Simplifying venous outflow: Prolonged venous transit as a novel qualitative marker correlating with acute stroke outcomes

Abstract

Background

Prolonged venous transit (PVT), defined as presence of time-to-maximum $\ge$ 10 s within the superior sagittal sinus (SSS) and/or torcula, is a novel, qualitatively assessed computed tomography perfusion surrogate parameter of venous outflow with potential utility in pretreatment acute ischemic stroke imaging for neuroprognostication. We aim to characterize the correlation between PVT and neurological functional outcomes in thrombectomy-treated patients.

Methods

A prospectively-collected database of large vessel occlusion acute ischemic stroke patients treated with thrombectomy was retrospectively analyzed. Spearman’s rank correlation coefficient and point-biserial correlations were performed between PVT status (i.e., no region, either SSS or torcula, or both), 90-day modified Rankin score (mRS), mortality (mRS 6), and poor functional outcome (mRS 4-6 vs 0-3).

Results

Of 128 patients, correlation between PVT and 90-day mRS ( $\rho$ = 0.35, p < 0.0001), mortality (r = 0.26, p = 0.002), and poor functional outcome (r = 0.27, p = 0.002) were significant.

Conclusion

There is a modest, significant correlation between PVT and severity of neurological functional outcome. Consequently, PVT is an easily-ascertained, qualitative metric that may be useful as an adjunct for anticipating a patient’s clinical course. Future analyses will determine the significance of incorporating PVT in clinical decision-making.

Introduction

Computed tomography perfusion (CTP) characterizes the physiological state of brain parenchyma. ¹ In cerebral ischemia, compromised perfusion dynamics, which drives the extent of infarction, are captured by CTP. ¹ Alongside altered arterial inflow and overall collateral status, venous outflow (VO) is thought to be strongly influential in stroke patient outcomes. ²

We propose a novel surrogate marker of delayed VO, reflecting the status of both superficial and deep VO drainage patterns. We define prolonged venous transit (PVT) as a multi-regional perfusion marker obtained from post-processed time-to-maximum (Tmax) maps derived from CTP. Prolonged transit delay, defined as Tmax $\ge$ 10s, ³ is thought to indicate compromised VO, and it can be qualitatively assessed from Tmax maps within specific venous structures. Superficial VO is approximated by assessing blood flow in the posterior superior sagittal sinus (SSS) ² ; deep VO is approximated by assessing perfusion in the torcula. ⁴ Severity of PVT is characterized by whether it is absent in both the SSS and torcula, is present in one of the regions, or is present in both regions. Consequently, PVT is a novel, qualitative imaging paradigm collectively capturing the statuses of both superficial and VO tracts.

To assess the potential utility of this novel marker, we aim to characterize the association between PVT and neurological functional outcomes in endovascular thrombectomy (EVT)-treated stroke patients.

Methods

De-identified data will be made available upon reasonable request. This study was approved by the Institutional Review Board with waiver of informed consent. This study was performed in accordance with the ethical standards outlined in the 1964 Declaration of Helsinki. The results were reported in accordance to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

A prospectively-recorded registry of acute ischemic stroke patients at three clinical sites was retrospectively reviewed for inclusion with the following criteria: (a) acute ischemic stroke by large vessel occlusion (distal intracranial internal carotid artery, M1 segment of the middle cerebral artery, and proximal M2 segment of the middle cerebral artery); (b) treated successfully with EVT; (c) available pretreatment CTP imaging; and (d) available 90-day modified Rankin score (mRS). Patients were excluded from analysis if significant premorbid disability (mRS >2) was present.

Additional demographic data collected included: age, sex, comorbidities associated with stroke risk (i.e., hypertension, diabetes mellitus, dyslipidemia, coronary artery disease, prior stroke/transient ischemic attack, atrial fibrillation, and tobacco use), admission National Institutes of Health Stroke Scale (NIHSS), premorbid mRS, occlusion location, occlusion laterality, treatment with IV thrombolysis (i.e., alteplase/tenecteplase), successful EVT (modified treatment in cerebral infarction [mTICI] 2 B/2 C/3), and discharge NIHSS.

