Abstract
Background
High-resolution magnetic resonance imaging (MRI) is a feasible method for identifying and measuring luminal thrombosis in the occluded middle cerebral artery (MCA) of stroke patients.
Purpose
To evaluate the relationship between thrombus length in the middle cerebral artery (MCA) and clinical prognosis in stroke patients using high-resolution MRI (T1-weighted [T1W]-CUBE).
Material and Methods
In this retrospective study, patients with MCA thrombi were identified via high-resolution MRI. Thrombus length was measured using T1W-CUBE sequences, and clinical prognosis was assessed using modified Rankin scale (mRS) scores at a 3-month follow-up. Patients were categorized into good or poor prognosis groups based on mRS scores. Statistical analyses compared thrombus length and diameter between groups. Restricted cubic spline (RCS) analyses were performed to examine the relationship between thrombus length, diameter, and mRS scores. Multiple logistic regression was used to explore the association between thrombus length and prognosis.
Results
The mean thrombus length was significantly longer in the poor prognosis group compared to the good prognosis group (25.69 ± 8.78 mm vs. 15.11 ± 6.91 mm; P <0.001). Thrombus diameter did not significantly differ between the groups (P = 0.961). RCS showed a monotonically increasing relationship between thrombus length and log OR of mRS. Multiple logistic regression indicated that each additional millimeter of thrombus length increased the risk of poor prognosis by 1.22 times (odds ratio = 1.22, 95% confidence interval = 1.11–1.40; P <0.05).
Conclusion
MCA thrombus length is a significant predictor of prognosis in stroke patients, with longer thrombi associated with worse clinical outcomes.
Introduction
Approximately 80% of strokes are ischemic, mainly caused by atherothrombosis (1). Currently, the diagnosis of cranial arterial thrombosis can be achieved through various imaging techniques, but quantitative thrombus measurement remains challenging without simple and accurate imaging tools.
The hyperdense sign on non-contrast computed tomography (CT) and the susceptibility vessel sign (SVS) on gradient-recalled echo magnetic resonance imaging (MRI) are established markers for detecting clots within the cerebral arterial tree (2). However, conventional CT and MRI scans experience low spatial resolution, making them inadequate for clearly visualizing intravascular lesions and providing detailed information about the location and length of thrombi (3,4). The SVS arises from deoxyhemoglobin within the clot, causing inhomogeneities in magnetic fields and resulting in a hypointense susceptibility artifact in T2*-weighted (T2*W) sequences. Nonetheless, the clinical utility of SVS is hindered by the blooming effect, complicating accurate thrombus length measurements (5). In addition, T2*W MRI techniques may underestimate the full clot extent due to variations in thrombus composition and age (4,6,7).
Some studies have explored thrombus length measurement using CT angiography (CTA) or MR angiography (MRA), but these methods have limitations. The current methods, MRA or CTA, with or without contrast enhancement, cannot directly visualize the thrombus itself. Instead, they evaluate the thrombus through indirect signs of impaired blood flow (8). Therefore, these methods are inadequate for reliably assessing thrombus size and characteristics, particularly for tracking scattered cerebral thromboemboli (9). For instance, recent research suggested that weak collateral circulation or short delays between contrast injection and imaging acquisition in CTA may overestimate thrombus involvement (10).
The measurement of cranial artery thrombosis has implications for the prognosis and treatment of stroke patients. Barreto et al. (11) found a significant association between high clot burden and poor clinical outcomes after 3 months. In another study, a middle cerebral artery (MCA) thrombus that was longer than 8 mm was found to have little possibility of recanalization by intravenous thrombolysis (12). Several studies have demonstrated that increased thrombus length predicted recanalization failure, while shorter thrombus length favored thrombus recanalization (13,14).
T1-weighted (T1W)-CUBE imaging displays the vessel wall by suppressing blood flow signals within the vessel lumen. This sequence ensures high-resolution isotropic imaging without altering the signal-to-noise or contrast-to-noise ratios (15,16). It has been demonstrated that T1W-CUBE not only accurately detects intraluminal thrombi in stroke patients but also provides reliable measurements of intraluminal thrombus length (17).
Therefore, the aim of the present study was to investigate the relationship between MCA thrombus length and clinical prognosis in stroke patients using T1W-CUBE imaging.
Material and Methods
Participants
The study was conducted in accordance with the Declaration of Helsinki. This prospective study was approved by the Shandong Provincial Qianfoshan Hospital Review Board (2024 No. S893). The requirement for informed consent was waived due to the retrospective nature of the study.
