Search

Myeloablative conditioning associated with hazardous immediate and late complications is considered as a mandatory first step in preparation for allogeneic blood or marrow transplantation (allogeneic BMT) for the treatment of malignant hematologic disorders and genetic diseases. Immune-mediated graft-versus-leukemia (GVL) effects constitute the major benefit of allogeneic BMT. Therefore, we have introduced the use of relatively nonmyeloablative conditioning before allogeneic BMT aiming for establishing host-versus-graft tolerance for engraftment of donor immunohematopoietic cells for induction of GVL effects to displace residual malignant or genetically abnormal host cells. Our preliminary data in 26 patients with standard indications for allogeneic BMT, including acute leukemia (n = 10); chronic leukemia (n = 8), non-Hodgkin's lymphoma (n = 2), myelodysplastic syndrome (n = 1), multiple myeloma (n = 1), and genetic diseases (n = 4) suggest that nonmyeloablative conditioning including fludarabine, anti–T-lymphocyte globulin, and low-dose busulfan (8 mg/kg) is extremely well tolerated, with no severe procedure-related toxicity. Granulocyte colony-stimulating factor mobilized blood stem cell transplantation with standard dose of cyclosporin A as the sole anti-graft-versus-host disease (GVHD) prophylaxis resulted in stable partial (n = 9) or complete (n = 17) chimerism. In 9 patients absolute neutrophil count (ANC) did not decrease to below 0.1 × 109/L whereas 2 patients never experienced ANC 20 × 109/L were achieved within 0 to 35 (median 12) days. Fourteen patients experienced no GVHD at all; severe GVHD (grades 3 and 4) was the single major complication and the cause of death in 4 patients, occurring after early discontinuation of cyclosporine A. Relapse was reversed by allogeneic cell therapy in 2/3 cases, currently with no residual host DNA (male) by cytogenetic analysis and polymerase chain reaction. To date, with an observation period extending over 1 year (median 8 months), 22 of 26 patients (85%) treated by allogeneic nonmyeloablative stem cell transplantation are alive, and 21 (81%) are disease-free. The actuarial probability of disease-free survival at 14 months is 77.5% (95% confidence interval, 53% to 90%). Successful eradication of malignant and genetically abnormal host hematopoietic cells by allogeneic nonmyeloablative stem cell transplantation represents a potential new approach for safer treatment of a large variety of clinical syndromes with an indication for allogeneic BMT. Transient mixed chimerism which may protect the host from severe acute GVHD may be successfully reversed postallogeneic BMT with graded increments of donor lymphocyte infusions, thus resulting in eradication of malignant or genetically abnormal progenitor cells of host origin.

Beta-thalassemia major (TM) is caused by any of approximately 150 mutations within the beta-globin gene. To establish the degree of chimerism after bone marrow transplantation (BMT), we have performed molecular analysis of beta-globin mutations in 14 patients with TM over a period of 10 years. All patients underwent T cell-depleted allogeneic BMT from HLA-identical related donors, using either in vitro T-cell depletion with CAMPATH 1M and complement or in vivo depletion using CAMPATH 1G in the bone marrow collection bag. To date, at different time periods after BMT, seven patients have some degree of chimerism; six of these patients, all blood transfusion-independent, have donor cells in the range of 70% to 95%, with stable mixed chimerism (MC). The seventh patient has less than 10% donor cells with, surprisingly, only minimal transfusion requirements. The detection of beta-globin gene point mutation, as used here, is a highly specific and sensitive marker for engraftment and MC in patients with thalassemia. In light of its specificity, the method is applicable in all cases of TM, as it is independent of sex and other non-globin-related DNA markers. The high incidence of MC found in our patients may be a consequence of the pre-BMT T-cell depletion. Because MC was associated with transfusion independence, complete eradication of residual host cells for effective treatment of TM and possibly other genetic diseases may prove not to be essential.

