We admitted 97 consecutive patients with diagnosis of brain AVM by angiography between December 2007 and January 2018 in a third level medical center at Mexico City. We excluded 22 patients because incomplete medical records or because they abandon medical counseling before medical treatment was started. The final sample analyzed was 75 patients. Total sample was stratified by Spetzler Martin Grading Scale in low grade (I-II) 38.6% (n = 29), transitional grade (III) 26.6% (n = 20), and giant AVM (IV-V) 34.6% (n = 26). According to performed treatment, low grade AVMs were treated with bimodal therapy in 72.4% (n = 21), transitional AVMs in 60% (n = 12), and giant AVMs in 73% (n = 19). Figure 1.
Outcome
We compared the MRS score before and 24 months after bimodal therapy, in low grade brain AVMs the mean MRS decreased from 2.5 ± 1.2 to 0.7 ± 0.9, in transitional grade brain AVMs the mean MRS decreased from 2.3 ± 1.7 to 0.9 ± 1, and in giant AVMs 2.4 ± 1.1 to. 2.3 ± 2.1. The difference of mean delta MRS score before and after bimodal therapy decreased as the grade of brain AVM increased (from 1.8 ± 0.3 to 1.4 ± 0.7 and 0.1 ± 1, for low, transitional, and large-giant AVMs respectively). (Table 2).
Table 2
Comparison of Modified Rankin Scale before and after endovascular embolization followed by surgical resection in 52 patients with brain AVM.
| Low grade (I-II) n = 21 | Transitional (III) n = 12 | Giant (IV-V) n = 19 |
| Basal | After 24 months | Basal | After 24 months | Basal | After 24 months |
Score, n (%) | | | | | | |
No symptoms | 0 (0) | 12 (57) | 0 (0) | 6 (50) | 0 (0) | 5 (26) |
No significant disability | 4 (19) | 3 (14) | 6 (50) | 2 (17) | 4 (21) | 2 (11) |
Slight disability | 9 (43) | 5 (24) | 2 (17) | 3 (25) | 7 (37) | 5 (26) |
Moderate disability | 2 (10) | 1 (5) | 1 (8) | 1 (8) | 5 (26) | 3 (16) |
Moderately severe disability | 4 (19) | 0 (0) | 0 (0) | 0 (0) | 2 (11) | 0 (0) |
Severe disability | 2 (10) | 0 (0) | 3 (25) | 0 (0) | 1 (5) | 0 (0) |
Dead | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 4 (21) |
mMRS, n ± SD | 2.5±1.2 | 0.7±0.9 | 2.3±1.7 | 0.9±1 | 2.4±1.1 | 2.3±2.1 |
∆ mMRS, n ± SD | 1.8 ± 0.3 | 1.4 ± 0.7 | 0.1 ± 1 |
AVM: arteriovenous malformation; mMRS: mean Modified Rankin Scale score; ∆ mMRS: mean delta Modified Rankin Scale; SD: standard deviation; Bold: difference between basal vs. after 24 months with p≤0.05. |
We described the procedural characteristics in 52 patients with brain AVM treated with bimodal therapy (Table 3). We observed an increasing number of embolization procedures in brain AVMs grades IV and V (1.2 ± 0.4 for grade I-II, 1.2 ± 0.4 for grade III, 2.4 ± 0.7 for grade IV, and 3.5 ± 0.7 for grade V), as well as an increase in the number of Onyx vials (2 ± 1.4 for grade I-II, 2.5 ± 1 for grade III, 9 ± 2.2 for grade IV, and 21.6 ± 6.5 for grade V). Similarly, when we analyzed the characteristics of only the first endovascular embolization procedure, we founded a progressive increase of Onyx vials (1.5 ± 0.7 for grades I-II, 2 ± 1 for grade III, 4.6 ± 1.5 for grade IV, and 5.2 ± 2.1 for grade V) there was no significant difference duration of first embolization procedure. In addition, the mean percentage of residual brain AVM after first embolization was 5 [0–20] for grades I-II, 10 [7.5–50] for grade III, 32.5 [27.5–50] for grade IV, and 60 [50–60] for grade V. To identify an association between procedural characteristics of bimodal therapy with clinical outcomes, we analyze the frequency of AVM ruptures according to the number of endovascular embolization procedures in 31 patients with transitional and giant brain AVMs, 6.4% (n = 2) of ruptures occurred in patients treated in one endovascular procedure, 6.4% (n = 2) with two sessions, 3.2% (n = 1) with three sessions, and none rupture with four and five embolization procedures.
Table 3
Procedural characteristics in 52 patients with brain AVM treated with endovascular embolization followed by surgical resection.
