Endovascular therapy versus medical management for mild strokes due to medium and distal vessel occlusions

Abstract

Background

Distal and medium vessel occlusion (DMVO) strokes account for up to 40% of all acute ischemic strokes, many of which have a low initial deficit severity. Increasingly, endovascular treatment (EVT) is considered for these patients. However, there is a paucity of data comparing potential safety and efficacy of EVT versus best medical management (BMT) in these patients, particularly when presenting with mild symptoms.

Methods

We retrospectively analyzed consecutive DMVO patients with NIHSS ≤6 who presented between January 2018 and December 2021 within 24 h from last seen well. We compared early and late clinical outcomes, as well as safety outcomes using multivariable linear and logistic regression analyses to determine whether EVT was independently associated with the outcomes of interest.

Results

We included 80 subjects that fulfilled our study criteria (n = 41 BMT and n = 39 EVT). On adjusted analyses, when comparing EVT and BMT groups, there was no statistical difference in the NIHSS at discharge (Coefficient = −0.6, 95%-CI = −3.45–2.26, p = 0.678) or in the modified Rankin scale (mRS) at discharge (Coefficient = 0.37, 95%-CI = −0.52–1.25, p = 0.408). Furthermore, there was no difference in the 3-month mRS defined as good (mRS 0–2) (OR = 0.56, 95%-CI = 0.17–1.83, p = 0.341) or excellent (mRS 0–1) (OR = 0.45, 95%-CI = 0.15–1.34, p = 0.152). Finally, there was no difference in the incidence of all-type parenchymal hemorrhage (OR = 0.9, 95%-CI = 0.22–3.62, p = 0.879).

Conclusions

We found that among DMVO patients with low admission NIHSS treatment with EVT versus BMT was associated with similar outcomes. Larger prospective studies are required to confirm these findings.

Introduction

Distal and medium vessel occlusion (DMVO) strokes account for approximately 25–40% of all acute ischemic strokes. ¹ Although outcome is generally better as compared to proximal large-vessel occlusion (LVO) strokes, prior studies indicated that DMVO strokes are associated with long-term disability in as many as 77% of patients^2,3 and one in three patients with mild strokes will not be functionally independent at 90 days. ⁴ In patients eligible for intravenous thrombolysis with recombinant tissue plasminogen activator (rtPA) lack of recanalization with treatment has been noted in up to 66% of DMVO strokes. ⁵

Whereas multiple randomized controlled trials showed the superiority of endovascular therapy (EVT) over best medical management (BMT) in patients presenting with proximal LVO involving the anterior circulation,^6,7 it is presently unclear whether this also translates to patients with DMVO, particularly in the presence of a low deficit burden. In patients with an M2 branch occlusion, clinical trial data suggests that EVT is associated with high reperfusion rates and better functional outcome with relatively low procedural risk.^8–10

However, data from retrospective studies and meta-analyses investigating efficacy and safety of EVT in all DMVO strokes combined is conflicting and limited as most studies did not compare outcomes to BMT.^11–15 Moreover, most of these studies included patients with an National Institute of Health Stroke Scale (NIHSS) of >6 and to our knowledge one study examined the association of EVT versus BMT with outcome in DMVO patients with low NIHSS. ¹⁶ A better understanding of this issue is important because approximately one-third of DMVO stroke patients present with a low NIHSS (≤6). ³

A better understanding as to whether EVT is beneficial in patients with DMVO in the setting of a low NIHSS is important because a significant subset of patients suffer disabling deficits, yet procedural risk may be increased due to the target vessels’ small caliber, tortuosity, and thinner walls. To better understand this issue, we conducted a retrospective single-center study to compare clinical and safety outcomes of DMVO patients with NIHSS ≤ 6 that were treated with EVT versus BMT.

Methods

Study cohort

We retrospectively analyzed consecutive patients who presented to the University of Massachusetts Memorial Medical Center as acute stroke activations between January 2018 and December 2021. Data collection and analysis was conducted between December 2021 and May 2022. Our investigation was approved by the local institutional review board with a waiver of informed consent.

