AIRCHILL Translational & Human Performance Development Pipeline

Development pipeline · Version 1.0 · 23 August 2026

From heat-transfer physics to human neurological benefit.

A stage-gated AIRCHILL development pipeline that separates engineering verification, preclinical validation, human factors, human performance, first-in-patient safety, randomized thermal-performance testing and confirmatory clinical efficacy. The programme is designed so that each stage answers one question well enough to justify the next, while keeping helium carrier-gas development distinct from any pharmacological gas-efficacy claim.

Programme logic

Do not ask efficacy before proving exposure.

1 · Mechanism

Prove heat removal

Bench and preclinical stages must establish that AIRCHILL creates a reproducible, quantifiable thermal exposure under realistic ventilation conditions.

2 · Human performance

Prove the device can deliver it

Human studies first establish usability, ventilation continuity, measurement validity, gas-path performance and achievable thermal separation—not neurological efficacy.

3 · Clinical signal

Then estimate benefit

Once exposure is reproducible and safe, randomized patient studies can estimate biomarker and patient-outcome effects in a prespecified phenotype.

4 · Confirmation

Only then power the pivotal trial

The confirmatory effect size comes from AIRCHILL-specific phase-2 data and current control rates, not from external cooling or gas studies.

Primary development thesis. The programme is designed to show and size whether very early controlled respiratory cooling produces a materially earlier and different thermal exposure than modern standard care, whether that exposure can be delivered without compromising ventilation or circulation, and whether it improves patient-relevant outcomes in the intended population.

Architecture rule · OEM-first outside the proprietary core. AIRCHILL does not need to re-invent every mature emergency-device function in order to offer a broad integrated platform. Proprietary engineering is concentrated on respiratory cooling, the safety-critical gas path, control logic and system integration. ECG/monitoring, defibrillation, suction, 5G/telemedicine and similar established functions should preferentially use qualified purchased components, OEM modules or standard interfaces where this reduces development burden without creating disproportionate classification or verification complexity. Those modules still remain inside the integrated system risk analysis, supplier controls, interface verification, EMC/electrical-safety assessment, software/cybersecurity controls and usability validation.

Stage-gated roadmap

The complete evidence ladder.

