Introduction
Stroke is the leading cause of adult disability in the developed world.1, 2 In ischemic stroke, a blood clot (thrombus) leads to an area of brain experiencing a significant drop in cerebral blood flow, which can ultimately result in death of this hypoperfused tissue. Collateral blood flow gives rise to the ischemic penumbra, which is defined as an area of hypoperfused tissue that is at risk of death (infarction) if hypoperfusion persists, and is separate from tissue that has already died (or programmed to die) soon after stroke onset (core).3 The concept of the ischemic penumbra led to the notion that opening an occluded blood vessel in order to re-establish blood flow to the ischemic area (reperfusion) would rescue the ischemic penumbra from death3,4 (Figure 1). Reperfusion can be achieved through the use of systemic thrombolytic drugs, although thrombolytic agents may theoretically be more effective if delivered directly into the clot (intra-arterial delivery) or if assisted with ultrasound insonation of the clot (sono-thrombolysis).
Thrombus formation and fibrinolysis
A thrombus can originate from several different causes, which may alter its susceptibility to thrombolytic drugs. The main causes of thrombus formation in ischemic stroke are atherosclerosis (in-situ thrombus or thromboembolic) or forming in the heart (cardioembolic). It was originally hypothesized that the source of thrombi had a large determining influence on clot composition and therefore potentially upon susceptibility to thrombolytic treatment as thromboembolic clots tend to be platelet and fibrin rich and were formed in high flow areas (often referred to white clots), whilst cardioembolic can contain tissue debris such as fat, air or bacteria interwoven inside the platelet and fibrin mesh which were formed in low flow areas (often referred to as red clots).5,6 However recent histological studies analyzing the thrombi extracted from patients using intra-arterial clot retrieval showed thrombi composition was similar between cardiac and arterial origins.7
Atherosclerotic plaque rupture or cardio-embolic thrombosis (typically due to atrial fibrillation) leads to activation of the coagulation cascade as well as platelet activation. In the cascade, zymogens (free floating inactive coagulation factor precursors) are converted to an activated coagulation factor by interaction with the atherosclerotic plaque/thrombus.8 Each active zymogene is able to activate nearby zymogens, leading to a large localized reaction of coagulation factors. Activated platelets then catalyze an interaction between activated coagulation factors (zymogens) to aid in the generation of thrombin by conversion of the soluble protein fibrinogen to insoluble fibrin, forming a blood clot. There are two blood coagulation cascades (intrinsic and extrinsic pathways), that have separate initial pathways but converge on a common pathway9 (Figure 2).10
Thrombolytic drugs
Thrombolytic drugs dissolve (lyse) thrombi in the vascular bed by activating plasminogen to form plasmin. Plasmin is a proteolytic enzyme that breaks the crosslinks between fibrin molecules to destabilise the structural integrity of blood clots. The main types of thrombolytic drugs used in ischemic stroke to activate plasminogen are urokinase/streptokinase and tissue plasminogen activators (eg, Alteplase). Thrombolytic drug development has undergone at least three generations with the aim of enhancing fibrin specificity or reducing inhibition of thrombolysis by plasminogen activator inhibitor type 1 (Table 1).
Tissue Plasminogen activator (tPA) is a serine protease found on the endothelial cells lining blood vessels and is involved in the breakdown of blood clots (thrombus). A thrombus is composed of fibrin monomers that are cross-linked through lysine side chains which tPA binds too. Lysine binding of tPA results in a activation of plasminogen only around a thrombus, which minimises activation of circulating plasminogen. The lysine side chains have a high affinity for binding with plasminogen, making the thrombus plasminogen rich. The tPA enzyme binds to fibrin components of a thrombus and catalyses plasminogen conversion to plasmin by cleavage of the arginine-valine bond at positions 560 and 561 to break down a clot by degrading the fibrin matrix of a thrombus. Plasmin then breaks the thrombus down into fibrin degradation products due to the plasmin's protease action in dissolving the thrombus.11 However, on fibrin bound plasmin, the inhibitory effect of alpha 2-antiplasmin and type 1 plasminogen activeator inhibitor restricts lysine binding.12
Streptokinase is a first generation thrombolytic agent. Streptokinase is an antigenic compound isolated and derived from purified streptococci bacteria. Therefore streptokinase is not a protease but binds to plasminogen for the generation of plasmin and is not restricted at the site of thrombus formation. Because of its non lysine specificity, Streptokinase, produces more fibrin degradation products (fibrinogensis) as a result of widespread lytic action in the body. Fibrinogensis occurs when there are high levels of fibrin in the blood and can cause thrombosis, haemorrhage or tissue oedema. Although much less expensive, this makes Streptokinase a less attractive agent for acute treatments in ischemic stroke than tPA products. Studies of streptokinase in acute stroke were stopped due to an increase in mortality compared to placebo due to increased haemorrhage rates.13,14 The specificity of tPA drugs (Alteplase, Retaplase, Tenecteplase and desmoteplase) for plasminogen bound fibrin means that conversion of plasminogen to plasma occurs in clots with minimal circulating plasma.
