WIN378 Asthma Trial: The Twice-Yearly TSLP Challenge

The latest WIN378 asthma data suggest that the biological effects of a single dose can last for months. Windward Bio has moved into Phase 3. Those trials must now answer the practical question: will patients have fewer serious attacks when treatment is given this far apart?^[1]

The WIN378 asthma programme moved forward on 8 September 2026, when Windward released interim POLARIS-1 findings and announced that the Phase 3 portion had begun. The antibody targets thymic stromal lymphopoietin, or TSLP. Windward intends to give it twice a year.^[1]

A twice-yearly asthma biologic is already approved in the United States. GSK’s Exdensur received FDA approval on 16 December 2025 for severe eosinophilic asthma in people aged 12 and older. Its target is IL-5, rather than TSLP.^[5][6]

That gives Windward a more specific task than introducing six-month asthma treatment for the first time. It wants to deliver TSLP blockade at a longer dosing interval. Tezspire is the comparison for the target, and Exdensur for the frequency. The results reported here include no head-to-head trial of WIN378 against either medicine.^[1][4][6]

01WIN378 asthma: what the interim analysis includes

POLARIS-1 is set up to move directly from Phase 2 into Phase 3 within one programme. Its Phase 2 portion follows people with uncontrolled, moderate-to-severe asthma for 48 weeks. The randomized, double-blind study compares three subcutaneous dose levels with placebo to help select the doses for further testing.^[1][9]

The first interim cohort includes 98 of the 147 people enrolled. These are the participants covered by this analysis. The difference between the two totals does not tell us how many people dropped out. Nor does the cohort total tell us how many measurements were available for each endpoint at each visit.^[1]

The full WIN378 asthma report needs to give separate counts for people enrolled, people treated and people with measurements available. Follow-up can differ even among participants in the same interim cohort. Each result needs its own denominator so readers can see how much information supports it.

The sponsor lists the programme as NCT07120503. The detailed registry record would not load during this review, so the study description below uses the dated sponsor disclosures. No additional Phase 3 enrolment target or dose strength has been assumed.^[1][12]

02The lung-function figures need careful reading

FEV1 is the amount of air someone can forcefully breathe out in the first second. It is a useful lung-function measure. An improvement in that test, though, does not tell us whether a patient avoided an exacerbation requiring additional treatment.^[7][13]

FEV1 change

Up to +174 mL from baseline; placebo-adjusted improvement up to +256 mL, p=0.033

Exhaled nitric oxide

Reduction up to 24 ppb, or 43%; p=0.006

Blood eosinophils

Reduction up to 200 cells/µL, or 51%; p<0.0001

Reported persistence

Effects reported to last through week 24 after a single dose

Sources: ^[1]

The release uses “up to” for these findings. That gives the upper end of the reported results, not a typical patient’s expected benefit or one estimate shared by all doses. Readers still need the complete results for each treatment arm, with confidence intervals and the number assessed at each visit.

The two FEV1 figures also cannot simply be subtracted to work out how much the placebo group declined. One describes change within a group; the other compares treatments. That calculation would first need confirmation that the dose groups, visits and statistical estimates match.

The WIN378 asthma analysis covers several doses and outcomes. FDA guidance explains why plans for multiple statistical tests need to be set in advance to control false-positive claims. That plan is not available here. The p-values can be reported as interim findings, but they do not settle a regulatory decision.^[10]

The falls in nitric oxide and eosinophils support a drug effect. They do not tell us the percentage reduction in asthma attacks. Attack reduction needs to be measured in a clinical comparison of its own.

03A long half-life helps, but does not settle the interval

WIN378 also appears under the names HBM9378 and SKB378. The earlier first-in-human study enrolled 50 healthy adults in five single-dose cohorts. Each cohort had eight people assigned to the antibody and two to placebo. The indexed paper describes the study as randomized and double-blind.^[2][3]

Harbour BioMed’s account of the publication reports mean half-lives of 55.0–65.8 days, with anti-drug antibodies in 2 of 40 exposed participants. These findings in healthy volunteers help explain why the developers pursued a long dosing interval. They cannot show whether the drug controls asthma.^[2]

The September WIN378 asthma update reports a half-life of up to 75 days. That supports testing less frequent treatment, although the drug concentration still changes between doses. What matters clinically is whether enough effect remains as the next scheduled dose approaches.^[1]

An average over the whole follow-up period could hide weaker protection towards the end of an interval. Results shown over time would help readers see whether that happens. Drug concentration, target engagement, symptoms and attacks each tell us something different. Repeated six-month dosing also needs follow-up of its own; observations after one dose cannot show how later treatment will perform.

There is another consequence of long exposure. Stopping future doses does not immediately clear the drug already in the body. The appeal of less frequent treatment therefore has to be considered alongside tolerability and how adverse reactions would be managed.

