Catheter-Directed Thrombolysis in Intermediate-High–Risk Pulmonary Embolism (PRAGUE-26)

Authors, Journal, Affiliations, Type, DOI

• Authors: J. Kroupa, M. Radvan, J. Mrozek, M. Sluka, S. Jirous, O. Hlinomaz, M. Pliva, J. Pudil, H. Voberkova, K. Bartoskova, M. Poloczek, M. Kamenik, M. Brabec, E. Lichnerova, M. Hutyra, I. Bernat, M. Novak, J. Horakova, L. Jelinkova, T. Knize, P. Tousek, C. Stechovsky, F. Varhanik, Z. Coufal, J. Jarkovsky, and V. Kocka (senior author)
• Affiliations: Department of Cardiology, Third Faculty of Medicine, Charles University, University Hospital Kralovske Vinohrady, Prague; and ten further tertiary cardiovascular centres across the Czech Republic
• Journal: New England Journal of Medicine (published online 31 August 2026; early-release article)
• Type: Investigator-initiated, multicentre, open-label, randomized trial
• DOI: 10.1056/NEJMoa2608012
• Registration: NCT05493163 (PRAGUE-26); EudraCT 2022-002218-18; EU CT 2024-516144-25-00
• Funding: Ministry of Health of the Czech Republic (NU23-02-00446); partnership between Charles University (Cardiovascular Research Area, "Cooperatio" programme) and CarDia (Programme EXCELES, LX22NPO5104; NextGenerationEU). Explicitly conducted without industry sponsorship

Overview

PRAGUE-26 tested conventional catheter-directed thrombolysis against anticoagulation alone in intermediate-high–risk acute pulmonary embolism, as defined by the 2019 ESC guidelines, and found a large reduction in the 7-day composite of death, recurrent PE, or cardiorespiratory decompensation or collapse (0.7% vs 6.8%; RR 0.10; 95% CI 0.02–0.44). The benefit was driven almost entirely by fewer episodes of cardiorespiratory decompensation rather than by fewer deaths or recurrences, and it was achieved without a significant excess of major or clinically relevant bleeding — the trade-off that had limited systemic thrombolysis in this population. Two haemorrhagic strokes occurred in the thrombolysis group, both with identifiable contributory factors. The trial's wider importance is that it is the second randomised trial to support catheter-directed reperfusion here, that it used conventional rather than ultrasound-facilitated catheters, that it enrolled the broader ESC intermediate-high–risk population rather than a additionally-selected subgroup, and that it was free of industry sponsorship.

Keywords

Pulmonary embolism, intermediate-high risk, catheter-directed thrombolysis, catheter-directed fibrinolysis, alteplase, low-dose fibrinolysis, cardiorespiratory decompensation, right ventricular dysfunction, right-to-left ventricular ratio, major bleeding, intracranial haemorrhage, PESI

Key Takeaways

The Clinical Problem

• Patients with intermediate-high–risk acute PE remain at risk of early haemodynamic deterioration despite therapeutic anticoagulation
• Systemic thrombolysis reduces clinical deterioration but its routine use is not recommended in this subgroup, because the benefit is offset by major bleeding including intracranial haemorrhage (the PEITHO precedent)
• Catheter-directed delivery of low-dose fibrinolytic therapy emerged as an alternative intended to preserve efficacy while limiting bleeding
• HI-PEITHO had already shown that ultrasound-facilitated catheter-directed fibrinolysis reduces early deterioration in a selected intermediate-risk population with additional features of impending decompensation; whether the same benefit holds for conventional catheter-directed thrombolysis in the standard ESC intermediate-high–risk definition was the open question PRAGUE-26 addressed

Design

• Investigator-initiated, multicentre, open-label, randomized trial at 11 tertiary cardiovascular centres in the Czech Republic, coordinated by Charles University
• 558 patients randomized 1:1: 280 to thrombolysis plus anticoagulation, 278 to anticoagulation alone
• Median age 64 years; 40.9% female; patients aged 18–80 (those over 80 were excluded)
• Simple stratified randomization with computer-generated sequences
• Patients were included whether directly admitted to a participating centre or transferred from a referring hospital
• Clinical events committee adjudicated all prespecified outcomes; an independent data and safety monitoring board reviewed safety and performed prespecified interim analyses of the primary outcome. Both committees were aware of trial-group assignments
• Because of interim analyses, a P value below 0.047 was considered significant

Population — the ESC Intermediate-High–Risk Definition

Inclusion required all of:

