Multiomics profiling for prediction of immunotherapy response in advanced pleural mesothelioma: Sub-study of the NIPU trial.

8086 Background: The combination of ipilimumab and nivolumab (IPI/NIVO) is a standard treatment for unresectable pleural mesothelioma. However, the objective response rate (ORR) is relatively low, and serious toxicity necessitating treatment cessation with or without steroid treatment is seen in 20%. Predictive biomarkers for IPI/NIVO in mesothelioma are needed to personalize treatment decisions. Methods: In the NIPU trial, 118 patients progressing after first-line chemotherapy were included in the study and were randomly assigned to IPI/NIVO alone or in combination with the telomerase UV1 vaccine. Whole-slide tumor tissues were available from 99 patients and analyzed using multiplex immunofluorescence (mIF) with a panel including CD8, CD20, CD66b, FoxP3, Granzyme-B, and pan-cytokeratin. Machine learning algorithms (XGBoost) were utilized and trained for immune cell subset classification and tissue subregion segmentation (tumor vs. stroma). Bulk RNA-sequencing (RNA-seq) was performed on 25 matched baseline fresh frozen tissues, followed by differential expression analysis (DESeq2), gene set enrichment analysis (GSEA) and immune cell deconvolution. Radiological evaluation was done by local assessment of immune version of the mesothelioma modified RECIST criteria. Disease control rate (DCR) was defined as the fraction of patients with partial response (PR) or stable disease (SD) compared to those with progressive disease (PD). Results: The DCR and ORR were 69% and 20%, respectively. From mIF analysis, stromal CD66b, CD20, and tumoral CD66b showed the highest area under the curve (AUC = 0.60 ± 0.1) for differentiating DCR groups. For ORR, tumoral CD8+FoxP3+ T-cells demonstrated the highest AUC (0.58) for identifying PR. Patients with tumoral CD66b% scores above the median (>0) had significantly longer progression-free survival (6.2 vs. 4.2 months; HR: 0.63, 95% CI: 0.42–0.97, P = 0.04) and showed a trend toward improved overall survival (HR: 0.65, 95% CI: 0.41–1.0, P = 0.07). These findings were consistent with RNA-seq-derived immune fraction scores. Lasso Cox regression identified natural killer cells, neutrophils, and CD4+ T cells as top predictive features for DCR groups. Additionally, GSEA hallmark analysis revealed significant enrichment of interferon-γ and -α pathways in the DCR PR/SD group (Q < 0.001) compared to PD. Conclusions: High/positive levels of tumoral CD66b+ neutrophils show promise as predictive biomarkers for immunotherapy efficacy in advanced pleural mesothelioma. Larger, independent studies are needed to confirm these findings and validate their clinical utility. Clinical trial information: NCT04300244 .