Adjuvant TG4050 was well tolerated and generated long-lasting neoantigen-specific T-cell responses in resected HPV-negative HNSCC.
Adjuvant treatment with the individualized, viral vector–based neoantigen vaccine TG4050 was safe and induced durable neoantigen-specific CD8-positive T-cell responses in patients with resected, human papHPV-negative, locally advanced head and neck squamous cell carcinoma (HNSCC), according to findings from the randomized phase 1 TG4050.02 trial (NCT04183166) published in Nature Communications.1 No patients who received the vaccine immediately after completing standard therapy experienced disease recurrence, whereas relapses occurred among patients assigned to observation.
No dose-limiting toxicities and no grade 3 or higher TG4050-related adverse effects (AEs) were observed, meeting the primary safety end point. In the safety population (n = 19), 94.7% of patients experienced grade 1 or 2 TG4050-related AEs.
At a median follow-up of 30 months, 0 of 16 patients in the per-protocol population who received TG4050 immediately (arm A) experienced relapse vs 3 of 16 patients (18.75%) in the observation arm (arm B). Relapses in arm B occurred at 6, 7, and 18 months. In an updated analysis at a median follow-up of 41 months (data cutoff, March 31, 2026), no additional relapses were reported in either arm. Neoantigen-specific immune responses, detected by ELISpot and/or tetramer analysis, were observed in 73.3% of evaluable arm A patients (n = 15), with a median of 3 responding neoantigens per patient (range, 1-16).
The investigators noted that the sample size and number of relapses were too small to draw conclusions about efficacy, and no hazard ratio or P value for disease-free survival was reported.
“[Single-agent TG4050] induces long-lasting tumor neoantigen-specific cytotoxic T cell responses that can prevent tumor recurrence,” the study authors wrote in their conclusion.
TG4050, developed by Transgene, is a Modified Vaccinia Ankara (MVA) viral vector encoding up to 30 patient-specific neoantigens selected from tumor sequencing using a prediction pipeline developed by NEC Bio. Earlier results from the TG4050.02 trial, presented at the 2024 AACR Annual Meeting at a median follow-up of 18.6 months, showed that all vaccinated patients remained in remission.2
The multicenter, open-label trial enrolled adults with newly diagnosed, locally advanced, HPV-negative, resectable stage III or IV HNSCC of the oropharynx, larynx, hypopharynx, or oral cavity who had an ECOG performance status of 0 or 1.1 Patients underwent surgery followed by adjuvant radiotherapy with or without cisplatin and were required to have a complete response on imaging 3 months after treatment.
A total of 33 patients were randomly assigned to receive TG4050 immediately after confirmed complete response (arm A; n = 17) or to undergo observation, with TG4050 offered alongside standard of care at relapse (arm B; n = 16). Randomization used dynamic minimization by pathologic stage, cisplatin use, and center. TG4050 was administered subcutaneously as 6 weekly induction doses followed by 14 maintenance doses every 3 weeks, for up to 20 injections.
Safety was the primary end point; feasibility and disease-free survival were secondary end points.
Vaccines were manufactured for 53 patients, each containing a median of 30 neoantigens (range, 10-30); 15 of 76 patients who started manufacturing relapsed before the 3-month cutoff and could not receive TG4050. Most patients had oral cavity primary tumors and pathologic stage IVA or IVB disease. The median tumor mutational burden was 3.02 mutations/Mb (range, 0.03-8.0), and 61% of tumors had an immune-deserted microenvironment.
By ELISpot, 61.5% of patients in arm A (n = 13) had a neoantigen-specific response at day 64, and 7 of those 8 patients had at least 1 de novo response. The median breadth was 2 neoantigens (range, 1-16), with a median magnitude of 106 spot-forming cells per 106 peripheral blood mononuclear cells (range, 17-5189). Tetramer analysis identified neoantigen-specific CD8-positive T-cell responses in 64.3% of patients (n = 14), all of which were de novo. MVA vector responses were seen in 84.6% of patients (n = 13).
Neoantigen-specific CD8-positive T cells persisted through month 24, 1 year after the last dose, with a median 3-fold decline from day 64. The breadth of response correlated with tumor mutational burden (P = .02), whereas tumor microenvironment subtype did not. None of the 7 TP53-derived vaccine neoantigens were immunogenic, and none of the 540 vaccine neoantigens were shared between any 2 patients.
Single-cell analyses showed that neoantigen-specific T cells were largely effector and tissue-resident cells with higher PDCD1 and TIGIT expression than MVA-specific cells. Of 68 neoepitope-specific clonotypes, 43 (63%) matched blood T-cell receptor sequences, and 91% of those expanded after vaccination; in 3 of 4 patients analyzed, at least 1 expanded clone was also present in the pretreatment tumor.
The safety population included all 17 patients in arm A and 2 patients in arm B who received TG4050 at relapse. TG4050-related AEs were predominantly injection-site reactions, including inflammation, pain, erythema, bruising, edema, and induration. One of the 2 arm B patients treated at relapse developed CD8-positive responses to 2 neoantigens; clinical outcomes for these patients were not reported because of the small numbers.
A randomized phase 2 portion of the trial is ongoing. The authors wrote that the findings “warrant further investigation of the role of TG4050 in preventing tumor recurrence,” adding that the adjuvant setting “remains the most promising window for monotherapy,” whereas additional strategies are likely needed for advanced disease.
