The efficacy of concurrent chemoradiotherapy with or without nimotuzumab in high‐risk patients with locally advanced head and neck squamous cell carcinoma post‐surgery: A real‐world study

Abstract

Purpose

Patients with locally advanced head and neck squamous cell carcinoma (LA‐HNSCC) harboring high‐risk pathological features after radical surgery face high recurrence rates. This real‐world study evaluated the efficacy and safety of concurrent chemoradiotherapy (CCRT) with and without nimotuzumab in this postoperative population.

Methods

Patients with LA‐HNSCC who had high‐risk pathological features after surgery were stratified as follows: those with extranodal extension (ENE) or positive/insufficient margins received nimotuzumab plus CCRT (NCCRT; extremely high risk), whereas those with other high‐risk factors (e.g., pT3–4 and N2–3) received CCRT alone (intermediate high risk). All patients underwent intensity‐modulated radiotherapy with weekly cisplatin. Overall survival (OS) was the primary endpoint, and event‐free survival (EFS) and safety were secondary endpoints. To minimize confounding bias due to non‐randomized treatment allocation, propensity‐score matching (PSM) was performed using a 1:1 nearest neighbor matching algorithm with a caliper width of 0.2 times the standard deviation of the propensity score.

Results

A total of 75 patients were enrolled in this study (CCRT, n = 50; NCCRT, n = 25). After PSM, 22 well‐balanced pairs were identified with no significant differences in baseline characteristics between the two groups. The median follow‐up time was 29.80 months (95% confidence interval [CI]: 25.83–33.77) in the CCRT group and 30.23 months (95% CI: 25.07–35.40) in the NCCRT group. In the overall cohort, the 2‐year EFS and OS rates were 74.1% vs 60.6% and 85.6% vs 83.6%, respectively. In the matched cohort, the 2‐year EFS and OS rates were 73.8% vs 58.3% (log‐rank p = 0.087) and 86.2% vs 82.5% (log‐rank p = 0.342), respectively. Univariate analysis revealed that the platelet‐to‐lymphocyte ratio (PLR) and hemoglobin‐albumin‐lymphocyte‐platelet (HALP) score were significantly associated with OS in the CCRT group (p = 0.019 and p = 0.009, respectively). Treatment‐related adverse events were predominantly grades 1–2, including radiodermatitis, oral mucositis, and hematological toxicities, with no significant increase in severe toxicities in the NCCRT group.

Conclusions

Risk‐stratified adjuvant CCRT, with or without nimotuzumab, was associated with promising outcomes and acceptable safety in patients with high‐risk postoperative LA‐HNSCC. Thus, PLR and HALP may serve as candidate prognostic biomarkers for intermediate‐risk patients and warrant further validation. These findings are hypothesis‐generating and require confirmation in prospective randomized studies with dedicated control groups.

Figure 1 Patient selection flowchart. CCRT, concurrent chemoradiotherapy; NCCRT, nimotuzumab and concurrent chemoradiotherapy.

1. INTRODUCTION

Head and neck squamous cell carcinoma (HNSCC) arises from the mucosal epithelium of the upper aerodigestive tract (oral cavity, pharynx, larynx, and nasal cavity) and is the seventh most prevalent malignancy worldwide. Recent global cancer statistics estimate approximately 890,000 new cases and 450,000 deaths annually, representing 4.5% and 4.6% of total cancer incidence and mortality, respectively. ¹ Despite advances in multimodal therapy combining surgery, radiotherapy, and chemotherapy, the prognosis for locally advanced HNSCC (LA‐HNSCC) remains suboptimal. ² Current National Comprehensive Cancer Network (NCCN) and Chinese Society of Clinical Oncology (CSCO) guidelines recommend surgical resection as the primary curative approach for resectable LA‐HNSCC, with adjuvant therapy determined by postoperative risk stratification. ³ , ⁴ Specifically: 1) Postoperative radiotherapy alone is indicated for patients with intermediate‐risk features (pT3–4 disease, close margins < 5 mm, N2–3 nodal status, level IV‐V lymph node involvement, or perineural/lymphovascular invasion); 2) Concurrent chemoradiotherapy (CCRT) is recommended for high‐risk patients (positive margins or extranodal extension [ENE]). ⁵ Nevertheless, treatment outcomes remain disappointing, with 5‐year overall survival (OS) rates below 40% for patients with LA‐HNSCC. ⁶ This persistent therapeutic challenge underscores the urgent need for novel treatment strategies to improve survival in patients with LA‐HNSCC.

