• The content of this article is based on an interview with Dr. Zhu Xun conducted on August 22, 2026

On August 19th, Moderna and Merck announced jointly that their Phase III clinical trial of the individualized mRNA cancer vaccine Intismeran (V940/mRNA-4157) combined with K drug for the treatment of high-risk melanoma had met both endpoints of recurrence-free survival (RFS) and distant metastasis-free survival (DMFS).

The news sent the capital market into a frenzy. Moderna's market capitalization surged by $44 billion in a single day, skyrocketing from $25 billion to nearly $70 billion. The following day, the biotech sector on both the A-share and Hong Kong stock markets experienced a broad-based upswing, with multiple vaccine stocks hitting their daily trading limit.

However, the euphoria didn't last long. After experiencing a surge on the first day, the share price of Moderna fell back by 23% the next day, with its market capitalization hovering around $55 billion. Domestic concept stocks also saw an immediate correction.

The swift transition of market sentiment from fervor to calmness amidst fluctuations reminds one of a story that has been repeated countless times over the years: a breakthrough news ignites expectations, capital pours in, and then the data is scrutinized inch by inch, causing the fervor to subside.

"History always repeats itself in a surprising way," said Zhu Xun, the founding chairman of the Tongxieyi New Drug Talent Club.

This immunologist, who has been deeply involved in the industry for over three decades, has witnessed the advancements in therapeutic antibodies and vaccines over the years, and has personally experienced the repeated "highlights" and "falls" in the field of tumor vaccines such as Provenge and MAGE-A3.

He is not only well-versed in the over-century exploration history of tumor vaccines, but also possesses his own systematic framework for judgment.

Zhu Xun believes that when evaluating any innovative project, the most crucial factors are three dimensions - the maturity of science, the acceptability of regulations, and the protectability of patents, all of which are indispensable. Beyond these three bottom lines, the accessibility of technology, the competitiveness of the market, the expansion of products, the achievability of the team, and the patience of capital should also be included as auxiliary considerations.

When the entire capital market was in a collective frenzy over a positive Phase III result, Zhu Xun chose to return to the most fundamental scientific logic, stating, "The complexity of tumor immunity far exceeds superficial data. We must ask whether the underlying biological mechanisms can withstand scrutiny." He combined immunological principles and mechanisms to dissect the three essential conditions for therapeutic tumor vaccines to truly function.

In his view, exploration should be encouraged, but it also requires calm objectivity and reliance on data, rather than being swayed by media hype and capital market speculation. What should be tracked is not company news or industry reports, but original papers. "One needs to examine the complete clinical trial protocol and FDA review report for every detail."

Moreover, scientific breakthroughs are merely the starting point. After that, there are a series of more practical questions that need to be answered: Can the success achieved in melanoma be replicated in other cancer types? How can highly personalized products be commercialized in the future? How can production be standardized, costs be controlled, and accessibility be ensured?

01. The century-long journey of tumor vaccines

The idea of using vaccines to fight cancer dates back to the late 19th century.

In 1891, William B. Coley, the father of cancer immunotherapy, discovered that the erysipelas toxin secreted by pyogenic streptococcus could cause tumor regression in patients with advanced sarcoma. Based on this discovery, he developed "Coley's Toxin" - a mixture of bacteria (including pyogenic streptococcus, Bacillus, etc.) that was heat-inactivated, laying the foundation for the research of tumor vaccines.

Over the next 30 years, Coley's toxin was used in the treatment of various tumors, and some patients achieved long-term survival or even complete cure. However, due to the poor reproducibility and unclear mechanism of this therapy, it was eventually replaced by radiotherapy and chemotherapy.

Throughout most of the 20th century, progress in tumor vaccine research was slow. Researchers continuously explored around two core issues: tumor antigenicity and the body's anti-tumor immune response, trying various technical approaches. However, the development of cancer vaccines was not smooth.