Ascertaining PVT

Specifications for CTP acquisition are in accordance with the protocol described previously in Yedavalli et al. ⁵ Qualitative Tmax maps were generated using the post-processing function of commercially-available RAPID software (IschemaView, Menlo Park, CA, USA). These maps were retrospectively reviewed by an experienced neuroradiologist to assess whether SSS and/or torcula regions had prolonged venous transit delay of Tmax $\ge$ 10s.

Statistical analysis

Patients were grouped depending on whether PVT was evidenced in (1) neither the SSS nor torcula, (2) either the SSS or torcula, and (3) both the SSS and torcula. The primary outcome measure was the 90-day mRS score. Secondary sensitivity analyses were for the outcomes of 90-day mortality (mRS (6) and poor functional outcomes (90-day mRS 4-6 vs 0-3). Continuous variables were summarized as medians with interquartile range (IQR), and categorical variables were reported as frequencies and percentages. Spearman’s rank correlation coefficient was calculated between PVT and 90-day mRS. Point-biserial correlation coefficients were performed between PVT, mortality, and poor functional outcomes. Distribution of data was visualized with scatterplots with nonparametric densities. Missing data was minimal and not imputed. Significance was prespecified at 0.05. Statistical analyses were performed using JMP v17 (SAS Institute Inc, Carey, NC, USA).

Results

Of 128 patients included, the median age was 72 years (IQR 63-81), and 59.4% were female (Table 1). The most common comorbidity was hypertension (80.5%) (Table 1). Median NIHSS upon admission was 15 (IQR 11–20), and median premorbid mRS was 0 (0-1) (Table 1). The PVT was not present in 68.0% of patients, while 17.2% had PVT in one region only and 14.8% had PVT in both regions (Table 1). All patients achieved successful reperfusion (mTICI 2 B/2 C/3) (Table 1). Median NIHSS upon discharge was 4 (IQR 2-10) (Table 1). At 90-day follow-up, mortality (mRS 6) occurred in 18.8% of patients, and 37.5% of patients had poor functional outcome (mRS 4-6) (Table 1).

Table 1.

Baseline demographic and clinical characteristics.

Variable	n = 128
Demographics
Age, median (IQR)	72 (63-81)
Female, no. (%)	76 (59.4%)
Past medical history, no. (%)
Hypertension	103 (80.5%)
Diabetes mellitus	39 (30.5%)
Dyslipidemia	68 (53.1%)
Coronary artery disease	69 (53.9%)
Prior stroke	26 (20.3%)
Atrial fibrillation	56 (43.8%)
Tobacco use	53 (41.4%)
Admission NIHSS, median (IQR)	15 (11-20)
Premorbid mRS, median (IQR)	0 (0-1)
Occlusion location, no. (%)
Distal intracranial carotid artery	4 (3.1%)
M1 middle cerebral artery	95 (74.2%)
Proximal M2 middle cerebral artery	29 (22.7%)
Left-sided occlusion, no. (%)	69 (53.9%)
Prolonged venous transit, no. (%)
Neither SSS or torcula	87 (68.0%)
SSS Only or torcula only	22 (17.2%)
Both SSS and torcula	19 (14.8%)
Intervention and Recovery
Thrombectomy alone, no. (%)	87 (68.0%)
IV Thrombolysis + thrombectomy, no. (%)	41 (32.0%)
mTICI, no. (%)
2B	35 (27.3%)
2C	23 (18.0%)
3	70 (54.7%)
Discharge NIHSS, median (IQR)	4 (2-10)
90-day mRS, no. (%)
0	20 (15.6%)
1	22 (17.2%)
2	25 (19.5%)
3	13 (10.2%)
4	19 (14.8%)
5	5 (3.9%)
6	24 (18.8%)

Abbreviations: IQR = interquartile range, NIHSS = National Institutes of Health Stroke Scale, mRS = modified Rankin score, SSS = superior sagittal sinus, mTICI = modified treatment in cerebral infarction score.

