Immunotherapy with CD25/CD71-allodepleted T cells to improve T-cell reconstitution after matched unrelated donor hematopoietic stem cell transplant: a randomized trial.

BACKGROUND AIMS: Delayed immune reconstitution is a major challenge after matched unrelated donor (MUD) stem cell transplant (SCT). In this randomized phase 2 multi-center trial, Adoptive Immunotherapy with CD25/71 allodepleted donor T cells to improve immunity after unrelated donor stem cell transplant (NCT01827579), the authors tested whether allodepleted donor T cells (ADTs) can safely be used to improve immune reconstitution after alemtuzumab-based MUD SCT for hematological malignancies. METHODS: Patients received standard of care or up to three escalating doses of ADTs generated through CD25+/CD71+ immunomagnetic depletion. The primary endpoint of the study was circulating CD3+ T-cell count at 4 months post-SCT. Twenty-one patients were treated, 13 in the ADT arm and eight in the control arm. RESULTS: The authors observed a trend toward improved CD3+ T-cell count at 4 months in the ADT arm versus the control arm (230/µL versus 145/µL, P = 0.18), and three ADT patients achieved normal CD3+ T-cell count at 4 months (>700/µL). The rates of significant graft-versus-host disease (GVHD) were comparable in both cohorts, with grade ≥2 acute GVHD in seven of 13 and four of eight patients and chronic GVHD in three of 13 and three of eight patients in the ADT and control arms, respectively. CONCLUSIONS: These data suggest that adoptive transfer of ADTs is safe, but that in the MUD setting the benefit in terms of T-cell reconstitution is limited. This approach may be of more use in the context of more rigorous T-cell depletion.

A novel anti-B7-H3 chimeric antigen receptor from a single-chain antibody library for immunotherapy of solid cancers.

B7-H3 (CD276) has emerged as a target for cancer immunotherapy by virtue of consistent expression in many malignancies, relative absence from healthy tissues, and an emerging role as a driver of tumor immune inhibition. Recent studies have reported B7-H3 to be a suitable target for chimeric antigen receptor-modified T cell (CAR-T) therapy using CARs constructed from established anti-B7-H3 antibodies converted into single-chain Fv format (scFv). We constructed and screened binders in an scFv library to generate a new anti-B7-H3 CAR-T with favorable properties. This allowed access to numerous specificities ready formatted for CAR evaluation. Selected anti-human B7-H3 scFvs were readily cloned into CAR-T and evaluated for anti-tumor reactivity in cytotoxicity, cytokine, and proliferation assays. Two binders with divergent complementarity determining regions were found to show optimal antigen-specific cytotoxicity and cytokine secretion. One binder in second-generation CD28-CD3ζ CAR format induced sustained in vitro proliferation on repeat antigen challenge. The lead candidate CAR-T also demonstrated in vivo activity in a resistant neuroblastoma model. An empirical approach to B7-H3 CAR-T discovery through screening of novel scFv sequences in CAR-T format has led to the identification of a new construct with sustained proliferative capacity warranting further evaluation.

Antitumor activity without on-target off-tumor toxicity of GD2-chimeric antigen receptor T cells in patients with neuroblastoma.

The reprogramming of a patient's immune system through genetic modification of the T cell compartment with chimeric antigen receptors (CARs) has led to durable remissions in chemotherapy-refractory B cell cancers. Targeting of solid cancers by CAR-T cells is dependent on their infiltration and expansion within the tumor microenvironment, and thus far, fewer clinical responses have been reported. Here, we report a phase 1 study (NCT02761915) in which we treated 12 children with relapsed/refractory neuroblastoma with escalating doses of second-generation GD2-directed CAR-T cells and increasing intensity of preparative lymphodepletion. Overall, no patients had objective clinical response at the evaluation point +28 days after CAR-T cell infusion using standard radiological response criteria. However, of the six patients receiving ≥108/meter2 CAR-T cells after fludarabine/cyclophosphamide conditioning, two experienced grade 2 to 3 cytokine release syndrome, and three demonstrated regression of soft tissue and bone marrow disease. This clinical activity was achieved without on-target off-tumor toxicity. Targeting neuroblastoma with GD2 CAR-T cells appears to be a valid and safe strategy but requires further modification to promote CAR-T cell longevity.

Tumor infiltrating lymphocytes expanded from pediatric neuroblastoma display heterogeneity of phenotype and function.

