Intracranial internal carotid artery (ICA) occlusion accounts for up to 30% of anterior acute ischemic stroke (AIS) patients eligible for mechanical thrombectomy (MT) [1,2]. This occlusion leads to large clot burden, poor collateral flow, large ischemic lesions, and unfavorable outcome [3]. Around 40–60% of patients remain disabled and one-third die [4,5].

To achieve rapid complete recanalization, preferably in a single pass, known as first-pass effect (FPE) [6], MT for large-vessel occlusion (LVO) involves direct aspiration first-pass technique (ADAPT), stent retrievers (SRs), and combined contact aspiration and SR (CA+SR) [7,8]. SR thrombectomy uses a device to capture thrombi and restore blood flow. ADAPT technique achieves direct thrombus aspiration using large-bore catheters [9-11]. The combined approach was introduced to achieve higher recanalization rates through synergistic effects [12,13]. While studies on middle cerebral artery (MCA) occlusion suggest that combined approach improves FPE compared to SR alone [14-16], others show no difference [1,17-19]. Data comparing ADAPT and combined approach remain limited. Studies shown ADAPT is not inferior to combined thrombectomy for LVO stroke recanalization [10,20-22], with some reporting better outcomes and faster procedures with ADAPT [21,23]. Others suggest combined strategy improves firstpass recanalization versus CA alone [24].

There remains no consensus on optimal MT strategy for ICA occlusion [25-29]. Although patients with higher initial National Institute of Health Stroke Scale (NIHSS) score and ICA occlusion may benefit more from SR thrombectomy [30], current evidence lacks consistent findings regarding ADAPT efficacy versus SR or combined thrombectomy [25-29]. Available cohorts with acute intracranial ICA occlusions have small samples, leaving unclear which technique offers better outcomes.

We aimed to compared efficacy and safety of ADAPT and non-ADAPT first-line thrombectomy for intracranial ICA occlusion using contemporary device technology in routine practice.

We included 85 patients (47% females; median age, 75 [IQR: 68, 83] years). The most common stroke etiology was cardioembolic (66%), morphology as L-shaped occlusion (78%), and occlusion as post-communicating ICA (54%). Intravenous thrombolysis was given to 32 (38%) patients.

Sixty (70.6%) patients received first-line ADAPT, and 25 (29.4%) received first-line non-ADAPT (18 [21.1%] with SR+CA and 7 [8.2%] with SR alone). Balloon guide catheter (BGC) was used in 5 cases of non-ADAPT group. Baseline characteristics, outcomes are summarized in Table 1. Groups were comparable in demographic, etiological, clinical, and radiological data.

While successful recanalization rate at first pass (mTICI 2b-3) was similar between groups (30% ADAPT vs. 20% non-ADAPT) and higher after first-line technique (mTICI 2b-3 rate: 65% ADAPT vs. 88% non-ADAPT, P=0.040), ADAPT achieved higher final recanalization (mTICI 2c-3 rate: 75% vs. 52%, P=0.038; mTICI 2b-3 rate: 98.3% vs. 88%, P=0.074) with shorter procedure time (median [IQR]: 32 [22, 53] minutes vs. 60 [42, 89] minutes, P=0.001) (Table 1, Fig. 2). ADAPT required more rescue techniques (50% vs. 12%, P=0.010). Passes with firstline techniques were lower in ADAPT (median [IQR]: 2 [1, 3] vs. 2 [1, 5], P=0.010). Final mTICI 2c-3 was associated with lower embolism to new territory (5.1% vs. 22.2%, P=0.026), lower 24h-NIHSS (median [IQR]: 11 [3, 18] vs. 18 [8, 20], P=0.043), and lower NIHSS at discharge (median [IQR]: 6 [0, 10] vs. 12 [4, 17], P=0.444) than mTICI 0-2b (Supplementary Table 1).

No significant group differences occurred in procedural or hemorrhagic complications. ADAPT showed higher mRS score ≤2 at 3 months (47% vs. 22%, P=0.039) (Tables 1, 2; Fig. 3). Patients with favorable outcomes at 3 months had lower NIHSS scores at baseline, fewer passes, lower 24h-NIHSS and discharge scores, and lower rates of ICH and sICH (Table 2).

