Outcomes of gamma knife treated large symptomatic arteriovenous malformations according to guidelines of Taiwan neurosurgical consensus

Abstract

Background: The treatment of large arteriovenous malformations (AVMs), in particular those unruptured, remains a topic of debate. Stereotactic radiosurgery has favorable outcomes for small to medium-sized AVMs. However, for large AVMs, the goal is to maximize obliteration rates and at the same time, to minimize radiation-induced complications. This study assessed outcomes of large symptomatic AVMs treated with Gamma Knife radiosurgery (GKRS) focusing on cases presenting with rupture or seizures. The study followed the guidelines of Taiwan Neurosurgical Consensus, a government-funded committee under the Central Bureau of Health Insurance that determines whether radiosurgery is an appropriate treatment.

Materials and methods: This retrospective study included 75 cases of large AVMs (> 10 cc) treated with GKRS during the period from June 2003 to January 2020. Inclusion criteria were as follows: a history of intracerebral hemorrhage (ICH) or seizures, no prior embolization, and periodic MRI examinations with clinical assessments post-GKRS. Treatment procedures were adapted based on the Taiwan Neurosurgical Consensus guidelines.

Results: The average patient age was 36.4 ± 16.1 years, with a median follow-up duration of 104 (range 82–150) months. Forty-six patients (61.3%) underwent single-stage treatment, while 29 patients (38.7%) received two-stage treatment. The mean AVM volume was 20.5 ± 11.7 cc, with an average peripheral radiation dose of 17.7 ± 1.2 Gy. Among the 32 cases with AVM volumes between 10 and 15 cc, 25 (78.1%) achieved total obliteration. For the 17 cases with volumes between 15 and 20 cc, 7 (41.2%) achieved total obliteration, while 8 out of 26 (30.8%) cases with volumes > 20 cc achieved total obliteration. Severe brain edema developed in 16 patients (21.3%) after an average follow-up of 105.4 ± 56.2 months, but 11 patients (14.6%) experienced symptoms. Only one patient (1.3%) suffered neurological disability. Seizure control in Engel classification I was achieved in 21 of 42 patients (50%). Eight patients (10.6%) experienced new hemorrhages, with 4 (12.1%) occurring in those with a prior history of hemorrhage (annual bleeding rate: 1.2%) and 4 (9.5%) in those patients with a history of seizures (annual bleeding rate: 1.1%). Univariate analyses showed that total obliteration was significantly associated with smaller nidus volumes (< 15 cc), single-stage radiosurgery, Radiosurgery-Based Grading Scale, first-stage volume, maximum dose, 12 Gy volume, and nidus coverage percentages at 16 Gy and 18 Gy. Multivariate analyses revealed that post-GKRS symptoms and severe brain edema were significantly correlated with the following: Virginia Radiosurgery AVM Score, Charlson Comorbidity Index, and mean radiation dose.

Conclusion: The obliteration rate of large AVMs is strongly correlated with their size. This approach appears to achieve the goals of obliteration and minimizing the risks of radiation-induced complications and hemorrhage. Further investigation is needed for adjuvant treatments in residual or refractory cases after GKRS.

Keywords: Large arteriovenous malformation; Stereotactic radiosurgery; Volume-stage.

Fig. 1

Gamma knife treatment algorithm. AVMs > 20 cc are treated in two stages; 15–20 cc in eloquent locations are treated in two stages; 15–20 cc in non-eloquent locations in a single stage; and AVMs < 15 cc in a single stage

