First-in-human study of GFH018, a small molecule inhibitor of transforming growth factor-β receptor I inhibitor, in patients with advanced solid tumors

The study was conducted at five sites in China, and the site list is provided in the Supplementary Material. Eligible patients were aged 18–75 years with Eastern Cooperative Oncology Group Performance Status (ECOG P.S. ) ≤ 1 and had adequate cardiovascular, liver, and renal functions. For the escalation part, patients with nonmeasurable lesions according to the response evaluation criteria in solid tumors (RECIST) 1.1 were included, while for expansion, patients had to have at least one measurable lesion. The full inclusion and exclusion criteria are listed in the Supplementary Material.

This study was conducted in accordance with the Declaration of Helsinki and the Good Clinical Practice guidelines of the International Council for Harmonization. The local Ethics Committee reviewed and approved the study protocol (Online Supplementary Table S1), and each patient provided written informed consent prior to the study procedure.

This was a phase I, open-label, multicenter study (NCT#05051241) of GFH018 in patients with advanced solid tumors who had failed the standard therapies. The study consisted of a dose-escalation part followed by an expansion part. In the dose escalation part, a 3-day pharmacokinetic (PK) lead-in period was conducted for a single dose in the first treatment cycle. Patients received GFH018 once on cycle 1 day 1 (C1D1) and underwent serial PK testing over the first 3 days. On C1D4, patients were administered the corresponding dose of GFH018 BID 14 days on/14 days off. The first cycle was defined as 31 days, including 3-day PK leading-in period with GFH018 single dosing and then 14d-on/14d-off dosing schedule. The subsequent cycles were 28 days. The starting dose of GFH018 was 5 mg, with eight planned dose levels up to 85 mg (5, 10, 20, 30, 40, 50, 65, and 85 mg). A modified 3+3 design was used to determine the maximum tolerance dose (MTD) and RDE, guided by safety/tolerability and PK data. Dose-limiting toxicities (DLTs) were evaluated during the first cycle at each dose level. The safety of 85 mg BID, 7 d-on/7 d-off, was explored after confirming the safety of 14 d-on/14 d-off. The DLTs were defined as hematological toxicities, including G4 neutropenia lasting for more than 5 days, G3 febrile neutropenia, G4 anemia, G4 thrombocytopenia, or G3 thrombocytopenia accompanied by bleeding, and ≥ G3 nonhematological toxicity (except for diarrhea, nausea, vomiting, and rash recovering ≤ G2 after supportive treatment within 3 days). The DLTs for cardiac toxicities included ≥ G2 cardiac valve abnormalities, ≥ G2 left ventricular ejection fraction (LVEF) decrease, any cardiovascular impairment shown in imaging, and abnormal highly sensitive troponin (hs-Tn) increased to ≥ twice the baseline in two consecutive tests with an interval ≥ 3 days. In the expansion part, patients were orally administered GFH018 RDE twice daily to further evaluate the safety of RDE and other dose regimens. Several types of tumors were predefined based on the biological mechanism of expansion, including nasopharyngeal carcinoma (NPC), biliary tract carcinoma (BTC), and head and neck squamous cell carcinoma (HNSCC). The treatment cycle lasted 28 days. A comprehensive RP2D would be determined based on the data totality.

All enrolled patients were monitored regularly after GFH018 administration. Safety assessments included adverse events (AEs), serious AEs (SAEs), laboratory assessments, vital signs, physical examinations, electrocardiography and echocardiography. AEs were graded according to the NCI Common Terminology Criteria for Adverse Events (NCI-CTCAE) v5.0. All AEs were followed up until they were stable or had recovered to baseline. Adverse events of special interest (AESIs) were predefined as echocardiographic abnormalities such as aggravated stenosis or regurgitation of the heart valves, clinically significant decreases in LVEF, brain natriuretic peptide (BNP) or N-terminal pro-brain natriuretic peptide (NT-proBNP) increases, and hs-Tn increases judged by investigators.

PK analysis included patients who received at least one dose of GFH018 and had measurable plasma concentrations. For patients enrolled in the escalation part of the study, PK blood samples were collected at C1D1 predose; 0.5 h, 1 h, 2 h, 4 h, 8 h, and 12 h postdose; d 2 (24 h), d 3 (48 h), d 4 (72 h), and d 10 predose; d 17 (for 14 d-on/14 d-off regimen) or d 24 (for 7 d-on/7 d-off regimen) predose; and 0.5 h, 1 h, 2 h, 4 h, 8 h, and 12 h postdose. For patients enrolled in the expansion part, PK samples were collected predose on the last administration day and 0.5 h, 1 h, 2 h, 4 h, 8 h, and 12 h postdose on the last administration day. GFH018 levels in plasma samples were analyzed using a validated liquid chromatography-tandem mass spectrometry method. GFH018 PK parameters were determined using noncompartmental analysis methods and calculated using Phoenix WinNonlin Version 8.3.1 (Certara, Princeton, NJ, USA). GFH018 plasma concentrations and PK parameters were summarized descriptively according to the dose level.

