After decades of failure, immunotherapy has recently begun to revolutionise the treatment approach taken for several cancer types. Harnessing the power of the immune system to seek and destroy transformed cells has the potential to achieve long-term remission and prevention of cancer recurrence.
CAR T-cell immunotherapy success in ALL
One of the roles of the immune system is to identify cancer cells through the recognition of tumour antigens and to eliminate these cells through a process of immunosurveillance. Unfortunately, due to the immunosuppressive nature of the tumour micro-environment, cancer cells can often escape this process, expand and become a poor target of immune responses. Immune checkpoint blockade is a logical approach that helps to re-invigorate tumour-specific immune responses, impacting meaningfully in the treatment of several solid tumours such as melanoma and non-small cell lung cancer.
Perhaps even more remarkable has been the recent success of cellular therapies in which patient derived T-cells are genetically retargeted using chimeric antigen receptors (CARs). When applied to the treatment of patients with refractory acute lymphoblastic leukaemia (ALL), CAR T-cell immunotherapy achieves complete remission of disease in over 80% of cases. In that setting, T-cells are re-targeted against the ubiquitous B-cell antigen, CD19. The efficacy of CAR T-cell immunotherapy of ALL is unprecedented for a new cancer medicine but a key question is whether comparable success can be achieved in other cancer types.
Construction of chimeric antigen receptors
Chimeric antigen receptors are membrane spanning fusion molecules that couple the direct binding of a specific cell surface-associated target to the delivery of an immune cell activating signal. Targeting is most commonly achieved using a single chain antibody fragment, although peptides and ligand derivatives may be used alternatively. This element is separated from the signalling domain by a hinge/spacer and transmembrane domain.
Chimeric antigen receptors are delivered by gene transfer, most commonly using retroviral or lentiviral vectors. The most commonly used host cells are patient-derived peripheral blood T-cells, although these fusions can also be expressed in other immune cell subsets (e.g. NK cells), or even in stem cells. Owing to their antibody-like binding properties, CAR T-cells bypass the need for HLA restriction and circumvent immune evasion mediated by tumours that downregulate HLA class I expression at the cell surface. T-cells that are re-targeted using CAR molecules mediate anti-tumour responses through direct cytotoxicity and the release of a panoply of immunomodulatory molecules.
Several other examples of on-target off tumour toxicity have been reported which have resulted in death or significant organ damage, highlighting the need for careful target validation.
Construction of CARs remains a largely empiric process since all elements of the fusion may require refinement and optimisation. Based upon the endodomain of these molecules, several CAR generations have been described in which a CD3 zeta (or functional equivalent) module is found alone (first generation) or is coupled to one (second generation) or more co-stimulatory elements.
First generation CARs were developed by Eshhar, establishing proof of concept for this emerging technology. However, clinical efficacy required the additional provision of co-stimulation by CD28 or 4-1BB, leading to enhanced proliferation and reduced anergy or activation induced cell death. Second generation CARs were originally described by Finney and colleagues and were first shown to be functional in human T-cells by Maher and Sadelain. The alternative 4-1BB-based second generation CAR platform was originally developed by Campana and Imai. Both second generation variants have demonstrated compelling efficacy in patients with ALL, treated at multiple independent clinical centres. Third generation fusions have also been described in which two co-stimulatory modules are present, although improved functionality compared with second generation iterations remains uncertain.
Limitations of CAR T-cell immunotherapy
Despite the success of this technology in patients with ALL, response rates have somewhat less impressive in patients with more indolent B-cell malignancies. Most disappointingly however, results in patients with solid tumours remain unimpressive. Lack of efficacy in this context has been ascribed to several challenges, including the paucity of safe tumour-specific targets, inadequate homing of CAR T-cells to tumour deposits and susceptibility of these cells to suppression within the tumour microenvironment.
The efficacy of CAR T-cell immunotherapy of ALL is unprecedented for a new cancer medicine but a key question is whether comparable success can be achieved in other cancer types.
CAR T-cell immunotherapy is also limited by potentially severe toxicity. In patients treated with CD19-targeted CAR T-cells, B-cell depletion commonly leads to impaired antibody-forming capacity and hypogammaglobulinemia. This predictable ‘on-target off tumour’ toxicity can be mitigated by immunoglobulin replacement therapy. However, many patients also develop cytokine release syndrome, which is a profound systemic inflammatory response driven by the interaction between CAR T-cells and cells of the mononuclear phagocyte lineage. Treatment of cytokine release syndrome may require intensive medical support including the administration of pressor agents and mechanical ventilation. In addition, biological agents such as the anti-interleukin 6 receptor antagonist, tocilizumab, can prove dramatically effective in some cases. Neurotoxicity is also emerging as a common and generally transient toxicity and may relate to the ability of CAR T-cells to enter the central nervous system. This can manifest with deficits such as impaired speech, vision, confusion or seizure activity. Several other examples of on-target off tumour toxicity have been reported which have resulted in death or significant organ damage, highlighting the need for careful target validation.
Immunotherapy using CAR T-cells imposes high manufacturing costs since cell products are generally autologous (e.g. patient-derived) and good manufacturing processes and manufacturing facilities are required. In general, patients undergo a leukapheresis to provide sufficient cells from which to generate a product; although some investigators have developed systems to allow the use of whole blood as an alternative. T-cells are generally activated ex vivo and, following gene transfer, cell products are expanded over a period of 10–14 days. Prior to administration, quality control assays must be performed on the CAR T-cell product, followed by a concentration and sometimes a cryopreservation step. This allows the shipping of cell products from centralised manufacturing facilities to the site of the patient.
Until recently, CAR T-cell immunotherapy has largely been an academic activity. However, there is increasing commercial interest in this sector, leading to the launching of several spin out companies and partnerships between academic centres and pharmaceutical companies. To facilitate more widespread rollout of this technology, there is increasing interest in the development of universal cell therapy products. Examples include the use of allogeneic CAR T-cells that have been genetically engineered to minimise risk of graft versus host disease and immune-mediated rejection.
We have reached a truly exciting stage in the development of this new modality of cancer therapy. To make major impact, it will be necessary to adapt this technology for the treatment of solid tumours. Success in that arena will certainly prove transformative to cancer medicine.
Research technician and GMP production scientist
Consultant and Senior Lecturer in Immunology
CAR Mechanics Group, Division of Cancer Studies, King’s College London
Image credit: Shutterstock