2022 Update of the Consensus on the Rational Use of Antithrombotics and Thrombolytics in Veterinary Critical Care (CURATIVE) Domain 6: Defining rational use of thrombolytics

Abstract Objectives To systematically review available evidence and establish guidelines related to the use of thrombolytics for the management of small animals with suspected or confirmed thrombosis. Design PICO (Population, Intervention, Control, and Outcome) questions were formulated, and worksheets completed as part of a standardized and systematic literature evaluation. The population of interest included dogs and cats (considered separately) and arterial and venous thrombosis. The interventions assessed were the use of thrombolytics, compared to no thrombolytics, with or without anticoagulants or antiplatelet agents. Specific protocols for recombinant tissue plasminogen activator were also evaluated. Outcomes assessed included efficacy and safety. Relevant articles were categorized according to level of evidence, quality, and as to whether they supported, were neutral to, or opposed the PICO questions. Conclusions from the PICO worksheets were used to draft guidelines, which were subsequently refined via Delphi surveys undertaken by the Consensus on the Rational Use of Antithrombotics and Thrombolytics in Veterinary Critical Care (CURATIVE) working group. Results Fourteen PICO questions were developed, generating 14 guidelines. The majority of the literature addressing the PICO questions in dogs is experimental studies (level of evidence 3), thus providing insufficient evidence to determine if thrombolysis improves patient‐centered outcomes. In cats, literature was more limited and often neutral to the PICO questions, precluding strong evidence‐based recommendations for thrombolytic use. Rather, for both species, suggestions are made regarding considerations for when thrombolytic drugs may be considered, the combination of thrombolytics with anticoagulant or antiplatelet drugs, and the choice of thrombolytic agent. Conclusions Substantial additional research is needed to address the role of thrombolytics for the treatment of arterial and venous thrombosis in dogs and cats. Clinical trials with patient‐centered outcomes will be most valuable for addressing knowledge gaps in the field.


INTRODUCTION
Thrombolytic agents (also known as fibrinolytics) are enzymatic plasminogen activators that convert plasminogen to plasmin, which in turn cleaves fibrin to form increasingly smaller protein fragments in the process of fibrinolysis. The endogenous fibrinolytic system is an important component of natural thromboresistance, and includes the natural plasminogen activators tissue plasminogen activator (tPA) and single-chain urokinase plasminogen activator (scu-PA; or urokinase plasminogen activator [uPA]). 1 Plasminogen activator inhibitor (PAI)-1 is the primary inhibitor of these endogenous fibrinolytic proteins. Pharmacological thrombolytics have been developed to facilitate the lysis of thrombi associated with disease states, with the goal of restoring blood flow through occluded vessels. The predominant pharmacological thrombolytics work by accelerating natural fibrinolysis, although direct thrombolytic enzymes have also been developed. 2 Three generations of thrombolytics are now used in clinical practice (Table 1).
Individual thrombolytics and generations of thrombolytics vary in their half-life, fibrin specificity, and susceptibility to inhibition by PAI-1, among other features. Function is directly related to their complex structure, with different components of the molecule conferring specific pharmacokinetic and pharmacodynamic properties. 2,3 First-generation thrombolytics (eg, streptokinase, urokinase) are naturally occurring compounds that have negligible fibrin specificity and are significantly inhibited by PAI-1. The lack of fibrin specificity increases the risk of hemorrhage associated with their use. 1 Fibrin specificity is improved with second-generation agents, specifically recombinant tissue plasminogen activator (rt-PA). Alteplase is the most widely available recombinant single-chain tPA. Third-generation products, such as reteplase and tenecteplase, were developed to reduce PAI-1 inhibition and increase half-life while maintaining or improving fibrin specificity. 4  The aim of Domain 6 was to systematically review available evidence and establish guidelines related to the use of thrombolytics for the management of small animals with suspected or confirmed thrombosis.

MATERIALS AND METHODS
The first CURATIVE guidelines were first published in 2019. 5 Domain 6 originated from feedback received during the initial CURATIVE consultation process that highlighted the need for guidelines regarding the use of thrombolytics. As previously described, the process of develop- Specific protocols for rt-PA were also evaluated. Outcomes assessed included efficacy and safety. Regarding efficacy, since endpoints vary among studies, outcomes were broadly considered as being patientcentered (eg, survival to hospital discharge, return of function) or not patient-centered (eg, revascularization documented via angiography).
Hemorrhage and complications of reperfusion were the specific safety outcomes of interest.
Relevant studies were assessed to determine the level of evidence (LOE) from 1 to 6, methodological quality, relevance to the PICO question, and magnitude of the observed effect supporting or opposing the PICO question. Randomized, controlled, clinical studies in companion animals were considered LOE 1. Controlled clinical studies in companion animals without randomization were considered LOE 2. Laboratory animal studies in dogs or cats were considered LOE 3, with further delineation based on the inclusion of randomization and controls (good quality), lack of randomized controls (fair quality), and studies without controls (poor quality). Retrospective clinical studies using controls but without randomization were considered LOE 4, while case series were LOE 5. Given differences between people and small animals regarding the fibrinolytic system, causes of thrombosis, clinical presentation, diagnostic capabilities, and hospital resources, the consensus decision of the CURATIVE Steering Committee was to exclude LOE 6 studies (human clinical studies). Note that quality assessments of the same study may vary by PICO question, since relevance to the PICO question is a key component of quality. Consistent with previous domains of CURATIVE, guidelines are written as "We recommend" where strong supportive evidence exists, and "We suggest" where the evidence is weak. 5 Preliminary guidelines were presented at the International Veterinary Emergency and Critical Care Symposium in September 2021.
Guidelines (Appendix A) and the proportion of the working group members reaching consensus via Delphi surveys, including the survey round in which consensus was reached, are reported.