Urokinase is formed by kidneys and is found in urine. Like Streptokinase, Urokinase has limited clinical use due to its fibrinogensis. The 1999 Prolyse in Acute Cerebral Thromboembolism II trial (PROACT II) tested the benefit of intra arterial delivery of 9 mg of recombinant prourokinasea (r-proUK) in 180 patients. Despite an increased rate of symptomatic intracranial haemorrhage the study found a significant benefit in patients treated with r-proUK at 90 days.15 To date, the PROACT II study is the only positive phase III intra-arterial trial. Some centres use r-proUK intravenously due to its low cost compared to other thrombolytic drugs.
The second generation thrombolytic drug Alteplase, is a recombinant form of human rtPA and has undergone the most study in the clinical stroke setting. Recombinant techniques mean that molecular cloning is used to bring genetic material together from multiple sources that would not otherwise be found together, to create new DNA sequences, which can be used to manufacture drugs. Alteplase is a purified glycoprotein (a protein with sugar chains covalently attached to polypeptide side chains) of 527 amino acids that is synthesised from the complementary DNA (cDNA) of natural human tissue-type plasminogen activator found in human melanoma cells and is made up of five structural components : a protease, epidermal growth factor (EGF) and two kringle domains (Figure 3). The lysine binding sites of alteplase are on the kringle 2 domain and are why alteplase has a high binding affinity to thrombi. In terms of pharmacodynamics, rtPA has a short half life of around 5 min, and therefore requires an infusion after a bolus injection in the acute ischemic stroke setting.
The majority of ischemic stroke clinical research has revolved around alteplase, as it is the only Food and Drug Administration (FDA) approved drug for ischemic stroke. FDA approval for Alteplase was licensed for the management of acute myocardial infarction in 1987, acute massive pulmonary embolism in 1990 and acute ischemic stroke in 1996.16 The total dose of Alteplase for acute ischemic stroke is divided into a 10% bolus and 90% as an infusion over 60 min. Small dose-range studies for Alteplase in human stroke found increasing neurological benefit up to a dose of 0.85 mg/kg, but an increased rate of haemorrhage with a dose of 0.95 mg/kg.17 The ground-breaking NINDS-rtPA study was the first positive phase III acute stroke trial. This trial showed that early treatment (within 3 hours) of an ischemic stroke with alteplase substantially improved clinical outcomes. An Alteplase dose of 0.9 mg/kg was used in the NINDS trial.18 and found no significant increase in the risk of haemorrhage.19 The ideal dose of alteplase is still under debate with the original alteplase dose-ranging studies being quite small. It is felt that Asian people may be more prone to haemorrhage from tPA, although this is not based on level one evidence. To this point, the standard dose of alteplase in Japan in 0.6 mg/kg based on a single Japanese study.20 On the other hand, there is justifiable concern that a reduced dose of alteplase will not successfully lyse larger clots, typically in more proximal locations. It has been reliably demonstrated that long clots on non contrast computed tomography (NCCT) show a reduced rate of reperfusion and subsequent clinical improvement.21 Indeed, a recent study of 130 patients using non-contrast CT to measure occlusion length identified that the tPA dose was independently predicative of thrombus resolution, with complete resolution in only 8% of patients. The average occlusion reduced by 20%, while 15% of patients had an increase in clot length following tPA administration. Therefore tPA dose was associated with thrombus resolution and improved clinical outcome.22
Following the NINDS study, which showed that early treatment (within 3 hours) was the key to improving ischemic stroke outcome, there have been several studies investigating extending this treatment time window. The European Cooperative Acute Stroke Study (ECASS) I and II looked at administering alteplase between 0 and 6 hours after ischemic stroke onset and did not find an overall benefit, with an increased rate of haemorrhage in ECASS I. However ECASS I and II did not use advanced imaging and may have had a high rate of patients with very little penumbra to salvage, especially in the 3-6 hours time window. ECASS III subsequently demonstrated evidence of good outcome following treatment with 0.9 mg/kg Alteplase between 3 and 4.5 hours in patients <80 years after symptom onset using only NCCT to exclude patients with haemorrhage or oedema. As a result of ECASS III, many countries have amended the license for stroke treatment with rtPA to 4.5 hours following NCCT assessment.