Serum samples were collected at baseline, 1 h after the first dosing on C1D1, and 1 h after the last dose of cycles 1, 4, 7, and 10. Serum TGF-β1 levels were analyzed using an enzyme-linked immunosorbent assay (R&D Systems, Cat #DB100B).

Peripheral blood mononuclear cells (PBMCs) were isolated from whole blood samples for pharmacodynamic assessment. Phosphorylated SMAD2 (pSMAD2) in PBMCs was measured using the AlphaLISA® SureFire® Ultra™ assay (Perkin Elmer, Cat #ALSU-PSM2-A-HV).

The endpoints for efficacy, including the objective response rate (ORR), disease control rate (DCR), progression-free survival (PFS), and time to progression (TTP), were assessed per RECIST 1.1. Tumor response was evaluated every two treatment cycles until disease progression, start of a new antitumor treatment, consent withdrawal, loss to follow-up, or study termination for other reasons.

For the safety and efficacy analyses, data from patients who received GFH018 at the same dose in the dose-escalation and expansion cohorts were pooled. The demographic and baseline characteristics were summarized using descriptive methods. Safety and efficacy data were summarized for all patients who received at least one dose of GFH018. The ORR and DCR were summarized by cohort and overall. When the sample size of the analyzed group was ten or more, the Kaplan–Meier method was used to estimate the median PFS and TTP. In this study, treatment-emergent AEs (TEAEs) were defined as AEs that occurred on or after the first dosing date of GFH018 and no later than 30 days after the last dose of GFH018. The incidence and severity of AEs were descriptively summarized according to dose level. The PK parameters of GFH018 were calculated using noncompartmental analysis methods. All plasma PK data were summarized by cohort and visit day or time using descriptive statistics.

In total, eighty-nine patients were screened, among whom 50 patients (14 d-on/14 d-off: 5 mg [n=4], 10 mg [n=3], 20 mg [n=4], 30 mg [n=7], 40 mg [n=4], 50 mg [n=4], 65 mg [n=6] and 85 mg [n=12]; 7 d-on/7 d-off: 85 mg [n=6]) were enrolled and received GFH018 treatment as single agent from August 2019 to August 2022. The median duration of drug exposure for all patients was 36.5 days (range, 5–213 days). The patient distribution is shown in Fig. 1. All 50 patients discontinued the study, with most patients discontinuing the study due to progressive disease (56.0%) and patient decision for end of treatment (22.0%).

The median age of the 50 enrolled patients was 52.5 years (range, 27–68 years). A wide range of solid tumors was observed, with colorectal cancer being the most common (13 patients, 26.0%), followed by non-small cell lung cancer (NSCLC, 8 patients, 16.0%), gastric and esophageal cancers, and endometrial cancer (3 patients, 6.0% each). Forty-seven patients (94.0%) had ECOG scores of 1. Regarding prior anti-tumor history, 37 (74.0%) patients had ≥3 prior systemic treatment lines, and 24 patients (48.0%) had progressed prior to PD-1/L1 therapies. Similar baseline demographic characteristics were observed at each dose level. The baseline characteristics of the enrolled patients are summarized in Table 1.

Approximately 48 patients (96.0%) experienced at least one TEAE, and 13 (26.0%) experienced at least one ≥ G3 TEAE. Treatment discontinuation occurred in two patients due to bilirubin increased and peripheral edema, none of which were related to GFH018, as judged by the investigator. Forty-three patients (86.0%) had at least one treatment-related adverse event (TRAE), and three patients (6.0%) were reported to have ≥ G3 TRAEs, which included anemia, proteinuria, lymphocyte count decrease, hypocalcemia, and blood creatine phosphokinase increase. No grade 4 or 5 TRAEs were observed, no DLTs were observed at any dose level, and the MTD was not reached. TEAEs with an incidence ≥ 10% are shown in Table S2. The most common TRAEs (all G/≥G3, Table 2) were AST increased (18.0%/0), proteinuria (14.0%/2.0%), anemia (14.0%/2.0%), ALT increased (12.0%/0), lymphocyte count decreased (12.0%/2.0%), urine protein present (12.0%/0), GGT increased, ALP increased, and LDH increased (all 10.0%/0). No bleeding events occurred. During the study, nine patients experienced at least one serious AE (SAE), none of which were related to GFH018. Six deaths were reported, all of which were judged by the investigators to be unrelated to GFH018. No significant cardiotoxicity was observed, and all AESIs were mild (Grade 1) without any clinical symptoms or signs.