Adoptive transfer of ex vivo expanded tumor infiltrating lymphocytes (TILs) has led to clinical benefit in some patients with melanoma but has not demonstrated convincing efficacy in other solid cancers. Whilst the presence of TILs in many types of cancer is often associated with better clinical prognosis, their function has not been systematically evaluated across cancer types. Responses to immunological checkpoint inhibitors in a wide range of cancers, including those for which adoptive transfer of expanded TILs has not shown clinical benefit, has clearly delineated a number of tumor type associated with tumor-reactive lymphocytes capable of effecting tumor remissions. Neuroblastoma is an aggressive childhood solid cancer in which immunotherapy with GD2-directed antibodies confers a proven survival advantage through incompletely understood mechanisms. We therefore evaluated the feasibility of ex vivo expansion of TILs from freshly resected neuroblastoma tumors and the potential therapeutic utility of TIL expansions. TILs were successfully expanded from both tumor biopsies or resections. Significant numbers of NKT and γδT cells were identified alongside the mixed population of cytotoxic (CD8+) and helper (CD4+) T cells of both effector and central memory phenotypes. Isolated TILs were broadly non-reactive against autologous tumor and neuroblastoma cell lines, so enhancement of neuroblastoma killing was attained by transducing TILs with a second-generation chimeric antigen receptor (CAR) targeting GD2. CAR-TILs demonstrated antigen-specific cytotoxicity against tumor cell lines. This study is the first to show reproducible expansion of TILs from pediatric neuroblastoma, the high proportion of innate-like lymphocytes, and the feasibility to use CAR-TILs therapeutically.

Redirection to the bone marrow improves T cell persistence and antitumor functions.

A key predictor for the success of gene-modified T cell therapies for cancer is the persistence of transferred cells in the patient. The propensity of less differentiated memory T cells to expand and survive efficiently has therefore made them attractive candidates for clinical application. We hypothesized that redirecting T cells to specialized niches in the BM that support memory differentiation would confer increased therapeutic efficacy. We show that overexpression of chemokine receptor CXCR4 in CD8+ T cells (TCXCR4) enhanced their migration toward vascular-associated CXCL12+ cells in the BM and increased their local engraftment. Increased access of TCXCR4 to the BM microenvironment induced IL-15-dependent homeostatic expansion and promoted the differentiation of memory precursor-like cells with low expression of programmed death-1, resistance to apoptosis, and a heightened capacity to generate polyfunctional cytokine-producing effector cells. Following transfer to lymphoma-bearing mice, TCXCR4 showed a greater capacity for effector expansion and better tumor protection, the latter being independent of changes in trafficking to the tumor bed or local out-competition of regulatory T cells. Thus, redirected homing of T cells to the BM confers increased memory differentiation and antitumor immunity, suggesting an innovative solution to increase the persistence and functions of therapeutic T cells.

Chimeric Antigen Receptor-Engineered Human Gamma Delta T Cells: Enhanced Cytotoxicity with Retention of Cross Presentation.

Gamma delta T (γδT) lymphocytes are primed for rapid function, including cytotoxicity toward cancer cells, and are a component of the immediate stress response. Following activation, they can function as professional antigen-presenting cells. Chimeric antigen receptors (CARs) work by focusing T cell function on defined cell surface tumor antigens and provide essential costimulation for robust activation. Given the natural tropism of γδT cells for the tumor microenvironment, we hypothesized that their transduction with CARs might enhance cytotoxicity while retaining their ability to migrate to tumor and act as antigen-presenting cells to prolong the intratumoral immune response. Using a GD2-targeting CAR as a model system, we showed that γδT cells of both Vδ1 and Vδ2 subsets could be expanded and transduced to sufficient numbers for clinical studies. The CAR added to the cells' innate cytotoxicity by enhancing GD2-specific killing of GD2-expressing cancer cell lines. Migration toward tumor cells in vitro was not impaired by the presence of the CAR. Expanded CAR-transduced Vδ2 cells retained the ability to take up tumor antigens and cross presented the processed peptide to responder alpha beta T (αßT) lymphocytes. γδ CAR-T cell products show promise for evaluation in clinical studies of solid tumors.

Avoidance of On-Target Off-Tumor Activation Using a Co-stimulation-Only Chimeric Antigen Receptor.

