HLX04-O: an ophthalmic bevacizumab that was "licensed out", Henlius's light move in the anti-VEGF red ocean

Before discussing HLX04-O, two easily confused things need to be separated.

Shanghai Henlius Biotech (HKEX: 2696) is no longer a biotech that "tells stories to raise money". In 2025 the company had revenue of about RMB 6.667 billion (up about 16.5% year on year), net profit of about RMB 827 million and R&D spending of about RMB 2.492 billion; it has 10 products approved globally, across China, Europe, the US and other markets, spanning oncology, autoimmunity and ophthalmology. Its commercial engine is in oncology — serplulimab (Hansizhuang), trastuzumab (Hanquyou, approved by FDA/EMA and exported), rituximab (Hanlikang) and bevacizumab (Hanbeitai). In other words, this is a mature pharma company with the full trio of "portfolio + own commercial platform + going global".

HLX04-O gives that "oncology bevacizumab" a different life: it is bevacizumab for intravitreal injection (an anti-VEGF-A humanized IgG1 monoclonal antibody, 1.25 mg/0.05 mL), aimed at wet age-related macular degeneration (wAMD). It shares the same active ingredient as the intravenous oncology product Hanbeitai (both bevacizumab), but its formulation, strength and manufacturing process have been specifically optimized for ophthalmology, and it has been developed independently as a new indication; it is also Henlius's first ophthalmic product filed for marketing. The "-O" after the code stands for Ophthalmic.

That last row is the key to understanding this whole article. For Henlius, HLX04-O is not a drug it "sells itself", but an asset it "licensed out to collect money on". So wherever the rest of the article touches on "commercialization, hospital access, sales teams", the protagonist is Essex, not Henlius itself.

Anti-VEGF has been the first-line standard of care for wAMD for nearly 20 years. Ranibizumab, aflibercept, conbercept, faricimab... there is no mechanistic uncertainty at all. This is both good and bad news for HLX04-O: zero target risk equals zero room for differentiation.

Both of its Phase 3 trials used ranibizumab 0.5 mg every 4 weeks as the comparator:

• China Phase 3 (AURA-1, NCT05003245): randomized, double-blind, active-controlled, non-inferiority design, about 60 sites, 407 patients; primary endpoint change in best-corrected visual acuity (BCVA) letters at Week 48; non-inferiority to ranibizumab achieved, with comparable safety.
• International multicenter Phase 3 (AURA-2, NCT04740671): about 90 sites across about 15 countries, 410 patients; primary endpoint BCVA at Week 36. On June 16, 2026, Essex announced that the study met its primary and key secondary endpoints — HLX04-O's BCVA improvement at Weeks 36 and 48 was non-inferior to ranibizumab, with broadly similar ocular and non-ocular safety. With this, both the China and international Phase 3 trials have read out positive.

More realistic competitive pressure comes from both sides:

Above, longer-acting products that are "less hassle". Faricimab (a VEGF-A × Ang-2 bispecific) and high-dose aflibercept 8 mg are both stretching dosing intervals to 12–16 weeks, whereas HLX04-O's Phase 3 used a q4w regimen (the final commercial dosing regimen is subject to the approved label) — even so, being "not inferior" to ranibizumab does not equal "having an advantage" over the long-acting generation.

Below, in-hospital compounded bevacizumab that is "almost free". Off-label split use of Avastin in ophthalmology has been going on for nearly 20 years at extremely low prices. HLX04-O's real difference is not really in efficacy or convenience, but in "compliant, standardized and traceable product accountability" — a story about quality and regulation, not clinical superiority.

The molecular barrier is also weak: bevacizumab's composition-of-matter patent expired long ago, and anyone can make an anti-VEGF biosimilar; the real moat is "the high cost of completing compliant registration for a sterile ophthalmic formulation", not "no one else can make this molecule".

If clinical differentiation is the weak point, then the clarity of the China registration pathway is HLX04-O's most solid piece.

It did not take the biosimilar route but filed the wAMD indication as a "new drug", supported by independent China Phase 3 data; the NDA was accepted by CDE on August 13, 2025. Henlius highlighted a key fact in its announcement: as of the acceptance date, none of the bevacizumab products marketed in mainland China was approved for wAMD. This means that if HLX04-O is approved, it will be the first bevacizumab in China to hold a formal wAMD indication — its most tangible local differentiation in the Chinese market.

But be clear-eyed: the value of this "first on-label bevacizumab" door is constrained by price, not efficacy. Its rivals are not "no available drugs" but "a pile of cheap anti-VEGFs already + almost-free in-hospital compounded bevacizumab". In this landscape, obtaining expedited pathways such as Breakthrough Therapy or priority review is almost unrealistic — those require "clear clinical benefit", while HLX04-O has only ever demonstrated "non-inferiority". So it will most likely have to plan its launch on a standard review timeline, and commercial value will have to be realized through pricing and accessibility rather than a regulatory "head start".