Kindlin-3 is a key lymphocyte function–associated antigen-1 (LFA-1) coactivator deleted in leukocyte adhesion deficiency-III (LAD-III). In the present study, we investigated the involvement of this adaptor in lymphocyte motility and TCR-triggered arrest on ICAM-1 or on dendritic cells (DCs). Kindlin-3–null primary T cells from a LAD-III patient migrated normally on the major lymph node chemokine CCL21 and engaged in normal TCR signaling. However, TCR activation of Kindlin-3–null T lymphocytes failed to trigger the robust LFA-1–mediated T-cell spreading on ICAM-1 and ICAM-1–expressing DCs that is observed in normal lymphocytes. Kindlin-3 was also essential for cytoskeletal anchorage of the LFA-1 heterodimer and for microclustering of LFA-1 within ventral focal dots of TCR-stimulated lymphocytes spread on ICAM-1. Surprisingly, LFA-1 on Kindlin-3–null lymphocytes migrating over CCL21 acquired normal expression of an epitope associated with the conformational activation of the key headpiece domain, β I. This activated LFA-1 was highly responsive to TCR-triggered ICAM-1–driven stop signals in normal T cells locomoting on CCL21, but not in their Kindlin-3–null T-cell counterparts. We suggest that Kindlin-3 selectively contributes to a final TCR-triggered outside-in stabilization of bonds generated between chemokine-primed LFA-1 molecules and cell-surface ICAM-1.

Leukocyte adhesion deficiency (LAD)-III is associated with homozygous stop codon mutations in Kindlin-3, the hematopoietic member of the Kindlin family of integrin coactivators. In addition, a subgroup of LAD-III patients has a homozygous splice junction mutation in and reduced expression of the Rap-1 guanine nucleotide exchange factor, CalDAG-GEFI (CDGI). In this study, we compared the adhesive properties of the leukocyte function-associated antigen-1 (LFA-1) and very late activation antigen-4 (VLA-4) integrins in both primary and activated leukocytes derived from these 2 LAD-III subgroups. Primary lymphocytes lacking both Kindlin-3 and CDGI lost all firm T-cell receptor-stimulated LFA-1 adhesiveness, in contrast to LAD-III lymphocytes deficient in Kindlin-3 alone. Effector T cells expanded from all tested LAD-III variants expressed normal CDGI, but lacked Kindlin-3. These Kindlin-3-null effector T cells exhibited total loss of inside-out LFA-1 activation by chemokine signals as well as abrogated intrinsic LFA-1 adhesiveness. Surprisingly, VLA-4 in Kindlin-3-null resting or effector lymphocytes retained intrinsic rolling adhesions to vascular cell adhesion molecule-1 and exhibited only partial defects in chemokine-stimulated adhesiveness to vascular cell adhesion molecule-1. Deletion of the putative beta(1) Kindlin-3 binding site also retained VLA-4 adhesiveness. Thus, our study provides the first evidence that Kindlin-3 is more critical to LFA-1 than to VLA-4-adhesive functions in human lymphocytes. Israel Science Foundation US-Israel Binational Science Foundation Minerva Foundation, Germany

Algorithms for grading acute graft-versus-host disease (GVHD) are inaccurate in assessing mortality risk. We developed a method to predict mortality by using data from 386 patients with acute GVHD. From the onset of GVHD to day 100, GVHD manifestations were scored for the skin, liver, and upper and lower gastrointestinal tract, and data were recorded for immunosuppressive treatment, performance, and fever. Logistic regression models predicting nonrelapse mortality (NRM) at day 200 were developed with data from 193 randomly selected patients and then validated in the remaining 193 patients. Clinical parameters were grouped to optimize predictive accuracy measured as the area under a receiver-operator characteristic (ROC) curve. The optimal model included the total serum bilirubin concentration, oral intake, need for treatment with prednisone, and performance score. When the overall burden of GVHD was measured by using average Acute GVHD Activity Index (aGVHDAI) scores for each patient in training and validation data sets, areas under ROC curves were 0.87 and 0.85, respectively. Contour lines were generated to reflect the predicted NRM at day 200 as a function of current aGVHDAI scores. These results demonstrate that clinical manifestations of GVHD severity can be used to accurately predict the risk of NRM in real time.

