Association of Venous Outflow Profiles and Successful Vessel Reperfusion After Thrombectomy

Abstract

Objective

Robust arterial collaterals are associated with successful reperfusion after thrombectomy treatment of acute ischemic stroke due to large vessel occlusion (AIS-LVO). Excellent venous outflow (VO) reflects excellent tissue perfusion and collateral status in patients with AIS-LVO. The goal of this study was to determine whether favorable VO profiles assessed on pretreatment CT angiography (CTA) images correlate with successful vessel reperfusion after thrombectomy in patients with AIS-LVO.

Methods

This was a multicenter retrospective cohort study of consecutive patients with AIS-LVO treated by thrombectomy. Baseline CTA was used to assess collateral status (Tan scale) and VO using the Cortical Vein Opacification Score (COVES). Favorable VO was defined as COVES ≥3. Primary outcome was excellent vessel reperfusion status (modified Thrombolysis in Cerebral Infarction 2c/3). Secondary outcome was good functional outcome defined as a score of 0 to 2 on the modified Rankin Scale after 90 days.

Results

Five hundred sixty-five patients met the inclusion criteria. Multivariable logistic regression analysis showed that favorable VO (odds ratio [OR] 2.10 [95% confidence interval (CI) 1.39–3.16]; p < 0.001) was associated with excellent vessel reperfusion during thrombectomy, regardless of good CTA collateral status (OR 0.87 [95% CI 0.58–1.34]; p = 0.48). A favorable VO profile (OR 8.9 [95%CI 5.3–14.9]; p < 0.001) and excellent vessel reperfusion status (OR 2.7 [95%CI 1.7–4.4]; p < 0.001) were independently associated with good functional outcome adjusted for age, sex, glucose, tissue plasminogen activator administration, good CTA collateral status, and presentation NIH Stroke Scale score.

Conclusion

A favorable VO profile is associated with reperfusion success and good functional outcomes in patients with AIS-LVO treated by endovascular thrombectomy.

Classification of Evidence

This study provides Class II evidence that a favorable VO profile is associated with reperfusion success and good functional outcomes in patients with AIS-LVO treated by endovascular thrombectomy.

Thrombectomy is an effective treatment for acute ischemic stroke due to large vessel occlusion (AIS-LVO),^1-4 and robust arterial collaterals have been associated with successful vessel reperfusion after treatment.^5,6 Arterial collaterals are typically assessed on CT angiography (CTA), but favorable CTA collaterals do not reflect tissue perfusion. It is most likely that the critical patterns of hypoperfusion are determined by the most distal arterial branches and associated venous microvascular changes.⁷

Robust collateral blood flow and tissue perfusion may be assessed with vascular imaging biomarkers of venous outflow (VO). Effective venous egress from ischemic brain implies robust pial arterial collaterals that deliver blood to the brain and effective transit of blood through the ischemic brain tissue.^8-11 However, whether favorable VO is a determinant of successful vessel reperfusion in patients treated with thrombectomy has not been described.

We hypothesized that favorable VO profiles would be associated with successful vessel reperfusion rates after mechanical thrombectomy. We further hypothesized that a favorable VO is associated with good functional outcomes after 90 days. We tested these hypotheses by determining the correlation between VO and excellent vessel reperfusion status (modified Thrombolysis in Cerebral Infarction (TICI) 2c/3 reperfusion) and functional outcomes in patients with AIS-LVO treated with thrombectomy.

Methods

Classification of Evidence

This study aimed to assess whether favorable VO profiles in patients with AIS-LVO are associated with excellent vessel reperfusion status after mechanical thrombectomy and whether favorable VO profiles are associated with good functional outcomes after thrombectomy treatment (Class II evidence).

Study Design

We performed a multicenter retrospective cohort study of consecutive patients undergoing thrombectomy triage for AIS-LVO treatment at 2 comprehensive stroke centers (University Medical Center Hamburg-Eppendorf, Germany, and Stanford University Hospital, CA) between January 2013 and December 2019.

Standard Protocol Approvals, Registrations, and Patient Consents

The study protocol was approved by the institutional review boards of both study centers, complied with the Health Insurance Portability and Accountability Act, and followed the guidelines of the Declaration of Helsinki. Patient informed consent was waived by our review boards for this retrospective study.

