Defibrotide (DF) in the Treatment of Severe Hepatic Veno-Occlusive Disease (VOD) with Multi-Organ Failure (MOF) Following Stem Cell Transplantation (SCT): Results of a Phase 3 Study Utilizing a Historical Control.

Phase 2 clinical trials of DF in the treatment of severe VOD/MOF have demonstrated a complete response (CR) in 36-46% of patients (pts) with encouraging overall survival and tolerability (Richardson Blood 2002; Richardson ASH 2006). Given the life-threatening nature of VOD/MOF, a trial randomizing pts to placebo or best supportive care was considered but rejected. A phase 3 trial, comparing DF in the treatment of VOD/MOF post-SCT to a contemporaneous historical control (HC) was therefore performed. The HC was created using a sequential review of medical charts starting 6 months prior to use of DF at each center.

Eligible pts met Baltimore VOD criteria by D+21 (total bilirubin ≥ 2.0 mg/dL with ≥ 2 of the following: hepatomegaly, ascites or 5% weight gain) and either renal and/or pulmonary failure by D+28. Exclusion criteria: severe GvHD involving liver or gut; clinically significant bleeding; or need for >1 pressors to maintain BP. As this is a non-randomized study, the primary efficacy analysis compared CR by D+100, adjusted by quintiles of propensity score based on 4 stratification variables, at an overall two-sided 0.01 significance level (Koch et al,1989). CR was defined as bilirubin < 2 mg/dL + resolution of MOF; stratification variables were allogeneic/autologous SCT, adult/pediatric, 1 or 2+ SCTs, and ventilator/dialysis dependence. A secondary endpoint was mortality at D+100. DF was given at 6.25 mg/kg IV q6h; treatment duration was recommended for at least 21d. To create the HC, 35 centers sequentially reviewed up to 266 cases. To determine HC eligibility, the Medical Review Committee (MRC, composed of 2 independent expert hematologists) assessed all pts who met VOD criteria with MOF. The MRC were provided data for each pt (a redacted medical chart or pt narrative, depending on the privacy laws for each center) only up to the date on which the pt met inclusion criteria. The MRC remained blinded to outcome data at all times. One interim analysis was planned.

For the HC, 6821 medical charts were screened, identifying 123 pts with features consistent with VOD in a setting of renal and/or pulmonary dysfunction that were reviewed for eligibility by the MRC. The MRC selected 32 cases as having an unequivocal diagnosis of VOD whose MOF was secondary to VOD, who met all protocol entry criteria; for all eligible pts, a confounding diagnosis of GvHD was ruled out. In the DF-treated group, 102 pts were enrolled. Following the interim analysis (comparing 61 DF pts to 32 HC pts), the DMC recommended an increase in HC sample size to 51 pts; given the large number of medical charts already reviewed, this was not considered feasible. In the final analysis (comparing 102 DF pts to 32 HC pts), the 2 groups were balanced regarding stratification variables. Baseline demographics (DF vs HC pts): median age 21 vs 18 yrs (43% and 44% pediatric); 63% vs 53% male, 88% vs 84% allogeneic SCT; 13% vs 3% with prior SCT; and 38% vs 38% ventilator/dialysis dependent. Median time post-SCT to VOD diagnosis was 13 and 11 days. Acute leukemia was the underlying diagnosis in 44% and 47%. Median duration of DF therapy was 22 days (range 1-60 days), with a median daily dose of 19 mg/kg/d. For the primary efficacy analysis, D+100 CR rate equaled 24% vs 9% (99%CI difference in CR rate: -1–35%; 95%CI difference in CR rate: 3–30%); p=0.015. D+100 mortality rate equaled 62% vs 75% (95%CI difference in rate -32–3%); p=0.051 by stratified log-rank. Consistent with prior studies, DF in children resulted in higher CR compared with HC (CR 36% vs 7%; p=0.04). Use of DF was associated with improved outcome in less sick pts (D+100 CR for pts without ventilator dependence equaled 40% vs 9%; p=0.051 and for pts without dialysis dependence equaled 34% vs 9 %; p=0.027). For pts receiving autologous SCT (n=12 and 5 pts in DF and HC arms), DF was associated with markedly improved CR (75% vs 0%, p=0.005). Hemorrhagic adverse events (any grade) were similar between the two groups (65% vs 69%); 18% of DF pts experienced a drug-related toxicity that led to discontinuation. D+100 CR strongly correlated with D+100 survival in both DF and HC groups (p<0.0001, p=0.0016).

DF improves CR by D+100 in pts with severe VOD/MOF post-SCT with a p value < 0.05. In addition, there was a trend towards improved D+100 survival in this critically ill population. DF-associated toxicities are consistent with prior studies, supporting the observation that DF is generally well tolerated.

Richardson:Gentium: Membership on an entity's Board of Directors or advisory committees. Arai:Gentium: Research Funding. Grupp:Gentium: Research Funding. Martin:Gentium: Research Funding. Corbacioglu:Gentium: Consultancy, Research Funding. Holler:Gentium: Consultancy. D'Agostino:Gentium: Membership on an entity's Board of Directors or advisory committees. Massaro:Gentium: Consultancy. Hannah:Gentium: Consultancy. Iacobelli:Gentium: Employment. Soiffer:Gentium: Consultancy, Membership on an entity's Board of Directors or advisory committees.