Abstract
Objectives:
Evidence guiding the optimal timing of mechanical thrombectomy for patients presenting with intermediate-risk pulmonary embolism (PE) is limited. We aimed to evaluate whether the timing of mechanical thrombectomy is associated with improved clinical outcomes in this patient population.
Design, Setting, and Patients:
This multicenter, retrospective cohort study was conducted at five large academic hospitals. In total, 290 patients presenting with intermediate-risk PE who were treated with mechanical thrombectomy were included for analysis. The primary outcome was in-hospital mortality. Using generalized estimated equation, we compared the odds of in-hospital mortality for patients undergoing early intervention (EI; mechanical thrombectomy < 12 hr after PE diagnosis) vs. those undergoing delayed intervention (DI; mechanical thrombectomy ≥ 12 hr after diagnosis) while comparing for PE severity and other confounders.
Interventions:
None.
Measurements and Main Results:
EI was performed in 179 patients (61.7%), while 111 patients (38.3%) received DI. Unadjusted mortality did not differ significantly between groups (7.3% [13/179] vs. 10.8% [12/111]; p = 0.39). After adjusting for the Pulmonary Embolism Severity Index and Composite Pulmonary Embolism Shock scores, timing of intervention did not influence mortality (odds ratio, 1.80; 95% CI, 0.82–3.95; p = 0.14). However, patients in the EI group had greater reductions in in pulmonary artery systolic pressure (–25.8% [17.0] vs. –18.9% [17.1]; p = 0.020 and mean pulmonary artery pressure, –26.8% [17.7] vs. –20.2% [19.7]; p = 0.016) and lower rates of intubation (8.9% [16/179] vs. 18% [20/111]; p = 0.028).
Conclusions:
In patients presenting with intermediate-risk PE, timing of mechanical thrombectomy did not influence in-hospital mortality. EI may result in greater reductions in pulmonary artery pressures and decreased incidence of intubation compared with DI.
Keywords: hemodynamics, mortality, pulmonary embolism, thrombectomy, timing
KEY POINTS.
Question: Does the timing of mechanical thrombectomy for patients presenting with intermediate-risk pulmonary embolism impact mortality?
Findings: This multicenter, retrospective study compared patients receiving mechanical thrombectomy less than 12 hours vs. greater than or equal to 12 hours from intermediate-risk pulmonary embolism diagnosis. There were no significant differences in mortality between the two groups. After adjusting for prognostic risk scores, delayed intervention was not inferior to early intervention.
Meaning: Early intervention with mechanical thrombectomy resulted in similar mortality compared with delayed intervention.
Pulmonary embolism (PE) is a life-threatening condition and the third most common cause of cardiovascular mortality after myocardial infarction and stroke (1, 2). Acute PE causes right ventricle (RV) failure as a result of acute pulmonary vasculature obstruction leading to a sudden increase in pulmonary arterial pressure and RV afterload. Unlike the left ventricle (LV), the RV is a thin-walled, highly compliant ventricle that cannot sustain an acute afterload increase, leading to RV dilatation, bowing of the interventricular septum, rise in wall stress, increased RV oxygen demand, and drop in systemic cardiac output. Eventually, this precipitates RV ischemia, cardiogenic shock, and hemodynamic collapse (3). Even in patients with intermediate-risk PE who are hemodynamically stable, the presence of RV dysfunction carries a 10–25% mortality rate (3, 4).
Therefore, catheter-directed therapies (CDTs) that rapidly relieve RV pressure overload have been developed as an adjunctive treatment to anticoagulation for acute PE. CDT offer a percutaneous approach that has been associated with low rates of bleeding and hemorrhagic strokes compared with systemic thrombolysis (5–10). CDT with large-bore aspiration devices has been shown to reduce the RV/LV ratio and have an immediate drop in mean pulmonary artery pressure (mPAP) (9–11). More recently, a randomized trial comparing large-bore thrombectomy to catheter-directed thrombolysis in intermediate-risk PE showed that the thrombectomy group had a lower incidence of clinical deterioration with similar rates of major bleeding and mortality. The data support catheter-based thrombectomy as an effective approach to reverse RV strain in acute PE while mitigating bleeding risk (12). Given this data, the 2019 European Society of Cardiology (ESC) guidelines provided a IIa recommendation for consideration of percutaneous revascularization in intermediate-risk PE with clinical deterioration on anticoagulation therapy (3). Similarly, the Pulmonary Embolism Response Team (PERT) Consortium recommends consideration CDT for patients with intermediate-PE with evidence of or risk for clinical decompensation based on vital signs, imaging, and laboratory values (13).
However, guidelines do not define the optimal timing of intervention in patients presenting with intermediate-risk PE. In clinical practice, timing of CDT patients with intermediate-risk PE varies and is often determined by clinical status and physician preference. A few retrospective studies of ultrasound-assisted thrombolysis (USAT) and mechanical thrombectomy treatments suggest improvement in patient outcomes in early compared with delayed treatment (14, 15). We hypothesized that earlier compared with delayed intervention (DI) with mechanical thrombectomy in patients with intermediate-risk PE would result in fewer in-hospital mortalities and less post-thrombectomy clinical decompensation.
