Search

Disease relapse is the major causes of treatment failure after allogeneic stem cell transplantation (SCT) in patients with acute myeloid leukemia (AML). As well as demonstrating significant clinical activity in AML, azacitidine (AZA) upregulates putative tumor antigens, inducing a CD8+ T cell response with the potential to augment a graft-versus-leukemia effect. We, therefore, studied the feasibility and clinical sequelae of the administration of AZA during the first year after transplantation in 51 patients with AML undergoing allogeneic SCT. Fourteen patients did not commence AZA either because of transplantation complications or withdrawal of consent. Thirty-seven patients commenced AZA at a median of 54 days (range, 40 to 194 days) after transplantation, which was well tolerated in the majority of patients. Thirty-one patients completed 3 or more cycles of AZA. Sixteen patients relapsed at a median time of 8 months after transplantation. No patient developed extensive chronic graft-versus-host disease. The induction of a post-transplantation CD8+ T cell response to 1 or more tumor-specific peptides was studied in 28 patients. Induction of a CD8+ T cell response was associated with a reduced risk of disease relapse (hazard ratio [HR], .30; 95% confidence interval [CI], .10 to .85; P = .02) and improved relapse-free survival (HR, .29; 95% CI, .10 to .83; P = .02) taking into account death as a competing risk. In conclusion, AZA is well tolerated after transplantation and appears to have the capacity to reduce the relapse risk in patients who demonstrate a CD8+ T cell response to tumor antigens. These observations require confirmation in a prospective clinical trial., Highlights • Azacitidine is well tolerated in the majority of patients with acute myeloid leukemia after allogeneic stem cell transplantation. • Administration of post-transplantation azacitidine is associated with a low risk of chronic graft-versus-host disease. • Patients who demonstrate a CD8+ T cell response to tumor antigens demonstrate a lower risk of relapse.

It is well-established that central nervous system activation affects peripheral blood mononuclear cell (PBMCs) function through the release of the catecholamines (Epi) and norepinephrine (NE), which act on ß2-adrenergic receptors (ß2AR). However, most studies have used non-specific stimulation of cells rather than antigen-specific responses. Likewise, few studies have parsed out the direct effects of ß2AR stimulation on T cells versus indirect effects via adrenergic stimulation of antigen presenting cells (APC). Here we report the effect of salmeterol (Sal), a selective ß2AR agonist, on IFN-γ+ CD4 and IFN-γ+ CD8 T cells following stimulation with Cytomegalovirus lysate (CMVL-strain AD169) or individual peptides spanning the entire region of the HCMV pp65 protein (pp65). Cells were also stimulated with Staphylococcal enterotoxin B. Additionally, we investigated the effect of Epi and Sal on cytotoxic cell killing of transfected target cells at the single cell level using the CD107a assay.The results show that Sal reduced the percentage of IFN-γ+ CD4 and IFN-γ+ CD8 T cells both when applied directly to isolated T cells, and indirectly via treatment of APC. These inhibitory effects were mediated via a ß2 adrenergic-dependent pathway and were stronger for CD8 as compared to CD4 T cells. Similarly, the results show that Sal suppressed cytotoxicity of both CD8 T and NK cells in vitro following stimulation with Chinese hamster ovary cell line transfected with MICA*009 (T-CHO) and the human erythromyeloblastoid leukemic (K562) cell line. The inhibitory effect on cytotoxicity following stimulation with T-CHO was stronger in NK cells compared with CD8 T cells.Thus, targeting the ß2AR on lymphocytes and on APC leads to inhibition of inflammatory cytokine production and target cell killing. Moreover, there is a hierarchy of responses, with CD8 T cells and NK cells inhibited more effectively than CD4 T cells.

Regulating appropriate activation of the immune response in the healthy host despite continual immune surveillance dictates that immune responses must be either self-limiting and therefore negatively regulated following their activation or prevented from developing inappropriately. In the case of antigen-specific T cells, their response is attenuated by several mechanisms, including ligation of CTLA-4 and PD-1. Through the study of the viral OX2 (vOX2) immunoregulator encoded by Kaposi's sarcoma-associated herpesvirus (KSHV), we have identified a T cell-attenuating role both for this protein and for CD200, a cellular orthologue of the viral vOX2 protein. In vitro , antigen-presenting cells (APC) expressing either native vOX2 or CD200 suppressed two functions of cognate antigen-specific T cell clones: gamma interferon (IFN-γ) production and mobilization of CD107a, a cytolytic granule component and measure of target cell killing ability. Mechanistically, vOX2 and CD200 expression on APC suppressed the phosphorylation of ERK1/2 mitogen-activated protein kinase in responding T cells. These data provide the first evidence for a role of both KSHV vOX2 and cellular CD200 in the negative regulation of antigen-specific T cell responses. They suggest that KSHV has evolved to harness the host CD200-based mechanism of attenuation of T cell responses to facilitate virus persistence and dissemination within the infected individual. Moreover, our studies define a new paradigm in immune modulation by viruses: the provision of a negative costimulatory signal to T cells by a virus-encoded orthologue of CD200.

