Expert Conversations: A Review of Data for Thrombolysis

In this podcast, Neurology Learning Network’s Stroke & Vascular Section Editor Amrou Sarraj, MD, interviews Professor Bruce Campbell, MBBS(Hons), BMedSc, PhD, FRACP, FAHMS, about data for thrombolysis. They discuss a historic perspective on thrombolytic therapies, as well as various trials including the WAKE-UP trial, EXTEND-IA TNK parts 1 and 2, SWIFT DIRECT trial, MR CLEAN trials, and more.

Discover more insights from your peers in our Stroke & Vascular Excellence Forum.

About the Speakers:

Amrou Sarraj, MD

Amrou Sarraj, MD, is an associate professor of neurology at McGovern Medical School at UTHealth in Houston, Texas.

Bruce Campbell

Bruce Campbell, MBBS(Hons), BMedSc, PhD, FRACP, FAHMS, is a professor, consultant neurologist, Head of Stroke, and Interim Head of Neurology in the Department of Neurology and Melbourne Brain Centre at Royal Melbourne Hospital in Australia. Dr Campbell is also a Dame Kate Campbell Professorial Fellow in the Department of Medicine & Radiology at the University of Melbourne in Australia.



Dr Amrou Sarraj: Greetings, this is Amrou Sarraj. I'm an associate professor of neurology at UT McGovern. Our guest today for the Neurology Learning Network podcast is Professor Bruce Campbell. Dr Campbell is a vascular neurologist, professor of neurology, and the head of the stroke unit at the Royal Melbourne Hospital.

Dr Campbell is a well‑renowned clinical trialist. He was the principal investigator for the extend IA trial, and the extend IA-TNK part I and II, among several other trials. Our subject, right within line with those, is thrombolysis for acute ischemic strokes. This is a subject with a lot of data recently. Thank you for joining us, Dr Campbell!

Dr Bruce Campbell: Great pleasure, Amrou.

Dr Sarraj: Wonderful. We will go over the data from back when, with the NINDS I and II trials, all the way to the most recent trials. We'll start by asking Dr Campbell to provide a brief perspective of the data on thrombolysis in acute ischemic stroke.

Dr Campbell: Thanks, Amrou. Thrombolysis came of age with the NINDS trial, which was published in 1995. Prior to that, there'd been some experimentation with streptokinase which was used in myocardial infarction. It didn't seem to have the safety profile that would be acceptable in ischemic stroke. NINDS was 2 trials, a part A and a part B, to get the FDA clearance.

It was successful in respect there's been a lot of evolution in trial design since then. The outcomes that we used, composites of NIHSS, MRS and a few other scales, are not the outcomes we'd necessarily choose at the primary outcome these days. Certainly, looking back with the lens of today's outcomes it was a very strongly positive trial.

That was in 0 to 3 hours after stroke onset. They had a very neat device where, to enroll a patient in the 90‑ to 180‑minute time window, the later part of that trial, they had to first enroll someone in the 0‑ to 90‑minute. There were some very early patients treated in that NINDS trial. There were then a number of other trials.

Fortunately, because of NINDS, people started treating patients with thrombolysis, but that… criteria of the later trials. They have trials like ATLANTIS, for instance, where the zero to 3‑hour window became impossible to recruit in.

Then, it became a 3- to 6‑hour trial and 5- to 6‑hour group, but then cut off because it concerns the safety. The nature of trials is that each trial then sets a benchmark for other trials, and you have to adapt rapidly. It's a lesson for some of our current trials as well.

There are a number of other trials, ECASS I and ECASS II. ECASS I, of course, use a slightly higher dose of alteplase, and probably didn't have the kind of non‑contrast CT exclusion of large established in facts that we'd now regard as standard.

They moved on to ECASS II, which wasn't positive on its primary pre‑specified outcome but did make one of the other ranking dichotomies. It's just another illustration of trials that if you pick a dichotomy on the modified Rankin Scale, you are taking a bet. We like to see a shift across the entire spectrum of the ranking in most cases.

ECASS III, of course, was the trial requested by European regulators to look at the 3 to 4.5‑hour window they'd done a meta‑analysis by that stage of the previous year at a 6‑hour trials. It really looked like they should be benefit 3 to 4.5 hours.

ECASS III was requested by the regulator to look at that specifically, and it was a positive trial. If you now put together all the 3 and a 4.5‑hour data, there's clearly a very strong signal benefit in that 3 to 4.5‑hour group. It's just not as strong as in the 3 hours. It's a very strong interaction between time and treatment as we all know time is brain.

