Over a 24-month period between March 2018 and March 2020), 75 hospitalized patients with acute PE were screened, 29 of whom were consecutive PE patients undergoing percutaneous mechanical thrombectomy for treatment of proximal location, intermediate-high- or high-risk acute PE, were enrolled (Figure 1: Flow chart). The study was approved by the institutional ethics committee, and the study protocol was written according to the ethical guidelines of the 1975 Declaration of Helsinki. Written informed consent was obtained from all patients before the percutaneous procedure.

Figure 1.

Treatment modalities recommended by our institutional Pulmonary embolism response team after its activation and flow chart of the study.

AC: anticoagulation; CDT: catheter-directed mechanical aspiration thrombectomy; FUP: follow-up; PE: pulmonary embolism; PERT: pulmonary embolism response team; pts: patients;

* Reasons for proposing a conservative strategy with isolated AC: PE with >14 days of onset of symptoms – acute on chronic PE (2 pts); intermediate-low-risk PE (1 pt); intermediate-high-risk PE without clinical signs of decompensation with isolated AC (23 pts); active bleeding or contraindication for AC (4 pts) and poor prognosis or cognitive impairment (2 pts).

(0.35MB).

Patients were referred to thrombectomy if aged 18 years or older. Acute PE was considered as symptoms duration equal to or less than 14 days. A computed tomography pulmonary angiography (CTPA) was performed for confirmation of diagnosis, and the presence of intraluminal filling defect in at least one main or lobar pulmonary artery defined the location as proximal.

Prognostic stratification as high- or intermediate-high-risk was performed according to the guidelines published jointly by the European Society of Cardiology (ESC) and the European Respiratory Society (ERS) on the diagnosis and management of acute PE in 2019.7

According to an institutional protocol, based on this ESC/ERS recommendations,7 patients were eligible for thrombectomy if they met one of the following criteria: high-risk PE and absolute or relative contraindication to systemic thrombolysis, high-risk PE and systemic thrombolysis failure (rescue reperfusion therapy) and intermediate high-risk PE with at least one early sign of hemodynamic decompensation. To identify a higher risk cohort of intermediate PE patients, we used a “3+1” rule, previously described, that recognizes patients who may benefit from invasive therapies but do not meet the threshold for systemic thrombolytic therapy17 (see Supplementary Figure 1). Patients were excluded if they presented contraindications for therapeutic anticoagulation (such as active bleeding) or had thrombus in the right cardiac chambers.

Patients were referred for percutaneous procedure from their own institution or other national hospitals. All the referred high-risk and intermediate-high-risk PE cases were discussed by the hospital pulmonary embolism response team (PERT), which is available 24 hours a day, seven days a week; the decision to treat was based on consensus from multidisciplinary PERT.

Data on patient demographics, medical history, symptoms and physical examination were obtained. Together with results of blood analysis, 12-lead electrocardiogram, transthoracic echocardiogram (TTE), CTPA, anticoagulation therapy, pulmonary angiographic findings, procedural characteristics and invasive hemodynamic measures were collected. In-hospital and three-month outcomes were obtained. All this information was provided by the referring physician and obtained from electronic medical records.

Based on recorded data, validated prognostic stratification scores from the original and simplified Pulmonary Embolism Severity Index (PESI) were computed.18 The Charlson Comorbidity Index score was also calculated for each patient.19

As a first step, we performed an invasive measurement of pulmonary artery pressures (PAP) followed by an initial pulmonary angiogram to demonstrate the filling defects. Mechanical thrombectomy with an 8-French continuous aspiration mechanical thrombectomy catheter (Indigo CAT8, XTORQ, Penumbra®, Alameda, California) connected to a suction pump (Pump MAX & MAX Canister, Penumbra®, Alameda, California) allowed clot aspiration through negative pressure (20–40 cm H2O) exertion (see Supplementary Table 1 for details of the procedure).

Use of adjunctive in situ thrombolytic therapy (combined method) was according to operator discretion. In these cases, a 5-French pigtail catheter (Cordis®, Miami Lakes, USA) was used for an intravenous drug bolus injection.

At the end of the percutaneous procedure, we performed a final selective pulmonary angiography and reassessed the hemodynamic parameters. Figures 2 and 3 illustrate different situations in which the technique was applied.

Figure 2.

This is the case of a 79-years-old woman admitted with an acute high-risk PE, with contraindication for fibrinolysis due to the presence of a paravertebral mass with medullary compression (A), later diagnosed as a lymphoma (*), underwent endovascular treatment. Axial contrast enhanced computed tomography at time of diagnosis illustrates central extensive bilateral thrombus (red arrows in B and C). Echocardiography at bedside shows an enlarged right ventricle and a flattened interventricular septum in parasternal short axis view (D); Left pulmonary angiography confirm nearly complete obstruction at the origin (E). Aspiration mechanical thrombectomy was performed with the Indigo® CAT8 with improvement of arterial perfusion (F) and venous return (G) of the left lung. Image H displays the extracted clots.

