mRNA cancer therapeutics advance from molecular design to clinical trials
核心洞察
A comprehensive review in Molecular Cancer (搜索), led by researchers at West China Hospital of Sichuan University (搜索), argues that mRNA cancer (搜索) therapies have moved beyond proof of concept into a programmable, integrated therapeutic system.
The field now focuses on precisely controlling the timing, location, dose, and duration of therapeutic protein production to attack tumors while sparing healthy tissue.
Delivery remains the central bottleneck, with conventional lipid nanoparticles accumulating in the liver and researchers pursuing extrahepatic targeting, improved endosomal escape, and repeat-dosing solutions.
Messenger RNA therapeutics, the technology that vaulted to global prominence through COVID-19 vaccines, is undergoing a decisive transformation in oncology, according to a comprehensive review published in the journal Molecular Cancer (搜索). The analysis, led by researchers at West China Hospital of Sichuan University (搜索), argues that mRNA cancer (搜索) therapies have crossed a conceptual threshold: the field is no longer asking whether messenger RNA can be used to produce therapeutic proteins inside the human body, but rather how the timing, location, dose, and duration of that protein production can be precisely controlled to attack tumors without harming healthy tissue.
The review frames mRNA cancer (搜索) medicine as an integrated, programmable system rather than a single drug class. Every therapeutic mRNA molecule is, in essence, a synthetic instruction sheet that co-opts the cell's own protein-making machinery. Chemically, these molecules are produced by in vitro transcription, a process that synthesizes RNA from a DNA template outside living cells. The resulting transcript is then engineered with a five-prime cap structure that allows ribosomes to recognize it, a polyadenylated tail that stabilizes the molecule, and untranslated regions at both ends that tune how efficiently and for how long the encoded protein is manufactured. Coding sequences themselves can be modified to favor particular amino acids, and nucleotide chemistries such as N6-methyladenosine can be incorporated to dampen unwanted immune recognition. Each of these design layers, the authors emphasize, independently shapes pharmacology, meaning that two mRNA drugs encoding the same protein can behave very differently in a patient depending on their molecular architecture.
The delivery bottleneck
Delivery remains the central engineering bottleneck. Synthetic mRNA is a large, negatively charged, fragile molecule that cannot simply cross cell membranes. The dominant solution is the lipid nanoparticle (LNP), the same class of carrier validated in billions of vaccine doses during the pandemic. LNPs encapsulate the RNA in a protective lipid shell containing an ionizable lipid that becomes positively charged in the cell's acidic environment, along with helper lipids, cholesterol, and polyethylene glycol-lipids that stabilize the particle. But a striking limitation, highlighted throughout the review, is that conventional LNPs accumulate overwhelmingly in the liver after intravenous administration, because the particles are captured by liver sinusoidal cells. For cancer (搜索) therapy, where tumors arise in the lung, pancreas, brain, and elsewhere, extrahepatic targeting is a critical frontier. Researchers are now tuning lipid composition, particle size, surface charge, and ligand decoration to redirect particles to lymph nodes, tumor tissue, and specific immune cell populations, and are exploring alternative platforms including lipoplexes, polymer carriers, extracellular vesicles, and virus-like particles.
Another obstacle is endosomal escape. When an LNP is engulfed by a cell, it first lands in an endosome, a membrane-bound compartment that typically routes its contents toward degradation. Only a fraction of delivered RNA molecules escape into the cytoplasm, where ribosomes can translate them. Improving this escape efficiency, the review notes, is one of the most active areas of delivery research, alongside the problem of repeat dosing. Repeated injections of PEG-containing nanoparticles can trigger accelerated blood clearance and hypersensitivity reactions, a serious concern for cancer (搜索) patients who may require months of treatment, unlike the two-dose vaccination paradigm.
Balancing immunity and expression
The immune system adds a further layer of complexity. mRNA molecules are intrinsically recognized by innate immune sensors such as Toll-like receptors 3, 7, and 8, retinoic acid-inducible gene I, melanoma differentiation-associated protein 5, and the cytosolic pathways involving protein kinase R and oligoadenylate synthetase. In vaccines, some degree of immune stimulation is a feature rather than a bug, acting as a built-in adjuvant that amplifies the response against the encoded antigen. In oncology, however, the calculus is subtle. Too little immune activation and the therapy fails to provoke a meaningful anti-tumor response; too much, and the RNA is degraded prematurely, inflammatory toxicity ensues, or the encoded therapeutic protein is neutralized before it can act. The review stresses that balancing transgene expression with immune activation is a defining design constraint across every mRNA cancer (搜索) modality.
