Search

Allogeneic hematopoietic stem cell transplantation (HSCT) is the most effective treatment in eradicating high-risk acute myeloid leukemia (AML). Here, we present data from a novel HLA-haploidentical HSCT protocol that addressed the 2 remaining major unmet medical needs: leukemia relapse and chronic graft-versus-host disease (cGVHD). Fifty AML patients were enrolled in the study. The conditioning regimen included total body irradiation for patients up to age 50 years and total marrow/lymphoid irradiation for patients age 51 to 65 years. Irradiation was followed by thiotepa, fludarabine, and cyclophosphamide. Patients received an infusion of 2 × 106/kg donor regulatory T cells on day −4 followed by 1 × 106/kg donor conventional T cells on day −1 and a mean of 10.7 × 106 ± 3.4 × 106/kgpurified CD34+ hematopoietic progenitor cells on day 0. No pharmacological GVHD prophylaxis was administered posttransplantation. Patients achieved full donor–type engraftment. Fifteen patients developed grade ≥2 acute GVHD (aGVHD). Twelve of the 15 patients with aGVHD were alive and no longer receiving immunosuppressive therapy. Moderate/severe cGVHD occurred in only 1 patient. Nonrelapse mortality occurred in 10 patients. Only 2 patients relapsed. Consequently, at a median follow-up of 29 months, the probability of moderate/severe cGVHD/relapse-free survival was 75% (95% confidence interval, 71%-78%). A novel HLA-haploidentical HSCT strategy that combines an age-adapted myeloablative conditioning regimen with regulatory and conventional T-cell adoptive immunotherapy resulted in an unprecedented cGVHD/relapse-free survival rate in 50 AML patients with a median age of 53 years. This trial was registered with the Umbria Region Institutional Review Board Public Registry as identification code 02/14 and public registry #2384/14 and at www.clinicaltrials.gov as #NCT03977103.

Allogeneic hematopoietic stem cell transplantation (HSCT) is the most effective treatment in eradicating high-risk acute myeloid leukemia (AML). Here, we present data from a novel HLA-haploidentical HSCT protocol that addressed the 2 remaining major unmet medical needs: leukemia relapse and chronic graft-versus-host disease (cGVHD). Fifty AML patients were enrolled in the study. The conditioning regimen included total body irradiation for patients up to age 50 years and total marrow/lymphoid irradiation for patients age 51 to 65 years. Irradiation was followed by thiotepa, fludarabine, and cyclophosphamide. Patients received an infusion of 2 × 106/kg donor regulatory T cells on day −4 followed by 1 × 106/kg donor conventional T cells on day −1 and a mean of 10.7 × 106± 3.4 × 106/kgpurified CD34+hematopoietic progenitor cells on day 0. No pharmacological GVHD prophylaxis was administered posttransplantation. Patients achieved full donor–type engraftment. Fifteen patients developed grade ≥2 acute GVHD (aGVHD). Twelve of the 15 patients with aGVHD were alive and no longer receiving immunosuppressive therapy. Moderate/severe cGVHD occurred in only 1 patient. Nonrelapse mortality occurred in 10 patients. Only 2 patients relapsed. Consequently, at a median follow-up of 29 months, the probability of moderate/severe cGVHD/relapse-free survival was 75% (95% confidence interval, 71%-78%). A novel HLA-haploidentical HSCT strategy that combines an age-adapted myeloablative conditioning regimen with regulatory and conventional T-cell adoptive immunotherapy resulted in an unprecedented cGVHD/relapse-free survival rate in 50 AML patients with a median age of 53 years. This trial was registered with the Umbria Region Institutional Review Board Public Registry as identification code 02/14 and public registry #2384/14 and at www.clinicaltrials.govas #NCT03977103.

Allogeneic hematopoietic cell transplantation from a human leukocyte antigen (HLA) haplotype mismatched donor (haploidentical transplantation) was not feasible for the treatment of hematologic malignancies until the early 1990s, due to the high risk of rejection and graft-versus-host disease (GVHD). The first successful protocol of haploidentical transplantation was based on a highly myeloablative and immunosuppressive conditioning regimen and the infusion of a “mega-dose” of T-cell-depleted hematopoietic stem cells. More than 90% of patients engrafted and <10% developed GVHD. The protocol did not include post-transplant immunosuppression, which favored the graft-versus-tumor effect mediated by alloreactive NK cells and residual alloreactive T cells. However, donor post-transplant immune reconstitution was slow with a high risk of infection-related mortality. More recently, T-cell-depleted haploidentical transplantation has become the platform for innovative cell therapies that aim to enhance T-cell immunity while preventing adverse reactions against host tissues. One strategy is adoptive immunotherapy with conventional T cells and regulatory T cells. Preclinical studies and clinical trials have proven that regulatory T cells control GVHD caused by co-infused conventional T cells while the graft-versus-tumor effect is retained. The use of regulatory T cells in the absence of any other form of immune suppression allowed for a conventional T cell-mediated full eradication of disease in the vast majority of high-risk acute leukemia patients.

Because activating killer cell immunoglobulinlike receptors (KIRs) are heterogeneously expressed in the population, we investigated the role of donor activating KIRs in haploidentical hematopoietic transplants for acute leukemia. Transplants were grouped according to presence vs absence of KIR-ligand mismatches in the graft-vs-host direction (ie, of donor-vs-recipient natural killer [NK]-cell alloreactivity). In the absence of donor-vs-recipient NK-cell alloreactivity, donor activating KIRs had no effects on outcomes. In the 69 transplant pairs with donor-vs-recipient NK-cell alloreactivity, transplantation from donors with KIR2DS1and/or KIR3DS1was associated with reduced risk of nonrelapse mortality, largely infection related (KIR2DS1present vs absent: hazard ratio [HR], 0.25; P= .01; KIR3DS1present vs absent: HR, 0.18; P= .006), and better event-free survival (KIR2DS1present vs absent: HR, 0.31; P= .011; KIR3DS1present vs absent: HR, 0.30; P= .008). Transplantation from donors with KIR2DS1and/or KIR3DS1was also associated with a 50% reduction in infection rate (P =.003). In vitro analyses showed that KIR2DS1 binding to its HLA-C2 ligand upregulated inflammatory cytokine production by alloreactive NK cells in response to infectious challenges. Because ∼40% of donors able to exert donor-vs-recipient NK-cell alloreactivity carry KIR2DS1and/or KIR3DS1, searching for them may become a feasible, additional criterion in donor selection.

