Symposium Summary Fourth Annual International Umbilical Cord Blood Transplantation Symposium, Los Angeles, California, May 19-20, 2006

Biology of Blood and Marrow Transplantation 12:1206-1217 (2006) 2006 American Society for Blood and Marrow Transplantation 1083-8791/06/1211-0001$32.00/0 doi:10.1016/j.bbmt.2006.08.030 Symposium Summary Fourth Annual International Umbilical Cord Blood Transplantation Symposium, Los Angeles, California, May 19-20, 2006 John Wagner, 1 Richard Champlin, 2 Lawrence D. Petz 3 1 University of Minnesota, Minneapolis, Minnesota; 2 MD Anderson Cancer Center, Houston, Texas; 3 StemCyte International Cord Blood Center, Arcadia, California The 34-member faculty for the Symposium included leaders of the major transplantation centers from around the globe. Attendees were from Japan, New Zealand, Australia, Singapore, France, the UK, Israel, Mexico, Colombia, Ecuador, Chile, Brazil, Venezuela, Argentina, Malaysia, India, Sweden, Italy, Spain, Slovak Republic, Taiwan, Saudi Arabia, the UAE, Kuwait, Greece, Netherlands, Canada, China, Hong Kong, Korea, Thailand and from across the USA. The program was divided into 8 sessions: (1) state of the art in cord blood transplantation, (2) optimal selection of stem cell products for hematopoietic cell transplantation, (3) umbilical cord blood transplantation for patients with transfusiondependent thalassemia and sickle cell anemia, (4) quality issues in cord blood banking, (5) immune reconstitution following UCBT: graft-versus-leukemia, infectious complications, and adoptive immunotherapy, (6) transplantation of children, (7) development of a national cord blood center, and (8) cord blood cell biology, ex vivo expansion and regenerative medicine. The following comments emphasize significant aspects of selected presentations. Further details are provided in the abstracts that follow. The Fourth Annual International Cord Blood Transplantation Symposium was presented by the California Blood Bank Society and the Cord Blood Forum, and was partially supported by unrestricted educational grants from StemCyte and the National Marrow Donor Program (NMDP). The Symposium was also partially financially supported by a contract between the Health Resources and Services Administration (HRSA), US Department of Health and Human Services, and the California Blood Bank Society. Any opinions, findings, conclusions or recommendations expressed or presented at this conference or those of the presenters and do not necessarily reflect the views of the US Government. SESSION 1. STATE OF THE ART IN CORD BLOOD TRANSPLANTATION Session 1A. Review lecture: umbilical cord blood transplantation: the state of the art Dr. John Wagner reviewed the history and current state of knowledge with regard to umbilical cord blood transplantation (UCBT). Based on an analysis of factors impacting leukemia free survival, data reported to the CIBMTR and New York Blood Center would suggest that the optimal donor source is an HLA matched unrelated UCB unit with all other donor sources being virtually identical (ie, HLA matched marrow, HLA 1 antigen mismatched marrow, HLA 1 or 2 antigen mismatched UCB). This finding supports the use of HLA 1-2 antigen mismatched UCB particularly for the estimated 15,000 patients each year who cannot locate a suitably HLA matched unrelated adult volunteer donor. However, there remains great impetus to overcome the obstacles of marrow donor searches, namely prolonged search time (several months), lack of HLA matched donors particularly for ethnic and racial minority recipients, and high risk of transplant-related mortality (TRM). Certain properties of UCB, especially immune properties and rapid availability, make it interesting to explore. Cell dose, however, is clearly the limiting feature of UCB transplantation to date. Therefore, great emphasis has been made on overcoming this 1 major obstacle. Multiple strategies have been proposed to address the cell dose limitation: namely, ex vivo expansion culture, co-infusion of haploidentical peripheral blood, use of multiple UCB units, co-infusion of mesenchymal stem cells, intra-bone marrow injection and methods for enhancing the homing of UCB hematopoietic stem cells and progenitors to the marrow microenvironment. In addition, novel strategies for reducing regimen related mortality (RRM) as well as enhancing the graft-versus-leukemia leukemia effect of the UCB graft have been explored. In antic- 1206

Summary of IUCBT Symposium 1207 ipation of the various presentations over the conference that would be discussing the most recent generation of UCB studies, Dr. Wagner reviewed published outcomes with conventional myeloablation and single UCB transplantation. The results of the NHLBI sponsored COBLT trial as well as published results from the University of Minnesota (n 102, from 1994-2001) were used to highlight a standard by which more recent trials might be compared. While initial studies in the mid 1990s used 1 10 7 TNC/kg as the minimum acceptable cell dose, it has now been shown that a dose of 2.5 10 7 kg is the lower limit. In the University of Minnesota analysis, CD34 cell dose had a definite effect on neutrophil recovery and survival with poorer results in recipients of CD34 doses 1.7 10 5 CD34/kg. Because large intra-laboratory variation occurs between labs, the general use of CD34 cell enumeration as a criterion for selecting UCB units for transplantation is prevented. Multiple regression analyses further indicate that HLA match, CD34 cell dose, and grade II-IV acute GVHD are significant factors associated with survival. In the COBLT study, consisting primarily of pediatric patients, HLA match, and TNC and CD34 cell doses were not factors associated with overall survival. Factors that did influence survival were primary disease, recipient CMV status, recipient gender (male recipients did better), donor race (Caucasian donors were better), and performance status. Higher survival rates were observed in recipients of high resolution HLA matched UCB units, although the strength of this conclusion was limited by a small number of patients. A CIBMTR retrospective analysis demonstrated lowest risk of treatment related mortality in recipients of 6/6 HLA matched units whereas relapse was lowest in recipients of an HLA 2 antigen mismatched unit. These findings explain the lack of effect of HLA and cell dose on leukemia free survival (in that higher TRM is countered by reduced risk of relapse in patients transplanted with HLA 2 antigen mismatched marrow). Studies in adults, notably those of Rocha et al and Laughlin et al (NEJM, 2004) demonstrate similar survival between recipients of at least 1 antigen mismatched BMT and UCBT, but the results overall are poor regardless of donor. While increasing the CB inventory may increase the likelihood of finding a closely HLA matched UCB donor, the