The Vaccine Revolution in Oncology Is Finally Here: Neoantigens, LNPs, and the New Paradigm
核心洞察
A 2026 Nature Medicine review by Haanen, Schumacher, Kagan, Fritsch, and Wu declares that cancer vaccines are now a clinical reality for melanoma (搜索), non-small cell lung cancer (搜索), and glioblastoma (搜索).
The field's inflection point came from neoantigen (搜索) discovery and personalised manufacturing, with private neoantigens replacing shared antigens as the consensus approach.
Lipid nanoparticles (LNPs) with innate agonists that drain to lymph nodes are now the gold standard delivery vehicle, as where and how antigen is delivered matters more than the antigen itself.
For decades, the phrase "cancer vaccine" was the punchline of oncology conferences. Billions were poured into stimulating T cells against tumours, only to watch those T cells get exhausted, ignored, or lost. But a landmark 2026 review published in Nature Medicine by Haanen, Schumacher, Kagan, Fritsch, and Wu makes one thing painfully clear: the joke is over. The field is not merely entering the era of cancer vaccines — it is realising that previous failures stemmed from treating vaccines like simple drugs rather than complex biological reprogramming events.
The review systematically dismantles the old approach: taking a handful of overexpressed self-antigens, mixing them with a blunt adjuvant, and injecting them into patients with late-stage, macro-metastatic disease. Those antigens were weakly immunogenic, the tumours had already perfected immune exclusion, and the patients had no T cells left to prime. No wonder it failed.
The Neoantigen (搜索) Revolution
Haanen et al. argue that the clinical inflection point came from two parallel revolutions: neoantigen (搜索) discovery and personalised manufacturing. By sequencing a patient's tumour and using algorithms to predict which mutant peptides will bind to their unique HLA molecules, researchers can now manufacture a truly personal vaccine. The 2026 consensus is unambiguous: shared antigens are out; private neoantigens are in.
This shift addresses one of the biggest challenges in oncology — cancers' so-called "stealth-mode." As Christopher Oelkrug, Director of Business Development for Advanced Clinical (搜索), notes, recent results indicate that therapeutic cancer vaccines can help researchers tackle this fundamental problem by pre-emptively training the immune system to increase the frequencies of tumour-reactive T cells.
Delivery Is Everything
Perhaps the most fascinating discussion in the paper concerns the delivery vehicle. For years, the field squabbled over RNA, peptides, or viral vectors. The authors decline to pick a winner. Instead, they highlight a more profound insight: the vaccine is only as good as the danger signal it carries.
Jonathan Kagan's contributions on innate immunity are critical here. A neoantigen (搜索) peptide injected alone is ignored. But when packaged into a lipid nanoparticle (搜索) (LNP) with an innate agonist, that same peptide becomes a red alert. The 2026 finding is that where and how you deliver the antigen matters more than the antigen itself. LNPs that drain to the lymph node and are picked up by resident dendritic cells are now the gold standard. Anything less is just expensive noise.
Timing Is Everything: Adjuvant and Neoadjuvant Settings
The authors do not shy away from the clinical trial data showing that vaccines in bulky, end-stage disease still largely flop. Why? Because the tumour microenvironment (TME) actively destroys T cells. You cannot prime an army into a battlefield that is already a graveyard.
The breakthrough trials cited in the review are the adjuvant (post-surgery, no detectable disease) and neoadjuvant (pre-surgery) settings. When you vaccinate a patient who has micrometastatic disease but no macroscopic immunosuppressive mass, the results are staggering. Pathologic complete responses are emerging. This flips the standard oncology algorithm on its head: don't cut first, vaccinate first.
The Next Frontier: Blurring Vaccines and Cell Therapy
The most provocative prediction from Haanen, Schumacher, and Wu is that the next five years will blur the line between vaccines and cell therapy. They point to in vivo reprogramming — using LNPs to deliver mRNA encoding not just antigens but also chimeric antigen receptors (CARs) directly into a patient's T cells.
The vision is striking: a single infusion that turns a patient's own lymph nodes into a factory for CAR-T cells, without the $500,000 price tag or the toxic chemotherapy preconditioning. The authors suggest that by 2028, "vaccine" will mean a multi-epitope mRNA-LNP that simultaneously expands neoantigen (搜索)-specific T cells and deploys a bispecific T cell engager to overcome the TME.
The Clinical Development Crisis
Despite the scientific brilliance, the review hints at a clinical development crisis. Personalised cancer vaccines take 30 to 60 days to manufacture. For a patient with metastatic melanoma (搜索) or pancreatic cancer (搜索), that is an eternity. The authors call for a new regulatory paradigm: "just-in-time" inventory of pre-manufactured neoantigen (搜索) libraries. Instead of making each vaccine from scratch, the field would maintain a warehouse of common neoantigen cassettes ready to match to a patient's HLA type within 48 hours.
This requires a level of coordination between pharma, regulators including the FDA and EMA, and hospital systems that currently does not exist. As the review starkly puts it: we have the science to cure many solid tumours; we lack the logistics.
Cancer vaccines are no longer a hypothesis. They are a clinical reality for melanoma (搜索), non-small cell lung cancer (搜索), and glioblastoma (搜索). But they demand a radical shift in how trials are run: stop testing on patients with ten prior lines of therapy, stop using aluminium-based adjuvants from the 1920s, and start measuring immune responses in the lymph node, not just in the blood. The immune system has remembered every virus humanity has ever fought. It is time to teach it to remember cancer. The tools are finally in hand. The only question is whether the clinical infrastructure is brave enough to use them.