Article Text
Abstract
Introduction Acute basilar artery occlusion (BAO) can result in extremely high disability and mortality. Stent retrievers (SRs) can achieve a high recanalization rate for BAO, therefore improving favorable outcomes. However, the efficacy of a direct aspiration first pass technique (ADAPT) to treat BAO is unclear. Our aim was to compare the efficacy and safety of firstline ADAPT with that of firstline SR for patients with acute BAO.
Methods Three databases were systematically searched for literature reporting outcomes on thrombectomy for acute BAO with both firstline ADAPT and firstline SR. The modified Newcastle–Ottawa scale was applied to assess bias risk. The random effects model was used.
Results Of 50 articles, 5 cohort studies (2 prospective and 3 retrospective) were included in our research. 193 cases were treated with firstline ADAPT and 283 cases received firstline SR. Successful recanalization rate was significantly higher in the firstline ADAPT group (OR=2.0, 95% CI 1.1 to 3.5). Procedure time (mean difference=−27.6 min, 95% CI −51.0 to −4.3) and the incidence of new territory embolic event (OR=0.2, 95% CI 0.05 to 0.83) was significantly less in the firstline ADAPT group. No significant difference was observed between the firstline ADAPT and firstline SR groups for rate of complete recanalization, rescue therapy, any hemorrhagic complication, favorable outcomes, or mortality at 90 days.
Conclusions Our meta-analysis suggested that for patients with acute BAO, firstline ADAPT might achieve higher and faster recanalization, comparable neurological improvement and safety compared with firstline SR. Further studies are needed to confirm these results.
- thrombectomy
- stroke
Statistics from Altmetric.com
Introduction
Acute basilar artery occlusion (BAO) can be catastrophic for patients, with disability and mortality in 70–80% of cases.1–3 Fortunately, early recanalization therapies can significantly improve the prognosis for patients with BAO, regardless of which approach is used, including intravenous thrombolysis, intra-arterial thrombolysis, and mechanical thrombectomy (MT).4 5 MT has been validated in several randomized controlled trials for patients with acute anterior circulation strokes.6 Additionally, two meta-analyses have shown that stent retrievers (SRs) could also achieve a high recanalization rate (>80%) and functional independence (>40%) in patients with BAO.7 8
Proposed by Turk et al,9 a direct aspiration first pass technique (ADAPT) with a large bore aspiration catheter has advanced rapidly and showed effectiveness and safety for recanalization for acute ischemic stroke (AIS),10 which may be better than SR.11 12 Recent randomized controlled trials (the Contact Aspiration vs Stent Retriever for Successful Revascularization (ASTER) trial13 and the Compass trial (conference presentation and not yet published)) were conducted to compare firstline ADAPT with firstline SR for patients with AIS. They demonstrated that ADAPT as a firstline strategy was similar to SR with regard to revascularization and functional independence, which was in line with another recently published meta-analysis.14 However, all of these studies focused on anterior circulation strokes, and the role of ADAPT for acute BAO is still unclear. In this study, our aim was to compare the efficacy and safety of firstline ADAPT with that of firstline SR for patients with acute BAO.
Methods
Protocol and guidance
This study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.15
Search strategy and eligibility criteria
Literature published before November 1, 2018, was systematically searched on PubMed, Embase, and Cochrane databases with no language restriction. The keywords used for the population were ‘basilar artery occlusion’ or ‘posterior circulation stroke’, while ‘aspiration*’ and ‘stent*’ were used as the keywords for interventions.
Two investigators (GY, JL) independently reviewed each retrieved literature. Disagreements were resolved by consensus or with the help of the senior investigator (DW). Studies were regarded as eligible if they reported both SR and ADAPT for patients with BAO. Sample size in each cohort was >5. For those studies with duplicate cases, only complete reports were included. Conference abstracts, case reports, and reviews were excluded.
