Background The preclinical evaluation of novel immune modulators for cancer treatment remains a challenge, as models require both, engraftment of human tumor cells and a compatible human immune cells. In previous experiments, we have demonstrated, that we can use either peripheral blood mononuclear cells (PBMC) or hematopoietic stem cells (HSC) to establish a humanized immune system with functional T-, B-, and NK cells, monocytes, and dendritic cells. However these models are limited by rarely matching HLA isotypes between tumor and immune cells. In this case study, we established a patient-derived xenograft (PDX) model from a patient with Head and Neck squamous cell cancer (HNSCC). After engraftment of HNSCC PDX, patients PBMC were used to humanize mice. By this procedure we successfully generated a patient-specific human tumor-immune cell model in mice with 100% HLA-match. Model development included the comparison of PDX engraftment on mice with either HLA-matching or non HLA-matching PBMC’s and purified T cells from different donors. Furthermore, these effects were investigated on humanized mice generated with HSC. Finally, we further validated the model by comparing treatment effects with the checkpoint inhibitor Nivolumab in the autologous immune cell PDX model with heterologous models.

Methods The HNSCC PDX was transplanted on NOG mice. After tumor engraftment mice were randomized in 6 groups, receiving PBMCs by i.v. transplantation either from the patient or from 5 well characterized donors (PDX patient PBMCs - 100% HLA matching, 5 donors with different HLA matching). In the last step, PDX were transplanted on humanized mice generated from 5 different HSC donors. Blood and tumor samples were analysed by FACS and IHC for immune cell infiltration and activation.

Results In the autologous huPBMC model, no interference with the proliferation of HNSCC PDX was seen. However, on mice humanized with donor PBMC’s with a high HLA match, a strong stimulation of tumor proliferation compared to non-humanized mice was observed. On humanized mice, generated from 5 different HSC donors, HLA-matching seem to have a lower influence on engraftment. On mice humanized with PBMC from different donors, we observed a correlation of treatment effects with HLA match, with strong tumor growth inhibition in the mice with the best match. In the PDX tumors, infiltrating immune cells were detected by FACS and IHC analyses.

Conclusions We developed a humanized immune-PDX model enabling appropriate preclinical translational research on tumor immune biology and the evaluation of new therapies and combinations, as well as the identification and validation of biomarkers for immune therapy. Furthermore, results showed a correlation between immune therapy effects and HLA matching in preclinical models.

Disclosure Information M. Stecklum: A. Employment (full or part-time); Significant; EPO - Experimental Pharmacology & Oncology Berlin-Buch GmbH. K. Klinghammer: None. A. Wulf-Goldenberg: A. Employment (full or part-time); Significant; EPO - Experimental Pharmacology & Oncology Berlin-Buch GmbH. B. Brzezicha: A. Employment (full or part-time); Significant; EPO - Experimental Pharmacology & Oncology Berlin-Buch GmbH. K. Jöhrens: None. J. Hoffmann: A. Employment (full or part-time); Significant; EPO - Experimental Pharmacology & Oncology Berlin-Buch GmbH.