I think guidelines around the world are very clear that 0- to 4.5‑hour thrombolysis is what we do based on a non‑contrast CT selection. The FDA license, of course, only goes 2 hours, but the EHA guidelines and all the international guidelines recommend 3 to 4.5 hours.

Dr Sarraj: Wonderful. Very‑well boiled in. As you said, the guidelines everywhere and probably the practice in most of the places deliver IV thrombolysis, alteplase, 0 to 4.5 hours based on simple imaging non‑contrast CT. Then, we had, after all of these trials that you mentioned, a good pose or paucity of evidence.

We continue to treat by time till recent data started coming out and utilizing advanced imaging, whether it's MRI, diffusion‑FLAIR mismatch or perfusion imaging that provided probably more than preliminary evidence on the potential benefit of IV thrombolysis beyond 4.5 hours so if you could briefly review those and tell us your thoughts on these findings. You had a good part of some of these analysis yourself.

Dr Campbell: Thanks, Amrou. We're going to start with the WAKE‑UP trial looking at chronological order, if you like. WAKE‑UP uses principle of diffusion FLAIR‑mismatch. What they're essentially trying to do with this, it's an MRI‑only technique, is to identify patients with an unknown onset who are actually within 4.5 hours.

It's a way of using MRI to tell what time. That's quite different to what will come to next, just perfusion mismatch. In the WAKE‑UP trial, about 500 patients who had this diffusion FLAIR‑mismatch indicating with reasonable specificity and sensitivity that they are within 4.5 hours of the onset of stroke even though we have no idea clinically when they had a stroke.

Those patients benefited from thrombolysis. Now, as a relative mild stroke population with a median NIH of 6 and obviously, you've got to be out to get an MRI acutely, which is a logistic challenge. It did show benefit and it's a useful technique if you can get MRI acutely.

One of the sub‑studies of that was interesting. They look at lacunar stroke which is a type of stroke that doesn't fit into the perfusion mismatch paradigm. Those patients also same benefit. Particularly if you're worried about BI‑CAP, what kind of stroke this would be, the way to go to try and treat that patient with thrombolysis.

As I mentioned, the EXTEND trial and the ECASS4 trial, and the EPITHET trial were all totally different. Now, we're using perfusion mismatch, using MRI or CT perfusion which, of course, is more readily available at most primary stroke centers.

The principle here is that if you don't know the time, what actually matters is the tissue, salvageability. These trials all use a paradigm where if you have a small ischemic course, small irreversibly injured region of brain and a much larger area of potentially salvageable ischemic penumbra that hybernating brain, it's not getting enough blood flow to function, but it's enough to keep it taking over metabolically.

If you re‑perfuse those patients, there's lots of evidence that those patients will have a much better outcome with great perfusion rather than not.

The original data came from the diffuse and EPITHET study. If you get out of place and you've got mismatch then rate of fusion leads to better outcome than not great perfusion. That will enter the EXTEND trial design. ECASS4 was almost exactly the same protocol in Europe, but using MRI visually assessed mismatch rather than it automated rapid‑based mismatch that we use in EXTEND.

Then we do the meta‑analysis putting those together with some EPITHET patients in the 4.5- to 6‑hour window as well. Both of this methodology produced statistically significant benefits. In EXTEND, they guess where EPITHET, if you have that automated mismatch profile, you have a benefit that certainly exceeds that of TPI and the 0- to 3‑hour window in non‑selected patients.

The odds ratio was over 2 compared to what we see with TPA during the 3 hours. The traumatic hemorrhage right was no different. There was a shift not just only in the MR‑01, being that 10% … nominated to treat of 10 to get someone back to over usual activities.

There was also a significant ordinal shift, the shift analysis of the MRS. The probability being one category O beta and your neurological recovery was better as well and independent functional recovery was better. A very clear CAT result across multiple endpoints with the meta‑analysis that we published in the "Lancet" that TPI 4.5 to 9 hours or within 9 hours of the midpoint of sleep.

That's an interesting concept. We've always taken the last known well time as the onset time, but on the EXTEND trial, and I guess 4. If you had a wake‑up stroke, we took, when you went to bed, when you wake‑up, the midpoint of that, 9 hours from the midpoint. That translates if you're using last night well to about 6‑day analysis.

A very diffused rate, 0‑ to 6‑day now window for treatment. It's interesting, with dawn and diffused rate proving this paradigm of perfusion mismatch, selecting patients beyond 6 hours in the case of thrombectomy. People often forget that those thrombolysis trials came way before. It's just they took a long time to recruit.

