The mRNA vaccines that became household names during the pandemic worked by handing cells a temporary instruction sheet: make this one protein, show it to the immune system, then let the message degrade. A first-in-human cancer study that posted to the U.S. registry today is built on a more ambitious version of that idea — an RNA that copies itself once inside the cell, stretching a single dose into a longer, larger production run, and aiming it not at a virus but at proteins that early-stage breast tumors quietly carry.
The trial, NCT07652242, is sponsored by Immunomic Therapeutics and carries the study name VITAL-TNBC. It enrolls people who have completed standard curative-intent treatment for stage II-III triple-negative breast cancer (TNBC) — a setting where the disease has been treated but carries a real risk of returning, and where there is room for a therapy that trains the immune system to patrol for any cancer cells left behind. The mechanism, as the registry describes it, is what makes the candidate worth a close read.
"ITI-5000 is a self-amplifying RNA (saRNA) vaccine that instructs the immune system to recognize and attack cancer cells expressing two proteins found on TNBC cells-HERV-K and CT83-fused with a molecule called LAMP-1 that helps the immune system respond more strongly. The vaccine is delivered inside lipid nanoparticles (LNPs), similar to other approved mRNA vaccines."— ClinicalTrials.gov, source
Each clause there encodes a design decision. The self-amplifying part means the RNA includes the machinery to replicate itself briefly after delivery, which can let a small dose generate a stronger, more sustained antigen signal than a conventional non-replicating message. The two targets — HERV-K and CT83 — are not random: they are proteins associated with tumor cells rather than healthy tissue, which is the entire safety logic of a cancer vaccine, since the goal is to point the immune system at the cancer and not at the patient. Fusing those antigens to LAMP-1 is a known trick to route the protein through the cell's antigen-presentation pathway more efficiently, sharpening the immune response. And the lipid-nanoparticle packaging is the same delivery format that carried approved mRNA vaccines, which is why the registry draws that comparison directly.
How the trial is built, and what it is built to prove
The study is a two-part, ascending-dose effort. Part A gives ITI-5000 alone using a modified 3+3 escalation across two cohorts — 1 microgram and 10 micrograms per vaccination — with each cohort starting from a single sentinel participant monitored for 28 days before others are enrolled. Part B then combines the vaccine at its best-tolerated dose with pembrolizumab, an approved anti-PD-1 immunotherapy, on the rationale that a vaccine which marks the cancer and a checkpoint inhibitor which removes the immune system's brakes may work better together than either alone. Planned enrollment is 60 participants.
The endpoint structure should govern how anyone reads early results. The primary objectives are to establish the dose-limiting toxicities and the maximum tolerated dose, and to characterize safety and tolerability — first for the vaccine alone, then in combination. Immune response and changes in circulating tumor DNA are explicitly exploratory. That is the correct posture for a first-in-human vaccine: prove it is safe and find a usable dose before asking whether it does anything to the disease. A clean safety profile and a defined maximum tolerated dose would be a success on this trial's own terms; an efficacy claim would not be supportable from it.
Why the platform choice is the headline
What makes this record notable beyond breast cancer is the platform statement it represents. Self-amplifying RNA is the next iteration of the modality that proved itself in infectious disease, and aiming it at tumor-associated antigens in the post-treatment, minimal-residual-disease setting is one of the most closely watched directions in cancer immunotherapy. The first-in-human label means there is no prior clinical safety read on this construct; every number that emerges from Part A will be new information about how a self-replicating RNA behaves in people.
The self-amplifying mechanism is the variable to watch most closely on the safety side. A conventional mRNA vaccine produces its protein in a brief, self-limiting burst as the message degrades. A self-amplifying RNA, by design, replicates itself first, which can extend and intensify protein production from a far smaller starting dose — note the microgram-scale cohorts here, well below the doses used for some conventional mRNA products. That efficiency is the platform's promise, but it also means the immune-activation profile can differ, and a single sentinel participant is monitored for 28 days before each cohort fills precisely because the kinetics of a self-replicating construct in humans are not yet fully characterized. The choice of tumor-associated rather than patient-specific antigens is its own deliberate bet: HERV-K and CT83 are shared targets that could, if validated, support an off-the-shelf vaccine rather than the bespoke, per-patient manufacturing that defines neoantigen approaches. Whether shared antigens generate a strong enough response is part of what the exploratory immune readouts are meant to probe.
The disciplined read of NCT07652242 today is therefore narrow and concrete. A self-amplifying RNA vaccine, delivered in lipid nanoparticles and engineered to present two tumor antigens fused to an immune-routing molecule, is entering its first human test in patients who have finished treatment for triple-negative breast cancer — first alone, then with pembrolizumab — to determine whether it is safe and at what dose. Whether it actually reduces recurrence is a question reserved for later, larger studies, and the exploratory immune and circulating-tumor-DNA readouts from this trial will only hint at the answer rather than settle it.