There was a decrease in systolic and mean pulmonary artery pressure of 54 and 33 mmHg pre-procedure (I), respectively, to 34 and 20 mmHg post-procedure (J).

(0.74MB).

Figure 3.

This is the case of an 86-years old woman who had a recent ischemic stroke and was admitted with an acute intermediate-high-risk PE, catheter-directed therapy proposed due to respiratory failure (paO2/FiO2 ratio=170), with no response to parenteral anticoagulation. a) and b) Axial contrast enhanced computed tomography at time of diagnosis illustrates central extensive bilateral thrombus (red arrows) and huge dilatation of right ventricle (RV/LV ratio of 2.1). c) Selective pulmonary angiography confirms obstruction of left inferior lobar artery and lingula. d) Continuous aspiration mechanical thrombectomy was performed with Indigo® CAT8. A Penumbra Indigo System Separator SEP8 device was used through the aspiration catheter to facilitate clot aspiration by preventing the catheter from clogging in the extensive thrombus. e) Post-procedure we observe a near restoration of the normal perfusion in the left lung. f) and g) There was a decrease in systolic and mean pulmonary artery pressure of 65 and 33 mmHg pre-procedure, respectively, to 42 and 25 mmHg post-procedure. PaO2/FiO2 ratio increased to 371 at 48 hours after intervention.

(0.75MB).

Continuous variables are expressed as mean ± standard deviation or median ± interquartile range (IQR) if distributions are skewed. Categorical variables are summarized as frequency and percentage for each data category. Missing data were not imputed for the efficacy and safety analyses. Comparisons of clinical, laboratorial, echocardiographic and hemodynamic characteristics before and after the procedure were conducted with the two-sided paired Student's t-test or Wilcoxon's non-parametric test (if continuous data), when appropriate, or the two-sided Fisher exact test and McNemar's test (if categorical variables). All reported p values were two-sided and a p<0.05 was considered statistically significant. Statistical analysis was performed with Statistical Package for Social Sciences (SPSS) software, version 25.0 (IBM Corp., Armonk, NY, USA).

Among the 29 patients, mean age was 67.2±14.4 years, and 72.4% were women. Hypertension (72.4%) was the most frequent comorbidity. Common risk factors for PE comprised obesity (body mass index of greater than 30 kg/m2 in 34.8%), immobility within 30 days of PE diagnosis (44.8%) and active cancer (20.7%). History of previous venous thromboembolism was present in 24.1%. The baseline demographics are outlined in Table 1.

Dyspnea was the main presenting symptom (58.6%). Average mean blood pressure, HR and peripheral oxygen saturation values were, respectively, 75.3±17.9 mmHg, 111.3±20.6 bpm and 86.0±18.5%. In 96.6% of patients, CTPA showed bilateral arterial pulmonary involvement. Regarding laboratory findings, mean lactate blood level was 2.6±1.4 mmol/L, median levels of high-sensitivity cardiac T troponin was 76.0 ng/L and NT-proBNP was 7898.0 pg/ml. The electrocardiogram showed repolarization changes consistent with RV overload in 72.4%, complete right bundle branch block in 31.0% and the typical pattern S1Q3T3 in 34.5%. The initial TTE revealed the presence of RV dilation (RV/LV ratio >0.9) in all patients and RV dysfunction in 79.3%. The mean systolic PAP was 61.8±17.3 mmHg and the mean tricuspid annular plane excursion (TAPSE) was 14.5±3.7 mm.

Acute intermediate-high-risk PE and high-risk PE were observed in 51.7% and 48.3% of patients, respectively. In the subgroup of patients with high-risk PE, 71.4% had absolute contraindications for systemic thrombolysis (history of hemorrhagic stroke in two patients, central nervous system neoplasm in two patients and major trauma, surgery or head injury in the previous three weeks in six patients); 28.6% had undergone failed systemic thrombolysis. Thirteen patients were receiving inotropic support before treatment, and three required mechanical ventilation.

More than half of the patients were in class IV or V of the original PESI score (58.6%) presenting a high or very high-risk of 30-day mortality. The median Charlson Comorbidity Index score for the cohort was 4.0 (IQR 3.0-6.0).

Clinical and pulmonary embolism characteristics are presented in Table 1.