It does complicate things because the EXTEND and ECASS4 studies were pretty much done before thrombectomy was proven in that extended time with this. We don't have the beta run when you had someone as eligible for thrombectomy, do you do thrombolysis before you do thrombectomy? It's something we can perhaps come back to to discuss more.

In Australia, I got lots of change. We have a living guideline model where we update it every month. Guideline is currently, if you're in that 4.5 to 9‑hour window or 9 hours from the midpoint of sleep, you should give thrombolysis unless you have immediate access to thrombectomy.

That's relevant because about 60 or 70% of these patients with diffusion mismatch had a large vessel occlusion that would be eligible, but DAWN or DEFUSE-3 criteria… A big chunk of those patients who are eligible for thrombectomy, a lot of them need transfer. Our recommendation is that they received thrombolysis of the primary center and then to have their thrombectomy.

If you're in a comprehensive stroke center just like within a 4.5‑hour window, we don't really know at the moment what the role of thrombolysis use in addition to thrombectomy.

Dr Sarraj: That's a comprehensive and wonderful review. Kudos to you all in Australia having guidelines that can be updated on the go so to speak. We have not updated our guidelines here.

Still, I believe some of the hospitals and academic centers, based on the available evidence, do have protocols to treat patients with WAKE‑UP and strokes up to 4.5 to 9 hours based on the perfusion deficit.

I would ask specifically, and you mentioned some of the points related to this. A trial like the WAKE‑UP trial—and not to criticize the trial—did not necessarily change the guidelines, did not change the practices widely.

You mentioned the difficulty getting the MRI, and the other point is that very mild deficit that 42+% of the patients in the control arm achieved functional independence. Any further thoughts on that?

Dr Campbell: I think that's exactly correct. The main issues are logistic, but most of us have difficulty getting an MRI very rapidly. Certainly, not really possible in any of the primary stroke centers I work with.

Yes, the population was a highly selected population. They were predominantly mild, so they weren't the large vessel occlusion patients in there. It's a useful technique to put someone within a time window.

I guess our Australian philosophy has always been...It's more important to look at what the tissue is doing and whether you adjust within 4.5 hours because our suspicion is that there are some patients within 4.5 hours who don't benefit, so it's not really about time. It's about tissue.

Dr Sarraj: That's an excellent leeway to the next question which is, if it's only about tissue, and we tried for our work to get rid of this line in the sand with the 4.5 hours, then why do we now restrict ourselves with 9 hours? Why if it's not 12? Why if it's not 14? I mean, treating by tissue any mismatch should be treated at any time.

Dr Campbell: We have the data pretty much for that with thrombectomy. The concern, of course, with thrombolysis is the symptomatic hemorrhage risk when you're thrombolysing people with mushy brain tissue that's been infected for a long time. It's taken us a while to get comfortable with these progressive extensions.

There are now trials extending to 24 hours. We had thrombolysis, it's a TIMELESS trial in the USA. There's the ETERNAL trial in other parts of the world moving at a 24‑hour window with tenecteplase.

Dr Sarraj: Absolutely. Now, within the same subject, thrombolysis is shifting from one agent to the other for the good old friend. Alteplase or tPA has been around since NINDS trials. Recently data, especially from Australia, came supporting tenecteplase. Tenecteplase is single shot vs the prolonged infusion, some data on good safety. It is cheaper, always follow the money.

If you can review the data on that for us, what are your thoughts? I am aware that things have already shifted in Australia. It seems that you are not only ahead of us in time zones but also in terms of utilizing data, the guidelines and practice.

Dr Campbell: Tenecteplase story goes back a long way to Clarke Haley's work with early dose finding studies. Then, in the Australian part of the story, Mark Parsons led a very small but very effective study that was published in the New England in 2012. The 75 patients looking at alteplase vs 2 doses of tenecteplase.

What that showed is that 0.25 milligram per kilogram dose of tenecteplase seemed more effective than 0.1, and both of them are more effective than the alteplase in getting reperfusion in early neurological recovery. It also translated into modified Rankin benefits of 3 months of functional outcome as a standard measure.

That then led on to a number of other trials. It was the ATTEST trial done in Scotland. That one was looking at an infant growth primary outcome which it didn't achieve.

When they met or analyzed those patients with the Australian patients, there was clearly a benefit in those with a large vessel occlusion which was actually everyone in the Australian team case study. That was well before the days of endovascular thrombectomy.

