Thrombolysis in high-risk patients with acute pulmonary embolism: underuse of a life-saving treatment in the real-world setting

This editorial refers to ‘Trends in thrombolytic treatment and outcomes of acute pulmonary embolism in Germany’†, by K. Keller et al., on page 522. Venous thrombo-embolism (VTE) is the third most common cause of death among thrombosis-related conditions (behind ischaemic heart disease and ischaemic stroke),1 and acute pulmonary embolism (PE) is the main contributor to morbidity and mortality in the majority of VTE cases. In the European Union, it has been estimated that the annual number of PE-related deaths may exceed 500 000 in the population, confirming the epidemiological relevance and the severity of this condition.2 Systemic thrombolytic therapy for acute PE has been used for almost 50 years and, in a meta-analysis of 15 randomized clinical trials (RCTs) involving a total of 2057 patients, thrombolysis reduced the overall mortality and the combined endpoint of death or treatment escalation.3 However, the decrease in overall mortality was not significant in haemodynamically stable patients with acute PE and an increased risk of major haemorrhage, and fatal or intracranial bleeding was observed in the overall population.3 Similar findings were observed in the PEITHO study that tested in an RCT setting the efficacy and safety of systemic thrombolysis in 1005 intermediate risk PE patients, reporting a reduction of the composite outcome endpoint of death or haemodynamic decompensation in the treated group but at the price of a 2.4% incidence of stroke.4 , 5 Based on the benefit to risk ratio, the 2014 ESC PE practice guidelines recommend to treat with systemic thrombolysis only high-risk PE patients with shock or hypotension, defined as systolic blood pressure <90 mmHg, or a systolic pressure drop by ≥40 mmHg, for >15 min if not caused by new-onset arrhythmia, hypovolaemia, or sepsis (Figure 1).6 Interestingly, the incidence of these haemodynamically unstable subjects is variable, with different case series ranging from 3.4% to 11.6% of the overall PE populations due to the heterogeneity of the recruitment characteristics of the studies and the definition of haemodynamic changes.7–9 The grade of recommendation and the level of evidence for systemic thrombolysis (IB) is the highest among the reperfusion methods recommended for high-risk PE patients, including surgical pulmonary embolectomy (IC) and percutaneous catheter-directed treatment (IIaB) (Figure 1).6 This reflects the relatively large number of RCTs performed in PE patients with systemic thrombolysis as compared with the other reperfusion methods. Acute pulmonary embolism management strategy.6 , 15 The management of PE patients starts with the suspicion of PE and the initial risk stratification based on clinical and haemodynamic conditions. Suspected high-risk and not high-risk PE patients will be assessed with specific diagnostic algorithms to confirm the PE diagnosis. Reperfusion treatment is indicated for confirmed high-risk PE patients including systemic thrombolysis, surgical embolectomy or percutaneous catheter treatment in addition to concomitant haemodynamic support and oxygen therapy if needed. Confirmed not high-risk patients require further risk stratification based on clinical scores (e.g. pulmonary embolism severity index; PESI), signs of right ventricular dysfunction (at echocardiography or CTPA), and cardiac laboratory biomarkers (e.g. troponin plasma levels). Intermediate- to high-risk PE patients will require anticoagulant treatment, monitoring of vital signs, and rescue reperfusion therapy in the case of progression to haemodynamic instability. Intermediate- to low-risk PE patients will require anticoagulant therapy and regular hospital stay. Low-risk PE patients are treated with anticoagulant therapy and may be candidates for early hospital discharge and home treatment in the absence of right ventricular dysfunction and increased troponin levels. A/C, anticoagulant therapy; CTPA, computed tomography pulmonary angiography; PE, pulmonary embolism. 1Initial risk stratification based on clinical and haemodynamic conditions. 2Shock or hypotension defined as systolic blood pressure <90 mmHg, or a systolic pressure drop by ≥40 mmHg, for >15 min if not caused by new-onset arrhythmia, hypovolaemia, or sepsis. 