Recently, we reported a rare leukocyte adhesion deficiency (LAD) associated with severe defects in integrin activation by chemokine signals, despite normal ligand binding of leukocyte integrins.1 We now report that the small GTPase, Rap1, a key regulator of inside-out integrin activation is abnormally regulated in LAD Epstein-Barr virus (EBV) lymphocyte cells. Both constitutive and chemokine-triggered activation of Rap1 were abolished in LAD lymphocytes despite normal chemokine signaling. Nevertheless, Rap1 expression and activation by phorbol esters were intact, ruling out an LAD defect in Rap1 guanosine triphosphate (GTP) loading. The very late antigen 4 (VLA-4) integrin abnormally tethered LAD EBV lymphocytes to its ligand vascular cell adhesion molecule 1 (VCAM-1) under shear flow due to impaired generation of high-avidity contacts despite normal ligand binding and intact avidity to surface-bound anti-VLA-4 monoclonal antibody (mAb). Thus, a defect in constitutive Rap1 activation results in an inability of ligand-occupied integrins to generate high-avidity binding to ligand under shear flow. This is a first report of an inherited Rap1 activation defect associated with a pathologic disorder in leukocyte integrin function, we herein term it “LAD-III.” (Blood. 2004;103:1033-1036)

Abstract 1143 Poster Board I-165 Introduction Cyclosporine (CSA) is the backbone of graft vs. host disease (GVHD) prophylaxis in the last decades. It is established that CSA levels in the 1st weeks after transplant are critical for the rate and severity of GVHD. Initially, we gave CSA starting on day -1 in all our protocols. However, 8 years ago, we have changed CSA initiation in most of our protocols to day -4 in order to have stable, controlled therapeutic blood levels of CSA prior to transplant. Since nowadays the use of CSA is the most widespread prophylaxis of GVHD, we found it essential to compare the initiation of CSA on day -4 to day -1, consecutively determining the preferred initiating timing of CSA for patients who undergo allogeneic transplantations. Methods Out of 1716 patients that underwent allogeneic transplantation, we identified 2 groups of patients that received T-cell repleted grafts in which only CSA was used for GVHD prevention, starting on days -1 or -4 (n=219 and 261 respectively). The guidelines for CSA cessation and DLI were uniform in both groups. Both groups were compared for age, sex, donor type and matching, disease, disease status upon transplant, graft type, engraftment, GVHD (both acute and chronic), GVHD associated death and overall survival. Results The groups were found to be equal for age (p=0.83), sex (p=0.58), donor type (p=0.54), matching (p=0.98), disease type (malignant or non-malignant; p=0.25), graft type (PBSC or BM; p=0.45) and disease status (remission or active; p=0.42). The median time to ANC>500 was 16 and 15 days in the CSA -1 and -4 groups respectively with a trend toward better engraftment with initiation of CSA on day -4 (figure 1A, P=0.07). However, platelet engraftment was significantly better with CSA -4, with a median of 14 and 12 days in the CSA -1 and -4 groups respectively (figure 1B, p=0.0005). 112 and 138 patients developed acute GVHD (aGVHD) of any grade, respectively. Out of them 54% and 44% had severe (grade 3-4) aGVHD (p=0.45). The median time to aGVHD was similar, with a median of 29 and 28.5 days in the CSA -1 and -4 groups respectively (p=0.54). However, 64 patients developed cGVHD in the CSA -1 group, while 102 did so in the CSA -4 group (figure 2A, p=0.0002. Hazard ratio 0.59, 95% CI 0.37 to 0.73). Of these patients, 46.8% and 40.2% of the patients had extensive cGVHD (p=0.70), respectively. Additionally, despite lower GVHD rate in the CSA -1 group, GVHD associated death occurrence was more frequent then in the CSA -4 group (41/148 and 17/132 patients, p=0.02). Kaplan Meier analysis of all cause mortality showed higher mortality in the CSA -1 group (67.6% and 50.5%, figure 2B, p=0.074. Hazard ratio 1.24, 95% CI 0.98 to 1.58). Conclusions The initiation of CSA on day -4 improves engraftment, conversely increases the risk for cGVHD of any grade (possibly through prevention of tolerance), but reduces the risk of GVHD associated death and improves overall survival. Disclosures No relevant conflicts of interest to declare.

