Bleeding risk with systemic thrombolytic therapy for pulmonary embolism: scope of the problem

Abstract

Acute pulmonary embolism represents a major complication of venous thromboembolism that is associated with high morbidity and mortality. Guidelines recommend the rapid initiation of anticoagulation and consideration of thrombolytic therapy in select patients, including those with hypotension or at high risk of developing hypotension. Evaluation for thrombolytic therapy should only be considered after assessment of contraindications and risk for major bleeding. The objective of this perspective article is to evaluate the bleeding risk associated with systemic thrombolytic therapy in the management of acute pulmonary embolism and discuss strategies to minimize this risk. Risk stratification of acute pulmonary embolism will be discussed to identify patient populations that warrant specific consideration of risk for major bleeding with thrombolytic therapy. In addition, the incidence, patient-specific risk factors, and pharmacologic characteristics, including concurrent anticoagulation and thrombolytic therapy, will be evaluated in the context of risk for major hemorrhage. Finally, supporting evidence for strategies to minimize risk of hemorrhage, including evaluation of contraindications, weight adjusted dosing, infusion strategy and catheter-directed thrombolytic administration will be evaluated. Despite published guidelines and review articles, select aspects to thrombolytic therapy for the management of pulmonary embolism remain controversial and under recognized, including risk of major hemorrhage. When making decisions about the role of thrombolytic therapy in pulmonary embolism, clinicians must be knowledgeable about the associated risks of thrombolytic therapy and individually evaluate patient risk factors prior to determining appropriate candidacy for thrombolytic therapy. For patients considered to be at high risk of major bleeding, strategies to minimize risk should be considered, including weight-adjusted doses and catheter directed therapy. Additional research is needed specific to the acute pulmonary embolism setting to validate risk factors and strategies to minimize major hemorrhage.

Introduction

Acute pulmonary embolism (PE) represents a major complication of venous thromboembolism (VTE) that is associated with high morbidity and mortality [Tapson, 2008]. Approximately 25% of patients with acute PE will never receive prompt treatment, as the initial clinical presentation is sudden death [Go et al. 2013]. Otherwise, the clinical presentation of PE may range from mild dyspnea to shock [Agnelli and Becattini, 2010]. In 2006, approximately 247,000 people were admitted to a United States hospital for a diagnosis of acute PE and approximately 22,000 expired secondary to a PE [Stein and Matta, 2010; Tsai et al. 2012]. Thirty-day survival following the diagnosis of a PE is only estimated to be 59.1% and despite advances in medical care, in-hospital mortality from acute PE has remained stable over the past decade [Stein and Matta, 2010; Heit et al. 1999]. Long-term complications of PE include venous stasis syndrome, venous ulcer and chronic thromboembolic pulmonary hypertension [Go et al. 2013].

Therefore, guidelines recommend the rapid initiation of anticoagulation (unfractionated heparin, low-molecular-weight heparin or fondaparinux) and consideration of thrombolytic therapy in select patients, including those with hypotension or at high risk of developing hypotension [Kearon et al. 2012; Konstantinides et al. 2014]. Yet, this recommendation is based on limited and controversial data. Evaluation for thrombolytic therapy should only be considered after assessment of contraindications and risk for major bleeding.

The objective of this article is to evaluate the bleeding risk associated with systemic thrombolytic therapy in the management of acute PE and discuss strategies to minimize this risk. Therefore, we performed a PubMed search of English-language literature from January 1990 through October 2014 using medical subject headings: pulmonary embolism, bleeding, thrombolytic therapy or fibrinolytic therapy, alteplase, streptokinase and urokinase as well as general search terms thrombolytics, fibrinolytics, pulmonary embolism and all combinations. All original prospective and retrospective studies, peer-reviewed guidelines, consensus statements, review articles and accompanying references were evaluated for inclusion. Relevance was determined considering the type of manuscript (excluding case series studies), sample size, evaluation of clinical outcomes and consideration safety of systemic thrombolytic therapy in the management of PE.

