# AIRCHILL-STROKE: Ischaemic Stroke Clinical Trial Synopsis > Protocol synopsis for staged AIRCHILL clinical validation in intubated large-vessel-occlusion stroke patients undergoing EVT, including eligibility, randomisation, endpoints, DSMB rules, interim analysis, sample size and… - Canonical: https://www.medicalcooling.com/protocol-airchill-ischaemic-stroke/ - Markdown: https://www.medicalcooling.com/protocol-airchill-ischaemic-stroke/index.md - Language: en-US - Last modified: 2026-08-29T14:25:39+00:00 .prot{--ink:#0B1226;--txt:#3D4756;--line:#E4E8EE;--tint:#F6F8FC;--blue:#13A2EE;--mag:#C11C7A;font-family:Inter,-apple-system,"Segoe UI",Arial,sans-serif;color:var(--ink)}.prot *{box-sizing:border-box}.prot .wrap{max-width:1180px;margin:auto}.prot .hero{padding:112px 24px 76px;background:radial-gradient(900px 520px at 10% 0%,rgba(193,28,122,.46),transparent 62%),radial-gradient(880px 520px at 92% 100%,rgba(19,162,238,.36),transparent 62%),#080E20;color:#fff}.prot .sec{padding:66px 24px}.prot .light{background:var(--tint)}.prot .dark{background:#0B1226;color:#fff}.prot .k{font-size:12px;font-weight:800;letter-spacing:.14em;text-transform:uppercase;color:#F3B8D8;margin:0 0 14px}.prot h1{font-size:clamp(40px,6vw,68px);line-height:1.02;letter-spacing:-.04em;margin:0 0 18px;max-width:17ch}.prot h2{font-size:clamp(28px,4vw,44px);letter-spacing:-.03em;margin:0 0 16px}.prot h3{font-size:19px;margin:0 0 8px}.prot p,.prot li{line-height:1.65}.prot .lead{font-size:18px;color:#C8D6E7;max-width:82ch}.prot .sec .lead{color:var(--txt)}.prot .grid{display:grid;grid-template-columns:repeat(auto-fit,minmax(min(100%,260px),1fr));gap:16px;margin-top:26px}.prot .card{background:#fff;border:1px solid var(--line);border-radius:17px;padding:22px}.prot .tag{font-size:11px;font-weight:800;letter-spacing:.11em;text-transform:uppercase;color:var(--mag);margin-bottom:10px}.prot .table{overflow:auto;border:1px solid var(--line);border-radius:17px;margin-top:26px;background:#fff}.prot table{border-collapse:collapse;width:100%;min-width:900px}.prot th,.prot td{padding:13px 14px;border-bottom:1px solid var(--line);text-align:left;vertical-align:top;font-size:13.5px;line-height:1.5}.prot th{background:#0B1226;color:#fff;font-size:11px;text-transform:uppercase;letter-spacing:.05em}.prot .call{margin-top:24px;border-left:4px solid var(--mag);background:#111A30;color:#C8D6E7;padding:19px 21px;border-radius:0 13px 13px 0;line-height:1.65}.prot .call strong{color:#fff}.prot .num{font-size:34px;font-weight:800;letter-spacing:-.04em}.prot a{color:#0B6FA6}.prot .dark a{color:#8FD8FA}.prot .small{font-size:12.5px;color:#6F7B8C}@media(max-width:760px){.prot .sec{padding:52px 18px}.prot .hero{padding-left:18px;padding-right:18px}} Protocol synopsis · development version 1.0 # AIRCHILL-STROKE: Ischaemic Stroke Clinical Trial Programme A staged, multicentre, randomised clinical-investigation programme designed to determine whether very early AIRCHILL respiratory cooling around endovascular thrombectomy improves neurological outcome in carefully selected large-vessel-occlusion stroke patients who already require general anaesthesia and invasive ventilation. This is a development synopsis, not a final approved CIP. Synopsis at a glance ## Primary confirmatory concept. Design ### Multicentre, randomised, controlled, assessor-blinded 1:1 AIRCHILL plus standard EVT care versus standard EVT care without active AIRCHILL respiratory cooling. Population ### Intubated LVO patients undergoing EVT Adults with acute anterior-circulation large-vessel-occlusion stroke undergoing thrombectomy who require general anaesthesia / invasive ventilation for clinical reasons independent of study participation. Primary endpoint ### 90-day ordinal mRS Blinded analysis of the full modified Rankin Scale distribution, with mortality, dichotomised disability thresholds and health status as key secondary outcomes. Planning N ≈1,600 Working Phase-III target for a common odds ratio around 1.35. Final N should be simulated from AIRCHILL Phase-II mRS distributions and the final