Abstract 2223 Poster Board II-200 Introduction: In general it is expected that strong immune suppression that alleviates GVHD, will increase the risk of relapse. Alefacept (Amevive®) is an immunosuppressive dimeric fusion protein that is used for psoriasis control. We showed its remarkable effect in acute steroid resistant/dependent and chronic extensive GVHD. Methods: To date, 42 patients with a median age of 31.5 years (range 3-66) were treated by us with alefacept due to acute (n=28) or chronic (n=14) steroid resistant/dependent GVHD (27 males, 15 females). Twenty were transplanted from HLA matched family members, 14 from HLA matched unrelated donors, 8 from mismatched donors and 2 from unrelated cord blood units. Pretreatment acute GVHD grade ranged 2-4 (median 3) and involved the skin (30), gut (19) and liver (7). All the patients with chronic GVHD had extensive involvement prior to therapy. Results: The median time from transplantation to alefacept was 42.5 days and 13 months in acute and chronic GVHD respectively (range 18-110d and 3-47.5m) and a median of 9 (range 1-25) injections that were given per patient. Thirty-four out of the 41 evaluable patients (83%) responded to the treatment (23/28 and 11/13 in the acute and chronic group respectively). Despite this high response rate, demonstrating the deep immunosuppressive and immunomodulative effect of alefacept, only 5/41 evaluable patients (figure 1) have relapsed (with a median follow-up of 30.8 months on the 17 survivors). Other than the 5 patients that relapsed (all treated with calcineurin inhibitor and steroids), full-donor peripheral blood chimerism (100% donor cells and no residual host-type DNA) was stable throughout the treatment period and later in all but one patient, that developed mixed chimerism under alefacept treatment. His chimerism returned to full-donor chimerism with taper-down of immune suppression. Currently, 17/42 (40.5%) patients are alive (figure 1), most with improved or stable GVHD. Twenty-six patients died due to either: GVHD progression (n=14), progression of the basic disease (n=4), infections (n=5) or other causes (n=3). Conclusions: Alefacept is effective and safe for the treatment of acute or chronic steroid resistant/dependent GVHD and may discriminate between GVHD and GVL. Disclosures: No relevant conflicts of interest to declare.

DBA is a congenital form of pure red cell anemia characterized by a macrocytic anemia, reticulocytopenia, and a block in erythroid differentiation at the proerythroblast stage, often in association with physical anomalies and growth retardation. About 25% of the patients carry mutations in genes that encode for proteins (RPS19, RPS24 and RPS17) that bind to the 40S subunit of the ribosome. The resultant defect in ribosomal biogenesis has been proposed to impair the initiation of globin translation, leading to mismatch between intracellular levels of heme and globin chains. It has been hypothesized that the transient excess of intracellular free heme resulting from the delay in globin synthesis exerts direct toxicity to erythroid precursors and plays a major role in pathogenesis of DBA through apoptosis of proerythroblasts (Keel et al., Science319;825,2008). Free hemin, however, is not necessarily toxic to developing erythroid precursors. Exogenously supplied hemin is readily taken up by erythroid cells in culture and its iron is incorporated into hemoglobin or stored in ferritin (Fibach et al., J Cell Physiol130;460,1987). Following addition of succinylacetone, a potent inhibitor of heme synthesis, exogenously supplied hemin can replace intracellularly synthesized heme and be incorporated into de novo formed hemoglobin (Fibach et al., Blood85;2967,1995). Hemin supplementation to semi-solid cultures promotes the growth of normal erythroid precursors (e.g., Lu and Broxmeyer, Exp Hematol11;721,1983). We showed in a two-phase liquid culture that exogenous hemin promotes normal erythropoiesis by accelerating the proliferation and hemoglobinization of erythroid precursors in the presence or absence of transferrin (Fibach et al., Blood85;2967,1995). This effect was particularly prominent during the early stages of hemoglobinization, when iron-uptake and heme synthesis are rate-limiting. In the present study we show that surplus hemin (10 - 50 mM) supplemented to cultures at early stage of erythroid development is well tolerated. Although the generation of reactive oxygen species (measured by staining with dichlorofluorescein diacetate) was modestly (50 ± 15%, N=4) increased, it was not associated with increased apoptosis, as measured by binding of annexin V, nor necrosis as measured by propidium iodide staining. Having demonstrated the growth and differentiation promoting potential of exogenous hemin on normal erythroid precursors and lack of overt toxicity, we studied the effect of exogenous heme in cultures of erythroid cells derived from six patients with DBA. We show that hemin, added as heme chloride or heme arginate, circumvented the primary defect and significantly stimulated (4 - 20-fold, p In conclusion, our results show that exogenous hemin is taken up by developing erythroid cells and can supplement or substitute endogenously synthesized heme; excess heme stimulates free radical generation moderately but does not cause apoptosis or necrosis; addition of hemin to cultured erythroid precursors derived from normal donors stimulates their growth and hemoglobinization, and in DBA, in contrast to the recently proposed scheme, heme can actually restore the growth and differentiation potential of the DBA-erythroid precursors. The beneficial effect of hemin on DBA erythroid precursors may be related to its effect on translation initiation factors, such as eIF-2 , and suggests a therapeutic potential.

