Evaluating the prognostic impact of hypoperfusion intensity ratio in acute ischemic stroke patients undergoing early-phase endovascular thrombectomy

Abstract

This research aimed to assess the prognostic relevance of the hypoperfusion intensity ratio (HIR) concerning 90-day outcomes in patients with acute ischemic stroke (AIS) managed within the early intervention window. A retrospective review was conducted on AIS patients who received pretreatment computed tomography perfusion imaging and endovascular thrombectomy due to large vessel occlusions in the anterior circulation between January 2020 and September 2022. Clinical data, including the Alberta Stroke Program Early Computed Tomography Score (ASPECTS) from non-contrast CT, along with perfusion metrics such as ischemic core, hypoperfusion extent, core-penumbra mismatch, and HIR, were analyzed. Patients were divided into groups with favorable (modified Rankin Scale score 0–2) and unfavorable outcomes (modified Rankin Scale score 3–6). Among the 187 patients evaluated, 95 (50.8%) had favorable outcomes. Univariate analysis showed significant associations between functional outcomes and variables like age, National Institutes of Health Stroke Scale score at admission, ASPECTS, HIR, ischemic core volume, and hypoperfusion volume (P < .05). Multivariate analysis revealed that younger age (odds ratio [OR] 1.064; 95% confidence interval [CI] 1.025–1.106, P = .001), lower National Institutes of Health Stroke Scale score at admission (OR 1.116; 95% CI 1.038–1.199, P = .003), smaller ischemic core volume (OR 1.017; 95% CI 1.002–1.033, P = .029), higher ASPECTS (OR 0.800; 95% CI 0.662–0.967, P = .021), and reduced HIR (OR 1.516; 95% CI 1.230–1.869, P = .001) independently predicted favorable outcomes at 90 days. Lower HIR was independently linked to improved functional outcomes in AIS patients receiving endovascular thrombectomy within the early intervention timeframe.

1. Introduction

Endovascular thrombectomy (EVT) has established itself as the primary treatment for acute ischemic stroke (AIS) patients suffering from large vessel occlusion (LVO) in the anterior circulation, particularly when administered in the early intervention window. Additionally, EVT can be considered in later stages for certain cases, contingent on computed tomography perfusion (CTP) imaging results in the extended therapeutic period.^[1–3] Despite these advancements, a considerable number of patients do not attain functional independence, even when EVT is performed within a 12-hour window from symptom onset and achieves successful recanalization. One of the key factors influencing posttreatment outcomes is the presence of robust collateral circulation, which is a significant predictor of sustained clinical recovery.^[4,5] The hypoperfusion intensity ratio (HIR), derived from CTP metrics, has emerged as a reliable indicator of collateral status in AIS patients with LVO, showing strong correlation with findings from both computed tomography angiography (CTA) and conventional angiography.^[4] Research has demonstrated that AIS-LVO patients undergoing EVT with CTP-based screening tend to have improved outcomes compared to those without such imaging selection.^[6–9] Although current guidelines do not universally endorse CTP imaging within the initial treatment window, recent studies suggest that early perfusion imaging selection may result in notably better functional outcomes for patients compared to non-imaging-based selection.^[1]

The HIR is determined by calculating the ratio of the Tmax > 10-second lesion volume to the Tmax > 6-second lesion volume,^[3] with a lower HIR indicating more favorable collateral circulation.^[4] Previous studies have shown that HIR is significantly linked to initial stroke severity, infarct progression, and even functional outcomes.^[3,10] A recent study indicated that a favorable HIR (HIR < 0.4) predicts better outcomes in AIS patients treated in the late therapeutic window.^[11]

However, the prognostic role of HIR in AIS patients undergoing EVT within the early intervention timeframe (0–6 hours) remains underexplored.

