Longitudinal Assessment of Stereotactic Radiosurgery Outcomes in Brain Arteriovenous Malformations: Continuous Monitoring of Treatment Effects

Abstract

Background

Arteriovenous malformations (AVMs) present significant treatment challenges due to their hemorrhage risk and associated neurological deficits. Stereotactic radiosurgery (SRS) is a technique designed to obliterate AVMs while minimizing complications.

Objective

This study aims to evaluate the obliteration rate and clinical outcomes of patients undergoing SRS using linear accelerators (LINAC) for the treatment of brain AVMs.

Materials and Methods

This retrospective and prospective study included 56 consecutive patients who underwent SRS for AVMs between 2008 and 2021, with the study period till 2023. The mean patient age was 29.21 years (range: 8–67 years). Complete obliteration rates and clinical improvements were assessed, and statistical analysis was conducted using SPSS.

Results

The mean planning target volume was 11.39 cc (median: 7.74 cc) and the average modified AVM score was 1.79 (median: 1.56). The obliteration rate was 83.92%, with 7.1% of patients exhibiting residual AVM and 5.3% requiring surgical excision. Post-radiation hemorrhage occurred in 7.14% of cases.

Conclusion

LINAC-based SRS proves to be a highly effective and safe treatment modality for small or surgically inaccessible brain AVMs.

 

MissingLinkText

Keywords

stereotactic radiosurgery - brain arteriovenous malformations - LINAC - complete obliteration - clinical improvement

Introduction

Arteriovenous malformations (AVMs) are congenital abnormalities characterized by direct arteriovenous communications without an intervening capillary bed, typically developing during the late somite stage (4th to 8th week of embryonic life).[1] AVMs pose a significant risk of nontraumatic intracerebral hemorrhage, with a prevalence of 18 per 100,000 population, resulting in a 10% mortality rate and a 30 to 50% morbidity rate per bleed.[2] The estimated annual hemorrhage risk ranges from 2 to 4%.

Treating AVMs involves several options, including microsurgery, endovascular embolization, and stereotactic radiosurgery (SRS). Microsurgical excision is often preferred due to its direct approach. However, for deep-seated AVMs, SRS is increasingly considered a compelling and definitive option. The choice of treatment depends on various factors, highlighting the necessity for a personalized and comprehensive strategy.

SRS employs a stereotactic three-dimensional coordinate system to precisely align a virtual target from diagnostic images with the corresponding target within the patient's body. Despite its name, SRS does not involve conventional surgical techniques. Instead, it delivers a focused single dose of radiation, offering a potential alternative to surgery in specific cases involving AVMs.[3]

The objective of this study is to evaluate the obliteration rate and clinical outcomes in patients undergoing SRS with linear accelerators (LINAC) for the treatment of brain AVMs.

 

MissingLinkText

Materials and Methods

This observational study, conducted both retrospectively and prospectively, focused on individuals who underwent SRS for AVMs at our institution between January 2008 and December 2021, with the study period till December 2023.

Patients were selected based on criteria that included targeting eloquent brain areas where there is a heightened risk of neurological deficits and those who chose radiosurgery over surgical intervention, even when the AVM was located outside an eloquent area. Inclusion criteria required patients to attend regular follow-up appointments and consent to providing medical data through modalities such as computed tomography (CT) brain, CT angiogram, magnetic resonance imaging (MRI) brain, MRI angiogram, and digital subtraction angiography (DSA).

Participants provided informed consent, and ethical approval was obtained from the institutional ethics committee. Out of 68 patients treated for AVMs, 12 were lost to follow-up. Consequently, the study included 56 patients to assess treatment outcomes.

The study analyzed data from 56 consecutive cases meeting inclusion and exclusion criteria that underwent single-fraction SRS. Prior to SRS, patients underwent MRI and DSA for planning.

MRI scans with a 1 mm slice thickness, no interslice gap, and a 256 × 256 matrix size were fused with CT planning images on a workstation. Our institute has employed a frameless technique for SRS since 2016, involving the fusion of MRI and pretreatment DSA with neurosurgeons and radiologists delineating the nidus target. Out of 56 consecutive patients, 20 patients underwent frameless technique since 2016.

