Defibrotide (DF) for the Treatment of Severe Veno-Occlusive Disease (VOD) and Multi-System Organ Failure (MOF) Post SCT: Final Results of a Phase II, Multicenter, Randomized Study and Preliminary Analyses of Surrogate Markers and Ultrasound Findings.

Introduction: DF, a polydisperse oligonucleotide, has anti-thrombotic, anti-ischemic and thrombolytic properties, especially on microvasculature. Studies have suggested that DF modulates endothelial injury in VOD.

Methods: A phase II randomized study of two doses, 25 mg/kg/d [arm A] or 40 mg/kg/d [arm B], was carried out in pts with severe VOD. Correlation with markers of vascular injury was undertaken. Endpoints included CR rate, toxicity and mortality at d+100 post SCT. VOD diagnosis was by Baltimore criteria and severity by ≥30% risk on the Bearman model or MOF. Abdominal ultrasound (US) was required prior to enrollment, repeated during therapy and at completion of treatment.Pts with ≥ grade II GVHD were excluded. Treatment arms were stratified for age (<18y) and cyclophosphamide-based conditioning. Treatment was planned for ≥14d. Pts receiving ≤3d of therapy were inevaluable for response. CR was defined as a bilirubin <2 mg/dl and resolution of VOD-related MOF.

Results: 102 pts have been treated; 51 pts on arm A and 51 pts on arm B: 87 pts underwent allo- and 15 auto-SCT. Median age was 33y (6 mos - 63y). At DF initiation, median bilirubin was 7.1 mg/dl; median wt gain 12%; ascites in 78%; RUQ pain in 65%; hepatomegaly in 71%; abnormal portal flow in 42%, and MOF in 97%. Median duration of therapy was 20d. There were no unexpected side effects and no treatment related deaths. In a subset of pts (n=14), central review of US was performed. Portal flow improved by wk 3 in pts with CR (n=8). Improvement was not seen in nonresponders (NR) (n=6), suggesting changes in portal flow may help predict outcome. Median values of markers of endothelial stress with DF treatment are summarized below. In pts with CR, median PAI-1 levels decreased and Protein C increased. Median NO declined regardless of response. Median thrombomodulin, tissue factor and TFPI increased in NR. CR was achieved in 50/93 pts (54%; 95% CI (43%,64%)), with survival to d+100 in 43/91 pts (47%; 95% CI (37%,58%)). No difference in outcome between the 2 dose arms was seen. An analysis using 38 controls matched for VOD severity confirmed a survival advantage (p=0.0004).

Conclusion: US findings and trends in endothelial stress markers may help predict successful DF treatment. The CR rate (54%) and d+100 survival (47%) in this first prospective, randomized trial of DF in pts with severe VOD and MOF confirm the encouraging results of prior studies, and a case control analysis shows a highly significant survival benefit with DF treatment.