During the study period, a total of 3930 patients were evaluated by our stroke team for acute neurological symptoms that had started within 24 h from the last see well. Of these, we excluded 2412 patients with stroke mimics and transient ischemic attacks and 1323 patients with either documented LVOs or no evidence of vessel occlusions on imaging. We excluded 115 patients with NIHSS score >6, as well as patients with ASPECT score of 6 or less. Of the included 80 subjects that fulfilled our study criteria of having NIHSS less or equal to 6 and evidence of a medium or distal vessel occlusion on imaging, with an ASPECT score of >6 and presenting within 24 h from last seen well, 41 were treated with BMT and 39 underwent EVT. Figure 1 summarizes the patient flow chart.

Figure 1.

Study patient flow chart.

For primary analysis, patients were stratified to BMT versus EVT as based on the original treatment plan after the initial head imaging and clinical assessment were completed (intent to treat). For secondary analysis, we re-assigned one EVT patient to the BMT group (the occlusion was considered too distal for safe intervention) and excluded one patient who required bailout intracranial stenting due to vessel re-occlusion within 24 h).

Clinical characteristics, stroke presentation and stroke etiology

Stroke-trained physicians certified in NIHSS assessed all patients. The early ischemic lesion burden was determined on the initial head CT using the Alberta Stroke Program Early CT Score (ASPECT). CT perfusion (CTP) was done at the treating physicians’ discretion. Where completed, we determined the presence of an ischemic penumbra, defined as an area with prolonged mean transit time or Tmax, and normal cerebral blood volume. ¹⁷ The final decision for EVT versus BMT was reached by consensus between the neurointerventional and stroke teams. Patients selected to the BMT group received standard treatment with intravenous rtPA (if applicable) followed by standard antiplatelet or anticoagulation treatment for secondary prophylaxis, depending on the stroke etiology.

Medical records were reviewed to collect information on demographics (age, gender), vascular risk factors (hypertension, hyperlipidemia, diabetes), atrial fibrillation, pre-admission medications (antiplatelet agents and/or anticoagulants), and admission glucose level. The stroke etiology was determined based on the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification. ¹⁸

Distal and medium vessel classification

In accordance with prior classifications, we defined DMVO stroke as a stroke caused by an occlusion of the M2–4 segments of the medium cerebral artery (MCA), A2-3 segments of the anterior cerebral artery, P1-2 segments of the posterior cerebral artery, posterior inferior cerebellar artery, anterior inferior cerebellar artery, or superior cerebellar artery. ⁵

Principles of EVT selection

Patients were selected for EVT by the treating team if they presented within 24 h from last seen well, had a DMVO causing a measurable disabling neurological deficit, ASPECT score > 6, and in select patients, evidence of ischemic penumbra on CTP. Patients who were evaluated with CTP but did not show evidence of penumbra received BMT. A significant neurological symptom was generally considered to be a deficit in the language and/or motor area. The technical feasibility and complexity of the case based on the occlusion location and patient anatomy were carefully considered by the neurointerventional team when deciding between treatment modalities. Ultimately, the decision to offer EVT to a patient was shared between the treating stroke neurologist and neurointerventionalist after considering the patient's symptoms and presumed technical feasibility and further informed by the expressed wishes of the patient or their legally authorized representative.

Outcomes

Outcomes of interest were the NIHSS at day 1 and at discharge, the change in NIHSS from admission to discharge (ΔNIHSS; whereby larger numbers indicate greater improvement in the NIHSS deficit and negative numbers indicate worsening from baseline). For patients dying in house or for whom the goals of care changed to withdrawal of life-sustaining therapy (WLST), the last documented NIHSS before death/WLST transition was reported as discharge NIHSS.