This study aims to investigate the association between baseline HIR observed on CTP and functional outcomes in AIS patients treated with EVT in the early therapeutic period. It is the first study to propose HIR as a prognostic marker specifically for the early intervention phase in AIS, providing a novel perspective for addressing the challenges associated with early treatment in this patient population.

2. Methods

2.1. Patients

This retrospective study included patients with LVO in the anterior circulation who underwent EVT at a single institution between January 2020 and September 2022. Ethical approval was obtained from the local review board (2021-SR-288). The inclusion criteria were as follows: (1) treatment initiated within 6 hours of symptom onset or the last known normal state; (2) occlusion in an anterior circulation artery, specifically the intracranial internal carotid artery (ICA) and/or the M1/M2 segment of the middle cerebral artery (MCA); (3) age > 18 years; (4) National Institutes of Health Stroke Scale (NIHSS) score ≥ 6 on admission; (5) high-quality CTP imaging without significant motion artifacts; (6) baseline mRS score of 0 to 2; and (7) no evidence of intracranial hemorrhage on initial cranial CT.

2.2. Imaging and clinical information

Data for this study were extracted from a prospective registry of acute stroke patients, encompassing demographic details (age, sex), stroke risk factors (such as hypertension, diabetes, coronary artery disease, atrial fibrillation, smoking, and history of ischemic stroke), and clinical information (NIHSS score at admission, use of intravenous thrombolysis with recombinant tissue-type fibrinogen activator [rt-PA], and time from stroke onset to puncture). Each patient underwent a structured telephone follow-up at 3 months poststroke by a trained nurse to assess all-cause mortality and functional outcomes, with a modified Rankin Scale score of 0 to 2 indicating a favorable outcome.

All AIS patients were managed under an institutional protocol using a 128-section multilayer CT system (Optima CT660, GE Medical Systems, Chicago, IL). CTP images were acquired via a periodic spiral method (four-dimensional adaptive spiral mode, 100 kVp, 200 mAs, rotation time 0.4 seconds, 0.984 pitch value).

Quantitative CTP parameters included: (1) ischemic core, defined by < 30% cerebral blood flow relative to the contralateral hemisphere; (2) Tmax > 6 and > 10 seconds lesion volumes; (3) mismatch volume, representing the difference between hypoperfusion volume and ischemic core; and (4) HIR, calculated as the ratio of Tmax > 10 seconds volume to Tmax > 6 seconds volume.

2.3. Favorable and unfavorable outcomes definition

In this study, patient outcomes at 90 days posttreatment were categorized as favorable or unfavorable based on the modified Rankin Scale. A favorable outcome was classified as an mRS score of 0 to 2, reflecting functional independence, while an unfavorable outcome ranged from 3 to 6, with a score of 6 indicating mortality. Outcome assessments were conducted through structured telephone interviews by trained research nurses.

2.4. Outcomes and statistical analysis

Continuous variables were presented as mean ± standard deviation or median and interquartile range (IQR). Categorical variables were summarized through frequency analysis. Comparisons between groups with favorable and unfavorable outcomes were made using independent samples t test or Mann–Whitney U test for continuous variables, and chi-square or Fisher exact test for categorical data. Variables with a significance level of P ≤ .05 in univariate analyses were included in multivariate linear or logistic regression models. CTP data was retrospectively analyzed using RAPID software, without influencing EVT decision-making. The predictive performance of clinical and imaging metrics, individually and in combination, for favorable outcomes was assessed using receiver operating characteristic (ROC) curve analysis. All statistical analyses were conducted using SPSS software (version 20.0, IBM).