3Based on transthoracic and transoesophageal echocardiography, venous compression ultrasound, and, if possible, CTPA. 4Based on clinical probability of pulmonary embolism, D-dimer assessment, CTPA, and, in specific cases, ventilation and/or perfusion lung scan. 5Further risk stratification based on clinical scores (e.g. PESI), signs of right ventricular disfunction (at echocardiography or CTPA), and cardiac laboratory biomarkers (e.g. troponin plasma levels). 6 Systemic thrombolysis is recommended (IB); surgical embolectomy (IC), and percutaneous catheter treatment (IIaC) are considered in the case of contraindication to thrombolysis according with the pulmonary embolism guidelines.5 A/C therapy and concomitant haemodynamic support and oxygen therapy as needed. 7In the case of clinical and haemodynamic deterioration on A/C therapy. 8Early hospital discharge and home treatment in the absence of right ventricular dysfunction and increased troponin levels. In this issue of the European Heart Journal, Keller et al. publish a study analysing the characteristics, comorbidities, treatment, and outcome of 885 806 PE patients in Germany between 2005 and 2015.10 The primary objective of the study was the analysis of the use of systemic thrombolysis for hospitalized patients with acute PE in Germany and its possible impact on early prognosis. In addition, the authors assessed trends in PE incidence rates, patients’ haemodynamic status, treatment, and outcomes during the 11 years of the study, and investigated predictors of in-hospital mortality. The database of the Federal Statistical Office of Germany (Statistisches Bundesamt) collecting nationwide treatment data from all inpatient cases was used. Diagnoses were encoded according to the International Classification of Diseases, 10th Revision with German Modification (ICD-10-GM) and diagnostic, surgical, or interventional procedures with OPS codes (Operationen- und Prozedurenschlüssel, German Procedure Classification). Hospitalized patients diagnosed with PE (ICD code I26) between 2005 and 2015 were included in the analysis. Haemodynamically stable PE patients were defined by the absence of cardiopulmonary resuscitation (CPR; OPS code 8-77) or shock (ICD code R57). Conversely, haemodynamically unstable patients were defined as those who needed CPR and/or presented with shock. The outcomes used for the analysis were all-cause in-hospital death and intracerebral bleeding (ICD code I61). The study was supported by the German Federal Ministry of Education and Research. PE incidence rates and the total annual number of patients diagnosed with PE increased over the 11-year period. This finding confirms previous data11–13 and may be related to a higher number of patients at PE risk in ageing populations and in neoplastic patients with extended survival on modern treatments, as well as to an improvement of imaging techniques, such as computed tomographic pulmonary angiography. However, minimal changes in mortality, and lower case fatality in some observational series may suggest overdiagnosis (e.g. subsegmental uncertain obstructions) and overtreatment as possible additional cause for the increased hospitalization rate.12 Interestingly, the in-hospital mortality rate in the study of Keller et al. 10 decreased from 20.4% in 2005 to 13.9% in 2015, and the duration of hospital stay decreased significantly from 12 days in 2005 to 8 days in 2015. The decrease in mortality and duration of hospital stay may be attributed alternatively to improvements in the overall treatment of PE patients, but also to a higher number of patients at low risk diagnosed thanks to the improvement of the imaging techniques. An additional cause for the reduction of the hospital stay may be related to a progressive wider use of non-vitamin K-dependent new oral anticoagulants (NOACs) which allow a more rapid achievement of a treatment effect, facilitating patients discharge.6 , 14 , 15 The primary endpoint of the study, which was the use of systemic thrombolysis, slightly increased from 3.1% in 2005 to 4.4% in 2015, whereas the rates of surgical embolectomy remained low and largely unchanged (between 0.16% and 0.15% for a total of 1394 patients) and only 450 (0.05%) patients were treated with percutaneous catheter interventions. However, the proportion of haemodynamically unstable high-risk patients in this study was 8.9%, which is intermediate as compared with previous literature data.6–8 Therefore, approximately only 50% of high-risk patients requiring systemic thrombolysis, according to the 2014 ESC guidelines (Figure 1),1 , 65 have received this potential life-saving therapy. Similar trends have already been detected in the RIETE registry (reperfusion therapy used in 20% of haemodynamically unstable PE patients)8 