Clonal Evolution

Jones, J. R., Weinhold, N., Ashby, C., Walker, B. A., Wardell, C., Pawlyn, C., Rasche, L., Melchor, L., Cairns, D. A., Gregory, W. M., Johnson, D., Begum, D. B., Ellis, S., Sherborne, A. L., Cook, G., Kaiser, M. F., Drayson, M. T., Owen, R. G., Jackson, G. H., Davies, F. E., Greaves, M., & Morgan, G. J. (2019). Clonal evolution in myeloma: the impact of maintenance lenalidomide and depth of response on the genetics and sub-clonal structure of relapsed disease in uniformly treated newly diagnosed patients. Haematologica, 104(7), 1440–1450.

Properties1
tagsprimary-research, clinical, selection, phylogenetics

11 min read

Bibliographic Reference

Jones, J. R., Weinhold, N., Ashby, C., Walker, B. A., Wardell, C., Pawlyn, C., Rasche, L., Melchor, L., Cairns, D. A., Gregory, W. M., Johnson, D., Begum, D. B., Ellis, S., Sherborne, A. L., Cook, G., Kaiser, M. F., Drayson, M. T., Owen, R. G., Jackson, G. H., Davies, F. E., Greaves, M., & Morgan, G. J. (2019). Clonal evolution in myeloma: the impact of maintenance lenalidomide and depth of response on the genetics and sub-clonal structure of relapsed disease in uniformly treated newly diagnosed patients. Haematologica, 104(7), 1440–1450. https://doi.org/10.3324/haematol.2018.202200

Core Argument

This study uses whole exome sequencing (WES) of paired presentation and relapse samples from 56 newly diagnosed multiple myeloma (NDMM) patients enrolled in the Myeloma XI trial to determine how depth of response and lenalidomide maintenance therapy shape the genetic and subclonal features of relapsed disease. All patients received uniform induction therapy (immunomodulatory agent + cyclophosphamide + dexamethasone) and were selected for high-risk disease, defined as relapse within 30 months of maintenance randomization. The central finding is that depth of response, not maintenance therapy, is the key determinant of evolutionary patterns at relapse: patients achieving a complete response (CR) showed branching evolution with an increased mutational load, altered mutational profile, and accumulation of deleterious structural lesions at relapse, whereas patients with partial responses (PR) showed predominantly stable clonal architecture with little change between presentation and relapse. Importantly, lenalidomide maintenance had no detectable impact on the mutational profile, copy number changes, or evolutionary patterns at relapse compared to observation, and no signal of selection for lenalidomide-resistance mutations was identified. The authors propose that effective therapy creates an evolutionary bottleneck — particularly in patients achieving deep responses — that selectively eliminates dominant clones and allows pre-existent resistant subclones to emerge at relapse via branching evolutionary pathways.

Methods

Patient cohort. Samples were selected from NDMM patients enrolled in the Myeloma XI trial (NCT01554852). A nested case-control analysis was performed on 56 uniformly treated patients who relapsed within 30 months of maintenance randomization, irrespective of classical genetic risk status or best response. Thirty patients received lenalidomide maintenance and 26 were observed. Prior to relapse, best responses were: CR in 21% (12/56), nCR in 21% (12/56), VGPR in 42% (23/56), and PR in 16% (9/56). Patients were grouped as complete responders (CR/nCR, n=24) and non-complete responders (VGPR/PR, n=32). Median time from trial entry to relapse was 19 months. Clinical characteristics were well matched between maintenance groups.

Sequencing. DNA was isolated from CD138+ plasma cells (bone marrow aspirate) using MACSorting. Control DNA from peripheral blood. WES libraries prepared with SureSelectQXT and SureSelect Clinical Research Exome kit, with additional baits covering immunoglobulin and MYC loci. Paired-end sequencing to median depth of 122x (tumor) and 58x (control) on HiSeq2500.

Variant calling and clonal reconstruction. SNVs called using Strelka (v.1.0.14) and MuTect. Variant allele frequency (VAF) distributions mapped with R package SciClone for mutational cluster identification. Cancer clonal fractions (CCF) were calculated for all mutations and plotted using kernel density estimation to infer clonal structure at presentation and relapse.

Copy number and structural variants. Copy number assessed by both MLPA (SALSA P425-B1 myeloma probemix) and Sequenza (v.2.1.2), with 94% consensus. Translocations determined using MANTA (v.0.29.3), with qRT-PCR validation for IGH translocations. All suspected bi-allelic CNA events confirmed by manual BAM file interrogation using IGV. Bi-allelic inactivation also called in cases with a non-synonymous mutation with mono-allelic loss or single mutation with VAF ≥80%.

Statistical analysis. Wilcoxon matched-pairs signed rank tests for paired comparisons; Fisher exact test for nominal variable differences.

Key Findings

  • “Branching evolution was the predominant pattern, seen in 66% (37 of 56) of all cases. Linear evolution characterized only by the gain of mutations at relapse was seen in 20% (11 of 56) of cases. The remaining 14% (8 of 56) had either the same mutational profile at both time points, and were classified as stable progression (n=7), or displayed a loss of a mutational cluster at relapse, classified as stable progression with clone loss (n=1).”

