InSilico's AI-Designed Rentosertib Enters Phase 3 in IPF, First for TNIK Target
The registrational trial tests whether an AI-generated drug against an undrugged target, TNIK, can slow lung function decline where a crowded antifibrotic field has struggled.
Executive Summary
- InSilico Medicine Hong Kong Limited registered a 320-patient Phase 3 trial of Rentosertib (INS018_055) in idiopathic pulmonary fibrosis, with a primary completion date of October 30, 2029, and no patients enrolled yet.
- Rentosertib inhibits TNIK, a kinase target for which the dossier finds no other clinical-stage competitor in any indication, making this the first disclosed Phase 3 test of the mechanism.
- The dossier contains zero completed, terminated, or withdrawn trials of Rentosertib in this target/indication/phase combination, so no prior human efficacy readout exists to calibrate expectations for the FVC endpoint.
- The trial enters an IPF landscape with 8 active trials and a moderate difficulty score of 46, competing for patients against nintedanib and pirfenidone as background standard of care, and against multiple other Phase 3 candidates with different mechanisms.
- The trial protocol carries a 'Stable' label with zero recorded change events and the sponsor holds a 100% completion rate across three prior trials, a favorable but small operational base heading into a multi-year registrational program.
The trial
NCT07687459 will enroll 320 adults with idiopathic pulmonary fibrosis (IPF) across sites in China, testing oral Rentosertib against a placebo comparator arm. The trial is registered as Not yet recruiting, with a planned start date of August 30, 2026. The primary endpoint measures the annual rate of forced vital capacity (FVC) decline, a standard lung-function measure of IPF progression, over 52 weeks. Secondary endpoints track diffusing capacity (DLCO), patient-reported quality of life, CT-based fibrosis burden, and time to disease progression events including hospitalization or death.
What is at stake
Rentosertib targets TNIK, a kinase for which AppliedXL's competitive-trials search found no other clinical-stage inhibitor in any indication. The dossier's comparator model flags the drug as potentially first-in-class, but on an unverified basis, meaning the finding reflects an absence of a matched comparator in the data rather than a confirmed absence of competition. No MoA taxonomy entry exists for 'TNIK Inhibitor,' underscoring that the mechanism sits outside established drug classes tracked by the system. There is no completed, terminated, or withdrawn trial of Rentosertib in this indication and phase in the historical-outcomes data, so no prior human efficacy signal exists to anchor expectations for the FVC endpoint.
The competitive field
IPF is not an empty field. The dossier's landscape sample counts 8 active trials, 60% of them Phase 3, with a moderate landscape difficulty score of 46 out of 100. Nintedanib (Boehringer Ingelheim) and pirfenidone (Genentech) remain the approved standard of care, and this trial's own eligibility criteria permit patients to stay on either as background therapy. Other Phase 3 candidates in the space include Boehringer Ingelheim's nerandomilast (PDE4B), Bristol-Myers Squibb's admilparant (LPAR1), Sunshine Lake's HEC585, and PureTech's deupirfenidone (TGF-beta); none target TNIK, so none serve as direct mechanistic comparators.
Operational signal
The trial's protocol shows zero recorded change events since registration, earning a 'Stable' label, though this reading is still a proxy metric since the study has not begun recruiting. InSilico Medicine holds a 100% completion rate across three prior completed trials globally, with zero terminations. That track record is thin, drawn from just three prior studies, but it is a positive baseline as the sponsor commits to a multi-year, 320-patient registrational program with a targeted primary completion date of October 30, 2029.
What the trial can and cannot show
This is a randomized, registrational Phase 3 design with a placebo comparator, the highest-rigor structure available for an IPF efficacy claim. It can, if completed as designed, establish whether Rentosertib slows FVC decline relative to placebo over 52 weeks. It cannot resolve anything before results post, given the trial has not enrolled a single patient and the readout sits more than three years away. The dossier supplies no interim data, no dose-finding result, and no regulatory designation (Breakthrough, Fast Track, Orphan, or Priority Review) for this program, so the current record offers no basis to size probability of success beyond the trial's design and the sponsor's operational history.
This analysis was produced using AI-assisted reporting systems, AppliedXL data, and official public records. These systems undergo editorial review, quality checks, and regular audits by human experts. Errors may still occur, as with any automated system. Always consult the linked primary sources. Read our AI Editorial Policy.