Baricitinib Shows Durable Responses in Refractory cGVHD

The Janus kinase (JAK) inhibitor baricitinib produced high, durable response rates without dose-limiting cytopenias in patients with severe, treatment-refractory chronic graft-vs-host disease (cGVHD), according to results of a study published in Blood Advances.¹

In the single-center phase 1/2 trial (NCT02759731), the 6-month overall response rate (ORR), the primary efficacy end point, was 76.2% (95% CI, 52.8%-91.8%) among evaluable patients, with all responses being partial responses (PRs); in the intent-to-treat analysis, ORR was 67% (95% CI, 51%-80%). No dose-limiting toxicities (DLTs) were observed at either dose level.

In an accompanying editorial, Julie R. Boiko, PhD, and Paul A. Carpenter, MD, of Fred Hutchinson Cancer Center, noted the failure of other treatments to replace steroids as first-line therapy, and proposed that if baricitinib’s efficacy is confirmed in larger phase 2 studies, the favorable safety profile “may place baricitinib as a strong contender for cGVHD prophylaxis or first-line therapy studies.”²

Trial Design and Patient Population

The trial used a single-arm, intrapatient dose-escalation design and enrolled 24 adults with moderate-to-severe cGVHD refractory to at least 2 prior lines of therapy, including corticosteroids. Patients received baricitinib at 2 mg daily for the first 3 cycles, escalating to 4 mg daily if no DLT occurred.¹

Eligible patients had severe cGVHD by National Institutes of Health global severity score at enrollment, with a median of 5 involved organs (range, 2-7) and sclerotic skin involvement in 92% of patients. Patients had a median of 5 prior lines of therapy (range, 2-11), including prior ruxolitinib (Jakafi) in 3 patients (12.5%), and 83% were on concurrent systemic immunosuppressive therapy (IST) at enrollment. After 4 weeks of baricitinib treatment, concurrent IST could be gradually tapered as clinically feasible.

Efficacy Findings

Among 13 evaluable patients at 12 months, responses were observed in 12 (93%), and the best ORR at any time was 95.6% (95% CI, 79%-97%), with a median time to response of 1.4 months. Response rates varied by organ, ranging from 91% in the joints/fascia and 100% in the lower gastrointestinal tract to 4% in the skin; 18 organ-specific complete responses occurred across 11 patients (46%). Among patients with sclerotic skin involvement, average sclerotic body surface area decreased by 5% at 6 months and 13.5% at 12 months, including 1 complete skin response. Seven of 13 patients (54%) on concurrent steroids were able to reduce the dose of steroids, with a median decrease of 5 mg (range, 1-16) of prednisone daily.

At a median follow-up of 35.8 months, median failure-free survival (FFS) was 19 months (95% CI, 11-26), with 1- and 2-year FFS rates of 71% and 37%, respectively. Overall survival was 100% at 1 year and 94.1% (95% CI, 65.0%-99.2%) at 3 years, a single death related to pneumonia occurred 20 months after completing study treatment, and no deaths or malignancy relapses occurred during active treatment. However, among 13 patients who completed the protocol-mandated 12 cycles and discontinued per protocol, 8 (62%) required initiation of new systemic therapy for progressive cGVHD at a median of 7.6 months after stopping baricitinib.

Safety Findings

The most common grade 2 or higher adverse events (AEs) were viral upper respiratory infection in 42%, hypophosphatemia in 21%, and lung infection in 21%. Serious AEs occurred in 8 patients (33%), 4 of which were considered possibly treatment related. Hematologic parameters remained stable throughout treatment, and dose reductions were required in 3 patients (13%), primarily because of neutropenia. Treatment was discontinued early in 11 patients (46%), most commonly because of progressive cGVHD or prolonged dose interruption.

Patient-Reported Outcomes and Immune Correlates

Clinically meaningful improvement on the Lee Symptom Scale of at least a 6-point reduction was observed in 53% of evaluable patients at 6 months and 73% at 12 months. Functional Assessment of Cancer Therapy–Bone Marrow Transplantation scores also showed significant improvement in total score and emotional well-being at 6 months. Baricitinib significantly reduced cytokine-induced STAT phosphorylation in CD4+ and CD8+ T cells and lowered plasma levels of proinflammatory cytokines including BAFF, IL-12/IL-23p40, MIG, and IP-10. A baseline absolute B-cell count below 0.82 × 10⁹ cells/L was identified as a potential predictor of response, with 80% of nonresponders and 75% of responders classified in the 21 evaluable patients by multivariable analysis.