The epidermal growth factor receptor (EGFR) is overexpressed in 80%–100% of patients with HNSCC, ⁷ , ⁸ making it an attractive therapeutic target. Cetuximab, the first FDA‐approved anti‐EGFR monoclonal antibody, has demonstrated clinical benefits in patients with advanced HNSCC. Studies have shown that cetuximab combined with CCRT is feasible and tolerable for stage III‐IV HNSCC patients with high‐risk features (positive margins, extracapsular extension, or ≥ 2 lymph node metastases). ⁹ Notably, the combination of cetuximab with docetaxel has shown superior OS and progression‐free survival (PFS) compared to historical studies. ¹⁰ However, meta‐analyses indicated no significant prognostic difference between weekly and triweekly cisplatin regimens in adjuvant chemoradiotherapy. ¹¹ Although the anti‐EGFR antibody cetuximab has demonstrated efficacy, its combination with CCRT was frequently hampered by severe mucocutaneous toxicities, leading to treatment interruptions that compromise local control. Nimotuzumab is a humanized monoclonal antibody that targets EGFR and inhibits EGFR‐dependent signaling by binding to its extracellular domain. This monoclonal antibody exhibits significant anti‐tumor activity, primarily by influencing biological processes, such as cell proliferation, growth, and angiogenesis, while also enhancing radiosensitivity and reducing drug resistance. Moreover, the drug demonstrates high target selectivity and a long half‐life, effectively inhibiting tumor progression. ¹² , ¹³ Importantly, previous studies suggested that nimotuzumab may cause milder skin and mucosal reactions compared to other anti‐EGFR antibodies. ¹⁴ A multicenter propensity score matched (PSM) study (n = 612) showed significant improvement in 3‐year OS (74.6% vs 63.3%, p = 0.038) and 3‐year PFS (57.7% vs 44.3%, p = 0.021) with nimotuzumab‐containing regimens, compared with radiotherapy or concurrent radiotherapy alone. ¹⁵ Meta‐analyses confirmed the superior efficacy of nimotuzumab plus radiotherapy/chemoradiotherapy in unresectable LA‐HNSCC, with comparable safety. ¹⁶ Therefore, nimotuzumab in combination with radiotherapy or chemoradiotherapy may represent a more effective treatment option for unresectable LA‐HNSCC. However, the adjuvant therapeutic value of nimotuzumab in postoperative LA‐HNSCC patients with high‐risk factors such as ENE remains unclear and warrants further exploration. Additionally, whether patients with LA‐HNSCC and non‐ENE high‐risk factors should receive cisplatin combined with radiotherapy alone merits further investigation. Therefore, to address these questions, we conducted a real‐world analysis of patients with LA‐HNSCC treated at Peking University Shenzhen Hospital. This study aimed to evaluate the prognostic impact of nimotuzumab‐enhanced CCRT in ENE‐positive high‐risk cases as well as the outcomes of CCRT alone in non‐ENE high‐risk patients. Our study aimed to provide evidence‐based guidance for risk‐adapted adjuvant therapy for patients with LA‐HNSCC.

2. MATERIALS AND METHODS

2.1. Patients

We retrospectively screened 212 patients with LA‐HNSCC treated at Peking University Shenzhen Hospital between March 2020 and June 2023. From this initial cohort, 55 patients were excluded due to a lack of radiotherapy, 67 did not meet the inclusion criteria, 7 had incomplete data, and 8 were lost to follow‐up (Figure 1). Eight patients lost to follow‐up were excluded from the survival analysis because their outcome data were not available. However, these patients initially received standard treatments based on different risk factors. This study was approved by the Institutional Review Committee of Peking University Shenzhen Hospital in accordance with the Declaration of Helsinki (No. 2025018). This study was prospectively registered with the Chinese Clinical Trial Registry (ChiCTR2500111656). The inclusion criteria included: (1) Patients were diagnosed with LA‐HNSCC confirmed by pathological biopsy; (2) Aged 20 to 70 years; (3) Eastern Cooperative Oncology Group (ECOG) score of 0–1; (4) Adequate hematologic, hepatic, and renal function; (5) All patients underwent radical surgical treatment; (6) The American Joint Committee on Cancer (AJCC) 8th Edition Staging System (2017) defines LA‐HNSCC as Stage III, IVA, IVB, and all patients have > 1 high‐risk pathological feature, including pT3–4 disease, close margins [5 mm], N2–3 nodal status, level IV‐V lymph node involvement, or perineural/lymphovascular invasion, positive or insufficient margins, and ENE; (7) The patient's clinical medical records are complete. The exclusion criteria were as follows: (1) nasopharyngeal carcinoma; (2) allergy to nimotuzumab or any of its components; and (3) synchronous primary tumors (assessed via panendoscopy and chest imaging).

