Longitudinal Analysis of Circulating Tumor DNA (ctDNA) in Diffuse Large B-Cell Lymphoma (DLBCL) Using KAPA HyperCap Design Share Non-Hodgkin Lymphoma (KAPA HyperCap DS NHL) Next-Generation Sequencing (NGS) Panel
Bibliographic record
Abstract
Introduction Studies have shown that dynamic changes in lymphoma ctDNA levels are associated with patient outcomes and that monitoring lymphoma ctDNA may help identify patients who are at risk of refractory disease or relapse. NGS panels can successfully identify patient-specific tumor reporter variants in plasma specimens using a tissue-naïve approach that avoids limitations of patient tissue availability. Recently, using the AVENIO Oncology Assay (AOA) NHL Test* validated by the Roche Molecular CAP/CLIA Laboratory, a pre-specified analysis plan using samples from ~800 DLBCL patients from the POLARIX study validated ctDNA as an early prognostic biomarker (Herrera et al., Blood 2022). To further support clinical research on early molecular response (EMR) to treatment and minimal residual disease (MRD) in NHL, we introduce a KAPA HyperCap DS NHL panel designed to cover 100% of the target regions in the AOA NHL Test used for the POLARIX study. We describe the NGS workflow and bioinformatics analysis suitable for identification and monitoring of ctDNA using a tissue-naïve approach. We further describe performance characterization studies and feasibility of ctDNA detection using plasma specimens from patients with DLBCL. Methods KAPA HyperCap DS NHL panel covers coding and/or untranslated regions of 383 genes, plus additional intergenic regions, for a total of 341 Kb. This panel is used in combination with KAPA HyperCap workflows and KAPA reagents on Illumina® platforms, to sequence plasma cfDNA and matched genomic (g)DNA to identify tumor-specific single-nucleotide variants (SNVs) and monitor the dynamics of ctDNA. ctDNA detection and monitoring are supported by open-source bioinformatics tools for a fully integrated MRD analysis solution. Contrived samples for workflow characterization were comprised of commercially available reference materials with known SNV allele frequencies (AF) mimicking cfDNA (SeraSeq® Complete Mutation Mix and Twist Pan-cancer Reference Standard), as well as pre-characterized healthy donor cfDNA. Clinically annotated DLBCL samples were characterized and serially diluted to demonstrate feasibility of longitudinal mutation analysis. Results Serially diluted cfDNA samples and high molecular weight gDNA samples were used to assess the performance of the plasma cfDNA and germline workflows. Library quality control metrics met the yield and size distribution criteria for sequencing. A median of 88 M raw reads were obtained across all libraries. De-duplication yielded a median coverage depth range from 5000-9100 across samples. Median on-target rate (% selected bases) was between 74% and 80%. Average error rate was between 0.00024 or 0.00031 mismatches/read depth. 9 and 3 variants were detected after germline and blocklist filtering in the two commercial reference samples. Variant calling sensitivity was 100% across replicates at 5, 1 and 0.5% AF. In the serial dilution analysis, ctDNA was detected in all replicates at 5, 0.1 and 0.05% mean AF. For samples at 0.01% mean AF, ctDNA detection sensitivity was 83%. Initial data demonstrate that 131 SNV reporters were detected at 23% mean AF in a plasma sample from a treatment-naïve DLBCL patient. 4 SNV reporters were detected at 1.9% mean AF in a plasma sample from an immunochemotherapy-treated DLBCL patient. After 10x and 500x dilution, ctDNA positivity was accurately called at 2.3% and 0.05% mean AF levels for the patient with 131 reporters, and at 0.19% mean AF level for the patient with 4 reporters (Monte Carlo p-value <0.0001). Conclusion The use of KAPA HyperCap DS NHL panel with KAPA HyperCap workflows and open-source bioinformatics tools enables the detection and monitoring of ctDNA for NHL research applications. In contrived samples, variants were detected with high reproducibility at AF as low as 0.05% and with good reproducibility at AF of 0.01%. For a DLBCL patient with 131 SNV reporters in plasma, ctDNA was detectable down to 0.05% mean AF levels, while for a DLBCL patient with 4 SNV reporters, ctDNA was detectable down to 0.19% mean AF levels, demonstrating that sensitivity of Monte Carlo-based ctDNA detection depends on the number of reporter variants. *AVENIO Oncology Assay (AOA) NHL Test and KAPA HyperCap Design Share panels are for Research Use Only, not for use in diagnostic procedures. AVENIO and KAPA are trademarks of Roche. All other product names and trademarks are the property of their respective owners.
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 | 0.001 |
| 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 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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".