Mutations Defining Patient Cohorts With Elevated PD-L1 Expression in Gastric Cancer

The immunotherapy agent pembrolizumab has been approved for gastric cancer (GC) patients with recurrent or advanced disease who are PD-L1 positive. Mutations in the primary lesion may drive the expression of immune targets thereby priming the tumor to therapeutic sensitivity. In this study, we aimed...

Full description

Bibliographic Details
Main Authors: Otília Menyhárt, Lőrinc Sándor Pongor, Balázs Győrffy
Format: Article
Language:English
Published: Frontiers Media S.A. 2019-01-01
Series:Frontiers in Pharmacology
Subjects:
TTK
Online Access:https://www.frontiersin.org/article/10.3389/fphar.2018.01522/full
Description
Summary:The immunotherapy agent pembrolizumab has been approved for gastric cancer (GC) patients with recurrent or advanced disease who are PD-L1 positive. Mutations in the primary lesion may drive the expression of immune targets thereby priming the tumor to therapeutic sensitivity. In this study, we aimed to uncover mutations associated with elevated PD-L1 expression in GC patients. Data from 410 GC patients were available, including the mutational spectrum of 39,916 genes and expression values of 20,500 genes. PD-L1 gene expression was compared to the mutational status of each gene separately by using a Mann-Whitney U-test and a Receiver Operating Characteristic test. Only mutations with a prevalence over 5% were considered. Significance was accepted in cases of p < 1E-05 and a fold change over 1.44. Mutations in 209 genes were associated with increased PD-L1 expression. These mutations were enriched in genes related to microtubule-based movement (p = 3.4E-4), cell adhesion (p = 4.9E-4), response to DNA-damage (p = 6.9E-4), and double-strand break-repair (p = 1.6E-3). Mutations in TTK (p = 8.8E-10, AUC = 0.77), COL7A1 (p = 2.0E-9, AUC = 0.74), KIF15 (p = 2.5E-9, AUC = 0.75), and BDP1 (p = 3.3E-9, AUC = 0.74) had the strongest link to elevated PD-L1 expression. Finally, we established a decision tree based on mutations in PIK3CA, MEF2C, SLC11A1, and KIF15 capable to separate patient sub-cohorts with elevated PD-L1 expression. In summary, we identified mutations associated with elevated PD-L1 expression that facilitate the development of better prognostic biomarkers for GC, and might offer insight into the underlying tumor biology.
ISSN:1663-9812