From Treatment to Prevention: How Far Are Cancer Vaccines?

Author tricision Time 2026-09-20

category：Media

Abstract: The Next Frontier of Cancer Vaccines

"Three transformative events have taken place in the cancer‑vaccine community over the past 20‑plus days," remarked Professor Youwen He, tenured full Professor in the Department of Immunology at Duke University Medical Center, specially appointed expert for the U.S. National Vaccine Program, and Chairman of tricision Biotherapeutics, in a recent interview with Tongxieyi.

First, on August 19, Merck and Moderna jointly announced that their personalized mRNA neoantigen therapy intismeran autogene (V940/mRNA‑4157) combined with Keytruda (pembrolizumab) met the primary endpoint in the Phase III INTerpath‑001 trial.

Also centered on mRNA and neoantigens, two weeks later, another major mRNA player BioNTech reported negative results in its Phase II adjuvant trial for its personalized mRNA cancer vaccine autogene cevumeran in post‑surgical colorectal cancer.

The third development is more likely to go unnoticed: on August 19, the U.S. Food and Drug Administration (FDA) released Potency Assessment of Active Immunotherapy Products Draft Guidance for Industry. "This landmark guidance is specifically tailored for personalized therapeutics."

In Professor He's view, taken together, these developments deliver a core insight: personalized therapeutics will become a central direction for treating major diseases.

"The one‑size‑fits‑all treatment model cannot cope with the high heterogeneity of cancer. Cancer is not a single disease but a collection of thousands of disease clones within a single patient. Therefore, treatment must be customized for each individual patient and adjusted dynamically as the disease evolves."

A scientist with 40 years of immunology research experience, Professor He has witnessed the full evolution of cancer immunotherapy. In this conversation with Tongxieyi, he examines the prerequisites for therapeutic cancer vaccines, the strategic value of preventive cancer vaccines, and future directions for paradigm shifts in oncology treatment from a macro perspective.

To draw an analogy, cancer vaccines function like the ignition of a car: ignition is required to start the car, yet ignition alone is insufficient. You also need to release the brake and step on the accelerator. "Releasing the brake corresponds to anti‑PD‑1 therapies, namely immune checkpoint inhibitors; stepping on the accelerator refers to cytokines such as IL‑12, IL‑2 and IL‑15."

何老师与程博士

Photo: Dr. Cheng and Professor Youwen He at BIOHK 2026

TONACEA

1、What Do One Success and One Failure Reveal?

Clinical research into cancer vaccines is a history written in failures. More than three decades ago, academia and industry worldwide began betting on this track — using vaccines to train the immune system to recognize and attack tumors.

The concept is compelling, yet real‑world outcomes have been harsh.

In 2010, Provenge (sipuleucel‑T), developed by Dendreon, gained FDA approval, becoming the world's first therapeutic cancer vaccine and the only approved one for more than a decade.

It marked a milestone, yet stood almost as an isolated case.

In its Phase III trial, Provenge extended the median overall survival of patients with advanced prostate cancer by merely 4.1 months (25.8 months versus 21.7 months). Its manufacturing workflow was extremely complex; driven by high costs and modest efficacy, Dendreon ultimately filed for bankruptcy.

Subsequently, more than 30 cancer vaccine candidates failed in Phase III clinical trials, and the FDA did not grant approval for any further therapeutic cancer vaccines for years.

Hope resurfaced in August 2026. The mRNA cancer vaccine intismeran combined with Keytruda demonstrated statistically significant and clinically meaningful improvements in both recurrence‑free survival (RFS) and distant metastasis‑free survival (DMFS) as adjuvant therapy for patients with completely resected Stage IIB‑IV melanoma.

Professor Youwen He notes that although full raw clinical data from Moderna have not yet been published, he remains optimistic, because this vaccine has identified the scientifically correct direction.

First, it selected the appropriate disease stage. Beyond tumor type, disease stage matters critically. Intismeran intervenes when tumor burden is at its lowest — after complete surgical resection, in the relapse‑prevention setting. At this point, the immune system remains fully capable of being activated, and residual minimal residual disease (MRD) represents the optimal window for vaccine action.

"Physical interventions such as radiotherapy, surgery and precision radiotherapy can control visible tumors. What truly concerns clinicians are invisible tumors undetectable by imaging, which can metastasize distantly and spread like dandelion seeds. This is where vaccines can play a role. Cancer vaccines must be deployed in the right clinical setting," Professor He elaborated.

Second, it uses combination therapy with PD‑1 blockade. Professor He offered a vivid analogy: "A cancer vaccine is a car's ignition. You have to ignite the engine before the car moves, but ignition alone will not suffice. You also need to release the brake — releasing the PD‑1‑imposed brake on T cells."

By contrast, BioNTech's colorectal cancer trial stumbled on exactly this point. The BNT122‑01 trial evaluated autogene cevumeran as monotherapy, without combining it with immune checkpoint inhibitors. Without the "brake‑release" effect of PD‑1 inhibitors, vaccine‑activated T cells are rapidly suppressed by the tumor microenvironment.