The inclusion criteria were as follows: (i) unilateral middle cerebral artery occlusion diagnosed by MRA, CTA or digital subtraction angiography (DSA); (ii) recurrent ischemic stroke or transient ischemic attack (TIA) related to the occlusive arteries; (iii) patients who did not receive timely thrombolytic therapy or mechanical thrombectomy in the acute phase due to a missed treatment window; (iv) exclusion of severe carotid artery stenosis; and (v) patient diagnosed with cranial artery thrombosis after T1W-CUBE imaging.
All retrospectively collected patients already diagnosed with cranial arterial thrombosis were re-diagnosed by three radiologists, each with more than 10 years of experience. Cases were included if there was a consensus on thrombus diagnosis and excluded if any of the three physicians provided a negative diagnosis.
After a screening of cases from between December 2017 and December 2022 in the Department of Neurology, a total of 67 patients with cerebral infarction (48 men, 19 women; median age = 51 years; age range = 21–72 years) were enrolled. The time from symptom onset to MRI was in the range of 2–14 days. Specifically, one patient underwent MRI on day 2, 23 patients on days 3–7, and 43 patients on days 8–14. All patients were treated with medication according to clinical guidelines.
On T1W-CUBE images, the signal characteristics of thrombi resemble those of hematomas: acute and subacute thrombi usually appear as heterogeneous high-signal areas within the lumen, with surrounding thickened walls exhibiting lower signal intensity, creating a clear boundary (Figs. 1a and 2a).
The differential diagnosis based on T1W-CUBE images includes: (i) atherosclerotic plaques, which often manifest as eccentrically distributed iso-signals or slightly high signals compared to the normal vessel wall, typically causing luminal stenosis rather than complete occlusion; (ii) as demonstrated in Fig. 2c, slow blood flow typically appears as a slightly hyperintense and relatively homogeneous intraluminal signal, which do not show occlusion on MRA or CTA (Fig. 2d) (16). The lumen diameter in these regions is preserved, with no evidence of narrowing, and the vessel wall remains thin and smooth, often making it indistinguishable from adjacent slow-flow signal.
Clinical data collection
Patients diagnosed with thrombus were followed up through telephone interviews 3 months after diagnoses to assess their mRS scores. The mRS score is divided into six grades: 0 = no symptoms; 1 = no significant disability; 2 = mild disability; 3 = moderate disability; 4 = severe disability (the patient cannot walk independently and needs assistance with basic activities of daily living); and 5 = severe disability (incontinence, bedridden, requiring continuous care). Relevant medical history of thrombus patients was obtained through a retrospective review of admission records, disease course, and relevant test results.
MRI scanning parameters
T1W-CUBE image acquisition was performed using a 32-channel head coil on a 3.0-T MRI scanner (Discovery 750w, GE Healthcare, Chicago, IL, USA). Whole-brain sagittal imaging was conducted without using contrast media, with a field of view of 240 × 240 mm. The scanning parameters were as follows: repetition time (TR)/echo time (TE) = 600/14.4 ms; slice thickness = 1 mm; no slice gap; and matrix size = 288 × 288. The resulting voxel size was 0.4 × 0.4 × 0.4 mm. Fat suppression was applied during image acquisition, and the scanning time was approximately 4 min 16 s.
Neuroimaging analysis
T1W-CUBE images were reconstructed and analyzed using the GE Advantage Workstation 4.6. Three radiologists, each with over 10 years of clinical experience, independently evaluated these images.
The length of the cranial arterial thrombus was measured by transferring the acquired sagittal T1W-CUBE images to the workstation. The corresponding axial and coronal images were reconstructed, and curve measurement for vessel contour was performed. To correct vessel twisting effects, measurements were made by combining images from three axial/sagittal/coronal planes and drawing a line parallel to the vessel wall along the center of the vessel (Fig. 1b, c). When multiple thrombi were present simultaneously, thrombus length measurements were performed and the length values were summed. Three radiologists independently performed thrombus measurements, and the results were averaged to one decimal place. Fig. 1a–d illustrates the process of thrombus identification and length measurement in a patient with a thrombus.
(a–d) A 38-year-old man with weakness of left limbs. (a) Higher T1 signal fills in the right MCA M1 segment in T1W-CUBE (arrow). (b) The thrombus length measurement was made using the workstation's curve measurement tool, and the measured thrombus length was 18.6 mm. (c) Magnified T1W-CUBE image shows thrombus measurement in the right MCA M1 segment. (d) DSA shows an occlusion in the M1 segment of the right MCA (arrow). DSA, digital subtraction angiography; MCA, middle cerebral artery.
