China's Order 818: How 1,700 Hospitals Can Now Legally Charge Patients for Cell and Gene Therapies Without an NMPA Drug Approval — A Regulatory Pathway That Exists Nowhere Else

SHANGHAI — A regulation that took effect on May 1, 2026 has quietly redrawn the line between clinical research and clinical practice in China, with downstream effects that will be felt by every international patient who has ever been told that a cell therapy, a gene therapy, or a gene-editing treatment is not available in their home country. Order 818 — formally the "Regulations on the Administration of Clinical Research and Clinical Translation and Application of Biomedical New Technologies" — gives roughly 1,700 top-tier (3A) hospitals in China the legal authority to clinically translate advanced therapeutics whose mechanisms of action sit at the cellular or molecular level, and to charge patients for those treatments, without first obtaining a National Medical Products Administration (NMPA) drug registration. The order is paired with a National Health Commission (NHC) consultation draft issued on May 8, 2026, which proposes to exclude most clinical data, imaging data, and protein data from China's human genetic resources export controls — a change that would, if enacted, make the data side of an Order 818 program exportable in a way that has not been possible in the past.

No other major regulatory jurisdiction has a comparable pathway. The U.S. Food and Drug Administration requires an IND (Investigational New Drug) application for any clinical use of an unapproved biologic, and a BLA (Biologics License Application) for any commercial use. The European Medicines Agency requires an ATMP (Advanced Therapy Medicinal Product) classification and a centralized marketing authorization. Both pathways can take 5 to 10 years from first-in-human use to commercial approval, with costs typically in the eight-figure range. Order 818 short-circuits this by recognizing that some advanced therapeutics — particularly autologous cell therapies engineered for a single patient, or allogeneic cell banks used to treat a small, well-defined patient population — cannot reasonably go through a mass-market regulatory track designed for small-molecule drugs. Under the new rules, a 3A hospital can take a cell or gene therapy that has cleared a hospital-internal ethics and safety review, treat paying patients with it, and accumulate the post-market evidence that would otherwise have to be collected in registration trials, all under a 1-to-5 year risk observation window before the technology can be shifted to the standard drug track.

The first major English-language report on the order was published by Genetic Engineering & Biotechnology News (GEN) on June 18, 2026, with on-the-record comments from Boyang Wang, founder of the Singapore-based global longevity fund Immortal Dragons; Todd Liao, a partner at Morgan Lewis Bockius's Singapore office; and Jeremy Levin, PhD, chairman of Ovid Therapeutics and Opthera, chairman emeritus of BIO, and author of the recently released Biotech in the Balance: Saving a Strategic Industry in an Age of Distrust. The three framed the order in similar terms: a clear signal that the Chinese government intends to standardize the environment for advanced therapy development, with a tilt toward local innovation and a structure that makes the country more attractive as a site for clinical translation, manufacturing, and partnership. The same article also flagged the contract-law friction that the order creates for existing license agreements — a wrinkle that is unlikely to be visible to most international patients but that will reshape which advanced therapy candidates reach which patients, in which order, over the next 24 to 36 months.

What Order 818 says, in plain language

Order 818 is short by Chinese regulatory standards — the published version runs to roughly 30 articles — but its core mechanism is a reclassification. Until May 1, 2026, any advanced therapy intended for clinical use in a Chinese patient had to clear the NMPA IND process, even if the therapy was autologous (made from the patient's own cells) or single-patient (made for a single named patient with a specific clinical indication). The NMPA IND process — originally designed for small-molecule drugs and later extended to biologics — is appropriate for therapies that will eventually be manufactured and distributed at scale, but it is a poor fit for therapies whose mode of action is inherently personalized. A CAR-T product engineered from a patient's own T cells, an iPSC-derived dopaminergic progenitor cell batch made for a small group of Parkinson's patients, or a base-editing treatment designed around a single patient's mutation profile, are not the kind of products the NMPA drug-registration pathway was built to evaluate. Forcing them through that pathway has historically meant a 5-to-10 year delay and a multi-million-dollar regulatory cost, with no guarantee of approval at the end — delays that, for patients with rapidly progressive disease, are often longer than the patient's expected survival.