Patient Inclusion and Population and Clinical Data

Patients were identified from prospectively maintained stroke databases at each center, and clinical, imaging, and demographic data were obtained from the electronic medical records.

Patients were included if they had (1) AIS-LVO and underwent thrombectomy treatment within 16 hours after stroke onset; (2) baseline head noncontrast CT (NCCT) and CTA with uniform opacification of the superior sagittal, transverse, and sigmoid sinuses to the level of the jugular bulb to allow VO determination; (3) interpretable CT or magnetic resonance perfusion imaging; (4) anterior circulation LVO of the internal carotid artery or first (M1) or second (M2) segment of the middle cerebral artery; or (5) interpretable digital subtraction angiography images that allowed final TICI score assignment after thrombectomy.

Exclusion criteria were (1) poor CTA image quality due to excessive patient motion or incomplete opacification of the dural venous sinuses to allow VO determination, as described above; (2) poor CT perfusion or magnetic resonance perfusion image quality due to excessive motion or failed contrast bolus; and (3) absence of at least 1 follow-up head NCCT/MRI examination within 72 hours of thrombectomy or hospital admission. An overview of the specific number of patients who were included or excluded from our analysis is provided in supplemental figure 1 (data available from Dryad: doi.org/10.5061/dryad.3j9kd51h2).

Imaging Analysis

The specific CT imaging protocols are listed in the supplemental material (additional supplemental Methods; data available from Dryad: doi.org/10.5061/dryad.3j9kd51h2). All CT and magnetic resonance perfusion studies were automatically analyzed with RAPID (iSchemaView, Menlo Park, CA). The ischemic core was defined as the volume of tissue with a 70% reduction in cerebral blood flow relative to the contralateral cerebral hemisphere on CT perfusion or the volume of tissue with restricted diffusion (apparent diffusion coefficient <620 ×10⁻⁶ mm²/s) on diffusion-weighted MRI.

VO was determined by the Cortical Vein Opacification Score (COVES) on single-phase CTA images, which quantifies venous opacification of the vein of Labbé, sphenoparietal sinus, and superficial middle cerebral vein on CTA as follows: 0 = not visible, 1 = moderate opacification, and 2 = full opacification (figure 1).¹² Therefore, COVES ranges from 0 (no opacification of the 3 venous pathways) to 6 (full opacification of the 3 venous pathways). Favorable VO (VO+) was defined as COVES of 3 to 6, and unfavorable VO (VO−) was regarded as a COVES of 0 to 2 according to previously published thresholds.¹³ Initially, VO was determined independently by 2 neuroradiologists (T.D.F., 8 years of experience; and J.J.H, 10 years of experience) on a subset of studies to determine interreader agreement and then assigned on the remaining studies by a single neuroradiologist (T.D.F.).

Figure 1. Assessment of COVES.

The 3 images exemplify the assessment of the Cortical Vein Opacification Score (COVES) on single-phase CT angiography. COVES is assessed by evaluating the contrast opacification within 3 major cortical veins representative for the drainage of the middle cerebral artery territory, namely the (A) superficial middle cerebral vein, (B) vein of Labbé, and (C) sphenoparietal sinus. COVES is a standardized 6-point score, ranging from 0 (no opacification), over 1 (moderate opacification) to 2 (full opacification), determined in each of the target veins. Green arrows in A–C point to the respective cortical veins on the left hemisphere, which showed full contrast opacification and consequently a COVES of 6.

All initial and final TICI scores were interpreted independently (by T.D.F and J.J.H.) on digital subtraction angiography images, and discrepancies were resolved by a consensus read. Excellent vessel reperfusion after thrombectomy was defined as TICI 2c/3 (95%–100% revascularization).¹⁴

The Alberta Stroke Program Early CT Score (ASPECTS)¹⁵ was determined on pretreatment NCCT images by a single reader (T.D.F.).

Arterial collaterals were assessed on CTA by consensus by 2 neuroradiologists (T.D.F. and J.J.H.) using the modified Tan scale,¹⁶ and good collaterals were defined as filling of ≥50% of the middle cerebral artery territory, whereas poor collaterals were defined as filling of <50% of the middle cerebral artery territory.