Variable | Total sample n = 52 | Grade I-II n = 21 | Grade III n = 12 | Grade IV n = 12 | Grade V n = 7 |
Characteristics of endovascular embolization | | | | | |
Number of embolization procedures, mean ± SD | 1.8±1 | 1.2±0.4 | 1.2±0.4 | 2.4±0.7 | 3.5±0.7 |
Vials of Onyx used, mean ± SD | 6.4±7.2 | 2±1.4 | 2.5±1 | 9±2.2 | 21.6±6.5 |
Characteristics of first endovascular embolization | | | | | |
Vials of Onyx used at first embolization procedure, mean ± SD | 2.8±1.9 | 1.5±0.7 | 2±1 | 4.6±1.5 | 5.2±2.1 |
Duration of first embolization procedure, min ± SD | 149±38 | 133±33 | 144±32 | 177±46 | 156±23 |
Percentage of residual brain AVM after first embolization procedure, (% ± SD) | 25 [3.7–50] | 5 [0–20] | 10 [7.5–50] | 32.5 [27.5–50] | 60 [50–60] |
Definitive surgical resection after embolization | | | | | |
Scheduled surgery, n (%) | 47 (90) | 21 (100) | 10 (83) | 12 (100) | 7 (100) |
bAVM rupture requiring emergency decompression, n (%) | 5 (10) | 0 (0) | 2 (17) | 2(17) | 1 (14) |
AVM: arteriovenous malformation; SD: standard deviation. |
Of total patient sample in this study, only seven (9.3%) patients from the total sample developed trans-procedural (during the embolization) rupture of brain AVM, none for low grade, three for transitional grade, and four for giant grade. After 24 months follow up 20 (26.6%) patients remains with neurological deficit, the majority, not added to that already presented by patients prior to bimodal treatment; and four (5.3%) died, three were grade IV Spetzler-Martin grade, the first was a left frontoparietal which was embolized in two sessions, after the second embolization with 95% obliterated presents rupture causing parenchymal hemorrhage, an emergency decompressive craniectomy was performed which end unsuccessfully; The second was right thalamic lesion which was embolized in two sessions without complications, during the planned surgical resection with 90% obliterated, died because rupture and uncontrollable bleeding; The third was a right parieto occipital, which was embolized in three session with 95% obliterated, and surgical resected successfully, unfortunately, during hospital staying acquired nosocomial pneumonia and died. The four patient was a right insular and thalamic grade V, which was embolized in three sessions with 95% obliterated, during the planned surgery presents rostro caudal deterioration, an emergency CT was performed showing no evidence of intracranial bleeding. (Table 4).
Table 4
Clinical outcomes in one year follow up of 75 patients with brain AVM categorized by Spetzler Martin Scale and performed treatment.
Variable | Total sample n = 75 | Grade I-II n = 29 | Grade III n = 20 | Grade IV-V n = 26 |
| | Bimodal n = 21 | Other n = 8 | Bimodal n = 12 | Other n = 8 | Bimodal n = 19 | Other n = 7 |
Trans-procedural complication, n (%) | | | | | | | |
AVM rupture | 7 (9.3) | 0 (0) | 0 (0) | 2 (17) | 1 (13) | 3 (16) | 1 (14) |
Complication at 12 months follow up, n (%) | | | | | | | |
Neurological deficit | 20(26.6) | 5 (24) | 2 (25) | 4 (33) | 2 (25) | 5 (26) | 2 (29) |
Death from any cause | 4 (5.3) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 4 (21) | 0 (0) |
AVM: arteriovenous malformation; Neurological sequel: neurological deficit, seizures, hemiparesis, aphasia. |
Illustrative Cases
Case 1
A 38-year-old woman with a right parieto-occipital arteriovenous malformation grade V that involved the basal ganglia, with a mean size of 68.7 millimeters and mixed venous drainage, was admitted in the neurological surgery department and treated by bimodal modality was proposed, the patient had experienced seizures and had history of intracranial hemorrhage in the childhood. Weakness involving the left side of her body was minimal. the lesion was embolized in 3 sessions with a conservative progression between each session, until a 5% residual lesion in the last procedure and scheduled for definitive surgical resection. the interval between each embolization was 6 weeks. A total of 34 onyx vials were used and there was no evidence of final neurologic deficit immediately either in the follow up. Figure 2.
Case 2
A 11-years-old female with history of seizures and focal neurological deficit, was diagnosticated a right temporo-parietal-occipital arteriovenous malformation with bi-lobulated aspect, involving eloquent areas, it had a mean size of 60.5 millimeters, and mixed venous drainage, because the arterial supply of each lobulation was practically dependent on the anterior and posterior cerebral circulation respectively, it was proposed from the beginning to treat each one of the lobulations as an independent AVM, first of all the bigger pseudo lobulation (temporo-parietal) was embolized in 3 sessions until 90% and then resected by surgery, with an interval of 10 weeks between each embolization session, this represented 70% of the entire lesion. then was embolized the parieto-occipital pseudo lobulation which represented the 30% of the entire lesion, in 2 sessions, in the first was embolized 70% and in the second until the 95% with an interval of 4 weeks. A total of 23 onyx vials were used, and there was no evidence of final neurologic deficit immediately either in the follow up. Figure 3.