Adoptive Cellular Immunotherapy

Stephen Gottschalk

Cliona M. Rooney

Malcolm K. Brenner

INTRODUCTION

Over the last thirty years, a multidisciplinary approach including chemotherapy, radiation, surgery, and/or hematopoietic stem cell transplantation (HSCT) has led to a dramatic improvement in the long-term survival of patients with pediatric malignancies. However, current treatments kill dividing cells indiscriminately, causing considerable short- and long-term adverse effects, which are of considerable public health concern since an estimated 1 in 640 young adults are long-term cancer survivors.¹^,² Adoptive cellular immunotherapy³^,⁴^,⁵ would be an appealing addition to the treatment armamentarium of pediatric malignancies because it may selectively kill malignant or diseased cells and thereby reduce short- and long-term side effects. In addition, it promises to benefit patients who currently fail multimodal therapy.

Adoptive immunotherapy of immune effector cells may have significant advantages over alternative immunotherapeutic approaches such as cancer vaccines since the phenotype, activity, and specificity of expanded cells can be analyzed prior to injection. However, the cells can be genetically engineered prior to infusion to modify the antigens they target, increase their potency, and reduce their susceptibility to tumor-mediated immune inhibition. Genetic modification also allows the cells to be tracked in vivo to learn their fate. Adoptive transfer of immune effector cells has been most successful after HSCT as prophylaxis and therapy for viral-associated disease.⁶ In addition, the adoptive transfer of tumor-infiltrating lymphocytes (TILs) in combination with lymphodepleting chemotherapy has shown promising long-term antitumor effects in patients with melanoma.³^,⁷^,⁸ Last, the adoptive transfer of T cells that have been genetically engineered to be tumor specific has resulted in impressive clinical responses, especially in patients with hematological malignancies.⁹^,¹⁰^,¹¹^,¹²^,¹³^,¹⁴^,¹⁵^,¹⁶^,¹⁷^,¹⁸^,¹⁹^,²⁰ Despite these successes, clinical studies have also highlighted current limitations of adoptively transferred immune cells, including limited expansion in vivo, lack of homing to tumor sites, and toxicities due to recognition of normal tissues. In addition, the complexities of ex vivo immune cell production have prevented their broader clinical application. While we will briefly discuss the adoptive transfer of natural killer (NK) cells, γδ thymus-derived T cells (γδ T cells), and invariant NKT cells (iNKT), the chapter is focused on the adoptive transfer of T cells, which have been used in the majority of clinical studies (Table 14.1). In addition, we will focus on genetic modifications intended to overcome current limitations of T-cell therapies and highlight streamlined cell production technologies that should facilitate larger scale clinical trials in the future.

NATURAL KILLER CELLS

NK cells are a component of the innate— or “front-line”—immune response. They recognize cell surface abnormalities, such as low expression of MHC class I or carbohydrate abnormalities present on virus-infected or cancerous cells. Full activation of NK cells is dependent on engagement of activating receptors, such as NKG2D, DNAM-1, and natural cytotoxicity receptors, such as NKp30, NKp44, and NKp46, and nonengagement of inhibitory receptors, notably the family of killer immunoglobulin-like receptors (KIRs). It has long been recognized that NK cells can contribute to control of malignancies, acute myeloid leukemia (AML) in particular. Studies showed that after HSCT, an increased degree of KIR mismatch between stem cell donor and recipient lead to prolonged relapse-free survival in patients with AML, most likely due to decreased NK-cell inhibition.²¹ In addition, adoptively transferred haploidentical, KIR-mismatched NK cells had anti-AML activity in patients with poor prognosis.²² NK cells also have antitumor activity in vitro against pediatric solid tumors such as Ewing sarcoma, osteosarcoma, and rhabdomyosarcoma.²³ Under physiological conditions, however, NK cells are inhibited by autologous KIRs and so have no activity against the patients’ own tumors. In addition, NK cells of cancer patients often express low levels of activating receptors or deficiencies in intracellular signaling molecules that inhibit their antitumor function.