• Proximal PE verified on CT angiography, symptom onset <14 days before randomization
• Intermediate-high–risk PE as defined by the 2019 ESC guidelines: haemodynamic stability, a simplified PESI score ≥1, right ventricular dysfunction, and elevated cardiac troponin or natriuretic peptide
• Age 18–80, and informed consent

Key exclusions shaped the population: active clinically significant bleeding; any haemorrhagic stroke or ischaemic stroke/TIA within 6 months; cranial trauma within 3 months or other active intracranial/intraspinal process; major surgery within 7 days; active cancer or severe illness with expected survival <2 years; haemoglobin <80 g/L, INR >2.0, platelets ≤100×10⁹/L, or creatinine >200 µmol/L (2.26 mg/dL); pregnancy or breastfeeding

• This is a broader population than HI-PEITHO enrolled. HI-PEITHO required additional features suggesting impending decompensation (respiratory rate >20/min or hypoxaemia within a 6-hour window); PRAGUE-26 required only the standard ESC intermediate-high–risk criteria

Intervention — Conventional Catheter-Directed Thrombolysis

• Initiation was to begin within 3 hours after randomization, whenever feasible
• Performed with a standard 4-French valved infusion catheter with a 10-cm infusion segment (e.g. Cragg–McNamara 4-French valved infusion catheter [Medtronic] or similar) — not an ultrasound-assisted system
• Alteplase dosing: 1-mg bolus through each catheter, then 1 mg per hour per catheter to a maximum total dose of 20 mg (10 mg in unilateral PE)
• This is genuinely low-dose: for context, systemic thrombolysis regimens use on the order of 100 mg of alteplase
• Unfractionated or low-molecular-weight heparin was given before randomization; during the alteplase infusion, thrombolysis-group patients received reduced-intensity heparin
• Comparator: anticoagulation alone, per standard care

Primary Outcome — Positive, Driven by Decompensation

Primary outcome: composite of death from any cause, recurrence of PE, or cardiorespiratory decompensation or collapse within 7 days of randomization (intention-to-treat)

• Risk difference −6.1 percentage points (95% CI −9.2 to −3.0; P<0.001), favouring catheter-directed thrombolysis
• The entire treatment effect is carried by the decompensation component. Deaths were 0 vs 4 and PE recurrence 2 vs 1 — neither individually significant. The trial therefore demonstrates prevention of deterioration, not a mortality or recurrence benefit
• Results were consistent in the per-protocol population
• Since the trial had few events (21 in total), the confidence interval around the primary result is wide in absolute terms even though it excludes 1 comfortably

Safety — No Bleeding Excess, But Two Haemorrhagic Strokes

• Clinically relevant bleeding (BARC type ≥2) through day 7: 13 (4.6%) vs 14 (5.0%); P=0.85 — no difference
• Between days 8 and 30: 3 (1.1%) vs 5 (1.9%)
• Major bleeding (GUSTO): 4 (1.4%) vs 6 (2.2%); P=0.54 — no difference
• Stroke within 7 days occurred in 3 patients: 2 (0.7%) haemorrhagic strokes in the thrombolysis group and 1 (0.4%) ischaemic stroke in the standard-care group. No strokes occurred between days 8 and 30
  • Of the two haemorrhagic strokes, one occurred shortly after completion of the alteplase infusion during a hypertensive episode, and one occurred on day 5 in the context of a documented overdose of low-molecular-weight heparin — both had identifiable contributing factors, but both occurred in patients who had received thrombolysis
• Deaths within 30 days: 1 in the thrombolysis group; 4 in the standard-care group died within 7 days
• This is the profile that distinguishes PRAGUE-26 from systemic thrombolysis: reduced deterioration without a demonstrable bleeding excess — although the trial was not powered to exclude small differences in rare serious events

Secondary Outcomes — Faster Early Recovery

• Right-to-left ventricular ratio at 24 hours (transthoracic echo): median 0.94 (IQR 0.83–1.05) with thrombolysis vs 1.07 (IQR 0.94–1.15) with standard care — more rapid relief of RV strain, consistent with ULTIMA
• Median ICU stay: 2.1 days (IQR 1.7–3.0) vs 2.8 days (IQR 1.7–4.0)
• Higher proportion in WHO functional class I at 7 and 30 days with thrombolysis
• The authors argue the clinical relevance of the composite is supported by this consistency across recovery-oriented secondary outcomes
• Secondary confidence intervals were not adjusted for multiplicity and, by the trial's own statement, should not be used to infer definitive treatment effects