3. Results

From January 2020 to September 2022, 187 AIS patients meeting all inclusion criteria and treated within the early therapeutic window were enrolled in the study. The median age was 72 years (IQR 64–81), with 42% being female. Occlusions were located in the ICA (n = 52), MCA-M1 segment (n = 94), MCA-M2 segment (n = 21), or a combination of ICA and M1 segment (n = 19). The time from stroke onset to groin puncture ranged from 85 to 360 minutes (median 247, IQR 160–306), while the onset-to-reperfusion time varied between 135 and 610 minutes (median 327, IQR 244–395). On CTP imaging, median lesion volumes were as follows: ischemic core 13 mL (IQR 0–43), hypoperfusion volumes for Tmax > 6 seconds 144 mL (IQR 96–203), and Tmax > 10 seconds 63 mL (IQR 34–120). The median HIR was 0.5 (IQR 0.3–0.6). At the 90-day mark, 95 patients (50.8%) achieved a favorable outcome.

In univariate analysis, patients with favorable outcomes were generally younger (P < .001), had lower NIHSS scores at admission (P < .001), higher Alberta Stroke Program Early Computed Tomography Score (ASPECTS) (P = .001), and smaller volumes for ischemic core, Tmax > 6 seconds, and Tmax > 10 seconds (all P < .05), along with a lower HIR (P < .001). No significant differences were observed between the groups regarding risk factors, stroke onset-to-puncture time, or vessel occlusion site. Detailed patient characteristics and univariate outcomes analysis are presented in Table 1.

Table 1.

Comparison of clinical variables and neuroimaging characteristic in patients with favorable versus unfavorable outcome at 90 days.

Variables	Favourable outcome (n = 95)	Unfavourable outcome (n = 92)	P-value
Patient data
Age, years, median (IQR)	69 (62–77)	76 (67–84)	<.001
Female, n (%)	39 (41.1%)	42 (45.7%)	.557
History
Hypertension, n (%)	50 (52.6%)	46 (50.0%)	.771
Diabetes mellitus, n (%)	16 (16.8%)	22 (23.9%)	.154
Hyperlipidemia, n (%)	8 (8.4%)	7 (7.6%)	.526
Atrial fibrillation, n (%)	47 (49.5%)	42 (45.7%)	.382
Prior stroke, n (%)	12 (12.6%)	14 (15.2%)	.101
Smoking, n (%)	13 (13.7%)	12 (13.0%)	.535
Coronary heart disease, n (%)	19 (20.0%)	21 (22.8%)	.385
Baseline NIHSS, median (IQR)	14 (11–17)	20 (14–24)	<.001
Baseline ASPECTS, median (IQR)	7 (6–9)	6 (4–7)	.001
Occlusion site			.314
ICA, n (%)	31 (32.6%)	21 (22.8%)
MCA-M1, n (%)	44 (46.3%)	50 (54.3%)
MCA-M2, n (%)	8 (8.4%)	13 (14.1%)
ICA + MCA-M1	10 (10.5%)	9 (9.8%)
Etiology of the occlusion
Cardioembolism, n (%)	52 (54.7%)	54 (58.7%)	.345
Large artery atherosclerosis, n (%)	28 (29.5%)	28 (30.4%)	.506
Others, n (%)	14 (14.7%)	9 (9.8%)	.210
IV tPA, n (%)	38 (40.0%)	30 (32.6%)	.185
Onset to puncture, median (IQR)	241 (170–303)	246 (160–307)	.656
Onset to reperfusion, median (IQR)	321 (234–382)	332 (226–400)	.322
CTP parameters
CBF < 30% volume in mL, median (IQR)	6 (0–20)	29 (9–68)	.001
Mismatch volume (mL), median (IQR)	115 (70–163)	118 (80–175)	.349
Tmax > 10 seconds, median (IQR)	45 (17–79)	97 (49–137)	<.001
Tmax > 6 seconds, median (IQR)	134 (80–184)	173 (113–233)	.001
Hypoperfusion intensity ratio, median (IQR)	0.4 (0.1–0.5)	0.6 (0.4–0.7)	<.001
mTICI 2b/3, n (%)	84 (88.4%)	73 (79.3%)	.068

Data were expressed as median, IQR [interquartile range presented as the 25th and 75th percentile].

