Withdraw, refile, fall behind: the localization trade-off behind eftansomatropin alfa

Eftansomatropin alfa’s early R&D code was GX-H9, and within the Tianjing system it has long been called TJ101. Its molecular design comes from Korea’s Genexine and its hyFc® platform — fusing human growth hormone to a specially designed human immunoglobulin fragment (a hybrid of the human IgD hinge and IgG4 Fc) to extend its duration of action in the body and allow once-weekly dosing.

There are three key nodes on this ownership chain.

The first node is in 2017. I-Mab (Tianjing Biotech’s predecessor) signed a license agreement with Genexine for China rights to four products including this molecule. According to I-Mab’s annual report filed with the US Securities and Exchange Commission, the combined upfront for the four licensed products was US$13 million, milestones related to this molecule totaled US$40 million and are tied to net sales targets, and the agreement term is 30 years.

The second node is in November 2021. I-Mab reached a strategic commercialization collaboration with Jichuan Pharmaceutical: Jichuan paid an upfront of RMB 224 million, plus up to RMB 1.792 billion in development, regulatory and sales milestones, for a total of up to RMB 2.016 billion; the two share mainland China commercialization profits 50/50, implemented via tiered low-double-digit sales royalties. The division of labor is that Tianjing acts as marketing authorization holder and supplies Jichuan at an agreed cost, while Jichuan is responsible for commercial promotion and new indication expansion in mainland China.

Note the timing of this deal: it was signed in November 2021, while the pivotal Phase 3 topline results were not released until August 2023. In other words, this deal was locked in about twenty-one months before the data readout. Locking in bought certainty, but gave up bargaining room after the readout.

The third node is in 2024. Nasdaq-listed I-Mab signed an agreement with Tianjing Biotech (Hangzhou) to divest its China assets and business operations for total consideration of up to US$80 million, tied to subsequent regulatory and sales milestones; the transaction closed in April 2024. Two independently operated companies resulted: I-Mab, which remained overseas, and Tianjing Biotech, which took over the China team and assets. Eftansomatropin alfa went to the latter along with the China assets.

Stringing the three nodes together, Tianjing’s real role in this asset is: a licensee upstream, a profit-sharing partner downstream, and in its own hands R&D, manufacturing and the marketing authorization. This position defines the boundaries of every strategic move that follows.

The pivotal Chinese Phase 3, code-named TALLER, was a randomized, open-label, active-controlled multicenter study that actually enrolled 168 children with growth hormone deficiency at 31 centers in China. Eligibility was boys aged 3–10 and girls aged 3–9, Tanner stage I, with height more than 2.0 standard deviations below the mean for age and sex. The experimental arm received 1.2 mg/kg once weekly, the control arm Norditropin 0.034 mg/kg once daily, observed over 52 weeks. The study started in early 2021 (the registered study start date is January 25, 2021), the company announced first patient dosed in February 2021, and the study was completed in July 2023.

This result needs to be read twice.

On the first read, it passes: the primary endpoint was met, and once-weekly dosing is not inferior to once-daily dosing in efficacy — exactly the form of evidence needed to register a long-acting formulation.

On the second read, the problem appears in that confidence interval — its lower bound is −0.06, crossing zero. This means “non-inferiority holds” and “superiority cannot be claimed” are both true at the same time. At the regulatory level this does not affect approval; but at the NRDL negotiating table and in medical promotion, it is a ceiling.

The ceiling is clearer against a comparator. Visen Pharmaceuticals’ lonapegsomatropin reported, in its Chinese Phase 3, an annualized height velocity of 10.66 cm/year at 52 weeks versus 9.75 cm/year for the daily comparator, a difference of 0.91, 95% confidence interval 0.37–1.45, p=0.0010 — a clear superiority result. The two studies differ in population, comparator and statistical setup, so the numbers cannot be compared directly side by side; but in label wording and in payers’ value arguments, “non-inferior” and “superior” are assets of two different grades.

There is also a gap that deserves more attention: immunogenicity. Clinical registry information shows the study listed immunogenicity as a secondary objective, but the related results have still not been publicly disclosed. Competitors have already turned this into a selling point — Xiamen Amoytop’s Y-shape pegpesomatropin (Yipeisheng) publicly touted at approval that its incidence of new anti-drug antibodies and neutralizing antibodies were both 0.0%, versus 26.4% and 7.9% for the short-acting comparator.

This is especially critical for eftansomatropin, because one of the original design goals of hybrid Fc carriers like hyFc is precisely to reduce immunogenicity while extending half-life. If the data are good, now is exactly the time to bring them out; if they keep being withheld, the market will interpret it in the least favorable way.

When eftansomatropin filed its import marketing application in December 2024, only one long-acting growth hormone was approved on the market. By the time it filed its localized application in August 2026, that number had become four. Almost all the change happened in those fourteen months.