There has been much interest in using advanced imaging to visualise the volume of the penumbra and infarct core, particularly with a view to extending the treatment window for thrombolysis. The EchoPlanar Imaging Thrombolytic Evaluation Trial (EPITHET) randomised ischemic stroke patients 3-6 hours after symptom onset to alteplase or placebo and used acute magnetic resonance imaging (MRI) to define the core and penumbra ('diffusion-perfusion mismatch'). The primary hypothesis of this phase II trial was that patients with perfusion-diffusion mismatch would have less infarct growth with alteplase than with placebo. Although the primary outcome was negative (mean geometric infarct growth) all secondary outcomes were positive and confirmed that alteplase attenuated infarct growth in mismatch patients. Other analyses from the EPITHET dataset have shown that the diffusion infarct core is the strongest predictor of response to thrombolysis, with a chance of good outcome dropping if the core >25 mL and there being virtually no benefit >70 mL.23,24 Pooled analysis from EPITHET and another trial of alteplase 3-6 hours after ischemic stroke, the Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution Study (DEFUSE), showed that patients in the 3-6 hours time window with a favourable imaging profile, i.e. a large penumbra, and small infarct core.25
As stroke occurs predominantly in the aged population, with >45% of patients receiving Alteplase being over 70 years old,26 there is a concern about decreasing benefit from thrombolysis with age. The main concerns are over the possibility of aged related renal and hepatic impairments affecting the metabolising and clearance of alteplase as well as the possibility of an increased risk of haemorrhagic transformation of the infarct following Alteplase administration. There is also concern that the elderly tend to have subtle pre-morbid cognitive dysfunction and also may have less capacity for brain plasticity and recovery after stroke. However, the International stroke trial 3 (IST3), recently found there was no direct association between Alteplase administration and haemorrhage rate in elderly patients selected for alteplase treatment using NCCT imaging.27 This was the largest stroke thrombolytic trial ever, and randomised 0-6 hours patients to 0.9 mg/kg alteplase or placebo (without using advanced imaging selection). Although the primary outcome was negative, IST3 confirmed that patients treated within 3 hours of symptom onset clearly benefited from alteplase, and the elderly had as much benefit as did younger patients.
Tenecteplase (TNK-tPA) is a third generation point mutation tissue plasminogen activator created by recombinant DNA technology from a mammalian cell line. Like Alterphase, Tenecteplase is a 527 amino acid glycoprotein. Tenecteplase has modifications at three sites of the protein structure on the complementary DNA template that differentiate TNK from Alterphase, such as substitution of threonine 103 with asparagine, and a substitution of asparagine 117 with glutamine, both within the kringle 1 domain, and a tetra-alanine substitution at amino acids 296-299 in the protease domain (Figure 4). These structural and functional changes to the cDNA mean that Tenecteplase has a longer half life and greater binding affinity for fibrin and better resistance to inactivation by endogenous inhibitor (PAI-1) compared to Alteplase. The amino acids that were replaced at the three positions are called T, N, and K according to the one letter code for amino acids, giving the drug the name TNK. Tenecteplase is expressed with carbohydrate side chains linked to the glycosylation sites of the polypeptide. The carbohydrate side chains enlarge the molecule, reducing its elimination and prolonging its plasma half life. Tenecteplase can be administered by IV bolus, without the need for follow-up infusion, with a half life of 17 minutes. Tenecteplase has been approved for acute myocardial infarction, and was demonstrated to be superior to Alteplase.28,29
In ischemic stroke, the tenecteplase dose used for myocardial infarction (0.5 mg/kg) initially did not show benefit when using standard clinical criteria to select patients for treatment, with an increase in symptomatic haemorrhage.30 However when patients were selected for treatment based on acute computed tomography perfusion imaging to identify the volume of the infarct core and penumbra within 3-6 hours of symptom onset, treatment with tenecteplase at low dose (0.1 mg/kg) appeared to be superior to alteplase in a non-randomised study.31 A subsequent randomised phase IIb study using multimodal CT to select patients (vessel occlusion, small core, large penumbra) found that moderate (0.25 mg/kg) and low (0.1 mg/kg) doses did lead to significantly better patient outcomes than standard dose Alteplase (0.9 mg/kg). Tenecteplase was associated with significantly increased reperfusion, early neurological improvement and improved 3 month functional outcome32 with a strong dose-dependent relationship, with a 0.25 mg/kg dose achieving significant reperfusion and neurological improvement compared to 0.1 mg/kg. Furthermore, despite the enhanced efficacy of tenecteplase for the larger proximal clots, there was not an increase in ICH (in fact a trend towards lower rates).
