Diagnostic Values of the “To and Fro” Conflict Sign on Intraoperative Indocyanine Green Video Angiography as a Warning Sign of the Focal Cerebral Hyperperfusion and Watershed Shift Phenomenon after Superficial Temporal Artery-Middle Cerebral Artery Bypass for Adult Patients with Moyamoya Disease Open Access

The superficial temporal artery (STA)-middle cerebral artery (MCA) bypass is the standard surgical treatment for ischemic-onset and posterior hemorrhage-onset adult moyamoya disease (MMD) [1‒3]. Despite favorable long-term outcomes, the focal cerebral hyperperfusion (CHP) syndrome is one of the major complications of STA-MCA bypass for MMD, which shows focal hyperperfusion at the site of anastomosis [4‒8]. The focal CHP can result in transient focal neurological deficits, local vasogenic edema, and fatal intracerebral hemorrhage [4‒8]. Despite recent advancements in perioperative management, delayed intracerebral hemorrhage due to the focal CHP remains a major clinical challenge [9, 10]. In addition, hemodynamic studies revealed that the focal CHP could accompany a paradoxical decrease in cerebral blood flow (CBF) adjacent to the vascular territory of the bypass, which can result in transient neurological deficits and infarctions [11]. This hemodynamic ischemia near the focal CHP was termed as the watershed shift (WS) phenomenon [11], which is evident in approximately 10% of adult MMD patients who underwent STA-MCA bypass [11]. Although various risk factors for CHP and the WS phenomenon have been reported (online suppl. Tables 1, 2; for all online suppl. material, see https://doi.org/10.1159/000546826), such as elderly, left side, increased cerebrovascular volume, STA/MCA mismatch, and impaired washout [4, 5, 12, 13], preoperative prediction of the focal CHP and WS phenomenon remains difficult.

Previous studies have investigated the usefulness of intraoperative indocyanine green video angiography (ICG-VA) for predicting the focal CHP, reporting longer ICG peak time [12], increased blood flow index [14], reduced microvascular transit time [15], and longer washout time of ICG [16] in MMD patients with the focal CHP compared with those without. However, it is difficult to recognize the risk of the focal CHP intraoperatively using these analytical methods. In this study, we investigated the predictive value of the “to and fro” conflict sign, which represents conflicting blood flow between donor and recipient arteries, for the development of the focal CHP and coexisting WS phenomenon.

This is a retrospective case-control study.

We assessed 109 consecutive patients with MMD (121 hemispheres, aged 19–72 years, mean age 44.2) who underwent STA-MCA (M4) anastomosis combined with encephalo-duro-myo-synangiosis by a single surgeon (M.F.) between July 2017 and October 2019 were assessed for eligibility. 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 [3]. Among the 109 patients (121 hemispheres), 6 pediatric patients (9 hemispheres), 7 patients (7 hemispheres) without quantitative ¹²³I-IMP-SPECT data during the preoperative period (online suppl. Fig. 1), 3 patients (3 hemispheres) who showed atypical magnetic resonance (MR) findings (e.g., the transient equivocal signal intensity of STA-MCA bypass on MR angiography [MRA] during the perioperative period) were excluded from this study. Finally, 93 patients (102 hemispheres) were enrolled in this study (Fig. 1).

STA-MCA bypass combined with encephalo-duro-myo-synangiosis was performed under general anesthesia as described previously [6‒8, 17]. In short, a 6-cm craniotomy was performed around the end of the Sylvian fissure. The stump of the STA frontal or parietal branch was anastomosed to the M4 segment of the MCA. We drilled out the inner layer of the bone to avoid compressing the brain surface by the temporal muscle. Intraoperative ICG-VA was performed using an OPMI Pentero 900 surgical microscope (Carl Zeiss Co., Oberkochen, Germany) equipped with integrated ICG technology. A standard dose of 10 mg ICG dye diluted in 4 mL of water was injected via a peripheral vein. The ICG signal was visualized once the ICG dye reached the brain. The “to and fro” conflict sign was defined as a constant alternating forward and backward movement in the cortical branches of the MCA within the operative field within 3 cm of the anastomosis site on intraoperative ICG-VA. Three independent neurosurgeons (M.F., R.T., and T.N.) assessed the ICG-VA findings. The patency of the STA-MCA anastomosis was confirmed intraoperatively using Doppler ultrasonography.

