Abstract
Purpose
The champagne bottle neck sign (CBNS) is regarded as a characteristic finding of the cervical internal carotid artery in patients with moyamoya disease (MMD). However, there is no universal definition, and its clinical significance remains unclear. Here, we aimed to propose a novel CBNS grading system using catheter angiography.
Methods
First, we proposed a novel grading system for CBNS using the lateral view of common carotid artery angiograms. Three grades were defined (grades 0,1, and 2). CBNS grade was defined as a ratio of the internal carotid artery diameter to the common carotid artery diameter of < 0.5 on the lateral view of the common carotid artery angiogram. CBNS grade 2 additionally required plump enlargement of the bulbous portion of the cervical internal carotid artery. We then conducted a retrospective case-control study including 134 patients (247 sides) with MMD and 94 patients (98 sides) with non-MMD intracranial large artery stenosis/occlusion. Finally, we analyzed the association between CBNS grade and clinical features, including Suzuki stage and hemodynamic impairment.
Results
CBNS grade 2 demonstrated a strong association with MMD, although grade 2 CBNS was infrequently observed among patients with non-MMD intracranial large-artery stenosis/occlusion (p < 0.001). Additionally, multivariate analysis demonstrated a significant association between CBNS grade 2 and hemodynamic impairment, which was present in 32.5% (13/40) and 55.9% (19/34) of MMD sides at CBNS grades 1 and 2, respectively (p < 0.001).
Conclusion
We propose a CBNS grading system for MMD patients. CBNS grade 2, characterized by pronounced enlargement of the bulbous portion of the cervical internal carotid artery, may serve as an imaging marker of advanced Suzuki stage and hemodynamic impairment in MMD.
Keywords: Moyamoya disease, Champagne bottle neck sign, Internal carotid artery, Hemodynamic compromise, Suzuki stage
Highlights
What is already known on this topic
Champagne bottle neck sign (CBNS) is characteristic finding of the cervical internal carotid artery in patients with Moyamoya disease, but its clinical significance remains unclear.
What this study adds
A novel CBNS grading system was developed; CBNS grade 2, characterized by pronounced enlargement of the bulbous portion of the cervical internal carotid artery, is significantly associated with advanced Suzuki stage and hemodynamic impairment.
How this study might affect research, practice or policy
Identification of CBNS grade 2 may allow clinicians to assess hemodynamic impairment prior to obtaining SPECT imaging.
Introduction
Moyamoya disease (MMD) is a cerebrovascular disorder characterized by progressive stenosis or occlusion of the terminal portion of the internal carotid artery (ICA) and the formation of an abnormal vascular network at the base of the brain [1]. The champagne bottle neck sign (CBNS) has been reported as a characteristic finding of the cervical ICA in patients with MMD [2, 3]. However, no universal definition of CBNS exists, and its clinical significance remains unclear.
Most previous studies have evaluated CBNS using sonography [2, 3]. For example, Yasaka and colleagues detected a rapid reduction in the proximal ICA diameter relative to the common carotid artery (CCA) diameter using carotid ultrasonography in patients with MMD [3]. In these studies, CBNS was defined as a rapid reduction in the luminal diameter of the proximal ICA, specifically an ICA-to-CCA ratio of < 0.5 on carotid sonography [2, 3]. Using this definition, CBNS has been reported in 56%-74% of patients with MMD [2, 3]. However, sonographic assessment is not routinely performed in daily clinical practice for MMD, and the reproducibility of ultrasonographic findings is limited. Therefore, a universal and clinically practical definition of CBNS is required.
Furthermore, the clinical significance of CBNS remains largely unknown. Previous studies have demonstrated that a significant proportion of patients exhibit CBNS, and that its presence is associated with hemodynamic compromise. Yasuda et al. reported that CBNS was associated with ipsilateral advanced MMD and impaired cerebral vasoreactivity [2]. Wang et al. demonstrated that patients with ipsilateral intracranial hemorrhage exhibited a significantly smaller luminal diameter ratio of the ICA to the CCA and a higher prevalence of CBNS [4]. Detecting hemodynamic compromise is clinically important as it may be associated with adverse outcomes, including an increased risk of stroke, potential cognitive decline, and the need for timely surgical intervention. Therefore, investigations into the clinical significance of CBNS based on a universal definition are warranted.
In this study, we aimed to propose a novel CBNS grading system using catheter angiography and to further evaluate its predictive value for Suzuki stage and hemodynamic compromise.
Materials and methods
Patients
This retrospective case-control study adhered to the Declaration of Helsinki on Ethical Principles for Medical Research Involving Human Subjects and was approved by the Ethics Committee of Kohnan Hospital (Sendai, Japan) under the approval number 2024-1120-03. Written informed consent was obtained from all participants.
We assessed patients with MMD and non-MMD intracranial large-artery stenosis, including isolated atherosclerotic stenosis or occlusion of the ICA and middle cerebral artery (MCA), who underwent catheter angiography between January 2021 and June 2024. Patients were assessed for eligibility using predefined inclusion and exclusion criteria.
The diagnosis of MMD was made based on the diagnostic criteria of the Research Committee on Spontaneous Occlusion of the Circle of Willis of the Ministry of Health, Labour and Welfare of Japan [5]. Newly diagnosed patients with MMD before revascularization surgery were included.
Non-MMD intracranial large-artery stenosis/occlusion was defined as angiographically confirmed 50% to 99% stenosis [6] or complete occlusion of the internal carotid artery (ICA) or the M1 segment of the MCA attributed to intracranial atherosclerosis. The presence of atherosclerotic etiology was supported by clinical vascular risk factors and characteristic neuroimaging findings (e.g., focal eccentric narrowing or plaque without typical basal moyamoya vessels). Alternative non-atherosclerotic etiologies were strictly excluded as follows:
Arterial dissection: Excluded based on angiographic and MRI findings, including double-lumen sign, luminal narrowing with “string sign”, or gradual tapering ending in total lumen occlusion (“flamelike” occlusion) [7].
