Skip to content

Investors

Development-stage · no CE mark · no FDA clearance · not for patient use

Last reviewed 18 August 2026 · Trust Center / change history

Financing tied to measurable value-creation milestones.

Medical Cooling is preparing a staged development programme for AIRCHILL. Every figure on this page is a planning assumption from the 2026 investor material and the 2027–2031 financial model — not a commitment, forecast or guarantee.

University Medical Center Hamburg-EppendorfBMBF Go-Bio 031A530Therapeutic Hypothermia and Temperature ManagementISMRM 2016 · Magna cum laude
€ 7.5mCurrent round30 months of runway
€ 9–13mRound 2, mid-2029Base case € 11.4m · carries the randomised protocol
€ 30 / 36 / 39mTotal to an outcome claim, low / base / highRun-rate built from a headcount plan, not a placeholder
€ 0Commercial revenue in the base case through 2035Deliberate modelling boundary

Round sizes are trial budgets plus an organisational run-rate. The trial budgets use the formula and per-patient assumptions published in our own study calculator; the run-rate is built from a headcount plan on 2026 German employer contribution rates and published salary surveys, and reproduces the current round to within 3.6 %. No quotations from a CRO, an investigational site or a notified body have been obtained — management planning, not a cost commitment.

Use of funds

What the € 7.5 million would buy.

Roughly € 6.75 million goes directly into the five technical and regulatory risks; € 0.75 million is contingency. The engineering budget is not based on re-developing mature monitoring, suction, ECG/defibrillation or telemedicine hardware from scratch: those functions are intended primarily as purchased/OEM subsystems where their regulatory impact is proportionate. Proprietary engineering spend is concentrated on controlled respiratory cooling, the gas path, control, interfaces and verification of the integrated system. The sixth risk — reimbursement — is deliberately not in the plan, because the base case has to carry without an additional payment. None of this is raised yet.

Team, QMS and regulatory affairs30 %€ 2.250m
Engineering and prototypes25 %€ 1.875m
Verification, validation, external labs20 %€ 1.500m
Preclinical and clinical preparation10 %€ 0.750m
IP, legal and insurance5 %€ 0.375m
Contingency10 %€ 0.750m

Round size and allocation from management planning, Medical Cooling investor deck 2026. Assigning one budget category to one risk is a presentation convention; the category shares themselves are unchanged. Amounts are rounded and are planning values, not quotations. The detailed AIRCHILL standards, regulatory, cost & time matrix shows the OEM-integration versus own-development planning ranges by subsystem.

Milestones

The gates that create measurable value.

Capital is released against evidence, and each gate has a positive target together with a pre-defined decision threshold.

Gate A · 2026–2028

Bench and airway safety

Temperature, flow, FiO₂, humidity, condensation, pressure, oxygen safety, transport and vibration, ISO 18562. To be funded from the current round.

Gate B · 2029

First-in-patient feasibility, randomised

40–80 patients, randomised 1:1 against standard care and deliberately not powered for clinical benefit. Technical performance, process integration, airway safety and temperature-related endpoints. Randomising this early yields our own control event rate and our own time-to-target distribution instead of borrowing both from other trials. Target of the current round and the transition into the follow-on.

Gate C · 2030–2031

Confirmatory study, adaptive and enriched

Start at 400, blinded interim, sample size re-estimated against the control rate observed by then, upper bound fixed in advance. Enrolled where device start within 20 minutes of EMS arrival is demonstrably achievable — that condition is part of the sample size, not a footnote to it, because 600 patients resolve an odds ratio of about 1.6, and that is defensible in the early-cooled population rather than an unselected one. Patient-relevant endpoints. To be funded from the follow-on.

Where these numbers come from. The 45 % control rate and the 1,022-participant comparison are the design assumptions of PRINCESS2. The case for enriching on time to device start is the PRINCESS time-to-cooling subanalysis — cooling started under 20 minutes, shockable rhythm, CPC 1–2, OR 3.25, 95 % CI 1.06–9.97 — which sits inside the PRINCESS trial that missed its own primary endpoint; that limit travels with the figure. The sample-size arithmetic is our own, two-sided α 0.05 at 80 % power, and it reproduces the PRINCESS2 plan to within 3 %. Trial budgets use the formula and per-patient assumptions published in our own study calculator, cross-checked against a Qserve benchmark for a pivotal MDR study and the disclosed ICECAP funding per patient. The run-rate behind the round sizes is built from a headcount plan on 2026 German employer contribution rates and published salary surveys; it reproduces the current round to within 3.6 %. No quotations from a CRO, an investigational site or a notified body have been obtained. All of this is management planning, not a cost commitment.