MATERIALS AND METHODS
Study Population
This study was a multicenter, retrospective study including patients presenting to five university quaternary care centers with an intermediate-risk PE between January 2014 to June 2024. Patients treated with mechanical thrombectomy were identified from departmental databases, PERT activations within the hospital systems, chart review of medical records, and CPT codes (pulmonary artery embolectomy [33910–33915]). Patient demographic and procedural data was extracted from the hospitals’ electronic medical records by trained clinicians and managed using Research Electronic Data Capture tools. Inclusion criteria for this study included adults (18 yr and older) presenting with an intermediate-risk PE who were treated with percutaneous mechanical thrombectomy. Exclusion criteria consisted of patients meeting high-risk PE criteria (cardiac arrest, hemodynamic instability requiring inotrope, vasopressor, or mechanical circulatory support) before mechanical thrombectomy, patients treated with USAT or anticoagulation without catheter-based thrombectomy, and patients treated on an outpatient basis. We compared patient outcomes between individuals receiving intervention within 12 hours of PE diagnosis (early intervention [EI]) to those receiving intervention 12 hours or later of PE diagnosis (DI). The 12-hour threshold reflected the clinical goal of rapidly reversing RV afterload in efforts to prevent early hemodynamic collapse. This threshold has been used previously (15), and reflects operational differences in intervention delivery (same-day/overnight vs. next-day). The procedures were followed in accordance with the ethical standards of each institutions’ Institutional Review Board (IRB) and in accordance with the Helsinki Declaration of 1975. This study was approved by the IRB of each participating site; please see the Supplemental Digital Content (https://links.lww.com/CCM/H955) for more information. A waiver of informed consent was granted, given that the research was retrospective and involved no more than minimal risk.
Study Definitions and Outcome Variables
Intermediate-risk PE was defined according to the ESC 2019 criteria (3). These criteria include hemodynamic stability defined as a sustained mean arterial pressure above 65 mm Hg without the use of inotropes or vasopressors, Pulmonary Embolism Severity Index (PESI) of III–IV or simplified PESI (sPESI) of greater than or equal to 1 (16, 17), evidence of RV dysfunction on transthoracic echocardiogram or CT imaging, or the presence of elevated troponin. Initial vitals and laboratory data were obtained at emergency department admission or at symptom onset for hospitalized patients. For patients with incomplete or discordant biomarker or imaging data, risk stratification relied on clinician documentation of PE severity. These individuals were included only when clinical assessment clearly supported an intermediate-risk classification, and they were not managed as low- or high-risk.
The primary outcome was in-hospital mortality. Secondary outcomes included post-thrombectomy cardiac arrest, extracorporeal membrane oxygenation (ECMO) use, need for mechanical ventilation, need for renal replacement therapy, percentage change in pulmonary artery pressures (PAPs), ICU length of stay (LOS), and hospital LOS.
Statistical Analysis
Patient demographic, clinical characteristics, and outcomes were summarized for the total cohort and by timing of intervention, using mean (sd) for continuous and count (%) for categorical variables. Fisher exact tests and t tests and were used for continuous and categorical variables, respectively, to compare these patient characteristics by intervention timing group. The odds of the primary outcome were compared by intervention timing groups using a generalized estimating equation (GEE) with exchangeable working correlation structure and logit link to account for correlation among patients treated at the same sites. Secondary outcome of percentage change in pressure was similarly evaluated by intervention timing group in GEE models with exchangeable working correlation structures, adjusted for the pre-intervention pressure measure for added precision. All primary and secondary models were adjusted for the potential confounders PESI (16) and Composite Pulmonary Embolism Shock (CPES) (18) scores (Supplemental Tables 1 and 2, https://links.lww.com/CCM/H955), as they incorporate variables that could influence both clinical decision-making and the timing of intervention. A p values are two sided and considered at the 0.05 level for statistical significance. Analyses were conducted in R, Version 4.5.0 (R Core Team, Vienna, Austria; 2025).
RESULTS
A total of 290 patients met the inclusion criteria. Of these, 179 patients (61.7%) were included in the EI group, while 111 patients (38.3%) were included in the DI group. Patient demographics and clinical characteristics are shown in Table 1; and Supplemental Table 3 (https://links.lww.com/CCM/H955). Clinical characteristics were recorded at time of emergency department admission or time of symptom onset if admitted earlier. The mean age of the entire cohort was 59.7 years (sd, 15.3 yr). One hundred fifty-five patients (53.4%) of the cohort were male. On presentation, the mean heart rate was tachycardic at 110 beats/min (sd, 21.2 beats/min), with a higher mean in the EI compared with the DI groups (111 vs. 106; p = 0.049). In both cohorts, 181 (62.4%) received supplemental oxygen. The mean lactic acid level was 2.91 (± 2.75) among the entire cohort, with no significant difference between the two groups. Patients in the EI group were more likely to have higher PESI scores (119 vs. 105; p = 0.008) and syncope on initial presentation (56 vs. 19; p = 0.009). The EI cohort also had higher prevalence of an elevated troponin level (91.2% vs. 81.1%; p = 0.038) and higher initial creatinine level (1.15 vs. 1.01; p = 0.012). Conversely, patients in the DI group had a higher prevalence of chronic kidney disease (13.5% vs. 5.0%; p = 0.015). The vast majority of the cohort was classified as intermediate-/high-risk (274, 94.5%), with no significant difference between the two groups.
