Introduction: PD-L1 is established as a predictive marker for therapy of non-small cell lung cancer with pembrolizumab. Furthermore, PD-L1 positive melanoma has shown more favorable outcomes when treated with anti-PD1 antibodies and dacarbazine compared to PD-L1 negative melanoma. However, the role of PD-L1 expression with regard to response to checkpoint inhibition with anti-CTLA-4 is not clear, yet. In addition, the lack of standardization in the immunohistochemical assessment of PD-L1 makes the comparison of results difficult. In this study, we investigated the PD-L1 gene expression with a new fully automated technique via RT-PCR and correlated the findings with the response to the anti-CTLA-4 antibody ipilimumab.
Materials and methods: Within a retrospective multi-center trial, PD-L1 gene expression was evaluated in 78 melanoma patients in a total of 111 pre-treatment tumor samples from 6 skin cancer centers and analyzed with regard to response to ipilimumab. For meaningful statistical analysis, the cohort was enriched for responders with 30 responders and 48 non-responders. Gene expression was assessed by quantitative RT-PCR after extracting mRNA from formalin-fixed paraffin embedded tumor tissue and correlated with results from immunohistochemical (IHC) stainings.
Results and discussion: The evaluation of PD-L1 expression based on mRNA level is feasible. Correlation between PD-L1 expression as assessed by IHC and RT-PCR showed varying levels of concordance depending on the antibody employed. RT-PCR should be further investigated to measure PD-L1 expression, since it is a semi-quantitative method with observer-independent evaluation. With this approach, there was no statistical significant difference in the PD-L1 expression between responders and non-responders to the therapy with ipilimumab. The evaluation of PD-L1 expression based on mRNA level is feasible. Correlation between PD-L1 expression as assessed by IHC and RT-PCR showed varying levels of concordance depending on the antibody employed. RT-PCR should be further investigated to measure PD-L1 expression, since it is a semi-quantitative method with observer-independent evaluation. With this approach, there was no statistical significant difference in the PD-L1 expression between responders and non-responders to the therapy with ipilimumab.
Keywords: Biomarker; Checkpoint inhibitors; Ipilimumab; Melanoma; PD-L1; Predictive marker; Prognostic marker.