In the multivariate analysis (Table 3), ADAPT was associated with better functional outcomes, with a trend toward statistical significance (OR 5.10, P=0.075). The NIHSS score at discharge was the only significant predictor of functional outcome at 3 months (OR 0.68, P<0.001).

Intracranial ICA occlusion is common and poorly contrasted with other LVOs. In an era where endovascular therapy has become protocol for AIS patients with LVO following favorable trial outcomes, no MT technique has proven more effective and safer than others for intracranial ICA occlusion. Comparing clinical and angiographic outcomes of endovascular approaches interests the medical-scientific community as technology has brought new-generation devices to market.

We found successful recanalization (TICI 2b-3) at procedure end in nearly all ADAPT group patients and 90% in the non-ADAPT group. These results exceed previous studies ranging between 56% and 89% [2,12], and contrast with others showing no differences among endovascular techniques [26,28]. In a recent meta-analysis of 759 patients from 4 studies with intracranial ICA occlusion, Li et al. [29] demonstrated that SRs outperformed direct aspiration in achieving final successful reperfusion (mTICI 2b-3). Although studies showed variability in populations, comorbidities, and treatment strategies, limiting comparability, 3 studies indicated SR superiority [26-28], while in the fourth, aspiration surpassed SR without BGC use [25]. No differences existed in mTICI 2c-3 and TICI 3 reperfusion rates. Puncture-to-reperfusion time, complication rates, intracerebral hemorrhage, and embolization to new territory were similar between groups. The authors noted limitations from unbalanced sample sizes between groups, potentially influencing results. The advancement of technology may have contributed to the improved outcomes of our cohort, which was treated with the most recent-generation devices over a period of 5–6 years. However, successful recanalization rate at first pass (mTICI 2b-3) rates reached only one-third of cases, likely due to high thrombus burden in intracranial ICA occlusion [2-5]. In our cohort the rate of recanalization achieved with first-line technique was higher in non-ADAPT group, though attempts were significantly greater in this group. While the results of the meta-analysis for ICA occlusion showed no differences in MT times for both procedures [29], our results show that faster recanalization was achieved with ADAPT and these differences persisted despite greater rescue technique needs in ADAPT (50%) versus non-ADAPT (12%). Therefore, our results suggest that the ADAPT technique used as a first-line approach achieves better and faster recanalization than other techniques, even when switching to a rescue technique is necessary.

The complication rate was lower than reported in previous studies. A prospective study from MR CLEAN registry showed 48% hemorrhage rate with 7% sICH in intracranial ICA occlusions [31]. Our study showed 33% overall rate and 4.7% PH2 hemorrhage. These results may be explained by intracranial ICA occlusions presenting large ischemic brain tissue volumes despite reperfusion therapy, increasing hemorrhagic transformation risk [32]. We found lower embolism incidence in the ADAPT group (8.3%) than non-ADAPT group (16%), though not significant. Our findings do not support that aspiration might disrupt the clot, causing downstream embolization, whereas SR could mitigate thrombus disruption [33].

A key finding was the ADAPT group’s more favorable outcome at 3 months. However, in the multivariate analysis, the association between ADAPT and functional outcomes did not reach statistical significance, although a near-significant trend was observed. We believe these results may have been influenced by insufficient sample size to identify outcome predictors. Nevertheless, the functional outcome outperformed those of other published studies, especially older series that reported 3-month functional independence rates of 28–31% [12,28]. Such results may be attributed to the shorter recanalization times and higher reperfusion rates achieved with ADAPT, both associated with better long-term outcomes and lower 90-day mRS scores [27,28].

We found that NIHSS score at discharge was the only independent predictor of decreased 3-month functional independence. This finding is aligned with recent studies [34,35] that demonstrated poor functional outcome was significantly associated with demographic factors, high mRS score at admission, living alone conditions, and high NIHSS score at discharge, among others, using machine learning models.

Study limitations include non-random group assignment, potentially influencing risk factor distribution. Our single-center design with small sample size may limit result generalizability, requiring cautious interpretation and future studies.

Higher recanalization and recovery rates were found in patients with acute intracranial ICA occlusion. ADAPT exceeded other strategies as first-line technique, achieving faster recanalization with comparable complications, and potentially impacting long-term functional outcome. Larger studies are needed to validate these results for intracranial ICA occlusions.