Neoadjuvant immunochemotherapy with adebrelimab combined with the TP regimen for locally advanced OSCC: early results of a single-arm phase II clinical trial

Abstract

Background:

This study aimed to investigate the efficacy of a PD-L1 inhibitor combined with chemotherapy in the treatment of locally advanced oral squamous cell carcinoma (OSCC).

Materials and Methods:

This study was designed as a single-arm phase II trial, and a total of 30 patients were enrolled. All patients were pathologically confirmed to have head and neck squamous cell carcinoma. Patients received paclitaxel, carboplatin, and adebrelimab every 3 weeks as a treatment cycle, followed by surgery after three cycles of treatment. The primary endpoint was the postoperative pathological complete response (PCR) rate. The secondary endpoints included the objective response rate (ORR), major pathological response (MPR) rate, 2-year disease-free survival rate, and 2-year and 5-year overall survival rates. This study was conducted without using artificial intelligence (AI) tools in accordance with the TITAN Guidelines 2025.

Results:

Among the 30 patients, 28 completed the full three cycles of immunochemotherapy, with an ORR of 60.7%. One patient dropped out, and one patient (3.3%) experienced grade 3–4 adverse reactions, leading to discontinuation after the first cycle. In the per-protocol population, 10 patients (35.7%) achieved a PCR in both the primary lesion and lymph nodes, whereas 18 patients (64.3%), including those with a PCR, achieved a MPR. Additionally, 20 patients (71.4%) achieved clinical to pathological downstaging. Common adverse reactions included alopecia (30; 100%), fatigue (7; 23.3%), anemia (24; 80%), hyperthyroidism (2; 6.7%), and pruritus (5; 16.7%).

Conclusions:

Preoperative neoadjuvant immunochemotherapy effectively improved the PCR rate in patients with locally advanced OSCC. PD-L1 inhibitors combined with chemotherapy have promising preliminary efficacy in patients with resectable locally advanced OSCC. However, additional follow-up is needed to obtain more comprehensive survival data and further validate these initial observations.

Introduction

According to the latest data from the International Agency for Research on Cancer (IARC), head and neck cancer ranked among the top 10 cancers in terms of global incidence in 2022[1]. For locally advanced head and neck squamous cell carcinoma (HNSCC), surgical resection followed by risk-adapted adjuvant radiotherapy with or without platinum-based chemotherapy or definitive concurrent chemoradiotherapy is recommended according to the standard treatment guidelines. However, despite aggressive multimodal therapy, the risks of recurrence, distant metastasis, and mortality remain high in patients with locally advanced HNSCC^[2,3]. With the introduction of the concept of preoperative induction chemotherapy, recent clinical studies have shown that patients who present a tumor response after preoperative induction chemotherapy have higher survival rates and lower risks of distant metastasis than those who do not respond to this treatment. Nevertheless, the postoperative pathological complete response (PCR) rate remains low, and satisfactory outcomes are not achieved with this approach[4].

Tumor immunotherapy represents a potential solution to the aforementioned challenges. Currently, immune checkpoint inhibitors (ICIs) are widely used to treat oral squamous cell carcinoma (OSCC). A clinical study (KEYNOTE-048) demonstrated that pembrolizumab (a PD-1 inhibitor) combined with chemotherapy significantly improved the survival of patients with recurrent or metastatic OSCC[5]; this regimen has become a first-line treatment option in the National Comprehensive Cancer Network (NCCN) guidelines. Additionally, in a retrospective study by Li[6] et al, neoadjuvant chemoimmunotherapy (NACI) exhibited a good safety profile and encouraging efficacy in patients with locally advanced resectable OSCC. Among the 104 patients who received NACI, the PCR rate was 47.1%, and the major pathological response (MPR) rate was 65.4%. The estimated 3-year disease-free survival (DFS) and overall survival (OS) rates reached 89.0% and 91.3%, respectively, indicating that NACI can effectively improve postoperative pathological response rates and the tumor prognosis.

Programmed death-ligand 1 (PD-L1) plays a role in suppressing the cancer–immunity cycle by binding to negative regulators of T-cell activation, such as PD-1 and B7.1[7]. In a meta-analysis[8], the overall incidence of grade 3 or higher adverse events was significantly higher with PD-1 inhibitors than with PD-L1 inhibitors (OR, 1.58; 95% CI, 1.00–2.54). This result may be attributed to the fact that PD-L1 inhibitors bind to PD-L1 on both tumor cells and antigen-presenting cells while preserving the normal function of PD-L2, thereby reducing the risk of potential autoimmune reactions and adverse events^[9,10]. Therefore, PD-L1 inhibitors may have greater therapeutic potential than PD-1 inhibitors.