Professor He states plainly that BioNTech's misstep lay in deploying the vaccine as monotherapy, reflecting an incomplete understanding of how the vaccine works. The trial termination merely confirms that the underlying scientific rationale was flawed from the outset.

"One failure and one success tell us we must select the right tumor type, disease stage and combination platform. mRNA is a promising technical modality, yet it must be deployed correctly. These serve both as lessons learned and cautionary take‑homes," Professor He summarized.

For the Moderna case, Professor He regards it also as a victory belonging to the AI era. Intismeran can encode up to 34 neoantigens.

"Tumors harbor numerous mutations; we only select a small subset of those mutations to construct neoantigens today. This work would be impossible without AI. For example, full exome sequencing analysis once took our team one week; now it can be completed within 10 minutes. The difference is enormous. Cancer vaccines could only emerge in the present era."

Professor He adds a word of caution: "Even though intismeran encodes 34 algorithm‑selected neoantigens, likely only one or two of them are functionally active. Even the developer cannot definitively identify which ones, pointing to the immense complexity of the underlying biological processes."

TONACEA

2、The Promise of Personalized Therapeutics

Among the three events highlighted by Professor He, the third has drawn limited public attention, yet he stresses it is "extremely important": a landmark regulatory guidance document.

Released by the FDA in August 2026, this draft guidance provides systematic recommendations for development, validation and quality‑control strategies for potency testing of Active Immunotherapy Products (ACTIMPs).

Traditional drug‑regulation frameworks are built around the concept of potency. For pharmaceuticals ranging from aspirin to Keytruda, the active ingredient and manufacturing process are fixed, and potency testing is performed for every batch before release to guarantee consistent quality across batches.

Personalized mRNA neoantigen vaccines upend this logic entirely. Each patient's vaccine is individually designed according to that patient's unique tumor mutational landscape; every single vaccine product is essentially one of a kind. In the precision‑medicine era, how should potency be evaluated for personalized therapeutics?

Professor He points out that this FDA draft guidance breaks new ground via two core principles:

First, potency is no longer assessed only through testing of the final finished product. Instead, evaluation is performed based on Critical Quality Attributes (CQA). Regulators assess whether bioinformatic analysis and biomanufacturing workflows are qualified, stable and well‑controlled. Regulatory oversight shifts from testing each individual vaccine lot to validating the entire production system, representing major regulatory progress for personalized therapeutics.

Second, personalized products are no longer restricted to manufacturing within a single fixed facility. Historically, FDA‑approved drugs had to be manufactured at predefined sites. This guidance explicitly recognizes that multiple manufacturing locations are permissible — a development critical for the commercialization of personalized vaccines.

Professor He holds that personalized therapeutics will become a core pillar for treating major diseases. "More than a decade ago I treated cancer patients in clinical practice and kept arguing that our existing treatments are far too outdated. How can one‑size‑fits‑all regimens work for such complex diseases as cancer? We need patient‑specific treatment strategies, and those strategies must be adjusted dynamically over time."

Professor He shared a striking real‑world example: glioblastoma, known as the "king of cancers." Tumor tissue is sequenced at initial diagnosis and again upon disease recurrence. The overlap in detected gene mutations between the two time points may be as low as 3%. "In practical terms, recurrent disease becomes an almost entirely distinct illness. Treatment eliminates mutation‑sensitive tumor clones, leaving behind resistant clones refractory to chemotherapy and radiotherapy."

Professor He emphasizes a second core principle in oncology: only the immune system can achieve tumor control. All other therapeutic modalities must serve to boost anti‑tumor immunity.

"It resembles the Terracotta Army: each warrior carries different weaponry. Our T lymphocytes operate by the same principle, targeting different mutations. Only by mobilizing the immune system can we achieve tumor control."

This clarifies the logic behind personalized tumor vaccines: they identify unique mutations from a patient's tumor and turn them into targets recognizable by the immune system so that T cells can recognize and attack tumor cells. Since mutation profiles differ across individuals, therapies must be customized per patient.

Nevertheless, industrial‑scale personalized therapeutics bring complex challenges. If each patient represents an independent manufacturing batch, accessibility and cost emerge as major bottlenecks. Professor He outlines three potential pathways to address these challenges.

First, bypass antigen‑specific components via innate immune activation. mRNA itself acts as an innate immune activator triggering Type I interferon pathway activation. Type I interferons (α, β) are already FDA‑approved for oncology indications, creating opportunities for universal therapeutic candidates.

Second, even though patient‑specific regimens represent the future, shared targets exist among tumor‑associated antigens and neoantigens. For instance, KRAS mutations occur across multiple tumor types and can be exploited for off‑the‑shelf vaccine candidates.

Third, AI‑driven fully automated manufacturing facilities will greatly reduce costs and shorten turnaround timelines. This is part of the intent behind the FDA guidance: in the future, on‑site devices near patients' hospital beds will manufacture personalized vaccines and therapeutics in real time. Instead of centralized production plus logistics transport, manufacturing occurs locally on demand. This unlocks tremendous potential for pharmaceutical innovation while clearing the path for regulatory adaptation.