Vasculitis and non-atherosclerotic vasculopathies: Excluded if patients showed clinical/laboratory evidence of systemic autoimmune or inflammatory conditions, smooth concentric wall enhancement on contrast-enhanced MRI, or diffuse multifocal arterial narrowing indicative of central nervous system vasculitis or reversible cerebral syndrome (RCVS) [8].
Moyamoya syndrome (quasi-Moyamoya): Excluded if arterial steno-occlusive lesions were secondary to known underlying conditions (e.g., prior cranial irradiation, Down syndrome, or neurofibromatosis type 1).
Patients with primary extracranial artery steno-occlusive disease were also excluded.
Following this comprehensive re-screening, a total of 134 consecutive patients with MMD (247 sides: mean age, 48.9 ± 13.0 years) and 94 consecutive patients with non-MMD intracranial large-artery stenosis/occlusion (98 sides; mean age, 58.9 ± 12.8 years) were enrolled. Clinical features and imaging data were obtained from medical records. Demographic features of the enrolled patients are summarized in Table 1.
Table 1.
Demographic characteristics of enrolled patients with Moyamoya disease and intracranial large artery stenosis/occlusion
| MMD (134 patients/247 sides) |
Non-MMD large artery stenosis/occlusion (94 patients/98 sides) |
|
|---|---|---|
| Age (yr) | 48.9 ± 13.0 | 58.9 ± 12.8 |
| Sex (M: F) | 43 patients:91 patients | 57 patients:37patients |
| Onset | ||
| Ischemia | 108 sides | 90 sides |
| Hemorrhage | 27 sides | 0 sides |
| Asymptomatic | 112 sides | 4 sides |
| Suzuki stage | ||
| Stage 1–2 | 43 sides | - |
| Stage 3 | 110 sides | - |
| Stage 4 | 76 sides | - |
| Stage 5 | 15 sides | - |
| Stage 6 | 3 sides | - |
MMD, Moyamoya disease; M, male; F, female. Age represents mean
standard deviation
Evaluation of angiographic findings
Diagnostic catheter angiography was conducted in all patients included in this study. Six-vessel angiography-including the CCA, ICA, external carotid artery, and vertebral artery- was routinely performed. A 5-Fr catheter was used in all cases. Angiographic images were acquired using an Innova 3131 IQPro X-ray system (GE Healthcare, Chicago, IL, USA). Suzuki stage [1] and CBNS were routinely evaluated in patients with MMD. The luminal diameters of the proximal ICA (below the mandible, 1.5–2.0 cm above the carotid bifurcation) and the CCA (proximal aspect to the bulb, 1.0–1.5 cm below the carotid bifurcation) were measured during the arterial phase on the lateral view of CCA angiograms (Fig. 1). Rapid luminal diameter reduction was defined as an ICA-to-CCA luminal diameter < 0.5.
Fig. 1.
Measurement of common carotid artery/internal carotid artery/external carotid artery diameters. The diameters of common carotid artery (CCA), internal carotid artery (ICA), and external carotid artery (ECA) were measured on the lateral view of the common carotid angiogram as follows; CCA diameter proximal to the bulb (1.0–1.5 cm below the carotid bifurcation); ICA diameter just distal to the bulb (1.5–2.0 cm above the carotid bifurcation); and ECA diameter at the main trunk
CBNS grading system
In daily clinical practice, catheter angiography is commonly used to diagnose MMD. We therefore established a grading system for CBNS as follows (Fig. 2): CBNS grade 1 was defined as ICA-to-CCA diameter ratio (ICA below the mandible to CCA proximal to the carotid bulb) of < 0.5 on the lateral view of the CCA angiogram. This criterion corresponds to the classical definition of CBNS reported previously [2, 3]. Furthermore, CBNS grade 2 was defined by the presence of plump enlargement of the cervical ICA on the lateral view of the CCA angiogram. Plump enlargement was defined as an angiographic appearance in which the origin of the ICA (carotid bulb) expands outward before narrowing into the distal ICA. Cases not meeting the criteria for CBNS grades 1 and 2 were classified as CBNS grade 0. Representative images are shown in Fig. 3.
Fig. 2.
Champagne bottle neck sign grading system. The lateral view of the common carotid angiogram (CCAG) was used to evaluate the champagne bottle neck sign (CBNS). Diameters of the common carotid artery (CCA), internal carotid artery (ICA) and external carotid artery (ECA) were measured as follows; CCA diameter proximal to the carotid bulb (1.0–1.5 cm below the carotid bifurcation), ICA diameter just distal to the bulb (1.5–2.0 cm above the carotid bifurcation); ECA diameter at the main trunk. An ICA-to-CCA diameter of < 0.5 on the lateral view of the CCAG was required for CBNS grades 1 and 2. In addition, CBNS grade 2 was defined by the presence of plump enlargement of the cervical ICA. Plump enlargement was defined as an angiographic appearance in which the origin of the ICA that gradually expands outward before narrowing into the distal ICA. Cases not meeting the criteria for CBNS grades 1 or 2 were classified as CBNS grade 0
Fig. 3.
Representative images of CBNS in patients with moyamoya disease. CBNS, champagne bottle neck sign
To evaluate the reproducibility and minimize interobserver variability of the proposed classification, the CBNS grade for each side was assessed independently by two neurosurgeons who were blinded to patient’s clinical histories and perfusion imaging results. Discrepancies in the initial grading between the evaluators were resolved through a joint review to reach a final consensus. Prior to the consensus-building process, interobserver agreement was statistically evaluated using Cohen’s kappa coefficient to validate the reliability of the qualitative criteria, including the “plump enlargement” definition for CBNS grade 2.
