2022 Update of the Consensus on the Rational Use of Antithrombotics and Thrombolytics in Veterinary Critical Care (CURATIVE) Domain 1‐ Defining populations at risk

Abstract Objectives To expand the number of conditions and interventions explored for their associations with thrombosis in the veterinary literature and to provide the basis for prescribing recommendations. Design A population exposure comparison outcome format was used to represent patient, exposure, comparison, and outcome. Population Exposure Comparison Outcome questions were distributed to worksheet authors who performed comprehensive searches, summarized the evidence, and created guideline recommendations that were reviewed by domain chairs. The revised guidelines then underwent the Delphi survey process to reach consensus on the final guidelines. Diseases evaluated in this iteration included heartworm disease (dogs and cats), immune‐mediated hemolytic anemia (cats), protein‐losing nephropathy (cats), protein‐losing enteropathy (dogs and cats), sepsis (cats), hyperadrenocorticism (cats), liver disease (dogs), congenital portosystemic shunts (dogs and cats) and the following interventions: IV catheters (dogs and cats), arterial catheters (dogs and cats), vascular access ports (dogs and cats), extracorporeal circuits (dogs and cats) and transvenous pacemakers (dogs and cats). Results Of the diseases evaluated in this iteration, a high risk for thrombosis was defined as heartworm disease or protein‐losing enteropathy. Low risk for thrombosis was defined as dogs with liver disease, cats with immune‐mediated hemolytic anemia, protein‐losing nephropathy, sepsis, or hyperadrenocorticism. Conclusions Associations with thrombosis are outlined for various conditions and interventions and provide the basis for management recommendations. Numerous knowledge gaps were identified that represent opportunities for future studies.

As in the previous iteration of the CURATIVE guidelines, the association between diseases and thrombosis was evaluated using a standard Population Exposure Comparison Outcome (PECO) question format.
For conditions considered to be potentially associated with thrombosis, the veterinary literature was examined to determine whether in dogs or cats (Population, P), the development of a disease (Exposure, E), as opposed to remaining disease free (Comparison C), was associated with the development of thrombosis (Outcome, O). Recommendations for or against antithrombotic therapy based on risk for thrombosis were further refined using a Delphi survey process where CURATIVE group members were asked to agree or disagree with guideline recommendations and to suggest alternative wording as necessary. The results of the Delphi surveys and the resulting draft guidelines are available at supplementary data (S1). It should be noted that the CURATIVE Steering Committee made some minor wording changes to the draft guidelines for clarity and consistency after the Delphi surveys were completed, resulting in the final recommendations discussed below. Where appropriate, knowledge gaps were included to highlight specific areas of limitation and to encourage further investigation.

PECO QUESTION: HEARTWORM DISEASE (DOGS)
In dogs (P), is the development of heartworm disease (E), as opposed to remaining disease-free (C), associated with the development of thrombosis (O)?

Heartworm (dogs)
a. Heartworm disease is associated with pulmonary artery thrombosis in dogs, with risk increasing with disease severity.
b. We recommend that antithrombotic therapy be considered in dogs with heartworm disease, particularly in thosewith more severe disease and those undergoing adulticide therapy.