04WIN378 asthma versus Tezspire and Exdensur

Tezspire and Exdensur are useful comparisons for different reasons. Tezspire already blocks TSLP. Exdensur is approved for use on a six-month schedule. For the WIN378 asthma programme, sharing one of those features does not make the medicines equivalent overall.^[4][5][6]

In NAVIGATOR, the annualized exacerbation rate was 2.10 with placebo and 0.93 with tezepelumab. The rate ratio was 0.44 (95% CI 0.37–0.53), and 1,061 participants were randomized. Those findings support tezepelumab and the relevance of TSLP as a target. They are not an estimate of WIN378’s efficacy.^[7]

FDA’s Exdensur review describes two placebo-controlled trials lasting 52 weeks that supported approval. Both the molecule and the intended population differ. Their schedules can be compared in a table, but separate trials cannot tell us which treatment is better.^[5]

A six-month anti-TSLP treatment could appeal to people who need that mechanism and find frequent administration difficult. Switching someone whose asthma is already controlled is a separate decision. That would need evidence about the benefits of switching, not just a longer half-life.

05What Phase 3 has to show

The pivotal study in severe asthma compares two dose levels with placebo. Both active-dose regimens are twice yearly; the two levels do not mean four administrations a year. The primary endpoint is annualized exacerbations. A second Phase 3 study, POLARIS-2, is also planned.^[1][9]

For the WIN378 asthma programme, a percentage reduction alone will not be enough to interpret the result. Readers need attack rates in both groups, the treatment difference and its confidence interval, and the follow-up behind those estimates. A relative reduction has less clinical impact when the starting attack rate is already low.

It also matters what treatment participants are already receiving. NAVIGATOR tested an add-on to existing asthma care. Any investigational regimen has to be judged by what it adds in the setting actually studied. Results should not be interpreted as though the patients had otherwise received no treatment.^[13]

Witfire has looked at this problem in Witfire’s analysis of background treatment and trial design. The diseases differ, but the reason to examine the background medicines is the same. What the trial compares sets the limits of the clinical claim its results can support.

The results also need to be read by phenotype. One overall average can cover patients with very different inflammatory profiles and starting risks. Subgroups should be defined in advance, with enough detail to judge the uncertainty in their estimates. A finding in one biomarker group cannot be assumed to apply to everyone.

The full protocol should also explain what happens to the analysis when the programme moves between phases. Calling it “seamless” tells us about its organization. It does not explain whether the populations are pooled, how doses were selected or how statistical error is controlled.^[10]

06Safety and the practical formulation questions

At the interim cutoff, Windward reported no treatment-related serious adverse events, withdrawals or discontinuations. The company put injection-site reactions below 1% and anti-drug antibodies at 2%. These are summaries of the dataset available so far. They do not rule out rare events or effects that appear later.^[1]

The next WIN378 asthma safety report needs to show how many participants received each dose and how long they were followed. Results after another administration will matter too. Without those details, an overall percentage tells us too little about long-term risk.

Several practical formulation details remain unanswered. The report does not establish the final commercial concentration, injection volume, device or number of needle injections per dose. Until those are known, fewer dosing occasions cannot be converted into a precise reduction in clinic visits or injection burden.

Tezspire shows why the device can change how a medicine is used. Its U.S. label allows an appropriately trained patient or caregiver to use the pre-filled pen; other presentations are administered by a healthcare provider. Exdensur’s original label specifies healthcare-provider administration. Those instructions belong to those products. They do not tell us how WIN378 would be given.^[4][6]

Manufacturing and delivery would still need work after a positive trial. Witfire’s report on Unicycive’s manufacturing-related rejection concerns a different drug, but shows why clinical evidence and a product ready for commercial use are separate milestones.

Patients should not read this as a reason to replace a rescue inhaler. The comparator labels describe add-on maintenance treatment, not acute relief, and warn against stopping corticosteroids abruptly. The subject here is drug development; the report does not recommend changing a patient’s treatment.^[4][6]

07WIN378 asthma: funding and the revised timetable

The WIN378 asthma antibody was co-developed by Harbour BioMed and Kelun-Biotech. Windward licensed broad development and commercial rights in 2025, excluding specified territories. The three development names belong to that same antibody. They do not represent three programmes that have succeeded independently.^[2][9]

Windward announced $165 million in crossover financing in May 2026. That brought the total raised since its January 2025 launch to $365 million. The money supports other pipeline work as well as WIN378. These fundraising totals do not show the current cash balance or how much has been budgeted for the asthma trials.^[8]

The statements show a change in timing. In May, Windward expected the first Phase 3 start in Q4 2026. By September, it reported that this stage had begun, ahead of the quarter it had forecast. The statement does not give the first participant’s actual dosing date.^[1][8]

The financing helps Windward carry out the work. But estimating how long it can keep spending requires its current cash, spending profile and programme budget. Dividing the amount raised by an assumed burn rate would give a number without that foundation. There is no forecast here of cash remaining or the need for another raise.

For the business, the useful question is which prescribing decision the evidence could change. Greater convenience might matter to a particular patient group. It does not, by itself, tell us the launch price, what payers would prefer, or future market share and sales.