Evaluation of hemodynamic compromise
Hemodynamic status was assessed using the auto-radiographic method by N-isopropyl-p-[123I] iodoamphetamine SPECT (123I-IMP SPECT) on a GCA9300R system (Canon, Tokyo, Japan). An acetazolamide challenge was performed in patients with MMD and, when available, in non-MMD control patients. Cerebral blood flow (CBF) was semiquantitatively measured at baseline and 15 min after intravenous administration of 17 mg/kg acetazolamide.
Regional CBF was expressed as a percentage relative to the institution-specific normal control values derived from volunteers without cerebrovascular disease.
Cerebrovascular reserve (CVR) in response to acetazolamide was calculated as follows:
CVR (%) = 100 × (CBFACZ - CBFrest)/CBFrest.
Resting CBF < 90% compared with normal control value and CVR < 10% was defined as hemodynamic compromise.
Statistical analysis
In line with the previous literatures [9–11], statistical analyses were performed using individual sides as the unit of observation, because the progression of steno-occlusive lesion and the development of collateral pathways in MMD often occur asymmetrically between the two hemispheres.
Continuous variables were expressed as mean 
standard deviation and were compared using the Mann-Whitney U test. Categorical variables were expressed as frequencies and percentages. For multi-group comparisons of categorical variables, the chi-square test was used; however, when the expected cell count was less than 5, Fisher’s exact test was applied when the expected cell count was less than 5.
To identify specific group differences, post-hoc pairwise comparisons were performed following the overall analysis, and the Bonferroni correction was applied to adjust for multiple comparisons. Multivariable logistic regression analysis was performed to determine independent predictors, adjusting for confounding factors as described above. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. All statistical analyses were two-sided, and a p-value less than 0.05 was considered statistically significant. Statistical analyses were conducted using JMP Student Edition 19 (JMP, Cary, NC, USA).
Results
Interobserver Agreement for the CBNS Grading system
The reproducibility of the proposed CBNS grading system was evaluated between the two evaluators, The analysis demonstrated high interobserver agreement, with a Cohen’s kappa coefficient of 
(95% confidence interval: 0.79–0.95), indicating strong reliability in distinguishing CBNS grades.
CBNS in patients with MMD and non-MMD intracranial artery stenosis
Understanding the incidence and distinct characteristics of CBNS in patients with MMD and those with non-MMD intracranial large artery stenosis is essential for improving diagnostic accuracy and patient management. We compared the proportions of CBNS grades 1 and 2 among the three groups: patients with MMD, non-MMD internal carotid artery occlusion/stenosis (ICAO/ICAS), and middle cerebral artery occlusion/stenosis (MCAO/MCAS). The overall comparison revealed a significant difference among the three groups (29.1% [72/247] in MMD, 36.1% [17/47] in ICAO/ICAS, and 0.0% [0/51] in MCAO/MCAS). To further investigate these differences, post-hoc pairwise comparisons were performed using Fisher’s exact test with Bonferroni correction (adjusted significance threshold: p < 0.025). The MMD group showed a significantly higher incidence of CBNS grades 1 + 2 compared with the MCAO/MCAS group, whereas there was no statistically significant difference in CBNS grades 1 and 2 rate between the MMD and ICAO/ICAS groups (Fig. 4A).
Fig. 4.
CBNS in non-MMD ICA/MCA stenosis/occlusion. (a) Comparison of the incidence of champagne bottle neck sign (CBNS) grades 1 and 2 among the patients with moyamoya disease (MMD), internal carotid artery occlusion/stenosis (ICAO/ICAS), and middle cerebral artery occlusion/stenosis (MCAO/MCAS) groups. Overall differences among the three groups were analyzed using the chi-square test. Post-hoc pairwise comparisons (MMD vs. ICAO/ICAS and MMD vs. MCAO/MCAS) were performed using Fisher’s exact test with Bonferroni correction to adjust for multiple comparisons (significance threshold set at p < 0.025 for two comparisons). ns, not significant; *, p < 0.025; **, p < 0.01; ***, p < 0.001. (b) Comparison of the incidence of CBNS grade 2 among patients with MMD, ICAO/ICAS, and MCAO/MCAS groups. Overall differences among the three groups were analyzed using the chi-square test. Post-hoc pairwise comparisons (MMD vs. ICAO/ICAS and MMD vs. MCAO/MCAS) were performed using Fisher’s exact test with Bonferroni correction to adjust for multiple comparisons (significance threshold set at p < 0.025 for two comparisons). ns, not significant; *, p < 0.025; **, p < 0.01; ***, p < 0.001
Similarly, the overall comparison revealed a significant difference in the incidence of CBNS grade 2 among the three groups (13.4% [33/247] in MMD, 2.1% [1/47] in ICAO/ICAS, and 0.0% [0/51] in MCAO/MCAS). The MMD group showed higher incidence of CBNS grade 2 compared to MCAO/MCAS group, whereas there was no statistically significant difference in CBNS grade 2 rates between the MMD and ICAO/ICAS groups (Fig. 4B).
CBNS prevalence increases with disease progression
Next, we examined the Suzuki angiographic stages at which CBNS was most frequently observed. CBNS was identified in 3.3% (1/33), 22.5% (18/80), 45.6% (26/57), 100.0% (14/14), and 3.3% (1/33) of patients with MMD at Suzuki stages 1–2, 3, 4, 5, and 6, respectively (p < 0.001; Fig. 5A). Patients with MMD at Suzuki stages 4 and 5 showed a significantly higher incidence of CBNS grades 1 + 2 compared with those at Suzuki stage 1–2 (Fig. 5A). Additionally, CBNS grade 2 was most frequently observed in patients with MMD at Suzuki stages 4 and 5 (p < 0.001 for Suzuki stage 4 vs. Suzuki stage 1–2, and Suzuki stage 5 vs. Suzuki stage 1–2) (Fig. 5B). Furthermore, higher CBNS grade was significantly associated with hemodynamic impairment (resting CBF < 90% and CVR < 10%), which was evident in 5.5% (7/127), 31.4% (11/35), and 48.0% (12/25) of patients with MMD with CBNS grades 0, 1 and 2, respectively (p < 0.001) (Fig. 6).