Introduction: Prolonged treatment of graft vs. host disease (GVHD) is extremely immunosuppressive. Local therapy with intra-arterial (IA) injection of steroids may induce remission with lower extent of systemic immune suppression. Here, we present our experience with IA treatment of gastrointestinal (GI) and/or hepatic steroid resistant/dependent GVHD with 2 consecutive protocols. Patients and methods: Thirty five patients (37 GVHD events (hepatic, n=15), (GI, n=16), (combined, n=6)) were treated with 53 IA sessions. Most side effects were minor. Results: We found that IA steroid therapy was associated with partial and complete remission among patients with steroid resistant/dependent hepatic or GI GVHD. Hepatic partial response was observed in 14 (66.6%) patients among whom 7 (33.3%) reached complete response. GI partial response was observed in 19 (86.4%) patients among whom 12 (54.4%) reached complete response. Early administration of the local therapy, female gender, myeloid basic disease, and a non-active status of the basic disease at the day of transplantation were found related for predicting a better response for the intra-arterial treatment. The use of high dose steroids in the hepatic IA protocol from was at least as good as intermediate dose steroids with methotrexate (table 1, figure 1) and may be safer. Conclusions: Intra-arterial catheter guided steroid therapy is safe and effective in steroid resistant/dependent GVHD. Hepatic artery treatment with methotrexate can be safely substituted with high dose IA methylprednisolone. Further research is warranted characterizing the patients benefit most. Table 1 - comparison between 1st and 2nd hepatic IA treatment protocols 1st protocol 2nd protocol Significance Median age (range) 25 years (7–42) 32 years (18–59) P=0.09 Sex (M:F) 6:1 8:5 NS family donor vs MUD 6:1 9:4 NS Median time in days SCT-GVHD (range) 27 (13–133) 45 (13–248) NS Median time in days GVHD-IA (range) 15 (6–218) 190 (12–2615) P=0.09 Median peak GVHD grade (range) 3 (3–4) 3 (2–4) NS Highest pre-IA treatment bilirubin level (in mmol/L; normal

Severe combined immunodeficiency (SCID) due to adenosine deaminase (ADA) deficiency is a fatal congenital disorder of the immune system associated with systemic toxicity due to accumulation of purine metabolites. We previously showed that retroviral-mediated ADA gene transfer into autologous hematopoietic stem/progenitor cells (HSC) allowed restoration of immune and metabolic functions. We have now enrolled eight ADA-SCID children (age: 7–67 months) in our phase I/II gene therapy trial in which HSC are combined with low intensity conditioning with busulfan (total dose 4 mg/Kg i.v.). Previous treatment included haploidentical bone marrow transplant (n=3) or long-term (>1 year) enzyme replacement therapy (PEG-ADA) (n=4) associated with insufficient immune reconstitution or severe autoimmunity. In the latter case, PEG-ADA was discontinued to favour the growth advantage for gene corrected cells. The patients received a median dose of 8.8x106/Kg bone marrow CD34+ cells (range 0.9–10.8), containing on average 26.2±9.6% transduced CFU-C. Five patients experienced ANC 1 year after gene therapy, we observed a progressive increase in lymphocyte counts which was sustained over time (median at 1.5 years 1.6x109/L), polyclonal thymopoiesis and normalization of T-cell functions in vitro. Serum Ig levels improved and evidence of antigen-specific antibodies was obtained, leading to IVIG discontinuation in five patients. All the children are currently healthy and thriving, and none of them showed severe infections. Sustained ADA activity in lymphocytes and RBC resulted in a dramatic reduction of RBC purine toxic metabolites (dAXP

