How are you taking part in this consultation?

You will not be able to change how you comment later.

You must be signed in to answer questions

    The content on this page is not current guidance and is only for the purposes of the consultation process.

    Interventional procedure overview of percutaneous thrombectomy for massive pulmonary embolism

    Closed for comments This consultation ended on at   Request commenting lead permission

    Evidence summary

    Population and studies description

    This interventional procedures overview is based on 531 patients and 67 'events' from 2 single-arm trials (1 trial included with its sub-study), 1 safety database review, 1 retrospective comparative study and 1 prospective registry. Of these 531 patients, 501 patients had the procedure and 67 events in the MAUDE database referred to this procedure. This is a rapid review of the literature, and a flow chart of the complete selection process is shown in figure 1. This overview presents 6 studies as the key evidence in table 2 and table 3, and lists 29 other relevant studies in table 5.

    Of the 6 included studies, all were from the USA. Of the 6 studies, all reported follow-up outcomes of various durations except the MAUDE database review. Four reported 30-day follow-up periods and 1 reported 6-month follow-up.

    All studies reported inclusion criteria, but these varied, as did terminology for the level of risk of PE in participants. Excluding the MAUDE database review which did not specify, 3 studies referred to intermediate or submassive PEs only and 2 included both intermediate and high-risk PEs. Across the 5 studies reporting age of participants, the average age ranged from 55.6 to 73.8 with a slight majority of male patients. Table 2 presents study details.

    Figure 1 Flow chart of study selection

    Table 2 Study details

    Study no.

    First author, date

    country

    Patients (male: female)

    Age

    Study design

    Inclusion criteria

    Intervention

    Follow-up

    1

    Toma, 2022

    USA

    N=250

    (131:119)

    Mean

    60.9

    Prospective multicentre registry in real-world population (single-arm)

    Age >18, clinical signs and symptoms of acute PE with evidence proximal filling defect in at least 1 main/lobar PA and who were undergoing treatment with FlowTriever.

    6.8% high-risk and 93.2% intermediate-risk

    FlowTriever System (Inari Medical)- percutaneous mechanical thrombectomy

    48 hours, 30 days and 6 months

    2

    Sista, 2021

    USA

    N=119

    66:53

    Mean 59.8

    Prospective, single-arm, multicentre investigational device exemption trial

    Submassive PE

    1. Clinical signs and symptoms consistent with acute PE with duration of 14 days or less. Evidence of PE must be from CTPA

    2. Systolic BP ≥90 mmHg with evidence of dilated RV with an RV/LV ratio >0.9

    3. 18 years of age or older

    Penumbra Indigo aspiration system- suction embolectomy device

    Intraprocedural, at 48 h, at discharge, and at 30 days

    3

    Tu, 2019

    USA

    N=104

    (56:48)

    Mean 55.6

    Prospective single-arm multicentre investigational device exemption trial

    Intermediate-risk patients.

    Ages 18 to 75, PE symptoms ≤14 days, symptomatic, CT-documented proximal PE, haemodynamically stable (no vasopressor requirement, heart rate <130, systolic blood pressure ≥90 mmHg at baseline assessment) and RV/LV ratios ≥0.9 (on basis of CT). Acute intermediate-risk PE.

    FlowTriever System (Inari Medical)- percutaneous mechanical thrombectomy

    48 hours and 30 days

    4

    Buckley, 2022

    USA

    N=58:

    Mechanical thrombectomy (MT) N=28 (46% Female)

    Routine care (RC) N=30 (67% Female)

    MT 68.8±14.3

    RC 73.8±12.7

    Retrospective single-centre comparative study

    Pulmonary Embolism Severity Index (PESI) score of 4 or 5 and European Society of Cardiology (ESC) classification of intermediate-high or high risk; acute, central PE (defined as thrombus within the pulmonary trunk, left/right main pulmonary artery, truncus anterior, or interlobar pulmonary artery); RV:LV ratio >1; and who were treated as inpatients.

    Additionally technical success (successful delivery of the device to the pulmonary arteries with extraction of clot) for the MT group.

    80% of RC group treated with anticoagulation alone, 13% with systemic thrombolysis and 7% with anticoagulation and CDT.