After the 1980s, with the advancements in molecular biology and immunology of tumors, the concept and technology of therapeutic tumor vaccines have rapidly developed. Research on tumor vaccines at the laboratory level has continuously progressed, and various small-scale clinical trial schemes have been gradually implemented.

In 2010, after repeated mediation, Provenge finally received FDA approval, becoming the world's first therapeutic tumor vaccine.

Zhu Xun said that looking back at history, the only therapeutic tumor vaccine that has been truly approved by the FDA is Provenge, but its "success" can only be regarded as "conditional approval". "When it was first approved in 2010, Wall Street and academic media were in a frenzy, with sales reaching over $400 million in the first year."

In the IMPACT trial, which served as the basis for FDA approval, the median survival time was 25.8 months in the Provenge group and 21.7 months in the placebo group – a mere 4.1-month extension. Moreover, each treatment session was priced at a whopping $99,000. With poor clinical feedback, Provenge sales plummeted, and the developer, Dendreon, ultimately went bankrupt, with the product changing hands several times.

The fate of Provenge reveals a harsh reality: even if successfully approved for marketing, tumor vaccines still face insurmountable challenges in terms of efficacy, convenience, and economics.

After Provenge, numerous vaccine candidates failed in phase III clinical trials: MAGE-A3 (melanoma and lung cancer), Stimuvax (NSCLC), Rintega (glioblastoma), Canvaxin (melanoma), Prostvac (prostate cancer), GVAX (pancreatic cancer)… A retrospective analysis of 23 completed phase III studies of tumor vaccines revealed that 74% (17/23) failed to demonstrate significant efficacy on the primary endpoint.

The reasons for failure are complex and intricate: "starting the race early" when there is insufficient understanding of biology or therapy, with commercial considerations overriding science; poorly designed clinical trials, with heterogeneous patient populations and a majority being in advanced stages; regulatory agencies lacking experience in evaluating tumor vaccine products; and significant challenges in production scaling.

"History always repeats itself in a surprising way," said Zhu Xun. "Just like how everyone is now chasing mRNA tumor vaccines, it's strikingly similar to the pursuit of angiogenesis inhibitors and Provenge back then. Of course, we need to encourage exploration, but we also need to be calm and objective."

Zhu Xun emphasized, "There have been too many cases that show people often overestimate the short-term impact that new technologies and products may have, while often ignoring the long-term benefits they could bring."

02. Three "hurdles" that cannot be bypassed

Looking back at the century-long exploration history of tumor vaccines, from Coley's toxins to Provenge, and now to mRNA vaccines, why are there far more failures than successes?

 (Note: This is a placeholder text and may not have a direct translation into English as it is likely a typo or an error in the original text. If you have any specific context or information about what should be translated, please provide it and I will update the translation accordingly.)

The answer is not complicated. Therapeutic tumor vaccines are not dealing with a healthy and intact immune system, but rather one that has been "tamed" or even "hijacked" by the tumor. And it is precisely within such a "modified" immune system that therapeutic tumor vaccines aim to reactivate anti-tumor responses, making the challenge quite conceivable.

In Zhu Xun's view, for therapeutic tumor vaccines to truly be effective, they must simultaneously meet three essential conditions, none of which can be dispensed with. Even if all three conditions are met, effectiveness cannot be guaranteed; however, if any one of them is missing, effectiveness is absolutely impossible.

Firstly, it is necessary to achieve correct immune recognition of one's own tumor.

The premise of immune recognition is the presence of recognizable "targets". Taking malignant melanoma as an example, after surgical resection, whole-genome analysis of tumor tissue can reveal a large number of gene mutations, which in turn produce so-called "neoantigens" - these mutation sites are lowly expressed or not expressed in normal cells, but specifically appear in tumor cells. A patient may typically carry dozens or even hundreds of neoantigen sites.

However, having antigens does not necessarily mean that the body can effectively recognize them, as most of these components still belong to the "self" category, and the immune system naturally tolerates their presence. This involves the antigen presentation process. Inducing anti-tumor immunity mainly relies on killer T cells, and their activation requires antigen-presenting cells - especially dendritic cells - to digest, process, and present antigens to T cells.

