Inclusion Criteria: (1) Patients with pathologically confirmed HNC via biopsy, newly diagnosed; (2) Patients meeting the indications for radiotherapy; (3) Age ≥16 years; (4) Absence of acute infection or severe systemic comorbidities.

Exclusion Criteria: (1) Prior use of EGFR inhibitor or similar drugs before planned concurrent chemoradiotherapy; (2) Prior surgical resection of the primary tumor or involved lymph nodes before radiotherapy or induction chemotherapy; (3) Radiotherapy interruption due to any reason; (4) History of other malignancies or diagnosis of double primary tumors; (5) Patients with non-evaluable efficacy. This study was approved by the hospital’s Ethics Committee (Approval No.: 2022-Science-091).

This study collected clinical data from 313 HNC patients diagnosed and treated at a tertiary hospital between 2019 and 2022. Follow-up ended at patient death or February 2024. Collected data included gender, age, pathological subtype, HNC type, T stage, N stage, M stage, and clinical stage, and the cohort comprised both nasopharyngeal carcinoma and non-nasopharyngeal head and neck cancer, with nasopharyngeal carcinoma constituting the majority of cases.

Induction Chemotherapy (IC): The interval between IC and radiotherapy (RT), treatment delays, and patient compliance were carefully monitored and recorded to assess their impact on treatment outcomes. The time gap between IC and RT was kept within standard guidelines, and any deviations from the protocol, including delays, were documented. Patients received one of three induction chemotherapy regimens: 1) The TP regimen, consisting of paclitaxel liposome (135–175 mg/m²) administered on day 1, combined with either cisplatin (30 mg/m² on days 1–3) or nedaplatin (80 mg/m² on day 1); 2) The DP regimen, which included docetaxel (75 mg/m² on day 1) along with the same cisplatin or nedaplatin options; 3) The TPF regimen, comprising docetaxel (60 mg/m² on day 1), fluorouracil (600 mg/m² on days 1–5), and the aforementioned cisplatin or nedaplatin schedule.

Radiotherapy: All patients received standard intensity-modulated radiotherapy (IMRT). The specific procedure was: (1) Patient positioning in the supine position with a mouthpiece, head resting on a customized foam pillow, and immobilized using a thermoplastic mask covering the head, neck, and shoulders. (2) CT simulation scan in the supine position, scanning from the vertex to 2 cm below the sternal notch via spiral CT (plain + contrast-enhanced), with a slice thickness of 3 mm. (3) Contouring of target volumes and organs at risk (OARs). MRI or CT was performed after induction chemotherapy and before starting radiotherapy to assess the tumor response and define the gross tumor volume (GTV). This allowed for precise planning of subsequent radiotherapy. Additionally, target volumes, including GTV and clinical target volumes (CTV), were delineated using pre-induction chemotherapy MRI with both non-contrast and contrast-enhanced sequences. The GTV was defined for the primary tumor and involved lymph nodes, while the CTV included areas at high risk for subclinical disease spread. OARs, such as the optic nerves, optic chiasm, lenses, temporal lobes, brainstem, spinal cord, inner ears, parotid glands, and temporomandibular joints, were delineated on the simulation computed tomography (CT) images. GTV-nx: Primary tumor region, including involved retropharyngeal lymph nodes. GTV-nd: Cervical metastatic lymph nodes. CTV-60 (High-risk clinical target volume): Area with high probability of subclinical spread around the primary site, defined as GTV-nx with a 5–10 mm margin, plus the entire nasopharyngeal mucosa and a 5 mm superior-inferior margin. CTV-54 (Low-risk clinical target volume): Area at potential risk of involvement and prophylactic neck nodal regions, including the clivus, parapharyngeal space, posterior nasopharynx, pterygopalatine fossa, etc., and a 5–10 mm expansion from CTV-60 in other directions. PTV: Planning target volumes created by expanding CTV-60 and CTV-54 by 3–5 mm in anterior, superior, inferior, left, and right directions, and 2–3 mm posteriorly, forming PTV-60 and PTV-54, respectively. The prescribed doses were as follows: PGTV-nx received 68–70 Gy, PGTV-nd received 60–70 Gy, PTV-60 received 60 Gy, and PTV-54 received 54 Gy, all delivered in 30 fractions. Treatment interruptions were recorded, and any delays in radiotherapy were documented and managed to minimize their impact on treatment efficacy. Additionally, adaptive radiotherapy replanning was performed in cases of significant tumor shrinkage during treatment to ensure optimal target volume coverage and minimize exposure to surrounding normal tissues. Dose-volume constraints for organs at risk adhered to the Radiation Therapy Oncology Group (RTOG) 0615 and 0225 guidelines.

Targeted Therapy: Patients in the observation group received concurrent nimotuzumab with radiotherapy. Nimotuzumab was administered at 200 mg per dose, diluted in 250 ml of 0.9% sodium chloride solution, and delivered via intravenous infusion. Targeted therapy was given once weekly for a total of 6 doses. Patients with distant metastases received systemic therapies, such as chemotherapy or targeted therapy. Local therapies, including radiotherapy, were considered for oligometastatic disease. Adaptive radiotherapy replanning was performed in cases of significant tumor shrinkage during treatment, ensuring optimal coverage of target volumes while minimizing exposure to surrounding normal tissues.