This is the center of gravity of the article. What makes HLX04-O special is not the molecule, but what kind of deal it was designed into.

As early as October 15, 2020, Henlius granted Essex Bio-Technology (an ophthalmology specialty company based in Macau/Hong Kong, with ophthalmic products such as Beifushu and an established presence in China's ophthalmology channels) exclusive rights to develop, manufacture and commercialize HLX04 in the global human ophthalmic field, all in one go. The division of labor: Henlius is responsible for preclinical and clinical development, and Essex for regulatory filings and global commercialization, bearing the bulk of development costs. After being signed in 2020, the agreement was amended in 2023, adjusting the cost cap and milestone structure — the terms below all reflect the amended version.

Where this deal is clever and where it is dangerous should be looked at together.

Clever as a "capital-light hedge". Henlius has neither an ophthalmic sales force nor a retina specialty hospital network — ophthalmology was never its home turf. Handing 80% of development costs and all commercialization execution to an ophthalmology specialist amounts to trading "giving up most of the upside" for "tying up almost no core resources and externalizing late-stage risk". For a company concentrating its ammunition on innovative oncology and autoimmune drugs, this is a reasonable trade-off.

To sum up this chapter in one sentence: for Henlius, HLX04-O is a "low-investment, low-participation-share" call option — most early risk is hedged away by the partner, and while Henlius's share is not high, it grows alongside product volume and is not a capped one-off. The success of execution and the maximum upside of this deal are concentrated more on Essex's side.

A biologics company's confidence rests on manufacturing, and here HLX04-O has both a trump card and a hard hurdle it must clear. First, clarify a frequently misread point: under the license Essex obtained manufacturing rights, but this does not mean the finished product will necessarily be made within Essex's system — in actual clinical supply, both the drug substance (DS) and drug product (DP) lines at Henlius's Songjiang plant have been involved in producing HLX04-O, and the specific manufacturing split at the commercial stage has not yet been settled.

On drug substance, Henlius is strong. The company has 10 approved biologics; its trastuzumab (Hanquyou) has passed FDA/EMA review and the associated on-site inspections, so its monoclonal antibody quality system and release capability are world-class. HLX04-O reuses its mature bevacizumab and biologics manufacturing platform, so drug substance scale-up risk is low. Add the very low ophthalmic dose (1.25 mg, a tiny fraction of the intravenous oncology dose), which consumes very little drug substance per vial, and the cost advantage from large capacity plus low per-unit consumption is the material basis for fighting a price war in a crowded market.

The real hurdle is the "ophthalmic drug product" stage, regardless of who ultimately does the fill-finish. Intravitreal injection has extremely high requirements for sterility, endotoxin, visible and sub-visible particles, protein aggregates and container closure integrity — metrics directly linked to serious local adverse events such as endophthalmitis and intraocular inflammation, and the most tightly regulated aspect of ophthalmic injectables.

On supply continuity, Henlius is the manufacturer of the drug substance and possibly of the drug product too, giving it considerable control over ophthalmic finished-product quality (actually an advantage relative to a pure licensor); what it needs to guard against more is any batch-related quality event on the ophthalmic side spilling over to damage the reputation and supply of the oncology product Hanbeitai.

Any single-product decision has to be seen within the company's portfolio priorities. Put HLX04-O back on Henlius's board and the conclusion is clear: this is a peripheral, financial, opportunistic asset, not a strategic focus.

• Non-core area: Henlius's main line is innovative oncology/autoimmune drugs + biosimilars going global + a self-built commercial platform; ophthalmology is not part of it, and has been licensed out.
• Very low resource usage: 80% of costs and late-stage execution sit with Essex, with almost no crowding-out of the core pipeline; opportunity cost is therefore very low.
• The real synergy is "upstream", not "downstream": it reuses the same bevacizumab drug substance and biologics manufacturing platform, so there is genuine synergy at the manufacturing level; but at the commercial, departmental and brand level, because in-house ophthalmology sales are zero, there is almost no synergy.
• What it validates is a "model": monetizing non-core assets through licensing to maximize platform value — consistent with Henlius's overall capital discipline.

So within the portfolio it looks more like passive cash flow that can be held long-term at low cost, rather than a growth engine requiring management to tilt attention and "bet" on it.

HLX04-O is a very "mature pharma" decision: knowing full well that the ophthalmic bevacizumab business has a limited ceiling — long-acting bispecifics above, almost-free in-hospital compounding below, and only a limited share of sales for itself — Henlius did not force itself to build ophthalmic capabilities, but turned an existing bevacizumab drug substance, through a capital-light early license, into a stream of passive cash flow that consumes almost no resources.

Its story value lies not in "yet another global first", but in demonstrating how a mature domestic pharma company does subtraction on non-core assets: self-commercialize what should be self-commercialized, license decisively what should be licensed, and save limited ammunition and management attention for the oncology and autoimmune lines that truly determine the company's fate. From this angle, the fact that HLX04-O was "licensed out" is precisely what makes it worth telling.