Cancer testis antigens exhibit physiological expression within germ cells and are frequently expressed in malignant tissue. Interestingly, immunological tolerance to cancer testis proteins does not appear to be established, and the expression of CTAg proteins within malignant cells can therefore lead to induction of cellular and humoral immunity. A considerable body of evidence now indicates that CD8-specific immunity plays an important role in the control of cancer cell growth, and a number of vaccine studies are in progress to boost CTAg-specific cellular immune responses. We have previously identified CTAg-specific immune responses in patients with multiple myeloma and reported that recognition of the MAGE-A1(289-298) peptide, which is described as being restricted by HLA-B*0702, was the most frequent response seen with our peptide panel. Here, we studied seven CD8+ T-cell clones specific for this peptide which were isolated from three patients with myeloma at several time-points. The affinity of peptide recognition was high with 50% maximal interferon-γ production observed at a peptide concentration of 10(-10) M and variation of only one order of magnitude between the affinities of the clones. Importantly, all the clones were able to recognise and kill multiple myeloma cell lines. Interestingly, one patient did not express HLA-B*0702, but three clones from this patient recognised the MAGE-A1(289-298) peptide on a lymphoblastoid cell line (LCLs) expressing HLA-Cw7, and we now show evidence that the MAGE-A1(289-298) peptide is expressed and recognised through Cw7. The T-cell receptor gene usage was determined in five clones and showed conserved features in both the α and the β chain genes indicating correlation between T-cell receptor usage and peptide specificity of cancer testis antigen-specific T-cell clones.

The expression of cancer germline antigens (CGAgs) is normally restricted to the testis but is also present in many types of malignant cells including plasma cells from patients with myeloma. Because T-cell immune responses to CGAg have been identified in patients with solid tumors, this may offer a novel target for immunotherapy in patients with myeloma. We have used 12 peptide epitopes from a range of CGAgs to screen for CGAg-specific T cells in blood from patients with multiple myeloma at various stages of their disease. T cells from 15 of 37 patients responded to one or more CGAg peptides and the magnitude of the CGAg-specific CD8+ T-cell response ranged between 0.0004% and 0.1% of the total CD8+ T-cell pool. Serial analyses showed that these immune responses were detectable in individual patients at multiple time points during the course of their disease. In patients undergoing treatment or in disease relapse, the magnitude of the CGAg-specific T-cell response was positively correlated with the level of paraprotein. Functional T cells specific for CGAgs are therefore present in a proportion of patients with multiple myeloma and offer the possibility of a novel approach for immunotherapy in this disease.

Deregulation of signaling pathways is a hallmark of malignant transformation. Signaling-associated phosphoproteins can be degraded to generate cancer-specific phosphopeptides that are presented by major histocompatibility complex (MHC) class I and II molecules and recognized by T cells; however, the contribution of these phosphoprotein-specific T cells to immune surveillance is unclear. We identified 95 phosphopeptides presented on the surface of primary hematological tumors and normal tissues, including 61 that were tumor-specific. Phosphopeptides were more prevalent on more aggressive and malignant samples. CD8(+) T cell lines specific for these phosphopeptides recognized and killed both leukemia cell lines and human leukocyte antigen-matched primary leukemia cells ex vivo. Notably, healthy individuals showed robust CD8(+) T cell responses against many of these phosphopeptides within the circulating memory compartment. This immunity was significantly reduced or absent in some leukemia patients. This reduction correlated with clinical outcome; however, immunity was restored after allogeneic stem cell transplantation. These results suggest that phosphopeptides may be targets of cancer immune surveillance in humans, and point to their importance for development of vaccine-based and T cell adoptive transfer immunotherapies.

The traditional paradigm suggests that during normal pregnancy maternal immunological tolerance of the allogenic fetus is association with a maternal T-lymphocyte shift from a Th1 to a Th2 phenotype, with the opposite effect reported in patients with recurrent miscarriage. However, studies on maternal peripheral blood are conflicting. In the present study, we characterized the maternal CD4 T-cell effector subsets, including the recently described Th17 subset, during normal pregnancy (cross-sectional cohort, n=71; longitudinal cohort, n=17) and contrasted this with women with recurrent miscarriage (n=24). Longitudinal analysis of peripheral blood from normal pregnancy demonstrated a fall in the percentage of Th17 cells between the first and second trimester (P≤0.05), but no significant changes were observed across gestation or the post-natal period in Th1 or Th2 subsets. In contrast, in women with a history of recurrent miscarriage, an elevated proportion of Th17 (0.314% compared with 0.097%; P=0.0009) and Th1 (12.4% compared with 5.3%; P=0.0002) cells was detected. The suggestion that Th17 cells may have a role in the normal events of implantation and early pregnancy requires further evaluation and mechanistic studies. The results of the present study, by conducting a careful longitudinal analysis, demonstrate that a peripheral Th1/Th2 shift is not a requirement for normal pregnancy. By contrast, the profound increase in Th1 and Th17 cells in women with recurrent miscarriage indicates that peripheral immunological dysfunction may be important in this group specifically, and these assays may be important in guiding therapeutic interventions in this group and warrant further investigation to determine whether they are predictive of outcome or responses to immunomodulatory therapy.