TABLE 1.
| Variable | Total (n = 290) | Early Intervention (n = 179) | Delayed Intervention (n = 111) | p |
|---|---|---|---|---|
| Age (yr), mean (sd) | 59.7 (15.3) | 60.4 (15.5) | 58.4 (14.9) | 0.28 |
| Sex (male), n (%) | 155 (53.4) | 101 (56.4) | 54 (48.4) | 0.23 |
| Chronic pulmonary disease, n (%) | 46 (15.9) | 26 (14.5) | 20 (18.0) | 0.51 |
| Chronic kidney disease, n (%) | 24 (8.3) | 9 (5.0) | 15 (13.5) | 0.015 |
| History of VTE, n (%) | 69 (23.8) | 42 (23.5) | 27 (24.3) | 0.89 |
| Cancer, n (%) | 91 (31.4) | 58 (32.4) | 33 (29.7) | 0.70 |
| Tobacco, n (%) | 119 (41.0) | 71 (39.7) | 48 (43.2) | 0.62 |
| Surgery within 3 mo, n (%) | 57 (19.7) | 38 (21.2) | 19 (17.1) | 0.45 |
| Hospitalization within 3 mo, n (%) | 69 (23.8) | 47 (26.3) | 22 (19.8) | 0.26 |
| Family history of VTE, n (%) | 34 (11.7) | 20 (11.2) | 14 (12.6) | 0.71 |
| History of COVID-19, n (%) | 39 (13.4) | 25 (14.0) | 14 (12.6) | 0.86 |
| Systemic estrogen use, n (%) | 15 (5.2) | 8 (4.5) | 7 (6.4) | 0.59 |
| Heart rate (beats/min), mean (sd) | 111 (21.2) | 106 (21.0) | 109 (21.2) | 0.049 |
| Mean arterial pressure (mm Hg), mean (sd) | 94.1 (18.9) | 93.5 (17.9) | 94.9 (20.5) | 0.54 |
| Respiratory rate (breaths/min), mean (sd) | 22.4 (5.7) | 22.1 (5.6) | 22.3 (5.7) | 0.70 |
| Peripheral oxygen saturation (%), mean (sd) | 92.6 (6.0) | 93.9 (4.3) | 93.1 (5.5) | 0.050 |
| Supplemental oxygen use, n (%) | ||||
| None | 109 (37.6) | 62 (34.6) | 47 (42.3) | 0.23 |
| Nasal cannula | 131 (45.2) | 84 (46.9) | 47 (42.3) | |
| High-flow nasal cannula | 14 (4.8) | 6 (3.4) | 8 (7.2) | |
| Nonrebreather mask | 9 (3.1) | 9 (5.0) | 0 | |
| Oxygen mask | 5 (1.7) | 3 (1.7) | 2 (1.8) | |
| Noninvasive positive pressure ventilation | 9 (3.1) | 7 (3.9) | 2 (1.8) | |
| Mechanical ventilation | 13 (4.5) | 8 (4.5) | 5 (4.5) | |
| Syncope, n (%) | 75 (25.9) | 56 (31.3) | 19 (17.1) | 0.009 |
| Creatinine (mg/dL), mean (sd) | 1.10 (0.44) | 1.15 (0.48) | 1.01 (0.37) | 0.012 |
| Lactic acid (mmol/L), mean (sd) | 2.91 (2.75) | 3.15 (3.01) | 2.45 (2.14) | 0.081 |
| Elevated troponin, n, (%) | 254 (87.6) | 164 (91.6) | 90 (81.1) | 0.038 |
| Elevated B-type natriuretic peptide/N-terminal pro-B-type natriuretic peptide, n (%) | 181 (62.4) | 114 (67.1) | 67 (64.4) | 0.69 |
| PESI, mean (sd) | 113 (41) | 119 (42) | 105 (38) | 0.008 |
| Simplified PESI, n (%) | ||||
| 0 | 31 (10.7) | 16 (8.9) | 15 (13.5) | 0.17 |
| 1 or 2 | 172 (59.3) | 111 (62.0) | 61 (55.0) | |
| 3 or 4 | 62 (21.4) | 34 (19.0) | 28 (25.2) | |
| 5 | 22 (7.6) | 17 (9.5) | 5 (4.5) | |
| Pulmonary embolism risk classification, n (%) | ||||
| Intermediate-low | 16 (5.5) | 7 (3.9) | 9 (8.1) | 0.13 |
| Intermediate-high | 274 (94.5) | 172 (96.1) | 102 (91.9) |
PESI = Pulmonary Embolism Severity Index, VTE = venous thromboembolism.
See Supplemental Table 3 (https://links.lww.com/CCM/H955) for additional variables.
See Supplemental Table 4 (https://links.lww.com/CCM/H955) for missing data.
Table 2 summarizes the imaging and electrocardiographic characteristics. There were no differences in CT imaging characteristics between the two groups. Additionally, there were no differences in initial RV cavity size, function, and presenting electrocardiographic features.
TABLE 2.
