Induction toripalimab plus chemotherapy sequential concurrent chemoradiotherapy with nimotuzumab followed by toripalimab maintenance in stage IVB hypopharyngeal squamous cell carcinoma: a single-arm phase 2 study

Abstract

Background

Optimal induction chemoimmunotherapy sequential concurrent chemoradiotherapy (cCRT) in locally advanced hypopharyngeal squamous cell carcinoma (HSCC) remains undefined. This study aimed to evaluate toripalimab plus chemotherapy sequential cCRT with nimotuzumab, followed by toripalimab maintenance in unresectable stage IVB HSCC.

Methods

In this single-arm, open-label, phase 2 study, untreated patients received two 21-day cycles of induction toripalimab 240 mg (d1) plus nab-paclitaxel 230 mg/m² (d1) and cisplatin/nedaplatin 40 mg/m² (d1-2) sequential cCRT (helical tomotherapy; nab-paclitaxel 230 mg/m² (d1, 22, 43)) with nimotuzumab 200 mg weekly for 7 weeks. Subsequently, toripalimab maintenance was administered for eight cycles. Primary endpoint was progression-free survival (PFS).

Results

From March 16, 2022 to April 24, 2024, 31 patients completed induction chemoimmunotherapy. With a median follow-up of 31 months, median PFS and overall survival (OS) were not reached; 2-year PFS and OS rates were 67.7% and 86.4%. The 2-year larynx preservation rate was 93.3%. 30 (96.8%) patients achieved the best objective response, with 28 (90.3%) complete responses and 2 (6.5%) partial responses. The most common grade 3–4 treatment-related adverse events (TRAEs) were radiation-induced oropharyngeal mucositis and decreased appetite (each 46.9%). No grade 5 and fatal TRAEs occurred. No difficulties in speech and swallowing functions were reported in 96.8% of patients. There was no significant association between PFS and PD-L1 combined positive score, epidermal growth factor receptor expression, tumor-infiltrating T-lymphocytes, tumor mutational burden, or genetic alterations. High baseline peripheral CD3⁺ and CD8⁺CD28⁺ T-cell levels were associated with favorable PFS. Exploratory biomarker analyses identified that patients without disease progression after maintenance exhibited significantly higher levels of CD8⁺CD28⁻ T cells with a memory phenotype and lower B-cell levels.

Conclusions

This exploratory study demonstrated preliminary efficacy and larynx preservation with manageable toxicity, possibly supporting further evaluation of induction toripalimab plus chemotherapy sequential cCRT with nimotuzumab, followed by toripalimab maintenance in stage IVB HSCC. Peripheral lymphocytes may warrant further investigation as candidate biomarkers in locally advanced disease.

Trial registration number

NCT05860335; ChiCTR2300074672.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Induction chemoimmunotherapy, sequential concurrent chemoradiotherapy (cCRT), and maintenance therapy in locally advanced hypopharyngeal squamous cell carcinoma (HSCC) preliminarily showed promising efficacy and manageable safety profiles. However, there were limited studies in stable IVB HSCC, a population with poor prognosis.

WHAT THIS STUDY ADDS

• Induction toripalimab plus chemotherapy, sequential cCRT with nimotuzumab, followed by toripalimab maintenance, preliminarily resulted in promising short-term and long-term outcomes, with manageable safety and promising larynx preservation. This study may provide candidate biomarkers warranting further investigation.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Induction toripalimab plus chemotherapy, sequential cCRT with nimotuzumab, followed by toripalimab maintenance, may be a potential regimen for an unresectable stage IVB population, which could also be applicable to most patients with locally advanced HSCC, allowing them to avoid invasive surgery.

Introduction

Hypopharyngeal squamous cell carcinoma (HSCC), an aggressive type of head and neck squamous cell carcinoma (HNSCC), is characterized by long asymptomatic or paucisymptomatic growth, enabling over 50% of patients to be diagnosed at a locally advanced stage.1 2 Despite intensive treatments such as surgery, locally advanced HSCC patients have a poor prognosis; in particular, stage IVB disease is generally unresectable due to extensive local invasion or extranodal extension, associated with a 5-year overall survival (OS) rate of less than 30%.3–5 Under the consensus on function-preservation strategies,6 pivotal studies have established the role of induction chemotherapy, sequential concurrent chemoradiotherapy (cCRT), in larynx preservation, locoregional control, and survival improvement,7–9 making it an evidence-based option for locally advanced HNSCC.10 Notably, our previous retrospective work also demonstrated the feasibility of induction chemotherapy, sequential cCRT, in stage IVB HSCC.11 12 More recently, several studies have revealed encouraging short-term and long-term efficacy and larynx preservation of incorporating immune checkpoint inhibitors (ICIs) into induction chemotherapy, sequential immunoradiotherapy, or cCRT with/without immunotherapy maintenance, with acceptable toxicity profiles in this disease.13–15 Despite potential utility, the optimal immunotherapy-based sequential therapy remains undefined due to limited evidence, especially for patients with stage IVB HSCC.

In this context, beyond the determination of the optimal regimen, an equally important issue revolves around establishing a low-toxicity, well-tolerated approach to reduce gastrointestinal toxicities, because clinically common cisplatin-based cCRT is associated with frequent severe nausea and vomiting (CINV), and thus leads to decreased treatment adherence and quality of life (QoL).16–18 To address this, we replaced cisplatin with nanoparticle albumin-bound paclitaxel (nab-paclitaxel) in the cCRT phase as a rational strategy to mitigate CINV, given that nab-paclitaxel rarely causes CINV.19 In parallel, nab-paclitaxel improved the effects of radiotherapy in a supra-additive manner,20 and exhibited therapeutic potency through the enhanced permeability and retention-mediated tumor-specific uptake.21 Meanwhile, preclinical data of epidermal growth factor receptor (EGFR)-driven resistance to radiotherapy and chemotherapy in up to 90% of HNSCC,22 23 and clinical evidence from our previous study demonstrating significantly improved progression-free survival (PFS) and OS,24 provide a rationale for its incorporation into the cCRT backbone to potentially enhance treatment efficacy.