Tolerance of the semiallogeneic fetus presents a significant challenge to the maternal immune system during human pregnancy. T cells with specificity for fetal epitopes have been detected in women with a history of previous pregnancy, but it has been thought that such fetal-specific cells were generally deleted during pregnancy as a mechanism to maintain maternal tolerance of the fetus. We used MHC-peptide dextramer multimers containing an immunodominant peptide derived from HY to identify fetal-specific T cells in women who were pregnant with a male fetus. Fetal-specific CD8+ T lymphocytes were observed in half of all pregnancies and often became detectable from the first trimester. The fetal-specific immune response increased during pregnancy and persisted in the postnatal period. Fetal-specific cells demonstrated an effector memory phenotype and were broadly functional. They retained their ability to proliferate, secrete IFN-γ, and lyse target cells following recognition of naturally processed peptide on male cells. These data show that the development of a fetal-specific adaptive cellular immune response is a normal consequence of human pregnancy and that unlike reports from some murine models, fetal-specific T cells are not deleted during human pregnancy. This has broad implications for study of the natural physiology of pregnancy and for the understanding of pregnancy-related complications.

Strategies that augment a GVL effect without increasing the risk of GVHD are required to improve the outcome after allogeneic stem cell transplantation (SCT). Azacitidine (AZA) up-regulates the expression of tumor Ags on leukemic blasts in vitro and expands the numbers of immunomodulatory T regulatory cells (Tregs) in animal models. Reasoning that AZA might selectively augment a GVL effect, we studied the immunologic sequelae of AZA administration after allogeneic SCT. Twenty-seven patients who had undergone a reduced intensity allogeneic transplantation for acute myeloid leukemia were treated with monthly courses of AZA, and CD8+ T-cell responses to candidate tumor Ags and circulating Tregs were measured. AZA after transplantation was well tolerated, and its administration was associated with a low incidence of GVHD. Administration of AZA increased the number of Tregs within the first 3 months after transplantation compared with a control population (P = .0127). AZA administration also induced a cytotoxic CD8+ T-cell response to several tumor Ags, including melanoma-associated Ag 1, B melanoma antigen 1, and Wilm tumor Ag 1. These data support the further examination of AZA after transplantation as a mechanism of augmenting a GVL effect without a concomitant increase in GVHD. The trial was registered at http://isrctn.org as #ISRCTN36825171.

BACKGROUND Human pregnancy offers an immunological challenge for the immunocompetent women accommodating an allogenic fetus, while continuing to combat potentially infectious disease. Cytomegalovirus (CMV) infects the majority of the human population and establishes lifelong persistence, which can lead to the oligoclonal expansion of differentiated T cells. Primary CMV infection and, less commonly, secondary infection during pregnancy can cause fetal disease and morbidity. The balance between maternal immune competence and viral pathogenicity is thus delicately poised. Our objective was to investigate the influence of CMV serostatus on maternal CD8+ T-cell phenotype and cytokine profile in an apparently healthy cohort of pregnant women. Furthermore, we assessed if CMV serostatus modulated changes in CD8 T cells during gestation. METHODS CD8+ T-cell phenotype was investigated in 87 pregnant women with samples obtained both during pregnancy [CMV immunoglobulin G (IgG) + n = 39, CMV IgG- n = 21] and in the early post-natal period (IgG+ n = 16, IgG- n = 11). Multiparameter flow cytometry was used to study T-cell phenotype and HLA-peptide tetramers identified CD8 T cells specific for CMV. Levels of 26 plasma cytokines, chemokines and chemokine receptors were assessed in a separate cohort of 20 women (IgG+ n = 10, IgG- n = 10) followed longitudinally during and after pregnancy. RESULTS CMV seropositivity profoundly influenced the T cell repertoire and its dynamics during pregnancy. Naive CD8+ T-cells (CCR7+CD45RA+) were reduced by 50% in CMV-seropositive women. The proportion of CD45RA effector cells was not increased in CMV-seropositive donors, although this population was more highly differentiated with reduced CD27 and CD28. However, there was a doubling in the proportion of CD45RA+ revertant memory cells (CCR7-CD45RA+) in seropositive donors. Moreover, seropositive women during late pregnancy demonstrated an accumulation of highly differentiated CMV-specific T-cells. T-cell activation independent of CMV was also seen in late pregnancy. No CMV-related changes in plasma cytokines, chemokines or their receptors were observed. CONCLUSIONS Thus, CMV serostatus is a crucial consideration in studies of T cell memory and differentiation during pregnancy. The reduction in maternal naive T cells in CMV-seropositive donors could have implications for the maternal response to infections during pregnancy. These findings shed light on the delicate balance between host, fetus and chronic infection during healthy pregnancy and will inform studies in relation to the importance of CMV on maternal and fetal health.