Evidence summary
Most available evidence supported the PECO question (17 studies), with 2 studies considered neutral and none opposed. Proving thrombosis in clinical settings, particularly in the pulmonary arteries, is challenging, and studies were only included if a thrombus was confirmed by angiography, ultrasound, or histopathology. Studies reporting thrombin generation (ie, D-dimers) and platelet reactivity were included to support or refute the association between dirofilariasis and a hypercoagulable state but with lower quality evidence scores. Overall, there is a strong association between clinical dirofilariasis and thrombosis and thromboembolism in dogs, with indications that this risk increases with disease severity, adulticide therapy and potentially with the presence of microfilaria.
Numerous studies document the presence of thrombi in dogs with dirofilariasis, universally in the same vascular bed as the worms. In a case series of 3 dogs with caval syndrome, 2 had extensive large and small vessel thrombosis throughout the pulmonary vasculature (LOE 5). 6 In an experimental infection study of 20 healthy dogs (LOE 3), lung histopathology showed varying degrees of thrombi in all dogs, frequently associated with worm fragments. 7 While most reports focus on the pulmonary vasculature, a small case series also documented wormassociated thrombi in 3/5 dogs suffering from aberrant worm migration involving the aorta, medial iliac, and femoral arteries. 8 Concentrations of D-dimer are increased in up to 40% of heartworm-infected dogs, 7,9 with higher concentrations seen in dogs with microfilaremia. 10,7 Similarly, plasma D-dimer concentrations correlated with disease severity (LOE 3-5), 9,11,12 increasing immediately after adulticide treatment (LOE 2), 11 and decreasing with disease resolution (LOE 3). 12,13 Platelet reactivity is also increased in infected dogs (LOE 3), 14,15 with enhanced 14 C-serotonin release and platelet aggregation in response to collagen and adenosine diphosphate (ADP) in infected dogs. Infected dogs are also more likely to experience thrombocytopenia (< 150,000 platelets/μL) than noninfected dogs.
Consistent with increased platelet reactivity, greater doses of aspirin, aspirin/dipyridamole or ticlopidine are needed to inhibit ADP-induced aggregation in infected dogs, particularly following embolization of dead worms. 16,17 Necropsy lesions were less severe in dogs receiving ticlopidine than in those receiving no antiplatelet agent in one report (LOE 3). 17 Antiplatelet therapy may diminish platelet adhesion, myointimal proliferation and vascular occlusion in infected dogs. 18,19 Adulticide treatment may increase thrombotic risk depending upon the protocol used, although older studies involving thiacetarsemide may be confounded by the thrombogenicity of the drug itself. 15 Preceding adulticide treatment with doxycycline with or without ivermectin may lessen the number of arterial lesions compared to the administration of melarsamine without pretreatment (LOE 5). 20 Treatment of dogs with patent infection following surgical worm implantation with a combination of imidacloprid, moxidectin, and doxycycline resulted in more thrombotic occlusion (on histopathology) than in untreated infected dogs (LOE 3). 21 Thrombi were frequently associated with worm fragments, suggesting that this therapy resulted in worm death and subsequent embolism. In some dogs with preexisting dirofilariasis, pulmonary artery insertion of a large number  of dead worms resulted in intimal proliferation of pulmonary arteries, dilation of main and lobar pulmonary arteries and obstruction of blood flow (LOE 3). 22 The infusion of homogenized Dirofilaria antigen into the pulmonary arteries of dogs led to thrombosis within 1 hour of administration, but these thrombi were not detectable 5 days later, suggesting that Dirofilaria are thrombogenic but that actual worms are required for persistent thrombosis. 23 The transient nature of these thrombi confirms the antigenicity of the heartworm but suggests that solid worm fragments are necessary to support prolonged thrombus residence.

Knowledge gap
Further investigation of the contribution of microfilariasis, especially to the thrombogenic risk of heart work disease, is warranted, particularly as the AHS does not currently recommend the use of antithrombotics in the management of dogs with heartworm disease.

PECO QUESTION: HEARTWORM DISEASE (CATS)
In cats (P), is the development of heartworm disease (E), as opposed to remaining disease-free (C), associated with the development of thrombosis (O)?

PECO QUESTION: IMMUNE-MEDIATED HEMOLYTIC ANEMIA (IMHA) (CATS)
In cats (P), is the development of immune-mediated hemolytic anemia (E), as opposed to remaining disease-free (C), associated with the development of thrombosis (O)?
b. There is no evidence that immune-mediated hemolytic anemia is a risk factor for arterial thromboembolism in cats.
c. We suggest that antithrombotic therapy be considered in cats with immune-mediated hemolytic anemia, where other risk factors for thrombosis exist.

Knowledge gap
The mechanism explaining the differences between the thrombogenicity of IMHA in dogs versus cats is unknown.

PECO QUESTION: PROTEIN-LOSING NEPHROPATHY (PLN) (CATS)
In cats (P), is the development of protein-losing nephropathy (E), as opposed to remaining disease-free (C), associated with the development of thrombosis (O)?
b. There is no evidence that protein-losing nephropathy is a risk factor for arterial thromboembolism in cats.
c. We suggest that antithrombotic therapy be considered in cats with protein-losing nephropathy, where other risk factors for thrombosis exist.