Based on these integrated rationales, we hypothesized that a sequential strategy as a whole combining chemoimmunotherapy induction with nab-paclitaxel-based cCRT plus nimotuzumab followed by immunotherapy maintenance could maximize therapeutic efficacy while maintaining a manageable safety profile in patients with stage IVB HSCC. Thus, this phase 2 study evaluated the efficacy, larynx preservation, and safety of toripalimab (anti-PD-1 inhibitor approved for multiple cancers)25 26 plus chemotherapy sequential nab-paclitaxel-based cCRT with nimotuzumab followed by toripalimab maintenance in stage IVB HSCC. In parallel, we profiled genetic alterations and reported on the biomarker analyses of PD-L1 combined positive score (CPS), tumor-infiltrating T lymphocytes, EGFR expression, tumor mutational burden (TMB), genetic alterations, and peripheral blood lymphocytes.

Methods

Study design and participants

This single-arm, open-label, phase 2 study was conducted at the Chinese People’s Liberation Army General Hospital and registered with ClinicalTrials.gov (NCT05860335) and ChiCTR.org.cn (ChiCTR2300074672).

Eligible patients were aged 18 to 70 years, with previously untreated stage IVB (according to the eighth TNM staging system of the American Joint Committee on Cancer) HSCC. Patients who had an Eastern Cooperative Oncology Group performance status of 0 to 1, no contraindications to chemotherapy/immunotherapy/radiotherapy, and adequate organ function were enrolled in this study. Key exclusion criteria were active or a history of autoimmune disease, uncontrolled cardiovascular disease, hypopharyngeal necrosis with bleeding risks, immune deficiency, or active infection. Full patient eligibility is provided in the study protocol.

Procedures

Enrolled patients received induction chemoimmunotherapy with toripalimab 240 mg on day 1 plus cisplatin 40 mg/m² on days 1–2 and nab-paclitaxel 230 mg/m² on day 1, with nedaplatin as an alternative to cisplatin for patients with elevated serum creatinine or grade ≥2 gastrointestinal toxicities. Induction chemoimmunotherapy was given intravenously every 3 weeks for two cycles, sequential cCRT with nab-paclitaxel 230 mg/m² on days 1, 22, and 43 and nimotuzumab 200 mg intravenously once a week for 7 weeks (figure 1A). Radiotherapy was delivered using helical tomotherapy with a 6 MV photon beam. High-risk clinical target volumes (CTV1) encompassed the hypopharynx, the entire larynx, and upper/mid-neck lymphatic drainage regions, or the full neck in cases with lower-neck nodal involvement. Low-risk CTV2 included lower-neck lymphatic regions in patients without lower-neck nodal metastasis and the tracheoesophageal region if the cervical esophagus was uninvolved. Planning target volumes (PTV) for the primary tumor and metastatic nodes (pGTVnx/pGTVnd) were generated by isotropically expanding the corresponding GTVs by 3 mm, while PTV1 and PTV2 were obtained by a uniform 3 mm expansion of CTV1 and CTV2, respectively. The prescribed doses were 66–69 Gy for pGTVnx/pGTVnd, 59.4–60 Gy for PTV1, and 54 Gy for PTV2, all delivered in 33 fractions. The target delineation and planning procedures were adapted from the methodology described previously.27 One month after completing cCRT, toripalimab maintenance was intravenously administered at 240 mg once every 3 weeks for eight cycles or until disease progression or intolerable toxicity. Dose reductions were allowed for chemotherapy but not for toripalimab and nimotuzumab.

Figure 1.

Study design and study flow chart. (A) Treatment procedure. The previously untreated patients with stage IVB hypopharyngeal squamous cell carcinoma received two 21-day cycles of intravenous induction chemoimmunotherapy (toripalimab 240 mg on day 1 plus cisplatin/nedaplatin 40 mg/m² on days 1–2 and nab-paclitaxel 230 mg/m² on day 1), sequential cCRT (helical tomotherapy and intravenously nab-paclitaxel 230 mg/m² on days 1, 22, and 43) plus concurrent nimotuzumab (200 mg intravenously once a week for 7 weeks), followed by immunotherapy maintenance (toripalimab 240 mg on day 1 every 3 weeks for eight cycles or until disease progression or intolerable toxicity). (B) Trial profile. cCRT, concurrent chemoradiotherapy; D, day; C, cycle; W, week.

Participants self-reported their sex at birth (male or female) and ethnicity. Tumors were assessed by the investigator using enhanced neck MRI and laryngoscopy as per Response Evaluation Criteria in Solid Tumors version 1.1 at baseline, 1 day before cCRT, after 20 fractions of radiotherapy, and 1 month after cCRT. Patients were followed up for survival every 3 months during the first year and every 4 months thereafter. Adverse events (AEs) were monitored until 3 months after the last dose of study treatment and graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events V.5.0.

Peripheral venous blood of 3–5 mL was collected using EDTA-K2 vacuum tubes at 1 day pretreatment, 1 day before cCRT, and after immunotherapy maintenance. The peripheral blood lymphocytes were examined using flow cytometry with fluorochrome-conjugated primary antibodies against CD3-APC (#A94680; Beckman Coulter, Brea, California, USA), CD4-PC5.5 (#C41174; Beckman Coulter), CD8-APC (#C41196; Beckman Coulter), CD16-PB (#302032; Biolegend, San Diego, California, USA), CD19-APC (#IM2470; Beckman Coulter), CD25-BV605 (#562660; BD, San Jose, CA), CD28-PE (#C41182; Beckman Coulter), CD56-PB (#C76801; Beckman Coulter), CD45RO-ECD (#B49192; Beckman Coulter), CD45RA-FITC (#C41158; Beckman Coulter), Foxp3-PE (#364704; Biolegend), and HLA-DR-BV785 (#307642; Biolegend).

Tumor-infiltrating T lymphocytes, EGFR expression, and PD-L1 CPS were analyzed by immunohistochemistry (IHC) in tumor samples collected at baseline. Primary antibodies included anti-CD3 (#ZM-0417; Zhongshan Golden Bridge Biotechnology, Beijing), anti-CD4 (#RMA-0620; Maixin Biotech, Fuzhou, Fujian), anti-CD8 (#RMA-0514; Maixin Biotech, Fuzhou, Fujian), and anti-EGFR (#ZM-0093; Zhongshan Golden Bridge Biotechnology, Beijing). The PD-L1 expression was centrally assessed using PD-L1 IHC 22C3 pharmDx (Agilent Technologies, Carpinteria, California, USA) and measured as the CPS (defined as the number of PD-L1 staining cells (tumor cells, lymphocytes, and macrophages) divided by the total number of viable tumor cells and multiplied by 100). Based on the clinical significance previously reported,13 14 28 CPSs of 1 and 20 were selected as cut-off points for analysis in the present study.