The correlation between PVT and 90-day mRS was significant ( $\rho$ = 0.35, p < 0.0001) (Figure 1). The significance of the correlation persisted when PVT was correlated with mortality (r = 0.26, p = 0.002) (Figure 1). The correlation between PVT and poor functional outcome was similarly significant (r = 0.27, p = 0.002) (Figure 1).

Figure 1.

Scatterplot showing prolonged venous transit (PVT) with (1) 90-day modified Rankin score (mRS), (2) 90-day mortality (mRS 6), and (3) poor functional outcome (mRS 4-6). The PVT is defined as time-to-maximum $\ge$ 10s volume in neither the superior sagittal sinus (SSS) nor torcula, in either the SSS or torcula, or in both regions.

Discussion

The PVT was statistically significantly correlated with 90-day neurological functional outcome. Hence, the PVT has potential utility in serving as a novel adjunct pretreatment imaging parameter for assessing functional recovery in stroke patients. Although the strength of the association was modest, it is important to note that several factors influence the post-treatment clinical course for stroke patients.^6,7 Further analyses are required to ascertain the strength and independence of PVT as a pretreatment prognostic marker. Nonetheless, the relationship established between PVT and functional recovery serves as a foundation for future explorations utilizing this new perfusion parameter.

The correlation between PVT and poor functional recovery is consistent with prior studies that examined superficial and deep regional VO separately. For example, Sue et al. demonstrated delayed VO peak time in the SSS was independently associated with unfavorable recovery despite successful reperfusion therapy with EVT. ² Similarly, Adusmulli et al. showed that VO peak time at the confluence of sinuses (i.e., torcula) was strongly correlated with cortical vein opacification scores (COVES), previously shown to correspond to post-EVT functional recovery.^4,8 Separately, analyzing perfusion dynamics in each region has been strongly associated with post-EVT functional recovery.^9–12

The PVT was derived from combining prior independent observations of the association between hemodynamics in the superficial and deep VO, respectively, and functional outcomes. Amongst the abundance of the ever-expanding list of clinical, imaging, and interventional parameters predicting clinical outcomes,^2,4,13–24 the PVT offers a paradigm that consolidates and simplifies VO assessment into a single, easy-to-ascertain measure. In the era of simplification, future studies with similar objectives are warranted to consolidate valuable assessment components into user-friendly, cumulative indices for enhanced practical implementation into clinical workflows governing acute stroke care management.

Limitations

Given the retrospective nature of the analysis, this study is subject to limitations inherent to the study design. Several factors affect the degree of neurorehabilitation in the 90-day period, including comorbidities and discharge disposition ⁶ ; hence, the unadjusted effect of confounding likely explains the mild-to-moderate strength of findings demonstrated in this analysis. Moreover, PVT does not account for compensatory venous drainage from the nonaffected side, which may affect the degree of prolonged venous transit on the affected side. Future analyses in larger cohorts will be able to adequately control for these confounders. Given that this was a demonstration of a novel concept in a sample of moderate size, external validation in a larger cohort is needed to validate generalizability of findings.

Conclusion

The PVT is a novel, qualitative perfusion imaging marker reflecting severity of compromised cerebral VO in key regions of both superficial and deep VO tracts. There is a mild, significant correlation between PVT and neurological functional outcome. Consequently, PVT may be a useful adjunct measure in neuroprognostication efforts. Future analyses will characterize the performance of PVT within current clinical decision-making paradigms.

Ethical statement

Ethical approval

This study was approved by the institutional review board of the Johns Hopkins School of Medicine.

Informed consent

This retrospective study was approved by the IRB with waiver of informed consent.

ORCID iDs

Vivek S Yedavalli https://orcid.org/0000-0002-2450-4014

Dhairya A Lakhani https://orcid.org/0000-0001-7577-1887

Manisha Koneru https://orcid.org/0000-0001-5012-6793

Licia Luna https://orcid.org/0000-0003-3539-4831

Adam A Dmytriw https://orcid.org/0000-0003-0131-5699

Max Wintermark https://orcid.org/0000-0002-6726-3951

Data availability statement

De-identified data will be made available upon reasonable request to the corresponding author.*