ASPECTS = Alberta Stroke Program Early Computed Tomography Scores, CBF = cerebral blood flow, HIR = hypoperfusion intensity ratio, ICA = internal carotid artery, MCA = middle cerebral artery, mTICI = modified Treatment in Cerebral Ischemia, NIHSS = National Institutes of Health Stroke Scale scores, rt-PA = recombinant tissue plasminogen activator, Tmax = time to maximum.

In the multivariate analysis, to address multicollinearity among independent variables (Tmax > 6 seconds, Tmax > 10 seconds, and HIR), only the most clinically relevant variables from each group were included in the regression model. Tmax > 6 seconds was selected, as mismatch volume calculation requires this measure (mismatch volume = Tmax > 6 seconds volume minus low ischemic core volume). Key independent predictors for a favorable 90-day outcome included younger age (odds ratio [OR] 1.064; 95% confidence interval [CI] 1.025–1.106, P = .001), lower NIHSS score at admission (OR 1.116; 95% CI 1.038–1.199, P = .003), smaller ischemic core volume (OR 1.017; 95% CI 1.002–1.033, P = .029), higher ASPECTS (OR 0.800; 95% CI 0.662–0.967, P = .021), and lower HIR (OR 1.516; 95% CI 1.230–1.869, P = .001) (see Table 2).

Table 2.

Multivariable analyses for predicting a favorable outcome.

Variables	Odds ratio (95% CI)	P-value
Age	1.064 (1.025–1.106)	.001
Baseline NIHSS	1.116 (1.038–1.199)	.003
Baseline ASPECTS	0.800 (0.662–0.967)	.021
CBF < 30% volume in mL	1.017 (1.002–1.033)	.029
Hypoperfusion intensity ratio	1.516 (1.230–1.869)	.001
Tmax > 6 seconds	1.000 (0.996–1.003)	.908

ASPECTS = Alberta Stroke Program Early Computed Tomography Scores, CBF = cerebral blood flow, CI = confidence interval, HIR = hypoperfusion intensity ratio, NIHSS = National Institutes of Health Stroke Scale scores, OR = odds ratio, Tmax = time to maximum.

The optimal HIR threshold for predicting favorable outcomes was determined to be 0.4, with a sensitivity of 75.0% and specificity of 70.6%. ROC curve analysis yielded the following area under the curve (AUC) values: 0.786 (95% CI 0.720–0.843) for HIR, 0.738 (95% CI 0.669–0.799) for ischemic core volume, 0.647 (95% CI 0.574–0.715) for age, 0.730 (95% CI 0.661–0.792) for NIHSS score at admission, and 0.718 (95% CI 0.648–0.781) for ASPECTS. Further analysis revealed that combining HIR with age, NIHSS score, ASPECTS, and ischemic core volume increased the AUC to 0.875 (Fig. 1). Additionally, Figure 2 highlights that both HIR and ischemic core volume were independent neuroimaging markers associated with favorable outcomes.

Figure 1.

The performance of the HIR for predicting a favorable outcome at 90 days in AIS patients within the early therapeutic window. The areas under the ROC curve are 0.786 (95% CI 0.720–0.843), respectively. AIS = acute ischemic stroke, CI = confidence interval, HIR = hypoperfusion intensity ratio, ROC = receiver operating characteristic.

Figure 2.