Imaging results within one month before IC served as the baseline for efficacy evaluation. At 12 weeks after completion of radiotherapy, follow-up imaging was performed. Efficacy was assessed independently by two radiologists blinded to treatment allocation. In cases of discrepancy, a consensus was reached after joint discussion. Response was evaluated using Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST 1.1) (8).

The therapeutic response was evaluated according to the following criteria:

To further assess the association between tumor shrinkage and prognosis, Depth of Response (DpR) was evaluated based on the maximum percentage reduction in the sum of target lesion diameters from baseline to the end of radiotherapy. Patients were classified into three groups:

The primary endpoint of the study was Overall Survival (OS), defined as the time from treatment initiation to death from any cause. Secondary endpoints included Progression-Free Survival (PFS), Objective Response Rate (ORR), and the safety profile of the treatment regimen.

Safety Assessment: All treatment-emergent adverse events (TEAEs) of grade 3 or above, deemed attributable to the study interventions and occurring either during the active treatment phase or within the 90-day post-treatment follow-up window, were systematically recorded and graded in accordance with the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE), version 5.0. In addition, late treatment-related toxicities (assessed ≥3 months post-radiotherapy) were also collected and graded using the same CTCAE criteria. Any-grade toxicities, including nimotuzumab-specific events such as skin rash, infusion-related reactions, and constitutional symptoms, were also documented and descriptively summarized. Quality-of-life outcomes were evaluated based on patient-reported data obtained from the EORTC QLQ-C30 and QLQ-H&N35 questionnaires administered at 6 and 12 months after treatment.

Response assessment after radiotherapy was performed using standard imaging techniques. For residual lesions, metabolic imaging (such as PET-CT) was not routinely performed. However, clinical evaluation and follow-up imaging, including MRI or CT, were used to monitor for progression or resolution. Locoregional control was assessed by monitoring the absence of progression in the primary tumor and regional lymph nodes. Follow-up imaging, including MRI and CT scans, was conducted at regular intervals to detect locoregional recurrence and assess the effectiveness of radiotherapy.

Data were entered into Microsoft Excel 2010 and analyzed using SPSS version 21.0. Continuous variables are presented as mean ± standard deviation (x̄ ± s) and compared using the Student’s t-test. Categorical variables are presented as frequencies (percentages) and compared using the Chi-square (χ²) test. For ordinal data, the non-parametric Mann-Whitney U test was used. Binary logistic regression analysis was employed to identify factors associated with survival prognosis, with T stage, N stage, clinical stage, and treatment modality as independent variables (Xi) and survival status as the dependent variable (Y). A two-sided P-value ≤ 0.05 was considered statistically significant. Given the limited number of death events relative to the number of covariates, the logistic regression model may be prone to overfitting, and its results should therefore be interpreted with caution as exploratory.

Survival Statistical Analysis: Kaplan-Meier survival curves were generated to estimate survival probabilities in the overall cohort and in the nasopharyngeal carcinoma subgroup, and intergroup comparisons were performed using the log-rank test. To determine independent factors influencing survival, both univariate and multivariate analyses were conducted via Cox proportional hazards regression models, which provided hazard ratios (HR) with 95% confidence intervals (CI). The multivariate model was adjusted for age, gender, T-stage, N-stage, clinical stage, histological type of HNC, treatment regimen, and induction chemotherapy regimen.

To explore the consistency of nimotuzumab efficacy across different patient populations, pre-specified subgroup analyses were conducted based on baseline characteristics (age, HNC type, T stage, N stage, clinical stage). Interaction tests were performed to assess the heterogeneity of treatment effects across subgroups.

Differences in overall survival OS across DpR subgroups were assessed using Kaplan-Meier analysis and the log-rank test. The independent prognostic significance of DpR for OS was examined through a multivariate Cox regression model that accounted for potential confounding variables, including age, gender, stage, and treatment modality.

A comparative analysis of patient characteristics demonstrated balanced distribution between groups for sex, T stage, M stage, and clinical stage (all p > 0.05). Although statistical analysis showed imbalances in age, pathological classification, tumor subtype, and N stage between the two groups, these factors, particularly in NPC patients, may heavily influence the observed survival benefit. A more detailed stratification of NPC patients, considering these confounding factors, was performed to minimize the impact of these baseline differences (Table 1). The distribution of induction chemotherapy regimens (TP, DP, and TPF) was comparable between the two groups. In the observation group, 60 patients (41.4%) received the TP regimen, 42 (29.0%) received DP, and 43 (29.7%) received TPF. In the control group, 67 patients (39.9%) received TP, 52 (31.0%) received DP, and 49 (29.2%) received TPF. No significant differences were observed in regimen distribution between the two groups (P = 0.96). The study flowchart is shown in Figure 1.

