Radiosurgery as a Stand-Alone Treatment Option for Cerebral Dural Arteriovenous Fistulas: The Vienna Series

Abstract

Background Gamma Knife radiosurgery (GKRS) has been demonstrated to be an effective and safe treatment method for dural arteriovenous fistulas (DAVFs). However, only few studies, mostly with limited patient numbers, have evaluated radiosurgery as a sole and upfront treatment option for DAVFs.

Methods Thirty-three DAVF patients treated with GKRS as a stand-alone management at our institution between January 1992 and January 2020 were included in this study. Obliteration rates, time to obliteration, neurologic outcome, and complications were evaluated retrospectively.

Results Complete overall obliteration was achieved in 20/28 (71%) patients. The postradiosurgery actuarial rates of obliteration at 2, 5, and 10 years were 53, 71, and 85%, respectively. No difference in time to obliteration between carotid–cavernous fistulas (CCFs; 14/28, 50%, 17 months; 95% confidence interval [CI]: 7.4–27.2) and non-CCFs (NCCFs; 14/28, 50%, 37 months; 95% CI: 34.7–38.5; p = 0.111) were found. Overall, the neurologic outcome in our series was highly favorable at the time of the last follow-up. A complete resolution of symptoms was seen in two-thirds (20/30, 67%) of patients. One patient with multiple DAVFs suffered from an intracranial hemorrhage of the untreated lesion and died during the follow-up period, resulting in a yearly bleeding risk of 0.5%. No complications after radiosurgery were observed in our series.

Conclusion Our results show that GKRS is a safe and effective stand-alone management option for selected DAVF patients.

Keywords

carotid–cavernous fistula - dural arteriovenous fistula - Gamma Knife radiosurgery

Introduction

Dural arteriovenous fistulas (DAVFs) are rare intracranial vascular malformations and consist of a pathologic shunt between meningeal arteries and a venous sinus or a pial cortical vein.[1] [2] The clinical presentation is highly variable and depends primarily on the localization and the venous drainage pattern of the fistula.[3] DAVFs without cortical venous drainage (CVD) are often asymptomatic or present with benign symptoms, such as pulsatile tinnitus or orbital symptoms (chemosis, exophthalmos, ocular pain, visual acuity, and cranial nerve palsy).[4] Fistulas with CVD often present with more aggressive symptoms that are mostly related to venous congestion. These include intracranial hemorrhage, seizures, progressive dementia, and other focal neurologic deficits.[5]

Current management options for DAVFs include microsurgery, endovascular embolization, stereotactic radiosurgery (SRS), and various combinations thereof.[1] [2] Endovascular embolization and microsurgery allow an immediate obliteration of the arteriovenous shunt, which results in an instant reduction of symptoms. Therefore, endovascular embolization or microsurgery has been considered the preferred treatment option for this pathology.[1] [2] Due to the prolonged time to obliteration, SRS has typically been reserved for lesions that cannot be cured with other approaches. Still, SRS has become an important treatment option for cerebral DAVFs, either as adjuvant or salvage therapy, especially for DAVFs not eligible for endovascular embolization or microsurgical treatment.[6] So far, SRS was shown to be an effective, alternative, adjuvant/salvage treatment method with high obliteration and low complication rates.[6] [7] [8] [9] [10] However, only a few studies, mostly with limited patient numbers, have evaluated radiosurgery as a sole and upfront treatment option.[11] [12] [13] Therefore, this study aims to present our consecutive series of Gamma Knife radiosurgery (GKRS) as a stand-alone treatment method for cerebral DAVFs.

Material and Methods

Patient Sample and Data Evaluation

Patient records, multimodal imaging at diagnosis, and radiosurgical plans from January 1992 to January 2020 were retrospectively reviewed to identify DAVF patients treated with GKRS at our department.

Overall, 86 patients were treated with GKRS. Thirty-three of the 86 patients (38%) patients were treated with SRS alone, 47/86 (55%) were treated with a combined radiosurgical and endovascular approach, and 6/86 (7%) were treated with a combination of GKRS, endovascular embolization, and microsurgery. Patients who underwent combined treatment (53/86 [62%]) were excluded from the study ([Fig. 1]). Moreover, DAVF characteristics, such as localization, CVD, and Borden/Cognard/Barrow classification, were evaluated retrospectively from digital subtraction angiography (DSA) performed at the time of diagnosis. Subsequently, DAVFs were divided into two groups: 17/33 (52%) DAVFs were carotid–cavernous fistulas (CCFs) and 16/33 (48%) were non-CCFs (NCCFs; [Table 1]).