PICO question: Thrombolysis in arterial thrombosis (dogs)
In dogs with suspected or confirmed arterial thrombosis (P), does use of a thrombolytic agent (I) compared to no thrombolytic agent (C) improve any outcomes (O)?

Evidence summary
Five LOE 3 good-quality studies supported the PICO question. [8][9][10][11][12] These studies varied in the mechanism of arterial thrombosis and the specific thrombolytic. Leach  Dogs treated with streptokinase experienced more bleeding from surgical sites (ie, site of femoral vein exposure, thoracotomy, and thrombin injection site in the coronary artery) than saline-treated dogs. 12 Rebello and colleagues also reported thrombolytic efficacy using an electrolytic injury canine coronary artery thrombosis model. 9 This study included various treatment groups, but the most relevant to the PICO question was the comparison between saline control and rt-PA.
All interventions occurred after 30 minutes of thrombotic occlusion.
IV rt-PA therapy resulted in a higher frequency of reperfusion (11/12; 92%) than in saline-treated control dogs (0/11; 0%). However, 2 dogs in the rt-PA group experienced ventricular fibrillation (VF) during reperfusion and subsequently died. 9 Additionally, 7 of 9 surviving dogs had reocclusion, leading to the authors' recommendation that antiplatelet and anticoagulant therapies be used to maintain coronary blood flow following reperfusion. 9 Feuerstein and colleagues investigated both anistreplase and tPA in an induced canine coronary artery thrombosis model. 10 20 Maki and colleagues compared catheter-directed intraarterial (ie, intracoronary) administration of urokinase (n = 6), or alteplase (n = 8), with saline (n = 5) in an endothelial injury model of coronary artery thrombosis. 16 Treatments were administered 30 minutes after thrombus formation with angioscopy used to assess thrombolytic efficacy at 60 minutes. All dogs in the alteplase group had complete thrombolysis, compared to zero dogs in the control group, while urokinase treatment resulted in incomplete lysis with residual thrombi in all animals. 16 Haberstroh and colleagues evaluated numerous thrombolytic agents in an autologous clot injection model of renal artery thrombosis. 17 Local and systemic application of thrombolytic agents resulted in complete recanalization, whereas the clot remained stable in control-treated animals with or without systemic heparinization.
Nonetheless, kidney injury, as demonstrated by increased serum urea or creatinine and decreased glomerular filtration rate, remained during 8 weeks follow-up, despite successful thrombolysis. 17 Tomaru and colleagues reported the use of thrombolytics in a model of bilateral iliac artery thrombosis where contralateral arteries acted as controls. 19 In 5 dogs, saline placebo was catheter-delivered into 1 thrombosed iliac artery, while low-dose tPA (tisokinase, 50,000 IU) was catheter-delivered into the other thrombosed artery. After 60 minutes, thrombotic stenosis decreased more in tPA-treated arteries compared to saline-treated controls. 19 Rote and colleagues published 2 studies that supported the PICO question using bilateral carotid artery thrombosis models, where contralateral carotid arteries acted as controls. 21,22 Local administration of anisoylated plasminogen streptokinase activator complex proximal to the occlusive thrombus restored blood flow, whereas flow was not restored in control arteries. A 70% reocclusion rate was reported, however. 21 Lu and colleagues reported the efficacy of rt-PA-induced thrombolysis with a platelet inhibitor in a combined model of arterial and venous thrombosis in dogs. 23 The arterial thrombosis component of this model In contrast, no dogs in the vehicle control group experienced thrombolysis. One of 6 dogs in the vehicle-treated group died of VF, while 2 dogs allocated to the rt-PA group died of VF during the occlusion or reperfusion. 24 Gu and colleagues also used rt-PA (intravenously or intraarterially) to induce thrombolysis in a coronary artery thrombosis model. 25 Thrombolysis occurred in dogs that received rt-PA, but not in control-treated dogs. 25 A large number of publications report the use of thrombolytics in experimental arterial thrombosis models in dogs (LOE 3), but lack suitable controls thereby precluding their use to address the PICO question.  In most of these studies, thrombolytics were administered within 60 minutes of thrombosis, 10 One case report (LOE 5) also supported the PICO question. 97 This report was considered to be of fair quality since it had objective outcome measures and long-term follow-up. This case report described an 8-year-old, intact female Maltese dog that was suspected to be hypercoagulable due to a malignant mammary adenocarcinoma. 97 The dog presented for evaluation of a 3-day history of unilateral (right) pelvic limb paralysis, and had a weak femoral pulse. No blood flow was detected in the artery by Doppler, and thermography was used to determine the location of the arterial thrombus. Initial IV administration of streptokinase was ineffective. Subsequent local, intraarterial, catheter-directed administration of streptokinase was effective; vessel patency was confirmed by thermography, and the right femoral pulse was palpable. Complications associated with hemorrhage or reperfusion were not evident. UFH and clopidogrel were also administered, and the dog was discharged with a near normal gait on the third day following thrombolysis. At 10-month follow-up, thrombosis had not recurred, and the dog was still receiving clopidogrel. 97 Another case report was considered neutral to the PICO question, since rt-PA was given early in the treatment course. 98 This case report described a 5.3-year-old male Yorkshire Terrier with distal aortic thrombosis secondary to protein-losing enteropathy. IV rt-PA was administered (1 mg/kg bolus every 60 min for a total of 10 doses), after which pelvic limb withdrawal responses returned, and pulses improved.
Due to a stagnation in clinical improvement, 2 additional doses of rt-PA were administered on each of the sixth and seventh days of hospitalisation; pulse quality was again noted to improve, no complications were evident, and the dog was discharged on the ninth day. 98 No studies were identified that opposed the worksheet question.