The performance of the HIR and clinical variables (age and NIHSS score at admission) and their combinations for predicting a favorable outcome at 90 days in AIS patients within the late therapeutic window. The areas under the ROC curve are 0.786 (95% CI 0.720–0.843), 0.738 (95% CI 0.669–0.799), 0.647 (95% CI 0.574–0.715), 0.730 (95% CI 0.661–0.792), and 0.718 (95% CI 0.648–0.781) for HIR, ischemic core volume, age, NIHSS score at admission, and ASPECTS, respectively. With the combination of age, NIHSS score at admission, ischemic core volume, and ASPECTS the area under the ROC curve increased to 0.837 (95% CI 0.776–0.887, sensitivity, 82.9%; specificity, 68.8%). With the combination of age, NIHSS score at admission, ischemic core volume, ASPECTS and HIR, the area under the ROC curve increased to 0.875 (95% CI 0.819–0.919, sensitivity, 80.9%; specificity, 79.6%), significantly higher than each variable (P < .05). AIS = acute ischemic stroke, ASPECTS = Alberta Stroke Program Early Computed Tomography Score, AUC = area under the curve, CI = confidence interval, HIR = hypoperfusion intensity ratio, NIHSS = National Institutes of Health Stroke Scale, ROC = receiver operating characteristic.

4. Discussion

In our retrospective analysis of a prospectively collected stroke registry, the CTP-derived HIR, which can be rapidly assessed using automated software, was found to be independently associated with favorable functional outcomes in patients undergoing EVT within the early therapeutic window. Specifically, a lower HIR (<0.4) correlated with better outcomes. Additional factors influencing clinical results included NIHSS score at admission, age, ASPECTS, and ischemic core volume. Moreover, combining HIR with clinical factors (age, NIHSS score at admission) and ASPECTS demonstrated high accuracy in predicting favorable neurological outcomes at 90 days.

The concept of HIR serves as an objective and continuous metric for assessing collateral flow, with lower HIR values indicating superior collateral status.^[3,12]Assessing the collateral status from pretreatment CT angiography images is straightforward, however, the quantification of the collateral status is not consistent and can be subjective and reviewer dependent, while the calculation of the HIR is more objective.^[13] Utilizing automated software to calculate the HIR can reduce potential discrepancies and increase its accuracy. As a continuous value, the HIR provides a quantitative approach to assess hypoperfusion severity, allowing for a more precise analysis of hypoperfusion severity.

Our primary finding reveals a strong correlation between pretreatment HIR and clinical outcomes post-EVT. Previous research has demonstrated that HIR is closely associated with collateral blood flow status^[4,14] and the formation of cerebral edema.^[15] Furthermore, patients with lower HIR values are more likely to qualify for thrombectomy.,^[16] exhibit favorable arterial collaterals, and experience less ischemic core expansion.^[12] Building on these findings, our study shows that patients treated within the early therapeutic window who achieve favorable outcomes tend to present with lower HIR values.

Consistent with other studies, our analysis identified younger age, lower NIHSS scores, and higher ASPECTS as independent predictors of a favorable outcome at the 3-month follow-up (mRS 0–2).^[17,18]. Specific perfusion deficit volumes, such as ischemic core volume, were also linked to clinical outcomes.^[19] Notably, our study found that larger infarct volumes correlate with poorer outcomes in patients who received EVT within the early therapeutic timeframe. Although ischemic core volume remains an essential factor in outcome determination, HIR appears to hold additional clinical relevance. Beyond traditional perfusion metrics, HIR may serve as an influential predictor of patient outcomes.

Within the therapeutic timeframe, several clinical prognostic tools are used to estimate outcomes in AIS patients undergoing EVT. These include the Pittsburgh Response to Endovascular Therapy score,^[20] the Totaled Health Risk in Vascular Events score, and the Houston Intra-Arterial Therapy score.^[21] These scoring systems primarily rely on clinical factors and basic neuroimaging parameters, such as age, NIHSS score, and ASPECTS derived from non-contrast CT scans. However, their predictive accuracy, as measured by the AUC, typically ranges from 0.60 to 0.75 for favorable outcomes. Our study has shown that age and NIHSS score remain crucial clinical indicators for predicting outcomes, consistent with prior research. Notably, our study introduced the HIR as a novel prognostic marker in AIS patients undergoing EVT. When combined with age, NIHSS score, ASPECTS, and ischemic core volume, the HIR yielded an AUC of 0.880, surpassing the performance of existing prognostic risk scores. This highlights the added value of incorporating HIR into prognostic models for more accurate outcome prediction. Early risk assessment of severity and prognosis is paramount for successful long-term treatment in AIS patients undergoing EVT. The clinical relevance of identifying high HIR in these patients provides valuable prognostic information posttreatment. Accurate prognostic information is essential for clinicians to effectively manage patients and guide treatment decisions. In cases where patients exhibit a higher HIR, closer follow-up after hospital discharge and/or more intensive therapies may be warranted to mitigate disability and improve outcomes. Furthermore, the availability of commercial CTP software platforms enables the automated derivation and reporting of HIR. This not only saves time but also facilitates the rapid delivery of results, expediting the decision-making process in clinical practice. Overall, our study underscores the importance of detailed discussions on the value of incorporating novel prognostic markers like HIR into existing prognostic models, ultimately enhancing their predictive accuracy and clinical utility in guiding patient management.^[22,23]