Fig. 1 Flowchart depicting study inclusion algorithm. Between 1992 and 2020, 86 patients with DAVFs were treated with GKRS at our department. Patients (53/86, 62%) treated with a combined treatment of GKRS and other treatment methods were excluded. The remaining 33/86 (38%) patients were divided according to DAVF localization into NCCF (16/33, 48%) and CCF (17/33, 52%) groups. Fifteen of 17 patients (88%) within the CCF group and 15/16 (94%) patients within the NCCF group had at least one clinical follow-up after the first GKRS and were included in the outcome evaluation. Radiologic follow-up was available in 14/16 (88%) NCCF patients and in 14/17 (82%) CCF patients. CCF; carotid–cavernous fistula; DAVFs, dural arteriovenous fistulas; GKRS, Gamma Knife radiosurgery; NCCF, non-CCF.

Table 1 Patient characteristics and Gamma Knife radiosurgery parameters for the first GKRS

	Total (n = 33)	CCF (n = 17)	NCCF (n = 16)	p value
Age (y)
Median (range)	65 (29–82)	66 (54–82)	61 (29–80)	0.090
Sex (F:M)	21:12	13:4	8:8	0.114
Risk factors, n (%)
Trauma	4 (12%)	2 (11%)	2 (13%)	0.622
CVST	3 (9%)	–	3 (19%)
Pregnancy	1 (3%)	–	1 (6%)
Factor V Leiden mutation	1 (3%)	1 (6%)	–
CVST + factor V Leiden mutation	1 (3%)	1 (6%)	–
Infection	1 (3%)	1 (6%)	–
None	22 (67%)	12 (71%)	10 (62%)
KPS at diagnosis (%)
Median (range)	90 (40–90)	90 (90–90)	90 (40–90)	0.006
mRS at diagnosis
Median (range)	1 (1–4)	1 (1–1)	1 (1–4)	0.006
Ophthalmic symptoms[a]	20 (61%)	17 (100%)	3 (19%)	<0.001
Cranial nerve VI palsy	12 (36%)	11 (65%)	1 (6%)
Chemosis	11 (33%)	11 (65%)	–
Exophthalmos	10 (30%)	10 (59%)	–
Retrobulbar pain	7 (21%)	6 (35%)	1 (6%)
Conjunctival injection	7 (21%)	7 (41%)	–
Visual acuity	6 (18%)	4 (24%)	2 (13%)
Cranial nerve III/IV palsy	3 (9%)	3 (18%)	–
Nonophthalmic symptoms[a]	18 (55%)	2 (12%)	16 (100%)	<0.001
Headache	4 (12%)	–	4 (25%)
Pulsatile tinnitus	7 (21%)	2 (12%)	5 (31%)
Other	11 (33%)	–	11 (69%)
Hemorrhage at diagnosis	4 (12%)	–	4 (25%)	0.044
GKRS parameters
Treatment volume (cm3)
Median (range)	0.9 (0.1–21.7)	0.6 (0.1–2.4)	4.4 (0.5–21.7)	<0.001
Isodose line (%)
Median (range)	50 (40–75)	50 (50–75)	50 (40–55)	0.011
Prescription dose (Gy)
Median (range)	18 (8–23)	18 (8–20)	20 (17–23)	0.002
Central dose (Gy)
Median (range)	36 (16–50)	34 (16–40)	42 (34–50)	<0.001

Abbreviations: CCF, carotid–cavernous fistula; CVST, cerebral venous sinus thrombosis; DAVF, dural arteriovenous fistula; GKRS, Gamma Knife radiosurgery; KPS, Karnofsky's Performance Status Scale; mRS, modified Rankin scale; NCCF, non-CCF.

Note: [Table 1] demonstrates an overview of the patient characteristics and GKRS parameters for the first GKRS. According to their location, DAVFs were divided into CCFs and NCCFs, and the differences between those groups were calculated.

^a If a patient presented with more than one clinical symptom at time of diagnosis, multiple symptom counts per patient were conducted. Thus, counts do not add up to 100%.

The study complied with the Declaration of Helsinki and was approved by the local ethics review committee (EK1176/2020).

