Background: Acute graft-versus-host disease (aGVHD) remains a major barrier to successful allogeneic hematopoietic stem cell transplantation (allo-HSCT). First-line corticosteroids yield suboptimal response rates of 30-60% in grade II-IV disease, with an increased risk of infection. Steroid-refractory aGVHD, particularly gastrointestinal involvement, has a poor prognosis with 50% 6-month survival. Effective immunomodulatory regimens enhancing regulatory T cell function without excessive immunosuppression are therefore warranted.

Belumosudil, a selective ROCK2 inhibitor approved for steroid-refractory chronic GVHD (cGVHD) based on the pivotal ROCKstar trial (ORR=74-77%), demonstrates dual immunomodulatory mechanisms: blockade of STAT3 phosphorylation to suppress IL-17/IL-21 signaling and enhancement of STAT5-mediated regulatory T cell (Treg) differentiation, restoring immune homeostasis. Preclinical evidence further establishes that ROCK1/2 inhibition attenuates aGVHD pathology by downregulating dendritic cell costimulatory molecules (CD80/CD86) and suppressing F-actin-dependent antigen-presenting cell trafficking, while preserving graft-versus-leukemia (GVL) effects. These mechanistic insights support the clinical investigation of belumosudil in aGVHD. We present real-world outcomes of belumosudil-treated aGVHD patients in this study.

Methods: We conducted a retrospective analysis of aGVHD patients treated with belumosudil following allo-HSCT in Union Hospital, Tongji Medical College, Huazhong University of Science and Technology between January 1, 2024 and March 31, 2025. Baseline characteristics, previous aGVHD treatment, and clinical outcomes and safety outcomes during belumosudil treatment, donor and recipient demographics, underlying hematological malignancies, prior and concomitant treatments, and clinical outcomes were collected. The clinical outcomes included overall response rates at day 28 (primary endpoint) and day 56, time to response (TTR; defined as time from belumosudil initiation to initial clinical response), failure-free survival (EFS) and safety.

Results: A total of 8 aGVHD patients (median age of 41.5 years; 75% male) were enrolled, with 87.5% haploidentical donors. All patients received myeloablative conditioning, with 75% of patients administered the fludarabine plus modified TBI regimen incorporating craniospinal, marrow, and nodal targets. Standard GVHD prophylaxis was cyclosporine/MMF/methotrexate + anti-CD25/ATG/ALG. Acute GVHD occurred at a median of 30 days post-HSCT, with grade III-IV disease detected in 63% (5/8) of cases. All patients manifested gastrointestinal involvement, 75% (6/8) had concurrent skin disease, and 13% (1/8) displayed liver involvement. Prior therapies included steroids, JAK inhibitors, and biologics (≥5 patients each), except for one treatment-naïve patient.

The median duration of belumosudil treatment was 90.5 days. Concomitant treatments included MMF (85.7%), cyclosporine (71.4%), and JAK inhibitors (57.1%). Within a median follow-up of 101 days, the overall response rate at day 28 (primary endpoint) was 100% (8/8), with 75% cases showing a complete response (6/8); median time to response was 13.5 days (95% CI not estimable). Responses were sustained through day 56 in all surviving patients (5/5 cases with a complete response). Median EFS was not reached due to insufficient protocol-defined events (n=2/8), comprising one hematologic relapse and one transplant-associated thrombotic microangiopathy (TMA). Three deaths occurred (37.5%), all adjudicated as disease-related complications: relapse (n=1), TMA (n=1), and pulmonary infection (n=1). No grade ≥3 treatment-related adverse events or dose modifications were observed.

Conclusion: This study provides the first disclosed real-world evidence of belumosudil in acute GVHD, demonstrating unprecedented 100% ORR (75% cases with a complete response) with rapid median time to response of 13.5 days and well tolerated. Despite the inherent limitations of this pilot investigation, these findings warrant validation in pivotal multicenter trials.

This content is only available as a PDF.
2025
Sign in via your Institution