and in the United States Nationwide Inpatients Sample study (thrombolytic therapy used in 30% of haemodynamically unstable PE patients).1 , 93 The underuse of systemic thrombolysis in potentially eligible haemodynamically unstable high-risk PE patients is confirmed in different areas of the western world and needs to be appropriately addressed. Possible causes include the increasing age and frailty of hospitalized PE patients, concerns about intracranial haemorrhage rate, and reduced overall use and experience of thrombolysis therapy in the era of percutaneous coronary interventions for ST-elevation acute myocardial infarction patients. On the other hand, an important finding of the study of Keller et al. is the reduction of the mortality rate in haemodynamically unstable PE patients (excluding those with CPR and/or mechanical ventilation) who received thrombolysis compared with those without this reperfusion treatment (28.6% vs. 49.9%; P < 0.001). The favourable association between thrombolysis and survival was preserved, although less pronounced, in patients needing CPR. Similar trends of reduction in mortality of high-risk patients treated with thrombolysis as compared with untreated patients was also observed in the RIETE registry (6.3% vs. 15%)8 and in the in the United States Nationwide Inpatients Sample study (8.4% vs. 42%).9 Despite all limitations of retrospective analyses and using non-randomized control group comparisons in large-scale real-life assessments, the concordant trends of reduction of mortality in high-risk PE patients treated with systemic thrombolysis represents a relevant confirmation of favourable RCT meta-analysis results.3 Interestingly, in the haemodynamically unstable patients of the study of Keller et al. 10 who underwent systemic thrombolysis, intracerebral bleeding was diagnosed in 1.5% compared with 0.7% in those who did not receive thrombolysis (P = 0.004). This compares well with a similar rate of haemorrhagic complications reported in RCT experiences3 , 4 and suggests a careful selection of patients for thrombolysis in the real-life setting. Despite the inclusion in PE practice guidelines already for several decades,6 alternative reperfusion methods in high-risk patients, such as surgical embolectomy or percutaneous catheter treatments, are performed in a very small fraction of patients (1.8% and 0.6% of the high-risk PE population, respectively) in the study of Keller et al. 10 Almost identical figures are also detected in other large-scale surveys performed in different periods in Europe and the USA,1 , 8 , 9 6 testifying to the minimal impact of these alternative reperfusion methods in the overall high-risk PE patient population. It is therefore mandatory to promote PE practice guideline implementation, in particular to increase the use of all reperfusion methods in haemodynamically unstable high-risk populations. The limitations of the study of Keller et al. 10 applies to all ‘administrative’ databases which include reduced clinical information and possible errors in coding of diseases or procedures. In addition, data are collected at patient discharge, the timing of events is missing, and vital signs are available as categorical definitions (e.g. tachycardia and shock) and not with continuous numbers. Finally, the exact cause of death may be uncertain and, paradoxically, correctness of the PE diagnosis also cannot be verified. Despite these limitations, the relevance of the study of Keller et al. 10 includes the demonstration of an evolutive pattern on the number, characteristics, and outcome of hospitalized PE patients over the past decades in a nationwide report. In addition, it confirms the underuse of a first-line reperfusion method such as systemic thrombolysis in high-risk PE patients, despite the confirmation of a favourable risk to benefit ratio also in real-life large-scale data collections. Furthermore, second-line alternative reperfusion methods such as surgical embolectomy and percutaneous catheter treatment are also largely underused in high-risk PE patients. This requires a call for an appropriate implementation of the PE practice guideline recommendations1 , 45 on the proper use of the reperfusion methods in haemodynamically unstable high-risk PE patients. This work was supported by the Department of Investigational, Diagnostic and Specialty Medicine, University of Bologna, Bologna, Italy and the National Institute of Biostructures and Biosystems, Rome, Italy. Conflict of interest: none declared. The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.