What is already known on this topic—Early and late thrombectomy time window in the anterior circulation influences rates of favorable functional outcomes.

Last known well (LKW) is a key time metric in stroke neurology. The phrase “time is brain” is a qualitative statement emphasizing the rapid loss of neurons in the human brain (1). Early randomized thrombectomy trials focused on LKW for inclusion criteria to establish the efficacy of thrombectomy in anterior circulation ischemic stroke (2). Furthermore, the DAWN and DEFUSE III trials extended this time window in patients with large penumbral volumes (3, 4). As such, the American Stroke Association/American Heart Association (ASA/AHA) guidelines established mechanical thrombectomy (MT) as class 1 and class 2a recommendations for patients presenting in the early and late time windows for anterior circulation large-vessel occlusions, respectively (5).

Basilar artery occlusion (BAO), however, was underrepresented in the aforementioned trials and remains a devastating form of ischemic stroke, accounting for up to 10% of large-vessel occlusion (LVO) (6). Recently, four randomized clinical trials evaluated the efficacy of thrombectomy compared to the best medical management in patients presenting with acute basilar artery occlusion (7–10). The BASICS trial enrolled patients within 6 h of the estimated time of BAO onset. This trial demonstrated higher rates of good and excellent clinical outcomes, basilar artery patency, and lower mortality in the endovascular group (7). The BEST trial evaluated patients presenting within 8 h of the estimated time of BAO and demonstrated that the endovascular group had a trend toward functional independence (8). The ATTENTION and BAOCHE trials established the superiority of thrombectomy compared to conservative management in the extended time windows of 0–12 h and 6–24 h of LKW, respectively (9, 10). Despite these high-level studies, however, the effect of time on the treatment effect of EVT compared to medical management has not been well studied. In this study, the outcomes of basilar artery thrombectomy as a function of LKW time to treatment were evaluated.

Several randomized controlled trials have recently proved mechanical thrombectomy benefits in BAO in the late time window (7–10). Our results demonstrate that there is a thrombectomy benefit regardless of the time window. Similar to the aforementioned clinical trials, we opted to use mRS of 0–3 to determine favorable outcomes given the high morbidity and mortality of BAO (9, 10). We determined that our early and late time window cohorts achieved similar favorable outcome rates (46% vs. 42%) and are comparable to the rate of outcomes in the recent RCTs (9, 10). This study reaffirms the findings of the BEST, BAOCHE, and ATTENTION trials, which have similar rates of favorable functional outcome and sICH despite patient presentation in different time windows. Within the early time window, however, our study still emphasizes the importance of timely thrombectomy. Saver et al. performed a meta-analysis of anterior circulation large-vessel occlusions and showed that for every 4-min delay in reperfusion the proportion of functional independence decreases by 1% (13). In our study, the rate of functional independence (mRS 0–2) at 3 months declined from 37.3% with an LKW-to-reperfusion time of 120 min to 20.9% with an LKW-to-reperfusion time of 480 min. Our results further stress this conclusion by demonstrating that for every 10-min groin-puncture delay leads to diminished favorable functional outcome and functional independence by 4% within the first 6 h from last seen normal. Many hypotheses for these minor differences exist including eloquence of brain tissue affected, level of macro/microvascular collateralization, density of cranial nerve motor neurons, and capacity of individual nerves to withstand ischemic insults (14–17).

Patients with intact AP collateralization are expected to react differently to an ischemic event compared to patients without AP collateralization (18, 19). Our study explored this hypothesis using ROC curves with respect to collateralization status and did not demonstrate a predictive value for our early vs. late time window cohorts (AUC: 0.550 and 0.495, respectively). This is most likely a reflection of patients within the cohort. “Fast” and “slow” progressors constitute our early time window population; however, as the “fast” progressors complete their stroke, they are effectively excluded as a thrombectomy candidate leaving the “slow” progressors in the late time window. This leaves an important unanswered question as to what differentiates “slow” vs. “fast” progressing posterior circulation strokes. It is hypothesized that outcomes could be a function of sub-occlusive/occlusive thrombi, clot location, microvascular collateralization, and LKW to recanalization (14, 20). Our subgroup analysis comparing clot location and early vs. late time window cohorts did not show appreciable differences in the shift toward improvement. Unfortunately, our registry did not have data points for sub-occlusive/occlusive thrombi or include measurement of cerebellar artery collaterals or CT perfusion studies to evaluate collateralization, and further studies may help to delineate these patient subtypes.

