Balinatunfib (SAR-441566) is a novel, oral, small-molecule inhibitor of TNF-α, belonging to the anti-inflammatory drug class and being developed as an alternative to traditional biologic TNF inhibitors. Rather than antagonizing TNF-α receptors, balinatunfib exerts its effect through an allosteric mechanism, stabilizing an asymmetrical and receptor-incompetent TNF-α trimer - a unique strategy that disrupts downstream inflammatory signaling [1].

***Figure 1. Crystal structure of the HUMAN TNF-ALPHA IN COMPLEX WITH CMPD 4 (PDB: 7JRA). Balinatunfib structure.***

Balinatunfib originated from Sanofi’s small-molecule immunology program and is being co-developed with UCB Pharma, leveraging fragment-based drug discovery insights from UCB alongside Sanofi’s translational and clinical development expertise.

Clinical progress has seen balinatunfib advance to Phase II trials across multiple inflammatory conditions - such as rheumatoid arthritis, Crohn’s disease, ulcerative colitis, and psoriasis [2]. While a recent Phase II trial in psoriasis did not meet its primary endpoint, Sanofi remains committed to the drug’s potential in other disease contexts and is exploring combination therapy strategies.

The Evolving Landscape of Direct TNF-α Inhibition

For decades, efforts to develop small-molecule inhibitors of TNF-α have faced steep hurdles due to the cytokine’s flat, trimeric protein–protein interaction surface, which resists classical ligand binding. Early molecules like SPD-304 demonstrated proof-of-concept by destabilizing the TNF-α trimer, but issues with potency, selectivity, and aggregation liability prevented clinical progression [4]. Landmark studies later demonstrated that small molecules could effectively “lock” TNF-α into a non-signaling conformation, preventing receptor binding and downstream inflammatory signaling [5–8].

***Figure 2. Representative Chemical Structures of TNF-α small-molecule inhibitors***

This breakthrough inspired multiple research groups, including teams at AbbVie and Bristol Myers Squibb, to apply fragment-based drug design approaches to probe the intermonomeric binding pockets of TNF-α. Through iterative optimization and scaffold-hopping strategies, these efforts yielded potent, selective inhibitors capable of binding deep within the trimer interface. Importantly, many of these compounds demonstrated robust efficacy in murine models of arthritis, providing compelling preclinical validation and setting the stage for the first clinical candidate, balinatunfib, to emerge from this evolving landscape (Fig. 2).

Patent Landscape and Synthetic Evolution

The chemical structure of SAR-441566 is claimed in UCB Biopharma/Sanofi patent WO2016050975A1 [9], which corresponds to the INN balinatunfib (proposed list 131, Aug. 2024).

Insights into the synthetic approaches to balinatunfib can be drawn from this foundational patent, as well as two subsequent patents filed by Sanofi/UCB Biopharma in 2017 (WO2017/167995 A1) [10] and 2018 (WO2018/197503 A1) [11].

The 2016 patent introduces a scaffold of fused pentacyclic imidazole derivatives as TNF-α modulators and describes the multi-step synthesis of key intermediates involving chiral resolution steps. The 2017 filing builds on this foundation, extending the chemistry to variant intermediates and cyclization strategies, while the 2018 patent consolidates these advances by integrating intermediates from earlier work into a complete synthetic route culminating in balinatunfib after final deprotection.

Synthetic Route and Fragment Assembly

The synthetic disclosures across these patents can be organized around the two major fragments of balinatunfib.

For Fragment 1, fused tricyclic benzimidazole (Compound N), synthesis begins from 2-bromo-6-hydroxybenzaldehyde (A) and proceeds through difluoromethylation, Ellman auxiliary imine formation (C), Reformatsky addition, and SNAr chemistry, followed by DIBAL-H reduction and cyanohydrin formation to generate cyclization precursor (H).

After SnCl₂-mediated cyclization and chiral resolution, the key (1R,3S) enantiomer (J) is isolated, with the undesired epimer (K) recycled via a Mitsunobu/deacetylation sequence (Fig. 3).