Assessment and treatment of PE

Outcomes associated with acute PE are variable; therefore, a risk stratification system has been developed to tailor medical and interventional therapies. Acute PE can be classified into three subgroups, including: massive or high risk, submassive or intermediate risk and stable or low risk [Jaff et al. 2011]. Massive PE is defined by hemodynamic instability (systolic blood pressure [SBP] less than 90 mmHg, fall from baseline greater than 40 mmHg, or cardiac arrest) and symptom manifestation is related to hypotension [Goldhaber, 2006]. Submassive PE is described when hemodynamically stable but evidence of right ventricular (RV) dysfunction (Table 1) [Piazza, 2013]. Patients with stable PE are hemodynamically stable and have no evidence of RV dysfunction. The presence of RV dysfunction significantly increases the risk of complications, including early mortality, compared with stable PE [Wolde et al. 2004].

Table 1.

Signs of right ventricular dysfunction.

Source	Definition
Electrocardiogram	Incomplete or complete right bundle-branch block
	T-wave inversions in leads V1-V4
	S wave lead I, Q wave in lead III, T-wave inversion lead III
Echocardiogram	Right ventricular hypokinesis and dilation
	Tricuspid regurgitation
	Pulmonary embolism
Chest computed tomography	Right ventricular enlargement (right ventricular diameter – left ventricular diameter ratio > 0.9)
Laboratory	Elevated cardiac troponins
	Elevated brain-type natriuretic peptide

Therefore, consideration of thrombolytic therapy and concern for hemorrhagic risk is warranted for patients classified with either a massive or submassive PE. These subgroups account for 5% and 20–25% of patients with acute PE, respectively [Piazza, 2013]. Thrombolytic therapy restores pulmonary perfusion more rapidly than anticoagulation alone. In the absence of hemodynamic instability, the clinical benefits of thrombolytic therapy have long been controversial. The likely benefit of thrombolysis compared with placebo in submassive PE is a reduced incidence of escalation of therapy (10.2% versus 24.6%; p = 0.004) or rate of hemodynamic collapse (1.6% versus 5%; p = 0.002), with no significant differences observed in mortality [Konstantanides et al. 2002; Meyer et al. 2014]. Guidelines and review articles are available elsewhere that discuss the efficacy of thrombolytic therapy in massive and submassive PE [Jaff et al. 2011; Kearon et al. 2012; Konstantanides et al. 2014; Daley and Lat, 2012]. The remainder of this perspective will focus on identifying patients at risk of major bleeding and strategies to minimize this risk.

Risk of major hemorrhage

Incidence

Determining the actual risk of major hemorrhage from published literature is complicated by several factors. Individual studies us several thrombolytic agents and dosing regimens which may have different risks. Sample sizes are often too small to assess the risk of a particular regimen. Definitions of major bleeding also vary among studies. Bleeding complications in early studies frequently involved vascular access sites for pulmonary angiography and need for blood transfusion. These rates may be lower in contemporary practice with noninvasive PE diagnostic testing.

The most concerning adverse effects of thrombolytic agents are major bleeding and intracranial hemorrhage; one registry of patients in Poland found that major bleeding was a significant predictor of in-hospital and 90-day mortality [Budaj-Fidecka et al. 2013]. As shown in Table 2, prior prospective trials with various thrombolytics for PE reported an incidence of major bleeding between 0% and 33% [Levine et al. 1990; Goldhaber et al. 1994; UPET, 1970] and intracranial hemorrhage between 0% and 7.4% [Levine et al. 1990; Goldhaber et al. 1994; UPET, 1970]; the wide range is due in part to the small sample sizes of many of these studies. A meta-analysis that included 16 studies and a total of 2115 patients found a 9.24% incidence of major bleeding and a significant increase in risk compared with treatment with anticoagulant alone (odds ratio [OR] 2.73; 95% confidence interval [CI] 1.91–3.91) [Chatterjee et al. 2014]. The incidence of intracranial bleeding was 1.46% and was also significantly increased compared with anticoagulant therapy (OR 4.63; 95% CI 1.78–12.04) [Chatterjee et al. 2014] and more conservative use of blood transfusions [DuPont-Thibodeau et al. 2014]

Table 2.

Major bleeding and intracranial bleeding in prospective trials.