PICO question: Thrombolytic agents in arterial thrombosis (dogs)
In dogs with suspected or confirmed arterial thrombosis (P), does use of 1 specific thrombolytic agent (I) compared to any other thrombolytic agent (C) improve any outcomes (O)?

Thrombolytic agents in arterial thrombosis (dogs)
Delphi consensus reached in 19/19, Round 1 a. In dogs with confirmed acute arterial thrombosis, there is insufficient evidence to support the use of one thrombolytic agent over another.
b. Of the currently available thrombolytic drugs, rt-PA has been used most widely, but when indicated the choice of thrombolytic agent will likely be dictated by availability. Of the 10 supporting studies, 4 demonstrated that a thirdgeneration thrombolytic agent was superior to a second-generation thrombolytic. Specifically, 2 studies supported that monteplase was superior to rt-PA, 53,56 1 noted that intraarterial reteplase was superior to IV alteplase, 35 and 1 that noted that IV reteplase was supe-rior to IV alteplase. 74 One study demonstrated that a secondary generation thrombolytic (alteplase) and a third-generation thrombolytic (reteplase) were both superior to the first-generation drug streptokinase. 99 Interestingly, although 3 studies demonstrated that alteplase was superior to the first generation drugs urokinase or single-chain urokinase, 16,17,75 other studies suggested superiority of urokinase 77 or pro-urokinase 32 over tPA. Each of the included studies is described in more detail below.

