PDF
Citation
Print
Introduction
The management of acute ischemic stroke (AIS) attributable to distal medium vessel occlusion (DMVO) presents a significant challenge in the field of neurointerventional therapy. Mechanical thrombectomy (MT), while established as the standard of care for large vessel occlusion (LVO) strokes [1-5], exhibits uncertain efficacy in DMVO cases [6]. This uncertainty is noteworthy as DMVOs are estimated to account for 25%–40% of AIS instances with discernible arterial blockage.
Traditionally, DMVOs have been scarcely represented in clinical trials and practice, leading to a significant gap in evidence-based guidelines for managing these patients. However, a number of studies are now investigating the safety and effectiveness of MT for DMVO strokes, including DISTAL (EnDovascular Therapy Plus Best Medical Treatment [BMT] vs. BMT Alone for MedIum VeSsel Occlusion sTroke, ClinicalTrials.gov Identifier: NCT05029414), DISTALS (Distal Ischemic Stroke Treatment With Adjustable Low-profile Stentriever, NCT05152524), DISCOUNT (Evaluation of Mechanical Thrombectomy in Acute Ischemic Stroke Related to a Distal Arterial Occlusion, NCT05030142), ESCAPE-MeVO (EndovaSCular TreAtment to imProve outcomEs for Medium Vessel Occlusions, NCT05151172), and FRONTIER-AP (Randomized controlled trial of the clinical outcome and safety of endovascular vs. standard medical therapy for stroke with medium sized vessel occlusion) [7,8], which explore various aspects of endovascular therapy in comparison to or alongside best medical treatment. Despite the potential benefits of MT, concerns arise over its application in DMVO due to the smaller size and complex structure of distal cerebral arteries, which may complicate the reperfusion process. Consequently, this raises significant safety and efficacy concerns regarding the use of MT in treating DMVOs [6,9].
A central point of uncertainty is whether MT can improve clinical outcomes compared to those attainable with intravenous thrombolysis (IVT) [10]. However, advancements in endovascular technologies, including retriever and aspiration devices, have expanded the application of MT in treating distal vessel occlusions [9].
In response to this clinical gap, our multicenter cohort study aimed to evaluate the safety and efficacy in DMVO patients undergoing treatment with either IVT alone or a combination of MT and IVT (MT-IVT).
Methods
Study design
This study represents a detailed analysis within the Multicenter Analysis of primary Distal medium vessel occlusions: effect of Mechanical Thrombectomy (MAD-MT) registry [8,11-14]. This research adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [15,16].
Ethical approval and patient consent
Ethical clearance for this retrospective study was obtained from the institutional review board or local ethical standards committee at each of the 37 participating centers across North America, Asia, and Europe. Considering the retrospective nature of our analysis, the requirement for informed consent was waived. The dataset supporting the conclusions of this article is available from the corresponding author upon reasonable request.
Study population and selection criteria
Our target population comprised patients who experienced AIS due to DMVO. Inclusion criteria were: (1) complete data for 3-month modified Rankin Scale (mRS) score, intervention type, and hemorrhagic complications including symptomatic intracerebral hemorrhage (sICH); and (2) patients treated with either MT-IVT or IVT alone. We excluded patients who received MT alone or did not receive either MT or intravenous tissue plasminogen activator (IV-tPA) treatment.
Data collection and review process
Data were collected from September 2017 to July 2023, encompassing consecutive patients treated with MT or IVT for primary medium proximal vessel occlusion (M2, A1, P1) or primary medium distal vessel occlusion (M3/M4, A2/A3, P2/P3) as defined by Saver et al. [9]. This included both prospective data collection and retrospective analysis. Local neurointerventionalists initially evaluated angiographic treatment success, subsequently contributing their data to the MAD-MT consortium. The data were self-reported by each participating center.