Chimeric antigen receptors (CARs) combine T cell activation with antibody-mediated tumor antigen specificity, bypassing the need for T cell receptor (TCR) ligation. A limitation of CAR technology is on-target off-tumor toxicity caused by target antigen expression on normal cells. Using GD2 as a model cancer antigen, we hypothesized that this could be minimized by using T cells expressing Vγ9Vδ2 TCR, which recognizes transformed cells in a major histocompatibility complex (MHC)-unrestricted manner, in combination with a co-stimulatory CAR that would function independently of the TCR. An anti-GD2 CAR containing a solitary endodomain derived from the NKG2D adaptor DAP10 was expressed in Vγ9Vδ2+ T cells. Differential ligation of the CAR and/or TCR using antibody-coated beads showed that pro-inflammatory cytokine response depended on activation of both receptors. Moreover, in killing assays, GD2-expressing neuroblastoma cells that engaged the Vγ9Vδ2 TCR were efficiently lysed, whereas cells that expressed GD2 equivalently but did not engage the Vγ9Vδ2 TCR were untouched. Differentiation between X-on tumor and X-off tumor offers potential for safer immunotherapy and broader target selection.

Effective combination treatment of GD2-expressing neuroblastoma and Ewing's sarcoma using anti-GD2 ch14.18/CHO antibody with Vγ9Vδ2+ γδT cells.

Gamma delta T lymphocytes (γδT cells) have pleiotropic properties including innate cytotoxicity, which make them attractive effectors for cancer immunotherapy. Combination treatment with zoledronic acid and IL-2 can activate and expand the most common subset of blood γδT, which express the Vγ9Vδ2 T cell receptor (TCR) (Vδ2 T cells). Vγ9Vδ2 T cells are equipped for antibody-dependent cell-mediated cytotoxicity (ADCC) through expression of the low-affinity FcγR CD16. GD2 is a highly ranked tumor associated antigen for immunotherapy due to bright expression on the cell surface, absent expression on normal tissues and availability of therapeutic antibodies with known efficacy in neuroblastoma. To explore the hypothesis that zoledronic acid, IL-2 and anti-GD2 antibodies will synergize in a therapeutic combination, we evaluated in vitro cytotoxicity and tumor growth inhibition in the GD2 expressing cancers neuroblastoma and Ewing's sarcoma. Vδ2 T cells exert ADCC against GD2-expressing Ewing's sarcoma and neuroblastoma cell lines, an effect which correlates with the brightness of GD2 expression. In an immunodeficient mouse model of small established GD2-expressing Ewing's sarcoma or neuroblastoma tumors, the combination of adoptively transferred Vδ2+ T cells, expanded in vitro with zoledronic acid and IL-2, with anti-GD2 antibody ch14.18/CHO, and with systemic zoledronic acid, significantly suppressed tumor growth compared to antibody or γδT cell-free controls. Combination treatment using ch14.18/CHO, zoledronic acid and IL-2 is more effective than their use in isolation. The already-established safety profiles of these agents make testing of the combination in GD2 positive cancers such as neuroblastoma or Ewing's sarcoma both rational and feasible.

OX40- and CD27-mediated costimulation synergizes with anti-PD-L1 blockade by forcing exhausted CD8+ T cells to exit quiescence.

Exhaustion of chronically stimulated CD8(+) T cells is a significant obstacle to immune control of chronic infections or tumors. Although coinhibitory checkpoint blockade with anti-programmed death ligand 1 (PD-L1) Ab can restore functions to exhausted T cell populations, recovery is often incomplete and dependent upon the pool size of a quiescent T-bet(high) subset that expresses lower levels of PD-1. In a model in which unhelped, HY-specific CD8(+) T cells gradually lose function following transfer to male bone marrow transplantation recipients, we have explored the effect of shifting the balance away from coinhibition and toward costimulation by combining anti-PD-L1 with agonistic Abs to the TNFR superfamily members, OX40 and CD27. Several weeks following T cell transfer, both agonistic Abs, but especially anti-CD27, demonstrated synergy with anti-PD-L1 by enhancing CD8(+) T cell proliferation and effector cytokine generation. Anti-CD27 and anti-PD-L1 synergized by downregulating the expression of multiple quiescence-related genes concomitant with a reduced frequency of T-bet(high) cells within the exhausted population. However, in the presence of persistent Ag, the CD8(+) T cell response was not sustained and the overall size of the effector cytokine-producing pool eventually contracted to levels below that of controls. Thus, CD27-mediated costimulation can synergize with coinhibitory checkpoint blockade to switch off molecular programs for quiescence in exhausted T cell populations, but at the expense of losing precursor cells required to maintain a response.

A highly compact epitope-based marker/suicide gene for easier and safer T-cell therapy.