greatest focus needs to be on finding ways for increasing the functional cell dose. Together, these strategies should have a profound positive impact on results in adults. As outlined in the Program, there are numerous approaches that can be explored for enhancing the GVL/GVT of UCB. Strategies include: multi-unit UCBT, co-infusion of UCB derived NK cells, or co-infusion of genetically modified UCB derived T cells targeting leukemic populations. Also, data from the NYBC suggest that an effective approach for improving engraftment is the use of fludarabine in the preparative regimen. Results with ex vivo expansion culture, novel growth factors, multiple CB units, intra-bone marrow injection, and non-myeloablative regimens will be discussed during the Symposium. Other new developments in the field include consideration of novel applications including the use of effector cell therapy and expansion of the use of UCB to new diseases. Finally, major efforts to increase the U.S. UCB inventory are being implemented, as a result of the Stem Cell Therapeutic and Research Act of 2005. Session 1B. Reduced-intensity pre-transplantation conditioning regimens Dr. Vanderson Rocha discussed reduced-intensity conditioning for single unrelated cord blood transplants for adults with hematological malignancies. This was an Eurocord-Netcord group analysis of 65 patients with a median age of 47 years (16-76) and median weight of 60 kg (40-110 kg). Sixty-three percent of patients were CMV and 39% had a previous autologous transplant. This was a very high risk group with 49% having advanced disease. The conditioning regimens generally included fludarabine, with the most common regimen being fludarabine endoxan TBI (2 Gy). Twenty-six percent of patients received anti-tcell antibodies and 87% received hematopoietic growth factors (day 8); GVHD prophylaxis most commonly (55%) consisted of CsA and MMF. The median TNC collected was 3.7 10 7 /kg; the median TNC infused was 2.4 10 7 /kg; the median cell loss after thawing was 15%. Chimerism analysis was available in 71% of the patients at 3 months and was full donor in 67%, mixed chimerism in 9% and autologous reconstitution in 24%. Neutrophil recovery was 87 7% of the 26 patients who received the Flu End TBI conditioning regimen and was 65 10% for patients receiving other regimens. Grade II agvhd was found in 13%, grade III in 7% and grade IV in 7%; the TRM was 45 7% overall, 50 9% in acute leukemia, 30 15% in lymphomas and 27 16% for other diagnoses. The TRM at 1 year for those receiving 2.4 10 7 TNC/kg was 53 9% and that for those receiving 2.4 10 7 TNC/kg was 39 10%. For patients receiving Flu End TBI the TRM at 1 year was 24 10% and that for those receiving other conditioning regimens was 60 9% (P 0.001). DFS at 1 year for lymphomas was 50 9%, for leukemias was 27 7%, and for other diagnoses was 0. When the number of HLA mismatches was 0-1, DFS at 1 year was 42 12% (n 28), for 2 disparities DFS was 27 9% (n 34) and for 3-4 disparities DFS was 0 (n 9). DFS was 43 11% for those receiving Flu End TBI

1208 Summary of IUCBT Symposium (n 26), and was 16 7% for patients receiving other conditioning regimens (n 38) (P 0.005). For patients receiving 2.4 10 7 TNC/kg the DFS was 31 12% and for patients receiving 2.4 10 7 TNC/kg the DFS was 14 8% (P 0.05). The authors concluded that the results of single unrelated UCBT with reduced intensity conditioning regimens are encouraging and that cell dose and HLA remain important factors in this setting. The type of conditioning (Flu End TBI) seems to be associated with decreased TRM and better DFS, but a multivariate analysis with a higher number of patients is needed. Dr. Shigesaburo Miyakoshi presented several sets of data regarding reduced-intensity unrelated cord blood transplantation (RI-UCBT) for patients with hematologic malignancies. First he summarized results of a study of 97 adult patients who received a conditioning regimen of fludarabine, melphalan and TBI (4Gy). The overall survival (OS) for standard risk patients (n 22) was 67% (95% CI, 47-87) and for high risk patients (n 81) was 31% (95% CI, 20-43). A second study of ALL in adults involved 22 cases. The OS was about 60% and disease-free survival (DFS) about 48% with better survival noted for patients who were transplanted in complete remission. In a study of 34 cases of relapsed and refractory malignant lymphomas, the OS and DFS at 2 years was 24% and 16%, respectively. For chemo-sensitive patients the OS was 73%, whereas for chemo-refractory patients, it was 13%. Deaths from transplant related mortality (TRM) were higher than for relapse. A further study compared the outcomes of related PBSCT, UBMT, and UCBT with reduced intensity regimens. Neutrophil and platelet engraftment were slower with RI-UCBT than with RI-PBSCT, but the incidence of acute GVHD, TRM and OS were similar among these 3 stem cell sources with reduced intensity regimens. However, relapse rate might be higher than others over 100 days after RI-UCBT. The most important problem after RI-UCBT is high TRM so that transplant physicians should consider means to decrease TRM. Dr. Prakash Satwani discussed reduced-intensity allogeneic cord blood hematopoietic cell transplantation in pediatric patients with malignant and nonmalignant diseases. Previously published data indicate that RI AlloSCT results in reduced grade III/IV neutropenia and thrombocytopenia, decreased transfusion requirements, decreased bacterial infections and reduced grade II acute GVHD. A study of 22 patients for whom the RI conditioning was fludarabine-based (150-180 mg/m 2 ) with busulfan and rabbit ATG (n 17) or fludarabine cyclophosphamide etoposide ATG (n 5). Seventeen patients had malignant disease whereas 5 had non-malignant diseases. The median nucleated cell dose was 3.8 10 7 /kg. The probability of 3-year OS in all patients was 63% and for standard-risk malignancy patients was 90%. However, the 1-year OS in poor-risk malignancy patients was 21% and 2 year survival was 0. Thus, RI AlloCBHCT is a promising transplant option in children and adolescents, particularly in those patients with malignancies that are not far advanced, and should be offered to pediatric patients whose condition at the time of transplantation puts them at a high risk of toxicity from myeloablative transplantation. It appears to be contraindicated in malignant conditions with progressive disease and/or high tumor burden. Session 1C. Review lecture: double umbilical cord blood transplants (UCBT) with myeloablative or reduced-intensity conditioning regimens Dr. Scott Baker pointed out that the rate of hematopoietic recovery is cell dose dependent as is the incidence of engraftment, TRM and survival; also, the negative impact of HLA mismatch can be partially reversed by cell dose. In a study at the University of Minnesota, fifty