Context Among JAK Inhibitors in cGVHD

Boiko and Carpenter situated the findings within a broader effort to identify better-tolerated JAK inhibitors for cGVHD. They noted that ruxolitinib remains the only category 1 second-line therapy in NCCN guidelines, a status earned through the REACH3 trial (NCT03112603),³ whereas 2 other JAK inhibitors have been investigated in cGVHD with mixed results: the phase 2/3 GRAVITAS-309 trial (NCT03584516)⁴ of itacitinib was suspended during dose-finding after itacitinib plus corticosteroids showed concerning rates of relapsed malignancy and all-cause mortality compared with corticosteroid monotherapy, and pacritinib (Vonjo) is under investigation (NCT05531786).⁵

They suggested that baricitinib’s pharmacologic profile distinguishes it from ruxolitinib: it was developed from the outset for once-daily oral dosing and provides “approximately equipotent, moderate JAK1/2 inhibition, with less JAK3 and TYK2 inhibition,” and its established safety profile in humans “excludes significant cytopenias,” unlike ruxolitinib’s more potent JAK2 inhibition, which suppresses hematopoiesis.² They noted the study authors selected baricitinib based on preclinical GVHD murine models showing survival superiority over ruxolitinib, itacitinib, and pacritinib.

Boiko and Carpenter also raised questions the trial leaves open, including whether ruxolitinib-intolerant or nonresponder patients would tolerate or respond to baricitinib or other JAK inhibitors, noting the cohort included relatively few ruxolitinib-exposed patients. They pointed to rheumatoid arthritis literature suggesting inadequate response to one JAK inhibitor may be overcome by switching to another.

Study Limitations

The study authors acknowledged that the single-center design, small sample size, and sequential dose-level assignment limit direct comparison between the 2 dose levels.¹ The cohort was heavily pretreated, with 92% of patients having sclerotic skin manifestations and some patients refractory to up to 11 prior lines of therapy; nonresponse was associated with a higher number of prior therapies. The study was not powered to assess efficacy specifically in patients with prior ruxolitinib exposure and only enrolled 3 patients who had received ruxolitinib. The authors identified this as a key question to address in future studies given the frequent use of ruxolitinib as a second-line cGVHD agent.

REFERENCES

1. Holtzman NG, Im A, Curtis L, et al. Baricitinib for refractory chronic graft-versus-host disease: results of a phase 1/2 study. Blood Adv. 2026;10(17):5751-5761. doi:10.1182/bloodadvances.2025018898

2. Boiko JR, Carpenter PA. Shuffling the JAKs: playing a new card in chronic GVHD. Blood Adv. 2026;10(17):5834-5835. doi:10.1182/bloodadvances.2026021099

3. Zeiser R, Russo D, Ram R, et al. Ruxolitinib in Patients With Corticosteroid-Refractory or Corticosteroid-Dependent Chronic Graft-Versus-Host Disease: 3-Year Final Analysis of the Phase III REACH3 Study. J Clin Oncol. 2025 Aug 10;43(23):2566-2571. doi: 10.1200/JCO-24-02477.

4. GRAVITAS-309: Itacitinib and Corticosteroids as Initial Treatment for Chronic Graft-Versus-Host Disease. ClinicalTrials.gov. Updated October 20, 2025. Accessed August 31, 2026. https://clinicaltrials.gov/study/NCT03584516

5. Phase I/​II Study of Pacritinib, A JAK2/​IRAK1/​CSF1R Inhibitor, in Refractory Chronic Graft-Versus-Host Disease (cGVHD) After Allogeneic Hematopoietic Stem Cell Transplantation (HSCT). ClinicalTrials.gov. Updated July 9, 2026. Accessed August 31, 2026. https://clinicaltrials.gov/study/NCT05531786