Classification of DAVFs

NCCFs (16/33 [48%]) were classified according to the Cognard[14] and Borden[15] classifications and CCFs (17/33 [52%]) were classified according to the Barrow[16] classification ([Table 2]). The Cognard and Borden classifications are based on the venous outflow architecture of the fistula and the presence or absence of CVD. The Borden classification describes three subtypes of DAVFs.[15] The Cognard classification further describes the direction of venous outflow and venous ectasia and distinguishes between five subtypes.[14]

Table 2 DAVF characteristics

n = 33	No. of patients (%)
Localization
Cavernous sinus	17 (52%)
Transverse/sigmoid sinus	10 (30%)
Tentorial	4 (12%)
Ethmoidal	2 (6%)
Side
Left	21 (64%)
Right	12 (36%)
Multifocal DAVFs
Yes	4 (12%)
No	29 (88%)
Cortical venous drainage
CVD	9 (27%)
No CVD	24 (73%)
Venous ectasia
Yes	3 (9%)
No	30 (91%)
Borden classification (NCCF, n  = 16)
I	7 (44%)
II	–
III	9 (56%)
Cognard's classification (NCCF, n  = 16)
I	7 (44%)
II	–
III	6 (37%)
IV	3 (19%)
Barrow's classification (CCF, n  = 17)
A	–
B	1 (6%)
C	–
D	16 (94%)

Abbreviations: CCF, carotid–cavernous fistula; CVST, cerebral venous sinus thrombosis; DAVF, dural arteriovenous fistula; GKRS, Gamma Knife radiosurgery; KPS, Karnofsky's Performance Status Scale; NCCF, non-CCF.

Note: [Table 2] provides an overview of DAVF characteristics. The Borden and Cognard classifications were evaluated for NCCFs (n = 16) only. The Barrow classification was used to classify CCFs (n = 17).

The Barrow classification classifies CCFs based on their arterial supply. Type A CCFs consist of a direct shunt between the internal carotid artery (ICA) and the cavernous sinus, type B CCFs are supplied by the meningeal branches of the ICA, type C CCFs receive arterial supply from the meningeal branches of the external carotid artery (ECA), and type D CCFs are supplied by the meningeal branches of the ICA and ECA.[16]

Radiosurgical Technique

Radiosurgical procedures were performed as described earlier.[17] Patients were treated with Leksell Gamma Knife (Model B until 2011/Perfexion from 2012 onward, Elekta AB, Stockholm, Sweden) and GammaPlan (Elekta AB) was used as the planning software. For treatment planning, magnetic resonance imaging (MRI) with high-resolution T2-weighted scans and time-of-flight MR angiography, contrast-enhanced computed tomography (CT) angiography together with bone window, and DSA were performed under stereotactic conditions. According to patients' compatibility, MRI was performed on either a 1.5- or a 3-T scanner.

The treatment target comprised the DAVFs on MRI, CTA, and angiographic coregistration imaging. The median treatment volume was 0.9 cm³ (range: 0.1–21.7 cm³). Among NCCF patients, a statistically significantly larger treatment volume was necessary compared with CCF patients (p < 0.001; [Table 1]). According to their smaller volume, CCFs were treated within a higher isodose line than NCCFs (p = 0.011). The median central dose was 36 Gy (range: 16–50 Gy) and the median prescription dose was 18 Gy (range: 8–23 Gy) for all patients ([Table 1]). Overall, 5/33 (15%) patients underwent repeated GKRS. Two of 17 (12%) patients with an NCCF and two of 16 (13%) patients with a CCF underwent a second radiosurgical treatment. One of 16 (6%) patients with an NCCF was treated three times. The median time between the first and second GKRS treatment was 28 months (range: 6.6–53.5 months). The median time between the second and third treatment was 31 months.

Clinical Assessment and Outcome Evaluation

A dedicated clinical assessment was performed at diagnosis, on the day prior to radiosurgery, and at every follow-up visit according to the standard management protocol at our department. Neurologic symptoms were evaluated, compared with clinical symptoms at the time of diagnosis, and categorized as complete resolution of symptoms, partial resolution, or no change of symptoms. Additionally, the Karnofsky Performance Scale (KPS) and the modified Rankin scale (mRS) were used to evaluate the patients' status.

