RATIONALE

Stroke represents the fourth leading cause of death in industrialized nations, after

  • heart disease,

  • cancer,

  • and chronic lower respiratory disease.

Each year approximately 795,000 people experience a new or recurrent ischemic stroke in the U.S/Europe. On average, every 40 seconds, someone in the United States has a stroke and approximately every four minutes someone dies of a stroke2,3. 

The overall burden of stroke will rise dramatically in the next 20 years due to an ageing population. The projections in the Burden of Stroke report indicate that between 2015 and 2035, overall there will be a 34% increase in total number of stroke events in the European Union from 613,148 in 2015 to 819,771 in 20354.

Proximal intracranial arterial occlusions cause the most disabling types of ischemic strokes and are predictive of poor neurological outcomes5. Stroke survivors constitute the majority of disable people worldwide. Approximately one-quarter of the patients suffering a stroke die within one year after the initial. Stroke brings a dramatic financial and personal burden to society. Direct medical costs related to stroke in the United States is an estimated $28.3 billion per year6. In the EU the total cost of stroke in 2015 was calculated as € 45 billion4.

Acute ischemic stroke (AIS) due to large vessel occlusion (LVO) is a potentially devastating event, with a poor prognosis in the absence of timely revascularization. 

Before 2015, IV t-PA administered within 4.5 hours after symptom onset was the only reperfusion therapy with proven efficacy in patients with AIS. Nevertheless, it is well-known that IV t-PA has many limitations, including a short therapeutic window, and a strong time-dependency9-12. In addition, the efficacy of IV t-PA is limited by the typically large thrombus burden that causes by proximal intra-cranial arterial occlusions13,14. Among patients with occlusions of the intracranial internal carotid artery or the first segment of the middle cerebral artery (or both), intravenous thrombolysis results in early reperfusion in only 13 to 50%7,8. 

Endovascular thrombectomy for AIS has evolved substantially since the beginning of the century. However only after the publication of several multicentre randomized controlled trials in 201515-19and 201620  (SWIFT-PRIME, REVASCAT, MR CLEAN, ESCAPE, EXTEND-IA and THRACE) carried out in comprehensive stroke centers located in different countries confirmed the superiority of MT combined with BMT    (which in the majority of cases included intravenous IV tPA) compared to BMT alone. Whether in patients with large anterior circulation stroke, direct mechanical intervention is equally effective, superior or inferior to bridging thrombolysis remains a matter of debate22

Thus, this procedure, is now the recommended standard of care for selective patients with LVO in the anterior circulation21. when performed with newergeneration devices (mainly stent retrievers), more stringent imaging selection criteria and more efficient workflow than in previous trials, significantly reduces disability rates after AIS caused by LVO in the anterior circulation15-19.

In 2015, a collaborative group was established by the investigators of the 5 first major trials aiming to pool patient-level data from these trials. Thus, the Highly Effective Reperfusion evaluated in Multiple Endovascular Stroke trials (HERMES) collaboration was formed, which allowed completion of a meta-analysis that ascertained the benefit of endovascular therapy(EVT) in a broader population compared to each individual trial23. In the HERMES meta-analysis, the number of patient needed to treat (NNT) with EVT to change a patient from a higher level of disability to a lower level of disability was 2.6, suggesting that EVT of large vessel stroke occlusion is one of the most efficacious procedural based therapy in medicine. 

The STAIR (Stroke Treatment Academic Industry Roundtable) meeting aims to advance acute stroke therapy development through collaboration between academia, industry, and regulatory institutions24. In pursuit of this goal and building on the available level I evidence of benefit from EVT in large vessel occlusion stroke, STAIR IX consensus recommendations were developed that outline priorities for future research in EVT. Three key directions for advancing the field were identified: 

(1) Development of systems of care for EVT in large vessel occlusion stroke

(2) Development of therapeutic approaches adjunctive to EVT

(3) Exploring clinical benefit of EVT in patient population insufficiently studied in recent trials and explore whether patient eligibility could be expanded 

The STAIR participants concluded that development of systems of care strategies should be geared, both toward ensuring broad access to EVT for eligible patients and toward shortening time to reperfusion to the minimum possible. Adjunctive therapy with potential for benefit and therefore identified as worth considering for development included neuro-protective approaches, adjuvant microcirculatory/collateral enhancing strategies, and peri-procedural management.