Whole-exome sequencing was carried out on pretreatment tumor tissue and blood samples to profile genetic alterations. TMB was calculated as the number of somatic mutations per coding region of a tumor genome.

Patient-reported QoL was assessed at baseline, after cCRT, and after immunotherapy maintenance using the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Head and Neck (EORTC QLQ-HN35). Questionnaire scores were postprocessed according to the respective manuals, resulting in scale scores ranging from 0 to 100. The EORTC QLQ-HN35 comprised symptom scales only, with higher scores indicating more severe complaints. The full treatment decision and assessment schedule are available in the study protocol.

Outcomes

The primary endpoint was PFS, which was defined as the time from enrollment to the first documented disease progression (including local, regional lymph node, metastasis, or second primary malignancy) or all-cause death. Secondary endpoints included objective response rate (ORR; proportion of patients with complete response (CR) and partial response (PR)), disease control rate (DCR; proportion of patients with CR, PR, and stable disease (SD)), OS (time from enrollment to death from any cause), 1-year/2-year/3-year PFS/OS rates, larynx preservation rate (LPR; proportion of patients without total laryngectomy or evidence of recurrence on primary site), QoL, and safety. The exploratory endpoint was the associations between outcomes and several factors including PD-L1 CPS, tumor-infiltrating T lymphocytes, EGFR expression, TMB, genetic alterations, and peripheral blood lymphocytes.

Statistical analysis

The sample size of 30 patients was determined based on the exploratory nature of this study. As no historical PFS data were available for patients with unresectable stage IVB HSCC at the time of study design, a formal power calculation based on a predefined PFS threshold could not be performed. The study was therefore designed as a hypothesis-generating investigation to provide preliminary estimates of efficacy and safety to inform future prospective controlled trials.

Efficacy was analyzed in treated patients who had baseline and postbaseline imaging assessments. The safety population was defined as all enrolled patients who received at least one dose of study treatment. Analyses of QoL and biomarkers were performed in patients with available data at baseline and follow-up.

The descriptive statistics were presented as median (IQR or range for continuous variables and frequency (percentage) for categorical variables. Unless otherwise specified, Mann-Whitney U test, unpaired/paired t-test, or Wilcoxon signed-rank test for continuous variables and χ² test or Fisher’s exact test for categorical variables were conducted to compare the differences between groups. The 95% CIs for response rates were calculated using the Clopper-Pearson method. The time-to-event endpoints and LPR were estimated using the Kaplan-Meier method; the log-rank test was used for between-group comparisons for PFS. Patients who were alive and free from disease progression were censored at the date of the last tumor assessment. Alive patients were censored at the last follow-up for OS calculations. To exclude the potential impact of second primary malignancies, a sensitivity analysis for PFS was performed in which five patients were censored at the date of the last tumor assessment. Patients who had missing data were not counted for biomarker and QoL analyses. Statistical analysis was performed using Prism V.9 and R V.4.5.0 software, and p<0.05 was considered statistically significant. Peripheral blood lymphocyte biomarker analyses were conducted using false discovery rate (FDR) correction, with FDR q-values <0.05 considered statistically significant; other biomarker analyses were exploratory and not adjusted for multiple comparisons. All analyses were deemed hypothesis-generating due to the exploratory design and small sample size of this study.

Results

Patient characteristics

Between March 16, 2022 and April 24, 2024, 32 of 33 screened patients were enrolled and received treatment, comprising the safety population (figure 1B). One patient was removed from the efficacy analysis due to no postbaseline imaging assessment. As of September 30, 2025, none of the patients were still on study treatment.

Baseline characteristics of 32 patients are presented in table 1. The median age was 57.5 years (range, 31–69) and 93.8% (30/32) of patients were male. The most common site of the primary tumor was the pyriform sinus (14/32 (43.8%)). All 32 patients had unresectable stage IVB disease, including 14 (43.8%) with T4bN1-2, 10 (31.3%) with T2-4aN3, and 8 (25.0%) with T4bN3. Three (9.4%) patients had human papillomavirus-positive disease as determined by p16 IHC. Totally, 96.9% (31/32) of patients had pretreatment samples to evaluate PD-L1 expression, with 43.8% (14/32) of CPS≥1 and 12.5% (4/32) of CPS≥20. Among 31 (96.9%) EGFR-evaluable patients, 27 (84.4%) were identified as high EGFR expression (IHC 3+) and 4 (12.5%) as low expression (IHC 2+) at baseline.

Table 1.

Baseline characteristics

Characteristics	Patients (n=32)
Age, years-median (range)	57.5 (31-69)
≥60 years	14 (43.8)
Male sex	30 (93.8)
Ethnicity
Han ethnic group	30 (93.8)
Others	2 (6.3)
Smoking status
Former	0
Current	29 (90.6)
Never	3 (9.4)
Drinking status
Former	26 (81.3)
Current	2 (6.3)
Never	4 (12.5)
ECOG performance status
0	16 (50.0)
1	16 (50.0)
Primary tumor site
Pyriform sinus	14 (43.8)
Posterior hypopharyngeal wall	9 (28.1)
Postcricoid	9 (28.1)
T stage by AJCC 8th edition
T2	5 (15.6)
T3	4 (12.5)
T4	23 (71.9)
N stage by AJCC 8th edition
N1	2 (6.3)
N2	12 (37.5)
N3	18 (56.3)
TNM staging
T4bN1-2	14 (43.8)
T2-4aN3	10 (31.3)
T4bN3	8 (25.0)
Differentiation
Low	13 (40.6)
Moderate	16 (50.0)
High	3 (9.4)
Lymph node necrosis	16 (50.0)
Carotid invasion	27 (84.4)
p16 positivity	3 (9.4)
EGFR expression
High (IHC3+)	27 (84.4)
Low (IHC2+)	4 (12.5)
Not available	1 (3.1)
PD-L1 combined positive score
≥1	14 (43.8)
≥20	4 (12.5)
Not available	1 (3.1)

Data are n (%) unless otherwise specified. Baseline characteristics were analyzed in 32 enrolled patients who received at least one dose of study treatment.