Variables and data recording
We meticulously documented a comprehensive set of baseline clinical and demographic characteristics, including sex, age, medical history (hypertension, hypercholesterolemia, diabetes mellitus, atrial fibrillation), smoking status, pre-stroke mRS score, location of the occluded vessel, National Institutes of Health Stroke Scale (NIHSS) score upon presentation, and baseline Alberta Stroke Program Early Computed Tomography Score (ASPECTS). Patients were also categorized based on their occlusion site during initial angiography into medium proximal (M2, A1, P1) and medium distal vessels (M3/M4, A2/A3, P2/P3).
We additionally recorded information on antiplatelet and anticoagulant medication use, treatment approach (mothership vs. drip-and-ship), timing from symptom onset to arterial puncture and recanalization, vital signs, glycemic levels, anesthesia type, access site, and post-MT imaging modalities.
Outcome measures
The primary outcome focused on functional independence (mRS 0–2), with secondary outcomes including an excellent outcome (mRS 0–1) and mortality (mRS 6). Safety outcomes centered on types of intracerebral hemorrhage (ICH) and sICH, defined in accordance with “The Heidelberg Bleeding Classification.” [17]
Procedural and technical aspects
Treatment modalities included MT-IVT and IVT alone. The choice of treatment, procedural techniques (aspiration, stent retriever, combined techniques like Solumbra), access site (femoral or radial artery), and number of passes were at the discretion of the treating clinicians, aligned with institutional protocols and guidelines.
Statistical analysis
Statistical analysis was conducted using R (version 4.2.2; R Foundation for Statistical Computing, Vienna, Austria). For categorical variables, we calculated frequencies and percentages and employed the χ2 test for comparison. In instances where cell counts were fewer than five, Fisher’s exact test was utilized. Continuous variables were summarized using median values and interquartile ranges (IQRs) and were compared using the Mann-Whitney U test to assess differences between groups.
Propensity score matching (PSM) controlled for potential confounders [18,19]. Propensity scores were estimated using logistic regression, and an optimal matching algorithm was employed with a 1:1 ratio. Variables likely to affect the outcomes according to previous studies [20-22] were included in the PSM, which included age, sex, baseline NIHSS, pre-admission mRS scores, history of antiplatelet and anticoagulant medication use, pertinent comorbidities (diabetes mellitus, hypertension, hypercholesterolemia, atrial fibrillation), and occlusion location. Covariate balance postPSM was assessed using the standardized mean difference, with <0.1 indicating effective balance. A P-value of 0.05 or less was considered statistically significant.
Missing data
In this study, we encountered a proportion of missing data, as documented in Supplementary Figure 1. Missing data were assumed to be either completely missing at random or randomly missing. To avoid bias and reduction in statistical power, we included cases with missing values by employing multiple imputation using chained equations for missing data handling [23], producing fifty imputed datasets with five iterations each. Imputations were performed on matched data only, with no imputation on cases with missing outcomes (90-day mRS, mortality, hemorrhagic complications).
Post-imputation, logistic regression models were applied to determine the relationship of IVT alone versus MT-IVT and observed outcomes. Multivariable models were adjusted to include age, sex, comorbidities (diabetes, hypertension, hypercholesterolemia, atrial fibrillation), baseline NIHSS, occlusion location, ASPECTS, and pre-stroke use of antiplatelet or anticoagulant agents.
Results
Patient characteristics
Before propensity score matching, our study involved 1,057 patients, with 715 undergoing MT-IVT and 342 receiving IVT alone (Figure 1). Prior to matching, the MT-IVT group displayed a median age of 74 years, similar to the IVT alone group (P=0.54). Notably, there was a lower prevalence of hypercholesterolemia in the MT-IVT group (34%) compared to the IVT alone group (51%; P<0.001). Similarly, the rate of atrial fibrillation was higher in the IVT alone group (37% vs. 30% in MT-IVT; P=0.02). Rates of hypertension were comparable, though the IVT alone group had a higher prevalence of diabetes (27% vs. 19% in MT-IVT; P=0.009). Smoking history and previous use of antiplatelet or anticoagulant drugs were also similar between the groups. Significant differences were noted in baseline mRS scores and site of initial occlusion, with more medium vessel (M2, P1, A1) occlusions in the MT-IVT group (81% vs. 74% in IVT alone; P=0.011). The median ASPECTS score was 9.00 in both groups, but the baseline NIHSS score was higher in the MT-IVT group (median 12 vs. 6 in IVT alone; P<0.001) (Supplementary Table 1).