To overcome these road blocks to clinical use investigators have expanded NK cells derived from healthy donors ex vivo using HLA-negative K562 cells engineered to express NK-cell stimulatory molecules such as CD137L and membrane-bound IL-15²⁴^,²⁵ or IL-21.²⁶ In preclinical studies, such expanded allogeneic NK cells have potent antitumor effects against Ewing sarcoma, rhabdomyosarcoma, and other pediatric malignancies.²³ Several clinical studies using these cells are currently in progress after matched-related or unrelated HSCT, or in patients who receive haploidentical NK cells derived from a related donor. Investigators are also pursuing several strategies to enhance the therapeutic potential of NK cells. These include epigenetic modifiers that increase expression of activating NK-cell ligands on cancer cells,²⁷ and combining NK cells with tumor-directed monoclonal antibodies (MAbs) to exploit the receptors for immunoglobulin constant regions (Fc receptors) that are present on the cells.²⁸ Other studies have armed NK cells with bispecific NK-cell engager,²⁹ or genetically modified NK cells, to express chimeric antigen receptors (CARs; see section “CAR T cells”).³⁰^,³¹

γδ THYMUS-DERIVED T CELLS

γδ T cells are a small subset of T cells. They express γδ T-cell receptors (TCRs), but unlike conventional αβ T cells that only recognize specific peptide antigens presented in the context of an MHC molecule, γδ T cells recognize a broader range of antigens in an MHC-independent fashion. These antigens include MHC-like stress-induced self-antigens such as NKG2D ligands, glycolipids presented by CD1c, and phosphoantigens produced as a byproduct of bacterial metabolic pathways. Ex vivo expansion of γδ T cells has been difficult in the past, and so far there is limited human experience using adoptive transfer of γδ T cells.³² One recent study described a method to rapidly generate a large number of clinical-grade, polyclonal γδ T cells using artificial antigen-presenting cells (APCs) in combination with IL-2 and IL-21.
This approach may facilitate future clinical studies for pediatric patients with malignancies that overexpress NKG2D ligands such as Ewing sarcoma and osteosarcoma.³³

TABLE 14.1 Clinical Experience with the Adoptive Transfer of Unmodified Immune Cells

Cell Product	Antigens	Disease/Comment	References
NK cells
Allogeneic	NKG2D ligands	Leukemia, pediatric solid tumors^a	22
Autologous	NKG2D ligands	Multiple myeloma^a	22
γδ T cells
Autologous	^b	Renal cell carcinoma	32
iNKT cells
Autologous	CD1d	Lung cancer	37
DLI
Unmanipulated	^c	Leukemia	69
Unmanipulated		Viral infections	66-68
Allodepleted		Postallogeneic transplant	72,73
Allodepleted, gene modified		iC9-modified; postallogeneic transplant	76,80
Gene-modified		HSV-tk or iC9^a gene-modified	75
Polyclonal-activated T cells
Allogeneic		Lymphoma; postallogeneic transplant	85
Autologous	Survivin, telomerase	Multiple myeloma, postautologous transplant; HCC	82,83,87
TILs	^d	Melanoma, renal cell carcinoma, cholangiocarcinoma	7,8,93
Antigen-specific T cells
Virus-specific T cells
Allogeneic	EBV, CMV, AdV, HHV6, BKV	Viral infections, EBV-positive lymphoma	95-108
Autologous	EBV, CMV	Viral infections, EBV-positive lymphoma and NPC, CMV-positive GBM	109-115
Tumor antigen-specific T cells
Allogeneic	mHAgs, BCR-ABL, PR1, WT1	Leukemia	142,143
Autologous	MART-1, gp100, NY-ESO-1, HER2	Melanoma, breast cancer	124,146-148,150
^aOngoing; ^bMultiple antigens recognized in MHC-independent manner; ^cMultiple antigens recognized in MHC-dependent manner; ^dSpecific for multiple melanoma-associated antigens (e.g., MART-1, gp100); patient-specific, immunogenic mutations in several proteins have also been identified (e.g., βKIF2C, POLA2).

INVARIANT NKT CELLS

iNKT cells, like γδ T cells, are a small subset of T cells. They express NK-cell markers and an invariant Vα24 TCR that recognizes the glycolipid CD1d.³⁴ Several studies have demonstrated that iNKT cells can contribute to the immune surveillance of malignancies.³⁵ While iNKT cells can directly kill CD1d-positive tumor cells, they also attack normal host cells such as tumor-associated macrophages (TAMs) that support the tumor microenvironment and that are CD1d positive even when the tumor cells themselves are CD1d negative.³⁶ Limited clinical experience with iNKT cells has shown safety but modest efficacy in advanced-stage lung cancer patients.³⁷ Preclinical studies in neuroblastoma xenograft models³⁸^,³⁹ have demonstrated that the effector function of iNKT cells can be dramatically improved by transgenic expression of either IL-15 or GD2-specific CARs but this approach has yet to be implemented clinically.

