McGranahan & Swanton (2017) — Clonal Heterogeneity and Tumor Evolution

Bibliographic Reference

McGranahan, N., & Swanton, C. (2017). Clonal heterogeneity and tumor evolution: Past, present, and the future. Cell, 168(4), 613—628. https://doi.org/10.1016/j.cell.2017.01.018

Core Argument

Intratumor heterogeneity (ITH) — the presence of genetically and phenotypically distinct subclonal populations within a single tumor — is a fundamental property of human cancers that arises through branching Darwinian evolution. ITH provides the substrate for adaptation under selective pressures (including therapy and immune surveillance), making it a critical determinant of drug resistance, metastasis, and patient outcome. Characterizing the clonal architecture of tumors — distinguishing truncal (clonal) mutations present in all cancer cells from branch (subclonal) mutations restricted to subsets — is therefore essential for understanding tumor biology and designing durable therapeutic strategies.

Methods

This paper is a Review published in Cell. It synthesizes observational data and conceptual frameworks from multiple sources, including:

• TRACERx (TRAcking Cancer Evolution through therapy (Rx)) — a prospective, multi-region, longitudinal sequencing study of non-small cell lung cancer (NSCLC) that systematically reconstructs tumor phylogenies across space and time.
• TCGA (The Cancer Genome Atlas) and other large-scale pan-cancer genomic datasets providing population-level estimates of ITH and clonal architecture.
• Multi-region and single-region bulk DNA sequencing studies across multiple cancer types.
• Phylogenetic reconstruction methods applied to somatic mutation and copy-number data from spatially separated tumor biopsies.
• Computational tools for clonal deconvolution (e.g., PyClone, ABSOLUTE) used to infer the cancer cell fraction (CCF) of mutations and distinguish clonal from subclonal events.
• Neoantigen prediction pipelines that integrate somatic mutation calls with HLA typing to estimate the immunogenicity of clonal versus subclonal mutations.

The review covers both DNA-level (point mutations, copy-number alterations, structural variants) and RNA-level (transcriptional heterogeneity) sources of ITH, with emphasis on the temporal distinction between early (truncal) and late (branch) events in tumor evolution.

Key Findings

• Intratumor heterogeneity is present across virtually all cancer types, with subclonal mutations detectable in the majority of solid tumors when analyzed with adequate sequencing depth and multi-region sampling; the degree of ITH varies both between and within cancer types.
• Truncal (clonal) mutations — including canonical driver events in genes such as EGFR, KRAS, TP53, and KEAP1 — occur early in tumor evolution and are shared by all tumor cells, while branch (subclonal) mutations arise later and are restricted to subsets of the tumor cell population.
• Clonal neoantigens derived from truncal mutations are expressed more homogeneously across the tumor and may represent superior targets for immunotherapy (including checkpoint blockade and adoptive cell therapy) compared to subclonal neoantigens, which are present in only a fraction of tumor cells.
• Chromosomal instability (CIN) and whole-genome doubling (WGD) events are associated with worse clinical outcome; subclonal copy-number alterations in particular are linked to poor prognosis in non-small cell lung cancer.
• APOBEC-family cytidine deaminase activity and other endogenous mutational processes (e.g., those associated with tobacco smoking) can remain active late in tumor evolution, generating ongoing subclonal mutational diversity that fuels adaptation to therapy and immune pressure.

Concepts Introduced or Used

• Intratumor Heterogeneity (ITH)
• Clonal vs. Subclonal Mutation
• Trunk-Branch Model of Tumor Evolution
• Cancer Cell Fraction (CCF)
• Clonal Neoantigen
• Subclonal Neoantigen
• Darwinian / Branching Tumor Evolution
• Chromosomal Instability (CIN)
• Whole-Genome Doubling (WGD)
• Mutational Signature Analysis
• APOBEC Mutagenesis
• Immune Editing / Immune Evasion via ITH
• Multi-region Sequencing
• Phylogenetic Reconstruction in Cancer
• Somatic Copy-Number Alteration (SCNA)
• Driver vs. Passenger Mutation
• TRACERx

Entities Referenced

• TRACERx (TRAcking Cancer Evolution through therapy (Rx))
• TCGA (The Cancer Genome Atlas)
• EGFR — early clonal driver in lung adenocarcinoma
• KRAS — early clonal driver in lung adenocarcinoma
• TP53 — early clonal tumor suppressor
• KEAP1 — early clonal event in NSCLC
• APOBEC cytidine deaminases (APOBEC3A, APOBEC3B)
• PyClone (Bayesian clustering for clonal deconvolution)
• ABSOLUTE (purity/ploidy estimation and clonality calling)
• Non-small cell lung cancer (NSCLC)
• Checkpoint blockade immunotherapy
• Adoptive cell therapy

Limitations

The authors note several limitations of current approaches to studying ITH:

• Sampling bias. Most studies rely on single-region or limited multi-region sampling. Under-sampling of a tumor can cause subclonal mutations to be misclassified as clonal, and low-frequency subclones may be missed entirely.
• Resolution limits of bulk sequencing. Standard bulk DNA sequencing detects mutations present at allele frequencies down to roughly 5—10%, meaning that rare subclones fall below the detection threshold. Single-cell sequencing approaches are needed to fully resolve clonal architecture.
• Driver vs. passenger discrimination at subclonal resolution. While truncal driver events can often be identified with reasonable confidence, determining whether a subclonal mutation is a functional driver or a neutral passenger is challenging; many subclonal mutations identified in sequencing studies are likely passengers.
• Lack of longitudinal data. Most genomic datasets are cross-sectional (single time-point), whereas ITH is a dynamic property. Repeated sampling over time and through therapy is essential to understand how clonal architecture evolves but is logistically difficult to obtain.
• Neoantigen prediction uncertainty. Computational prediction of which mutations generate immunogenic neoantigens remains imprecise; the relationship between predicted clonal neoantigen burden and actual anti-tumor immune responses requires further validation.
• Tissue-context dependence. The functional impact of ITH and specific subclonal alterations is likely context-dependent — varying by tissue of origin, tumor microenvironment, and treatment history — yet most studies have been performed in limited cancer types.

Relevance to Clonal Evolution

This review is a landmark synthesis of the clonal evolution paradigm in cancer. It bridges population-genetic models of Darwinian selection with clinical oncology by demonstrating that ITH is not merely an epiphenomenon but a therapeutically actionable feature of tumor biology. The paper’s emphasis on distinguishing clonal from subclonal events — and its argument that truncal targets may yield more durable therapeutic responses — has shaped the design of clinical trials, neoantigen-targeting strategies, and the interpretative frameworks used in large-scale cancer genome projects such as TRACERx.