A compact marker/suicide gene that utilizes established clinical-grade reagents and pharmaceuticals would be of considerable practical utility to T-cell cancer gene therapy. Marker genes enable measurement of transduction and allow selection of transduced cells, whereas suicide genes allow selective deletion of administered T cells in the face of toxicity. We have created a highly compact marker/suicide gene for T cells combining target epitopes from both CD34 and CD20 antigens (RQR8). This construct allows selection with the clinically approved CliniMACS CD34 system (Miltenyi). Further, the construct binds the widely used pharmaceutical antibody rituximab, resulting in selective deletion of transgene-expressing cells. We have tested the functionality of RQR8 in vitro and in vivo as well as in combination with T-cell engineering components. We predict that RQR8 will make T-cell gene therapy both safer and cheaper.

Neuroblastoma killing properties of Vδ2 and Vδ2-negative γδT cells following expansion by artificial antigen-presenting cells.

PURPOSE: The majority of circulating human γδT lymphocytes are of the Vγ9Vδ2 lineage, and have T-cell receptor (TCR) specificity for nonpeptide phosphoantigens. Previous attempts to stimulate and expand these cells have therefore focused on stimulation using ligands of the Vγ9Vδ2 receptor, whereas relatively little is known about variant blood γδT subsets and their potential role in cancer immunotherapy. EXPERIMENTAL DESIGN: To expand the full repertoire of γδT without bias toward specific TCRs, we made use of artificial antigen-presenting cells loaded with an anti γδTCR antibody that promoted unbiased expansion of the γδT repertoire. Expanded cells from adult blood donors were sorted into 3 populations expressing respectively Vδ2 TCR chains (Vδ2(+)), Vδ1 chains (Vδ1(+)), and TCR of other δ chain subtypes (Vδ1(neg)Vδ2(neg)). RESULTS: Both freshly isolated and expanded cells showed heterogeneity of differentiation markers, with a less differentiated phenotype in the Vδ1 and Vδ1(neg)Vδ2(neg) populations. Expanded cells were largely of an effector memory phenotype, although there were higher numbers of less differentiated cells in the Vδ1(+) and Vδ1(neg)Vδ2(neg) populations. Using neuroblastoma tumor cells and the anti-GD2 therapeutic mAb ch14.18 as a model system, all three populations showed clinically relevant cytotoxicity. Although killing by expanded Vδ2 cells was predominantly antibody dependent and proportionate to upregulated CD16, Vδ1 cells killed by antibody-independent mechanisms. CONCLUSIONS: In conclusion, we have demonstrated that polyclonal-expanded populations of γδT cells are capable of both antibody-dependent and -independent effector functions in neuroblastoma.

Characterization of innate lymphoid cells in human skin and blood demonstrates increase of NKp44+ ILC3 in psoriasis.

Innate lymphoid cells (ILCs) are increasingly appreciated as key regulators of tissue immunity. However, their role in human tissue homeostasis and disease remains to be fully elucidated. Here we characterize the ILCs in human skin from healthy individuals and from the inflammatory skin disease psoriasis. We show that a substantial proportion of IL-17A and IL-22 producing cells in the skin and blood of normal individuals and psoriasis patients are CD3-negative innate lymphocytes. Deep immunophenotyping of human ILC subsets showed a statistically significant increase in the frequency of circulating NKp44+ ILC3 in the blood of psoriasis patients compared with healthy individuals or atopic dermatitis patients. More than 50% of circulating NKp44+ ILC3 expressed cutaneous lymphocyte-associated antigen, indicating their potential for skin homing. Analysis of skin tissue revealed a significantly increased frequency of total ILCs in the skin compared with blood. Moreover, the frequency of NKp44+ ILC3 was significantly increased in non-lesional psoriatic skin compared with normal skin. A detailed time course of a psoriasis patient treated with anti-tumor necrosis factor showed a close association between therapeutic response, decrease in inflammatory skin lesions, and decrease of circulating NKp44+ ILC3. Overall, data from this initial observational study suggest a potential role for NKp44+ ILC3 in psoriasis pathogenesis.

TLRs to cytokines: mechanistic insights from the imiquimod mouse model of psoriasis.