We've currently got quite a few trials going on to try and establish a phase III positive result with tenecteplase. In the non‑thrombectomy category, there's the TASTE trial. It's been going on for many years in Australia. There's ATTEST 2 trial in the UK particularly. There's the ACT trial. It started out more recently in Canada and NOR‑TEST 2 in Norway.

I didn't mention NOR‑TEST 1, that was a very large trial, 1,100 patients, exceptionally mild stroke patients had a median NIH stroke scale of 4 and quite a lot of mimics. About 78% of the patients included had stroke mimics.

It's interesting to think about what proportion of some of our other tPA trials may have had mimics that weren't reported on, but certainly, when you've got mild stroke patients and you're putting them in quickly, you have a high rate of mimics and that certainly dilutes the effect that you can show with a thrombolytic.

I think there's lots of data around—IST-3 another example—where they looked at patients with and without an identifiable vessel occlusion. It's very hard to show a benefit of a thrombolytic agent without evidence of the vessel occlusion.

I think that NOR‑TEST population, it's not too surprising that there wasn't a big difference between tenecteplase and alteplase, but the safety of that higher 0.4 dose looked reasonable in that study.

Moving to using it as an adjunct to thrombectomy, it's an area where I've had a big interest, so we did the EXTEND‑IA TNK trial which was published in 2018, and that showed superiority.

We designed it as a sequential first test non‑inferiority and then test superiority because we felt that if tenecteplase was at least as good for the reasons that Amrou mentioned like it's cheaper and more convenient with a single bolus, then people would probably be interested in using it.

As it turned out, we met non‑inferiority. We also have added superiority doubling the rate of endovascular reperfusion at the initial angiogram from 10% to 22%. That was very positive for EXTEND‑IA TNK.

There was also a shift towards more better functional outcomes on the rank and scale, the value 0.4. We then took that forward to see if there was any benefit in increasing beyond that 0.25 milligrams per kilogram dose we used in part one to the 0.4 milligram per kilogram dose they've been using in Norway.

Part 2 of EXTEND‑IA TNK case study showed there was really no advantage of going up in dose. It's still the same reasonably safe, but there was no benefit in reperfusion, whatsoever.

We have stuck with the 0.25 milligram per kilogram dose of tenecteplase. It is an our guidelines as a potential alternative to alteplase mainly in patients with a large vessel occlusion, but also as an option if you don't have alteplase for instance.

A lot of our country hospitals used tenecteplase for myocardial infarction, and they don't have alteplase in stock, but clearly, in that situation, you'd be much better to have tenecteplase than nothing.

That's pretty much the tenecteplase story so far. There's been a number of meta‑analyses that lump the data together. The one particularly by Burgos and Saver, suggested non‑inferiority of tenecteplase. It's crept into the AHA guidelines as a 2B recommendation, particularly in patients with large vessel occlusion. There are some people starting to use it.

It remains off‑label in every country, not just the USA. We need those phase III trials to be coming through.

Dr Sarraj: Where do you see this progressing? Would TNK potentially take over with that? Should timeless show benefit over placebo 4.5 to 24 hours, do we still need a 0 to 4.5 hours trial of TNK vs tPA, or this would be sufficient?

Dr Campbell: That's a difficult question. I'm probably a little too close to the data to get that overall perspective, particularly in the United States, which is not my climate. In Australia, people are pretty comfortable to move over in most parts of the country, but not everywhere.

There's still a big push to complete these phase III trials. It is important that we complete these phase III trials. The TIMELESS trial is obviously looking at a 4.5‑ to 24‑hour window in people with large vessel occlusion who are mostly going to have a thrombectomy.

Adding in thrombolysis to that context is certainly something that may be beneficial, but it's predominantly being done in endovascular comprehensive stroke centers. The benefit is probably mostly in patients who need to be transferred in. We'll have to wait and see how that all pans out.

Dr Sarraj: A goodly way to thrombolysis before thrombectomy. This is a very important subject, especially with recent data from 3 randomized trials showing no adjunctive benefit of thrombolysis prior to thrombectomy in improving clinical outcomes or recanalization.

The question shifted. Initially, should we do thrombectomy after thrombolysis? All of the wonderful data from the early‑window clinical trials and the late‑window clinical trials showed that thrombectomy is superior to best medical management in patients with anterior circulation LVO, and now, the question shifted. Should we do thrombolysis prior to thrombectomy?

