Research report: PF chemotherapy combined with a PD-1/CTLA-4 bispecific antibody as first-line treatment in patients with recurrent/metastatic head and neck squamous cell carcinoma: a prospective, single-arm, phase II clinical study

In a prospective, single-arm phase II clinical study at a single cancer center in China, first-line treatment with the PD-1/CTLA-4 bispecific antibody cadonilimab plus platinum and 5-fluorouracil chemotherapy produced an objective response in 9 of 16 patients with recurrent or metastatic head and neck squamous cell carcinoma. The exploratory results, reported by Liao and colleagues in Frontiers in Oncology, included a 56.3% objective response rate, a 50.0% one-year progression-free survival rate and a 75.0% one-year overall survival rate. Median progression-free survival was 9.8 months and median overall survival was 17.7 months after a median follow-up of 18.4 months.

These estimates came with wide confidence intervals and the trial stopped after enrolling 16 of its planned 30 participants because of slow accrual, so the findings require cautious interpretation and confirmation in randomized studies. The trial enrolled adults with histologically confirmed recurrent or metastatic squamous cell carcinoma arising in the oral cavity, oropharynx, larynx or hypopharynx. Disease was not amenable to curative surgery or radiotherapy, and participants had not received systemic therapy for recurrent or metastatic disease.

Other eligibility requirements included at least one measurable lesion under Response Evaluation Criteria in Solid Tumors version 1.1, an Eastern Cooperative Oncology Group performance status of 0 or 1, an estimated life expectancy of at least six months and adequate organ function. Patients were excluded if they had previously received PD-1, PD-L1 or CTLA-4 inhibitors, or if they had active autoimmune disease, another malignancy within two years, systemic immunosuppression, symptomatic central nervous system metastases or active infection. Treatment had an induction phase consisting of intravenous cadonilimab at 10 mg per kilogram on day one, combined every three weeks with platinum and 5-fluorouracil, or PF, chemotherapy for four to six cycles.

Cisplatin was administered at 80 mg per square meter on days one through three, while carboplatin was given at an area under the curve target of five on day one. Each regimen also included a continuous infusion of 5-fluorouracil at 1,000 mg per square meter per day for four days. Patients without progression or unacceptable toxicity then continued cadonilimab alone every three weeks until progression, unacceptable toxicity or withdrawal of consent.

All 16 participants received induction treatment for a median of five cycles, ranging from two to six, and 10 transitioned to maintenance cadonilimab. Six patients stopped during induction, four because of disease progression and two because of patient decision or treatment fatigue. Of the 10 who entered maintenance, seven later discontinued because of progression, one because of patient decision or fatigue and two remained on treatment with ongoing responses at the data cutoff.

The enrolled group had a median age of 59.5 years, ranging from 47 to 71, and 14 participants were male. Most patients had a favorable performance status, with 12 of 16 scoring ECOG 0, while four scored ECOG 1. The oropharynx was the most common primary site, accounting for six patients, followed by the oral cavity with five, larynx with three and hypopharynx with two.

Eight patients had local recurrence, six had regional lymph-node recurrence and four had distant metastasis; some categories overlapped, so their cumulative percentages exceeded 100%. Three of the six patients with oropharyngeal tumors were p16 positive. PD-L1 combined positive scores were at least 20 in six patients, between 1 and 19 in four and below 1 in five, with one result unavailable.

The cohort also had relatively favorable prognostic features, including a high proportion with ECOG 0, p16 positivity among half of patients with oropharyngeal primary tumors and PD-L1 expression of at least 1 in 62.5% of the group. Objective responses were assessed every six to nine weeks, and all nine responses were partial responses; no complete responses were recorded. Five patients, or 31.3%, had stable disease, two had progression at the best assessment, and the resulting disease control rate was 87.5%, with a 95% confidence interval of 61.7% to 98.4%.

The objective response rate was 56.3%, with a 95% confidence interval of 29.9% to 80.2%. The one-year progression-free survival estimate was 50.0%, with a 95% confidence interval of 30.6% to 81.6%, while one-year overall survival was 75.0%, with a 95% confidence interval of 56.5% to 99.5%. Median progression-free survival was 9.8 months, with a 95% confidence interval of 2.8 months to not reached, and median overall survival was 17.7 months, with a 95% confidence interval of 12.3 months to not reached.

Eight patients were alive at the end of follow-up, and four of the nine patients with partial responses had ongoing responses at the data cutoff on June 3, 2026. Exploratory biomarker analyses did not reveal a clear response pattern based on PD-L1 expression. Objective responses occurred in three of six patients with a combined positive score of at least 20, two of four with scores from 1 to 19 and three of five with scores below 1, corresponding to response rates of 50.0%, 50.0% and 60.0%, respectively.

