300P - Correlation of PD-L1 protein and mRNA expression and their prognostic impact in triple-negative breast cancer

Background

PD-L1 expression is determined by immunohistochemical (IHC) analyses. PD-L1 positivity can differ depending on the used antibody. The aim of this study was to evaluate the correlation of IHC and mRNA PD-L1 expression and their associations to prognosis.

Methods

For this analysis, 85 TNBC samples from a prospective cohort of breast cancer patients (n=1,270, PiA Prognostic Assessment in routine application, NCT 01592825) were eligible. IHC was performed using CAL10 antibody (BioCare), ≥1% staining was considered PD-L1 positive for immune cell score (IC), tumour proportion score (TPS) and combined positive score (CPS). For analysis of mRNA expression, microarray analysis was performed (Affymetrix®, HG U133 Plus 2.0, probesets #1: 223824, #2: 227458, detecting different transcripts of the PD-L1 gene); maximum likelihood method was used for cut off determination. Correlations with IHC and TILs were tested using Spearman´s rank correlation. Survival analysis included recurrence free interval (RFI) and overall survival (OS), considering only those patients who received chemotherapy (n=76). Median follow up was 73 months (20-127).

Results

In IHC analysis, half of the samples were classified PD-L1 positive for IC and CPS (50.6% and 49.4%) and only 23.5% considering TPS. In probeset #1, 63.5% were determined PD-L1 positive, no correlation to the IHC scores, TILs and probeset #2 was shown. In probeset #2 only 34.1% showed a PD-L1 positivity and had a strong correlation to the IHC scores and to TILs (p<0.01). PD-L1 positivity had no impact on survival when determined by IHC scores or probeset #2. In contrast, patients with high PD-L1 expression in probeset #1 had a more favourable 7-years RFI probability (84.6% vs 64.3%). Low PD-L1 expression showed higher risk for recurrence in univariate (2.68, 95%CI 1.089-7.532) and multivariate analysis (3.43, 95%CI 1.294-9.080, adjusted to nodal status). Considering OS only a trend was shown (HR 1.445, 95%CI 0.570-3.662).

Conclusions

In our cohort, the PD-L1 mRNA analysis detected additional PD-L1 positive tumours compared to IHC analysis. For validation of the prognostic impact and to examine the predictive value considering therapy with immune checkpoint inhibitors, further studies are warranted.

Clinical trial identification

NCT01592825 release date 16.12.2009.

Editorial acknowledgement

Legal entity responsible for the study

M. Vetter, E. Kantelhardt.

Funding

Wilhelm Roux Program of the Medical Faculty Martin Luther University Halle-Wittenberg (grant number FKZ 25/36), German Federal Ministry of Education and Research (grant number Med FKZ 031A429).

Disclosure

C. Thomssen: Financial Interests, Personal, Speaker, Consultant, Advisor: Amgen, AstraZeneca, Aurikamed, Daiichi Sankyo, Gilead, Jörg Eickeler, Hexal, Lilly, Medupdate, MSD, Nanostring, Novartis, Onkowissen, Pfizer, Roche, Seagen, Vifor; Financial Interests, Personal, Financially compensated role: Forum Sanitas; Non-Financial Interests, Personal, Member: AGO Breast Committee, ASCO, DGGG (Germ Soc OB/GYN), DGS (Germ Soc Senology), DKG (Germ Cancer Soc), EORTC PathoBiomarker Group; Non-Financial Interests, Personal, Member of Board of Directors: AGO -B Breast Study Group; Non-Financial Interests, Personal, Officer: BIG; Non-Financial Interests, Personal, Invited Speaker: ESO; Non-Financial Interests, Personal, Steering Committee Member: ESMO. All other authors have declared no conflicts of interest.