    FlowTriever System (Inari Medical)- percutaneous mechanical thrombectomy

    To 30 days

    5

    Jaber, 2020

    USA

    (sub-study of Tu, 2019)

    N=76

    44:32

    Median age 56

    Multi-centre single arm prospective trial: sub-study

    Acute intermediate-risk PE. Emergency department (ED) patients with acute PE (that is diagnosed with PE in ED) and RV/LV ratio ≥0.9 enrolled in the FLARE study (Inclusion criteria for this study seen listed in Tu, 2019).

    FlowTriever System (Inari Medical)- percutaneous mechanical thrombectomy

    48h, 30 days

    6

    Sedhom, 2021

    USA

    N=67 events (over 20 months)

    Not documented

    Manufacturer and User Facility Device Experience (MAUDE) FDA database review (real world data)

    Reports related to the use of the device in the pulmonary vasculature. MAUDE database reporting is either mandatory (for manufacturers and device user facilities) or voluntary (for healthcare professionals, patients, and consumers)

    Penumbra Indigo aspiration system- suction embolectomy device

    No follow-up

    Table 3 Study outcomes

    First author, date

    Efficacy outcomes

    Safety outcomes

    Toma, 2022

    Secondary endpoints: intra-procedural changes in haemodynamics, markers of cardiac size and function at follow up as measured by echocardiography, thrombectomy time, estimated blood, length of stay, and dyspnoea:

    mPAP: statistically significant intra-procedural reduction of 7.1 mmHg (22.2%, p<0.001).

    Cardiac index: overall no statistically significant change (n=202) pre- and post-procedure, but there was a statistically significant change in the low baseline cardiac index group: statistically significantly improved by 13.3% on‐table from 1.7±0.2 to 1.9±0.4 l/min/m2 (p=0.005).

    Heart rate: decreased statistically significantly during hospitalisation by 13.5 bpm (12.6%, p<0.0001).

    Dyspnoea: decreased statistically significantly in those dyspnoeic before thrombectomy from 2.9±1.1 preprocedure to 1.4±1.3 at 48 h (p<0.001).

    Mean length of stay: 3 days post-procedure. 56.8% did not require ICU post-procedure.

    Mean RV/LVratio: statistically significantly decreased by 0.36±0.76 (28.3%) from 1.27±0.26 pre‐procedure to 0.91±0.75 at follow‐up (p<0.001) (n=86).

    RV systolic pressure: statistically significantly decreased by 19.1±15.6 mmHg (35.8%) from 53.3±14.2 mmHg pre‐procedure to 33.5 ± 11.9 mmHg at

    follow‐up (p<0.001).

    Primary endpoint: Major adverse event rate (MAE)- composite of MAE within 48 h of the index procedure consisting of device‐related death, major bleeding, and device‐ or procedure‐related adverse events:

    3 MAEs (1.2%): all major bleeds that resolved without sequalae following transfusion.

    Intraprocedural device‐ or procedure‐related adverse events:

    No device-related injuries, clinical deteriorations or deaths at 48 h.

    Median estimated blood loss 255.0 ml (100 to 425ml).

    Secondary endpoints:individual components of the MAE composite, major access-site complications, all‐cause mortality through 30 days, and device‐related serious adverse events (SAE) within 30 days:

    All-cause mortality was 0.4% at 30 days (single death unrelated to PE).

    One access-site complication (0.4%): haematoma in patient who received thrombolytics as well.

    12 other non-device-related SAEs at 30 days.

    30-day readmission rate: 13/216 (6.0%), only one of which (0.5%) related to PE.

    Sista, 2021

    Primary endpoint: change in RV/LV ratio from baseline to 48 h post-procedure (computed tomography angiography)

    Mean RV/LV ratio reduction 0.43 (95% CI 0.38 to0.47; p<0.0001), representing 27.3% reduction (95% CI 24.83 to 29.67%).

    Secondary endpoints:

    Intraprocedural thrombolytics used in 2 patients (1.7%)

    73 required ICU stay (61%) and median stay was 1.0 day

    mPAP reduction post-aspiration: 4.3 mmHg (95% CI: 2.6 to 5.9 mm Hg; 7.9% reduction; p<0.0001)

    mPAP reduction post-procedure: 4.7 mmHg (95% CI: 3.0 to 6.4 mmHg; 8.7% reduction; p<0.0001)

    Mean reduction in CT obstruction index from pre-procedure to 48-h follow-up was 11.3% (p<0.0001).