The focal CHP and the WS phenomena were diagnosed based on previously described definition [6‒8]. Perioperative CBF values, measured preoperatively and on postoperative days (PODs) 1 and 7, were quantified using the autoradiographic method by ¹²³I-IMP SPECT on an Infinia II (GE Healthcare, Japan). CBF was quantitatively measured by manually defining a region of interest (ROI) with a 1-cm diameter in the vascular territory supplied by the bypass and in the ipsilateral cerebellar hemisphere (online suppl. Fig. 1), following our established method [6‒8]. The definition of the focal CHP included all of the following criteria: (1) a greater than 150% increase in CBF at the vascular territory supplied by the bypass, (2) obvious visualization of the STA-MCA bypass on MRA, (3) the absence of other pathologies, including compression of the brain surface by the temporal muscle inserted for indirect bypass (online suppl. Fig. 2). In addition, the WS phenomenon was defined as a decrease in CBF (postoperative CBF/preoperative CBF <1) in the cortex adjacent to the CHP site (online suppl. Fig. 3).

All patients were prospectively subjected to a standardized intensive management strategy, as we described previously [5‒7, 17, 18]. Shortly, prophylactic antihypertensive medications were administered to maintain a systolic blood pressure of 100–130 mm Hg to prevent symptomatic focal CHP. Minocycline hydrochloride (200 mg/day) was administered intraoperatively and postoperatively until POD 7 to prevent unfavorable CHP outcomes. Based on the temporal changes in ¹²³I-IMP-SPECT and MRI findings, a gradual return to normotensive blood pressure was permitted between POD 7 and 10.

A logistic regression model was employed to calculate odds ratio (ORs) with 95% confidence interval (CI) to investigate the association between focal CHP/WS phenomenon and various clinical characteristics, such as age, sex, side, mode of onset, Suzuki angiographic stage, preoperative CBF values, the presence of “to and fro” conflict sign, and recipient artery diameter. Additionally, a multiple logistic regression model was utilized to explore the association among the focal CHP/WS phenomenon, “to and fro” conflict sign, age, sex, and relevant clinical factors that had p values below 0.05 in univariate analysis. Two-sided p values below 0.05 and 95% CIs that did not include 1.0 were considered statistically significant. All statistical analyses were conducted by JMP Student Edition ver 18 (SAS Institute Inc., Cary, NC, USA).

The association between perioperative hemodynamic changes and clinical features, including the “to and fro” conflict sign, is outlined in Tables 1 and 2. Specifically, Table 1 highlights the association between the “to and fro” conflict sign and perioperative hemodynamics. The incidence of the focal CHP and WS phenomenon was 29.4% (30/102) and 10.8% (11/102), respectively. The “to and fro” conflict sign was evident in 40.0% (12/30) of those with MMD exhibiting focal CHP and 63.6% (7/11) demonstrating the WS phenomenon. In contrast, only 2.8% (2/72) and 7.7% (7/91) of those with MMD who did not present with the focal CHP and WS phenomenon showed a “to and fro” conflict sign. The positive predictive values of the “to and fro” conflict sign were 85.7% (12/14) and 50.0% (7/14), respectively. First, a logistic regression model revealed the focal CHP/WS phenomenon and clinical features, including “to and fro” conflict sign. The “to and fro” conflict sign was significantly associated with both the focal CHP (p = 0.002, OR 26.3, 95% CI: 2.96–558.0) and the WS phenomenon (p = 5.3 × 10⁻⁴, OR 33.7, 95% CI 4.34–486.2). A multiple regression analysis was performed to explore the association between the focal CHP/WS phenomenon and the “to and fro” conflict sign while controlling for confounding factors, which revealed strong association between “to and fro” conflict sign and the development of the focal CHP (p < 0.001)/WS phenomenon (p < 0.001). Representative cases are shown in online supplementary materials.

