Only a few weeks have passed since I last wrote about negative trials for invasive therapy in pulmonary embolism. You wouldn’t think I would need to spend more time talking about negative research, but the HI-PETHO trial is different, because it is reported as positive, and some people seem to agree. Let’s review the details so you don’t fall into the same trap.
The question
Does ultrasound-facilitated, catheter-directed fibrinolysis combined with anticoagulation improve patient outcomes as compared with anticoagulation alone in patients with intermediate risk pulmonary embolism?
The paper
The HI-PEITHO trial: Rosenfield K, Klok FA, Piazza G, Sharp ASP, Ní Áinle F, Jaff MR, Barco S, Goldhaber SZ, Kucher N, Lang IM, Schmidtmann I, Sterling KM, Araszkiewicz A, Arora V, Cires-Drouet R, Coghlan J, Hobohm L, Ito WD, Jacobson K, Kaiser C, Kopec G, Marx K, McElwee S, Meneveau N, Monteleone P, Montero-Cabezas JM, Olivier CB, Park J, Roik M, Sakhuja R, Tego A, Theurl M, Visveswaran G, Vos JA, Young MN, Asch FM, Konstantinides SV; HI-PEITHO Investigators. Ultrasound-Facilitated, Catheter-Directed Fibrinolysis for Acute Pulmonary Embolism. N Engl J Med. 2026 Mar 28. doi: 10.1056/NEJMoa2516567. PMID: 41910345
The methods
HI-PEITHO is a multinational, adaptive-design, open-label, randomized trial from 59 hospitals in the United States and Europe.
Patients
Adult patients (18-80 years) with acute intermediate risk pulmonary embolism (PE) in at least one main or proximal lobar pulmonary artery.
- Patients had to have and RV:LV ratio of 1.0 or higher, abnormal troponin, and at least 2 of:
- Systolic blood pressure less than 111
- Heart rate over 99
- Respiratory rate over 20
Patients with persistent hemodynamic instability were excluded.
Intervention
Ultrasound-facilitated, catheter-directed fibrinolysis with alteplase plus anticoagulation.
The alteplase dose, for some reason excluded from the manuscript, was a bolus of 2 mg per catheter, followed by infusion of 1 mg per catheter per hour for seven hours, for a total dose of 9 mg for unilateral therapy and 18 mg for bilateral therapy.
Comparison
Anticoagulation alone.
Outcome
The primary outcome was a composite of pulmonary embolism–related death, cardiorespiratory collapse or decompensation, or nonfatal, symptomatic recurrence of pulmonary embolism.
Cardiorespiratory collapse or decompensation included:
- Cardiac arrest or CPR
- Signs of shock
- Placement of ECMO
- Intubation
- Use of non-invasive ventilation
- NEWS (national early warning score) that rose to 9 or higher
The results
They enrolled 544 patients (out of 4313 screened), with a mean age of 58, and 43% were female.
The composite outcome favoured the intervention group (10% vs 4%, RR 0.39, 95% CI 2.0-7.7, p=0.005).
There was no difference in PE related mortality. It occurred in 3 patients (1.1%) in the intervention group and 1 patient (0.4%) in the control group. Recurrent PE rates were identical (1 in each group). The difference was entirely in cardiorespiratory decompensation (3.7% vs 10.3%).
Cardiorespiratory decompensation is itself a composite outcome. When you dig into the supplementary data, essentially the entire difference is in the NEWS score 9 or more, with the things you actually care about (CPR/intubation) looking the same or worse with intervention.
The NEWS score did influence clinical behavior. 3% of the intervention group went on to have rescue therapy, as compared to 9% of the control group.
Bleeding events were obviously a little bit higher with thrombolysis, but not dramatically (and also not statistically) so.
My thoughts
Although this is a positive trial with clinical outcomes that sound somewhat patient oriented, I think it is very clear that there is no actual benefit here. I certainly wouldn’t want to be given thrombolysis, whether by catheter or peripherally, if I fell into the intermediate risk group they defined here. This trial essentially proves that a ‘wait and see’ approach, which preserves the option for rescue therapy if you get worse, is the better way to go.