    Primary endpoint: composite of 48-h MAEs: device-related death, major bleeding, and device-related serious adverse events (clinical deterioration, pulmonary vascular, or

    cardiac injury)

    2 (1.7%, 95% CI 0.0 to 4.0%) patients experienced 3 MAEs:

    1 patient experienced haemoptysis and access-site bleed, post-procedure death (ventricular tachycardia likely related to RV ischaemia from RV overload and haemorrhage).

    1 patient experienced access-site bleeding only.

    Primary endpoint: Device-related SAEs within 48 h = 0.8% (95% CI 0.0 to 2.5%), a composite of:

    device-related clinical deterioration within 48 h 0.8% (1/119), device-related pulmonary vascular injury within 48 h 0.8% (1/119), device-related cardiac injury within 48 h 0%.

    Secondary safety endpoints:

    At 48 h rates of:

    cardiac injury 0%, pulmonary vascular injury 1.7%, clinical deterioration 0.8%, major bleeding 1.7%, and device-related death 0.8%.

    At 30 days rates of:

    Any cause-mortality 2.5% (95% CI 0.0 to 5.3%), device-related SAEs 1.7%, symptomatic recurrence of PE 0%.

    73.1% of patients had an estimated overall blood loss <400ml, 26.9% had >400ml but none required transfusion.

    3 (2.5%) patients required transfusion related to the procedure.

    Tu, 2019

    Primary effectiveness endpoint: change in RV/LV ratio from baseline to 48h (±8 h or discharge):

    RV/LV ratio 1.56 at baseline, 1.15 at 48 h, average reduction of 0.38 (p<0.0001) that is 25.1%.

    Mean pulmonary artery pressure (mPAP): 29.8 mmHg pre-procedure vs 27.8 post-procedure (p=0.001). Effect driven by patients with pulmonary hypertension on presentation (n=70): 3.2 mmHg reduction (p<0.0001).

    Refined modified Miller score (thrombus burden):

    Pre-procedural score 20.8±2.4 vs 18.9±2.9 at post-procedure; p<0.001.

    Length of ICU stay 1.5±2.1 days

    Forty-three patients (41.3%) did not require any intensive care unit stay.

    Length of hospital stay 4.1±3.5 days

    Primary safety endpoint: composite MAE rate (any of following within 48 h: device-related death, major bleeding,

    treatment-related clinical deterioration, treatment-related pulmonary vascular injury, and treatment-related cardiac injury):

    Composite MAE rate within 48 h 3.8% (n=4 patients experiencing 6 MAEs):

    • Clinical deterioration n=4

    • Major bleeding event n=1

    • Pulmonary vascular injury n=1

    Major bleeding rate 0.9%.

    Secondary safety endpoint: All primary safety events as well as mortality, device-related SAEs and symptomatic recurrence of embolism within 30 days:

    An additional 10 patients experienced SAEs within 30 days. Total = 14 patients experienced 26 SAEs within 30 d, 5 experienced multiple SAEs.

    1 death at 23 d (respiratory failure from undiagnosed cancer).

    Buckley, 2022

    Secondary endpoints: ICU length of stay, total hospital length of stay:

    Average ICU length of stay statistically significantly lower for mechanical thrombectomy (MT) group vs routine care (RC) (2.1 ±1.2 vs 6.1±8.6 days, p<0.05).

    No statistically significant difference in total hospital length of stay (7.7±6.9 days in MT, 6.8±6.9 in RC).

    Primary endpoint: in-hospital mortality:

    In-hospital mortality was statistically significantly lower for MT group vs routine care RC (3.6% vs 23.3%, p<0.05).

    Secondary endpoints: 30-day readmission rate:

    No statistically significant difference (11% in MT, 13% in RC).

    RC group: 3 self-limited bleeding complications not requiring transfusion (10%) and 1 case of haemodynamically significant bleeding requiring transfusion and endoscopy (3.3%).