Response was observed in two of three p16-positive patients and four of nine p16-negative patients. The small subgroup counts prevent reliable inference about treatment selection or biological determinants of response. The investigators also described a patient whose p16-positive oropharyngeal carcinoma recurred about 16 months after chemoradiotherapy and who then received four cycles of cadonilimab with chemotherapy.

Follow-up imaging showed substantial tumor shrinkage and a partial response, and the regression lasted 15.2 months. This individual observation documented the response seen in that patient, but the study as a whole was not designed to establish effectiveness in an individual case. Safety findings were based on all 16 participants, and every patient experienced at least one treatment-related adverse event.

Four patients, or 25.0%, had grade 3 or higher treatment-related events. These were primarily blood-cell abnormalities: leukopenia occurred in four patients, neutropenia in three, thrombocytopenia in two and anemia in one, with some patients experiencing several events simultaneously. Hypothyroidism was the only immune-related adverse event reported, affecting seven patients, or 43.8%, but all cases were grade 1 or 2.

Patients with grade 2 hypothyroidism were asymptomatic or mildly symptomatic and were managed with oral levothyroxine without interrupting treatment. The investigators reported no grade 3 or higher immune-related adverse events, treatment-related deaths, immune-related cardiac toxicity, colitis, hepatitis, pneumonitis or dermatitis. They cautioned that the absence of some toxicities may reflect the limited sample size and follow-up rather than establish that they cannot occur.

Cadonilimab, also known as AK104, is a tetravalent bispecific antibody designed to block PD-1 and CTLA-4 simultaneously. Its structure eliminates Fc-gamma receptor binding, which the authors say reduces Fc-mediated effector functions and may improve the therapeutic index. The proposed biological rationale is that PD-1 blockade can restore the function of exhausted tumor-reactive T cells already present in the tumor microenvironment, while CTLA-4 blockade can support T-cell priming and expansion by reducing regulatory T-cell-mediated suppression.

Chemotherapy may also contribute through immunogenic tumor-cell death, release of damage-associated molecular patterns and tumor antigens, and changes in the tumor immune environment. These mechanisms provide context for the observed activity, but the single-arm trial cannot determine whether benefit came from cadonilimab, chemotherapy, their combination or an interaction between them. The response rate appears encouraging when placed beside results from other immunotherapy-based studies in recurrent or metastatic head and neck cancer, but cross-trial comparisons are vulnerable to differences in patient selection, trial design, baseline characteristics and standards of care.

The investigators cited a 36.0% response rate for pembrolizumab plus chemotherapy in KEYNOTE-048 and 24.2% for nivolumab plus ipilimumab in CheckMate 651. These comparisons do not establish superiority. The present cohort had several features associated with better outcomes, including ECOG performance status 0 in 75.0% of patients, p16 positivity in half of those with oropharyngeal primary tumors and PD-L1 combined positive scores of at least 1 in 62.5% of the cohort.

The study was also single-center, lacked an anti-PD-1 comparator, did not formally assess health-related quality of life and examined only p16 and PD-L1 as biomarkers. With 16 participants and early termination, the study offers preliminary evidence of activity and manageable observed toxicity, not confirmation of clinical benefit. The authors conclude that adequately powered randomized trials with biomarker stratification are needed to determine whether dual PD-1/CTLA-4 blockade adds benefit to standard PD-1-based regimens.

Subject of Research: Preliminary antitumor activity and safety of first-line cadonilimab plus PF chemotherapy in recurrent or metastatic head and neck squamous cell carcinoma

Article Title: PF chemotherapy combined with a PD-1/CTLA-4 bispecific antibody as first-line treatment in patients with recurrent/metastatic head and neck squamous cell carcinoma: a prospective, single-arm, phase II clinical study

Article References: Liao, W., Li, Q., Deng, Y., Zhang, J., Zhao, Y., & Zhang, S. (2026). PF chemotherapy combined with a PD-1/CTLA-4 bispecific antibody as first-line treatment in patients with recurrent/metastatic head and neck squamous cell carcinoma: a prospective, single-arm, phase II clinical study. Frontiers in Oncology, 16, Article 1961360. https://doi.org/10.3389/fonc.2026.1961360

Image Credits: AI Generated

DOI: 10.3389/fonc.2026.1961360

Keywords: cadonilimab, head and neck cancer, PD-1, CTLA-4, chemoimmunotherapy, phase II trial, recurrent cancer, metastatic cancer, bispecific antibody, platinum chemotherapy, chemotherapy, combined