Outcomes in patients with high-risk pulmonary embolism deteriorating into cardiogenic shock or cardiac arrest receiving VA-ECMO therapy

Introduction: Pulmonary embolism (PE) leading to cardiogenic shock and cardiac arrest prevails as a common cause for death in adults. Venoarterial extracorporeal membrane oxygenation (VA-ECMO) has recently shown promising results in patients (pts) undergoing cardiopulmonary resuscitation for refractory cardiac arrest. While some predisposing characteristics for favourable outcome have been investigated, the influence of the primary cause of cardiac arrest remains widely unclear. In current consensus opinion, VA-ECMO therapy is a valid therapeutic option in pts with high risk PE deteriorating into cardiogenic shock or cardiac arrest. VA-ECMO support is known to provide adequate systemic oxygenation, while disencumbering the right ventricle, therefore potentially reversing the obstructive shock caused by PE. We hypothesized, that the use of VA-ECMO in pts with high-risk PE might be associated with improved outcome, when compared to pts with other indication for VA-ECMO therapy.

Methods: We retrospectively analysed data of 441 consecutive pts treated with VA-ECMO for intrahospital cardiac arrest (IHCA), out of hospital cardiac arrest (OHCA), as well as refractory cardiogenic shock in our cardiac arrest center. Pts where categorized following whether PE or a different pathology was defined as the primary cause of deterioration and the subgroups compared. Primary outcome was defined as survival of the primary hospital admission with favourable neurological outcome (cerebral performance category [CPC]-score ≤2). Furthermore, in a subgroup including only pts undergoing eCPR for refractory cardiac arrest (N=368) we consecutively performed an interaction analysis to assess, whether the apparent differences were mediated by cardiac arrest.

Results: Of the 441 studied pts, 37 (8%) suffered a massive PE as cause of deterioration or cardiac arrest. Regarding the VA ECMO indication, OHCA was less common in pts with PE (35% vs. 59%; p=0.007), while prevalence of IHCA (38% vs. 25%; p=0.13) did not differ significantly. Overall, pts with pulmonary embolism received intravenous thrombolytic therapy in 35% of cases (N=13), 35% (N=13) had known absolute contraindications for thrombolysis. Catheter directed therapy was performed in one case (2.7%). Bleeding complications requiring transfusion were observed more frequently in the PE group (38% vs. 20%; p=0.042). The subgroup of pts with PE showed significantly higher rates of overall (43% vs. 23%, p=0.011) and neurologically favourable survival (32% vs. 14%, p=0.022). The effect was slightly less pronounced for pts undergoing eCPR for IHCA or OHCA (37% vs. 20%; p=0.065). Effect analysis revealed PE to be directly associated with survival and favourable neurological outcome in all pts (HR: 1.62, 95%CI: 1.07 – 2.45), as well as pts undergoing eCPR (HR: 1.59, 95%CI: 0.99 – 2.57). Interaction analysis ruled out a direct mediation by cardiac arrest (OHCA and IHCA, p=0.88/p=0.7).

Conclusion: Our analysis suggests a survival benefit for pts with PE undergoing VA-ECMO therapy regardless of preceding cardiac arrest. The higher rates of bleeding complications in pts with PE might be explained by the higher rates of prior administration of thrombolysis. These results must be evaluated under the constraint of lower rates of OHCA and higher rates of IHCA in the PE subgroup, which might account for improved outcome. Nonetheless, these findings warrant further investigation of VA-ECMO therapy in pts with high-risk PE.