The modified Rankin scale (mRS) at 90 days was assessed by a stroke-trained physician or stroke study nurse certified in mRS via in-person or phone interview using a simplified mRS questionnaire. ¹⁹ In one patient lost to follow-up, we imputed the 90-day mRS by carrying forward the discharge mRS. We defined a good functional outcome as mRS 0–2 and an excellent functional outcome as mRS 0–1 or return to baseline.

Clinical deterioration was defined as an increase in the NIHSS by 4 or more points during the hospital admission.

Intracranial hemorrhage evaluation

The occurrence of intracerebral hemorrhage (ICH) was recorded if noted on a follow-up image within the first 36 h. Intracranial hemorrhage was classified on follow-up head CT or MRI obtained within 36 h using the European Cooperative Acute Stroke Study II (ECASS II) definition. ²⁰

Statistics

Continuous variables are reported as median (25th–75th percentile). Categorical variables are reported as proportions. Between group comparison for continuous and ordinal variables were made with Mann–Whitney U test. Categorical variable were compared using the χ²-test and Fischer exact test as appropriate. A two-sided p < 0.05 was considered statistically significant.

To determine whether EVT (versus BMT) was independently associated with each outcome of interest, we created separate multivariable logistic (90-day mRS 0–1, 90-day mRS 0–2, death, parenchymal hemorrhage) and linear (mRS at discharge, mRS at 90 days, NIHSS at day one and at discharge, change in NIHSS between day 1 and day 0 and between discharge day and day 0) regression models. All models were adjusted for history of diabetes, antiplatelet use, oral anticoagulant use, pre-stroke mRS, ASPECTS, DMVO branch (MCA versus other), age, and use of intravenous thrombolysis. To create parsimonious multiple logistic regression models, co-variates were sequentially removed (likelihood ratio) from the models at a significance level of 0.1. Collinearity diagnostics were performed (and its presence rejected) for all models. Model calibration was assessed by Hosmer-Lemeshow test and model fit determined by examining the −2 log-likelihood statistic and its associated chi-square statistics.

All statistical analyses were performed using IBM SPSS Statistics Version 22 (IBM, Armonk, NY, USA).

Results

Study cohort

Figure 1 summarizes the patient flow chart. Overall, we included 80 subjects that fulfilled our study criteria. Of these, 41 were treated with BMT and 39 were selected to undergo EVT.

Patient characteristics associated with selection for EVT

Table 1 summarizes the baseline characteristics of the included patients stratified according to EVT versus BMT. Overall, clinical characteristics were similar between groups except for more frequent diabetes (p = 0.035) and antiplatelet use (p = 0.045) as well as less frequent anticoagulation (p = 0.019) in the BMT group. Moreover, subjects selected for EVT had a lower pre-stroke mRS (p = 0.025), a lower ASPECT score on the admission head CT (p = 0.044) and more frequently had an MCA branch occlusion (p = 0.036).

Table 1.

Baseline characteristics.