After propensity score matching, a total of 640 patients were analyzed, and the cohorts were well balanced across all measured variables. Sex distribution, median age, and prevalence of hypercholesterolemia, hypertension, diabetes, atrial fibrillation, smoking history, and previous medication use showed no significant differences between the two groups. Baseline mRS scores, site of initial occlusion, ASPECTS scores, and baseline NIHSS scores were also similar (Table 1).
Periprocedural details
Table 2 provides comprehensive periprocedural data for the study cohort post-matching. In the MT-IVT group, the first-line procedural techniques were distributed as follows: aspiration in 23%, a combination of techniques in 64%, and stentretriever in 13%. The median times from onset to IVT needle were similar in both groups, recorded as 155 minutes and 164 minutes, respectively. Additionally, in the MT-IVT group the median time from symptom onset to recanalization was 283 minutes. Periprocedural data before PSM are documented Supplementary Table 2.
Outcomes after matching
Following the matching process (Table 3), the median number of passes required for thrombectomy in the MT-IVT group was 1.00. On the first day post-admission, the median NIHSS score was 3 (IQR 1–8) across the entire cohort (IQR: 1–11 in MT-IVT vs. 1–6 in IVT alone; P=0.061). In terms of angiographic outcomes, a high rate of successful recanalization (modified Thrombolysis in Cerebral Infarction [mTICI] 2b–3) was achieved in 91% of the MT-IVT group. Complete reperfusion (mTICI 2c–3) was accomplished in 61% of this group.
Functional outcomes at 90 days, assessed by mRS scores, showed no significant difference between the two groups for achieving mRS 0–1 (45% in MT-IVT vs. 50% in IVT alone; P=0.21) or mRS 0–2 (62% in MT-IVT vs. 63% in IVT alone; P=0.74). Mortality rates at 90 days were 14% in the MT-IVT group and 11% in the IVT alone group, with no statistically significant difference (P=0.4) (Figure 2A).
The incidence of sICH was similar between the two groups (8.3% in MT-IVT vs. 6.9% in IVT alone; P=0.59). However, any type of ICH was significantly higher in the MT-IVT group (32% vs. 17% in IVT alone; P<0.001). Notably, hemorrhagic infarction type 1 (HI1) and subarachnoid hemorrhage (SAH) were significantly more common in the MT-IVT group (P<0.001 for both). Other ICH types, including HI2, PH1, and PH2, did not significantly differ between the groups (Figure 2B). Perforations were observed solely in the MT-IVT group, with a total of 14 events. These perforations were associated with 12 instances of any type of ICH, including three cases of sICH and five cases of SAH. Other complications, such as embolization in new territories and artery dissection, were infrequent and did not differ significantly between the groups. Outcomes before matching were documented in Supplementary Table 3.
Logistic regression analysis
The unadjusted and adjusted logistic regression models (Table 4) reinforced these findings. In the adjusted model, the odds ratios for 90-day mRS 0–1 and mRS 0–2 were 1.21 (95% CI 0.81 to 1.79; P=0.35) and 1.00 (95% CI 0.66 to 1.51; P>0.99), respectively, indicating no significant difference in functional outcomes between the two treatment modalities. Similarly, the adjusted model showed no significant differences in mortality rates (OR 0.75, 95% CI 0.44 to 1.29; P=0.3) and rates of sICH (OR 0.78, 95% CI 0.42 to 1.45; P=0.44).
However, the risk of any type of ICH was significantly lower in the IVT alone group, with an OR of 0.43 (95% CI 0.29 to 0.63; P<0.001). Specifically, hemorrhagic infarction type 1 (HI1) showed a significantly reduced risk in the IVT alone group (OR 0.17, 95% CI 0.08 to 0.36; P<0.001), as did subarachnoid hemorrhage (SAH) (OR 0.17, 95% CI 0.07 to 0.41; P<0.001). Conversely, hemorrhagic infarction type 2 (HI2) was more often in the IVT alone group (OR 4.37, 95% CI 1.42 to 13.5; P=0.01), whereas parenchymal hematoma types 1 and 2 (PH1 and PH2) showed no significant difference in odds between the groups.