Treatment was administered using Siemens ONCOR Impression Plus LINAC with micromultileaf collimators for 40 patients, while 16 patients were treated with Varian TrueBeam STx LINAC. Planning utilized Brainlab's iPlan Software version 4.1 and Eclipse Software version 16.1 (Varian Inc). Post-fusion, the nidus and organs at risk (including the eye lens, optic nerve, optic chiasm, brain stem, and normal brain) were contoured. The planning target volume (PTV) was generated with a margin of 1 mm around the nidus. A single dose of radiation was delivered to treat the AVMs. Following SRS, patients were followed with serial MRI/CT angiography at regular intervals followed by cerebral angiography (DSA) for the final assessment of complete obliteration.

A modified radiosurgery-based AVM grading score (Modified Pollock–Flickinger Scoring System) was applied to all patients preprocedure. Statistical analysis was conducted using IBM SPSS Statistics software version 24.0. The modified AVM score is a refined tool used to evaluate the severity and treatment outcomes of AVMs.

 

MissingLinkText

Results

The study included 56 consecutive patients who met the specified inclusion/exclusion criteria, providing a comprehensive view of the patient population. The mean age of the patients was 29.2 years, with an age range of 8 to 67 years. This wide range indicates the diverse nature of the patient group and reflects the broad applicability of the findings across different age groups. In terms of gender distribution, 40 of the patients were male (71.42%) and 16 were female (28.57%), giving a male-to-female ratio of 2.5:1. This skewed distribution suggests a higher prevalence or detection rate of AVMs in males.

Regarding the size and location of the AVMs, 21 cases had a diameter of less than 3 cm, while 35 cases had diameters between 3 and 6 cm. Venous drainage was predominantly superficial in 49 cases, with deep drainage observed in seven cases. Of the 56 patients, 33 had AVMs located in eloquent brain areas, while 23 were not in such regions.

The Spetzler–Martin (SM) grading revealed that 15 cases were classified as grade 1, 20 cases each as grades 2 and 3, and 1 case as grade 4. Obliteration rates were stratified according to SM grades. Patients with SM grade I and II AVMs demonstrated higher obliteration rates of 92% (13/14) and 85% (17/20), respectively. In contrast, grade III AVMs had a moderate obliteration rate of 70% (14/20), while the single patient with a grade IV AVM did not achieve complete obliteration. A negative correlation was observed between increasing SM grade and obliteration rate, indicating that higher grade AVMs were less likely to be completely obliterated. However, this trend did not reach statistical significance (p = 0.072). These findings are consistent with prior literature, where increased AVM complexity and size, as reflected by higher SM grades, are associated with lower radiosurgical success rates.

The mean PTV was 11.39 cc, with a median of 7.74 cc, indicating that half of the patients had a PTV of less than 7.74 cc. The highest recorded PTV was 63.4 cc.

The mean modified AVM score in the study was 1.79, with a median of 1.56, suggesting a relatively low AVM score for most patients, which correlates with better prognostic outcomes. This correlation was statistically significant, with a p-value of 0.0004. In terms of treatment outcomes, the overall obliteration rate was 83.92%, indicating that a significant majority of patients achieved complete obliteration of the AVM posttreatment. During the follow-up period of ≤3 years, 76.60% of patients showed complete obliteration ([Fig. 1]), while 7.1% of patients had residual AVMs, meaning they did not achieve full obliteration with the initial treatment. Additionally, 5.3% of patients required surgical excision due to hemorrhage, indicating a minority required further surgical intervention.

Fig. 1 CT angiography of pre (left) and post (right) SRS of AVM.

Post-radiation hemorrhage occurred in 7.14% of the cases, representing a relatively low but notable risk of hemorrhage following radiation treatment. Out of the 56 patients, 2 cases resulted in death due to coronavirus disease, preventing clinical evaluation. Among the remaining patients, 47 were asymptomatic and showed clinical improvement, while 7 exhibited neurological deficits at follow-up.

Among the seven patients (12.5%) who developed neurological deficits following SRS, the majority had AVMs located in eloquent brain areas (five out of seven cases). The average nidus volume in these patients was 18.6 cc, ranging from 8.2 to 63.4 cc. Motor weakness was the most common deficit, observed in four patients, followed by visual disturbances in two patients and cognitive slowing in one.