Treatment	n	Major bleedingrate (%)	ICH rate (%)	Reference
Urokinase 50,000 units/lb	82	32.9	NR	[UPET 1970]
t-PA 0.6 mg/kg	33	0	0	[Levine et al. 1990]
t-PA 40–80 mg	9	11.1	0	[PIOPED Investigators, 1990]
r-PA 20 units	23	4.3	0	[Tebbe et al. 1999]
t-PA 100 mg	13	7.7	0	[Tebbe et al. 1999]
t-PA 100mg	20	15	NR	[Dalla-Volta et al. 1992]
Urokinase 3 million units	45	10.9	2.2	[Goldhaber et al. 1992]
t-PA 100 mg	44	15.9	4.5	[Goldhaber et al. 1992]
Urokinase 57,200 units/kg	29	27.6	3.4	[Meyer et al. 1992]
t-PA 100 mg	34	20.6	0	[Meyer et al. 1992]
t-PA 100 mg	46	NR	2.2	[Goldhaber et al. 1993]
t-PA 0.6 mg/kg	60	3.3	0	[Goldhaber et al. 1994]
t-PA 100 mg	27	7.4	7.4	[Goldhaber et al. 1994]
t-PA 0.6 mg/kg	36	8	0	[Sors et al. 1994]
t-PA 100 mg	17	6	0	[Sors et al.1994]
STK 1.45 million units	25	12	NR	[Meneveau et al. 1997]
t-PA 100 mg	25	16	NR	[Meneveau et al. 1997]
STK 1.5 million units	43	8	NR	[Meneveau et al. 1998]
t-PA 100 mg	23	20	NR	[Meneveau et al. 1998]
t-PA 100 mg	118	0.8	0	[Konstantides et al. 2002]
t-PA 100 mg	7	0	0	[Muhl et al. 2007]
STK 9 million units	8	0	0	[Muhl et al. 2007]
TNK 30–50 mg	525	NR	2.7	[Bottiger et al. 2008]
TNK 30–50 mg	28	7.1	3.6	[Becattini et al. 2010]
t-PA 50 mg	55	3.6	1.8	[Wang et al. 2010]
t-PA 100 mg	48	10.4	0	[Wang et al. 2010]
t-PA 100 mg	37	5.4	0	[Fasullo et al. 2011]
t-PA 50 mg	61	0	0	[Sharifi et al. 2013]
STK 2.65–5.05 million units	75	1.3	1.3	[Patra et al. 2013]
TNK 30–50 mg	25	0	0	[Patra et al. 2013]
TNK 30–50 mg	506	11.5	2.0	[Meyer et al. 2014]
STK 2.65–5.05 million units	105	1.9	NR	[Patra et al. 2014]
TNK 30–50 mg	25	0	NR	[Patra et al. 2014]
TNK 30–50 mg	40	2.5	2.5	[Kline et al. 2014]
t-PA 10–20 mg	30	0	0	[Kucher et al. 2014]

t-PA, alteplase; TNK, tenecteplase; STK, streptokinase; r-PA, reteplase; ICH, intracranial hemorrhage; NR, not reported; ICH, intracranial hemorrhage.

Recently, a multicenter randomized controlled trial with tenecteplase compared with placebo found increased rates of major bleeding (11.5% versus 2.4%) and intracranial hemorrhage (ICH) (2% versus 0.2%), correlating with a number-needed-to-harm of 11 and 55, respectively [Meyer et al. 2014]. However, the bleeding rates seen in prospective trials may also underestimate the risk due to strict inclusion and exclusion criteria; patients treated in real-world settings may receive thrombolytics despite potential contraindications [Kasper et al. 1999]. One international registry also examined bleeding related events in 2454 patients treated for PE [Goldhaber et al. 1999]. This registry also found an increase in major bleeding (21.7% versus 8.8%) and intracranial bleeding (3.0% versus 0.3%) compared to those who did not receive thrombolytics. A Polish registry found a similar rate of major bleeding in 19% of patients who received thrombolytics [Budaj-Fidecka et al. 2013].