T CELLS

Target Antigens for T-cell Therapy

Viral infections induce robust activation of innate immune responses that in turn promote the activation and expansion of T-cells specific for foreign viral antigens. By contrast, most tumors can develop without activating the innate immune response and therefore tumor-antigen-specific T-cells are activated only weakly or not at all. To compound this problem, high-affinity self-antigen-specific T-cells are deleted in the thymus during development. Nevertheless, many TAAs can be recognized by the cellular and humoral immune systems because they are normally expressed only during fetal development or at immunoprivileged sites; are expressed by the tumor at higher levels than by normal tissues; contain a novel peptide sequence created by gene mutation or rearrangement; or are derived from viral antigens.⁴⁰ Over the last two decades, numerous TAAs have been identified, many of which are relevant for pediatric malignancies and are listed in Table 14.2.⁴¹^,⁴²^,⁴³^,⁴⁴^,⁴⁵^,⁴⁶^,⁴⁷^,⁴⁸^,⁴⁹^,⁵⁰^,⁵¹^,⁵²^,⁵³^,⁵⁴^,⁵⁵^,⁵⁶^,⁵⁷^,⁵⁸^,⁵⁹^,⁶⁰^,⁶¹^,⁶²

Since many TAA are also expressed at low levels in normal tissues, there is a continuing need to discover new antigens. Mining
publically available gene expression array data sets, Orentas et al.⁶³ recently described several TAAs that are shared among multiple pediatric tumors including melanoma cell adhesion molecule (MCAM) and glypican-2. In addition, whole-exome sequencing has successfully identified protein mutations that produce novel epitopes.⁶⁴

TABLE 14.2 Selected Tumor-associated Antigens Expressed in Pediatric Malignancies

Antigen	Malignancy	References
Mutations/novel epitopes in oncogenic fusion proteins
BCR-ABL	CML	41
DEK-CAN	AML	42
ETV6-AML1	ALL	43
EWS-FLI-1	EWS	44
PML-RARα	PML	45
SYT-SSX2	Synovial sarcoma	46
Cancer testis antigens
SSX family	OS	47
BAGE family	RMS	48
GAGE family	Brain tumors, RMS	48
MAGE family	Brain tumors, NB, OS, RMS	48-50
XAGE family	EWS	51
NY-ESO-1	OS, synovial sarcoma	50
Overexpressed antigens
B7-H3	EWS, NB, OS, RMS	52
CLUAP1	OS	53
HER2	High-grade glioma, MB, OS	54-56
IL-11Rα	OS	57
IL-13Rα2	High-grade glioma, MB	56
EphA2	High-grade glioma	58
Papilloma virus-binding factor	OS	59
STEAP1	EWS	60
Survivin^a	Brain tumors, hematopoietic malignancies, NB	58
Telomerase^a	HB	61
WT1	Hematopoietic malignancies, Wilms tumor	62
^aUniversal tumor antigens, most likely overexpressed in other pediatric malignancies.
ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CML, chronic myelogenous leukemia; EWS, Ewing sarcoma; OS, osteosarcoma; NB, neuroblastoma; MB, medulloblastoma; PML, promyelocytic leukemia; RMS, rhabdomyosarcoma; SS, synovial sarcoma.

Requirements for the Activation and Expansion of Antigen-specific T Cells

The activation of antigen-specific T cells is a multistep process requiring antigen-specific triggering of the TCR complex on the T-cell and additional signaling via co-stimulatory molecules and cytokines (Fig. 14.1).⁶⁵ The TCR is triggered by the specific recognition of peptides complexed with MHC class I or class II molecules at the cell surface. CD8-positive cytotoxic T cells classically recognize peptides presented on MHC class I molecules and CD4-positive helper T cells recognize peptides in the context of MHC class II molecules. MHC class I peptide loading occurs in the endoplasmic reticulum and requires proteasome-mediated antigen processing in the cytosol (Fig. 14.2). Thus, for class I presentation, antigens must gain access to the cytosol. This is usually a prerogative of endogenously expressed proteins. “Professional” APCs, however, have the capacity to phagocytose soluble antigens and present them for MHC class I molecules, a process called “cross-priming.” In contrast to MHC class I peptide loading, MHC class II peptides are predominantly derived from phagocytosed soluble antigens. In addition to the described “classical” MHC class I and II presentation pathways, recent studies have highlighted the role of autophagy, the degradation of a cell’s own components, in antigen presentation.

TCR recognition of peptide/MHC complexes results in the formation of an immunological synapse between T cells and the target cells or APCs. The immunological synapse consists of a central supramolecular activation cluster (SMAC) containing TCR and peptide MHC complexes, and a peripheral SMAC consisting of cell adhesion molecules such as LFA-1 and its counterpart ICAM-1. For effective T-cell activation and expansion, additional co-stimulatory signals are necessary. These include receptors belonging to the immunoglobulin superfamily, such as CD28 and ICOS, as well as members of the tumor-necrosis factor receptor (TNFR) superfamily such as OX40 (CD134) and 4-1BB (CD137). Inhibition of the CD28 pathway in the presence of antigenic stimulation results in T-cell anergy while blocking of the TNFR superfamily limits the expansion of antigen-specific T cells and reduces the frequency of memory T cells.