Evidence summary
Martin and colleagues (LOE 3, good) demonstrated that alteplase and reteplase were superior to streptokinase. 99  streptokinase), but cumulative patency time was significantly longer in the reteplase group compared to the streptokinase group. Additionally, the postmortem weight of the residual thrombus was significantly lower in the alteplase and reteplase groups than the streptokinase group. Regarding adverse effects, the incidence of rebleeding from 1and 2-day-old ear incision sites was highest in the alteplase group, but differences were not statistically significant. 99 Saito and colleagues (LOE 3, good) demonstrated superiority of monteplase to native t-PA and urokinase. They randomized dogs (n = 6/group) to receive either monteplase (E6010), native t-PA, or urokinase 60 minutes after LAD coronary artery occlusion. 56 Monteplase was administered as a single IV bolus (0.2 mg/kg). 56 The total t-PA dose was 0.6 mg/kg, with 10% administered as a bolus and the remaining 90% administered as an IV CRI over 60 minutes. The total dose of urokinase was 60,000 IU/kg administered as an IV CRI, again with 10% of the dose given as a bolus and the remaining 90% over 1 hour. Time to complete reperfusion was not different among groups.
The rate of reperfusion, however, was more gradual in the monteplase group, than the t-PA or urokinase groups (P < 0.01). Reperfusion with native t-PA and urokinase resulted in significantly more ventricular premature contractions per minute compared to baseline; however, this was not evident in the monteplase group. Monteplase-treated dogs also had significantly fewer ventricular premature contractions at 15 minutes after reperfusion than the urokinase group (P < 0.05).
While mortality due to VF was not statistically different between groups, it was numerically higher in the t-PA (3/6) and urokinase (2/6) groups, than the monteplase group (0/6). As such, monteplase was considered superior to native t-PA and urokinase in this study. 56 The same group compared the effects of monteplase-induced thrombolysis to rt-PA and urokinase on left ventricular function in the same model of coronary artery thrombosis (LOE 3, fair); again demonstrating some evidence for superiority of monteplase. 53 Thrombolysis was commenced 30 minutes after occlusion with either monteplase (0.2 mg/kg IV bolus), rt-PA (0.6 mg/kg total dose, 10% bolus, then 90% as IV CRI over 1 h), or urokinase (0.38 mg/kg IV CRI over 1 h).
Reperfusion time was not significantly different among groups, and no reocclusion was observed over a 4-hour period. Monteplase was superior to rt-PA and urokinase with regard to earlier recovery of left ventricular ejection fraction and regional wall motion. 53 Interestingly, another study by the same group (LOE 3, fair), this time using a femoral artery thrombosis model, was considered neutral to the PICO question in that monteplase had equivalent thrombolytic activity to an IV CRI of rt-PA at the same dose. 20 Gurewich and colleagues (LOE 3, fair) compared IV administration of rt-PA with the M5 mutant of pro-urokinase in a femoral artery thrombosis model. 32 Pro-urokinase (2 mg/kg IV) was comparably effective to rt-PA (1.4 mg/kg IV over 60 min, with 20% given as a bolus) when assessed by radioisotope counts over the femoral artery at 90 minutes after therapy, as well as postmortem examination. Rethrombosis developed in 1 dog in the rt-PA group. Safety was superior in the pro-urokinase group, with much lower blood loss from standardized incisions on the abdomen and ear (mean: ∼4 ml blood loss) than the rt-PA group (mean: 40 ml, P = 0.026). One dog in the rt-PA group lost >110 ml blood (>10% body weight). 32 Qureschi and colleagues (LOE 3, poor) performed a randomized comparison of intraarterial reteplase to IV alteplase in a dog model of acute basilar artery thrombosis. 35 It was considered of poor quality to answer the PICO question because although 2 different thrombolytic drugs were compared in a randomized fashion, they were administered by different routes. Thrombolysis was performed 2 hours after verification of arterial occlusion, and serial angiography was used to monitor for recanalization over a 6-hour period. Additionally, a postmortem brain MRI was performed to assess for hemorrhage or infarction.
Alteplase was administered at a dose of 0.9 mg/kg infused over 90 minutes IV, compared to catheter-directed reteplase injected into the proximal portion of the clot (0.09 U/kg over 20 min). The reteplase dose is considered to be equivalent to half the alteplase dose. At 6 hours, there was no significant difference in the rates of partial or complete recanalization between groups (2/6 in the alteplase group, 5/7 in the reteplase group, P = 0.2), and reocclusion occurred in 1 dog in each group. Intraarterial reteplase was considered superior since it was not associated with intracerebral hemorrhage (0/7) in contrast to IV alteplase (4/6, P = 0.02), including 1 dog that experienced hemorrhagic transformation of the cerebellar infarct. 35 Maki and colleagues compared the catheter-directed intracoronary administration of urokinase (480,000 units, n = 6) with tPA (alteplase, 12,000,000 units, n = 8) 30 minutes after the formation of an occlusive thrombus in the LAD coronary artery of dogs (25-30 kg body weight). 16 Alteplase was superior to urokinase in that all dogs in the alteplase group experienced complete thrombolysis when assessed angiographically and angioscopically 60 minutes after treatment. In contrast, thrombolysis was incomplete in the urokinase-treated dogs, and angioscopy demonstrated residual adherent thrombi in all dogs. 16 Haberstroh and colleagues compared the thrombolytic efficacy of Additionally, the lower rt-PA dose resulted in a greater thrombolysis than the lower urokinase dose (P < 0.05).
Although some studies have shown that rt-PA is superior to urokinase, a study by Fitzgerald and colleagues documented the opposite in a coronary artery occlusion model (LOE 3, fair). 77 In their study, groups a, c, d, and f were most relevant to the PICO question, and they compared rt-PA to urokinase and pro-urokinase. In each group, the thrombolytic was commenced 2 hours after complete coronary occlusion and administered as a CRI via a peripheral vein until 10 minutes after reperfusion. The rt-PA rate was 10 µg/kg/min (n = 10). Urokinase was administered at either 1000 IU/kg/min (n = 8) or as a 6600 IU/kg bolus, followed by an infusion of 75 IU/kg/min (n = 12); these groups were combined for analysis given that reperfusion rates were not different. Pro-urokinase was administered at 20 µg/kg/min (n = 7). Time to reperfusion was similar for rt-PA, urokinase, and pro-urokinase; however, the rate of complete reocclusion was significantly higher in the rt-PA group (9/10) than the urokinase group (1/20, P < 0.001).
Nonetheless, cyclic flow variations did occur in the urokinase groups.
The rate of bleeding from a standardized thoracic incision was not different among groups. 77 Another study was neutral to the PICO question, demonstrating no difference in efficacy between comparable doses (0.25, 0.5, and 1 mg/kg) of thrombolytic (rt-PA and rscu-PA) as measured by frequency and rate of recanalization and persistence of patency over 2 hours, in a femoral artery eversion graft model of thrombosis. 23

PICO question: Thrombolysis in venous thrombosis (dogs)
In dogs with suspected or confirmed venous thrombosis (P), does use of a thrombolytic agent (I) compared to no thrombolytic agent (C) improve any outcomes (O)?  Additionally, groups of 3 dogs received 1 mg/kg rscu-PA or 1 mg/kg rt-PA without ridogrel, a control group of 5 dogs received no thrombolytic but ridogrel, while another control group of 3 dogs received no thrombolytic and no ridogrel. The degree of clot lysis was lower in the 2 control groups (22% ± 1% without ridogrel, 28% ± 4% with ridogrel) compared to the 1 mg/kg rscu-PA group (80% ± 6% without ridogrel, 75% ± 6% with ridogrel) and the 1 mg/kg rt-PA group (84% ± 6% without ridogrel, 78% ± 5% with ridogrel). 23 110 This study was considered neutral to PICO question in that efficacy of urokinase and rt-PA was similar, but intrathrombic injection was more effective than parathrombic infusion.