Risk factors

Identifying risks for bleeding could help determine whether individuals should receive thrombolytic therapy (Box 1). Unfortunately, few studies exist analyzing bleeding risk with thrombolytics for PE. One early single-center study of PE use for massive PE found the only significant risk factor for major bleeding was femoral vein access for pulmonary angiography [Meyer et al. 1998], which is not helpful in current practice. Another, more recent, study [Curtis et al. 2014] identified risks for major bleeding in patients who received treatment for PE with alteplase 100 mg at a single institution. Significant risk factors in this analysis included major surgery within the prior 3 weeks (OR 9.00, 95% CI 1.01–79.99), international normalized ratio (INR) > 1.7 (OR 13.20, 95% CI 1.54–113.52), weight < 100 kg (OR 1.18 for each 10 kg below 100 kg, 95% CI 1.01–1.37), and at least one of the characteristics (OR 5.02, 95% CI 1.78–18.55): internal bleeding in previous four weeks, hypertension, acute myocardial infarction, stool occult positive, presence of intra-aortic balloon pump, African-American race, gastrointestinal bleeding in prior 3 months, aortic dissection, acute pancreatitis, cardiopulmonary resuscitation exceeding 10 minutes, bilirubin > 3 mg/dL, or dementia. The wide confidence intervals and single-center nature of this study likely warrant validation of these risk factors, but this information does offer some benefit in assessing risk. Within the environment of a controlled trial, advanced age greater than 75 years (OR 2.8; 95% CI 1–7.86) and female gender (OR 11.49; 95% CI 2.67–49.53) were independently associated with rates of major extracranial bleeding, with rates of 11.1% and 8%, respectively [Meyer et al. 2014].

Box 1. Known risk factors for major bleeding following thrombolytic therapy for acute pulmonary embolism.

• Demographic characteristics

  Advanced age (>75 years)

  Female gender

  African American race

  Low body weight (risk inversely related with each 10 kg below 100 kg)

• Medical history

  Acute myocardial infarction

  Hypertension (poorly controlled at baseline)

  Aortic dissection

  Acute pancreatitis

  Dementia

  Cardiopulmonary resuscitation exceeding 10 minutes

• Surgical history

  Major surgery within prior 3 weeks

• Bleeding history

  Stool occult positive

  Internal bleeding in previous 4 weeks

  Gastrointestinal bleeding in prior 3 months

• Laboratory

  Elevated bilirubin (> 3 mg/dl)

  Coagulopathy (defined as INR >1.7)

• Invasive device

  Presence of intra-aortic balloon pump

  Femoral venous access

Previous studies in patients who received thrombolytics for myocardial infarction also identified older age, female sex, African-American race, and lower body weight as risk factors for major bleeding [Berkowitz et al. 1997]. These risks may warrant further validation specifically assessed in PE patients but still warrant consideration as the absence of evidence is not the evidence of absence.

Concurrent anticoagulation

Uncertainty exists over which anticoagulant to administer to patients who receive thrombolytics and whether to administer concomitantly. Parenteral anticoagulant options include intermittent subcutaneous fondaparinux and low-molecular-weight heparin or continuous infusion unfractionated heparin. Whether bleeding risk varies with these different agents in PE is unknown. In patients with myocardial infarction, the use of enoxaparin with thrombolytics increased major bleeding compared with unfractionated heparin (1.4% versus 1.0%, p = 0.004) [Antman et al. 2006]. Although no specific studies exist, European guidelines recommend administering unfractionated heparin with t-PA but not with streptokinase or urokinase while the CHEST guidelines state it is acceptable to either withhold or continue anticoagulation therapy [Konstantinides et al. 2014; Kearon et al. 2012]. Owing to the possibility of bleeding during thrombolytic and anticoagulant administration, unfractionated heparin is specifically recommended because of its shorter half-life and reversibility. These guidelines also recommend continuing unfractionated heparin for several hours after thrombolytic administration while monitoring for bleeding before changing to another anticoagulant.

Thrombolytic agents

The primary mechanism of all thrombolytics is the conversion of plasminogen to the active form, plasmin, which then degrades fibrin. This proteolysis can occur with fibrin-bound plasminogen on the surface of thrombi and the unbound form within the plasma. The unbound plasmin generated degrades fibrin but also fibrinogen, factor V, and factor VIII [Haire, 1992]. Thrombolytic agents such as t-PA, tenecteplase, reteplase, and desmoteplase are referred to as fibrin-specific lytics because of a higher affinity for fibrin-bound plasminogen compared with free plasminogen. Theoretically, higher fibrin specificity may reduce bleeding complications due to direct action at desired site of action and less depletion of circulating procoagulant factors. No adequately powered trials have been conducted in PE patients comparing bleeding differences with fibrin-specific versus nonspecific agents. However, in myocardial infarction patients, major bleeding was lower with fibrin-specific lytics than with streptokinase [Giraldez et al. 2007; GUSTO Investigators, 1993]. Thus increased fibrin-specificity remains a potential means to reduce bleeding without affecting efficacy.