Figure 14.1 Requirements for T-cell activation and expansion. Step 1—Adhesion: Leukocyte function antigen 1 (LFA1) on T cells and intracellular adhesion molecule 1 (ICAM1) on target cells/antigen-presenting cells facilitate the interaction between both cells. Step 2—T-cell receptor (TCR) engagement: The TCR binds to its cognate peptide presented by the appropriate MHC class molecule. Step 3—Co-stimulation: The T-cell receives co-stimulation, commonly by binding of CD28 to CD80/CD86. In addition, other co-stimulatory molecules are often required for optimal stimulation including CD134 (OX40) or CD137 (4-1BB). Step 4—Expansion: Cytokines like IL-2, IL-7, and/or IL-15 are required for expansion. γpr, private γ cytokine chain; γc, common γ cytokine chain.

Figure 14.2 MHC class I and II processing pathways. MHC class I processing pathway: CD8 positive CTL recognize peptide presented on MHC class I molecules. For MHC class I molecule loading, intracellular proteins are degraded in the cytosol by the proteasome. The generated peptides are transported into the endoplasmic reticulum (ER) by transporter-associated proteins (TAP) 1 and 2. In the ER, the peptides are loaded onto MHC class I molecules and transported to the cell surface. MHC class II processing pathway: CD4 positive helper T cells recognize peptides on MHC class II molecules. Peptides for MHC class II molecule loading are usually derived from extracellular proteins that are phagocytosed by APCs. Phagocytosed proteins are degraded in endosomal/lysosomal compartments, where the resulting peptides are loaded onto MHC class II molecules and transported to the cell surface. Helper T cells provide cognate help to CTL for optimal expansion.

In addition to the co-stimulatory and adhesion molecules described above, the CD40/CD154 (CD40L) pathway is important for continued T-cell expansion after the initial stimulation. Moreover, the adhesion molecule LFA-1 also contributes to T-cell activation and differentiation by increasing the affinity of the initial TCR/MHC interaction. Finally, CD8-positive T cells must receive appropriate help from antigen-specific T helper cells with Th1 activity. Th1 cells secrete cytokines, such as IFN-γ and IL-2, which are important for T-cell activation and expansion. However, cytokine-mediated help may also derive from other cell types such as monocytes. Once a productive T-cell response is initiated, immune check points such as CTLA-4 and PD-1 prevent uncontrolled T-cell proliferation and mitigate collateral tissue damage.

Uses of T-cell Therapy

Donor Lymphocyte Infusions Following Stem Cell Transplantation

Investigators have attempted to treat posttransplant viral diseases and relapsed malignancies by infusion of donor lymphocytes (DLI) since this population contains virus-reactive and potentially tumor-specific T cells.

Donor Lymphocytes to Treat Viral Infections. Serious viral infections contribute significantly to morbidity and mortality following HSCT, and the most frequent of these include CMV, EBV, AdV, BK polyoma virus (BKV), human herpes virus 6 (HHV6), respiratory syncytial virus (RSV), and metapneumovirus (MPV). While pharmacological therapies such as ganciclovir (GCV) or foscarnet are effective treatment for CMV, many other viruses lack established therapies. Even for CMV, treatment may be associated with toxicities such as myelosuppression or renal dysfunction and may lead to the development of drug-resistant viral strains. Under physiological conditions, T cells play a critical role in control of these viral infections, and it is the diminished number and function of these cells after HSCT that contributes to the increased incidence of viral disease. Indeed, recipients who receive ex vivo T-cell depleted stem cell products or T-cell depleting antibodies as part of their conditioning regimen have a particularly high morbidity and mortality from these disorder. Although DLIs have successfully been employed to treat EBV, CMV, and AdV infections post-HSCT,⁶⁶^,⁶⁷^,⁶⁸ their use is limited by the presence of alloreactive T cells with the DLI product that can cause graft-versus-host disease (GVHD), a potentially life-threatening complication. Several strategies have been developed to reduce the risk of GVHD from DLI, either by depleting alloreactive T cells, genetically modifying T cells with “suicide genes,” or enriching cell products for virus-specific T cells (see section “Virus-specific T cells”).