Neuroimaging follow-up was performed at the same time as clinical follow-up, every year until obliteration of the fistula and every 3 to 5 years after obliteration. Obliteration of the arteriovenous shunt was defined as an absence of flow voids on T1- and T2-weighted MRI.[18] [19] If a residual fistula was suspected on follow-up MRI 2 years after GKRS, angiography was planned under stereotactic conditions and if confirmed, another GKRS was performed.

Statistical Analysis

Categorical data were presented as counts and percentages, and continuous parameters as median and range. To compare patient groups, the chi-squared test and Mann–Whitney U test were performed as appropriate. The median time to obliteration and obliteration rates were calculated with Kaplan–Meier estimators and life tables. The log-rank test was used to evaluate group differences. A p-value less than 0.05 was considered to be statistically significant. IBM SPSS Statistics for Windows (version 26.0, IBM Corp., Armonk, New York, United States) was used.

Results

Patients' Characteristics

The most common symptoms at the time of diagnosis were ophthalmic symptoms (20/33, 61%; [Table 1]). Those included cranial nerve VI palsy in 12/33 (36%) patients, chemosis in 11/33 (33%) patients, exophthalmos in 10/33 (30%) patients, retrobulbar pain in 7/33 (21%) patients, conjunctival injection in 7/33 (21%) patients, visual acuity in 6/33 (18%) patients, and cranial nerve III/IV palsy in 3/33 (9%) patients. More than half (18/33, 55%) of the patients presented with nonophthalmic symptoms. The most common nonophthalmic symptoms were pulsatile tinnitus in 7/33 (21%) patients and headache in 4/33 (12%) patients. Other nonophthalmic symptoms noted in 11/33 (33%) patients included hemiparesis, aphasia, sensory disturbance, nausea, and vertigo. CCFs were more likely to present with ophthalmic symptoms than NCCFs (p < 0.001). Overall, 4/33 (12%) patients suffered from intracranial bleeding at the time of diagnosis; all of them were diagnosed with an NCCF (p = 0.026). In 11/33 (33%) patients, risk factors associated with the genesis of DAVFs, such as trauma, cerebral venous sinus thrombosis (CVST), pregnancy, infection, or factor V Leiden mutation, could be identified.

DAVF Characteristics

About half of the DAVFs (17/33, 52%) were located within the cavernous sinus, while 10/33 (30%) lesions were located in the region of the transverse/sigmoid sinus, 2/33 (6%) were ethmoidal, and 4/33 (12%) tentorial fistulas. In 4/33 (12%) patients, two DAVFs were diagnosed before intervention. In all four patients, the second lesion was treated with another treatment method or was managed conservatively. In 9/33 (27%) patients, CVD was present, and in 3/33 (9%) patients, the draining veins were ectatic.

According to the Borden classification, 7/16 (44%) NCCFs were classified as Borden type I DAVFs and 9/16 (56%) were classified as Borden type III DAVs. Similarly, 7/16 (44%) NCCFs were classified as Cognard type I, 6/16 (37%) as type III, and 3/16 (19%) as Cognard type IV. The Barrow classification was used to classify CCFs. One of 17 (6%) patients presented with a Barrow type B CCF and 16/17 (94%) patients were diagnosed with a Barrow type D CCF ([Table 2]).

Overall Outcome and Complications after Radiosurgery

Overall, 3/33 (9%) patients were lost to follow-up. Thus, the median follow-up time after the first intervention was 5.8 years (range: 1.0–17.1 years). The median follow-up period did not show any statistically significant differences between the NCCF (6.4 years) and CCF groups (4.2 years; p = 0.908). The total observation period was 192.6 years.

For all patients, a death register comparison was performed to evaluate the complete outcome. Overall, 7/33 (21%) patients died. While six patients died due to other causes, one patient died due to a cerebral hemorrhage from the DAVF. This resulted in a mortality rate of 3% (1/30 with follow-up available) and a yearly bleeding risk of 0.5%. However, this patient was diagnosed with two DAVFs; one was treated with SRS alone and the other lesion, which could not be treated due to its large extension (Borden type III/Cognard type IV) and localization, was followed up conservatively. Unfortunately, the untreated DAVF bled during the follow-up period and caused the patient's death. Aside from this one cerebral hemorrhage, no complications related to GKRS occurred during the follow-up period.