  • “Stable progression was only seen in the non-CR series (25%, 8 of 32, vs. 0%, 0 of 24; P=0.008). Breaking down the non-CR series further showed that stable progression was predominantly associated with a PR (56%, 5 of 9 of PR; 13%, 3 of 23 of VGPR patients). No CR or nCR patients (0 of 24) had evidence of stable progression, with all patients showing branching or linear evolution (P=0.002).”

  • “Patients achieving a CR had a significantly higher non-synonymous mutational load at relapse, with a median of 59 mutations compared to 40 at presentation (P<0.001). Non-CR patients had a similar mutational load at relapse, with a median of 39 mutations at both time points (P=0.63).”

  • “We did not identify a mutational signal consistent with the selection of clones carrying mutations associated with acquired resistance to lenalidomide. Five patients had mutations in DDB1 (n=2), SLC16A1 (n=2), and CRBN (n=1), but these were not confined to the cases of lenalidomide maintenance, nor were they seen exclusively at relapse.”

  • “Bi-allelic inactivation of tumor suppressor genes located at sites of recurrent CNA are likely to be relevant mediators of relapse. We show that bi-allelic inactivation events of RB1, TRAF3, and TP53 are important with 18% (10 of 56) of patients having evidence of bi-allelic inactivation of ≥1 gene at relapse, in comparison to 14% (8 of 56) at presentation.”

Additional key result: Gain(1q) was the most common new event at relapse (13% of patients). Secondary translocations to chromosome 8q (MYC) increased from 21% at presentation to 27% at relapse. Five patients had evidence of two separate MYC translocations at relapse, consistent with MYC translocations being late events.

Concepts Introduced or Used

  • Evolutionary patterns at relapse — The study classifies relapse-associated clonal dynamics into three patterns: branching (gain and loss of mutational clusters, predominant at 66%), linear (gain of mutations only, 20%), and stable (preserved mutational profile, 14%). These map directly to existing wiki concepts of branching-evolution, linear-evolution, and clonal stability.
  • Treatment-induced evolutionary bottleneck — Effective therapy (especially deep response leading to CR) creates a strong selective bottleneck that eliminates dominant clones, allowing low-level pre-existent resistant subclones to emerge at relapse. This is the central evolutionary mechanism proposed (population-bottleneck, clonal-sweep).
  • Depth of response as a determinant of clonal dynamics — CR vs. non-CR status is framed as the primary variable governing evolutionary outcome, not the maintenance regimen used. Deep responders undergo the bottleneck-driven branching path; non-responders maintain the same clonal architecture (therapy-resistance).
  • Clonal extinction and subclonal emergence — The loss of mutational clusters at relapse (clone loss) and gain of new mutations in dominant clones at relapse (subclonal emergence) are the operational signatures of branching evolution (subclonal-architecture, subclonal-reconstruction).
  • Bi-allelic inactivation as a relapse mechanism — Bi-allelic inactivation of tumor suppressor genes (RB1, TRAF3, TP53) via mutation + copy number loss is identified as an important mediator of relapse, especially in the CR series (driver-mutation).

Entities Referenced

Genes and pathways:

  • NRAS, KRAS, BRAF — RAS-MAPK pathway genes; the most commonly mutated pathway in the cohort (57% at some point during disease course); mutations gained and lost at relapse (kras)
  • TP53 — Tumor suppressor; bi-allelic inactivation via mutation + del(17p) at relapse; present in >10% of patients at relapse (tp53)
  • RB1 — Tumor suppressor at 13q; bi-allelic inactivation at relapse (rb1)
  • TRAF3 — Tumor suppressor at 14q; bi-allelic inactivation at relapse
  • MYC — Oncogene at 8q24; translocations to MYC increased at relapse; five patients had two separate MYC translocations at relapse
  • DIS3 — Mutated in 13% at presentation
  • PRDM1 — Most common new mutation at relapse (5% of patients)
  • CRBN — Cereblon, the direct target of lenalidomide; one patient developed a novel CRBN mutation (p.Cys326Gly) after 8 months of lenalidomide maintenance, achieving MRD negativity before relapsing
  • DDB1, SLC16A1 — Additional genes in the lenalidomide mechanism-of-action pathway; mutations rare and not exclusive to lenalidomide-treated patients
  • NF1, EGFR, TET2, FANCA, FAM46C, CDH2, FAF1 — Other recurrently mutated myeloma genes with gain/loss at relapse
  • CDKN2C — Tumor suppressor at 1p32.3, region of recurrent CNA

Drugs:

  • Lenalidomide — Immunomodulatory drug (IMiD) used as maintenance therapy; no significant impact on clonal evolution patterns detected (lenalidomide)
  • Thalidomide — Immunomodulatory drug used as induction therapy in 52% of patients
  • Cyclophosphamide — Alkylating agent used in induction
  • Dexamethasone — Corticosteroid used in induction
  • Immunomodulatory drugs (IMiDs) — Class of agents including thalidomide, lenalidomide, and pomalidomide