	All patients	BMT	EVT
Characteristics	(n = 80)	(n = 41)	(n = 39)	p-value
Age	67 (59–78)	68 (60–80)	65 (58–78)	0.350
Sex
Male	42 (52.5)	20 (48.8)	22 (56.4)	0.512
Female	38 (47.5)	21 (51.2)	17 (43.6)
Hypertension	52 (65)	29 (70.7)	23 (59)	0.350
Hyperlipidemia	57 (71.3)	29 (70.7)	28 (71.8)	1.000
Diabetes	19 (23.8)	14 (34.1)	5 (12.8)	0.035
Atrial fibrillation	26 (32.5)	11 (26.8)	15 (38.5)	0.341
Antiplatelet use	34 (42.5)	22 (53.7)	12 (30.8)	0.045
Oral anticoagulant use	14 (17.5)	3 (7.3)	11 (28.2)	0.019
Admission glucose	114 (97–133)	118 (97–168)	113 (97–129)	0.506
SARS-CoV-2	1 (1.3)	1 (2.4)	0 (0)	1.000
NIHSS on arrival	4 (2.25–5)	4 (2–4.5)	4 (3–5)	0.146
Systemic thrombolysis	16 (20)	6 (14.6)	10 (25.6)	0.270
Pre-stroke mRS	0 (0–1)	0 (0–1)	0 (0–0)	0.025
ASPECTS	8 (8–10)	9 (7–10)	10 (9–10)	0.044
CTP	33 (41.2)	19 (46.3)	14 (35.9)	0.49
Presenting symptoms				0.204
Speech	29 (36.3)	12 (29.3)	17 (43.6)
Motor	14 (17.5)	8 (19.5)	6 (15.4)
Speech and motor	30 (37.5)	15 (36.5)	15 (38.5)
Other	7 (8.8)	6 (14.6)	1 (2.6)
Site of arterial occlusion				0.188
M2	34 (42.5)	15 (36.6)	19 (48.7)
M3	17 (21.3)	7 (17.1)	10 (25.6)
M4	1 (1.3)	0 (0)	1 (2.6)
A2	2 (2.5)	2 (4.9)	0 (0)
A3	1 (1.3)	0 (0)	1 (2.6)
P1	11 (13.8)	7 (17.1)	4 (10.3)
P2	14 (17.5)	10 (24.4)	4 (10.3)
DMVO Branch				0.036
Middle cerebral artery	52 (65)	22 (53.7)	30 (76.9)
Other	28 (35)	19 (46.3)	9 (23.1)
Time from last seen well to groin puncture (minutes)			455 (231–780) a
Index stroke discovery				0.481
Unknown	34 (32.5)	19 (46.3)	15 (38.5)
Wake up	16 (20)	6 (14.6)	10 (25.6)
Witnessed	30 (37.5)	16 (39)	14 (35.9)
Stroke etiology				0.246
Cardioembolic	49 (61.3)	25 (61)	24 (61.5)
Large artery atherosclerosis	7 (8.8)	6 (14.6)	1 (2.6)
Other	11 (13.8)	5 (12.2)	6 (15.4)
Undetermined	13 (16.3)	5 (12.2)	8 (20.5)
Recanalization Score (TICI score)
TICI0	0		0
TICI 2A	1		1
TICI 2B	8		8
TICI 2C	9		9
TICI 3	20		20

Data are n (%) or median (25th–75th quartile). Bolded characteristics and numbers represent statistically significant values. BMT: best medical therapy, EVT: endovascular therapy, NIHSS: National Institutes of Health Stroke Scale, mRS: modified Rankin Scale, ASPECTS: Alberta Stroke Program early CT score, CTP: CT Perfusion, DMVO: distal and medium vessel occlusion, TICI: Thrombolysis in cerebral infarction.

^a Value includes patient who did not undergo intervention.

Clinical outcomes in EVT versus BMT groups

Overall, there was no statistical difference between NIHSS at day 1 (p = 0.654), NIHSS at discharge (p = 0.244), and the ΔNIHSS from day 0 to day 1 (p = 0.340) or the ΔNIHSS from day 0 to discharge day (p = 0.080) (Table 2). There was no difference in the discharge (p = 0.895) and 3-month (p = 0.957) mRS between groups. Likewise, there was no between-group difference in the frequency of excellent (p = 0.655) and good (p = 1.000) outcomes at 3 months (Table 2). Two (4.9%) BMT and four (10.3%) EVT subjects died by 3 months (p = 0.426). All patients that died had WLST without neurological worsening from their index stroke.

Table 2.

Patient outcomes.