patients received myeloablative double UCBT and the results were compared with 21 patients who received a single UCBT. In the double UCBT the units were HLA matched at least at the 4/6 level with the patient and with each other; the combined cell dose was 3.5 10 7 TNC/kg. There was no difference in cell dose between the 2 groups. Sustained donor engraftment was seen in 88% of doubles compared to 95% in singles (P 0.08) at a median of 24 days. Grade II-IV agvhd was higher in doubles (66%) compared to singles (38%) (P 0.02) but grade III-IV agvhd was similar (20% in doubles vs 14% in singles; P 0.51). TRM was similar in the two groups (doubles 22% vs singles 19%) as was DFS (59% doubles vs 56% singles) and OS (63% doubles vs 56% singles) singles. The conclusion reached was that the co-infusion of 2 partially HLA matched UCB units in the myeloablative setting offers the opportunity of safer and more successful UCBT for larger patients and provides a high incidence of engraftment, low TRM, and potentially a lower risk of relapse. In another study, 95 patients received a non-myeloablative double UCBT and the results were compared with 17 patients receiving a single UCBT. Results were comparable to 5 published series by other investigators. Sustained neutrophil engraftment was 85% with doubles vs 94% with singles (P 0.08); ATG was shown to improve engraftment for patients lacking recent chemotherapy, but was associated with increased risk of EBV viremia or PTLD. TRM was 15% in doubles vs 29% in singles (P 0.17); OS at 3 years was 45% for doubles and 40% for singles (P 0.30). Multivariate analysis indicated that independent predictors of inferior survival were lack of chemotherapy or prior auto-transplant within the previous 3 months, grade III-IV agvhd, and presence of co-morbidity (Karnofsky 50-60%, major organ dysfunction, or major prior life-

Summary of IUCBT Symposium 1209 threatening infection). The conclusions were that NMA UCBT is associated with a high probability of chimerism, a low incidence of TRM despite older age, a low risk of agvhd despite HLA mismatch and older age, and promising DFS and OS results. Session 1D. Double cord blood transplants Dr. Karen Ballen discussed double cord blood transplantation using a reduced intensity chemotherapy only conditioning regimen. Twenty-one adult patients with good risk hematologic malignancies were transplanted with 2 cord blood units with a total precryopreservation dose of 3.7 10 7 TNC/kg. The conditioning regimen consisted of fludarabine, melphalan, and rabbit ATG. ANC 500 was reached at a median of 20 days, and median days to platelet 20,000 was 41 days. Forty percent had grade II-IV agvhd but no patient had grade IV agvhd and there were no deaths from agvhd; 5 patients had cgvhd which was extensive in 2 and caused death in 1. The 100-day TRM was only 14%, OS at 1 year was 71%, and DFS at 1 year was 66%. Infection was a significant problem and was treated aggressively. New information was presented including the fact that parathyroid hormone post-transplant appears to improve survival in a mouse competitive transplant model. The conclusions reached were that double UCBT using a reduced intensity conditioning regimen provides an alternative stem source for patients without matched related or unrelated adult donors, and achieves engraftment in adults. Emphasis in continuing studies should be on improving immune reconstitution. Dr. Michael Verneris discussed the risk of relapse after UCBT in patients with acute leukemia with emphasis on a comparison of outcomes following single vs double unit grafts. A retrospective study was performed comparing the outcomes in 109 patients; 67 patients received a single and 42 patients received a double UCBT. The patients had AML or ALL and all received myeloablative conditioning. Two different conditioning regimes were used: regimen A consisted of cyclophosphamide, TBI, and ATG, followed by CSA/methylprednisolone immunosuppression (n 53), and regimen B consisted of cyclophosphamide, fludarabine, TBI, and CSA/MMF immunosuppression (n 43). Potential risk factors for relapse that were evaluated included: age, gender, recipient CMV sero-status, diagnosis, disease risk, HLA disparity, conditioning regimen, and TNC dose of the unit responsible for sustained engraftment. Outcomes reviewed were neutrophil recovery, TRM, acute and chronic GVHD, relapse and overall survival. In Cox regression analysis 2 factors were associated with lower relapse risk: disease risk (CR1/CR2 vs CR3 / REL; RR, 0.32; P 0.02) and transplantation of two UCB units (RR, 0.3; P 0.03). The diagnosis, UCB graft cell dose, and presence of agvhd had no demonstrable impact. Patients who received regimen B in CR1/2 showed a significantly lower risk of relapse with 2 UCB units compared to 1 unit (11% vs 54%, P 0.01). This is the first report suggesting that double unit UCB grafts may be associated with reduced relapse risk in acute leukemia. Dr. Michael Creer discussed a retrospective case control study which compared single cord transplants (SCT) vs dual cord transplants (DCT) in adults with acute leukemia undergoing myeloablative transplant conditioning. Twenty DCT cases were compared with 20 comparably matched SCT recipients with the largest TNC dose. Patients were matched for age, sex, diagnosis, disease status, ethnicity and body weight excluding patients with graft failure or previous transplants. The administered TNC dose for DCT (3.7 10 7 TNC/kg) was nearly 2-fold greater than SCT (1.9 10 7 TNC/kg). Overall survival (OS) at 12 months was 18% for SCT and 72% for DCT (P 0.0001) with median duration of follow-up of 18 months for both SCT and DCT. Although the TNC dose was approximately 2-fold greater in DCT recipients, the difference in OS for DCT recipients is greater than that expected on the basis of increased cell dose alone. In the first 100 days post-transplant, infection rates for SCT and DCT were similar; however, infection was the primary cause of death at 100 days for SCT (9/20), whereas for DCT there were no infection-related deaths at 100 days post-transplant. Conclusions reached were that OS of adult patients receiving DCT is significantly greater than that following SCT. There were no reported deaths from infection in DCT recipients in the first 100 days post-transplant although the frequency of infection was similar to that in recipients of SCT. The TNC dose was related to infection frequency and mortality from infection in SCT recipients, but infection frequency was not related to cell dose in DCT recipients. Accordingly, earlier recovery of neutrophil production and other potential benefits arising from infusion of cord blood units from 2 different donors may contribute to reduced early transplant-related mortality that is independent of increased TNC dose in DCT recipients. SESSION 2. OPTIMAL SELECTION OF STEM CELL PRODUCTS FOR HEMATOPOIETIC CELL TRANSPLANTATION Dr. Dennis Confer presented a talk entitled HLA matching for transplantation: lessons from bone marrow. Available HLA loci are HLA-A, B, C, DRB1; DRB3, 4 and 5; HLA-DQA and B; and HLA-DPA and B. One may type at different levels of resolution: low (broad antigens ), intermediate (split antigens ), and