AJCC, American Joint Commission on Cancer; ECOG, Eastern Cooperative Oncology Group; EGFR, epidermal growth factor receptor; IHC, immunohistochemistry; PD-L1, programmed death ligand 1; TNM, tumor/node/metastasis.

Treatment compliance

Among 32 patients treated with induction chemoimmunotherapy, one received a single cycle and subsequently underwent cCRT at a local hospital; the remaining 31 patients completed planned two cycles of induction chemoimmunotherapy. 71.9% (23/32) of patients received cisplatin, while 28.1% (9/32) received nedaplatin due to elevated serum creatinine in the first cycle (n=2) and elevated serum creatinine or grade 2 vomiting in the second cycle (n=7). During the induction chemoimmunotherapy, the median relative dose intensity of cisplatin/nedaplatin was 100% (range, 50–100); the median cumulative dose of toripalimab was 480 mg (range, 240–480; online supplemental table S1).

Subsequently, 31 patients then received helical tomotherapy and 30 completed the total dose of radiotherapy. 3.2% (1/31) of patients died from pneumonia and intracranial infection after receiving five fractions of radiotherapy. Five patients had a temporary interruption of radiotherapy because of COVID-19 (n=4) and radiation-induced oropharyngeal mucositis (n=1). The median PTV dose was 59.4 Gy (IQR, 59.4–59.4). The concurrent nab-paclitaxel and nimotuzumab were both administered in 31 patients. The median relative dose intensity of nab-paclitaxel was 100% (range, 33.3–100) and of nimotuzumab was 100% (range, 42.9–100).

Of 30 patients who received maintenance, 28 completed planned administrations of 8 cycles; 2 discontinued treatments due to death, with one from major hemorrhage after recurrence and the other from pneumonia and intracranial infection. The median duration of toripalimab maintenance was eight cycles (range, 4–8). The median cumulative dose of toripalimab was 1920 mg (range, 960–1920).

Survival outcomes and larynx preservation

During a median follow-up of 31 months (range, 6–42), 11 (35.5%) of 31 patients experienced disease progression or death. The median PFS has not yet been reached (figure 2A), with the estimated PFS rate of 74.2% (95% CI 60.3% to 91.3%) at 1 year, 67.7% (95% CI 53.1% to 86.4%) at 2 years and 62.1% (95% CI 46.1% to 83.6%) at 3 years. 16.1% (5/31) of OS events were observed, and the median OS has not yet been reached (figure 2B). The 1-year, 2-year, and 3-year OS rates were estimated as 90.3% (95% CI 80.5% to 100.0%), 86.4% (95% CI 74.8% to 99.8%), and 81.0% (95% CI 66.8% to 98.1%), respectively. Subgroup analysis showed that the T4bN3 stage was significantly associated with poor PFS (p=0.016; figure 2C). Besides, the patient subgroup who responded well to treatment (tumor shrinkage ≥70%) had a significantly improved OS (p=0.004). Throughout the study treatment, no patients received a total laryngectomy. The 1-year, 2-year, and 3-year LPR rates were all 93.3% (95% CI 84.8% to 100.0%; figure 2D).

Figure 2.

Time-to-event outcomes and larynx preservation. Kaplan-Meier estimates of PFS (A) and OS (B). Tick marks indicate censoring of data. (C) Forest plot of PFS and OS subgroup analysis. MPR presented here was defined as tumor shrinkage of ≥70% to <100%; minorPR presented here was defined as tumor shrinkage of ≥30% to <70%. (D) LPR. Outcomes were assessed in 31 treated patients who had baseline and postbaseline imaging assessments. CR/MPR, complete response/major partial response; ECOG, Eastern Cooperative Oncology Group; LPR, larynx preservation rate; MinorPR/SD, minor partial response/stable disease; OS, overall survival; PD-L1, programmed death ligand 1; PFS, progression-free survival; TNM, tumor/node/metastasis.

Of 11 patients with disease progression, six received postprogression therapy and were alive; three patients were disease-free and three were alive with disease at the data cut-off (online supplemental table S2). The causes of progression in these six patients included second primary esophageal carcinoma (n=2), second primary esophageal carcinoma with liver metastasis (n=1), distant liver metastasis (n=1), second primary liver cancer (n=1), and regional recurrence (n=1). The remaining 5 patients experienced disease progression and died: one died due to distant liver and lung metastases during subsequent therapy; the other 4 patients died prior to subsequent therapy, with causes of death including pneumonia and intracranial infection (n=2), respiratory failure due to tracheoesophageal fistula (n=1), and major hemorrhage after recurrence (n=1). The results of sensitivity analysis for PFS were consistent with the primary analysis (median, not reached (NR); 1 year rate, 83.9% (95% CI 71.9% to 97.9%); 2-year rate, 80.1% (95% CI 66.9% to 95.8%); 3-year rate, 80.1% (95% CI 66.9% to 95.8%; online supplemental figure S1).

Radiological response to treatment

After induction chemoimmunotherapy, 27 (87.1%; 95% CI 70.2% to 96.4%) of 31 patients achieved objective response, with 8 (25.8%) CR and 19 (61.3%) PR (table 2). Among partial responders, 14 (45.2%) achieved significant tumor shrinkage of 70% or more (defined as major partial response (MPR); figure 3A). An additional 4 (12.9%) patients had SD, for a DCR of 100.0% (95% CI 88.4% to 100.0%). After 20 fractions of radiotherapy, the ORR was 100.0% (30/30; 95% CI 88.8% to 100.0%), including 24 (80.0%) patients having a CR and 6 (20.0%) having a PR (figure 3B). At 1 month after cCRT, CR and PR were respectively observed in 93.3% (28/30) and 6.7% (2/30) of patients, resulting in an ORR of 100.0% (30/30; 95% CI 88.8% to 100.0%). Overall, the protocol treatment was effective in significant tumor shrinkage in patients with a high disease burden (figure 3C,D).

Table 2.