Psoriasis is an inflammatory disease of the skin affecting 2-3% of the population, characterized by a thickening of the epidermis and immune infiltrates throughout the dermis and epidermis, causing skin lesions that can seriously affect quality of life. The study of psoriasis has historically been hampered by the lack of good animal models. Various genetically induced models exist, which have provided some information about possible mechanisms of disease, but these models rely mostly on intrinsic imbalances of homeostasis. However, a mouse model of psoriasiform dermatitis caused by the repeated topical application of Aldara™ containing 5% imiquimod was described in 2009. The mechanisms of action of Aldara™ are complex. Imiquimod is an effective ligand for TLR7 (and TLR8 in humans) and also interferes with adenosine receptor signaling. In addition, isostearic acid present in the Aldara™ vehicle has been shown to be biologically active and of importance for activating the inflammasome. Interestingly, the repetitive application of Aldara™ reveals a complex aetiology involving multiple cell types, cytokines, and inflammatory pathways. In this review, we will dissect the findings of the imiquimod model to date and ask how this model can inform us about the immunological aspects of human disease.

What on "irf" is this gene 4? Irf4 transcription-factor-dependent dendritic cells are required for T helper 2 cell responses in murine skin.

Interferon regulatory factors play an important role in the transcriptional regulation of immunity. In this issue of Immunity, Kumamoto et al. (2013) and Gao et al. (2013) identify an Irf4-dependent migratory dendritic cell subset required for T helper 2 cell polarization following cutaneous challenge.

Cell-intrinsic regulation of murine dendritic cell function and survival by prereceptor amplification of glucocorticoid.

Although the inhibitory effects of therapeutic glucocorticoids (GCs) on dendritic cells (DCs) are well established, the roles of endogenous GCs in DC homeostasis are less clear. A critical element regulating endogenous GC concentrations involves local conversion of inactive substrates to active 11-hydroxyglucocorticoids, a reduction reaction catalyzed within the endoplasmic reticulum by an enzyme complex containing 11ß-hydroxysteroid dehydrogenase type 1 (11ßHSD1) and hexose-6-phosphate dehydrogenase (H6PDH). In this study, we found that this GC amplification pathway operates both constitutively and maximally in steady state murine DC populations and is unaffected by additional inflammatory stimuli. Under physiologic conditions, 11ßHSD1-H6PDH increases the sensitivity of plasmacytoid DCs (pDCs) to GC-induced apoptosis and restricts the survival of this population through a cell-intrinsic mechanism. Upon CpG activation, the effects of enzyme activity are overridden, with pDCs becoming resistant to GCs and fully competent to release type I interferon. CD8α(+) DCs are also highly proficient in amplifying GC levels, leading to impaired maturation following toll-like receptor-mediated signaling. Indeed, pharmacologic inhibition of 11ßHSD1 synergized with CpG to enhance specific T-cell responses following vaccination targeted to CD8α(+) DCs. In conclusion, amplification of endogenous GCs is a critical cell-autonomous mechanism for regulating the survival and functions of DCs in vivo.

Xenogeneic graft-versus-host-disease in NOD-scid IL-2Rγnull mice display a T-effector memory phenotype.

The occurrence of Graft-versus-Host Disease (GvHD) is a prevalent and potentially lethal complication that develops following hematopoietic stem cell transplantation. Humanized mouse models of xenogeneic-GvHD based upon immunodeficient strains injected with human peripheral blood mononuclear cells (PBMC; "Hu-PBMC mice") are important tools to study human immune function in vivo. The recent introduction of targeted deletions at the interleukin-2 common gamma chain (IL-2Rγ(null)), notably the NOD-scid IL-2Rγ(null) (NSG) and BALB/c-Rag2(null) IL-2Rγ(null) (BRG) mice, has led to improved human cell engraftment. Despite their widespread use, a comprehensive characterisation of engraftment and GvHD development in the Hu-PBMC NSG and BRG models has never been performed in parallel. We compared engrafted human lymphocyte populations in the peripheral blood, spleens, lymph nodes and bone marrow of these mice. Kinetics of engraftment differed between the two strains, in particular a significantly faster expansion of the human CD45(+) compartment and higher engraftment levels of CD3(+) T-cells were observed in NSG mice, which may explain the faster rate of GvHD development in this model. The pathogenesis of human GvHD involves anti-host effector cell reactivity and cutaneous tissue infiltration. Despite this, the presence of T-cell subsets and tissue homing markers has only recently been characterised in the peripheral blood of patients and has never been properly defined in Hu-PBMC models of GvHD. Engrafted human cells in NSG mice shows a prevalence of tissue homing cells with a T-effector memory (T(EM)) phenotype and high levels of cutaneous lymphocyte antigen (CLA) expression. Characterization of Hu-PBMC mice provides a strong preclinical platform for the application of novel immunotherapies targeting T(EM)-cell driven GvHD.