Procedural characteristics are summarized in Table 2. Right femoral venous access was the most frequently used (82.8%). Bilateral pulmonary intervention was performed in 72.4% of patients. In situ thrombolytic therapy (bolus injection of alteplase) via catheter was administered to four patients (13.8%) with a median total dose of 10.0 mg (range from 5.0-15.0 mg). This was given at the investigator's discretion because of perceived major thrombus burden and not because of clinical decompensation.

The mean fluoroscopy time was 40.5±15.8 minutes and intra-procedural blood loss was 150-900 ml.

The Indigo CAT8 XTORQ catheter was successfully positioned within the thrombus in 28 patients, achieving 96.6% technical success. In one patient, the catheter was not correctly placed in the right main pulmonary artery owing to catheter kinking at the level of the main pulmonary artery bifurcation due to its marked angulation.

Miller angiographic obstruction index score decreased from 21.9±7.2 at baseline to 16.4±6.5 at procedure completion, revealing a significant reduction in both obstruction and perfusion indexes (p<0.001) (Figure 4). The average of invasively measured systolic PAP decreased from 63.2±19.8 mmHg at baseline to 53.0±17.7 mmHg at procedure completion (mean difference, -10.2 mmHg; p<0.001). Acute procedural success was achieved in 75.9%. A relative reduction of at least 50% of MS was only achieved in 10.3% of the patients.

Figure 4.

Changes in the obstruction index, perfusion index and Miller Score before and after treatment evaluated by angiography analysis.

(0.13MB).

Considering in-hospital efficacy outcomes, a statistically significant improvement was observed in all of the hemodynamic and oxygenation parameters (Table 3). There was a significant elevation of mean paO2/FiO2 ratio and a significant reduction of shock index. The need for aminergic support significantly reduced from 44.8% to 10.3% of patients in the first 48 hours after the procedure (p=0.006) and RV function improved in 66.6% of the 14 patients who underwent TTE 48 hours after the procedure (p=0.008). Globally, clinical success was 75.9% (10 patients with high-risk and 12 with intermediate high risk PE).

Concerning primary safety outcomes (Table 4), the major adverse event rate was 10.3%, resulting from the occurrence of two potential procedure-related deaths (as related procedural complications cannot be excluded) and one pulmonary macroembolization during the device progression, followed by cardiorespiratory arrest with need for cardiopulmonary resuscitation, mechanical ventilation and systemic thrombolysis (this patient survived the procedure, but died on day 12 of hospitalization with active urinary cancer). Regarding intra-procedure deaths: One patient died during the procedure due to multiorgan failure in the evolution of cardiogenic shock over several hours; the other patient died at the end of the procedure; PE recurrence was considered the probable cause of death. There were no other treatment-related events, nor any severe nor life-threatening bleeding. Secondary safety outcome occurrence was 17.2%: three patients needed a red cell transfusion within 48 hours without visible bleeding, and two patients had mild hemoptysis during the procedure with no need of percutaneous or surgical intervention or mechanical ventilation.

Five patients died while hospitalized (in-hospital mortality rate of 17.2%). Three of these deaths are described above, however two additional patients died due to mesenteric ischemia and pneumonia on days 15 and 34 of hospitalization, respectively.

The median length of hospital stay was 12.0 (IQR 16.0) and the median intensive care unit stay was 5.0 days (IQR 6.5).

Clinical follow-up was completed in 95.8% of patients who survived to hospital discharged (23 patients). Three patients died before completing three months from hospital discharge due to active cancer (Table 5).

At three months post-procedure, one patient was lost to follow-up and 20 patients were alive. Of these patients, 15 (75.0%) were in World Health Organization functional class I or II. Median value of NT-proBNP was significantly reduced compared to the value before the procedure: median 6833 (IQR 9206) vs 176 (IQR 230); p=0.003. Four patients (20.0%) were diagnosed with pulmonary hypertension confirmed by right heart catheterization, one of them was diagnosed with pulmonary arterial hypertension (group 1) related to an ostium secundum atrial septal defect; the remaining three were diagnosed with chronic thromboembolic pulmonary hypertension (CTEPH, group 4). Of these three patients, two had a mean PAP of 50 mmHg before the mechanical thrombectomy and other features of pre-existing pulmonary hypertension. Only one patient (5.0%) developed chronic thromboembolic disease without rest pulmonary hypertension, and another one (5.0%) presented recurrent PE owing to anticoagulation therapy non-compliance. One temporary inferior vena cava filter was implanted after the procedure in one patient with myeloid angiopathy who had a contraindication to long-term anticoagulation therapy.

The present study had several limitations, including a limited sample size from a single center, its observational nature and lack of information about long-term clinical outcomes. In addition, echocardiographic evaluation at 48 hours post-procedure was not performed in all patients due to the retrospective nature of the study and because several cases were referred from other hospitals.