 

Future research priorities seeking to expand the eligible patient population identified at STAIR IX were to determine benefit of EVT in patients presenting beyond conventional time windows, in patients with large baseline ischemic core lesions, and in other important subgroups24.

 

Consistent with recommendations by STAIR IX, this trial will seek to answer the question of whether patients with AIS due to Large Vessel Occlusion who present with large baseline infarcts (defined as ASPECTS 0-5), who can be treated within 7 hours after stroke onset derive benefit from thrombectomy. This patient population was largely excluded from prior thrombectomy trials for stroke. 

The study has been designed with subject safety in mind, in order to address the concerns around potential unknown harms to enrolled subjects. If positive this study, more patients in the future could receive endovascular treatment (either in addition to or instead of IV t-PA).  

 

Large Ischemic Core Volume

 

The LASTE hypothesis is that patient harbouring a large ischemic core stroke at presentation, may still benefit from mechanical thrombectomy, given the very poor natural history of patient managed by a standalone medical therapy. 

The threshold of ASPECTS 5 to set a benefit for treatment was inspired by initial large medical trials using IV t-PA25-27(Samourai-CT ALBERTA Score IV cohort). In the paradigm of acute and complete reperfusion, radiographic infarct stroke volume may be partially reversible and thus the resulting final infarct may be smaller compared to the one in non-re-perfused patients who will consequently have poorer clinical outcomes. By now, large core patients defined as ASPECTS 0-5 were excluded from most randomized clinical trials resulting in a lack of unequivocal evidence of benefit in this patient population.

  

Analyses of several prospective cohorts suggest signals of benefit in favor of thrombectomy in patients with large baseline core. In the prospective cohort ETIS28, Lapergue et al reported a rate of good outcome of 34% in the subgroup of patient with ASPECT 4-5. In the prospective cohort analysis RECOST29, a mean mRS score of 34% in a group of patient presenting ASPECT 0-5 treated with mechanical thrombectomy, versus 9% of good outcome in the medical group, suggesting a benefit to treat LVO harbouring a large core.

The strongest argument in favor of benefit of EVT even in patients with large core is provided by the six completed randomized endovascular stroke trials. These trials did prove that Combined Approach Mechanical + IV t-PA, is superior to standalone IV lytics, mRs score ≤ 2 ranging from 33% and up to 72%, but Large Core Infarction were largely excluded.Nonetheless a pooled analysis of these trials revealed that a minority of patients with large baseline core were still randomized.In MR CLEAN17, despite no exclusion criteria defined in the initial protocol, the median ASPECT score of the cohort was 9 out of 10. In SWIFT-PRIME15, the mean ASPECT score of the cohort was 7, and in REVASCAT it was 6.816. 

 

In SWIFT-PRIME only ASPECT 6-10 or RAPID volumetric core evaluation < 50 ml and a proved mismatch profile, was required prior to randomisation. In EXTEND-IA, all patients underwent a systematic RAPID evaluation in order to exclude large core. In REVASCAT per the study protocol large core were excluded based on CT-ASPECT score given that patients with ASPECTS < 6 were excluded. Even though REVASCAT investigators score all baseline scans as ASPECTS > 5, core lab analysis revealed that 25% of enrolled patients had an ASPECTS score of 5 and below). In ESCAPE only 6% of the cohort was included with a CT-ASPECT below 6. In the HERMES meta-analysis out of 1764 patients 65 underwent mechanical thrombectomy with an ASPECT score < 5 on MRI and 61 patients who underwent MT had an ASPECTS score of < 5 on baseline CT. Both in the MRI and in the CT group there was a trend in favor of benefit with EVT while in the combined analysis of patients who had ASPECTS < 5 either by MRI or by CT (n=126) the benefit of EVT reached statistical significance (OR 2.15, 95% CI 1.06-4.37). In addition patients with evidence of hypodensity involving greater than 1/3 MCA territory (n=228) by MRI (n=96) or by CT (n=232), another measure of large baseline infarct, also derived benefit from EVT in HERMES (OR 1.70, 95%CI 1.04-2.78). Nonetheless, the AHA Guidelines therefore included in the level 1A of evidence, patient harbouring a LVO and a limited core volume defined as an ASPECT score >5. 

