Prospective Multicenter International Registry of Ultrasound-Facilitated Catheter-Directed Thrombolysis in Intermediate-High and High-Risk Pulmonary Embolism (KNOCOUT PE).
Jan 24, 2024DOI: 10.1161/circinterventions.123.013448Observational StudyAHA
Written by: Keith M SterlingSamuel Z GoldhaberAndrew S P SharpNils KucherNoah JonesRobert MaholicNicolas MeneveauDavid ZlotnickSameh SayfoStavros V KonstantinidesGregory Piazza
The KNOCOUT PE study (EKoSoNic Registry of the Treatment and Clinical Outcomes of Patients With PE) was a prospective multi-center international registry across North America (ClinicalTrials.gov: NCT03426124) and Europe (EudraCT: 2018-001235-46). KNOCOUT PE was designed by the study steering committee in collaboration with the study sponsor (BTG/Boston Scientific Corporation, Marlborough, MA). The aim of the study was to evaluate a large cohort of patients with intermediate-high and high-risk PE treated with USCDT with regard to the dose and duration of r-tPA delivered and frequency of major bleeding.Enrollment for the Prospective arm of KNOCOUT PE took place from March 2018 through June 2020; each patient’s participation in the registry was expected to last 12 months. A total of 489 patients were enrolled across 64 sites throughout North America (United States and Canada) and Europe (Austria, France, Germany, the Netherlands, Switzerland, and the United Kingdom). The protocol was approved by each site’s institutional review board or ethics committee, and patients at each site were required to provide informed consent before enrollment in the registry. The data that support the findings of this study are available from the corresponding author upon reasonable request.Patients enrolled were risk-stratified and determined to have either intermediate-high risk or high-risk PE. Intermediate-high risk PE was defined by evidence of RV dysfunction (RV/LV >1.0) on either echocardiogram or computed tomography, and troponin elevation per the site’s clinical laboratory protocol. High-risk PE was defined as having sustained hypotension (systolic blood pressure <90 mm Hg for at least 15 minutes or requiring inotropic support). The study included patients who had already been selected for treatment with the EKOS device and were 18 to 80 years old, had RV/LV >1.0 measured by CTA or echocardiogram, PE symptom duration ≤14 days, troponin elevation ≥upper limit of normal per the institution’s assay, and signed informed consent. Exclusion criteria were life expectancy <1 year and the clinician deeming the patient high-risk for catastrophic bleeding, per International Society on Thrombosis and Haemostasias criteria (defined as fatal or associated with any of the following: [1] a fall in Hb of 2 g/dL or more or transfusion of at least 2 units of red blood cells or [2b] involvement of a critical anatomic site).11Of note, sites were allowed to complete inform consent in lieu of patients in KNOCOUT PE following USCDT if the patient had a severe presentation (such as high-risk PE) such that delay for obtaining consent was clinically inappropriate. Sites were encouraged to maintain screening logs but were not required to do so.
USCDT was performed by a trained endovascular specialist in an angiographic suite or hybrid operating room with digital angiographic equipment using the EkoSonic (EKOS) Endovascular System (EKOS/BTG/Boston Scientific Corporation, Marlborough, MA) in conjunction with r-tPA (recombinant tissue-type plasminogen activator) administration. Treatment dose and duration were determined by the individual investigator’s judgment. The EKOS catheter was inserted into the pulmonary artery using standard ultrasound and fluoroscopic-guidance techniques; the venous access site was selected by the treating physician. In the case of a unilateral filling defect in 1 main or proximal lobar pulmonary artery, only 1 device was placed. In the case of a bilateral filling defect in both main or proximal lobar pulmonary arteries, as determined by CTA, 2 devices were placed (one in each of the involved main or proximal pulmonary arteries).