Imaging and Electrocardiographic Characteristicsa
| Variable | Total (n = 290) | Early Intervention (n = 179) | Delayed Intervention (n = 111) | p |
|---|---|---|---|---|
| CT imaging | ||||
| RV strain on CTA, n (%) | 275 (94.8) | 172 (96.1) | 103 (92.8) | 0.31 |
| CTA RV/left ventricle ratio, mean (sd) | 1.52 (0.45) | 1.53 (0.47) | 1.50 (0.43) | 0.81 |
| Pulmonary embolism location, n (%) | ||||
| Saddle | 107 (36.9) | 69 (38.5) | 38 (34.2) | 0.29 |
| Bilateral | 173 (59.7) | 106 (59.2) | 67 (60.4) | |
| Unilateral | 6 (2.1) | 2 (1.1) | 4 (3.6) | |
| Acute deep vein thrombosis, n (%) | 198 (68.3) | 124 (69.3) | 74 (66.7) | 1.00 |
| Echocardiographic parameters | ||||
| RV dilatation, n (%) | 0.31 | |||
| Normal | 38 (13.1) | 29 (16.2) | 9 (8.1) | |
| Mild | 54 (18.6) | 26 (14.5) | 28 (25.2) | |
| Mild-to-moderate | 10 (3.4) | 7 (3.9) | 3 (2.7) | |
| Moderate | 75 (25.9) | 37 (20.7) | 38 (34.2) | |
| Moderate-to-severe | 21 (7.2) | 13 (7.3) | 8 (7.2) | |
| Severe | 31 (10.7) | 18 (10.1) | 13 (11.7) | |
| RV function, n (%) | ||||
| Normal function | 42 (14.5) | 27 (15.1) | 15 (13.5) | 0.64 |
| Mild dysfunction | 43 (14.8) | 24 (13.4) | 19 (17.1) | |
| Mild-to-moderate dysfunction | 12 (4.1) | 6 (3.4) | 6 (5.4) | |
| Moderate dysfunction | 69 (23.8) | 37 (20.7) | 32 (28.8) | |
| Moderate-to-severe dysfunction | 26 (9.0) | 12 (6.7) | 14 (12.6) | |
| Severe dysfunction | 38 (13.1) | 24 (13.4) | 14 (12.6) | |
| McConnell sign, n (%) | 73 (25.2) | 40 (22.3) | 33 (29.7) | 0.44 |
| Estimated pulmonary artery systolic pressure (mm Hg), mean (sd) | 44.6 (13.5) | 43.4 (13.7) | 46.2 (13.2) | 0.18 |
| Thrombus-in-transit, n (%) | 8 (2.8) | 5 (2.8) | 3 (2.7) | 0.72 |
| Electrocardiogram data, n (%) | ||||
| Sinus tachycardia | 164 (56.6) | 105 (58.7) | 59 (53.2) | 0.54 |
| Atrial fibrillation/flutter | 17 (5.9) | 13 (7.3) | 4 (3.6) | 0.30 |
| Right bundle branch block | 51 (17.6) | 33 (18.4) | 18 (16.2) | 0.75 |
| S1Q3T3 | 79 (27.2) | 47 (26.3) | 32 (28.8) | 0.89 |
| Anterior T-wave inversion | 74 (25.5) | 40 (22.3) | 34 (30.6) | 0.31 |
CTA = CT angiography, RV = right ventricle.
See Supplemental Table 5 (https://links.lww.com/CCM/H955) for missing data.
Table 3 summarizes the cohort’s procedural characteristics. Mean time to intervention was 6.0 hours in the EI group and 34.6 hours in the DI group (p < 0.001). The Inari FlowTriever was the most commonly used mechanical thrombectomy device. In total, 223 patients (76.9%) had invasive hemodynamic measurements recorded before intervention, and 207 (71.4%) after intervention. Of the total cohort, pre- and post-thrombectomy pulmonary artery systolic pressure (PASP) measurements were available in 141 patients (48.6%), while mPAP measurements were available in 96 patients (33.1%). Pre-thrombectomy PASP and mPAP were similar between groups. Post-thrombectomy, there was a greater percent reduction in PASP (–25.8 [17.0] vs. –18.9 [17.1]; p = 0.020) and mPAP (–26.8% [17.7] vs. –20.2% [19.7]; p = 0.016) in the EI group. After adjusting for PESI and CPES score, DI was associated with a significantly smaller percent reduction in PASP compared with EI (adjusted difference, 5.82%; 95% CI, 0.27–11.38; p = 0.040), but not mPAP (3.52%; 95% CI, –2.3 to 9.33; p = 0.24). There were no differences in adjuvant therapy (e.g., systemic thrombolysis, catheter-directed thrombolysis) or inferior vena cava (IVC) filter insertion.
TABLE 3.
Procedural Characteristics
| Variable | Total (n = 290) | Early Intervention (n = 179) | Delayed Intervention (n = 111) | p |
|---|---|---|---|---|
| Time of diagnosis to intervention (hr), mean (sd) | 17.0 (31.1) | 6.0 (6.4) | 34.6 (44.4) | < 0.001 |
| Mechanical thrombectomy, n (%) | ||||
| Inari FlowTriever | 241 (83.1) | 154 (86.0) | 87 (78.4) | 0.11 |
| Penumbra Indigo | 42 (14.5) | 22 (12.2) | 20 (18.0) | 0.23 |
| Other/not reported | 9 (3.1) | 4 (2.2) | 5 (4.5) | 0.31 |
| Adjuvant therapy, n (%) | ||||
| Systemic thrombolysis | 7 (2.4) | 4 (2.2) | 3 (2.7) | 0.94 |
| Catheter-directed thrombolysis | 9 (3.1) | 6 (3.4) | 3 (2.7) | |
| Other | 5 (1.7) | 3 (1.7) | 2 (1.8) | |
| Inferior vena cava filter placement, n (%) | 47 (16.2) | 32 (17.9) | 15 (13.5) | 0.41 |
| Pre-thrombectomy hemodynamics (%) | 223 (76.9) | 133 (74.3) | 90 (81.1) | 0.15 |
| Pre-PASP (mm Hg), mean (sd) | 52.3 (13.0) | 51.7 (12.1) | 53.2 (14.1) | 0.45 |
| Pre-mPAP (mm Hg), mean (sd) | 32.6 (8.6) | 32.4 (7.6) | 33.0 (10.0) | 0.58 |
| Post-thrombectomy hemodynamics (%) | 207 (71.4) | 125 (69.8) | 82 (73.9) | 0.51 |
| Post-PASP (mm Hg), mean (sd) | 40.6 (14.1) | 38.4 (13.4) | 43.6 (14.5) | 0.023 |
| Post-mPAP (mm Hg), mean (sd) | 24.6 (8.5) | 23.8 (8.4) | 25.9 (8.6) | 0.086 |
| Percent change in PASP (%), mean (sd) | n = 141 (48.6), –22.7 (17.3) | n = 93 (52.0), –25.8 (17.0) | n = 48 (43.2), –18.9 (17.1) | 0.020 |
| Percent change in mPAP (%), mean (sd) | n = 96 (33.1), –24.2 (18.7) | n = 63 (35.2), –26.8 (17.7) | n = 33 (29.7), –20.2 (19.7) | 0.016 |
mPAP = mean pulmonary artery pressure, PASP = pulmonary artery systolic pressure.