Fig. 2

A 35-year-old female with the focal seizure and diagnosed as a huge AVM with volume of 31 cc under the guidance of AVM volume > 20 cc treated in two stage gamma knife treatment. The post GKRS imaging showed total obliteration without any neurological deficits. (A) Anteroposterior view of cerebral angiography following right common carotid contrast injection showed a large AVM located in the occipital lobe. The AVM was supplied by branches from the right external carotid artery and internal cerebral artery, with venous drainage into the straight, transverse, and superior sagittal sinuses. (B) Lateral view of cerebral angiography after right common carotid contrast injection demonstrated a large AVM in the occipital lobe, with arterial supply from the right external carotid artery and internal cerebral artery. Venous drainage was observed into the straight, transverse, and superior sagittal sinuses. (C) Photograph illustrating dose planning for the first stage of Gamma Knife radiosurgery (GKRS). Treatment parameters are shown on the right side of the image. The yellow line indicates the 50% isodose line. (D) Photograph showing dose planning for the second stage of GKRS. Treatment parameters are displayed on the right side of the image. The yellow line represents the 50% isodose line. (E) No definite nidus was observed in the lateral view of cerebral angiography 4 years after GKRS. (F) No definite nidus was noted in the anteroposterior view of cerebral angiography 4 years after GKRS. (G) A small residual abnormality in the right occipital region was detected on FLAIR sequence imaging 14 years and 2 months after GKRS. (H) Small residual abnormalities in the right occipital region were observed on T2-weighted MRI 14 years and 2 months after GKRS. (I) Mild residual contrast enhancement was seen in the right occipital region on T1-weighted MRI with contrast. (J) No definite nidus was detected on MRA 14 years and 2 months after GKRS

Fig. 3

A 28-year-old male presented with focal seizure and diagnosed as a huge AVM with volume of 15.2 cc. The treatment was separated into 2 stages under the guidance of AVM volume 15–20 cc in eloquent locations treated in two stages treatment. The post GKRS imaging showed the tiny residual nidus and cyst formation mandatory of s-p shunt with clinical symptoms of left leg clumsy movement. (A) Anteroposterior view of cerebral angiography showed an AVM located in the right parietal lobe, with arterial supply from the right anterior and middle cerebral arteries and venous drainage into the deep venous system and the superior sagittal sinus. (B) Lateral view of cerebral angiography showed an AVM in the right parietal lobe, supplied by the right anterior and middle cerebral arteries, with drainage into the deep venous system and the superior sagittal sinus. (C) Photograph showing dose planning for the first stage of Gamma Knife radiosurgery. Treatment parameters are displayed on the right side of the image. The yellow line represents the 50% isodose line. (D) Photograph showing dose planning for the second stage of Gamma Knife radiosurgery. Treatment parameters are shown on the right side of the image. The yellow line indicates the 50% isodose line. (E) A small cyst was detected at the targeted lesion on T2-weighted MRI, 5 years and 6 months after the first GKRS. However, no definite symptoms were noted at that time. (F) A small nidus, indicated by the arrowhead, was identified in the lateral view of cerebral angiography 5 years and 6 months after the first GKRS. (G) A small nidus, indicated by the arrowhead, was detected in the anteroposterior view of cerebral angiography 5 years and 6 months after the first GKRS. (H) A large cyst in the right parietal region was observed on T2-weighted MRI 5 years and 9 months after GKRS. Neurological examination revealed clumsiness in the left leg. (I) A large cyst in the right parietal region was noted on T2-weighted MRI 6 years and 3 months after GKRS. Neurological examination showed continued clumsiness in the left leg without further deterioration (J) A large cyst in the right parietal region was detected on T2-weighted MRI 7 years and 3 months after Gamma Knife treatment. Neurological examination revealed worsening clumsiness in the left leg. A cyst-peritoneal shunt was scheduled at this time. (K) A small residual cyst was detected on T2-weighted MRI 10 months after the cyst-peritoneal shunt. Neurological examination showed improvement in left leg movement

