Mosliciguat: PHocus randomized evidence and the Phase 3 test

Timing and source

Company-guided data window: H2 2026. Source dated 2026-08-06; checked September 16, 2026. [1]

Program and indication

Mosliciguat · Pulmonary hypertension associated with ILD · Phase 2. [1]

The company guides Phase 2 data in PH-ILD during the second half of 2026.

What this event represents

This record tracks an expected disclosure. Enrollment, study completion, database lock and public release are different milestones. Only a confirmed disclosure should move this event into History.

What to check next

At release, distinguish prespecified primary results from exploratory analyses. Check the comparator, participant counts, effect size and uncertainty, missing data, follow-up and safety before judging the result.

Timing limits

Quarter, half-year and year labels retain the disclosed precision. “Early” and “late” are not converted into invented months. Timeline grouping uses a broad sorting bound, not a confirmed readout day. A window overlapping the next two years is retained in full even when its end falls later.

September 8 result / History correction

Roivant disclosed the positive randomized PHocus readout on September 8, earlier than this report’s preserved broad second-half window. Verification on September 24 corrects a missed completed event; it does not change the occurrence date to the audit date. Mosliciguat remains investigational. Initiation of the separate PHrontier Phase 3 trial is a development milestone, not an approval or a second Phase 2 result. [2] [3] [4]

The principal evidence comes from sponsor topline materials and the original data presentation, rather than a recovered full clinical study report. These materials support precise numerical extraction but leave some statistical and safety details unresolved. The registry’s May 26 primary-completion date describes trial operations; September 8 is the public readout date. The estimated 2028 study completion includes later follow-up and does not defer the reported primary result.

Biology / Pulmonary vascular resistance

Interstitial lung disease can damage the lung parenchyma and its vascular bed. Associated WHO Group 3 pulmonary hypertension adds right-ventricular load and may impair exercise capacity. Pulmonary vascular resistance measures a hemodynamic burden; it is not itself a direct measure of fibrosis reversal or survival. An intervention that improves vessel tone may therefore require separate confirmation of functional and clinical benefit. [2] [5] [6]

Mosliciguat is an inhaled soluble guanylate cyclase activator, designed to work independently of nitric oxide and heme. Activation increases cGMP signaling, providing a rationale for vasodilation. An activator is not interchangeable with an sGC stimulator. Inhaled delivery aims to concentrate pharmacological action in the lung, but does not guarantee zero systemic effects or uniform deposition across severely diseased lungs.

PHocus titrated dry-powder inhalation toward 4 mg once daily. The report does not have a validated individual exposure–response threshold or complete systemic PK distribution. The dose schedule is therefore described as tested, without inferring equivalence to oral drugs or other inhaled products. Practical dose delivery, adherence and tolerability remain part of the next-stage question.

Prior evidence / ATMOS is a different population

The earlier ATMOS program evaluated a single inhaled dose in 38 people with WHO Group 1 or Group 4 pulmonary hypertension, not the PH-ILD population. It was open label and nonrandomized. The disclosed per-protocol analysis involved 20 participants across 1, 2 and 4 mg cohorts; exact cohort sizes were not provided in the release. Mean maximal PVR reductions were 25.9%, 38.1% and 36.3%, respectively. [5]

Those acute within-person changes support biological activity, not a chronic comparative treatment effect. Selecting the largest post-dose change differs from a prespecified week-16 endpoint. No placebo-adjusted estimate, responder denominator or survival effect should be reconstructed from these numbers. PHocus supplies a stronger randomized test in the intended disease setting, while preserving the earlier experiment’s limited role.