Desmoteplase is extracted from the saliva of vampire bats (desmodus rotundus). It was noted in 1932 that vampire bat saliva interfered with the haemostatic mechanisms of the host.33 The DNA sequencing of four plasminogen activators in vampire bat saliva were completed in 1991, with activator alpha 1 (rDSPAα1, desmoteplase) being the most active34 and shows a 72% homology to human t-PA.35 Desmoteplase is structurally similar to Alteplase, but is missing the plasmin sensitive cleavage site and lysine binding kringle 2 domain. This results in Desmoteplase being more selective for fibrin (2×increase in catalytic activity), and having no known effect on the blood brain barrier. Desmoteplase has a half life of 4 hours (compared to Alteplase 5 minutes, Reteplase 13 minutes, Tenecteplase 17 minutes).
A small phase three trial of Desmoteplase in Acute Ischemic Stroke (DIAS 2) has been completed and but demonstrated no benefit in patients selected for treatment biased on acute advanced neuroimaging findings.17,36 The trial randomised patients to desmoteplase or placebo 3-9 hours after stroke onset using imaging selection with CT perfusion or MRI, assessed visually for mismatch between core and penumbra volume of >20% by the local investigator. This patient selection paradigm followed the promising phase 2 studies, DIAS and DEDAS, which demonstrated increased reperfusion and strong trends to improved outcome with desmoteplase compared to placebo. However, the primary clinical outcome DIAS 2 was negative. Apart from a very small sample size for a phase III study, it is thought the predominant reason for lack of benefit seen with desmoteplase was the lack of standardisation of the advanced imaging assessment. Additionally, DIAS 2 did not have a minimum or maximum volume of infarct or penumbra volume as inclusion criteria, as such, many patients had small perfusion lesions with a favourable natural history that would have made detecting a difference with desmoteplase treatment challenging even with a larger sample size. Subsequent analyses of pooled data from DIAS and DIAS-2 identified that patients with a proximal cerebral vessel occlusion or high grade stenosis had much greater mismatch tissue volumes and a positive response to desmoteplase compared to placebo.37
Enhancing thrombolysis delivery
The effectiveness of tPA treatment decreases the more proximal an occlusion becomes. The rate of early (effective) recanalization using alteplase is 25% in patients with a proximal middle cerebral artery occlusion and 10% in patients with an internal carotid occlusion.38 Additionally, the rate of re-occlusion, a situation where a thrombus moves down a vascular tree but does not dissolve and re-occludes a distal blood vessel, is as high as 30% with alteplase.39 Therefore, methods to enhance recanalization are required.
