Business

Yatiri Bio options global rights to Oscotec's multikinase inhibitor denfivontinib for AML

Yatiri Bio, a San Diego-based precision medicine company, has entered an exclusive global option to in-license agreement with South Korea's Oscotec for denfivontinib (SKI-G-801), a multikinase inhibitor, for the treatment of acute myeloid leukemia. Under the terms of the Oscotec in-license agreement, Yatiri Bio holds the exclusive option to license denfivontinib for development and commercialization worldwide. The deal was announced on March 24, 2026. Oscotec originated the compound and retains underlying ownership until the option is exercised.

Financial terms of the agreement were not disclosed. The option-to-license structure implies staged commitments, with an option exercise fee likely triggered at a future decision point tied to clinical or biomarker data. No upfront payment, milestone values, royalty rates, or equity components were made public. This is consistent with deal announcements from private, pre-revenue companies. The deal is indication-specific to AML, suggesting Oscotec retains rights to denfivontinib in other indications.

Deal context

Denfivontinib is a small molecule multikinase inhibitor targeting FLT3 and the CLK family of kinases. FLT3 is a receptor tyrosine kinase frequently mutated in AML and a validated therapeutic target. CLK kinases regulate RNA splicing, and their inhibition represents a less established but distinct mechanism in AML. The dual FLT3/CLK profile is what Yatiri Bio frames as the compound's differentiating feature. Oscotec completed a Phase I trial in relapsed/refractory AML (NCT03564288), enrolling 14 patients between 2018 and 2021. A separate Phase I trial in advanced solid tumors (NCT05971862) was conducted at Yonsei University in Seoul. No clinical data from either trial have been published in peer-reviewed journals. No US FDA regulatory designations have been reported for the compound.

Yatiri Bio plans to apply its proprietary ProteoCharts platform to identify a biomarker profile for denfivontinib AML treatment that extends beyond standard FLT3 mutation status. The platform uses deep neural networks to integrate proteomic profiling with clinical data and functional testing in patient-derived models. The company states it has identified a responder population that is FLT3-independent, potentially broadening the eligible patient pool beyond mutation-targeted therapies. Yatiri Bio intends to develop a laboratory determined test at its CLIA/CAP-accredited facility to stratify patients for upcoming trials. The company has described this as a capital-efficient approach to precision medicine oncology, designed to enrich trial enrollment with patients most likely to respond.

The AllSci BriefSystematic R&D and deal news. Daily.

For Oscotec, the deal provides a development path for an asset that completed Phase I but had not advanced further. For Yatiri Bio, denfivontinib appears to be the first publicly disclosed application of ProteoCharts in a licensing context. The company has stated plans to expand the platform across hematologic and solid tumors in 2026, including MDS, ovarian cancer, prostate cancer, breast cancer, and colorectal cancer.

The AML treatment landscape for FLT3-targeted therapies includes several approved agents and established acute myeloid leukemia biomarkers based on genomic mutation testing:

  • Midostaurin (Rydapt) — Novartis — approved for FLT3-mutant AML
  • Gilteritinib (Xospata) — Astellas — approved for relapsed/refractory FLT3-mutant AML
  • Quizartinib (Vanflyta) — Daiichi Sankyo — approved 2023 for newly diagnosed FLT3-ITD-positive AML

All three rely on DNA-based companion diagnostics for patient selection. Yatiri Bio's thesis is that proteomic stratification can identify responding populations that genomic testing misses. Acrivon Therapeutics operates the closest competing platform, AP3, which uses phosphoproteomics for drug-specific biomarker signatures and is further along in clinical development with its own pipeline assets.


Spot something wrong? Report an issue with this article