Fig. 5.
Association between CBNS grade and clinical features. (a) Association between Suzuki stage and champagne bottle neck sign (CBNS) grades 1 and 2 among the patients with moyamoya disease (MMD). Overall differences among the five groups were analyzed using the chi-square test. Post-hoc pairwise comparisons (Stage 3 vs. Stages 1–2, Stage 4 vs. Stages 1–2, Stage 5 vs. Stages 1–2, and Stage 6 vs. Stages 1–2) were performed using Fisher’s exact test with Bonferroni correction to adjust for multiple comparisons (significance threshold set at p < 0.0125 for four comparisons). ns, not significant; *, p < 0.0125; **, p < 0.01; ***, p < 0.001. (b) Association between Suzuki stage and CBNS grade 2 among the patients with MMD. Overall differences among the five groups were analyzed using the chi-square test. Post-hoc pairwise comparisons (Stage 3 vs. Stages 1–2, Stage 4 vs. Stages 1–2, Stage 5 vs. Stages 1–2, and Stage 6 vs. Stages 1–2) were performed using Fisher’s exact test with Bonferroni correction to adjust for multiple comparisons (significance threshold set at p < 0.0125 for four comparisons). ns, not significant; *, p < 0.0125; **, p < 0.01; ***, p < 0.001
Fig. 6.
Association between CBNS grade and hemodynamic impairment. Association between hemodynamic impairment and CBNS grades among the patients with moyamoya disease (MMD). Overall differences among the three groups were analyzed using the chi-square test. Post-hoc pairwise comparisons (CBNS grade 1 vs. CBNS grade 0, and CBNS grade 2 vs. CBNS grade 0) were performed using Fisher’s exact test with Bonferroni correction to adjust for two comparisons (significance threshold set at p < 0.025 for four comparisons). ns, not significant; *, p < 0.025; **, p < 0.01; ***, p < 0.001
Multivariate Analysis for Independent Associations
In the multivariate logistic regression analysis adjusting for age, sex, Suzuki stage, onset, vascular risk profiles, such as hypertension, diabetes mellitus, the CBNS grade had a strong association with Suzuki stage in both Model 1 and 2 (Table 2). In particular, Suzuki stage 4 and 5 remained independently associated with CBNS grade with odds ratio of 41.5 (95%CI: 5.16–333.1, p < 0.0001) and 1175.0 (95%CI: 96.2-14345.0, p < 0.0001) in Model 2, respectively (Table 2). Similarly, in the multivariate logistic regression analysis adjusting for age, sex, Suzuki stage, and onset, the CBNS grade remained significantly and independently associated with hemodynamic impairment in both Model and 2 (Table 3). CBNS grades 1 and 2 were independently associated with CBNS grade with odds ratio of 4.23 (95%CI: 1.6–10.7), p < 0.0001) and 15.1 (95%CI: 4.85–46.7) in Model 2, respectively (Table 3). These findings indicate that the CBNS grading system provides independent predictive value beyond established clinical parameters and disease markers.
Table 2.
Multivariate analysis between CBNS grade and clinical features, including Suzuki stage
| n | CBNS Grade1 + 2 |
CBNS Grade 2 |
Crude OR (95% CI) |
Model 1 OR (95% CI) |
Model 2 (95% CI) |
||
|---|---|---|---|---|---|---|---|
| Age | < 35 | 27 |
5 (18.5%) |
3 (11.1%) |
1.00 (reference) |
||
| 35–44 | 55 |
14 (25.4%) |
7 (12.7%) |
1.46 (0.46–4.52) |
1.12 (0.27–4.54) |
||
| 45–54 | 86 |
27 (31.3%) |
15 (17.4%) |
1.97 (0.68–5.68) |
1.56 (0.40–6.08) |
||
| 55–64 | 51 |
17 (33.4%) |
5 (9.8%) |
1.98 (0.66–6.03) |
2.01 (0.48–8.45) |
||
| 65< | 28 |
10 (35.7%) |
4 (14.3%) |
2.25 (0.66–7.62) |
1.88 (0.38–9.34) |
||
| Sex | |||||||
| 0 | 80 |
28 (35.0%) |
18 (22.5%) |
1.00 (reference) |
1.00 (reference) |
||
| 1 | 167 |
47 (26.9%) |
16 (9.6%) |
0.61 (0.35–1.06) |
0.66 (0.33–1.29) |
||
| Hypertension | |||||||
| 0 | 145 |
41 (28.3%) |
19 (13.1%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 1 | 102 |
32 (31.4%) |
15 (14.7%) |
1.16 (0.67–1.99) |
0.95 (0.47–1.91) |
0.83 (0.39–1.72) |
|
| Diabetes | |||||||
| 0 | 205 |
59 (28.8%) |
31 (15.1%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 1 | 42 |
14 (33.3%) |
3 (7.1%) |
1.07 (0.52–2.18) |
0.27 (0.09–0.76) |
0.24 (0.08–0.72) |
|
| Dyslipidemia | |||||||
| 0 | 195 |
56 (28.7%) |
26 (13.3%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 1 | 52 |
17 (32.7%) |
8 (15.4%) |
1.20 (0.63–2.29) |
