Thrombolysis for Pulmonary Embolism and Risk of All-Cause Mortality, Major Bleeding, and Intracranial Hemorrhage – A Meta-analysis

Chatterjee. 2014 JAMA:311(23);2414-2421.

Clinical Question

• In patients with PE, does the addition of thrombolysis to anticoagulation, compared with anticoagulation alone, affect mortality and bleeding complications?

Design

• Meta-analysis
• Followed PRISMA statement
• Systematic search using MeSH and keywords in the following databases:
  • PubMed, Cochrane, EMBASE, EBSCO, Web of Science, CINAHL
  • No language restriction

Population

• Trial Inclusion: Randomised trials comparing thrombolysis vs. anticoagulant therapy in patients with PE
• Trial Exclusion: Comparisons of different thrombolytic agents against one another or different doses of the same thrombolytic drug
• Trial Populations: Adult patients; stratified into high-, intermediate-, low- and unknown-risk according to haemodynamic stability and RV dysfunction (by echogardiogram or biomarkers).
  • 16 trials comprising 2115 patients were identified, 4 of these trials accounted for 74% of the total patients
    • PEITHO: 1005 patients
    • MAPPETT: 256 patients
    • UPETSG: 180 patients
    • MOPETT: 121 patients
  • 8 trials comprising 1775 patients with intermediate-risk PE for sub-group analysis
    • Haemodynamically stable (BP>90mmHg) with objective evidence of right ventricular dysfunction on echocardiogram, and/or Troponin/brain natriuretic peptide
      • PEITHO: 1005 patients
      • MAPPETT 3: 256 patients
      • MOPETT: 121 patients
      • Goldhaber: 101 patients
      • TOPCOAT: 83 patients
      • FASULLO: 72 patients
      • ULTIMA: 59 patients (catheter directed thrombolysis)
      • TIPES: 58 patient

Intervention

• Thrombolysis and anticoagulation
  • various types and doses used

Control

• Anticoagulation alone
  • Heparin, low molecular weight heparin or fondaparinux

Outcome

 

Measure	Thrombolytic	Anticoagulants	OR	ARR	NNT/NNH	p
Primary Outcome
All-cause mortality	2.17%	3.89%	0.53 (95% CI 0.32–0.88)	1.72%	59	0.01
Major bleeding	9.24%	3.42%	2.73 (95% CI 1.91–3.91)	-5.82%	18	<0.001
Secondary Outcomes
Recurrent PE	1.17%	3.04%	0.40 (95% CI 0.22–0.74)	1.87%	53	0.03
ICH	1.46%	0.19%	4.63 (95% CI 1.78–12.04)	-1.27%	79	0.002
Pre-specified sub-group analyses
Major bleeding age >65 years	12.93%	4.1%	3.10 (95% CI 2.10–4.56)	-8.83%	11	0.001
Major bleeding age ≤65 years	2.84%	2.27%	1.25 (95% CI 0.50–3.14)	-0.57%	175	0.89
All-cause mortality Intermediate-risk PE	1.39%	2.92%	0.48 (95% CI 0.25–0.92)	1.53%	65	0.03
Major bleeding Intermediate-risk PE	7.74%	2.25%	3.19 (95% CI 2.07–4.92)	-5.49%	18	0.001
OR = Odds Ratio; ARR = Absolute Rate Reduction; NNT/NNH = Number needed-to-treat/harm; ICH = Intra-cerebral haemorrhage

Authors’ Conclusions

• Among patients with PE, including those who are haemodynamically stable, thrombolysis was associated with lower mortality and more major bleeding

Strengths

• Clear inclusion criteria for studies included
• Results of individual studies presented clearly
• Trial registered with PROSPERO

Weaknesses

• Definitions for haemodynamic isntability, major and minor bleeding and RV dysfunction were not standardised between studies
• Varying doses and types of thrombolysis used in different studies.
• One study utilised catheter directed thrombolysis which may not be possible in most centres. However with this study excluded the results for the primary outcome were still statistically significant
• The authors report that study methods were assessed for risk of bias, but this is not reported in the results. MAPPETT-3 and PEITHO provide 1/3 of patients for this meta-analysis, and both have weaknessness of concern.

The Bottom Line

• Thrombolysis for PE is associated with a lower all cause mortality when compared to anticoagulation. This improvement is maintained in the subgroup of ‘intermediate risk’ PE with a NNT of 65. This is despite a significant increase in major bleeding in the thrombolysis group, with a NNH of 18. Of note, patients who were ≤65 years had no increase in major bleeding with thrombolysis.
• With the MOPETT and TOPCOAT studies finding improved longterm outcomes with regards to pulmonary hypertension, and self assessment of overall health in the thrombolysis groups, this meta-analysis provides a convincing argument for the use of thrombolysis in submassive PE.
• If I have a submassive PE, I want thrombolysis. (I am under 65 years). I would want half-standard dose as per the MOPETT trial. If my dad (he is over 65 years) had a sub-massive PE I would not recommend thrombolysis as the 12.9% chance of major bleeding means that the risks may outweigh the benefits.

Links

abstract / doi: 10.1001/jama.2014.5990

Editorial, Commentaries or Blogs

• pulmcrit

Metadata

Summary author: @DavidSlessor and @DuncanChambler
Summary date: 29 June 2014
Peer-review editor: @stevemathieu75