Currently, PD-L1 inhibitors are predominantly used to treat lung cancer, liver cancer, triple-negative breast cancer, and other malignancies^[11–13]. Adebrelimab, a humanized monoclonal antibody targeting PD-L1, significantly improved the median overall survival of patients (15.3 months [95% CI: 13.2–17.5]) in a phase III clinical trial evaluating its combination with chemotherapy for small cell lung cancer[14]. However, research on its application in OSCC remains limited, with only phase I clinical studies confirming its safety and antitumor activity in this context[15]. Therefore, this study aimed to further validate the efficacy of PD-L1 inhibitors combined with chemotherapy in the treatment of locally advanced OSCC. This study has been reported in line with the STROCSS guidelines[16].

Materials and methods

Study design and patients

This study was designed as a single-arm phase II trial, and the primary endpoint was the postoperative PCR rate. The secondary endpoints included the objective response rate (ORR), MPR rate, 2-year disease-free survival (DFS) rate, and 2-year and 5-year OS rates. This study aimed to investigate the efficacy and long-term survival outcomes of a PD-L1 inhibitor (adebrelimab) combined with chemotherapy in the treatment of locally advanced HNSCC. This study was conducted without using artificial intelligence (AI) tools in accordance with the TITAN Guidelines 2025[17].

HIGHLIGHTS

• PD-L1 inhibitor combined with chemotherapy significantly improved postoperative pathological complete response rate in patients with locally advanced resectable oral squamous cell carcinoma.

• The efficacy of neoadjuvant immunochemotherapy has a certain persistence.

• PD-L1 inhibitors demonstrated a low incidence of grade 3–4 adverse events and exhibited a favorable safety profile.

• Neoadjuvant immunochemotherapy effectively achieved clinical to pathological downstaging.

After the initiation of this study, all patients were required to undergo a pathological biopsy to confirm the diagnosis and undergo a preliminary screen based on the inclusion and exclusion criteria. Eligible patients were between 18 and 75 years of age with a pathologically confirmed diagnosis of HNSCC (oral cavity, oropharynx, hypopharynx, or larynx) according to the 8th Edition of the American Joint Committee on Cancer (AJCC) staging guidelines. Patients with stage III–IVB nonoropharyngeal cancer and HPV-negative oropharyngeal cancer or stage II–III HPV-positive oropharyngeal cancer were included. Additionally, for all included patients, the resectability of the tumor was evaluated by a head and neck surgeon prior to enrollment, and patients with clinical evidence of distant metastasis were excluded. According to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1, patients who had at least one measurable tumor lesion and an Eastern Cooperative Oncology Group (ECOG) performance status score of 0–1 were included. Other inclusion criteria included meeting the following baseline laboratory test values: hematology – white blood cell (WBC) count ≥ 3.0 × 10⁹/L, absolute neutrophil count (ANC) ≥ 1.5 × 10⁹/L, platelet (PLT) count ≥ 100 × 10⁹/L, and hemoglobin level (HGB) ≥ 9.0 g/dL (without supportive treatments such as blood transfusion or leukocyte enhancement within 7 days); liver function – aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels ≤2.5 times the upper limit of normal (ULN) for patients without liver metastasis and albumin (ALB) levels ≥ 30 g/L; renal function – serum creatinine levels ≤ 1.5 times ULN or creatinine clearance (CrCl) ≥ 50 mL/min (calculated using the Cockcroft–Gault formula), and urine protein (UPRO) < (++) or 24-h urine protein <1.0 g.

For all included patients with oropharyngeal cancer, the HPV status was determined via p16 immunohistochemistry (IHC) to assess whether the criteria for clinical staging were met. None of the included patients had participated in other clinical trials within the past 30 days, and all patients voluntarily provided written informed consent. Patients with abnormal blood tests or liver/kidney function, who were assessed by a multidisciplinary team to be unable to tolerate the study treatments, were excluded. Additionally, patients with a history of other malignancies who had undergone surgery, chemotherapy, or radiotherapy were also excluded. Patients who were unable to complete the entire clinical study process for personal, social, or economic reasons or those with severe systemic diseases that are currently incurable or uncontrolled by medication were also excluded.

Procedures

All enrolled patients who met the inclusion criteria and signed the informed consent form received intravenous infusions of albumin-bound paclitaxel (260 mg/m²), carboplatin (AUC = 5), and adebrelimab (1200 mg) on the first day of each 3-week cycle. Routine antiemetic prophylaxis was administered during each cycle, and dexamethasone was administered before each infusion to prevent allergic reactions. If patients experienced adverse events during treatment, they were assessed according to the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. For grade 2 adverse events, treatment was temporarily suspended and resumed only after the event had resolved to grade 0–1. For grade 3 adverse events, treatment was permanently discontinued, and serious adverse events were recorded and reported. If grade 3 adverse events occurred in 30% of the participants, the study was halted, and the safety of the treatment was re-evaluated. During the study suspension period, follow-up was continued, and the follow-up data of the original participants were maintained. If tumor hyperprogression occurred during the three treatment cycles, imaging studies were performed to confirm the condition, followed by a preoperative evaluation and surgical treatment. Postoperative management was conducted in the oncology department.