Therefore, the first essential element is the presence of suitable tumor antigens that can be effectively presented by antigen-presenting cells (APCs), thereby activating specific T cells.

"In this process, AI can play a certain auxiliary role, such as predicting which new antigen epitopes are more likely to break immune tolerance and stimulate a stronger immune response, but its effect is limited. The mutation spectrum of each patient is completely different, and even for the same patient, it may change at different stages of the disease. Currently, the average number of mutations in melanoma is less than 100, while the known mutations in lung cancer have exceeded a thousand. How to screen targets is a huge challenge in itself. ”

Secondly, after immune recognition, the initial immune recognition should induce a sufficiently powerful immune killing effect capable of eliminating tumors.

Having immune recognition does not necessarily equate to having therapeutic effect. The intensity of the immune response determines whether the tumor can be truly eliminated. Zhu Xun vividly demonstrated this point by citing an image: in an electron microscope photo, a tumor cell is huge in size, surrounded by a large number of T cells - no matter how strong an individual T cell is, if the number is insufficient, it is difficult to win.

"A tumor cell is huge in size, while T cells are relatively small, and solid tumors often have a large amount of stroma and inhibitory microenvironment 'protection'. It's like a lion facing a pack of wolves. Sometimes, it may not be because I'm not strong enough, but because the enemy is too powerful," said Zhu Xun.

He also used CAR-T therapy as an analogy, 'Why is CAR-T relatively effective?'? Because we extensively expand and arm T cells into teeth in vitro before reinfusion. Currently, the majority of tumor vaccines induce a much lower number and activity of T cells compared to CAR-T. ”

Therefore, the "intensity" of the immune response is crucial - not only to initiate, but also to amplify enough to combat tumor burden.

Thirdly, immune suppressive factors in the body must be eliminated.

Tumors originate from mutations in normal autologous cells. In addition to the distinction between "cold" and "hot" tumors, depending on the stage of development, the tumor and the body's immune system generally go through multiple stages such as immune tolerance, immune escape, immune suppression, and immune exhaustion.

、
Therapeutic tumor vaccines are often used in the middle and late stages of tumors. Even if the vaccine activates T cells, these T cells may still be rapidly "disarmed" after entering the tumor microenvironment. The tumor microenvironment is filled with inhibitory cytokines such as TGF - β and IL-10, as well as inhibitory immune cells such as regulatory T cells and myeloid derived suppressor cells, which actively weaken the anti-tumor immune response.

The most commonly used strategy currently is the combination of PD-1/PD-L1 checkpoint inhibitors, but even in most solid tumors, the effectiveness of PD-1 monoclonal antibodies is only about 20% to 30%, which is not a panacea.

03. Moderna's Five Challenges

After understanding these foundations, let's take a look at Moderna's data - Zhu Xun admits that technically speaking, this is certainly a breakthrough. But he is more inclined to 'let the bullet fly a little longer'. He focuses on details beyond the data. Here are the five levels he dismantled one by one:

Firstly, the data is not yet complete.

At present, only the "top line results" - recurrence free survival and distant metastasis free survival - have been disclosed, but overall survival (OS) data has not yet been released.

More importantly, the experimental design: how was the control group set up for the experiment? In theory, the experimental group analyzed the specific epitopes after surgery, synthesized mRNA, delivered it through LNP, and administered K drugs to patients. What about the comparison? If we want to prove the value of these selected epitopes, the control group should also use this person's tumor sample, but randomly select a few epitopes, synthesize mRNA, and use LNP. ”

In his view, LNP itself has adjuvant effects and can stimulate the body to produce some immune responses. If the contribution of LNP is not excluded, it cannot be proven that the antigen of the vaccine is at work.

There are too many factors that can cause changes here, "Zhu Xun said." These details need to be seen in the complete clinical trial protocol and FDA review report, not in media summaries. ”

Secondly, it does not necessarily mean that other types of cancer are effective.