    MT group: 3 procedure-related complications (10.7%; 1 self-limited haemoptysis, 2 post-procedure transfusions due to aspiration-related blood loss).

    Jaber, 2020

    Primary endpoint: change in the RV/LV ratio from baseline to 48 h post-procedure

    Reduction in median RV/LV ratio of 0.37 from 1.50 pre-procedure (range 0.88 to 2.52) to 1.13 post-procedure (range, 0.66 to 1.81), (p<0.001).

    Secondary endpoints:

    Change in RV/LV ratio of patients with non-elevated cardiac troponin and zero simplified PE Severity Index (sPESI) score i.e. normal cTn-sPESI: intermediate-low risk (n=17, 22.4%)

    Reduction in mean RV/LV ration of 0.27 (p<0.001)

    ICU stay: 53% admitted to ICU post-procedure, median length of stay 1 day (range 0 to 11 days).

    Heart rate: median 91 bpm (55 to 123) to 89 (62 to 118) at 48 h.

    Median PAP: 30 mmHg (range, 7 to 57 mmHg) at presentation and 27 mmHg (range, 9 to 50 mmHg) after the procedure (p=0.533, NS).

    In 52 patients with elevated PAP (68.4%): statistically significant reduction in median PAP (34 to 31 mmHg, p=0.003).

    Primary endpoint: composite MAEs including major bleeding, device-related death or clinical deterioration, and vascular or cardiac injury within 48 h.

    3 MAEs (4%): 2 periprocedural respiratory deterioration requiring intubation, 1 major bleeding (leading to lobectomy), 1 patient experienced pulmonary vascular injury. All adjudicated as procedure-related rather than device-related.

    Secondary endpoints:

    All-cause mortality

    100% survival to 30 days

    Symptomatic recurrence PE within 30 d

    None

    Sedhom, 2021

    None reported

    Primary outcome: mechanisms of failure of Penumbra Indigo aspiration system

    Most common failure mode: Lightning Unit malfunction (35.8%, n=24) (tubing with dual pressure sensors with a built-in microprocessor for real-time blood flow monitoring).

    Rotating haemostasis valve malfunction (31.3%, n=21)

    Resistance during use (15%, n=10)

    Aspiration failure (11.9%, n=8)

    Engine malfunction (9%, n=6)

    Catheter clogging with thrombi (7.5%, n=5)

    Catheter kinking (6%, n=4)

    Engine canister malfunction (4.5%, n=3)

    Catheter broken (4.5%, n=3).

    Secondary outcome: clinical consequences of device failure

    Death (4.5%, n=3); 2 from fatal pulmonary vessel perforation (3%) and 1 from fatal right-sided heart failure (1.5%).

    Pericardial effusion (1.5%, n=1)

    Procedure aborted (6%, n=4)

    No need for ECMO, no haemoptysis, no intracranial bleeding and no blood transfusions.

    Procedure technique

    Of the 6 studies, all detailed the devices used and 5 detailed the procedure technique. Four studies used the Inari FlowTriever device for mechanical thrombectomy (Tu 2019, Toma 2022, Buckley 2022, Jaber 2020). Two studies used the Penumbra Indigo aspiration system for aspiration thrombectomy (Sedhom 2021, Sista 2021).

    Aspiration thrombectomy involves the aspiration of thrombi through an aspiration catheter. Mechanical thrombectomy involves mechanical engagement of the thrombus and removal. Rheolytic thrombectomy includes the use of high-pressure jets of saline to disperse the thrombus followed by aspiration. The jets can consist of saline, but local thrombolytic agents can also be used. There are multiple different devices with other mechanisms associated with this procedure, some of which are no longer in use.

    Efficacy

    RV/LV ratio

    This outcome was reported in 4 studies. All 4 found a statistically significant reduction in RV/LV ratio from baseline to follow-up. With the FlowTriever, Tu (2019) found a mean reduction of 0.38 (p<0.0001) at 48 hours and Toma (2022) found an average reduction of 0.36 (p<0.001) at follow-up (variable follow-up times, median 32 to 33.5 days). In the Jaber (2020) sub-study of FLARE population, reduction was 0.37 at 48 hours (p<0.001). Jaber (2020) also reported the change in RV/LV ratio for the intermediate-low risk sub-group of the study (n=17) and found a smaller but still statistically significant reduction of 0.27 (p<0.001). Sista (2021) reported a reduction of 0.43 (p<0.0001) at 48 hours with the Indigo system.

