Coupling of response biomarkers between tumour and peripheral blood in mesothelioma patients undergoing chemoimmunotherapy

Platinum-based chemotherapy in combination with anti-PD-L1 antibodies has shown promising results in mesothelioma. However, the immunological mechanisms underlying its efficacy are not well understood and there are no predictive biomarkers of clinical outcomes to guide treatment decisions. Here, we combine time-course RNA sequencing of peripheral blood mononuclear cells with pre-treatment tumour transcriptome data from the 54-patient cohort in the single arm phase II DREAM study. The identified immunological correlates are predictive of response and provide further evidence for the additive nature of the interaction between platinum-based chemotherapy and PD-L1 antibodies. Our study highlights the complex, but predictive interactions between the tumour and immune cells in peripheral blood during the response to chemoimmunotherapy. Fifty-four participants were recruited to the DREAM clinical trial of durvalumab plus chemotherapy in malignant mesothelioma. Tumour biopsy tissue and matched bloods were collected. To analyse tumour tissue, archival FFPE tissue (blocks or slides) was obtained from participating hospital sites and mRNA was extracted. We retrieved mRNA of sufficient quantity and quality to perform gene expression analysis using the nanoString nCounter platform from 46 of these patients. The nanoString PanCancer IO 360 kit provided gene expression data for 770 immune-oncology related genes. Differential expression of genes were compared between two patient groups based on Progression at 6 months (PFS6) for the purpose of identifying predictive biomarkers. Samples were run as a single replicate, with each nCounter casette containing a manufacturer's QC Standard sample plus up to 11 experimental samples.