Donor Lymphocytes to Treat Malignancies. Adoptive immunotherapy with DLI after HSCT effectively augments the graft-versus-leukemia response and can eliminate residual disease, especially in patients with chronic myelogenous leukemia (CML).⁶⁹^,⁷⁰ In the original studies, three patients with relapsed CML attained complete cytogenetic remissions after treatment with DLI and IFNα post-HSCT. In larger series, approximately 70% of all relapsed CML patients treated in chronic phase achieved complete cytogenetic remission but in accelerated phase or blast crisis just 11% had a CR. DLI are less effective treatment for patients who relapse with other hematological malignancies after HSCT, with a response rate of 29% for AML and 5% for ALL.⁶⁹

Although this DLI-mediated graft-versus-leukemia (GVL) effect may simply be another manifestation of GHVD, recent studies indicate that minor histocompatibility antigen (mHAgs)-specific
T cells present in DLIs can induce durable remissions of malignancies without causing GVHD. Nonetheless, the application of DLI remains limited by the frequency with which it induces GVHD. Thus, investigators have attempted to ameliorate GVHD while sparing GVL by infusing allodepleted and suicide gene-modified donor lymphocytes. In addition, investigators have isolated donor-derived T cells that recognize virus-specific T cells (see section “Virus-specific T cells”) or TAA-specific T cells as discussed in section “Donor-derived, antigen-specific T cells” (Fig. 14.3).

Figure 14.3 Strategies to decrease the GVHD risk of DLI. The risk of GVHD or its complication can be reduced by (1) genetically modifying donor lymphocytes with suicide genes, (2) depleting alloreactive T cells, or (3) expanding or selecting antigen-specific T cells.

Allodepleted and Suicide Gene-modified Donor Lymphocytes. Infusing CD8-depleted lymphocytes reduces the incidence of GVHD with preservation of “graft versus tumor” effects.⁷¹ However, due to varying degrees of alloreactivity with different donor-recipient pairs, it has proven difficult to define a T-cell dose that has antitumor activity but does not cause GVHD. Several approaches have therefore been developed to reduce the risk of GVHD or to specifically eliminate infused T cells once GVHD develops.

One means of overcoming the problem of alloreactivity is to remove alloreactive cells from the T-cell product on the basis of these cells’ upregulation of activation markers in response to stimulation by alloantigens. Several studies have evaluated this strategy using an immunotoxin directed against the activation marker CD25. This procedure enables the in vitro depletion of alloreactive cells while preserving T cells reactive to viruses and tumor-associated antigens such as the minor histocompatibility antigen HA1 and primary granule enzyme proteinase 3. In two Phase I clinical studies, patients received T cells depleted of alloreactive cells, and early T-cell expansion was seen with improvements in immune reconstitution against viral pathogens without significant GVHD.⁷²^,⁷³ Besides CD25 immunotoxin, a photoactive rhodamine derivate, TH9402, has also been used to deplete alloreactive T cells ex vivo prior to T-cell infusion.⁷⁴

An alternative approach relies on the introduction of a suicide gene into T cells so that cell death can be induced in vivo once GVHD develops. For example, donor-derived T cells transduced with a retroviral vector expressing human herpes simplex virus thymidine kinase (HSV-tk), which renders cells sensitive to GCV, have been infused into HSCT recipients. These modified T cells were safe and accelerated immune reconstitution, and any GVHD was successfully controlled with GCV.⁷⁵ The limitations of this approach are the inherent immunogenicity of HSV-tk, which may lead to unwanted elimination of the modified cells, and the requirement of nucleoside analogs such as GCV for T-cell killing, which precludes their use as antiviral agents. As an alternative, killing may be induced by introducing dimerizable death molecules such as inducible caspase 9 (iC9) that can be triggered by a small-molecule dimerizing drug or by using monoclonal antibodies directed to cell surface antigens, such as CD20 (using rituximab) or EGFR (using cetuximab) introduced into the T cells.⁷⁶^,⁷⁷^,⁷⁸^,⁷⁹ iC9 gene-modified, allodepleted donor-derived T cells were evaluated in haploidentical HSCT recipients.⁷⁶ A single dose of a small-molecule drug (AP1903) that dimerizes and activates the iC9 transgene eliminated more than 90% of iC9 gene-modified T cells within 30 minutes of administration and eliminated GVHD without recurrence. Long-term follow-up of these patients revealed accelerating immune reconstitution, resulting in protection against major viral pathogens including EBV, CMV, AdV, and BKV.⁸⁰

