Surgical pulmonary embolectomy and catheter-based therapies for acute pulmonary embolism: A contemporary systematic review.

OBJECTIVES: Mortality in acute pulmonary embolism (PE) is believed to be principally due to the subgroup of PEs that are massive. Systemic thrombolysis is the therapeutic mainstay for acute massive PE, despite evidence suggesting limited survival benefits. Both catheter-based therapies (CBT) and surgical pulmonary embolectomy (SE) are well-accepted alternatives to treat acute PE. However, the comparative effectiveness of these approaches is difficult to study. We conducted a systematic review of CBT and SE for acute PE. METHODS: The PubMed database was queried for CBT- and SE-related publications between January 1998 and June 2017. A minimum of 10 patients undergoing intervention(s) was required for inclusion, and studies must not have excluded patients with massive PE. End points examined included hospital mortality, and additionally for CBT, procedural success rate. RESULTS: A total of 75 studies (41 of CBT, 34 of SE) were identified, with 1650 patients undergoing CBT and 1101 undergoing SE. Patients undergoing SE were more critically ill than those undergoing CBT (massive PE, 545 out of 975 [55.9%] for SE vs 742 out of 1553 [47.8%] for CBT). Cardiopulmonary resuscitation (CPR) was required in 217 out of 1015 patients undergoing SE (21.4%) versus 38 out of 983 patients undergoing CBT (4.0%). The hospital mortality of SE was 14.0%, versus 5.6% for CBT, in the entire patient group. However, the hospital mortality of SE in patients with pre-SE CPR was 46.3%, whereas it was 6.8% in those patients without pre-SE CPR. Although CPR was associated with an increased risk of mortality both for CBT and SE, it accounted for all of the mortality effect on SE (the adjusted odds ratio for CPR in a random effects model with treatment considered was 9.79 (95% confidence interval, 4.98-19.17; P < .0001). The adjusted odds ratio for mortality for SE relative to CBT was 1.36 (95% confidence interval, 0.80-2.32; P = .84). Moreover, CBT was associated with a procedural failure rate of 8.3%. CONCLUSIONS: Both CBT and SE were associated with satisfactory published outcomes. SE is associated with greater absolute postprocedure mortality than CBT, but has been undertaken in more critically ill populations. The markedly higher incidence of CPR in SE accounts for the differential mortality between the patients undergoing SE and those undergoing CBT. Decision making with respect to best therapy must take into account potential needs for periprocedure artificial mechanical right ventricle and lung support, institutional experience and outcomes, anticipated therapeutic efficacy and benefit, and approach-specific risks.