Final infarct volume was acquired by manual segmentation on follow-up NCCT images or b1000 diffusion-weighted MRIs using the software package Horos (Horos Project, version 3.3.6) by a single neuroradiologist (T.D.F.). Final infarct volumes were acquired on NCCT images 48 to 72 hours after thrombectomy.

Both readers were blinded to the clinical and outcome information of each patient during the scoring process.

Outcome Measures

Primary outcome was excellent vessel reperfusion status (TICI 2c/3). Secondary outcome was a good functional outcome 90 days after treatment (modified Rankin Scale [mRS] score 0–2). The mRS score is a standardized 6-point score that measures functional independency of patients with acute stroke 90 days after treatment¹⁷ and was obtained from each patient by a stroke neurologist or specialized study nurse.

Statistical Analysis

Patient demographics, clinical variables, and neuroimaging data were compared between 2 groups with the χ² and Wilcoxon rank-sum tests. Clinical and imaging variable association with excellent vessel reperfusion status and good functional outcome was assessed with multivariable binary logistic regression models that were fitted to the primary and secondary outcomes. The regression models were adjusted for prespecified factors. The model for primary outcome was adjusted for age, good CTA collaterals, administration of IV tissue plasminogen activator (tPA), and favorable VO. The secondary outcome regression model was adjusted for age, excellent vessel reperfusion status after thrombectomy, admission blood glucose, favorable VO, and presentation NIH Stroke Scale (NIHSS) score. In addition, we adjusted the models for the baseline characteristics that were imbalanced between VO− groups at least at p < 0.1 in the univariate analysis. Additional tests for collinearity were also performed using the c statistic. We set α at the 0.05 level for significance, and all reported results are 2 sided. Statistical analysis was done with IBM SPSS statistics version 26.0 (Armonk, NY).

Data Availability

The data that support the findings of this study and additional data are available from the corresponding author on reasonable request.

Results

A total of 740 patients were evaluated for thrombectomy, and 565 met inclusion criteria (supplementary figure 1; data available from Dryad: doi.org/10.5061/dryad.3j9kd51h2). There were 221 patients in the VO+ group and 344 patients in the VO− group. Interrater agreement for VO determination was substantial (κ = 0.64 [95% confidence interval (CI) 0.46–0.78]).

Compared to VO− patients, VO+ patients were younger (median 72 [IQR 62–80] years vs 76.5 [IQR 66–83.5] years; p < 0.001), were more likely to have received IV tPA (70% vs 40.9%; p < 0.001), and presented with less severe symptoms (median presentation NIHSS score 12 [IQR 7–17] vs 17 [IQR 13–20]; p < 0.001). These results are summarized in table 1.

Table 1.

Patient Demographics

On baseline imaging evaluation, VO+ patients had higher ASPECTS (median 8 [IQR 7–9] vs 7 [IQR 6–8.5]; p < 0.001), smaller baseline ischemic core volumes (median 3 [IQR 0–13] mL vs 19 [IQR 5–49] mL; p < 0.001), and smaller penumbra volumes (median 99 [IQR 58–151] mL vs 140 [IQR 88.5–195] mL; p < 0.001). A higher proportion of VO+ patients exhibited a good CTA collateral status (88% vs 60%; p < 0.001) compared to VO—patients. These findings are shown in table 2.

Table 2.

Baseline Imaging Characteristics

During thrombectomy treatment, the time from groin puncture to vessel reperfusion was shorter in VO+ (median 39 [IQR 22.5–65.5] minutes vs median 58 [31.5–94.5] minutes; p < 0.001) compared to VO− patients. VO+ patients were more likely to achieve excellent vessel reperfusion after thrombectomy (54% vs 39%; p < 0.001). At 90 days after treatment, VO+ patients had significantly better functional outcomes compared to VO− patients (median 1 [IQR 0–3] vs 5 [IQR 3–6]: p < 0.001) These data are presented in table 3.

Table 3.

Thrombectomy and Clinical Outcomes

For the primary outcome analysis, 537 patients were included in a multivariable binary logistic regression model, which was adjusted for imbalances in VO groups and known predictors of vessel reperfusion (table 4). VO+ profiles (odds ratio [OR] 2.10, 95% CI 1.39–3.16; p < 0.001) were independently associated with excellent vessel reperfusion status after thrombectomy (figure 2), although discrimination of outcome by the model was moderate with a c statistic of 0.60 (95% CI 0.55–0.65).