Polyclonal-activated T Cells

T cells in cancer patients are often anergic, that is, unresponsive when they are stimulated through their T-cell receptor. One potential strategy to reverse anergy is to nonspecifically activate T cells ex vivo outside the immunosuppressive environment of the cancer patient prior to reinfusion. Ex vivo activation of T cells has been most extensively evaluated using CD3-antibody-stimulated T cells. Most clinical studies have been in patients with metastatic renal cell carcinoma, but initially encouraging results have not been replicated in larger studies.⁸¹ Similarly, a study using autologous activated T cells as adjuvant therapy after complete resection of hepatocellular carcinoma showed no influence on overall survival, although infused patients had significantly longer recurrence-free survival.⁸² Activation of T cells with anti-CD3- and anti-CD28-coated beads ex vivo may have greater potential to overcome disease-induced anergy and augment CD4-positive T-cell responses. Several clinical studies have been conducted with anti-CD3/anti-CD28 autologous, activated T cells,⁸³^,⁸⁴^,⁸⁵^,⁸⁶ in which the patients received activated T cells after autologous HSCT. T-cell infusions induced a rapid recovery of lymphocyte counts and reversed cytokine activation deficits in vitro. In a subset of patients, T-cell infusions were associated with a clinical picture indistinguishable from acute GVHD.⁸⁴

In one study, donor-derived, activated T cells were given to patients after allogeneic HSCT. Infusions were safe and not associated with an increased risk of GVHD. The impact on disease and relapse was unclear, and further studies will be needed to determine if this approach is superior to DLIs. In an effort to increase the frequency of antigen-specific T cells in autologous, activated T-cell
products, multiple myeloma patients have been vaccinated with an influenza virus vaccine, pneumococcal vaccine, and/or a peptide vaccine targeting the TAAs survivin and telomerase prior to leukapheresis and ex vivo T-cell activation. This resulted in improved recovery of influenza virus-specific or pneumococcal-specific immune responses post-infusion in comparison with patients, who did not receive the vaccine.⁸³^,⁸⁶^,⁸⁷ While approximately one-third of patients who received the survivin and telomerase peptide vaccine developed survivin- and/or telomerase-specific immune responses, the responses were in general significantly lower when compared with the coadministered pneumococcal vaccine, most likely explaining why the induction of surviving- and/or telomerase-specific immune responses did increase event-free survival rate in comparison with nonresponders.⁸⁷

Tumor-infiltrating Lymphocytes

TILs can be isolated from 30% to 65% of human solid tumors and are usually CD4-positive T cells except in melanoma, in which CD8-positive T cells predominate.⁸⁸ As the number of TILs isolated from tumor biopsies is insufficient for adoptive immunotherapy protocols, they must be expanded ex vivo prior to infusion, using cytokines such as IL-2. Expanded CD8-positive TIL have cytolytic activity against the original tumor and, unlike LAK cells, their killing is MHC class I restricted.³ The varied success of clinical trials using TIL likely reflects differences in tumor immunogenicity and tumor burden at the time of TIL infusion.⁸⁹ To increase the antitumor efficacy of TILs, patients have been lymphodepeleted with fludarabine (Flu) and cyclophosphamide (Cy) with 0, 2, or 12 Gy of total body irradiation (TBI) prior to TIL transfer.⁷^,⁹⁰ Lymphodepletion in 93 patients with metastatic, refractory melanoma increased the levels of IL-7 and IL-15 in peripheral blood and increased the objective response rate from 49% in patients receiving only Flu/Cy to 72% in patients receiving Flu/Cy and 12 Gy TBI. Telomere length and the expression of cell surface markers such as CD27 have been correlated with antitumor activity of infused TILs.⁹¹

The specificity of T cells in TILs is actively being investigated and studies have identified mutated antigens within tumors, such as kinesin family member 2C (KIF2C)-mutated DNA polymerase alpha subunit B (POLA2) or erbb2 interacting protein (ERBB2IP) as TIL targets.⁹²^,⁹³ While the majority of clinical responses have been observed in melanoma patients receiving TILs enriched in CD8-positive T cells, a recent case report highlights the antitumor activity of CD4-positive TILs in a patient with metastatic cholangiocarcinoma.⁹³ Overall, the variable success of isolation and expansion of TILs from tumor samples other than melanoma means that adoptive transfer of TILs is not currently a viable option for pediatric malignancies.⁹⁴

Antigen-specific T Cells

Virus-specific T Cells.