Intra Arterial (IA) thrombolysis and more recently clot retrieval have become increasingly popular acute stroke treatments, albeit without strong evidence. The PROACT II trial was the first IA thrombolysis trial to show benefit (compared to placebo). Subsequent phase II trials such as Interventional Management of Stroke (IMS) I and II suggested benefit of a 'bridging' IV/IA approach when compared to IV rtPA alone. With evolution of technology, there has been an increased interest in mechanically removing clots.40 However, three recent IA clot retrieval tials using first generation endovascular devices have recently been published including, IMS III, SYNTHESIS and MR RESCUE, all failed to show benefit compared to IV rtPA. This mostly likely relates to delays in revascularisation compared to IV treatment, and also to poor patient selection. For example, MR RESCUE not only had a very late time to endovascular treatment (median >6 hours), but the pre-treatment baseline infarct core volume was exceptionally large (median 60 mL).41,42 These trials with the first generation clot retrieval devices (MERCI and Penumbra) had quite poor recanalization rates, which may be another reason why they failed to show improved clinical outcomes compared to IV tPA alone. However, there is unbridled enthusiasm for the newer 'thrombectomy' devices such as Solitaire43 and Trevo,44 with phase II studies showing much better recanalization than with the first generation devices. However, it is unlikely any device will prove superiority to IV tPA unless clot retrieval can occur at a similar time point to IV treatment. However there is still a trend towards increased haemorrhage in patients undergoing IA treatments regardless of time to treatment.
Sono enhanced thrombolysis has been shown in vitro and in vivo to accelerate the enzymatic activity of thrombolysis.45 One proposed method of action, is that ultrasound pops the microbubbles in a targeted blood vessel to generate micro-streaming of blood to the occlusion as a delivery mechanism for tPA. Additionally, ultrasound causes enlargement of the fibrin mesh to allow better binding and penetration of tPA into a thrombus.46
Phase 2 clinical trials have shown that continuous transcranial ultrasound at 2 MHz significantly increased the rate of early recanalisation in patients treated with alteplase.47 However these studies did not use the gold standard of angiography to assess recanalisation, rather they relied on transcranial ultrasound changes, which are less well validated. A phase III trial of sonolysis (CLOTBUSTER) is soon to commence. There has also been interest in administering intravenous micro-bubbles as a contrast agent in order to increase the available volume of micro-bubbles for ultrasound to affect, however this approach may increase risk of ICH.48
Glycoprotein IIb/IIIa inhibitors may prevent platelet activation induced by thrombolysis and promote a more complete and rapid action of thrombus breakup.49 Glycoprotein inhibitors are used to prevent platelet activation to stop new blood clots forming to avoid re-occlusion. Platelet activation by Adenosine diphosphate (ADP, stored in blood platelets and is released on platelet activation which can be blocked by clopidogrel) leads to a conformational change in platelet gpIIb/IIIa receptors that induces binding to fibrinogen. However recent myocardial studies have shown that gpIIb/IIIa inhibitors lead to more effective reperfusion, but also increased the rate of intracranial haemorrhage.50 For ischemic stroke there is currently little clinical data, with only small case series indicating a low rate of symptomatic intracranial haemorrhage after duel alteplase and gpIIb/IIIa administration.51,52
Nanotherapeutics aim to increase tPA delivery to the vessel occlusion. The current concept of thrombolysis is to inject a thrombus dissolving drug into the blood stream and hope that it reaches the occlusion site in a high enough concentration to totally dissolve the thrombus. The chosen doses of clot-lysing drugs that can be administered are a trade off between the potential risk of bleeding and the required dosage to have an effect. However, the more distal an occlusion is, the less likely there is to be effective drug delivery regardless of dose. One recent approach to guide tPA delivery is through the use of shear-activated nanotherapeutics (SA-NTs).53 However nanoparticles are only in phase one clinical trials.
Treatment of ischemic stroke with thrombolysis is a trade off between the risk of haemorrhage and the possible benefit. Clinical trials have shown that treatment with tPA in selected patients results in highly clinically significant improvement. However the criteria on how we best identify these highly treatmentresponsive patients are still undergoing investigation. There is much interest in the use of neuroimaging to enhance our ability to identify these responders but this approach is yet to be proven in a phase III clinical trial.54 While better thrombolytics and drug delivery mechanisms such as sono-thrombolysis may increase the breakdown of a thrombus, there is still the major risk of haemorrhage brought about by damaged to the blood brain barrier by the stroke. It is unknown if a resulting haemorrhage is caused by interactions of thrombolytic drugs and the blood brain barrier, or extensive brain damage brought on by an established infarct. Therefore it is unknown if better drugs or mechanisms will reduce the risk of haemorrhage after stroke, or if appropriately selecting patients with a large penumbra and without a large core using imaging reduces the rate of post lysis haemorrhage. The recent failure of the endovascular trials to show benefit compared to IV alteplase demonstrate that we should focus on early treatment with lytics and concentrate on either enhancing their delivery, and/or on better agents such as tenecteplase.