	All patients	BMT	EVT	p-value
Characteristics	(n = 80)	(n = 41)	(n = 39)
NIHSS day 1	3 (1–4)	2 (1–4)	3 (1–4)	0.654
ΔNIHSS between day 1 and day 0	1 (0–2)	1 (0–2)	1 (0–3)	0.340
ICH at 36 h (all combined)	25 (31.3)	10 (24.2)	15 (38.5)	0.229
Parenchymal hemorrhage (PH1 + PH2)	9 (11.3)	4 (9.8)	5 (12.8)	0.734
Clinical deteriorationa	7 (8.8)	2 (4.9)	5 (12.8)	0.258
NIHSS at discharge	2 (0–4)	2 (0.5–4)	1 (0–4)	0.244
ΔNIHSS between discharge day and day 0	1.5 (0–3)	0 (0–2)	2 (0–4)	0.080
mRS at discharge	2 (1–3)	2 (1–3)	2 (0–3)	0.895
mRS at 3 months	1 (0–3)	1 (0–3)	2 (0–3)	0.957
Good outcome (mRS 0–2) at 3 months				1.000
mRS 0–2	56 (70)	29 (70.7)	27 (69.2)
mRS 3–6	24 (30)	12 (29.3)	12 (30.8)
Excellent outcome (mRS 0–1) at 3 months				0.655
mRS 0–1	42 (52.5)	23 (56.1)	19 (48.7)
mRS 2–6	38 (47.5)	18 (43.9)	20 (51.3)
Death at 3 months	6 (7.5)	2 (4.9)	4 (10.3)	0.426

Data are n (%) or median (25th–75th quartile). BMT: best medical therapy, EVT: endovascular therapy, NIHSS: National Institutes of Health Stroke Scale, ICH: intracranial hemorrhage, PH: parenchymal hemorrhage (classified according to the ECASS II classification). ΔNIHSS difference in NIHSS (larger numbers signify greater improvement), mRS: modified Rankin Scale. ^aClinical deterioration: worsening by NIHSS of 4 points or more during hospital admission.

There was no statistical difference in outcomes of interest between patients with MCA branch occlusion versus a non-MCA branch occlusion (p > 0.05, each). Among patients with stroke due to large artery atherosclerosis, the improvement of the NIHSS from day 0 to discharge day was significantly greater in patients undergoing EVT (p = 0.020). However, no significant difference was found in any of the other outcomes of interest.

We constructed separate multivariable linear and logistic regression models to determine whether EVT was independently associated with the outcomes of interest. All models were adjusted for diabetes, antiplatelet use, oral anticoagulant use, pre-stroke mRS, ASPECTS, DMVO branch (MCA versus other), age, and use of intravenous thrombolysis. In these analyses, we found no independent association of EVT with the outcomes of interest (Supplemental Table 1).

Clinical outcomes in EVT versus BMT groups based on ischemic penumbra on CTP

Overall, 33 included patients underwent CTP (n = 14 in the EVT and n = 19 in the BMT group, p = 0.495) (Table 1). Patients without a penumbra (n = 4) did not undergo EVT. We found that among patients that underwent CTP, EVT was associated with a significantly greater reduction in the NIHSS by discharge (p = 0.018), whereas no difference between groups was present among subjects that did not undergo CTP. More important, among patients with penumbra, both the discharge NIHSS (p = 0.031) and the NIHSS improvement from admission to discharge (p = 0.047) were greater in EVT versus BMT.

Clinical outcomes and EVT recanalization status

Among patients who underwent EVT, we did not find a significant correlation between the TICI score and any of the studied outcomes of interest (Spearman rank correlation; p > 0.1, each). Also, we found no significant difference between the outcomes of interest when we dichotomized the TICI score to 0–2a vs 2b-3 and 0–2b vs 2c-3 (p > 0.05, each).

Incidence of ICH and clinical deterioration

With regard to the safety outcomes, there was no difference in the number of all types of ICHs combined between the two groups (p = 0.229). Furthermore, there was no difference specifically in the number of parenchymal hemorrhages solely (p = 0.734) between the EVT and BMT groups. No difference was noted in the number of patients who had clinical deterioration p = 0.258 (Table 2).

In secondary analyses excluding the subject that underwent repeat endovascular intervention and reassigning, one patient to BMT the overall results did not meaningfully change with exception that there was no longer a significant difference in the baseline ASPECT between groups and the ΔNIHSS between day 0 and discharge day was higher in EVT patients (p = 0.019) (data not shown).