Because activating killer cell immunoglobulinlike receptors (KIRs) are heterogeneously expressed in the population, we investigated the role of donor activating KIRs in haploidentical hematopoietic transplants for acute leukemia. Transplants were grouped according to presence vs absence of KIR-ligand mismatches in the graft-vs-host direction (ie, of donor-vs-recipient natural killer [NK]-cell alloreactivity). In the absence of donor-vs-recipient NK-cell alloreactivity, donor activating KIRs had no effects on outcomes. In the 69 transplant pairs with donor-vs-recipient NK-cell alloreactivity, transplantation from donors with KIR2DS1 and/or KIR3DS1 was associated with reduced risk of nonrelapse mortality, largely infection related (KIR2DS1 present vs absent: hazard ratio [HR], 0.25; P = .01; KIR3DS1 present vs absent: HR, 0.18; P = .006), and better event-free survival (KIR2DS1 present vs absent: HR, 0.31; P = .011; KIR3DS1 present vs absent: HR, 0.30; P = .008). Transplantation from donors with KIR2DS1 and/or KIR3DS1 was also associated with a 50% reduction in infection rate (P = .003). In vitro analyses showed that KIR2DS1 binding to its HLA-C2 ligand upregulated inflammatory cytokine production by alloreactive NK cells in response to infectious challenges. Because ∼40% of donors able to exert donor-vs-recipient NK-cell alloreactivity carry KIR2DS1 and/or KIR3DS1, searching for them may become a feasible, additional criterion in donor selection.

Posttransplant relapse is still the major cause of treatment failure in high-risk acute leukemia. Attempts to manipulate alloreactive T cells to spare normal cells while killing leukemic cells have been unsuccessful. In HLA-haploidentical transplantation, we reported that donor-derived T regulatory cells (Tregs), coinfused with conventional T cells (Tcons), protected recipients against graft-versus-host disease (GVHD). The present phase 2 study investigated whether Treg-Tcon adoptive immunotherapy prevents posttransplant leukemia relapse. Forty-three adults with high-risk acute leukemia (acute myeloid leukemia 33; acute lymphoblastic leukemia 10) were conditioned with a total body irradiation–based regimen. Grafts included CD34+cells (mean 9.7 × 106/kg), Tregs (mean 2.5 × 106/kg), and Tcons (mean 1.1 × 106/kg). No posttransplant immunosuppression was given. Ninety-five percent of patients achieved full-donor type engraftment and 15% developed ≥grade 2 acute GVHD. The probability of disease-free survival was 0.56 at a median follow-up of 46 months. The very low cumulative incidence of relapse (0.05) was significantly better than in historical controls. These results demonstrate the immunosuppressive potential of Tregs can be used to suppress GVHD without loss of the benefits of graft-versus-leukemia (GVL) activity. Humanized murine models provided insights into the mechanisms underlying separation of GVL from GVHD, suggesting the GVL effect is due to largely unopposed Tcon alloantigen recognition in bone marrow.

Posttransplant relapse is still the major cause of treatment failure in high-risk acute leukemia. Attempts to manipulate alloreactive T cells to spare normal cells while killing leukemic cells have been unsuccessful. In HLA-haploidentical transplantation, we reported that donor-derived T regulatory cells (Tregs), coinfused with conventional T cells (Tcons), protected recipients against graft-versus-host disease (GVHD). The present phase 2 study investigated whether Treg-Tcon adoptive immunotherapy prevents posttransplant leukemia relapse. Forty-three adults with high-risk acute leukemia (acute myeloid leukemia 33; acute lymphoblastic leukemia 10) were conditioned with a total body irradiation–based regimen. Grafts included CD34+ cells (mean 9.7 × 106/kg), Tregs (mean 2.5 × 106/kg), and Tcons (mean 1.1 × 106/kg). No posttransplant immunosuppression was given. Ninety-five percent of patients achieved full-donor type engraftment and 15% developed ≥grade 2 acute GVHD. The probability of disease-free survival was 0.56 at a median follow-up of 46 months. The very low cumulative incidence of relapse (0.05) was significantly better than in historical controls. These results demonstrate the immunosuppressive potential of Tregs can be used to suppress GVHD without loss of the benefits of graft-versus-leukemia (GVL) activity. Humanized murine models provided insights into the mechanisms underlying separation of GVL from GVHD, suggesting the GVL effect is due to largely unopposed Tcon alloantigen recognition in bone marrow.

Today human leukocyte antigen-haploidentical transplantation is a feasible option for patients with high-risk acute leukemia who do not have matched donors. Whether it is T-cell replete or T-cell depleted, it is still, however, associated with issues of transplant-related mortality and posttransplant leukemia relapse. After reports that adoptive immunotherapy with T-regulatory cells controls the alloreactivity of conventional T lymphocytes in animal models, tomorrow’s world of haploidentical transplantation will focus on new “designed” grafts. They will contain an appropriate ratio of conventional T lymphocytes and T-regulatory cells, natural killer cells, γ δ T cells, and other accessory cells. Preliminary results of ongoing clinical trials show the approach is feasible. It is associated with better immune reconstitution and a quite powerful graft-versus-leukemia effect with a low incidence of graft-versus-host disease and no need for posttransplant pharmacological prophylaxis. Future strategies will focus on enhancing the clinical benefit of T-regulatory cells by increasing their number and strengthening their function.

Hastening posttransplantation immune reconstitution is a key challenge in human leukocyte antigen (HLA)–haploidentical hematopoietic stem-cell transplantation (HSCT). In experimental models of mismatched HSCT, T-regulatory cells (Tregs) when coinfused with conventional T cells (Tcons) favored posttransplantation immune reconstitution and prevented lethal graft-versus-host disease (GVHD). In the present study, we evaluated the impact of early infusion of Tregs, followed by Tcons, on GVHD prevention and immunologic reconstitution in 28 patients with high-risk hematologic malignancies who underwent HLA-haploidentical HSCT. We show for the first time in humans that adoptive transfer of Tregs prevented GVHD in the absence of any posttransplantation immunosuppression, promoted lymphoid reconstitution, improved immunity to opportunistic pathogens, and did not weaken the graft-versus-leukemia effect. This study provides evidence that Tregs are a conserved mechanism in humans.