Epigenetic therapies, including DNA methyltransferase and histone deacetylase inhibitors, represent important new treatment modalities in hematologic malignancies, but their mechanism of action remains unknown. We reasoned that up-regulation of epigenetically silenced tumor antigens may induce an immunologically mediated antitumor response and contribute to their clinical activity. In this study, we demonstrate that azacitidine (AZA) and sodium valproate (VPA) up-regulate expression of melanoma-associated antigens (MAGE antigens) on acute myeloid leukemia (AML) and myeloma cell lines. In separate studies, we observed that prior exposure to AZA/VPA increased recognition of myeloma cell lines by a MAGE-specific CD8+ cytotoxic T-lymphocyte (CTL) clone. We therefore measured CTL responses to MAGE antigens in 21 patients with AML or myelodysplasia treated with AZA/VPA. CTL responses to MAGE antigens were documented in only 1 patient before therapy; however, treatment with AZA/VPA induced a CTL response in 10 patients. Eight of the 11 patients with circulating MAGE CTLs achieved a major clinical response after AZA/VPA therapy. This is the first demonstration of a MAGE-specific CTL response in AML. Furthermore, it appears that epigenetic therapies have the capacity to induce a CTL response to MAGE antigens in vivo that may contribute to their clinical activity in AML. This study was registered at http://isrctn.org as #ISCTN68418952.

Background Allogeneic stem cell transplantation is associated with a powerful ‘graft-versus-leukemia’ effect that is generally considered to result from an alloreactive T-cell immune response. However, disease remission can also be observed after syngeneic transplantation and we investigated whether a T-cell immune response to cancer-testis antigens can be detected in patients in the post-transplant period.Design and Methods The T-cell immune response against cancer-testis antigens was studied in a cohort of 41 patients who underwent allogeneic stem cell transplantation for the management of acute myeloid leukemia or multiple myeloma. The cytokine secretion assay was combined with magnetic selection to allow detection of an interferon-γ-secreting T-cell response to a panel of cancer-testis antigen peptides.Results A cancer-testis antigen-specific CD8+ T-cell immune response was observed in the peripheral blood of five patients with an average magnitude of 0.045% of the CD8+ T-cell repertoire. Four of these patients had undergone reduced intensity conditioning transplantation with alemtuzumab for the treatment of acute myeloid leukemia and three remain in long-term remission. T-cell immunity was focused against peptides derived from MAGE proteins and was markedly increased within the bone marrow.Conclusions Functional cancer-testis antigen-specific CD8+ T-cell immune responses develop in the early period following reduced intensity allogeneic stem cell transplantation and are preferentially localized to bone marrow. These immune responses are likely to contribute to the cellular basis of the graft-versus-leukemia effect.

The factors that determine progression from monoclonal gammopathy of undetermined significance (MGUS) to multiple myeloma are unclear but may include the breakdown of immune surveillance. Cancer testis antigens (CTAgs) are expressed by the majority of myelomas and MGUS tumors and are a potential immune target. We have characterized CD4+ and CD8+ T-cell immune responses to MAGE-A1/A2/A3 in these patients. CD4+ T-cell immunity to MAGE proteins is stronger and more frequent in MGUS compared with myeloma with a predominantly CD45RA−CCR7− effector memory profile and cytotoxicity against MAGE-positive cell lines. In contrast CD8+ T-cell immune responses were present almost exclusively in patients with multiple myeloma, correlating with disease, with a CD45RA+CCR7− memory phenotype, localizing poorly to the bone marrow but were able to lyse myeloma cell lines in vitro. This suggests that the CD4+ CTAg-specific immune response may play a role in controlling tumor growth, whereas the efficacy of the CD8+ T-cell response appears to be limited in vivo. Despite this, patients with evidence of a CTAg-specific immune response had a 53% reduction in mortality over a median follow-up of 4 years. These findings have important implications for clinical approaches to CTAg-specific immunotherapy in patients with cancer.

Multiple myeloma is a malignant tumour of plasma cells that remains incurable for the vast majority of patients, with a median survival of 2-3 years. It is characterized by the patchy accumulation of tumour cells within bone marrow leading to variable anaemia, bone destruction, hypercalcaemia, renal failure and infections. Immune dysfunction is an important feature of the disease and leads to infections that are both a major cause of morbidity and mortality and may promote tumour growth and resistance to chemotherapy. Numerous defects of the immune system have been described in multiple myeloma although the relative clinical importance of these remains elusive. There has been considerable interest in the identification of an autologous response against myeloma. Although T cells and humoral responses directed against myeloma-associated antigens have been described, it is uncertain if the immune system plays a role in preventing or controlling myeloma cell growth. There is increasing interest in the potential role of immunotherapy but the success of these interventions is likely to be modified by the immunologically hostile environment associated with multiple myeloma. This review attempts to summarize the current knowledge relating to the immune defects found in multiple myeloma.

