Branching Evolution
Definition
Branching evolution is the pattern of clonal-evolution in which multiple genetically distinct subclonal lineages diverge from a common ancestor, coexist within the tumor, and may each undergo further diversification. It produces a phylogenetic-tree with multiple contemporaneous branches rather than a single linear succession of clones.
This pattern is the norm in solid cancers. Gerlinger et al. (2012) provided the first direct demonstration through multi-region exome sequencing of renal-cell carcinoma: clonal ordering of 128 mutations revealed a branching phylogenetic tree with one branch evolving into metastatic clones and the other diversifying into primary-tumor regions (gerlinger2012-ith-branched-evolution). Greaves & Maley (2012) subsequently noted that “the evolutionary trajectories are complex and branching, exactly as Nowell proposed and in parallel with Darwin’s iconic evolutionary speciation tree” (p. 309).
Distinction from Linear Evolution
In linear-evolution, only one lineage survives at each stage — the fossil record gives the appearance of stepwise progression because intermediate forms are extinct. Branching evolution, by contrast, produces contemporaneous subclones with distinctive mutational and phenotypic profiles that may be territorially segregated within the tumor (Greaves & Maley, 2012).
Turajlic et al. (2019) caution that evolution is always branched at the genotypic level, because mutation and cell division continuously produce divergence. “Branching of a tumour phylogenetic tree does not always imply clonal selection, as branching is the natural product of mutational processes” (p. 405). What distinguishes branching driven by selection from branching driven by neutral drift is whether the branches show differential expansion — evidence of positive-selection acting on some lineages but not others.
Evidence
Direct evidence for branching evolution comes from:
- Single-cell sequencing: Navin et al. (2011) used single-cell sequencing of breast cancers to reveal branching clonal structures, and Anderson et al. (2011) showed repeated independent acquisition of copy number changes within the same leukemia (Greaves & Maley, 2012)
- Multi-region sequencing: Spatially separated biopsies frequently contain different subclones with distinct mutational profiles, demonstrating territorial segregation of branches (McGranahan & Swanton, 2017)
- Metastatic seeding: Metastatic lesions are clonal in origin yet genetically unique, with ancestry traceable to different branches of the primary tumor’s phylogenetic tree (Greaves & Maley, 2012)
Quantitative Evidence from Uniformly Treated Cohorts
Jones et al. (2019) provided the first precise quantification of evolutionary patterns at relapse in a uniformly treated cohort, using WES of paired diagnosis-relapse samples from 56 NDMM patients in the Myeloma XI trial (jones2019-clonal-evolution-myeloma):
| Pattern | Frequency | Definition |
|---|---|---|
| Branching | 66% (37/56) | Gain and loss of mutational clusters at relapse |
| Linear | 20% (11/56) | Gain of mutations only, no loss |
| Stable | 14% (8/56) | Same mutational profile at both time points (n=7) or clone loss without new mutations (n=1) |
The key finding is that depth of response, not maintenance therapy, determines the evolutionary pattern. All CR/nCR patients showed branching or linear evolution — none had stable progression (0/24 vs. 25% in non-CR, P=0.008). Stable progression was concentrated in PR patients (56% of PRs stable). This establishes a clinical partition: deep responders undergo bottleneck-driven branching at relapse; shallow responders maintain their pre-existing clonal architecture. The absence of any impact of lenalidomide maintenance on evolutionary patterns — despite its known PFS benefit — suggests that the clinical benefit operates through a mechanism (possibly immune-mediated) that does not exert strong direct selective pressure on clonal composition (Jones et al., 2019).
Clinical Implications
Branching evolution means that a single biopsy may sample only one branch of a tumor’s phylogeny, missing clinically relevant subclones elsewhere. This is why multi-region-sequencing is essential for accurate subclonal reconstruction. It also means that targeting a subclonal driver mutation — one present in only one branch — leaves other branches untouched, providing a reservoir for relapse (McGranahan & Swanton, 2017).
Revision history
- 2026-07-28 — Added quantitative clinical evidence. Jones et al. (2019) myeloma cohort: 66% branching, 20% linear, 14% stable. Depth of response (CR vs. PR) determines evolutionary pattern at relapse. (jones2019-clonal-evolution-myeloma)