2.03 (0.87–4.72) |
1.75 (0.08–0.72) |
|
| Smoking | |||||||
| 0 | 169 |
48 (28.4%) |
25 (14.8%) |
1.00 (reference) |
1.00 (Reference) |
1.00 (reference) |
|
| 1 | 78 |
25 (32.0%) |
9 (11.6%) |
1.10 (0.62–1.96) |
0.96 (0.49–1.89) |
0.98 (0.49–1.96) |
|
| Alcohol | |||||||
| 0 | 168 |
45 (26.8%) |
21 (12.5%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 1 | 79 |
28 (35.4%) |
13 (16.5%) |
1.48 (0.84–2.60) |
1.93 (0.99–3.74) |
1.64 (0.82–3.28) |
|
| Onset | |||||||
| Asymptomatic | 112 |
25 (22.5%) |
11 (9.8%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| Ischemia | 108 |
40 (37.0%) |
19 (17.6%) |
2,01 (1.12–3.60) |
1.14 (0.58–2.23) |
1.15 (0.58–2.27) |
|
| Hemorrhage | 27 |
8 (29.4%) |
4 (14.8%) |
1.47 (0.58–3.70) |
1.15 (0.42–3.14) |
1.11 (0.40–3.08) |
|
| Suzuki stage | |||||||
| 1–2 | 43 |
1 (2.3%) |
0 (0.0%) |
1.0 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 3 | 110 |
23 (20.9%) |
7 (6.4%) |
11.1 (1.44–85.6) |
11.5 (1.46–90.1) |
10.8 (1.35–85.5) |
|
| 4 | 76 |
33 (43.6%) |
16 (21.1%) |
34.4 (4.49–264.3) |
42.2 (5.36–333.3) |
41.5 (5.16–333.1) |
|
| 5 | 15 |
15 (100.0%) |
11 (73.3%) |
471.4 (45.8-4852.6) |
1083.7 (90.8-12941.2) |
1175.0 (96.2-14345.0) |
|
| 6 | 3 |
1 (33.3%) |
0 (0.0%) |
18.0 (0.77–418.2) |
35.1 (1.125-988.1) |
38.4 (1.40-1058.6) |
Model 1 : Age-sex adjusted multivariate analysis
Model 2 : Multivariate analysis using all explanatory variables in the table
CBNS, Champagne bottle neck sign; OR, Odds ratios; CI, confidence intervals
Table 3.
Multivariate analysis between hemodynamic impairment and clinical features, including CBNS grade
| n | Hemodynamic impairment | Crude OR (95% CI) |
Model 1 OR (95% CI) |
Model 2 OR (95% CI) |
||
|---|---|---|---|---|---|---|
| Age | < 35 | 27 | 3 (11.1%) | 1.00 (ref) | ||
| 35–44 | 55 | 11 (20.0%) |
2.00 (0.51–7.87) |
1.73 (0.37–8.10) |
||
| 45–54 | 86 | 16 (18.6%) |
1.83 (0.49–6.83) |
1.52 (0.35–6.67) |
||
| 55–64 | 51 | 13 (25.5%) |
2.74 (0.71–10.6) |
2.69 (0.57–12.6) |
||
| 65< | 28 | 5 (17.9%) |
1.74 (0.37–8.12) |
1.89 (0.32–11.2) |
||
| Sex | ||||||
| 0 | 80 | 17 (21.2%) |
1.00 (reference) |
1.00 (reference) |
||
| 1 | 167 | 31 (18.6%) |
0.84 (0.44–1.64) |
1.13 (0.51–2.54) |
||
| Hypertension | ||||||
| 0 | 145 | 28 (19.3%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 1 | 102 | 20 (19.6%) |
1.02 (0.54–1.93) |
1.15 (0.54–2.43) |
0.98 (0.45–2.15) |
|
| Diabetes | ||||||
| 0 | 205 | 43 (21.0%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 1 | 42 | 5 (11.9%) |
0.51 (0.19–1.37) |
0.79 (0.25–2.48) |
0.75 (0.23–2.43) |
|
| Dyslipidemia | ||||||
| 0 | 195 | 40 (20.5%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 1 | 52 | 8 (15.4%) |
0.70 (0.31–1.62) |
0.55 (0.20–1.52) |
0.57 (0.20–1.61) |
|
| Smoking | ||||||
| 0 | 169 | 30 (17.8%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 1 | 78 | 18 (23.1%) |
1.39 (0.72–2.68) |
1.57 (0.73–3.37) |
1.52 (0.69–3.38) |
|
| Alcohol | ||||||
| 0 | 168 | 32 (19.1%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 1 | 79 | 16 (20.3%) |
1.07 (0.55–2.11) |
0.76 (0.35–3.37) |
0.71 (0.31–1.61) |
|
| Onset | ||||||
| Asymptomatic | 112 | 20 (17.9%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| Ischemia | 108 | 22 (20.4%) |
1.00 (0.46–2.16) |
1.02 (0.46–2.22) |
||
| Hemorrhage | 27 | 6 (22.2%) |
0.92 (0.28–3.01) |
0.97 (0.29–3.17) |
||
| Suzuki stage | ||||||
| 1–2 | 43 | 5 (11.6%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 3 | 110 | 21 (19.1%) |
1.79 (0.63–5.11) |
1.06 (0.33–3.29) |
1.06 (0.33–3.39) |
|
| 4 | 76 | 18 (23.7%) |
2.23 (0.76–6.54) |
0.78 (0.22–2.76) |
0.75 (0.20–2.78) |
|
| 5 | 15 | 4 (26.7%) |
2.76 (0.63–12.1) |
0.25 (0.04–1.74) |
0.267 (0.04–1.90) |
|
| 6 | 3 | 0 (0.0%) |
0 NA |
1.2≉10− 6 NA |
1.1≉10− 6 NA |
|
| CBNS | ||||||
| 0 | 174 | 19 (10.9%) |
1.00 (reference) |
1.00 (reference) |
1.00 (reference) |
|
| 1 | 39 | 12 (30.8%) |
3.63 (1.58–8.31) |
4.40 (1.76–11.0) |
4.23 (1.6–10.7) |
|
| 2 | 34 | 17 (50.0%) |
8.16 (3.58–18.6) |
14.8 (4.89–44.8) |
15.1 (4.85–46.7) |
Model 1 : Age-sex adjusted multivariate analysis
Model 2 : Multivariate analysis using all explanatory variables in the table
CBNS, Champagne bottle neck sign; OR, Odds ratios; CI, confidence interval; NA, not available.