Data extraction
Two investigators (GY, JL) independently extracted baseline data and main outcomes from each eligible study: demographic profiles (total number of patients, age, and sex), lesion location, stroke etiology, baseline National Institutes of Health Stroke Scale (NIHSS) score, baseline Alberta Stroke Program Early CT score (ASPECTS), number of patients receiving intravenous tissue plasminogen activator (IV tPA), time from onset to groin puncture, specific types and number of devices used, procedure time, successful recanalization (defined as Thrombolysis in Cerebral Ischemia (TICI) 2b or 3), complete recanalization (TICI 3), number of rescue therapies used, favorable outcome/functional independence defined as a modified Rankin Scale score (mRS) of 0–2 at 90 days, new territory embolic events, any hemorrhagic complications, and mortality at 90 days. Disagreements were resolved by consensus or by the senior investigator (DW).
Risk of bias assessment
To evaluate the risk of bias, two investigators (GY, JL) independently assessed each article according to the modified Newcastle–Ottawa scale, which is designed to assess the quality of non-randomized studies included in a systematic review and meta-analysis.16 Each study was assessed based on three aspects: selection of cohorts (four items); comparability of cohorts (one item); and assessment of outcome (three items). Factors giving a low risk of bias were as follows: (1) well defined selection criteria; (2) comparable baseline characteristic, such as NIHSS and lesion location; (3) and independent assessment of neurologic and angiographic outcomes.
Statistical analysis
The main outcomes were as follows: successful recanalization (TICI 2b/3), complete recanalization (TICI 3), favorable outcomes/functional independence (mRS 0–2 at 90 days), procedure time, rescue therapy, any hemorrhagic complication, new territory embolic event, and mortality at 90 days. OR values were calculated for categorical variables and the mean difference (MD) was calculated for continuous variables (procedure time). A random effects model was used to pool the main outcomes.17
Publication bias was evaluated qualitatively via the asymmetry of a funnel plot or quantitatively by the Begg test, defined as significant when the P value was <0.1. Heterogeneity between studies was assessed by calculating the Q statistic and I2. A P value of the Q statistic <0.05 was considered to be statistically significant. The value of I2 was used to estimate the magnitude of heterogeneity; I2 >50% indicated moderate to high heterogeneity.18 Sensitivity and subgroup analyses were used to analyze the sources of heterogeneity. We attempted to conduct subgroup analyses according to ethnicity, using Caucasians and Asians.
All statistical analyses were conducted with Review Manager (V.5.3) and Stata (V.12.0).
Results
Characteristics of included studies
Of 50 articles, 5 cohort studies19–23 (2 prospective and 3 retrospective) were included in our research (figure 1). All studies had a low risk of bias according to the modified Newcastle–Ottawa scale (see online supplementary data). Baseline characteristics are shown in table 1, Forest plots are shown in figure 2, and the pooled results of the main outcomes are shown in table 2.
Supplemental material
All studies were published between 2014 and 2018, and a total of 476 patients (291 men and 185 women) with acute BAO were included. Specifically, 193 cases (40.5%) were treated with firstline ADAPT and 283 cases (59.5%) received firstline SR. Mean patient age ranged from 63.0 to 71.0 years, and mean time from onset to groin puncture was 283.5–562.0 min. Mean initial NIHSS score ranged from 17.0 to 23.8, with a mean ASPECTS score (two studies) of 7. The proportion of patients receiving IV tPA (four studies) ranged from 30.7% to 45.2%. Successful recanalization was 77.8–93.5%, while complete recanalization (four studies) was 42.0–63.6%. The proportion of rescue therapies (three studies) ranged from 14.0% to 22.2%, and mean procedure duration was 40.0–78.9 min. Additionally, favorable outcome at 90 days (four studies) ranged from 35.0% to 44.8%. The incidence of new territory embolic event (two studies) was 11.0% and 15.2%, respectively, while the incidence of any hemorrhagic complications (three studies) ranged from 15.7% to 42.4%. Finally, mortality during hospitalization (three studies) ranged from 22.6% to 30%, while 90 day mortality (three studies) ranged from 16% to 44.2%.