Table 4 summarizes the primary and secondary outcomes. In-hospital mortality did not differ significantly between groups in either the unadjusted (13 [7.3%] vs. 12 [10.8%] events; p = 0.39) or adjusted analyses (odds ratio [OR], 1.80; 95% CI, 0.82–3.95; p = 0.14). The DI group experienced higher rates of post-thrombectomy cardiac arrest (5 [2.8%] vs. 9 [8.1%] events; p = 0.050) and intubation (16 [8.9%] vs. 20 [18.0%] events; p = 0.028); however, after adjustment for CPES and PESI scores, only intubation was significant (adjusted OR [aOR], 2.63; 95% CI, 1.38–5.01; p = 0.003). ICU admission was more frequent in the EI group (140 vs. 74 events; p = 0.039), although this association was attenuated after adjustment (aOR, 0.58; 95% CI, 0.32–1.08; p = 0.086). A sensitivity analysis including syncope was not associated with increased odds of in-hospital mortality (aOR, 1.75; 95% CI, 0.78–3.91; p = 0.17). Additionally, a sensitivity analysis setting the intervention cutoff to 24 hours did not result in differences in mortality (aOR, 1.00; 95% CI, 0.99–1.01; p = 0.93). No significant differences were observed between groups in the use of venoarterial ECMO, vasopressors/inotropes, major bleeding, post-thrombectomy renal failure, or ICU and hospital LOS.
TABLE 4.
Primary and Secondary Outcomes
| Outcomes | Total (n = 290) | Early Intervention (n = 179) | Delayed Intervention (n = 111) | p | Adjusted OR (95% CI)a | p |
|---|---|---|---|---|---|---|
| Binary outcomes (OR), n (%) | ||||||
| In-hospital mortality | 25 (8.6) | 13 (7.3) | 12 (10.8) | 0.389 | 1.80 (0.82–3.95) | 0.140 |
| Venoarterial extracorporeal membrane oxygenation | 9 (3.1) | 4 (2.2) | 5 (4.5) | 0.311 | 1.90 (0.52–6.86) | 0.329 |
| Cardiac arrest | 14 (4.8) | 5 (2.8) | 9 (8.1) | 0.050 | 3.02 (0.97–9.40) | 0.056 |
| Intubation | 36 (12.4) | 16 (8.9) | 20 (18.0) | 0.028 | 2.63 (1.38–5.01) | 0.003 |
| Vasopressor/inotrope | 66 (22.8) | 45 (25.1) | 21 (18.9) | 0.250 | 0.77 (0.45–1.33) | 0.347 |
| ICU admission | 214 (73.8) | 140 (78.2) | 74 (66.7) | 0.039 | 0.58 (0.32–1.08) | 0.086 |
| Major bleeding | 28 (9.7) | 16 (8.9) | 12 (10.8) | 0.683 | 1.22 (0.58–2.59) | 0.598 |
| Renal failure | 5 (1.7) | 1 (0.6) | 4 (3.6) | 0.073 | b | b |
| Continuous outcomes (mean difference, d), mean (sd) | ||||||
| ICU LOS | 9.06 (53.7) | 11.12 (66.4) | 5.38 (11.9) | 0.431 | –2.16 (–10.40 to 6.09) | 0.608 |
| Hospital LOS | 8.32 (14.7) | 7.15 (9.4) | 10.19 (20.5) | 0.088 | 2.98 (–0.99 to 6.96) | 0.141 |
LOS = length of stay, OR = odds ratio.
Adjusted for Composite Pulmonary Embolism Shock and Pulmonary Embolism Severity Index scores, reference group is early intervention.
Insufficient numbers for analysis.
DISCUSSION
This multicenter, retrospective study demonstrated that timing of thrombectomy did not influence in-hospital mortality rate in intermediate-risk PE after adjusting for both PESI and CPES scores. However, EI was associated with greater reductions in PAP and intubation rates as well as a numerically lower cardiac arrest events post-intervention.
When comparing early vs. delayed mechanical thrombectomy for this patient presenting with intermediate-risk PE, our study did not demonstrate differences in mortality (13 [7.3%] vs. 12 [10.8%] events; p = 0.39). This finding remained nonsignificant after adjusting for two important risk prediction scores (aOR, 1.80; 95% CI, 0.82–3.95; p = 0.14). Presently, there is no conclusive data that has demonstrated mortality benefit of mechanical thrombectomy in intermediate-risk PE. While mortality rates were low in several mechanical thrombectomy device trials and registries, these were single-arm interventions, thus unable to establish a comparative mortality benefit (9, 10, 19). The Large-Bore Mechanical Thrombectomy vs Catheter-Directed Thrombolysis for Treatment of Intermediate-Risk Pulmonary Embolism (PEERLESS) trial compared FlowTriever to catheter-directed thrombolysis, but found no differences in mortality (12). It is possible that our study is underpowered and or the results reflect confounding given the observation study design. Ongoing randomized control trials evaluating the effect of catheter-directed interventions on mortality compared with anticoagulation will further help to define the role of these therapies in the intermediate-risk PE population (20–22).
