SAR444200 Demonstrates Safety, Signals of Clinical Activity in Advanced GPC3+ HCC, Other Solid Tumors

The anti-GPC3/T-cell engager SAR444200 demonstrated a manageable safety profile across all evaluable dose levels in patients with GPC3-expressing advanced solid tumors, as well as preliminary signals of antitumor activity in a subset of patients with heavily pretreated hepatocellular carcinoma (HCC), according to initial data from the monotherapy dose escalation cohort (Part 1A) of a phase 1/2 study (NCT05450562).

Findings presented at the 2025 ESMO Targeted Anticancer Therapies Congress showed that, among patients in Part 1A (n = 33), the total incidence of treatment-emergent cytokine release syndrome (CRS) was 9.1%. Three grade 3 or higher instances of CRS were reported as dose-limiting toxicities (DLTs) with 70 mg of SAR444200. Specifically, 2 CRS events were observed in patients treated at dose level 6A (DL6A; n = 3), and 1 occurred at dose level 7A (DL7A; n = 4) during lead-in dosing. These events were resolved completely with corrective treatment, including steroids and vasopressors, with or without tocilizumab (Actemra), and led to the permanent discontinuation of SAR444200.

Furthermore, following treatment, 27.8% of patients (n = 18) exhibited at least a 50% decrease in levels of alpha-fetoprotein (AFP), a commonly used serological biomarker for HCC.

“A reduction in the HCC marker AFP in a subset of heavily pretreated patients [with HCC] suggests preliminary antitumor activity,” lead study author Khaldoun Almhanna, MD, MPH, and colleagues stated in a presentation of the data. “Our data provide initial clinical evidence for SAR444200-mediated targeting of GPC3-expressing solid tumors.” Almhanna is a gastrointestinal oncologist at the Brown University Health Cancer Institute, as well as an associate professor of Medicine, clinician educator at The Warren Alpert Medical School of Brown University.

Rationale and Study Overview

GPC3 is a glycoprotein involved in early development and is largely absent in normal adult tissues. However, GPC3 expression is prevalent in several pediatric malignancies and adult solid tumors, including HCC.

The novel Nanobody® molecule SAR444200 is designed to co-engage T cells and GPC3-expressing tumor cells, leading to T-cell–dependent cellular cytotoxicity. Preclinical studies have shown that SAR444200 generated specific and potent antitumor activity in vitro, as well as significant tumor regression in GPC3-expressing xenograft mice models.

This first-in-human, dose-escalation/-expansion study investigates the safety, pharmacokinetics, and early efficacy of SAR444200 for patients with previously treated GPC3-expressing advanced solid tumors. To be eligible for Part 1 of the study, patients are required to be at least 18 years of age, have positive GPC3 expression on tumor tissue with an H score of 1 or greater; have metastatic and/or advanced HCC and non-HCC; and have an ECOG performance status (PS) of 0 or 1. They cannot be amenable to available standard therapies.

Upon enrollment, patients are assigned to 1 of 2 cohorts. In Part 1A (n = 33), following lead-in dose administration, patients are intravenously (IV) administered SAR444200 monotherapy at weekly doses of 3 mg (DL1; n = 4), 1 mg (DL1A; n = 4), 2.5 mg (DL2A; n = 4), 4.5 mg (DL3A; n = 4), 18 mg (DL4A; n = 4), 36 mg (DL5A n = 6), DL6A, or DL7A over a 21-day cycle.

In Part 1B, patients will go on to receive SAR444200 plus atezolizumab (Tecentriq), provided that biological evidence of T-cell activation was observed.

The incidence of DLTs and adverse effects (AEs) serve as primary end points in Part 1A, and key secondary end points include overall response rate, duration of response, pharmacokinetics, and the incidence of anti-drug antibodies (ADA).

