Background: Metronomic chemotherapy with frequent, low dose administration may not only retain the anticancer feature of conventional, periodic maximum tolerated dose (MTD)-based, parenteral chemotherapy, but also possess other novel properties such as immune enhancement and angiogenesis suppression. D07001 is an absorption, enhanced oral gemcitabine originally developed as liquid form (D07001-F4), then further refined into D07001-softgel capsule to enhance therapeutical applicability. Herein, we report the results of two phase I studies that evaluated the dose-limiting toxicity (DLT), pharmacokinetics (PK), bioavailability (BA) and maximum tolerated dose (MTD) of D07001-F4 and D07001-softgel capsules in subjects with refractory, advanced solid tumors. Methods: Both studies applied 3+3 dose escalation design. Referring to a similar trial by EIi Lilly, in the first Phase I study (NCT01800630), eligible subjects received a bolus intravenous injection of 5 mg gemcitabine in run-in period then D07001-F4 thrice per week for 2 weeks followed by 1 week rest with escalating doses from 2 to 80 mg. Since no dose-limiting toxicity (DLT) was observed even at maximum 80mg dose-level (DL) in this study, another Ib study (NCT03531320) with the uses of refined D07001-softgel capsule and a modified regimen without the resting period was conducted in subjects with refractory gastrointestinal malignancies to investigating the DLT, maximum tolerated dose (MTD) and PK. The test dose of this Ib study was from 40 to 120 mg (40, 60, 80, 120), in which the bridging dose of 40 mg was employed based on its equivalence to the dose of 80 mg in the first Phase I study. Results: A total of 53 eligible patients were enrolled, 34 (10 DL) and 19 (5+1 DL) in the Phase I and Ib studies, respectively. Comparing to parenteral gemcitabine, the mean oral bioavailability of D07001-F4 was about 39%. No DLT was observed in all planned DL, until the 120 mg DL of D07001-softgel capsules, in which 2/6 patients had DLTs (grade 3 hepatotoxicity and grade 3 anorexia, respectively). An addition 100 mg DL was tested (without DLT in a 3 patient cohort) and determined as MTD. Cmax and AUC_(0-48h) of gemcitabine metabolite (dFdU) on C1D1 were largely dose-dependent and comparable for two formulations at bridging dose levels. Despite similar trend for dFdC, more inter-individual variation was observed. The mean dFdC Cmax at MTD of D07001-softgel capsules was 24 ng/mL (0.09 mM) and half-life of 5 hours. Grade >3 anorexia (10.5%) and diarrhea (10.5%) were observed in Phase Ib extension study. Conclusions: The results suggest the feasibility of oral gemcitabine treatment with D07001-softgel capsules, which is well-tolerated for continuous dosing and deserves further clinical investigation. Clinical trial information: NCT01800630, NCT03531320.