Discussion
CBNS lacks a clear diagnostic framework, which creates a significant gap in effective identification and treatment decision-making. CBNS is a well-recognized angiographic finding: however, only a limited number of studies have investigated its clinical significance. We aimed to establish a clear definition of CBNS to improve diagnostic accuracy and to clarify its clinical relevance.
To achieve these aims, we proposed a novel visual grading system for CBNS (Fig. 2). Unlike previous ultrasonographic definitions, which lacked a standardized grading system, we proposed a novel classification comprising three grades: CBNS grades 0, 1, and 2. A key advantage of this grading system is its simplicity and reproducibility. In contrast, the reproducibility of ultrasonography is limited by inter-rater variability, often resulting in inconsistency and poor agreement. Ultrasonographic evaluation is particularly challenging in cases with high carotid bifurcation. In such cases, transoral carotid sonography has been used to accurately measure the internal diameter of the extracranial ICA [12–15]. Furthermore, the current CBNS grading system incorporates plump enlargement, a feature overlooked by previous ultrasonographic criteria, which focused solely on cervical ICA shrinkage. While the current definition of “plump enlargement” in CBNS grade 2 relies primarily on qualitative assessment, our results demonstrated a high interobserver agreement (
), suggesting that the criteria are sufficiently intuitive and reproducible for clinical use. CBNS grade 2 is observed typically in MMD patients at Suzuki stages 4 and 5, whereas CBNS grade 1 was observed in both MMD and non-MMD large artery stenosis/occlusion (Fig. 4). Therefore, CBNS grade 2 may serve as an imaging marker for differentiating MMD from non-MMD intracranial large-artery stenosis/occlusion.
The potential role of CBNS as an imaging marker was explored in this study. First, our results showed that a higher CBNS grade remains characteristic of MMD and is rarely seen in non-MMD intracranial stenosis, even after adjusting for potential confounders (Fig. 4; Table 2). This finding strongly suggests that the CBNS grading system possesses high diagnostic specificity for MMD. Second, we demonstrated that the CBNS grading system corresponds to hemodynamic status (Fig. 5c). A previous study also reported that CBNS was associated with an ipsilateral advanced MMD stage (Suzuki stages 4–6) and impaired cerebral vasoreactivity [2]. Furthermore, several studies have investigated the association between CBNS and cerebrovascular events. For example, Wang et al. demonstrated that patients with ipsilateral intracranial hemorrhage had a significantly smaller luminal diameter ratio of the ICA to the CCA and a higher prevalence of CBNS [4]. When hemispheres with intracranial hemorrhage were compared, patients with ipsilateral carotid artery CBNS had a significantly higher prevalence of hemorrhage in posterior territories than patients without CBNS [4].
Collectively, CBNS grade 2 may serve as an imaging marker of hemodynamic impairment and may reflect an increased risk of cerebrovascular events. In this study, CBNS grade was correlated with advanced Suzuki stage and hemodynamic impairment, which suggests the potential value of CBNS as a marker of hemodynamic adaptation in the course of steno-occlusive changes in MMD [1]. Clinically, the presence of CBNS grade 2 may suggest an advanced Suzuki stage and impaired cerebrovascular reserve prior to obtaining SPECT data. Crucially, our multivariate analysis demonstrated that the association between CBNS grade and hemodynamic impairment remains robust even after rigorous adjustment for potential confounders, including age, sex, and conventional risk profiles (Table 2). This underscores that the CBNS grading system is not merely a reflection of systemic demographic or vascular risk but rather serves as an independent imaging marker that directly correlates with hemodynamic impairment. In this study, the association between CBNS and cerebrovascular events was not evident (Tables 2 and 3), since most of the patients had diagnostic angiography after cerebrovascular events. Therefore, the clinical utility of the CBNS grading system lies in its potential to assist clinicians in estimating preliminary hemodynamic status immediately from routine angiographic or imaging findings, even before definitive functional imaging like SPECT data becomes available. This may provide helpful, adjunctive insights during the initial stages of surgical planning, particularly in identifying patients who warrant prioritized hemodynamic evaluation.
The mechanism underlying CBNS formation remains largely unknown. CBNS refers to a reduction in the diameter of the proximal ICA that resembles a champagne bottle neck. CBNS primarily reflects the shrinkage of the ICA, while the ICA diameter at the bulbous level remains relatively intact. Previous histopathological studies have revealed that the bulbous level corresponds to the transition from elastic to muscular arteries [16, 17]. Accordingly, CBNS may reflect shrinkage of the muscular ICA in the distal cervical ICA, while the elastic CCA and the proximal ICA remain relatively intact.
Crucially, the development of CBNS should not be viewed as an isolated or incidental angiographic phenomenon, but rather as an imaging biomarker that reflects extensive extracranial and skull base vascular remodeling in response to progressive intracranial ischemia. It is important to consider the additional clinical value of the CBNS grading system over established disease markers, such as the Suzuki stage. While the Suzuki stage directly reflects intracranial steno-occlusive changes, our multivariate analysis clearly demonstrated that a higher CBNS grade remains independently associated with hemodynamic compromise even after adjusting for the Suzuki stage. This independence strongly suggests that CBNS formation does not merely mirror the progression of intracranial vascular remodeling and collateral adaptation. Previous literature has well-documented that progressive intracranial steno-occlusion triggers prominent morphological adaptations in the external carotid artery branches and skull base structures [1]. For instance, morphological changes in the foramen spinosum and the enlargement of the middle meningeal artery directly correlate with ischemia progression and spontaneous trans-dural collateral development [18]. In line with these established findings on progression and collateral development in the course of MMD, “plump enlargement” captured by the CBNS grading system may represent a hemodynamic adaptation occurring at the extracranial level.