Order 818 creates a parallel "tech track" for these products. The hospital that wishes to offer the therapy must be a 3A hospital — a designation that covers roughly the top 10% of Chinese hospitals, including the academic medical centers most international patients already know by reputation: Peking Union Medical College Hospital, Huashan Hospital (affiliated with Fudan University), West China Hospital (Sichuan University), Tongji Hospital (Huazhong University of Science and Technology), and the Cancer Hospital of the Chinese Academy of Medical Sciences, among many others. The hospital's institutional review board and clinical research office must approve the use, and the therapy must enter a 1-to-5 year risk observation window during which post-market data is collected and reported. Once that window is complete, the therapy can be shifted to the standard drug track for mass-market approval — or it can stay on the tech track indefinitely, with the hospital continuing to offer it as an in-house clinical service to patients who meet the indication.

The 1-to-5 year observation window is the part of the order that matters most to international biotech partners. It gives the Chinese hospital system a window of Chinese exclusivity during which the therapy is being clinically translated and used in Chinese patients, before the same therapeutic developer can file for approval with regulators outside China. For Chinese developers, this is a clear competitive advantage: they get the early commercial data and the first-mover experience with a 1-to-5 year head start before the same therapy needs to be filed in the U.S. or EU. For multinational developers partnering with Chinese companies, it triggers a review of existing license agreements — a review that, in the words of Todd Liao, has already begun in earnest. "Legacy agreements were written for a single-pathway world that no longer fully describes the Chinese landscape," Liao told GEN. "If milestones are defined solely by NMPA events, a licensee commercializing [a therapeutic] through hospitals may never trigger them." His recommended fix is to redefine milestones around clinical and commercial outcomes — "first fee-paying patient" rather than "first NMPA-approved commercial sale."

The May 8, 2026 NHC consultation draft: the data side of the equation

Two weeks after Order 818 took effect, the National Health Commission released a separate consultation draft, dated May 8, 2026, that addresses the data side of advanced-therapy development in a way that is at least as significant as the order itself for international partners. The draft proposes to exclude three categories of data from China's human genetic resources (HGR) export controls: pure clinical data, imaging data, and protein data. Only nucleic acid sequence data would remain in scope. The draft also introduces same-day filing confirmations for international clinical trial applications — a process that today can take 60 to 90 days for a formal reply, with the new system promising an answer on the day the application is filed.

The HGR reform matters because China's existing human genetic resources regulations, in force since 2019, require governmental approval before any human genetic material or the data derived from it can be transferred outside China, and the approval process has historically been slow and unpredictable. The practical effect has been that any clinical trial in China that collected genetic data from patients — which is most oncology, rare-disease, and cell-therapy trials — had to be very careful about what was shared with international partners, and was often unable to ship raw data out of the country at all. Under the May 8 draft, clinical readouts, imaging, and protein biomarker data could be shared freely; only sequencing data would still require the HGR approval. For an international biotech running a global trial with Chinese sites, this is a major operational change: it means the Chinese site's contribution to the global dataset can be analyzed alongside the U.S. and EU data in real time, rather than being quarantined behind a separate approval process.

The caveat is that the draft is a consultation draft, not yet law. Todd Liao, the Morgan Lewis partner, framed the broader direction this way to GEN: "The overall regulatory direction is actually loosening, not tightening," citing the new "centralized oversight for clinical research and clinical transformation applications of biomedical new technologies." Jeremy Levin, the BIO chairman emeritus, added a separate observation: "China is clearly trying to standardize and industrialize that environment" — and noted that "if implemented consistently, that could make China a more attractive environment for advanced therapeutics development, and potentially accelerate local innovation, manufacturing, and partnerships." The "if" matters: until the boundary guidelines are published, the order's scope is still being negotiated in practice. But the direction of travel is the part that has the industry paying attention.

The personalization vs. mass-market split: where the boundary will sit

The most important operational question for any biotech considering the Order 818 path is whether the therapy in question will be classified as personalized (and thus eligible for the tech track) or as mass-market (and thus still required to go through the NMPA). The boundary delineation guidelines that will specify this in detail have not been published as of this article's writing, but the working interpretation — reflected in the GEN article and consistent with the language of the order itself — is that therapies intended for mass manufacturing and wide distribution will continue to be governed by the NMPA, while personalized therapies will use the Order 818 pathway.

The personalization criterion is doing a lot of work in that sentence. In practice, "personalized" in this context covers a much wider range of products than the word might suggest. An autologous CAR-T product — engineered from a patient's own T cells, with manufacturing turnaround measured in weeks and a cost that has historically run into six figures per patient — is clearly personalized. So is an iPSC-derived dopaminergic progenitor cell batch made for a small group of Parkinson's patients who share an HLA type, even if the manufacturing process is more standardized than an autologous product. So is a base-editing treatment designed around a single patient's mutation profile — a class of therapies that is moving from "theoretically possible" to "clinically tested" in 2026. So, in many cases, is a gene therapy for an ultra-rare disease, where the patient population is small enough that a mass-market launch is commercially infeasible.