Post-procedure, patients were followed per the investigator’s usual clinical practice. At a minimum, health status checks were completed during the hospital stay, before discharge, and at 3 and 12 months post-treatment (Table 1). Safety data were to be collected and reported through 12 months post-treatment. An on-site visit with echocardiography was recommended at 3 months for patients with an abnormal echocardiogram post-treatment to allow for evaluation of chronic pulmonary hypertension, and at 12 months for all patients. Additionally, matched echocardiograms were required (ie, a patient could not have an echocardiogram at 1 visit and a CTA at the next). The availability of imaging-based assessment could vary across patients and centers due to differences in follow-up protocol employed at each center (Figure S1).
Post-USCDT Follow-Up
Data were collected via an electronic data capture system (DATATRAK International Inc, Mayfield Heights, OH) and were site-reported.
Safety outcome measures collected included frequency and severity of serious adverse events related to the device or procedure; recurrent venous thromboembolism, International Society on Thrombosis and Haemostasias–defined major bleeding, and mortality during index hospitalization and the first 12 months post-procedure. A third-party independent physician medical monitor, blinded to details of the dosing and duration protocol, adjudicated safety outcomes and reviewed all major bleeding events and serious adverse events.
An additional outcome measure of interest was a change in RV/LV ratio, as assessed by matched echocardiogram or CTA from baseline to between 24 and 48 hours after the start of the USCDT procedure. Additional outcomes are in Table S1.
Quality of life data were collected, as measured using the PE quality of life, the visual analog scale, and EuroQoL 5-dimension 5-level instruments at 3 and 12 months posthospitalization visits (Misery, Utility, and Composite scores; Table S2).12–15
In this single-arm study, sample size was based on the expected enrollment per site, based on estimated cases per physician across total number of study sites with no formal hypothesis testing. Up to 550 patients would be enrolled to collect data on current EKOS use at up to 100 sites. For baseline and procedure characteristics and safety outcomes, all patients who received the USCDT procedure were included. For the outcomes of RV/LV ratio, tricuspid annular plane systolic excursion, estimated RV systolic pressure and patient-reported outcomes, all patients who met eligibility criteria and received the USCDT procedure were included and changed from baseline was assessed using 2-sided P values from a t test. Sensitivity analyses were conducted post hoc to determine whether significant differences existed in major bleeding rate and mortality between patients who were consented before treatment versus after treatment by excluding the patients who consented after treatment; 2-sided P values were produced from Fisher exact test.
Continuous data were summarized with means, medians, SDs, minima and maxima, unless otherwise specified. Categorical data were summarized with observed counts and percentages for each category. Percentages were based on the total number of patients for all analysis. There was no imputation for missing values. Posthoc multivariable logistic regression models were performed to assess if the following fixed effects had an impact on mortality and major bleeding events: treatment center, body mass index, age, sex, whether the patient had active cancer and whether the patient was at a high or intermediate risk for PE. A P<0.05 would indicate the effect was impactful. In the event there were too few patients (<5) in a treatment center, the center’s data was combined with other center(s) to achieve at least 5 patients. The combining of treatment centers was accomplished by ranking the sites that did not have at least 5 patients, with ties broken by treatment center number. The centers with the largest number of patients were combined with the smallest center as necessary until there were 5 patients. The process was repeated among the remaining centers. Centers were only combined within geographic region (United States versus outside the United States). A sensitivity analysis of the post hoc multivariable logistic regression models was also performed by excluding any centers with <5 patients.
RESULTS
A total of 489 patients were included in the prospective KNOCOUT PE cohort (Table 2; Table S3). Of the enrolled patients, 93.3% (456/489) were consented before enrollment; 6.7% (33/489) were consented post-procedure. Eleven sites had received the waiver to consent patients post-procedure in an effort to include more high-risk PE patients. Of the 33 patients consented post-procedure, 24 of the 33 were from 2 sites (1 United States and 1 non-United States).
