Dimitrios Goundis spent his career on peptides that cross membranes other chemists call impassable. On 29 September 2026, one of them reached a human being. Kither Biotech dosed its first COPD patient with KIT2014, a 42-amino-acid molecule that blocks two enzymes at once. The design goal was not potency. It was staying in the lung.
Fourteen months earlier Merck agreed to pay roughly $10 billion for Verona Pharma, whose lead asset, ensifentrine, hits the same two enzymes. Same pathway, opposite engineering problem. Verona's molecule travels to the lung in a nebuliser cup and is exhaled away. Kither's is built not to leave. Investors should read those as one bet and two delivery strategies, and the second one is still unproven.
The pathway Verona and AirNexis already paid for
PDE3 and PDE4 are two phosphodiesterases that sit next to each other in airway smooth muscle. Both relax it. Blocking both at once produces bronchodilation that neither enzyme delivers alone, which is why the combination has drawn commercial interest for a decade. The clinical proof is recent and unusually clean. In Verona's ENHANCE-1 and ENHANCE-2 trials, ensifentrine delivered improvements in 12-hour FEV1 AUC of 87 mL and 94 mL versus placebo at week 12, both at p<0.001. The FDA approved Ohtuvayre on 26 June 2024. Merck agreed to buy the company in July 2025 for about $10 billion and closed the deal that October.
AirNexis Therapeutics ran the same idea as a dry powder. Its AN01 study, NCT07108699, enrolled 195 patients in China and reported treatment differences of 116 mL and 127 mL at week 4. The two datasets are not comparable and must never be read as a head-to-head. Different countries, different background therapy, different endpoints, different weeks.
What both programmes establish is narrow and important. Dual PDE3/4 inhibition produces a real, measurable, investable effect on lung function in COPD. What neither establishes is whether a third delivery route can match it. That is the only question KIT2014 is being asked.
The pathway is de-risked. The delivery is not. Merck's $10 billion bet on Verona validated the mechanism. Nothing has validated a cell-permeable peptide in this indication.
The Phase 1 result is pharmacokinetic. 56 healthy volunteers, doses from 0.1 to 2 mg, plasma below the limit of detection at every stop. Distribution, not efficacy.
36 patients over seven days is the next test. NCT07668518 targets March 2027 for primary completion. That is a mechanism readout on schedule, not a clinical one.
Three routes to the same two enzymes
Engineers will recognise this shape of problem. The delivery question has a settled pattern of failure. A molecule works in a dish, works in an animal, then arrives in humans at the wrong concentration in the wrong compartment. Kither's answer is structural. A 42-amino-acid sequence is short enough to synthesise cheaply and long enough to hold a shape, and the sequence is chosen so the molecule enters airway epithelial cells instead of dissolving in the mucus above them.
The healthy-volunteer data are consistent with that design. Across doses from 0.1 to 2 mg, given for up to seven days per cohort, NCT06659757 recorded no plasma concentration above the limit of detection. The molecule went in and did not come out. Whether it went where it needed to go is the unmeasured part, and it is the whole argument.
| Parameter | Ensifentrine | AN01 | KIT2014 |
|---|---|---|---|
| Owner | Verona, now Merck | AirNexis | Kither |
| Modality | Small molecule | Small molecule | 42-aa peptide |
| Device | Nebuliser, liquid | Dry powder | Dry powder |
| Stage | ✔ Approved 2024 | ◐ Phase 2b, China | ✗ Phase 1b/2a, Australia |
| Lung function reported | ✔ +87 mL, +94 mL vs placebo | ◐ +116 mL, +127 mL vs control | ✗ None yet |
| Main trade-off | Per-dose nebuliser burden | Dry powder handling | Airway exposure unproven in patients |
Stage and lung-function figures from sponsor disclosures and ClinicalTrials.gov, 2024–2026. Not a head-to-head comparison.
KIT2014 chain length
One inhaled peptide inhibiting both PDE3 and PDE4, in a single covalent chain · Kither Biotech, 2026
What Kither is actually betting on
Kither is not claiming a better bronchodilator than Verona's. It is claiming a different pharmacokinetic profile, and the difference is commercial. A nebulised bronchodilator is a device-plus-drug product. Patients buy time and equipment, and adherence decays with the inconvenience. A dry powder inhaler is cheaper and portable, which is why AirNexis has a real market to attack. A peptide that stays in the airway and still comes from a dry powder device would remove both constraints at once. If that holds, the delivery advantage outlives any single molecule.
That is the shape of a durable asset. It is also, today, a shape drawn on a slide. Three things have to go right, and each is separable from the others.
