Divergent Modes of Tumor Evolution Underlie Histological Transformation and Early Progression of Follicular Lymphoma
Bibliographic record
Abstract
Abstract Introduction: Follicular lymphoma (FL) remains a significant clinical burden as it is an incurable disease and most patients will eventually suffer from disease progression. Two clinical events are associated with poor outcomes for patients with FL: (1) histological transformation (TFL) of their original FL into a high-grade, aggressive lymphoma subtype (2-3% of patients per year) and (2) early disease progression (PFL) where patients experience treatment failure within 2 years of receiving therapy (20% of patients). Despite recent high-throughput sequencing studies, the nature of tumor clonal dynamics leading to TFL or PFL is poorly understood and it is unknown if similar, or contrasting, modes of selection underpin these FL clinical events. Materials & Methods:We assembled a study cohort consisting of 21 patients: 15 experiencing TFL and 6 PFL. For each TFL and PFL patient, we obtained primary biopsies (T1; taken at the time of the initial FL diagnosis), biopsies at transformation/progression (T2) and matched normal samples. We performed whole genome sequencing on each specimen and identified single point mutations and copy number alterations using MutationSeq and TITAN, respectively. We compared T1 to T2 somatic mutation profiles and identified mutations associated with extinction of T1 clones and expansion of T2 clones. To validate these patterns, we selected 192 positions from each patient for deep-targeted sequencing validation (~10733X) in their T1, T2, and normal samples. We applied a statistical model (PyClone) to estimate cancer cell fraction (CCF) of each validated mutation. These CCF estimates were used to construct clonal phylogenies (Citup) and infer clonal dynamic patterns during their evolutionary histories. The Wright-Fisher model of genetic drift was used to model tumor evolution. Results: Temporal analysis of mutational burden revealed that mutational burden was significantly higher in T2 (8162 mutations ± 2146) than in T1 (6373 ± 2630) tumors for both TFL and PFL patients (Wilcox P < 0.001). This was independent of time interval between sampling (Spearman R2 = 0.029, P = 0.456). Mutation variant allelic fraction (VAF) distributions revealed that all distributions showed evidence of shared clonal ancestry between T1 and T2 tumors accompanied by substantial numbers of T1 and T2-specific mutations. We selected ≥ 192 mutations per patient from these distributions and performed deep-targeted amplicon sequencing, validating 96.3% of mutations and acquiring precise VAFs to infer clonal dynamics. In 13 of 15 TFL patients (87%), we observed dramatic clonal dynamics, characteristic of T2 tumors dominated by clones (or phylogenetic lineages) that were absent or extremely rare in T1 tumors (< 1% CCF). Digital droplet PCR was used to confirm the existence of both scenarios (confirming a clone as rare as 2 out of approximately 105 cells). Tumor evolution modeling demonstrated that this mode of evolution was driven through positive selection for mutations that confer fitness advantages and not by genetic drift. In contrast, PFL patients exhibited markedly different patterns of clonal dynamics compared to TFL patients. 4 of 6 PFL patients (67%) harbored readily detectable clones at T1, which expanded to full clonal prevalence during treatment with immuno-chemotherapy. Tumor evolution modeling demonstrated that this mode of evolution could be explained under neutral evolutionary dynamics (drift). Conclusions: We have shown that histological transformation and early progression manifest through divergent modes of tumor evolution. As the transformation phenotype may arise after diagnosis, more frequent monitoring of these patients will be required to determine the exact timing of the evolutionary inflection point that elicits transformation. In comparison, prediction of early treatment resistance should be achievable through comprehensive characterization of the genetic and clonal composition at diagnosis; this would ultimately identify patients who may benefit from upfront alternative therapies without the need to first endure predictable early treatment failure. Disclosures Sehn: roche/genentech: Consultancy, Honoraria; amgen: Consultancy, Honoraria; seattle genetics: Consultancy, Honoraria; abbvie: Consultancy, Honoraria; TG therapeutics: Consultancy, Honoraria; celgene: Consultancy, Honoraria; lundbeck: Consultancy, Honoraria; janssen: Consultancy, Honoraria. Connors:Millennium Takeda: Research Funding; Seattle Genetics: Research Funding; F Hoffmann-La Roche: Research Funding; Bristol Myers Squib: Research Funding; NanoString Technologies: Research Funding. Scott:Janssen: Consultancy; Celgene: Consultancy; Roche: Honoraria; BC Cancer Agency: Patents & Royalties: Inventor on a patent licensed to NanoString Technologies.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".