Stage	Question	Population / setting	Primary output	Indicative scale	Gate to next stage
0 Intended-use & jurisdiction freeze	What exactly is AIRCHILL claiming, and is helium a device-performance constituent, medicinal gas, or combination-product element?	Regulatory / engineering programme	Frozen intended purpose, gas architecture, claims matrix, risk classification hypothesis, FDA Pre-Sub / EU classification questions.	No subjects	Regulatory route is sufficiently defined to design nonclinical and human evidence without creating avoidable rework.
1 Bench metrology	Does the system deliver the intended gas temperature, flow, pressure and FiO₂ accurately?	Test lung, thermal lung, environmental chamber	Calibration, repeatability, condensation, alarm behavior, sensor accuracy, gas consumption, heat-removal curves.	Dozens–hundreds of repeated runs	Stable performance across the intended operating envelope and worst-case conditions.
2 Ex-vivo / airway model	What happens at the airway and lung interface before live-animal testing?	Validated airway / lung models, ex-vivo tissue where appropriate	Mucosal temperature gradients, condensation, resistance, humidity, local thermal load, gas-path deposition/icing risk.	Design-driven	No unresolved local thermal or circuit hazard that requires live-animal discovery.
3 Acute large-animal thermal study	Can AIRCHILL reproducibly cool brain / target-organ temperature while maintaining ventilation?	Randomized porcine physiology model	Direct brain temperature, core temperature, cooling rate, gas exchange, airway pressure, haemodynamics, histology/safety.	~20–30 evaluable animals after variance-based justification	Prespecified thermal advantage plus ventilation and safety gates.
4 Disease-model replication	Does the selected configuration perform in cardiac arrest / ROSC conditions?	Randomized porcine arrest model	Thermal separation during CPR/after ROSC, ROSC performance, haemodynamics, early biomarkers and pathology.	Variance-based	Reproducible exposure and no material resuscitation penalty.
5 Formative human factors	Can intended users deploy the system correctly in realistic EMS conditions?	Paramedics, emergency physicians, ICU staff; manikin/simulation	Critical-task failures, setup time, alarm handling, circuit changes, transport/handover errors, training needs.	Iterative cohorts	Critical use errors reduced to acceptable residual risk.
6 Summative usability validation	Does the near-final interface support safe use by representative users without coaching?	Representative users; simulated use environments	Human-factors validation against critical tasks and hazard-related use scenarios.	Risk-based representative sample	No unacceptable unresolved use-related risk; design sufficiently frozen for clinical use.
7 Human performance feasibility	Can the near-final system deliver stable ventilation and controlled respiratory cooling in humans who already require invasive ventilation?	Stable, already-intubated adults in a controlled hospital environment; no intubation solely for research	Device functionality, delivered gas variables, ventilation stability, airway/gas-exchange safety, modest reversible thermal response, workflow.	~10–20 subjects, staged/sentinel	No device-related serious safety signal; prespecified ventilation and performance success rate met.
8 First-in-patient early feasibility	Can AIRCHILL be started early in the intended emergency population and create meaningful thermal separation?	Selected intubated OHCA patients; narrow phenotype	Time-to-start, brain/core surrogate trajectory, ventilation continuity, safety, protocol adherence, handover, device deficiencies.	~10–30 patients	Thermal separation and safety/adherence gates justify randomized performance testing.
9 Randomized clinical performance trial	Does AIRCHILL reliably produce earlier / greater thermal exposure than standard care?	Multicentre OHCA; same downstream guideline-based care	Prespecified thermal-performance endpoint; key safety endpoints; blinded neurological follow-up as secondary.	~40–120, depending variability	Clear exposure contrast with acceptable safety; variance and control rates adequate for phase-2 design.
10 Phase-2 dose / timing / phenotype study	Which timing, thermal dose and phenotype should enter pivotal testing?	Randomized multicentre intended-use population	Thermal dose-response, biomarkers, safety, 90-day neurological signal, event rates, recruitment/adherence.	~100–300	Prospectively selected regimen/phenotype, clinically credible effect estimate and acceptable benefit-risk.
11 Confirmatory efficacy trial	Does AIRCHILL improve a patient-relevant endpoint?	Multicentre randomized intended-use population	Blinded 90-day neurological outcome (e.g. ordinal mRS or prespecified favourable outcome), mortality and safety; health economics alongside.	Likely several hundred to >1,000; final n from phase-2 data	Predefined statistical success and acceptable benefit-risk support claim expansion / market authorization.
12 Post-market clinical follow-up	Does performance persist in broader real-world use?	Registries / PMCF / implementation cohorts	Rare safety events, subgroup performance, workflow, utilization, long-term outcomes and economics.	Ongoing	Feeds CER, risk management, labeling and future indication expansion.

Human performance programme

Three human-performance studies before a pivotal efficacy question.

HP-1 · Simulated-use performance

Clinician / EMS usability

Representative users run the complete workflow in ambulance, transfer and ICU simulations. Primary outputs: successful setup, time-to-therapy, critical-task failure, alarm recovery, cylinder/gas change, condensation handling, transport fixation and handover.

HP-2 · Controlled human device performance

Already-intubated stable adults

Short, reversible exposure in subjects already requiring invasive ventilation for clinical reasons. The study should not intubate healthy volunteers solely for research. Primary aim: confirm delivered temperature/flow/FiO₂, airway pressures, PaCO₂/PaO₂, humidification and device stability under real human respiratory loads.

HP-3 · Intended-population early feasibility

Emergency workflow + thermal separation

Selected OHCA patients test whether the system can be started early enough and maintained through transport/handover to generate the intended exposure. The neurological endpoint remains exploratory at this stage.

Why not a classic healthy-volunteer intubation study? The clinically relevant AIRCHILL configuration is an invasive ventilatory intervention. Intubating healthy volunteers solely to test device performance would add substantial procedure risk without answering the emergency-workflow question. A more defensible route is simulated human factors first, then short device-performance exposure in people who are already invasively ventilated for clinical reasons, followed by a tightly controlled early-feasibility study in the intended population.

Clinical performance to efficacy

Human patient trials should separate exposure, signal and benefit.