Evidence summary
Two studies suggest an association between PLN and PTE (venous thromboembolism) in cats. 30,31 The prevalence of PLN in these 2 reports of feline PTE was 6-14% (4/29 cats and 1/17 cats). By comparison, neoplasia (34-35%) and cardiac disease (6-41%) were more commonly associated with PTE in cats in the same reports. Several of the reports describing histopathologic changes in cats with PLN or glomerulonephritis had evidence of fibrin deposition; however, none of the studies that reported histopathology described any micro-or macrovascular thrombosis. 84,85 Numerous studies were considered neutral to the PECO question because they did not include a control group, they described arterial thrombosis and did not mention PLN, or they were reports focused on PLN that did not discuss or describe thrombosis. 45

PECO QUESTION: LIVER DISEASE (DOGS)
In dogs (P), is the development of hepatic disease (E), as opposed to remaining disease-free (C), associated with the development of thrombosis (O)?

Liver disease (dogs)
a. Liver disease is associated with thrombosis in a small subset of dogs only, independent of the specific underlying disease.
b. We suggest that antithrombotic therapy be considered in dogs with liver disease following an assessment of the risk and benefit in individual patients or where other risk factors for thrombosis exist.

Evidence summary
Of the 32 reports included for review, 13 suggest that liver disease may be associated with the development of thrombosis (1 LOE 5, good; 1 LOE 2, fair; 8 LOE 5, fair; 3 LOE 5, poor).
Hepatic disease might be preferentially associated with thrombosis of certain vessels, particularly the portal and splenic veins. Hepatic disease was documented in 29/140 (21%) dogs with portal or splenic vein thrombosis [111][112][113][114] but not in dogs with pulmonary or aortic thrombosis. [115][116][117][118][119][120] While this suggests an association between hepatic disease and thrombosis, comorbidities, including glucocorticoid administration, neoplasia, kidney disease or recent surgery, were common, and their contribution to overall thrombotic risk in these dogs could not be determined. 111,112,114 Three studies employed thromboelastography (TEG) and documented a hypercoagulable state (but not thrombosis) in dogs with liver diseases, including extrahepatic biliary tract obstruction (10/10 dogs, 100%); 121  In conclusion, because of the retrospective or in vitro nature of most of the studies, the impact of potential comorbidities on the association between liver disease and thrombosis is difficult to ascertain. Overall, the veterinary literature supports an association between liver disease and thrombosis, in particular PVT and splenic vein thrombosis, in a subset of dogs with liver diseases. We suggest that antithrombotic therapy should be considered in dogs with liver disease following an assessment of the risk and benefit in individual patients recognizing that bleeding disorders may be present due to severe liver dysfunction. The presence of other risk factors for thrombosis should also prompt clinicians to consider antithrombotics for dogs with liver disease.

Knowledge gap
Prospective controlled studies are needed to better ascertain the risk of thrombosis in dogs with liver disease of various etiologies and the contribution of comorbidities to the overall prothrombotic risk.

PECO QUESTION: LIVER DISEASE (CATS)
In cats (P), is the development of hepatic disease (E), as opposed to remaining disease free (C), associated with the development of thrombosis (O)?

Liver disease (cats)
No guidelines for this PECO question were generated during the current CURATIVE iteration. This important question will be addressed in a future iteration of the CURATIVE guidelines.

PECO QUESTION: CONGENTIAL PORTOSYSTEMIC SHUNTS (CPSS) (DOGS)
In dogs (P), is the presence of a congenital portosystemic shunt (E), as opposed to remaining disease-free (C), associated with the development of thrombosis? c. We recommend against routine use of antithrombotic therapy in dogs with cPSS.

Evidence summary
Twelve reports met the criteria for review. Four reports supported the PECO question, with 7 considered neutral and 1 in opposition. In a ret- Overall, there is some suggestion in the literature that cPSS increases the risk of thrombosis, particularly following surgical attenuation. These patients may be at risk of bleeding due to liver dysfunction; hence, antithrombotic therapy should be considered only after an assessment of the risk and benefit in individual patients, and routine administration of antithrombotic drugs for dogs with cPSS is not recommended.

PECO QUESTION: CONGENITAL PORTOSYSTEMIC SHUNTS (CPSS) (CATS)
In cats (P), is the presence of a congenital portosystemic shunt (E), as opposed to remaining disease free (C), associated with the development of thrombosis (O)?

Portosystemic shunts (cats)
a. Congenital portosystemic shunts (cPSS) may be associated with thrombosis in cats.
b. We recommend against routine use of antithrombotic therapy in cats with cPSS.
c. We suggest that antithrombotic therapy be considered in cats with cPSS, following an assessment of the risk and benefit in individual patients, when additional risk factors for thrombosis exist.