 

The intent of this study is to support the use of EVT below the currently limitation of Stroke Volume, who actually have no other option besides medical management of their symptoms. 

 

Time Window

Patient harbouring a large stroke may present a “fast progressor” profile that make essential the speed of in hospital workflow as previously described in REVASCAT or HERMES meta-analysis. It has been shown that in patient populations not selected based on mismatch, the benefit of thrombectomy declines in a time dependent fashion such that when the procedure starts beyond 7.3 hours from time last known well, the benefit of EVT can no longer be demonstrated with high degree of certainty. 

While recently the DAWN trial30-32demonstrated the efficacy of mechanical thrombectomy in the 6-24h time windows with a NNT of 2.8, which is likely to expand the consequently therapeutic time window for stroke due to proximal LVO, patients included in DAWN have been carefully selected based on the presence of clinical core mismatch and therefore the results of DAWN are unlikely to be applicable to the entire LVO stroke population presenting beyond 6 hours. Therefore a timewindow of 7 hours in a study aiming to establish the benefit of EVT in patients with large infarcts seems to be the most reasonable33.

Moreover DAWN was restricted to low progressor profile, patient with a small core that does not correspond to the patient profile we have in LASTE, more likely to be fast progressor profile patient. The time window is therefore restricted to 7 hours between symptoms and randomization, in order to achieve complete recanalization within the limit of 8 hours34. For the same reason, the time between randomization and groin puncture will be reduced to 30 minutes in the LASTE protocol. 

Imaging selection criteria and Age

Diffusion-weighted MRI (DWI) is useful for patient selection during the first 6 hours after stroke onset. The main aim of one Montpellier's Stroke Unit study was to investigate the relationship between the time from stroke symptom onset and stroke volume assessed using DWI.

 

Patients were classified according to the number of hours after the onset of symptoms that DWI was performed. No relationship was found between temporal groups and the DWI-ASPECTS. The number of patients who reached each of the 3 classified stroke volumes was not different between the temporal groups then, the study concluded that there is no correlation in the author’s findings between the time of stroke symptom onset and the DWI-ASPECT score during the first 6 hours from stroke onset35.

 

In the dynamic setting of ischemia, there is continuous growth of the infarct core at the expense of the penumbral tissue until either the infarct is completed or reperfusion is achieved. The pace of expanding cerebral ischemia is highly variable between individuals and is likely dependent on multiple factors, including the presence of collateral circulation, ischemic pre-conditioning, cerebral perfusion pressure, and cerebral blood volume as well as serum glucose, body temperature, and oxygen delivery capacity. 

 

Several studies have been reported in the literature demonstrating the general safety and effectiveness of using the ratio of “core infarct” to “salvageable penumbra” concept to select patients for reperfusion therapies. The methods of measuring and/or defining “core infarct” and “salvageable penumbra” however vary from study to study36-44.  However, in view of the most recent findings from HERMES, suggesting that benefit from EVT exists even in patients with large baseline core, the concept of imaging guided selection of patients in the early (0-6 hours’ time window) may need to be reappraised. It is increasingly evident that the more sophisticated imaging paradigm used to select patients the most of the time delays necessarily the completing imaging and multiple studies have by now established that time delays have deleterious effects on patient outcomes. Given that the HERMES data suggests a strong time dependent benefit of EVT when initiated within 0-7 hours from time last seen well in patients encompassing the entire spectrum of baseline infarct. Additional time consuming imaging studies seeking to establish a large baseline infarct beyond a CT or MRI does may introduce more harm than benefit through the inherent delays associated with imaging and is not recommended. 

 

While still controversial, some data show that infarct core volume is a better predictor of outcomes than perfusion based imaging selection45-47. DEFUSE 2 pre-procedure infarct volume along with age were the only independent predictors of outcome and core infarct volumes of less than or equal to 15 cc is the best discriminator of good versus bad outcome. Perfusion MR in addition to DWI did not add a higher predictive value to this model48. In another retrospective analysis of 201 endovascularly treated patients, age and final infarct volume were found to be independent predictors of outcome49.