Patients who successfully completed three cycles of treatment returned for surgical intervention 21 days after the third cycle, and adjuvant radiotherapy or chemotherapy was administered based on the postoperative pathological stage. Patients who refused surgery were eligible to receive alternative treatment plans in the oncology department or undergo follow-up observation according to their preferences. Throughout the study, all adverse events and surgical complications were assessed and recorded according to the CTCAE version 5.0 and the Clavien–Dindo complication grading system[18].

Endpoints

In this study, the primary endpoint was the postoperative PCR rate; PCR was defined as the absence of viable tumor cells in both the primary lesion and the lymph nodes. The secondary endpoints included the MPR rate, which is defined as the proportion of patients with ≤10% viable tumor cells in the primary lesion specimen; the ORR, which is defined as the proportion of patients who achieve complete remission (CR) or a partial response (PR) after three cycles of immunochemotherapy, as assessed by RECIST version 1.1; the 2-year DFS rate, defined as the proportion of patients who remained free of recurrence or disease progression within 2 years after completing all treatments; and the 2-year and 5-year OS rates, defined as the proportions of patients who survived for 2 years and 5 years, respectively, after completing all treatments. Based on the 8th Edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual and postoperative pathological findings, each patient was re-evaluated for ypTNM staging to analyze the extent of pathological downstaging.

Imaging analysis

Throughout the treatment process, each patient was required to undergo magnetic resonance imaging (MRI) before the initiation of treatment and after completing three cycles. Efficacy was assessed by two radiologists according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1, as shown in Table 1. In cases of disagreement between the two radiologists, a third, more senior radiologist made the final decision that was used to determine the radiological response rate. Moreover, all radiologists interpreting the images were blinded to the patients’ treatment history and regimen, ensuring an objective assessment of tumor changes.

Table 1.

Evaluation of the target lesions

Complete response (CR)

Disappearance of all target lesions. Any pathological lymph nodes (whether target or nontarget) must have a reduction in short axis to <10 mm.

Partial response (PR)

At least a 30% decrease in the sum of diameters of target lesions, taking the baseline sum diameters as the reference.

Progressive disease (PD)

At least a 20% increase in the sum of diameters of target lesions, taking the smallest sum of the diameters recorded as the reference (this value includes the baseline sum if that is the smallest recorded). In addition to the relative increase of 20%, the sum must also demonstrate an absolute increase of at least 5 mm. (Note: the appearance of one or more new lesions is also considered progression).

Stable disease (SD)

This outcome is defined as neither sufficient shrinkage to qualify for PR nor a sufficient increase to qualify for PD, taking the smallest sum diameters while on the study medication as the reference.

Pathology analysis

All patients underwent pathological tissue biopsy before enrollment to confirm the diagnosis. An IHC analysis, including the evaluation of the combined positive score (CPS) and levels of PD-L1, EGFR, Ki67, CD4, CD8, CD20, p16, P53, and other markers, was conducted by pathologists. In this study, the DAKO 22C3 assay was used on the DAKO Autostainer Link 48 platform for the detection and assessment of PD-L1 expression, with a minimum requirement of 100 tumor cells in the examined area. The CPS was calculated as the number of PD-L1-stained cells (tumor cells + lymphocytes + macrophages) divided by the total number of viable tumor cells. After patients underwent surgical treatment, the resected tumor samples were fixed and sectioned. Two pathologists then examined the slides to determine the presence of residual viable tumor cells. If residual tumor cells were present, the proportion of residual tumor cells was calculated to assess treatment efficacy, as shown in Table 2. In cases of disagreement between the two pathologists, a third, more senior pathologist made the final decision. All pathologists were blinded to patients’ treatment regimens and performed objective assessments of residual tumor rates. Additionally, each patient was re-evaluated for the ypTNM stage to analyze the extent of pathological downstaging based on the 8th Edition of the AJCC Cancer Staging Manual and postoperative pathological findings.

Table 2.

Evaluation of pathological specimens

Pathological complete response (PCR)

No viable tumor cells detected in either the primary tumor specimen or lymph nodes.

Major pathological response (MPR)

Viable tumor cells in the primary tumor specimen ≤10%.

Partial pathological response (PPR)

Viable tumor cells in the primary tumor specimen ranging from >10% to <90%.

No pathological response (NPR)

Viable tumor cells in the primary tumor specimen ≥90%.

Monitoring and management

During the study period, patients’ laboratory test results and physical symptoms were monitored throughout each treatment cycle. Adverse events were recorded and graded according to the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0, with appropriate symptomatic management provided. Additionally, postoperative complications were assessed and documented using the Clavien‒Dindo complication grading system during the perioperative period to ensure timely intervention and treatment. This work has been reported in accordance with the STROCSS criteria[19].