Alloreactive donor lymphocytes can induce potent graft vs malignancy (GVM) effects in conjunction with allogenenic stem cell transplantation or when given as donor lymphocyte infusion (DLI) following allogeneic stem cell transplantation. Unfortunately GVM effects are frequently accompanied by graft vs host disease (GVHD) and in many cases, despite severe GVHD, no adequate control of tumor progression can be achieved. In order to improve GVM effects while avoiding GVHD, we have attempted to use natural killer (NK) cells activated with rIL-2 isolated from the blood of haploidentically mismatched donors following engraftment of T-cell depleted donor stem cells. A pilot trial was conducted in patients with hematological malignancies 3 MDS; 2 AML; 1 ALL; 1 biphenotypic leukemia; 5 lymphoma (2 Burkitt’s; 1 anaplastic (ALK) & 1 Hodgkin’s), all resistant to chemotherapy, 4 failing autologous and 1 allogeneic transplant. Eleven received haploidentically mismatched graft. Prevention of rejection was accomplished by activation-induced clonal deletion following infusion of donor buffy coat 24 hours before administration of cytoxan 120–180mg/kg to eliminate host anti-donor T lymphocytes or by using total body irradiation 750cGy following conditioning that was fludarabine based. Positively selected CD34+ cells or CD3 depleted stem cells mobilized with G-CSF were infused on day 0. Based on our murine data, mismatched rIL-2 activated NK cells cause no GVHD. Consequently, 22 procedures were done involving administration of purified NK cells 1–15 months post transplantation. Donor mononuclear cells were activated 4 days at 37°C in 5% CO2 in air with rIL-2 (6,000 IU/ml) and administered on day 4. NK cells were positively selected with anti-CD56, or by negative selection of anti-CD3 using Miltenyi’s immunomagnetic beads. Using this procedure engraftment of mismatched allografts can be accomplished with no or minimal GVHD. Of 11 treated recipients, 5 entered CR. The first patient failing 2 prior transplants from a matched sibling is alive and well 4 years out. Out of 10 patients, 3 died from unrelated causes with no evidence of disease, but no GVHD developed. In conclusion, following engraftment of haploidentically mismatched related stem cell, donor-derived rIL-2 activated NK cells positively selected with anti-CD56 or negatively selected anti-CD3 immunomagnetic beads can be used to induce GVM effects while avoiding GVHD.

