Radiation-induced bystander and abscopal effects: important lessons from preclinical models

Abstract

Radiotherapy is a pivotal component in the curative treatment of patients with localised cancer and isolated metastasis, as well as being used as a palliative strategy for patients with disseminated disease. The clinical efficacy of radiotherapy has traditionally been attributed to the local effects of ionising radiation, which induces cell death by directly and indirectly inducing DNA damage, but substantial work has uncovered an unexpected and dual relationship between tumour irradiation and the host immune system. In clinical practice, it is, therefore, tempting to tailor immunotherapies with radiotherapy in order to synergise innate and adaptive immunity against cancer cells, as well as to bypass immune tolerance and exhaustion, with the aim of facilitating tumour regression. However, our understanding of how radiation impacts on immune system activation is still in its early stages, and concerns and challenges regarding therapeutic applications still need to be overcome. With the increasing use of immunotherapy and its common combination with ionising radiation, this review briefly delineates current knowledge about the non-targeted effects of radiotherapy, and aims to provide insights, at the preclinical level, into the mechanisms that are involved with the potential to yield clinically relevant combinatorial approaches of radiotherapy and immunotherapy.

Fig. 1. Schematic overview of local and distant effects triggered by tumour irradiation.

At the heart of the primary lesion that is irradiated (panel on the left), two local effects can be distinguished: first, bystander effects occur between high-dose-targeted cells (dark orange) or low-dose-targeted cells (light orange) and non-irradiated cells (blue); second, cohort effects occur between high-dose-targeted cells and low-dose-targeted cells. Whether/how non-irradiated cells can influence the outcome of irradiated cells (depicted with a question mark) remains to be determined. Irradiation induces immunogenic cell death in cancer cells and the subsequent release of tumour-associated antigens (TAAs) (pink dots), thereby activating the immune system, especially antigen-presenting cells (APC, in purple) and macrophages (in pink). APCs then cross-present TAAs to T cells in draining lymph nodes. As a result, polyclonal antigen-specific T cells are primed to attack tumours located within the irradiated field as well as those in distant locations. This distant radiation-induced effect is termed an abscopal effect (panel on the right). Exosomes (in green) are novel mediators thought to participate in these non-targeted effects locally and at distant sites.

Fig. 2. Preclinical experimental strategies for efficient radio-immunotherapy combinations in clinical routine.

a In vitro investigations using 2D cell cultures and next-generation sequencing: (i) comparative analysis between co-culture assays without contact and solely monocultures (one irradiated [blue] and one non-irradiated [dark orange]); (ii) co-culture assays with contact using partial irradiation through grids (black boxes) and following cell sorting; (iii) co-culture assays with contact using partial irradiation through grids and with two doses (yellow and orange ‘thunders’). b In vitro investigations using 3D tumour models (in blue) containing endogenous immune cells (antigen-presenting cells [purple stars], lymphocytes [green circles]) for testing the radio-immunotherapy combination, in order to decipher the molecular mechanisms by RNA sequencing, to identify the immune repertoire by mass cytometry (CyTOF) and to develop high-throughput technologies for drug screening, treatment schemes, etc. c In vivo investigations relying on (i) orthotopic, (ii) local autochthonous and (iii) disseminated autochthonous tumour mouse models for testing the radio-immunotherapy combination.

Fig. 3. Applications of the 2D, 3D and mouse models in discovery cancer research and translational oncology.

a Co-cultures, b mouse or patient-derived tumour organoids and c genetically engineered mouse models or patient-derived tumour xenografts. Advantages (+) and disadvantages (−) of each strategy are highlighted in green and red boxes, respectively. Human specimens might also serve as a study model using tumour organoids (condition (b)) or xenografts (condition (c)).