Identification of potential predictive biomarkers of rapid progression and rapid response to anti-PD1 treatment by gene profiling analysis in metastatic melanoma patients.

Abstract

e22068 Background: Anti-PD-1 agents represent a standard treatment for melanoma patients. However, most patients fail to respond, showing in some cases very rapid disease progression. At moment, there are no effective biomarkers that can predict patient's clinical benefit. The aim of this study is to retrospectively identify gene profiling biomarkers that could help to select melanoma patients who most likely respond to anti-PD-1 therapy. Methods: We defined as fast responder (FR) or fast progressor (FP) patients who got clinical response or clinical progression after two cycles of therapy. We collected data from 44 metastatic melanoma patients (21 FR and 23 FP) treated in first-line with anti-PD1 monotherapy (nivolumab or pembrolizumab) at National Cancer Institute of Naples, Italy. Gene expression profiling analysis was performed using NanoString IO 360 panels on PBMCs collected at baseline from 18 patients (10 FR and 8 FP). Patients with ECOG≥2 were excluded. They were all IV stage (5 M1a, 1 M1b, 12 M1c) of which 15 were B-RAF wild-type (83%) and 3 were B-RAF mutated (17%). Statistical associations between treatment response and gene score variables were estimated by Student’s T tests and correction for multiple comparisons by the Benjamini-Hochberg method. Results: Patterns of gene expression were assessed for correlation to response. We compared PBMCs nanostring analysis between FR and FP patients. We found a higher expression of KRas, CD39, IFI16, IL18, FCGR2A, IL1RN, MAP3K8, TLR5, TLR8, MyD88 and NF-kB in FP patients (all with p-value ≤0.005), most of them related to cell proliferation and immunosuppressive mechanism. Instead we found a higher expression of PRF1, PIK3R1, HLA-DPA1, HLA-DRB1, HLA-DOA, CD45RA, LDHB, KIR3DL2, CD2, CD28, CD7, CD27 in FR patients (all with p-value ≤0.01), most of them related to priming and cytolysis. Conclusions: These preliminary data obtained through gene profiling analysis on baseline PBMCs of melanoma patients suggest that a specific gene signature may discriminate FR or FP patients. Our study provides rationale for further investigating gene profiling signature as a potential association for response to immunotherapy.