A multiomic T cell differentiation index (TEDI) for non-invasive early detection of lung and renal cancer

Background Tumour-specific T cell differentiation occurs during early carcinogenesis, leading to marked changes in the host TCR repertoire, T cell phenotype and T cell epigenetics.

However, it remains unclear whether this process can be measured in liquid biopsies as a novel immunological marker of preinvasive disease or early-stage lung and renal cancer. We recently reported a program of neoantigen-driven, intra-tumoral T cell differentiation in lung cancer.

We have since discovered that the presence of primary or recurrent lung cancer (Reading et al in preparation) and high-grade preinvasive lesions of the airways (Reading & Janes laboratories, unpublished) is marked by a significant shift in CD4 and CD8 T cell differentiation in the blood.

Our preliminary data suggests that systemic T cell differentiation is also skewed in the presence of malignant vs benign renal cell masses.

Aims Here, we will examine if systemic T cell differentiation skewing can be used as the basis for a novel, non-invasive multi-cancer early detection tool for high and low mutational burden tumours. To do so we will 1. Validate that systemic T cell differentiation skewing can detect high-grade pulmonary neoplasia. 2.

Test and validate whether systemic T cell differentiation skewing detects LDCT screen-detected lung nodules, small renal masses or renal cell carcinoma. 3.

Use multi-omic T cell differentiation data from 1-2 to train and validate a multivariate classifier of preinvasive/ invasive lesions and assess whether this can be used to forecast disease progression.

Methods This project will leverage cryopreserved PBMCs from the UCLH surveillance (High-grade lung neoplasia), SUMMIT/NHS lung screening study (lung nodules) and NEST/Renal cancer Biobank (RCC) cohorts.

In each setting we will measure T cell differentiation state via; i) T cell phenotype (31 parameter spectral cytometry), ii) TCR sequencing and iii) epigenetics (ATAC-seq). Metrics from each assay will be ranked by feature importance.

Top ranking features will be used with orthogonal clinical data to train/validate a multivariate machine learning classifier of disease in each histology.

Impact These data will be used to determine the sensitivity, specificity, and AUC of univariate and multivariate systemic T cell differentiation analysis for renal and lung cancer early detection.

If preliminary results are confirmed/extended, we will patent the classifier and initiate prospective studies to formally evaluate efficacy in lung and renal cell carcinoma.

We aim to spin out a biotech company that will use these results to engineer a clinically viable, immune-focused mutli-cancer early detection blood test.