In terms of peri-procedural outcomes, our early vs. late time window cohorts did not influence an operator’s ability to achieve successful recanalization (83% vs. 79%) or recanalization at the first pass (39% vs. 36%). This was similar to the recent posterior circulation randomized controlled trials, which had 88–91% successful recanalization (albeit vessel patency was defined differently in these studies) (9, 10). Mokin et al. retrospectively evaluated 100 patients who received mechanical thrombectomy within 12 h and demonstrated that time metrics are associated with both improved successful recanalization rates and long-term (12 months) functional outcomes (21). The main difference in this study is that stroke symptom onset to femoral artery puncture was the main time metric, and the primary outcome was mRS 0–2. The differences in the aforementioned study can account for the variable results seen between BAOCHE and ATTENTION trials as well as in this study.

Despite the early time window presenting with a higher presenting NIHSS, the 24-h NIHSS (10 vs. 9) was similar between the early and late cohorts following thrombectomy. Compared to the ATTENTION trial, our dataset had a lower presenting NIHSS and 24-h NIHSS (9). As our dataset includes patients treated before the publication of the BAO endovascular trials, it is likely that patients presenting with a worse PC-ASPECT were not offered thrombectomy resulting in selection bias. However, it is apparent that thrombectomy results in an improvement in NIHSS despite the time window presentation.

Interestingly, our study demonstrated higher rates of in-hospital complications in the early time window group. Although not statistically significant, it was observed that these patients presented with a higher NIHSS. As the NIHSS has limitations in the posterior ischemic stroke, patients in the early time window presented with examinations affecting consciousness compared to those in the late time window which had a higher predominance of hemiparesis and sensory loss. This hypothesis is reflected in our data set in which early time window patients were more likely to require intubation for neurological failure (47.9% vs. 33.9%) and suffered from higher rates of infection (13.4% vs. 4.6%) compared to patients in the late time window.

Strengths of this study include its heterogeneous North American population, variety of stroke etiologies, and reasonably high rates of successful recanalization and recanalization at first pass. This study has several limitations. This study derived data which were obtained from a retrospective registry. Patients with low pc-ASPECTS were likely excluded from intervention, which may have resulted in a higher proportion of patients that are “slow progressors.” We did not have NIHSS inclusion criteria, which means that patients with low NIHSS would possibly have been excluded as well.

This study shows that in the North American population, select basilar artery thrombectomy patients in the late time window achieve similar rates of favorable outcomes compared to patients within the early time window. We also demonstrated that approximately every 10 min of delayed recanalization results in a 4% relative reduction in favorable functional outcome and functional independence within the first 8 h from last seen normal. Further studies are required to determine whether outcomes may vary in BAO patients with varying presentations with different pc-ASPECTS or NIHSS on presentation and also in the time window beyond 24 h.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by the University of Toledo Bioethics Committee. The studies were conducted in accordance with the local legislation and institutional requirements. The ethics committee/institutional review board waived the requirement of written informed consent for participation from the participants or the participants’ legal guardians/next of kin because this was an observational retrospective design.

AM: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Visualization, Writing – original draft, Writing – review & editing. RR: Writing – original draft, Writing – review & editing. SA: Writing – review & editing. AN: Writing – review & editing. SO: Writing – review & editing. JV-S: Writing – review & editing. MF: Writing – review & editing. AJ: Writing – review & editing. SD: Writing – review & editing. GT: Writing – review & editing. AA: Writing – review & editing. TN: Writing – review & editing. PK: Writing – review & editing. MA: Writing – review & editing. HS: Writing – review & editing. AP: Writing – review & editing. ZW: Writing – review & editing. SK: Writing – review & editing. NV: Writing – review & editing. NA: Writing – review & editing. KG: Writing – review & editing. EA: Writing – review & editing. SZ: Writing – review & editing. MJ: Writing – review & editing.