Statistical analysis

A phase III clinical trial reported a PCR rate of only 13.4% with preoperative traditional chemotherapy, specifically the TPF regimen (docetaxel, cisplatin, and 5-fluorouracil)[4]. Compared with traditional preoperative chemotherapy (TPF), the combination of TP and immunotherapy in this prospective, single-arm, phase II clinical trial is expected to increase the postoperative PCR rate by approximately 20%, reaching 34%. For this study, we hypothesized that the PCR rate in the experimental group would be superior to that of the group receiving the TPF regimen. The study parameters were set as follows: α = 0.025 (one-sided) and power = 80%. Based on the preliminary research results, the PCR rate in the experimental group was expected to increase by approximately 20%, reaching 34%. Using PASS 15.0 software for the calculation, 27 patients needed to be enrolled. Considering a 10% dropout rate, a total of 30 patients were ultimately included.

The research data in this study were statistically analyzed using the Statistical Package for the Social Sciences (SPSS) and Statistical Analysis System (SAS) software. Continuous variables were analyzed and compared using the medians, and categorical variables were described and compared using percentages. Additionally, the Mann‒Whitney test was used for intergroup comparisons, and the Kaplan‒Meier method was used to estimate patient survival rates.

Results

Patient recruitment for this study began in October 2023, and by December 2024, a total of 30 patients were enrolled (Fig. 1). The patients’ demographic information and clinical characteristics are summarized in Table 3. The median age of the patients was 52 years (range: 34–71), and 23 patients (76.7%) were male. Among the 30 patients, 18 (60%) had tongue cancer, 6 (20%) had buccal cancer, 4 (13.3%) had gingival cancer, 1 (3.3%) had floor-of-mouth cancer, and 1 (3.3%) had HPV-negative oropharyngeal cancer. In terms of the clinical stage, 19 patients (63.3%) had stage III tumors, and 11 patients (36.7%) had stage IVA tumors.

Figure 1.

Flow chart of the study.

Table 3.

Demographic and clinical characteristics

Characteristics	Patients (n = 30)
Age, median (range; years)	52 (34, 71)
Sex, n (%)
Male	23 (76.7)
Female	7 (23.3)
Height, median (range; cm)	165.75 (151, 183)
Weight, median (range; kg)	63.65 (45.6, 95.1)
BMI, median (range)	23.86 (17.99, 31.77)
Smoking, n (%)
No	12 (40.0)
Yes	18 (60.0)
Alcohol use history, n (%)
Never	18 (60.0)
Ever	12 (40.0)
History of areca nut chewing, n (%)
Never	25 (83.3)
Ever	5 (16.7)
Tumor site, n (%)
Tongue	18 (60.0)
Gingiva	4 (13.4)
Oropharynx	1 (3.3)
Floor of mouth	1 (3.3)
Cheek	6 (20.0)
Clinical T stage, n (%)
T1	0 (0.0)
T2	0 (0.0)
T3	23 (76.7)
T4	7 (23.3)
Clinical N stage, n (%)
N0	21 (70.0)
N1	3 (10.0)
N2	6 (20.0)
N3	0 (0.0)
Clinical TNM stage, n (%)
I	0 (0.0)
II	0 (0.0)
III	19 (63.3)
IV	11 (36.7)
PD-L1 combined positive score, n (%)
<1	0 (0.0)
1–19	11 (36.7)
≥20	19 (63.3)

As of 20 January 2025, 28 of the 30 enrolled patients had completed all three planned immunochemotherapy cycles, with adverse event data comprehensively documented for the entire cohort (n = 30). During the treatment period, all patients experienced treatment-related adverse events of varying severity (Table 4), with 1 patient (3.3%) developing grade 3–4 adverse events, leading to the discontinuation of adebrelimab after the first cycle. According to the statistical analysis, the most common adverse events included alopecia (30; 100%), fatigue (7; 23.3%), anemia (24; 80%), hyperthyroidism (2; 6.7%), and pruritus (5; 16.7%). Further analysis revealed that alopecia and anemia consistently manifested within the first week posttreatment. Notably, all patients who developed anemia exhibited normal baseline hemoglobin levels, suggesting that these adverse effects were attributable primarily to chemotherapy (paclitaxel/cisplatin). In contrast, immune-related adverse events (irAEs), including hyperthyroidism and pruritus, typically emerged during weeks 2–3, aligning with the characteristic delayed onset of PD-1 inhibitor-associated toxicities.

Table 4.

Treatment-related adverse events

TRAEs, n (%)	Patients (n = 30)
	Grade 1	Grade 2	Grade 3	Grade 4
Anemia	18 (60.0)	6 (20.0)	0	0
Hyperthyroidism	0	2 (6.7)	0	0
Fatigue	7 (23.3)	0	0	0
Pruritus	5 (16.7)	0	0	0
Alopecia	0	30 (100.0)	0	0
Tumor-related hemorrhage	2 (6.7)	0	0	0
Elevated creatinine levels	5 (16.7)	1 (3.3)	0	0
Leukopenia	0	1 (3.3)	0	0
Elevated alanine aminotransferase levels	0	3 (10.0)	0	0
Elevated creatine phosphokinase levels	5 (16.7)	0	0	0
Elevated aspartate aminotransferase levels	0	1 (3.3)	1 (3.3)	0
Elevated blood bilirubin levels	1 (3.3)	0	0	1 (3.3)
Elevated lipase levels	0	0	0	1 (3.3)
Elevated serum amylase levels	0	0	0	1 (3.3)
Liver failure	0	0	1 (3.3)	0

