Table 2.2 summarizes experimental and basic studies of rt-PA for ischemic stroke. There was intense interest in establishing the efficacy of thrombolytics, especially tissue rt-PA for experimental cerebral ischemia, primarily using the autologous clot cerebral embolism model in rabbits [4, 5].

If rt-PA is administered immediately after experimental embolic occlusion, significant reduction in neurological damage occurs [4]. rt-PA may reduce neurological damage in rabbit embolic stroke models as late as 45 min after the cerebral embolic occlusion. rt-PA-related ICH did not occur when therapy was started 4 h after the onset of vascular occlusion. However, there was no benefit when treatment was delayed for 1 h. In both a small- and a large-clot rabbit embolic stroke model, there was no evidence that rt-PA changed the histological appearance of lesions compared with untreated controls. Zivin et al. [4], in a landmark 1985 Science study documented for the first time that rt-PA could, in fact, substantially improve neurological function after embolization with artificially made clots. It was primarily these data that paved the path for eventual FDA approval of IV rt-PA for stroke within 3 h in 1996 that was more recently extended to use within 4.5 h [4, 7–11].

Using awake baboons, Del Zoppo et al. [5] studied rt-PA-induced hemorrhagic transformation of ischemic brain within 3.5 h after MCA occlusion and 30 min of reperfusion. Three doses of rt-PA were infused over 1 h and compared with normal saline infusion. Peripheral (non-intracranial) hemorrhages were related to rt-PA dose, significant for the two highest doses. Peak plasma rt-PA levels were directly related to dose. No significant differences in the incidences or volumes of infarction-related hemorrhage occurred in any group compared with saline-treated animals that were sacrificed at 14 days. Their data suggest that rt-PA alone does not increase the risk (incidence or volume) of hemorrhagic infarction if administered within 3.5 h after MCA occlusion and reperfusion in baboons. Their data also suggest that rt-PA does not substantially decrease the infarct volume at any of the doses administered early after symptom onset.

rt-PA administered later [12] rather than earlier [6], albeit in different models, was more often associated with ICH. Lyden et al. 14 found no difference in the frequency of ischemic brain hemorrhagic transformation, however, when rt-PA was administered 10 min, 8 h, or 24 h after injection of autologous emboli. Del Zoppo et al. [2] infused rt-PA (0.3 mg/kg or 1.5 mg/kg) in baboons following 3 h of reversible acute MCA occlusion. Five of six animals at each dose (10 of 12 total) had petechial hemorrhagic infarction at 14 days compared with 7 of 12 control animals and infarct size did not differ between treated and untreated animals.

Vaugh et al. [14] demonstrated that rt-PA combined with IV aspirin synergistically and markedly prolonged the template bleeding time with a significant bleeding tendency. Administering reactivated PAI-1 can rapidly reverse this bleeding time prolongation.

Slivka and Pulsinella [12] investigated the hemorrhagic potential of both rt-PA (200,000 U; 10 % bolus remainder over 4 h) and SK; 10,000 U/kg bolus or 32,000 U/kg bolus, remainder over 4 h) initiated 24 h after experimental stroke in rabbits using a tandem common carotid and ipsilateral MCA occlusions with 2 h of halothane anesthesia. In addition, six rabbits were administered SK (10,000 U/kg bolus) 1 h after occlusion. Microscopic hemorrhage was frequently present in infarct tissue irrespective of treatment. Gross hemorrhagic infarction did not occur in rabbits either untreated or administered SK 1 h after occlusion but did occur in the other groups of treated animals. Two of 12 animals administered SK 24 h after injection had gross hemorrhages within the infarct. Only the rt-PA treated rabbits showed a significantly greater incidence of gross ICH than controls. Their data suggest that the use of thrombolytic agents may increase the risk of microscopic hemorrhage unless the agents are administered early enough after onset of the insult.

