Clonal Evolution

Passenger Mutation

10 min read

Passenger Mutation

Definition

A passenger mutation is a somatic alteration that has no effect on cell fitness — it is evolutionarily neutral. Passengers accumulate due to the background mutation rate and genetic-instability, and they increase in frequency not because of selection but because they “hitchhike” on the clonal-expansions driven by driver-mutations (Greaves & Maley, 2012). The term “hitchhiker mutation” is the direct evolutionary biology equivalent.

Prevalence

The vast majority of somatic mutations in a cancer are passengers. Individual cancers can contain hundreds to tens of thousands of mutations, and “the great majority of these are assumed to be neutral mutations arising from genetic instability” (Greaves & Maley, 2012, p. 308). Only a modest number are functionally relevant drivers.

Bozic et al. (2010) provided a quantitative relationship: a typical solid tumor contains 40–100 coding gene alterations, including 5–15 driver mutations. In their branching process model, passenger mutations accumulate linearly with time — n(t) = v × t/T — where v is the neutral mutation rate and T is the cell division time. The expected number of passengers in a tumor with k drivers is given by:

n = (v / 2s) × log(4ks² / u) × log(k)

where s is the selective advantage per driver and u is the driver mutation rate. This formula links observable quantities (total mutations, estimated drivers) to evolutionary parameters — a bridge between genomic data and evolutionary theory.

Bozic et al. (2016) derived the frequency spectrum for passenger mutations during neutral clonal expansion in a multitype branching process with cell death. The expected number of subclonal passenger mutations present above frequency α is:

m_s = u(1 − α) / ((1 − δ)α)

and the expected number of clonal passenger mutations is:

m_c = δu / (1 − δ)

where u is the exome-wide passenger mutation rate (~0.015 per cell division) and δ = d/b is the death-birth ratio. These formulas show that the familiar 1/f frequency spectrum is the δ = 0 (no cell death) special case. When δ is close to 1 — the relevant regime for real tumors — far more mutations appear at high frequencies, and multiple passenger mutations reach fixation (become clonal) during expansion without ever being present in the founding cell. This means that clonal passenger mutations are not necessarily truncal; many were collected during clonal expansion. Each such fixation event represents a bottleneck where competing intermediate-clones went extinct — the clonal passenger is the sole survivor, and the losers are invisible to standard sequencing. For δ = 0.99, there are ~150 subclonal passengers at >1% frequency, ~15 at >50%, and ~1.5 clonal; for δ = 0.999 (premalignant), these numbers rise tenfold.

flowchart TD
    Founder["Founding Cell<br>────────<br>1 cell"] -->|"all descendants inherit"| Truncal["TRUNCAL Passengers<br>CCF = 1.0<br>Founder-inherited<br>(true founder mutations)"]

    Founder -->|"clonal expansion<br>δ = d/b"| Pop["Expanding<br>Population"]

    Pop -->|"early divisions"| ClonalNotTruncal["CLONAL (not truncal) Passengers<br>CCF = 1.0<br>m_c = δu/(1−δ)<br>~1.5 clonal passengers at δ=0.99"]
    Pop -->|"mid divisions"| Subclonal["SUBCLONAL Passengers<br>CCF 10−50%<br>~15 at δ=0.99"]
    Pop -->|"late divisions"| LowFreq["LOW-FREQUENCY Passengers<br>CCF 1−10%<br>~150 at δ=0.99"]
    Pop -->|"recent divisions<br>(~last 7 doublings)"| Invisible["INVISIBLE Passengers<br>CCF < 1%<br>Below detection floor<br>at 100x depth"]

    ClonalNotTruncal -.->|"fixation via drift<br>when δ ≈ 1"| FixNote["ρ_k ≈ (u/(u−log δ))^k<br>Drift-fixated clonal passengers<br>can be indistinguishable from<br>selected drivers by frequency alone"]

Figure: Passenger accumulation during clonal expansion. Mutations present in the founding cell are truncal (CCF = 1.0, inherited by all cells). Mutations arising during early expansion can also reach CCF = 1.0 via drift when the death-birth ratio δ = d/b is close to 1 — these are clonal but not truncal. Later-arising passengers remain subclonal at progressively lower frequencies, and the most recent ~7 doublings are below the detection floor. The dashed arrow highlights the formal tension: drift-fixated clonal passengers can masquerade as selected drivers by frequency alone. Synthesized from Bozic et al. (2010, 2016), Turajlic et al. (2019), and Nik-Zainal et al. (2012).

Scientific Value

Paradoxically, passenger mutations are among the most informative features of a cancer genome. Because they are unaffected by selection, they preserve a faithful record of the mutational processes that have been active throughout the tumor’s lifetime.

Mutational signatures. Nik-Zainal et al. (2012) exploited this property: “Most somatic mutations in cancers are thought to be ‘passenger’ events that do not contribute to cancer development. These bystanders bear the imprints of the DNA damage and repair processes operative during the development of the cancer, unmodified by selection” (p. 980). By analyzing thousands of passenger mutations per genome, they extracted distinct mutational-signatures reflecting specific processes such as aging, BRCA deficiency, and APOBEC-mutagenesis.