1210 Summary of IUCBT Symposium high ( allele level or sequence level). Questions that arise about HLA typing for hematopoietic cell transplantation (HCT) include: How many loci should I type? How many loci should I match? What level of resolution should be used? What about DQ and DP? Are some loci more important than others? Answers for some of these questions are available for bone marrow transplantation, but much less information is available regarding cord blood transplantation. An analysis by the NMDP of overall survival (OS) following 3559 bone marrow transplants indicated that a mismatch at the antigen level for HLA-A, -B, -C and DRB1 was associated with a statistically significant increased risk for mortality as was high resolution mismatch for HLA-A. An analysis focusing an allele mismatching indicated that mismatches at A, B, C or DR reduced OS but mismatch at DQ or DP had no demonstrable effect. Conclusions about HLA typing for bone marrow transplantation were: 8 of 8 allelelevel matching at HLA-A, -B, -C and -DRB1 is preferred; when a single mismatch is present, antigen mismatches fare worse than allele mismatches; the role of HLA-DQ is uncertain; and the weighting of multiple mismatches is unknown. Dr. Juliet Barker discussed an analysis of 607 umbilical cord blood transplants facilitated by the New York Blood Center for the treatment of acute leukemia or myelodysplasia. With the recognition that cell dose (TNC) is a critical determinant of hematopoietic recovery, and with the use of new conditioning regimens and the use of double unit grafts, engraftment is no longer a major limitation to successful UCBT for many patients. However, transplant related mortality (TRM) remains a barrier to UCBT. Therefore, the aim of the present study was to analyze the impact of TNC dose and HLA match upon post-engraftment TRM (days 30-180) in engrafting patients. Results indicated that while low cell dose 2.5 10 7 TNC/kg was associated with day 30-180 TRM, increased cell dose above a dose of 2.5 10 7 TNC/kg was not beneficial. In contrast, HLA-mismatch was independently predictive of TRM at all levels of match. The group was then divided into those patients with grade 2-4 agvhd and those without GVHD. Analysis of each of these groups revealed that the adverse impact of HLA-mismatch upon TRM was most pronounced in the absence of agvhd (grade 0-1), with 63% of the deaths in this group being due to infection. Conclusions reached were that cell dose is a critical determinant of engraftment and TRM, but that in engrafting patients, cell dose 2.5 10 7 TNC/kg had no impact on day 30-180 TRM. In contrast, HLA-mismatch was an independent risk factor for post-engraftment TRM at all levels of match even in the absence of significant agvhd, possibly due to an adverse impact of mismatch on immune reconstitution. An important implication of these data is that improved UCBT outcomes will depend on the ability to obtain units of both sufficient size and match, and that this will require an increase in the global UCB inventory. Dr. Priya Kumar presented data regarding allogeneic hematopoietic stem cell transplantation in the treatment of acute lymphocytic leukemia (ALL). In a single center study, outcomes were compared among 4 different donor sources: HLA-matched related donor (MRD) (n 85), HLA-matched unrelated donor (URD:M) (n 15), HLA-mismatched unrelated donor (URD:MM) (n 14), and HLA-0-2 mismatched umbilical cord blood (UCB) (n 12). The patients underwent myeloablative conditioning (a cyclophosphamide 120 mg/kg and total body irradiation 1320-1375 cgy based regimen in 92%) followed by allogeneic HSCT. As a consequence of low transplant related mortality (TRM), overall survival (OS) and leukemia free survival (LFS) were superior after UCB transplants. In multiple regression analysis, 5 independent risk factors were significantly associated with poorer OS: use of URD:M and URD:MM, CR3 at HSCT, WBC 30 10 9 /L, cytomegalovirus (CMV) seropositive recipient and donor, and 2 induction regimens to achieve initial CR. GVHD was associated with improved LFS, suggesting a significant GVL effect across all donor sources. In conclusion, these results support the use of UCB as an acceptable donor source for adult ALL with results comparable to MRD, but findings must be confirmed using studies with larger numbers of patients. Dr. Eliane Gluckman presented the results of a study of the impact of the number and type of HLA incompatibilities and cell dose on outcomes of unrelated cord blood transplants for patients with malignant and non-malignant disorders. One thousand patients with hematological malignancies were analyzed. The results indicated that cell dose is the most important factor for engraftment and TRM. A minimum of 3 10 7 nucleated cells/kg at collection and 2 10 7 /kg at infusion must be obtained. HLA mismatches increase the risk of engraftment delays, TRM and cgvhd, but decreases the risk of relapse resulting in an absence of the role of HLA mismatches for survival. Increasing cell dose partially abrogates the influence of HLA mismatches. In 268 patients with non-malignant diseases, requirements for cell dose and HLA matching were different compared to malignant diseases. In nonmalignant diseases, cell dose is higher at 4.9 10 7 /kg at collection and 3.5 10 7 /kg at infusion. HLA mismatches play a major role for engraftment, survival and GVH. It is partially abrogated by increasing cell dose. Two HLA DRB1 mismatches resulted in an adverse outcome. Conclusions were that HLA mismatches play a major role for engraftment and survival in non-malig-

