The safety and efficacy of systemic versus catheter-based therapies: application of a prognostic model by a pulmonary embolism response team

The decision by pulmonary embolism response teams (PERTs) to utilize anticoagulation (AC) with or without systemic thrombolysis (ST) or catheter-directed therapies (CDT) for pulmonary embolism (PE) is a balance between the desire for a positive outcome and safety. Our primary aim was to develop a predictive model of in-hospital mortality for patients with high- or intermediate-risk PE managed by PERT while externally validating this model. Our secondary aim was to compare the relative safety and efficacy of ST and CDT in this cohort. Consecutive patients hospitalized between June 2014 and January 2020 at the Cleveland Clinic Foundation and The University of Rochester with acute high- or intermediate-risk PE managed by PERT were retrospectively evaluated. Groups were stratified by treatment strategy. The primary outcome was in-hospital mortality, and secondary outcome was major bleeding. A logistic regression model to predict the primary outcome was built using the derivation cohort, with 100-fold bootstrapping for internal validation. External validation was performed and the area under the receiver operating curve (AUC) was calculated. Of 549 included patients, 421 received AC alone, 71 received ST, and 64 received CDT. Predictors of major bleeding include ESC risk category, PESI score, hypoxia, hemodynamic instability, and serum lactate. CDT trended towards lower mortality but with an increased risk of bleeding relative to ST (OR = 0.42; 95% CI [0.15, 1.17] and OR = 2.14; 95% CI [0.9, 5.06] respectively). In the multivariable logistic regression model in the derivation institution cohort, predictors of in-hospital mortality were age, cancer, hemodynamic instability requiring vasopressors, and elevated NT-proBNP (AUC = 0.86). This model was validated using the validation institution cohort (AUC = 0.88). We report an externally-validated model for predicting in-hospital mortality in patients with PE managed by PERT. The decision by PERT to initiate CDT or ST for these patients had no impact on mortality or major bleeding, yet the long-term efficacy of these interventions needs to be elucidated.

Automated summary

The study aimed to develop a predictive model for in-hospital mortality in patients with high- or intermediate-risk PE managed by PERT and externally validate it, while comparing the safety and efficacy of ST and CDT. CDT trended towards lower mortality but with increased bleeding risk. Key predictors of in-hospital mortality identified by the logistic regression model included age, cancer, hemodynamic instability, and elevated NT-proBNP levels.