Cell types:

  • Plasma cells — Malignant cell of origin in multiple myeloma; CD138+ selection used for sample purification
  • Multiple myeloma — Plasma cell malignancy; described as having intermediate mutational load between CML and solid cancers

Methods and tools:

  • Whole exome sequencing (WES) — Primary assay (subclonal-reconstruction)
  • SciClone — R package for VAF-based mutational cluster inference (subclonal-reconstruction, variant-allele-fraction)
  • Cancer clonal fraction (CCF) — Mutation cancer cell fraction; calculated via VAF, purity, and copy number correction (cancer-cell-fraction)
  • Kernel density estimation — Used to plot CCF distributions and infer clonal structure
  • MLPA — Multiplex ligation-dependent probe amplification for copy number
  • Sequenza — Bioinformatic copy number caller from sequencing data
  • MANTA — Structural variant caller
  • Myeloma XI trial — Large UK multicenter phase 3 trial (Jackson et al., 2017)

Limitations (as stated by authors)

  • “It does remain important to evaluate the impact of maintenance on low-risk cases who are long-term survivors, a question not addressed in this study.” — The study cohort was restricted to high-risk patients (early relapse within 30 months), and results may not generalize to low-risk or long-remission patients.
  • “Clonal structure may be further assessed using single cell analysis. However, for the moment, the challenges of obtaining individual malignant plasma cells from large series of patients at multiple disease time points has proved to be a significant barrier.”
  • The study focused on early-relapsing patients, selected because “it had previously been suggested that exposure to lenalidomide could enhance progression of such cases post maintenance.” — Selection bias toward high-risk biology.
  • “The maintenance lenalidomide dose” (10 mg) may be a factor in the lack of observed selective pressure, as noted in the companion editorial by miething2019-clonal-evolution-myeloma.
  • The authors acknowledge that “it is likely that the [CRBN] mutation occurred by chance” in the single patient who developed it after lenalidomide — individual case, not a population-level signal.
  • Larger analyses using patients with prolonged remissions are required to fully assess the impact of maintenance on mutational patterns.
  • Sampling was from a single site (pelvis), raising the possibility of spatial sampling bias, though the authors argue this is mitigated by consistent biopsy location.
  • Specific mutations associated with relapse could not be identified with confidence; “higher patient numbers will be required” for this analysis.

Relevance to Clonal Evolution

This study is one of the most directly relevant clinical investigations of clonal evolution under a defined therapeutic selective pressure in a uniformly treated cohort. Several findings bear directly on the wiki’s core concepts:

The bottleneck paradox. The study provides a clear clinical demonstration that deep response (CR) creates a tight population-bottleneck: dominant clones are eradicated, but resistant subclones survive and re-emerge via branching-evolution at relapse. The very success of therapy selects for more complex clonal architecture — an instance of the compression-entrenchment dynamic described in compression-progress-evolution and visible in the relationship between CR status and increased mutational load at relapse (median 59 vs. 40 mutations, P<0.001).

Selection under therapy vs. innate resistance. The finding that PR patients showed stable clonal architecture (56% of PRs stable) reveals a distinct evolutionary path: these patients have treatment-resistant dominant clones from disease onset that require no genetic change to drive relapse. This maps onto the distinction between acquired resistance (via bottleneck + branching) and innate resistance (via pre-existing resistant clone expansion) — a key axis in therapy-resistance.

Dissociation between clinical benefit and clonal impact. Lenalidomide maintenance improves progression-free survival in the Myeloma XI trial yet had no detectable effect on clonal evolution patterns at relapse. This dissociation, also highlighted by miething2019-clonal-evolution-myeloma, suggests that the clinical benefit of maintenance operates through a mechanism (possibly immune-mediated) that does not exert strong direct selective pressure on clonal composition. This is a critical nuance for the therapy-resistance concept: not all therapeutic effects operate through clonal selection.

Bi-allelic inactivation as a convergent relapse mechanism. The enrichment of bi-allelic TP53, RB1, and TRAF3 inactivation at relapse (18% vs. 14%) identifies a convergent evolutionary strategy — complete loss of tumor suppressor function — that may be a general feature of relapse across cancer types, consistent with findings in other uniformly treated cohorts (driver-mutation, clonal-sweep).

Absence of a lenalidomide-resistance mutational signal. Despite the large sample size (n=56 with WES), no evidence of selection for mutations in the CRBN-DDB1 pathway was found, except for a single case. This contrasts with findings in heavily pre-treated patients and cell lines, suggesting that resistance mutations to IMiDs are rare in the upfront setting and that clinical resistance in NDMM may operate through alternative, non-mutational mechanisms (e.g., clonal selection of pre-existent resistant subclones without new mutations).

Methodological contribution. The combination of SciClone (VAF clustering), kernel density estimation (CCF distributions), and manual IGV validation of bi-allelic events provides a rigorous analytical pipeline for subclonal-reconstruction in paired diagnosis-relapse studies that serves as a methodological template for future work.

Graph View

Backlinks