The neurologic outcome could be evaluated in 30/33 (91%) patients and showed highly favorable outcome for both groups. A complete resolution of symptoms was seen in two-thirds (20/30, 67%) patients, a partial resolution of symptoms was observed in 6/30 (20%) patients, and in 4/30 (13%) patients' symptoms did not improve after GKRS treatment. At the time of the last follow-up, the median KPS was 100% (range: 40–100%) and the median mRS was 0 (range: 0–4) for patients with CCF. For NCCF patients, the median KPS at the last follow-up was 90% (range: 40–100%) and the median mRS was 1 (range: 0–4).

Obliteration after Gamma Knife Radiosurgery

Radiologic follow-up imaging was available for 28/30 (93%) patients. Follow-up MRI was performed in 26/28 (93%) patients and 2/28 (7%) patients underwent follow-up angiography ([Fig. 2]). In the majority of our patients (20/28, 71%), a complete DAVF obliteration could be achieved. In the remaining patients, either a reduction in the size of flow (7/28, 25%) or no change in flow (1/28, 4%) was seen.

Fig. 2 A dural arteriovenous fistula (DAVF) located at the transverse sinus before and 1 year after Gamma Knife treatment. (A) A dural fistula fed by multiple small feeders arising from the occipital artery shunting directly into the transverse sinus. (B) The angiographic result without any residual feeders 1 year after radiosurgery.

Thus, total obliteration of NCCFs and CCFs was achieved in 8/14 (57%) and 12/14 (86%) patients, respectively (p = 0.209). No progression or proliferation of a DAVF after radiosurgery was observed. Overall, the median time between the first GKRS and DAVF obliteration was 35 months (95% confidence interval [CI]: 19.7–51.1).

Furthermore, obliteration rates after 2, 5, and 10 years were 53, 71, and 85%, respectively. The median time to obliteration was 17 months (14/28, 50%, 95% CI: 7.4–27.2) for CCFs and 37 months (14/28, 50%, 95% CI: 34.7–38.5) for NCCFs ([Fig. 3A]). Although a trend toward NCCFs taking longer to obliterate than CCFs could be demonstrated, the estimated time to obliteration did not show any differences between the two groups (p = 0.111). However, time to obliteration was significantly longer in DAVFs with CVD (8/28, 29%; 119 months, 95% CI: 0.0–240.9) than in DAVFs without CVD (20/28, 71%, 22.9 months, 95% CI: 6.7–39.2; p = 0.037; [Fig. 3B]). Additionally, a trend toward a shorter obliteration time in DAVFs classified as Borden type I (6/14, 43%, 28 months, 95% CI: 18.4–37.3; p = 0.057) compared with DAVFs classified as Borden type III (8/14, 57%, 119 months, 95% CI: 0.0–240.9) was demonstrated.

Fig. 3 Kaplan–Meier plot: time to obliteration. (A) The difference in the estimated time to obliteration between CCFs and NCCFs. The median time to obliteration was 17 months (14/28, 50%, 95% CI: 7.4–27.2) for CCFs and 37 months (14/28, 50%, 95% CI: 34.7–38.5) for NCCFs. Although a trend toward NCCFs taking longer to obliterate than CCFs could be demonstrated, no significant difference was found between the groups (p = 0.111). (B) The difference in time to obliteration between DAVFs with and without CVD. In patients with CVD (8/28, 29%, 119 months; 95% CI: 0.0–240.9), time to obliteration was significantly longer than in patients without CVD (20/28, 71%, 22.9 months; 95% CI: 6.7–39.2; p = 0.037). CI; confidence interval, CCF; carotid–cavernous fistula, CVD, cortical venous drainage; DAVFs, dural arteriovenous fistulas; GKRS, Gamma Knife radiosurgery; GKRS1, first Gamma Knife radiosurgery; NCCF, non-CCF.