Summary of IUCBT Symposium 1211 nant diseases, while in malignant disease mismatches have a beneficial role for preventing leukemic relapse resulting in an absence of effect on LFS. HLA mismatches can be partially overcome by increasing cell dose, and matching for class II is better than for class I. Dr. Dennis Confer moderated a panel discussion (Drs. Barker, Gluckman, Wagner, Kurtzberg, and Stevens) with the goal of developing a consensus statement on the optimal selection criteria for cord blood stem cell products. The panelists agreed that when the cell dose is marginal, HLA match becomes more critical. With a HLA 5/6 match, a cell dose above 2.5 to 3.0 10 7 TNC is adequate above which there is no additional benefit. For a HLA 4/6 match, a minimal dose 2.5 10 7 TNC/kg may be adequate, although a preferable number might be as high as 4.0 10 7 TNC/kg, and there is a continuum of effectiveness with increasing cell doses. A somewhat lower cell dose may be adequate for 6/6 matches. When TNC and CD34 cell dose values seem discrepant, there is no way to resolve the problem and a suggestion was to not use such units. In regard to CD34 cell dose, the panelists agreed that CD34 cell dose probably correlates better with transplant outcome than TNC, but the reality is that the present status of CD34 testing is such that one cannot compare results from one laboratory to another. Thus, except in laboratories that develop their own data with 1 method for measurement, CD34 cell dose cannot be used to select units. SESSION 3. UMBILICAL CORD BLOOD TRANSPLANTATION FOR TRANSFUSION-DEPENDENT THALASSEMIA AND SICKLE CELL ANEMIA Review lecture. Dr. Mark Walters pointed out that more than 2500 hematopoietic cell transplants (HCT) have been done for -thalassemia major, primarily with HLA-identical sibling marrow donors. Recent data indicate that in Lucarelli class 1 and class 2 patients transplanted with marrow from sibling donors, the thalassemia-free survival rate is 87% and 84%, respectively. For class 3 patients 17 years of age who were treated with a conditioning regimen providing more intensive chemosuppression, the DFS was 85% and OS was 93%. Unrelated BMT in children with class 1 or 2 thalassemia have an OS of 92% and DFS of 82%, results that are similar to those obtained with sibling BMT. Also of note is that post- BMT phlebotomy has been shown to result in improvement in liver fibrosis and iron overload. In sickle cell disease (SCD), studies of survival of patients managed only with supportive care indicate a mean age of death in males of 42 years and in females of 48 years, a 30-year decrement compared to normal. Following BMT using sibling donors, the OS is 93% and the DFS is 85%. Importantly, patients who have been cured of SCD with BMT have no painful crises, no episodes of acute chest syndrome, no episodes of splenic or hepatic sequestration, no need for RBC transfusions and do not develop stroke, TIA or hemorrhagic CNS events. A sibling donor program established in the late 1990s has led to the transplantation of 28 patients with umbilical cord blood (UCB) units for SCD or thalassemia (a minority of the transplants was supplemented with bone marrow from the sibling). With about an 8-year follow-up, the OS is 89% and the DFS is 86%. Further, the GVHD rate is low and no patient has died of GVHD, whereas this is the leading cause of death after BMT. Conclusions were that HCT for SCD and thalassemia is curative in the majority of children who receive this treatment and that UCB transplantation using sibling-donors produces results similar to those of BMT. The risk/ benefit considerations favor HCT in the long term for SCD and thalassemia, and high-risk patients with sickle cell anemia may also benefit in the short-term. Dr. Tang-Her Jaing discussed unrelated umbilical cord blood transplantation (UCBT) for hemoglobinopathies. Between July 1991 and April 2006, 39 unrelated UCBTs were performed in 38 pediatric patients with hemoglobinopathies (36 transfusion-dependent thalassemia and 2 sickle cell disease) using 45 umbilical cord blood units (6 double cords and 1 re-transplant) after myeloablative therapy at the authors institutions and 11 other centers using mostly plasma depleted cord blood units (36/45) that were not washed after thawing (38/45). Eighteen patients were Pesaro class 1, 5 class 2, and 1 class 3, with 12 status unknown. The median age of patients was 6 years (range, 0.3-20 years) with a median weight of 19 kg (range, 8-76 kg). The cumulative incidence of unadjusted ANC 500 with donor chimerism and platelet 20,000 engraftment was achieved in 81 7% and 79 8% of the cases, respectively. Median times to ANC 500 engraftment, and platelet 20,000 engraftment were 17 days (range, 11-33 days) and 40 days (range, 16-135 days) after transplantation, respectively. Eight patients died with 2 early deaths prior to day 20 and 1 traumatic accident. All remaining 30 patients are alive (25 are thalassemia-free and 5 have autologous recovery) for a 2-year OS of 77 7% and DFS of 65 5%, with a mean and median follow up time of 392 and 257 days, respectively (range, 7-1760 days). An analysis was undertaken to compare experienced centers ( 5 cases) with less experienced centers, as well as to compare results with units that were washed post-thaw versus units that were not washed. The overall survival and disease free survival at experienced centers that did not perform post-thaw wash of the CB unit were 87 7% and 77 9%, respectively. At experienced centers that do not practice post-thaw

1212 Summary of IUCBT Symposium wash, outstanding results were obtained with plasma depleted CBU - ANC 500 93 7%, platelet 20,000 91 8%, and 2-year DFS 86 9%. When cell dose and other factors are optimal, unrelated CBT is a promising approach for curative therapy of transfusion-dependent thalassemia major. Dr. Shalini Shenoy presented data regarding reduced intensity conditioning in stem cell transplants performed for non-malignant disorders of childhood. The conditioning regimen consisted of campath-1h (days 21 to 19), fludarabine (days 8 to 4) and melphalan (day 3); GVHD prophylaxis consisted of CSP/tacrolimus, methyl prednisone, and a short course of MTX. Supportive care emphasized infection surveillance and prophylaxis, and disease specific care. Diagnoses included metabolic disorders (n 13), immune function disorders (n 25), bone marrow failure (n 7), hemoglobinopathy (n 6) and Evans syndrome (n 1). Twenty-six patients received an unrelated BMT or PBSCT, 11 an unrelated UCBT, and 15 a related donor BMT (HLA-matched). Graft rejection occurred in 5% of patients; myeloid engraftment was achieved at a median of 13 days (range, 10-36 days) and platelet engraftment ( 50,000) at a median of 26 days (range, 12-82 days). OS was 93% and 83% and event-free survival was 86% and 73% in related and unrelated transplants, respectively. Acute GVHD grade 2 occurred in 15% and cgvhd in 11%. Infectious complications occurred during the first 6 months. Of the 6 patients with hemoglobinopathies, 4 achieved 100% donor chimerism, none developed acute GVHD, and 1 patient developed cgvhd of the skin; Lansky scores were 90%-100%. Two patients with sickle cell disease who received lower doses of melphalan failed to achieve donor chimerism and had autologous recovery. They are alive with Lansky scores of 100%. Conclusions were that the conditioning regimen was well tolerated, successful engraftment occurred with BM, PB, and UCB, and that the rates of TRM and GVHD were acceptable. SESSION 4. QUALITY ISSUES IN CORD BLOOD BANKING Dr. Emer Clarke discussed standardization of CFU assays for cord blood products. A program was developed to assess variability in fresh CB samples. A standardized protocol was provided