FIGURE 1.

Patient selection flowchart. CCRT, concurrent chemoradiotherapy; NCCRT, nimotuzumab and concurrent chemoradiotherapy.

2.2. Treatment

All patients underwent radical surgical treatment, including radical resection of the primary HNSCC lesion, neck lymph node dissection, and skin flap repair surgery, if necessary. Intraoperative frozen‐section analysis was performed to confirm adequate surgical margins. Radiotherapy was initiated within six weeks of surgery. Patients with positive or insufficient surgical margins and ENE were assigned to the NCCRT group, which was defined as an extremely high‐risk group owing to a higher risk of recurrence or metastasis. In contrast, patients with other common high‐risk factors (T3–4, N2–3, lymph nodes in regions IV or V, vascular invasion, and perineural invasion) were assigned to the CCRT group, which was defined as intermediate–high risk.

Radiotherapy details: All patients underwent intensity‐modulated radiation therapy (IMRT) using a linear accelerator (Varian Medical Systems). The prescribed radiation dose to the postoperative tumor bed and high‐risk nodal regions was 60–66 Gy in 2.0 Gy per fraction, delivered once daily for 5 days/week for a total of 6 weeks. Clinical target volumes and planning target volumes (PTVs) were delineated according to institutional protocols and international contouring guidelines. Chemotherapy details: Cisplatin (Qilu Pharmaceutical Group Co., Ltd.) was administered intravenously at a dose of 30–40 mg/m² once weekly, concurrently with radiotherapy for a total of six cycles. Dose adjustments or delays were permitted based on renal function, hematologic toxicity, or other clinically significant adverse events, according to the institutional guidelines.

Nimotuzumab administration: In the NCCRT group, nimotuzumab (200 mg per dose; Biotech Pharmaceutical Co. Ltd.) was administered via intravenous infusion once weekly for six doses, concurrently with cisplatin and radiotherapy. Antihistamine premedication was not routinely required. Treatment completion: In the CCRT group, 50/50 patients (100.0%) completed all planned cycles of cisplatin and radiotherapy. In the NCCRT group, 25/25 patients (100.0%) completed the full course of nimotuzumab, cisplatin, and radiotherapy. Treatment discontinuation was primarily due to grade 3–4 toxicities or patient refusal. Adverse reactions were monitored and recorded during the treatment.

2.3. Assessments

The primary endpoint was OS, and event‐free survival (EFS) and safety were the secondary endpoints. OS was defined as the time from the first day of CCRT with and without nimotuzumab to death from any cause or last follow‐up. The EFS was defined as the time from the first day of CCRT with and without nimotuzumab to disease progression, locoregional recurrence, distant metastasis, second primary cancer, or death from any cause. Inflammatory and nutritional indicators included: neutrophil‐to‐lymphocyte ratio (NLR), platelet‐to‐lymphocyte ratio (PLR), and derived neutrophil‐to‐lymphocyte ratio (dNLR) = neutrophil count (×10⁹/L) divided by the difference between white blood cell count (×10⁹/L) and lymphocyte count (×10⁹/L); systemic immune‐inflammation index (SII) = platelet count (×10⁹/L) multiplied by PLR; hemoglobin‐albumin‐lymphocyte‐platelet score (HALP) = albumin level (g/L) × hemoglobin level (g/L) × lymphocyte count (×10⁹/L) / platelet count (×10⁹/L); prognostic nutritional index (PNI) = albumin value (g/L) plus five times the lymphocyte count (×10⁹/L). Body mass index (BMI) was calculated by dividing the body weight (kg) by the square of height (m²). Based on the World Health Organization (WHO) standards, patients were categorized into three groups: underweight (BMI < 18.5 kg/m²), normal weight (18.5 kg/m² ≤ BMI ≤ 24.9 kg/m²), and overweight (BMI ≥ 25 kg/m²). Inflammatory and nutritional biomarkers, including the PLR and HALP, were selected based on their established roles in reflecting systemic inflammation and host immunonutritional status, which are associated with prognosis in multiple cancer types. These biomarkers can be readily obtained from routine blood tests, which supports their potential utility in clinical practice. Baseline weight and height were obtained from the patients' medical records within 7 days treatment initiation.