Most importantly, the primary outcome sucks. Not only does the composite include potentially irrelevant outcomes, but it also uses disease specific death rather than all cause mortality in an unblinded trial. The risk of bias is through he roof. They then make things even more complicated by including a composite (cardiorespiratory collapse) within their overall composite, obscuring the true clinical outcomes.
The only thing that actually differed between the groups was the NEWS score. The NEWS score is a meaningless monitor oriented outcome. They don’t even tell us what aspects of the score changed. However, they do tell us very clearly that whatever the NEWS was measuring, it was meaningless. Table 4 presents the outcomes that should have been the primary outcomes of this trial.
There was no difference in all cause mortality. There was no difference in recurrent PE. There was no difference in serious adverse events. Honestly, the editors should be embarrassed to have published this as if it was a positive trial demonstrating good outcomes for patients.
The use of (or “need for”) rescue therapy is an outcome that can easily be misleading in unblinded trials, especially when the rescue therapy being used is completely unproven. This outcome is so biased from the outset that I don’t think it can provide you with any valuable information. Even if two patients look clinically identical, clinicians will always intervene more often on patients who have not received advanced therapy than those who have a fancy catheter in place. In an unblinded trial, we expect to see more rescue therapy in the control group. Its presence here represents the biased methodology, not an important clinical outcome.
It is even more problematic when the ‘rescue’ therapy is the therapy being studied. You are basically arguing that by giving 100% of patients an invasive procedure up front, you can avoid giving 5% of them the same invasive procedure later. That approach is illogical, and clearly results in more harm than necessary. (The results in this trial are slightly more complicated, as there were a variety of rescue therapies given, but the bulk of the rescues were just the exact same ultrasound guided fibrinolysis that was being studied.)
This gets even more complicated when the ‘rescue’ therapy has not been proven to help. You end up with circular logic. Why does catheter directed thrombolysis work? Well, because it prevents the need for rescue therapy. But what was that rescue therapy? Catheter directed thrombolysis. But how do we know it helped? Well, there is evidence it decreases the need for rescue therapy…
Using the need for rescue therapy as an outcome in an unblinded trial, I can basically prove that any therapy is effective, so long as enough clinicians believe in the therapy. In the 1700s, you might compare blood letting to no blood letting for a septic patient, and of course there would be no benefit. However, if you allowed the doctors to use ‘rescue therapies’ if the patient wasn’t improving, you would see that a huge number of patients in the no blood letting group ‘needed’ this ‘rescue therapy’. Therefore, blood letting is great because using it prevents the need for rescue therapy.
At the end of the day, we know for sure that clinical outcomes were identical. Death, recurrence, and adverse events were all identical between the two groups. In one group, 100% of the patients received this invasive therapy with known harms. In the other group, only 9% received it as ‘rescue therapy’. If the clinical outcomes are identical whether you use an intervention in 100% or 9% of patients, that therapy is clearly not providing patients with benefit.
We are, however, left with questions of generalizability. In this dataset, the wait and see approach is clearly better. Rather than jumping to invasive therapy with everyone, you can wait and see which patients get worse, treat only them, and the outcomes are identical. However, the results might not generalize. These are big research institutions. I don’t know, but I could imagine that there is a doctor covering the ICU 24/7. There is probably someone around to repeat your echo at 3 am. There is probably someone around to identify that you are getting worse. You might not have the same level of monitoring available in the average community hospital, and so it is possible that the handful of patients who deteriorate will not be noticed until too late, resulting in worse outcomes.
As they are presented, I think it is pretty clear that this should be interpreted as a negative trial. However, it is important to remember that this is an unblinded trial, and therefore at incredibly high risk of bias. Everything they were measuring was measured by unblinded clinicians. Decisions to start ECMO are going to be biased. If you know the patient is in the middle of getting thrombolytic therapy, you are going to hold off just a little longer to see if they improve. Same for decisions about intubation, rescue therapy, and basically every other decision in this trial. The trial is going to be heavily biased in favour of the intervention group, and so the fact that there is clearly no benefit – even in that context – is concerning.