and participants measured TNC, viability and plated a specified number of viable cells; assays for BFU-E, CFU-GM, CFU- GEMM and total CFC were performed on day 14. Results on fresh CB indicated the following coefficients of variation (CVs): TNC 6.1%, viability 3.2%, and viable TNC 6.4%. Results for colony assays were more variable with CVs as follows: BFU-E 40.6%, CFU-GM 65% and total CFC 34.4%. On frozen CB samples CVs were: TNC 16.5%, viability 35.7% and viable TNC 40.9%. In 1 survey 17 of 41 sites reported no growth in their colony assays and the investigation of the lack of growth is presently under investigation. Conclusions were that cell number and viability assessment from fresh cord blood samples was excellent; variability in the recognition of colonies is increased with RBC background; and the variability in cell and viability assessment in frozen samples was significantly greater than in fresh samples. Dr. Cladd Stevens discussed HLA matching for cord blood transplantation with emphasis on an empirical analysis of the chance of finding an HLA matched CBU in the NY Blood Center cord blood bank. The analysis included 14,378 search requests for patients with high resolution DRB1 typing using a CB inventory of 25,917 CBUs with high resolution DRB1 typing. The ethnic distribution of patients was 4.4% Asian, 13.3% African-American, 15% Hispanic and 67.3% Caucasian. Among CBU donors, 6.6% were Asian, 16.4% were African-American, 17% were Hispanic, 50.4% were Caucasian and 9.6% were mixed. The chances of 1 matched CBU were as follows: Ethnic Group 2 MM 1 MM 0 MM Asian 98.9% 55.7% 6.9% African-American 98.8% 48.4% 2.4% Hispanic 98.9% 59.7% 8.3% Caucasian 99.5% 70.6% 13.1% All patients 99.3% 65.0% 10.6% The effect of inventory size on chance of finding 1 5/6 or 6/6 CBU match for Caucasians is as follows: Inventory Size Probability of Finding 5/6 or 6/6 Matched Unit 5000 36.0% 10,000 47.9% 15,000 55.3% 20,772 61.2% 25,917 65.2% Other data indicated that patients with a 6/6 match had a lower rate of relapse than those with HLA mismatches (in contrast to some data presented by others). Among patients with 0 HLA mismatches, TRM was not affected by increasing cell doses. When 2 or 3 HLA mismatches were present, the TRM was increased above the level found with 0 and 1 HLA mismatch. The interaction of HLA mismatch and TNC dose was important: compared to transplants with 1 mismatch and a high cell dose ( 5.0 10 7 TNC/kg), TRM was progressively increased when there were 2 mismatches but high cell dose, 1 mismatch with a low cell dose and 2 mismatches with a low cell dose. Dr. Robert Chow presented data on the outcome of cord blood transplantation (CBT) using 2 approaches to maximize cord blood (CB) cell dose: depletion of

Summary of IUCBT Symposium 1213 plasma (PD) but not of red blood cells during CB processing, and no post-thaw washing (NW). Outcomes of all CBTs using 237 PD CB units in 215 patients with audited engraftment and/or survival data from 2 cord blood banks up to 3/2006 were analyzed. Patient s median age was 9 years (0.3-59) with 33% 16 years; median weight 30 kg (4.5-112) with 35% 50kg; median number of HLA ABDR matches 4.0; median TNC dose 5.3 10 7 /kg; transplant center reported median post-thaw TNC dose of 4.4 10 7 / kg; median CD34 dose 1.8 10 5 /kg; malignant indications 70%; transplants outside US 39%; double transplant 27%; and non-myeloablative 16%. Unadjusted ANC 500, platelet 20,000 and 50,000 engraftment (all deaths prior to engraftment as graft failures) were 88 3%, 82 4%&76 4%. With a median follow-up of 325 days, 1-year overall survival (OS) and disease-free survival (DFS) were 59 4% and 54 4%, respectively. Stratification analysis showed worse engraftment and survival outcome at cord blood bank s CD34 cell dose below 0.7 10 5 /kg. There were 113 washed (W) and 95 non-washed (NW) PD CBTs. No significant adverse events occurred when the recommended DMSO threshold of 1g per kg recipient weight was not exceeded. TNC recovery after thawing as reported by transplant centers was higher for NW (median, 89% vs 75%). Unadjusted engraftment rates were higher and median time to engraftment was earlier for NW than W PD CB: 92 3% and 20 days for NW vs 88 3% and 24 days for W for ANC500 (with RR of W at 1, Cox RR of NW for engraftment 1.46, P 0.02), 89 5% and 44 days for NW vs 79 5% and 58 days for W for platelet 20K (Cox RR 1.79, P 0.002), 88 5% and 57 days for NW vs 74 5% and 75 days for W for platelet 50,000 (Cox RR 1.73, P 0.006). Relapse rates were 16 4% for NW and 28 5% for W (Cox RR 0.55, P 0.09), with TRM at 26 5% for NW and 33 5% for W (Cox RR 0.92, P 0.75). Oneyear OS was 63 5% vs 54 5% (Cox RR for death 0.87, P 0.54), and 1-year DFS was 62 5% vs 48 5% for NW and W (Cox RR for relapse or death 0.77, P 0.23), respectively. The authors concluded that outcomes using PD CBT compare favorably to published data of outcomes using RBC depleted CBT. There appears to be no clear benefit for post-thaw washing of PD CB, and results in this study indicated better results with NW CB units with respect to neutrophil and platelet engraftment rate and speed to engraftment; favorable trends were noted regarding TRM, relapse rate, 1-year OS and DFS. Dr. Michael Creer discussed a comparison of cord blood product thawing methods on cell recovery and progenitor integrity. A study was designed to investigate the effects of an albumin reconstitution approach (AR) (albumin dilution, no wash) to conventional washing (CW) following thawing of the frozen hematopoietic progenitor cells utilizing the following endpoints: cell recovery and viability, progenitor cell function, and transplant outcome. Results of the study indicated that following AR, the stability of the reconstituted HPC population is as good as that seen following CW, TNC recovery is significantly improved and the procedure is faster and much easier to perform. Also, AR may be superior to CW in terms of safety and viable CD34 cell recovery. AR and CW provide greater stability of viable CD34 cells and functional HPC than direct thaw (infusion after thawing without AR or CW) where significant loss of cell viability and function is seen after 4 hours and may occur earlier as a result of exposure of the product to 10% DMSO. AR has no significant adverse impact on transplant outcome and may offer a significant benefit over CW particularly in single unit transplants in adults where cell dose is limited and cell recovery post-thaw is critical. Recommendations made included the use of AR in all settings were the postprocessing TNC dose is 3.5 10 7 /kg. CW should be considered if the TNC dose is 3.5 10 7 /kg and there is concern about DMSO toxicity (DMSO dose 1 g/kg) or major ABO incompatibility ( 30 ml incompatible RBC stroma). SESSION 5. IMMUNE RECONSTITUTION FOLLOWING UCBT: GRAFT-VERSUS-LEUKEMIA, INFECTIOUS