Radiological response to treatment

	After induction (n=31)	20th cCRT (n=30)	1 month post-cCRT (n=30)
Overall tumor site
CR	8 (25.8)	24 (80.0)	28 (93.3)
PR	19 (61.3)	6 (20.0)	2 (6.7)
Reduction of ≥70% in tumor size	14 (45.2)	3 (10.0)	0
Reduction of ≥30% to <70% in tumor size	5 (16.1)	3 (10.0)	2 (6.7)
SD	4 (12.9)	0	0
PD	0	0	0
ORR	27 (87.1; 70.2 to 96.4)	30 (100.0; 88.8 to 100.0)	30 (100.0; 88.8 to 100.0)
DCR	31 (100.0; 88.4 to 100.0)	30 (100.0; 88.8 to 100.0)	30 (100.0; 88.8 to 100.0)
Response at primary sites
CR	22 (71.0)	27 (90.0)	29 (96.7)
PR	7 (22.6)	3 (10.0)	1 (3.3)
Reduction of ≥70% in tumor size	3 (9.7)	1 (3.3)	0
Reduction of ≥30% to <70% in tumor size	4 (12.9)	2 (6.7)	1 (3.3)
SD	2 (6.5)	0	0
PD	0	0	0
Response at lymph nodes
CR	14 (45.2)	25 (83.3)	29 (96.7)
PR	13 (41.9)	5 (16.7)	1 (3.3)
Reduction of ≥70% in tumor size	11 (35.5)	2 (6.7)	0
Reduction of ≥30% to <70% in tumor size	2 (6.5)	3 (10.0)	1 (3.3)
SD	4 (12.9)	0	0
PD	0	0	0

Data are n (%) or n (%, 95% CI). Efficacy was analyzed in treated patients who had baseline and post-baseline imaging assessment after induction chemoimmunotherapy, after 20 fractions of radiotherapy, and 1 month after cCRT.

cCRT, concurrent chemoradiotherapy; CR, complete response; DCR, disease control rate; ORR, objective response rate; PD, disease progression; PR, partial response; SD, stable disease; 20th cCRT, 20 fractions of radiotherapy.

Figure 3.

Tumor response and treatment duration. (A) Waterfall plot of percentage reduction of tumor diameter from baseline as measured by Response Evaluation Criteria in Solid Tumors V.1.1. (B) Swimming plot of treatment duration. MPR presented here was defined as tumor shrinkage of ≥70% to <100%; minorPR presented here was defined as tumor shrinkage of ≥30% to <70%. (C) Pretreatment and post-treatment enhanced neck MRI images from a stage T2N3M0 patient. A 47-year-old man with PD-L1 CPS<1 showed a primary lesion in the right aryepiglottic fold and piriform sinus with bulky right cervical lymphadenopathy (largest, 9.0×4.5×8.0 cm) at baseline, encasing carotid artery. After induction chemoimmunotherapy, the primary lesion completely resolved and cervical lymph nodes markedly regressed (largest, 1.8×1.3×2.2 cm), achieving an MPR. CR was observed after 20 fractions of radiotherapy, with no evidence of recurrence or metastasis 1 month postradiotherapy, and sustained tumor-free survival at 37 months of follow-up. (D) Pretreatment and post-treatment enhanced neck MRI images from a stage T4bN2cM0 patient. A 54-year-old man with PD-L1 CPS of 7 had a mass (3.1×4.5×5.7 cm) involving the right piriform sinus, postcricoid region, thyroid/cricoid cartilage, and soft tissue at baseline, with carotid artery encasement, extending superiorly to the aryepiglottic fold and inferiorly to the esophagus, accompanied by multiple bilateral cervical lymph nodes. After induction chemoimmunotherapy, both the primary lesion and cervical lymphadenopathy decreased in size, achieving PR. After 20 fractions of radiotherapy, a marked tumor regression was observed, achieving MPR. 1 month after completion of radiotherapy, complete tumor resolution (CR) was confirmed. At the data cut-off, the patient remained tumor-free with a follow-up of 40 months. cCRT, concurrent chemoradiotherapy; CPS, combined positive score; CR, complete response; MPR, major partial response; MinorPR, minor partial response; PD, progressive disease; PD-L1, programmed death ligand 1; SD, stable disease; TNM, tumor/node/metastasis.

Safety

During the induction chemoimmunotherapy, 32 (100.0%) of 32 patients experienced at least one treatment-related AE (TRAE) of any grade, with 16 (50.0%) at severity of grade 3–4. The most common TRAE at grade 3 was neutropenia (21.9% (7/32)) and at grade 4 was neutropenia (6.3% (2/32); table 3; online supplemental table S3). No serious TRAEs were observed during the induction chemoimmunotherapy.

Table 3.