During pregnancy there is the transplacental ‘traffic’ of allogenic fetal cells into the maternal circulation and this could provoke a maternal cellular immune response. Transgenic murine models suggest during pregnancy cytotoxic T cells (CTL) specific for fetal antigen are deleted or rendered anergic. We investigated the maternal cellular immune response to fetal (HY) antigens during human pregnancy. Pregnant women without medical complications (n=90) were studied longitudinally prenatally and within the puerperium at 6 timepoints and non-pregnant multiparous controls (n=65) investigated. We used human leucocyte antigen-peptide (HLA) ‘dextramers’ in mothers who were HLA-A*0201 or HLA-B*0702 to detect fetal specific T cells; directly ‘ex-vivo’ with phenotyping of fetal-specific cells by 9 colour flow cytometry and following expansion in culture. Fetal specific CTL were cloned and functional assessment performed. Maternal CTL responses against fetal antigen were detected in 51% of multiparous women (14/27), even decades after pregnancy. Such responses were also detected prenatally in 54% (18/33) of women. T cell clones specific for fetal cells were isolated from maternal blood during pregnancy and the puerperium (n=16). Fetal-specific clones are cytotoxic (>50% killing at a 1:1 ratio) and release IFN-α (mean 1600 pg/ml IFN-α) in response to fetal antigen, with no difference in function between clones isolated prenatally or after pregnancy. A maternal cellular immune responses to fetal antigens can commonly be detected during and after human pregnancy. Functional investigations demonstrated that they retain cytotoxicity and their ability to produce cytokines. This novel finding has important implications for understanding the immunobiology of pregnancy, spontaneous abortion and transplantation.

Harnessing the power of adaptive immunity to combat cancer has been a long-term goal of translational immunotherapy. Tumor-specific immunity, where present, is typically at low frequency and affinity with compromised effector function. By contrast, immunity against persistent herpesviruses in man is characterised by high affinity cytotoxic T-lymphocytes (CTL) at high frequency with potent effector function. Furthermore, the immunosubversive mechanisms employed by herpesviruses show striking parallels to tumors, yet the associated anti-viral immunity limits these to life-long asymptomatic infections. We reasoned that the delivery of immunodominant viral peptide epitopes to the tumor surface might facilitate passive-loading of peptides into empty MHC class-I molecules, effectively mimicking viral infection, rendering tumors susceptible to lysis by anti-viral immunity. To address this we developed a new class of targeting antibodies: APEC (Antibody Peptide Epitope Complexes) that are able to deliver an antigenic payload at the cell surface through proteolytic release of covalently-coupled peptide antigens. As a proof-of-concept we used clinically-validated antibodies cetuximab (anti-EGFR) and rituximab (anti-CD20) to develop APECs that are able to target human tumors. We screened 15 HLA-A*0201+ EGFR-expressing NCI-60 cell lines, CD20+ lymphoma cell lines, 20 primary CD20+ CLL tumor samples and four healthy B-cells against a library of 190 cetuximab-APECs (cAPEC) or rituximab (rAPEC) incorporating the immunodominant cytomegalovirus (CMV) pp65495-503 epitope and candidate protease cleavage sites following co-incubation with CMV-specific CTL (CMV-CTL). The most effective cAPEC and rAPEC were those incorporating MMP2, MMP9, Cathepsin B and Cathepsin D protease recognition domains. Very few (2/190) rAPEC were able to redirect CMV-CTL against healthy cells. Heterogeneity was observed for primary CLL tumors but a limited number of rAPEC were effective in all cases (5/190). Mechanistic studies demonstrated that: (i) peptide loading occurred at the cell surface, (ii) required the expression of target antigens at the cell surface and (iii) T-cell recognition could be inhibited by unconjugated antibody (92%) or by incubation with protease inhibitors (83%). T-cell specificity was examined using rAPEC treated tumor targets co-incubated with various HLA-matched effector T-cell populations. No activation of CD4+ was observed including CD4+CD25hi regulatory T-cell populations. Incubation with CD8+ T-cells revealed that only pp65495-503-specific CD8+ T-cells engaged with APEC-treated tumor cells. Lastly, xenograft studies using EGFR+ and CD20+ tumor cell lines demonstrated efficacy of both cAPEC and rAPEC to eliminate tumors in vivo by redirecting anti-viral CTL. These data indicate that APECs represent a powerful new approach to combat cancer. Citation Format: David G. Millar, Laura Morton, Manuela Carvalho Gaspar, Punam Mistry, Hugo De La Peña, Ricky Joseph, Sarah Penny, Oliver C. Goodyear, Margaret Goodall, Guy E. Pratt, Mark Cobbold. Engineered release and presentation of antibody-bound viral antigens: A highly specific and novel immunotherapeutic approach to target cancer in vivo. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2893. doi:10.1158/1538-7445.AM2014-2893

