EM – Critical Care Journal Club

Topic: Thrombolysis in Sub-massive PE – Part II

Article: Chatterjee, Saurav, et al. “Thrombolysis for pulmonary embolism and risk of all-cause mortality, major bleeding, and intracranial hemorrhage: a meta-analysis.” Jama 311.23 (2014).

Resident: Thiago Oliveira, MD, MPH

Case / Problem

There are over 100,000 incident cases of PE annually in the US, with as many as 25% of patients presenting with sudden death.

Does thrombolytic therapy confer a mortality benefit that outweighs bleeding risk in patents with sub-massive PE (hemodynamically stable patients with evidence of right ventricular dysfunction) when compared to the standard of care (anticoagulation therapy)?

How was article selected?

Numerous FOAM resources were reviewed to identify commonly referenced sources for this clinical question. This meta-analysis was selected to compliment two the of the most widely-referenced RCTs on the topic and an article following long-term outcomes for one of the RCTs. This article aims to achieve statistical power and determine whether thrombolytic therapy in acute PE is associated with improved survival.

Study Description and Research Question

A meta-analysis of randomized control trials comparing thrombolytic therapy vs anticoagulant therapy in PE patients to answer the question:

Is thrombolytic therapy associated with lower rates of all-cause mortality when compared to anticoagulant therapy in patients with pulmonary embolism, including the subset of patients with intermediate-risk (sub-massive) PE? Is thrombolytic therapy associated higher rates of major bleeding versus anticoagulation in the aforementioned population?


Thrombolytic therapy may offer greater therapeutic benefit than the current standard of care in some patients with PE. Systemic thrombolysis is an intervention that could be offered in the ED if such a benefit exists.


Population: 16 RCTs comparing thrombolysis vs anticoagulation in PE – total 2115 individual subjects. Of these, 8 trials comprising 1775 patients specified inclusion of patients with intermediate-risk PE.  

Intervention: Thrombolytic therapy for PE (one of trials was catheter directed, remainder were systemic administration)

Comparison: Anticoagulation with any of the following – LMWH, VKA, fondaparinux, unfractionated heparin

Outcomes: Primary outcomes of all-cause mortality and major bleeding. Secondary outcomes recurrent PE and ICH. Peto OR estimates and 95% CIs calculated using fixed-effect models.

Appraisal of Internal Validity

  • Authors followed conventional practices for meta-analysis.
  • Meta-Analysis has clear and focused questions describing target population, intervention, comparison group and outcomes of interest.
  • Used trial level data, not patient level data which would have been gold standard.
  • Comprehensive search through relevant databases. Search was not limited to English. Search strategy included both controlled vocabulary terms (MeSH) and text words. There was no mention in methods section of searches through reference lists from relevant studies, contact with experts, or searches for unpublished studies.
  • Inclusion criteria of studies were clearly defined a priori. A primary outcome in the meta-analysis was major bleeding, but reporting of this outcome in RTCs was not specified as a required criterion for inclusion in meta-analysis.
  • definitions of HD instability, shock, major bleeding, and minor bleeding were not standardized, sometimes not provided in included studies
  • varying does and types of thrombolytic therapy combined in meta-analysis
  • primary outcomes reported at varying time intervals
  • Systematic review to identify pertinent studies done using PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) statement framework. Methodological quality of included studies was assessed by bias assessment guided by Cochrane Handbook of Systematic Reviews.
  • 4 trials ~75% of patients; MAPPETT-3 and PEITHO about one-third of patients
  • Authors report that outcomes were associated with negligible heterogeneity (I2 < 25%) and no publication bias was observed with funnel plots or Egger regression test. Analysis not provided.
  • Potential COI. Individual authors did disclose relationships with pharmaceutical companies, financial support from EKOS Corp, Embolitech, Boston Scientific, and Cordis Corp, and equity interest in Johnson&Johnson.
  • Planned exploratory sub-analyses to investigate outcomes in patients > 65 y/o, an analysis excluding ULTIMA trial (only trial not using systemic thrombolysis).
  • Results reported using Peto Odds Ratios, which are appropriate given low expected outcome proportions (2-3%). Actual proportions (deaths / total number of patients per treatment group) were between 2% and 4%. Peto OR can lead to bias when treatment effect is large or groups are unbalanced. Groups are well balanced in this meta-analysis (1061 vs 1054 for all PEs; 866 vs 889 for sub-massive PEs).


Primary Results

Use of thrombolytics vs AC therapy was associated with lower all-cause mortality OR 0.53 (95% CI 0.32 – 0.88) and greater risk of major bleeding vs AC therapy OR 2.73 (95% CI 1.91 – 3.91), among patients with sub-massive PE, use of thrombolytics associated with lower all-cause mortality OR 0.48 (95% CI 0.25 – 0.92) and greater risk of major bleeding OR 3.19 (95% CI 2.07 – 4.92).

 Primary result of decreased mortality remained significant with exclusion of ULTIMA trial (only trial using catheter directed thrombolysis instead of systemic administration) for both overall analysis and intermediate-risk PE sub-analysis. No association between thrombolysis and increase in major bleeding in patients < 65 years of age (OR 1.25; 95% CI 0.50 – 3.14).

Some wide confidence intervals, and CI for all-cause mortality in intermediate risk PE approaches OR of 1 (0.25-0.92). Sensitivity analysis done with aim of demonstrating that inclusion of further studies unlikely to change overall conclusions.

External Validity

Given heterogeneity across included studies in terms of patient characteristics collected, invention offered, and follow-up period – it is difficult to come to decisive conclusion of generalizability. Meta-analysis cannot replace RCT to inform practice, but that may be difficult to achieve for this clinical question.

Some argument that mortality benefit is potentially under-represented given inclusion of all sub-massive PE patients and patients symptomatic for several days.

Utility to Practice

Results suggest that intermediate risk, sub-massive acute PE patients may benefit from systemic thrombolysis. Difficult to change practice based on this meta-analysis alone.

However, with RCT data, would consider thrombolytics in aforementioned patients who are age 65 and younger – especially lower-dose thrombolytics as alternative to conventional anticoagulation therapy.

Resident Reviewer: Thiago Oliveira, MD, MPH
Review: Alex Sheng, MD