Article Text

Download PDFPDF
Future acute ischemic stroke trials should randomize on the angio table
  1. Mahesh V Jayaraman1,
  2. Ryan A McTaggart2,3
  1. 1Departments of Diagnostic Imaging and Neurosurgery, Warren Alpert School of Medicine at Brown University, Rhode Island Hospital, Providence, Rhode Island, USA
  2. 2Departments of Radiology and Neurosurgery, Cleveland Clinic Florida, Weston, Florida, USA
  3. 3Cerebrovascular Center, Cleveland Clinic, Cleveland, OH, USA
  1. Correspondence to Mahesh V Jayaraman, Warren Alpert School of Medical at Brown University, 593 Eddy St, Room 377, Providence, RI 2903, USA; MJayaraman{at}Lifespan.org

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Perhaps in order to move forward we need to take a lesson from the past. The only two positive randomized trials of intra-arterial therapy to date are PROACT II and MELT.1 ,2 In both of those cases, randomization occurred on the table, after angiographic confirmation of proximal vessel occlusion. Of patients in PROACT II randomized to a saline infusion, 25% had a good functional outcome (modified Rankin Scale 0–2) at 90 days, compared with 40% of those who received a urokinase infusion. Similarly, in MELT, 23% randomized to saline infusion had modified Rankin Scale 0–1 at 90 days compared with 42% of those with urokinase infusion. Both these differences were statistically significant, with a total of just 294 patients randomized in both trials

We have better tools now. Our Thrombolysis in Cerebral Infarction (TICI) 2b/3 recanalization rates with modern stentriever and aspiration technology approach 90%,3–5 and we can perform these procedures quickly. So although our tools have improved, we are still unable to replicate the improvement in a randomized trial of simple lytic infusion.

Additional imaging has been added to our armamentarium. Has it helped us? PROACT II and MELT used non-contrast CT alone, excluding only those with large early infarcts. Many centers now routinely use ‘advanced imaging’, including CT angiography, CT perfusion, and MRI with diffusion weighted imaging. It is the hope that these additional studies will enable better selection of patients. Yet in the late time window, imaging selection was not helpful.6 The extra imaging also adds time, potentially mitigating any benefit of additional imaging,7 and CT perfusion may not change clinical decision-making above and beyond CT angiography.8 In addition, if we wait long enough in the time window, we may be selecting patients with good collaterals (small infarct cores) late in the time window who may not progress to infarction.

Clearly we are at a crossroads. We believe that thrombectomy does improve outcomes in properly selected patients. However, we have been unable to prove that in a randomized trial. Perhaps the answer is to randomize on the table, as PROACT II and MELT did. This would eliminate the dilution of the endovascular signal created by patients who improve after IV tissue plasminogen activator (IV tPA) alone, and would also simulate ‘real-world’ practice. In every day work, if we obtain an angiogram of a patient who has received IV tPA and lysed their proximal occlusion before angiography, we would not do any additional thrombectomy, and that patient is likely to do well. So why would we want those patients to cloud the data in a trial?

Patients are currently being enrolled in large, multicenter, randomized controlled trials which are using a variety of advanced imaging techniques, different thrombectomy devices, and patient selection criteria. But what if the key to a positive trial were to take a page from the past and randomize on the angio table? Many investigators would be reluctant, but we believe the benefit of thrombectomy would be shown, with small sample sizes, if we did so.

References

View Abstract

Footnotes

  • Contributors MVJ and RAM contributed to the idea and content for this editorial.

  • Competing interests None.

  • Provenance and peer review Not commissioned; internally peer reviewed.