Discussion

Cerebral DAVFs represent a heterogeneous group of vascular malformations. Symptoms and prognosis highly depend on the presence or absence of CVD. Studies have demonstrated that DAVFs with CVD commonly present with more severe symptoms and carry a significantly higher risk of future neurologic events.[5] [20] [21] A yearly bleeding risk between 6 and 19% for DAVFs with CVD has been reported.[3] [5] [20] [21] DAVFs without CVD, on the other hand, often present with mild symptoms and carry a low risk of bleeding (0.0–1%).[3] [4] As various as their clinical symptoms are, as various are the available therapeutic options for DAVFs. These include microsurgery, endovascular embolization, SRS, or combinations of these options.[1] [2] Endovascular embolization or microsurgery allows an immediate obliteration of the arteriovenous shunt, which results in the termination of the bleeding risk. Therefore, embolization and microsurgery have become the first-line management options for DAVFs.[22] However, despite continued advances, high complication rates have been reported.[23] [24] [25] In terms of treatment choice, the risks of the intervention should always be weighed against the expected clinical course of the DAVF. Since DAVFs without CVD are mostly benign and the risk of a cerebral hemorrhage is very low, some authors even suggest conservative treatment for asymptomatic, incidentally found lesions.[4] [26] It is noteworthy that DAVFs without CVD still carry a low risk (2%) of progression into a higher-grade fistula.[26] Therefore, those patients should undergo regular clinical and radiologic follow-up examinations. Furthermore, DAVFs without CVD can cause intolerable clinical symptoms and may require some form of treatment.[8]

Moreover, not all patients with high-grade DAVFs may be eligible for microsurgery or endovascular treatment. Due to its low complication and high obliteration rates, radiosurgery has become an important alternative management option for DAVFs. Several studies have demonstrated its efficacy and safety.[6] [8] [9] [10] Thus, SRS was shown to be an effective, alternative, adjuvant/salvage treatment method. However, there are only few studies, mostly with limited patient numbers, that evaluate radiosurgery as a sole and upfront treatment option.[12] [27] Our department looks back on three decades of experience in the radiosurgical treatment of cerebral vascular malformations.

For our patients with vascular pathologies, our interdisciplinary vascular board decides the most accurate treatment. Those decisions are reached in agreement of the hybrid endovascular–vascular neurosurgeon, radiosurgically trained neurosurgeon, and radiologist. SRS has often been recommended due to the patient's wish or for high-risk patients not eligible for endovascular or microsurgical treatment due to comorbidities or anatomical and technical difficulties.

In this retrospective series, we evaluated data obtained from 33 patients with a DAVF who were managed by GKRS as an upfront treatment. The complete obliteration rate was 71% and in another 25% of DAVFs a reduction in flow or size was seen on follow-up imaging. These results are in the upper range of previously published obliteration rates of 63 to 77%.[6] [8] [9] [10]

Overall, the median time to obliteration in our cohort was 35 months. Obliteration rates after 2, 5, and 10 years were 53, 71, and 85%, respectively. Although a trend toward a shorter time to obliteration in CCFs could be shown, differences between the NCCF and CCF groups were not statistically significant. However, the median time to obliteration was significantly longer in DAVFs with CVD. This may be explained by the more complex angioarchitecture of DAVFs with CVD compared with DAVFs simply shunting into a venous sinus. As a result, the exact definition of the treatment target is more challenging and may lead to a longer obliteration time. Moreover, in our series, more DAVF patients with CVD underwent repeated radiosurgical treatment compared with those without CVD. In addition to a high obliteration rate, clinical outcome was highly favorable in our series.[6] [8] [9] [10] [28] A complete resolution of symptoms was seen in two-thirds of all patients and a partial resolution of symptoms in 20% of patients. No improvement after GKRS was observed in only 13% of patients.

Due to their localization within the cavernous sinus, CCFs represent a distinct subtype of DAVFs. Clinical presentation varies from NCCFs and is mostly benign with a lower bleeding risk.[29] The increased venous flow within the cavernous sinus typically causes ophthalmic symptoms. These include chemosis, exophthalmos, ocular pain, visual acuity, and cranial nerve palsy.[6] As expected, our data showed that CCFs were more likely to present with ophthalmic symptoms. However, in some CCFs, symptoms can be severe, and therapy may prevent vision loss or intracranial hemorrhage.[11] In our series, none of the patients diagnosed with a CCF suffered from an intracranial hemorrhage. Several studies consider radiosurgery as an effective treatment method for low-flow CFFs (Barrow types B–D).[6] [9]

Pan et al reported a complete obliteration rate of 70% in one of the few existing larger series of 156 low-flow CCFs.[8] In our series, a rather high obliteration rate of 86% within the CCF group was achieved. Some studies suggest that CCFs may also auto-thrombose with time. However, this spontaneous resolution of CCFs occurs in only 1 to 13% of cases according to the published literature.[30] [31]

Overall, no complications related to GKRS were observed in our series. One patient had an intracranial hemorrhage after GKRS. However, this patient was diagnosed with two DAVFs. One lesion was treated with GKRS and the other one, which could not be treated due to its large extension (Borden type III/Cognard type IV) and localization, was followed conservatively. Unfortunately, the untreated DAVFs bled during the follow-up period and the patient died following this hemorrhage.