Thrombolysis in venous thrombosis (dogs)
Lu and colleagues investigated a novel chimera of uPA and tPA, with standard rscu-PA or rt-PA. 23 Although the chimeric molecule is total clot lysis was 80% ± 6% in the rscu-PA and rt-PA groups, compared to 22%-28% ± 1%-4% in the control groups without and with ridogrel, respectively. 23

PICO question: Anticoagulants with thrombolysis (dogs)
In dogs with suspected or confirmed venous or arterial thrombosis (P), does use of a combination of an anticoagulant and a thrombolytic agent (I) compared to use of a thrombolytic agent alone (C) improve any outcomes (O)?  There was no difference between saline and hirudin in the time to reperfusion (34 ± 4 vs 37 ± 5 min) or the percentage experiencing occlusion (89% saline, 50% hirudin). The only statistically significant difference was a longer time with 100% flow restoration in the hirudin group (20 ± 6 min, compared to 7 ± 2 min in the saline group). As such, the authors concluded that these doses of hirudin delayed but did not prevent rethrombosis. 51 Nonetheless, this study was considered to provide some evidence in support of the PICO question. (60 ± 23 min). Thrombus mass was also significantly lower in the enoxaparin group (6.0 ± 1.3 mg) than that in the saline group (11.8 ± 3.2 mg, P < 0.05). This study provides support for the PICO question in that some outcome measures were improved by the addition of enoxaparin as an anticoagulant to the thrombolytic compared to the thrombolytic alone.

Anticoagulants with thrombolysis (dogs)
Jun and colleagues (LOE 3, good) also compared various anticoagulants as adjuncts to coronary thrombolysis with alteplase and aspirin in dogs. 55  supports the PICO question that use of anticoagulants in addition to thrombolytics improves some outcomes. 40 Stassen and colleagues (LOE 3, good) also evaluated LMWH (enoxaparin) and UFH as adjuncts to rt-PA (alteplase) in a combined femoral arterial and femoral venous thrombosis model in dogs. 64 All dogs were treated with a 5 mg/kg IV bolus of aspirin and 0.5 mg/kg alteplase (with 0.05 mg/kg as an IV bolus, followed by 0.45 mg/kg as an infusion over 1 h). Dogs were randomized to 1 of 7 groups (n = 4/group) to receive either saline, enoxaparin (low-, medium-, and high-dose groups, 1.5, 3, and 6 mg/kg, respectively), or UFH (low-, medium-, and highdose groups, 0.5, 1, and 2 mg/kg, respectively). Fifty percent of the dose of each anticoagulant was administered as a bolus, followed by the remainder as a 2-hour infusion. Statistically significant improvements in outcomes in the arterial thrombosis model were seen only in the high-dose anticoagulant groups when compared to control. Specifically, the time to reflow was significantly shorter in the 6 mg/kg enoxaparin group (19 ± 5 min) and 6 mg/kg UFH group (22 ± 5 min) than the saline group (120 ± 36 min) (P < 0.03). Additionally, the total time of arterial patency during the 3-hour observation period was sig-nificantly longer in the 6 mg/kg enoxaparin group (140 ± 13 min) and 6 mg/kg UFH group (120 ± 24 min) than the saline group (9 ± 5 min, P < 0.01). 64 Another study by Nicolini (1994, LOE 3, good)

PICO question: Antiplatelet agents with thrombolysis (dogs)
In dogs with suspected or confirmed venous or arterial thrombosis (P), does use of a combination of an antiplatelet agent and a thrombolytic agent (I) compared to use of a thrombolytic agent alone (C) improve any outcomes (O)?  and hirudin (10 µg/kg/min for 90 min) resulted in stable and sustained reflow after rt-PA, significantly reducing the rate of reocclusion. 89 Prager and colleagues (LOE 3, good) conducted a similar study to that of Nicolini, but it was considered neutral to the PICO question since only the combination of hirudin (1.5 mg/kg IV bolus followed by 1.5 mg/kg/h) and aspirin (5 mg/kg IV bolus), but neither agent alone, shortened the time to reperfusion and reduced recurrent thrombosis after rt-PA (1 mg/kg total dose) induced thrombolysis in a canine coronary arterial thrombosis model. 61 Yao and colleagues (LOE 3, fair) demonstrated that clopidogrel (10 mg/kg IV bolus, followed by 2.5 mg/kg/h CRI) was more effective than aspirin (5 mg/kg IV bolus) as an adjunctive treatment when UFH alone (72 ± 11 min) (P < 0.01). 70 Clopidogrel use was associated with mild to moderate bleeding around surgical incisions that was subjectively greater in the high-versus low-dose clopidogrel group, and greater than in the aspirin-treated groups. 70 In the final study that supported the PICO question, Roux and colleagues (LOE 3, fair) demonstrated some benefit of aspirin (10 mg/kg IV), with or without UFH (200 U/kg followed by 50 U/kg/h), in reducing coronary reocclusion after rt-PA-induced thrombolysis in dogs. 69 This benefit, however, was lost when the thrombogenic stimulus was enhanced. 69 Five additional studies were neutral to the PICO question. Dommke and colleagues (LOE 3, good) reported that abciximab did not significantly improve clot lysis in dogs treated with the thrombolytic microplasmin after coronary artery thrombosis. 27 Rote and colleagues (LOE 3, fair) described that pretreatment with the antiplatelet agent ramatroban (BAY U 3405) did not reduce the incidence of coronary artery rethrombosis after anisoylated plasminogen streptokinase activator complex-induced thrombolysis. 21 Additionally, Chen and colleagues (LOE 3, poor) reported that aspirin did not potentiate the effects of low-dose inogatran (a direct thrombin inhibitor) in dogs treated with rt-PA after coronary artery thrombosis. 52 Similarly, Leadley and colleagues (LOE 3, good) did not identify differences between dogs receiving UFH alone versus UFH with aspirin (5 mg/kg IV once) with regard to incidence or time to reperfusion, incidence or time to reocclusion, and thrombus mass. 46 McAuliffe and colleagues (LOE 3, fair) showed no benefit of aspirin over saline placebo in the times to coronary thrombolysis with rt-PA or rate of reocclusion. 58