The programme

Three gates, three questions, three answers.

Gate A is bench work. Gate B is first in patient and is randomised, so it returns a control event rate of our own. Gate C confirms, adaptively and in an enriched population. Each is a package with a scope, a date and an endpoint set — and each needs funding, none of it is raised.

The three development gatesA timeline from 2026 to 2036 with three gates. Gate A, bench and airway safety, runs from 2026 to 2028 and is funded by round one of seven and a half million euro. Gate B slash B2 is one randomised protocol from 2029 to 2032 with 211 patients, of which the first 60 form a prespecified feasibility and safety pilot; it costs five point six million euro within round two of nine to thirteen million. Gate C, the adaptive confirmatory trial, runs from 2032 to 2035, starting at 400 patients with the sample size re-estimated at a blinded interim and an upper bound of 700, enrolling only where cooling can start within twenty minutes; it costs eleven point seven million euro within round three of thirteen to nineteen million. These are management planning targets, not regulatory or clinical commitments.20262027202820292030203120322033203420352036GATE ABench and airway safetyBench, no patientsTemperature, flow, FiO₂, humidity, pressure, oxygen safety, ISO 18562 · cold-air dose ladder, each step released by a safety committeeRound 1 · € 7.5mGATE B/B2One randomised protocol211patientsFirst 60 a prespecified pilot: feasibility and safety, efficacy not analysed · then dose, signal, phenotype — and the effect estimate for Gate CRound 2 · € 5.6m of 9–13mGATE CConfirmatory, adaptive400–700patientsStart 400, blinded interim, sample size re-estimated against our own control rate · enrolled only where cooling can start inside 20 minutesRound 3 · € 11.7m of 13–19mManagement planning from the 2026 investor deck — target values, not regulatory or clinical commitments.Dates, sample sizes and the claim path update after the intended-purpose decision and notified-body feedback.

Swipe the timeline sideways →

Three gates across two protocols. Gate A is bench, preclinical and human performance work, funded by the current round. Gate B/B2 is one randomised protocol whose first 60 participants form a prespecified feasibility and safety pilot before it continues for dose and signal — running them as one study rather than two saves € 1.2m and 22 months. Gate C then starts at 400 and re-estimates its own sample size instead of assuming one. Every gate is an investment opportunity with a scope, a date and a defined endpoint set, and none of it is paid for. The narrow performance and safety claim is reachable after the pilot gate, at roughly € 12.6m; the neurological outcome claim needs the full path. Sources: trial budgets are computed with the formula and per-patient assumptions published in our own study calculator, cross-checked against a Qserve benchmark for a pivotal MDR study and the disclosed ICECAP funding per patient; no quotations obtained. The sample-size arithmetic and the headcount plan behind the round sizes are recorded in our internal assumption register (Annahmen.md, chapters 11a, 11b and 12) — the register is not public, it is the document a data room receives, and every figure on this page is traceable to a numbered chapter in it.Management planning · target values

Market conditions

Nearly two thirds of the manufacturers in our field have already pulled products off the market.

This is the environment the round is raised into, and it is measured rather than asserted: 393 manufacturers under the MDR, surveyed by the German Chambers of Commerce together with MedicalMountains and SPECTARIS. In pneumology, sleep medicine, anaesthesia and intensive care, 63 % of the companies working in the field had taken at least individual products off the EU market — and the examples the study names for that field include emergency ventilators.

