Date of Award

January 2025

Document Type

Thesis

Degree Name

Medical Doctor (MD)

Department

Medicine

First Advisor

David A. Braun

Abstract

Renal cell carcinoma (RCC) is classically considered an immunogenic tumor, with high immune cell infiltration and both historic and contemporary responses to immune-based therapies. Currently, treatment regimens involving immune checkpoint inhibitors (ICIs) are the standard of care for advanced disease, but many patients do not benefit from therapy and the molecular mechanisms influencing response to ICIs in RCC are not completely understood. Recent molecular characterization of tumors from patients with clear cell RCC enrolled in the CheckMate 009, 010, and 025 clinical trials (evaluating nivolumab, an anti-PD-1 ICI) has indicated that typical correlates of immunotherapy response across other cancer types, such as somatic mutation burden, neoantigen load, and CD8+ T cell infiltration, do not appear to predict treatment outcomes.

The R2000 project is an ongoing multi-institutional effort to perform a molecular analysis of RCC with unprecedented statistical power by aggregating and harmonizing the analysis of whole exome and RNA sequencing data from more than 2,000 patients across various clinical studies, providing new insights into RCC biology and illuminating potential mechanisms underlying therapeutic response and resistance. Initial analysis of R2000 exome data has identified 118 significantly mutated genes and 57 chromosomal regions subject to recurrent copy number changes, many of which had not been previously associated with RCC.

In this thesis, we present an exploratory analysis of tumor transcriptomes (analyzed in a rigorous, harmonized fashion starting with raw RNA-Seq read files) from 2,026 patients in the R2000 cohort, 1,155 of whom have corresponding tumor mutational and copy number profiles available from prior exome analysis. We leveraged immune deconvolution methods, pathway enrichment analysis, and machine learning-based classification of tumors into established molecular subtypes to characterize tumor biology and investigate associations between somatic variants, phenotypes, and clinical outcomes.

Our analysis revealed that tumors with 14q loss, a common genomic event in clear cell RCC and established driver of aggressive disease, had numerous changes in intratumoral immune cell populations and displayed features of CD8+ T cell exhaustion and dysfunction. Although we reproduced the finding that tumoral CD8+ T cell infiltration did not generally predict sensitivity to ICIs in the original CheckMate studies, we observed that increased CD8+ T cell populations did correlate with improved ICI response across more modern, first-line immunotherapy clinical trials; this observation suggests that, while the effects of cytotoxic T cell infiltration on ICI response are complex and heterogeneous, RCC at least somewhat aligns with the classical paradigm that infiltration potentiates response to immune checkpoint blockade. Additionally, we identified that a molecular subtype of RCC enriched for metabolic and cell cycle dysregulation was significantly associated with loss of PRDM10, a gene that was recently implicated in an extremely rare hereditary RCC syndrome but had not been previously identified as a recurrently mutated gene in sporadic RCC prior to the R2000 project. This molecular subtype was also profoundly depleted of canonical clear cell RCC drivers (including VHL mutations and 3p deletions) despite displaying clear cell histology, suggesting that it may represent a novel, variant class of RCC. These findings illuminate potential determinants of therapeutic response in RCC and highlight the power afforded by this harmonized multi-cohort analytical approach.

Comments

This thesis is restricted to Yale network users only. This thesis is permanently embargoed from public release.

Download

Share

COinS