Strategies to minimize risk

The decision to administer fibrinolytic therapy requires practitioners to be knowledgeable about areas with limited or conflicting evidence. Treatment approaches are especially challenging due to the variability of available literature in regards to risk factors, dosing regimens and administration techniques. Risk minimization strategies should be employed whenever possible.

Contraindications

The American College of Chest Physicians (ACCP) recommends individualized assessment of benefits and risk when fibrinolytic therapy is being considered for PE [Kearon et al. 2012]. Potential benefits of thrombolysis include hemodynamic stabilization, rapid symptom resolution and decreased complications from PE. Conversely, the risks consist of major bleeding, including intracranial hemorrhage, minor bleeding and prolonged hospital length of stay. Given these risks, patients should be actively screened for contraindications prior to fibrinolysis. Absolute contraindications include history of intracranial hemorrhage, known intracranial neoplasm, arteriovenous malformation or aneurysm, significant head trauma, active internal bleeding, known bleeding diathesis, intracranial or intraspinal surgery within 3 months and cerebrovascular accident within 2 months. Relative contraindications include age greater than 75, pregnancy, current use of anticoagulation, prolonged cardiopulmonary resuscitation, recent bleeding, history or current severe uncontrolled hypertension, remote ischemic stroke (3 months), and/or major surgery within 3 weeks [Kearon et al. 2012]. It should be noted that due to the heterogeneous nature of PE studies, several of the absolute and relative contraindications are derived from ST-elevated myocardial infarction guidelines. Also, in the setting of high-risk PE, absolute contraindications should be interpreted as relative contraindications [Konstantinides et al. 2014]. Ultimately, the decision to proceed with therapy lies in the hands of the provider based on individual patient risk factors.

Dosing considerations

The Food and Drug Administration label approved dose for alteplase in the setting of PE is 100 mg infused over 2 hours [Genentec, 2005]. National guidelines also reference this dosing strategy as the standard of care for patients with PE that qualify for fibrinolysis [Jaff et al. 2011; Kearon et al. 2012; Konstantinides et al. 2014]. This dosing strategy has been shown to be effective but has also demonstrated a high rate of bleeding complications. In addition, fibrinolytics carry a significant dose-dependent risk of bleeding [Wang et al. 2010]. Although fixed dosing is standard for the PE indication, weight-based dosing is utilized in the setting of myocardial infarction and ischemic stroke. An ideal fibrinolytic dose should maximize fibrinolysis and minimize bleeding complications.

There is a lack of strong evidence in regards to weight-adjusted dosing strategies in the setting of PE. One prospective, randomized, multicenter trial from China evaluated 118 patients with acute PE with evidence of hemodynamic compromise or anatomically massive pulmonary embolism with right ventricular dysfunction [Wang et al. 2010]. Patients were randomized to low-dose alteplase at 50 mg/2 hours (n = 65) or 100 mg/2 hours (n = 53) followed by low-molecular-weight heparin. Patients in each arm were stratified by low (<65 kg), medium (65–74 kg) and high (>74 kg) weight groups. Patients were excluded if they received anticoagulation within 72 hours or had absolute contraindications to fibrinolysis. Efficacy was assessed based on improvements of pulmonary artery pressure and right ventricular function on echocardiography, lung perfusion on ventilator-perfusion scan, and pulmonary artery obstruction on computed tomographic pulmonary angiography. The 50 mg alteplase regimen was similar in efficacy to the 100 mg regimen as assessed by right ventricular dysfunction on qualitative scoring systems. The low dose alteplase resulted in less major bleeding (3% versus 10%), especially in low-body-weight patients <65 kg (14.8% versus 41.2%, p = 0.049; relative risk [RR] 0.19, 95% CI 0.04– 0.092). Mean baseline weight was similar amongst groups but notably lower than the mean weight in multicenter studies (82.5 ± 17.9 kg versus 71.9 ± 12.6 kg) [Meyer et al. 2014]. Analysis on patients greater than 100 kg or BMIs above 30 kg/m² was not evaluated due to sample size. This study substantiates the need for further research on safe and effective dosing strategies in PE.