Products withdrawn from the EU market under the MDR, by fieldTwo panels from one survey. The upper panel gives the share of manufacturers active in a field who took at least individual products off the EU market: surgical instruments seventy per cent, dentistry sixty-seven, pneumology together with sleep medicine, anaesthesia and intensive care sixty-three, thoracic surgery sixty, trauma and orthopaedic surgery fifty-eight, radiology fifty-eight. The lower panel gives, for the discontinued products, the share with no equivalent alternative on the EU market, just under twenty per cent, a further forty-five per cent not fully compensable, and the remainder compensable.PANEL A · SHARE OF MANUFACTURERS ACTIVE IN THE FIELDwho took at least individual products off the EU market0%20%40%60%Surgical instruments70%Dentistry67%Pneumology, sleep medicine,anaesthesia, intensive care63%Thoracic surgery60%Trauma and orthopaedic surgery58%Radiology58%Named in the survey for our field: ventilator stands, emergency ventilators, sleep diagnostics.PANEL B · OF THE DISCONTINUED PRODUCTSdifferent denominator — cases, not manufacturersjust under 20 %45 %No equivalent alternativeNot fully compensablethe remainder

Swipe sideways to see the full chart →

What the MDR did to the field we are entering. In a survey of 393 manufacturers under the MDR, 63 % of the companies working in pneumology, sleep medicine, anaesthesia and intensive care had taken at least individual products off the EU market — the third-highest rate of all twenty-one fields surveyed, and the examples the study names for that field are ventilator stands, emergency ventilators and sleep diagnostics. For just under a fifth of the discontinued products there is no equivalent alternative on the EU market. That is not a statement about price. It is a statement about regulatory load — and it is the same unmet need the FDA and MDCG 2025-9 breakthrough criteria are written around. The half that points at us: the survey measures incumbents withdrawing existing products, not the prospects of a new entrant, and the same load applies to us — certification costs up 124 %, procedures 150 % longer, and 67 % of companies under ten employees reporting product stops. We plan against that, we do not plan around it.DIHK / MedicalMountains / SPECTARIS 2023, n = 393
What the withdrawal leaves behind

A gap that the regulators themselves have a name for.

For just under a fifth of the discontinued products, the surveyed companies report no equivalent alternative available on the EU market; a further 45 % are not fully compensable. That is the precise condition both accelerated regulatory routes are written around — the FDA’s Breakthrough criterion “no approved or cleared alternatives exist”, and the unmet-need limb of the EU BtX criteria under MDCG 2025-9. The market argument and the regulatory argument are the same argument, reached from two directions.

How both routes work, and what we would file →

The same load applies to us

We are exactly the company size the survey shows getting hit hardest.

The same respondents report the cost of technical documentation up 111 %, notified-body certification up 124 % and procedures running 150 % longer. Product stops are reported by 67 % of companies with fewer than ten employees, against 48 % of those above 250. A quarter of firms intend to move R&D out of the EU. None of that spares a new entrant, and we are on the small side of every one of those splits.

What we do about it is written down rather than hoped for: an open-loop design that keeps the device out of MDR class III, one engineering evidence core reused across MDR, FDA and NMPA, the two breakthrough filings placed early because they are answered in sixty days, and notified-body cost and duration carried in the plan at the levels above — not at pre-MDR levels.

How far this number carries — stated before anyone asks

The survey measures incumbents withdrawing existing products. It is evidence that the field is thinning; it is not evidence that we can fill the space, and we do not present it as such. It is a self-reported industry survey rather than a random sample, so affected companies are likelier to have answered than unaffected ones. It was fielded in mid-2023, and relief measures after that point are not captured. And the 63 % covers the combined field of pneumology, sleep medicine, anaesthesia and intensive care — it is not a figure for emergency ventilators on their own, which is why we quote the study’s own examples rather than a narrower claim.

Source: DIHK, MedicalMountains and SPECTARIS, Befragung zur EU-Medizinprodukteverordnung, December 2023 — 514 questionnaires returned, of which 393 qualified as manufacturers under the MDR; fielded June to August 2023. Field-level figures p. 10 with footnote 8; alternatives p. 11; costs and durations pp. 8–9. Checked at the source document on 17 August 2026.

Market

Built bottom-up, because the top-down number does not exist.

There is no official count of ambulances in Germany — not at the Federal Statistical Office, not at the Federal Highway Research Institute, not at the Motor Transport Authority, and not at the relief organisations. What is documented is deployment density per 100,000 population. We publish the derivation instead of buying a market-research figure.

Structure

Just under 300 districts

Independent emergency-service districts in Germany, about 240 dispatch centres, roughly 85,000 staff. A fragmented market, not key-account business.