    Pulmonary artery pressure

    This outcome was reported in 4 studies. Tu (2019) reported a statistically significant reduction in mPAP (29.8 to 27.8 mmHg), p=0.001. The effect was driven by patients with raised mPAP on presentation in which subgroup there was a 3.2 mmHg reduction (p<0.0001). Toma (2022) reported a statistically significant intraprocedural reduction of 7.1 mmHg (p<0.001). In the Jaber (2020) sub-study, the median PAP demonstrated a statistically non-significant reduction of 3 mmHg (from 30 to 27 mmHg, p=0.533), although in the 52 patients with elevated PAP on presentation, there was a statistically significant reduction of 3 mmHg (34 to 31 mmHg, p=0.003). Sista (2021) reported a reduction in mPAP of 4.3 mmHg post-aspiration (p<0.0001) and of 4.7 mmHg post-procedure (p<0.0001), both statistically significant.

    Length of stay

    This outcome was reported in 5 studies. Tu (2019) reported an average length of ICU stay of 1.5±2.1 days, and hospital stay of 4.1±3.5 days. Toma (2022) reported a mean length of hospital stay of 3 days post-procedure and in the Jaber 2020 sub-study, a median length of ICU stay of 1 day is reported. In Sista (2021), 61% required ICU care and the median stay was 1.0 day.

    Buckley (2022) reported an average length of ICU stay of 2.1±1.2 days which is statistically significantly lower than the ICU length of stay for the routine care group (vs 6.1± 8.6 days, p<0.05). There was, however, no statistically significant difference in hospital stay between the 2 groups, with the mechanical thrombectomy group staying in hospital on average 7.7±6.9 days overall. These hospital length of stay figures are higher than reported by the other studies.

    Modified Miller score

    This outcome was reported in 1 study. Tu (2019) use the refined modified Miller score and reported a pre-procedural score of 20.8±2.4 vs a post-procedural score of 18.9±2.9, representing an average reduction of 1.9 (p<0.001).

    CT obstruction index

    This outcome was reported in 1 study. Sista (2021) reported a mean reduction in CT obstruction index of 11% at 48 hours (p<0.0001).

    Cardiac index

    This outcome was reported in 1 study. Toma (2022) reported overall no statistically significant change in cardiac index between pre- and post-procedure. However, there was a statistically significant improvement in cardiac index in the sub-group with low baseline cardiac index (that is, impaired cardiac function). In this group cardiac index improved by 13% on-table (p=0.005).

    Safety

    Major adverse event rate

    Four studies reported a composite MAE rate as their primary safety endpoint. The composition of this MAE rate varied by study but was similar, usually including major bleeding, device-related death or clinical deterioration and injury to vessels/the heart/the lungs. The composite measure is discussed followed by the individual measures.

    Tu (2019) reported a composite MAE rate (device-related death, major bleeding, treatment-related clinical deterioration, treatment-related pulmonary vascular injury, and treatment-related cardiac injury) of 4% (n=4) at 48 hours. Toma (2022) reported an MAE rate (device‐related death, major bleeding, and device‐ or procedure‐related adverse events) of 1% (n=3) at 48 hours, all major bleeding events but with no device-related injuries, clinical deteriorations or deaths at 48 hours. Jaber (2020), in their sub-study of the FLARE population (Tu 2019), reported an MAE rate (major bleeding, device-related death or clinical deterioration, and vascular or cardiac injury) of 4% (n=3) at 48 hours. Sista (2021) reported an MAE rate (device-related death, major bleeding, and device-related serious adverse events [clinical deterioration, pulmonary vascular, or cardiac injury]) of 3 MAEs in 2 patients (2%) at 48 hours.