The mass-market side, by contrast, is where the global CAR-T products that have already cleared the NMPA — the satri-cel solid-tumor CAR-T that the NMPA approved in June 2026, the Yescarta-class products already on the Chinese market, and the next wave of allogeneic off-the-shelf cell therapies now moving toward Chinese approval — will continue to live. Those products are manufactured at scale, distributed to multiple hospitals, and priced for a patient population measured in thousands per year. For them, the NMPA pathway remains the right fit, and Order 818 is not the route.

For everything in between — the autologous products that are too personalized for mass manufacturing, the early-stage cell therapies that are too small for a Phase 3 registration trial, the rare-disease gene therapies that will never have a market large enough to justify a full BLA — the Order 818 tech track is the new operating reality. The biotech and hospital partners who understand the boundary well, and who can structure their clinical translation program around the 1-to-5 year observation window, will be the ones who get their therapies to patients first.

The free trade zone wrinkle: where foreign investment can and cannot go

A separate layer of regulation intersects with Order 818 in ways that matter for international biotech partners. China's 2024 Foreign Investment Negative List, the document that specifies which industries are closed or restricted to foreign capital, currently prohibits multinational companies from investing in cell and gene therapy development at a national level. Foreign investment in the cell and gene therapy space is allowed, however, within the free trade zones of Beijing, Shanghai, Guangdong, and Hainan — but only under the NMPA product registration pathway. The Negative List has not yet been updated to reflect the Order 818 tech track, which means that, in practice, a multinational company cannot invest in an Order 818 clinical translation program inside a free trade zone, even though the order itself permits the program to exist.

That gap is, almost certainly, temporary. Boyang Wang of Immortal Dragons, who has been tracking the order since its enactment, framed the regulatory direction in terms that suggest the negative list will be updated in the next 12 to 18 months. "The clear signal is that biotech, healthcare, and biomedical technologies are at the top of senior government officials' priority list," he told GEN. "They want to develop this sector. They want investment, and they want to leverage international capital and expertise. If any of the terms of this [order] will stifle innovation development, they will likely make modifications." Jeremy Levin, the BIO chairman emeritus, agreed: "China intends to compete at the highest levels of biotechnology over the long term."

For now, the practical effect is that international biotech companies with an interest in the Order 818 tech track are watching for the negative list update before they make capital commitments. The Hainan Free Trade Zone — which already hosts the Lecheng medical tourism zone (covered in our 2026-04-21 Hainan 865K visitors article) and a cluster of imported-drug and imported-device pathways — is the most likely first site where the negative list update will be felt. The Hainan provincial government has been the most active of the four free trade zone jurisdictions in courting advanced therapy developers, and the existing Lecheng infrastructure — the special medical-use device pathway, the real-world data collection program, the medical visa system — is a natural anchor for an Order 818 tech-track program.

What this means for international patients today, in concrete terms

For a patient outside China who has been told that an advanced therapy is not available in their home country, Order 818 changes the conversation in three concrete ways. The first is access: a 3A hospital in China can, today, offer a cell or gene therapy that has cleared the hospital's own ethics and safety review, without waiting for NMPA drug approval. This includes autologous CAR-T products for indications that have not yet been approved in the patient's home country, allogeneic iPSC-derived cell therapies in early-phase use, gene-editing treatments for rare genetic diseases, and single-patient therapies engineered for a specific tumor or mutation profile. The hospital-based pathway is not a clinical trial in the conventional sense — the patient is paying for a clinical service, not enrolling in a research study — but it is also not a fully approved commercial product. The patient and the treating physician are operating in a space that sits between the two.

The second change is the cost structure. The hospital-based pathway removes the multi-year development cost that is normally priced into a commercial-stage advanced therapy, which is why the cost of a CAR-T treatment in China has historically been one-third to one-fifth of the U.S. list price (covered in our 2026-03-25 China CAR-T comparison article). For an Order 818 program, the cost is set by the hospital and is typically reflective of the actual cost of the cell or gene therapy product, plus a margin for the hospital's clinical infrastructure. For international patients who are paying out of pocket, the cost is more transparent and more predictable than a U.S. hospital billing cycle, and is usually substantially lower than the equivalent treatment at a U.S. academic medical center.

