In the first part of the study, the researchers isolated and sequenced ctDNA from blood samples of six patients with metastatic melanoma, breast cancer, head and neck cancer, or colorectal cancer. Two additional patients, one with breast cancer and one with head and neck cancer, yielded no ctDNA — consistent, Gros explained, with the fact that some tumors shed very little DNA.

The comparison with tissue analysis produced two notable results:

• ctDNA analysis identified 63.25% to 97.4% of the neoantigens found by standard tissue testing across all six patients;
• it also detected neoantigens that tissue analysis missed entirely.

That second point matters for patients with advanced disease. "Patients with advanced disease have tumors in different organs, so a biopsy of one tumor may not capture the neoantigens found in other lesions," Gros said. "By accessing the blood, we can identify neoantigens present across different tumor lesions, as well as T cells able to recognize them. This gives us a broader picture of the cancer throughout the body and could help us develop T-cell therapies that target multiple tumor lesions, making it harder for the cancer to escape treatment."

The researchers also showed the approach captures immune responses, not just targets: T cells isolated from the blood of six of the eight patients recognized and reacted to neoantigens identified by ctDNA, tumor tissue, or both.

Which cancer types worked best?

To test broader applicability, the team expanded the analysis to a separate cohort of 69 patients with various metastatic solid tumors. ctDNA was detectable in 32 of them — 46.4% — suggesting the blood-based strategy could work for roughly half of patients.

Detection rates varied sharply by cancer type. Among 17 patients with colorectal cancer, ctDNA was detectable in 14, or 82.4%. That group included patients with mismatch repair-deficient tumors, which typically produce more neoantigens and are often treated with neoantigen-targeting immunotherapies.

What happens next?

"Our study suggests that blood-based neoantigen identification has the potential to replace or complement the traditional tissue-based approach in many patients and cancer types," Gros said. Because blood is easier and faster to collect than tumor tissue, the method could reach patients with inaccessible tumors and could let others start treatment sooner, without waiting for a biopsy.

Gros also sees potential in tracking how neoantigens and immune responses shift during treatment, which could help clinicians monitor responses and understand tumor evolution and resistance.

The results, though promising, are preliminary. The study involved small patient numbers, and some cancer types were sparsely represented. Gros emphasized that validation in larger cohorts is needed before the approach reaches routine clinical use.

Publication: Peripheral Blood as the Sole Source for Cancer Neoantigen and reactive T cell discovery, Cancer Discovery (2026). DOI: 10.1158/2159-8290.CD-25-1779