Designing and combining chimeric antigen receptors for improved cellular immunotherapy of hematological malignancies

Cancer remains the primary cause of premature death in the world. Hematological malignancies collectively account for 1.27 million of those cases, of which the large majority are of B-cell origin. Despite significant therapeutic advancements over the past decades, many patients still experience relapse and treatment resistance. Chimeric antigen receptor (CAR-)T-cell therapy, in which T cells are armored with a synthetic tumor-specific receptor, has been able to provide hope for many of these patients. The first part of this thesis provides the rationale behind and development of a multi-target CAR therapy for B-cell malignancies. We summarize the clinical outcomes of pivotal trials of the most advanced CD19 and BCMA CAR-T-cell products that led to the commercial approval of several CAR-T-cell products. Moreover, a meta-analysis was performed on clinical efficacy and safety data of BCMA CAR-T cells. We discovered that both patient and product characteristics had a notable impact on therapeutic potency and safety. These findings could support the design of future BCMA CAR-T-cell products and clinical trials. Unfortunately, our analysis also highlighted the lack of a plateau in progression-free survival curves of BCMA CAR-T-cell therapy, highlighting the need for additional or different interventions. For this reason, we investigated a BCMA/CD19 dual-CAR strategy to overcome antigen escape. As product quality, and manufacturing cost and length of autologous CAR-T cells are suboptimal, we chose to use the FDA-approved NK-92 cell line as a homogeneous, readily available cell source to test our dual-CAR strategy. we demonstrated efficient production of off-the-shelf BCMA/CD19 dual-CAR NK-92 that show potency against antigen escape tumor models. The second part of the thesis explores the modular design of CARs. This modularity theoretically allows quasi-infinite combinations of CAR domains, yet CARs currently used in the clinic are composed of only a handful of components. Hypothesis-driven, low throughput exploration of novel domains is resource-intensive and time-consuming. Existing high throughput workflows largely focusses on antigen binding domains and costimulatory domains, leaving a large knowledge gap on structural domains such as the hinge domain. Therefore, we show the preliminary development of bulk and arrayed screens of medium-size CAR hinge domain libraries. We believe that the combination of these methodologies will expand our understanding of structural and functional properties of the hinge domain. The development of these high throughput screens is expected to ultimately facilitate rational design of novel CAR therapies.