Hastening posttransplantation immune reconstitution is a key challenge in human leukocyte antigen (HLA)–haploidentical hematopoietic stem-cell transplantation (HSCT). In experimental models of mismatched HSCT, T-regulatory cells (Tregs) when coinfused with conventional T cells (Tcons) favored posttransplantation immune reconstitution and prevented lethal graft-versus-host disease (GVHD). In the present study, we evaluated the impact of early infusion of Tregs, followed by Tcons, on GVHD prevention and immunologic reconstitution in 28 patients with high-risk hematologic malignancies who underwent HLA-haploidentical HSCT. We show for the first time in humans that adoptive transfer of Tregs prevented GVHD in the absence of any posttransplantation immunosuppression, promoted lymphoid reconstitution, improved immunity to opportunistic pathogens, and did not weaken the graft-versus-leukemia effect. This study provides evidence that Tregs are a conserved mechanism in humans.

Introduction

No relevant conflicts of interest to declare.

Much progress has been made in the clinical, biological and technical aspects of the T-cell-depleted full-haplotype mismatched transplants for acute leukemia. Our experience demonstrates that infusing a megadose of extensively T-cell-depleted hematopoietic peripheral blood stem cells after an immunomyeloablative conditioning regimen in acute leukemia patients ensures sustained engraftment with minimal graft-vs-host disease (GvHD) without the need of any post-transplant immunosuppressive treatment. Since our first successful pilot study, our efforts have concentrated on developing new conditioning regimens, optimizing the graft processing and improving the post-transplant immunological recovery. The results we have so far achieved in more than 200 high-risk acute leukemia patients show that haploidentical transplantation is now a clinical reality. Because virtually all patients in need of a hematopoietic stem cell transplant have a full-haplotype mismatched donor, who is immediately available, a T-cell depleted mismatched transplant should be offered, not as a last resort, but as a viable option to high risk acute leukemia patients who do not have, or cannot find, a matched donor.

In this study we determined whether human stromal cells could be engineered with a retroviral vector carrying the interleukin 7 (IL-7) gene and investigated the effects on T cells in vitro and in vivo in a murine model. Transduced mesenchymal cells strongly express CD90 (98.15%), CD105 (87.6%), and STRO-1 (86.7%). IL-7 production was 16.37 (±2 SD) pg/ml, which remained stable for 60 days. In vitro-immunoselected naive T cells maintained the CD45RA+CD45RO− naive phenotype (4.2 times more than controls) after 7 days of culture with IL-7-engineered stromal cells. The apoptosis rate (4.7%) of the naive T cells cultured with transduced stromal cells overlapped with that of freshly isolated cells. Immunohistological analysis detected stromal cells in bone marrow, spleen, and thymus. Cotransplantation of IL-7-engineered stromal cells with CD34+ cells improved engraftment in terms of CD45+ cells and significantly increased the CD3+ cell count in peripheral blood, bone marrow, and spleen. These data demonstrate the following: (1) human stromal cells can be transduced, generating a normal layer; (2) transduced stromal cells in vitro maintain the naive T cell phenotype; and (3) IL-7-transduced stromal cells in vivo home to lymphoid organs and produce sufficient IL-7 in loco, supporting T cell development in a cotransplantation model. Because of their efficient cytokine production and homing, IL-7-engineered stromal cells might be an ideal vehicle to hasten immunological reconstitution in T cell-depleted hosts.

Patients who undergo transplantation with haploidentical “three-loci” mismatched T-cell-depleted bone marrow (BM) are at high risk for graft failure. To overcome the host-versus-graft barrier, we increased the size of the graft inoculum, which has been shown to be a major factor in controlling both immune rejection and stem cell competition in murine models. Seventeen patients (mean age, 23.2 years; range, 6 to 51 years) with end-stage chemoresistant leukemia were received transplants of a combination of BM with recombinant human granulocyte colony-stimulating factor-mobilized peripheral blood progenitor cells from HLA-haploidentical “three-loci” incompatible family members. The average concentration of colony-forming unit-granulocyte-macrophage in the final inoculum was sevenfold to 10-fold greater than that found in BM alone. The sole graft-versus-host disease (GVHD) prophylaxis consisted of T-cell depletion of the graft by the soybean agglutination and E-rosetting technique. The conditioning regimen included total body irradiation in a single fraction at a fast dose rate, antithymocyte globulin, cyclophosphamide and thiotepa to provide both immunosuppression and myeloablation. One patient rejected the graft and the other 16 had early and sustained full donor-type engraftment. One patient who received a much greater quantity of T lymphocytes than any other patient died from grade IV acute GVHD. There were no other cases of GVHD ≥grade II. Nine patients died from transplant-related toxicity, 2 relapsed, and 6 patients are alive and event-free at a median follow-up of 230 days (range, 100 to 485 days). Our results show that a highly immunosuppressive and myeloablative conditioning followed by transplantation of a large number of stem cells depleted of T lymphocytes by soybean agglutination and E-rosetting technique has made transplantation of three HLA-antigen disparate grafts possible, with only rare cases of GVHD.