Whereas some patients with chemotherapy-resistant hematological malignancies may be successfully treated by allogeneic stem cell transplantation (SCT) through induction of graft-versus-tumor (GVT) effects, others may fail to respond or succumb to uncontrolled acute and/or chronic GVHD. We have previously documented that GVT effects induced by intentionally mismatched lymphocytes are much more potent and occur much faster, however, control of GVHD is mandatory. A pilot clinical trial was conducted in 19 patients (age 4–63): 6 AML; 2 ALL; 2 biphenotypic leukemia; 3 RAEB-t; 4 NHL; 2 Hodgkin’s; all fully resistant to chemotherapy (4 failing prior autologous and 1 prior allogeneic SCT). A total of 15 received haploidentical related transplant; 4 received matched sibling and 1 received matched unrelated donor. Prevention of rejection of mismatched allografts was accomplished by activation induced clonal deletion following infusion of donor with buffy coat 24 hours before administration of Cytoxan 120–180mg/kg to eliminate host anti-donor activated T lymphocytes. Additional conditioning included Fludarabine 30mg/m2 x 6; Busulfex 3.2mg/kg x 2 or low dose TBI 200–400 cGy ± Melphalan 70–140mg/m2 and ATG 40mg/kg or Alemtuzumab (60mg). Positively selected CD34+ cells mobilized with G-CSF were infused on day 0. Purified CD34+ cells were isolated by Miltenyi’s cliniMACS following mobilization with G-CSF. Based on our preclinical work suggesting that fully mismatched isolated NK cells cause no GVHD, 22 procedures were done involving administration of purified NK cells 1–15 months post SCT. Donor mononuclear cells were activated for 4 days at 37oC in 5% CO2 in air with rIL-2 (6,000 I.U./ml) and administered on day 4. NK cells were positively selected with Miltenyi’s anti-CD56 (n=15) or by anti-CD3 immunomagnetic beads (n=3). In 4 cases, CD56+ cells were separated before rIL-2 culturing. All 4 patients with matched donors received treatment of overt relapse despite GVHD. One, receiving the treatment prophylactically after transplant, is alive and well >29 months. Of 15 recipients of haploidentically mismatched allografts, 5 entered CR. One is alive and well after failing 2 prior transplants >27 months with no GVHD at all. Three patients died with no evidence of disease: 1 due to hepatitis B; 1 due to aspergillosis that was present before transplant and 1 committed suicide for unrelated causes. One patient died of CNS disease. Overall, GVHD developed in 4/19 cases (1 grade 2; 1 grade 4; 1 limited chronic; 1 extensive chronic), however, these patients had excessive T cell contamination with NK cells. In conclusion, carefully purified IL-2 activated intentionally mismatched NK cells can be most effective against resistant tumor cells and can be administered with minimal or no risk of GVHD. Immunotherapy with mismatched NK cells can be a practical approach to induce GVT effects with no GVHD.

Background: Hematopoietic stem cell transplantation (HSCT), the sole cure for β-thalassemia major (TM), is plagued by graft rejection or disease recurrence (10–30%). The risk of recurrence is influenced by age, type of allograft, presence of graft-versus-host-disease (GVHD), ferritin levels and hepatic iron overload. Complete donor hematopoiesis was believed essential for sustained marrow engraftment, but according to pre-clinical/clinical data, stable mixed chimera is also curative. We report our experience using T cell depletion (TCD) as prevention of GVHD versus fludarabine-based conditioning in unmanipulated HSCT. Material & Methods: 35 patients (6 mo-7 yr; median 29 mo; 10 female and 25 male) with TM were transplanted (1984–95) with HLA-matched donors siblings (29), parents (2), grandparents (2), cousin (1) and aunt (1). Conditioning encompassed total lymphoid irradiation (TLI) of 2 cGy/dx4 followed by busulfan 4 mg/kg/dx4 and cyclophosphamide 50 mg/kg/dx4; TCD was performed in vitro with CAMPATH1M (21) or CAMPATH1G “in the bag” (14). No GVHD-prophylaxis was given. Based on our results with non-myeloablative regimens, we used since 1996 fludarabine-based conditioning for TM. A total of 24 patients (2–23 yr; median 5 yr; 14 female and 10 male) were transplanted with unmanipulated inocula from HLA-matched siblings (20), matched grand-father (1), mismatched sibling (1) and MUD (2). For 21 patients conditioning consisted of fludarabine 30 mg/m2/dx6, busulfex (BU) 3.2 mg/kg/dx4; 2 patients received BU 3.2 mg/kg/dx2, and 1 patient received TBI 2 cGy/dx1 without BU. All patients received ATG (Frisenius) 10 mg/kg/dx4 and Cyclosporine from day -4 for GVHD-prophylaxis. Results: In the TCD group, no GVHD occurred; 5 patient (14.3%) died from transplant related complications, 5 patients (14.3%) rejected the graft and in 3 patients (8.6%) recurrence of TM occurred; 11 patients (31.4%) were 100% donor, 10 patients (28.6%) had stable mixed chimera (70–96% donor); one patient had 5% of donor cells 5 years after transplant and with allogeneic HSCT from the same donor following ambulatory non-myeloablative conditioning, complete displacement of host cells was accomplished (R.Or et al. Br J of Haem1996:94;285–7). In the fludarabine based group no transplant related mortality occurred; 4 patients (16.7%) rejected the graft (2 patients with BUx2; 1 patient transplanted from a mismatched sibling; 1 patient with unrelated donor), but all survived with autologous rescue. All 7 patients (29.2%) suffering from acute GVHD responded to treatment. 11 patients (45.8%) had 100% donor cells; 9 patients (37.5%) had stable mixed chimera and overall survival up to 9 years after HSCT is 100%. Conclusions: TCD is a feasible conditioning preventing completely GVHD. Unstable mixed chimera can be converted to full donor with donor lymphocyte infusion. TCD requires more intensive conditioning to prevent allograft rejection, but is associated with increased transplant related mortality. Fludarabine based conditioning is well tolerated with minimal procedure related morbidity and no mortality. The incidence of acute GVHD was low and treatable. Low dose BU is not sufficient for consistent engraftment; mixed chimera was frequent, yet stable and associated with functionally normal hematopoiesis. With reduced intensity conditioning, age seems no longer to be contraindication for HSCT for TM.