In pioneering work that set the stage for well-conducted clinical trials, Zivin et al. [6] found that rt-PA-induced ICH did not occur more commonly than controls when therapy was started within 4 h after the onset of vascular occlusion in rabbit emboli model of ischemic stroke. rt-PA was administered at 1.0 mg/kg at 15, 30, or 60 min after small-clot embolization (24-h aged clot) and at 2 mg/kg at 45 and 60 min after small-clot embolization. rt-PA was also administered 30 min or 4 h after large-clot (1 mm³) embolization. Evaluations were performed blinded to the treatment group. The effective dose of clots required to produce a clinically apparent neurological disorder in 50 % of a group of animals was significantly greater in the rt-PA-treated rabbits when rt-PA was administered at 15, 30, or 45 min but not at 60 min post-embolization in either the 1 mg/kg or 2 mg/kg dose. In controls, grossly apparent ICH was present in 3 of 10 (30 %) animals.

When rt-PA was administered 30 min after embolization, 8 of 14 rabbits had gross hemorrhage. When rt-PA was delayed 4 h, 2 of 10 animals had such hemorrhages (p = NS). In the small-clot model, ICH was uncommon, visible only microscopically, and only found in association with relatively large infarcts. The presence of microscopically visible intravascular clots was a function of the time between clot injection and animal death. In the large-clot model, the microscopy of the lesions in control- and rt-PA-treated rabbits was indistinguishable, and no difference was noted in neurological functions [13–15].

Lyden et al. [13] demonstrated in the rabbit emboli stroke model that ICH rates in the rt-PA (3 mg/kg or 5 mg/kg) or saline-treated group did not differ. rt-PA was infused 10 min, 8 h, and 24 h after emboli were instilled. Lyden et al. [15] also evaluated saline, rt-PA, or SK infusion in a similar rabbit embolic stroke model at various times of infusion to assess the rate of thrombolysis and ICH 24 h later. Only SK was associated with a significant increase in the rate of ICH compared with the saline controls. In animals administered 3 mg/kg, 5 mg/kg, or 10 mg/kg of rt-PA, there was no clear dose response for ICH, but there was for thrombolysis. However, only in rabbits who achieved thrombolysis was rt-PA associated with twice (24 %) the ICH rate as saline controls (12 %), whereas the hemorrhages were nearly identical in animals without thrombolysis.

Benes et al. [16] found that treatment with either rt-PA or UK significantly reduced the number of emboli present in a rabbit stroke model but only rt-PA significantly reduced the incidence of infarction. No animal suffered an ICH.

In summary, the studies by Phillips et al. [17, 18], Zivin et al. [4–6], Kissel et al. [19], Del Zoppo et al. [5] and Papadopolous et al. [20] taken together suggest that rt-PA reliably opens cerebral arteries occluded with either autologous or non-autologous embolic clots. Further, there is experimental data suggesting that rt-PA is more effective in inducing thrombolysis within precerebral vessels than systemic vessels [21]. At present, IV rt-PA at 0.9 mg/kg is licensed in many countries and appears to represent best practice and other drugs (see new thrombolytics below), doses or routes of administration should only be used in randomised controlled trials [22].

Although in vitro clot systems have been used to understand the effect of adjunts in combination with rt-PA on thrombolysis, to understand their roles in hemorrhagic conversion, the in vivo system of thromboembolic stroke with rt-PA is required, Further, reproducible models of hemorrhagic transformation associated with delayed administration of rt-PA similar to that observed in humans are required [40, 41]. Although there are unquestionable benefits from rt-PA thrombolytic activity, rt-PA also has deleterious effects on the ischemic brain including cytotoxicity and the increased permeability of the neurovascular unit with the development of cerebral edema and hemorrhagic transformation when it is delivered late after stroke [42]. Models that can study in detail the mechanisms of rt-PA-induced hemorrhagic conversion are also important in the evaluation of adjuncts to rt-PA [43].