In their announcement, the companies describe it as “China’s first and only fusion protein long-acting growth hormone”. The claim holds, but on two qualifiers: “China” and “fusion protein”. Of the four approved products, two take the PEGylation route, one is a sustained-release prodrug and one is an albumin-binding analogue; indeed none is a fusion protein. But remove the qualifiers and it is wrong — globally, somatrogon, which uses CTP fusion technology, was approved in the US, EU and Japan back in 2023, although there is no public record of approval in China. So the only accurate formulation is “China’s first fusion protein long-acting growth hormone” — not “the first long-acting”, nor “the world’s first fusion protein”.

Even more damaging than the number of competitors is price.

The shock of price restructuring showed up immediately in the leading company’s financial statements. Changchun High-Tech’s 2025 revenue was RMB 12.083 billion, down 10.27% year on year, with net profit attributable to shareholders of RMB 155 million, down 94%; within this, GeneScience, the growth hormone business entity, had revenue of RMB 9.819 billion, down 7.98%, and net profit attributable to shareholders of RMB 487 million, down 81.83% — whereas GeneScience’s revenue in 2024 was still RMB 10.671 billion.

The demand-side base has not collapsed. China has nearly 7 million children with short stature, with a standardized diagnosis and treatment rate below 5%, a clear gap from the 20%+ level in Europe and the US; the burden of about 313 injections a year for daily formulations is indeed the reason long-acting formulations exist. But the long-term trend on the other side is equally clear: births have fallen from 18.83 million in 2016 to 9.54 million in 2024.

Put the two sides together and the conclusion is blunt: the volume is still there; the price is gone. Eftansomatropin is entering a market where penetration still has several-fold room, but the central unit price has already dropped by about 70%.

And it may not fall only once. Growth hormone has been touched by volume-based procurement before — the 2022 inter-provincial alliance procurement led by Guangdong already included recombinant human growth hormone. In 2026, nationwide biologic alliance procurement is also advancing; public reports indicate the first batch is mainly monoclonal antibodies, and there is no public information showing growth hormone has been included, so this can only be regarded as a directional risk requiring ongoing tracking, not a fait accompli.

But for a product that is still more than a year from possible launch, this risk exists in a very concrete way: its cost structure must be able to withstand the next price cut, not just this one.

The clinical risk for this product has essentially been cleared — Phase 3 is done and the primary endpoint was met. The remaining gates lie in CMC, timelines and the calendar. Ranked from high to low risk, there are seven places to watch.

For many domestic assets, “China first” is a strategic choice: get it working in China first, then talk about going abroad. Not for eftansomatropin. For it, China is not the first choice but the only choice.

The reason is written in the rights structure. Upstream, Genexine’s license limits the territory to China, with global rights still held by the licensor and its Korean partners; downstream, mainland China commercialization rights were granted to Jichuan Pharmaceutical in 2021. On this chain Tianjing can neither go outward nor control the end market. What it controls is the middle section: R&D, manufacturing and the marketing authorization.

Acknowledging this structure actually makes the problem clearer. Since there is no option to go abroad, the asset’s valuation anchor can only be sales cash flow within China; since the end market is in the partner’s hands, only two variables remain for Tianjing to optimize — make costs low, make the label broad.

“Making costs low” means localization, which has already been done. “Making the label broad” has not started: idiopathic short stature, small for gestational age, Turner syndrome and adult growth hormone deficiency are all mature expansion directions for same-class products, but each requires new registration studies. With the price of the lead indication already capped, indication expansion is the only remaining source of scale growth for this asset.

China’s patient base is more than enough to support this — 31 centers and 168 enrolled patients have already proven execution capability, and the center network can be reused directly. What is missing is not feasibility but a start date.

Back to the anomalous move at the start. Lay out the public timeline, and the logic is actually quite clear.

August 2024: Hangzhou manufacturing base obtains a drug manufacturing license

December 9, 2024: Import marketing application accepted (JXSS2400105)

December 29, 2025: Full localization strategy announced: JXSS2400105 voluntarily withdrawn; bioequivalence trial between imported and local product approved and started; technology transfer and process optimization fully completed

August 12, 2026: Localized marketing application accepted (CXSS2600122), about eight months after the strategy announcement

What this trade sold was time. What can be established is that twenty months separate the two acceptance dates: the import application was accepted on December 9, 2024 and the localized application on August 12, 2026. As for “what would have happened without withdrawal”, that can only be conjecture, not a conclusion — with reference to review timelines for same-class products, the import application might have landed within 2026 and caught that year’s NRDL application window, but review time is affected by rounds of supplementary requests, on-site inspection scheduling and many other factors, so this counterfactual path cannot be verified. Only the outcome itself is certain: the current timeline has been pushed beyond 2027.