This study identified the “to and fro” conflict sign in intraoperative ICG-VA, which represents the conflicting blood flow between donor and recipient arteries, in 36.7% (11/30) and 54.5% (6/11) of adult MMD patients with the focal CHP and the WS phenomenon, respectively. In contrast, only 2.7% (2/73) and 7.6% (7/92) MMD patients without the focal CHP and WS phenomenon showed the “to and fro” conflict sign. The “to and fro” conflict sign in intraoperative ICG-VA was significantly associated with the focal CHP and the WS phenomenon, highlighting its potential as a predictive marker for the “to and fro” conflict sign for these complications after STA-MCA bypass for adult patients with MMD. One of the advantages of the “to and fro” conflict sign is its ease of intraoperative recognition, in contrast to the ICG time-intensity curve analysis, which requires postoperative assessment. A previous study also demonstrated that the ICG-VA effectively detects early venous filling and arteriovenous shunts after revascularization surgery for MMD [19‒21]. Therefore, the “to and fro” conflict sign serves as a valuable intraoperative warning for the potential development of the focal CHP and WS phenomenon after STA-MCA bypass surgery in adult patients with MMD.

The mechanisms underlying the focal CHP and WS phenomenon remain poorly understood. These conditions may be explained by conflicting blood flow between the donor artery and the intrinsic antegrade flow of the MCA (Fig. 2). Ideally, the flow from the donor artery is distributed extensively immediately after anastomosis. However, when the antegrade flow of the recipient artery conflicts with the flow from the donor artery, the bypass flow is restricted to a limited area near the anastomosis site due to flow instability by competitive blood flow between donor and recipient arteries, resulting in the focal CHP. However, the development of the focal CHP and WS phenomenon cannot be explained simply by the competition between the donor and recipient arterial flow. Localized donor arterial flow due to increased vascular resistance caused by impaired autoregulatory mechanisms and impaired venous reflux could be the key element for the development of the focal CHP. This localization of the donor arterial flow can result in competition between the donor and recipient arterial flows, which can be observed as the “to and fro” conflict sign. Our previous study supported this hypothesis, by demonstrating an association between decreased peripheral vascular resistance and the focal CHP [22]. Furthermore, preexisting collaterals affect retrograde recipient arterial flow, thereby being critical modifying factors for the focal CHP and WS phenomenon. Also, the angulation of the donor artery might be the key element for the development of the focal CHP and WS phenomenon (online suppl. Fig. 4). The distribution of the donor artery’s flow relies on the balance among residual antegrade flow, retrograde flow, and the donor artery flow. If the angle between the donor artery and the recipient artery is greater than 90 degrees, the flow tends to favor the proximal portion of the M4 segment, avoiding distribution to distal portion (online suppl. Fig. 4a). Conversely, when the angle is less than 90 degrees, the donor artery flow is more likely to be directed toward the distal part of the M4 segment (online suppl. Fig. 4b). In this scenario, the conflict between the donor artery flow and retrograde flow may become more pronounced. In our case series, the angulation of the donor artery relative to the recipient artery was nearly vertical. We did not observe any correlation between the angulation of the donor artery and the focal CHP/WS phenomenon. Collectively, the “to and fro” conflict sign on ICG-VA post-anastomosis for MMD reflects a failure to adapt to rapid hemodynamic changes caused by STA-MCA bypass, which may reflect the impaired autoregulatory mechanisms, microcirculation, and venous reflux in patients with MMD.