Progress has been made in our understanding of the mechanisms of hemorrhagic conversion with rt-PA use. The significant involvement of oxidative stress/free radicals, inflammatory cytokines and proteins, and vascular/parenchymal neuroinflammation in the evolution of stroke-induced brain injury is now well-documented in animals and man [44, 45]. The extracellular matrix degrative protein matrix metalloproteinase (MMP)-9 is now recognized as a key player in both early stroke injury and the hemorrhage-producing effects of rt-PA in both animals and man [46, 47]. In addition to its role in clot lysis, rt-PA is also an extracellular protease and signaling molecule in the brain. rt-PA interacts with the NMDA-type glutamate receptor and facilitates excitotoxicity in the brain (i.e., can amplify excitotoxic calcium currents) and can also be vasoactive. Also, by augmenting matrix MMP further up-regulation after stroke, rt-PA induces further degradation of the extracellular matrix, increases neurovascular cell death, blood–brain barrier leakage, edema, and hemorrhagic conversion. Similar rt-PA toxic effects have been demonstrated in both human and animal stroke research [48–50]. Understanding the pathology of the evolution of brain injury in stroke and the pleiotropic toxic actions of rt-PA has provided the rationale for much of the adjuvants to thrombolysis molecules tested with rt-PA as discussed below. New data from these studies are expected to provide novel therapeutic opportunities for combination stroke therapy.

Experimental models need to incorporate the additional comorbidities, other diseases, and of course aging in order to mimic the human stroke situation where these variables are important. Failure to include such variables important in the stroke patient population can result in poor translational value from preclinical models. For example, rt-PA-induced hemorrhage is dependent on blood pressure, and reduction of hypertension during fibrinolysis can reduce the risk of hemorrhagic transformation [51]. Diabetes and high glucose exhibit increased risk of hemorrhage with rt-PA use in stroke. rt-PA treatment in diabetic stroked rats significantly enlarged brain hemorrhage, augmented BBB leakage, and failed to decrease lesion volume and improve functional outcome [52]. However, early insulin glycemic control for ischemic stroke with diabetes mellitus can significantly improve outcome in rt-PA thrombolysis [53]. Similarly, work with aged animals needs to be carried out to understand the opportunity of adjuncts to thrombolytics as well. For example, in diabetic rats with embolic stroke, combination therapy with minocycline plus rt-PA may be beneficial in ameliorating inflammation and reducing infarction, brain swelling, and hemorrhage after ischemic stroke with diabetes mellitus/hyperglycemia [54]. These data emphasize the importance of making translational modification of experimental models that can better reflect and predict the clinical situation.

Anticoagulants are highly effective for preventing cardioembolic stroke, but their effectiveness in non-cardioembolic stroke is unproven. Direct thrombin inhibitors (DTIs) are more effective than heparin in inhibiting platelet deposition and thrombus formation, and also show promise in preventing reocclusion after thrombolysis for both experimental thrombotic and embolic stroke. Of those being evaluated, argatroban will be discussed here, as it is the most advanced thrombin inhibitor and is being studied in stroke treatment trials.

Argatroban is used clinically because of its safe and effective antithrombotic action [56]. Thrombin is a blood-borne coagulation factor that contributes to neurovascular injury during acute focal ischemia. Thrombin mediates severe vascular disruption and neuronal damage during ischemia. Thrombin activation is a key step in coagulation, and thrombin also has been shown to mediate endothelial permeability and cellular toxicity. Thrombin also can enter brain parenchyma due to increased ischemic vascular permeability, and thus augments neurotoxicity during ischemia. Thrombin’s proteolytic activity is specifically associated with ischemia. Protease activated receptor-1, the presumptive thrombin receptor, appears to mediate ischemic neurovascular injury [57, 58]. Endogenous thrombin may play a role in this disturbance of microcirculation following cerebral ischemia [59].