T1W-CUBE and MRA images from two patients with left limb weakness. (a, b) A 71-year-old male patient with an 8-day history of left limb weakness. (a) T1W-CUBE image demonstrates thrombus formation (arrow) in the distal segment of the right MCA M1 segment. (b) MRA source image reveals a localized filling defect (arrow) in the corresponding region. (c, d) A 48-year-old male patient presenting with 14 days of episodic left-sided limb weakness. (c) T1W-CUBE image shows slow blood flow (arrow) in vessels surrounding the right MCA M2 segment. (d) MRA source image shows preserved blood flow signals (arrow) in the corresponding area. MCA, middle cerebral artery; MRA, magnetic resonance angiography.
Thrombus diameter was independently measured at its largest visible diameter by three radiologists, and the results were averaged to one decimal place.
Statistical analysis
R software version 4.1.2 was used for statistical analyses. Continuous variables were expressed as mean ± SD and compared using t-tests or Wilcoxon tests, depending on their distribution. Categorical variables were presented as frequencies (percentages) and compared using the chi-square test. The inter-observer agreement among the three readers was evaluated using Kendall's W test. The associations of thrombus length and diameter were separately tested with the prognosis using restricted cubic spline (RCS). The patients were categorized into good prognosis (mRS ≤2) and poor prognosis (mRS >2) groups. A multiple logistic regression model was established using stepwise regression with thrombus length as the independent variable (X) and mRS score as the dependent variable (Y).
Results
Inter-observer analysis
The inter-observer agreement in measuring thrombus lengths was excellent (Kendall's coefficient of concordance: 0.999; P <0.05). In addition, the measurement of thrombus diameter on T1W-CUBE images showed excellent agreement between the three observers (Kendall's coefficient of concordance: 0.898; P <0.05).
Comparison of clinical information by sex
The mean age of male patients was 50.71 years, with a mean thrombus length of 18.76 mm and a mean thrombus diameter of 2.47 mm. For female patients, the mean age was 49.95 years, with a mean thrombus length of 15.91 mm and a mean thrombus diameter of 2.35 mm (Table 1). No statistically significant differences were observed between these groups. Furthermore, the mean mRS scores were 1.60 for men and 1.21 for women, with no statistically significant differences noted (P = 0.282). The proportions of men who smoked and drank alcohol were 52.08% and 39.58%, respectively, significantly higher than those of women who smoked (5.26%; P = 0.001) and drank alcohol (10.53%; P = 0.044). There were no statistically significant differences between men and women in the incidence of stroke-related medical history, family history, and the total number of risk factors.
Comparison of clinical information of participants grouped by sex.
Values are given as n (%) or mean ± SD.
*Indicates a significant difference between male and female groups (P <0.05).
mRS, modified Rankin scale.
Characteristics of patients with different mRS score classifications
Patients were categorized into two groups based on the mRS score grading: favorable prognosis (mRS ≤2) and poor prognosis (mRS >2). Significant differences were observed in mean thrombus lengths between the two groups (15.11 mm for favorable prognosis and 25.69 mm for poor prognosis; P <0.001). In contrast, no significant difference was found in the mean thrombus diameter (2.44 mm for both groups; P = 0.961) (Table 2). The difference in diabetes rates between the two groups is statistically significant, with a higher proportion of diabetic patients in the group with poor prognosis. No significant differences were observed in sex, age, the distribution of other risk factors, or the count of risk factors among patients with different prognoses.
Characteristics of study participants graded according to mRS scores.
Values are given as n (%) or mean ± SD.
*Indicates a significant difference between favorable and poor prognosis groups (P <0.05).
mRS, modified Rankin scale.
RCS assessment of the association between thrombus length, diameter, and prognosis
RCS showed a monotonically increasing relationship between thrombus length and log OR of mRS, indicating that the longer the thrombus, the worse the prognosis (Fig. 3). Thrombus diameter and prognosis have an inverted U-shaped relationship (Fig. 4). Because the confidence interval always includes 0, there is no significant linear correlation between thrombus diameter and prognosis.
Association between thrombus length and prognosis. The vertical axis represents the logarithmic value of the odds ratio value.
Association between thrombus diameter and prognosis. The vertical axis represents the logarithmic value of the odds ratio value.
Multivariate logistic regression models to estimate the association coefficient between thrombus length and prognosis
A multiple logistic regression model included four variables: thrombus length (X), age, smoking, and diabetes mellitus (Table 3). Thrombus length was identified as a risk factor (odds ratio [OR] = 1.22), suggesting that for every 1 mm increase in thrombus length, the risk of poor patient prognosis increased by 1.22 times while holding other variables constant. Diabetes is also a risk factor for poor prognosis.