Abstract 324 Disease relapse and graft-versus-host disease (GVHD) remain the commonest causes of treatment failure after allogeneic stem cell transplantation (SCT) for Acute Myeloid Leukemia (AML). Azacitidine (AZA) possesses inherent anti-leukemic activity but also expands immunomodulatory T regulatory cells in animal models and up-regulates the expression of tumor antigens in vitro. Reasoning that AZA might selectively augment a graft-versus-leukemia (GVL) without a concomitant increase in GVHD we have studied the tolerability and immunological sequelae of AZA administration in patients who have undergone a reduced intensity allogeneic SCT for AML. All patients were transplanted using a conditioning regimen consisting of fludarabine (30 mg/m2 IV × 5 days), melphalan (140 mg/m2IV), and alemtuzumab (10 mg IV × 5 days) with cyclosporine GVHD prophylaxis. Patients received AZA (36 mg/m2) × 5 days every 28 days commencing after sustained neutrophil and platelet engraftment at day +42 with the aim of administering monthly cycles of AZA until 12 months post-transplant. Numbers of CD4+CD25+CD127loFoxP3+ T regulatory cells were measured by flow cytometry. Tumor specific cytotoxic T lymphocytes (CTL) recognising members of the cancer testis antigen (CTA) family and WT1 were quantitated using a CD137 expression and enrichment assay. We report results on 27 patients (median age 59 years) who commenced treatment with AZA (follow-up 3–21 months). 11 patients were transplanted from an HLA identical related donor and 16 from a volunteer unrelated donor. Disease status at the time of transplant was: CR1 n=18; CR2 n=7; first relapse n=2. Post-transplant AZA was well tolerated and 24 patients tolerated at least three cycles of AZA. Haematological toxicity was modest with only two patients experiencing treatment delay for neutropenia, or thrombocytopenia. Three patients developed Grade 2 acute GVHD and no patient developed >Grade 2 acute GVHD. Two patients developed limited chronic GVHD. To date seven patients have relapsed at a median time of 6 months (4–15 months) post-transplant. Administration of AZA had no impact on absolute lymphocyte counts or CD8+ and CD4+ cell numbers compared with a control population of 17 patients transplanted using an identical conditioning regimen. However AZA administration was noted to increase the number of CD4+CD25+CD127loFoxP3+ regulatory T cells within the first 3 months post transplant compared with a time-matched control population (p=0.017) although there was no difference detectable at 6 or 12 months. AZA administration also induced a cytotoxic CD8+ T cell response to candidate tumor antigens MAGE, GAGE, RAGE and WT-1 in the peripheral blood. A CD8+ T cell response to these candidate tumor antigens was only detectable in 1/22 patients pre-transplant but circulating CTA- or WT1 specific CD8+ T cells were detected in 14/16 patients who had received at least six cycles of AZA with a frequency between 0.001–1.4% (mean 0.25%). In four patients with paired peripheral blood and bone marrow samples the size of the CD8+ T cell response was noted to be up to 100 fold greater in the bone marrow. Ex vivo characterisation of the CTA response using dextramers demonstrated the CD8+ T cell response to be effector memory (CD45RA-CCR7-) and functional assays confirmed by mobilization of CD107a and secretion of interferon-g, TNF-a and IL-2 in response to peptide. These data confirm the tolerability of adjunctive AZA post-transplant and its administration appears to be associated with a low incidence of both acute and chronic GVHD. The demonstration that AZA induces both early expansion of T regulatory cells and a tumor specific CD8+ CTL response highlight its potential use as a strategy to epigenetically manipulate a graft-versus-leukemia reaction after allogeneic SCT. Disclosures: Dennis: Celgene: Honoraria, Research Funding. Craddock:Celgene: Honoraria, Research Funding.

Abstract 5076 Multiple Myeloma (MM) is a malignant disease of bone marrow plasma cells characterised by immune dysfunction and increased chemokine levels. We investigated chemokine receptor expression on T cells, particularly the Th1 and Th2-associated receptors, CXCR3 and CCR4, in patients with paraproteinaemia and compared this to controls and correlated this to the level of related chemokines in blood and bone marrow. The frequency of CCR4+ T cells within the bone marrow was markedly increased in patients with MM and the normal distribution between blood and marrow was reversed. CXCR3+ Th1 Cells which were rare within normal marrow were increased relative to blood. Th17 CD4+ T cells were markedly increased within marrow with increased Th17 CD4+ CCR4+ and Th17 CD4+ CXCR3 populations and the suppressor CD25highCd127low FoxP3positive subset reduced within blood. The expression of the ligands for CCR4 and CXCR3 was markedly increased within the marrow of patients with MGUS and MM. Mesenchymal stem cells isolated from the bone marrow of myeloma patients supported T cell adhesion and migration with a preferential recruitment of CCR4 and CXCR3-positive cells. Chemokine gradients within the tumour microenvironment alter the distribution of major T cells subsets between blood and marrow. This pattern is established early in the development of paraproteinaemia and may offer a potential target for new forms of treatment. Disclosures: No relevant conflicts of interest to declare.

