Clonal Evolution

Metastasis

6 min read

Metastasis

Metastasis — the spread of cancer to distant sites — is the fundamental definition of malignancy (Nowell, 1976) and the cause of most cancer deaths. It is an evolutionary process: a clone acquires the capacity to invade, survive in circulation, colonize a distant tissue, and proliferate in a new microenvironment.

Timing of Metastatic Divergence

Al Bakir et al. (2023) systematically quantified the timing of metastatic divergence in 126 NSCLC patients from the TRACERx cohort using multi-region whole-exome sequencing. Divergence timing was defined relative to the last clonal-sweep in the primary tumour:

  • Late divergence (74.6%): The metastatic clone departed after the last clonal sweep. All primary clonal mutations are present in the metastasis. This was the predominant pattern.
  • Early divergence (25.4%): A complete clonal sweep occurred in the primary after the metastatic clone departed. Primary-ubiquitous clonal mutations are absent from the metastasis.

Even in early divergence cases, a median of 92.1% of primary-ubiquitous mutations were shared with the metastasis — early divergence was still relatively late in molecular evolutionary time. Whole-genome doubling in the primary provided further granularity: metastatic divergence occurred after clonal WGD in 81.0% of WGD cases, and pre-WGD mutations were significantly more likely to be maintained in the metastasis (P = 0.003), suggesting they make better therapeutic targets.

Clinical correlates

Early divergence was independently associated with smoking at the time of primary resection (P = 0.016). Simulations demonstrated that early divergence is more likely when the primary tumour diameter is <8 mm — the actionable threshold in CT screening protocols. At <1 mm simulated diameter, 78% of cases diverged early. This suggests that for early-diverging tumours, metastasis may already be established before radiological detection.

Sampling is critical

When only a single primary tumour region was used to classify divergence timing, 83.3% of late-divergence cases were incorrectly classified as early. Multi-region sampling of the primary tumour is essential for accurate evolutionary inference of metastasis timing.

Modes of Dissemination

Al Bakir et al. (2023) classified dissemination patterns at the case level relative to the primary tumour phylogeny:

  • Monoclonal dissemination (68.3%): A single primary tumour clone seeded all sampled metastases. This is probably an overestimate due to undersampling of metastases.
  • Polyclonal dissemination (31.7%): Multiple genetically distinct primary clones seeded metastases. Subclassified as monophyletic (single branch, 21/40), polyphyletic (multiple branches, 16/40), or mixed (3/40).

Polyclonal dissemination was associated with extrathoracic disease recurrence (P = 0.0056), suggesting that higher clonal diversity among metastatic seeds enables rapid adaptation to distant environmental niches. In 38% of cases with multiple metastases, metastases were seeded from other metastases rather than directly from the primary tumour — evidence of metastasis-to-metastasis seeding.

Role of Lymph Node Metastases

Primary lymph node metastases rarely seed further recurrences. In only 3 of 19 cases with both primary LN and subsequent recurrence did the LN seed the recurrence. In 13 of 19 cases, dissemination occurred solely from the primary tumour. LN involvement is a marker of metastatic propensity — a hallmark of an aggressive primary tumour — rather than an anatomical gateway to distant disease.

Selection in Metastasis-Seeding Clones

Paired analysis of multi-region primary tumours and their metastases revealed that seeding clones are not random — they show evidence of positive selection within the primary tumour:

  • Greater abundance: Seeding clusters had significantly higher maximum cancer cell fraction than non-seeding clusters (P = 6.4 × 10
  • Greater spatial dispersion: Seeding clusters were more broadly distributed across primary tumour regions (P = 1.6 × 10
  • Signatures of positive selection: In LUAD, seeding clusters showed dN/dS = 1.97 for lung cancer genes. In LUSC, positive selection was observed only in seeding clones (dN/dS = 2.03); subclonal mutations in non-metastasizing primaries showed no significant selection (dN/dS = 0.89).

Two categories of metastasis-associated somatic alterations emerge:

CategoryCharacteristicsExamples
Truncal/maintainedClonal in the primary, maintained in metastases; associated with metastatic propensityTP53 mutations (LUAD and LUSC), MDM2 amplification (LUAD)
Metastasis-favouredSubclonal in the primary, enriched in metastases; may confer niche-specific advantageHIST1H3B amplification (LUAD)

KRAS, TP53, and KEAP1 mutations were significantly maintained in LUAD metastases (q < 0.05). TP53 was also significantly maintained in LUSC. Notably, NRAS, RB1, and EGFR exon 19 deletions/L858R were always shared between primary and metastasis — never primary-unique.

Platinum Chemotherapy as an Evolutionary Force

The platinum mutational signature (SBS31/35) was detected in 81.8% (9/11) of treated recurrence samples, demonstrating that adjuvant chemotherapy leaves a mutagenic footprint on the cancer genome. In one case, two brain metastases showed differential platinum-signature exposure, timing their divergence to the chemotherapy window — direct evidence that treatment shapes the evolutionary trajectory of metastatic disease.

Evolutionary Mechanisms

In TRACERx Renal, metastasis-competent clones were distinguished from non-metastasizing clones by the degree of aneuploidy and chromosome complexity. Specific copy-number-alterations — loss of 9p and loss of 14q — were enriched in metastasizing clones. Critically, no evidence of selection for small-scale SNV mutations was found, suggesting that the large-scale genomic changes driven by chromosomal-instability are the primary facilitators of metastatic capacity (Turajlic et al., 2019).

Two Evolutionary Routes

Metastatic spread patterns reflect the evolutionary dynamics of the primary tumor (Turajlic et al., 2019):

  • Punctuated tumors (early clonal aneuploidy, low ITH): Metastasize early and widely from a single dominant clone (monophyletic seeding). Metastases show limited intermetastatic heterogeneity.
  • Gradual/Darwinian tumors (ongoing subclonal diversification): Produce oligometastases from later-arising subclones. When they spread to multiple sites, they may do so polyphyletically (different subclones seed different sites), resulting in intermetastatic heterogeneity.

Metastasis and Early Detection

The observation that metastatic competency can be acquired at the earliest stages of cancer evolution — tumors “born to be bad” — has implications for cancer-early-detection. As Turajlic et al. (2019) note, preclinical models show metastatic dissemination before frank malignancy is histologically detectable, and “as the latency between the emergence of the invasive clone and metastatic spread can be short, the window for early detection could be very limited” (p. 413).

Graph View

Backlinks