Radiomic response evaluation of recurrent or metastatic head and neck squamous cell cancer (R/M HNSCC) patients receiving pembrolizumab on KEYNOTE-012 study.
Bibliographic record
Abstract
6545 Background: Immunotherapy has become a standard of care in the treatment of R/M HNSCC, however only a subset of patients respond, highlighting the need for predictive and prognostic biomarkers. Radiomics is a non-invasive method to quantitatively analyze tumors through conventional imaging. Methods: The pre-treatment and first-on-treatment (after 8 weeks) computed tomography (CT) scans from 132 R/M HNSCC patients treated with single-agent Pembrolizumab (10mg/kg Q2W or 200mg Q3W IV) on the KEYNOTE-012 study were analyzed. Identified target lesions, per RECIST 1.1, were manually contoured, and radiomic features from the tumor and peritumoral region (3 mm expansion of the tumor) were extracted using PyRadiomics. All combinations of image filters and feature classes, not including shape descriptors of peritumoral region, were extracted. Feature space dimensionality was reduced by clustering features (hierarchical clustering using Pearson-based distance and complete linkage) and selecting the medoid of each cluster. Correlation with lesion-level response (LLR) at first-on-treatment CT and overall response (OR) was evaluated using concordance index (CI) with Benjamini-Hochberg multiple testing correction. Results: A total of 406 lesions were included (45 head & neck (HN), 207 lung, 57 liver, 86 lymph nodes (LN), 11 other). 3562 features were extracted from pre-treatment scans (2246 tumor, 1316 peritumor). Considering all lesion sites collectively, 27 of 110 feature clusters were significantly correlated with LLR (false discovery rate (FDR) < 0.05) but not with best overall RECIST response per patient on study. However, when grouped by organ, a number of feature cluster medoids were significantly associated (FDR < 0.05) with LLR (HN: 1, lung: 28, liver: 8, LN: 1) and OR (liver: 18). Feature clusters predictive of LLR and OR included descriptors of both tumor-specific and tumor/peritumoral gray-level intensity and texture (e.g. 74% tumor and 26% peritumoral features in clusters significantly associated with OR in liver). Conclusions: Tumor and peritumoral radiomic features at baseline correlate with LLR and OR to immunotherapy in R/M HNSCC. Despite significant heterogeneity in lesion site, both global and site-specific significant feature clusters could be identified.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.009 | 0.039 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 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 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".