for cerebral arteriovenous malformations. J Clin Neurosci. 2015;22(6):945–950.
- 8↑
Kaufmann TJ, Huston J III, Mandrekar JN, Schleck CD, Thielen KR, Kallmes DF. Complications of diagnostic cerebral angiography: evaluation of 19,826 consecutive patients. Radiology. 2007;243(3):812–819.
- 9↑
Khandanpour N, Griffiths P, Warren D, Hoggard N. Prospective comparison of late 3T MRI with conventional angiography in evaluating the patency of cerebral arteriovenous malformations treated with stereotactic radiosurgery. Neuroradiology. 2013;55(6):683–687.
- 10↑
RStudio Team. RStudio: Integrated development environment for R. Accessed May 16, 2022. http://www.rstudio.com/
- 11↑
R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.Accessed May 16, 2022. https://www.R-project.org/
- 12↑
Sheehan J, Yen CP, Steiner L. Gamma Knife surgery–induced meningioma Report of two cases and review of the literature. J Neurosurg. 2006;105(2):325–329.
- 13↑
Maesawa S, Flickinger JC, Kondziolka D, Lunsford LD. Repeated radiosurgery for incompletely obliterated arteriovenous malformations. J Neurosurg. 2000;92(6):961–970.
- 14↑
Kano H, Kondziolka D, Flickinger JC, et al. Stereotactic radiosurgery for arteriovenous malformations, Part 3: outcome predictors and risks after repeat radiosurgery. J Neurosurg. 2012;116(1):21–32.
- 15↑
Karlsson B, Kihlström L, Lindquist C, Steiner L. Gamma knife surgery for previously irradiated arteriovenous malformations. Neurosurgery. 1998;42(1):1–6.
- 16↑
Ilyas A, Chen CJ, Ding D, et al. Radiation-induced changes after stereotactic radiosurgery for brain arteriovenous malformations: a systematic review and meta-analysis. Neurosurgery. 2018;83(3):365–376.
- 17↑
Ilyas A, Chen CJ, Ding D, et al. Cyst formation after stereotactic radiosurgery for brain arteriovenous malformations: a systematic review. J Neurosurg. 2018;128(5):1354–1363.
- 18↑
Bunevicius A, Pikis S, Schlesinger D, Sheehan J. Editorial: Radiosurgical induced malignancy associated with stereotactic radiosurgery. Acta Neurochir (Wien). 2021;163(4):969–970.
- 19↑
Wolf A, Naylor K, Tam M, et al. Risk of radiation-associated intracranial malignancy after stereotactic radiosurgery: a retrospective, multicentre, cohort study. Lancet Oncol. 2019;20(1):159–164.