Demographics and Baseline Characteristics of the Patients With Intermediate-High and High-Risk Pulmonary Embolism Treated With Ultrasound-Facilitated Catheter-Directed Thrombolysis in the KNOCOUT PE Prospective Registry Study (N=489)
Of the 489 patients, 463 (94.7%) were classified as intermediate-high risk and 26 (5.3%) were classified as high-risk (Table 3). Simplified PE severity index (sPESI) and Bova scores were calculated for patients based upon information in the database. Within the study population, 55.6% had an sPESI score greater than or equal to 1 (median=1.0; interquartile range, 0.0–1.0). For patients with systolic blood pressure ≥90 mm Hg (N=477), the mean Bova score was 4.4 (SD=0.61).
Clinical Presentation of Patients With Intermediate-High and High-Risk Pulmonary Embolism Treated With Ultrasound-Facilitated Catheter-Directed Thrombolysis in the KNOCOUT PE Prospective Registry Study (N=489)
Procedural success, defined as having all catheters successfully placed and treatment administered, was 99.8% (488/489 patients; Table 4). Median length of hospital stay was 4.0 days (interquartile range, 3.0–6.0; Table 4). Of the 393 patients treated in the United States, 90.8% had a stay in the intensive care unit (ICU); of the 90 patients treated outside of the United States, 45.6% had an ICU stay as part of their treatment. For the entire patient population, the mean time in the ICU was 48.9 hours (SD=47.4). Decisions regarding triage post-procedure (ICU, step-down unit, general floor) were made according to each site’s policies and procedures along with the clinical team.
Procedural Characteristics From Patients With Intermediate-High and High-Risk Pulmonary Embolism Treated With Ultrasound-Facilitated Catheter-Directed Thrombolysis in the KNOCOUT PE Prospective Registry Study (N=489)
The mean r-tPA dose was 18.1 mg (SD=7.4). Three patients outside of the United States received tenecteplase with doses between 19.8 and 20 mg; these patients were not included in r-tPA dose calculations. Of those who received r-tPA for whom complete dosing data were available, 69.5% received 20 mg or less; 31.0% of patients received 12 mg or less (Table 4; Figure S1). Mean duration of thrombolytic therapy overall was 10.5 hours (SD=5.4). In the United States, mean duration of thrombolytic therapy was 9.8 hours (SD=4.9); outside of the United States, it was 13.4 hours (SD=6.4).
Among the 489 patients, there were 8 major bleeding events within the first 72 hours post-treatment (1.6%; Table 5). One patient with bilateral subdural hematomas underwent bilateral craniotomies and developed postoperative PE. However, the patient had residual subdural hematomas; s/he was treated with both USCDT and therapeutic anticoagulation and developed worsening of the left-sided subdural hematomas, which ultimately required repeat surgical decompression. Other than this event, no other ICH events were reported. A sensitivity analysis showed no significant difference in major bleeding rates within the first 72 hours post-treatment when patients who were consented after intervention were removed from the population (1.5% [7/456] versus 1.6% [8/489] for the total study population (2-sided P=1.0). Additionally, there were no meaningful differences in major bleeding in United States and outside of US patients. A post hoc regression analysis and a sensitivity analysis showed that none of the collected risk factors impacted the number of major bleeding events (Tables S4 and S5).
Major Bleeding Events in Patients With Intermediate-High and High-Risk Pulmonary Embolism Treated With Ultrasound-Facilitated Catheter-Directed Thrombolysis in the KNOCOUT PE Prospective Registry Study (N=489)
Two of the 489 patients suffered recurrent PE (0.4%; Table 5). Both occurred within the first 30 days following treatment; one of the patients was noted to be subtherapeutic on anticoagulation while transitioning from unfractionated heparin to oral anticoagulation; the other patient with a preexisting iliofemoral deep vein thrombosis experienced recurrence 3 days after completion of USCDT.