Our study demonstrates that HIR is an independent predictor of favorable outcomes in patients with AIS treated with EVT in the early therapeutic window. In the future, HIR could potentially be incorporated into prognostic scoring systems to better identify high-risk patients. By combining HIR with established clinical parameters such as age, NIHSS score, and ASPECTS, a more comprehensive and accurate prognostic tool could be developed. The objective nature of HIR, which is derived from automated CTP software, offers a distinct advantage over subjective imaging assessments. Our analysis shows that integrating HIR with other clinical and imaging variables significantly improves predictive performance, as evidenced by the increase in AUC when these factors are combined. Incorporating HIR into future scoring systems could enhance early risk stratification and help guide clinical decision-making.

In the past, researchers such as Nitin have designed various frameworks based on methods like the internet of things and Fog computing for real-time monitoring of health information. Such systems also provide assistance in clinical detection of disease complications. These systems can be integrated with clinical practices and combined with indicators like HIR to contribute jointly to improving patients’ clinical prognosis.^[24–26] The researchers’ future direction will be closely intertwined with internet of things, artificial intelligence, and other emerging technologies. By integrating with secure Internet frameworks and the clinical indicators we use, it will generate greater impact.^[27–29] Our study has a few limitations. Firstly, our study was a retrospective research design, which may not have taken into account some confounding factors. Secondly, the number of patients from a single center was small. Furthermore, the HIR value, obtained from automated software, was an objective finding; however, the cutoff point of the HIR for a favorable outcome and its relevance may be restricted to the particular study population. Moreover, this cohort exclusively treated all AIS patients with EVT, thus limiting the applicability of the findings. Additionally, HIR values can be affected by the imaging technique employed (e.g., computed tomography), technical issues (e.g., head movement or incorrect selection of the arterial/venous input function) and the software platform used for perfusion imaging analysis. Therefore, larger prospective and randomized trials are needed to validate the role of HIR in determining EVT eligibility and predicting outcomes for AIS patients treated within the early therapeutic window.

5. Conclusion

In our study of thrombectomy patients within the early therapeutic window, evaluated with CTP, we found that HIR was independently associated with 90-day functional independence. This novel association highlights the potential significance of HIR in predicting outcomes in this patient group. Our findings contribute valuable insights to discussions on treatment and prognosis, and may inform future studies exploring the role of CTP in thrombectomy patients with an early therapeutic window.

Author contributions

Conceptualization: Aicheng Sun, Yuezhou Cao, Haibin Shi.

Data curation: Aicheng Sun, Yuezhou Cao, Linbo Zhao, Haibin Shi.

Funding acquisition: Zhenyu Jia, Linbo Zhao.

Investigation: Yuezhou Cao, Zhenyu Jia.

Validation: Haibin Shi.

Visualization: Haibin Shi, Sheng Liu.

Writing – original draft: Aicheng Sun, Zhenyu Jia, Linbo Zhao, Haibin Shi, Sheng Liu.

Writing – review & editing: Aicheng Sun, Sheng Liu.