PHocus / Design and population

The global, blinded Phase 2 trial randomized 135 participants 2:1: 91 received mosliciguat and 44 placebo. The registry specifies masking of participants, care providers, investigators and outcome assessors. Eligible adults had CT-confirmed interstitial lung disease and right-heart-catheter-confirmed pulmonary hypertension, with PVR at least 4 Wood units in the presentation and ability to walk at least 100 meters in six minutes. [2] [3] [4] [6]

The registry excludes WHO Groups 1, 2, 4 and 5, recent lung-disease exacerbation requiring intensified treatment or emergency hospitalization, and resting oxygen above 10 L/min. Baseline mean PVR was 7.0 versus 7.3 Wood units and six-minute walking distance 285.0 versus 261.6 meters. Background PDE5 inhibitor use was approximately 26% versus 25%, and antifibrotic use 54% versus 59%. These imbalances and concomitant drugs belong in interpretation.

The primary endpoint assessed PVR percentage change at week 16; walking distance and NT-proBNP were key secondary measures. The presentation uses nonparametric ANCOVA for PVR and MMRM for walking distance. Detailed intercurrent-event estimands, complete missing-data sensitivity analyses and the full multiplicity procedure have not been independently recovered. A small p value does not fill those reporting gaps.

Efficacy / What remains unproven

The placebo-adjusted PVR effect had a reported 95% interval of −66.2% to −46.4%. Week-24 walking distance improved by an adjusted 52.7 meters, but that later analysis was exploratory with nominal p<0.0001. It should not be relabeled as the primary endpoint. The same distinction applies to the week-24 NT-proBNP result. Longer follow-up can be informative while carrying a different inferential status. [2] [3]

Exploratory time to clinical worsening had HR 0.63 and p=0.0955. This does not establish a statistically significant reduction in clinical worsening, let alone mortality. The composite includes hospitalization, walking decline, transplant, rescue therapy and death. Full component counts and uncertainty are needed before interpreting which process drives that estimate. Cross-trial promotional comparisons cannot establish superiority to treprostinil or another competitor.

Safety / Cough is not the whole profile

The original safety table shows fewer overall TEAEs and less cough proportionally, but more serious events and treatment discontinuations proportionally on active treatment. With unequal 2:1 randomization, percentages and denominators must accompany counts. Three versus one deaths are all-cause observations in the table; they are not automatically treatment-related, nor does the small sample exclude a safety difference. [3]

Grade 3 or higher totals, total treatment-related serious events, exact cumulative exposure and complete adjudication details were not recovered. These remain gaps. Headache was 16/91 versus 3/44 and upper respiratory infection 13/91 versus 1/44; the sponsor attributed none of the latter to treatment. Attribution is a clinical judgment and is distinct from the observed imbalance.

Capital and execution / The confirmatory test

Pulmovant develops the program within Roivant following the Bayer-origin license. Parent-company resources are not a dedicated project budget; this review has not recovered a complete current program-level funding commitment or fully diluted capitalization. A large confirmatory trial and longer follow-up require additional execution and spending even after a positive Phase 2 result. No implied valuation is assigned to the readout. [2] [3] [5]

PHrontier was reported underway, with approximately 375 planned participants randomized 1:1 and a week-24 walking-distance primary endpoint. Background treprostinil is permitted, making combination tolerability and the incremental benefit particularly important. Its planned sample and endpoint are not results. A future readout date is not invented from estimated enrollment speed.

The next evidence threshold is replication of a functional benefit with credible missing-data handling and a more mature safety profile. Hemodynamics, exercise, worsening events and patient experience should remain separate measures. Positive surrogate physiology is a reason to conduct the next trial, not a substitute for its outcome.

Sources / References

1. Roivant Sciences · company disclosure · 2026-08-06
2. Roivant · PHocus results · September 8, 2026 SEC exhibit; checked September 24
3. Roivant · Original PHocus topline presentation · September 8, 2026, slides 8–20 and 23
4. ClinicalTrials.gov · PHocus NCT06635850 · posted September 10, 2026; retrieved September 24
5. Roivant · ATMOS proof-of-concept disclosure · September 10, 2024
6. Ghofrani et al. · PHocus study-design poster · ERS 2025