Five patients died during the initial 30 days post-treatment (1.0%); all were intermediate-high risk patients (Table 5). Of the 26 patients with high-risk PE who were enrolled in KNOCOUT PE, there were no deaths. A sensitivity analysis showed no significant difference in mortality rates within the first 30 days post-treatment, when patients who consented after intervention were removed from the population (0.9% [4/456] versus 1.0% [5/489] for the whole population; 2-sided P=1.0). Additionally, there were no meaningful differences in mortality in United States and outside of US patients. A post hoc regression analysis and a sensitivity analysis showed that none of the collected risk factors impacted the mortality rate (Tables S4 and S5).
Of the patients included in the prospective KNOCOUT PE registry, 176 (36%) had postprocedure-matched echocardiograms before discharge (Figure; Table 6). Mean baseline RV/LV ratio was 1.35 (SD=0.32). Of the 176 patients, 77 had a matched echocardiogram between 24 hours and up to 48 hours post-procedure, with a mean RV/LV ratio of 0.98 (SD=0.27); the mean relative reduction from this postprocedure time point as compared with baseline was 24.57% (2-sided P<0.0001; Table 6).
Imaging Assessments (RV/LV, TAPSE, and eRSVP) in Patients in the KNOCOUT PE Prospective Registry Who Had Postprocedure-Matched Echocardiograms Before Discharge
Change in right ventricular/left ventricular (RV/LV) ratio on echocardiogram in patients with intermediate-high and high-risk pulmonary embolism treated with ultrasound-facilitated catheter-directed thrombolysis in the KNOCOUT PE (The EKoSoNic Registry of the Treatment and Clinical Outcomes of Patients With Pulmonary Embolism) prospective registry study. For each boxplot, the upper and lower part of the blue box show the upper and lower quartiles; the upper line shows the maximum observation below the upper fence (+1.5×IQR); the lower line shows the minimum observation below the lower fence (−1.5×IQR); the horizontal line within the box shows the median; the diamond shape within the box shows the mean.
Another exploratory parameter was tricuspid annular plane systolic excursion. At baseline, mean tricuspid annular plane systolic excursion was 16.50 (SD=17.91) among the 67 patients evaluated (Table 6). The mean tricuspid annular plane systolic excursion among 36 patients assessed 24 to 48 hours post-procedure was 18.74 (SD=6.4); mean relative change was 82.58% (SD=232.95, 2-sided P=0.04; Table 6).
We also assessed the mean change in estimated right ventricular systolic pressure (mm Hg) on ECHO. At baseline, the mean estimated RV systolic pressure in the 62 patients evaluated was 50.35 mm Hg (SD=15.34). Mean estimated RV systolic pressure between 24 and 48 hours post-procedure in the 31 patients evaluated was 35.77 mm Hg (SD=16.73); mean relative reduction was 28.55% (SD=31.37; 2-sided P<0.0001; Table 6).
Patient-reported data were available for 243 (49.7%) patients at 3 months post-treatment and for 153 (31.3%) patients at 12 months post-treatment (Table 7). The mean PEmb-Quality of life score 3 months post-procedure in 243 patients evaluated was 16.0 (SD=17.7), compared with 38.5 (SD=22.1) at baseline (41.1% change; 2-sided P<0.0001; Table 7; Table S2). Similarly, mean visual analog scale score improved from 63.1 (SD=23.0) to 75.5 (SD=19.8) from post-procedure to the 3-month time point in 242 patients evaluated (56.0% change; 2-sided P=0.0008; Table 7). The composite EuroQoL 5-dimension 5-level Misery score was reduced from a mean of 10.2 (SD=4.2) immediately post-procedure to 7.4 (SD=3.5) at 3 months (17.8% change, 2-sided P<0.0001) in 243 patients (Table 7). The composite EuroQoL 5-dimension 5-level Utility score also showed improvement, from a mean of 0.64 (SD=0.31) immediately post-procedure to 0.83 (SD=0.24) at 3 months (0.2, 2-sided P<0.0001) across the 243 patients assessed (Table 7). Improvement across all instruments was maintained through the 12-month follow-up period (Table 7).