During human pregnancy an allogeneic fetus is accommodated within an immunocompetent host. Human cytomegalovirus (CMV) infection during pregnancy can cause significant fetal disease. The balance between maternal immune competence, fetal tolerance and viral pathogenicity is thus delicately poised. We studied the CD8+ T cell (CTL) repertoire and CMV-specific T cell dynamics during healthy pregnancy and related this to CMV serostatus and gestation. Six human leucocyte antigen-peptide tetramers identified CMV-specific CTL. Phenotyping of CTL was carried out by 10 colour flow cytometry at a range of gestations (n=105). Serostatus was determined by CMV IgG ELISA. Maternal plasma levels of 27 cytokines were quantified by luminex assay. There was a significant reduction in naive T cells in CMV seropositive women (21% vs 48%, p≤0.001) whereas the memory pool was markedly expanded with significantly more highly differentiated (CD45RA+CCR7-) CTL (34% vs 16%, p≤0.001). Seropositive women in the third trimester showed a significant increase in highly differentiated CTL (CD45RA+CCR7-) compared to the second trimester, which was not observed in seronegative women and was due to CMV specific CTL. CMV seropositivity had a marked impact on maternal CTL phenotype, with reduced naive T cells and a more differentiated T cell phenotype. CMV seropositivity also altered the dynamics of the maternal immune response during pregnancy. Thus, CMV serostatus is a crucial confounder for studies of CTL during pregnancy. Also, reduced naive CTL may impair responses to other infections. The differentiated CTL repertoire and altered dynamics of the immunological response are likely to have a clinical impact on pregnancy.

Progesterone (P4) has immunomodulatory effects during pregnancy, including shifting CD4 T cell clones from a Th1 toTh2 profile. Effects on CD8+ T cell (CTL) clones have not been described. Maternal CTL specific for fetal (HY) antigen are present during human pregnancy. Mechanisms of tolerance, including the action of progesterone, may regulate their action in-vivo. We studied the effect of P4 on the function of fetal-specific CTL. Maternal CTL specific for fetal antigen were cloned during human pregnancy. Clones were stimulated by antigen expressing targets in media containing natural P4 at a range of concentrations. The production of interferon-α (IFN-α) was measured by ELISA. All experiments were conducted in triplicate. Results show IFN-α release by fetal-specific T cell clones was significantly reduced (median 36% reduction, p= IFN-α release by fetal-specific CTL clones is reduced by concentrations of progesterone similar to those found in the human decidua. Work is ongoing to isolate the cellular mechanisms of this effect and determine why fetal-specific clones respond differently. Understanding the mechanism of action of progesterone on fetal-specific CTL is important as progesterone is used therapeutically and could be manipulating fetal-specific T cell responses.

Abstract 2086 Poster Board II-63 The mechanism by which epigenetic therapies exert an anti-tumor effect in Acute Myeloid Leukemia (AML) and myeloma is unknown. We have studied immune responses to cancer testis antigens (CTA) a family of immunodominant tumor-associated antigens, in 12 patients with AML treated with the DNA methyltransferase inhibitor 5-Azacitidine (AZA) and the histone deacetylase inhibitor sodium valproate (VPA). The frequency of CTA-specific CD8+ T cells was studied in 12 patients with high risk AML prior to treatment with AZA (75 mg/m2 × 7 days per 28 day cycle) and sodium valproate (1-2.5 g daily administered continuously as tolerated) on a Phase II clinical trial. CTA specific CTLs were detectable in 4 of 12 patients with relapsed AML undergoing treatment with AZA and VPA. In one patient CTLs specific to the peptide MAGE-A2157-166 were detectable prior to commencement of trial therapy at a frequency of 0.001%.and persisted throughout treatment. In three patients CTLs were only detectable after commencement of AZA therapy. A polyclonal T cell line was subsequently generated from one patient and tetramer staining of the line revealed the presence of 1% MAGE-A2157-166-specific T cells and this could be enriched using anti-PE magnetic beads (Miltenyi Biotec). Two patients in whom AZA induced a CTL response to MAGE achieved a partial response to trial therapy by Cheson criteria. Treatment with AZA resulted in up to 100fold increased levels of expression of MAGEA4, MAGEA1, HAGE, SSX1, NY ESO1, MAGEC1, LAGE and XAGE mRNA in AML cell lines: HL60, KG1a, U937 and NB4 and multiple myeloma cell lines: U266 and JJN3. We next studied whether prior treatment with AZA augmented recognition of hematopoietic tumors by MAGE specific T cell clones. Prior treatment with AZA increased recognition of two myeloma cell lines (JJN3 and U266) by a MAGEA1 specific T cell clone. This is the first demonstration that AZA induce a CTL response to MAGE in patients with AML. Taken together our data are consistent with the possibility that epigenetic manipulation of cancer testis antigen expression and induction of a tumor specific CTL response may contribute to the clinical activity of AZA in hematologic malignancies. Disclosures: No relevant conflicts of interest to declare.

