Background
Phase 2, open-label, dose-randomized study (ROCKstar, KD025-213). 132 patients with steroid-refractory/steroid-dependent chronic GVHD (SR-cGVHD) after allo-HCT who had received ≥2 prior lines of systemic therapy. Belumosudil (KD025) selectively inhibits ROCK2 (Rho-associated coiled-coil–containing protein kinase 2), regulating STAT3/STAT5 balance in T and B cells — a mechanism distinct from calcineurin, JAK, or HDAC inhibitors.
Interventions and follow up
Arm A: belumosudil 200mg PO once daily (QD)
Arm B: belumosudil 200mg PO twice daily (BID)
Primary endpoint: Overall response rate per 2014 NIH cGVHD Consensus criteria
Median follow up: 14.7 months
Arm B: belumosudil 200mg PO twice daily (BID)
Primary endpoint: Overall response rate per 2014 NIH cGVHD Consensus criteria
Median follow up: 14.7 months
Results
ORR (200mg QD): 74% (95% CI 58–87%)
ORR (200mg BID): 77% (95% CI 61–89%)
CR/PR (200mg QD): CR 6%, PR 68%
Median duration of response: not reached (62% maintained response at 1 year)
Failure-free survival at 1 year (QD): 55%
Prior ibrutinib subgroup ORR: 73% — active in post-ibrutinib population
ORR (200mg BID): 77% (95% CI 61–89%)
CR/PR (200mg QD): CR 6%, PR 68%
Median duration of response: not reached (62% maintained response at 1 year)
Failure-free survival at 1 year (QD): 55%
Prior ibrutinib subgroup ORR: 73% — active in post-ibrutinib population
Adverse events
Overall/infections: grade ≥3 AEs 57% (QD) vs 67% (BID); grade ≥3 infections 36% (QD) vs 45% (BID)
Constitutional/GI: fatigue 28% vs 44%; diarrhea 23% vs 31%; nausea 23% vs 25%
Metabolic: phosphate increase 7% vs 15%; no cumulative toxicity with QD, BID not more effective but more toxic
Constitutional/GI: fatigue 28% vs 44%; diarrhea 23% vs 31%; nausea 23% vs 25%
Metabolic: phosphate increase 7% vs 15%; no cumulative toxicity with QD, BID not more effective but more toxic
Conclusions
Belumosudil 200mg QD demonstrated a 74% ORR in heavily pretreated SR-cGVHD, including post-ibrutinib patients, with durable responses and a manageable safety profile, supporting FDA approval as a novel ROCK2-selective inhibitor for SR-cGVHD.
Key Limitations
Single-arm dose-comparison design rather than randomized vs active comparator; no head-to-head data vs ibrutinib, ruxolitinib, or axatilimab. CR rate is low (6%) — most responses are PRs, so continued therapy is needed. The 14.7-month median follow-up limits OS and durability assessment. Selection for ≥2 prior lines enriches for patients who tolerated prior toxicity. Optimal sequencing with axatilimab (AGAVE-201) is undefined.
Clinical Context
FDA approved belumosudil (Rezurock) in August 2021 for SR-cGVHD in adults and pediatric patients ≥12 years after ≥2 prior lines. Its unique ROCK2 mechanism makes it a rational option after calcineurin inhibitor, steroid, ibrutinib, or ruxolitinib failure. The landscape now includes axatilimab (AGAVE-201, NEJM 2024) with similarly high ORR; sequencing between belumosudil and axatilimab is guided by individual toxicity profiles and organ involvement per ASTCT practice.