Abstract 1314P
Background
Immunotherapy based on immune checkpoint inhibitors (ICIs) has made a paradigm shift in oncology. However, only a small proportion of patients respond to treatment. Major efforts are being made to identify patients who will benefit from ICI therapy. Previously, we showed that host responses to cancer therapies may promote tumor progression. Here, we used blood-based proteomic profiling and a machine-learning approach to identify host response biomarkers for predicting clinical outcome in ICI-treated non-small cell lung cancer (NSCLC) patients.
Methods
Plasma samples were obtained at baseline (T0) and early on treatment (T1) from a cohort of stage IV NSCLC patients receiving anti-PD-1 therapy (n=134). Proteomic profiling of plasma samples was performed using antibody arrays. Treatment response determination was based on prolonged stability. Machine learning algorithms were applied to data from a subset of the cohort (training set, n=52) to identify a proteomic signature that distinguishes between responders and non-responders. The predictive signature was then validated in an independent cohort (validation set, n=82). Advanced bioinformatic tools were used to identify biological pathways and driver proteins unique to responders and non-responders.
Results
We identified a 3-protein signature that accurately distinguishes between responders and non-responders with an area under the curve (AUC) of the receiver operating characteristics plot of 0.82, and sensitivity and specificity of 0.9 and 0.59, respectively, in the training set. This signature was successfully validated in the independent cohort, with an AUC of 0.83, and sensitivity and specificity of 0.89 and 0.65, respectively. Pathway enrichment analysis revealed activation of inflammatory and metastatic biological pathways in non-responders. Key proteins that drive such pathways and potential combination therapies that may be beneficial for non-responders were identified.
Conclusions
Our study demonstrates the potential clinical value of analyzing the host response to ICI therapy for the discovery of novel predictive biomarkers for NSCLC patient stratification and combination therapy development.
Clinical trial identification
NCT04056247.
Editorial acknowledgement
Legal entity responsible for the study
OncoHost.
Funding
OncoHost.
Disclosure
Y. Shaked: Advisory/Consultancy, Shareholder/Stockholder/Stock options, Officer/Board of Directors: OncoHost. M. Harel, C. Lahav, E. Jacob, E. Issler, O. Sharon: Full/Part-time employment: OncoHost. A.P. Dicker: Advisory/Consultancy, Leadership role: OncoHost; Advisory/Consultancy: Roche; Advisory/Consultancy: EMD Serono; Advisory/Consultancy: Celldex; Advisory/Consultancy: Janssen; Advisory/Consultancy: Cybrexa; Advisory/Consultancy: Self Care Catalysts; Advisory/Consultancy: ThirdBridge; Advisory/Consultancy: Noxopharm; Advisory/Consultancy: Varian; Advisory/Consultancy: Accordant; Advisory/Consultancy: Moleculin; Advisory/Consultancy: Envisino Health Partners . H. Bar: Advisory/Consultancy: OncoHost. All other authors have declared no conflicts of interest.