Even if all the results of this vaccine are true, we still need to objectively look at it - malignant melanoma is effective, while other tumors may not be effective

Melanoma is a "highly immunogenic" tumor - high mutation load, located in the skin, and relatively open immune microenvironment, while lung cancer, pancreatic cancer, and colorectal cancer have more complex mutation spectrum and stronger immunosuppression, which may not replicate the same effect.

Thirdly, the cost and accessibility are difficult to be optimistic about.

Due to its high degree of individualization, commercialization in the future will face great difficulties, "said Zhu Xun." Accessibility and cost are both issues
Provenge's lesson is right in front of us - each treatment costs $99000, with peak annual sales of over $400 million, ultimately leading to the company's bankruptcy due to poor sales. The complexity of personalized mRNA vaccines far exceeds Provenge, where a single patient is an independent production batch, resulting in higher costs and greater difficulty in achieving economies of scale.

More importantly, in the later stages, the patient's physical condition has deteriorated significantly. It will not only compete with similar therapies in the future, but also with various existing standard treatment plans - in terms of clinical value and economics. Its true value is still uncertain

Fourthly, there are numerous regulatory and patent barriers.

"There is no patent problem for LNP to make COVID-19, because it is authorized for emergency use. When used for other indications, you will definitely face patent challenges. ”Zhu Xun said.

The greater challenge lies in the quality control standards. Every patient is different, how can we protect them? "He said." CAR-T is taking out cells and adding fixed CARs, and everyone adds almost the same CARs. Isn't this - after each patient comes out, do we make 30 epitopes or 60? Is the mRNA length consistent? There is no standardized thing to determine these. ”

It's a long way to go to make it a qualified, pre approved standardized product 

Fifth, the issue of treatment window.

This type of vaccine requires patients to undergo surgical removal of the tumor before sequencing and vaccine production can begin. And the requirement is for patients in the later stage of treatment, who have already used various therapies before. At this stage, what is the body in? In the stage of immune suppression or even immune exhaustion

Zhu Xun pointed out, "What should be the most valuable time for a true tumor vaccine? In the early stages of immune evasion. But at this time, the vast majority of people will not undergo surgery because they have not yet found where the tumor is. Preventive vaccines are only given when there is no infection, and if given after infection, the effect will be greatly reduced

- Write at the end -

As early as 1988, during his doctoral studies, Zhu Xun participated in the compilation of the book "Outline and Technology of Immune Bioengineering". The cover of the book featured a schematic diagram he had drawn himself, depicting "genes transcribed into mRNA, translated into proteins, and secreted to function." He said, "At that time, talking about mRNA vaccines was still a fantasy, but scientific knowledge had already been established. ”

More than thirty years later, mRNA went from a schematic to a reality. But Zhu Xun does not believe that the maturity of technology means that everything will come naturally. "Feasible technology does not mean feasible medicine. There are far more pitfalls on this road than imagined

Regarding the future of personalized tumor vaccines, Zhu Xun believes that they are likely to be used as adjuvant therapy in some high mutation burden cancers, but becoming a universal "miracle drug" still requires overcoming more challenges.

Its highly individualized nature determines that it cannot be produced and sold on a standardized scale like monoclonal antibodies. We may need to design new regulatory pathways specifically for this type of product - such as managing it as a 'therapy' rather than a 'drug', allowing for multi center, small sample, and dynamically adjusted approval models

As emphasized by Zhu Xun in the report of the first Global Conference on New Drugs for Major Powers on July 22, China is already the country with the largest use of preventive vaccines in the world, and therapeutic vaccines, especially personalized therapeutic vaccines, have huge development space in the future. However, industry consensus or guidelines have not yet been formed on their evaluation criteria, clinical trial pathways, etc. Proactively establishing and promoting "Chinese standards" for new drugs in emerging and advantageous fields in China is of great significance for China's pharmaceutical innovation and leapfrog development.