Historically metastatic melanoma has been associated with poor prognosis. The advent of immune checkpoint inhibitors targeting CTLA-4 and PD-1 have greatly improved patient survival in metastatic melanoma, but a significant proportion of patients fail to respond to therapy. To explore the effect of genetic background on immune checkpoint inhibitor efficacy, we have developed a tumor immunotherapy model using genetically heterogeneous Diversity Outbred (DO) mice and the C57BL/6 syngeneic B16F0 melanoma model. To implement genetic diversity, DO mice were crossed with C57BL/6 to generate (C57BL/6xDO)F1 mice. Untreated (C57BL/6xDO)F1 mice (n=34) reliably develop B16F0 tumors after subcutaneous inoculation, with some variation in tumor latency. To test the role of genetics in response to ICI, (C57BL/6xDO)F1 mice (n=142) were treated with combined anti-PD1/anti-CTLA-4 on days 3, 6, and 10 after inoculation with 2×105 B16F0 cells. Mice receiving therapy show wide variation in tumor development of up to 65 days (mean 20.86 +/− 11.04, CV 52.94%), with 19 never developing tumor by day 88. Re-challenge of tumor free mice confirmed adaptive response to tumor. Preliminary Genome Wide Association shows suggestive associations on chromosomes 8, 12, and 13. The locus on chromosome 13 contains a family of related genes with known function in autoimmune conditions and expression in several immune cell types. QTL effects show the NZO genotype is a positive driver at this locus and C57BL/6 is a negative driver, reflecting known ICI resistance in the C57BL/6 strain. These results demonstrate host genetic background significantly contributes to response to ICI and identifies genomic loci associating with primary resistance. NIH R37 CA220482.

Historically metastatic melanoma has been associated with poor prognosis. The advent of immune checkpoint inhibitors targeting CTLA-4 and PD-1 have greatly improved patient survival in metastatic melanoma, but a significant proportion of patients fail to respond to therapy. To explore the effect of genetic background on immune checkpoint inhibitor efficacy, we have developed a tumor immunotherapy model using genetically heterogeneous Diversity Outbred (DO) mice and the C57BL/6 syngeneic B16F0 melanoma model. To implement genetic diversity, DO mice were crossed with C57BL/6 to generate (C57BL/6xDO)F1 mice. Untreated (C57BL/6xDO)F1 mice (n=17) reliably develop B16F0 tumors after subcutaneous inoculation, with some variation in tumor onset (8.2 +/− 4.1 days, CV 49.9%). In a pilot immunotherapy study, 20 (C57BL/6xDO)F1 mice were treated with anti-PD-1/anti-CTLA4 on days 12, 14, and 16, after tumors were palpable (mean onset 6.8 +/− 4.6 days, CV 68.2%). Tumor tissue was collected on day 19 and CD8 T cell infiltration was analyzed by IHC and qPCR. Mice with reduced tumor growth showed an increase presence of CD8+ cells and IFN-g expression. To further test the role of genetics in response, a second cohort of C57BL/6xDO)F1 mice (n=95) was treated with combined anti-PD1/anti-CTLA-4 on days 3, 6, and 10 after B16F0 inoculation. In this larger trial, mice receiving therapy show wide variation in tumor onset up to 61 days (mean 13.6 +/− 8.9, CV 65.2%), with 10 never developing tumor by day 88. This data further suggests that the genetic background significantly influences response to treatment and validates further testing of this model and subsequent Quantitative Trait Locus analysis. Supported by NIH R37 CA220482.

Advances in immune modulation for cancer and inflammation warrant parallel development of imaging technologies to gauge in situ immune activity. We have developed a toolbox of full-length monoclonal antibody (mAb) positron emission tomography (immunoPET) tracers by desferrioxamine conjugation and 89Zr radiolabeling. Tracer uptake is reported as % injected dose/gram (%ID/g) tissue. Because mAbs to cell surface molecules often facilitate depletion, we tested enzymatic removal of the Fc glycan of CD8 mAb 2.43 to disrupt Fc:FcR interaction and prevent depletion. In the CT26 tumor model, CD8 immunoPET showed similar uptake between WT and deglycosylated tracers. However, unlike WT 2.43, the deglycosylated tracer did not deplete peripheral CD8 T cells, supporting this approach. In addition to surface markers, tracers that define cell function may have great utility. Checkpoint blockade therapy (anti-PD-1/anti-CTLA-4) in CT26-bearing mice shows IFNg immunoPET tracer uptake in 5/5 treated (16.5 ± 3.6 %ID/g) and 2/5 untreated (12.1 ± 4.8 %ID/g) tumors, possibly indicating baseline CT26 immunogenicity. To evaluate antigen presenting cell activation, an IL-12-specific tracer was also developed. Mouse mammary tumor TUBO-bearing mice showed enhanced tracer uptake after intratumoral injection of adenovirus encoding GM-CSF (2.6 ± 0.7 %ID/g) versus controls (0.4 ± 0.4 %ID/g). IL-12 PET was also tested for applications outside of cancer by mimicking inflammation with i.m. LPS injection. LPS-treated mice showed localized, elevated uptake at the site of injection (3.7 ± 0.7 %ID/g) versus controls (0.7 ± 0.1 %ID/g). These results support the use of immunoPET imaging for non-invasive monitoring of immune activity in situ. NIH R37 CA220482.