Human Sialidase Fusion Protein

HLX79 is Shanghai Henlius Biotech's recombinant human sialidase fusion protein, now in a small Phase 2 trial combined with HLX01 (Henlius's rituximab biosimilar, an anti-CD20 B-cell depleting antibody) versus placebo plus HLX01 in patients with active glomerulonephritis - autoimmune inflammation of the kidney filters [1]. This is a novel-mechanism biologic for autoimmune kidney disease, layered on top of standard B-cell depletion, and one of the first sialidase fusion proteins to reach the clinic outside oncology [2]. The bet is that stripping sialic acid sugars off circulating autoantibodies and immune cells reshapes the autoimmune attack on the kidney; the evidence supporting that bet, in this indication, is thin and rests almost entirely on oncology-derived mechanism rather than published autoimmune preclinical data.

HLX79 is a first-in-class novel compound, not a repurposed approved drug. It has no FDA designations because the program is run in China by Henlius and is currently China-only on ClinicalTrials.gov registration [1]. There is no public indication of an FDA IND or EMA CTA for HLX79, and Henlius has not announced ex-China development plans for this asset. The combination partner HLX01 is already approved in China as a rituximab biosimilar, so the combo arm has a known safety baseline. The Phase 2 study (NCT07038382) is RECRUITING with a target of just 24 patients, and the primary endpoint is safety and tolerability, not efficacy [1]. That framing matters: this is functionally a Phase 1b/2a signal-seeking study dressed as Phase 2. Route of administration and full dosing regimen are not disclosed in the public registry summary. Expect a first safety readout in 2026 and any preliminary efficacy color in 2027 at the earliest. There is no public expected approval timeline; this is years away from a registrational program, and a global filing would require an entirely separate development path through the FDA or EMA. Henlius itself is best known as a biosimilars company (Hanquyou/trastuzumab biosimilar is their flagship), and novel-biologic development is a newer muscle for them [3].

Sialic acids are a family of sugar molecules that sit on the outer surface of most human cells and on the tails of antibodies. Think of them as a sugar coating that tells the immune system 'don't attack - I'm self.' Immune cells read this coating through receptors called Siglecs, which act as 'off switches' that calm the immune response. HLX79 is an enzyme (sialidase) fused to a protein scaffold; the enzyme part chews off these sugar caps. The therapeutic logic in cancer - pioneered by Palleon Pharmaceuticals with E-602 - is to strip sialic acids off tumors so the immune system stops ignoring them [4]. In autoimmune disease, the logic is harder to defend on its face, because stripping the 'don't attack me' signal off healthy tissue should theoretically worsen autoimmunity. The resolution depends on a kinetic and substrate-preference window: pathogenic IgG autoantibodies have a serum half-life of roughly three weeks but are continuously produced and re-deposited in the kidney glomerulus. If HLX79 desialylates circulating IgG fast enough to alter their Fc receptor engagement - Fc receptors are the docking sites on immune cells that read antibody tails - and complement activation before re-deposition, the autoimmune attack should attenuate even as the underlying B-cell production continues. This is the glycosylation lever: small sugar changes on the antibody tail flip the antibody from inflammatory to anti-inflammatory signaling. A secondary hypothesis is that desialylating B cells makes them more visible to rituximab-mediated killing. Both are plausible; neither has been validated in a Phase 2 readout in autoimmune disease. Critically, sialidase is not target-specific - it cleaves sialic acid wherever it finds it, including on red blood cells, platelets, vascular endothelium, and the kidney's own glycocalyx. That non-specificity is the core mechanistic safety risk, and the Phase 2 trial is, in part, an empirical test of whether systemic sialidase exposure produces tolerable on-tissue effects in patients who already have inflamed kidneys. No peer-reviewed preclinical data has been published for HLX79 in autoimmune kidney disease - no mouse nephritis model, no in vitro desialylation data on patient-derived autoantibodies, no kidney biopsy data are in the public record. The IND rationale appears to rest on the oncology-derived mechanism plus general sialoglycan biology rather than on a published preclinical autoimmune package. That is a real evidence gap for a first-in-class biologic entering patients with active disease.