The onset of symptoms ranged between 3 and 18 months after treatment. MRI findings in these patients revealed perilesional edema or signs of radiation-induced changes. All patients responded well to corticosteroid therapy, with symptoms resolving over time.

The follow-up duration ranged from 24 to 180 months, with a median follow-up of 114 months and a mean of 113.14 months. This extensive follow-up period demonstrates the long-term monitoring of patients posttreatment to manage any late-onset complications.

The study further explored the relationship between PTV and obliteration rates. Patients with PTVs less than 10 mL had higher obliteration rates (32 patients) compared with those with PTVs greater than 10 mL (15 patients), though the p-value of 0.458 suggests no statistically significant association between PTV and obliteration. Notably, even cases with large PTVs, such as the highest recorded PTV of 63.4 cc, were successfully obliterated. Age also played a role in obliteration rates, with patients aged 35 years or younger showing higher obliteration rates (35 patients) than those older than 35 years (12 patients), with a statistically significant p-value of 0.017 ([Fig. 2]).

Fig. 2 Graph comparing the obliteration rate with various variables (PTV, Age, and SRS dose).

There was a strong correlation between the modified AVM score and obliteration rates. With an average modified AVM score of 1.79 and a median of 1.56, the study showed a high obliteration rate of 83.92%, highlighting the effectiveness of LINAC-based SRS, especially in cases with lower scores indicating smaller or less complex AVMs. The p-value of 0.00004 further emphasizes the robustness of this correlation, suggesting that lower modified AVM scores are predictive of higher obliteration success rates, validating the modified AVM score as a useful prognostic tool in clinical practice.

The effectiveness of different doses of SRS was also analyzed. Patients receiving an SRS dose of less than 18 Gy had the highest rate of complete obliteration (90.0%). Those who received doses between 18 and 20 Gy had an obliteration rate of 83.8%, while those who received doses greater than 20 Gy had a lower obliteration rate of 66.7%. Despite these differences, the p-value was greater than 0.05, indicating that the differences in obliteration rates across different SRS doses were not statistically significant.

In terms of obliteration rates and the impact of partial embolization, the study found that 47 out of 56 AVMs were completely obliterated after treatment. Of these, 19 had undergone partial embolization, while 28 had not. Interestingly, a higher percentage (50%) of AVMs achieved complete obliteration without prior embolization compared with those that had undergone partial embolization (33.9%). However, a chi-square test revealed a p-value of 0.544, indicating no statistically significant difference in obliteration rates between embolized and nonembolized AVMs.

 

MissingLinkText

Discussion

AVMs are among the most commonly detected vascular malformations. The main goal in their treatment is to maximize the obliteration rate of AVMs while minimizing treatment-associated neurological deficits and complications, a balance achievable through SRS. This study demonstrates that AVM obliteration can be achieved with minimal risk of complications posttreatment.

Originally believed to be primarily congenital, it was later discovered that AVMs could also form de novo.[4] On imaging, AVMs appear as serpiginous vessels on contrast CT scans and as a honeycomb-like structure on T1- and T2-weighted MRI images.[5] [6] The precise pathogenesis remains unclear but may involve a developmental arrest in the resorption of plexiform anastomoses between arteries and veins, resulting in the persistence of abnormal arteriovenous connections.[2] AVMs are the most commonly detected symptomatic vascular malformations[7] and account for 2% of all strokes and 38% of intracranial bleeding cases in individuals aged 15 to 45 years.[8] They predominantly occur supratentorially but are occasionally found in the cerebellum, brainstem, and ventricles.

High-dose radiation in AVM treatment damages endothelial cells, stimulates smooth muscle cell growth, thickens vascular walls, and results in gradual vessel lumen obliteration. SRS procedures utilize diverse radiation sources, including LINAC, Gamma Knife, and proton beam therapy.

