Peripheral CD4 T cell differentiation in lung cancer development

Late diagnosis of non-small cell lung cancer (NSCLC) is a major cause of cancer mortality. Current lung cancer screening methods are burdensome and/or ineffective, underscoring the need for improved non-invasive early detection technologies. The T cell response represents a naturally amplified, measurable and specific signal which could be exploited for early cancer detection. We and others recently discovered that neoantigens elicit distinct patterns of T cell differentiation within NSCLC tumours, implying that pre-invasive T cell dynamics could be leveraged to detect pulmonary carcinogenesis. CD4 T cells specifically have been shown to be amongst the earliest and most dynamically remodelled lymphocytes in pre-invasive lung neoplasia. In this thesis, I ask how and when the peripheral immune system senses and responds to the earliest stages of lung carcinogenesis, and whether these changes are measurable in peripheral blood to guide development of immune-focused early detection technologies. To address this, I profiled systemic immune remodelling across cohorts of pre-invasive lung squamous carcinoma (LUSC) and early-stage lung adenocarcinoma (LUAD), via flow cytometry, TCR sequencing, and DNA methylation. Flow cytometry revealed that early carcinogenesis is associated with skewing of the blood CD4 T cell compartment, including a consistent enrichment of effector CD39+ Tregs across all cohorts. In LUAD, this was concomitant with increases in markers of tissue adaptation and tumour experience (CD103, CCR8, CXCR4, GzmK) and parallel shifts in early to late conventional CD4 T cell differentiation. DNA methylation profiling demonstrated widespread hypomethylation at immune-regulatory and activation-associated loci with stable epigenetic imprinting of effector Treg and recirculating S1PR1+ Treg programmes in LUAD, consistent with an immune state that is activated yet functionally restrained. TCR sequencing indicates clonal reshaping most apparent in pre-invasive and early LUSC. Finally, a multi-omic machine learning framework integrating flow cytometry, methylation and TCR features yielded a compact, highly discriminative signature dominated by effector Treg phenotypes and methylation markers that distinguish early LUAD from high-risk controls. Together, these results define systemic immune remodelling by CD4 T cell differentiation skewing and effector Treg expansion as hallmarks of early lung cancer. These data support a model in which tumour-derived signals propagate beyond the lesion to reshape circulating immunity, establishing a tolerogenic environment that may constrain anti-tumour responses even before overt invasion. Crucially, these changes are measurable in peripheral blood, highlighting translational potential for non-invasive early detection of not only lung cancer but pre-invasive disease, and reframing early lung cancer development as a systemic immunological process in which CD4 T cell regulatory dominance emerges early and tracks carcinogenesis. These signatures may provide a new window for interception of disease years prior to typical clinical diagnosis.