NCT07038382 is a randomized, placebo-controlled Phase 2 study of HLX79 plus HLX01 versus placebo plus HLX01 in active glomerulonephritis, with an enrollment target of 24 patients and a primary endpoint of treatment-related adverse events and abnormal laboratory values [1]. The indication mix is broad - membranous nephropathy and lupus nephritis sit under the 'glomerulonephritis' umbrella but are biologically distinct: membranous is driven largely by anti-PLA2R autoantibodies (a blood test that measures the pathogenic autoantibody driving membranous nephropathy - falling titers mean the autoimmune attack is weakening), while lupus nephritis is a multi-pronged systemic autoimmune attack. Lumping them together in n=24 means efficacy signals will be hard to interpret by subtype. Secondary and exploratory endpoints are not fully detailed in the public registry summary; for trials in this space, the standard efficacy readouts are proteinuria (protein leakage into urine - the primary signal that kidney filters are damaged), eGFR (estimated glomerular filtration rate, a measure of kidney function), and disease-specific autoantibody titers (anti-PLA2R for membranous nephropathy; complement and anti-dsDNA for lupus nephritis). The placebo + HLX01 comparator is appropriate - it isolates the contribution of HLX79 above standard B-cell depletion, which is the only clinically meaningful question. The trial is open and recruiting. The honest read: this is a safety-led first-in-patient study with exploratory efficacy. Don't expect it to define a Phase 3 path on its own. A positive readout buys a properly powered Phase 2b in one chosen subtype, probably lupus nephritis given the larger market and clearer regulatory roadmap.

Our model estimates a 9% chance this drug is eventually approved. That starting point is the historical approval rate for Phase 2 drugs in this area - about 30% - which then gets adjusted up or down based on ten facts about this trial and its sponsor. The number is pulled up by more secondary endpoints than usual, but pulled down by the sponsor's thin approval record, weak earlier-phase results, and heavier-than-usual blinding. The remaining factors are close to average for this stage, so they leave the estimate near where those three downward adjustments landed it.

Efficacy risk is the dominant concern. The mechanism is biologically interesting but unproven in autoimmune disease; n=24 across two distinct glomerulonephritis subtypes is too small to resolve a real efficacy signal even if one exists, and the primary endpoint is safety, not proteinuria or remission. Safety risk is non-trivial and mechanistically expected. Sialic acids are everywhere - on red blood cells, platelets, vascular endothelium, neurons - and a systemically active sialidase could in principle hit any of these. Watch for hemolysis, thrombocytopenia, and infusion reactions. On top of that, layering any immunomodulator on rituximab adds infection risk, particularly for hepatitis B reactivation, which is a known rituximab problem and more prevalent in the Chinese trial population [6]. Execution risk: Henlius is scaling from biosimilars into novel biologics and has limited experience running global registrational autoimmune trials. Commercial risk if it ever gets approved: lupus nephritis already has voclosporin (Aurinia) and belimumab (GSK) approved, with obinutuzumab (Roche) reporting positive Phase 3 REGENCY data showing complete renal response benefit over standard of care, putting it on a near-term filing path [7]. A novel-mechanism add-on to rituximab faces a crowded second-line market and will need either differentiated efficacy or a defined biomarker-selected subpopulation to win payer coverage.

Watch, don't chase. HLX79 is the most clinically advanced sialidase fusion protein in autoimmune disease, which makes it scientifically interesting, but a 24-patient safety-led Phase 2 in two different kidney diseases will not produce a definitive efficacy signal. The signal that would change the assessment: a clean safety readout combined with a biomarker effect - anti-PLA2R titer drop in membranous nephropathy patients, or proteinuria reduction in lupus nephritis above the placebo-plus-rituximab arm. Check back in late 2026 for the first interim safety look and again in mid-2027 for any efficacy color. Henlius (HKEX: 2696) remains independently listed after H-shareholders rejected Fosun Pharma's HKD 5.4B take-private offer in January 2025; the company is repositioning from biosimilars into novel biologics without the organizational shelter of a private structure, and HLX79 is one of the first real tests of that thesis [3]. Commercial framing: lupus nephritis is a ~$1.5-2B near-term US market with two approved agents (voclosporin, belimumab) and obinutuzumab on a near-term filing path on REGENCY data [7]; membranous nephropathy is a smaller orphan market (~30,000 US prevalence) where rituximab is used off-label as standard of care, so a rituximab-add-on faces an awkward positioning question. A positive readout would be more meaningful for the Henlius corporate narrative than for the immediate commercial opportunity, which remains years and several trials away. For investors tracking the sialidase mechanism, Palleon's E-602 oncology readouts (GLIMMER-01, NCT05259696) are the more informative near-term datapoints [4].