Physicians AbstractsEBMT 2010

Physicians Abstracts EBMT 2010 Van Bekkum Award Presidential Symposium 70 Defibrotide prevents hepatic VOD and reduces significantly VOD-associated complications in children at high risk: final results of a prospective phase II/III multicentre study S. Corbacioglu, S. Cesaro, M. Faraci, D. Valteau-Couanet, B. Gruhn, J.J. Boelens, C. Uderzo, N. Hewitt, A. Rovelli, J. Schrum, I. Müller, A. Schulz, J. Stein, R. Wynn, J. Greil, S. Matthes, K. Sykora, M. Albert, A. O’Meare, J. Massaro, R. D’Agostino, M. Pfitzenmaier, P. Sedlacek, P. Schlegel, J. Arvidson, J. Toporski, A. Groll, A. Fasth, J. Winiarski, H. Ozsahin, A. Schrauder, P. Bader, M. Hoyle, M. Iacobelli, G. Dini, C. Peters on behalf of the Pediatric Disease Working Party O84 Basic Science Award CD40-activated B-cells migrate towards secondary lymphoid organs and interact dynamically with T- cells N. Klein-González (1), S. Balkow (2), E. Kondo (1), T. Liebig (1), A. Shimabukuro-Vornhagen (1), S. Grabbe (2), W. Bloch (3), M. von Bergwelt-Baildon (1) (1)Stem Cell Transplantation Program Cologne (Cologne, DE); (2)Johannes Gutenberg University (Mainz, DE); (3)German Sport University Cologne (Cologne, DE) B cells have been demonstrated to present antigen to T cells in vivo. CD40-activation dramatically improves antigen presentation by normal and malignant B cells and has therefore been studied as an approach to generate autologous “non-artificial” antigen presenting cells for active immunotherapy. Furthermore, CD40-B cells have recently been shown to expand tumorantigen and viral specific CTL as well as regulatory T cells and are therefore of great interest for post-transplant immunotherapy. Human B cells when activated via CD40-L/IL-4 can be expanded from small amounts of peripheral blood in 1214 days. CD40-activated B cells can prime naïve CD4 + and CD8 + T cells, expand memory T cells and express important surface homing molecules. Nevertheless, it remains unclear whether CD40-activated B cells migrate to secondary lymphoid organs (SLO) in vivo to attract and interact with T cells. To address this question we established a methodology to generate murine CD40-activated B cells. At day 14 of culture, these cells are >95% CD19 + and CD80/86/MHCI/MHCIIhi. Murine CD40-activated B cells present a ‘homing phenotype’; migrate toward SLO chemokines such as CCL19, CCL21 and CXCL13; and induce T-cell chemotaxis in vitro. Upon CD40L activation, B cells up-regulate CCR7 while down-regulating CXCR5 expression which suggests direction of activated B cells toward the B-zone/T-zone boundary. We compared the homing of GFP + CD40-activated B cells to resting GFP + B cells and show for the first time that CD40-activated B cells home to SLO significantly more efficiently than resting B cells. Furthermore, CD40activated B cells localize in the B-cell areas, and a significant fraction move to the B-T boundary close to the T-cell zone over time. To dissect T-cell-APC interactions on a single cell we analyzed three-dimensional migration in collagen matrix using time-lapse videomicroscopy. Interestingly, antigen-loaded CD40-activated B cells differ from immature and mature DC by displaying a rapid migratory pattern undergoing highly dynamic, short-lived (7.5 min) and sequential interactions with cognate T cells. Taken together, these data revealed that CD40-activated B cells can home to secondary lymphoid organs and interact dynamically with T cells thus underlining their potential as cellular adjuvant for cancer immunotherapy. Background: Hepatic VOD in children following SCT is common and often associated with substantial morbidity and early death. DF has been effective in small trials to prevent VOD. Methods: Pts < 18 years with the following eligibility criteria and myeloablative SCT were enrolled: conditioning with BU/Mel, liver disease, 2nd myeloablative SCT, allo-SCT for leukemia in 2nd relapse, MAS, abdominal irradiation, gemtuzumab, osteopetrosis, and ALD. Pts were randomized to receive DF from conditioning until D + 30 post SCT or to the control arm. Pts diagnosed with VOD received DF for treatment. Primary objective: incidence of VOD by D + 30 according to the modified Seattle criteria. An independent review committee, blinded to the treatment, adjudicated the diagnosis of VOD. Secondary endpoint: morbidity and mortality assessed by a composite score incorporating VOD-associated MOF and mortality. Incidence and severity of GvHD was assessed. Results: The final sample size of 360 pts was recruited between 2006 and 2009 by 28 centers. The Intent-to-Treat (ITT) analysis was performed on all randomized patients (356 pts). Mean age: 6.6 years; 24% infants, 52% children and 23% adolescents; 41% female, 59% male. Transplants were 69% allogeneic, 31% autologous. Both groups were balanced for disease types and risk factors. In the ITT analysis, 12% of the pts in the DF arm and 20% of the control group developed VOD by D + 30 (Competing Risk (CICR) P = 0.049; Kaplan Meier (KM) P = 0.051). In the PP analysis, the VOD incidence was 11% vs. 20% (CICR P = 0.022; KM P = 0.023). In the control group VOD was experienced by 27% of the infants, 17% of the children and 20% of the adolescents. The composite score for morbidity and mortality was significantly in favor of the DF arm (P = 0.034); renal failure was observed in 1% of DF pts vs. 6% of the control (P = 0.017). Excluding auto SCT pts, the incidence (45% vs. 63%; P = 0.004) and severity by grade of aGvHD by D + 100 (P = 0.003) was significantly reduced in the DF pts. VOD, independent of severity, led to a 4x higher mortality compared to patients without VOD (24.6% vs. 6%). SAEs were experienced by 61% of the DF pts vs. 59% of the control. Conclusions: This Phase II/III study clearly demonstrates the efficacy and safety of DF in preventing VOD in pediatric pts at high risk, reducing the incidence by 40%. Both renal failure and aGvHD were significantly reduced in the DF arm. DF can be recommended for the prevention of VOD in this high risk population. S1 mortality (NRM) and OS. We measured by sequential ELISA, levels of these five proteins in plasma collected prospectively from 485 allogeneic HCT patients randomly divided into training (149 GVHD-, 175 GVHD + ) and validation (74 GVHD-, 87 GVHD + ) sets. We obtained samples 3-14 days before onset of GVHD (median day 29) and at equivalent time points in patients without GVHD. There were no significant differences between sets in age, conditioning intensity, donor source, HLA match or GVHD grade between training and validation sets. The median day of sample acquisition was day 20 and day 21, respectively. Logistic regression determined that a linear combination of the five proteins levels predict the occurrence of acute GVHD in the training set. The Receiver Operating Characteristic curves of each of the five individual biomarkers are shown in Figure 1 with an area under the curve (AUC) for the composite panel of 0.77 (95%CI: 0.72-0.82). When t (...truncated)