    Major bleeding

    Six studies reported on bleeding-related complications. Tu (2019) reported 1 major bleeding event in 104 patients at 48 hours. Toma (2022) reported a 1% major bleeding rate (n=3). Buckley (2022) did not classify bleeding complications by severity but reported 3 procedure-related complications in the mechanical thrombectomy group (11%): 1 self-limited haemoptysis and 2 post-procedure transfusions due to aspiration-related blood loss. In the routine care group the rate was similar with 3 self-limited bleeding complications not requiring transfusion (10%) and 1 significant bleed requiring transfusion (3%). Jaber (2020) reported 1 major bleeding event which was procedure-related at 48 hours in a study of 76 patients. Sista (2021) reported 2 instances of major bleeding (2%) and 3 patients requiring transfusion related to the procedure (3%). Sedhom (2021)reported no haemoptysis episodes, no intracranial bleeding and no episodes requiring blood transfusion.

    Clinical deterioration

    Four studies specifically reported a clinical deterioration rate. Tu (2019) reported a clinical deterioration rate of 4% (n=4) at 48 hours and in the sub-study, Jaber (2020), reported 4% (n=3) at 48 hours. Toma (2022) reported a clinical deterioration rate of 0% at 48 hours. Sista (2021) reported 1 episode of clinical deterioration at 48 hours in 119 patients.

    Pulmonary vascular injury

    Five studies reported on instances of pulmonary vascular injury. Tu (2019) reported 1 pulmonary vascular injury in 104 patients at 48 hours and Jaber (2020) also reported 1 pulmonary vascular injury in their sub-study of 76 patients, which likely represents the same patient. Sedhom (2021) reported 2 pulmonary vessel perforations (3%). Sista (2021) reported 2 instances of pulmonary vascular injury (2%) at 48 hours, 1 of which (1%) was device-related. Toma (2022) reported 0 device-related pulmonary vascular injuries.

    Cardiac injury

    Five studies reported on cardiac injury. Tu (2019), Jaber (2020), Toma (2022), Sista (2021) and Sedhom (2021) all reported no episodes of cardiac injury.

    Mortality

    All 6 studies reported measures of mortality.Tu (2019) reported 1 death during follow-up (1%) at 23 days post-procedure, attributed to respiratory failure from undiagnosed cancer and so unrelated to the procedure. Toma (2022) reported 1 death in a study of 250 participants at 30 days which was unrelated to the procedure. Buckley (2022) reported in-hospital mortality of 4% for the mechanical thrombectomy group, which was statistically significantly lower than the routine care group (23%, p<0.05). Sedhom (2021) reported 3 deaths (5%): 2 from pulmonary vessel perforation (3%) and 1 from right-sided heart failure (2%). Jaber (2020) reported 100% survival at 30 days. Sista (2021) reported 1 post-procedure device-related death (1%) related to haemorrhage and right ventricular overload at 48 hours and a 3% rate of any cause mortality at 30 days (n=3).

    Access-site complications

    Access-site complications may include bleeding, nerve injury and damage to other structures, for example arteries. Two studies reported access-site complications. Toma (2022) reported 1 access-site haematoma in a patient who also received thrombolytics in 250 participants. Sista (2021) reported 2 patients (2%) experiencing access-site bleeding.

    30-day readmission rate

    Two studies discussed 30-day readmission rates. Toma (2022) reported a 30-day readmission rate of 6% (n=13/216), however only 1 was related to PE. When comparing with other treatments, Buckley (2022) found no statistically significant difference in 30-day readmission rate between the mechanical thrombectomy group (11%) and the routine care group (13%).

    Device failure

    One study, Sedhom 2021, specifically reviewed device failure reports for the Indigo aspiration system used for aspiration thrombectomy. It found that the most common failure mode was a Lightning Unit malfunction (36%, n=24) followed by rotating haemostasis valve malfunction (31%, n=21), resistance during use (15%, n=10) and aspiration failure (12%, n=8).

    Anecdotal and theoretical adverse events

    Expert advice was sought from consultants who have been nominated or ratified by their professional Society or Royal College. They were asked if they knew of any other adverse events for this procedure that they had heard about (anecdotal), which were not reported in the literature. They were also asked if they thought there were other adverse events that might possibly occur, even if they have never happened (theoretical).