Although allogenenic stem cell transplantation may provide a cure for a growing number of patients with hematologic malignancies and several metastatic solid tumors, several problems remain to be solved. In routine medical practice transplant can be offered for patients with a matched donor available whereas the large majority of patients in need have no matched donor available. Although alloreactive lymphocytes may eliminate residual malignant cells, such an effect is accompanied by acute and chronic GVHD which may be hazardous even in recipients with perfectly matched allografts, and prohibitive in recipients treated with haploidentically mismatched allografts. On the other hand immunotherapy with intentionally mismatched allografts could provide a much more effective tool for eradication of tumor cells resistant to chemotherapy. We have pioneered a new approach for treatment of patients with resistant hematological malignancies (AML/MDS 5; ALL 1; Biphenotype 2; NHL 3; HD 1) using matched siblings (n=4), matched unrelated donor (n=1) or haploidentically mismatched donors (n=7). Prevention of rejection of mismatched allografts was accomplished by combination of fludarabine and deletion of donor reactive host lymphocytes by infusion of donor mononuclear blood cells and elimination of alloreactive lymphocytes susceptible to high-dose cyclosphosphamide (60mg/kgx3) one day later. Prevention of GVHD following infusion of G-CSF mobilized, haploidentically mismatched blood stem cells was accomplished using Miltenyi’s immunomagentic beads coupled with anti-AC133 (n=6) or using anti-CD3 (n=1). No other anti-GVHD prophylaxis was used. Following transplantation, patients were treated with rIL-2 activated donor peripheral blood lymphocytes activated for 4 days at 37°C in 5% C02 in air incubator with rIL-2 6,000 IU/ml. T cell depletion was accomplished either by positive selection of CD56+ (n=10) or negative selection of CD3 (n=2) for optimal induction of graft vs leukemia (GVL) effects by mismatched and fully activated NK cells. One patient with resistant leukemia became disease free for 8 months but died of resistant aspergilosis which was evident prior to transplantation. Five out of 12 patients with intractable and fully resistant leukemia are alive with no GVHD and no evidence of disease 1–18 (median 13) months post transplantation. Based on our ongoing preliminary study we conclude that patients with resistant hematological malignancies may benefit from cell therapy mediated by rIL-2 activated donor lymphocytes, and most likely from intentionally mismatched haploidentical allografts following elimination of host anti-donor alloreactive lymphocytes and prevention of GVHD by positively or negatively selected stem cells, followed by immunotherapy with rIL-2 activated CD3 depleted NK cells. Intentionally mismatched rIL-2 activated NK cells represents a safe approach for elimination of residual tumor cells, aiming for induction of GVL while avoiding GVHD.

To investigate whether platelet ABO antigens are intrinsic to platelets or adsorbed from plasma, the amount of A-antigen on washed platelets was determined and related to the Lewis and secretor phenotypes, as these genes control the amount A, B. or H antigen present in plasma. The A-antigen on platelets was quantitated using a plate-let-erythrocyte mixed agglutination assay in 28 A, individuals of different Lewis and secretor phenotype. There was a direct correlation (r = 0.6, p