Success recanalization (TICI 2b–3)
Five studies reported successful recanalization, and the corresponding Forest plot is shown in figure 2A. Although all of the studies revealed a tendency to be more favorable for the firstline ADAPT group, there was no significant difference between the two techniques. However, the pooled results suggested that firstline ADAPT achieved a significantly higher successful recanalization rate than firstline SR in patients with BAO (OR=2.0, 95% CI 1.1 to 3.5; P=0.02). No significant heterogeneity (P=0.82, I2=0%) or publication bias (P=0.22) was detected between studies (see online supplementary data).
Complete recanalization (TICI 3)
Four studies reported complete recanalization and the corresponding Forest plot is shown in figure 2B. These studies tended to favor firstline ADAPT (two significant and one non-significant); the remaining study did not favor ADAPT. The pooled results showed no significant difference between the two strategies, but firstline ADAPT revealed a tendency to be more favorable (OR=2.38, 95% CI 0.92 to 6.13; P=0.07). Remarkably, higher heterogeneity was detected between studies (P=0.02, I2=71%). There was no significant publication bias observed according to the Begg test (P=0.31). Sensitivity analysis indicated that heterogeneity might come from the article published by Kang et al, while this study was of relatively high quality. Subgroup analysis was conducted by ethnicity (Caucasians and Asians) and the pooled results within the Asian subgroup also showed severe heterogeneity (see online supplementary data).
Favorable outcome (mRS 0–2) at 90 days
Functional independence at 90 days was reported in four studies and showed a relatively similar effect between the two strategies. The remaining study reported that firstline ADAPT had more favorable outcomes (mRS 0–3) at discharge. Our pooled results of mRS 0–2 at 90 days are shown in figure 2C. There was no significant difference between the two techniques in each study, and the pooled results also did not differ between firstline ADAPT and firstline SR (OR=0.93, 95% CI 0.62 to 1.38; P=0.71). No significant heterogeneity (P=0.75, I2=0%) or publication bias (P=0.73) was detected between studies (see online supplementary data).
Procedure duration
Procedure duration was a continuous variable, while all of the other outcomes were categorical variables. One study reported total procedure time (median 40 min (IQR 25 to 70 min)) without grouping. Another study showed that procedure duration was significantly lower in the ADAPT group (median 45 min (IQR 34 to 62 min) vs 56 min (40 to 90 min); P=0.05). The other three studies reported a procedure time of MD±SD so that it could be pooled, and the corresponding Forest plot is shown in figure 2D.
All three studies revealed that firstline ADAPT tended to have faster reperfusion than SR (two significant and one non-significant). Pooled results showed that procedure duration was significantly lower in the firstline ADAPT group than in the SR group (MD=−27.6, 95% CI −51.0 to −4.3; P=0.02). However, moderate heterogeneity was detected between studies (P=0.05, I2=66%). There was no significant publication bias observed according to the Begg test (P=0.30). Sensitivity analysis suggested that heterogeneity might originate from the study by Mokin et al, while this study had relatively moderate quality according to the modified Newcastle–Ottawa scale. Subgroup analysis was not conducted due to only three studies being included (see online supplementary data).
Rescue therapy
Three studies reported the proportion of rescue therapies and the Forest plot is shown in figure 2E. The study from Gory et al suggested that more patients in the firstline ADAPT group received rescue therapy (P=0.001). In contrast, Mokin et al found that although there was no significant difference, the proportion receiving rescue therapy in the firstline SR group was higher (P=0.44). Another study indicated that rescue therapy among both two groups was equivalent. Finally, pooled results showed no significant difference between the two techniques (OR=1.51, 95% CI 0.42 to 5.45; P=0.53). However, high heterogeneity was detected between studies (P=0.01, I2=77%). No significant publication bias was observed (P=1.0). Sensitivity analysis did not suggest the source of heterogeneity and subgroup analysis was not conducted because only three studies were included (see online supplementary data).
