Abstract PD5-03: PD5-03 Comparison of a personalized sequencing assay and digital PCR for circulating tumor DNA based Molecular Residual Disease detection in early-stage triple negative breast cancer in the cTRAK-TN trial
Bibliographic record
Abstract
Abstract Background: Detection of circulating tumour DNA (ctDNA) in patients (pts) who have completed treatment for early-stage breast cancer is associated with a high risk of future relapse. Identifying those at high risk of subsequent relapse may allow tailoring of further therapy to delay or prevent recurrence. Previous analysis of this cohort showed that tools capable of detecting ctDNA at lower concentrations are needed to increase sensitivity and lengthen the lead time between ctDNA detection and relapse. We compared ctDNA detection via a personalised sequencing assay to dPCR in patients from the cTRAK TN clinical trial. Methods: The cTRAK-TN trial recruited 161 pts into prospective ctDNA surveillance with dPCR, with ctDNA positive pts randomised to 1) CT staging plus pembrolizumab therapy for patients without relapse or 2) observation. Pts had serial post-treatment surveillance plasma samples collected every 3 months for up to 2 years. Whole exome sequencing (WES) was performed on tumor DNA from FFPE samples to design personalised Residual Disease and Recurrence (RaDaR®) multiplex PCR based NGS assays. Retrospectively, plasma DNA extracted from a minimum of 2mls banked plasma, was sequenced with personalised RaDaR assays, and ctDNA detection identified with a proprietary algorithm. dPCR assays tracked 1-2 mutations, as previously described. Primary endpoint was rate of positive ctDNA detection by 12 months from start of surveillance in both assays. Secondary endpoints were agreement in ctDNA detection between RaDaR and dPCR assays and lead-time between ctDNA detection and disease recurrence. Results: Overall, 147 pts and 241 tissue samples were subject to WES, and RaDaR assays were developed for 142 pts with sufficient plasma for testing. RaDaR assays tracked a median of 47 variants (range 33-56) per patient, and a total of 907 timepoints were analysed (median 6 timepoints per pt, range 1-11). With RaDaR, 39.4% (56/142) patients tested ctDNA positive during follow-up, with a median ctDNA detected level of 0.081% estimated variant allele fraction (eVAF). With dPCR, 35.2% (50/142) pts tested ctDNA positive. The ctDNA detection rate by 12 months from the start of ctDNA surveillance was 36.2% (95% CI; 27.6% – 43.7%) with RaDaR and 29.9% (95%CI; 21.6% – 37.3%) with dPCR. The overall test agreement between RaDaR and dPCR assays was 92.7% (95%CI; 90.7% – 94.4%). From a patient perspective, 58.7% pts were ctDNA negative for both assays, 32.9% ctDNA were positive for both assays and 8.6% presented discrepancies. ctDNA was detected by RaDaR but not by dPCR in 9 pts and it was detected by dPCR but not by RaDaR in 3 pts. Among ctDNA positive pts, 55.2% were first detected positive by RaDaR, 5.2% by dPCR, and 39.6% were detected at the same time-point (test of proportions, p< 0.001). The median lead time from ctDNA detection to relapse was 7.1 months (95% CI 5.9 – 15.9%) with RaDaR and 5.7 months (95% CI 3.2% – 7.4%) with dPCR. Conclusion: The RaDaR personalised multi-mutation sequencing assay detected MRD with a longer median lead time prior to relapse, and with higher sensitivity, than dPCR mutation tracking assays. These findings have implications for the choice of ctDNA assay in clinical trials designed to treat patients at the point of MRD detection. Citation Format: Maria Coakley, Prithika Sritharan, Guillermo Villacampa, Claire Swift, Kathryn Dunne, Lucy Kilburn, Katie Goddard, Patricia Rojas, Andy Joad, Warren Emmett, Charlene Knape, Karen Howarth, Peter S. Hall, Catherine Harper-Wynne, Tamas Hickish, Iain Macpherson, Alicia F. Okines, Andrew M. Wardley, Duncan Wheatley, Simon Waters, Rosalind Cutts, Isaac Garcia-Murillas, Judith Bliss, Nicholas Turner. PD5-03 Comparison of a personalized sequencing assay and digital PCR for circulating tumor DNA based Molecular Residual Disease detection in early-stage triple negative breast cancer in the cTRAK-TN trial [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr PD5-03.
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 | 0.005 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.005 | 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".