The third change is the regulatory question. A treatment that is being delivered under an Order 818 tech-track program in China is not, in most cases, a treatment that can be continued at home. Most advanced therapies administered in China cannot be replicated in a U.S. or EU hospital without the home-country regulator's approval. International patients considering an Order 818 program need to be clear with their home physician about what happens after the treatment — the follow-up monitoring, the management of any late effects, the logistics of returning home. The hospital's international patient services office will usually coordinate this in writing before the patient travels, but the patient should expect to commit to a follow-up schedule that may include remote monitoring, periodic imaging, and re-admission to the Chinese hospital for any complications. Our 2026-06-22 Raffles Medical 37,000 foreign patients article covers the broader international patient infrastructure, including the three Raffles hospitals in Beijing, Shanghai, and Chongqing that have built out dedicated international patient departments for this kind of cross-border treatment planning.

The contract-law ripple effect: what changes for biotech license agreements

For international biotech executives and their in-house counsel, the most immediate operational consequence of Order 818 is the need to review existing license agreements with Chinese partners. The standard milestone structure for a cross-border license — an upfront payment, development milestones tied to IND and BLA events, and royalties on commercial sales — was written for a world in which the only path to commercial revenue ran through the NMPA drug-registration track. Under Order 818, a therapy can be clinically translated and sold (charged for) at a 3A hospital before it ever reaches the NMPA, and the revenue from those hospital charges does not, in the existing contract structure, trigger a milestone payment or a royalty.

Todd Liao, the Morgan Lewis Bockius partner, framed this as a contract-amendment wave that is already underway. "I expect a wave of proactive contract amendments," he told GEN. "It becomes contentious only if ignored." His recommended fix — redefining milestones around "first fee-paying patient" rather than "first NMPA-approved commercial sale" — is straightforward in concept but requires careful drafting to avoid disputes over what counts as a "fee-paying patient" (hospital clinical translation program, investigator-initiated trial with a hospital copay, or full commercial launch) and how the milestone value scales across indications and geographies. The first disputes are likely to arise in late 2026 or early 2027, when the first Order 818 hospital programs begin charging for therapies that were already licensed to multinational partners under legacy agreements.

For biotech founders evaluating a new China partnership, the message is to build the contract structure around the tech track from the start — and to plan for the possibility that the most important commercial revenue for the first 1-to-5 years will come from Chinese hospital charges rather than from a global commercial launch. The same planning is also a useful hedge against the alternative scenario, in which the boundary guidelines, when published, end up routing a given therapy back to the NMPA pathway after all.

What to watch over the next 12 to 18 months

The next year is going to be defined less by Order 818 itself — which is already in effect — than by the boundary guidelines, the human genetic resources consultation draft, and the free trade zone negative list update. The boundary guidelines are the most important single document to watch: they will determine, in detail, which therapies can use the tech track and which must go through the NMPA. The NHC consultation draft, if enacted in something close to its current form, will make the data side of cross-border partnerships much more straightforward. The free trade zone update will determine whether international capital can flow into Order 818 programs inside the four designated zones, or whether the tech track remains, for now, a Chinese-domestic capital story.

Operationally, the next 12 months are also likely to see the first wave of Order 818 hospital programs open to international patients. The 3A hospitals that have the existing cell-therapy and gene-therapy infrastructure — the major academic medical centers in Beijing, Shanghai, Guangzhou, Chengdu, and Hangzhou, plus the Cancer Hospital of the Chinese Academy of Medical Sciences — are the most likely first sites. The 1-to-5 year observation window — which is the same window for every therapy on the tech track — means that the first therapies to enter the tech track in 2026 will not see their first commercial NMPA-style approval until 2027 to 2031; in the meantime, they will be available to paying patients in China, including international patients, on a hospital-by-hospital basis.

For a U.S. or EU patient who has been told that a particular cell or gene therapy is not available to them, the practical next step is to ask the treating physician in China whether the therapy in question is being delivered under an Order 818 tech-track program at a 3A hospital, or whether it is being delivered under a clinical trial or an NMPA-approved commercial product. The three pathways have different cost structures, different consent frameworks, and different follow-up requirements — and a hospital's international patient services office can usually clarify which pathway applies within a single email exchange. The pathway that did not exist before May 1, 2026 is the Order 818 one, and the next 18 months will see it become a regular part of the menu for international patients who can travel to China for treatment.