Any hemorrhagic complications
Three studies reported the incidence of any hemorrhagic complications and the Forest plot is shown in figure 2F. There was no significant difference between the two strategies observed in these studies. Two studies with relatively large sample sizes tended to show lower hemorrhagic events in the firstline ADAPT group. Our pooled results indicated less hemorrhagic complications in the firstline ADAPT group but this was not statistically significant (OR=0.74, 95% CI 0.31 to 1.73; P=0.48). Mild yet acceptable heterogeneity was detected (P=0.16, I2=46%) with no publication bias between studies (P=0.30) (see online supplementary data).
New territory embolic events
Kang et al and Son et al reported a total incidence of new territory embolization of 6/212 (2.9%) and 0/31 (0%), respectively. Only two other studies had data available for the meta-analysis and the Forest plot is shown in figure 2G. Both studies revealed that firstline ADAPT tended to have less new territory embolization than SR (one significant and one non-significant). The pooled results also suggested that firstline ADAPT could significantly reduce new territory embolic events compared with firstline SR (OR=0.20, 95% CI 0.05 to 0.83; P=0.03). No significant heterogeneity (P=0.35, I2=0%) or publication bias (P=1.0) was detected between studies (see online supplementary data).
Mortality at 90 days
All five studies reported mortality. Three studies reported mortality at 90 days and three reported mortality during hospitalization (Son et al reported both). Mortality was equivalent between the two techniques and no significant difference was observed in each study. The Forest plot for mortality at 90 days is shown in figure 2H and the pooled results showed no significant difference between firstline ADAPT and firstline SR (OR=1.06, 95% CI 0.63 to 1.79; P=0.83). There was no significant heterogeneity (P=0.85, I2=0%) or publication bias (P=1.0) detected between studies (see online supplementary data).
Discussion
This meta-analysis (five cohort studies with 476 cases) compared firstline ADAPT with firstline SR in the treatment of acute BAO and showed that the former strategy could achieve a more successful recanalization rate (OR=2.0, 95% CI 1.1 to 3.5; P=0.02), less procedure time (MD=−27.6, 95% CI −51.0 to −4.3; P=0.02), and less new territory embolic events (OR=0.20, 95% CI 0.05 to 0.83; P=0.03). However, favorable outcomes (OR=0.93, 95% CI 0.62 to 1.38; P=0.71) and mortality (OR=1.06, 95% CI 0.63 to 1.79; P=0.83) in these patients did not show any significant improvement. In addition, there were no significant differences between these two strategies in terms of complete recanalization (OR=2.38, 95% CI 0.92 to 6.13; P=0.07), rescue therapies (OR=1.51, 95% CI 0.42 to 5.45; P=0.53), and any hemorrhagic complications (OR=0.74, 95% CI 0.31 to 1.73; P=0.48).
Our pooled results suggested that the firstline ADAPT technique might achieve higher and faster recanalization for patients with acute BAO compared with SR. ADAPT is technically easier than SR; the aspiration catheter needs to be placed up to the proximal end of the thrombus without the micro guidewire and catheter passing the clot, as is the case for SR, which might optimize the final step by streamlining the process through rapid reperfusion, hence leading to better outcomes.24 Although it had a tendency to need more rescue therapy, firstline ADAPT was also simpler for preparing another pass or for switching to other rescue therapies. In addition, our study showed that the firstline ADAPT strategy might be safer than SR with a lower risk of new territory embolic events and comparable hemorrhagic complications. This might also be attributable to the simple operative procedure with no need to cross the thrombus, because the increased use of a micro guidewire and catheter can lead to a greater risk of thrombus fragment and damage to the vessel wall. Also, a recent animal study suggested that SR seemed to be more harmful to all vessel layers compared with ADAPT, especially in the endothelium.25 Although firstline ADAPT appeared to be more effective and safer in terms of revascularization for acute BAO, the pooled rate of functional independence and mortality at 90 days in the firstline ADAPT group was not greater compared with the SR group, which might be due to the limited sample size.