Three radiation protocols currently used in treatment of leukemia patients before bone marrow transplantation (BMT) were investigated in a murine model (C57BL/ 6 → C3H/HeJ) for BM allograft rejection. These include (a) a single dose of total body irradiation (8.5 Gy TBI delivered at a dose rate of 0.2 Gy/min). (b) fractionated TBI (12 Gy administered in six fractions, 2 Gy twice a day in 3 days, delivered at a dose rate of 0.1 Gy/min, and (c) hyperfraction-ated TBI (14.4 Gy administered in 12 fractions, 1.2 Gy three times a day in 3 days, delivered at a dose rate of 0.1 Gy/min). Donor-type chimerism 6 to 8 weeks after BMT and hematologic reconstitution on day 12 after BMT found in these groups were compared with results obtained in mice conditioned with 8 Gy TBI delivered at a dose rate of 0.67 Gy/min, routinely used in this murine model. The results in both parameters showed a marked advantage for the single dose 8.5 Gy TBI over all the other treatments. This advantage was found to be equivalent to a three- to fourfold increment in the BM inoculum when compared with hyperfractionated radiation, which afforded the least favorable conditions for development of donor-type chimerism. The fractionated radiation protocol was equivalent in its efficacy to results obtained in mice irradiated by single-dose 8 Gy TBI, both of which afforded a smaller but not significant advantage over the hyperfractionated protocol. This model was also used to test the effect of radiation dose rate on the development of donor-type chimerism. A significant enhancement was found after an increase in dose rate from 0.1 to 0.7 Gy/min. Further enhancement could be achieved when the dose rate was increased to 1.3 Gy/min, but survival at this high dose rate was reduced. These results demonstrated indirectly that dose rate affects the expression of host-type pluripotent stem cells, the progeny of which appear 3 to 6 weeks after treatment with 8 Gy TBI delivered at a dose rate of 0.1 Gy/min, but which are eradicated if radiation is delivered at a dose rate of 1.3 Gy/min.

Peripheral blood T cell subsets were evaluated in 11 patients during the reconstitution phase after allogeneic bone marrow transplantation and compared with 11 age-matched controls. The proportion of cells coexpressing Leu7 and CD11b (C3bi receptor) markers was determined within the CD4+(T-helper) and the CD8+(T-suppressor) subsets by two-color immunofluorescence analysis. CD4+and CD8+T cells reached normal or near-normal values within the first year posttransplant. In contrast to normal controls, however, most of the cells in both subsets coexpressed the Leu7 and CD11b markers. T cells with such phenotype display the morphological features of granular lymphocytes (GLs) and a functional inability to produce interleukin 2 (IL 2). These T cell imbalances were not related to graft v host disease (GvHD) or to clinically detectable virus infections and may account for some defects of cellular and humoral immunity that occur after bone marrow transplantation.

A major problem in using murine models for studies of bone marrow allograft rejection in leukemia patients is the narrow margin in which graft rejection can be analyzed. In mice irradiated with >9 Gy total body irradiation (TBI) rejection is minimal, whereas after administration of 8 Gy TBI, which spares a significant number of clonable T cells, a substantial frequency of host stem cells can also be detected. In current murine models, unlike in humans, bone marrow allograft rejection is generally associated with full autologous hematopoietic reconstitution. In the present study, we investigated the effect of the myeloablative drug dimethyl myleran (DMM) on chimerism status following transplantation of T cell-depleted allogenic bone marrow (using C57BL/6 donors and C3H/HeJ recipients, conditioned with 8 Gy TBI). Donor type chimerism 1 to 2 months post-transplant of 1 to 3 x 106bone marrow cells was markedly enhanced by using DMM one day after TBI and prior to transplantation. Conditioning with cyclophosphamide instead of DMM, in combination with 8 Gy TBI, did not enhance engraftment of donor type cells. Artificial reconstitution of T cells, after conditioning with TBI plus DMM, by adding mature thymocytes, or presensitization with irradiated donor type spleen cells 1 week before TBI and DMM, led to strong graft rejection and consequently to severe anemia. The anti-donor responses in these models were proportional to the number of added T cells and to the number of cells used for presensitization, and they could be neutralized by increasing the bone marrow inoculum. These results demonstrate the potential of DMM to facilitate engraftment in unsensitized mice in which the host stem cells may compete with donor type cells; the use of DMM to create models in which mechanisms of immune rejection can be studied without interference due to stem cell competition; and that bone marrow allograft rejection may be overcome by increasing the bone marrow inoculum in these stringent models.

T cells recovering after bone marrow transplantation (BMT) were analyzed for their phenotypic and functional features by two-color immunofluorescence and a high efficiency cloning technique. A predominance of cells co-expressing natural killer (NK)-related surface antigens, such as Leu 7 (CD57) and CD11 b, was detected within both the CD4+ and CD8+ subsets from 5 months postgrafting onward. Such cells are virtually absent among normal circulating CD4+ cells and account for a minority (≈30%) of normal CD8+ cells. Postgrafting T cells representative of the whole range of NK-related antigen co-expression were selected from six patients for clonal analyses. In control subjects, 63% and 41% of the CD4+ and CD8 + clones, respectively, produced interleukin-2 (IL-2) whereas %30% of either CD4 + or CD8+ control clones produced interferon (IFN)-gamma. At variance, and irrespective of their CD4+/CD8+ phenotype, lower proportions of BMT recipient-derived clones produced IL-2 (20% and 12%, respectively), whereas the majority of both CD4+ and CD8+ clones (75% and 71%, respectively) released high amounts of IFN-gamma. Purified populations of CD57 + / CD11 b + vnegative cells from two BMT recipients and two control subjects were cloned and subsequently evaluated for IL-2 and IFN-gamma production. CD57 + /CD11b+ cell-derived clones were poor IL-2 producers in both normal subjects and BMT patients. In contrast, IL-2-producing clones were frequent (62% to 79%) among those derived from CD57-/CD11b– cells from normal subjects, whereas they were still represented at lower than normal proportions, ie, 25% to 41 %, among clones generated from BMT recipients. CD57+/CD11b+ cells gave rise to comparably high proportions of IFN-gamma producing clones in both normal subjects and BMT recipients (~80%). In contrast, IFN-gamma producing clones were ~25% to 50% of CD57 – /CD11b– cell-derived clones in both normal subjects and BMT patients. Therefore, while the predominance of NK-related antigen-positive T cells may be predictive of poor IL-2 and high IFN-gamma production, the immune derangement in long-term BMT recipients is further enhanced by the finding that all T cells may be poor IL-2 producers. It is also suggested that IL-2 production is a preferential function of T cells that do not express CD57 and CD11 b, whereas IFN-gamma production is attributable to T cells that express CD57 and CD11 b. e1989 by Grune & Stratton, Inc.

Background:

No relevant conflicts of interest to declare.

No relevant conflicts of interest to declare.