The current analysis includes data from patients enrolled onto Part 1A of the study. The median age was 61 years (range, 29-76) and the majority of patients were male (75.8%), Asian (63.6%) had an ECOG PS of 0 (69.7%). The median GPC3 expression membranous H-score was 31 (range, 0-300), and the median number of prior therapies was 4 (range, 2-8).

The most common tumor type in this patient population was HCC (69.7%), followed by gallbladder adenocarcinoma (6.06%), gastrointestinal stromal tumors (6.06%), breast cancer, cecum adenocarcinoma, ascending colon adenocarcinoma, gastric adenocarcinoma, lung adenocarcinoma, and esophageal adenocarcinoma (3.03% each).

The duration of target infusion was approximately 1 hour for cycles 1 and 2, after which it was reduced to 30 minutes. Several premedications were used. For DL1, 100 mg of IV prednisone or equivalent were administered; this was replaced by 15 mg of IV dexamethasone 15 mg or equivalent for the remaining dose levels. Orally administered acetaminophen at 650 mg to 1000 mg was also permitted, as was 25 mg to 50 mg of either IV or oral diphenhydramine.

Additional Safety and Pharmacokinetic Data

Any-grade AEs were experienced by 97% of patients in Part 1A, 48.5% of whom experienced grade 3 or higher treatment-emergent AEs (TEAEs). The most common grade 3 or higher TEAEs included anemia (9.1%), disease progression (9.1%), and pneumonitis (6.1%).

Any-grade treatment-related AEs (TRAEs) were reported in 93.9% of patients, 15.2% of which were grade 3 or higher. Serious TRAEs occurred in 24.2% of patients; 2 patients (6.1%) at DL1 and DL2A, respectively, experienced fatal TEAEs.

When further broken down according to dose level, any-grade AEs were reported in 75% of patients at DL1A, and 100% of patients from all other dose levels. Grade 3 or higher TEAEs were reported in 50%, 25%, 25%, 50%, 50%, 50%, 100%, and 50% of patients at DL1, DL1A, DL2A, DL3A, DL4A, DL5A, DL6A, and DL7A, respectively. Any-grade TRAEs were observed in 50% of patients at DL1A and all patients at the other dose levels. Of these, 25%, 100%, and 25% were grade 3 or higher at DL1A, DL6A, and DL7A. Serious TRAEs were reported in 25%, 50%, 66.7%, and 75% of patients at DL1A, DL2A, DL6A, and DL7A.

Grade 3 or higher anemia was reported in 25%, 25%, and 16.7% of patients at DL1A, DL3A, and DL5A; grade 3 or higher disease progression occurred in 25%, 25%, and 16.7% of patients at DL1, DL2A, and DL5A; and grade 3 or higher pneumonitis was experienced by 25% and 33.3% of patients at DL1A and DL6A, respectively.

At the data cutoff of October 15, 2024, the incidence of ADAs was 84%, with a median time to ADA development of 3 weeks (lower limit of quantification = .154). Mean SAR444200 pharmacokinetic exposure increased with dose escalation; however, by the sixth administration on day 1 of cycle 2, a decline in exposure was observed at lower dose levels in association with high ADA titers. In contrast, sustained exposure was maintained from the DL5A onward, where ADA titers remained low.

During lead-in doses across DL1 to DL7A, an increase in interleukin-6 and interferon-gamma levels was observed, with maximum concentrations reaching 1326 pg/mL and 461 pg/mL, respectively. This supported the diagnosis of CRS, study authors added. Additionally, a dose-dependent increase in mean SAR444200 pharmacokinetics was noted.

Dr Almhanna did not report any relevant disclosures.

Reference

Almhanna K, Hong JY, Chenard-Poirier M, et al. Phase 1/2, open-label, first-in-human study of the anti-GPC3 T cell engager SAR444200 in patients with advanced solid tumors: Updated safety and pharmacokinetic analysis. ESMO Open. 2025;10(suppl 2):1-7. doi:10.1016/esmoop/esmoop104162