Clinical evaluation of an automated pan-organ combined PD-L1 scoring using artificial intelligence on immunostained whole-slide images
Bibliographic record
Abstract
Background Programmed death-ligand 1 (PD-L1) inhibitors have shown remarkable results in oncology; however, many patients fail to respond, highlighting the need for reliable assessment of PD-L1 expression for patient selection. PD-L1 scoring, especially the combined positive score (CPS), is hindered by inter- and intraobserver variability, complex staining patterns, and technical discrepancies, all of which can impact therapeutic decisions. Artificial intelligence (AI) offers a solution by standardizing PD-L1 evaluation. This study evaluates a PD-L1 CPS AI, designed for reproducible and robust PD-L1 scoring across various tumor types and conditions. Materials and methods AI performance was validated on 142 samples spanning multiple tumor types (gastrointestinal, head and neck, breast, and uterine cervix) and sourced from four centers, reflecting diverse staining protocols. Routine scores were available. A gold standard was established through independent retrospective scoring by three senior pathologists enabling the assessment of variability. The scoring process was followed by collegial discussions to resolve discordant cases and ensure medical consensus. After a washout period, cases were reassessed with AI assistance. AI and routine manual scores were compared with the gold standard using organ-specific cut-offs. Results AI assistance improved interobserver agreement among pathologists, increasing intraclass correlation coefficient (ICC) from 62% to 74%, with a particularly pronounced effect in challenging cases with CPS < 20 ( n = 91), where ICC improved from 19% to 62%, underscoring the value of AI in reducing variability near clinical decision thresholds. Based on clinical cut-offs, AI-based scoring outperformed routine manual scoring in accuracy (88% versus 75%) and sensitivity (96% versus 78%), while maintaining a comparable positive predictive value (88% versus 87%), indicating an improved ability to detect true-positive cases. Conclusions This study highlights the potential of an AI-driven tool—DiaKwant PD-L1 algorithm—to improve PD-L1 scoring accuracy and reduce observer variability, particularly near clinical thresholds, across various solid carcinomas, independently of pre-analytical and digitization platforms. Its integration into clinical workflows could enhance efficiency and optimize patient eligibility for immunotherapy.
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 | 0.002 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| 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".