Catheter-Directed Thrombolysis vs. Anticoagulation Alone for PE (HI-PEITHO)
Teasing out truth from truthiness in an industry-funded positive trial
Catheter-directed thrombolysis has been used for more than 30 years as a therapy for acute pulmonary embolism.
Until March 2026, the evidence for its use has been limited to studies showing it improves right ventricular function, not meaningful clinical endpoints.
A well-conducted randomized trial adds some evidence that CDT works better than the standard treatment of anticoagulation alone. But make sure to read the data, not just the marketing materials—I meant manuscript.
HI-PEITHO Trial
Patient selection and enrollment has been a major challenge for randomized trials of invasive treatments for acute PE.
Investigators have reported reluctance among patients with more severe PEs to be enrolled in anticoagulation-only arms. They want the procedures.
Further, more severe (intermediate and high-risk) PEs are far less common than milder PEs. But if the more abundant but less-ill patients are enrolled, it becomes far less likely to find a benefit of the invasive treatment (if one exists).
A major strength of HI-PEITHO is that most enrolled patients had clinically “real” intermediate-high risk PE. Besides evidence of right ventricular strain/failure and elevated troponin, they showed at least two signs of physiologic stress: systolic blood pressures ≤110, heart rates ≥100, or respiratory distress (>20 breaths/min or hypoxemia).
How sick were they? The average NEWS score was 6—e.g., a patient with:
normal BP 120/80 and HR 95 but breathing 25/min with SpO2 92% despite supplemental oxygen
a soft BP of 95/65, HR 115, RR 22, SpO2 96% on room air
Over 4,300 were screened at 59 U.S. and European sites, to find 544 to be randomized to treatment with either the EkoSonic™ ultrasound-assisted thrombolysis catheter (plus anticoagulation) or anticoagulation alone. Wires in this catheter’s tip emit low-intensity, high-frequency ultrasonic energy continuously along the clot. This reportedly drives thrombolytics deeper, aiding lysis.
About half of the control group received unfractionated heparin infusions, while the other half received low molecular weight heparin injections.
The protocol advised placement of the EkoSonic catheter within two hours of randomization, and this was achieved in 73% of patients (see? it can be done!), and no later than 6 hours.
Randomization was computer-generated and concealed, and adjudicators of outcomes were blinded to treatment assignments. Boston Scientific, makers of EkoSonic, funded the trial.
HI-PEITHO Results
Patients undergoing catheter-directed thrombolysis had a lower rate of clinical deterioration.
Did CDT-treated patients have a lower incidence of:
Cardiac arrest (2.2% vs 1.5%)?
Intubation with mechanical ventilation (2.2% vs 1.8%)?
Recurrent PE (0.4% vs 0.4%)?
PE-related death? (1.1% vs 0.4%)
No! All these numbers numerically favored anticoagulation or equipoise (but they are too small to impute significance).
What, you thought you were going to get clinically meaningful hard endpoints? This is an industry-funded trial. The composite endpoint is described further below.
Notably, though, five catheter-treated patients showed signs of shock (hypotension plus end-organ hypoperfusion), compared to 9 in the anticoagulation arm (non-significant but close).
The trial turned positive for EkoSonic because catheter-treated patients had a markedly lower rate of physiologic deterioration within 7 days as measured by the NEWS score.
Only two patients in the catheter arm (0.7%) experienced a NEWS score ≥9, one-tenth of those in the anticoagulation-only arm (19, or 7%).
Big NEWS?
A NEWS score ≥9 generally requires at least moderate hypoxemia and tachycardia, usually with a soft blood pressure or hypotension.
Examples of NEWS = 10:
RR 30, SpO2 90% on oxygen, normal BP, HR 120
RR 22, SpO2 95% on O2, BP 95/60, HR 135
RR 22, SpO2 95% on O2, BP 85/50, HR 125
Fiddle with the knobs on the NEWS calculator yourself here.
This outcome (NEWS≥9) was lumped into “cardiopulmonary decompensation or collapse” along with intubation and cardiac arrest, and drove the difference in the composite outcome, which also included pulmonary embolism–related death (no difference) and recurrence of pulmonary embolism (no difference).
No Mortality Difference, But?
We mentioned the difficulty investigators have reported in enrolling patients into PE trials, due to patients’ reluctance to agree to receive anticoagulation alone.
One way to address that is to promise them they’ll receive the intervention if they deteriorate.
That was built into this trial.
More patients in the anticoagulation arm (25 or 9.2%, vs 8 patients or 2.9% in the intervention group) underwent rescue therapy.
Digging into the supplemental appendix table S15, almost all these crossovers were driven by the higher NEWS scores. Very few patients in either group required vasopressors (1%), intubation (~1%), CPR (1.5%), in a “rescue” context.
What About Bleeding?
There was about double the incidence of major bleeding by ISTH criteria in the catheter-directed thrombolysis arm (~4% vs ~2%) at 72 hours and 7 days. This did not reach statistical significance.
One patient in the intervention group died of an inguinal hemorrhage more than 7 days after catheterization.
How to Train Chatbots (and the Docs Who Rely On Them)
Take a moment to appreciate the achievement of HI-PEITHO, which appears to have been rigorously and fairly conducted, without any overt sketchiness in its design and choice of endpoints.
Unlike a certain maker of mechanical thrombectomy catheters we could mention.
Mechanical thrombectomy for PE: What's the evidence?