More recently, two large observational studies examined the effects of timing of intervention on outcomes in this sample. Zhang et al (15) compared early (n = 57, < 12 hr) vs. late (n = 76, ≥ 12 hr) catheter-directed intervention in intermediate-risk PE. This study noted that EI was associated with a significantly lower rate of the primary composite endpoint, which included 30-day mortality, resuscitated cardiac arrest, hemodynamic instability, and 90-day readmission (4% vs. 18%; p < 0.001). This difference was driven by a higher rate of hemodynamic decompensation in the delayed treatment arm (0 vs. 7%; p = 0.048). Leiva et al (23) retrospectively compared patients in the Nationwide Readmissions Database undergoing early (≤ 1 d) compared with delayed catheter-based therapies (> 1 d). In contrast to our findings, this study reported a lower risk of 90-day mortality in intermediate-risk PE undergoing treatment within 1 day (hazard ratio, 0.55; 95% CI, 0.46–0.66). Notably, we elected to restrict our primary outcome to in-hospital mortality to provide a direct and interpretable measure of treatment timing effect, and to reduce bias introduced by secondary endpoints, loss to follow-up, and post-discharge care variability.
In our cohort, DI was associated with a higher rate of post-thrombectomy cardiac arrest (5 [2.8%] vs. 9 [8.1%]; p = 0.050) and peri-procedural mechanical ventilation (16 [8.9%] vs. 20 [18.0%]; p = 0.028). However, after adjustment only rate of intubation was significantly higher in the DI group (aOR, 2.63; 95% CI, 1.38–5.01; p = 0.003). A multicenter, retrospective study examining early vs. late mechanical thrombectomy for intermediate-risk PE found fewer adverse clinical events in a composite outcome (2 [4%] vs. 14 [18%]; p = 0.009) that included cardiac arrest, although cardiac arrest alone was not significantly different between the groups (0 vs. 1 [1%]; p = 0.38) (15). Rates of intubation and mechanical ventilation in intermediate-risk PE are sparsely reported in the literature, whereas hemodynamic decompensation is often a reported outcome. We were unable to adjudicate vasopressor and inotrope initiation relative to thrombectomy timing, which precluded a robust clinical-decompensation composite outcome. Intubation and mechanical ventilation are generally avoided in acute PE unless absolutely clinically necessary, and a prior randomized control trial suggested it occurred in only 2–3% of intermediate-risk patients (24). In contrast, a higher proportion of our entire cohort required intubation (36 [12.4%]). Thus, our sample may represent the higher acuity subgroup within intermediate-risk PE, as the majority were categorized as intermediate-/high-risk with elevated PESI and sPESI scores. The higher prevalence intubation in our cohort may indicate a population with greater clinical instability, although causality cannot be inferred. More data is needed on the incidence and clinical implications of mechanical ventilation in this population.
EI was associated with a greater percentage decrease in both PASP (–25.8% vs. –18.9%; p = 0.020) and mPAP compared with DI (–26.8% vs. –20.2%; p = 0.016). After adjusting for important covariates, DI had a significantly smaller percent reduction in PASP (5.82%; 95% CI, 0.27–11.38; p = 0.040). PAP, RV/LV ratio, and RV echocardiographic parameters have been used as clinical surrogate markers in PE treatment literature. The Ultrasound Accelerated Thrombolysis of Pulmonary Embolism (ULTIMA) trial compared USAT with anticoagulation alone and demonstrated improvements in RV dilatation in patients treated with USAT within 4 hours of diagnosis (25). Edla et al (26) evaluated 41 patients treated with USAT for submassive PE and noted patients receiving intervention within 24 hours of diagnosis had greater improvements in cardiac index, pulmonary vascular resistance, and shorter LOS, but not RV/LV diameter ratio and mean PAP, when compared with later intervention. Lehr et al (14) found catheter-directed interventions within 24 hours of emergency department admission to be associated with greater reductions in PASP (14 vs. 6 mm Hg) following CDT compared with delayed treatment. An analysis from the FlowTriever All-Comer Registry for Patient Safety and Hemodynamics (FLASH) study also noted that shorter time to thrombectomy was associated with greater reduction in PAP, RV/LV ratio, and associated with longer 6-month 6-minute walk test distance (27). Consistent with these findings, we found EI was associated with greater reductions in PAPs. However, this finding did not translate to reductions in mortality. Further large cohort studies are required to fully assess this interaction.
This study is novel for several reasons. Prior investigations of intervention timing in PE have been limited by significant heterogeneity in treatment modality. For example, in Lehr et al (14), only 19 patients (29.7%) underwent mechanical thrombectomy, while in Zhang et al (15), 92 patients (69.2%) received this therapy. Neither study was powered to evaluate whether outcomes differed by catheter-directed thrombolysis vs. mechanical thrombectomy. In contrast, our study is the first to specifically examine the timing of mechanical thrombectomy in intermediate-risk PE. Defining the optimal timing of distinct interventional modalities is critical, as their therapeutic effects may differ in immediacy and durability. To our knowledge, this is also the largest multicenter study using direct chart review across five large academic medical centers, which enabled accurate adjudication of exposures and outcomes, incorporation of validated risk scores, and precise classification of EI vs. DI. Last, selecting in-hospital mortality as the primary outcome provided the most clinically meaningful measurement of treatment effect. While imaging outcomes such as RV strain or secondary outcomes like change in PAP are important, they do not carry the same weight as mortality and could dilute this effect. Taken together, these strengths support the robustness of our findings and underscore the unique contribution of this work to the evolving literature on PE management.