Administration techniques

Early studies in myocardial infarction, pulmonary embolism and venous thrombosis administered alteplase via continuous infusion and were associated with significant bleeding [Meyer et al. 1998; Daley and Lat, 2012]. Continuous infusion dosing strategy was employed to account for short half-life of 4-5 minutes and therefore maintain adequate concentrations at the thrombus site. Despite the pharmacodynamics of alteplase, several studies suggest continued fibrinolysis after the medication is cleared from circulation [Jaff et al. 2011; Levine et al. 1990]. Although promising, bolus dosing with alteplase for PE requires validation.

Guidelines do not provide clear and consistent viewpoints on optimal administration techniques [Jaff et al. 2011; Kearon et al. 2012; Konstantinides et al. 2014]. The British Thoracic Society suggests utilizing 50 mg bolus of intravenous alteplase when cardiac arrest secondary to massive PE occurs or is strongly suspected [British Thoracic Society Standards of Care Committee Pulmonary Embolism Guideline Development Group, 2003]. ACCP guidelines recommend 2 hour infusion, however bolus therapy may indicated in actual or imminent cardiac arrest [Kearon et al. 2012]. Only two of the five existing bolus dose alteplase trials are outcomes driven. A meta-analysis determined that alteplase infusion was more effective than alteplase bolus dosing (RR 1.27, 95% CI 1.09–1.47) and lower mortality rates (RR 0.16, 95%, CI 0.05–0.59) and decreased bleeding risk (RR 0.62, 95% CI 0.43–0.96) [Capstick and Henry, 2005]. Overall, this was a crude analysis based on limited data making it difficult to apply to real-world scenarios but bolus dosing may be preferred when cardiac arrest is imminent.

Reperfusion strategies

Reperfusion therapy for PE is not limited to systemic thrombolysis. The extensive list of absolute and relative contraindications to systemic fibrinolysis rules out up to two-thirds of patients presenting with massive PE [Engelberger and Kucher, 2011]. The American Heart Association therefore references catheter-based reperfusion treatment and surgical embolectomy as additional treatment strategies in the setting of PE [Jaff et al. 2011]. Catheter-based interventions may be considered if systemic thrombolysis is contraindicated or urgent recanulization of PE is required for patient stabilization and surgical embolectomy is unavailable. Patients who do not have absolute contraindications to thrombolytic therapy may qualify for conventional catheter-directed thrombolysis (CDT) [Piazza and Goldhaber, 2006]. Additional treatment modalities such as high-frequency ultrasound, fragmentation and aspiration my complement CDT.

Conventional catheter-directed thrombolysis is oftentimes utilized in patients with relative contraindications to thrombolysis secondary to the lower rate of bleeding complications [Engelberger and Kucher, 2011]. Assessing the efficacy of CDT is difficult due to the lack of existing large clinical trials and the low likelihood of future trials comparing catheter therapy to thrombolytic therapy or surgical embolectomy [Engelberger and Kucher, 2011; Jaff et al. 2011]. Clinical success rate defined as survival, hemodynamic stability or hypoxia resolution was 86% from pooled study data of 594 PE patients who underwent catheter intervention with or without thrombolysis. Overall success appeared higher in patients who received CDT compared with catheter intervention without thrombolysis (91% versus 83%) [Kuo et al. 2009]. Goals of CDT are to accelerate clot lysis and cause rapid reperfusion of the pulmonary arteries. The catheter must be positioned within the embolus and is initiated with bolus of a fibrinolytic agent followed by continuous infusion. Various dosing regimens for this reperfusion strategy have been reported including urokinase 250,000 IU/h over 2 hours, followed by 100,000 IU/h for 12–24 hours; alteplase 10 mg bolus followed by 20 mg/h over 2 hours or 100 mg over 7 hours [Engelberger and Kucher, 2011].

Conclusion

Major hemorrhage following thrombolytic therapy for acute PE is a common complication that warrants specific evaluation of patient risk factors prior to determining appropriate candidacy for thrombolytic therapy. For patients considered to be at high risk of major bleeding, strategies to minimize risk should be considered, including weight-adjusted doses and catheter directed therapy. Additional research is needed specific to the acute PE setting to validate risk factors and strategies to minimize risk.