Limitations

Limitations of our study were its retrospective design and the rather small patient cohort. Still, our series represents a homogenous cohort of DAVF patients treated with radiosurgery as a stand-alone management option. Another limitation to our study was the radiologic follow-up. Our institution's follow-up regime after radiosurgery did not routinely include a follow-up angiography. However, in our opinion, DAVFs without CVD and especially CCFs may be sufficiently assessed by MRI and the resolution of clinical symptoms. If a residual fistula was suspected on follow-up MRI, angiography was planned under stereotactic conditions and if confirmed, another GKRS was performed.

Conclusions

In the past, GKRS was often used as an alternative to treat DAVFs that could not be obliterated with endovascular or surgical approaches. Our results demonstrate that GKRS is an effective and safe stand-alone treatment option for selected DAVF patients.

Conflict of Interest

None declared.

^* These authors contributed equally to this work.

• References

• 1 Gandhi D, Chen J, Pearl M, Huang J, Gemmete JJ, Kathuria S. Intracranial dural arteriovenous fistulas: classification, imaging findings, and treatment. AJNR Am J Neuroradiol 2012; 33 (06) 1007-1013
• 2 Miller TR, Gandhi D. Intracranial dural arteriovenous fistulae: clinical presentation and management strategies. Stroke 2015; 46 (07) 2017-2025
• 3 Gross BA, Du R. The natural history of cerebral dural arteriovenous fistulae. Neurosurgery 2012; 71 (03) 594-602 , discussion 602–603
• 4 Davies MA, Saleh J, Ter Brugge K, Willinsky R, Wallace MC. The natural history and management of intracranial dural arteriovenous fistulae. Part 1: benign lesions. Interv Neuroradiol 1997; 3 (04) 295-302
• 5 Davies MA, Ter Brugge K, Willinsky R, Wallace MC. The natural history and management of intracranial dural arteriovenous fistulae. Part 2: aggressive lesions. Interv Neuroradiol 1997; 3 (04) 303-311
• 6 Hung YC, Mohammed N, Kearns KN. et al. Stereotactic radiosurgery for cavernous sinus versus noncavernous sinus dural arteriovenous fistulas: outcomes and outcome predictors. Neurosurgery 2020; 86 (05) 676-684
• 7 Dmytriw AA, Schwartz ML, Cusimano MD. et al. Gamma Knife radiosurgery for the treatment of intracranial dural arteriovenous fistulas. Interv Neuroradiol 2017; 23 (02) 211-220
• 8 Pan DH, Lee CC, Wu HM, Chung WY, Yang HC, Lin CJ. Gamma Knife radiosurgery for the management of intracranial dural arteriovenous fistulas. Acta Neurochir Suppl (Wien) 2013; 116: 113-119
• 9 Park SH, Park KS, Kang DH, Hwang JH, Hwang SK. Stereotactic radiosurgery for intracranial dural arteriovenous fistulas: its clinical and angiographic perspectives. Acta Neurochir (Wien) 2017; 159 (06) 1093-1103
• 10 Starke RM, McCarthy DJ, Chen CJ. et al. Evaluation of stereotactic radiosurgery for cerebral dural arteriovenous fistulas in a multicenter international consortium. J Neurosurg 2019; 132 (01) 114-121
• 11 Chen CJ, Lee CC, Ding D. et al. Stereotactic radiosurgery for intracranial dural arteriovenous fistulas: a systematic review. J Neurosurg 2015; 122 (02) 353-362
• 12 Gross BA, Ropper AE, Popp AJ, Du R. Stereotactic radiosurgery for cerebral dural arteriovenous fistulas. Neurosurg Focus 2012; 32 (05) E18
• 13 Söderman M, Edner G, Ericson K. et al. Gamma knife surgery for dural arteriovenous shunts: 25 years of experience. J Neurosurg 2006; 104 (06) 867-875
• 14 Cognard C, Gobin YP, Pierot L. et al. Cerebral dural arteriovenous fistulas: clinical and angiographic correlation with a revised classification of venous drainage. Radiology 1995; 194 (03) 671-680
