RESCUE-3 — Temporary Transvenous Diaphragm Neurostimulation for Weaning

1. Publication

• Title: Temporary Transvenous Diaphragm Neurostimulation for Weaning from Mechanical Ventilation (RESCUE-3): A Randomized Clinical Trial
• Acronym: RESCUE-3
• Year & Journal: American Journal of Respiratory and Critical Care Medicine, epublished June 11, 2025 (print: 2026;212(1):86-94)
• Citation: Dres M, Ewert R, Conrad SA, et al; RESCUE-3 Trial Investigators. Temporary Transvenous Diaphragm Neurostimulation for Weaning from Mechanical Ventilation (RESCUE-3): A Randomized Clinical Trial. Am J Respir Crit Care Med. 2026;212(1):86-94. doi:10.1164/rccm.202505-1056OC

2. Context & Rationale

Background: Diaphragm dysfunction impedes weaning from mechanical ventilation. Temporary transvenous diaphragm neurostimulation (using the Lungpacer/AeroPace system) can increase diaphragm strength, and an earlier feasibility trial (RESCUE-1) and RESCUE-2 (which did not meet its primary endpoint in a difficult-to-wean population) maintained interest in whether patient selection, timing, or trial design could reveal a clinically meaningful effect.

Research Question/Hypothesis: In patients requiring mechanical ventilation ≥96 hours who met readiness-to-wean criteria but failed ≥2 weaning attempts, does twice-daily transvenous diaphragm neurostimulation increase successful weaning at day 30 compared with standard of care?

Why This Matters: Tests a novel device-based approach targeting a specific, mechanistically plausible cause of weaning failure (diaphragm dysfunction) in a difficult-to-wean population where standard approaches have already failed at least twice.

3. Design & Methods

• Study Type: International, multicenter, open-label, randomized clinical trial with a Bayesian primary analysis borrowing prior information from RESCUE-2
• Setting & Centers: 33-48 centers (sources vary: 33 per one summary, 48 per company release), United States and Europe (France, Germany)
• Population:
• Inclusion: Adults requiring mechanical ventilation ≥96 hours, meeting readiness-to-wean criteria, having failed ≥2 weaning attempts
• Exclusions: Not detailed in available trial text
• Intervention: Twice-daily transvenous diaphragm neurostimulation (up to 120 stimulations/day for up to 30 days) plus standard of care
• Comparator: Standard of care alone
• Randomization: 223 patients
• Blinding: Open-label
• Statistical Power & Follow-Up: Primary outcome: successful weaning at day 30. Secondary: duration of ventilation to day 30, mortality at day 30. Bayesian primary analysis borrowed (downweighted) prior information from RESCUE-2. Analysis populations: ITT, modified ITT (successful catheter placement + ability to stimulate ≥1 phrenic nerve, prespecified as the primary analysis population), and per-protocol. Trial stopped early due to slow enrollment.

4. Key Results

223 patients randomized; trial stopped early for slow enrollment.

5. Internal Validity Assessment

• Randomization & Allocation: International randomization; specific concealment mechanics not detailed in available trial text.
• Protocol Adherence & Separation: Prespecified modified ITT population (successful catheter placement + ability to stimulate ≥1 phrenic nerve) as the primary analysis population is a methodologically important choice — excludes patients where the intervention could not be delivered as intended, but also means the ITT (all-randomized) population, including failed catheter placements, is not the headline result.
• Blinding & Detection Bias: Open-label (inherent to a device-based intervention); an unblinded Clinical Events Committee independently adjudicated all serious adverse events — a partial mitigation for detection bias in safety outcomes.
• Missing Data & Sensitivity Analyses: Trial stopped early due to slow enrollment (not for efficacy or futility) — a meaningful limitation, though the Bayesian design with informative prior borrowing from RESCUE-2 was specifically intended to preserve some statistical power despite a smaller-than-planned sample.
• Overall Internal Validity Conclusion: Moderate — the Bayesian approach with prior borrowing is a sophisticated way to extract information from a smaller-than-planned sample, and the high posterior probabilities of benefit (97.9% for weaning, 97.1% for duration) are genuinely promising; however, early stopping for enrollment (not efficacy), the modified-ITT (not strict ITT) primary population, and a notably higher serious-adverse-event rate (36% vs 24%) all warrant caution before considering this a definitive positive trial.

6. External Validity Assessment

• Population Representativeness: Specifically difficult-to-wean patients (≥96h ventilation, ≥2 failed weaning attempts) — not generalizable to routine, straightforward weaning, only to this more refractory subgroup.
• Practice Context: Requires specialized transvenous catheter placement and neurostimulation device capability (Lungpacer/AeroPace system) — a significant procedural and equipment requirement limiting broad, immediate implementation.
• Overall External Validity Conclusion: Moderate — relevant specifically to difficult-to-wean patients in centers with device and procedural capability; the requirement for successful catheter placement and phrenic nerve stimulation capability (defining the primary modified-ITT population) further narrows real-world applicability relative to the full difficult-to-wean population.

