Abstract PS9-05: Somatic Structural Variation in Breast Cancer and its Application in Longitudinal Analysis of Circulating Tumor DNA in Early Breast Cancer
Bibliographic record
Abstract
Abstract Background: Genomic structural variation (SV) is a recognized property of cancer cells, contributing to genomic instability and oncogenesis. SV breakpoints and rearrangement patterns are often tumor-specific and can reflect underlying tumor biology. The landscape and implications of SV in breast cancer has been incompletely characterized. Furthermore, the utility of using SVs for prospective circulating tumor DNA (ctDNA) detection and monitoring in early breast cancer (EBC) has not been evaluated. Methods: SV was evaluated through whole genome sequencing (WGS) analysis in two independent breast cancer datasets: (1) The 100,000 genomes project [n=3044 patients; Genomics England, GEL], made possible through access to data and findings in the National Genomic Research Library via the Genomics England Research Environment and (2) a cohort of patients with EBC treated with neoadjuvant chemotherapy [ctDNA evaluation in early breast cancer (TRACER; n=210 patients) Princess Margaret Cancer Centre, Canada]. SV burden and type was evaluated in the GEL dataset using MANTA and in TRACER using an in-house pipeline (SV size cut-off of >50 bp). Survival correlates within the GEL dataset were evaluated using Cox proportional hazard models with cases censored after 6 years. The performance of an SV-based ctDNA assay was evaluated in BT-474 (cell line, contrived samples) and in FFPE tumors (TRACER, clinical samples) using shallow depth (∼15X) WGS followed by SV detection (up to 16) via proprietary multiplex digital PCR. Longitudinal ctDNA detection was performed in the TRACER cohort. Results: SV was common across all breast cancer subtypes (GEL data; median SV burden: 108, range: 4-1448; median aggregate copy number of the top 16 SV: 57, range: 16-160). A higher proportion of inversions were seen in HER2-positive tumors, while deletions were common in TNBC. In a multivariate model (including clinical stage, subtype, and tumor mutational burden), there was a significant association between higher SV count and worse overall survival (OS) in ER+/HER2- breast cancer (HR: 2.31, p=0.0207). Patients with ER+/HER2- breast cancer who experienced a clinical recurrence had higher SV copy number in the top 16 variants than those who did not (p<0.0001). In silico analyses demonstrated that personalized SV-based ctDNA panels (fingerprints) could be successfully designed for 97.1% of GEL cases. To assess the characteristics of an SV-based ctDNA assay, a limit of detection (LoD) study was performed with BT-474 using contrived cfDNA (70 ng). The LoD95 was estimated at 0.00052% tumor fraction (5 PPM) with variants detected as low as 1 in 10 million (0.00001% or 0.1 PPM in 31% of cases). A specificity of 100% was seen in 134 healthy donors using 24 different fingerprints (assessment of 1600 SVs). In an initial cohort of 55 patients (TRACER; ER+/HER2-:19, HER2+:23, TNBC:13), shallow WGS and fingerprint design were successful in all patients. The median number of SVs was 336 (range: 73-1345) with the SV type distributed in a similar fashion (more inversions in HER2+ tumors, deletions in TNBC). A trend towards higher SV copy number based on tumor-only WGS was also seen in those with ER+/HER2- disease who experienced subsequent recurrence (TRACER; SV in no recurrence vs. recurrence: 54.3 vs. 134.1, p=0.099). Shallow WGS and SV-based ctDNA assay design is underway for additional patients in the TRACER cohort. Conclusion: SVs are prevalent in breast cancer and associated with prognosis. Personalized SV-based panels permitted ultrasensitive ctDNA detection with high sensitivity and specificity in contrived samples, supporting assay feasibility for early breast cancer. Further analysis of the prognostic impact of SV-burden and type as well as on-treatment and adjuvant ctDNA detection in patients with EBC (TRACER) using SV tracking will be presented at the meeting. Citation Format: Mitchell Elliott, Karen Howarth, Sasha Main, Jesús Fuentes Antrás, Philippe Echelard, Aaron Dou, Eitan Amir, Michelle B. Nadler, Elizabeth Shah, Celeste Yu, Scott Bratman, Taylor Bird, June Roh, Elza C. de Bruin, Christopher Rushton, Sofia Birkeälv, Miguel Alcaide, Lucia Oton, Sergii Gladchuk, Yilun Chen, Anthony George, Girish Putcha, Samuel Woodhouse, Philippe L. Bedard, Lillian L. Siu, Hal K. Berman, David W. Cescon. Somatic Structural Variation in Breast Cancer and its Application in Longitudinal Analysis of Circulating Tumor DNA in Early Breast Cancer [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr PS9-05.
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 | 0.003 | 0.008 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.009 | 0.002 |
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".