MR CLEAN and the changing landscape of intra-arterial stroke therapy


David Sacks analyses the results of the MR CLEAN study and finds that with three other randomised intra-arterial stroke trials (ESCAPE, SWIFT-PRIME, and EXTEND-IA) presenting positive results for the intra-arterial arm, the treatment landscape is poised for change.

In February 2013, intra-arterial acute ischaemic stroke revascularisation was discredited by the simultaneous publication in the New England Journal of Medicine of three randomised trials (IMS III, MR RESCUE, and SYNTHESIS) all of which concluded that there was no clinical benefit compared to best medical therapy. These trials were criticised for including patients who lacked a large artery intracranial occlusion (who could not possibly benefit from intra-arterial therapy), having prolonged times to treatment and poor revascularisation rates, and using obsolete first generation technology. Despite the criticisms, these studies led to widespread scepticism of intra-arterial therapy and several health insurance and technology assessment panels determined mechanical thrombectomy to be investigational and ineligible for reimbursement.

The recently published Dutch MR CLEAN trial used this scepticism to scientific advantage and linked reimbursement to trial participation, facilitating rapid study of 500 patients randomised to best medical therapy (89% received intravenous tissue plasminogen [IVtPA]) with or without intra-arterial treatment. Unlike the three previous trials, MR CLEAN found a very significant benefit from intra-arterial treatment. The absolute improvement in independent function increased from 19% to 33% with intra-arterial treatment. While there was no increased risk of mortality or symptomatic intracranial haemorrhage, the benefits were not risk free. In the intra-arterial group, 9% of patients suffered emboli into previously unaffected territory, 1.7% had a vessel dissection, and 0.9% had a vessel perforation. 

Why did MR CLEAN have better outcomes than the three previous recent trials? Better patient selection and better revascularisation. MR CLEAN selected only those patients with large vessel occlusions. IMS-III was started before the common emergency availability of CTA or MRA and included some patients who did not have large vessel occlusions. A subgroup analysis of IMS-III patients with large vessel occlusions did demonstrate a significant (p=0.01) shift to good outcomes in the intra-arterial group, confirming the importance of patient selection. Further, MR CLEAN successfully revascularised at least 59% of patients (TICI 2b or 3) compared to about 40% in IMS-III and 27% in MR RESCUE. 

MR CLEAN included some patients not usually treated—patients with mild stroke symptoms (NIHSS>2), the very aged, and patients with large completed strokes. A National Institute of Health Stroke Score (NIHSS) of 10 was used in IMS-III, since this increases the likelihood of the patient having a large vessel occlusion. However, a large clot can be present with mild symptoms if collateral flow is robust and mild symptoms may worsen in the presence of a large clot. A recent retrospective study (Mokin et al, Journal of Neurointerventional Surgery) found that 40% of patients with an NIHSS of 2 and a large vessel clot who did not receive revascularisation either died or were discharged to hospice or a facility and over 45% were unable to ambulate at discharge. MR CLEAN found that the aged also benefited significantly from intra-arterial treatment, but doubling the likelihood of a very small chance of a good outcome is still a low likelihood. For patients with large established strokes, there was a marginally significant benefit.

Are the results from MR CLEAN likely to be reproducible? Yes. Three other randomised intra-arterial stroke trials (ESCAPE, SWIFT PRIME, and EXTEND-IA) were presented at the International Stroke Conference in February 2015 and confirmed the benefit of intra-arterial revascularisation. Their results provide more information on the benefit of intra-arterial treatment for subgroups of patients such as the elderly, patients with time of symptom onset >6 hours, and the role of perfusion imaging to help exclude patients who have irreversible deficits. 

Can the results of MR CLEAN be improved? Most likely, yes. Devices might be improved to have better revascularisation rates with a lower risk of emboli. However, the largest opportunity for improvement is speed. MR CLEAN had a median time of 175 minutes from start of IVtPA to groin puncture. The IMS studies have demonstrated that there is roughly a 10% drop in good outcomes for every 30-minute delay in revascularisation. The international multisociety intra-arterial stroke quality measures have a target of 120 minutes from door to arterial puncture and the ESCAPE trial requires puncture within 60 minutes of starting the non-contrast head CT. With faster times to treatment, ESCAPE, SWIFT PRIME, and EXTEND-IA all demonstrated better outcomes than MR CLEAN.

Can the results of MR CLEAN be generalised outside of the trial environment? Probably, but not certainly. Stroke facilities and individual stroke interventionists must track their data (, both to improve processes of care and confirm acceptable outcomes as defined in the published quality guidelines. Patients should be directed to those facilities that have acceptable documented outcomes. In the USA facilities can become accredited as stroke centres, but the accreditation does not yet include comparison of clinical outcomes to accepted benchmarks.

In summary, MR CLEAN, ESCAPE, SWIFT PRIME, and EXTEND-IA have confirmed that intra-arterial stroke therapy is beneficial. Optimal results will depend not only on device effectiveness, but also on patient selection, speed of response, and operator skill. Expeditious stroke processes of care and documentation of outcomes are essential.

David Sacks is at the Interventional Radiology section of The Reading Hospital and Medical Center in West Reading, USA. He is a past president of the Society of Interventional Radiology (SIR) and the Intersocietal Commission for the Accreditation of Carotid Stent Facilities (ICACSF). He is a board director of INSTOR, the stroke registry mentioned in the article