Volume

13.1m missions a year

Of which 8.09 million emergency missions, 61.7 %. Ambulances account for 55.8 % of response trips.

Density

4.6–7.3 per 100,000

Daytime ambulance provision, cities versus rural districts. This is the only defensible basis for a fleet estimate; any absolute number is our own extrapolation and is labelled as such.

Payer

€ 8.4bn transport costs

Statutory health insurance expenditure on transport costs, 2022. Devices are financed through user charges, not from a separate investment budget.

Federal Highway Research Institute report M 345 (2024, survey 2020/21) · Government Commission, 9th statement, 7 September 2023 · Wolff, Breuer, Dahmen et al., Notfall + Rettungsmedizin 2024 (50 emergency-service districts, 21.5 million population) · Statutory health insurance statistics via the Government Commission, based on the Federal Audit Office report 2018.

The scale

How many people this is about.

Germany alone, from the registry’s own projection down to the people who reach hospital with a circulation.

How many people this is about, in Germany, each yearA funnel of four stages. Between 54,000 and 67,000 out-of-hospital resuscitations a year in Germany. Circulation returns at some point in 40.2 per cent. 30.3 per cent reach hospital with circulation, roughly 16,000 to 20,000 people a year. 10.9 per cent are discharged alive. A shockable rhythm is present in 22.5 per cent, which is a cross-cut of the total rather than a further step of the funnel.Out-of-hospital resuscitationsGerman Resuscitation Registry, annual report 2024 · own projection54,000–67,000100.0%Return of circulation at any point40.2 % of all resuscitations40.2%≈ 22,000–27,000Admitted to hospital with circulation30.3 % — reference group 33.6 %30.3%≈ 16,000–20,000Discharged alive / 30-day survival10.9 % in the reference group, 8.8 % across the whole dataset10.9%≈ 5,900–7,300CROSS-CUT, NOT A FURTHER STEPA shockable rhythm is present in 22.5 % of all resuscitations — roughly 12,000–15,000 a year. It is a slice of the total, not a subset of the patients above.Around 5,900 ambulances across just under 300 independent services, roughly 20 vehicles each — a fragmented market, not a key-account business.

Swipe sideways →

How many people this is about. Percentages from the German Resuscitation Registry, public annual report 2024; the 54,000–67,000 projection is the registry’s own. The absolute figures are our arithmetic on that range and are marked as such in Annahmen.md, chapter 1, where the calculation is written out. This internal assumption register is not a public source; it is the traceability document provided in the investor data room, and every planning figure on this page maps to a numbered chapter there. We use the reference group throughout, because the full dataset is under-reported. Fleet and market structure from the BASt report M 345 and the Government Commission’s ninth statement.Registry data · own arithmetic marked

Business model

Device, service and consumables — stated as assumptions.

One finding shapes the model more than any other: a single-use ventilation circuit and a reusable one cost almost the same, so the single-use item pays for itself against reprocessing after one cycle. That supports a razor-and-blade structure. Pricing, gross margin, procurement route and adoption still have to be validated with customers and health systems.

Device

Transport ventilator platform

Capital equipment assumption, subject to product configuration and procurement rules.

Service

Training and maintenance

A service layer for uptime, competency and lifecycle support.

Consumables

Procedure-linked recurring revenue

Dependent on the final design, the clinical workflow and reimbursement economics.

Single-use ventilation circuit € 87.89 gross versus reusable circuit € 91.90 gross, dealer prices August 2026. For comparison, the RhinoChill consumable was priced at £ 1,440 per treatment in the 2014 NICE briefing — roughly nineteen times higher. Peer-set gross margin in medical technology runs 37–50 % (Dräger, Getinge, ICU Medical, financial year 2025); for a start-up with a single device the lower half is realistic.

Intellectual property

Granted in Europe and the United States — with clear, concentrated founder ownership.

Both granted patents are held solely by inventor and founder Fabian Temme. The university is no longer part of the patent ownership structure. For investors, that means a concentrated chain of title with no institutional co-owner and a direct, structurally simple path to assign or exclusively license the IP to the operating company as part of the financing.

