Background
Phase III, double-blind, placebo-controlled RCT. 372 pediatric patients (≤17 years) undergoing allo-HCT at high risk for sinusoidal obstruction syndrome/veno-occlusive disease (SOS/VOD). Defibrotide is a polydisperse single-stranded DNA with endothelial-protective, thrombolytic, and anti-inflammatory properties — FDA-approved for treatment of SOS/VOD. HARMONY tested whether prophylactic defibrotide prevents SOS/VOD in high-risk pediatric HCT recipients.
Interventions and follow up
Arm A: Defibrotide 25mg/kg/day IV in 4 divided doses, beginning the day before conditioning and continuing for 30 days post-HCT
Arm B: Matching placebo IV × same 30-day schedule
Primary endpoint: Incidence of SOS/VOD by day +30
Median follow up: 12 months
Arm B: Matching placebo IV × same 30-day schedule
Primary endpoint: Incidence of SOS/VOD by day +30
Median follow up: 12 months
Results
SOS/VOD by day +30 (primary): 12.0% (defibrotide) vs 19.7% (placebo) — OR 0.56, 95% CI 0.28–1.12, P=.10 — NOT significant
Severe SOS/VOD: 4.3% vs 9.5% — trend only, P=.09
Day-100 OS: 83.9% vs 78.5% — NS
All-cause mortality: 18.7% vs 22.0% — NS
Severe SOS/VOD: 4.3% vs 9.5% — trend only, P=.09
Day-100 OS: 83.9% vs 78.5% — NS
All-cause mortality: 18.7% vs 22.0% — NS
Adverse events
Hemorrhagic (grade ≥3): 24.0% (defibrotide) vs 19.1% (placebo) — numerically higher (NS); procedural hemorrhage 12.3% vs 9.1% (NS)
Infusion reactions (grade ≥3): 1.6% vs 0.5%
Discontinuation due to AEs: 19.0% vs 8.1%
Infusion reactions (grade ≥3): 1.6% vs 0.5%
Discontinuation due to AEs: 19.0% vs 8.1%
Conclusions
HARMONY was a NEGATIVE trial — defibrotide prophylaxis did not significantly reduce SOS/VOD incidence or mortality in high-risk pediatric allo-HCT vs placebo. A numerical trend toward lower SOS/VOD did not reach significance. Defibrotide prophylaxis is not supported by this evidence and should not be routine practice.
Key Limitations
Likely underpowered for the observed effect size — placebo-arm SOS/VOD (19.7%) was lower than the assumed 30%, reducing study power. Definition heterogeneity (Baltimore vs Seattle vs 2017 EBMT criteria) affects ascertainment. Subgroup signals in the highest-risk patients (busulfan-based conditioning) are hypothesis-generating only. The hemorrhagic safety signal warrants caution.
Clinical Context
HARMONY is the definitive negative trial against routine defibrotide prophylaxis in pediatric allo-HCT. Defibrotide (Defitelio) remains FDA and EMA approved for TREATMENT of SOS/VOD. The prophylaxis indication is not supported. EBMT guidelines recommend against routine prophylaxis but acknowledge possible benefit in very high-risk patients, left to institutional protocols.