Relationship to HI-PEITHO — Four Differences That Matter

The authors note the treatment-response pattern was "similar to that reported in the HI-PEITHO trial", in which the benefit was likewise driven predominantly by fewer decompensation episodes. The trials differ in ways that bear on how the evidence should be read:

• Ultrasound facilitation is not the active ingredient. PRAGUE-26 achieved a comparable result with a plain 4-French valved infusion catheter, which matters for cost and for applicability where ultrasound-assisted systems are unavailable
• The indication is broader than HI-PEITHO's. PRAGUE-26 enrolled patients meeting the standard ESC intermediate-high–risk definition without requiring additional decompensation features
• The industry-sponsorship asymmetry is now resolved. HI-PEITHO was sponsored by a manufacturer of the catheter system it tested; PRAGUE-26, testing the same class of intervention, was independently funded and found the same direction of effect

Limitations of the document

• The primary composite is dominated by a soft component. Cardiorespiratory decompensation or collapse accounted for 17 of the 21 events; death (0 vs 4) and PE recurrence (2 vs 1) were individually non-significant. The trial establishes prevention of deterioration, not improved survival
• Only 21 primary-outcome events in 558 patients, so the absolute numbers behind the headline relative risk are small
• 7-day primary outcome window — the trial says nothing about benefit or harm beyond day 7, and the composite is designed to capture early deterioration only
• Not powered to exclude small differences in rare but serious adverse events. The authors state this explicitly. With two haemorrhagic strokes in the thrombolysis arm and none in the control arm, the safety conclusion of "no bleeding excess" rests on underpowered comparisons
• Both haemorrhagic strokes had contributing factors (a hypertensive episode after infusion; a documented LMWH overdose), which the authors use to argue for careful patient selection and protocol adherence — but this also means the observed ICH rate may be neither the intrinsic rate nor an artifact
• Single country (Czech Republic) with a predominantly White population, limiting generalisability to other racial and geographic populations
• Screening logs were not collected, so the representativeness of enrolled patients relative to all screened patients cannot be assessed
• Patients over 80 were excluded, precluding conclusions in the elderly — a group in which both PE risk and bleeding risk are highest
• Open-label design, with the primary composite partly subjective in its decompensation component; the clinical events committee was aware of group assignment
• Mortality was very low (0 vs 4), which the authors attribute to close trial monitoring and the tertiary-care setting with immediate access to rescue reperfusion. They note that a substantial proportion of PE patients are treated in smaller regional hospitals, where decompensation may carry higher mortality — so the trial may understate the real-world benefit of preventing decompensation, while also being unrepresentative of where most such patients are treated
• Transferred patients were included, which broadens applicability but introduces heterogeneity in time-to-treatment and pre-randomisation care
• No comparison against systemic thrombolysis or against ultrasound-facilitated catheter therapy — the comparator is anticoagulation alone, so the trial does not establish that catheter-directed delivery is preferable to systemic lysis, only that it beats anticoagulation alone

Key Concepts Mentioned

• concepts/Acute-PE-Clinical-Categories — the risk stratification framework within which intermediate-high–risk is defined; PRAGUE-26 provides the second RCT supporting catheter-directed lysis in this category
• entities/Pulmonary-Embolism — the condition
• concepts/Right-Ventricular-Failure — RV dysfunction is both an inclusion criterion and the mechanism of decompensation; the RV/LV ratio improvement at 24 h is the mechanistic readout
• concepts/Venous-Thromboembolism-Anticoagulation — the comparator arm; the interaction between thrombolysis and heparin intensity (including the reduced-intensity heparin during infusion) is directly relevant

Key Entities Mentioned

• Medtronic — manufacturer of the Cragg–McNamara 4-French valved infusion catheter named as an example in the protocol. Mentioned as a device example only; no wiki page exists and none is warranted
• Ministry of Health of the Czech Republic; Charles University / CarDia — funders, with no industry involvement

Wiki Pages Updated

• wiki/sources/CDT-PE-PRAGUE-26-NEJM-2026.md — created (this page)
• wiki/concepts/Acute-PE-Clinical-Categories.md — the "CDL evidence gap resolved by HI-PEITHO" contradiction updated to two trials; category rows updated for the broader population
• wiki/entities/Pulmonary-Embolism.md — catheter-directed thrombolysis evidence updated
• wiki/sourceindex.md — index row
• log.md — ingest entry