***Figure 3. Synthetic Sequence for Fragment 1***

Subsequent DPPA/PMe₃ chemistry, intramolecular carbonylative Buchwald–Hartwig amination, and methylation complete the bicyclic lactam fragment (Compound N) (Fig. 4).

***Figure 4. Synthetic Sequence for Fragment 1 (continued)***

Fragment 2 (Boc-protected cyclobutylamine–bromopyrimidine) (S) originates from cyclobutanone (O) and involves Ellman auxiliary addition, lithiation/coupling with bromopyrimidine, auxiliary removal, and Boc-protection in a four-step sequence (Fig. 5).

For the final stage, both fragments are coupled in a one-pot Miyaura borylation/Suzuki–Miyaura cross-coupling reaction, followed by Boc deprotection, to yield balinatunfib in 18 total steps (longest linear sequence: 14 steps) (Fig. 5).

***Figure 5. Synthetic Sequence for Fragment 1 and final cross-coupling/deprotection sequence that led to balinatunfib***

Together, these disclosures establish the foundation for applying our retrosynthesis platform to streamline and optimize the synthetic pathway for balinatunfib.

ChemAIRS-Driven Synthesis of Balinatunfib: Streamlined Access to Key Intermediates

Retrosynthetic analysis of balinatunfib using ChemAIRS identified 10 potential synthetic routes within 20 minutes. Analysis of these routes revealed that Route R001 directly recapitulates the published synthesis of balinatunfib (Fig. 6), demonstrating the platform’s ability to source vendors for both simple building blocks and advanced intermediates with defined stereochemistry.

For example, the known precursor 1a (intermediate J in Fig. 3) leading to the fused tricyclic benzimidazole fragment 4a is commercially available for under $10/gram, while the protected cyclobutylamine–bromopyrimidine fragment 4b is available at a much higher cost of $1078/gram.

Notably, R001 was the only route incorporating the one-pot Miyaura borylation/Suzuki–Miyaura cross-coupling strategy from the 2018 patent. This not only highlights ChemAIRS’ ability to align with patented, state-of-the-art methods but also demonstrates the platform’s diversity in generating alternative routes that balance novelty, efficiency, and practicality.

***Figure 6. ChemAIRS-Driven Synthesis of Balinatunfib. Route 001***

ChemAIRS-Driven Synthesis of Balinatunfib: Human–AI Synergy in Retrosynthetic Planning

Moving beyond the known synthesis, routes R002 and R006, originally suggested by ChemAIRS, were evaluated and strategically merged under expert guidance to maximize synthetic efficiency, yielding a particularly compelling hybrid route (Fig. 7).

***Figure 7. ChemAIRS-Driven Synthesis of Balinatunfib. Hybrid route combining R002 and R006***

This hybrid route featured several key highlights, including:

It retained the inexpensive 1a precursor and, unlike R001, replaced the stoichiometric titanium-mediated amination with a continuous-flow azidation/reduction sequence (12), offering improved workup and scalability.
While ChemAIRS proposed cyclobutanol 2a as a starting point, chemists’ strategic oversight optimized the route by selecting and integrating the cheaper precursors 1a and 1b (Fig. 8).

***Figure 8. Two-Step Synthesis of Mesylate 3a from Cyclobutanone 1b***

Mesylation of 2a enabled a flow-based azidation/reduction to intermediate 5a (Fig. 9), minimizing the handling of hazardous intermediates and improving both safety and scalability.
Subsequent Boc protection and Miyaura borylation furnished 10b for Suzuki–Miyaura coupling with 10a, and Boc deprotection of the resulting 11a delivered balinatunfib(Fig. 10) in 10 total steps with a longest linear sequence of 7 ( including the continuous-flow sequence).