A combination of various risk factors and diagnostic tools may help in the preoperative and/or intraoperative prediction of postoperative focal CHP and the WS phenomenon. Several epidemiological features are associated with the development of the focal CHP, such as older age and adult-onset cases [23], hemorrhagic onset [23], and the involvement of the dominant side or left hemisphere [5]. Preoperative ¹²³I-IMP SPECT can also identify risk factors for the focal CHP and the WS phenomenon, with increased cerebral blood volume and reduced CBF associated with the focal CHP and WS phenomenon, respectively [4, 11]. Additionally, a decreased cerebrovascular reserve is associated with delayed CHP [24]. Evaluating angioarchitecture using MRA and catheter angiograms may further help identify the risk of the focal CHP. Zhang et al. [25] identified that the anterograde flow of perisylvian arteries was linked to the focal CHP. Our group has also reported the utility of 3D time-of-flight MRA [7]. By calculating hemispheric MRA scores based on the signal intensities of major intracranial arteries (anterior, middle, and posterior cerebral arteries), we found that a higher MRA score, indicating decreased signal intensities of major intracranial arteries, was associated with the focal CHP development. Intraoperative findings are critical for assessing the risk of the focal CHP and WS phenomenon. Microscopic findings of redness of the superficial cortical vein, also termed as arterialized vein, just after STA-MCA anastomosis can be found in MMD patients who exhibited the focal CHP [19‒21]. This early venous filling or arterialized vein may be caused by an arteriovenous shunt [20]. Our group also demonstrated that the waveform analysis of the STA-MCA bypass graft using a flowmeter to measure time from peak to 50% decay is associated with vascular resistance and can predict the focal CHP [22]. Genetic testing has also shown that the RNF213 c.14576G>A (rs112735431, p.R4810K) polymorphism, a susceptibility variant of MMD, is associated with prolonged and delayed CHP [8]. Therefore, multi-modality diagnostic tools could help establish the predictive measures for the development of focal CHP and the WS phenomenon.

This study has several limitations. First, the manual definition of ROIs in ¹²³I-IMP SPECT may introduce bias. Despite efforts to detect hemodynamic ischemia due to the WS phenomenon by referring “to and fro” conflict sign on ICG-VA, the incidence of the WS phenomenon might be underestimated due to the manual definition of small ROIs near the CHP site. Second, the “to and fro” conflict sign is not capable of predicting the focal CHP and WS phenomenon in all cases. For instance, the intraoperative ICG-VA cannot detect the “to and fro” conflict sign if the conflict between the bypass flow and MCA flow occurs outside the operative field [18]. In addition, the “to and fro” conflict sign may be found around the anastomosis site, even if it is not necessarily adjacent to it. Third, perioperative changes around the anastomosis site are not fully understood due to the lack of pre-anastomosis ICA-VA. Also, comparison of postoperative catheter angiograms would facilitate the understanding of how the “to and fro” conflict between the donor and recipient arteries affects the postoperative hemodynamic status. Fourth, this is a single-centered, retrospective study. Further studies are warranted to validate the prognostic value of the “to and fro” conflict sign on intraoperative ICG-VA to predict the focal CHP and WS phenomenon after STA-MCA anastomosis for MMD.

The authors wish to thank Editage (www.editage.com) for the English language editing.

This study protocol was reviewed and approved by the Ethics Committee of Kohnan Hospital (Sendai, Japan), approval No. 2020-0520-03. Written informed consent was obtained from the participants to participate in the study. Written informed consent was obtained from the individuals who were presented as representative cases.

Miki Fujimura was a member of the journal’s editorial board at the time of submission. The other authors have no interest to declare.

This study was supported by the Japan Society for the Promotion of Science Grant-in-Aid for Scientific Research A (22K16675, 23K08491, 23H00433). The funder had no role in the design, data collection, data analysis, and reporting of this study.

Conception and design: M.F. and R.T. Acquisition of data: M.F. Analysis and interpretation of ICG video angiogram: R.T., M.F., T.N., and K.T. Statistical analysis and drafting: R.T. Critical revision of the article: M.F., T.N., K.T., A.K., and H.E. Study supervision: H.E. All authors reviewed the manuscript.

All data generated or analyzed during this study are included in this article and its supplementary material files. Further inquiries can be directed to the corresponding author.