2.4. Statistical analyses

Statistical analyses were conducted using IBM SPSS Statistics (version 26.0; IBM Corp) and GraphPad Prism (version 6.0; GraphPad Software). The optimal cut‐off values for continuous variables were determined using receiver operating characteristic (ROC) curve analysis based on Youden's index. Notably, ROC‐derived cut‐offs were generated and evaluated within the same cohort; therefore, these threshold values should be considered exploratory and require external validation before clinical application. Categorical variables are summarized as frequency counts (n, %). Between‐group comparisons for unordered categorical variables were performed using the χ ² test or Fisher's exact test (for expected frequencies < 5). Ordinal or ranked variables were analyzed using the Mann–Whitney U test or Kruskal–Wallis test. Continuous variables were compared using Student's t‐test or analysis of variance (ANOVA) for normally distributed data and the Wilcoxon rank‐sum test for non‐normally distributed data. OS and EFS were estimated using the Kaplan–Meier method, with log‐rank tests used for between‐group comparisons. Multivariate Cox proportional hazards regression models were employed to identify the independent prognostic factors after adjusting for covariates. The assumption of proportional hazards was verified using the Schoenfeld residuals. All tests were two‐sided, and statistical significance was set at p < 0.05.

To mitigate the potential confounding bias inherent in nonrandomized treatment allocation, PSM was performed. The propensity score was estimated using a logistic regression model that included the following clinically relevant baseline covariates: age, sex, T stage, N stage, tumor location, depth of invasion, perineural invasion, and lymphovascular invasion. Patients were matched using a 1:1 nearest neighbor matching algorithm with a caliper width of 0.2 times the standard deviation of the propensity score. Standardized mean differences (SMDs) were calculated to evaluate the balance of covariates between the two groups after matching, with an SMD < 0.1 indicating a negligible difference. Notably, due to the absence of an ENE‐positive cohort treated with CCRT alone, direct attribution of outcomes to nimotuzumab intensification was not feasible, and the PSM analysis primarily served to balance non‐ENE‐related covariates between the two risk‐defined groups.

3. RESULTS

3.1. Patients characteristics

This study enrolled 75 patients with LA‐HNSCC: 50 in the CCRT group and 25 in the NCCRT group. The entire cohort was predominantly male (46/50 in the CCRT group and 24/25 in the NCCRT group) and elderly (60 years: 40/50 in the CCRT group and 25/25 in the NCCRT group). Most patients presented with advanced disease, including T3–4 classification (35/50 in the CCRT group and 19/25 in the NCCRT group), N2–3 nodal involvement (25/50 in the CCRT group and 18/25 in the NCCRT group), and stage IV disease (29/50 in the CCRT group and 19/25 in the NCCRT group). The tongue was the most common tumor location in both groups (26/50 in the CCRT group; 19/25 in the NCCRT group, respectively). The buccal mucosa and oropharyngeal primaries were exclusively present in the CCRT group (16/50 and 8/50, respectively), whereas gingival primaries were only observed in the NCCRT group (6/25). In the NCCRT group, a depth of invasion ≥ 0.95 cm (22/25) and perineural invasion (19/25) were most prevalent in the NCCRT group. Notably, all patients in both groups had negative surgical margins (100% in each group), confirming that margin status was not a distinguishing factor between the two cohorts. Consequently, treatment allocation was solely based on the presence of ENE: ENE was present in all NCCRT patients (100%) and absent in all CCRT patients (0%). Human papillomavirus (HPV) status was predominantly negative in both cohorts (45/50 in the CCRT group and 23/25 in the NCCRT group). Consequently, the NCCRT group exhibited a higher baseline risk profile.

