One Injection, Cancer Gone? The Four Phrases That Decide Whether You Qualify

A widely shared Chinese medical post in July 2026 asked whether a single injection could make cancer disappear. It was promoting a literature app, and most of its text described subscription features. But the question it borrowed is a real one, and it deserves a better answer than a marketing hook. In June 2026 China approved the first CAR-T cell therapy anywhere in the world for a solid tumour. That is a genuine milestone. It is not, however, a treatment that makes tumours vanish in most people who receive it.

The gap between those two statements is where patients and families get hurt. A headline about a disappearing tumour travels across borders in hours. The eligibility criteria that determine whether a specific person can even be assessed travel much more slowly. This article is about that slower, more useful document: the approved indication, read phrase by phrase, and what it means for someone considering travelling to China for treatment.

What was actually approved

In June 2026 the Chinese regulator approved 恺力美 (satricabtagene autoleucel, often shortened to satri-cel and known in trials as CT041). It is an autologous CAR-T therapy, meaning each dose is manufactured from the patient's own T cells. Its target is Claudin 18.2, a protein exposed on the surface of some gastric cancer cells. Before this approval, every CAR-T product licensed anywhere targeted a blood cancer. This was the first to cross into a solid organ.

The full indication is a single long sentence, and every clause in it is a filter: CLDN18.2-positive, HER2-negative, advanced gastric or gastro-oesophageal junction adenocarcinoma that has progressed after at least two prior lines of therapy. Patients who fail any one of those four tests are outside the approved population, regardless of how encouraging the headline sounds.

Gate one: CLDN18.2-positive

The first filter is a laboratory result, not a clinical judgement. A pathologist stains a tumour sample and scores how many cancer cells carry Claudin 18.2 on their surface, and how strongly. Only tumours meeting the trial's threshold count as positive. This is an immunohistochemistry test on tissue from a biopsy or a resection, so a blood draw cannot substitute for it.

This matters practically. Somewhere around 35 to 40 percent of advanced gastric cancers express the target at a level likely to qualify. That single fact excludes the majority of patients before any other criterion is considered. It also means the first concrete step for an interested family is not booking flights but confirming that a recent, adequate tumour block exists and can be stained, or that a re-biopsy is feasible.

Expression is not uniform across a tumour or stable over time. A sample taken at diagnosis may not reflect the disease after two lines of chemotherapy. Clinicians commonly want a reasonably recent specimen, and differences between testing platforms and scoring cutoffs are one reason a result from one laboratory may need repeating at the treating centre.

Gate two: HER2-negative

The second filter is an exclusion. HER2-positive gastric cancer has its own established targeted treatment path, and patients in that group were not part of the pivotal trial. HER2 status is therefore determined before CAR-T is even discussed. A patient who is HER2-positive is not eligible for this product, though they may have other approved options.

Families sometimes read an exclusion as a minor technicality. It is not. It reflects how the evidence was generated: the trial compared satri-cel against chemotherapy in a defined population, and the approval covers that population only. Extending the result to a different biological subgroup without data is speculation, not medicine.

Gate three: after at least two lines of therapy

The third filter concerns treatment history. The label requires that the cancer has progressed after at least two prior systemic regimens. CAR-T here is a later-line option, not a first treatment. A newly diagnosed patient, or someone still responding well to first-line therapy, is not in the approved population regardless of biomarker results.

There is a logic to this beyond regulation. Later-line gastric cancer has few effective options, so the risk and cost of an intensive cell therapy are weighed against a poor alternative. In earlier lines, where chemotherapy and targeted agents still work, that balance has not been tested. Any centre offering this product as a first-line treatment would be operating outside the label, and that should prompt a careful question about the evidence being relied on.

The phrase also has a documentation consequence. "Two lines" means two distinct systemic regimens with recorded progression, not two drugs within one regimen, and not necessarily two cycles of the same treatment. International patients should expect to provide full treatment records, including what was given, when, and the imaging or clinical evidence that the disease progressed.