Quality of Life Outcomes Assessed in Patients With Intermediate-High and High-Risk Pulmonary Embolism Treated With Ultrasound-Facilitated Catheter-Directed Thrombolysis in the KNOCOUT PE Prospective Registry Study Over 12 Months Posttreatment Follow-Up
DISCUSSION
The prospective multicenter, international KNOCOUT PE registry was designed to study a larger cohort of patients treated with USCDT following the OPTALYSE PE trial, with an emphasis on safety and dosing patterns. Across all patients enrolled in KNOCOUT PE, the mean total r-tPA dose administered was 18.1 mg, with a mean infusion duration of 10.5 hours. Both major bleeding within 72 hours post-procedure and all-cause mortality within 30 days of intervention were low (1.6% and 1.0%, respectively). Aside from the worsening of a preexisting subdural hematoma in 1 patient after initiation of anticoagulation and USCDT, no other ICH events were reported. The echocardiographic outcome measure of change in RV/LV ratio demonstrated a mean relative reduction of 22.6% post-procedure and 41.8% 1-month post-procedure.
Previously published bleeding complication rates in some studies of USCDT have reported major bleeding rates approaching 10%.7 In KNOCOUT PE, an International Society on Thrombosis and Haemostasias major bleeding rate of 1.6% was observed—lower than those previously reported in the ULTIMA and SEATTLE II studies, and comparable to that observed in OPTALYSE PE.6–8,10 This is notable because the specific dosing and duration regimen were not prescribed, and there was no compulsory anticoagulation protocol. The low frequency of major bleeding in KNOCOUT PE reflected the procedure’s safety in practice and was similar to rates of major bleeding reported with mechanical thrombectomy.16–18 The reduction in the frequency of these serious adverse events may be attributable to multiple changes in practice, including the increased adoption of ultrasound for venous access, greater experience with the procedural technique, and the reduction in dose and duration of thrombolytic therapy.
In intermediate-high and high-risk patients, Bova and sPESI are 2 scoring systems used to predict short-term mortality at 30 days post-PE.19–22 In KNOCOUT PE, 55.6% of patients had an sPESI score ≥1, which correlates with a 30-day mortality rate of 10.9%. Of the 477 patients in KNOCOUT PE with intermediate-high risk PE for whom Bova scores could be calculated, the mean score was 4.4 (SD=0.61); a Bova score of >4 correlates with 42% of patients experiencing PE-related complications and a 10% PE-related mortality. The 30-day mortality rate in the KNOCOUT PE prospective study was 1.0%, which is lower than would be predicted by the Bova and sPESI scoring systems; however, this may reflect the study population enrolled and selection bias. Furthermore, none of the patients who presented with high-risk PE died. These low mortality rates occurred during a time of increased implementation of organized PE response teams, increased use of catheter-based reperfusion strategies, and local protocolization of anticoagulation regimens.23,24 Time in the ICU and length of hospital stay are also key considerations in evaluating interventional therapies for PE. In KNOCOUT PE, 90.8% of US patients and 45.6% of patients outside of the US required an ICU stay, with a mean ICU stay of 48.9 hours for the entire patient cohort. Although the patients in KNOCOUT PE had a mean thrombolytic infusion time of 10.5 hours, their longer stay in the ICU may be attributable to having clinically more severe PE (intermediate-high or high-risk PE) Additionally, decisions regarding patient placement in an ICU setting are subject to policies of individual hospitals and nursing competencies and not solely related to reperfusion therapy.
The mean RV/LV reduction of 22.6% on postprocedure echocardiogram was similar to observations in other single-arm prospective thrombectomy and catheter-directed thrombolysis trials (22.6%–34.9%).6–10,16,18 In the stratified analyses by time of imaging post-intervention, the RV/LV ratio continues to improve and then stabilizes. Further reduction in RV/LV ratio was noted at 1-month post-intervention (41.8% mean improvement).