Discussion

Current guidelines recommend EVT for patients with acute occlusion of the internal carotid artery and/or MCA M1 segment, and an NIHSS ≥ 6.^6,21,22 Several prior analyses suggest benefit of EVT in patients with occlusions of the MCA M2 segment.^5,8,23 However, little it is known about the benefits of EVT for patients who present with all DMVO strokes and particularly those with a low NIHSS score. The most important findings from our study were that early and 3-month clinical efficacy and safety outcomes were comparable between EVT and BMT. These observations are in line with several prior studies that suggested that EVT in patients with acute DMVO in the anterior and posterior circulation combined across all NIHSS severities is safe and feasible but results in comparable outcomes as compared to BMT.^11–13,24 Our results expand on the only study that directly compared the outcomes of patients presenting with DMVO and low NIHSS strokes treated with EVT versus BMT (specifically rtPA). ¹⁶ Notably, in our patient cohort, the distribution of patients to BMT versus EVT was almost equal, which may reflect the uncertainties regarding the best treatment approach in this particular patient population at a time when randomized controlled trial results are still pending to answer which treatment approach in this specific patient population may be preferred. Our analysis includes in the control BMT group patients both eligible and ineligible for rtPA, which is important since many DMVO and low NIHSS stroke patients are ineligible for rtPA administration at the time of presentation.

Nevertheless, the comparable clinical outcome between BMT and EVT groups needs to be interpreted cautiously because subjects selected for EVT had more frequent involvement of MCA branches. Thus, it remains possible that these patients were at increased risk for secondary deterioration, which was averted by EVT. Conversely, EVT patients had a numerically greater utilization of rtPA (26% versus 15%) as compared to patients selected for BMT. Accordingly, it remains uncertain as to whether there is a subpopulation of DMVO patients with low NIHSS that would benefit from EVT (e.g. M2 occlusion ineligible for rtPA).

With respect to safety outcomes, we noted no difference in the incidence of all parenchymal hemorrhages combined between the EVT and BMT groups. Two patients in the BMT group and five patients in the EVT group experienced clinical deterioration. However, among the EVT subjects, one had a clot too distal and did not undergo EVT and the other had bailout stenting due to vessel re-occlusion. However, when we reanalyzed our dataset by reassigning the first patient to the BMT group and removing the second patient, results did not change meaningfully and there was no significant difference in the incidence of all-type ICH, parenchymal hemorrhage, as well as the discharge and 3-month mRS.

Limitation of our study relate to its single center, retrospective design. Although our results are in general agreement with recently published data on DMVO patients treated with EVT, the modest number of subjects in our study limits the statistical power of our analyses and results should be considered as hypothesis generating. For example, we noted that more patients with aphasia symptoms were treated with EVT. While adjustment of our logistic regression model for the type of neurological deficit on presentation (speech, motor, or both) did not change the result of our analysis it is conceivable that a larger, adequately powered study could show that patients with certain neurological deficits may benefit from EVT. Moreover, it is important to recognize that the 90-day mRS may not fully reflect the degree of recovery from a specific neurological deficit. Thus, it is very well possible that patients subjected to EVT had a more complete recovery trajectory, which we were unable to determine in our study.

Further large-scale studies are required to clarify efficacy and safety of EVT versus BMT in DMVO patients with low clinical deficit severity. While single-institution retrospective analysis generally has the challenge of not generating high enough subject numbers for a powerful statistical analysis, a randomized control trial could offer a solution to this problem. Current randomized control trials are aiming to enroll between 200 and 500 patients. Our results may inform power calculations of future studies aiming to determine efficacy and safety of EVT in patients with DMVO and mild deficit severity. In addition, implementing measures to ensure equal allocation of specific neurological deficits such as aphasia or motor weakness to the study arms should be considered. Some ongoing trials expected to shed light on EVT vs BMT in patients with medium and distal vessel occlusions include ENDOLOW, ESCAPE-MeVO, DISCOUNT, and DISTAL.