In terms of perioperative and postoperative management, among the 30 enrolled patients, 1 patient was hospitalized due to severe adverse events after the first cycle of treatment, and all subsequent study medications were immediately discontinued. Seven days after the first cycle of immunochemotherapy, this patient developed jaundice and gastrointestinal symptoms, including loss of appetite. By day 9, the symptoms worsened, with periumbilical pain, anorexia, and fatigue, prompting immediate transfer to the intensive care unit. Laboratory tests revealed acute liver failure (according to the CTCAE 5.0 criteria), with markedly elevated liver enzyme levels: total bilirubin (310.4 μmol/L), ALT (138 U/L), and AST (210 U/L). Immunotherapy was promptly discontinued, and the patient received methylprednisolone alongside a comprehensive liver-protective regimen (adenosylmethionine, reduced glutathione, and magnesium isoglycyrrhizinate). Within one month, the patient’s liver function had normalized. The patient subsequently underwent successful radical surgery for oral cancer and was discharged after recovery. Another patient withdrew from the study for personal reasons and refused surgical treatment and other alternative options. The remaining 28 patients experienced no treatment-related surgical delays. The average postoperative hospital stay for the 29 patients was 8.1 days, with a median hospital stay of 8 days (range: 4–14). No grade 3 or 4 surgical complications were observed, and no patient deaths occurred. The most common postoperative complications included fever (7; 24.1%), pain (6; 20.7%), and swelling (5; 17.2%), which were alleviated within 3 to 4 days after symptomatic treatment. No patients experienced prolonged hospitalization due to severe postoperative complications such as wound infection, pneumonia, bleeding, or flap necrosis (Table 5).

Table 5.

Postoperative complications

Symptoms, n (%)	Patients (n = 29)
	Grade 1	Grade 2	Grade 3	Grade 4
Fever	7 (24.1)	0	0	0
Swelling	5 (17.2)	0	0	0
Postoperative hemorrhage	0	0	0	0
Pain	6 (20.7)	0	0	0
Dysphagia	3 (10.3)	0	0	0
Pneumonia	0	0	0	0
Delayed wound healing	0	0	0	0
Flap necrosis	0	0	0	0

Among the 28 patients who completed the study, imaging revealed that 15 patients (53.6%) achieved a PR, 2 patients (7.1%) achieved a CR, 10 patients (35.7%) had stable disease (SD), and 1 patient (3.6%) experienced disease progression (PD). The ORR was 60.7% (95% CI: 40.58–78.50%). The postoperative pathological evaluation revealed that 10 patients (35.7%; 95% CI: 18.64–55.93%) achieved a PCR, 18 patients (64.3%; 95% CI: 44.07–81.36%) achieved an MPR (including PCR), 9 patients (32.1%) achieved a partial pathological response (PPR), and 1 patient (3.6%) achieved no pathological response (NPR). Additionally, 20 patients (71.4%; 95% CI: 51.33–86.78%) achieved clinical to pathological downstaging, and R0 resection was achieved in all patients who underwent surgery (28/28, 100%). Among the 16 patients who achieved a PR on imaging, 13 achieved an MPR on the postoperative pathology examination. These findings are summarized in Table 6, and a waterfall plot of imaging and pathological responses for each patient is shown in Figure 2. In addition, Figure 3 shows the relationships between pretreatment PD-L1 expression in primary tumor samples and the pathological response associated with neoadjuvant chemoimmunotherapy.

Table 6.

Tumor response

Imaging evaluation, n (%)	Patients (n = 28)
Partial response (PR)	15 (53.6)
Complete remission (CR)	2 (7.1)
Stable disease (SD)	10 (35.7)
Progression of disease (PD)	1 (3.6)
Pathological evaluation, n (%)
Pathological complete response-TN (PCR)	10 (35.7)
Major pathological response (MPR)	18 (64.3)
Pathological partial remission (PPR)	9 (32.1)
No pathological response (NPR)	1 (3.6)
Pathological T stage, n (%)
T0	12 (42.9)
T1	12 (42.9)
T2	2 (7.1)
T3	2 (7.1)
T4	0 (0.0)
Pathological N stage, n (%)
N0	21 (75.0)
N1	3 (10.7)
N2	4 (14.3)
N3	0 (0.0)
Pathological TNM stage, n (%)
0	10 (35.7)
I	8 (28.6)
II	1 (3.6)
III	5 (17.9)
IV	4 (14.3)

Figure 2.

Waterfall plot of tumor responses (n = 30).

Figure 3.

