Shuffling the JAKs: playing a new card in chronic GVHD Authors: Julie R Boiko, Paul A Carpenter Published: 2026 In this issue of Blood Advances, Holtzman et al1 report the results of a phase 1/2 study of the JAK inhibitor (JAKi) baricitinib for the treatment of refractory chronic graft-versus-host disease (cGVHD). Acknowledging the pilot study context, they demonstrate that baricitinib is safe and effective in patients who had failed prior cGVHD-directed systemic immunosuppressive therapy (IST). Moderate and severe cGVHD affects up to 30% of allogeneic hematopoietic cell transplantation survivors, in which the donor-derived immune system becomes dysregulated, leading to inflammatory and, subsequently, fibrotic manifestations in one or more host tissues, including the eyes, mouth, skin, lungs, joints/fascia, gastrointestinal tract, genitalia, and liver.2,3 Although cGVHD response to first-line IST (usually corticosteroids [CS]) varies widely, a US claims analysis determined that 71% of first-line therapy recipients progress to second-line therapy.4 Advances in the knowledge of causal immune pathways have led to 4 US Food and Drug Administration (FDA) approvals over the past decade for cGVHD after failure of at least 1 line of IST, including ibrutinib (a Bruton tyrosine kinase inhibitor), ruxolitinib (JAK1/2i), belumosudil (a ROCK2 inhibitor), and axatilimab (a colony-stimulating factor-1 receptor [CSF-1R] inhibitor). These therapeutics are standardly incorporated alongside CS, but patients with cGVHD may require years of sequential CS-sparing agents during their individual IST journeys. JAKs are a family of cytoplasmic tyrosine kinases associated with intracellular tails of numerous cytokine and other receptors. JAK activation recruits STAT proteins, which migrate to the nucleus and regulate expression of various immunomodulatory genes.5,6 At least 14 JAKis have been approved for clinical use globally for rheumatologic/autoimmune disorders, myeloproliferative neoplasms, and dermatologic conditions. Currently, ruxolitinib (a selective JAK1/2i that inhibits downstream STAT1/3 signaling and tempers alloreactive T-cell activity) is the only category 1 second-line cGVHD therapy in the current National Comprehensive Cancer Network guidelines, a status justifiably earned based on the results of the relatively rigorous REACH3 study.7 To date, 3 other JAKis have been, or currently are being, investigated as cGVHD therapeutics, including pacritinib (a selective JAK2i with additional FLT3, IRAK1, CSF-1R, and ACVR1/ALK2 targets; ClinicalTrials.gov identifier: NCT05531786), itacitinib (a nonclinically approved JAK1i), and another JAK1/2i baricitinib, which is evaluated in the study by Holtzman et al. The phase 2/3 GRAVITAS-309 trial of itacitinib for the initial treatment of cGVHD was suspended during the dose-finding portion of the study after 103 patients were enrolled because CS + twice-daily itacitinib dosing led to concerning rates of relapsed malignancy and all-cause mortality (mainly infections) compared with the CS-monotherapy control group.8 Baricitinib was developed from the outset as an oral therapeutic. Unlike ruxolitinib, its pharmacokinetics were specifically tailored for once-daily dosing. In addition, its approximately equipotent, moderate JAK1/2 inhibition, with less JAK3 and TYK2 inhibition, was considered potentially suitable for autoimmune diseases. It is now also FDA-approved for adults hospitalized with COVID-19-associated pneumonia. Notably, its established safety profile in humans excludes significant cytopenias. Conversely, the more potent JAK inhibition of ruxolitinib, specifically the JAK2i component, more effectively suppresses the hematopoietic abnormalities found in myeloproliferative disorders, consistent with its original clinical indication, although this effect is typically undesirable in cGVHD therapy. Unsurprisingly, Holtzman et al selected baricitinib for their study based on more optimal JAK1/2 inhibition, along with preclinical GVHD murine models showing baricitinib survival superiority compared with treatment with ruxolitinib, itacitinib, or pacritinib. The study by Holtzman et al addresses 2 important barriers associated with next-line cGVHD therapies. First, despite solid overall response rates (ORRs) to second-line or later ISTs, responses are often incomplete, and some patients must switch IST because of nondurable responses, nonresponses, or dose-limiting toxicity. More effective agents are thus needed. Second, the de facto standard second-line IST, ruxolitinib, is associated with infections (particularly viral), and up to 18% of patients with GVHD discontinue ruxolitinib because of cytopenias.9 Therefore, it is first noteworthy that the Holtzman et al study included 24 heavily pretreated adults, all with longstanding severe cGVHD, including 92% of patients with sclerotic skin manifestations. The ORR at the 6-month primary endpoint was 76.2% (95% confidence interval, 52.8-91.8), which was also encouragingly durable with a median failure-free survival of 19 months. Second, very low rates of dose-limiting anemia, thrombocytopenia, or infections were observed, which, if confirmed in larger controlled studies, would be highly beneficial in treating cGVHD. Likely reflecting the expected lengthy total course of IST required to achieve cGVHD quiescence, many responders required resumption of IST after the mandated end to baricitinib study therapy at 12 months after enrollment. Validated patient-reported outcomes also corroborated GVHD clinical responses. Laboratory immune studies suggested a shift toward regulatory T cells in responders with plasma cytokine shifts, hinting at a dampening of a B-cell–monocyte/macrophage axis. Conclusions from the results of the 5 REACH trials conducted in adults and children have led to ruxolitinib becoming a staple second-line cGVHD therapy. In this context, the study by Holtzman et al provokes several questions. First, would ruxolitinib-intolerant or nonresponder patients potentially tolerate and/or respond to other JAKis? Although ruxolitinib naïveté was not an inclusion criterion, the cohort in the study by Holtzman et al included relatively few ruxolitinib-exposed patients. Subsequent baricitinib (or other JAKi) cGVHD trials would need to consider including JAKi-exposed and -unexposed patients to address this pragmatic detail. Perhaps relevant, rheumatoid arthritis literature suggests that inadequate responses to 1 JAKi may be overcome by switching to an alternative JAKi.10 Next, as with other contemporary cGVHD therapies, the question is where these agents should sit in the treatment sequence for cGVHD. Currently, no contemporary cGVHD drug is approved for first-line therapy, with CS unfortunately remaining the most common first agent. If baricitinib’s efficacy is confirmed in larger-scale phase 2 studies, the favorable safety profile observed in the pilot study by Holtzman et al may place baricitinib as a strong contender for cGVHD prophylaxis or first-line therapy studies. This work is a reminder that shuffling JAKis by thoughtfully exploiting their differential JAK isoform selectivity, additional kinase targets, and myelosuppressive profiles may help determine which JAKi to investigate next as we collectively try to advance the cGVHD therapeutic portfolio through the next generation of clinical trials. Conflict-of-interest disclosure: The authors declare no competing financial interests. References