estimand. Development sequence ## Use imaging to de-risk the functional-outcome trial. StageTarget NMain purposeGo / no-go output ST-1 · Peri-EVT feasibility60–100Workflow, start timing, ventilation stability, thermal separation, haemodynamic and airway safety during EVT.Show the intended exposure can be delivered without compromising thrombectomy or anaesthesia. ST-2 · Randomised signal finding200–300Imaging core-lab endpoints, early NIHSS, 90-day mRS, treatment timing and phenotype effect estimation.Freeze dose, population, endpoint model and pivotal effect assumption. ST-3 · Confirmatory efficacy≈1,300–1,600 for common OR ~1.35–1.40; >2,000 if ~1.30Definitive patient-relevant functional benefit at 90 days.Multiplicity-controlled confirmatory ordinal analysis. Research question & estimand ## What the pivotal stroke trial is intended to prove. Population Adults with acute ischaemic stroke due to protocol-defined anterior-circulation LVO, selected for EVT and requiring invasive ventilation/general anaesthesia as part of routine care. Treatment AIRCHILL respiratory cooling initiated after airway control and as early as feasible before or during thrombectomy, then continued through a prespecified early reperfusion period. Comparator Standard anaesthesia, ventilation, EVT and temperature management without active AIRCHILL respiratory cooling. Estimand Treatment-policy estimand for the shift in 90-day mRS attributable to assignment to AIRCHILL strategy, with death retained at the worst disability category and crossover/interruption handled explicitly. Core hypothesis. The intended therapeutic contrast is peri-reperfusion neuroprotection delivered through an already secured airway, not elective intubation to administer cooling. The pivotal study must therefore preserve EVT speed and reperfusion quality while moving cooling earlier than conventional post-procedure temperature management. Eligibility framework ## Draft inclusion and exclusion criteria. DomainDraft criterionRationale Age≥18 years; upper age restriction only if prospectively justified.Maintain generalisability while controlling competing risks. Stroke typeAcute anterior-circulation large-vessel occlusion eligible for EVT according to contemporaneous guideline-based practice.Creates a coherent first efficacy population. Airway / anaesthesiaGeneral anaesthesia and invasive ventilation clinically indicated independent of trial enrolment.Prevents study-driven intubation and isolates the intervention to patients already receiving ventilation. ImagingProtocol-defined ASPECTS/core/penumbra criteria appropriate to the selected early and extended-window EVT population; final thresholds frozen after ST-2.Controls baseline injury burden and supports imaging effect modification analyses. TimingRandomisation before arterial puncture or at the earliest feasible peri-EVT point, with maximum randomisation-to-device-start delay prespecified.Protects the early peri-reperfusion hypothesis. Major exclusionsPrimary intracranial haemorrhage; known severe pre-stroke disability above the prespecified threshold; refractory severe haemodynamic/respiratory instability incompatible with study gas delivery; known terminal illness; contraindication identified by device risk analysis; inability to obtain required emergency/deferred consent pathway; other competing interventional trial where prohibited.Safety and interpretability. ConsentProspective representative consent or emergency/deferred consent where legally available, followed by participant consent if capacity returns.Time-critical stroke research requires jurisdiction-specific consent planning. Intervention & co-interventions ## Cooling must never delay reperfusion. AIRCHILL start ### Parallel, not sequential, workflow Device setup occurs in parallel with anaesthesia and EVT preparation. A prespecified maximum allowed workflow delay should trigger protocol deviation or omission rather than delaying groin puncture/reperfusion. Thermal dose ### Reconstruct every exposure Patient-side gas temperature, absolute