PICO question: Thrombolytic protocols-Alteplase (dogs)
In dogs with suspected or confirmed venous or arterial thrombosis (P), does use of a specific protocol (dose, frequency, route) for use of alteplase (I) compared to any other protocol (C) reduce the risk of complications (eg, fatal or nonfatal hemorrhage) or improve any outcomes (O)?

Thrombolytic protocols-Alteplase (dogs)
Delphi consensus reached in 16/16, Round 2 a. We suggest that 0.5-1 mg/kg rt-PA delivered (systemically or catheter-directed) over 60-90 minutes is associated with successful thrombolysis in dogs with confirmed acute arterial or venous thrombosis.
b. There is insufficient evidence to determine if a specific tPA dosing protocol confers a safety benefit.

Evidence summary
Four LOE 3 studies support the PICO question that a specific protocol for rt-PA improves the chance of successful thrombolysis (2 good quality, 23 2 fair quality 125 ). One LOE 3 study (good) 88 and 4 LOE 5 studies 98,106,126,127 were neutral to the PICO question (discussed in depth in Supporting Information S1).

In a combined model of arterial and venous thrombosis in 75 dogs,
Lu and colleagues documented that at least 0.5 mg/kg tPA IV is necessary for consistently successful lysis of femoral arterial thrombi. 23 More effective thrombolysis of femoral vein thrombi created by injection of whole blood clots was evident at a dose of 1 mg/kg IV (compared to 0.25 or 0.5 mg/kg). 23 Another LOE 3 good-quality study demonstrated a dose-dependent effect of tPA in a canine model of completely occlusive, radiolabeled, femoral arterial thrombosis. 128 Of note they did not use alteplase specifically, but the product they sourced had equivalent thrombolytic activity to commercially available alteplase. Specifically, 0.10 and 0.20 mg/kg tPA administered IV over 60 minutes resulted in greater thrombolysis (35% and 49%, respectively) than 0.05 mg/kg (15%, P < 0.01) as determined by decreased thrombus radioactivity. While clinically relevant efficacy and safety endpoints were not assessed, tPA at these doses did not affect measured prothrombin time, aPTT, thrombin time, hematocrit, platelet count, or fibrin degradation product concentration. 128 An LOE 3, fair-quality study evaluated the efficacy of intrathrombic versus parathrombic injection of highly concentrated rt-PA in dogs in a subacute model of iliac venous thrombosis in 6 dogs (6 additional dogs were treated with urokinase). 110 Thrombi were created bilaterally such that 1 side could be treated with intrathrombic thrombolytic, while parathrombic infusion was performed on the other side. Intrathrombic injection was performed through a steel catheter, with multiple fenestrations, under high pressure. Thrombi subject to intrathrombic rt-PA injection all lysed in a median time of 64 ± 26 minutes, while those subject to parathrombic infusion had more variable lysis (3 complete lysis, 1 partial lysis, 2 no lysis). 110 One LOE 3, fair-quality study determined that a 60-minute IV infusion of rt-PA at 30 µg/kg/min (1.8 mg/kg) was more effective at producing recanalization than 15 µg/kg/min (0.9 mg/kg) in a model of coronary artery thrombosis, with concurrent high-grade stenosis. 125 Specifically, 6 of 8 dogs in the high-dose rt-PA group achieved recanalization compared to 0 of 4 dogs in the low-dose group. Nonetheless, 4 of 6 dogs in the high-dose rt-PA group experienced reocclusion during or shortly after completing the rt-PA infusion. The clinical applicability of the model used in this study is limited, however, and the study did not specify what form of tPA was used. 125 Prewitt and colleagues published an LOE 3 (good) study 88 in a coronary artery thrombosis model that was neutral to the PICO question.
They found no difference in the thrombolytic efficacy of 3 different protocols for the administration of the same total dose of 1.25 mg/kg of IV rt-PA in heparinized dogs: either a single bolus, 2 boluses administered 15 minutes apart, or a "front loaded" protocol (15% as a bolus, 60% over 30 min, and 25% over 30 min). 88

PICO question: Thrombolysis in arterial thrombosis (cats)
In cats with suspected or confirmed arterial thrombosis (P), does use

PICO question: Thrombolytic agents in arterial thrombosis (cats)
In cats with suspected or confirmed arterial thrombosis (P), does use of 1 specific thrombolytic agent (I) compared to any other thrombolytic agent (C) improve any outcomes (O)?

Thrombolytic agents in arterial thrombosis (cats)
Delphi consensus reached in 16/16, Round 2 a. In cats with confirmed arterial thrombosis, there is insufficient evidence to support the use of one thrombolytic agent over another.
b. Of the currently available thrombolytic drugs, tPA has been used most widely for arterial thrombosis in cats, but when indicated the choice of thrombolytic agent will likely be dictated by availability.

