Rentosertib (INS018_055): The world's first AI-discovered TNIK inhibitor to enter the clinic

Rentosertib (development codes INS018_055 / ISM001-055; USAN name announced in 2025-03) is an oral small-molecule TNIK inhibitor. TNIK (TRAF2- and NCK-interacting kinase) is over-activated in IPF lesions and simultaneously regulates multiple downstream fibrotic pathways — TGF-β / SMAD, EMT/FMT, NF-κB and YAP/TAZ — a "multi-pathway anti-fibrotic" logic of hitting several signals at once, theoretically harder to bypass through compensation than single-pathway inhibition.

The asset now has more than one oral formulation. The inhaled solution formulation received a CDE IND on 2026-04-28, becoming the world's first AI candidate to enter "direct-to-lung" clinical research. The inhaled route is both an engineering solution for reducing systemic exposure / optimizing liver safety and a differentiation chip reserved for lifecycle management. But note — oral remains the main line and inhaled the secondary line, and the allocation of resources between them must not be reversed.

1.1 Indication positioning and differentiation claims

The lead indication is idiopathic pulmonary fibrosis (IPF) in adults. IPF is a progressive, fatal interstitial lung disease — with 5-year survival comparable to many solid tumors. Current SOC consists of just two anti-fibrotic drugs, nintedanib and pirfenidone, whose mechanisms are non-selective kinase inhibition (the former) or phenotypic inhibition (the latter); their benefit is mainly "slowing FVC decline", without stopping disease progression or reversing fibrosis, and discontinuation rates are high with notable liver/GI side effects. Disease modification remains a recognized unmet need.

Rentosertib's differentiation has four angles: (1) a first-in-class TNIK target with a new mechanism; (2) a signal of FVC improvement (not merely slowed decline) observed in the 60 mg QD arm of the China Phase IIa — something the market had not seen before; (3) dual oral + inhaled formulation paths; (4) an end-to-end AI platform narrative that can be told to capital markets and BD at the same time. But the "quality" of each differentiator still awaits long-term data — 12 weeks is not the scale of an IPF registration study; long-term FVC, acute exacerbations and liver safety are the keys to setting the label.

2.1 Phase 1 healthy volunteer study

The Phase 1 SAD/MAD (NCT05154240) was completed in healthy volunteers in New Zealand, confirming basic PK, single/multiple-dose tolerability and a preliminary safety profile. FIH risk was absorbed overseas, and China went straight to patient PoC — for Chinese biotechs, this structure of "FIH overseas, PoC in China" is more economical in both time and money.

2.2 China Phase IIa GENESIS-IPF (NCT05938920)

21 centers in China, 71 IPF patients, 12-week double-blind randomized placebo-controlled, three dose arms (30 mg QD / 60 mg QOD / 60 mg QD), with continuation of existing anti-fibrotic therapy allowed. Topline data were released on 2024-11-12 and the full paper published in Nature Medicine in 2025-06, with the abstract indexed in PubMed in 2025. This is a rare case of a domestic biotech publishing positive Phase IIa IPF data in full in Nature Medicine.

This is a sample of "relatively high design strength for an early PoC" — randomized, double-blind, placebo-controlled, multi-dose, 21 centers, ICH-GCP compliant, with clinical supply packaged and labeled under GMP. But it must be acknowledged that 12 weeks is not the scale of an IPF registration study; a 71-patient safety database is insufficient to support a marketing decision; FVC improvement rather than "slowed decline" is a highlight, but whether it can be reproduced in a large sample within a 12-week window is the core question Phase IIb/III must answer.

2.3 US Phase IIa and global supplementary evidence

The US Phase 2a (NCT05975983) is currently shown as ongoing on ClinicalTrials.gov, with leading IPF research centers participating. FDA granted orphan drug designation (ODD) for IPF in 2023-02. "China PoC read out + US PoC running + dual regulatory incentives in place" — these are the objective conditions for subsequently taking a China-origin, global-confirmed route.

This signal is not fatal — elevated liver enzymes with concurrent nintedanib is a known phenomenon — but it turns "how Phase IIb/III handles combination with SOC" into an unavoidable design question. If the pivotal still allows patients to continue nintedanib/pirfenidone (the clinical reality of IPF in China and globally), the liver signal must be managed to an explainable level using stratification, dose adjustment rules, discontinuation rules and DDI assessment; if it is done as a monotherapy comparison (off-SOC), it will face double pressure on ethics and enrollment speed. Neither road is easy, but "a road must be chosen".

