AJ1-11095, a potent and highly selective type-II JAK2 inhibitor, shows enhanced therapeutic efficacy as compared with type-I JAK2 inhibitor ruxolitinib in models of myeloproliferative neoplasms (MPNs)

Abstract Therapy with JAK2 inhibitors (JAK2i) is a standard of care for myelofibrosis (MF). All FDA-approved JAK2i are type-I inhibitors, and while these agents provide symptomatic relief and reduce splenomegaly, they all fail to effectively eliminate the mutant MPN clone, and many patients discontinue type-I JAK2i treatment within 3-5 years of starting therapy. Type-I JAK2i bind the active conformation of JAK2, which can result in disease persistence due to continued JAK2 activation through heterodimerization with JAK1 or TYK2. Recent studies have shown that type-II JAK2i, which bind the inactive JAK2 kinase conformation, can overcome type-I JAK2i persistenceand enhance therapeutic efficacy in MPN animal models. However, until now, the development of type-II JAK2i has been limited by off-target toxicity and poor bioavailability. We investigated the pre-clinical activity of AJ1-11095, a potent, orally administered, highly selective type-II JAK2i which is the first type-II JAK2i to be tested in a multi-institutional Phase 1 clinical trial for MF patients refractory to type-I JAK2i (NCT06343805). AJ1-11095 was identified using a multi-parameter in silico design strategy using physics-based computationally guided structure-based drug design (SBDD) tools and medicinal chemistry ideation. A co-crystal structure of AJ1-11095 bound to the JAK2 kinase domain revealed that AJ1-11095 binds in the ATP binding pocket with a DFG-out configuration of the kinase activation loop, consistent with a type-II inhibitor binding mode. Kinome selectivity profiling revealed high JAK2 selectivity over other JAK family kinases with minimal off-target inhibition. Potent down-regulation of JAK/STAT signaling was observed in ruxolitinib (RUX)-naïve human SET2 cell lines (IC50: 45nM), and AJ1-11095 effectively restored JAK/STAT signaling inhibition in RUX-persistent SET2 cells at an IC50 similar to naïve cells consistent with the ability to overcome type-I JAK2i mediated persistence. We first tested the in vivo efficacy of AJ1-11095 using a Cre-inducible human JAK2VF transgenic competitive transplant model. After disease was established, mice were treated with AJ1-11095 (60mg/kg QD), RUX (60mg/kg BID), or vehicle (VEH) for 56 days. Dose-dependent improvements in hematocrit (VEH 58%, RUX 50%, AJ1-11095 35%, p0.0005), platelets, and spleen volumes (mg: VEH 223, RUX 98, AJ1-11095 57, p0.0005) were observed with AJ1-11095, and to a greater degree than mice treated with RUX. Flow analysis showed potent reductions in mutant allele fraction throughout peripheral blood, spleen and bone marrow (BM) in AJ1-11095-treated mice, including down to lin-cKit+Sca1+ (LSK; VEH 51.5%, RUX 45.6%, AJ1-11095 8.4%, p0.005) BM stem cell fractions not observed with either VEH or RUX. We also observed significant mutant allele fraction reductions in mice initially treated with RUX for 28 days followed by 28 days of treatment with AJ1-11095 (LSK 21.9%, p0.05) suggesting AJ1-11095 can reduce mutant allele fraction even in mice pre-exposed to type-I JAK2i. We next investigated AJ1-11095 in the hMPLW515L adoptive BM transfer model of MF. Similar to the JAK2VF model, we not only saw improvements in hematologic parameters, but also observed further reductions in BM fibrosis, spleen weights (mg: VEH 430, RUX 393, AJ1-11095 112, p0.005), and restoration of splenic architecture compared to that of RUX alone. In line with the potent in vivo efficacy of AJ1-11095, we also observed significant reductions in serum pro-fibrotic cytokine levels, including TNFa, IL13, and IP-10, to a degree not seen with RUX treatment, suggesting AJ1-11095 more effectively inhibits pro-inflammatory cytokine production compared to type-I JAK2i. Notably, in both models, no significant weight loss was observed suggesting AJ1-11095 can enhance therapeutic effect without added toxicity.In sum, AJ1-11095, which is currently being tested in a Phase 1 clinical trial in patients with MF refractory to type-I JAK2i, is a potent, orally administered, highly selective type-II JAK2i that demonstrates superior pre-clinical efficacy compared to RUX, including reductions in mutant allele burden and reticulin fibrosis, with an enhanced safety profile compared to previous type-II JAK2i. These data further demonstrate the preclinical utility and clinical rationale for type-II JAK2i, including AJ1-11095, for the treatment of MPNs.