• 15 Borden JA, Wu JK, Shucart WA. A proposed classification for spinal and cranial dural arteriovenous fistulous malformations and implications for treatment. J Neurosurg 1995; 82 (02) 166-179
• 16 Barrow DL, Spector RH, Braun IF, Landman JA, Tindall SC, Tindall GT. Classification and treatment of spontaneous carotid-cavernous sinus fistulas. J Neurosurg 1985; 62 (02) 248-256
• 17 Hirschmann D, Goebl P, Witte FH. et al. Evaluation of the radiosurgical treatment of cerebral arteriovenous malformations: a retrospective single-center analysis of three decades. J Neurointerv Surg 2020; 12 (04) 401-406
• 18 Kwon BJ, Han MH, Kang HS, Chang KH. MR imaging findings of intracranial dural arteriovenous fistulas: relations with venous drainage patterns. AJNR Am J Neuroradiol 2005; 26 (10) 2500-2507
• 19 Meckel S, Maier M, Ruiz DS. et al. MR angiography of dural arteriovenous fistulas: diagnosis and follow-up after treatment using a time-resolved 3D contrast-enhanced technique. AJNR Am J Neuroradiol 2007; 28 (05) 877-884
• 20 Söderman M, Pavic L, Edner G, Holmin S, Andersson T. Natural history of dural arteriovenous shunts. Stroke 2008; 39 (06) 1735-1739
• 21 van Dijk JM, terBrugge KG, Willinsky RA, Wallace MC. Clinical course of cranial dural arteriovenous fistulas with long-term persistent cortical venous reflux. Stroke 2002; 33 (05) 1233-1236
• 22 Ertl L, Brückmann H, Kunz M, Crispin A, Fesl G. Endovascular therapy of low- and intermediate-grade intracranial lateral dural arteriovenous fistulas: a detailed analysis of primary success rates, complication rates, and long-term follow-up of different technical approaches. J Neurosurg 2017; 126 (02) 360-367
• 23 Gross BA, Albuquerque FC, Moon K, McDougall CG. Evolution of treatment and a detailed analysis of occlusion, recurrence, and clinical outcomes in an endovascular library of 260 dural arteriovenous fistulas. J Neurosurg 2017; 126 (06) 1884-1893
• 24 Hiramatsu M, Sugiu K, Hishikawa T. et al. Results of 1940 embolizations for dural arteriovenous fistulas: Japanese Registry of Neuroendovascular Therapy (JR-NET3). J Neurosurg 2019; 133 (01) 166-173
• 25 Kim DJ, Willinsky RA, Krings T, Agid R, Terbrugge K. Intracranial dural arteriovenous shunts: transarterial glue embolization–experience in 115 consecutive patients. Radiology 2011; 258 (02) 554-561
• 26 Satomi J, van Dijk JM, Terbrugge KG, Willinsky RA, Wallace MC. Benign cranial dural arteriovenous fistulas: outcome of conservative management based on the natural history of the lesion. J Neurosurg 2002; 97 (04) 767-770
• 27 O'Leary S, Hodgson TJ, Coley SC, Kemeny AA, Radatz MW. Intracranial dural arteriovenous malformations: results of stereotactic radiosurgery in 17 patients. Clin Oncol (R Coll Radiol) 2002; 14 (02) 97-102
• 28 Park KS, Kang DH, Park SH, Kim YS. The efficacy of gamma knife radiosurgery alone as a primary treatment for intracranial dural arteriovenous fistulas. Acta Neurochir (Wien) 2016; 158 (04) 821-828
• 29 Ellis JA, Goldstein H, Connolly Jr ES, Meyers PM. Carotid-cavernous fistulas. Neurosurg Focus 2012; 32 (05) E9
• 30 Iampreechakul P, Tirakotai W, Tanpun A, Wattanasen Y, Lertbusayanukul P, Siriwimonmas S. Spontaneous resolution of direct carotid-cavernous fistulas: case series and literature review. Interv Neuroradiol 2019; 25 (01) 71-89
• 31 Kim DJ, terBrugge K, Krings T, Willinsky R, Wallace C. Spontaneous angiographic conversion of intracranial dural arteriovenous shunt: long-term follow-up in nontreated patients. Stroke 2010; 41 (07) 1489-1494

Publication History

Received: 16 February 2023

Accepted: 20 December 2023

Accepted Manuscript online:
27 December 2023

Article published online:
04 March 2024