The larger the mismatch between size of core and size of “at risk” territory  the greater the treatment effect is likely to be with reperfusion therapy, and the more substantial the infarct growth is likely to be without reperfusion therapy. No mismatch signals that the subject is not going to grow their infarct and thus uncertain with regards to benefit from reperfusion39,43.

Stricter inclusion criteria may be necessary for subjects greater than 80 years of age in order to demonstrate treatment effect. In one study of IV t-PA treated patients, the overall rate of symptomatic intracranial haemorrhage (SICH) in the octogenarians was 6.9%, compared with 5.3% in younger patients. The use of MRI to select octogenarians for thrombolytic therapy seemed to decrease the risk of SICH, but did not influence the overall outcome after 3 months50. In another published study comparing outcomes in IV t-PA treated and non-treated subjects ≥ 80 and < 80 years old, although age was associated with poorer outcomes the association between thrombolysis treatment and improved outcomes was maintained in the very elderly subjects, and their conclusion is that age alone should not be considered a barrier to treatment51.Recently, Danière et al52reported a comparative analysis of patient treated with a large vessel occlusion among class of age in patient ASPECT 6-10 underlining a clear cut off in term of mortality and outcome for patient beyond 70 years old, with a mortality of 26% (Seventies) and 35% (Eighties) at 3 months versus 9% for younger patient. In the meantime, the good outcome rate was ranging from 28% to 19% for patients >70 versus 72% to 58% in the younger patients. In this cohort, patient included after 80 years old with an ASPECT score of 4 presented a 12% rate of good outcome. However, in order to mitigate the potential risks associated with endovascular treatment in the elderly as well as to maximize the chance of a good outcome, the core size in subjects who are ≥ 80 years will be restricted to ASPECTS > 3.

In the absence of a “gold standard” to define salvageable penumbra, LASTE subject selection is based on a Clinical Imaging Mismatch using baselineASPECTS on CT and MRI to estimate baseline infarct volumes, across all study sites.

A Clinical Imaging Mismatch is the observed difference between the size of the core infarct and the magnitude of the neurological deficit, in the presence of confirmed LVO by CTA/MRA, and has been validated in DAWN as a reliable and efficient surrogate for assessing salvageable brain tissue. Using this method we aim to select subjects at risk of further infarct growth without rapid reperfusion.

The HERMES collaborative group aimed to investigate baseline-imaging factors on CT/CTA and MRI/MRA and their relationship to outcomes after endovascular treatment. In this pre-specified secondary study of the HERMES (Highly Effective Reperfusion evaluated in Multiple Endovascular Stroke Trials) Collaboration, baseline imaging studies (except for CT perfusion and MR perfusion) from patients from MR CLEAN, ESCAPE, REVASCAT, SWIFT PRIME, EXTEND IA, PISTE and THRACE trials were analyzed to determine imaging predictors as effect modifiers treatment in acute large vessel stroke. The conclusion was that treatment effect was favorable for thrombectomy in all the neuroimaging baseline categories across all baseline infarct sizes, extent of collaterals, presence of hyper-dense MCA and thrombus size. Although overall, no imaging feature predictive of harm was found, an increased risk of symptomatic intracranial haemorrhage (but not PH2 or mortality) was noted in patients with ASPECTS 0-4 or > 1/3 MCA territory involvement. Although overall the  effect of EVT in patients with large core was in favor of benefit, the higher rates of symptomatic haemorrhage in the EVT group justifies further studies aiming to definitively clarify safety and benefit of endovascular treatment in these patients. 

MRI-DWI is the preferred method to measure the core infarct volume using the ASPECT score, due to logistic barriers such as unavailability of MRI equipment or technicians and potential subject contra-indication for MR, sites are permitted to use CT to measure the ASPECT score. 

Since it is recognized that any image of the brain is a “snapshot in time”, LASTE requires that the corresponding clinical “mismatch” be evaluated using the baseline NIHSS obtained within 1 hour, of the processed neuro-images used to qualify the subject for the study.