Evidence summary
Four reports met the criteria for review. Evidence from 1 (LOE 5, fair) documented a possible association of cPSS with PVT in cats. 62 In this case series, 3/6 cats had congenital PSS at the time of PVT identification. One of these cats had recently undergone shunt ligation that may have increased thrombotic risk. In a multicenter retrospective study of 34 cats with cPSS, 11 cats developed complications, and 6 of these died. However, thrombosis was not described in any cat, and hence, this study was judged to oppose the PECO question. 135 Two other studies (LOE 5, fair) identified hemostatic abnormalities in cats with cPSS, but neither described thrombosis. 136,137 These studies were considered neutral to the PECO question.
Overall, given the paucity of data to review, the risk of thrombosis associated with cPSS in cats cannot be clearly determined. As such, we recommend against routine use of antithrombotic therapy in cats with cPSS. Where more detailed assessments of risk versus benefit conducted in individual cats with cPSS are suggestive of thrombotic risk, where other risk factors for thrombosis are present, then we suggest antithrombotic therapy may be considered.

Knowledge gap
Studies are needed to investigate the potential association of thrombosis and PVT in cats prior to shunt ligation to remove the confounding effect of shunt ligation on the overall hemostatic state.

PECO QUESTION: CARDIAC ARRHYTHMIAS (DOGS)
In dogs (P), is the development of cardiac arrhythmias (E), as opposed to c. We recommend against the use of antithrombotic therapy in dogs with arrhythmias other than atrial fibrillation, unless other risk factors for thrombosis exist.

Evidence summary
Several reports of experimental models of rapid pacing-induced atrial fibrillation (AF) in dogs were considered relevant to the PECO question. Two reports comparing the coagulability of blood sampled from the right atrium to that of peripheral blood showed evidence of hypercoagulability in atrial, but not peripheral, blood after induction of AF. 138,139 Another report demonstrated decreased trans-mitral and left atrial appendage flow velocities during AF and following conversion to sinus rhythm compared to baseline measurements; this blood stasis may predispose to thrombosis. 140 Induced AF in dogs also upregulates gene expression for some procoagulant mediators. 141 One experimental study was classified as neutral to the PECO question, as it did not detect significant changes in von Willebrand factor or P-selectin over time after the onset of induced AF, but the sample size was very small, and other coagulation parameters were not assessed. 142 The only study deemed to oppose the PECO question showed no increase in the rate of thrombosis in 3 different canine models of AF following radiofrequency ablation compared to control dogs. 143

Knowledge gap
The contribution of underlying structural cardiac changes to a prothrombotic state in dogs with cardiac arrhythmias remains unclear.
Studies investigating hemostatic changes in dogs with cardiac arrhythmias in the absence of significant underlying cardiac disease and prior to the development of thrombosis are needed.

PECO QUESTION: CARDIAC ARRHYTHMIAS (CATS)
In cats (P), is the development of cardiac arrhythmias (E), as opposed to remaining disease free (C), associated with the development of thrombosis (O)?  (5), AF (4), ventricular tachycardia (3), first-degree AV block (2), and supraventricular tachycardia (1). It is not possible to determine from the report whether the cats with thrombosis had arrhythmias detected. An early case report of ATE in 5 cats identified an arrhythmia consistent with AF in one cat. 156 In summary, arrhythmias in cats with structural cardiac disease are associated with arterial thromboembolism, although it is difficult to determine the isolated contribution of the arrhythmia to thrombotic risk in these cats. Given the strong association between structural cardiac disease, arrhythmias and thromboembolism in cats, we recommend antithrombotic therapy for all affected cats.

Knowledge gap
Further studies are needed to differentiate the contribution of arrhythmias to changes in coagulation status from those of underlying cardiomyopathy, cardiac dysfunction and blood flow alterations.

PECO QUESTION: SEPSIS (CATS)
In cats (P), is the development of sepsis (E), as opposed to remaining disease-free (C), associated with the development of thrombosis (O)?

Sepsis (cats)
a. Sepsis is associated with the development of thrombosis in a small subset of cats.
b. We recommend against routine use of antithrombotic therapy in cats with sepsis.
c. We suggest that antithrombotic therapy be considered for cats with sepsis, following an assessment of the risk and benefit in individual patients or where other risk factors for thrombosis exist.