Study	Primary question	Illustrative primary endpoint	Key secondary endpoints	Decision use
FIH / EFS	Can the device be used safely and early enough?	Composite performance success: protocol start achieved + stable ventilation + no device-related serious safety event.	Time-to-start, thermal trajectory, airway/circuit events, haemodynamics, gas exchange, interruptions, handover.	Device and protocol iteration.
Randomized performance	Does AIRCHILL create a reproducible treatment contrast?	Difference in prespecified thermal exposure metric within the early treatment window.	Time below/above target, cooling rate, ventilation metrics, safety, biomarkers, exploratory 90-day mRS/CPC.	Proves the intervention is biologically distinct from standard care.
Phase 2	Which dose/timing/phenotype should be tested?	Protocol-specific biomarker or intermediate clinical endpoint plus safety, with neurological outcome prespecified.	90-day ordinal mRS/CPC, mortality, adverse events, ICU/resource use, treatment interactions.	Locks pivotal regimen, phenotype and effect-size prior.
Pivotal	Does treatment improve patient-relevant outcome?	Blinded 90-day neurological outcome using a prespecified estimand and analysis model.	Mortality, safety, ICU/hospital use, QoL, EQ-5D, long-term care and cost-effectiveness.	Clinical claim / regulatory submission.

Recommended cardiac-arrest phenotype strategy. Start narrow rather than broad: intended invasive-ventilation population, clearly recorded initial rhythm and witnessed status, very early randomisation, exact treatment clocks, and downstream temperature care standardized across groups. Expansion to other arrest phenotypes should be prospectively tested rather than assumed transferable.

Helium branch

Carrier-gas engineering first; therapeutic-gas claims later.

Branch A · physical carrier gas

Helium selected for heat transfer

Development claim: gas composition improves the physical performance of respiratory cooling. Evidence focuses on heat-transfer efficiency, ventilation, sensor accuracy and safety. Regulatory classification still requires authority confirmation because administered helium is a regulated medical gas in major jurisdictions.

Branch B · therapeutic gas claim

Helium claimed as biologically neuroprotective

This is a materially different programme. It would require a pharmacological rationale, dose/exposure justification, gas-quality controls and a dedicated clinical efficacy strategy, potentially invoking medicinal-gas / combination-product requirements.

Decision

Do not mix the hypotheses

Until human data justify otherwise, the AIRCHILL helium programme should test whether helium improves device heat-transfer performance. Any independent neuroprotective claim should remain a separate protocol and regulatory workstream.

Regulatory & ethics gates

Human studies begin only after a frozen risk-based evidence package.

MDR clinical investigations

Human device investigations should be designed under MDR Annex XV and ISO 14155:2026, with a plan capable of confirming or refuting safety/performance claims and scientifically valid observations.

Pre-Sub → IDE / EFS

Use FDA Q-Submission interaction early. A significant-risk early-feasibility study requires appropriate IRB approval and, where applicable, an approved IDE before enrolment.

Gas jurisdiction

Resolve before HP-2 / FIH

Helium is a designated medical gas under US law; if device and gas are jointly required for the intended effect, combination-product questions may arise. EU medicinal-gas implications likewise require explicit authority/NB assessment rather than assumption.

Human protection

Staged exposure

Sentinel enrolment, stopping rules, independent safety review where appropriate, prespecified rescue/reversion to standard ventilation and no research-only intubation of healthy volunteers.

Regulatory sequencing recommendation. Seek formal feedback before the first human gas-path exposure: one EU classification / Notified-Body–authority interaction and one FDA Pre-Sub covering intended purpose, device class, helium role, nonclinical package, first-human population, endpoints and whether the programme sits solely under device regulation or crosses into a medicinal-gas / combination-product pathway.

Submission-ready evidence architecture

Every stage should produce a reusable dossier component.

Design history

Engineering evidence

Requirements, hazard analysis, verification, gas-path characterization, environmental testing, alarm verification, software/controls and human-factors traceability.

Nonclinical

Biological & performance evidence

Bench thermal performance, airway/lung safety, animal thermal dose, pathology, gas exchange, haemodynamics and disease-model replication.

Clinical

Performance → efficacy

Human factors validation, human device-performance report, EFS report, randomized performance trial, phase-2 dose/phenotype decision record and pivotal clinical investigation report.

Economics

Collect once, reuse later

EMS time, consumables, gas consumption, ICU/ward days, ventilation days, complications, rehabilitation, EQ-5D, discharge destination and long-term care should be built into later randomized studies.

Core go/no-go rule. AIRCHILL should not progress because a study is merely “positive.” It progresses only when the result answers the stage-specific question, the confidence interval excludes a development-trivial effect, the safety gate is met, and the result changes a concrete engineering, protocol or regulatory decision.

Development planning document, Version 1.0. Indicative sample-size ranges are planning envelopes, not protocol commitments. Final numbers, endpoints and regulatory route must be frozen from stage-specific variance, current control rates, device configuration and competent-authority / ethics feedback. Related protocol: AIRCHILL Helium Carrier-Gas Translational Study Protocol · Clinical Study Planner · Therapeutic-gas decision records.