Limitations include those typically associated with an observational study, such as susceptibility to bias and confounding as well as uncertainty about causality. A major limitation is that, while screening logs were encouraged in KNOCOUT PE, they were not mandated. Without complete screening logs from each site, we were limited in ascertaining the extent of selection bias that may have occurred. The lack of follow-up echocardiographic data from some institutions was also a limitation. Conclusions regarding outcomes in improvement in echocardiographic measures are therefore exploratory. An additional limitation is that there was no core laboratory adjudication of imaging. Finally, the enrolled study population was predominantly White (60.5%), limiting the ability to generalize the study findings to other demographic populations.
CONCLUSIONS
USCDT is part of an increasingly diverse scope of interventional therapies for PE. This prospective study provides data on current practice and safety, which may be used by clinicians as part of the decision-making and consent processes when considering reperfusion for PE. Prospective randomized controlled trials, such as the currently enrolling HI-PEITHO (NCT04790370) and PE-TRACT trials (NCT05591118), support the evolution of evidence-based guidelines and clinical practice.25
ARTICLE INFORMATION
The authors thank Janiece Rutherford, MBA (Boston Scientific Corporation) and Alissa Garman for study support; Binal Patel, MPH (Boston Scientific Corporation) for statistical support; and Alexandra J. Greenberg-Worisek, PhD, MPH, CMPP (Boston Scientific Corporation) for medical writing assistance.
KNOCOUT PE (The EKoSoNic Registry of the Treatment and Clinical Outcomes of Patients With Pulmonary Embolism) was funded by the EKOS Corporation, a BTG International group company, a wholly owned subsidiary of Boston Scientific Corporation.
At the time of KNOCOUT (The EKoSoNic Registry of the Treatment and Clinical Outcomes of Patients With Pulmonary Embolism) and writing of this article, Dr Sterling was a consultant to Boston Scientific, Access Vascular Inc, and vizai. Dr Sterling is now an employee of Boston Scientific. Dr Piazza has received research support from Bristol-Myers Squibb/Pfizer Alliance, Bayer, Janssen, Alexion, Amgen, and Boston Scientific Corporation, and consulting fees from Bristol-Myers Squibb/Pfizer Alliance, Boston Scientific Corporation, Janssen, NAMSA, Penumbra, Prairie Education and Research Cooperative, Boston Clinical Research Institute, and Amgen. Dr Sharp was a Consultant to Medtronic, Boston Scientific, Recor Medical, Penumbra, Philips. Dr Kucher received consulting and lecture fees from Concept Medical, Boston Scientific, Bayer, and INARI; institutional grants from the Swiss National Foundation, Bard, Bayer, Boston Scientific, Concept Medical, and Sanofi; not related to the present work. Dr Goldhaber received Research Support from Bayer, BMS, Boston Scientific, Janssen, NHLBI, and Pfizer. Dr Meneveau was a consultant at Abbott Medical, Bayer Health Care, Bristol-Myers Squibb, Edwards Lifesciences, Terumo, Inari, Boston Scientific. Fees: Abbott Medical, AstraZeneca, Bayer Health Care, Bristol-Myers Squibb, Pfizer, Terumo. Dr Zlotnick participated in Speaker’s Bureau: Abiomed, Inari Medical; Consultant/Speaker’s Bureau: Angiodynamics. Dr Sayfo was consultant with Boston Scientific. Dr Konstantinides received institutional grants from Bayer AG, Daiichi-Sankyo, and Boston Scientific; consulting or lecture fees from Bayer AG, Daiichi Sankyo, Boston Scientific, Penumbra, MSD, Pfizer, and Bristol-Myers Squibb. The other authors report no conflicts.
Tables S1–S5
Figure S1
KS
Dr. Keith Sterling
Radiology · Interventional RadiologyInova Health Systemkeiths