As compared to the last 2 PE trials I discussed, this trial at least tried to find the right population, by focusing on clinical markers of cardiorespiratory distress. They might have come a little bit closer, but unfortunately with only 1 PE related death (0.4%) in the control group, there was basically no room for thrombolysis to improve outcomes. And that is what they ultimately found. Death was actually a little bit higher in the intervention group. This was a step in the right direction, but it is clear that they are still failing to find a population of PE patients that is sick enough to warrant invasive management or thrombolysis. I still think the bulk of the data suggests that thrombolysis will help some patients who have been labelled ‘submassive’ in the past, but these modern trials are just looking at a population that is far too healthy at baseline.
I think we have more than enough data to know that ultrasound adds absolutely nothing but cost to this procedure. I think the STRATIFY trial (and just basic physiology) suggests very strongly that catheters do not add anything over intravenous thrombolysis. (Kjaergaard 2026) At this point, we should probably abandon these invasive procedures, and future research should focus just on finding a sick enough population of intermediate-high risk patients to benefit from thrombolysis.
As an aside: it is absolutely infuriating to me when modern trials, which are clearly only going to be read in electronic format, hide critical information in the supplementary appendices alone. For example, this manuscript simply does not state the dose of alteplase that was used. How can you possibly publish a scientific paper that randomized patients to alteplase or placebo and not say clearly the dose that you use?! Even if you put it in the supplement, it costs nothing to make the supplement part of the same PDF. Over the last few years, I have noticed journals hiding more and more essential information in supplements. It is not just infuriating, but it also makes appraising science very difficult. There are many ways for doctors to access the primary PDFs, but the supplements are usually only available on the journal website to people who pay for access. My sense is they are doing this on purpose, to try to fight piracy, but it is very bad for science and needs to stop.
Bottom line
You will hear this trial reported as positive, but it very clearly demonstrates no real clinical benefit from catheter directed thrombolysis in this group of intermediate risk pulmonary embolism patients.
More work is needed to identify which higher risk PE patients, if any, benefit from thrombolysis or mechanical thrombectomy.
Morgenstern, J. HI-PEITHO: More negative data on invasive therapy for higher risk PEs, First10EM, May 25, 2026. Available at:
https://doi.org/10.51684/FIRS.145986
Other FOAMed
I debate Weingart about this trial on EMCrit here
Higher risk PE management updates (The STRATIFY and STORM-PE trials)
There are two deep dives into PE management on EM Cases:
Ep 204 High Risk Pulmonary Embolism Management
Ep 203 Intermediate Risk Pulmonary Embolism Risk Stratification, Management and Algorithm
Evidence based medicine is easy
Evidence based medicine resources
References
Kjaergaard J, Bang LE, Sonne-Holm E, Wiberg S, Holmvang L, Lassen JF, Sørensen R, Høfsten DE, Ulriksen PS, Jawad S, Palm P, Søe C, Ersbøll MK, Boesgaard S, Møller JE, Thune JJ, Hassager C, Tilsted HH, Lønborg J, Egstrup M, Kristiansen OP, Seven E, Lindholm MG, Eskesen K, Fanø S, Carlsen J. Randomized trial of low-dose -, ultrasound assisted thrombolysis or heparin for pulmonary embolism. Cardiovasc Res. 2026 Jan 29:cvag038. doi: 10.1093/cvr/cvag038. PMID: 41610160
Rosenfield K, Klok FA, Piazza G, Sharp ASP, Ní Áinle F, Jaff MR, Barco S, Goldhaber SZ, Kucher N, Lang IM, Schmidtmann I, Sterling KM, Araszkiewicz A, Arora V, Cires-Drouet R, Coghlan J, Hobohm L, Ito WD, Jacobson K, Kaiser C, Kopec G, Marx K, McElwee S, Meneveau N, Monteleone P, Montero-Cabezas JM, Olivier CB, Park J, Roik M, Sakhuja R, Tego A, Theurl M, Visveswaran G, Vos JA, Young MN, Asch FM, Konstantinides SV; HI-PEITHO Investigators. Ultrasound-Facilitated, Catheter-Directed Fibrinolysis for Acute Pulmonary Embolism. N Engl J Med. 2026 Mar 28. doi: 10.1056/NEJMoa2516567. PMID: 41910345