Furthermore, the conceptual novelty of our system lies in the inclusion of the “plump enlargement” of the cervical ICA, moving beyond the simple ICA shrinkage described in previous literature [2–4]. Although our cross-sectional analysis did not show a direct correlation between CBNS grades and specific symptomatic presentation types-likely due to a selection bias where most catheterized patients were already symptomatic-CBNS may still serve as a convenient, adjunctive imaging biomarker. It can alert clinicians to potential hemodynamic impairment directly from routine angiograms before functional perfusion data like SPECT is obtained, thereby refining initial risk stratification.
This study has some limitations. First, the study cohort was relatively small due to inclusion from a single institution. Second, the definition of plump enlargement was based on qualitative assessment. Establishing objective, quantitative parameters is essential to further reduce interobserver variability and enhance the clinical utility of this classification. To address this limitation, we are currently undertaking a quantitative validation study using geometric measurements, including diameter ratios, curvature indices, and cross-sectional area parameters of the cervical ICA. Integrating these objective metrics with our visual grading system will be a key focus of our future work to establish a more robust and standardized diagnostic framework. Third, hemodynamic evaluation using SPECT was not uniformly performed across all included patients, particularly within the non-MMD cohort, which introduces a potential selection bias. In our clinical workflow, preoperative SPECT was routinely obtained for stable MMD patients as part of their comprehensive evaluation. However, many patients in the non-MMD group (those with ICAO/ICAS or MCAO/MCAS) presented with acute or crescendo ischemic strokes that demanded emergent surgical revascularization, leaving no time for preoperative hemodynamic imaging. As a result, the non-MMD patients who received SPECT may represent a biased subpopulation in a more stable or chronic state, potentially underestimating the overall hemodynamic compromise of the non-MMD group. This non-uniform evaluation limits our ability to definitively compare hemodynamic correlations across both cohorts. Future prospective studies utilizing a standardized imaging protocol for all patients are warranted to minimize this selection bias and validate the generalizability of the CBNS grading system. Fourth, angiographic features and hemodynamic changes might not be entirely independent between the hemispheres. Development of collateral vessels is one of the hallmarks in MMD as the steno-occlusive changes progress. Collateral vessels include leptomeningeal anastomoses from other major arteries, periventricular anastomoses (known as moyamoya vessels), and anterior and posterior communicating arteries [1]. Most of the collateral development originates from ipsilateral systems. Consequently, many previous studies analyzing patients with MMD have examined each hemisphere separately [9–11]. In general, the angiographic progression and hemodynamic changes can be assessed independently in each hemisphere. Fifth, statistical analysis treated each side as an independent unit of observation (n = 187 sides from 98 patients), which introduces a potential within-patient correlation. Pathophysiologically, the angiographic progression and collateral remodeling captured by the Suzuki staging system and CBNS occur predominantly on an ipsilateral basis, often resulting in marked asymmetry between the right and left sides within the same individual. Nevertheless, from a strictly statistical perspective, bilateral data from the same patient may share systemic genetic or environmental background factors and cannot be considered entirely independent. This limitation may influence the statistical power of conventional comparative tests. To further validate our findings and fully control for this hierarchical data structure, future large-scale studies utilizing advanced statistical modeling, such as generalized estimating equations (GEE) or mixed-effects models are warranted. Collectively, future prospective case-control studies with larger, multicenter cohorts are warranted to validate the diagnostic and screening utility of CBNS in the clinical practice of MMD.
Conclusion
We proposed novel CBNS grading system for patients with MMD based on the lateral view of CCA angiograms. CBNS grade 2, which reflects plump enlargement of the bulbous portion of the cervical ICA, may serve as an imaging biomarker of hemodynamic impairment in patients with MMD. Identification of CBNS grade 2 may assist in estimating preliminary hemodynamic status prior to obtaining SPECT imaging.
Author contributions
Author Contributions Conceptualization: R. T., M. F., H. E.; Methodology: R. T., A. K., M. F., H. E.; Formal analysis and investigation: R. T., A. K., H. S., H. U., S. O., H. E.; Writing-original draft preparation: R. T.; Writing-Review and editing: A. K., M. F., H. E.; Funding acquisition: H. E.; Supervision: M.F., H. E.
Funding
This study was supported by the Japan Ministry of Health, Labour and Welfare (MHLW) grant (26FC1015).
Data availability
The Data supporting the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author upon reasonable request.