Sensitivity analysis
Sensitivity analysis using pre-imputation and imputed models showed similar trends, confirming the robustness of the findings (Supplementary Table 4).
Subgroup analysis
Additional subgroup analyses comparing treatment outcomes between the “mothership” scenario (n=366) and “drip-and-ship” scenario (n=166) showed no significant differences in achieving mRS 0–1, mRS 0–2, mortality, and sICH. However, there was a significant difference in the rate of ICH between these subgroups (P=0.018), with the “drip-and-ship” group having an increased risk of ICH in the MT-IVT treatment (Figure 3). Subgroup analysis based on occlusion location medium (n=492) and distal (n=148) showed no differences in outcomes between the subgroups (Figure 4).
Discussion
In this study, we sought to determine the safety and efficacy of MT-IVT compared to IVT alone in patients with AIS due to DMVO. We found similar rates of 90-day good and excellent functional recovery and mortality between the two treatment strategies. In addition, there was an increased incidence of any ICH in the MT-IVT group, despite no significant difference in sICH.
DMVOs represent a challenging subset of AIS, with existing recommendations primarily targeting larger vessel occlusions [24]. Current recommendations from the American Stroke Association predominantly address large vessel occlusion strokes. They suggest that MT for MCA M3 occlusions may be reasonable, while no specific recommendations are provided for M4 occlusions or DMVOs [16,25]. This is echoed in European guidelines, which do not provide specific MT recommendations for DMVO [26].
Notably, our findings align with previous research. A recent meta-analysis of 2,469 patients, comparing endovascular therapy with best medical therapy in medium vessel occlusion AIS, found no significant difference in functional independence, mortality rates, and incidence of sICH between the two groups [27]. The observational study by Saber et al. [28] indicated a slightly higher rate of excellent mRS outcomes in the MT-treated group for primary DMVO stroke, without notable differences in complications. This is consistent with our findings, which suggest comparable efficacy of the two treatment strategies but without significant difference in sICH. Notably, the older age (median: 74, IQR: 63–83) in our study cohort compared to that in the study by Saber et al. [28] might have contributed to the higher mortality rate observed in our cohort, despite lower baseline NIHSS score. Consequently, the odds of mortality were similar in both studies.
Recent observational studies have also explored the bridging treatment approach (IVT followed by MT) and noted an enhanced reperfusion rate in MT without a significant difference in complication rates [29,30]. Consistent with these studies, we observed high successful (mTICI 2b–3) and excellent (mTICI 2c–3) recanalization rates with MT-IVT. Nevertheless, since we lack comparable data for the IVT alone group, further study is required to determine whether combined treatment is superior to IVT alone.
The similar outcomes in function and mortality observed across the treatment modalities may be attributed to the generally less severe clinical presentations of DMVO strokes in comparison with LVOs. This distinction is reflected in our PSM model, which demonstrated a tendency to select patients presenting with lower NIHSS scores at admission for the MT-IVT group. This selection bias suggests that our findings may primarily apply to patients experiencing mild to moderate DMVO strokes and may not be generalizable to those with severe strokes characterized by higher NIHSS scores. Furthermore, the applicability of the commonly utilized mRS scores in evaluating outcomes after DMVO strokes warrants further discussion, raising questions about the adequacy of these measures in capturing the full spectrum of patient experiences and outcomes following such interventions [28,31]. In a previous study, we demonstrated that there is no significant difference in the ICH rate (OR, 1.10; 95% CI 0.96–1.25), successful recanalization rate (OR, 2.22; 95% CI 0.61–8.03), or 90-day functional mRS outcomes between aspiration and stent-retriever thrombectomy in DMVO stroke patients [12], indicating that the choice of MT technique likely did not influence our findings.
