The National Cancer Institute is recruiting 86 participants into a Phase 2 trial testing whether a personalized synthetic long peptide neoantigen vaccine adds benefit to durvalumab, tremelimumab, and nab-paclitaxel in metastatic triple-negative breast cancer. Registered as NCT03606967 and last…
The National Cancer Institute NCI , the trial's lead sponsor, is recruiting 86 participants into a Phase 2 trial testing whether adding a personalized synthetic long peptide neoantigen vaccine to durvalumab , tremelimumab , and nab-paclitaxel improves treatment of metastatic triple-negative breast cancer . The trial is registered as NCT03606967 , was last updated on August 10, 2026, and is listed as currently recruiting.
The study compares the chemotherapy plus dual checkpoint inhibitor backbone with or without the individualized vaccine. Durvalumab is a monoclonal antibody against PD-L1, tremelimumab is a monoclonal antibody against CTLA-4, and nab-paclitaxel is a taxane chemotherapy. The vaccine is synthesized for each patient from neoantigens, meaning peptide sequences derived from mutations that are unique to that patient's tumor.
The question the trial is built to answer is narrow and specific: does a vaccine-induced T cell response add clinical benefit on top of a regimen that already contains two immune checkpoint inhibitors and chemotherapy? That question has not been answered in this disease, and the NCI has not reported any efficacy or safety results because enrollment is still underway.
The trial enrolls adults with anatomic stage IV breast cancer as defined by the AJCC version 8 staging system, invasive breast carcinoma, or metastatic triple-negative breast carcinoma. In practical terms, the population is patients with triple-negative breast cancer that has spread from the primary site to other places in the body. Triple-negative means the tumor lacks estrogen receptor, progesterone receptor, and HER2 amplification, which removes endocrine therapy and anti-HER2 drugs from the available treatment options. The available options in the metastatic setting are limited, and most patients in this situation will have received prior systemic therapy.
The planned sample size is 86 participants. The treatment comparison is nab-paclitaxel, durvalumab, and tremelimumab with or without the personalized synthetic long peptide vaccine. The registry record does not specify randomization, blinding, or a control arm, so the structure of the comparison is not fully described in the public record.
The trial record lists a set of accompanying procedures. Biopsy procedure and biospecimen collection appear as interventions, reflecting the need to obtain tumor tissue for sequencing and neoantigen identification. Computed tomography and magnetic resonance imaging are also listed; they are the imaging modalities used to assess disease burden and response. The record further lists carboplatin, gemcitabine hydrochloride, and sacituzumab govitecan among the agents associated with the study, along with Poly ICLC , an innate immune adjuvant routinely paired with peptide vaccines to strengthen the response to the injected peptides. Carboplatin and gemcitabine are cytotoxic agents used in this disease, and sacituzumab govitecan is an antibody-drug conjugate directed against Trop-2 that is used in this setting. The registry summary does not clarify whether these agents are protocol-defined therapy or part of the broader treatment context for the enrolled population.
Phase 2 trials are designed to detect signals of activity and to characterize safety in a relatively small number of patients, not to establish definitive superiority. With 86 participants, the trial is large enough to estimate whether the vaccine-added regimen produces responses worth pursuing, but it is not powered in the way a registration trial would be. The absence of randomization details in the record means the groups being compared, if there are two, may not be directly comparable, and any difference observed could reflect patient selection rather than the vaccine itself.
Interpretation will also depend on what the "without vaccine" side of the comparison actually consists of. If patients are assigned to two concurrent groups, baseline characteristics such as number of prior lines of therapy, visceral disease burden, and performance status will need to be balanced or accounted for in the analysis. If the comparison is against a historical benchmark instead, the results will carry the usual limitations of cross-trial comparison: differences in staging, prior treatment, and supportive care across eras can mimic or mask a treatment effect.
Because the trial is still recruiting, there is no efficacy or safety data to report. That is a limitation of the current record, not a flaw in the study: it is simply where the investigation stands. Any claim that the vaccine works in this setting would be premature, and any claim that it does not work would be equally unsupported. The value of the trial at this stage is in the question it poses and the infrastructure it requires to answer it.
This is, however, a difficult setting for a vaccine. Metastatic tumors have had time to evolve, to lose antigen presentation, and to establish an immunosuppressive microenvironment. Many vaccine trials in advanced disease have failed not because the vaccines failed to induce T cell responses but because the tumor microenvironment suppressed those responses before they could act. The combination with two checkpoint inhibitors is, in part, an attempt to address that problem: the vaccine primes the response, and the checkpoint blockade is meant to keep that response from being shut down.
Nab-paclitaxel is a microtubule inhibitor. It stops tumor growth by killing dividing tumor cells, blocking their division, or preventing their spread. Beyond that direct cytotoxic effect, taxane chemotherapy can induce immunogenic cell death, releasing tumor antigens and damage-associated molecular patterns into the microenvironment, and can deplete immunosuppressive cell populations such as regulatory T cells and myeloid-derived suppressor cells, which may make the tumor more permissive to an incoming T cell response.