Logistic regression model of patients’ prognoses.
CI, confidence interval; OR, odds ratio.
Discussion
The appearance of SVS depends on the presence of deoxyhemoglobin within thrombi (18). Cardioembolic thrombi are primarily fibrin-rich (white thrombi) and may not produce a prominent SVS due to low paramagnetic content (19). In addition, factors such as scanning time, hematocrit, field strength, and flow dynamics influence T2* signal intensity, limiting its diagnostic reliability. Studies have reported that SVS is detectable in only about 53.1% of acute thrombosis cases (18). Unlike T2*/SWI, T1W-CUBE is not dependent on magnetic susceptibility effects. T1W-CUBE sequence's high sensitivity and specificity in thrombosis detection, as reported in previous studies, provided a solid foundation for its use as a diagnostic tool in our research (17,20). Due to differences in hemoglobin content and its degradation products, the temporal evolution of T1W MRI signal intensity in arterial thrombi does not precisely mirror that of intracerebral hematomas. The MR appearance of arterial thrombi and the changes detected over time result from the combination of different oxygenation states of the hemoglobin, changes in intracellular and matrix content of proteins and the hydration of the cellular components (21). In a porcine model of carotid artery thrombosis, Corti et al. (21) demonstrated that thrombi exhibited increased T1 signal intensity as early as 6 h after formation, peaking at approximately 1 week and plateauing by 6 weeks while remaining mildly elevated. In our cohort, all patients underwent MRI more than 48 h after symptom onset, at which point thrombi are typically in the subacute phase and appear hyperintense with clear contrast relative to surrounding tissues. These imaging characteristics support the suitability of T1W-CUBE sequences for reliable thrombus visualization and accurate length measurement in this population.
Our cohort had a higher proportion of male patients compared to female patients, yet no significant differences were observed in thrombus length, diameter, or other clinical parameters between the sexes, except for smoking and alcohol consumption history. Consequently, sex was not considered a confounding factor in the analysis of thrombus morphological characteristics.
The detrimental impact of longer thrombi on clinical outcomes can be attributed to several factors:
(1) Collateral flow compromise due to longer occlusions: Longer thrombi are more likely to obstruct the origins of vital perforating arteries, such as the lenticulostriate and insular perforators, which have limited collateral support. Moreover, the strength of pial collateral circulation is inversely associated with thrombus length (22).
(2) Delayed spontaneous recanalization in longer thrombi: The force required to remove a thrombus is influenced by both friction and adhesion between the thrombus and the vessel wall. Longer thrombi experience greater frictional forces and have increased surface area interactions through adhesion (14,23). This may make spontaneous recanalization difficult.
(3) Infarct growth kinetics (core vs. penumbra): The size of the infarct core is a key determinant of clinical outcome. Longer thrombi are associated with poorer collateral circulation, which promotes the development of larger infarct cores (7,24). Moreover, impaired collateral flow may facilitate thrombus propagation, further reducing cerebral perfusion in the ischemic penumbra and exacerbating ischemic injury (22).
Our study found no significant correlation between thrombus diameter and clinical prognosis. The diameter of an intracranial arterial thrombus is largely constrained by the vessel's internal diameter, which remains relatively stable during the thrombotic process. In MCA thrombi, particularly within the M1 segment, diameters typically do not exceed 3 mm, and variation in diameter across cases is minimal (25). This lack of variability likely accounts for the absence of a significant relationship between thrombus diameter and clinical outcomes.
Diabetes was specifically linked to poor functional outcomes at discharge and 90 days, as well as increased mortality at 90 days (26,27). Among acute ischemic stroke (AIS) patients with diabetes, those with acute hyperglycemia face an elevated risk of poor outcomes, possibly due to heightened infarct growth risk, which could compromise penumbral vulnerability (27). Previous studies have reported longer reperfusion times in diabetic patients, attributed to the additional medical attention required for hyperglycemia or diabetes management before reperfusion therapy (28,29).
The present study has one limitation that warrants discussion. The exclusion of patients who received thrombolysis or thrombectomy (due to missed treatment windows) introduces a potential selection bias. This subgroup represents a key population in which thrombus length significantly influences recanalization success and clinical outcomes. Future studies focusing on this population may further clarify the prognostic value of thrombus length, thereby enhancing the clinical applicability of our findings.
In conclusion, MCA thrombus length is a significant predictor of poor prognosis in stroke patients. These findings underscore the importance of early and effective interventions in patients with longer thrombi to improve clinical outcomes.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this project was provided by the Tianjin Key Medical Discipline (Specialty) Construction Project (TJYXZDXK-001A).
