ABSTRACT
UNLABELLED: AKT (a serine/threonine-specific protein kinase) regulates many cellular processes contributing to cytotoxic drug resistance. This study's primary objective examined the relationship between GSK2141795, an oral, pan-AKT inhibitor, and (18)F-FDG PET markers of glucose metabolism in tumor tissue to determine whether (18)F-FDG PET could be used to guide personalized dosing of GSK2141795. Biomarker analysis of biopsies was also undertaken. METHODS: Twelve patients were enrolled in 3 cohorts; all underwent dynamic (18)F-FDG PET scans and serial pharmacokinetic sampling at baseline, week 2, and week 4 with tumor biopsies before treatment and at week 4. Response was evaluated by RECIST v1.1 and Gynecologic Cancer Intergroup criteria. Biopsy samples were analyzed for mutations and protein expression. RESULTS: GSK2141795 did not significantly influence blood glucose levels. No dose-response relationship was observed between GSK2141795 pharmacokinetics and (18)F-FDG PET pharmacodynamic measures; however, an exposure-response relationship was seen between maximum drug concentrations and maximal decrease in (18)F-FDG uptake in the best-responding tumor. This relationship also held for pharmacokinetic parameters of exposure and 1,5-anhydroglucitol (a systemic measure of glucose metabolism). Phospho-AKT upregulation at week 4 in biopsies confirmed AKT inhibition by GSK2141795. Single-agent activity was observed with a clinical benefit rate of 27% (3/11) and 30% (3/10) CA125 response in the study's platinum-resistant ovarian patients. AKT pathway activation by PIK3CA/PIK3R1 mutation did not correlate with clinical activity, whereas RAS/RAF pathway mutations did segregate with resistance to AKT inhibition. CONCLUSION: GSK2141795 demonstrated an exposure-response relationship with decreased (18)F-FDG uptake and is active and tolerable. This study's design integrating (18)F-FDG PET, pharmacokinetics, and biomarker analyses demonstrates the potential for clinical development for personalized treatment.