Relationships between pretreatment PD-L1 expression in primary tumor samples and pathological responses associated with neoadjuvant chemoimmunotherapy The combined positive score (CPS) before treatment was as follows: PCR (n = 10), MPR (n = 8), and PPR (n = 9). A, PCR vs. PPR. B, MPR vs. PPR. C, PCR + MPR vs. PPR. P-values were determined using the Mann‒Whitney test.

In the pathological images of PCR patients, fibrotic changes were observed in areas of tumor regression, and significant inflammatory cell infiltration was noted. In the metastatic lymph nodes, foamy cell hyperplasia and multinucleated giant cell reactions were observed, with local areas showing abundant keratin pearl formation. Figures 4 and 5 show the pathological images of the primary lesion and lymph nodes from a patient with T3N2aM0 buccal cancer who achieved a remarkable response after three cycles of treatment. Imaging revealed the disappearance of the primary lesion and shrinkage of the lymph nodes, and postoperative pathology confirmed the absence of residual tumor cells, indicating a PCR.

Figure 4.

Radiological changes in a patient diagnosed with T3N2aM0 buccal squamous cell carcinoma following three cycles of immunochemotherapy. (A and B) Morphological alterations in the left buccal mucosal lesions before and after treatment. (C and D) Contrast-enhanced magnetic resonance imaging (MRI) findings of the primary tumor site before and after therapeutic intervention, revealing the near-complete resolution of the left buccal lesion following three cycles of immunochemotherapy. (E and F) MRI characteristics of cervical lymph node metastases at baseline and after treatment, showing a significant volumetric reduction in the metastatic cervical lymphadenopathy after the completion of three immunochemotherapy cycles.

Figure 5.

Histopathological evaluation of the treatment response in a patient with T3N2aM0 buccal squamous cell carcinoma following three cycles of neoadjuvant therapy. (A) Histology of the pretreatment incisional biopsy specimen showing the characteristic features of squamous cell carcinoma, × 20. B, Illustration of the posttreatment surgical specimen from the primary tumor site, showing a complete pathological response and the absence of residual malignant cells in the tumor bed, × 20. C, Histological findings from the dissected regional lymph node specimen, showing a treatment-induced pathological complete response and lymphocyte infiltration without viable tumor cells, × 20.

As of 28 February 2025, the median follow-up time was 295.5 days (range: 85–492), and all 30 patients were alive. However, one patient whose postoperative pathology revealed no lymph node metastasis experienced a neck recurrence 4 months after surgery. Another patient, who refused radiotherapy despite confirmed lymph node metastasis via postoperative pathology, also developed a neck recurrence 7 months after surgery. Currently, the 2-year DFS rate and the 2-year and 5-year OS rates remain undetermined, and follow-up is ongoing.

Discussion

OSCC is the most common malignant tumor of the head and neck. According to global cancer statistics, over 380 000 new cases of oral cancer were diagnosed in 2022, and oral cancer accounted for approximately 180 000 deaths[20]. Numerous retrospective studies have shown that the TNM stage is closely associated with the tumor prognosis[21]. Higher stages correlate with lower 5-year survival rates[22]. Patients with lymph node metastasis have significantly reduced 5-year survival rates and are more prone to recurrence and even death^[23,24]. Therefore, for patients with locally advanced OSCC, TNM downstaging, minimizing the recurrence risk, and improving the prognosis have become critical challenges to address.

With the advent of immune checkpoint inhibitors (ICIs), particularly PD-1 inhibitors, the overall survival of patients with locally advanced OSCC has significantly improved[25]. Pembrolizumab and nivolumab were approved in 2016 for the treatment of recurrent or metastatic head and neck squamous cell carcinoma (HNSCC)^[26,27], with pembrolizumab becoming a first-line treatment option in the National Comprehensive Cancer Network (NCCN) guidelines for recurrent or metastatic HNSCC[28]. Several studies have demonstrated that neoadjuvant therapy combining PD-1 inhibitors with chemotherapy can effectively improve the ORR and survival rates, thereby improving patients’ prognosis^[29,30]. PD-1 inhibitors work by specifically binding to the PD-1 receptor, preventing the interaction between PD-1 and PD-L1/PD-L2 and thereby blocking inhibitory signaling. This process restores the ability of T cells to recognize and attack tumor cells, enhancing the immune response[31]. However, since PD-1 inhibitors eliminate immune suppression, overactivation of the immune system can lead to indiscriminate attacks on normal tissues, resulting in adverse events such as skin toxicity, thyroid dysfunction, pneumonitis, and hepatitis^[32,33]. Therefore, reducing immune-related adverse events (irAEs) has become an urgent challenge.

