Although risk factors for incident CRC are relatively well-established, less is known about factors associated with CRC survival. At present, the strongest known predictor of CRC prognosis is stage; however, there is considerable heterogeneity in survival among individuals with the same stage at diagnosis. To extend our understanding of CRC pathogenesis and to potentially direct treatment, there remains a need to identify markers of CRC prognosis. Recent studies have suggested that the presence of tumor infiltrating lymphocytes (TILs) in CRC tissue, indicative of a host immune response to the tumor, may be associated with favorable survival outcomes. While biologically plausible, most prior studies exploring this hypothesis have been limited to evaluations of T-cell counts and density based on standard immunohistochemical (IHC) assays. TIL clonality is likely to be of even greater prognostic significance than TIL density, but cannot be measured based on IHC. Using the novel ImmunoSEQ assay, Robins et al. have recently shown that TIL clonality is an independent biomarker of clinical outcome for ovarian cancer and melanoma, and have demonstrated that the diversity of T-cell receptor sequences from TILs is markedly lower in CRC tissue relative to peripheral blood. This latter finding implies that CRC induces a specific adaptive immune response that merits further investigation.

The objective of this proposal is to characterize the prognostic and epidemiologic implications of TILs in colorectal. To address this objective, we propose the following aims:
Aim 1. To quantify the presence and clonality of TILs in colorectal cancer
Aim 2. To evaluate the association between TIL quantity and clonality measures with:
2a. Colorectal cancer survival outcomes (i.e., overall and disease-specific survival)
2b. Basic patient and tumor attributes of colorectal cancer known to be associated with prognosis (i.e., age at diagnosis, tumor site, stage at diagnosis)

In pursuit of these Aims, we will utilize the rich resources of the Genetic Epidemiology of Colorectal Cancer Consortium (GECCO) which include the cohort of colorectal cancers from PLCO. Data from studies participating in GECCO has previously been harmonized for a variety of patient characteristics (e.g., sex, age at diagnosis, smoking history), basic tumor attributes (e.g., tumor site, stage at diagnosis) and, as available, survival outcomes (i.e., disease-specific and overall survival). In the evolution of the GECCO resource, increasing research focus is being placed on the tumor genome and tumor microenvironment. The existing and accumulating resources of GECCO provide a unique context for quantifying and characterizing the prognostic and epidemiologic impact of TILs in CRC in a well-powered fashion. The proposed project would mark the largest study of TIL quantity and clonality in CRC to date. Insights gained through this project could have significant clinical and translational implications, and could ultimately be used in guiding targeted cancer therapy regimens and post-diagnostic surveillance strategies for individuals with CRC.

Andrew Chan (Massachusetts General Hospital and Harvard Medical School)
Amanda Phipps (Fred Hutchinson Cancer Research Center)
Ulrike Peters (Fred Hutchinson Cancer Research Center)
Yin Cao (Massachusetts General Hospital and Harvard Medical School)