Rentosertib: The First AI-Discovered TNIK Inhibitor Shows Promise in Idiopathic Pulmonary Fibrosis Phase 2a Trial

Highlight

• Rentosertib is the first AI-discovered small-molecule inhibitor targeting TNIK, a novel pathway in idiopathic pulmonary fibrosis (IPF).
• The phase 2a multicenter, double-blind, randomized placebo-controlled trial showed rentosertib to be safe and well tolerated over 12 weeks.
• At the highest dose (60 mg once daily), rentosertib improved forced vital capacity (FVC), suggesting potential to attenuate lung function decline.
• These results validate the use of generative AI in identifying new therapeutic targets and compounds for difficult-to-treat diseases like IPF.

Study Background and Disease Burden

Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive lung disease characterized by scar tissue formation within the pulmonary interstitium, leading to irreversible lung function decline, respiratory failure, and poor prognosis. IPF predominantly affects older adults and currently has limited treatment options; available antifibrotic agents, such as pirfenidone and nintedanib, can slow disease progression but do not reverse fibrosis or restore lung function. Novel therapeutic targets and interventions are urgently needed to address this high unmet medical need.

Recent advances in artificial intelligence (AI), specifically generative chemistry, have created new frontiers in drug discovery by enabling rapid identification and rational design of novel compounds targeting previously unexplored biological pathways. The present study investigates rentosertib, a first-in-class small-molecule inhibitor of Traf2- and Nck-interacting kinase (TNIK), discovered using generative AI methods. TNIK is implicated in fibrotic processes and represents a novel mechanistic target in IPF.

Study Design

This study was designed as a multicenter, double-blind, randomized, placebo-controlled phase 2a clinical trial (ClinicalTrials.gov identifier: NCT05938920) assessing the safety and preliminary efficacy of rentosertib administered over 12 weeks. A total of 71 patients with diagnosed IPF were randomized into four groups: 30 mg rentosertib once daily (QD, n=18), 30 mg rentosertib twice daily (BID, n=18), 60 mg rentosertib QD (n=18), and placebo (n=17).

The primary endpoint was the incidence of treatment-emergent adverse events. Secondary endpoints included detailed pharmacokinetic profiling (maximum concentration [Cmax], trough concentration [Ctrough], time to maximum concentration [tmax], area under the concentration-time curve [AUC], and half-life [t1/2]), changes in lung function (forced vital capacity [FVC], diffusion capacity for carbon monoxide [DLCO], forced expiratory volume in one second [FEV1]), symptom assessment via the Leicester Cough Questionnaire (LCQ), 6-minute walk distance (6MWD), and incidence and duration of acute IPF exacerbations requiring hospitalization.

Key Findings

Safety profiles were comparable across treatment and placebo arms. Treatment-emergent adverse events occurred in 72.2% (13/18) of patients in the 30 mg QD group, 83.3% (15/18) in the 30 mg BID group, 83.3% (15/18) in the 60 mg QD group, and 70.6% (12/17) in the placebo group. Serious adverse events related to treatment were infrequent and did not differ significantly among groups. The primary reasons for treatment discontinuation involved manageable liver toxicities and diarrhea.

Pharmacokinetic analyses confirmed dose-proportional increases in exposure with predictable pharmacodynamics consistent with once or twice-daily dosing schedules.

Regarding efficacy, the 60 mg QD cohort demonstrated a statistically significant increase in FVC compared to placebo, with a mean change of +98.4 mL (95% confidence interval [CI], 10.9 to 185.9) over 12 weeks, whereas the placebo group showed a mean decline of -20.3 mL (95% CI, -116.1 to 75.6). Changes in DLCO, FEV1, LCQ scores, and 6MWD showed trends favoring rentosertib but did not reach statistical significance within the study duration. The number and length of hospitalizations due to acute exacerbations were numerically reduced in treatment arms but sample sizes limit definitive conclusions.

Expert Commentary

The demonstration of clinical safety and potential efficacy of rentosertib marks a significant milestone for AI-based drug discovery in IPF and broader fibrotic diseases. TNIK inhibition offers a novel and rational mechanistic approach targeting pathways involved in fibroblast activation and extracellular matrix deposition.

Nevertheless, the study’s relatively short duration and limited sample size impose constraints on long-term safety and efficacy interpretations. Extended phase 3 trials with larger patient cohorts and longer follow-up are warranted to confirm the durability of lung function benefits, impact on quality of life, and survival outcomes.

The tolerability profile is encouraging, particularly given the challenges of adverse effects seen with other antifibrotic agents. Liver toxicity and diarrheal events must be carefully monitored. Mechanistically, TNIK’s role in non-canonical signaling emphasizes the importance of continued biomarker development to optimally select patients likely to derive benefit and enhance mechanistic understanding.

This trial also exemplifies how generative AI tools can expedite the drug discovery pipeline, from target identification to candidate optimization and clinical testing, heralding a new era of precision medicine in chronic progressive lung diseases.

Conclusion

Rentosertib, the first AI-generated small-molecule TNIK inhibitor, shows a favorable safety profile and promising signals of efficacy in IPF patients. The phase 2a data open avenues for larger, longer trials to validate clinical benefit and establish this novel agent as a part of the IPF therapeutic armamentarium. Furthermore, this study reinforces the transformative potential of AI-driven drug discovery to address complex, unmet medical needs.

Integration of AI in early and translational drug development could significantly shorten timelines and improve success rates for innovative therapies targeting intractable diseases.

References

Xu Z, Ren F, Wang P, et al. A generative AI-discovered TNIK inhibitor for idiopathic pulmonary fibrosis: a randomized phase 2a trial. Nat Med. 2025 Aug;31(8):2602-2610. doi:10.1038/s41591-025-03743-2. Epub 2025 Jun 3. PMID: 40461817; PMCID: PMC12353801.

Raghu G, Remy-Jardin M, Myers JL, et al. Diagnosis of Idiopathic Pulmonary Fibrosis. An Official ATS/ERS/JRS/ALAT Clinical Practice Guideline. Am J Respir Crit Care Med. 2018 Sep 1;198(5):e44-e68.

King TE Jr, Bradford WZ, Castro-Bernardini S, et al. A phase 3 trial of pirfenidone in patients with idiopathic pulmonary fibrosis. N Engl J Med. 2014 May 29;370(22):2083-92.

Noble PW, Albera C, Bradford WZ, et al. Pirfenidone in patients with idiopathic pulmonary fibrosis (CAPACITY): two randomised trials. Lancet. 2011 May 21;377(9779):1760-9.