An immunologically mediated graft-versus-tumor (GVT) underlies the curative effect of reduced intensity allografts in acute myeloid leukaemia (AML) and other haematologic malignancies. Cancer testis antigens (CTA) represent a family of immunodominant proteins that are variably expressed in haematological malignancies and represent a potential target of a GVT response. Importantly a number of CTA genes demonstrate promoter hypermethylation in solid tumours which can be reversed using demethylating agents such as azacitidine. We have therefore examined whether donor T cell responses to CTAs are observed in patients allografted for AML and multiple myeloma (MM) and whether epigenetic therapies can be used to manipulate T cell mediating killing of haemopoietic targets. We screened 37 patients with AML and 8 patients with MM, who had not received demethylating agent treatment, for T cell responses to 25 peptides derived from 10 CTA genes, including BAGE, LAGE, MAGE-A1, MAGE-A2, MAGE-A3, MAGE-A4, MAGE-C2, RAGE-1, by interferon-γ cytokine secretion assay (IFN-γ CSA). We found CTA specific T cells in 11.1% of patients (5 of 45) with frequencies ranging from 0.0005% to 0.2% (median 0.024%). Subsequently, we studied expression of 15 cancer-testis antigens (CTA) at the RNA and protein levels in AML, MM and Hodgkins’ lymphoma cell lines before and after exposure to demethylating agent 5-aza-2′-deoxycytidine (Aza) and histone deacetylase inhibitor sodium valproate (VPA), as single agents and in combination. We found that expression of CTAs HAGE, SACA3, SPANXB, LAGE, XAGE, MAGEA3 could be induced in a dose dependent manner by Aza alone but not with VPA alone. We also observed that expression induced by Aza treatment was increased further by combination treatment with VPA. Induction of CTAs was confirmed in vitro in primary AML cells and in vivo in 2/6 patients on VPA/Aza trial. Furthermore, using interferon-γ ELISA assay we observed that Aza-induced expression of the CTA MAGEA3 in MM cell lines was accompanied by increased recognition by MAGEA3-specific T cells. These studies confirm the importance of members of the CTA family as targets of a T cell mediated immune response. Our data demonstrate that the expression of these putative immunodominant antigens in haematological malignancies in disease such as AML and myeloma, can be significantly up-regulated by epigenetic therapies with functional increases in target cell killing and support the use of adjunctive epigenetic therapies after allogeneic transplantation with the aim of augmenting a GVL response. A Phase II clinical trial combining post transplant azacitidine with a reduced intensity allograft is currently ongoing in patients with AML.

The expression of Cancer Germline Antigens (CGAgs) is normally restricted to the pre-meiotic spermatogonia cells of the testis. The testis is an immunologically privileged site and so immunological tolerance to CGAg is not established. However, CGAg expression is also detected in many types of malignant disease including plasma cells from patients with multiple myeloma. CGAg expression has been shown to prime a T cell immune response in many patients with solid tumours and this may offer a novel target for immunotherapy in patients with myeloma. We have used immunodominant peptide epitopes from a range of CGAgs to screen for CGAg-specific T cells in the blood of patients with multiple myeloma at various stages of their disease. Initial studies demonstrated that T cells from 15 out of 37 patients responded to one or more CGAg peptides and the magnitude of the CGAg-specific CD8+ T cell response ranged between 0.0004% and 0.1% of the total CD8+ T cell pool. Serial analysis showed that these immune responses were detectable in individual patients at multiple time-points during the course of their disease. A further 13 peptides have now been obtained including several CD4 peptide. We have subsequently cloned CD4 T cells specific to a MAGE 3 peptide and have shown them to be functional. In some patients we determined the membrane phenotype of the CGAg-reactive cells as CD45RA+ and CCR7−, an effector memory differentiation state. CGAg-specific responses have also been detected in patients with clinically benign forms of paraproteinaemia indicating that T cell immunity may play a role in the control of disease progression. Plasma cells are localised to bone marrow and we are now focussing on the study of immunity to CGAg at this site. Initial findings indicate a higher proportion of CGAg-specific T cells within bone marrow and the phenotypic profile of these cells is being determined. Functional T cells specific for CGAg are therefore present in a large proportion of patients with multiple myeloma and offer the possibility of a novel approach for immunotherapy in this disease.

Proteins from the family known as ‘cancer-testis antigens’ (CTAg) are expressed in some cases of multiple myeloma and subsets of acute myeloid leukaemia. CTAg can stimulate CD8+ T cell responses in patients with melanoma but there are no reports of CTAg-specific immune response in patients with haematological malignancy. Such information is critical to assess whether or not these antigens act as targets for tumour-specific immunity or if they could be used as targets for immunotherapy. We have used twelve peptide epitopes from a range of cancer-testis antigens which have been previously defined as epitopes for CD8+ T cells. These were used to screen for tumour-specific T-cells in blood of patients with multiple myeloma at various stages of their disease. The IFNγ cytokine secretion assay was used to detect functional responses and magnetic selection was employed to increase the sensitivity of detection. FACS analysis was used to quantitate the frequency of responding cells. 37 patients were screened with an age range of between 45 and 88 years. Blood samples were taken at monthly intervals and the percentage of CD8+ T cells responding to each peptide was calculated. 13 patients responded to 1 or more of the peptides with a range between 0.01% and 0.7% of the total CD8+ T cell pool. The frequency of the tumour-specific response fluctuated during treatment in individual patients. Analysis of the CTAg-specific immune response in relation to disease course revealed that the immune response was generally correlated with tumour burden as revealed by the paraprotein level. CTAg HLA-peptide tetramers incorporating peptides from LAGE-1 and MAGE-2 were able to directly visualize CTAg-reactive T cells in PBMC. CTAg-specific CD8+ T cells may have been primed and expanded by expression of CTAg on tumour cells or following ‘cross presentation’ through dendritic cells. In conclusion, T cells specific for cancer-testis antigens are present in the blood of a subset of patients with multiple myeloma. The clinical significance of this observation is currently being addressed.