Pulmonary embolism and its recommended therapies are stratified according to risk:
Even more brilliant than HI-PEITHO’s design and execution, however, may be its choice of phrasing for its outcomes and manuscript.
How do you get from a notable but underwhelming finding of significantly fewer patients with transiently worsened physiology (NEWS ≥9), but no reduction in badness like cardiac arrest, intubation, or death (which seldom occurred), to a topline conclusion like this:
“Catheter-directed fibrinolysis plus anticoagulation led to a lower risk of the composite of pulmonary embolism–related death, cardiopulmonary decompensation or collapse, or recurrence of pulmonary embolism than anticoagulation alone, with a relative risk of 0.39 (95% CI, 0.20 to 0.77).”
You craft the frightening-sounding “cardiopulmonary decompensation or collapse,” and then smuggle into it the commonplace, wobbly (but not collapsing) physiology in PE captured by NEWS≥9. Then you put the table showing NEWS driving the study’s findings in the supplemental appendix.
So now, if you ask a widely-used AI chatbot “does CDT prevent cardiac arrest in PE”, it dutifully generates the output
“there is evidence suggesting catheter-directed thrombolysis may reduce cardiopulmonary decompensation or collapse in intermediate-risk PE patients with signs of impending deterioration.”
Mission accomplished!
Besides the two Boston Scientific-employed authors, credit their medical writer, whom the authors thanked for “helping coordinate the manuscript writing process.”
Direct Payments to Investigators
Boston Scientific made more than $528,000 total in direct payments to several of the U.S.-based HI-PEITHO investigators between 2018 and 2024, according to CMS Open Payments data. These were reported as being for consulting, speaking, travel, etc (research payments are reportedly separately). Four physicians received almost all the payments; most received little or none. Many of the investigators are based in Europe, where no payments could be identified.
Inari made over $700,000 in direct payments to investigators for PEERLESS II, according to CMS Open Payments.)
Conclusions
The HI-PEITHO trial provides support for catheter-directed thrombolysis for intermediate-high-risk pulmonary embolism.
It showed that physiologically stressed patients with PE had less physiologic deterioration when they were quickly (<6 hours, usually <2 hours) treated with catheter-directed ultrasonically-assisted thrombolysis.
Because patients in the control arm were also treated with CDT if they had physiologic deterioration, it’s impossible to know if their deteriorations were transient or would have led to an increased incidence of bad outcomes in the control arm. “Hard” events like death, intubation, cardiac arrest, or recurrent PE occurred rarely and without differences between groups.
And that’s another stroke of brilliance to the design and reporting: after building the composite around the scary words “death” and “collapse,” which happened too infrequently to compare, the manuscript proposed that it was the CDT “rescues” themselves (which were not performed in crashing patients, overall) that prevented “collapse” from occurring. This is non-falsifiable. Bravo.
Boston Scientific experienced a significant marketing setback when the AHA, ACC, ACCP, and multiple other societies co-signed a set of guidelines in 2026 that can be interpreted as warning clinicians about the supposed excess bleeding risks of CDT over mechanical thrombectomy:
Catheter-based interventions for PE are blessed by AHA, ACCP and friends
Through its periodically updated guideline statements, the American College of Chest Physicians has traditionally defined standard care for acute pulmonary embolism. (PulmCCM is not affiliated with any specialty society.)
Major bleeding at 7 days occurred at ~4% vs ~2% in controls in HI-PEITHO; although not statistically significant, the figure doesn’t help CDT. But the top-line conclusions of this unequivocally positive trial certainly do.
It’s good to know CDT can improve cardiopulmonary physiology, not just measures of RV function. Are harder clinical endpoints too much to hope for? Given the limitations of research on people with more severe PEs, there will probably never be a large RCT testing invasive therapies vs anticoagulation without high crossover. It’s ethically dicey and unreasonable to ask patients to take that kind of risk today. By now, most intensivists believe that these treatments help sometimes; it’s patient selection that remains a challenge.
Until that hypothetical trial is performed, interventionalists will continue to have wide discretion in the use of catheter-based therapies for acute PE—and in selecting the patients they believe might benefit.
Reference
Rosenfield K, Klok FA, Piazza G, et al. Ultrasound-Facilitated, Catheter-Directed Fibrinolysis for Acute Pulmonary Embolism. New England Journal of Medicine. Published online March 28, 2026. doi:https://doi.org/10.1056/nejmoa2516567
I love your analysis. Way to sleuth out the mildly interesting, transient worsening of NEWS score. Most of all, I love your use of "truthiness". I always enjoyed that on the Colbert Report but never thought of using it in my day to day existence. I will guarantee that I use it on rounds sometime today. And ever other day that it applies.
And the editorial was like way to enthusiastic too - written by a guy deep in thrombosis research and an interventional radiology dude.
It’s a composite outcome trial where the composite outcome was powered a lot by what sounded like the best form of p-fishing they could get from a wow that seems high NEWS of 9.0.
Your breakdown on that was excellent!
Between the intervention oriented gang creating the “new”
AHA/accp guidelines and this very shady trial - the rot appears deep.
And you know what might have helped? A tpa arm. Especially in this one - what is catheter directed lysis other than tpa given pretty much on the clot? Does the sonic aided distribution really help? Can it when given peripherally do the same? We know from trials that have better longer term outcomes like pulmonary hypertension at 90d or such that tpa might have this benefit.
The harm signal from the intervention has to be concerning - like a basic rule of the Hippocratic thumb should to be skeptical about cooked up benefits but be wary of harms (even if statistically not strong) harms.