No relevant conflicts of interest to declare.

Allogeneic hematopoietic transplantation is the most powerful therapy for patients with acute leukemia at high risk of relapse. For HLA haploidentical transplantation, we recently applied a protocol consisting of a high-intensity TBI-based conditioning regimen and a peripheral blood CD34+ hematopoietic stem cell graft combined with the adoptive transfer of naturally occurring regulatory T cells (Tregs) and conventional T cells (Tcons). No post-transplant pharmacologic immune suppressive GvHD prophylaxis was given. Such a protocol was associated with remarkably low GvHD and relapse rates (Martelli MF et al., Blood. 124:638-644, 2014; reviewed in Mancusi A et al., Blood 128:2616-2623, 2016).

Bonini: MolMed S.p.A: Consultancy.

Introduction and Aim:HLA-haploidentical hematopoietic stem cell transplantation (haplo-HSCT) is increasingly offered to patients with high-risk acute myeloid (AML) or lymphoid leukemia (ALL). Unfortunately, graft manipulation employed to overcome the HLA barrier significantly delays immune reconstitution, posing the patients at risk of infections. Accordingly, non-relapse mortality after haplo-HSCT clearly extends beyond day 100 post-transplant. Over the years, different approaches have been investigated to speed-up immune reconstitution. In the absence of validated immune biomarkers, it is however difficult to evaluate the clinical impact of accelerated immune reconstitution. The aim of this EBMT retrospective study is to explore immune-cell counts early after haplo-HSCT as predictive of its overall outcome.

No relevant conflicts of interest to declare.

Posttransplant relapse is still a major cause of treatment failure in high-risk acute leukemia. Evidence from experimental bone marrow transplantation (BMT) showed coinfusion of conventional T lymphocytes (Tcons) with regulatory T lymphocytes (Tregs) suppressed lethal Graft-versus-Host Disease (GvHD) without impairing the Graft-versus-Leukemia (GvL) effect (Edinger et al., Nat Med. 2003;9:1144-1150). In HLA haploidentical transplantation for acute leukemia we demonstrated that donor-derived natural Tregs, coinfused with Tcons, protected recipients against GvHD (Di Ianni et al., Blood 2011, 117:3921-3928) and largely prevented posttransplant leukemia relapse, as only 5% of evaluable patients relapsed (Martelli et al., Blood 2014, 124:638-44).

CMML patients even when heavily pretreated, can be responsive to decitabine which can be administered for several cycles with good tolerability. Our data are confirming reported data on ORR and OS after decitabine therapy (Braun et al Blood 2011). OS of decitabine- responding patients is significantly prolonged. No clinical or cytogenetic characteristics seem to be predictive of response to decitabine. No significant difference in enzyme expression was demonstrated, although a trend in lower expression of hENT1 was noted for non responsive cases.Santini: Celgene: Honoraria; Janssen : Honoraria; Novartis: Honoraria; gsk: Honoraria.

No relevant conflicts of interest to declare.

In HLA-haploidentical stem cell transplantation we showed adoptive immunotherapy with naturally occurring T regulatory cells (nTregs) followed by conventional T cells (Tcons) prevented acute and chronic graft-versus-host-disease (GvHD), favoured lymphoid reconstitution and immunity against pathogens (Di Ianni et al., Blood 2011). One major concern is whether a graft-versus-leukemia (GvL) effect is maintained since FoxP3+Tregs can also suppress immune response against tumour. However, we showed in immunodeficient NSG mice engrafted with human myeloid and lymphoblastic acute leukemia cells (7x106iv), Treg (3x106)+Tcon(3x106) infusion rescued mice from leukemia death in the absence of GvHD. In contrast, controls died of leukemia (when not treated or infused with Tregs only) or of GvHD (when infused with Tcons only).Since September 2008, 45 consecutive patients (21 male; 24 female; median age 40, range 20-59) with high-risk acute leukemia ( 33 AML, 12 ALL) were transplanted. The conditioning regimen included TBI, thiotepa, fludarabine and cyclophosphamide (25 patients in the 1stclinical trial) or alemtuzumab/ATG (20 patients in the 2ndclinical trial). On day -4, patients received an infusion of freshly isolated donor Tregs (mean 2.5/kg ±1; Foxp3 81.01%±16.47; Helios/FoxP3 54%±8.4; CD127 11.72%±7.653) followed on day 0 by a megadose of purified CD34+ cells (9.7x106/kg ± 3.1) and Tcons (mean 1.1 /kg ±0.6; CD3 90.72%± 9.6; CD4 57.77% ± 8.85; CD8 31.21% ± 8.59). No post-transplant GvHD prophylaxis was given. 43 of the 45 patients achieved primary, sustained full-donor type engraftment. Only 6/43 evaluable patients developed ≥ grade II acute GvHD (3 alive, treated and now off-therapy). At a median follow-up of 28 months (range 4-58) no patient has developed chronic GvHD. CD4 and CD8 counts reached 100/µL medianly on days 40 (range 25-150) and 45 (range 18-100) and 200/µL on days 55 (range 45-160) and 60 (range 50-140). In KIR-ligand mismatched transplants (Ruggeri et al., Science 2002; Ruggeri et al., Blood 2007), donor versus recipient alloreactive NK cells were also preserved. At a median follow-up of 46 months (range 7-58) 23/43 patients are alive and disease free with a probability of DFS of 0.56. The cumulative incidence of TRM is 0.37. The cumulative incidence of relapse was 0.06 as only 2 AML patients transplanted from non-NK alloreactive donors have relapsed to date. This incidence is extremely low considering the high-risk patient population (9 with FLT-3/ITD; 8 with complex karyotypes; 4 with t(9:22); 2 PIF; 1 secondary AML; 1CR after second-line induction; 1 with CNS and skin localization at diagnosis; 19 ≥ 2CR). In conclusion, adoptive immunotherapy with Tregs and Tcons exerts a powerful GVT activity in both AML and ALL patients in the absence of GvHD.

No relevant conflicts of interest to declare.

Abstract 154

Abstract 4

No relevant conflicts of interest to declare.

