Inter‐ and intraobserver agreement of programmed death ligand 1 scoring in head and neck squamous cell carcinoma, urothelial carcinoma and breast carcinoma
Bibliographic record
Abstract
AIMS: Programmed death-ligand 1 (PD-L1) expression by tumour cells (TC) is a mechanism for tumour immune escape through down-regulation of antitumour T cell responses and is a target for immunotherapy. PD-L1 status as a predictor of treatment response has led to the development of multiple biomarkers with different reference cut-offs. We assessed pathologist consistency in evaluating PD-L1 immunopositivity by examining the inter- and intraobserver agreement using various antibody clones and different cancer types. METHODS AND RESULTS: PD-L1 expression in TC and immune cells (IC) was manually scored in 27 head and neck squamous cell carcinoma (HSCC), 30 urothelial carcinoma (UC) and breast carcinoma (BC) using three commercial clones (SP263, SP142, 22C3) and one platform-independent test (E1L3N). For interobserver agreement, PD-L1 status was evaluated blindly by three pathologists. For intraobserver agreement, PD-L1 expression was re-evaluated following a wash-out period. Intraclass correlation coefficient (ICC), overall percentage agreement (OPA) and κ-values were calculated. Using clinical algorithms, the percentage of PD-L1-positive cases in HSCC, BC and UC were 15-81%, 47-67% and 7-43%, respectively. The percentage of PD-L1 positive cases relied heavily on the algorithm/cut-off values used. Almost perfect interobserver agreement was achieved using SP263 and E1L3N in HSCC, 22C3, SP142 and E1L3N in BC and 22C3 in UC. The SP142 clone in UC and HSCC showed moderate agreement and was associated with lower ICC and decreased intraobserver concordance. CONCLUSIONS: Excellent inter- and intraobserver agreement can be achieved using SP263, 22C3 and E1L3N, whereas PD-L1 scoring using SP142 clone is associated with a higher level of subjectivity.
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 | 0.032 | 0.038 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".