EP3509683B1

Transportable device and system for lowering the body temperature of a mammal over the airways, in particular of a human, by means of a cooled, oxygen-containing gas flow

  • StatusGranted
  • Grant date24 January 2024
  • ApplicationEP17790957.9
  • A1 publication17 July 2019
  • OfficeEuropean Patent Office
US11395900B2

Transportable device, system and method for providing a cooled, oxygen-containing gas flow

  • StatusGranted
  • Grant date26 July 2022
  • ApplicationUS 16/331,782
  • A1 publicationUS2020/0164173A1
  • OfficeUSPTO
2016Priority application LU93200, 9 September
2017PCT/EP2017/001065 filed, 9 September
2018Published as WO2018046128A1
2022US patent granted
2024European patent granted
Clear founder-controlled IP structure

The patents are currently held solely by founder and inventor Fabian Temme; there is no university co-owner in the patent ownership structure. For the financing, the IP can therefore be assigned or exclusively, worldwide and sublicensably licensed to the operating company directly by the sole owner. This is a defined transaction step rather than a multi-party institutional ownership issue, and the € 0.375 million IP and legal line of the current round includes the documentation and legal work required for that structure.

Checked against the public patent registers on 8 August 2026. Inventor and sole patent holder: Fabian Temme. The ownership position is concentrated with the founder; the university is not part of the patent ownership structure. The corporate IP transfer or exclusive-licence step can therefore be documented directly between the sole owner and the operating company. Records: EP3509683B1 · US11395900B2 · WO2018046128A1. Term runs to 9 September 2037 subject to fees and legal validity. Validation states, annuity payments, opposition status and the scope of the granted claims against the planned product have not been assessed.

Protection

One family. Granted in Europe and the United States.

From the priority filing in 2016 to the expected term end in 2037 — including the part a buyer will find in the register anyway.

The patent family, from priority to expected term endA timeline from 2016 to 2038. The filing chain runs from the priority filing LU93200 on 9 September 2016 through the PCT filing, the WO publication in March 2018 and the EP A1 publication in July 2019. Two patents are granted: US 11395900 on 26 July 2022 and European patent 3509683 on 24 January 2024. The expected term ends 9 September 2037. The registered applicant is a natural person, not the operating company; transfer or an exclusive licence is a condition before closing.’16’18’20’22’24’26’28’30’32’34’36’38US grantedUS11395900B226 Jul 2022EP grantedEP3509683B124 Jan 2024Expected term end9 Sep 2037FILING CHAINPriorityLU932009 Sep 2016PCT filingPCT/EP2017/0010659 Sep 2017WO publishedWO2018046128A115 Mar 2018EP A1 publishedEP17790957.917 Jul 2019WHAT EVERY DUE DILIGENCE FINDSThe granted patents are registered to a natural person,not to Ethicalsaving. Transfer or an exclusive licence is acondition before closing. Validation states, annuities,opposition status and claim scope are not yet checked.

Swipe the timeline sideways →

The patent family. All entries verified in the public patent registers on 8 August 2026 — Level A, primary source. EP3509683B1 granted 24 January 2024 and US11395900B2 granted 26 July 2022, both from priority LU93200 of 9 September 2016, expected term to 9 September 2037. Public records: patents.google.com/patent/EP3509683B1 · /US11395900B2 · /WO2018046128A1. The box on the chart carries the part a buyer will find anyway, so they find it here first. A freedom-to-operate search against BrainCool and BeneChill is still outstanding.Register · Level A

Rights at closing

The licence is not a negotiation after the round. It is signed the same day.

The patents are held by the founder personally and were never assigned to an operating company. That is deliberate, and the practical consequence for an investor is speed: there is one signature between this company and an exclusive, worldwide, sublicensable licence — and it is executed simultaneously with the closing, in the same document set as the financing.

One counterparty

No consent to collect.

The registers name the inventor as sole proprietor of the family. There is no co-applicant, no institutional co-owner and no third party whose consent is needed before rights can move. A licence negotiation that would normally run across several parties runs across one.

No residual obligations

Owned outright, on the owner’s statement.

The rights passed from the university to the inventor and carry, on the owner’s statement, no further obligation to a university or a funding body — no march-in right, no revenue share, no reserved licence. We state that as what it is: an owner’s statement. The assignment deed and the final grant notice belong in the data room, and they are on the list below.

No restructuring first

The rights can move without anything else having to move.