In this study, 56 cases met the inclusion/exclusion criteria with a mean age of 29.2 years and an age range of 8 to 67 years. Among the 56 cases, 40 (71.42%) were male and 16 (28.57%) were female, yielding a male-to-female sex ratio of 2.5:1. This preponderance toward males is comparable to studies by Al-Shahi et al[9] and Luo et al[10] although some reports indicate that AVMs affect males and females equally.[11]

In this study, 66% of AVMs had a maximum diameter of less than 4 cm and 34% were between 4 and 6 cm, compared with Zacest et al,[12] where SRS was performed on AVMs with diameters of 0.5 to 4.6 cm. Additionally, 88% of patients in this study had superficial venous drainage, whereas Da Costa et al[13] reported predominantly deep venous drainage in 53.8% of cases.

The complete obliteration rate in this study was 84%, higher than the 64% reported by Park et al,[14] 77% reported by Tamura et al,[15] and 74% by Chen et al[16] ([Table 1]). Notably, 64% of patients achieved complete obliteration by the end of the third year, with a 7.14% rate of post-radiation hemorrhage comparable to other studies with rates of 2.6 to 10%.[17] [18] [19]

Table 1 Comparative analysis of obliteration rates in AVM stereotactic radiosurgery studies

Author	Patient number (n)	Median follow-up (mo)	Obliteration rate (%)	Modality
Tamura et al[15]	22	47	77	LINAC
Skjøth-Rasmussen et al[24]	50	24	78	GAMMAKNIFE
Ding et al[25]	502	62	80	GAMMAKNIFE
Bowden et al[26]	171	74	67	GAMMAKNIFE
Mohan et al[23]	15	60	77.14	LINAC
Kiran et al[27]	103	26	86	GAMMAKNIFE
Javalkar et al[28]	15	23	46.7	GAMMAKNIFE
Mark et al[29]	85	57	61	LINAC
Gawish et al[30]	71	26	66	LINAC
Our series	56	114	83.92	LINAC

Abbreviation: AVM, arteriovenous malformation.

Our study findings imply that partial embolization alone may not substantially affect the complete obliteration of AVMs, given the nonsignificant p-value. However, a nuanced perspective emerges from Nadeem et al's study,[20] indicating that partial embolization achieved satisfactory obliteration rates. Crucially, the timing between these interventions significantly influences treatment efficacy, with shorter intervals correlating with superior outcomes.

Our study demonstrated that single-dose radiosurgery achieved an obliteration rate of 79% for large AVMs (PTV > 10 cc). In our study, the largest AVM—measuring 63.4 cc—was treated with a single session of SRS. This decision was made after the patient declined multiple treatment visits due to personal, financial, and logistical reasons. We delivered a dose of 16 Gy, aiming to balance safety with treatment effectiveness, especially considering the lesion's large size and its location near critical brain structures. While multi-session treatments are usually preferred for large AVMs to reduce risks, this was not feasible in this case due to system constraints and the patient's situation. The AVM was completely obliterated, which was a positive outcome. This case reflects both the promise and the challenges of treating large AVMs with a single high-dose session.

This finding contrasts with the results from Garg and Singh[21] and Mukherjee et al,[22] which highlighted the benefits of fractionated radiation in treating large AVMs. Garg and Singh reported high obliteration rates and fewer complications with volume fractionation SRS, while Mukherjee et al achieved a 43% obliteration rate using dose-fractionated Gamma Knife radiosurgery. These studies suggest that fractionated approaches might be more effective in managing large AVMs by allowing higher cumulative doses and reducing the risk of complications compared with single-dose treatments.

In our study, patients who received less than 18 Gy had better obliteration rates than those who received more than 20 Gy. This might seem surprising, as we usually expect higher doses to work better. But this difference likely reflects the nature of the AVMs being treated. Higher doses were often given to larger, deeper, or more complex AVMs, which are harder to treat successfully. Similar results have been seen in other studies, like the one by Chen et al,[16] showing that treatment success depends not just on the dose, but also on the size, shape, and location of the AVM.