The safety and efficacy of allogeneic and autologous virus-specific T cells has been evaluated in numerous clinical trials.⁹⁵^,⁹⁶^,⁹⁷^,⁹⁸^,⁹⁹^,¹⁰⁰^,¹⁰¹^,¹⁰²^,¹⁰³^,¹⁰⁴^,¹⁰⁵^,¹⁰⁶^,¹⁰⁷^,¹⁰⁸^,¹⁰⁹^,¹¹⁰^,¹¹¹^,¹¹²^,¹¹³^,¹¹⁴^,¹¹⁵

CMV-specific T Cells

Donor-derived CMV-specific T Cells. CMV is a human beta herpes virus associated with a short febrile illness during primary infection of immune-competent individuals. Like other herpes viruses, CMV establishes latency and an intact cellular immune system is critical for preventing CMV reactivation. Hence, HSCT and solid organ transplant (SOT) recipients are at great risk of CMV reactivation, which is associated with life-threatening infections including pneumonia, hepatitis, encephalitis, and gastroenteritis. Riddell et al.¹¹⁶ pioneered the infusion of CMV-specific T cells by preparing donor-derived CD8-positive T-cell clones that were activated and expanded by coculture with CMV-infected, autologous fibroblasts. Infusion of these T-cell clones proved safe and protected HSCT recipients against the reactivation of CMV. However, persistence of infused CD8-positive T-cell clones was dependent on the recovery of endogenous CD4-positive CMV-specific T cells.¹¹⁷ Later studies showed that co-infusion of CD4- and CD8-positive CMV-specific T-cell clones was sufficient to ensure persistence of the latter. In another clinical study, CMV-specific CD4-positive T-cell clones, generated with irradiated peripheral blood mononuclear cells (PBMCs) pulsed with CMV antigens as APCs, were given to haploidentical HSCT recipients.¹¹⁸ Infusions were safe, caused no GVHD, and induced the recovery of endogenous CMV-specific CD8-positive T-cell responses. Several other groups have prepared polyclonal CMV-specific T cells by different ex vivo culture methods and have documented their safety and efficacy.⁹⁵^,⁹⁶^,⁹⁷ In addition, polyclonal CMV-specific T cells have been given as part of multivirus-specific T cells with similar results (see section “Donor-derived multivirus-specific T cells”). Autologous CMV-specific T Cells. CMV-specific T cells may also be valuable for the treatment of CMV-associated malignancies. Several studies have demonstrated the CMV proteins pp65 and IE1 in 80% to 100% of GBM by immunochistochemistry.¹¹⁹^,¹²⁰ In addition, CMV DNA and/or viral particles were detected in some pp65- and IE1-positive cells.¹¹⁹^,¹²¹ CMV may promote the malignant phenotype of GBM cells by enhancing cell invasiveness and activating telomerase, but irrespective of any contribution to the disease process, these antigens should be targets for CMV-specific T cells. Several studies are exploring CMV vaccines for GBM, and one clinical study has evaluated the use of CMV-specific T cells as adjuvant post-chemotherapy for recurrent GBM.¹¹⁵ T-cell infusions were well tolerated and the median overall survival was 403 days, which is longer than expected for this incurable brain tumor. There are limited data about CMV antigen expression in pediatric high-grade glioma patients, but CMV-specific T cells can be readily prepared from glioma patients and there are few concerns in regard to “on target/off cancer” toxicity. Studies of the approach are therefore imminent.

EBV-specific T Cells

Donor-derived EBV-specific T Cells. EBV-associated malignancies provide an excellent model system to test and optimize cellular immunotherapies. EBV is a latent gamma herpes virus, and more than 90% of the world’s population is EBV seropositive, indicating prior exposure to the virus. During primary infection, EBV establishes lifelong latency in the memory B-cell compartment and the number of latently infected B cells within an individual remains stable over years. Healthy individuals mount a vigorous humoral and cellular immune response to primary infection. Although EBV-specific antibodies neutralize virus infectivity, the cellular immune response, consisting of CD4- and CD8-positive T cells, is essential for controlling primary and latent EBV infection. The virus produces B-lymphoproliferative disorders, leading to immunoblastic lymphoma in the absence of such control. In order to control viral reactivation in the immunocompromised host after HSCT and prevent EBV-PTLD, investigators prepared and then administered EBV-specific cytotoxic T cells from the stem cell donor to the recipient. In initial studies, PBMCs from the donor were infected with B95-8, the laboratory strain of EBV, to generate an EBV-transformed B-cell-derived lymphoblastoid cell line (EBV-LCL), which express all EBV-associated latency antigens and are phenotypically identical to PTLD cells. EBV-specific T cells were then generated by weekly stimulations of a second aliquot of PBMCs with these irradiated EBV-LCL. Our group has used these cells as prophylaxis in more than 100 HSCT recipients. None developed EBV-PTLD over 15 years of follow-up, compared with 5 of 42 (11%) patients enrolled on the same transplantation protocols who did not receive EBV-specific T cells.¹²² These T-cell infusions were safe and did not induce de novo GVHD. Moreover, of 13 patients treated with EBV-specific T cells for active EBV-PTLD, 11 achieved complete remission with no recurrence. Of the two nonresponders, one patient had extensive central nervous system disease and died 8 days after T-cell infusion. Analysis of the second patient who progressed after T-cell infusion revealed
a deletion in the EBV protein EBNA3B that deleted two of the dominant epitopes recognized by the infused T-cell product, highlighting a rare and previously unseen mechanism of immune escape to T-cell therapy in humans, which has subsequently been observed in other malignancies.¹²³^,¹²⁴ The safety and efficacy of donor-derived EBV-specific T cells has been confirmed by other investigators.⁶⁶^,¹²⁵