What it bought back was a cost structure. The companies’ stated reason was “comprehensive advantages in manufacturing supply, cost optimization and commercialization”, and they said localization “has become an inevitable trend in the development of the growth hormone industry”. Translated into business terms: domestic manufacturing means lower unit costs, a shorter release chain, and freedom from the rhythms of import customs clearance and port inspection.

In 2024, this trade would have looked foolish. In 2026, it may be the only correct choice.

Imagine the other path: eftansomatropin is approved in 2026 as an import, and then, with the cost structure of an imported product, faces two NRDL rivals already down to RMB 853.2 and about RMB 900. It would either quote a price in negotiation that it cannot sustain, or give up on the NRDL and fight four long-acting products for the self-pay market. That would be the real dead end. From this angle, the December 2025 withdrawal was not a mistake but a stop-loss.

On execution, results have indeed been delivered: it took only eight months from strategy announcement to re-acceptance, completing technology transfer, process optimization, localized production preparation and all registration filing work in that time. This speed is top-tier among domestic companies, and the most convincing part of this project so far.

But for the whole trade to truly stand up, there is an as-yet unverified premise: local manufacturing costs must be significantly lower than import costs. For now this rests only on the companies’ statements, with no verifiable public data. If the yield and unit cost of the local process do not open up a gap, then what twenty months bought would be nothing more than a more cumbersome filing path.

For most domestic biotechs at this stage, the first constraint is cash. Not for Tianjing Biotech.

On September 30, 2025, the company completed a Series C2 round of nearly RMB 600 million, led by a fund under CICC Capital. On April 20, 2026, the company reached an agreement with Biogen for exclusive Greater China rights to felzartamab, receiving a US$100 million upfront and up to US$750 million in milestones (up to US$850 million in total), plus tiered royalties from mid-single-digit to low-double-digit percentages. On August 13, 2026, the two jointly announced that felzartamab (Jingfei®) had been approved by the NMPA in combination with lenalidomide and dexamethasone for adult patients with multiple myeloma who have received at least one prior line of therapy; per the announcement this is the product’s “first approved indication globally”, with Biogen responsible for commercialization and Tianjing supplying commercial product from its Hangzhou base. In addition, the second marketing application for TG103, the GLP-1 fusion protein partnered with CSPC, was accepted on June 22, 2026.

This set of facts has two layers of meaning.

The first layer is good news, but should be viewed by nature: the C2 round is equity funding already received; what has actually been banked from the Biogen deal is the US$100 million upfront, while the remaining up to US$750 million are contingent milestones that can only be realized once conditions are met, and the cost advantage currently rests only on company statements with no public data behind it. Even counting only the confirmed portions, the remaining investment for the eftansomatropin project — responding to review, manufacturing validation and launch preparation — is far smaller in magnitude than in the clinical stage. In this sense, money is most likely not this project’s binding constraint.

The second layer needs to be viewed calmly: within the company, eftansomatropin is no longer the flagship. The company’s major events in 2026 are concentrated on felzartamab and TG103. Eftansomatropin is closer to a “steady cash-flow asset” — its value lies in moving the company from “one product on the market” to “multiple products on the market”, not in generating valuation upside.

So this project’s real constraint is not money but the calendar. The NRDL is a gate that opens and closes once a year, and every miss means a full year of price and channel disadvantage. And competitors will not stand still: Weizhengao has already written its first prescription nationwide, Sogroya is fully launching in summer 2026, and both will fight for their positions in the next round of negotiations.

In this race timed by the calendar, Jichuan Pharmaceutical is the heaviest card in Tianjing’s hand. Growth hormone is a category highly dependent on physician education, patient follow-up and long-term adherence management; prescribing is concentrated in the relatively narrow department of pediatric endocrinology, and out-of-hospital self-pay and patient management systems have long mattered as much as in-hospital access. Jichuan’s channel depth in pediatric medicines lands precisely where this category is most demanding — which is what that deal worth up to RMB 2.016 billion really bought.

Conversely, this card also means Tianjing neither needs nor should build its own growth hormone sales force. With four products on the market and two on the NRDL, self-commercialization has no room for a return. Here, partnered commercialization is not a second-best option but the only one whose numbers add up.

The story of eftansomatropin alfa is not a story about innovation, but a story about timing.

Its Phase 3 data passed in 2023, and still pass in 2026. What has changed is not the molecule but the world around it — four long-acting products entering one after another, the NRDL cutting unit prices by half to three-quarters, and the industry leader’s profits falling by more than 80%. The market no longer pays for “passing”.

Facing this change, Tianjing and Jichuan made a not-so-pretty but possibly correct decision: withdraw an application already halfway through the queue and go back to redo the costs. They gave up a launch timing destined not to fetch a good price, in exchange for a cost structure that might still make money.

Whether this decision was right will ultimately be revealed in two places: one is the data from the bioequivalence study, the other is the number at the 2027 NRDL negotiating table.

Until then, every claim of “first” and “only” is just a game of qualifiers.