In this study, 12% of patients still had neurological deficits while no deficits were observed in 84% of patients who achieved complete obliteration, consistent with findings by Park et al.[14]

To the best of our knowledge, this is the largest single-center study from India evaluating LINAC-based SRS for brain AVMs. Our cohort included 56 patients, with a median follow-up of 114 months—nearly a decade—which is notably longer than most reported series. Indian studies, such as the one by Mohan et al[23] from All India Institute of Medical Sciences Jodhpur, included 15 patients with a follow-up of 2 to 5 years. What sets our study apart is not only the sample size and follow-up duration but also the inclusion of complex cases, including AVMs located in eloquent brain areas and those with large nidus volumes up to 63.4 cc. We went beyond reporting obliteration rates by analyzing how factors like AVM size, SM grade, and radiation dose influenced outcomes and complications. Importantly, our findings reflect the realities of treating AVMs in the Indian health care setting, where socioeconomic and logistical challenges often shape treatment decisions. While our results are in line with international experience, they offer much-needed long-term data from an Indian context and help fill a significant gap in the global AVM radiosurgery literature.

 

MissingLinkText

Conclusion

SRS using LINAC technology is an exceptionally effective and safe treatment modality for surgically inaccessible brain AVMs. This method achieved a remarkable 84% complete obliteration rate coupled with significant clinical improvement, all while maintaining minimal side effects.

 

MissingLinkText

 

MissingLinkText

Conflict of Interest

None declared.

Acknowledgment

The authors wish to acknowledge the support of the Radiosurgery and Radiation Oncology Department for providing the necessary facilities for the preparation of this article.

Authors' Contribution

P.K.R.G.: study design, data collection, manuscript writing, and data analysis. A.B.: study design, manuscript proof reading, and data analysis. A.H.: study design, manuscript proof reading, and data analysis. A.K.: study design and manuscript writing. R.S.: study design, manuscript proof reading, and data interpretation. A.K.: manuscript writing and data collection. C.P.: manuscript writing and data collection. N.D.: study design, manuscript writing, and data analysis. S.K.V.: manuscript writing and proof reading, and data analysis. R.H.: manuscript writing and data analysis.

