Multimodal detection in plasma of molecular residual disease (MRD) in locally advanced head and neck squamous cell carcinoma (LA-HNSCC).
Bibliographic record
Abstract
6065 Background: Despite intensive therapy, 30% of patients (pts) with LA-HNSCC relapse. Published data shows that MRD detection during follow up (FU) may predict relapse, e.g. human papillomavirus (HPV) DNA in p16+ pts post radiation (RT) or chemoradiation (CRT); or circulating tumor DNA (ctDNA) post surgery. MRD detection using multiple assays after definitive RT/CRT has not been reported. Methods: We enrolled pts with high risk LA-HNSCC (stage III HPV+, III-IVB HPV-) treated with curative intent: surgery ± adjuvant therapy, RT or CRT. Plasma was collected pre-treatment at baseline (B), at 4-6 weeks (FU1) and 8-12 weeks (FU2) post-treatment. RaDaR, a personalized liquid biopsy assay that targets patient specific somatic variants identified in whole exome sequencing (WES) from matched tumor tissue, was used to detect ctDNA, reported as estimated Variant Allele Fraction (eVAF). Cancer Personalized Profiling by deep sequencing (CAPP-seq) was used as a tumor naïve assay for ctDNA, reported as mean VAF. HPV sequencing (HPV-seq) in all and digital PCR (dPCR) in p16+ pts were used to detect HPV DNA. Relapse free survival (RFS) was estimated using Kaplan Meier and associations of ctDNA or HPV DNA with RFS using log-rank test. Assays were compared with Spearman correlation. Results: A total of 89 plasma samples from 35 pts were collected prospectively; 32 pts with at least a FU sample were evaluable: 26 had both. Median age was 63 years (34-71). Most pts were stage III HPV + (N = 16, 50%) and received CRT (N = 25, 78%). No pts had clinical or radiological residual disease at FU2. Median FU was 18.3 months (5.1-25.9), there were 7 clinical relapses. RaDar was applied in 17 pts with WES in matched tissue, including 6 with relapse. ctDNA at B was detected in 15/17 (88%). eVAF at B was not associated with RFS (p = 0.98). Two pts relapsed < 1 year after RT/CRT and had eVAF > 0.001% in FU2 sample; lead time to relapse was 100 and 245 days. FU1 sample of a pt who relapsed > 1 year post-CRT was close to the threshold for ctDNA+ with eVAF 0.0001%; lead time was 494 days. Three pts who received CRT relapsed > 500 days from FU2 sample (ctDNA-). CAPP-seq was applied to 29 pts; mean VAF correlated with eVAF at B (r = 0.75) but not at FU. ctDNA+ at FU using CAPP-seq was not associated with RFS (p = 0.25). HPV-seq was applied to 32 pts; specificity and sensitivity was 100% at B. Among p16+ pts with FU sample (N = 15), HPV DNA was detected in 4/4 with relapse and not in those without, predicting shorter RFS (p < 0.01). dPCR detected HPV DNA in 2/4 pts with relapse. HPV-seq and dPCR correlation was high at B (r = 0.99), lower at FU2 (r = 0.66). Conclusions: HPV DNA and ctDNA can be detected in LA-HNSCC before and after definitive therapy. RaDaR but not CAPP-seq may detect MRD in pts who relapse within 1 year after RT/CRT with a significant lead time. HPV-seq may be more sensitive than dPCR to detect HPV DNA in MRD. Validation in an interception study is planned (NCT 04599309 ).
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 | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 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".