First, the peptide has to reach the airway epithelium at a concentration that blocks both enzymes. Plasma data are consistent with retention and equally consistent with poor absorption. A molecule that dissolves in mucus and is expectorated would look identical in blood.
Second, local exposure has to translate into smooth-muscle relaxation. A 42-residue peptide inhibiting PDE3 and PDE4 inside the cell is a reasonable hypothesis. Reasonable is the operative word.
Third, tolerability has to hold in patients rather than in healthy volunteers. The healthy study ran seven days per cohort in people with no airway disease. The COPD population is where the tolerability question actually lives. It is also a group on background therapy, with comorbidities the healthy-volunteer study never had to account for.
Kither's framing is that delivery is solved and the molecule is now being tested. The registry tells a more cautious version of the same story. NCT07668518 lists an estimated 36 participants across three ascending-dose cohorts in Australia, with primary completion estimated for 31 March 2027. Thirty-six people cannot establish efficacy in COPD. Verona's own registrational base was 1,553 patients across ENHANCE-1 and ENHANCE-2, and the approved endpoint was 12-hour FEV1 AUC at week 12. A study this size is built to inform dose selection and tolerability. The honest read is that 2027 produces a mechanism readout on schedule.
The distribution claim cuts both ways. Plasma below the limit of detection is the right endpoint for a delivery thesis. It is also the endpoint that says nothing about the airway. A molecule can be absent from blood and inert in lung tissue. Tissue exposure and adverse events separate those two cases, and NCT07668518 gives each cohort seven days.
A 36-patient study has one job
That job is safety, tolerability and dose selection in the population that matters. If KIT2014 clears it, Kither will have a delivery platform with a Phase 1b/2a scaffold and something concrete to discuss with investors about a registrational path. If the tolerability signal is noisy, the whole class of inhaled PDE3/4 peptides compresses toward Merck's asset and Kither's premium evaporates. The binary sits inside a very small sample.
For a principal underwriting private life sciences, the useful asymmetry is that being wrong is cheap. A 42-residue peptide costs little to manufacture, the trial is small, and the readout arrives inside roughly two quarters. Being right is not cheap in the same way. Local delivery to the airway is a platform claim, and platforms get re-rated once, early, and hard.
NCT07668518 enrolment
Three ascending-dose cohorts, seven days per cohort, primary completion estimated 31 March 2027 · ClinicalTrials.gov, 2026
What a clean result from NCT07668518 would and would not settle
Would not establish: any lung-function benefit over placebo, comparative efficacy against ensifentrine or AN01, or device superiority over existing nebuliser and dry powder options.
Does a peptide delivery route earn its own premium in COPD?
Probability: 55% — retention holds but without a functional edge over a cheap dry powder, leaving Kither a well-funded specialist rather than a platform.
✅ Arguments for the platform thesis
Confirmation criteria: disclosed airway or sputum exposure, a named Phase 2b protocol, and a large-pharma licensing or partnership announcement.
❌ Arguments against it
Refutation criteria: no disclosed tissue exposure, a noisy tolerability profile, or an early switch to a non-inhaled formulation.
Three ways this resolves by 2027
🟢 Platform re-rating (45%)
Consequences: the dual PDE3/4 field restructures around device burden rather than enzyme choice, and inhalation becomes a delivery market rather than a bronchodilator market.
🟡 Mechanism proven, economics ordinary (35%)
Consequences: Kither proceeds as a well-funded specialist programme, subscale against the inhaled market's economics, and the delivery thesis survives as a footnote.
🔴 The delivery claim fails (20%)
Consequences: inhaled peptides in this indication are shelved, and Kither's value resets to its preclinical dataset.
Airway tissue exposure. Any disclosure of lung or sputum KIT2014 concentrations would resolve the central ambiguity faster than 2027.
Merck's post-acquisition conduct. Pricing, device investment or a late-stage trial in nebulised ensifentrine reveals how much Merck thinks the delivery problem is worth solving.
AirNexis regulatory timing. A filing in China would put a second approved dual PDE3/4 product on a dated timeline and reset every comparator.
Kither's next financing. Round structure and whether existing investors follow is the cheapest live signal on the platform thesis.
Editor's note: Kither's earlier KIT2014 data in healthy volunteers were reported in June 2026. The COPD registry entry, NCT07668518, was first posted in June 2026 and last updated in August 2026, before the 29 September dosing announcement. Where the two dates differ, the registry governs the trial record.
¹ — Cross-trial comparison only. The ensifentrine, AN01 and KIT2014 figures come from separate studies, in separate countries, with different background therapies and different endpoints. No head-to-head trial exists.