Specific protocol-level management measures that can be brought forward: (1) enrollment stratification — by baseline SOC use, baseline liver function and CYP phenotype; (2) dynamic liver function monitoring — every 2 weeks for the first 12 weeks, monthly thereafter, with clear pre-specified dose adjustment thresholds; (3) a DDI sub-study — a PK sub-study on co-administration with nintedanib / pirfenidone to quantify exposure changes; (4) discontinuation and rechallenge rules — defining clearly what liver enzyme elevations require permanent discontinuation and when low-dose rechallenge is possible. Do these well, and the credibility of a positive pivotal will rise significantly; do them poorly, and the same problem from Phase IIa may recur in Phase III, only at larger scale with more visible impact.

This also directly determines the shape of the post-marketing pharmacovigilance system — liver function, acute exacerbations, electrolytes and adherence will be the same set of key monitoring items for both the last mile before launch and post-marketing real-world studies. If China takes the "BTD + rolling review" path, CDE's depth of questioning on this dimension is usually higher than for an ordinary NDA, so the pharmacovigilance plan should take shape at the same time as the pivotal protocol, not be added just before launch.

On the China registration dimension, rentosertib has already accumulated a rare "combination punch":

Looking at the registration path, the biggest advantage is that the China data package is currently stronger than the overseas one — the core Phase IIa was completed in China, CDE BTD has been granted, and the inhaled formulation also received its IND in China first — so the logic of advancing Phase IIb/III China-first is sound. But BTD is not marketing authorization. To translate "BTD + inhaled IND" into an NDA, three things still need resolving:

1. Pivotal design must be fixed early — the question of adding to existing anti-fibrotics (combination) vs monotherapy, the choice of primary endpoint (FVC? or a broader composite?) and duration, and how to handle the liver safety subgroup. The design of this pivotal directly determines success or failure and the shape of the final label.
2. EOP2 communication must be done well — use the CDE breakthrough therapy channel to confirm the pivotal design, comparator and endpoints in one pass. BTD's value lies not just in review speed but in rolling communication; early discussion avoids a costly rerun at the protocol stage.
3. The inhaled formulation's independent track must not slow down the oral formulation — the inhaled IND has been obtained; the strategy should be to advance formulation research and early bridging studies with a small team and small budget, reserving its right as a second-mover differentiated chip. The main resources must be concentrated on the oral pivotal.

IPF in China is not a big-pool race — prevalence is significantly lower than metabolic or cardiovascular disease — but value density is high: annual treatment cost per patient can exceed RMB 50,000, 5-year survival is comparable to most solid tumors, and disease severity gives drugs strong pricing power and bargaining chips in NRDL negotiation. Insilico's public materials cite millions of IPF patients globally, with prevalence rising in China — a combination of disease severity + a not-small patient pool + dissatisfaction with existing SOC.

5.1 Same-target race and mechanism race

5.2 Side-by-side comparison of global IPF candidates

Placing rentosertib among existing global IPF SOC and pipeline candidates makes its mechanistic differentiation and clinical position more intuitive:

5.3 "Early and late" in access and commercialization

On access, it is still "too early": short-term FVC improvement is not a complete payer story; NRDL / national negotiation access requires long-term FVC, acute exacerbation, hospitalization and (possibly) mortality data plus HEOR models; patient affordability and PAP (patient assistance) models have not been disclosed; channel strategies need to be designed separately for oral and inhaled — oral through long-term specialty chronic prescriptions, inhaled requiring device training and drug–device coordination. The "build vs partner" decision for a commercial team need not be made now, but RWE / HEOR design should be brought forward to Phase III and pre-launch, not left to be patched after launch.

Insilico's asset page explicitly states wholly-owned and available for licensing — a fairly proactive BD signal. Combined with the China PoC read out, the US Phase 2a ongoing, FDA ODD, CDE BTD, two Nature-family papers and the Pharma.AI platform narrative, the completeness of diligence materials ready for the table is high. But to win a "big deal" with a large global upfront + high-percentage milestones, three pieces still need to be added: long-term FVC and liver safety data, complete IP/FTO documentation (public leads only go as far as WO2022179528A1), and commercial-grade CMC (especially for the inhaled formulation and CDMO relationships).