Evidence summary
Hemostatic abnormalities are commonly identified in cats with sepsis, but thrombosis is infrequently reported. Three retrospective studies including 160 cats with sepsis described just 2 cats (1.3%) with pulmonary thrombosis noted at necropsy. [157][158][159] In aggregate from 2 studies of cats with PTE, 6.5% cats (3/46) had sepsis. 30 Four studies were considered neutral to the PECO question. A study investigating hemostatic changes in cats with sepsis showed changes consistent with coagulation activation, such as reduced protein C activity and increased D-dimers, but thrombosis was not described; hence, the study was deemed neutral to the PECO question. 162 Another report described 10 cats undergoing adrenalectomy for hyperadrenocorticism. Two cats developed septic complications, and 1 experienced a fatal thromboembolic event. However, it is unclear whether thrombotic events occurred in cats with sepsis. 163 Several case reports describe thrombotic complications of sepsis in cats, but multiple potential causes of thrombosis were present. 164,165 Overall, cats with sepsis commonly develop hemostatic abnormalities consistent with the activation of coagulation, but the incidence of thrombosis in cats with sepsis is low. In cats with sepsis that do develop thrombi, PTE is most described. The low overall incidence of thrombosis in cats with sepsis argues against routine use of antithrombotic therapy in these animals. We suggest that antithrombotic therapy be considered for cats with sepsis, particularly where other risk factors for thrombosis exist. Because cats with sepsis can develop DIC and experience clinical bleeding, antithrombotic medication should only be initiated after an assessment of the risk and benefit in individual patients.

Knowledge gap
Studies specifically investigating the development of clinically relevant thrombosis in cats (microvascular and macrovascular) are needed to better understand the risk factors for thrombotic complications in this patient population.

PECO QUESTION: PROTEIN-LOSING ENTEROPATHY (PLE) (DOGS)
In dogs (P), is the development of protein-losing enteropathy (E), as opposed to remaining disease free (C), associated with the development of thrombosis (O)?

Knowledge gap
The differences in coagulation status between dogs with PLE experiencing thrombosis and those that do not develop thrombosis are not clear, as most studies that describe thrombosis in dogs with PLE do not include hemostatic testing. The contribution of corticosteroid therapy to the overall prothrombotic risk is also not known.

PECO QUESTION: PROTEIN-LOSING ENTEROPATHY (PLE) (CATS)
In cats (P), is the development of protein-losing enteropathy (E), as opposed to remaining disease-free (C), associated with the development of thrombosis (O)?

Knowledge gap
Prospective studies with control groups are needed to better ascertain any association between PLE and thrombosis in cats.

PECO QUESTION: HYPERADRENOCORTICISM (CATS)
In cats (P), is the development of hyperadrenocorticism (E), as opposed to remaining disease-free (C), associated with the development of thrombosis (O)? b. We suggest that antithrombotic therapy should not be routinely used in cats with hyperadrenocorticism.

Evidence summary
The literature describing hyperadrenocorticism in cats is scarce, with most studies describing case series or case reports (LOE 5). [251][252][253][254][255][256][257][258][259][260][261][262][263][264][265][266] Some reports included necropsy results or the findings of CT or abdominal ultrasound imaging, but none were focused on identifying thrombosis. Across all studies reviewed, a total of 75 cats were reported to have hyperadrenocorticism, and only 1 cat was diagnosed with PTE, as described in a retrospective study of 7 cats with hyperadrenocorticism. 260 The scant literature suggests that the incidence of hyperadrenocorticism in cats is seemingly low but precludes evidence-based recommendations. As such, we suggest that antithrombotic therapy should not be routinely used in cats with hyperadrenocorticism.

PECO QUESTION: GLUCOCORTICOID ADMINISTRATION (CATS)
In cats (P), is glucocorticoid administration (E), as opposed to no glucocorticoid administration, (C), associated with the development of thrombosis (O)?

PECO QUESTION: INTRAVENOUS CATHETERS (DOGS)
In dogs (P), is the presence of an intravenous catheter (E), as opposed to no intravenous catheter (C), associated with the development of thrombosis (O)?

Intravenous catheters (dogs)
a. The risk of thrombosis associated with IV catheters in dogs is unknown.
b. We suggest that antithrombotic therapy should be considered in dogs with IV catheters only where other risk factors for thrombosis exist.