    They listed the following anecdotal or theoretical adverse events not reported in the literature:

    • DVT

    • Infection

    • Cardiac complications; tamponade, myocardial infarction, valvular dysfunction

    • Iodine anaphylaxis

    • Contrast nephropathy

    • Haemothorax

    • Puncture site pseudoaneurysm

    • Stroke (clot transit through patent foramen ovale).

    Three professional expert questionnaires were submitted for this procedure. Find full details of what the professional experts said about the procedure in the specialist advice questionnaires for this procedure.

    Validity and generalisability

    The main evidence summary included 6 studies, including 4 prospective single-arm studies (3 trials and 1 registry), 1 retrospective comparative study and 1 database review. All 6 studies were in USA populations, included small sample sizes with the largest being 250 participants, and had short term follow-up. Variable inclusion criteria were used and many of the studies included intermediate-risk or sub-massive PEs either solely or in combination with high-risk or massive PEs. All of these factors may impact on the outcomes and on the generalisability to the UK population. The FLARE-ED sub-study by Jaber (2020) is less generalisable as it has a smaller sample of patients presenting via the Emergency Department only. The longest follow-up period was 6 months and so there is a lack of data on long-term outcomes in this group. The included studies generally reported fairly consistent outcomes, with statistically significant improvements in markers of right ventricular function and low MAE and mortality rates in the short-term follow-up.

    Various techniques and devices are used for mechanical thrombectomy. The main evidence summary includes 4 studies which use the FlowTriever device and 2 which use the Indigo aspiration system. There are other devices with varying techniques available for this procedure, some of which are covered in the additional studies in table 5. This variability means that the outcomes may not be generalisable to other techniques and devices, and caution should be taken in interpreting outcomes for this procedure.

    Due to the non-randomised nature of the studies and lack of comparator, the study samples may be subject to selection bias as those included are not randomly selected. Buckley (2022) was the only study to include a comparator arm but was performed retrospectively and was not randomised. The comparator of routine care was heterogenous in terms of treatments used and the study did not include any measures of right ventricular function. There may be information bias in terms of the way the cases are managed and the outcomes recorded when participants are in a trial. Loss to follow-up may also introduce bias in the outcome reporting. Sedhom (2021) was a review of the MAUDE safety database which relies on voluntary reporting of events, it therefore will not capture all events so introduces selection bias. It also will not capture all mechanical thrombectomy events being performed so an incidence rate can't be calculated as there is no denominator available. The review was retrospective in nature and doesn't correlate device failure to clinical outcomes.

    The FLARE study and FLARE-ED sub-study were sponsored by Inari medical, the manufacturers of the FlowTriever device, and they were involved in study design and data collection. Various authors of Toma (2022) are consultants for Inari medical and various other manufacturers including Penumbra. A couple of authors for Sedhom (2021) are consultants for various manufacturers including Inari Medical. Sista (2021) was funded by Penumbra and the authors have consultant roles for various companies including Penumbra and Inari medical.

    There are gaps in the evidence overall and for particular patient populations such as those with Patent Foramen Ovale (PFO). Further evidence is needed from randomised controlled trials with larger populations and longer-term follow-up. There are some ongoing trials which may help to address some of the evidence gaps:

    • STRIKE-PE: A Prospective, Multicentre Study of the Indigo™ Aspiration System Seeking to Evaluate the Long-Term Safety and Outcomes of Treating Pulmonary Embolism NCT04798261. N=600, prospective cohort study. Study completion March 2026.

    • FlowTriever for Acute Massive Pulmonary Embolism (FLAME). NCT04795167. N=250, prospective parallel-group cohort. Completion date May 2023.

    • The PEERLESS Study. NCT05111613. N=550, Prospective, multicentre, RCT of the FlowTriever System compared to Catheter-Directed Thrombolysis (CDT) for acute intermediate-high-risk pulmonary embolism (PE). Completion date October 2023.

    • Evaluating the Safety and Efficacy of the AlphaVac Multipurpose Mechanical Aspiration (MMA) F1885 PE for Treatment of Acute Pulmonary Embolism. NCT05318092. N=122, single arm trial. Completion date November 2023.

    • FlowTriever All-Comer Registry for Patient Safety and Hemodynamics (FLASH). Observational (patient registry), n=1300. Completion date July 2023.