This study has several limitations. First, our retrospective design inherently limits our ability to account for unmeasured biases and confounders relevant to the timing of thrombectomy. Specifically, we were unable to control for all clinical variables that influence the decision-making process of early vs. late intervention, such as renal dysfunction or institutional workflow variations. PE risk was defined at time of diagnosis, and our study could not account for progression in PE severity, which could influence timing of intervention. Additionally, variability in radiographic and echocardiographic image acquisition, interpretation, and reporting across providers and institutions may have introduced heterogeneity and bias. Presence of early echocardiographic findings of right heart strain may have influenced treatment decisions, but we could not account for this variable. Furthermore, because echocardiographic parameters were variably reported across centers, this limited our ability to evaluate the relationships between RV function, intervention timing, and outcomes. Second, incomplete data limits the interpretability of these results. Numerous patients did not have both pre- and post-thrombectomy PAPs, which may limit the interpretation of this secondary outcome. Third, while infrequent, adjuvant therapies such as catheter-directed thrombolysis and systemic thrombolytics may confound the results. Fourth, we were unable to consistently determine timing of vasopressor and inotrope use across all sites, which limited the ability to use clinical deterioration as an outcome. Fifth, time from symptom onset to thrombectomy may be a more relevant timeframe for analysis. However, documentation of symptom onset was often imprecise; time from imaging diagnosis or PERT activation were often more measurable time points and thus used for this analysis. Last, while our mortality rates are comparable to prior reports of intermediate-risk PE outcomes, the CI around our estimates were wide, suggesting that our study may have been underpowered to detect such mortality effects. A larger cohort would be needed to determine the effect of timing on mortality.
CONCLUSIONS
In patients presenting with intermediate-risk PE, early compared with delayed mechanical thrombectomy did not result in significant different in-hospital mortality. Early treatment may result in a lower incidence of peri-procedural intubation and greater reductions in PASP and mPAP compared with delayed treatment. Further large-scale randomized trials are necessary to determine the optimal timing of mechanical thrombectomy for acute intermediate-risk PE.
Supplementary Material
Footnotes
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health’s National Center for Advancing Translational Sciences.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (http://journals.lww.com/ccmjournal).
This research was supported by the National Institutes of Health’s National Center for Advancing Translational Sciences, grant UM1TR004405. Supported by grant from the Jennifer Luft Award—National Blood Clot Alliance.
Drs. Chiang’s and Bernal’s institutions received funding from the National Blood Clot Alliance. Dr. Chiang received support for article research from the National Center for Advancing Translational Sciences (UM1TR004405). Dr. Sanfilippo’s institution received funding from the National Heart, Lung, and Blood Institute (R01) and the American Society of Hematology (ASH); she received funding from the ASH and Quinn Johnston. Drs. Schaefer and Patell received funding from Sanofi. Dr. Schaefer received funding from Novartis, Pharmacosmos, and Pfizer. Dr. Patell received funding from Merck Research and Octapharma. Dr. Baumann Kreuziger’s institution received funding from CSL Behring, Takeda, Veralox, and Sanofi. Drs. Freese and Bernal received support for article research from the National Institutes of Health. The remaining authors have disclosed that they do not have any potential conflicts of interest.
Contributor Information
Caleb J. Chiang, Email: calebjchiang@gmail.com.
Kelsey E. Bria, Email: bria.k@wustl.edu.
Pavlina Chrysafi, Email: chrysap1@mskcc.org.
Virginia Falvello, Email: vfalvello@gmail.com.
Kristen M. Sanfilippo, Email: ksanfilippo@wustl.edu.
Jordan K. Schaefer, Email: jschaef@med.umich.edu.
Kenneth Bauer, Email: kbauer@bidmc.harvard.edu.
Rushad Patell, Email: rpatell@bidmc.harvard.edu.
Lisa Baumann Kreuziger, Email: lisakreuziger@versiti.org.
Rebecca L. Freese, Email: frees048@umn.edu.
Alejandra Gutierrez Bernal, Email: agb@umn.edu.
REFERENCES
- 1.Office of the Surgeon General (US), National Heart, Lung, and Blood Institute (US): The Surgeon General’s Call to Action to Prevent Deep Vein Thrombosis and Pulmonary Embolism. Rockville, MD, Office of the Surgeon General (US), 2008 [PubMed] [Google Scholar]
- 2.Giri J, Sista AK, Weinberg I, et al. : Interventional therapies for acute pulmonary embolism: Current status and principles for the development of novel evidence: A scientific statement from the American Heart Association. Circulation 2019; 140:e774–e801 [DOI] [PubMed] [Google Scholar]
- 3.Konstantinides SV, Meyer G, Becattini C, et al. ; The Task Force for the diagnosis and management of acute pulmonary embolism of the European Society of Cardiology (ESC): 2019 ESC guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS): The task force for the diagnosis and management of acute pulmonary embolism of the European Society of Cardiology (ESC). Eur Respir J 2019; 54:1901647. [DOI] [PubMed] [Google Scholar]
- 4.Bĕlohlávek J, Dytrych V, Linhart A: Pulmonary embolism, part I: Epidemiology, risk factors and risk stratification, pathophysiology, clinical presentation, diagnosis and nonthrombotic pulmonary embolism. Exp Clin Cardiol 2013; 18:129–138 [PMC free article] [PubMed] [Google Scholar]
