A systematic review and meta-analysis of five studies and more than 11,000 patient records has indicated that each hour saved between acute ischaemic stroke symptom onset and initiating a mechanical thrombectomy procedure is associated with a 22–25% increase in the odds of achieving functional independence.
The review and analysis in question—dubbed ‘SWIFT-EVT’—has been published in the Journal of Stroke and Cerebrovascular Diseases (JSCVD) by Brittney Legere (University of Guelph, Guelph, Canada), Ashfaq Shuaib (University of Alberta, Edmonton, Canada) and colleagues.
“While other systematic reviews have similarly analysed previous onset to EVT [endovascular therapy] time effects on patients, SWIFT-EVT provides a more recent and thorough analysis evaluating good and excellent functional independence,” Legere, Shuaib et al note. “Quantifying this substantial time-dependent benefit, relevant during the entire first six hours post-acute ischaemic stroke onset, may be useful to inform patient-specific and systems planning decisions.”
The researchers initially state that data through 2014 from the HERMES meta-analysis demonstrated that—particularly over the first six hours post-acute ischaemic stroke—the rapid administration of thrombectomy treatment “markedly improves outcomes”. However, while a subsequent 2016 meta-analysis and additional studies over the past few years have further elucidated how ‘time to thrombectomy’ affects outcomes, “no recent systematic reviews have quantified patient outcomes based on time intervals, and the time-saving benefit of EVT”.
Against this backdrop, SWIFT-EVT set out to calculate precise, patient-centred outcome benefits for incremental pre-thrombectomy time savings within six hours of stroke onset—a “broadly accepted” and guideline-based window for prioritising EVT timings in large vessel occlusion (LVO) stroke patients. The investigators’ aim was to provide “an updated metric summarising latest estimates for modified Rankin scale [mRS] improvements accrued by streamlining time to EVT”.
Legere, Shuaib et al conducted a systematic review and meta-analysis using electronic databases in which they considered eligible studies to be those reporting a time-benefit slope with times from stroke onset/time last known normal to thrombectomy commencement, whereby ‘onset-to-groin-puncture’ time was the predictor. They ultimately included five studies; four large, national/international registries—the 2017 STRATIS, 2018 MR CLEAN-R, 2018 Trevo Retriever and 2019 Get With The Guidelines (GWTG)-Stroke registries—as well as the 2020 Triveneto registry from a single Italian region. In combination, these studies comprised data from 11,343 patients.
The researchers’ primary endpoint was the rate of functional independence (mRS 0–2) at 90 days, while rates of 90-day ‘excellent function’ (mRS 0–1) and mortality, and level of occlusion, were among key secondary endpoints.
Their results ultimately revealed an increased chance of a good functional outcome, defined as mRS 0–2, with each hour of pre-EVT time savings—a trend that was present within an earlier time window of 0–270 minutes (odds ratio [OR], 1.25; 95% confidence interval [CI], 1.16–1.35; I2, 40%) and a later timeframe of 271–360 minutes (OR, 1.22; 95% CI, 1.12–1.33; I2, 58%). Legere, Shuaib et al report that, for each hour saved within the earlier timeframe, the odds of achieving functional independence improved by 25%, with these odds only diminishing slightly to 22% in the later window.
In addition, within those studies assessing excellent function, the authors go on to note that comparable trends were observed at timeframes of both 0–270 minutes (OR, 1.34; 95% CI, 1.19–1.51; I2, 27%) and 271–360 minutes (OR, 1.20; 95% CI, 1.03–1.38; I2, 60%), as the odds of achieving mRS 0–1 were found to increase by 34% and 20%, respectively, with each hour saved.
Legere, Shuaib et al also detail that—across their primary and secondary endpoint results on mRS 0–2 and mRS 0–1—”Cochrane’s Q, a reassuring Galbraith graph, and fixed-effects modelling, provided support for a finding that heterogeneity was not present to a degree precluding generation of a pooled effect estimate”.
“While not a primary focus of this study, some studies reported on the effect of time savings on mortality,” the authors add in a supplement accompanying their JSCVD paper. “However, high I2 for both timeframes precluded generation of a pooled effect estimate. Cochrane’s Q assessment for the mortality endpoint also rejected the null hypothesis of homogeneity (p<0.01 for both the 0–270-minute and the 271–360-minute timeframes). Unlike the case with either the primary or secondary endpoint, the mortality endpoint meta-analysis did not reject the null hypothesis of no effect of time savings on the endpoint (survival). This failure to identify an effect of time savings on mortality was noted for both the early (p=0.68) and late (p=0.71) timeframes. Given the heterogeneity of study findings, results for any meta-analysis pooled effect estimate are not presented.”
The investigators relay that three studies reported on the level of occlusion in relation to time savings, leading them to perform a sensitivity analysis to determine whether their overarching findings would be altered when including only these studies. On this point, they note that heterogeneity remained acceptable (27.45%) and the overall estimate was that each hour saved was associated with increased survival (OR, 1.35; 95% CI, 1.14–1.33).
“Quantifying the benefit of faster time to EVT is of critical importance for design and improvement of stroke systems of care,” Legere, Shuaib et al write, discussing their results in greater detail. “Many patients who experience stroke are not near EVT-capable hospitals—over a third of the US population has EMS [emergency medical services] transport time of greater than one hour to a comprehensive or EVT-capable stroke centre. This has contributed to disparities in access to thrombectomy, particularly for patients in lower-income and rural communities.
“Our systematic review and meta-analysis estimated that—within the initial six hours—for every 20 cases of 60-minute reduction in time from LVO acute ischaemic stroke symptom onset to EVT start time, there was one additional case of functional independence.
“Importantly, our findings were based on calculations using a denominator of all EVT cases, so SWIFT-EVT estimations can be broadly useful in application to triage and transfer decisions that are necessarily made without full knowledge (or foreknowledge) of variables such as advanced imaging results or reperfusion success. While not definitive, SWIFT-EVT’s calculations may be useful for system-wide organisation and planning of medical care, as well as for those making individual patient triage decisions.”
The authors conclude by positing that notable strengths of their review and analysis are its large number of patients, allowing for high levels of precision, as well as its focus on all patients undergoing EVT—rendering their data “potentially more useful for system-wide planning”. However, they also concede that their use of only mRS 0–2 and 0–1 to evaluate EVT benefit—despite providing “a comprehensive overview of patient independence and outcome post-EVT”—may be considered a limitation of SWIFT-EVT.
In addition to highlighting the fact that more research is required to understand the time-savings benefit of EVT beyond six hours from stroke symptom onset, Legere, Shuaib et al aver that additional study endpoints—ranging from mRS shift, discharge National Institutes of Health stroke scale (NIHSS) scores, and Montreal cognitive assessment (MoCA) scores, to the relevance of more severe stroke cases, symptomatic intracranial haemorrhage (sICH) and door-to-groin-puncture times—may also warrant future investigation.
“Further studies should evaluate alternative endpoints, such as those mentioned in the discussion, to ensure patient treatment optimisation and prompt patient selection for EVT,” they add.