People make the argument that it delays thrombectomy. It can dilute the treatment effect with a delay. It may end up taking the patient to the wrong center, the closest center that can deliver thrombolysis. It prevented the evolution, so to speak, with the treatment paradigms to take the patients to the closest thrombectomy center.

It, of course, add to the cost. It can have higher respite. If we can review this recent data from the direct MT to SKIP and the DVT trials, and also look at the future with MR CLEAN‑NO IV and directing and your take on the results of the initial trials, and where this would be moving in the future.

Dr Campbell: Thanks, Amrou. Really interesting area. The first thing to say is that these trials only apply to patients presenting directly to an endovascular‑capable center that is able to treat immediately.

None of them are including transferred patients or patients where you might be after hours. The team needs to be called in, and there's going to be a prolonged delay to getting to therapy. You get the procedural standing there, scrubbed, with an open table ready to go, do you give the thrombolysis as you're going out to thrombectomy?

I think that's important thing to say up front. You mentioned a number of important issues. It's concerned that thrombolysis might delay thrombectomy. I must say it's not my experience. In a nice, well‑organized, parallel‑processing system which is what we emphasized, you shouldn't be thinking, "I'm going to give thrombolysis now, do I give thrombectomy?"

The decision should be made in parallel. In that way you don't have a delay particularly if you're using tenecteplase. There's no complications of setting up infusions and things. That is another potential advantage. Of course, all these trials we use alteplase.

If you think about the way you've got to do these trials, you can't give the thrombolysis until you confirmed that they are candidate for thrombectomy, and randomized them. To consent to them randomizing. That's very different to the normal workflow where you get a non‑contrast CT say, "I'm going to give thrombolysis now. Carry on with the extra imaging," and then go to thrombectomy.

You're introducing an artificial delay when you can give the thrombolysis. For instance in the SKIP trial, I reported this that they only had 8 minutes between when they started their alteplase and when they punctured the groin.

There's no real, very fair comparison compared to if they've been able to give the alteplase when the patient arrived after the non‑contrast CT. This is contortion in the workflow that's very unfavorable to alteplase. They're also using alteplase rather than tenecteplase which we've shown to be superior in this population.

There may be more advances to come in the intravenous thrombolysis space. We need to bear all that in mind. The system issue of at the moment as distance is designed towards getting patients towards primary… for thrombolysis is true. There is a lot of emphasis now on those LVO triage tools by paramedics.

It is starting to change that if you've got a high probability of a large vessel occlusion, you'll go to a comprehensive center first stop, but yes, that has been an issue.

Then, the cost, of course, in our system, I don't think the cost of the thrombolysis is a relevant consideration, but in some parts of the world, the patient has to pay or the relatives have to physically pay for the elder care or whatever before they can be treated, of course, in those systems.

It's a very relevant consideration of how much extra bang for buck do you get when people are shelling out a lot of their yearly income for the sake of this treatment.

Let's go through the trials. DIRECT‑MT, SKIP, and the EVT, they're all non‑inferiority trials. Two of them meet the margin and SKIP did not. It was a much smaller trial though.

If you look at just the point estimates, they're all suggesting very similar outcomes and depending on which ranking CAT you take, slightly better or slightly worse.

The non‑inferiority margins are important to consider like DIRECT‑MT had a 20% non‑inferiority margin, and that really blows the effective tPA even on its own. Most people think that that was too generous in non‑inferiority margin, and it only just scraped into that at 19%.

DIRECT, EVT had a more conventional 10% non‑inferiority margin, but as the conclusion of the paper says, you got to still judge whether you think that's an acceptable non‑inferiority margin given again that's probably about as much benefit as you could possibly expect to get from tPA‑free standing, let alone in the context of thrombectomy.

There are more trials coming through. There's a SWIFT DIRECT trial, which is predominantly in Europe. There is the DIRECT‑SAFE trial, which is in Australia and some places in Asia. There's the MR CLEAN‑NO IV trial which has actually completed recruitment, and we'll hear results at the ISC. Lots of exciting data coming through.

All these existing trials have been in pretty much exclusively Asian populations, and that is an important difference because there's a lot of intracranial atherosclerosis in those populations that may behave differently. There's going to have to be a higher rate of stenting and use of other agents like eptifibatide, for instance.

We do need data for those of us practicing in Western countries that are more relevant to our population. As I said, the use of tenecteplase with distorted workflow is something that is very hard to disentangle from the results of these trials.