Evidence summary
A review of the literature yielded one report (LOE 2, fair quality) that was supportive of the PECO question. 268 In that study, 50 dogs with cephalic peripheral catheters placed to enable repeated radiotherapy were monitored daily using vascular ultrasound. Eighteen dogs developed phlebitis, suggesting an incidence of 36%. Ten dogs with phlebitis developed evidence of local thrombosis (58%), suggesting an over-

PECO QUESTION: INTRAVENOUS CATHETERS (CATS)
In cats (P), is the presence of an IV catheter (E), as opposed to no IV catheter (C), associated with the development of thrombosis (O)?

PECO QUESTION: ARTERIAL CATHETERS (DOGS)
In dogs (P), is the presence of an arterial catheter (E), as opposed to no arterial catheter (C), associated with the development of thrombosis (O)?

Arterial catheters (dogs)
a. The risk of thrombosis associated with arterial catheterization in dogs appears to be low.
b. No evidence-based recommendations can be made regarding the use of antithrombotic therapy in dogs with arterial catheters.
c. We suggest that antithrombotic therapy should not be routinely used in dogs with arterial catheters. c. We suggest that antithrombotic therapy should not be routinely used in cats with arterial catheters.

Evidence summary
Only 3 studies (LOE 5) were identified that specifically evaluated arterial catheters in cats and aimed to evaluate complications such as thrombosis. 308,309,312 None of these studies included a control group without an arterial catheter; hence, all 3 were considered neutral to the PECO question. As with dogs, the complication 'catheter occlusion' was not further characterized to determine or differentiate the underlying cause. This lack of confirmation of thrombosis precludes true determination of the risk of thrombosis posed by the use of arterial catheters. In most patients, arterial catheters were placed for monitoring arterial blood pressure during anesthesia with continued use postoperatively. The association between arterial catheters and dwell time was partially explored in these 3 studies. The dwell times were median (mix-max) 12 h (3.5-35); 308 3 h (1-117); 312 and 7.7 h (0.9-42.5, dogs and cats combined). 309 Complication rates were related to dwell time, but the exact nature of the complications was not characterized.
In a study of 13 cats, arterial catherization sites were checked following catheter removal in only 4 cats. 309 In each case, a pulse was detected distally, indicating that the arteries remained patent, but the remaining catheter sites were not evaluated.
In a study that included 29 arterial catheters in cats, 2 catheters (7%) were removed due to a complication. 308 Loss of catheter function where the catheter could not be flushed and failed to aspirate occurred 8 times (28%), while for 4 catheters (14%), a complication of "cold paw" was noted. Thrombosis is a possible cause for these various complications but was not confirmed.
The most likely occurrence of thrombosis associated with an arterial catheter in a cat was noted in the study by Mooshian et al. (LOE 5,fair), in which 1 cat suffered ischemic injury secondary to a coccygeal arterial catheter resulting in tail amputation. 312 Histopathology was not conducted to confirm thrombosis; however, the underlying condition of this cat was not reported, which precludes assessment of the contribution of other risk factors.
One additional study identified arterial thrombosis in a cat associated with an infected arterial catheter. 313 The infection was confirmed by bacterial culture, and thrombosis was confirmed by histopathology following amputation. However, in this case report, the relative con-

PECO QUESTION: VASCULAR ACCESS PORTS (DOGS)
In dogs (P), is the presence of a vascular access port (E), as opposed to no vascular access port (C), associated with the development of throm-

PECO QUESTION: VASCULAR ACCESS PORTS (CATS)
In cats (P), is the presence of a vascular access port (E), as opposed to no vascular access port (C), associated with the development of thrombosis (O)?

Vascular access ports (cats)
a. There is no evidence that the use of vascular access ports in cats is associated with an increased risk of thrombosis.
b. We suggest that antithrombotic therapy should not be routinely used in cats with vascular access ports.

Evidence summary
Six studies describing the use of VAPs in cats were reviewed, the largest of which included 46 cats (and 126 dogs) receiving radiotherapy for various forms of neoplasia. 317

PECO QUESTION: EXTRACORPOREAL CIRCUITS (DOGS)
In dogs (P), is the use of an extracorporeal circuit (E), as opposed to no c. We suggest that antithrombotic therapy should not be routinely used between extracorporeal therapy cycles in dogs unless indicated by other risk factors for thrombosis.