Fig. 4

A 23-year-old male presented with focal seizure and diagnosed as huge AVM with volume of 17.4 cc. The treatment was conducted in a single stage according to the guidance of AM volume 15–20 cc in no-eloquent locations treated in single stage. The patient suffered severe headache after Gamma Knife radiosurgery due to the hemorrhage and received the craniotomy for hematoma removal. The patient gained full recovery without neurological deficits with Engle I of seizure control. (A) A large AVM located in the left parietal lobe was detected in the anteroposterior view of cerebral angiography, with arterial supply from the anterior cerebral artery and venous drainage into the superior sagittal sinus. (B) A large AVM in the left parietal lobe was observed in the lateral view of cerebral angiography, with feeding arteries from the anterior cerebral artery and drainage into the superior sagittal sinus. (C) Photograph showing dose planning for Gamma Knife radiosurgery (GKRS). Treatment parameters are displayed on the right side of the image. The yellow line indicates the 50% isodose line. (D) A small abnormal signal in the left parietal region was observed on FLAIR imaging one year and 10 months after GKRS. (E) A small abnormal signal was detected in the left parietal region on T2-weighted MRI one year and 10 months after GKRS (F) Mild abnormal enhancement resembling a vascular structure was seen in the left parietal region one year and 10 months after GKRS. (G) A small nidus with early draining veins in the left parietal region was detected on MRA one year and 10 months after GKRS. (H) The patient experienced a severe headache and was diagnosed with intracerebral hemorrhage in the left parietal region by CT scan 2 years and 3 months after GKRS. (I) A small abnormal vascular structure in the left parietal region was identified by CTA 2 years and 3 months after GKRS. (J) A small residual intracerebral hemorrhage at the surgical bed was seen on brain CT the day after surgery. Neurological examination revealed no definite deficits. (K) No definite nidus was observed in the lateral view of MRA 3 years and 10 months after GKRS. (L) No definite nidus was seen in the anteroposterior view of MRA 3 years and 10 months after GKRS

Fig. 5

A 39-year-old male presented with right frontal hemorrhage and diagnosed as a huge AVM with volume of 14.7 cc. The treatment was conducted in single stage treatment according to the guidance of AVM volume 10–15 cc procedure in single stage. The post GKRS angiography showed the total obliteration after gamma knife and MRI showed a small cyst over the previous hemorrhage cavity. (A) A large right frontal AVM was detected in the anteroposterior view of cerebral angiography, with arterial supply from the anterior cerebral artery and venous drainage into the superior sagittal sinus (B) A large right frontal AVM was observed in the lateral view of cerebral angiography, with feeding arteries from the anterior cerebral artery and drainage into the superior sagittal sinus. (C) Photograph showing dose planning for Gamma Knife radiosurgery (GKRS). Treatment parameters are displayed on the right side of the image. The yellow line indicates the 50% isodose line. (D) A small abnormal signal surrounding the cyst cavity in the right frontal region was observed on T2-weighted MRI 3 years and 1 month after GKRS. (E) A small abnormal signal surrounding the cyst cavity in the right frontal region was seen on FLAIR imaging 3 years and 1 month after GKRS. (F) Mild abnormal enhancement resembling a vascular structure was noted in the right frontal region on contrast-enhanced T1-weighted MRI 3 years and 1 month after GKRS. (G) A small nidus with early venous drainage in the right frontal region was detected on MRA 3 years and 1 month after GKRS. (H) Cerebral angiography in the anteroposterior view showed no definite nidus 3 years and 3 months after GKRS. (I) Cerebral angiography in the lateral view revealed no definite nidus 3 years and 3 months after GKRS. (J) A small cyst without progression was observed in the right frontal region on T2-weighted MRI 10 years and 10 months after GKRS. (K) A small cyst without progression was also observed on FLAIR imaging 10 years and 10 months after GKRS. (L) No definite abnormal enhancement surrounding the cavity was seen 10 years and 10 months after GKRS. Neurological examination revealed no definite neurological deficits