• References

• 1 Ramamurthi B, Tandon P. Textbook of Neurosurgery. 3rd ed.. New Delhi: JP Brothers Medical Publishers; 2012: 1052
• 2 Maruyama K, Kawahara N, Shin M. et al. The risk of hemorrhage after radiosurgery for cerebral arteriovenous malformations. N Engl J Med 2005; 352 (02) 146-153
• 3 Chen JC, Girvigian MR. Stereotactic radiosurgery: instrumentation and theoretical aspects-part 1. Perm J 2005; 9 (04) 23-26
• 4 Michelsen WJ. Natural history and pathophysiology of arteriovenous malformations. Clin Neurosurg 1979; 26: 307-313
• 5 Gonzalez LF, Bristol RE, Porter RW, Spetzler RF. De novo presentation of an arteriovenous malformation. Case report and review of the literature. J Neurosurg 2005; 102 (04) 726-729
• 6 Mahajan A, Manchandia TC, Gould G, Bulsara KR. De novo arteriovenous malformations: case report and review of the literature. Neurosurg Rev 2010; 33 (01) 115-119
• 7 Furlan AJ, Whisnant JP, Elveback LR. The decreasing incidence of primary intracerebral hemorrhage: a population study. Ann Neurol 1979; 5 (04) 367-373
• 8 Gross CR, Kase CS, Mohr JP, Cunningham SC, Baker WE. Stroke in south Alabama: incidence and diagnostic features–a population based study. Stroke 1984; 15 (02) 249-255
• 9 Al-Shahi R, Fang JS, Lewis SC, Warlow CP. Prevalence of adults with brain arteriovenous malformations: a community based study in Scotland using capture-recapture analysis. J Neurol Neurosurg Psychiatry 2002; 73 (05) 547-551
• 10 Luo J, Lv X, Jiang C, Wu Z. Brain AVM characteristics and age. Eur J Radiol 2012; 81 (04) 780-783
• 11 Mascitelli JR, Yoon S, Cole TS, Kim H, Lawton MT. Does eloquence subtype influence outcome following arteriovenous malformation surgery?. J Neurosurg 2018; 131 (03) 876-883
• 12 Zacest AC, Caon J, Roos DE, Potter AE, Sullivan T. LINAC radiosurgery for cerebral arteriovenous malformations: a single centre prospective analysis and review of the literature. J Clin Neurosci 2014; 21 (02) 241-245
• 13 da Costa L, Wallace MC, Ter Brugge KG, O'Kelly C, Willinsky RA, Tymianski M. The natural history and predictive features of hemorrhage from brain arteriovenous malformations. Stroke 2009; 40 (01) 100-105
• 14 Park HR, Lee JM, Kim JW. et al. Timestaged gamma knife stereotactic radiosurgery for large cerebral arteriovenous malformations: a preliminary report. PLoS One 2016; 11 (11) e0165783
• 15 Tamura N, Hayashi M, Chernov M. et al. Outcome after Gamma Knife surgery for intracranial arteriovenous malformations in children. J Neurosurg 2012; 117 (Suppl): 150-157
• 16 Chen JC, Mariscal L, Girvigian MR. et al. Hypofractionated stereotactic radiosurgery for treatment of cerebral arteriovenous malformations: outcome analysis with use of the modified arteriovenous malformation scoring system. J Clin Neurosci 2016; 29: 155-161
• 17 Bir SC, Ambekar S, Maiti TK, Nanda A. Clinical outcome and complications of gamma knife radiosurgery for intracranial arteriovenous malformations. J Clin Neurosci 2015; 22 (07) 1117-1122
• 18 Ding D, Yen CP, Starke RM, Xu Z, Sun X, Sheehan JP. Outcomes following single-session radiosurgery for high-grade intracranial arteriovenous malformations. Br J Neurosurg 2014; 28 (05) 666-674
• 19 Friedman WA, Bova FJ, Bollampally S, Bradshaw P. Analysis of factors predictive of success or complications in arteriovenous malformation radiosurgery. Neurosurgery 2003; 52 (02) 296-307 , discussion 307–308
• 20 Nadeem M, Shah K, Modi M. et al. Gamma Knife Radiosurgery in partially embolised arteriovenous malformations: management dilemmas and outcomes. Neurol India 2023; 71 (Supplement): S90-S99
• 21 Garg K, Singh M. Volume fractionation stereotactic radiosurgery for large volume intracranial arteriovenous malformations. Neurol India 2023; 71 (Supplement): S82-S89
• 22 Mukherjee KK, Kumar N, Tripathi M. et al. Dose fractionated gamma knife radiosurgery for large arteriovenous malformations on daily or alternate day schedule outside the linear quadratic model: proof of concept and early results. A substitute to volume fractionation. Neurol India 2017; 65 (04) 826-835
• 23 Mohan A, Tiwari S, Pareek P. et al. Linear Accelerator (LINAC) radiosurgical management of brain arteriovenous malformations: an experience from a tertiary care center. Cureus 2024; 16 (12) e76232
• 24 Skjøth-Rasmussen J, Roed H, Ohlhues L, Jespersen B, Juhler M. Complications following linear accelerator based stereotactic radiation for cerebral arteriovenous malformations. Int J Radiat Oncol Biol Phys 2010; 77 (02) 542-547
• 25 Ding D, Yen CP, Xu Z, Starke RM, Sheehan JP. Radiosurgery for low-grade intracranial arteriovenous malformations. J Neurosurg 2014; 121 (02) 457-467
• 26 Bowden G, Kano H, Caparosa E. et al. Stereotactic radiosurgery for arteriovenous malformations of the postgeniculate visual pathway. J Neurosurg 2015; 122 (02) 433-440
• 27 Kiran NA, Kale SS, Vaishya S. et al. Gamma Knife surgery for intracranial arteriovenous malformations in children: a retrospective study in 103 patients. J Neurosurg 2007; 107 (6, Suppl): 479-484
• 28 Javalkar V, Pillai P, Vannemreddy P, Caldito G, Ampil F, Nanda A. Gamma knife radiosurgery for arteriovenous malformations located in eloquent regions of the brain. Neurol India 2009; 57 (05) 617-621
• 29 Mark F, Jin AH, Zacest A. et al. LINAC stereotactic radiosurgery for brain arteriovenous malformations: an updated single centre analysis of outcomes. J Clin Neurosci 2022; 102: 54-59
• 30 Gawish A, Röllich B, Ochel HJ, Brunner TB. Linac-based stereotactic radiosurgery for brain arteriovenous malformations. Radiat Oncol 2022; 17 (01) 161

Publication History

Article published online:
28 July 2025