Background: GNS561 belongs to a novel generation of drug blocking cancer cell proliferation by inhibiting late-stage autophagy and dose-dependent accumulation of enlarged lysosomes by interacting with palmitoyl-protein thioesterase-1 (PPT1). Methods: This phase I, multicenter, open-label, dose-escalation trial (3+3 design) explored two dosing schedules: one single oral intake three times a week and twice daily (BID) continuous oral intake of GNS561 in patients with advanced primary and secondary liver cancers (NCT03316222). The primary objective was to determine recommended phase II dose (RP2D) and schedule for further clinical development. The secondary objectives included a preliminary evaluation of the safety, pharmacokinetic (PK), pharmacodynamics (PD), and antitumor activity of GNS561. Results: Nineteen treatment-refractory patients were enrolled and were evaluable for primary endpoint: intrahepatic cholangiocarcinoma (iCCA) (9), hepatocellular carcinoma (HCC) (7), pancreatic ductal adenocarcinoma (PDAC) (2) and colorectal cancer (CRC) (1). Median age was 60, 89% were male and 37% had received 3 or more lines as prior cancer therapies. Dose escalation ranged from 50 mg three times a week to 200 mg BID. No dose-limiting toxicity were observed. Treatment-related adverse events were grade 1-2 gastrointestinal toxicity, primarily nausea/vomiting, occurring in 8 patients (42%) and diarrhea in 4 patients (21%). Occurrence of nausea/vomiting despite antiemetic prophylaxis prevented increasing doses above 200 mg BID. GNS561 displayed favorable bioavailability with interpatient variability (CV%: 13 to 223% and 21 to 98.2% on plasma concentrations on cycle 1 day 1 and cycle 2 day 1 respectively), and dose proportional exposure in plasma. GNS561 concentrations accumulated after multiple administration (2.60 - 9.00-fold) and exhibited a long half-life. Plasma and liver concentrations at doses ranging 100-200 mg BID were comparable to therapeutic exposures in preclinical models. Five patients (3 HCC and 2 iCCA) experienced tumor stabilization according to RECIST 1.1 criteria, including a minor response (-23%). Conclusions: GNS561 RP2D single agent was set at 200 mg BID based on this favorable safety profile and plasma exposure, GNS561 will be next further evaluated in monotherapy and in combination with checkpoint inhibitors considering the autophagic activity restriction of major histocompatibility complex-1 promotion of immune invasion. Clinical trial information: NCT03316222