Granting the licence does not depend on a prior reorganisation, on a holding company being formed, on a transfer between entities, or on where the founder is resident. The family sits with one natural person and can be licensed from there directly. That is the reason “secure the rights” is not a workstream inside this financing — it is a signature inside it.

What an investor should read into this — and what not

Read into it that the rights question is short: one owner, no consents, a licence that can be drafted alongside the term sheet and signed at the same table. Do not read into it that the company owns the patents today — it does not, and nothing on this site says otherwise. Until the licence is executed, the correct description is that the company has access to the family through its founder. The scope of the granted claims against the planned product, the validation states, the annuity status and any opposition history are separately not yet examined; they sit on the diligence list below, where they belong.

Diligence

What the next financing is designed to resolve.

Institutional capital can underwrite specific, documentable value-creation milestones rather than an undifferentiated science bet.

Clinical-design gate

Differentiate AIRCHILL from prior cooling strategies

PRINCESS was neutral overall, while its later shockable-rhythm signal helped motivate the prospectively enriched PRINCESS2 programme. TTM2 tested a much later systemic strategy. AIRCHILL’s 2017 endotracheal work defined a separate engineering target: stronger lower-airway heat transfer. Together these results specify the next value-creation study — demonstrate a materially earlier and reproducible thermal exposure, verify airway/ventilation safety, prospectively define the phenotype and then test and size the patient effect.

Execution risk

Founder capacity

Fabian Temme holds a full-time job alongside Medical Cooling. Weekly hours, compensation, a transition date and the secondary-employment and IP terms of that contract are not yet documented. Until they are, neither unpaid full-time availability nor a founder salary is modelled.

Regulatory risk

Class IIb, and notified-body capacity

Classification is preliminary. Planning horizon 18–24 months from application. MDR cost is roughly 90 % internal personnel and only 7 % notified-body fees — a plan built mainly on fees understates it by about an order of magnitude.

IP position

Granted patents with concentrated founder ownership

EP3509683B1 and US11395900B2 are granted from a 2016 priority and are held solely by founder and inventor Fabian Temme. The university is no longer part of the patent ownership structure, avoiding institutional co-ownership. The remaining corporate step is to document an assignment or exclusive licence to the operating company. Validation states, annuities, opposition status and claim scope against the planned product remain separate diligence workstreams.

Model boundaries

What is set to zero, and why

Opening liquidity, debt and non-committed grants are all set to € 0 until a bank statement, contract or decision letter exists. Development spending is not capitalised. The model is explicitly screen-grade and internal.

MedTech Europe Survey 2024 (n = 211): notified-body fees averaging € 136,981 for the quality management system and € 176,202 for technical documentation, ongoing MDR maintenance € 99,648 a year, cost structure 90 % personnel / 7 % fees / 3 % maintenance · Team-NB Survey 2025 · Investor materials are shared directly because the current deck is confidential.

Closing readiness

What must be true before institutional capital closes.

This is a diligence checklist, not a representation that the items are already complete. It separates the current state from the documentary or operating condition that should close the risk.

Requirement
Current state
Closing condition
Evidence / path
Operating-company IP rights
Granted EU/US patents remain founder-controlled.
Executed assignment or exclusive worldwide sublicensable licence to the operating company.
Public patent records; transaction document restricted.
Founder capacity
Weekly hours, compensation, transition date and employment/IP terms are not yet documented publicly.
Documented transition, availability, secondary-employment clearance and IP obligations.
Investor Data Room · execution-risk register.
QA / Regulatory ownership
QMS and regulatory work are funded workstreams, not completed certifications.
Named accountable QA/RA leadership, controlled implementation plan and budget.
Regulatory pathway
EU Class IIb is a planning assumption; US 510(k) versus De Novo remains open.
Written classification/pathway rationale informed by regulator or qualified regulatory interaction.
Technical Gate A
Development-stage platform; verification package is not represented as complete.
Bench, airway-safety and verification evidence against explicit go/stop criteria.
Transaction package
Round planning exists; cap table and transaction records are not public.
Agreed financing terms, cap table, approvals and documentary closing set.

Start a diligence conversation.

The business plan exists in bank-standard, development-bank and venture versions, alongside the 2027–2031 financial model with low, base and high cases. We share them on request.