Abstract 3489
Background
PD-L1/PD-1 checkpoint inhibitors have shown clinical activity in UBC. It has been shown that PD-L1 expression on TC and/or IC correlates with clinical efficacy. In this study (ML39708) we examined technical comparability and inter-reader agreement of 4 clinically relevant assays for PD-L1 expression on IC and TC in locally advanced UBC.
Methods
Archived formalin-fixed paraffin-embedded sections from 30 patients with locally advanced UBC (70% cystectomies, 30% transurothelial resections) were selected from 150 cases based on PD-L1 status per VENTANA SP142 IC 5% (10 cases each). The study cohort was stained for PD-L1 using SP142, SP263 (VENTANA), 22C3, and 28-8 (DAKO) assays at two sites according to manufacturer protocols. Stainings were blinded and scored at 5 sites for the PD-L1 expression on IC (% per tumor area) and TC (%). All readers were trained on scoring IC with SP142.
Results
Percentage of IC cells staining for PD-L1 varied from 6.54 to 8.18%, and TC from 5.46 to 15.85%, depending on the assay used (Table). For each assay, IC staining varied slightly to moderately between readers, with small non-significant differences between assays. Results for TC were comparable except for significantly lower staining with SP142. Pairwise comparison revealed –0.3 to 1.6% differences in adjusted means between assays for IC, and for TC, –10.5 to –7.8% (SP142 vs other assays) and –1.9 to 2.7% (other comparisons). Retrospective allocation to binary cut-offs (1%, 5% and 10%) for IC or TC only predominantly showed substantial or high Kappa agreement scores (0.6–0.8) for IC and TC between assays for each reader.Table:
1175P
| Assay | PD-L1 on IC % | Reader ICC† | PD-L1 on TC % | Reader ICC† |
|---|---|---|---|---|
| (95% CI)* | (95% CI)* | |||
| VENTANA SP142 | 8.18 (7.32–9.03) | 0.699 | 5.46 (2.85–8.07) | 0.609 |
| VENTANA SP263 | 7.08 (6.22–7.94) | 0.729 | 15.85 (13.24–18.47) | 0.805 |
| DAKO 22C3 | 6.54 (5.68–7.39) | 0.532 | 13.19 (10.57–15.80) | 0.883 |
| DAKO 28-8 | 6.88 (6.02–7.74) | 0.573 | 15.15 (12.54–17.77) | 0.845 |
Adjusted means for each assay;
†Intra-class correlation per test between 5 readers
Conclusions
This is the first multicenter study for analytical comparison of PD-L1 IHC staining on IC and TC in UBC. High concordance rates across all assays were achieved between trained readers for scoring PD-L1 on IC and TC.
Clinical trial identification
ML39708
Legal entity responsible for the study
Roche Pharma AG
Funding
Roche Pharma AG
Disclosure
A. Hartmann: Membership of an advisory board: Roche, MSD, AstraZeneca Corporate-sponsored research: Novartis, Biontech, Illumina, Nanostring, Quiagen. G. Baretton: Membership of an advisory board Roche, Bristol-Myers Squibb, AstraZeneca Corporate-sponsored research Roche, Bristol-Myers Squibb. F. Lasitschka: Membership of an advisory board: Roche NSCLC regional advisory board. Bristol-Myers Squibb NSCLC regional advisory board Corporate-sponsored research: Roche PLACU study. P. Schirmacher: Roche (Research, Honorarium) Bristol-Myers Squibb (Research, Honorarium), MSD (Research, Honorarium), AstraZeneca (Research, Honorarium), Novartis (Research, Honorarium). T. Braunschweig: Membership of an advisory board and invited guest speaker: Bristol-Myers Squibb Invited guest speaker: MSD Corporate-sponsored research: Roche. R. Tauber: Membership of an advisory board: Roche, Sanofi, Bristol-Myers Squibb) Corporate-sponsored Research: conduct as subinvestigator of clinical trials. S. Hieke-Schulz: Employee Roche Pharma AG. J. Ammann: Stock ownership: Roche Pharmaceuticals Other substantive relationships: Employee of Roche Pharma AG. W. Weichert: Conflicts of interest Advisory boards for Roche, AstraZeneca, MSD, Bristol-Myers Squibb, Pfizer, Novartis, Boehringer. Collaborative research with Roche, Novartis, AstraZeneca, Boehringer. All other authors have declared no conflicts of interest.