Dr Sarraj: I absolutely agree. Many things that you mentioned is in order for us from thrombolytic to work, at least, it should be completed. Good part, if it's completed prior to the trial. Good part is from thrombectomy and 8 minutes prior to do that does not give thrombolysis justice 87% of the patients in DIRECT‑MT did not complete thrombolysis prior to the initiation of thrombectomy.

All of these patients were DIRECT presenters. Maybe, there is a potential benefit with given time for the thrombolysis and the drip‑and‑ship model to show benefit. Hopefully, the future trials should light on that. Would it be possible that this is the same as we learned in thrombectomy?

You select identify subpopulations that may benefit, those with milder stroke, maybe, those with more distal occlusion. Things are different with milder stroke vs distal occlusion because milder stroke, it can triage in the field. Would this be a possibility?

Dr Campbell: I think you're right. It's probably going to be...In broad terms, they look similar. Then, we'll try and identify subpopulations where there's benefit with one strategy vs the other.

I also would hope that we would have better intravenous strategies as well. I mean, tenecteplase looks better now with alteplase, but we could probably do better still. It will be an evolving field and the background of thrombolytic care will probably evolve to something better, I suspect.

Dr Sarraj: The next trial that should come, should test that thrombectomy is better than tenecteplase, or we think we are sitting in that area. We don't need to repeat what we already had, right?

Dr Campbell: I don't think we need to repeat that. EXTEND, that was great, but 80% of the patients still needed the thrombectomy after the tenecteplase. I don't think there's any challenge to thrombectomy at the moment.

Maybe, if we have some wonderful cocktail that addresses all the components of thrombus and stats getting 85, 90% recanalization in the first few half an hour, something that's a different matter, but we're nowhere near that yet.

Dr Sarraj: We do not have that. We mentioned many of the future trials already, and this is great too. People know what is coming. There is also the delivery of acute treatment in the field and recent data showing their outcomes in patients being treated in the mobile stroke units from Berlin. There is the BEST‑MSU here in Houston and there are other trials in Australia.

How do you see this faring with the overall acute management of ischemic strokes and thrombolysis?

Dr Campbell: That's a great point, Amrou. We didn't discuss the mobile stroke unit, the impact on these direct trials in the previous response either it does totally change the paradigm. Mobile stroke units might seem a boutique solution, a very expensive up front solution.

We've kept rolling stuff in cost, but it’s essentially been shown to be quite cost‑effective in our initial analysis in Australia and probably will be in other context as well. Basically, there'd be proud results from Berlin, a very exciting. That's the first demonstration of improved functional outcome with MSU treatment.

There's been lots of study showing that if you give TPI faster. In Australia, we've shown that thrombectomy can be significantly faster as well. Clearly, the relationship between rate of fusion, timing and outcome is very strong. That should transfer to benefits, but the BEST‑MSU trial, of course, will be presented soon.

We are eagerly awaiting those results and if they're congruent with the Berlin results, I think that's a very strong place for mobile stroke units in metropolitan areas around the world.

Dr Sarraj: That is great. Within not delivery but within adjunctive treatments to thrombolysis, there has been attempts to add anti‑platelets and other agents to alteplase in thrombolysis to improve recanalization, achieve faster recanalization, and subsequently, clinical improvement and better outcomes.

Currently, the MOST trial is ongoing and what are your thoughts on that? Do you think that this is something that can potentially add or improve the efforts to thrombolyze the patients and improve their outcomes?

Dr Campbell: I'm optimistic that we will do better than just a single fibrinolytic strategy. Most for those who don't know, I was using eptifibatide and argatroban in a very sophisticated design to play those 2 off against each other.

I think there are a host sleuth of trials I'm aware of in the planning stage looking at lots of novel agents to try and attack different paths to clot, different ways of trying to improve thrombolysis, either standalone or in addition to fibrinolysis with alteplase or tenecteplase.

It's a really exciting field and I'm quite confident that over the next 5 or 10 years, we will see an advance in this space.

Dr Sarraj: That's great. I believe this was a wonderful and comprehensive review. We didn't only review the historic and the current data on thrombolysis, we rather went on to potential near agents with tenecteplase, the bridging vs direct thrombectomy and the adjunctive benefit vs not prior to thrombectomy, delivering methods with the mobile stroke units and additional agents.

More than one thrombolic to be currently ongoing MOST trial and maybe other trials in the future. Thank you, Bruce, and hopefully we'll have you in the future for other topics. You're multi‑talented and this is not the only topic where you have deep and well‑thought data so to speak.

Dr Campbell: Thanks very much, Amrou. It's quite a pleasure to be with you.