The Copy Number Landscape of Relapsed and Refractory Diffuse Large B-Cell Lymphoma
Bibliographic record
Abstract
Introduction Patients diagnosed with diffuse large B-cell lymphoma (DLBCL) are treated with standard frontline immunochemotherapy (R-CHOP). However, for cases where R-CHOP fails (relapsed-refractory DLBCL, rrDLBCL), prognosis is extremely poor, with 2-year overall survival of 20-40%. The successful development of new therapies may be hampered by our limited understanding of the genetic and molecular mechanisms underpinning treatment resistance. For example, recent data from our group has highlighted novel mutations that emerge following treatment with R-CHOP. The contribution of copy-number variations (CNVs) towards treatment resistance has not yet been thoroughly explored. A more complete characterization of these genetic alterations may lead to new prognostic biomarkers or treatment strategies. Methods We analyzed exome sequencing data from 59 rrDLBCL cases derived from either tissue biopsies or liquid biopsies collected after relapse, including both unpublished and previously published cases (Schmitz et al. (2018) NEJM 378:1396-1407 and Morin et al. (2016) Clin Can Res 22(9)). We separately performed low-pass whole-genome sequencing (lpWGS, 0.1-1x coverage) on 45 rrDLBCL liquid biopsies with ctDNA levels insufficient for exome-based analysis, for a total of 104 cases with copy-number information. We identified CNVs from exome and lpWGS data using Sequenza and ichorCNA, respectively. Next, we identified significant peaks of recurrent gains and losses using GISTIC2. Comparison of these peaks to CNVs in a previously published diagnostic DLBCL cohort (Schmitz et al. (2018) NEJM 378:1396-1407) enabled the identification of events that were significantly more prevalent in rrDLBCL. Results Overall, the landscape of CNVs in rrDLBCL is reminiscent of diagnostic DLBCL, with recurrent amplifications of chromosome 7 (43/104, 41.3%) and 18q (42/104, 40.4%) and recurrent deletions of 6q (25/104, 24.0%) and 17p13 (39/104, 37.5%). We identified nine regions enriched for recurrent amplifications or deletions among rrDLBCLs. These include deletions of 17p13.1 (20.4% in diagnostic biopsies vs 41.3% of rrDLBCLs, q=8.53x10-5) and recurrent amplifications of 8q24 (18.5% vs 42.3%, q=5.72x10-7) and 7p22 (27.2% vs 57.9%, q=6.29x10-8). Many of these peaks represent focal events that are exceedingly rare in diagnostic DLBCL and do not contain established lymphoma-associated genes, including amplifications affecting 700kb of 6p11.2 (2.03% vs 7.69%, q=0.0178) and 500kb of 19p13.3 (6.7% vs 31.7%, q=9.99x10-10). Notably, the 6p11.2 amplifications were associated with inferior progression-free survival following R-CHOP (p=0.02), with most tumors harboring this alteration relapsing within 12 months. We also identified a novel, recurrent deletion affecting a 20mb region of 5q (2.78% vs 10.6%, q=0.00604) which was significantly deleted in rrDLBCL. For tumors with additional samples collected prior to R-CHOP and following salvage therapy, deletions of 5q appeared to emerge following frontline therapy and persisted after subsequent treatments, suggesting they may contribute to treatment resistance. Discussion The 17p13.1 deletion enriched in rrDLBCL encompasses TP53, which is a common target of somatic point mutations in rrDLBCL and associated with inferior treatment outcomes. The amplification of 8q24 and 7p22 include MYC and GNA12/CARD11, respectively, although these large events encompass numerous additional genes which may be the target of such events. Curiously, the focal 6p11.2 amplification only overlaps a handful of genes including miR_598, which has been predicted to target CD27 and CD38 and whose expression is upregulated in B-cell cell lines (Lawrie et al. (2008) Leukemia 22:1440-2446). Further investigation and validation of these events and their corresponding targets will provide insight into the biology of rrDLBCL and may reveal novel therapeutic targets. Disclosures Michaud: Epizyme: Current Employment. Daigle:Epizyme: Current Employment. Jain:Kite/Gilead: Consultancy; Novartis: Consultancy. Kuruvilla:Merck: Consultancy, Honoraria; Bristol-Myers Squibb Company: Consultancy; Celgene Corporation: Honoraria; AstraZeneca Pharmaceuticals LP: Honoraria, Research Funding; AbbVie: Consultancy; Gilead: Consultancy, Honoraria; Karyopharm: Consultancy, Honoraria; Roche: Consultancy, Honoraria, Research Funding; Seattle Genetics: Consultancy, Honoraria; Janssen: Honoraria, Research Funding; Amgen: Honoraria; Antengene: Honoraria; Novartis: Honoraria; Pfizer: Honoraria; TG Therapeutics: Honoraria. Crump:Servier: Consultancy; Roche: Consultancy; Kite/Gilead: Consultancy. Assouline:BeiGene: Consultancy, Honoraria, Research Funding; AbbVie: Consultancy, Honoraria, Speakers Bureau; Janssen: Consultancy, Honoraria, Speakers Bureau; Takeda: Research Funding; Pfizer: Consultancy, Honoraria; AstraZeneca: Consultancy, Honoraria, Speakers Bureau; F. Hoffmann-La Roche Ltd: Consultancy, Honoraria, Research Funding. Steidl:Juno Therapeutics: Consultancy; Seattle Genetics: Consultancy; Roche: Consultancy; Bristol-Myers Squibb: Research Funding; AbbVie: Consultancy; Bayer: Consultancy; Curis Inc: Consultancy. Johnson:AbbVie: Research Funding; Roche/Genentech, Merck: Honoraria; Roche/Genentech, Merck, Bristol-Myers Squibb, AbbVie: Consultancy. Scott:NanoString: Patents & Royalties: Named inventor on a patent licensed to NanoString, Research Funding; Janssen: Consultancy, Research Funding; Roche/Genentech: Research Funding; NIH: Consultancy, Other: Co-inventor on a patent related to the MCL35 assay filed at the National Institutes of Health, United States of America.; Celgene: Consultancy; Abbvie: Consultancy; AstraZeneca: Consultancy. Morin:Celgene: Consultancy.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
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".