humidity, FiO₂, flow, airway pressure, minute ventilation, exposure duration, systemic temperature and rewarming are captured to support dose-response analysis. Anaesthesia ### Standardised core principles Prespecify blood-pressure, oxygenation, ventilation and temperature targets and capture anaesthetic agents. Avoid protocol choices that materially differ between study arms except for AIRCHILL. EVT quality ### Protect the dominant treatment effect Capture puncture-to-reperfusion time, final eTICI score, number of passes, rescue therapy and procedural complications. AIRCHILL benefit cannot be interpreted if EVT performance differs materially between groups. Endpoints ## Functional outcome as the proof, imaging as the bridge. LevelEndpoint set Primary confirmatoryBlinded 90-day ordinal modified Rankin Scale shift. Key secondary clinical90-day mortality; mRS 0–2 and 0–1; early and discharge NIHSS; EQ-5D-5L; symptomatic intracranial haemorrhage; malignant oedema/decompressive surgery; ICU/hospital length of stay. Phase-II imagingFinal infarct volume, infarct growth, penumbral salvage where available, oedema metrics, and prespecified imaging-core adjudication. Reperfusion quality remains a procedural covariate/outcome, not an AIRCHILL mechanism endpoint. Thermal performanceTime to first measurable thermal effect, rate/depth of cooling, peri-reperfusion temperature separation, cumulative respiratory thermal dose and rewarming. SafetyDevice-related SAE; haemodynamic instability; arrhythmia; bronchospasm/ventilation compromise; hypoxaemia/hypercapnia; pneumonia; intracranial haemorrhage; procedure delay attributable to study intervention. Health economicsAcute-care cost, ICU/ward days, rehabilitation, discharge destination, readmissions, long-term care and EQ-5D-5L/QALY inputs. Randomisation & blinding ## Central allocation, blinded outcomes and imaging. Allocation Central concealed 1:1 randomisation with variable undisclosed blocks. Stratification Site and one or two high-value baseline variables such as time window and baseline imaging severity; final set selected from ST-2 data without over-stratification. Outcome blinding 90-day mRS assessor masked to allocation using standard structured interview methods. Imaging core laboratory fully blinded. Treatment team Anaesthesia/EVT team cannot be blinded to device assignment. Treatment deviations and reasons are logged contemporaneously. Statistics ## Statistical analysis plan skeleton. TopicPrespecified approach Primary populationIntention-to-treat. Separate safety/exposure set for device-related analyses. Primary modelProportional-odds ordinal logistic regression for 90-day mRS with adjusted common OR and 95% CI, plus prespecified assessment of proportional-odds assumption. Robust alternatives such as generalised odds / win-ratio or utility-weighted sensitivity analysis may be prespecified. CovariatesSmall prespecified prognostic set: age, baseline NIHSS, baseline ASPECTS/core burden, onset/last-known-well to randomisation, and reperfusion-related variable if appropriate and not post-randomisation causal adjustment in the primary model. Imaging analysisCore-lab blinded continuous modelling of infarct volume/growth with appropriate transformation or robust regression; handle death/missing scan with prespecified strategy to avoid survivorship bias. Missing 90-day mRSCentral follow-up programme, multiple imputation under prespecified assumptions and conservative sensitivity analyses. Death coded in the worst outcome category. MultiplicityPrimary ordinal mRS tested first; hierarchical key-secondary sequence for mortality/selected disability threshold if desired; imaging and biomarkers exploratory or Phase-II-specific unless promoted prospectively. SubgroupsPredefined interaction analyses for cooling-start time, successful reperfusion, baseline core size/ASPECTS, collateral status if collected, age and sex. Subgroup findings remain supportive unless separately powered. Sample size & interim ## Power from AIRCHILL data, not from spectacular pilot effects. Working common ORApprox. pivotal NUse 1.30~2,000+Conservative small shift; may require >2,000 depending baseline mRS distribution and covariate adjustment. 