Table 4.

Multivariable Logistic Regression Analysis for Primary Outcome (Excellent Vessel Reperfusion [TICI 2c/3] After Thrombectomy)

Figure 2. Association Between VO and Reperfusion Success.

Two patients with acute ischemic stroke and large vessel occlusion (AIS-LVO) due to occlusion of the M1 segment of the middle cerebral artery (MCA), (A–D) 1 with a favorable venous outflow (VO) profile and (E–H) 1 with an unfavorable VO. Red arrows indicate poor venous microperfusion, and green arrows display robust venous egress on baseline CT angiography imaging. Blue arrows indicate vessel occlusion locations on digital subtraction angiography imaging. In the first patient, favorable VO was found in the (A) superficial middle cerebral vein and (B) vein of Labbé in the ischemic left hemisphere. After 1 retrieval attempt, (C) the initial proximal MCA1 occlusion was (D) successfully reperfused with a Thrombolysis in Cerebral Infarction (TICI) of 3. The second patient exhibited a poor VO profile in the (E) superficial middle cerebral vein and (F) vein of Labbé within the infarctious right hemisphere. (G) The proximal MCA1 thrombus on the right side (H) failed to be fully reperfused, resulting in a final TICI of 0.

In the secondary outcome multivariable binary logistic regression model, both favorable VO+ profiles (OR 8.9, 95% CI 5.3–14.9; p < 0.001) and excellent vessel reperfusion status after thrombectomy (OR 2.7, 95% CI 1.7–4.4; p < 0.001) were independently associated with favorable clinical outcomes (mRS score 0–2) after adjustments (table 5). The model provided a strong discrimination of favorable functional outcomes with a c statistic of 0.89 (95% CI 0.87–0.92).

Table 5.

Multivariable Logistic Regression Analysis for Secondary Outcome (Good Functional Outcome [mRS Score 0–2] After 90 Days)

Discussion

In this study, we found that favorable VO profiles were associated with excellent vessel reperfusion (TICI 2c/3) in patients with AIS-LVO treated by thrombectomy. Patients with favorable VO and excellent vessel reperfusion status were also likely to achieve good functional outcomes after 90 days following thrombectomy treatment. Our study underscores the importance of a greater understanding of the cerebral collateral cascade. A favorable VO profile (VO+) represents excellent cerebral blood flow in the distal-most component of the cerebral collateral cascade, and the finding that VO+ is strongly associated with reperfusion after thrombectomy suggests that blood flow preservation influences the likelihood of reperfusion.

Prior studies have focused on the influence of arterial collaterals on reperfusion after thrombectomy,^5,18-22 and we similarly found that a large proportion of patients with favorable VO exhibited good single-phase CTA collateral profiles. We found favorable VO to be independently associated with excellent vessel reperfusion after thrombectomy regardless of CTA collaterals. This finding is consistent with the theory that venous biomarkers of cerebral blood flow may be more sensitive for the assessment of tissue perfusion and collateral robustness because they reflect blood flow after permeating ischemic brain tissue.^11,12,23,24

We found a notable number of patients with good CTA collaterals to have unfavorable VO profiles. This finding is comparable to prior studies that found that patients with AIS-LVO with favorable CTA collaterals do not necessarily exhibit robust venous egress.^12,25 It has been speculated that blood transit through ischemic brain may be hindered from reaching the VO secondary to impaired autoregulation or other mechanisms.¹² Ultimately, single-phase CTA collateral assessment inherently does not measure tissue perfusion and may not reliably determine pial and parenchymal arteries and arterioles that still allow retrograde blood flow, even when the CTA collateral status appears to be poor.²⁶ Furthermore, this uncoupling of favorable arterial collaterals from VO may also partially explain why some patients with favorable arterial collaterals on CTA fail to achieve successful reperfusion and favorable outcomes after thrombectomy.

