Comparing the first period with the second period, the number of patients with each S–M grade was as follows: grade I: 29 vs. 13 patients (total: 42 patients); grade II: 46 vs. 33 (total: 79); grade III: 39 vs. 22 (total: 61); grade IV: 31 vs. 14 (total: 45); and grade V: 9 vs. 6 (total: 15), respectively. The mean age was 33.1 ± 16.3 years in the first period and 33.3 ± 17.5 years in the second period. There were 82 men and 72 women in the first period, and 46 men and 42 women in the second period. The number of hemorrhagic-onset patients was 78 in the first period and 36 in the second period. There was no significant difference between periods for the proportion of patients with each S–M grade, age, sex, and hemorrhagic-onset.
The number of patients receiving intervention was 37 (88.1%) grade I, 71 (89.9%) grade II, 44 (72.1%) grade III, 14 (31.1%) grade IV, and 3 (20.0%) grade V. The proportion of hemorrhagic-onset patients was 50.0% (21/42) grade I; 51.9% (41/79) grade II; 45.9% (28/61) grade III; 42.2% (19/45) grade IV, and 33.3% (5/15) grade V, with 109 patients (70.8%) patients undergoing intervention in the first period and 60 patients (68.2%) undergoing intervention in the second period. There was no statistical difference in the intervention rate between the first and second periods, or for each grade.
Intervention by microsurgery alone decreased significantly comparing the first and second periods. The use of combined microsurgery and endovascular embolization increased significantly, and the use of stereotactic radiosurgery alone also increased significantly. Overall, interventions including microsurgery decreased significantly, and interventions including stereotactic radiosurgery increased significantly. The morbidity rate associated with intervention was 0.92% in the first period and 0% in the second period, and the mortality rate was 0.92% in the first period and 1.67% in the second period (Table 1).
Comparison between patients with hemorrhagic and nonhemorrhagic AVM
We found no significant differences for S–M grade, and no difference in the proportion of each S–M grade, age, and sex when comparing hemorrhagic (n = 114) and nonhemorrhagic (n = 128) patients. Regarding the intervention rate, there was no difference for patients with S–M grade I; however, patients with S–M grade II had high intervention rates for both hemorrhagic (97.6%) and nonhemorrhagic (81.6%) AVM, but the rate of intervention was significantly lower in nonhemorrhagic AVM (p = 0.025). In S–M grade III and IV patients, the rate of intervention was significantly lower in nonhemorrhagic AVM (III: p = 0.044 and IV: p = 0.011, respectively). For S–M grade V patients, intervention was often difficult, and therefore there was no difference between hemorrhagic and nonhemorrhagic patients. Overall, there were significantly fewer interventions by microsurgery alone and fewer interventions including microsurgery for nonhemorrhagic AVM. The proportion of patients undergoing medical treatment only was significantly greater for nonhemorrhagic AVM. Morbidity associated with intervention was 1.05% (1/95 intervention cases) in hemorrhagic AVM and 0% (0/74) in nonhemorrhagic AVM. Mortality associated with intervention was 1.75% (2/95) in hemorrhagic AVM and 0% (0/74) in nonhemorrhagic AVM. There were no statistical differences in morbidity and mortality between patients with hemorrhagic and nonhemorrhagic AVM (Table 2).
Table 2
Hemorrhagic vs nonhemorrhagic AVM.
|
|
(n = 114)
|
(n = 128)
|
|
|
Hemorrhagic
|
Non-hemorrhagic
|
p value
|
No. of Pt.
|
|
|
|
Gr. I n (%)
|
21 (18.4)
|
21 (16.4)
|
0.735
|
Gr. II n (%)
|
41 (36.0)
|
38 (29.7)
|
0.337
|
Gr. III n (%)
|
28 (24.6)
|
33 (25.8)
|
0.883
|
Gr. IV n (%)
|
19 (16.7)
|
26 (20.3)
|
0.511
|
Gr. V n (%)
|
5 (4.4)
|
10 (7.8)
|
0.299
|
Age
|
31.7 ± 16.7
|
34.4 ± 16.6
|
0.219
|
Sex(M, F)
|
57, 57
|
70, 58
|
0.520
|
Interventional cases (%)
|
|
|
|
Gr. I
|
19 (90.5)
|
18 (85.7)
|
1.000
|
Gr. II
|
40 (97.6)
|
31 (81.6)
|
0.025
|
Gr. III
|
24 (85.7)
|
20 (60.6)
|
0.044
|
Gr. IV
|
10 (52.6)
|
4 (15.4)
|
0.011
|
Gr.V
|
2 (40.0)
|
1 (10.0)
|
0.242
|
|
95
|
74
|
|
Treatment n (%)
|
|
|
|
surgery (S) only
|
54 (47.4)
|
30 (23.4)
|
< 0.001
|
S + embolization (E)
|
15 (13.2)
|
26 (20.3)
|
0.170
|
S + radiation (R)
|
9 (7.9)
|
1 (0.8)
|
0.007
|
S + E + R
|
1 (0.9)
|
1 (0.8)
|
1.000
|
R only
|
10 (8.8)
|
12 (9.4)
|
1.000
|
E + R
|
3 (2.6)
|
2 (1.6)
|
0.669
|
E only
|
3 (2.6)
|
2 (1.6)
|
0.669
|
medical
|
19 (16.7)
|
54 (42.2)
|
< 0.001
|
include surgery
|
79 (69.3)
|
58 (45.3)
|
< 0.001
|
include radiation
|
23 (20.2)
|
16 (12.5)
|
0.117
|
mortbidity, n (%)
|
1 (1.05)
|
0 (0)
|
1.000
|
mortality, n (%)
|
2 (1.75)
|
0 (0)
|
0.505
|
Baseline characteristics of patients with hemorrhagic vs nonhemorrhagic AVM.