Evidence summary
Extracorporeal circuits, specifically those used for hemodialysis and therapeutic apheresis, are associated with activation of platelets, neutrophil-platelet aggregation, and activation of coagulation. Extracorporeal circulation leads to reduced and turbulent blood flow within the circuit, high shear stress, and blood contact with both air and artificial surfaces (eg, filter membranes, tubing, IV catheter lumens). Platelet activation in response to shear stress leads to platelet-neutrophil aggregate formation, 321 and a study utilizing radiolabeled canine platelets circulated through a dialyzer showed that platelets exposed to the circuit were significantly more thrombogenic than healthy control platelets. 322 Leukocyte activation triggered by contact between blood and extracorporeal circuits contributes to a procoagulant state. 323 In some instances, contact activation by the artificial membrane may also occur. 323 Given these circuit-induced procoagulant effects, anticoagulation (systemic or regional) during extracorporeal therapy is required to avoid clotting within the circuit. We therefore recommend that anticoagulation (systemic or regional) be used during the procedure unless otherwise contraindicated.

PECO QUESTION: EXTRACORPOREAL CIRCUITS (CATS)
In cats (P), is the use of an extracorporeal circuit (E), as opposed to no extracorporeal circuit (C), associated with the development of throm-

Evidence summary
Two studies investigating dialysis catheter performance in cats were reviewed. 330,99 One report assessed patency of a nitric oxide releasing extracorporeal circuit using an in vivo feline model, 330

Knowledge gaps
No studies were identified that directly addressed the PECO question, despite ample evidence supporting the high incidence of thrombosis in people undergoing extracorporeal circulation as opposed to those who do not.

PECO QUESTION: TRANSVENOUS PACEMAKER (DOGS)
In dogs (P), is the presence of a transvenous pacemaker (E), as opposed c. We suggest that antithrombotic therapy be considered in all dogs following transvenous pacemaker implantation.

Evidence summary
Thirty-five references that included ≥1 dog with transvenous pacemaker implantation were reviewed. 244 336 The other studies were focused on pacemaker complications and included large sample sizes (total n = 380), and thrombosis was not reported . 328,[350][351][352] One case series of dogs in which pacemakers were implanted for the management of suspected myocarditis described 74 dogs with prothrombotic comorbidities, including protein-losing disease, immune-mediated disease, the presence of spontaneous echocontrast, hepatic disease, and hyperadrenocorticism. Thrombosis was not reported as a complication, with a median survival time of 1079 days. 359 Overall, the available literature suggests that the presence of a transvenous pacemaker in dogs is associated with the development of thrombosis, with an estimated overall prevalence of ∼5% and an estimated prevalence of symptomatic thrombosis of ∼1%. Cranial vena cava syndrome is the most commonly associated thrombotic complication of transvenous pacemakers in dogs, with outcomes ranging from resolution to death. The risk of thrombosis is not uniform, however, because some dogs with other comorbidities that might predispose to thrombosis did not develop thrombotic complications from their pacemaker. It seems reasonable, however, that all dogs with pacemakers and prothrombotic comorbidities should receive antithrombotic therapies for prophylaxis, while antithrombotic therapy can be considered in every dog following transvenous pacemaker implantation to minimize the risk of symptomatic thrombosis development.

Knowledge gap
Additional studies are needed to better characterize the role of comorbidities in the development of thrombosis in dogs with transvenous pacemakers.

PECO QUESTION: TRANSVENOUS PACEMAKER (CATS)
In cats (P), is the presence of a transvenous pacemaker (E), as opposed to no transvenous pacemaker (C), associated with the development of thrombosis (O)?

Evidence summary
Most studies of cardiac pacemaker implantation in cats describe epicardial lead placement only (total n = 52) and were not further evaluated. Five studies consisting of isolated case reports or small case series (LOE 5, fair) describing cats with implanted transvenous pacemakers were reviewed, 365-369 although 2 studies included cats with epicardial leads. 366,367 Four studies were considered neutral to the worksheet question since thrombosis was not described, while 1 was considered neutral because chylothorax developed as a complication, and thrombosis could not be ruled out as the cause. 369 The case report that was considered neutral to the PECO question described a cat with a three-week history of syncope due to third-degree atrioventricular block. 369 A permanent transvenous pacemaker was placed in the left jugular vein, and the cat was asymptomatic for 3 months following implantation, when chylothorax was detected that required repeated thoracocentesis and ultimately prompted euthanasia. The cause of chylothorax was not determined, but thrombosis was a potential explanation. Overall, since most literature describes epicardial lead placement and there is no definitive evidence of thrombosis in cats following transvenous pacemaker placement, we were unable to make evidence-based recommendations on the use of antithrombotic medications for these cats.