Treatment-related adverse events in the safety population

	Safety population (n=32)
	Any grade	Grade 3	Grade 4
Induction chemoimmunotherapy phase (n=32)
Any events	32 (100.0)	14 (43.8)	2 (6.3)
Alopecia	32 (100.0)	0	0
Decreased appetite	32 (100.0)	4 (12.5)	0
Nausea	32 (100.0)	0	0
Hemoglobin decreased	30 (93.8)	0	0
Pharyngodynia	24 (75.0)	0	0
Fatigue	22 (68.8)	1 (3.1)	0
Dysphagia	22 (68.8)	0	0
Leukopenia	20 (62.5)	3 (9.4)	0
Hoarseness	19 (59.4)	0	0
Neutropenia	17 (53.1)	7 (21.9)	2 (6.3)
Peripheral neuropathy	16 (50.0)	0	0
γ-glutamyltransferase increased	15 (46.9)	1 (3.1)	0
Constipation	15 (46.9)	0	0
Alanine aminotransferase increased	14 (43.8)	0	0
Vomiting	13 (40.6)	1 (3.1)	0
Cough	12 (37.5)	0	0
Pruritus	11 (34.4)	0	0
Fever	11 (34.4)	1 (3.1)	0
Hypokalemia	10 (31.3)	0	0
Rash	10 (31.3)	0	0
Arthralgia	10 (31.3)	1 (3.1)	0
Blood bilirubin increased	7 (21.9)	0	0
Creatinine increased	7 (21.9)	0	0
cCRT with nimotuzumab phase (n=31)
Any events	31 (100.0)	23 (74.2)	4 (12.9)
Decreased appetite	31 (100.0)	15 (48.4)	0
Nausea	31 (100.0)	3 (9.7)	0
Dry mouth	31 (100.0)	4 (12.9)	0
Subcutaneous tissue fibrosis	31 (100.0)	3 (9.7)	0
Radiation-induced oropharyngeal mucositis	31 (100.0)	14 (45.2)	0
Dermatitis radiation	31 (100.0)	11 (35.5)	0
Hemoglobin decreased	31 (100.0)	2 (6.5)	0
Fatigue	29 (93.5)	7 (22.6)	0
Hoarseness	29 (93.5)	3 (9.7)	0
Pharyngodynia	28 (90.3)	8 (25.8)	0
Dysphagia	28 (90.3)	5 (16.1)	0
Leukopenia	27 (87.1)	8 (25.8)	3 (9.7)
Alopecia	24 (77.4)	0	0
Peripheral neuropathy	23 (74.2)	3 (9.7)	0
Cough	19 (61.3)	0	0
Constipation	18 (58.1)	0	0
Neutropenia	18 (58.1)	6 (19.4)	3 (9.7)
γ-glutamyltransferase increased	17 (54.8)	3 (9.7)	0
Pruritus	17 (54.8)	0	0
Fever	17 (54.8)	2 (6.5)	0
Thrombocytopenia	15 (48.4)	3 (9.7)	0
Rash	15 (48.4)	1 (3.2)	0
Vomiting	15 (48.4)	2 (6.5)	0
Arthralgia	15 (48.4)	3 (9.7)	0
Alanine aminotransferase increased	13 (41.9)	1 (3.2)	0
Creatinine increased	12 (38.7)	0	0
Hypokalemia	12 (38.7)	0	0
Blood bilirubin increased	10 (32.3)	0	0
Pneumonitis	9 (29.0)	0	1 (3.2)
Diarrhea	8 (25.8)	0	0
Aspartate aminotransferase increased	7 (22.6)	1 (3.2)	0
Hyponatremia	4 (12.9)	2 (6.5)	0
Immunotherapy maintenance phase (n=30)
Any events	30 (100.0)	10 (33.3)	2 (6.7)
Decreased appetite	30 (100.0)	6 (20.0)	0
Nausea	30 (100.0)	1 (3.3)	0
Fatigue	28 (93.3)	4 (13.3)	0
Hemoglobin decreased	27 (90.0)	1 (3.3)	0
Cough	15 (50.0)	0	0
Constipation	15 (50.0)	0	0
Pruritus	13 (43.3)	0	0
Fever	13 (43.3)	1 (3.3)	0
Hypothyroidism	12 (40.0)	0	0
Rash	12 (40.0)	1 (3.3)	0
Vomiting	9 (30.0)	0	0
Pneumonitis	8 (26.7)	1 (3.3)	1 (3.3)
Leukopenia	7 (23.3)	0	0
Arthralgia	7 (23.3)	1 (3.3)	0
γ-glutamyltransferase increased	5 (16.7)	2 (6.7)	0
Neutropenia	2 (6.7)	1 (3.3)	0
Hyponatremia	2 (6.7)	1 (3.3)	1 (3.3)

Data are n (%). Safety population included all enrolled patients who received at least one dose of study treatment. The patient proportion of adverse events was calculated within the safety population for each treatment phase. Listed are adverse events of any grade with an incidence >20% and all grade 3–4 events in either group.

cCRT, concurrent chemoradiotherapy.

During the cCRT with nimotuzumab, the incidence of any-grade and grade 3–4 TRAEs was 100.0% (31/31) and 87.1% (27/31), respectively. The frequent grade 3–4 TRAEs included decreased appetite (48.4% (15/31)) and radiation-induced oropharyngeal mucositis (45.2% (14/31)). 3.2% (1/31) of patients experienced serious pneumonia and intracranial infection.

During the immunotherapy maintenance, grade 3–4 TRAEs were documented in 12 (40.0%) of 30 patients and the most frequent TRAE was decreased appetite (20.0% (6/30)). Serious TRAEs were observed in 3.33% (1/30) of patients and were pneumonia and intracranial infection. Throughout the study treatment, no grade 5 TRAEs, treatment-related discontinuation, and fatal TRAEs were observed.

Biomarker analyses

Given the limited number of OS events, analyses assessing the association between OS and biomarkers would be underpowered and were not conducted. Therefore, we focused on the associations of biomarkers with PFS and treatment response.

Association of biomarkers with PFS

High baseline levels of CD8⁺CD28⁺ T cells (NR vs 22.5 months, p=0.03; figure 4A) and CD3⁺ T cells (NR vs 29.0 months, p=0.021; figure 4B) in the peripheral blood were associated with significantly improved PFS. Besides, no significant associations were observed between PFS and other peripheral blood lymphocyte subsets (online supplemental figure S2), PD-L1 expression (online supplemental figure S3A and B), EGFR expression (online supplemental figure S3C), and tumor-infiltrating T lymphocytes (online supplemental figure S3D–F).

Figure 4.

Exploratory biomarker analysis. Kaplan-Meier estimates of PFS stratified by baseline levels of CD8⁺CD28⁺ T cells (median, 12.0; A) and CD3⁺ T cells (median, 74.5; B) in peripheral blood. (C) Genetic alteration spectrum in baseline tumor tissue samples. Frequency distribution of common genetic alterations between CR/MPR and minorPR/SD or between PD/death and non-PD patients. MPR presented here was defined as tumor shrinkage of ≥70% to <100%; minorPR presented here was defined as tumor shrinkage of ≥30% to <70%. (D) Box plots showing a significant difference in the change of peripheral blood lymphocytes. Between-group comparisons were performed using the false discovery rate (FDR) correction, and biomarkers with FDR q-values <0.05 were considered statistically significant; however, these analyses were still deemed hypothesis-generating only due to the exploratory design and small sample size of this study. 1st indicated peripheral blood lymphocytes at baseline; 3rd indicated peripheral blood lymphocytes after immunotherapy maintenance. cCRT, concurrent chemoradiotherapy; CPS, combined positive score; CR/MPR, complete response/major partial response; MinorPR/SD, minor partial response/stable disease; mo, month; NK, natural killer; NR, not reached; PD, progressive disease; PD-L1, programmed death ligand 1; PFS, progression-free survival; TMB, tumor mutational burden; TNM, tumor/node/metastasis.