In a large meta-analysis[34], compared with PD-L1 inhibitors, PD-1 inhibitors were associated with a higher risk of irAEs of any grade, including rash (OR: 4.34; 95% CI: 2.25–8.39), elevated ALT levels (OR: 4.64; 95% CI: 1.36–15.83), and colitis (OR: 2.53; 95% CI: 1.29–4.95). Additionally, PD-1 inhibitors were associated with a higher risk of certain grade 3 or higher irAEs, including colitis (OR: 3.79; 95% CI: 1.92–7.51), hypothyroidism (OR: 2.85; 95% CI: 1.15–7.04), and rash (OR: 4.38; 95% CI: 1.62–11.82). Currently, no definitive evidence explains how PD-L1 inhibitors reduce the risk of irAEs, which may be related to their mechanism of action. PD-L1 inhibitors preserve the normal physiological function of PD-L2 during blockade, which may be a key factor in reducing the incidence of autoimmune reactions[35]. However, research on the application of PD-L1 inhibitors in OSCC remains limited, with only a few phase I and II clinical studies confirming their safety and antitumor activity[36].

This study is the first to report the therapeutic efficacy of a PD-L1 inhibitor (adebrelimab) in patients with locally advanced OSCC. The ORR was 62.1% (95% CI: 40.58–78.50%), the PCR rate was 35.7% (95% CI: 18.64–55.93%), and the clinical to pathological downstaging rate was 71.4% (95% CI: 51.33–86.78%). R0 resection was achieved in all patients who underwent surgery (28/28, 100%). Further analysis revealed that among patients who achieved a PCR, only 20% (2/10) had systemic comorbidities (such as hypertension and diabetes), whereas this proportion was significantly higher (50%, 5/10) in patients with a partial pathological response (PPR) or no pathological response (NPR). These findings suggest that patients with underlying systemic diseases may have greater difficulty achieving PCR outcomes. Additionally, the statistical analysis revealed that among patients with a CPS ≤20, 61.11% (11/18) achieved PCR or MPR, whereas 38.89% (7/18) experienced a PPR or NPR. Fisher’s exact test yielded a P-value of 0.7029 (well above the 0.05 threshold), indicating no significant association between the CPS stratification (≤20 vs. > 20) and pathological response categories (PCR + MPR vs. PPR + NPR). Although the PCR/MPR rate was numerically higher in the CPS ≥20 group (66.67% vs. 61.11%), the difference was not statistically significant, possibly because of the limited sample size and insufficient statistical power.

In this study, a discrepancy was observed between the ORR assessed by imaging (60.7%) and the PCR rate (35.7%), highlighting certain limitations. The RECIST criteria rely on morphological tumor shrinkage, whereas immunotherapy can lead to pseudoprogression (immune cell infiltration mimicking tumor growth) or tumor fibrosis, despite the absence of viable malignant cells. In contrast, the direct cytotoxic effects of chemotherapy are more readily captured by imaging. Additionally, the delayed effects of immunotherapy may not be reflected in early radiographic assessments. Immunotherapy induces profound remodeling of the tumor microenvironment (characterized by lymphocyte infiltration, necrosis, and inflammation), which can lead to significant radiographic tumor shrinkage – or even complete radiological disappearance – despite the persistence of residual viable tumor cells. This biological phenomenon accounts for the observation that PCR rates are consistently lower than ORRs according to imaging criteria. Notably, accumulating evidence has shown that the PCR serves as a more robust predictor of long-term survival outcomes, particularly in the neoadjuvant treatment setting. This predictive superiority stems from the direct assessment of tumor eradication at the cellular level via the PCR, unlike the imaging-based ORR, which may reflect treatment-related inflammatory changes rather than true antitumor activity. These findings underscore the critical need for a comprehensive response evaluation incorporating both radiographic (ORR) and pathological (PCR) assessments in clinical practice. This dual evaluation provides complementary information that enhances the accuracy of treatment response interpretation and helps prevent potential misjudgments arising from relying solely on imaging modalities.

During the period of medication, the most common adverse events included alopecia (30; 100%), fatigue (7; 23.3%), anemia (24; 80%), hyperthyroidism (2; 6.7%), and pruritus (5; 16.7%), with only a 3.3% incidence of grade 3 or higher adverse events. However, documented cases of alopecia and anemia all occurred within the first week of treatment. Notably, patients who developed anemia had confirmed normal baseline hemoglobin levels prior to treatment, effectively ruling out preexisting anemia. Given the timing of onset and known profiles of chemotherapeutic agents, we concluded that these adverse effects (alopecia and anemia) were primarily induced by chemotherapy. In contrast, irAEs, such as hyperthyroidism and pruritus, manifested later, typically emerging during weeks 2–3 of treatment. This delayed presentation aligns with the characteristic temporal pattern of immune checkpoint inhibitor-related toxicity. Our findings indicate that the PD-L1 inhibitor used in this study exhibited a manageable safety profile, with adverse event rates remaining within acceptable limits. Although one grade ≥3 adverse event was observed, this outcome may reflect the limitations of our small sample size rather than inherent drug toxicity. As an approach to address these limitations, we will continue longitudinal follow-up of the enrolled patients while planning expanded cohort studies to further evaluate the safety profile of PD-L1 inhibitors.