Durvalumab and tremelimumab target different checkpoints. Durvalumab blocks PD-L1 expressed on tumor cells, preventing it from engaging PD-1 on T cells and thereby removing a brake that operates at the tumor site. Tremelimumab blocks CTLA-4, a receptor that dampens T cell activation early in the immune response, largely during priming in lymphoid tissue. Used together, the two antibodies intervene at two different stages of the cancer immunity cycle, one acting during T cell activation and the other during effector function in the tumor.
The vaccine is the antigen-specific component. Synthetic long peptides are typically 20 to 30 amino acids long, which distinguishes them from short minimal epitopes. A long peptide cannot bind directly to MHC class I on non-professional cells; it must be taken up, processed, and presented by professional antigen-presenting cells. That processing engages both CD4 helper T cells and CD8 cytotoxic T cells, and it is less likely to induce tolerance than a short peptide that can bind MHC directly. Because neoantigens arise from tumor-specific mutations, they are not expressed on normal tissue, so T cells that recognize them are less constrained by self-tolerance and are more likely to attack tumor cells specifically. Because the vaccine is personalized, each patient's peptide set is selected from the mutations present in their own tumor, which makes the product different in kind from an off-the-shelf antigen vaccine.
Poly ICLC, listed among the agents in the trial record, is a synthetic double-stranded RNA analog that activates Toll-like receptor 3 and related cytosolic sensors such as MDA5 and RIG-I. In peptide vaccine regimens it serves as an adjuvant, providing the innate immune signals, including type I interferon, that are required for a robust adaptive response to the injected peptides. The logic of the regimen is therefore complementary: chemotherapy creates antigen release and reduces suppressive cells, the vaccine provides a defined set of tumor-specific antigens plus adjuvant, and the checkpoint inhibitors remove the brakes that would otherwise limit the vaccine-induced T cells.
The sequence and timing of the components will matter when the results are read. Chemotherapy can be immunosuppressive in some schedules and immunostimulatory in others, and the same agents that release tumor antigen can also damage lymphocytes. A regimen containing a taxane, two checkpoint inhibitors, and a vaccine has to be sequenced so that antigen release, antigen presentation, and checkpoint release converge in time. Those details sit in the protocol, and the eventual report will need to describe them for the results to be interpretable.
The triple-negative label carries direct therapeutic consequences. Endocrine therapy is ineffective without hormone receptors, and anti-HER2 agents are ineffective without HER2 amplification. Chemotherapy has long been the mainstay, and in the metastatic setting the disease carries a poor prognosis relative to hormone receptor positive breast cancer.
Immunologically, however, triple-negative breast cancer is one of the more inflamed breast cancer subtypes. It tends to carry higher mutation burdens and more tumor-infiltrating lymphocytes than hormone receptor positive disease, and PD-L1 expression is common, which is why checkpoint blockade has shown benefit in some patients. A vaccine that expands tumor-specific T cells is a rational addition in principle: the disease appears to be one in which T cells can matter, and the checkpoint inhibitors provide a mechanism for keeping those T cells active.
The obstacle is the stage. In patients whose cancer has spread to other places in the body, tumors have had years to edit themselves under immune pressure. They can lose HLA expression, downregulate antigen processing machinery, and recruit suppressive cells into the stroma. Vaccines have historically performed better in settings where the antigen load is low and the immune system is not exhausted, such as adjuvant or minimal residual disease. This trial is therefore a demanding test of the peptide vaccine concept: if an individualized vaccine adds measurable benefit in stage IV triple-negative breast cancer, it would suggest that antigen-specific priming can work even in a hostile environment. If it does not, the result would point toward earlier disease as the more promising setting for this class of products.
For peptide scientists, the trial is a direct test of whether individualized synthetic long peptide vaccines can function in a metastatic solid tumor where checkpoint inhibitors are already part of the regimen. The logistical requirements are substantial. A personalized vaccine demands a tumor biopsy, sequencing, neoantigen prediction, peptide synthesis under good manufacturing practice, and formulation and release testing for each individual patient. The biopsy procedure and biospecimen collection listed in the trial record are not incidental details; they are the supply chain inputs for the product being tested.
Speed is a clinical variable in this setting. Patients with metastatic disease can progress during the weeks required to sequence the tumor, predict neoantigens, synthesize peptides, and release the final product. A vaccine that arrives after the patient's condition has deteriorated is a vaccine that cannot be evaluated fairly, so the manufacturing timeline is part of the scientific design rather than a mere operational detail. The trial will generate practical evidence on whether that pipeline can run reliably for 86 patients under protocol…
Related reading: GLP-1 Analogs for Psoriatic Arthritis With Obesity and Diabetes, Peptide Nanostructures Kill Resistant Bacteria and Restore Meropenem, ALXN2420 Phase 1: GH receptor peptide antagonist lowers IGF-1, Tirzepatide tied to lower predicted 10-year CVD risk over 3 years.