Baseline tumor tissues from 31 patients were eligible for whole-exome sequencing analyses to profile the genetic alterations. The mean TMB was 8.5 mut/Mb, with tTMB-high status (defined as tumor tissue TMB of ≥10 mut/Mb) observed in 12 patients. Patients with a tTMB-high status exhibited a numerically prolonged PFS versus those with a tTMB-low status (NR vs NR, p=0.20; online supplemental figure S3G). The most common genetic alteration was TP53 (100.0% (31/31)), followed by NOTCH1 (38.7% (12/31)), AJUBA (19.4% (6/31)), MUC16 (19.4% (6/31)), and NAV3 (19.4% (6/31); figure 4C). However, genetic alterations were not associated with PFS (online supplemental figure S4).

Association of biomarkers with treatment response

In this study, patients with available data at the corresponding treatment phase were divided into two groups according to the depth of response after induction chemoimmunotherapy (CR/MPR and minorPR/SD groups). No significant differences in peripheral blood lymphocyte subsets between the CR/MPR and minorPR/SD groups at either baseline or postinduction (online supplemental table S4).

Patients were stratified by the disease control status after immunotherapy maintenance (disease progression (PD)/death and non-PD groups). The significantly higher baseline levels of CD3⁺ T cells and CD3⁻CD16⁺/CD56⁺ natural killer cells in peripheral blood were seen in the non-PD group (online supplemental table S5). After immunotherapy maintenance, patients who remained progression-free by data cut-off exhibited higher levels of CD8⁺CD28^- T cells, CD3⁺ T cells, CD3⁺CD8⁺ T cells, and CD8⁺/CD4⁺ ratio; while patients who experienced PD or death had higher levels of CD3^-CD19⁺ B cells in peripheral blood (online supplemental table S5). There was a significantly higher level of CD8⁺CD28^- T cells and CD3⁺CD8⁺ T cells in the non-PD group and CD3⁻CD19⁺ B cells in the PD/death group (figure 4D). Besides, there were no between-group differences in baseline PD-L1 expression, EGFR expression, tumor-infiltrating T lymphocytes, tTMB, and mutation patterns of TP53 (online supplemental figures S5 and S6).

Patient-reported QoL

The assessments of speech and swallowing function were performed in 32 patients. During a 31-month follow-up, one patient underwent tracheotomy and one patient underwent gastrostomy after treatment; the remaining 30 patients reported no difficulty in phonatory and swallowing functions. 32 patients completed at least one QoL assessment. There was no significant change in the EORTC QLQ-HN35 symptom scales from baseline after immunotherapy maintenance (online supplemental figure S7).

Discussion

To our knowledge, this was among the first prospective studies to demonstrate the feasibility of comprehensive treatment in previously untreated, locally advanced HSCC. Specifically, induction toripalimab plus chemotherapy, followed by nab-paclitaxel-based cCRT with nimotuzumab and subsequent toripalimab maintenance, preliminarily demonstrated beneficial signals of tumor shrinkage, sustained disease control, and long-term survival. Besides, mirroring the pattern previously reported,12 29–31 patients who achieved CR/MPR after induction chemoimmunotherapy exhibited promising OS, whereas those with T4bN3 disease were associated with poor PFS. Alongside promising effects on larynx preservation and QoL, and manageable toxicities, our sequential regimen may be considered a potential nonsurgical approach for patients with unresectable stage IVB HSCC carrying a poor prognosis, requiring further confirmation in a randomized study.32

There have been three recent prospective efforts adopting a similar concept of comprehensive treatment.13–15 With exclusive enrollment of patients with stage cT3-4aN0-2M0 HSCC, induction camrelizumab plus TPF chemotherapy sequential radiotherapy plus concurrent camrelizumab or surgery, followed by camrelizumab maintenance resulted in a PFS rate of 77.1% and OS rate of 83.0% at 2 years.13 Tislelizumab-based comprehensive treatment without immunotherapy maintenance demonstrated a 1-year PFS of 69.8% and 1-year OS of 77.4% in stage T1N+M0 or T2-4N0-3M0 HSCC.15 Besides, despite 48.1% of patients with stage IVB disease, induction toripalimab plus chemotherapy followed by cCRT or surgery and toripalimab maintenance achieved 1-year OS and PFS rates of 84.7% and 77.6%, respectively, in a resectable cohort that also enrolled laryngeal carcinoma with a relatively favorable prognosis.14 By contrast, our study focused on a more challenging population of unresectable, stage IVB HSCC and yielded clinically meaningful survival benefits (2-year/3-year PFS, 67.7%/62.1%; 2-year/3-year OS, 86.4%/81.0%). However, the cross-trial comparisons should be interpreted cautiously due to the differences in study design and patient selection.

In this study, the promising efficacy may be the result of the cumulative effect of the entire treatment sequence with the unique design of the dosing regimen. Notably, compared with our previous studies evaluating induction chemotherapy sequential cCRT with nimotuzumab in stage IVB HSCC,11 12 the numerically improved outcomes observed in the present study may largely relate to the addition of immunotherapy, although the single-arm design precluded definitive determination of the relative contribution of each treatment component. In line with the original study assumption, the synergy of chemoimmunotherapy in both innate and adaptive immune activation33 during the induction phase and the unique pharmacological advantages of nab-paclitaxel contributed to the considerable reductions in tumor burden. Such therapeutic potency was amplified by sequential cCRT substituting platinum with nab-paclitaxel and adding nimotuzumab, as evidenced by the increased CR over time at the targeted lesions (25.8% to 93.3%), primary site (71.0% to 96.7%), and lymph nodes (45.2% to 96.7%). In addition, unlike the aforementioned studies that employed intensity-modulated radiotherapy,13–15 helical tomotherapy used in the present study may also serve as a contributor to the favorable outcomes, consistent with its positive role in the improved locoregional control and survival.34 35 The long-term benefits were further maintained with subsequent administration of immunotherapy maintenance 1 month after cCRT—a treatment sequence allowing for immune recovery and supporting more robust antitumor immunity—contrasting with the negative outcomes of trials where ICIs were initiated immediately before or concurrently with radiotherapy.36–38 The last argument was that our treatment advantage may be driven by full target dose exposure in most patients (87.5%) owing to high tolerability. Overall, these efficacy signals may have highlighted the clinical implications of the proposed sequential regimen to improve patient outcomes, although the exploratory nature of this study rendered these findings primarily hypothesis-generating.