The clock mechanism. Passengers are not merely usable as a clock — they are the molecular-clock mechanism. Every temporal inference in cancer genomics relies on counting passenger mutations per branch of the phylogenetic-tree. The clock is not a separate instrument applied to the genome; it is the genome’s own record of neutral mutation accumulation. This principle enables four major classes of timing: (1) branching order — comparing passenger loads across branches reveals the relative timing of clonal divergences; (2) driver timing — the ratio of clonal to subclonal passengers surrounding a driver indicates when it arose (Gerstung et al., 2020); (3) copy-number timing — whether a passenger appears on one or both copies of a gained segment reveals whether it predates or postdates the gain; (4) signature activity timing — changes in mutational-signature composition between early (clonal) and late (subclonal) passengers reveal the temporal sequence of mutagenic process exposures (Nik-Zainal et al., 2012). In each case, the measured quantity is a count of passenger mutations. “The number of passenger mutations unique to a lineage is a measure of the molecular age of that clone” (Turajlic et al., 2019, p. 407). This principle underlies the finding that many driver mutations precede diagnosis by years or decades (Gerstung et al., 2020) — the latency is read directly from the passenger clock. See the full molecular-clock concept page for caveats on rate constancy, detection limits, and single-timepoint limitations.

Phylogenetic markers. Passengers provide genetic marks to distinguish different clones. The proportion of passengers shared between clones reveals their ancestry, and the variant-allele-fraction of passenger mutations determines clone abundance (Turajlic et al., 2019).

Formal Tension: Neutral Recorders vs. Drift-Fixated Passengers

A productive tension runs through the literature on passenger mutations, centered on whether the assumption of strict neutrality can be maintained in frequency-based analyses.

The neutral-recorder paradigm (Nik-Zainal et al., 2012) treats passengers as unconditionally neutral — their value lies precisely in being “unmodified by selection” (p. 980). Under this view, any passenger’s frequency reflects only the timing of its origin and the subsequent expansion of its lineage. Frequency-based methods for distinguishing drivers from passengers (such as recurrence at CCF = 1.0 across independent tumors) rest on this assumption: if passengers are strictly neutral, clonal passengers arise only through founder inheritance or through the stochastic low-frequency persistence inherent in 1/f dynamics.

The drift-fixation paradigm (Bozic et al., 2016) challenges this picture quantitatively. When the death-birth ratio δ = d/b is close to 1 — the regime relevant for most human cancers — passengers can reach fixation (CCF = 1.0) through drift alone. The fixation probability of the k-th surviving passenger is:

ρ_k ≈ (u / (u − log δ))

For δ = 0.997 (MSS colorectal cancer), ρ_1 ≈ 0.80, meaning ~80% of early passengers reach fixation. The expected number of clonal passengers not present in the founder is m_c = δu/(1 − δ) ≈ 5 for typical parameter values. These drift-fixated passengers are clonal (CCF = 1.0) but not truncal (not in the founding cell). By frequency alone, they are indistinguishable from selected drivers that reached fixation through positive selection.

The tension. If a passenger can reach CCF = 1.0 by drift in a high-δ tumor, then the presence of a clonal mutation — even across multiple samples — is not, by itself, sufficient evidence of selection. The Bozic et al. (2016) framework shows that ~1.5 clonal passengers per tumor are expected at δ = 0.99 under strictly neutral drift. This means that any given clonal mutation in a tumor may be a drift-fixated passenger rather than a selected driver. Methods that rely on frequency alone (or on statistical recurrence without accounting for the δ-dependent fixation rate) risk conflating the two.

Partial resolution. The tension is not irreconcilable. Recurrence at the same genomic locus across many independent tumors remains strong evidence of selection, because drift-fixated passengers are expected at random loci — the probability that the same locus fixates by drift in multiple independent tumors is vanishingly small. The rank-and-cut method (see below) addresses this by combining recurrence, functional consequence, and excess above background, rather than relying on frequency alone. However, for rare drivers or drivers in genes with low background mutation rates, the distinction between a single drift-fixated passenger and a rare selected driver remains ambiguous when only one tumor is examined.

Implications for the 1/f test. The standard neutrality test (1/f frequency distribution) is a joint test of neutrality and pure birth (δ = 0). Bozic et al. (2016) showed that δ close to 1 shifts the passenger frequency spectrum toward higher frequencies, meaning that a neutral tumor with high δ may deviate from 1/f and be misclassified as showing selection. The δ correction is therefore essential for correctly calibrating frequency-based neutrality tests — without it, drift-fixated passengers are misinterpreted as evidence of selection.

Rank-and-Cut Method

The PCAWG Consortium (2020) developed a ‘rank-and-cut’ approach to systematically distinguish drivers from passengers. Mutations in significantly mutated genomic elements are ranked by recurrence, estimated functional consequence, and expected driver pattern. The excess burden of mutations above the background mutation rate is then estimated, and the ranked list is cut at this level — mutations above the threshold are probable drivers; those below are probable passengers. Using this method, ~5% of tumours across 2,658 cases had no identified drivers, suggesting that current knowledge of driver genes and elements remains incomplete.

Distinction from Drivers

Driver status can be ambiguous or context-dependent. Greaves & Maley (2012) note examples where a mutation is functionally relevant only in the context of therapeutic response involving that gene, or where monoallelic loss only affects function when the second allele is also lost. “Passenger lesion status can also be ambiguous or context-dependent” (p. 307).

Graph View

Backlinks