Clinical strategies across the field
The clinical landscape surveyed in the review spans several distinct therapeutic strategies. Cancer (搜索) vaccines built on mRNA typically encode tumor-associated antigens or, in the personalized medicine paradigm, patient-specific neoantigens. Neoantigens arise from mutations unique to a patient's tumor, making them genuine molecular fingerprints that the immune system has not been trained to tolerate. Personalized mRNA vaccines are manufactured by sequencing a patient's tumor, predicting which mutated peptides will bind the patient's human leukocyte antigen molecules, and synthesizing a bespoke mRNA encoding up to dozens of these neoantigens. Combined with immune checkpoint inhibitors such as antibodies targeting PD-1 (搜索) or PD-L1 (搜索), these vaccines aim to expand T cell populations capable of recognizing and destroying tumor cells, with trials underway in pancreatic cancer, melanoma, colorectal cancer, and other solid tumors. Universal vaccines, by contrast, target shared antigens applicable to broader patient populations, trading personalization for speed, cost, and manufacturability.
Beyond vaccines, mRNA can encode fully functional therapeutic proteins in their own right. The review catalogs clinical programs delivering messenger RNAs for cytokines such as interleukin-12 (搜索) and granulocyte-macrophage colony-stimulating factor (搜索), which are injected directly into tumors to convert the local tumor microenvironment from immunologically cold to inflamed. Other candidates encode immune agonists such as CD40 (搜索), OX40 (搜索), and 4-1BB (搜索) ligands, designed to stimulate anti-tumor T cells, as well as encoded antibodies and bispecific T-cell engagers, which direct T cells toward tumor cells without requiring the ex vivo manufacturing steps of conventional biologic drugs. Intratumoral delivery is emerging as a particularly attractive strategy, allowing potent immune modulators to be confined to the tumor site and limiting systemic toxicity that has hampered recombinant cytokine therapy for decades.
In vivo cell engineering and next-generation RNA formats
Perhaps the most technologically ambitious application is in vivo cell engineering. Rather than removing a patient's T cells, reprogramming them to express a chimeric antigen receptor in a laboratory, and reinfusing them, as standard CAR-T therapy requires, researchers are exploring mRNA delivered directly into the body to instruct immune cells to build their own receptors. LNPs functionalized with targeting ligands can, in principle, home to T cells, natural killer cells, or macrophages and deliver mRNA encoding a CAR, a T-cell receptor, or a B-cell maturation antigen-binding construct. Because mRNA is transient, the engineered state lasts days rather than years, which the review suggests may offer a safety advantage over permanently integrated viral vectors, potentially reducing risks such as cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome, though it may also require repeated dosing to sustain activity.
The review also surveys the expanding RNA chemistry toolbox beyond conventional linear mRNA. Self-amplifying RNA incorporates an RNA-dependent RNA polymerase, typically derived from alphaviruses, allowing the transcript to replicate itself inside the cytoplasm, which dramatically reduces the dose required per administration. Trans-amplifying RNA divides this machinery between two separate molecules for greater design control. Circular RNA, produced by joining the ends of a linear transcript into a covalently closed loop, lacks the exposed ends that cellular exonucleases attack, conferring remarkable stability and enabling protein expression that persists far longer than linear mRNA. Each platform carries trade-offs in manufacturing complexity, immune stimulation, and duration of expression, and the authors argue that clinical indications will ultimately dictate which RNA format is optimal.
Looking across the field, the authors conclude that mRNA cancer therapeutics (搜索) are diverging into modality-specific solutions rather than converging on a single dominant design. Clinical efficacy, they contend, depends on the coordinated optimization of four interlocking elements: the RNA construct itself, the delivery vehicle, the pharmacology of the encoded payload, and the biology of the tumor-immune interaction. Advances in good manufacturing practice, quality control, and chemistry and manufacturing controls are simultaneously driving down production timelines, a crucial consideration for personalized vaccines that must be synthesized within weeks of a patient's diagnosis. What began as a technically constrained modality has matured into a validated platform with dozens of clinical programs, and the pace at which molecular design translates into approved cancer (搜索) medicines may now be limited less by RNA chemistry than by the intricacy of the tumor microenvironments these programmable molecules are being sent to reprogram.