Autologous EBV-specific T Cells. The success of adoptively transferred EBV-specific T cells for the prophylaxis and treatment of EBV-PTLD in HSCT recipients has led to evaluation of this approach in SOT recipients who are at risk of developing EBV-PTLD. Infusions of autologous EBV-specific T cells into SOT recipients with elevated EBV-DNA load reduced viral loads and had antitumor effects in association with a rise in EBV-specific cellular immune responses; no patient had evidence of graft rejection.¹⁰⁹^,¹²⁶^,¹²⁷ Unlike treated HSCT recipients, EBV-specific T cells in SOT recipients persisted only transiently, likely due to the chronic immunosuppressive environment of these patients. Investigators are now genetically modifying T cells to render them resistant to immunosuppressive drugs, such as FK506 or rapamycin to overcome this obstacle.¹²⁸^,¹²⁹^,¹³⁰

EBV is also associated with subset of non-Hodgkin and Hodgkin lymphomas in immunocompetent individuals. In addition, more than 95% of nasopharyngeal carcinomas (NPC) are EBV positive. In contrast to the EBV antigens expressed in EBV-PTLD, only a limited number of EBV-derived antigens, including EBNA1, LMP1, LMP2, and BARF1, are expressed by EBV-positive lymphomas and NPC but these may nevertheless provide targets for T-cell immunotherapy.

Autologous EBV-specific T cells generated using EBV-LCLs as antigen-presenting cells have been given to patients with multiply relapsed EBV-positive Hodgkin lymphoma and to individuals with minimal residual disease after autologous HSCT.¹¹¹ No immediate toxicities were seen, and infused T cells localized to tumor sites. Immunological studies showed an increase of LMP2- and EBV-specific cellular immunity in peripheral blood after T-cell infusion in some patients. A subset of patients received T cells that were gene marked and could be detected in blood and at tumor sites for up to 12 months post-infusion. Because EBV-specific T cells generated with EBV-LCLs contain relatively few T cells specific for the viral antigens actually expressed in lymphoma, subsequent studies focused on enriching the frequency of T cells specific for the subdominant EBV antigens LMP1 and LMP2.¹¹² Fifty patients with EBV-positive non-Hodgkin or Hodgkin lymphoma have been infused with LMP2- or LMP1/LMP2-specific T cells. Twenty-eight of 29 patients who were at high risk for recurrent disease and received T cells as adjuvant therapy remained in remission for more than 3 years post-T-cell infusion. Of 21 patients with relapsed or resistant disease at the time of T-cell infusion, 13 had clinical responses, including 11 complete responses (CRs). Activation of T cells specific for endogenously expressed, nonviral TAA (i.e., epitope spreading), which has been shown to correlate with the antitumor activity of cancer vaccines, was observed only in patients achieving clinical responses.

EBV-specific T cells have also been given as therapy for NPC. Comoli et al.¹¹⁰ reported 10 patients who received between 2 and 23 T-cell infusions. Two patients had a partial response for 3 to 4 months, and 4 patients had stable disease for up to 15 months. We treated 15 NPC patients with active disease; 5 had a CR/CRu, 2 patients had a PR, and 2 patients had stable disease.¹¹³ More recently, in a Phase II clinical study in Singapore, up to six doses of EBV-specific T cells were given as consolidation therapy to 35 patients with advanced-stage NPC, who had received four cycles of gemcitabine and carboplatin.¹³¹ With a median follow-up of 29.9 months, the 2-year overall survival was 62.9%, which is the highest published overall survival rate for this patient population. Clearly, these findings need to be confirmed in a larger, randomized clinical study. On another clinical study, NPC patients received a cell product that was enriched in T cells specific for the EBV antigens LMP1, LMP2, and EBNA1.¹¹⁴ When cells were given as an adjuvant, infused patients (n = 16) had prolonged median overall survival in comparison with patients (n = 8), who did not receive cells (523 vs. 220 days).

Only gold members can continue reading. Log In or Register to continue