However, the results of our meta-analysis seemed to be inconsistent with those in anterior circulation strokes. Two randomized controlled trials (ASTER and Compass trials) and the Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in The Netherlands (MR CLEAN) registry have compared the efficacy and safety of ADAPT and SR in anterior circulation strokes. All three studies drew a similar conclusion: although significantly less procedure time was needed for ADAPT, comparable clinical outcomes were observed between the two strategies in terms of recanalization, neurological improvement, and safety.13 26 Interestingly, relatively more patients achieved an mRS score of 0–1 in the SR group (78/182) than in the ADAPT group (59/181) of the ASTER trial, although there was no statistically significant difference. Notably, more large bore balloon guide catheters (BGCs) were used in the SR group than in the ADAPT group in all of these studies. Moreover, a BGC was even routinely used in the SR group of the ASTER trial. A recent meta-analysis has shown that MT with BGCs could achieve higher firstpass, successful and complete recanalization, as well as lower mortality and procedure time than the non-BGC group.27 This finding suggested that the performance of the SR group in these studies might be partially attributed to the role of BGCs. Hence the results of the ASTER trial showed that the firstline SR+BGC group could achieve comparable recanalization and relatively better prognosis than the firstline ADAPT group. However, posterior circulation stroke is another condition where BGCs might not be available. As the basilar artery is supplied by bilateral vertebral arteries, the unilateral position of BGCs cannot effectively arrest proximal flow. Under these conditions, all studies included in this meta-analysis adopted SR alone without BGCs, which might explain to some extent the better performance of firstline ADAPT in posterior circulation strokes .
Limitations
This study has some limitations that should be noted. There was no randomized controlled trial on this topic and only five cohort studies with 476 cases were included. Although the risk of bias was low in each study, selection bias cannot be avoided due to the study designs. The inclusion criteria between studies differed, so some pooled results showed high heterogeneity.
In addition, some important baseline characteristics were missing from each cohort, such as etiology, lesion location, specific device used, baseline NIHSS and ASPECTS, etc. These factors might have confounded the clinical outcome of MT. Firstly, stroke etiology might determine the nature of the thrombus and thus affect the difficulty of thrombectomy,28 as cardiogenic clots seemed to be more reluctant to MT.28 29 Meanwhile, lesion location might reflect the source of the thrombus to some extent and the racial proportion in each cohort was also not reported, which might influence the distribution of the etiologies. Additionally, the specific bore of the device was of great importance, as a catheter with a larger inner diameter might achieve higher aspiration force and better outcomes.30 Initial NIHSS and ASPECTS scores could also indicate the degree of cerebral ischemia or infarction and thus impact on prognosis. As a consequence, further subgroup analysis or meta-regression was difficult due to the missing information. Moreover, the main outcomes assessed at each center might have differed. For instance, Gerber et al 18 used the arterial occlusive lesion scale to assess recanalization instead of TICI, and only reported functional independence at discharge (not at 90 days).
In summary, studies on this topic are still lacking and further studies with better design and larger sample sizes are needed to provide more evidence.
Conclusion
In conclusion, our meta-analysis suggested that for patients with acute BAO, firstline ADAPT might achieve higher and faster recanalization while being relatively safe compared with firstline SR. However, these effects did not translate into improvements in prognosis for these patients, but such an effect might have been limited by the sample size. Further high quality studies with large sample sizes are needed.
References
Footnotes
Contributors GY designed the study, performed the literature search and selection, extracted and analyzed the data, and wrote and revised the manuscript. JL performed the literature search and selection, extracted the data, and revised the manuscript. PQ performed the literature search and selection, and revised the manuscript. XY and LW analyzed the data and revised the manuscript. DW designed the study, monitored the study inclusion and data extraction, and revised the manuscript. He is the guarantor.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial, or not-for-profit sectors.
Competing interests None declared.
Patient consent Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement All data are available in the manuscript.