08Witfire Risk Score: 6.4 out of 10

Witfire’s score reflects editorial judgement about the WIN378 asthma development programme. A higher score means more uncertainty. The five dimensions keep their fixed weights; the ratings are judgements rather than measured probabilities.

Scale: below 4.0 is low; 4.0 to below 7.0 is moderate; 7.0 to 10 is elevated. The index is not a validated prediction model. It does not give the probability of approval, trial failure or patient harm.

Competition is the largest practical challenge in this assessment. A long dosing interval is no longer a category of its own. The results need to show which patients benefit from combining that interval with TSLP blockade.

09WIN378 asthma: the next evidence checkpoints

The next updates need to fill the gaps in what is known. A full dataset would help readers judge the interim findings. The pivotal programme then has to answer the clinical question.

• Full Phase 2 presentation. Check the estimates for each dose, their confidence intervals, the numbers assessed at each visit and the analysis methods.
  Future congress; no date stated
• Repeated-dose follow-up. Look for protection late in the interval, adverse events and immunogenicity after additional treatment.
  Follow-up needed; result date not assumed
• POLARIS-2 initiation. Check the design and confirm that the trial has started, rather than treating the forecast as a completed event.
  H1 2027 plan
• Pivotal exacerbation results. Read the absolute rates and rate ratios with their uncertainty estimates and the background treatment used.
  No topline date assumed
• Regulatory or commercial step. Keep trial progress, submission, approval and product availability separate.
  Each requires a separate future decision

Sources: ^[1][9]

Each WIN378 asthma update should make clear which milestone has changed. A new forecast is still a forecast. An announcement that a trial has begun does not tell us whether the treatment works.

10Witfire verdict

The early WIN378 asthma findings support taking the programme into the next comparison. There is a reason to continue development, but the case against existing treatments needs more than the promise of “only two doses a year”.^[1][4][5]

What is established

Interim signs of biological activity and improved lung function; a Phase 3 start reported by the sponsor.

What is not established

Pivotal evidence of fewer attacks, long-term benefit–risk or superiority to another biologic.

The credible opportunity

TSLP blockade given less often, with useful protection lasting until the next dose.

The commercial question

Which patients would benefit enough to justify starting or switching treatment?

WIN378 would not need to beat every alternative to be useful. A result might support its use in a particular patient group, or show less treatment burden while maintaining meaningful control. Those possibilities can be tested. A claim that it will take over the market would get ahead of the evidence.

Attack rates and tolerability should decide how the final story is told, with dosing convenience considered alongside them. For now, WIN378 remains a development candidate with a potentially useful schedule. It has not been shown to replace current asthma care.

11For Academics: questions behind the headline

The WIN378 asthma announcement can be cited when discussing the interim findings and the decision to move ahead in development. It is not a full pivotal efficacy report. The earlier publication in healthy volunteers and the asthma announcement concern different populations at different stages.^[1][2][3]

1. Who contributes to each estimate? Check the enrolled, treated and analysed populations separately, along with the reasons measurements are missing.
2. Which decisions were made in advance? Review dose selection, interim-analysis rules, multiplicity control and any pooling across phases before calling a result confirmatory.
3. Does protection last until the next dose? Look at the effect through the dosing interval as well as overall. Small exploratory time windows need cautious interpretation.
4. What part does background treatment play? Check controller use, rescue therapy, adherence and changes in treatment in each arm.
5. Would the benefit matter to the intended patients? Read absolute attack rates, patient-reported control, serious adverse events and the precision of subgroup estimates together.

The next report would be much more useful with patient flow and the prespecified analysis beside the outcome table. Those details would help readers judge whether a promising estimate supports a dependable treatment claim.^[10][13]

12WIN378 asthma: seven reader questions

What is WIN378?

WIN378 is a long-acting antibody that targets TSLP and is still investigational. HBM9378 and SKB378 are other names for the same molecule.^[2][9]

What did the WIN378 asthma trial report?

The interim report describes better lung function and lower inflammatory biomarkers. A pivotal reduction in asthma attacks has not yet been established.^[1]

Is WIN378 approved or available for routine treatment?

WIN378 remains investigational. Moving into Phase 3 continues its development; it does not mean the drug has marketing approval or has been launched.^[1][9]

Would this be the first twice-yearly asthma biologic?

No. Exdensur already has FDA approval for severe eosinophilic asthma and is given every six months. Its target is IL-5; WIN378 targets TSLP.^[5][6]

Is WIN378 better than Tezspire?

These results cannot tell us. Tezspire is an approved anti-TSLP biologic, but the interim trial did not use it as an active comparator.^[1][4]

What happens next in the WIN378 asthma programme?

The pivotal programme needs to test whether the drug reduces exacerbations. Windward plans to start POLARIS-2 in H1 2027. That is the company’s forecast.^[1][9]

Could twice-yearly treatment replace daily inhalers?

The announcement does not show that WIN378 can replace inhalers. The approved comparator biologics are add-on maintenance treatments, not rescue medicines. Changes to treatment need clinical supervision.^[1][4][6]