The heterogeneous nature of myelodysplastic syndromes (MDS), age- and disease-related factors (complications and progression to acute leukemia) are associated with the complexity of quality of Life (QoL). Patients are offered mainly non-curative or experimental drugs and/or supportive care. The exploration of QoL in MDS is a prerequisite for adequate therapeutic choice.We designed an observational study in MDS patients with IPSS risk score ≤2 to evaluate determinants of QoL and its correlates. Clinical and laboratory data were collected up to 18 months and QoL instruments (QOL-E v.2, LASA scale, and EQ-5D) were completed by patients and physicians (both blind to each other's responses) at baseline, months 1, 3, 6, 12, and 18. After diagnosis, treatment was assigned based on investigators' judgment.Of 148 patients enrolled (mean age 72 years, 56% males), 115 (78%) patients were anemic at diagnosis and 38 (26%) had already received transfusions. Mean (± SD) Hb was 10.3 ± 2.1, ANC 2.3 ± 1.9×103/μL and PLT count 155 ± 118×103/μL. Charlson's Comorbidity Score was > 1 in 33 (22%). Physical and functional QoL, health, energy, activity and general states were generally poor (<65). In 94 cases ECOG PS was assigned a 0 (best) value, though 36 patients had a physical QoL-E score < 60. Surprisingly, there were significant correlations between patients' and physicians' QoL-E scores in all but those regarding the disturbances due to dependency on hospital and staff and to the inability to travel. However, MDS-specific well-being was systematically overestimated by physicians in the MDS-specific domain (p<0.0001). After 12 months, 94 patients have completed the QoL instruments. There was a transient increase in Hb with parallel changes in QoL and no significant change at 12 months (only MDS-specific QoL-E scores worsened from 79 ± 17 to 73 ± 22, p=0.005). At univariate analysis, Hb correlated with all QoL-E and LASA activity scores at baseline. After 12 months, Hb correlated with social, fatigue, MDS-specific, health status and LASA scales, while transfusion dependence with MDS-specific QOL-E scores (p<0.01). At multivariate analysis (Table), major determinants of QoL at baseline were comorbidities, IPSS, and Hb; at 12 months QoL was independently predicted by age, comorbidities, transfusion-dependence and mainly by Hb, while thrombocytopenia (PLT < 50000 microliters) and neutropenia did not seem to have a relevant impact.QoL, together with prognostic scores, should guide therapeutic choice in MDS. Avid research in MDS tends to focus on younger and fit patients to offer more aggressive or novel therapies with the aim of increasing survival. However, patients of advanced age with comorbidities complicated with severe anemia and transfusion-dependence represent a fragile category with particularly poor QoL that actually require special therapeutic attention to favor transfusion-independence and possibly an hemoglobin response.No relevant conflicts of interest to declare.

Abstract 3822

INTRODUCTION: Patients with myelodysplastic syndromes (MDS) are at continuous risk of complications and must face the reality of a relatively low survival while still mainly depending on supportive care. The chronic nature of MDS, the risk of progressive evolution, advanced age, anemia and the rapid changes in physical status make the MDS patient unique within the cancer population. Quality of Life (QoL) in MDS is undoubtedly compromised by functional, psychological, and disease-specific components. Treatment options are still scarce in this elderly population and hematologists must select candidates for experimental therapy. Patients' preferences are requisites for therapeutic choice so that it is essential that clinicians understand patients' perceptions.

INTRODUCTION: Patients with myelodysplastic syndromes (MDS) are at continuous risk of complications and must face the reality of a relatively low survival while still mainly depending on supportive care. The chronic nature of MDS, the risk of progressive evolution, advanced age, anemia and the rapid changes in physical status make the MDS patient unique within the cancer population. Quality of Life (QoL) in MDS is undoubtedly compromised by functional, psychological, and disease-specific components. Treatment options are still scarce in this elderly population and hematologists must select candidates for experimental therapy. Patients’ preferences are requisites for therapeutic choice so that it is essential that clinicians understand patients’ perceptions. STUDY DESIGN: We designed a multicenter Italian 18-month prospective observational study in 150 MDS patients from diagnosis to evaluate the correlation between QoL changes and changes in Hb and other MDS-related variables. Comparison between the patients’ and the physicians’ perception of QoL, QoL changes related to adaptation, progression and treatment response, and feasibility, consistency and validity of the QoL instruments were evaluated. Selection criteria included adult age, MDS at diagnosis with IPSS score < 2, at least 1 cytopenia, ability to complete the QoL questionnaires and ECOG PS < 3. METHODS: The QoL instruments (QOL-E v.2, LASA scale, and EQ-5D) were completed by patients and physicians (both blind to each other’s responses) at baseline, months 1, 3, 6, 12, and 18. Clinical and laboratory data were collected throughout the study. RESULTS: Median age was 74 (IQ range 67–79); 85 were males (57%). Patients were classified according to WHO: 48 RA (32%), 30 RCMD (20%), 21 RAEB-1 (14%), 7 RAEB-2 (5%), 6 5q-syndrome (4%), 4 RARS (3%), 3 CMML (2%), 1 RCMD/RS (1%), 30 MDS-U (20%). Fifty-six patients (37%) had IPSS low risk, 62 (41%) Int-1 and 14 (9%) Int-2. In 18 (12%) patients IPSS was not evaluable due to karyotype. Thirty-nine patients (26%) were transfusion-dependent (TD) and 28 (72%) had at least 1 additional cytopenia. In the transfusion-free (TF) patients, median Hb was 10.5 g/dL (IQ range 9.2–12.6); 77 patients (69%) were anemic (Hb < 12.0 g/dL) and 48 (43%) had at least 1 additional cytopenia. ECOG PS was significantly better in TF patients (p=0.0061). Charlson’s comorbidity index was 0 in 78 patients (52%), 1 in 38 patients (25%), 2 in 20 patients (13%) and ≥ 3 in 14 patients (9%). Median serum epo was 50.4 U/L (N=94). Of the 29 QOL-E items, 25 items had a response rate > 90%. Though 26 of the 29 QOL-E items showed significant correlations between patients’ and physicians’ scores, there were significant score differences in 9 (31%) of the items with a general trend of physicians to overestimate patients’ problems. Several QoL scores correlated with Hb levels and were significantly lower in TD patients than in TF patients (Table). Table. Patient QoL scores in transfusion-dependent (TD) and transfusion-free (TF) patients and correlations with Hb values. QOL domain Whole group Correlation with Hb r (p) TD patients (n = 39) TF patients (n = 111) p QOL-E, scores Q1 33 (33–67) 0.186 (0.023) 33 (0–33) 33 (33–67) 0.036 Q2 67 (67–100) −0.032 (0.70) 67 (33–100) 67 (67–67) 0.11 Physical 63 (38–75) 0.286 (0.0004) 50 (38–63) 63 (50–75) 0.0072 Social 50 (25–63) 0.197 (0.017) 44 (17–63) 50 (38–63) 0.064 Sexual 100 (50–100) 0.250 (0.011) 50 (0–100) 100 (75–100) 0.0011 Fatigue 81 (71–86) 0.236 (0.0038) 78 (67–86) 81 (71–90) 0.14 MDS-specific 81 (71–90) 0.287 (0.0005) 74 (63–83) 83 (74–93) 0.0005 LASA, scores Energy 48 ± 25 0.285 (0.0004) 42 ± 25 50 ± 25 0.081 Activity 54 ± 27 0.226 (0.0055) 50 ± 29 56 ± 26 0.24 General 53 ± 27 0.304 (0.0002) 44 ± 25 57 ± 27 0.0093 EQ-5D, scores Mobility 50 (50–100) 0.119 (0.15) 50 (50–100) 75 (50–100) 0.45 Self-Care 100 (100–100) 0.148 (0.074) 100 (100–100) 100 (100–100) 0.14 Usual Activity 100 (50–100) 0.129 (0.12) 50 (50–100) 100 (50–100) 0.0050 Pain/Discomfort 50 (50–100) −0.027 (0.75) 100 (50–100) 50 (50–100) 0.48 Anxiety/Depression 50 (50–100) 0.010 (0.90) 50 (50–100) 50 (50–100) 0.88 VAS Health State 60 ± 20 0.212 (0.010) 52 ± 19 63 ± 19 0.0020 Data are reported as mean ± standard deviation or median and interquartile range as appropriate. CONCLUSIONS: QoL is compromised in MDS and is associated with Hb values and transfusion-dependence. Patient preferences must be evaluated when considering treatment options since hematologists seem to overestimate patients’ perception of healthrelated problems. The disease-specific questionnaire, QOL-E, and the LASA scale are valid tools for the evaluation of QoL in MDS patients.

