First-in-human phase I study of HCB101, a 3.5th-generation CD47-SIRPα fc fusion protein: Cytopenia-sparing safety, broad pharmacologic window, and low-dose clinical activity in relapsed/refractory non-Hodgkin lymphoma

Abstract Introduction: Non-Hodgkin lymphoma (NHL) comprises a heterogeneous group of lymphoproliferative malignancies arising from B, T, or natural killer (NK) lymphocytes with diverse biological behaviors. Despite advances in chemo-immunotherapy and cell-based therapies, patients with relapsed/refractory (R/R) NHL after ≥ 2 prior lines face limited options and poor prognosis. CD47 delivers a “don't eat me” signal via signal regulatory protein alpha (SIRPα), suppressing macrophage-mediated phagocytosis. First-generation anti-CD47 antibodies and second-generation wild-type SIRPα-Fc fusions have been constrained by high red blood cell (RBC) binding, on-target anemia/thrombocytopenia, dose-limiting cytopenia, low therapeutic index, and inconsistent results. HCB101 is a 3.5th-generation Fc fusion biologic with a modified SIRPα domain and an IgG4 Fc backbone (using the proprietary FBDB™ platform) designed to selectively block CD47–SIRPα while minimizing RBC binding. This reduces hematologic toxicity and enhances antibody-dependent cellular phagocytosis, bridging innate and adaptive immunity. Preclinical studies demonstrated robust antitumor activity across > 80 CDX/PDX models (including both solid and hematologic malignancies) and a favorable toxicology profile without RBC or platelet binding. We report initial safety, pharmacokinetics (PK), pharmacodynamics (PD), and efficacy in R/R NHL. Methods This ongoing, multinational, open-label Phase I dose-escalation study (NCT05892718) administers HCB101 intravenously once weekly using a Bayesian Optimal Interval (BOIN)-guided 3+3 design across cohorts from 0.08 to 18.0 mg/kg. Eligible adults had histologically confirmed advanced solid tumors or R/R hematologic malignancies, an Eastern Cooperative Oncology Group (ECOG) performance status of 0-2, and adequate organ function. Dose-limiting toxicities (DLTs) are assessed during Cycle 1 (28 days). The primary endpoints are safety and tolerability. Secondary endpoints include PK characteristics, objective response rate (ORR), duration of response (DOR), and progression-free survival (PFS). Exploratory endpoints include CD47 receptor occupancy (RO), immune profiling, and biomarker analyses. This analysis focuses on the R/R NHL subset. Results As of July 17, 2025, 49 patients were enrolled; 12 had R/R NHL and received doses ranging from 1.28 to 18.0 mg/kg (cohorts of 24.0 and 30.0 mg/kg are ongoing). Median prior therapies were 4 (range 2-7). No dose-limiting toxicities (DLTs) or treatment-related serious adverse events were reported. Treatment-related adverse events (TRAEs) were all Grade 1-2, with the most common being leukopenia (25.0%), neutropenia (22.9%), anemia (20.8%), and thrombocytopenia (6.3%), confirming the expected cytopenia-sparing profile. PK analysis revealed dose-proportional exposure. Peripheral CD47 RO was ≥90% at doses of ≥1.28 mg/kg and ≥99% at doses of ≥5 mg/kg. A maximum tolerated dose (MTD) was not achieved through 24.0 mg/kg at the data cutoff, indicating a broad pharmacologic window (low-dose target saturation with preserved high-dose tolerability). One patient with marginal zone B-cell lymphoma achieved a confirmed partial response (PR) at 8.00 mg/kg by Week 8, which deepened to a near-complete response by Week 16 at the same dose (–89.5% tumor reduction on CT, excluding stable orbital lesions); the response is ongoing. Two additional NHL patients treated at 18 mg/kg are pending their first radiographic tumor assessment. Conclusion HCB101, a 3.5th-generation CD47-SIPRα Fc fusion protein, demonstrates a favorable, cytopenia-sparing safety profile (no DLTs, no Grade ≥3 TRAEs), a low-dose PD saturation (≥90% RO at 1.28 mg/kg and ≥99% RO at 5 mg/kg; MTD not reached at >24 mg/kg), and durable monotherapy activity at 8.00 mg/kg in heavily pretreated R/R NHL. Compared with earlier CD47-targeting modalities, including first-generation antibodies (e.g., Gilead, I-MAB) and second-generation wild-type SIRPα-Fc fusions (e.g., Pfizer, ImmuneOnco), HCB101 achieves superior receptor occupancy, safety beyond 24.0 mg/kg, and early clinical efficacy at 8.00 mg/kg – well below the upper dose levels – supporting a broad pharmacologic window. The selective CD47-SIRPα fusion design distinguishes HCB101 from first-generation antibodies, which are limited by anemia and dosing complexity. These data support the dose-expansion in hematologic malignancies and rational combination strategies to leverage macrophage/T-cell synergy.