6.1 The BD value of the platform narrative

At the BD negotiating table, rentosertib is not a single-point asset — it also carries the implicit identity of "the clinically validated showcase of end-to-end AI discovery". The Pharma.AI platform has so far produced 13 IND-cleared programs and about 30 PCC nominations; rentosertib is the "first ray of light" of this pipeline — a positive pivotal means a step-change up in platform valuation, while a failed pivotal means the whole platform narrative is discounted. This dual identity has concrete implications for choice of BD structure:

The public partner pool (Eli Lilly, Qilu, CMS, Fosun, Sanofi, etc.) is itself an implicit chip for platform credibility. Among MNCs that are active buyers in IPF and anti-fibrotics, the main candidates include Boehringer Ingelheim (with existing OFEV and nerandomilast, there may be pipeline synergy or substitution), Roche (a gap after PRM-151's failure), Pfizer, Takeda, Regeneron and others — any large deal would significantly reshape the company's valuation and platform narrative.

6.2 Supplementary overseas evidence and the MRCT path

Overseas regulators (FDA/EMA) generally will not accept a single body of Chinese evidence for an NDA/MAA, so "fully China-first then file overseas" rarely works in practice. The current US Phase 2a and China Phase 2a are separate structures rather than an MRCT; whether Phase IIb/III can later be upgraded to a China–US / China–EU MRCT is one of the most important strategic choices of the next 12 months. This not only determines the speed of overseas registration but also BD valuation at the time — the earlier an MRCT takes shape, the more willing partners are to negotiate global rights in one go.

Global SoC alignment is also a "structural problem" that must be solved. Usage rates and guideline positions of nintedanib/pirfenidone differ somewhat by region; Phase 2a subgroups suggest SOC combination interacts with the FVC signal. If the China pivotal design is out of step with IPF clinical practice overseas, overseas regulatory acceptance will be discounted.

Oral-line CMC has already gone from 0 to 1: Phase 2a clinical supply was packaged and labeled under GMP, 12 weeks of enrollment across 21 centers saw no supply interruptions, and pharma-grade stability, PK analysis and IRT dispatch have worked. What remains is scale-up and commercialization — public information here is limited, with CDMO relationships, technology transfer capability and commercial manufacturing costs undisclosed, making them high-priority questions in BD diligence and regulatory communication.

7.1 The engineering logic of an inhaled formulation in IPF

Why did Insilico push the inhaled solution formulation to a China IND immediately after the positive Phase 2a? This is not simply "formulation expansion", but has a clear engineering motive:

1. Reducing systemic exposure — delivering drug directly to the lung increases local concentration while lowering systemic exposure, potentially optimizing liver safety
2. Dose reduction potential — lung-targeted delivery may achieve equivalent efficacy at lower doses
3. Lifecycle management — a second formulation extends patent life and market exclusivity
4. Patient compliance — for patients with GI intolerance to oral drugs, an inhaled route offers an alternative

But to turn these four engineering motives into commercial reality still requires mid- to long-term formulation stability data, nebulizer device compatibility (which type of nebulizer, what particle size distribution, adherence with home use), a path for local manufacturing / import licensing, and most importantly — whether it can be clinically proven in IPF patients that the inhaled formulation is superior to the oral one on some dimensions. Each step along this path needs dedicated resources, so "the priority of the inhaled formulation" is a matter of governance discipline.

The cash runway looks sufficient on its own; stacked with a pivotal it will tighten, inevitably driving an external financing or partnership move. "Using a BD upfront to ease pivotal funding pressure" — this path generally works for domestic Chinese biotechs, but only if the BD timing is right and the deal structure reasonable, without selling core rights cheaply at a low valuation.

Compressing all signals into one sentence: the next leg is not "can it continue", but "with what sample, what comparator, what endpoints and whose money it will produce disease-modifying-grade evidence". The inflection points on this path form a staircase:

9.1 Main risk list

The fork between the two scripts is highly concentrated in three things: key choices in the Phase IIb/III design, protocol-level management of liver safety, and aligning BD timing with the funding window. These are precisely the highest-priority actions for the next 12–18 months.