COMPLICATIONS, AND ADOPTIVE IMMUNOTHERAPY Review lecture. Dr. Mary Eapen reviewed the topic of graft-versus-leukemia (GVL) effect after allogeneic transplantation for acute leukemia. She indicated that allografts initiate immune responses which are strongest when major HLA antigens of the donor and recipient are incompatible, and are weaker when involving minor HLA antigens. A study by the IBMTR of 2,000 recipients of HLA-matched sibling donor transplants for ALL, AML, or CML (CR1 or CP1) indicated that there was a lower relapse rate after non-t-cell depleted BMT in patients with agvhd, cgvhd or agvhd and cgvhd. Further, the relapse rates correlated with GVHD severity. A GVL effect may also occur in the absence of GVHD following BMT as suggested by the fact that there is a lower relapse rate after twin transplants when patients receive 3 10 8 cells/kg. A GVL effect is also found when there is a donor/ recipient sex mismatch, as indicated by a lower risk of relapse in male recipients of female donors even if they do not develop GVHD. In unrelated donor transplantation, GVHD rates are higher, relapse rates are lower but TRM is higher. Consequently, the influence of GVHD on overall survival is the balance of higher TRM and lower relapse rates.

1214 Summary of IUCBT Symposium Dr. Maria S. Albano discussed CMV infections in cord blood transplantation and screening strategies for cord blood (CB) donors. She indicated that donor CMV-antibody status is relevant for BMT donor selection, but the strategy for screening CMV among CB donors is not clear. A positive test for antibody to CMV in an adult BM donor indicates active or latent infection, but a positive antibody test in a newborn may detect maternal antibody and is not necessarily evidence of infection in the infant donor. A number of studies carried out at the New York Blood Center National Cord Blood Program indicated the following: Regarding CB recipients for HSCT, the incidence of post-transplant CMV infection was higher in patients with pre-transplant CMV-antibody positive test. CMV infection after CB transplantation was a significant cause of post-transplant morbidity and mortality and CB donor-maternal CMV-antibody positive test (total and IgM) was not associated with CMV infection in the recipients. Regarding CB donors: the incidence of CMV infection among CB donors (saliva culture and CB-PCR) was low (0.12%). Maternal serology (total antibody and IgM) had a low positive predictive value for diagnosis of CMV infection in infant donors. Only 1/3 of infants with confirmed CMV infection had a CMV-IgM positive mother. This means that 2/3 of infants with confirmed infection would be missed by IgM screening. Detection of CMV-DNA in CB by PCR is a good alternative to viral culture for screening among CB donors. Dr. Jeffrey Miller discussed NK cells and KIR mismatch in cord blood transplantation. He reviewed the results of a study of the significance of KIR ligand (KIR-L) mismatches in 243 patients with hematologic malignancies who were transplanted with CB units. There was no difference in grades II-IV or III-IV agvhd rates based on KIR-L mismatch, there was a trend to less relapse if KIR-L mismatch was present (but this was not statistically significant), and there was a slight increase in TRM if KIR-L mismatch was present. In double cord transplants, the investigators hypothesized that NK cell alloreactivity may predict the winning unit; their data suggested that NK cell alloreactivity of unit 2 against the recipient predicts engraftment of unit 1. (Variables that did not predict the winner were infusion order, mononuclear cell dose, CD34 dose, CD3 dose and CFU-GM dose.) In a study of adoptive transfer of haploidentical NK cells, 19 patients with AML were treated with allogeneic NK cells which were infused on transplant day 0 and then expanded in vivo with IL-2 3 /week 6 doses. A KIR-L mismatched donor correlated with achieving AML CR (P 0.04). Conclusions were that NK cells are important in transplantation; KIR reconstitution after UCB transplant may predict the winner in double CBT and provide a clinically important functional advantage for NK cell alloreactivity in advanced leukemias. Also, UCB may provide a useful source of cells for adoptive NK cell transfer. Dr. Prakash Satwani discussed the optimal ex vivo expansion of cord blood natural killer (NK) cells, NK cytotoxicity and NK receptor expression and the potential for CB adoptive cellular immunotherapy. He pointed out that UCBT is limited by the absence of available donor effector cells (NK, CTL, LAK and NKT) for DLI for the treatment of hematological relapse and/or PTLD. Remissions have been produced by DLI in patients with hematologic malignancies in relapse. Also, donor-versus-recipient NK cell alloreactivity reduced the risk of leukemia relapse in a study of 57 AML patients at high risk of relapse, while improving engraftment and protecting against GVHD. Studies in the authors laboratory indicated that ex vivo expansion of CB with IL-2, IL-7, IL-12 and anti-cd3 (AB/CY) for 7-10 days resulted in a significant increase in NK cells expressing C-lectin receptor CD94 /NKG2D ; an increase in NK cells (dim and bright) NK KAR KIR2DS4; a decrease in NK cell NCR NKp44 dim subset; and a decrease in NK cell NCR NKp46 dim subset. In summary, there is an increase in NK cytotoxicity, an increase in LAK cytotoxicity, an increase in granzyme B expression, and a decrease in perforin expression. Dr. Laurence Cooper discussed the differentiation of naive cord blood T cells into CD19-specific cytolytic effectors for post-transplant adoptive immunotherapy. One can produce chimeric immunoreceptors which couple recognition domains with TCR-signaling domains to redirect specificity of T cells to cell-surface antigens independent of MHC. The investigators developed chimeric antigen receptors (CARs) to redirect the specificity of T cells for CD19, a B-lineage antigen, and expressed these transgenes in UCB-derived T cells. In vitro experiments using a chromium release assay of primary B-ALL cells demonstrated the effectiveness of the CD19-specific T cells and in vivo experiments indicated their value for adoptive immunotherapy. A second generation immunoreceptor was developed termed CD19RCD28 to provide T-cell costimulation for CD19 CD80 CD86 targets; this provided improved T-cell persistence and efficacy. A major objective of the research is to assess the safety and feasibility of cellular immunotherapy following UCB transplant by administering ex vivo expanded donor-derived T cells (genetically modified to express a CD19-specific chimeric immunoreceptor) into research participants with CD19 ALL after non-myeloablative or ablative conditioning regimens. SESSION 6. TRANSPLANTATION OF CHILDREN Review lecture. Dr. Joanne Kurtzberg discussed progress in UCBT for inborn errors of metabolism.