Gate four: advanced gastric or gastro-oesophageal junction adenocarcinoma

The fourth filter confines the approval to a specific histology and anatomical site. Adenocarcinoma of the stomach or the junction where the oesophagus meets the stomach qualifies. Squamous cell cancers, other histological types, and tumours at other primary sites do not. Earlier-stage disease that can be surgically removed is not the population studied.

Because Claudin 18.2 also appears in pancreatic, ovarian and other tumours, the target attracts interest well beyond the label. That interest is legitimate research, but it is not approval. A patient with CLDN18.2-positive pancreatic cancer is not covered by this indication, and any access would be through a clinical trial rather than a licensed product.

What "cancer gone" would actually mean

The phrase in the headline implies complete disappearance of disease. In trials, that is called a complete response, and it is rare in this setting. The more relevant endpoint is the objective response rate, which counts both complete and partial responses. In the randomised phase II trial supporting the approval, satri-cel produced responses in 22 percent of patients assigned to it, against 4 percent for physician's choice of chemotherapy. Median progression-free survival was 3.25 months versus 1.77 months.

Those are meaningful numbers in a disease with few options. They are not a description of tumours vanishing. Roughly four out of five patients did not achieve an objective response. Even among responders, a response is not a cure, and median overall survival was 7.92 months versus 5.49 months, with a hazard ratio whose confidence interval crossed one. In plain terms, the survival difference was encouraging but did not reach conventional statistical significance in that analysis.

The approved-drugs page sets out the full endpoint table and the product list prices.

The point for a family reading a viral headline is simpler: the honest headline is that a later-line therapy extended control of the disease by about one and a half months on average, with a higher chance of tumour shrinkage, in a narrowly defined group of patients.

What the trial did not show

Beyond the numbers, several limits deserve attention. The trial was randomised but relatively small, conducted in a single country, and compared against physician's choice rather than a single standard regimen. Longer-term survival, quality of life, and outcomes in non-Chinese populations remain less well described. There is no evidence yet that giving the therapy earlier improves on existing treatment, and no evidence supporting its use in other CLDN18.2-positive tumour types.

A single approved product also does not establish a class effect. Other CLDN18.2 CAR-T candidates in development have their own constructs, conditioning regimens and toxicity profiles. Results from one cannot be transferred to another. When a clinic quotes an impressive percentage, the first question is which product and which trial produced it.

The price of being eligible

Meeting all four filters still leaves the practical question of cost. The publicly reported list price for 恺力美 is around ¥990,000 for the product alone. Apheresis, bridging therapy, conditioning chemotherapy, the hospital stay, management of cytokine release syndrome, ICU support and follow-up are billed separately and vary by centre.

For international patients the payment picture is different from that of Chinese residents. China's national insurance and city-level supplementary schemes generally require local residency and enrolment, so visitors normally pay the published product price plus hospital service costs. Some Chinese CAR-T products appear in the commercial insurance innovative drug catalogue that took effect in January 2026, but eligibility for those arrangements should never be assumed for an overseas patient.

The international patients page explains the three payment tiers and the questions to ask before accepting a quotation.

What this means if you are considering treatment in China

The useful sequence is the reverse of the exciting one. Before discussing travel, confirm the four filters with documents. Then ask the treating centre which product it uses, whether the planned use is within the approved label or inside a registered trial, and what the itemised quotation covers. Only after that is it worth comparing hospitals.

1. Is there a recent tumour tissue sample, and has CLDN18.2 staining been performed and scored to the treating centre's threshold?
2. Has HER2 status been formally determined and documented as negative?
3. Can the oncology records show at least two distinct systemic regimens, with documented progression after each?
4. Does the pathology report confirm gastric or gastro-oesophageal junction adenocarcinoma at an advanced stage?
5. Is the proposed treatment within the approved indication, or part of a registered clinical trial with ethics approval?
6. Does the written quotation separate the product price from hospital, conditioning, ICU and follow-up costs?
7. What is the centre's plan for managing cytokine release syndrome and neurological toxicity, and is ICU backup on site?

This article is educational and is not medical advice. Eligibility can only be determined by a treating specialist after reviewing pathology, imaging and treatment records.