***Figure 9. Synthesis of 5a via Flow-Based Azidation/Reduction and Subsequent Two-Step Elaboration to 10b***

***Figure 10: Final Steps in the Synthesis of Balinatunfib***

The hybrid route combining R002 and R006 for balinatunfib exemplifies the synergistic power of AI-driven retrosynthesis and human expertise. While ChemAIRS rapidly identified multiple viable synthetic pathways, expert chemists critically evaluated and strategically merged the strengths of R002 and R006 to optimize cost, safety, and scalability. This collaboration enabled the replacement of costly precursors, elimination of stoichiometric titanium reagents, and incorporation of continuous-flow azidation/reduction, ultimately delivering a streamlined 10-step synthesis with a longest linear sequence of just seven steps. The result highlights how AI can accelerate route discovery while human insight refines and adapts these routes into practical, efficient, and industrially relevant processes.

Conclusion

The case study of balinatunfib (SAR-441566) illustrates how AI-powered retrosynthesis and human expertise can converge to overcome synthetic complexity in modern drug discovery. By rapidly generating diverse synthetic pathways, ChemAIRS accelerates the early stages of route planning, while chemists bring essential domain knowledge to refine and adapt these routes into practical, cost-effective, and scalable solutions.

At Chemical.AI, we aim to amplify the creativity and problem-solving skills of chemists by pairing their expertise with the predictive power of our AI-driven retrosynthesis platform, ChemAIRS. By rapidly generating and evaluating synthetic strategies, ChemAIRS helps research teams navigate from initial molecular concepts to optimized, scalable routes, accelerating the journey from discovery to drug candidate.

Reference

Vugler, A.et al. An orally available small molecule that targets soluble TNF to deliver anti-TNF biologic-like efficacy in rheumatoid arthritis. Front. Pharmacol. 2022, 13, 1037983.
Nassr, N.et al. First-in-Human Single and Multiple Ascending Dose Studies of Balinatunfib, a Small Molecule Inhibitor of TNFR1 Signaling in Healthy Participants. Clin. Pharmacol. Ther. 2025, 118 (1), 164-176.
Guide to Pharmacology. balinatunfib (SAR-441566), Ligand-ID 13583. IUPHAR/BPS Guide to PHARMACOLOGY. guidetopharmacology.org
He M. M.et al., Small-Molecule Inhibition of TNF-α. Science 2005, 310 (5750), 1022–1025.
O’Connell, J. et al. Small Molecules That Inhibit TNF Signalling by Stabilising an Asymmetric Form of the Trimer. Nat. Commun. 2019, 10, 5795.
Xiao, H.-Y.et al. Biologic-like In Vivo Efficacy with Small Molecule Inhibitors of TNFα Identified Using Scaffold Hopping and Structure-Based Drug Design Approaches. J. Med. Chem. 2020, 63(23), 15050-15071.
McMillan, D.et al. Structural Insights into the Disruption of TNF-TNFR1 Signalling by Small Molecules Stabilising a Distorted TNF. Nat. Commun. 2021, 12 (1), 582.
Dietrich, J. D.et al. Development of Orally Efficacious Allosteric Inhibitors of TNFα via Fragment-Based Drug Design. J. Med. Chem. 2021, 64 (1), 417-429.
UCB Biopharma Sprl; Sanofi. Fused Pentacyclic Imidazole Derivatives, Modulators of TNF-α Signalling. WO Patent WO2016050975A1. Priority Date: 3 October 2014; Published: 7 April 2016.
UCB Biopharma Sprl; Sanofi. Fused Hexacyclic Benzimidazole Derivatives and Analogues Thereof, Potent Modulators of Human TNF-α Activity. WO Patent WO2017/167995 A1, published 25 April 2017.
Sanofi; UCB Biopharma SPRL. Fused Pentacyclic Imidazole Derivatives as Modulators of TNF-α Activity. WO Patent WO2018/197503 A1, published 1 November 2018.
Tissot, M.; Jacq, J.; Pasau, P. Stereospecific Amination of Mesylated Cyclobutanol in Continuous Flow. Org. Process Res. Dev. 2019, 24 (4), 802-806.

Human–AI Synergy in Retrosynthetic Analysis and Route Optimization of Balinatunfib