Historically, the 1-year LPR of 66.7%–88.7% reported in similar studies may result from the application of larynx-preservation surgery for non-responders alongside induction chemoimmunotherapy.13–15 By contrast, a potential strength of this study may lie in the achievement of a 1-year LPR of 93.3% in more advanced patients—a sole result of postinduction tumor shrinkage in all patients with disease control—thereby allowing them to preserve the laryngeal structure using sequential cCRT alone without surgical intervention. Equally noteworthy, beyond the preservation of anatomical structure, promising speech and swallowing functions and stable QoL indicated that protocol treatment appeared to maximize the retention of larynx function. While initially designed for an unresectable stage IVB population, our sequential regimen could be applicable to most patients with locally advanced HSCC, allowing them to avoid invasive surgery.

The relatively high proportion of second primary malignancies, predominantly esophageal carcinoma, observed among PFS events in this study was consistent with prior studies and may reflect the co-occurrence of hypopharyngeal and esophageal squamous cell carcinoma attributable to shared carcinogenic exposures, including tobacco, alcohol, and gastroesophageal reflux.39 40 Although carcinogenic exposure may be partially mitigated following primary tumor treatment, the risk of metachronous second primary malignancy persists and increases with time.39 Overall, these findings suggested that, in the clinical management of hypopharyngeal and esophageal cancers—particularly in patients with a heavy history of smoking and alcohol consumption—attention should be paid to the potential occurrence of second primary malignancies, and appropriate mutual screening should be considered.

The predictive value of widely studied conventional biomarkers for immunotherapy, such as PD-L1 CPS and TMB, remained controversial in locally advanced head and neck cancers.13–15 In this study, the similar survival outcomes between subgroups appeared to suggest that the use of immunotherapy in stage IVB HSCC was not mainly driven by PD-L1 CPS or TMB. Besides, no prognostic values of tumor-infiltrating T lymphocytes, EGFR expression, and TP53 alterations were reported here. The analysis in peripheral blood samples demonstrated that high baseline levels of peripheral CD3⁺ and CD8⁺CD28⁺ T cells were associated with longer PFS, suggesting their potential role as candidate biomarkers of therapeutic efficacy.

An important observation of exploratory longitudinal analysis of dynamics in peripheral blood lymphocytes was the significant enrichment of CD8⁺CD28⁻ T cells—exhibiting functional features of memory T cells beyond their role of immunomodulation41 42—in non-PD patients after immunotherapy maintenance, along with a marked increase compared with baseline levels. Consequently, we speculated that with decreasing tumor burden, a subset of CD8⁺ T cells was differentiated to effector memory T cells and entered the peripheral circulation, a process considered essential for maintaining sustained immune responses and improved outcomes in patients receiving ICIs.43 44 In support of this hypothesis was most patients in this study remained progression-free over a 2-year follow-up after treatment completion. Besides, patients without disease progression after immunotherapy maintenance exhibited lower levels of peripheral CD3^-CD19⁺ B cells, consistent with a prior report.45 Such findings suggested that peripheral B cells may impair the efficacy of PD-1-based immunotherapy, mechanistically by inhibiting the induction of T-cell-dependent antitumor immune responses.46 Overall, an increase in B cell frequency may serve as a potential biomarker for identifying patients at higher risk of progression following an initial response to ICIs in locally advanced HSCC. Further clinical and translational analyses are needed to validate these exploratory biomarker findings.

The observed safety profiles were consistent with toxicity profiles of the individual treatment components and historical studies,13–15 47–50 without new safety signals. Despite numerically higher incidence of grade 3–4 TRAEs after induction chemoimmunotherapy (50% vs 26%–51%) than in similar studies,13–15 it remained acceptable when contextualized by the exclusive inclusion of more advanced stage IVB HSCC patients. Importantly throughout the entire sequence of therapy, most TRAEs were manageable with clinically common strategies, including dose modifications or supportive care; no patients required treatment-related discontinuation and none died of TRAEs. The absence of CINV during the cCRT may support our initial hypothesis favoring a non-platinum regimen, where platinum was replaced with nab-paclitaxel.19 Besides, considering a median follow-up of over 2 years, we anticipated that late toxic effects, particularly swallowing dysfunction, would not be expected. Overall, the clinical benefits achieved here were not accompanied by an increased risk of treatment-related toxicities.

Common limitations for phase II studies were a relatively small sample size, a single-arm design, and the absence of an independent central review committee. The relative contribution of each component of the treatment sequence cannot be determined, and the observed efficacy may reflect the cumulative effect of the entire regimen; meanwhile, this study was not powered for meaningful statistical analysis of subgroup analysis and biomarkers, but shed valuable insights into identifying patients who may benefit most from immunotherapy-based strategies—an underexplored area in locally advanced HSCC. Besides, the analysis of tumor tissue-centered biomarkers may also be limited by the heterogeneity across biopsy sites. In addition, as this was an exploratory study without a formal statistical hypothesis or historical PFS benchmark, the findings should be interpreted as preliminary and hypothesis-generating rather than demonstrating superiority over existing therapies. Finally, mature OS data are needed and the survival results might be updated with continuous follow-up.

Conclusions

In conclusion, induction toripalimab plus chemotherapy sequential nab-paclitaxel-based cCRT with nimotuzumab followed by toripalimab maintenance preliminarily resulted in a promising tumor response and long-term survival, with a manageable safety profile and stable QoL. Combined with a 1-year LPR, our sequential strategy may represent an active, low-toxicity, and well-tolerated treatment option for the individualization of definitive therapy in stage IVB HSCC patients desiring laryngeal preservation, warranting further confirmation in well-designed trials. In parallel, peripheral blood lymphocytes may represent potential biomarkers of immunotherapeutic efficacy, and, if validated in larger trials, could serve to guide treatment management.

Data availability statement

Data are available on reasonable request. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. If you encounter any issues, please contact the corresponding authors (XZ, xinxinzhang66@hotmail.com).

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

This study was performed in accordance with the Declaration of Helsinki, the international standards of good clinical practice, and applicable local laws and regulations. The study protocol was approved by the Ethics Committee of the Chinese People’s Liberation Army General Hospital (approval number: S2021-339-02). All patients provided written informed consent before enrollment.