Immunotherapy Modeling in Organoids Co-preserving Tumor and Infiltrating Immune Compartments

PROJECT SUMMARY The immune system remarkably distinguishes between self and non-self/self-aberrant antigens, affording exquisite anti-tumor specificity and inhibition of tumorigenesis. However, tumor immunosurveillance is unfortunately opposed by tumor cell evasion of the immune response. Immune checkpoint blockade (ICB)

targeting PD-1, PD-L1, CD40 and others, as well as adoptive cell transfer (CAR-T, bulk TILs) favorably modulate this equilibrium for therapeutic benefit. However, response rates are often incomplete, progressive disease is common, and predictive biomarkers are suboptimal. The development of next-generation immunotherapies has been hindered by a lack of in vitro models

that functionally recapitulate syngeneic interactions between tumor and infiltrating immune cells. In response, we have developed organoid methods that culture primary human tumor biopsies together with their infiltrating immune components as a cohesive unit without reconstitution. These “patient-derived tumor organoids” (PDO)

preserve tumor cells alongside endogenous T, B, NK cells and macrophages, robustly recapitulate the T cell receptor clonotype repertoire of the original tumor, and crucially, manifest tumor-infiltrating lymphocyte (TIL) expansion, activation and tumor cell killing in response to anti-PD-1/PD-L1 therapeutic antibodies (Cell, 2018).

The PDO system thus represents a holistic organoid model of human tumor-immune interactions. Here, we leverage the PDO technique to investigate immunotherapeutic mechanisms and treatments in PD-1-responsive cutaneous squamous cell carcinoma (cSCC) and melanoma, exploiting pre- and post-treatment human biopsies and mouse models. Aim 1 hypothesizes that checkpoint inhibition

induces a complex and sequential network response involving immune-tumor and immune-immune crosstalk. Thus, Aim 1 employs the ability to perform serial time-course sampling of PDOs to define a single cell RNA-seq network cellular crosstalk model of the early anti-PD-1-stimulated anti-tumor immune response over multiple

acute time points typically inaccessible to clinical biopsies performed after months. Importantly, comparison of this immune propagation in responding versus non-responding mouse and human organoids will define nodal points conferring resistance. Aim 2 improves bulk TIL adoptive transfer immunotherapy by using PDOs as living

bioreactors to enrich tumor-reactive mouse and human melanoma TILs by anti-PD-1 checkpoint inhibition, followed by testing of enhanced anti-tumor activity in vitro and in vivo. Lastly, Aim 3 performs a co-treatment trial comparing anti-PD-1 responses of pre-treatment biopsy cSCC PDOs to clinical outcomes. Further, post-

treatment biopsy PDOs are re-challenged with anti-PD-1 and a novel agent inactivating PD-1 by dephosphorylation. We thus utilize the holistic PDO model preserving endogenous tumor epithelial and immune components en bloc to investigate and improve cancer immunotherapy via our team of Calvin Kuo (organoids),

Mark Davis and Chris Garcia (tumor immunology) and Anne Chang and Dimitri Colevas (cSCC clinicians).