- 5.Patel N, Patel NJ, Agnihotri K, et al. : Utilization of catheter-directed thrombolysis in pulmonary embolism and outcome difference between systemic thrombolysis and catheter-directed thrombolysis. Catheter Cardiovasc Interv 2015; 86:1219–1227 [DOI] [PubMed] [Google Scholar]
- 6.Kuo WT, Banerjee A, Kim PS, et al. : Pulmonary embolism response to fragmentation, embolectomy, and catheter thrombolysis (PERFECT): Initial results from a prospective multicenter registry. Chest 2015; 148:667–673 [DOI] [PubMed] [Google Scholar]
- 7.Piazza G, Hohlfelder B, Jaff MR, et al. ; SEATTLE II Investigators: A prospective, single-arm, multicenter trial of ultrasound-facilitated, catheter-directed, low-dose fibrinolysis for acute massive and submassive pulmonary embolism: The SEATTLE II study. JACC Cardiovasc Interv 2015; 8:1382–1392 [DOI] [PubMed] [Google Scholar]
- 8.Zhang RS, Maqsood MH, Sharp ASP, et al. : Efficacy and safety of anticoagulation, catheter-directed thrombolysis, or systemic thrombolysis in acute pulmonary embolism. JACC Cardiovasc Interv 2023; 16:2644–2651 [DOI] [PubMed] [Google Scholar]
- 9.Tu T, Toma C, Tapson VF, et al. ; FLARE Investigators: A prospective, single-arm, multicenter trial of catheter-directed mechanical thrombectomy for intermediate-risk acute pulmonary embolism: The FLARE study. JACC Cardiovasc Interv 2019; 12:859–869 [DOI] [PubMed] [Google Scholar]
- 10.Sista AK, Horowitz JM, Tapson VF, et al. ; EXTRACT-PE Investigators: Indigo aspiration system for treatment of pulmonary embolism: Results of the EXTRACT-PE trial. JACC Cardiovasc Interv 2021; 14:319–329 [DOI] [PubMed] [Google Scholar]
- 11.Sławek-Szmyt S, Stępniewski J, Kurzyna M, et al. : Multicentre, real-world data of next-generation computer-assisted vacuum aspiration thrombectomy in acute pulmonary embolism. Respir Res 2025; 26:87–11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Jaber WA, Gonsalves CF, Stortecky S, et al. ; PEERLESS Committees and Investigators*: Large-bore mechanical thrombectomy versus catheter-directed thrombolysis in the management of intermediate-risk pulmonary embolism: Primary results of the PEERLESS randomized controlled trial. Circulation 2025; 151:260–273 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Rivera-Lebron B, McDaniel M, Ahrar K, et al. ; PERT Consortium: Diagnosis, treatment and follow up of acute pulmonary embolism: Consensus practice from the PERT consortium. Clin Appl Thromb/Hemost 2019; 25:1076029619853037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lehr A, Guichet P, Garimella B, et al. : Impact of time to intervention on catheter-directed therapy for pulmonary embolism. Crit Care Explor 2023; 5:e0828. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zhang RS, Yuriditsky E, Zhang P, et al. : Impact of time to catheter-based therapy on outcomes in acute pulmonary embolism. Circ Cardiovasc Interv 2025; 18:e014499. [DOI] [PubMed] [Google Scholar]
- 16.Aujesky D, Obrosky DS, Stone RA, et al. : Derivation and validation of a prognostic model for pulmonary embolism. Am J Respir Crit Care Med 2005; 172:1041–1046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Jiménez D, Aujesky D, Moores L, et al. ; RIETE Investigators: Simplification of the pulmonary embolism severity index for prognostication in patients with acute symptomatic pulmonary embolism. Arch Intern Med 2010; 170:1383–1389 [DOI] [PubMed] [Google Scholar]
- 18.Zhang RS, Yuriditsky E, Zhang P, et al. : Composite pulmonary embolism shock score and risk of adverse outcomes in patients with pulmonary embolism. Circ Cardiovasc Interv 2024; 17:e014088. [DOI] [PubMed] [Google Scholar]
- 19.Toma C, Jaber WA, Weinberg MD, et al. : Acute outcomes for the full US cohort of the FLASH mechanical thrombectomy registry in pulmonary embolism. EuroIntervention 2023; 18:1201–1212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Klok FA, Piazza G, Sharp ASP, et al. : Ultrasound-facilitated, catheter-directed thrombolysis vs anticoagulation alone for acute intermediate-high-risk pulmonary embolism: Rationale and design of the HI-PEITHO study. Am Heart J 2022; 251:43–53 [DOI] [PubMed] [Google Scholar]
- 21.Giri J, Mahfoud F, Gebauer B, et al. : PEERLESS II: A randomized controlled trial of large-bore thrombectomy versus anticoagulation in intermediate-risk pulmonary embolism. J Soc Cardiovasc Angiogr Interv 2024; 3:101982. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.ClinicalTrials.gov: STORM-PE: A Prospective, Multicenter, Randomized Controlled Trial Evaluating Anticoagulation Alone vs Anticoagulation Plus Mechanical Aspiration With the Indigo® Aspiration System for the Treatment of Intermediate High Risk Acute Pulmonary Embolism. Penumbra. 2024. Available at: https://clinicaltrials.gov/study/NCT05684796. Accessed December 31, 2023
- 23.Leiva O, Rosovsky RP, Alviar C, et al. : Early versus delayed catheter-based therapies in patients hospitalised with acute pulmonary embolism. EuroIntervention 2025; 21:e463–e470 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Meyer G, Vicaut E, Danays T, et al. ; PEITHO Investigators: Fibrinolysis for patients with intermediate-risk pulmonary embolism. N Engl J Med 2014; 370:1402–1411 [DOI] [PubMed] [Google Scholar]
- 25.Kucher N, Boekstegers P, Müller OJ, et al. : Randomized, controlled trial of ultrasound-assisted catheter-directed thrombolysis for acute intermediate-risk pulmonary embolism. Circulation 2014; 129:479–486 [DOI] [PubMed] [Google Scholar]
- 26.Edla S, Rosman H, Neupane S, et al. : Early versus delayed use of ultrasound-assisted catheter-directed thrombolysis in patients with acute submassive pulmonary embolism. J Invasive Cardiol 2018; 30:157–162 [PubMed] [Google Scholar]
- 27.Balderman J: Shorter time to mechanical thrombectomy is associated with improved outcomes in patients with acute pulmonary embolism: Results from the FLASH registry. J Vasc Surg Venous Lymphat Disord 2024; 12:101787 [Google Scholar]