Introduction: Randomized data on clinical outcomes of large-bore mechanical thrombectomy (LBMT) versus catheter-directed thrombolysis (CDT) for pulmonary embolism (PE) have recently been reported. This post hoc analysis provides additional data on the trial’s hospital and intensive care unit (ICU) utilization trends.
Methods: The PEERLESS randomized controlled trial (NCT05111613) enrolled 550 acute intermediate-risk PE patients randomized to LBMT (n=274) or CDT (n=276). Additionally, a non-randomized thrombolytics contraindication cohort (CC) enrolled 142 patients treated with LBMT. The primary endpoint was a win-ratio of all-cause mortality, intracranial hemorrhage, major bleeding per ISTH definition, clinical deterioration (CD) and/or bailout (BO), and post-procedural ICU stay through discharge or 7 days. Patients were followed through the 30-day visit.
Results: The primary endpoint win ratio was 5.01 (95% CI: 3.68–6.97; P< 0.001) favoring LBMT over CDT. This finding was driven by significantly lower rates of CD/BO and ICU utilization in the LBMT arm. CD/BO occurred in 1.8% (5/274) of LBMT patients vs 5.4% (15/276) of CDT patients (P=0.0378), and in 3.5% (5/142) of CC patients. All CD/BO events in the LBMT (5/5) and 4/5 events in the CC patients occurred on the procedure day, with the fifth CC event occurring the next day. In contrast, the 15 CD/BO events in the CDT group began a mean of 2.1 ± 1.7 days after CDT initiation and were more clinically severe. Less than half (41.6%) of LBMT patients had post-procedure ICU stays (mean 14.2 ± 25.4 hours), with 50% of stays due to institutional standard of care only. In contrast, 98.6% of CDT patients had post-procedure ICU stays (P < 0.0001 vs LBMT) with mean stay of 39.3 ± 28.0 hours (P < 0.0001 vs LBMT). In CC patients, who were sicker with higher bleeding risks, less than half (45.8%) had post-procedural ICU stays with a mean stay of 20.3 ± 36.4 hours. LBMT patients had shorter total hospital stays than CDT patients (4.5 ± 2.8 vs 5.3 ± 3.9 overnights; P=0.0022), with 7.8 ± 7.4 overnights for CC patients.
Conclusions: Less frequent ICU admissions, shorter hospital and ICU stays, and fewer CD/BO events after the index procedure suggest that LBMT may use fewer hospital resources to manage PE. Additional analyses incorporating specific costs are needed.