Morbidity and mortality
Only one patient in the first period developed neurological deterioration with an mRS score ≥ 2 following intervention. Two patients died after intervention, both of which had S–M grade IV AVM with repeated hemorrhage; one patient died in the first period, and the other patient died in the second period. Although the deterioration was not mRS ≥ 2, six patients developed visual field defects. Overall morbidity associated with intervention was 0.92% (1/109 intervention cases) in the first period and 0% (0/60 intervention cases) in the second period. Mortality associated with intervention was 0.92% (1/109) in the first period and 1.67% (1/60) in the second period.
Comparing treatment for patients with each S–M grade between the first and second periods
Grade I
Patients with S–M grade I AVMs were an almost homogeneous group. We intervened in 24/29 patients (82.8%) in the first period and all 13 patients (100%) in the second period. Among patients undergoing intervention, treatment included microsurgery in 23 patients in the first period and 12 patients in the second period. Treatment with microsurgery alone was conducted in 19 patients in the first period and 4 patients in the second period. The use of microsurgery alone decreased significantly as a treatment strategy (p = 0.049), and the combination of microsurgery and endovascular embolization increased significantly (p < 0.001). Although a high proportion of patients with S–M grade I AVMs underwent treatments including microsurgery, the number of microsurgical procedures increased after the introduction of endovascular embolization (Figure 1A). Mortality and morbidity rates were both 0% following intervention for grade I AVMs throughout the study.
Grade II
In patients with S–M grade II AVMs, the treatment strategy changed from the first to the second periods (Figure 1B). The rate of microsurgery alone decreased (p < 0.001) (Figure 2A), the combination of microsurgery and endovascular embolization increased (p = 0.035) (Figure 2A), stereotactic radiosurgery alone increased (p = 0.035) (Figure 2A), treatment including microsurgery decreased (p = 0.005) (Figure 2B), and treatment including stereotactic radiosurgery increased (p = 0.036) (Figure 2C).
Patients with S–M grade II AVMs appeared as three subtypes. The first group had a medium-sized nidus (3–6 cm) (size 2, eloquent 0, deep drainage 0: S2E0D0); the second group had a small-sized (< 3 cm) nidus located in the eloquent area (S1E1D0); and the third group had deep venous drainage with a small-sized nidus (S1E0D1).
In the S2E0D0 subgroup, there was no significant change in the selected treatment (Figure 3A). In the S1E1D0 subgroup, the rate of microsurgery alone decreased significantly (p = 0.002) (Figure 3B); however, the rates of stereotactic radiosurgery alone and treatment including stereotactic radiosurgery increased significantly (p = 0.006, 0.038) (Fig 3B). In the S1E0D1 subgroup, treatment with microsurgery alone decreased significantly (p = 0.009) (Figure 3C). Only one patient with hemorrhagic-onset grade II AVM experienced neurological deterioration after microsurgery during the first period, thus mortality and morbidity associated with intervention were 0% and 2.3% (1/43 intervention cases) in the first period, respectively. In the second period, mortality and morbidity according to the intervention were both 0%.
Grade III
In patients with S–M grade III AVMs, microsurgery alone decreased significantly (p < 0.001) (Figure 1C and 4A), the combination of endovascular embolization and stereotactic radiosurgery increased significantly (p = 0.043) (Figure 4A), treatment including microsurgery decreased significantly (p = 0.003) (Figure 4B), and treatment including stereotactic radiosurgery increased significantly (p = 0.011) (Figure 4C).
Patients with grade III AVMs were subdivided into four subtypes. The first group had a small-sized nidus in the eloquent area with deep venous drainage (S1E1D1); the second group had a medium-sized nidus located in the eloquent area (S2E1D0); the third group had deep venous drainage with a medium-sized nidus (S2E0D1); and the fourth group had a large-sized nidus (S3E0D0). We saw only one patient with the S3E0D0 subtype in the first period and none in the second period; therefore, comparisons between periods were not possible. In the S1E1D1 subtype, microsurgery alone decreased significantly (p = 0.006) (Figure 5A), and treatment including microsurgery also decreased significantly (p = 0.033) (Figure 5A). For the other subgroups, the number of patients was small, and we saw no significant difference as a result (Figure 5B and 5C). Both mortality and morbidity according to intervention were 0% throughout the study period.
Grade IV
In patients with S–M grade IV AVMs, we saw no change in the selected treatment between the two periods. Patients with S–M grade IV AVMs were divided into three subtypes. The first group had a medium-sized nidus in the eloquent area with deep venous drainage (S2E1D1); the second group had a large-sized nidus located in the eloquent area (S3E1D0); and the third group had a large-sized nidus with deep venous drainage (S3E0D1). The S2E1D1 subgroup consisted of a relatively high number of patients; however, the treatment strategy remained unchanged. The other subgroups consisted of only a small number of patients, and no statistical difference was observed (Figure 1D). Although no patients developed neurological deterioration in either period, two patients developed repeated hemorrhagic leading to death despite intervention (1 patient in the first period and 1 patient in the second period). Mortality and morbidity according to intervention were, respectively, 10.0% (1/10) and 0% in the first period, and 25.0% (1/4) and 0% in the second period, respectively.
Grade V
In patients with S–M grade V AVMs, we saw no significant difference in the selected treatment between the first and second periods. Microsurgery was not selected in either period, and medical treatment was selected for 80% of patients. Among 15 patients with S–M grade V AVMs, only three patients underwent intervention (2 underwent endovascular embolization alone and 1 underwent stereotactic radiosurgery alone; Figure 1E). No patients developed neurological deterioration secondary to intervention, and no patients died. Both mortality and morbidity according to intervention were 0% throughout the study period.