Similar to our findings, good collaterals on single-phase CTA predicted neither the success of endovascular therapy nor the functional outcomes of patients with AIS-LVO in the Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke 3 (DEFUSE 3) study,²⁶ which selected patients for thrombectomy on the basis of the presence of a target mismatch profile between the baseline ischemic core and the salvageable penumbra. This similar uncoupling of favorable CTA collaterals from reperfusion and clinical outcome suggests that although pial arterial collaterals are an important tool that predicts thrombectomy success, a more comprehensive understanding of collateral flow to and through the ischemic brain may be a superior predictor of treatment success. The assessment of alternative vascular pathways such as the venous microvascular system or the use of more advanced imaging techniques such as perfusion imaging-based tissue-level collateral assessment may enable more comprehensive assessments of collateral blood flow compared to single-phase CTA.^12,27-29

The mechanism by which favorable VO promotes excellent reperfusion after thrombectomy remains to be determined. A favorable VO reflects robust cerebral blood flow within ischemic brain, which implies more rapid and nonstatic movement of blood within the microvascular circulation.^10-12,23 Slow or static blood flow promotes thrombosis,^30,31 so it is possible that the robust transit of blood through the brain in VO+ patients promotes thrombolysis and prevents distal vessel occlusion after thrombectomy. Furthermore, robust retrograde blood flow immediately distal to the thrombus was found to be associated with vessel reperfusion after thrombectomy.⁵ In addition, it has been presumed that the maintenance of collateral blood flow in the vicinity of the thrombus serves as a strong lytic.^{5,6,20,21,32-34} Thus, we speculate that VO+ patients have more robust collateral blood flow distal to the occlusive thrombus and through the ischemic tissue, which likely may explain the association between favorable venous collaterals and successful vessel reperfusion after thrombectomy. These hypotheses require further investigation.

Our finding that favorable VO profiles were independently associated with good functional outcomes of patients with AIS-LVO treated with thrombectomy is in accordance with prior studies.^10,12,35 Similar to these studies, we found that patients with a favorable VO had higher ASPECTS, lower ischemic core volumes, and less severe initial symptoms (lower NIHSS score), all of which are robust biomarkers associated with good functional outcomes in patients with AIS-LVO.^{10,12,25,27,32,33,36-41} These findings should be further validated in larger, prospective studies.

Our study has several limitations. The retrospective design may introduce bias. The inclusion of patients from 2 comprehensive stroke centers that routinely perform perfusion imaging may limit the generalizability of our results. Venous opacification has been reported to be affected by the selected imaging protocol, including acquisition timing and rate of contrast injection.¹⁰ Although only patients with complete opacification of the sigmoid sinuses were included in our study, measurement of VO with single-phase CTA technique may limit the complete assessment of VO throughout the entire venous phase. The use of perfusion imaging for endovascular treatment triage may introduce selection bias to the enrolled patients and limit the generalizability of our findings.

Robust VO profiles were associated with excellent vessel reperfusion (TICI 2c/3) and functional outcomes in patients with AIS-LVO treated by endovascular thrombectomy. The assessment of alternative blood flow pathways such as the venous microvascular system may adequately reflect the collateral status and tissue microperfusion and may be an additional useful vascular biomarker associated with endovascular treatment success and good functional outcomes.

Glossary

AIS-LVO

acute ischemic stroke due to large vessel occlusion

ASPECTS

Alberta Stroke Program Early CT Score

CI

confidence interval

COVES

Cortical Vein Opacification Score

CTA

CT angiography

mRS

modified Rankin Scale

NCCT

noncontrast CT

NIHSS

NIH Stroke Scale

OR

odds ratio

TICI

Thrombolysis in Cerebral Infarction

tPA

tissue plasminogen activator

VO

venous outflow

Appendix. Authors

Study Funding

The authors report no targeted funding.

Disclosure

T. Faizy and R. Kabiri report no disclosures relevant to the manuscript. S. Christensen reports equity and consulting for iSchemaView. M. Mlynash, G. Kuraitis, and M. Mader report no disclosures relevant to the manuscript. G. Albers reports equity and consulting for iSchemaView and consulting from Medtronic. M. Lansberg reports no disclosures relevant to the manuscript. M. Marks reports equity in ThrombX Medical. J. Fiehler reports grants and personal fees from Acandis, Cerenovus, Microvention, Medtronic, and Stryker; grants from Route 92; and personal fees from Phenox and Penukbra, outside of the submitted work. M. Wintermark reports no disclosures relevant to the manuscript. J. Heit reports consulting for Medtronic and MicroVention and Medical and Scientific Advisory Board membership for iSchemaView. Go to Neurology.org/N for full disclosures.