TONACEA

3、The "Crown Jewel" Atop the Pyramid

Compared with therapeutic cancer vaccines, Professor He is more optimistic about preventive cancer vaccines, corresponding to vaccines in the conventional public understanding.

Renowned medical scientist Dr. Eric Topol, Founder and Director of the Scripps Research Translational Institute, published an immunotherapy pyramid. Immune checkpoint inhibitors sit at the base, followed upward by oncolytic viruses, CAR‑T and other modalities. Positioned at the very apex of the pyramid are interceptive vaccines — preventive vaccines that intercept cancer onset.

Figure: Dr. Eric Topol's Cancer Immunotherapy Pyramid

Compared with therapeutic cancer vaccines, Professor He is more optimistic about preventive cancer vaccines, corresponding to vaccines in the conventional public understanding.

Figure: Dr. Eric Topol's Cancer Immunotherapy Pyramid

Professor He concurs with this framework. "For one thing, the target population is vastly larger: these approaches aim at cancer prevention rather than treating established malignancy. Intervention at early preventive stages represents the ideal clinical setting. For another, our technology has matured enough to make this feasible."

Recent clinical progress validates this outlook. Preventive cancer vaccines are first being evaluated in high‑risk individuals with well‑defined genetic predispositions who have not yet developed cancer.

Take Lynch syndrome as an example. Patients carry germline mutations in the MMR gene that impair DNA‑repair function. Errors accumulate within cells, substantially elevating cancer risk. The lifetime risk of colorectal cancer reaches 70%–80%, with disease onset occurring significantly earlier, at a median age in the mid‑40s.

In January 2026, investigators at MD Anderson Cancer Center published Phase Ib/II clinical data for NOUS‑209 in Nature Medicine. This off‑the‑shelf immunotherapeutic vaccine targets shared frameshift‑mutation neoantigens in patients with Lynch syndrome. Forty‑five subjects were enrolled; all mounted T‑cell immune responses against cancer‑relevant targets, with enhanced immune reactivity following annual booster vaccination. After one year of treatment, reductions in precancerous lesions were observed and no new advanced polyps emerged.

In June of the same year, Moderna together with the University of Oxford announced that the mRNA vaccine mRNA‑4194 for Lynch syndrome had obtained MHRA approval for a Phase I/II clinical trial in the United Kingdom. It encodes specific neoantigens linked to early‑stage Lynch‑syndrome‑associated carcinogenesis. Using mRNA technology, it activates the immune system and trains T cells to recognize and eliminate precancerous cells to block cancer initiation at source.

Beyond Lynch‑syndrome‑related indications, Professor He disclosed that tricision Biotherapeutics' in‑house Survivin‑targeting mRNA‑DC therapeutic cancer vaccine has entered Phase II clinical trials and has obtained approval for clinical application at Weiyuan Hospital within the Beidaihe Life Science & Industry Innovation Demonstration Zone. "Survivin represents an excellent tumor antigen, and we plan to expand its indications toward preventive vaccines in the future."

In Professor He's view, preventive vaccines will initially target high‑risk populations: Lynch‑syndrome‑associated colorectal cancer, BRCA‑mutation‑associated breast and ovarian cancer, and lung cancer among long‑term smokers. Risk profiles are well‑defined, benefit‑to‑risk profiles can be quantified, and clinical evaluation complexity is far lower than many anticipate.

Safety nevertheless remains an unavoidable consideration for preventive vaccines. Professor He's stance is clear: benefit‑to‑risk profile is paramount.

"Preventive cancer vaccines can be considered for populations with clear clinical indications, for example cirrhosis patients at high risk for hepatocellular carcinoma where the benefit‑to‑risk profile is well‑established. Caution should be exercised when administering liver‑cancer preventive vaccines to entirely healthy individuals with normal liver function. Broad‑population deployment of preventive cancer vaccines demands extreme prudence."

From a technical‑evolution perspective, Professor He notes that current mRNA‑LNP platforms belong to the first‑generation technology; long‑term safety datasets remain incomplete. Multiple research groups, including his own team, are advancing several lines of development at a rapid pace.

These include biodegradable LNPs and highly‑targeted LNPs capable of delivering payloads selectively to defined cell subsets such as specific dendritic cell subpopulations. "Targeted delivery greatly improves safety. First, required dosages decrease, eliminating the need for high‑dose systemic administration. Second, payloads are delivered to defined cell populations."

Professor He anticipates that second‑ and third‑generation delivery technologies will enter preventive‑vaccine R&D pipelines soon, potentially within this year and next year.

l This article reproduces verbatim the original Chinese text published on the WeChat Official Account "Tongxieyi" (同写意). Reprinted with permission from Tongxieyi. The English-language content is a full translation of the original Chinese article. The copyright of the original Chinese article belongs to Tongxieyi.