In the observation group, CR was achieved in 41.38% of patients, PR in 53.79%, SD in 3.45%, and PD in 1.38%. At the final follow-up, 86.90% of patients were alive and 13.10% had died from any cause. In the control group, the rates were as follows: CR 25.00%, PR 58.33%, SD 10.71%, and PD 5.95%, with 77.98% alive and 22.02% deceased. These overall survival and mortality rates reflect all-cause outcomes over the entire follow-up period and should not be interpreted as representing acute treatment-related mortality. The absolute between-group differences were: CR + 16.38%, PR –4.54%, SD –7.26%, PD –4.57%, overall survival +8.92%, and mortality –8.92%.

Comparative analysis of treatment efficacy demonstrated a significantly higher CR rate in the observation group compared to the control group (41.38% vs. 25.00%, P = 0.001), along with notably lower rates of SD and PD (Table 2). Although no significant difference was observed in PR between the two groups (P > 0.05), the shift from non-response (SD/PD) to complete response suggests improved tumor eradication with the addition of nimotuzumab.

Furthermore, the objective response rate (ORR, defined as CR + PR) was significantly higher in the observation group than in the control group (95.17% [138/145] vs. 83.33% [140/168], P ≈ 0.001), indicating an overall enhancement of antitumor activity with the combined regimen.

It should be noted that the absolute CR rates observed in both groups were lower than those typically reported for homogeneous NPC cohorts treated with IMRT, where CR rates often range between 80% and 95%. This discrepancy may be attributed to the inclusion of non-NPC head and neck cancers in our study cohort, the relatively early timing of response assessment at 12 weeks after radiotherapy completion, and the application of stringent RECIST 1.1 criteria, which may underestimate late radiographic resolution of residual lesions in locally advanced disease.

Based on their final survival status, patients were stratified into two distinct cohorts: a Survival Group and a Death Group. Among the 313 patients, 56 (17.89%) had died. There were no significant differences between the two groups in terms of gender, age group, pathological subtype, HNC type, or T stage (P > 0.05). Significant differences were observed in N stage, clinical stage, and treatment modality (P < 0.05) (Supplementary Table S1).

Using survival status as the dependent variable and T stage, N stage, clinical stage, and treatment modality as independent variables (Supplementary Table S2), the logistic regression analysis revealed that treatment with radiotherapy combined with nimotuzumab following induction chemotherapy was a significant protective factor for survival prognosis in HNC patients (P < 0.05). In contrast, T2 stage, T3 stage, and clinical stage II were identified as significant risk factors (P < 0.05) (Supplementary Table S3).

Kaplan–Meier curves for OS are presented in Figure 4. Censoring marks are indicated on the curves, and the numbers at risk are provided at predefined time points to enhance transparency. As shown, a clear separation between the two treatment groups is evident during the first 3–4 years of follow-up. However, the tail of the curves beyond approximately 50 months should be interpreted with caution due to the declining number of patients still under observation, particularly in the radiotherapy-alone group, which results in less stable survival estimates in the later follow-up period. The stepwise pattern observed in the late portion of the curves reflects the actual timing of events within a context of limited long-term follow-up rather than an artifact of the analytical method.

Subgroup analyses based on OS demonstrated that the survival benefit associated with nimotuzumab was particularly notable in patients with nasopharyngeal carcinoma (HR = 0.55, 95% CI: 0.36–0.85), those younger than 60 years (HR = 0.58, 95% CI: 0.38–0.89), and those with extensive lymph node involvement (N2–N3; HR = 0.61, 95% CI: 0.40–0.93). A significant interaction was observed between treatment modality and HNC type (NPC vs. non-NPC) (P for interaction = 0.032), suggesting a stronger treatment effect among NPC patients.

In contrast, no significant OS benefit was observed in the non-nasopharyngeal head and neck cancer subgroup (HR = 1.34, 95% CI: 0.82–2.19; P = 0.241). To address potential bias introduced by tumor-type heterogeneity, we performed a sensitivity analysis limited to NPC patients, which confirmed a consistent survival benefit with nimotuzumab within this population (HR = 0.55, 95% CI: 0.36–0.85). Together, these findings indicate that the overall survival advantage observed in the full cohort is predominantly driven by the NPC subgroup, and caution should be exercised when extrapolating these results to non-NPC head and neck cancers.

We also examined the association between the degree of tumor shrinkage observed at the conclusion of post-induction radiotherapy—measured as the maximal percent reduction in the total diameter of target lesions from baseline—and subsequent long-term survival outcomes. Patients were divided into three groups based on response depth: Deep Response (shrinkage ≥60%), Moderate Response (shrinkage 30%–59%), and Stable Disease/Progressive Disease (SD/PD, shrinkage <30%). Kaplan-Meier analysis showed a significant difference in OS among the three groups (Log-rank P < 0.001). After adjusting in multivariate analysis, Deep Response remained an independent favorable predictor of OS (HR = 0.42, 95% CI: 0.25–0.71, P = 0.001). Notably, the proportion of patients achieving a Deep Response was significantly higher in the observation group compared to the control group (35.2% vs. 18.5%, P = 0.001), providing clinical evidence supporting the mechanism that “nimotuzumab improves survival by enhancing tumor response.” (Figure 5; Table 5).