Memory lapses in graft-versus-host disease.

"Faster, better, more" is the conventional benchmark used to define responses of memory T cells when compared with their naïve counterparts. In this issue of the European Journal of Immunology, Mark and Warren Shlomchik and colleagues [Eur. J. Immunol. 2011. 41: 2782-2792] make the intriguing observation that murine memory CD4(+) T-cell populations enriched for alloreactive precursors are fully capable of rejecting allogeneic skin grafts but yet are incapable of inducing significant graft-versus-host disease. These observations add to the emerging concept that memory CD4(+) T-cell development is more nuanced and complex than predicted by conventional models. In particular, the data suggest that it may be just as important to consider what naïve or effector cells have "lost" in their transition to memory.

Langerhans cells regulate cutaneous injury by licensing CD8 effector cells recruited to the skin.

Langerhans cells (LCs) are a distinct population of dendritic cells that form a contiguous network in the epidermis of the skin. Although LCs possess many of the properties of highly proficient dendritic cells, recent studies have indicated that they are not necessary to initiate cutaneous immunity. In this study, we used a tractable model of cutaneous GVHD, induced by topical application of a Toll-like receptor agonist, to explore the role of LCs in the development of tissue injury. By adapting this model to permit inducible and selective depletion of host LCs, we found that GVHD was significantly reduced when LCs were absent. However, LCs were not required either for CD8 T-cell activation within the draining lymph node or subsequent homing of effector cells to the epidermis. Instead, we found that LCs were necessary for inducing transcription of IFN-γ and other key effector molecules by donor CD8 cells in the epidermis, indicating that they license CD8 cells to induce epithelial injury. These data demonstrate a novel regulatory role for epidermal LCs during the effector phase of an inflammatory immune response in the skin.

Specificity for the tumor-associated self-antigen WT1 drives the development of fully functional memory T cells in the absence of vaccination.

Recently, vaccines against the Wilms Tumor antigen 1 (WT1) have been tested in cancer patients. However, it is currently not known whether physiologic levels of WT1 expression in stem and progenitor cells of normal tissue result in the deletion or tolerance induction of WT1-specific T cells. Here, we used an human leukocyte antigen-transgenic murine model to study the fate of human leukocyte antigen class-I restricted, WT1-specific T cells in the thymus and in the periphery. Thymocytes expressing a WT1-specific T-cell receptor derived from high avidity human CD8 T cells were positively selected into the single-positive CD8 population. In the periphery, T cells specific for the WT1 antigen differentiated into CD44-high memory phenotype cells, whereas T cells specific for a non-self-viral antigen retained a CD44(low) naive phenotype. Only the WT1-specific T cells, but not the virus-specific T cells, displayed rapid antigen-specific effector function without prior vaccination. Despite long-term persistence of WT1-specific memory T cells, the animals did not develop autoimmunity, and the function of hematopoietic stem and progenitor cells was unimpaired. This is the first demonstration that specificity for a tumor-associated self-antigen may drive differentiation of functionally competent memory T cells.

Nonhematopoietic antigen blocks memory programming of alloreactive CD8+ T cells and drives their eventual exhaustion in mouse models of bone marrow transplantation.

Allogeneic blood or BM transplantation (BMT) is the most commonly applied form of adoptive cellular therapy for cancer. In this context, the ability of donor T cells to respond to recipient antigens is coopted to generate graft-versus-tumor (GVT) responses. The major reason for treatment failure is tumor recurrence, which is linked to the eventual loss of functional, host-specific CTLs. In this study, we have explored the role of recipient antigen expression by nonhematopoietic cells in the failure to sustain effective CTL immunity. Using clinically relevant models, we found that nonhematopoietic antigen severely disrupts the formation of donor CD8+ T cell memory at 2 distinct levels that operate in the early and late phases of the response. First, initial and direct encounters between donor CD8+ T cells and nonhematopoietic cells blocked the programming of memory precursors essential for establishing recall immunity. Second, surviving CD8+ T cells became functionally exhausted with heightened expression of the coinhibitory receptor programmed death-1 (PD-1). These 2 factors acted together to induce even more profound failure in long-term immunosurveillance. Crucially, the functions of exhausted CD8+ T cells could be partially restored by late in vivo blockade of the interaction between PD-1 and its ligand, PD-L1, without induction of graft-versus-host disease, suggestive of a potential clinical strategy to prevent or treat relapse following allogeneic BMT.