As Chronic Lymphocytic Leukemia (CLL) is associated with several defects in the T cell compartment, the impact of tumour burden on the autologous immune system was studied. Gene expression profiles (using Applied Biosystem Human Genome Microarray) identified 237 genes with significantly increased expression and 221 genes with significantly decreased expression (p<0.05) in CD3+ cells from CLL patients compared with healthy donors. Panther software analysis identified 34/237 upregulated genes and 26/221 downregulated genes that were involved in specific pathways, mainly cell differentiation and proliferation, survival, apoptosis, cytoskeleton formation, vesicle trafficking and T cell activation. The 26 dowregulated genes included Zap70, a member of the syk family protein tyrosine kinase, which is involved in T-cell activation. Zap-70 results were validated by mRNA quantification by RT-PCR (−1.77 fold in comparison with healthy controls) and by flow-cytometric analysis (Mean Intensity Fluorescence=33±12 vs 80±23.62 in controls, p<0.05). To test the hypothesis that activation with OKT3 /IL-2 could bypass these T cell deficiencies, activated T cells from 20 patients with CLL were tested in vitro for cytotoxicity (using the 51chromium release assay) against mutated and unmutated (according to IgVH mutational status) autologous B cells, DAUDI, K562 and P815 cell lines. After 10 days’ culture, the T cell count remained unchanged; CD8 cells expanded more than CD4; TCR spectratyping analysis indicated no differences in TCR repertoires. Activation restored the ZAP-70 mRNA (+1.67 fold). The 51chromium release cytotoxicity assay showed an index > 30% in 5/20 patients. The other 15 were partially cytotoxic against P815, K562 and Daudi. Cell line analysis in all 20 confirmed prevalently T cell-mediated cytotoxicity and poor NK/LAK activity. Cytotoxicity did not correlate with B cell mutational status. We tested the cytotoxic activity of autologous activated T cells in NOD/SCID mice co-transplanted with leukaemic B cells. Only activated T cells exerting cytotoxicity vs autologous B-cell CLL prevent CLL in human-mouse chimera, as confirmed by PCR and FACS analysis which visualised only CD3+ cells. In conclusion, in patients with CLL, activating autologous T cells with OKT3 /IL-2 bypasses, at least in part, the T cell immunological deficiencies. These in vitro and in vivo findings might serve to throw light on new mechanisms that could be exploited in immunotherapy designed to exert disease control.

As Chronic Lymphocytic Leukemia (CLL) is associated with several defects in the T cell compartment, the impact of tumour burden on the autologous immune system was studied. Gene expression profiles (using Applied Biosystem Human Genome Microarray) identified 237 genes with significantly increased expression and 221 genes with significantly decreased expression (p<0.05) in CD3+ cells from CLL patients compared with healthy donors. Panther software analysis identified 34/237 upregulated genes and 26/221 downregulated genes that were involved in specific pathways, mainly cell differentiation and proliferation, survival, apoptosis, cytoskeleton formation, vesicle trafficking and T cell activation. The 26 dowregulated genes included Zap70, a member of the syk family protein tyrosine kinase, which is involved in T-cell activation. Zap-70 results were validated by mRNA quantification by RT-PCR (−1.77 fold in comparison with healthy controls) and by flow-cytometric analysis (Mean Intensity Fluorescence=33±12 vs 80±23.62 in controls, p<0.05). To test the hypothesis that activation with OKT3 /IL-2 could bypass these T cell deficiencies, activated T cells from 20 patients with CLL were tested in vitro for cytotoxicity (using the 51chromium release assay) against mutated and unmutated (according to IgVH mutational status) autologous B cells, DAUDI, K562 and P815 cell lines. After 10 days' culture, the T cell count remained unchanged; CD8 cells expanded more than CD4; TCR spectratyping analysis indicated no differences in TCR repertoires. Activation restored the ZAP-70 mRNA (+1.67 fold). The 51chromium release cytotoxicity assay showed an index > 30% in 5/20 patients. The other 15 were partially cytotoxic against P815, K562 and Daudi. Cell line analysis in all 20 confirmed prevalently T cell-mediated cytotoxicity and poor NK/LAK activity. Cytotoxicity did not correlate with B cell mutational status. We tested the cytotoxic activity of autologous activated T cells in NOD/SCID mice co-transplanted with leukaemic B cells. Only activated T cells exerting cytotoxicity vs autologous B-cell CLL prevent CLL in human-mouse chimera, as confirmed by PCR and FACS analysis which visualised only CD3+cells. In conclusion, in patients with CLL, activating autologous T cells with OKT3 /IL-2 bypasses, at least in part, the T cell immunological deficiencies. These in vitro and in vivo findings might serve to throw light on new mechanisms that could be exploited in immunotherapy designed to exert disease control.