Declarations
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Suzuki J, Takaku A (1969) Cerebrovascular moyamoya disease. Disease showing abnormal net-like vessels in base of brain. Arch Neurol 20(3):288–299 [DOI] [PubMed] [Google Scholar]
- 2.Yasuda C, Arakawa S, Shimogawa T, Kanazawa Y, Sayama T, Haga S, Morioka T (2016) Clinical Significance of the Champagne Bottle Neck Sign in the Extracranial Carotid Arteries of Patients with Moyamoya Disease. AJNR Am J Neuroradiol 37(10):1898–1902 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Yasaka M, Ogata T, Yasumori K, Inoue T, Okada Y (2006) Bottle neck sign of the proximal portion of the internal carotid artery in moyamoya disease. J Ultrasound Med 25(12):1547–1552 quiz 1553 – 1544 [DOI] [PubMed] [Google Scholar]
- 4.Wang J, Chen G, Yang Y, Zhang B, Jia Z, Gu P, Wei D, Ji J, Hu W, Zhao X (2018) Association Between Champagne Bottle Neck Sign of Internal Carotid Artery and Ipsilateral Hemorrhagic Stroke in Patients with Moyamoya Disease. World Neurosurg 118:e18–e24 [DOI] [PubMed] [Google Scholar]
- 5.Kuroda S, Fujimura M, Takahashi J, Kataoka H, Ogasawara K, Iwama T, Tominaga T, Miyamoto S, Research Committee on Moyamoya Disease of the Ministry of Health L, Welfare J (2022) Diagnostic Criteria for Moyamoya Disease – 2021 Revised Version. Neurol Med Chir (Tokyo) 62(7):307–312 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chimowitz MI, Lynn MJ, Howlett-Smith H, Stern BJ, Hertzberg VS, Frankel MR, Levine SR, Chaturvedi S, Kasner SE, Benesch CG, Sila CA, Jovin TG, Romano JG, Warfarin-Aspirin Symptomatic Intracranial Disease Trial I (2005) Comparison of warfarin and aspirin for symptomatic intracranial arterial stenosis. N Engl J Med 352(13):1305–1316 [DOI] [PubMed] [Google Scholar]
- 7.Pezzini A, Del Zotto E, Archetti S, Negrini R, Bani P, Albertini A, Grassi M, Assanelli D, Gasparotti R, Vignolo LA, Magoni M, Padovani A (2002) Plasma homocysteine concentration, C677T MTHFR genotype, and 844ins68bp CBS genotype in young adults with spontaneous cervical artery dissection and atherothrombotic stroke. Stroke 33(3):664–669 [DOI] [PubMed] [Google Scholar]
- 8.Ryoo S, Cha J, Kim SJ, Choi JW, Ki CS, Kim KH, Jeon P, Kim JS, Hong SC, Bang OY (2014) High-resolution magnetic resonance wall imaging findings of Moyamoya disease. Stroke 45(8):2457–2460 [DOI] [PubMed] [Google Scholar]
- 9.Kuroda S, Yamamoto S, Funaki T, Fujimura M, Kataoka H, Hishikawa T, Takahashi J, Endo H, Nariai T, Osato T, Saito N, Sato N, Hori E, Ito YM, Miyamoto S, Group AS (2023) Five-Year Stroke Risk and Its Predictors in Asymptomatic Moyamoya Disease: Asymptomatic Moyamoya Registry (AMORE). Stroke 54(6):1494–1504 [DOI] [PubMed] [Google Scholar]
- 10.Kuroda S, Hara S, Funaki T, Araki Y, Saito R, Hamano E, Kataoka H, Murai S, Hishikawa T, Ito M, Uchino H, Fujimura M, Tashiro R, Kanoke A, Endo H, Yamaguchi K, Nomura S, Kawashima A, Kawamata T, Yamamoto S, Kashiwazaki D, Hori E, Nishimura A, Yoshimoto K, Takagi Y, Miyawaki S, Saito N, Takahashi JC, Maehara T, Miyamoto S (2026) Unique clinical and radiological features in patients with moyamoya disease under 4 years of age: a multicenter MACINTOSH Study in Japan. J Neurosurg Pediatr 37(5):429–439 [DOI] [PubMed] [Google Scholar]
- 11.Tashiro R, Fujimura M, Katsuki M, Nishizawa T, Tomata Y, Niizuma K, Tominaga T (2021) Prolonged/delayed cerebral hyperperfusion in adult patients with moyamoya disease with RNF213 gene polymorphism c.14576G > A (rs112735431) after superficial temporal artery-middle cerebral artery anastomosis. J Neurosurg 135(2):417–424 [DOI] [PubMed] [Google Scholar]
- 12.Koga M, Kimura K, Minematsu K, Yasaka M, Isa K, Yamaguchi T (2000) Transoral carotid ultrasonographic findings in internal carotid artery dissection–a case report. Angiology 51(8):699–703 [PubMed] [Google Scholar]
- 13.Kishikawa K, Kamouchi M, Okada Y, Inoue T, Ibayashi S, Iida M (2002) Evaluation of distal extracranial internal carotid artery by transoral carotid ultrasonography in patients with severe carotid stenosis. AJNR Am J Neuroradiol 23(6):924–928 [PMC free article] [PubMed] [Google Scholar]
- 14.Kamouchi M, Kishikawa K, Okada Y, Inoue T, Ibayashi S, Iida M (2005) Poststenotic flow and intracranial hemodynamics in patients with carotid stenosis: transoral carotid ultrasonography study. AJNR Am J Neuroradiol 26(1):76–81 [PMC free article] [PubMed] [Google Scholar]
- 15.Kishikawa K, Kamouchi M, Okada Y, Inoue T, Ibayashi S, Iida M (2004) Transoral carotid ultrasonography as a diagnostic aid in patients with severe carotid stenosis. Cerebrovasc Dis 17(2–3):106–110 [DOI] [PubMed] [Google Scholar]
- 16.Hori E, Hayashi N, Hamada H, Masuoka T, Kuwayama N, Hirashima Y, Origasa H, Ohtani O, Endo S (2008) A development of atheromatous plaque is restricted by characteristic arterial wall structure at the carotid bifurcation. Surg Neurol 69(6):586–590 discussion 590 – 581 [DOI] [PubMed] [Google Scholar]
- 17.Janzen J, Lanzer P, Rothenberger-Janzen K, Vuong PN (2001) Variable extension of the transitional zone in the medial structure of carotid artery tripod. Vasa 30(2):101–106 [DOI] [PubMed] [Google Scholar]
- 18.Matsukawa H, Fujii M, Murakata A, Shinoda M, Takahashi O (2015) Foramen spinosum and middle meningeal artery in moyamoya disease: Preliminary results of a pilot study. Brain Inj 29(10):1246–1251 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The Data supporting the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author upon reasonable request.