The safety profile in patients treated with the 7 d-on/7 d-off regimen was not significantly different to that observed in patients treated with the 14 d-on/14 d-off regimen. Higher incidences of ALT/AST and GGT increased were observed in the 14d-on/14d-off than 7d-on/7d-off. However, all of them were Grade 1 or 2 per CTCAE (see Table S3).

GFH018 was rapidly absorbed after oral administration. The median T_max following single dosing was approximately consistent across cohorts ranging from 0.49 to 0.96 h. The geometric mean t_1/2 ranged from 3.11 to 8.60 h. Plasma exposure to GFH018 (AUC_0-t and C_max) increased in an approximately proportional manner within the dose range of 5–85 mg.

Following administration of multiple doses for 7 or 14 d, a steady-state concentration of GFH018 was reached. The median T_max at the steady state ranged from 0.5 to 1.8 h. The geometric mean of t_1/2 ranged from 2.25 to 4.09 h, which did not correlate with the dose level or dosing regimen. The PK profiles after multiple doses were similar to those observed after a single dose. Similar to single dosing, exposure to GFH018 increased in an approximately proportional manner in the dose range of 5–85 mg. No obvious accumulation of GFH018 was observed in any cohort. The geometric mean of the accumulation ratio calculated using the AUC and R_acc across all dose groups was 1.02–1.32. A summary of PK concentrations can be found in Table 3. For the PK profile of GFH018, please refer to Fig. 2.

Out of the 50 enrolled patients, 33 underwent at least one tumor assessment after receiving treatment. No complete response (CR) or partial response (PR) was observed. Nine patients (18.0%) treated with doses ranging from 10 to 85 mg BID achieved stable disease (SD). Among these patients, one with thymic carcinoma receiving 65 mg BID achieved significant tumor shrinkage (the maximum target lesion decreased by 18.4%). Twelve patients receiving 85 mg BID, 14 d-on/14 d-off, showed a DCR of 25.0%, and the median PFS was 1.8 months (90% CI, 0.82–3.75) (Figure S1). The efficacy data and swimming plots are shown in Table 4 and Fig. 3, respectively.

Baseline serum TGF-β1 data were obtained from 49 patients. The median baseline TGF-β1 level was 27881 pg/mL (range: 10442–58205) in patients who achieved SD and 29213 pg/mL (range: 8645–61434) in patients with progressive disease (Figure S2). This indicated that baseline serum TGF-β1 levels were not associated with clinical efficacy. On-treatment serum TGF-β1 levels were also measured. There was no clear upward or downward trend observed for TGF-β1 levels in either the SD or PD group (Figure S3). pSMAD2 in PBMCs was originally designated as a pharmacodynamic marker of TGF-β pathway inhibition. However, quality control issues associated with the assay preclude data interpretation.

The RP2D was determined to be 85 mg BID, 14 d-on/14 d-off based on the safety, PK, and efficacy data. There was no obvious correlation between the incidences of TRAEs and dosages. GFH018 PK-PD analysis in mouse H22 models showed that the efficacy was observed with approximately 50% inhibition of pSMAD3 at a concentration reaching 125 ng/ml after the 1-h dose. The geometric mean of the Cmin of GFH018 was 189.9 ng/ml at the steady state for 85 mg BID, 14 d-on/14 d-off, indicating that after repeated administration, the concentration of GFH018 could be continuously maintained above the effective concentration from preclinical studies. With respect to safety, we have noted that proteinuria occurred across different dosages (Table S4) although generally mild in severity. It seems that no clear dose-dependency for proteinuria, however, one patient experienced grade 3 proteinuria after GFH018 treatment. Considering the future clinical development strategy, i.e., in combination with immune checkpoint inhibitors (ICIs), there might be increasing risk due to overlap toxicity using GFH018 in combination with ICIs at higher dose levels than 85 mg BID, 14d-on/14d-off. Based on the data totality, 85 mg BID 14 d-on/14 d-off is considered as an appropriate dosing regimen for phase II.