1.35~1,600Base planning anchor after positive ST-2. 1.40~1,300Stronger but still far below the very large effect assumptions used in some Phase-II cooling protocols. Formal interim One Phase-III interim near 50% information for overwhelming efficacy and non-binding futility, with alpha-spending fixed in the final SAP. Blinded re-estimation Permit nuisance-parameter/sample-size update based on the observed pooled mRS distribution and missingness if prospectively specified. Why Phase II matters Current cooling trials such as FOCUS and COOLHEAD-2b are valuable design benchmarks but are powered for much larger effects or imaging endpoints. AIRCHILL should not transfer those effect sizes directly. DSMB & stopping domains ## Protect both the airway and the thrombectomy pathway. Device / respiratory Unexpected device-related SAE, clinically significant airway pressure/resistance increase, persistent gas-exchange deterioration, bronchospasm or treatment interruption signal. Haemodynamic Clinically meaningful excess hypotension, vasopressor requirement or arrhythmia temporally associated with treatment. Neurological / procedural Excess symptomatic intracranial haemorrhage, malignant oedema/mortality imbalance, or material delay to arterial puncture/reperfusion attributable to the study intervention. Oversight cadence Early staged safety review in ST-1, then regular unblinded DSMB review by information fraction. Numeric stopping boundaries are finalised after preclinical/ST-1 data and regulatory review. Recruitment model ## What it takes to enrol ~1,600 pivotal stroke patients. ScenarioThrombectomy centresRandomised per centre / yearAnnual enrolmentApprox. active recruitment Conservative2510250~6.4 years Base case4015600~2.7 years High-performance network50201,000~1.6 years Actual recruitment depends on EVT volume, proportion managed under general anaesthesia, eligible vascular territory/time window, competing trials and ability to randomise before reperfusion without delaying care. Site qualification Minimum EVT volume, 24/7 research/randomisation capacity, anaesthesia participation, blinded 90-day outcome pathway, imaging-core transfer and AIRCHILL competency. Run-in dashboard Screening volume, eligible fraction, randomisation rate, device-start delay, puncture/reperfusion times, protocol interruptions and completeness of core imaging/90-day mRS. Quality threshold A site that repeatedly delays reperfusion or cannot deliver the planned thermal exposure should be retrained, paused or removed rather than compensated by additional low-quality enrolment. ST-2 → ST-3 go/no-go ## Commit to Phase III only when mechanism, workflow and patient signal align. Proceed if ST-2 demonstrates acceptable airway/haemodynamic safety, no material EVT delay, reproducible thermal separation, a coherent imaging or early-neurological signal, and a 90-day mRS distribution whose effect estimate and uncertainty support a feasible pivotal sample size. A spectacular small-study effect is not required; a consistent mechanistic and clinical pattern with a realistic confirmatory N is the decision target. Protocol standards & benchmarks ## Framework used for the synopsis. ISO 14155:2026 · EU MDR 2017/745 Annex XV · SPIRIT protocol principles · ICH E9/E9(R1) estimand/statistical principles where applicable · FDA pivotal medical-device / IDE guidance · contemporary thrombectomy-cooling protocols including FOCUS and COOLHEAD-2b, with COTTIS treated only as hypothesis-generating evidence. [Back to the cardiac-arrest & stroke validation plan →](https://www.medicalcooling.com/cardiac-arrest-stroke-clinical-validation-plan/) ## Operational protocol package [Open the full AIRCHILL-STROKE Clinical Investigation Plan →](https://www.medicalcooling.com/cip-airchill-ischaemic-stroke/) [Open the complete Clinical Investigation Package →](https://www.medicalcooling.com/airchill-clinical-investigation-package/)