Summary of IUCBT Symposium 1215 Particular emphasis was placed on transplantation of patients with the severe phenotype of Hurler syndrome (PMS I) and Krabbe disease, or globoid-cell leukodystrophy. In Hurler patients, UCBT resulted in durable engraftment, normalization of alpha-l-iduronidase levels and improvement in cognitive function with an overall survival of 85%. Krabbe disease is an autosomal recessive disorder due to deficiency of the lysosomal enzyme galactocerebrosidase, characterized by failure of the process of myelination in the central and peripheral nervous systems, rapidly progressive neurologic deterioration, and death. In the infantile form, symptoms appear before 6 months of age and include irritability, dysphagia, progressive spasticity, mental deterioration, blindness, deafness, seizures, and death, usually before 2 years of age. Forty-two patients have been transplanted, 17 of which were neonatal transplants. Results indicated that surviving patients showed durable engraftment of donor-derived hematopoietic cells with restoration of normal blood galactocerebrosidase levels. Infants who underwent transplantation before the development of symptoms showed progressive central myelination and continued gains in developmental skills, and most had age-appropriate cognitive function and receptive language skills, but a few had mild-to-moderate delays in expressive language and mild-to-severe delays in gross motor function. Children who underwent transplantation after the onset of symptoms had minimal neurologic improvement. Conclusions were that UCB transplantation is uniquely suited for pediatric patients with genetic diseases, inborn errors of metabolism, bone marrow failure syndromes, hemoglobinopathies, and immunodeficiency diseases. Dr. Mary Eapen discussed Unrelated donor hematopoietic stem cell transplantation in children with acute leukemia: risks and benefits of umbilical cord blood vs bone marrow. A study compared the outcome of 492 unrelated BM and 508 CB transplants in patients younger than 16 years of age having ALL or AML. Median follow-up of survivors was 59 months for BM and 45 months for CB transplants. Recipients of matched CB transplants had significantly lower transplant related mortality (TRM), treatment failure and overall mortality than recipients of matched BM; recipients of high cell dose CB mismatched at 1 locus had risks similar to recipients of matched BM. Recipients of low cell dose CB mismatched at 1 locus and CB mismatched at 2 loci (any cell dose) had higher risks of TRM, treatment failure and mortality than matched BM recipients during the early post transplant period. However, among those surviving the early period, subsequent risks were similar to matched BM recipients. The data support use of CB matched or mismatched at 1 locus with a high cell dose for children needing HSCT whether or not a matched BM donor is available; low cell dose or 2-loci mismatched CB grafts may provide a reasonable alternative when a matched BM donor is not available or for those whose disease requires immediate transplantation. Dr. Vanderson Rocha discussed outcomes after unrelated cord blood transplantation in children with acute lymphoblastic leukemia; An Eurocord-Netcord survey. Few data are available on the outcomes after UCBT in children with ALL. An analysis of results was performed on 361 UCB transplants in children with ALL receiving an UCBT from 01/1995 to 01/2005 in 99 transplant centres in 24 countries, mostly in Europe. The children received a single non-expanded CB unit and a myeloablative conditioning regimen, and the majority (67%) received CsA corticoids as GVHD prophylaxis. The median age was 6.5 years, median cell dose infused was 4.1 10 7 /kg and the median follow-up time was 22 months (3-96). The cord blood was HLA identical (6/6) in 12% of the cases, 5/6 in 46%, 4/6 in 39% and 3/6 in 3%. Overall 2 year-lfs was 36 3%. In a multivariate analysis, only CR1 or CR2 were associated with better LFS (HR 1.8; P 0.0001). A separate analysis was performed according to the disease status. Results indicated that status of the disease at transplant was a major risk factor for outcomes. The most important risk factor associated with relapse incidence and LFS after UCBT for children in CR2 was the duration of CR1 (first relapse on or off therapy). Conclusions were that UCBT should be proposed as alternative source of allogeneic transplantation for children with high risk ALL in an earlier status of the disease for those children lacking an HLA identical donor. SESSION 7. DEVELOPMENT OF A NATIONAL CORD BLOOD PROGRAM Robert Baitty, representing the Health Resources and Services Administration (HRSA), discussed implementation of the Stem Cell Therapeutic and Research Act of 2005. The aims of the Act are to increase the number of unrelated donor transplants, to increase the public inventory of high quality CBUs from diverse populations, and to increase the number of CBUs available for research. A target of 150,000 new units was set. Cord blood banks will be funded for 3 years through 1-time 10-year contracts awarded competitively; in addition, there will be annual competitions for new cohorts of banks. A national cord blood coordinating center (CBCC) will also be developed and its functions will include the facilitation of transplants with cord blood, coordination of contingency planning with banks, coordination of inter-bank proficiency testing programs and compilation of data on CBUs released for research. An Advisory Council will be established which may make recommendations in