PICO question: Thrombolysis in venous thrombosis (cats)
In cats with suspected or confirmed venous thrombosis (P), does use of a thrombolytic agent (I) compared to no thrombolytic agent (C) improve any outcomes (O)?

PICO question: Thrombolytic agents in venous thrombosis (cats)
In cats with suspected or confirmed venous thrombosis (P), does use of 1 specific thrombolytic agent (I) compared to any other thrombolytic agent (C) improve any outcomes (O)?

Thrombolytic agents in venous thrombosis (cats)
Delphi consensus reached in 16/16, Round 2 a. In cats with confirmed venous thrombosis, there is insufficient evidence to support the use of one thrombolytic agent over another.
b. When indicated, the choice of thrombolytic agent will likely be dictated by availability.

Evidence summary
No studies were identified comparing different thrombolytic agents in cats with suspected or confirmed venous thrombosis. The use of streptokinase 145 and urokinase 146 has been reported for the treatment of cats with experimentally induced venous thrombosis. Marked differences in efficacy of thrombolysis between these models are reported, but this likely reflects model design, particularly the abil-ity of the thrombolytic drug to reach the thrombus, rather than the thrombolytic agents used.

PICO question: Anticoagulants with thrombolysis (cats)
In cats with suspected or confirmed venous or arterial thrombosis (P), does use of a combination of an anticoagulant and a thrombolytic agent (I) compared to use of a thrombolytic agent alone (C) improve any outcomes (O)?

Guidelines
Anticoagulants with thrombolysis (cats) The timing of the antiplatelet agents relative to thrombolytics in the case series is unclear, but it appears that antiplatelet agents were commenced either before or concurrently with thrombolytics in each of the case reports. These studies are briefly described here since they provide some evidence of the way in which antiplatelet agents have been used in combination with thrombolytics, which may inform clinical decision-making.
All of the tPA-treated cats and most of the standard-of-care-treated cats in the case control study by Guillaumin et al received antiplatelet agents. 139 Cats in the tPA group (n = 16) received either clopidogrel as the sole antiplatelet agent (9) or clopidogrel and aspirin (7). Since this was not a focus of the study, doses were not reported. 139

PICO question: Thrombolytic protocols-Alteplase (cats)
In cats with suspected or confirmed venous or arterial thrombosis (P), does use of a specific protocol (dose, frequency, route) for use of alteplase (I) compared to any other protocol (C) reduce the risk of complications (eg, fatal or nonfatal hemorrhage) or improve any outcomes (O)?

CONCLUSIONS
Generation of guidelines for the use of thrombolytics in dogs and cats is hampered by overall low levels of evidence in the literature. Substantial additional research is needed to address the role of thrombolytics for the treatment of arterial and venous thrombosis in dogs and cats.
Clinical trials with patient-centered outcomes will be most valuable for addressing knowledge gaps in the field.

PICO question: Thrombolytic agents in arterial thrombosis (dogs)
In dogs with suspected or confirmed arterial thrombosis (P), does use of 1 specific thrombolytic agent (I) compared to any other thrombolytic agent (C) improve any outcomes (O)?

Thrombolytic agents in arterial thrombosis (dogs)
a. In dogs with confirmed acute arterial thrombosis, there is insufficient evidence to support the use of one thrombolytic agent over another.
b. Of the currently available thrombolytic drugs, rt-PA has been used most widely, but when indicated the choice of thrombolytic agent will likely be dictated by availability.

PICO question: Thrombolysis in venous thrombosis (dogs)
In dogs with suspected or confirmed venous thrombosis (P), does use of a thrombolytic agent (I) compared to no thrombolytic agent (C) improve any outcomes (O)?

Thrombolysis in venous thrombosis (dogs)
a. In dogs with confirmed acute venous thrombosis, we suggest use of a thrombolytic agent can be considered following an assessment of the risk and benefit in individual patients.
b. We suggest the thrombolytic agent be delivered in a catheterdirected manner if feasible.
c. No evidence-based recommendations can be made regarding the use of thrombolytic agents for treatment of chronic venous thrombosis in dogs.

PICO question: Thrombolytic agents in venous thrombosis (dogs)
In dogs with suspected or confirmed venous thrombosis (P), does use of 1 specific thrombolytic agent (I) compared to any other thrombolytic agent (C) improve any outcomes (O)?

Thrombolytic agents in venous thrombosis (dogs)
a. In dogs with confirmed venous thrombosis, there is insufficient evidence to support the use of one thrombolytic agent over another.
b. Of the currently available thrombolytic drugs, rt-PA has been used most widely, but when indicated the choice of thrombolytic agent will likely be dictated by availability.

PICO question: Anticoagulants with thrombolysis (dogs)
In dogs with suspected or confirmed venous or arterial thrombosis (P), does use of a combination of an anticoagulant and a thrombolytic agent (I) compared to use of a thrombolytic agent alone (C) improve any outcomes (O)?

Anticoagulants with thrombolysis (dogs)
a. We suggest that combining an anticoagulant with a thrombolytic agent can be considered for treatment of dogs with confirmed arterial or venous thrombosis, where other risk factors for thrombosis exist.
b. Strong evidence for improved patient-centered outcomes is lacking and careful consideration of the potential increased risk of bleeding is indicated.
c. No evidence-based recommendations can be made with respect to the timing of anticoagulant administration in dogs undergoing thrombolysis.