Fig. 6

A 24-year-old female with the generalized tonic seizure and diagnosed as a huge AVM with volume of 72.4 cc under the guidance of AVM volume > 20 cc treated in two stage gamma knife treatment. The post GKRS imaging showed the no definite obliteration with Engle II in seizure control. (A) A large AVM located in the left parietal-occipital region was observed in the anteroposterior view of cerebral angiography. It was supplied by feeders from the left anterior, middle, and posterior cerebral arteries, with venous drainage into the internal cerebral veins and the superior sagittal sinus (B) A large AVM in the left parietal-occipital region was visualized in the lateral view of cerebral angiography, with arterial supply from the left anterior, middle, and posterior cerebral arteries, and venous drainage into the internal cerebral veins and the superior sagittal sinus. (C) Photograph showing dose planning for the first stage of Gamma Knife radiosurgery. Treatment parameters are displayed on the right side of the image. The yellow line indicates the 50% isodose line. (D) Photograph showing dose planning for the second stage of Gamma Knife radiosurgery. Treatment parameters are illustrated on the right side of the image. The yellow line represents the 50% isodose line. (E) No definite abnormal signal surrounding the nidus was detected on FLAIR MRI 5 years and 2 months after GKRS. (F) Persistent vascular structure–like enhancement in the left parietal region was observed on contrast-enhanced T1-weighted MRI 5 years and 2 months after GKRS. (G) No definite abnormal signal surrounding the nidus was seen on FLAIR MRI 7 years and 3 months after GKRS. (H) Persistent vascular structure–like enhancement in the left parietal region was detected on contrast-enhanced T1-weighted MRI 7 years and 3 months after GKRS. (I) A persistent large AVM was identified in the anteroposterior view of cerebral angiography 10 years and 5 months after GKRS. (J) A persistent large AVM was again visualized in the anteroposterior view of cerebral angiography 10 years and 5 months after GKRS

Fig. 7

A 12-year-old female presented with generalized seizure and diagnosed as huge AVM with volume of 15.8 cc. The treatment was separated into 2 stages under the guidance of AVM volume 15–20 cc in eloquent locations treated in two stages treatment. The imaging showed the total obliteration but severe brain edema mandatory of craniotomy to lessen the symptom. (A) A large AVM was detected in the right occipital region on the anteroposterior view of cerebral angiography, with arterial supply from the posterior cerebral artery and venous drainage into the superior sagittal sinus. (B) A large AVM was observed in the right occipital region on the lateral view of cerebral angiography, with feeders from the posterior cerebral artery and drainage into the superior sagittal sinus. (C) Photograph showing dose planning for the first stage of Gamma Knife radiosurgery. Treatment parameters are illustrated on the right side of the image. The yellow line indicates the 50% isodose line. (D) Photograph showing dose planning for the second stage of Gamma Knife radiosurgery. Treatment parameters are illustrated on the right side of the image. The yellow line indicates the 50% isodose line (E) An abnormal signal surrounding the nidus was detected in the right occipital region on FLAIR MRI one year and 6 months after GKRS. (F) A persistent abnormal signal, without increase in size, surrounding the nidus was seen on FLAIR MRI 9 years and 5 months after GKRS. (G) An increase in the abnormal signal surrounding the nidus was noted on FLAIR MRI 10 years and 10 months after GKRS. (H) A vascular structure–like enhancement in the right occipital region was detected on contrast-enhanced T1-weighted MRI 10 years and 10 months after GKRS. (I) No definite nidus was observed in the anteroposterior view of cerebral angiography 10 years and 10 months after GKRS. (J) No definite nidus was seen in the lateral view of cerebral angiography 10 years and 10 months after GKRS. The patient experienced intractable headache and underwent craniotomy for removal of the enhancing lesion. (K) No evidence of intracerebral hemorrhage (ICH) was found on brain CT one day after the operation. The patient experienced significant improvement in headache symptoms

Fig. 8

Plot of incidence of hemorrhage related to time in huge AVM treated with GKRS. Y axis: rate of free hemorrhage (ICH/IVH) after GK X: Time presented in months