A total of 13 (26.0%) patients in the CCRT group and 4 (16.0%) patients in the NCCRT group had a BMI ≥ 25 kg/m². Conversely, there was a higher incidence of patients exhibiting white blood cells (WBC) < 5.305 ×10⁹/L (26/50 in the CCRT group, 14/25 in the NCCRT group), LYM ≥ 1.265 ×10⁹/L (38/50 in the CCRT group, 17/25 in the NCCRT group), NLR ≥ 2.007 (27/50 in the CCRT group, 16/25 in the NCCRT group), and PLR ≥ 149 (27/50 in the CCRT group, 17/25 in the NCCRT group). Furthermore, an increased number of patients presented with dNLR < 0.9633 (48/50 in the CCRT group, 14/25 in the NCCRT group), SII < 678.9 (36/50 in the CCRT group, 16/25 in the NCCRT group), HALP < 44.17 (38/50 in the CCRT group, 22/25 in the NCCRT group), and PNI ≥ 45.2 (29/50 in the CCRT group, 19/25 in the NCCRT group). Detailed clinical characteristics are presented in Table 1. As shown in Table 1, there were significant baseline differences between the two groups, particularly in terms of tumor location, N stage, depth of invasion, ENE, and several inflammatory biomarkers. After propensity‐score matching, 22 well‐balanced pairs were identified. The baseline characteristics in the matched cohort were well‐balanced between the two groups, with no significant differences observed in any of the included covariates.

TABLE 1.

Patient clinical characteristics.

	Treatment (before matching)			Treatment (after PSM)
Variable	CCRT (n, %) N = 50	NCCRT (n, %) N = 25	p	CCRT (n, %) N = 22	NCCRT (n, %) N = 22	p
Gender			0.639			1.000
Male	46 (92.0)	24 (96.0)		20 (90.9)	21 (95.5)
Female	4 (8.0)	1 (4.0)		2 (9.10)	1 (4.50)
Age (years)			0.020
> 60	40 (80.0)	25 (100.0)		22 (100.0)	22 (100.0)	1.000
≤ 60	10 (20.0)	0 (0.00)		0 (0.00)	0 (0.00)
Location			< 0.001			0.106
Tongue	26 (52.0)	19 (76.0)		14 (63.6)	16 (72.7)
Buccal mucosa	16 (32.0)	0 (0.00)		5 (22.7)	0 (0.00)
Gingiva	0 (0.00)	6 (24.0)		0 (0.00)	6 (27.3)
Oropharynx	8 (16.0)	0 (0.00)		3 (13.6)	0 (0.00)
T stage			0.581			1.000
T1‐T2	15 (30.0)	6 (24.0)		6 (27.3)	6 (27.3)
T3‐T4	35 (70.0)	19 (76.0)		16 (72.7)	16 (72.7)
N stage			0.065			1.000
N0‐N1	25 (50.0)	7 (28.0)		7 (31.8)	7 (31.8)
N2‐N3	25 (50.0)	18 (72.0)		15 (68.2)	15 (68.2)
Pathological stage			0.121			0.749
III	21 (42.0)	6 (24.0)		5 (22.7)	6 (27.3)
IV	29 (58.0)	19 (76.0)		17 (77.3)	16 (72.7)
Pathological differentiation			0.564
High	20 (40.0)	8 (32.0)
Median	29 (58.0)	15 (60.0)
Low	1 (2.0)	2 (8.0)
Depth of invasion (cm)			0.005			1.000
< 0.95	23 (46.0)	3 (12.0)		4 (18.2)	3 (13.6)
≥ 0.95	27 (54.0)	22 (88.0)		18 (81.8)	19 (86.4)
Lymph vascular invasion			0.092
No	39 (78.0)	15 (60.0)
Yes	11 (22.0)	10 (40.0)
Perineural invasion			0.063			0.752
No	23 (46.0)	6 (24.0)		7 (31.8)	5 (22.7)
Yes	27 (54.0)	19 (76.0)		15 (68.2)	17 (77.3)
Perifocal invasion			1.000
No	22 (44.0)	11 (44.0)
Yes	28 (56.0)	14 (56.0)
Surgical margin
Negative	50 (100.0)	25 (100.0)		22 (100.0)	22 (100.0)
Lymph node metastasis			0.087
No	7 (14.0)	0 (0.00)
Yes	43 (86.0)	25 (100.00)
ENE			< 0.001
No	50 (100.00)	0 (0.00)		22 (100.0)	0 (0.00)
Yes	0 (0.00)	25 (100.0)		0 (0.00)	22 (100.0)
HPV			0.834
Negative	45 (90.0)	23 (92.0)
Positive	4 (8.0)	2 (8.0)
Unknown	1 (2.0)	0 (0.00)
BMI (kg/m2)			0.319
< 18.5	6 (12.0)	1 (4.0)
18.5–24.9	31 (62.0)	20 (80.0)
≥ 25	13 (26.0)	4 (16.0)
WBC (×109/L)			0.745
< 5.305	26 (52.0)	14 (56.0)
≥ 5.305	24 (48.0)	11 (44.0)
LYM (×109/L)			0.466
< 1.265	12 (24.0)	8 (32.0)
≥ 1.265	38 (76.0)	17 (68.0)
NLR			0.409
< 2.007	23 (46.0)	9 (36.0)
≥ 2.007	27 (54.0)	16 (64.0)
PLR			0.250			0.757
< 149	23 (46.0)	8 (32.0)		8 (36.4)	7 (31.8)
≥ 149	27 (54.0)	17 (68.0)		14 (63.6)	15 (68.2)
dNLR			< 0.001
< 0.9633	48 (96.0)	14 (56.0)
≥ 0.9633	2 (4.0)	11 (44.0)
SII			0.480
< 678.9	36 (72.0)	16 (64.0)
≥ 678.9	14 (28.0)	9 (36.0)
HALP			0.224			0.706
< 44.17	38 (76.0)	22 (88.0)		17 (77.3)	19 (86.4)
≥ 44.17	12 (24.0)	3 (12.0)		5 (22.7)	3 (13.6)
PNI			0.123
< 45.2	21 (42.0)	6 (24.0)
≥ 45.2	29 (58.0)	19 (76.0)