Argatroban may be useful In cardioembolic stroke, increasing the improvement of recovery of stroke severity without increasing the risk of hemorrhage in one study [60]. The combination of Argatroban and IV rt-PA is potentially safe in patients with moderate neurological deficits due to proximal intracranial arterial occlusions and may produce more complete recanalization than rt-PA alone [61]. Importantly, argatroban may be an effective and safe for treatment of acute ischemic stroke presenting beyond 6 h of ischemic symptom onset [62]. In preclinical, rat studies, argatroban and rt-PA extends the window of opportunity for treatment of stroke to 4 h without increasing hemorrhagic transformation [63]. Currently, the combination of IV rt-PA with either low or high dose argatriban is being tested in a phase II randomized clinical trial (ARTSS-2).

rt-PA is a serine protease found on endothelial cells, the cells that line the blood vessels. As an enzyme, it catalyzes the conversion of plasminogen to plasmin, the major enzyme responsible for clot (fibrinogen) breakdown. Thrombolytics in clinical use include alteplase (rt-PA; approved for ischemic stroke) and reteplase and TNK (approved for other peripheral clot conditions/myocardial ischemia). In stroke alteplase is substantially underused because of time to treatment limitations and concerns regarding adverse bleeding risk. This limitation has fuelled the search for other thrombolytic agents, which display greater fibrin dependence and selectivity, but lack detrimental effects within the central nervous system (which has been largely unsuccessful as discussed above).

The many failures of neuroprotective therapies in clinical trials can suggest that neuroprotection alone without restoration of tissue perfusion and vascular integrity may not be adequate for successful treatment of acute stroke. Many have suggested combination strategies are necessary to optimize the brain protection already provided by early thrombolysis/reperfusion [64–67].

Some studies evaluate compounds given concomitantly with alteplase to reduce the haemorrhage conversion rate and/or protect brain cells from injury associated with ischemic toxicity mechanisms at thrombolysis. For example, MMP inhibitors have been evaluated as potential combination therapy candidates because they prevent MMP-induced production (and visa versa) of the inflammatory cytokine tumour necrosis factor-alpha (TNFα). MMP-9 plays a very significant role in rt-PA toxicity because it is upregulated in ischemia and responsible for degradation of the extracellular matrix. This upregulation and degradation is further upregulated by rt-PA, thus resulting in leaky vessels/vasogenic edema in stroke and results in hemorrhagic conversion when rt-PA is used later than recommended to lyse the occlusive cerebrovascular clot.

Antioxidant reagents also have been put forward due to their free radical scavenging capabilities that can be protective and reduce expression of inflammation and toxic mediators that also upregulate MMPs and stimulate apoptosis. Earlier work in the rabbit large clot embolism model demonstrated the rt-PA effectively lysed blood clots and improved outcome but rt-PA also increased haemorrhage rate as in man. In rabbit or rodent studies, alteplase-induced haemorrhage rate was reduced significantly by administration of the MMP inhibitor batimastat (BB-94) or the spin trap (antioxidant) agent alpha-phenyl-N-t-butylnitrone (PBN) [65]. Other antioxidants have also demonstrated a potential use in conjunction with rt-PA ( [67] and as listed below)

The focus here is “adjuncts to thrombolysis”. Adjuncts include therapeutic agents that have been used in pre-clinical, animal model studies in conjunction with rt-PA. Table 2.3 is a summary of research in animal models. All studies listed in this table used rats, except where indicated otherwise. Some of them have also been tested in clinical trials as a monotherapy and/or adjuncts to rt-PA, and this is also included. In addition, we have described the “adjuncts” mechanism(s) of action to help understand the actions and outcomes for consideration with thrombolysis. The data from preclinical/animal model studies with monotherapy and rt-PA adjunct therapy, and the results of clinical trials, if any, are also summarized in Table 2.3. Of interest here were treatment strategies that might reduce cerebral vascular and brain injury, decrease adverse rt-PA effects (e.g., hemorrhage), and extend the therapeutic time window for rt-PA use (i.e., thus are expected to improve access to rt-PA therapy for more patients). Thrombolysis may improve delivery of neuroprotectant molecules to the penumbral region increasing beneficial effect. Many neuroprotective agents exhibit synergistic effects with rt-PA preclinically, again including MMP inhibitors, free radical scavengers-antioxidant agents, NMDA and AMPA receptor antagonists, and anti-inflammatory and antiplatelet agents.