Clonal Evolution and Genomic Instability in Cancer

Clonal Dynamics: Trunk, Branches, and Leaves

Multiregion sequencing of primary tumours revealed that cancer evolution follows a branching phylogenetic model. Early driver mutations (KRAS, TP53) present in all tumour regions — the 'trunk' — arose before spatial separation of subclones. Later mutations occurred in specific regions — 'branches' — creating subclonal heterogeneity. This architecture has immediate therapeutic implications: trunk mutations are present in virtually all cancer cells and represent the most rational therapeutic targets; branch mutations may be present in only a fraction of cells and targeting them causes subclonal selection pressure but incomplete tumour elimination.

Chromosomal instability (CIN) — the ongoing mis-segregation of chromosomes during mitosis — dramatically accelerates clonal evolution by generating large-scale copy number changes, loss of heterozygosity, and chromothripsis (catastrophic chromosomal shattering and reassembly in single mitotic events). CIN is paradoxically associated with both more aggressive tumours (more rapid evolution to therapy resistance) and increased immunogenicity (more neoantigens from structural variants). TP53 loss is the most common enabler of CIN, by eliminating the G2/M checkpoint that would otherwise eliminate mitotically aberrant cells.

Genomic Instability Pathways in Cancer

Genomic instability — the elevated rate of mutation or chromosomal rearrangement in cancer — arises through several distinct mechanisms, each leaving a characteristic genomic signature. Microsatellite instability (MSI) from mismatch repair deficiency produces a hypermutator phenotype with thousands of small insertions/deletions at repetitive microsatellite sequences; MSI-high tumours are the most immune-checkpoint-sensitive cancer subtype. Homologous recombination deficiency (HRD) from BRCA1/2 loss produces large deletions, tandem duplications, and balanced translocations (the 'genomic scarring' signature), identifying a population sensitive to PARP inhibitors.

APOBEC-driven mutagenesis — cytidine deaminase activity creating C→T transitions in TC motifs — is the most common mutational signature in breast, bladder, and cervical cancers, driven by replication stress-induced ssDNA exposure. UV photoproducts (melanoma), tobacco carcinogens (lung, head/neck), aflatoxin B1 (hepatocellular), and other environmental mutagens each leave characteristic trinucleotide mutation signatures detectable by sequencing and used to infer carcinogen exposure and DNA repair pathway status.

Clonal Haematopoiesis and Pre-Malignant Clonal Dynamics

Clonal haematopoiesis of indeterminate potential (CHIP) describes the age-related accumulation of somatic mutations in haematopoietic stem cells (HSCs) that drive clonal expansion without meeting diagnostic criteria for haematological malignancy. The most commonly mutated genes in CHIP — DNMT3A, TET2, ASXL1, and JAK2 — are the same genes mutated in myeloid malignancies, identifying CHIP as a pre-malignant state. CHIP mutations are detectable in 10–20% of people over 70 years and confer a 10-fold increased risk of haematological malignancy over 10 years, in addition to unexpected increases in cardiovascular mortality through TET2-driven macrophage inflammatory signalling.

CHIP provides a tractable model for studying early clonal dynamics before cancer develops. Longitudinal sequencing studies show that CHIP clones grow at 5–20% per year, with growth rates dependent on mutation type (JAK2 V617F clones grow fastest), clone size, and the presence of additional hits. The observation that some CHIP clones persist for decades without progression while others rapidly evolve to myeloid malignancy highlights the stochastic nature of clonal evolution and the role of additional somatic hits, inflammatory microenvironment, and immune surveillance in determining malignant transformation timing.

Liquid Biopsy: Tracking Clonal Evolution Non-Invasively

Circulating tumour DNA (ctDNA) — tumour-derived DNA fragments released into the bloodstream through apoptosis, necrosis, and active secretion — captures the clonal composition of a tumour in real time and enables longitudinal tracking of clonal evolution without repeated tissue biopsies. Serial ctDNA analysis during targeted therapy reveals emergence of resistant subclones months before radiological progression: EGFR T790M emergence in osimertinib-treated patients, KRAS secondary mutations in G12C inhibitor-treated patients, and ESR1 mutations in aromatase inhibitor-treated breast cancer are all detectable in liquid biopsy 3–6 months before clinical progression.

ctDNA as a minimal residual disease (MRD) marker post-curative surgery has become a major clinical development focus. Multiple studies across colorectal, lung, and breast cancers demonstrate that detectable ctDNA after resection predicts relapse with sensitivity and lead time over conventional imaging, potentially guiding adjuvant therapy decisions in ctDNA-positive patients while sparing ctDNA-negative patients from unnecessary treatment. FDA clearance of ctDNA-based MRD assays is expanding across cancer types, with trials evaluating ctDNA-guided adjuvant therapy escalation and de-escalation strategies.

Key Takeaways

• ·Cancer evolution follows a branching phylogenetic model: trunk mutations (KRAS, TP53) present in all tumour regions arise before spatial separation; branch mutations create intratumoral heterogeneity and represent subclonal therapeutic targets.
• ·Chromosomal instability (CIN) accelerates clonal evolution through continuous copy number variation and chromothripsis, with TP53 loss being the most common enabler by eliminating the G2/M checkpoint that removes mitotically aberrant cells.
• ·MSI-high tumours (mismatch repair deficiency) and HRD tumours (BRCA1/2 loss) each leave characteristic genomic signatures — the 'genomic scarring' detectable by NGS — with direct therapeutic implications for checkpoint inhibitors and PARP inhibitors respectively.
• ·CHIP (clonal haematopoiesis) affects 10–20% of people over 70, conferring 10-fold increased haematological malignancy risk and elevated cardiovascular mortality — a pre-malignant state that models early clonal dynamics.
• ·Serial ctDNA liquid biopsy detects resistant subclone emergence 3–6 months before radiological progression and identifies minimal residual disease after curative surgery with higher sensitivity than conventional imaging.