To date, one patient who underwent bilateral neck lymph node dissection (with a pathologically confirmed pN0 status) and received postoperative maintenance monotherapy developed cervical lymph node recurrence at 4 months post surgery. This recurrence may be attributed to either occult micrometastases undetected by conventional pathology or immune escape mediated by immunoediting mechanisms. Another patient experienced recurrence at 7 months postoperatively; this patient had pathologically confirmed lymph node metastasis (pN +) but declined adjuvant radiotherapy for personal reasons. The latter recurrence likely resulted from inadequate local control and an insufficient systemic treatment intensity. These cases highlight the need for improved micrometastasis detection methods and underscore the importance of patient compliance education, suggesting that personalized adjuvant strategies should be developed based on individual recurrence risk profiles. Notably, in one patient who refused surgery after completing three treatment cycles, serial imaging showed continued tumor regression for 4 months following treatment cessation (Fig. 6). This observation suggests potential durable treatment effects or delayed immunological activity, although the confirmation of this result requires larger cohorts and longer follow-up.

Figure 6.

One patient with gingival cancer (T4aN0M0) declined surgery after three cycles of treatment. (A and B), (D and E) Tumor regression was marked after three cycles. (C) Ten months after the discontinuation of the drug, the clinical examination still revealed no signs of tumor progression. (F) Five months after discontinuation of the drug, imaging revealed a further reduction in the tumor volume.

The current study is limited by relatively short follow-up (median 295.5 days), precluding meaningful survival endpoint analyses. Preliminary survival data should be interpreted cautiously, and all enrolled patients should continue protocol-specified follow-up for subsequent survival analyses.

As an exploratory phase II trial, this study provides initial efficacy and safety data for adebrelimab combined with the TP regimen (paclitaxel + cisplatin) for locally advanced OSCC, establishing a foundation for future phase III randomized controlled trials (RCTs). While the single-arm design has inherent limitations, it allows an efficient preliminary evaluation of promising treatment strategies. The TP regimen was selected as the backbone chemotherapy because of its well-established role as standard neoadjuvant therapy for locally advanced OSCC, which is supported by extensive studies. The decision against including a TP-only control arm was intentional to avoid depriving potential beneficiaries of immunotherapy. Additional limitations include the absence of significant correlations between immunohistochemical markers (e.g., CD8 +/EGFR + tumor samples) and treatment outcomes, likely due to the limited sample size. We plan to address this limitation by expanding the study cohort and conducting comparative analyses between 2-cycle regimens and 3-cycle regimens. Furthermore, whole-genome sequencing of paired pre- and posttreatment samples will be performed to identify potential predictive biomarkers.

Conclusions

This report is the first to describe the therapeutic efficacy of a PD-L1 inhibitor (adebrelimab) combined with chemotherapy for locally advanced OSCC. The PD-L1 inhibitor combined with chemotherapy have promising preliminary efficacy in patients with resectable locally advanced OSCC. However, additional follow-up is needed to obtain more comprehensive survival data and further validate these initial observations. These findings highlight the research value of PD-L1 inhibitors. We look forward to the further validation of the effectiveness of PD-L1 inhibitors in larger clinical studies, as these agents may represent new treatment options for patients with locally advanced OSCC.

Ethical approval

This study was approved by the Medical Ethics Committee of Sun Yat-sen Memorial Hospital, Sun Yat-sen University (SYSKY-2023-616-04) and has been registered at clinicaltrials.org (NCT06016413).

Consent

The study complied with the Good Clinical Practice (GCP) guidelines, the Management Measures for Clinical Research Initiated by Researchers in Medical and Health Institutions, and the Declaration of Helsinki. All the participating personnel and data analysts signed confidentiality agreements, ensuring that no patient personal information or disease-related data would be disclosed to any individuals or organizations unrelated to this study. All patients provided written informed consent, and patient enrolment was completed between 26 October 2023 and 5 December 2024.

Sources of funding

This work was supported by Jiangsu Hengrui Pharmaceutical Co., Ltd. In this study, they provided partial free drug support and purchased health insurance for patients. This work also was supported by grants from the National Natural Science Foundation of China (#82303563), the Science and Technology Program of Guangdong (#2023A1515010567, #2021A1515111121), the Guangzhou Science and Technology Project (#202103000093).

Author contributions

Y.H., J.G., Y.W., and Z.H.: Conceptualization, Writing-original draft, review and editing. Z.K., Z.W., Y.C., W.L., and S.F.: Data curation, Formal analysis, Investigation and Methodology. Z.H.: Funding acquisition, project management, Supervision and Visualization. Y.W. and X.Z.: Resources, Supervision, Validation and Writing-review and editing.

Conflicts of interest disclosure

All authors have no conflicts of interest to declare.

Guarantor

The data generated in this study are available upon request from the corresponding author at: hzhquan@mail.sysu.edu.cn.

Research registration unique identifying number (UIN)

This study was approved by ClinicalTrials.gov (NCT06016413).

Provenance and peer review

Not commissioned, externally peer-reviewed.

Data availability

The data sets used for the current study are available from the corresponding author upon reasonable request. All the data generated or analyzed during this study are included in this published article or its Supplementary Information files.