Despite advances in chemotherapy for acute leukemia, survival is poor when patients have unfavourable prognostic features at diagnosis, when they do not achieve CR after the first induction cycle and when they are in second or later remission. In these circumstances an allogeneic HSCT is preferred. The chance of finding a matched unrelated donor depends on the HLA diversity and although molecular analysis achieves closer matches it reduces the probability of finding a donor. Furthermore, many patients relapse while waiting for transplant. Transplantation of HSCs from a one-haplotype mismatched family member offers an immediate source of HSCs to almost all leukemia patients who urgently need an allogeneic transplantation because of the high-risk of leukemia relapse and who do not have a matched, either related or unrelated, avaible donor. Over the past decade, our group has shown the two major obstacles to mismatched transplants, that is severe acute GVHD in T-cell-replete transplants and graft rejection in T-cell-depleted transplants, can be overcome by infusing a megadose of extensively T-cell-depleted HSCs after an immuno-myelo-ablative conditioning regimen. Since our first reports (Aversa et al. Blood 1994 and NEJM 1998), the main modifications to our original approach were: a) in October 1995, fludarabine was substituted for cyclophosphamide in our TBI-based conditioning regimen; b) peripheral blood cells were positively selected by using initially the Ceprate device and then, since January 1999, the Clinimacs instrument which ensures a 4.5 log T-cell depletion in a one-step procedure with no E-rosetting; c) in the 138 patients transplanted since January 1999 post-transplant G-CSF administration was stopped so as to improve immune recovery. The patient population included 90 AML and 48 ALL, median age 28 years (range 9–62), 40 (29%) in bad-risk CR I, 43 (31%) in second or later CR and 55 (40%) in relapse at transplant. Primary full-donor engraftment was achieved in 125/134 evaluable patients (93%); 8 patients engrafted after second transplants. Overall engraftment was achieved in 133 patients (96%). Without any post-transplant immunosuppressive prophylaxis, grade II-IV acute GvHD occurred in 7/133 evaluable patients and 5/106 developed chronic GvHD. Cumulative incidence (C.I. 95%) of non-leukemia mortality was 36% (19%–53%) and 40% (19%–66%) for patients who were respectively in CR or in relapse at transplant. 38/51 deaths were infection-related. Disease status was the major risk factor for relapse and EFS. Cumulative incidence of leukemia relapse was 27% (12%–45%) and 60% (30%–80%), p=0.006, for ALL patients in CR and relapse respectively; 17% (8%–29%) and 46% (29%–61%), p=0.0001, for AML in CR and relapse respectively. ALL and AML patients transplanted in relapse have, respectively, a 6% and 13% probability of surviving event-free. For those transplanted in remission, EFS is respectively 38% and 50% for ALL and AML patients in any CR at transplant. These results indicate the mismatched transplant should be offered to high-risk acute leukemia patients without a HLA-identical donor not as a last resort, but as a viable option in the early stages of the disease.

Despite advances in chemotherapy for acute leukemia, survival is poor when patients have unfavourable prognostic features at diagnosis, when they do not achieve CR after the first induction cycle and when they are in second or later remission. In these circumstances an allogeneic HSCT is preferred. The chance of finding a matched unrelated donor depends on the HLA diversity and although molecular analysis achieves closer matches it reduces the probability of finding a donor. Furthermore, many patients relapse while waiting for transplant. Transplantation of HSCs from a one-haplotype mismatched family member offers an immediate source of HSCs to almost all leukemia patients who urgently need an allogeneic transplantation because of the high-risk of leukemia relapse and who do not have a matched, either related or unrelated, avaible donor. Over the past decade, our group has shown the two major obstacles to mismatched transplants, that is severe acute GVHD in T-cell-replete transplants and graft rejection in T-cell-depleted transplants, can be overcome by infusing a megadose of extensively T-cell-depleted HSCs after an immuno-myelo-ablative conditioning regimen. Since our first reports (Aversa et al. Blood 1994 and NEJM 1998), the main modifications to our original approach were: a) in October 1995, fludarabine was substituted for cyclophosphamide in our TBI-based conditioning regimen; b) peripheral blood cells were positively selected by using initially the Ceprate device and then, since January 1999, the Clinimacs instrument which ensures a 4.5 log T-cell depletion in a one-step procedure with no E-rosetting; c) in the 138 patients transplanted since January 1999 post-transplant G-CSF administration was stopped so as to improve immune recovery.

Teicoplanin in combination with amikacin and ceftazidime was used as empirical therapy in treating 46 febrile episodes in 34 consecutive patients undergoing allogcnic bone marrow transplantation. All but four of these febrile episodes occurred in neutropenic patients and 50% of them proved to be bacteraemias (23/46). Cure was achieved in 90% of Gram-positive bacteraemias (18/20) and in six of them teicoplanin was the only antibiotic with activity in vitro against the infecting strain. All (3/3) Gram-negative bacteraemias were cured. Central venous catheter removal was required in five patients (three tunnel infections, one exit-site infection and one thrombophlebitis). Two failures occurred among Gram-positive bacteraemias and in one case the patient died of infection. Four instances of side effects were documented but only one was severe (hearing loss).