← ResourcesPublications
DIAKWANT · PD-L1 · ECDP GRAZ
Impact of Artificial Intelligence Assistance on Combined PD-L1 Scoring on Whole Immunohistochemical Slide Image of tumors
C. Boulagnon-Rombi, C. Bazille, H. Gil, R. Picot, M. Four, L. Dibombe, Y. Salhi, F. Thomas, Y. Bellahsen-Harrar, C. Bossard
17 JUN 2026
Abstract
Background & Objectives
PD-L1 blockade has revolutionized cancer therapy, highlighting the need for accurate PD-L1 evaluation using the Combined Positive Score (CPS), which remains limited by inter-observer variability. We evaluated the AI-based DiaKwant PD-L1 CPS solution against an expert consensus across multiple tumor types to address these challenges.
Methods
A reference CPS score was defined by three expert pathologists on 149 digital slides from gastrointestinal and head and neck tumors. Four different pathologists independently reviewed all cases without AI. Following a washout period of two months, the cases were reassessed with AI assistance. For each slide, the algorithm provided CPS values along with a PD-L1 expression heatmap to support result interpretability. Pathologist assessments were compared with the reference score. Performance was assessed using accuracy, sensitivity and specificity for clinically relevant binary cutoffs, and intraclass correlation coefficient (ICC) to evaluate concordance between pathologists. Scoring time was also recorded.
Results
Using organ-specific cut-offs, the CPS accuracy increased from 77.6% to 83.9% with AI assistance (p < 0.01), sensitivity increased from 78.5% to 89.6% and specificity increased from 73.4% to 76.2%, while the average scoring time per case decreased from 109s to 78s (p < 0.01). The ICC between pathologists rose from 0.71 to 0.90 when introducing AI assistance.
Conclusion
AI-assisted PD-L1 assessment significantly improved agreement with the reference standard, enhanced inter-pathologist reliability, and reduced scoring time while maintaining clinically relevant performance across tumor types. These findings support the potential of AI assistance to exhaustively quantify PD-L1 CPS in routine practice.