PICO question: Antiplatelet agents with thrombolysis (dogs)
In dogs with suspected or confirmed venous or arterial thrombosis (P), does use of a combination of an antiplatelet agent and a thrombolytic agent (I) compared to use of a thrombolytic agent alone (C) improve any outcomes (O)?

Antiplatelet agents with thrombolysis (dogs)
a. We suggest that combining an antiplatelet agent with a thrombolytic agent can be considered for treatment of dogs with confirmed arterial or venous thrombosis, where other risk factors for thrombosis exist.
b. Strong evidence for improved patient-centered outcomes is lacking and careful consideration of the potential increased risk of bleeding is required.
c. No evidence-based recommendations can be made with respect to the timing of antiplatelet agent administration in dogs undergoing thrombolysis.

PICO question: Thrombolytic protocols-Alteplase (dogs)
In dogs with suspected or confirmed venous or arterial thrombosis (P), does use of a specific protocol (dose, frequency, route) for use of alteplase (I) compared to any other protocol (C) reduce the risk of complications (eg, fatal or nonfatal hemorrhage) or improve any outcomes (O)?

Thrombolytic protocols-Alteplase (dogs)
a. We suggest that 0.5-1 mg/kg rt-PA delivered (systemically or catheter-directed) over 60-90 minutes is associated with successful thrombolysis in dogs with confirmed acute arterial or venous thrombosis.
b. There is insufficient evidence to determine if a specific tPA dosing protocol confers a safety benefit.

PICO question: Thrombolysis in arterial thrombosis (cats)
In cats with suspected or confirmed arterial thrombosis (P), does use of a thrombolytic agent (I) compared to no thrombolytic agent (C) improve any outcomes (O)?

Thrombolysis in arterial thrombosis (cats)
a. No evidence-based recommendations can be made regarding the use of systemic or catheter-directed thrombolytic agents for treatment of acute (<6 h) arterial thromboembolism in cats.
b. We suggest that thrombolytic agents can be considered for treatment of acute (<6 h) arterial thromboembolism following an assessment of the risk and benefit in individual patients.

PICO question: Thrombolytic agents in arterial thrombosis (cats)
In cats with suspected or confirmed arterial thrombosis (P), does use of 1 specific thrombolytic agent (I) compared to any other thrombolytic agent (C) improve any outcomes (O)?

Thrombolytic agents in arterial thrombosis (cats)
a. In cats with confirmed arterial thrombosis, there is insufficient evidence to support the use of one thrombolytic agent over another.
b. Of the currently available thrombolytic drugs, tPA has been used most widely for arterial thrombosis in cats, but when indicated the choice of thrombolytic agent will likely be dictated by availability.

PICO question: Thrombolysis in venous thrombosis (cats)
In cats with suspected or confirmed venous thrombosis (P), does use of a thrombolytic agent (I) compared to no thrombolytic agent (C) improve any outcomes (O)?

Thrombolysis in venous thrombosis (cats)
a. We suggest that in cats with confirmed acute venous thrombosis (<6 h), use of a thrombolytic agent (administered systemically) can be considered when the potential benefits of thrombolysis outweigh the risks of bleeding.

PICO question: Thrombolytic agents in venous thrombosis (cats)
In cats with suspected or confirmed venous thrombosis (P), does use of 1 specific thrombolytic agent (I) compared to any other thrombolytic agent (C) improve any outcomes (O)?

Thrombolytic agents in venous thrombosis (cats)
a. In cats with confirmed venous thrombosis, there is insufficient evidence to support the use of one thrombolytic agent over another.
b. When indicated, the choice of thrombolytic agent will likely be dictated by availability.

PICO question: Anticoagulants with thrombolysis (cats)
In cats with suspected or confirmed venous or arterial thrombosis (P), does use of a combination of an anticoagulant and a thrombolytic agent (I) compared to use of a thrombolytic agent alone (C) improve any outcomes (O)?

Anticoagulants with thrombolysis (cats)
a. We suggest that combining an anticoagulant with a thrombolytic agent can be considered for treatment of cats with confirmed arterial thrombosis.
b. Strong evidence for improved patient-centered outcomes is lacking and careful consideration of the potential increased risk of bleeding is required.

PICO question: Antiplatelet agents with thrombolysis (cats)
In cats with suspected or confirmed venous or arterial thrombosis (P), does use of a combination of an antiplatelet agent and a thrombolytic agent (I) compared to use of a thrombolytic agent alone (C) improve any outcomes (O)?

Antiplatelet agents with thrombolysis (cats)
a. We suggest that combining an antiplatelet agent with a thrombolytic agent can be considered for treatment of cats with confirmed arterial thrombosis.
b. Strong evidence for improved patient-centered outcomes is lacking and careful consideration of the potential increased risk of bleeding is required.

PICO question: Thrombolytic protocols-Alteplase (cats)
In cats with suspected or confirmed venous or arterial thrombosis (P), does use of a specific protocol (dose, frequency, route) for use of alteplase (I) compared to any other protocol (C) reduce the risk of complications (eg, fatal or nonfatal hemorrhage) or improve any outcomes (O)?

Thrombolytic protocols-Alteplase (cats)
a. No evidence-based recommendations can be made for a specific protocol for use of alteplase in cats.