Outcomes in high-risk pulmonary embolism patients undergoing FlowTriever mechanical thrombectomy: the FLAME study in perspective

Pulmonary embolism (PE) is a common and life-threatening acute cardiovascular problem. In some patients, an abrupt increase in right ventricular (RV) afterload and impaired gas exchange result in acute RV failure, hypoperfusion, and severe hypoxemia due to V/Q mismatch. State-of-the-art management of acute PE includes early detection, risk stratification based on clinical, imaging, and biomarker criteria, and multidisciplinary decision-making regarding therapy.1,2 The European Society of Cardiology guidelines for management of PE recommend selection of therapy based on clinical severity of the presentation with class I recommendations for anticoagulation and systemic thrombolysis in patients with high-risk PE and for surgical thrombectomy in such patients in whom thrombolysis is contraindicated or has failed.3 Catheter-directed interventions for acute PE are aimed at increasing the cross-sectional area of patent pulmonary vasculature, thereby lowering resistance and alleviating V/Q mismatch.1,2 Percutaneous catheter-directed treatments are recommended as a reasonable alternative (Class IIa) in patients for whom thrombolysis is contraindicated or has failed. Compared with catheter-directed fibrinolytic administration, catheter-based thrombectomy has the potential advantage of immediate removal of thrombus and restoration of pulmonary blood flow. Catheter-based thrombectomy approaches include rheolytic devices (e.g. AngioJet), aspiration thrombectomy (e.g. Indigo or AngioVac), or mechanical thrombectomy, using devices such as the FlowTriever. Previous observational studies of percutaneous thrombectomy devices have enrolled patients with intermediate-risk PE and demonstrated reductions in the RV/LV ratio at 48 h after treatment along with improvements in hemodynamic parameters immediately post-procedure. For example, in the FLARE study, the mean reduction in RV/LV ratio from baseline to 48 h was 0.38 (P < 0.0001) and there was one major bleeding event among 106 patients.4 The FlowTriever for Acute Massive pulmonary Embolism (FLAME) study was a prospective cohort observational study designed to evaluate treatment outcomes in patients with high-risk PE being treated with the FlowTriever catheter at the discretion of their primary providers. The primary endpoint, evaluated through hospital discharge or 45 days (whichever came first), was a composite of all-cause mortality, clinical deterioration, bailout, and major bleeding. Secondary endpoints included individual elements of the primary composite, stroke, device-related complications, and access site injury.5 Eligible patients were at least 18 years of age with shock as the result of PE. Key exclusion criteria included intracardiac thrombus or clot-in-transit, comatose state after cardiac arrest, or witnessed arrest with >30 min of cardiopulmonary resuscitation, active COVID-19, hematocrit <28%, or platelet count <25 000/µL.5 The investigators reported a pre-specified interim analysis at enrollment of 50 patients with a pre-specified stopping rule in comparison to a performance goal of 32% for the primary endpoint based on a published meta-analysis.6,7 Data were reported for a ‘context’ group that included all patients with high-risk PE primarily managed without FlowTriever during the study period. The findings as presented at the American College of Cardiology Scientific Sessions 2023 were as follows.6 Between March 2021 and December 2022, 11 centers in the United States enrolled 53 patients with shock due to PE who were treated with transcatheter pulmonary embolectomy using the FlowTriever catheter. The mean age was 64.8 years and 49% were women. Qualifying inclusion criteria for high risk were hypotension in 64%, need for vasopressor support in 60%, and resuscitation after cardiac arrest in 21%. Of enrolled patients, 55%, 9%, and 11% had Society for Cardiovascular Angiography and Interventions (SCAI) stages C, D, or E shock, respectively. The remainder was reported as SCAI stages A or B (pre-shock). The distribution of PE included central (93%), lobar (68%), and segmental (51%) thrombus. The primary outcome occurred in 17% of patients with the following rates of the individual elements of the endpoint: death 1.9%; clinical deterioration 15.1%; bailout 3.8%; major bleeding 11.3%. Device-related complications occurred in 22.6%, with individual components of hemoglobin decrease (15.1%), vascular access hemorrhage (7.5%), and hypotension (1.9%).There were no reported pulmonary vascular injuries or hemorrhagic strokes. The non-randomized ‘context’ group had 75% that required vasopressors and 33% that had a preceding cardiac arrest. SCAI stage E shock was present in 33%. Systemic thrombolysis was administered in 69% and surgical thrombectomy in 1.6%. No patients in the context group received mechanical thrombectomy of any type as the primary treatment. Major bleeding occurred in 24.6%. Their in-hospital mortality was 29.5%. The pre-specified in-hospital mortality performance goal was 28.5%. In this modestly sized observational study of a percutaneous mechanical thrombectomy device in patients with high-risk PE, mortality was qualitatively very low. Vascular access hemorrhage occurred with moderate incidence (1 in 13–14 patients), major bleeding in ∼1 in 10 patients and there were no reported pulmonary vascular injuries. These data contrast favorably with observational data from a reported meta-analysis of patients with high-risk PE treated with other strategies7 and are encouraging for the possibility of a favorable profile of efficacy and safety compared with current guideline-directed care. Further, these findings are reasonably aligned with data reported to date from other registries. However, in this non-randomized study, patients in the ‘context’ group had qualitatively more cardiac arrest, need for vasopressors, and stage E cardiogenic shock. These baseline characteristics highlight the selection pressures that may exist in this non-randomized design. While the absolute rates of mortality and major bleeding support some optimism for the studied strategy, the observational study design does not support a claim of efficacy. In addition, because of the small sample size, confidence limits around individual event rate estimates are wide. Nonetheless, the findings in FLAME provide useful preliminary information for multidisciplinary teams making decisions regarding therapy for patients with high-risk PE. Additional data are needed to reliably determine the safety and efficacy of percutaneous mechanical thrombectomy as a possible first-line therapeutic alternative. Moreover, comparative efficacy vs. other catheter-based approaches should be evaluated. The ongoing much larger FLASH registry is intended to enroll 1300 all-comers with acute PE and RV/LV ratio ≥0.9 undergoing mechanical thrombectomy. PEERLESS is an ongoing randomized controlled trial comparing FlowTriever to catheter-directed thrombolysis (NCT05111613) in patients with acute intermediate-risk PE. The HI-PEITHO randomized trial (NCT04790370) with a planned enrollment of ∼400 patients with intermediate-risk PE will compare clinical outcomes with catheter-directed fibrinolytic vs. systemic anticoagulation alone. These additional studies are needed for rigorous assessment of transcatheter management strategies for moderate and high-risk PE. David A.