Abstract
Introduction: Preclinical data suggest that the single nucleotide polymorphism substituting a glycine for an arginine in codon 388 of the FGFR4 transmembrane domain may increase the proliferation of xenografted neuroendocrine cell lines and decrease their sensitivity to everolimus by modulating STAT3 signaling and the mTOR pathway. Aim: To evaluate the prognostic and predictive values of this polymorphism on everolimus efficacy in patients treated for digestive neuroendocrine tumor (NET). Patients and Methods: This monocentric retrospective cohort included patients with small bowel NET (SBNET) and pancreatic NET (PNET) treated with everolimus (2006-2013). The patients were genotyped by classical sequencing, and mTOR pathway activity was assessed by immunochemistry on formalin-fixed paraffin-embedded samples (PTEN/pPTEN/pAKT/pmTOR/pS6/p4EBP1). Results: Forty-one patients (21 males, median age 57 years) with PNET (n = 28), SBNET (n = 12) or NET of unknown origin (n = 1), grade 1 (n = 8), 2 (n = 27), 3 (n = 3) or unknown grade (n = 3), were studied. At least one 388Arg allele was found in 14/23 PNET and 10/11 SBNET. Progression-free survival in the whole population and the PNET subgroup was not influenced by the presence of one or two 388Arg alleles [HR = 1.31 (0.58-2.99), p = 0.52 and HR = 1.11 (0.45-2.73), p = 0.82, respectively]. Similarly, overall survival was not influenced. Finally, mTOR pathway molecule expression was not modified by the presence of at least one 388Arg allele. Conclusion: The Gly388Arg FGFR4 polymorphism does not seem to have a prognostic value in digestive NET. In addition, it neither predicts the response to everolimus nor modifies the activation of the mTOR pathway.