Our findings indicate that while the rate of ICH in the MT-IVT group is elevated, it remains comparable to bleeding rates observed in prior studies focused on MT for acute M2 occlusion [32]. These hemorrhagic incidents are predominantly mild and asymptomatic. This observation is supported by the lack of significant difference in the occurrence of sICH between the treatment cohorts. The observed increase in any ICH in DMVO treated with MT may be attributed to the fragility of smaller-sized vessels. This risk is particularly notable in patients eligible for IVT, aligning with a meta-analysis that identified an elevated risk of ICH in patients treated with MT for distal MCA occlusions [31-34]. Moreover, certain MT-IVT subpopulations, such as patients treated in a “drip-and-ship” scenario, might be at increased risk of ICH [35,36]. Our subgroup analysis revealed a significant interaction (P=0.018), indicating that the MT-IVT group in the “drip-and-ship” scenario had a significantly increased risk of any ICH (OR 0.15, 95% CI 0.02–0.58; P=0.017). In contrast, there was no significant difference in the “mothership” scenario (OR 0.99, 95% CI 0.59–1.66; P=0.98). This increased risk is, in part, likely due to the longer time interval between IVT administration and MT. The subgroup analysis based on occlusion site, as shown in Figure 4, revealed no significant interaction differences between medium and distal occlusions across various outcomes, including 90-day mRS scores, mortality, and ICH complications. A notable finding was the significance of ICH in medium occlusions (OR 0.38, 95% CI 0.24–0.6, P<0.001) but not in distal vessel occlusions (OR 0.5, 95% CI 0.18–1.31, P=0.16). This discrepancy is likely attributable to the smaller sample size of the distal occlusion subgroup (n=148 vs. n=492 in medium occlusions), which may have limited the statistical power to detect significant differences. Consequently, the P-value for interaction was not significant (P=0.37). A similar observation applies to the 90-day mRS 0–1 outcome, where there was a nonsignificant interaction (P=0.18).
This study’s strengths include its extensive scope, incorporating large-scale data from multiple nations and centers, thereby enhancing the generalizability of our findings to real-world settings. However, there are limitations inherent to its retrospective design. Treatment decisions based on physician discretion and institutional protocols may introduce biases. Even with the application of propensity score matching, the potential for unmeasured confounders to influence outcomes remains. The variability in procedural techniques and the range of operator experience across participating centers might also limit the universality of our results. A notable limitation is the absence of long-term outcome data beyond the 90-day mark, which constrains our ability to comprehensively understand the sustained effects of MT on patients’ functional recovery and quality of life. Additionally, this study did not assess reperfusion rates in the IVT alone group, an aspect that could offer further insight into treatment effectiveness. Moreover, lack of perfusion scans and lack of data on the location of the ICH and perforations are other limitations. It is important also to highlight that the baseline NIHSS scores in our matched cohorts were relatively low in both groups. Specifically, the 25th percentile for the MT-IVT group was 4, indicating that some patients with NIHSS score of <4 were included. Clinical trials largely excluded these patients. According to current evidence including PRISMS (The Potential of rtPA for Ischemic Strokes With Mild Symptoms) and ARAMIS (Antiplatelet vs. R-tPA for Acute Mild Ischemic Stroke) trials, the standard of care for patients with low NIHSS scores, particularly up to an NIHSS of 5 is not clear yet, and might be better managed with dual-antiplatelet therapy rather than IVT [37,38]. Given these limitations, there is a clear need for prospective randomized clinical trials to better identify DMVO patient subgroups that are most likely to benefit from MT.
Conclusions
In conclusion, this comprehensive retrospective study provides valuable insights into the efficacy and safety of MT-IVT compared to IVT alone in DMVO acute ischemic stroke patients. The analysis indicated comparable 90-day mortality and functional outcomes (mRS 0–1 and 0–2) between the two treatment groups. However, a higher rate of any hemorrhagic transformation was noted in the MT-IVT group. These findings suggest that MT in combination with IVT for all DMVO stroke patients may not always be beneficial. Further randomized clinical trials are warranted to identify DMVO patients most likely to benefit from MT-IVT treatment.
XML Download