7. Strengths & Limitations

Strengths:

• Addresses a genuine, difficult clinical problem (refractory weaning failure) with a novel, mechanistically targeted device-based approach
• Sophisticated Bayesian design with informative prior borrowing from RESCUE-2, maximizing information from a challenging-to-recruit population
• Independent, unblinded Clinical Events Committee adjudication of safety events
• Follows an evidence-building pathway (RESCUE-1 feasibility → RESCUE-2 → RESCUE-3) reflecting iterative, responsible device development

Limitations:

• Stopped early due to slow enrollment, not achieving the planned sample size
• Notably higher serious adverse event rate with active treatment (36% vs 24%) — a real safety signal requiring careful risk-benefit consideration
• Open-label design
• Modified ITT (not strict ITT) as the primary analysis population
• Predecessor RESCUE-2 trial did not meet its primary endpoint, and this trial's Bayesian framework explicitly borrows from and downweights that less favorable prior result
• Device manufacturer (Lungpacer Medical) issued promotional statements describing results as "breakthrough" and "game-changing," a framing that should be weighed against the more measured academic publication conclusion ("high probability of potential benefit... but with a possible increase in serious adverse events")

8. Interpretation & Practice Impact

• Clinical Implications: Transvenous diaphragm neurostimulation shows a high Bayesian posterior probability of benefit for weaning success and reduced ventilation duration in difficult-to-wean patients, but this must be weighed against a notably higher serious adverse event rate — supports cautious, selective consideration rather than routine adoption pending further safety characterization.
• Mechanistic Coherence: Consistent with the biological rationale that diaphragm neurostimulation could counteract ventilator-induced diaphragm dysfunction, a recognized contributor to weaning failure.
• Systems-Level Takeaway: Given the early stopping and safety signal, further confirmatory data (ideally from a completed, adequately powered trial) would strengthen confidence before broader clinical adoption; the honest academic conclusion explicitly flags the adverse-event tradeoff, in contrast to the more enthusiastic industry framing.

9. Controversies & Subsequent Evidence

• Editorial Commentary/Debates: The device manufacturer's press release framed results in strongly promotional terms ("breakthrough," "game-changing," "transform outcomes") — a notable contrast with the peer-reviewed publication's more measured conclusion ("high probability of potential benefit for weaning success but with a possible increase in serious adverse events"). This handbook flags this discrepancy as an important consideration for readers evaluating device-industry-funded trial communications.
• Guideline Integration: Follows directly from RESCUE-2, which did not meet its primary endpoint — RESCUE-3's Bayesian design explicitly incorporates that less favorable prior data, and the resulting more favorable posterior estimate should be interpreted within that context rather than as an entirely independent confirmation.

10. Summary & Executive Takeaway

Summary: RESCUE-3, stopped early due to slow enrollment, randomized 223 difficult-to-wean patients (≥96h ventilation, ≥2 failed weaning attempts) to twice-daily transvenous diaphragm neurostimulation plus standard care or standard care alone. Using a Bayesian analysis incorporating prior RESCUE-2 data, successful weaning at day 30 was higher with neurostimulation (70% vs 61%, adjusted HR 1.34, 97.9% posterior probability of superiority), with reduced ventilation duration (97.1% posterior probability of superiority), but a notably higher serious adverse event rate (36% vs 24%).

Overall Takeaway: Transvenous diaphragm neurostimulation shows a promising, Bayesian-supported signal for improving weaning success and reducing ventilation duration in difficult-to-wean patients, but early trial termination and a real safety signal (higher serious adverse events) mean this should be interpreted as encouraging rather than definitive — particularly given the gap between measured academic conclusions and more enthusiastic industry promotional framing of the same results.

11. Bibliography

• Evans D, Shure D, Clark L, et al. Temporary transvenous diaphragm pacing vs. standard of care for weaning from mechanical ventilation: study protocol (RESCUE-3). Trials. 2019;20(1):60.
• Dres M, de Abreu MG, Merdji H, et al; RESCUE-2 Study Group Investigators. Randomized clinical study of temporary transvenous phrenic nerve stimulation in difficult-to-wean patients (RESCUE-2). Am J Respir Crit Care Med. 2022;205(10):1169-1178.
• Temporary transvenous diaphragmatic neurostimulation in prolonged mechanically ventilated patients: a feasibility trial (RESCUE-1). Crit Care Explor. 2020;2(4):e0106.