Evolution and impact of subclonal mutations in chronic lymphocytic leukemia

Authors

Landau DA

Carter SL

Stojanov P

McKenna A

Stevenson K

Lawrence MS

Sougnez C

Stewart C

Sivachenko A

Wang L

Wan Y

Zhang W

Shukla SA

Vartanov A

Fernandes SM

Saksena G

Cibulskis K

Tesar B

Gabriel S

Hacohen N

Meyerson M

Lander ES

Neuberg D

Brown JR

Getz G

Wu CJ

Doi

10.1016/j.cell.2013.01.019

PMID: 23415222 · DOI: 10.1016/j.cell.2013.01.019 · Journal: Cell (2013)

TL;DR

This study performed whole-exome sequencing of 149 chronic lymphocytic leukemia (CLL) tumor/normal pairs to characterize intratumoral heterogeneity and clonal evolution. By integrating somatic mutations with copy number data using the ABSOLUTE algorithm, the authors classified mutations as clonal or subclonal, identified 20 putative CLL driver genes falling into 7 core signaling pathways, inferred temporal ordering of driver acquisition (MYD88, trisomy 12, del(13q) as early events; SF3B1, TP53 as later events), and demonstrated that the presence of subclonal driver mutations is an independent risk factor for rapid disease progression (adjusted HR 3.61, Cox P=0.007).

Cohort & data

  • 160 matched CLL tumor/normal pairs subjected to whole-exome sequencing (~130X average depth); 149 with both WES and copy number data analyzed by ABSOLUTE.
  • Cancer type: chronic lymphocytic leukemia (CLLSLL).
  • Dataset: lcll_broad_2013.
  • 111 samples also profiled with Affymetrix SNP 6.0 arrays for somatic copy number analysis.
  • 18 patients sampled at two timepoints (median 3.5 years apart); 12 received intervening chemotherapy, 6 untreated.
  • Methods: whole-exome-seq, affymetrix-snp6, GISTIC, MutSig, ABSOLUTE algorithm.

Key findings

  • Identified 20 putative CLL driver genes (q<0.1) and 5 recurrent cytogenetic alterations (del(8p), del(13q), del(11q), del(17p), trisomy 12).
  • Mean somatic mutation rate: 0.6 +/- 0.28 per megabase; average 15.3 nonsynonymous mutations per patient.
  • 54% of detected mutations were clonal (average 10.3/sample); 46% were subclonal (average 8.5/sample), detected in 146/149 samples.
  • Driver mutations classified as predominantly early/clonal: MYD88 (clonal in 80-100% of affected samples), trisomy 12, hemizygous del(13q).
  • Driver mutations classified as predominantly later/subclonal: SF3B1, TP53, ATM.
  • Prior chemotherapy associated with significantly higher number of subclonal mutations (P=0.048); number increased with number of prior therapies (P=0.011).
  • Longitudinal analysis: 10/12 treated CLL cases showed clonal evolution vs. 1/6 untreated (P=0.012, Fisher exact test). Five treated cases showed branched evolution.
  • Presence of subclonal driver mutation associated with shorter failure-free survival from sample (FFS_Sample, P<0.001) and shorter failure-free survival from next treatment (FFS_Rx, P=0.006).
  • Adjusted HR for subclonal driver presence: 3.61 (CI 1.42-9.18), Cox P=0.007, adjusting for IGHV status, prior therapy, and high-risk cytogenetics.

Genes & alterations

  • TP53 — predominantly subclonal/later event; mutations in 17 samples; expanded over time post-treatment.
  • ATM — predominantly subclonal; 9 mutations in 6 samples.
  • SF3B1 — predominantly subclonal/later event; 19 mutations in 19 samples; expanded over time post-treatment.
  • MYD88 — predominantly clonal/early event; n=12, clonal in 80-100% of harboring samples.
  • NOTCH1 — recurrently mutated CLL driver.
  • DDX3X — recurrently mutated CLL driver (previously reported).
  • ZMYM3 — recurrently mutated CLL driver (previously reported).
  • FBXW7 — recurrently mutated CLL driver (previously reported).
  • XPO1 — newly identified CLL driver (confirmed by other studies).
  • CHD2 — newly identified CLL driver (confirmed by other studies).
  • POT1 — newly identified CLL driver (confirmed by other studies).
  • NRAS — novel CLL driver candidate; mutations at highly conserved sites.
  • KRAS — novel CLL driver candidate; mutations at highly conserved sites.
  • BCOR — novel CLL driver candidate; known role in cancer biology.
  • EGR2 — novel CLL driver candidate; induces apoptosis.
  • MED12 — novel CLL driver candidate.
  • RIPK1 — novel CLL driver candidate; cell cycle/apoptosis.
  • SAMHD1 — novel CLL driver candidate; immune pathway.
  • ITPKB — novel CLL driver candidate; B cell receptor signaling.
  • H1-4 — novel CLL driver candidate (histone gene, formerly HIST1H1E); clonal in 5/5 affected samples.

Clinical implications

  • Presence of subclonal driver mutations is an independent prognostic biomarker for shorter time to retreatment or death in CLL (adjusted HR 3.61, P=0.007), comparable in strength to IGHV status and high-risk cytogenetics.
  • Pre-treatment subclonal drivers anticipate the dominant genetic composition of relapsing tumor, potentially guiding therapy selection.
  • Chemotherapy acts as an evolutionary bottleneck promoting expansion of fit subclones with driver mutations — supporting the “watch and wait” paradigm in CLL.
  • Detection of subclonal drivers may provide a new prognostic approach testable in larger clinical trials.

Limitations & open questions

  • Cross-sectional inference of temporal ordering assumes a common evolutionary trajectory across patients.
  • Detection limited to subclones present in >10% of cancer cells (~130X coverage); very low-frequency subclones undetectable.
  • Longitudinal cohort limited to 18 patients; treated vs. untreated groups were small (12 vs. 6).
  • Cannot definitively distinguish post-therapy diversification from outgrowth of pre-existing minor subclones.
  • Indels in genes other than NOTCH1 and X-chromosome sSNVs excluded from ABSOLUTE analysis.
  • Validation of novel candidate drivers (EGR2, RIPK1, SAMHD1, ITPKB, H1-4) in independent cohorts needed.

Citations from this paper used in the wiki

  • “We identified driver mutations as predominantly clonal (e.g., MYD88, trisomy 12 and del(13q)) or subclonal (e.g., SF3B1, TP53), corresponding to earlier and later events in CLL evolution.” (Summary)
  • “Ten of 12 CLL cases treated with chemotherapy (but only 1 of 6 without treatment) underwent clonal evolution, predominantly involving subclones with driver mutations (e.g., SF3B1, TP53) that expanded over time.” (Summary)
  • “Presence of a subclonal driver mutation was an independent risk factor for rapid disease progression.” (Summary)
  • “Adjusted hazard ratio (HR) of 3.61 (CI 1.42-9.18), Cox P=0.007” (Results, p.7)

This page was processed by paper-compiler on 2026-05-06.