Abbreviations: BMI, body mass index; CCRT, concurrent chemoradiotherapy; dNLR, neutrophil‐to‐lymphocyte ratio; ENE, extra‐nodal extension; HALP, hemoglobin‐albumin‐lymphocyte‐platelet score; HPV, human papillomavirus; LYM, lymphocyte; NCCRT, nimotuzumab and concurrent chemoradiotherapy; NLR, neutrophil‐to‐lymphocyte ratio; PLR, platelet‐to‐lymphocyte ratio; PNI, prognostic nutritional index; PSM, propensity‐score matching; SII, systemic immune‐inflammation index; WBC, white blood cell.

Inflammatory and nutritional biomarkers were evaluated in the subsequent analyses. The cut‐off values for each parameter were determined via ROC analysis, with all areas under the curve (AUC) values exceeding 0.5. In the CCRT group, the cut‐off values of NLR, PLR, dNLR, SII, HALP, PNI, WBC, LYM, and depth of invasion were 2.007 (AUC = 0.6225), 149 (AUC = 0.6475), 0.963 (AUC = 0.5925), 678.9 (AUC = 0.5750), 44.17 (AUC = 0.6875), 45.2 (AUC = 0.5925), 5.305 (AUC = 0.6300), 1.265 (AUC = 0.5738), and 0.95 (AUC = 0.5763), respectively (Figure 2). Similarly, the cutoff values for NLR, PLR, dNLR, SII, HALP, PNI, WBC count, LYM, and depth of invasion in the NCCRT group were the same as those in the CCRT group. However, the AUC values were 0.5351, 0.6228, 0.5044, 0.5965, 0.5877, 0.5526, 0.5265, and 0.5175, respectively (Figure 2).

FIGURE 2.

ROC curve of inflammation and nutritional indicators. A: CCRT group; B: NCCRT group. AUC, area under the curve; CCRT, concurrent chemoradiotherapy; dNLR, derived neutrophil‐to‐lymphocyte ratio; HALP, hemoglobin‐albumin‐lymphocyte‐platelet score; LYM, lymphocyte; NCCRT, nimotuzumab and concurrent chemoradiotherapy; NLR, neutrophil‐to‐lymphocyte ratio; PLR, platelet‐to‐lymphocyte ratio; PNI, prognostic nutritional index; ROC, receiver operating characteristic; SII, systemic immune‐inflammation index; WBC, white blood cell.

3.2. Survival outcomes and prognostic factors

The median follow‐up time was 29.80 months (95% CI: 25.83–33.77) in the CCRT group and 30.23 months (95% CI: 25.07–35.40) in the NCCRT group, respectively. The 2‐year EFS and OS rates were 74.1% vs 60.6% and 85.6% vs 83.6% in the CCRT and NCCRT groups, respectively (Figure 3A, B) In this study, the NCCRT group, which exclusively consisted of patients with ENE‐positive disease, demonstrated 2‐year survival rates that were numerically similar to those of the CCRT group, which comprised intermediate‐risk patients without ENE. However, due to the absence of an ENE‐positive cohort treated with CCRT alone, direct comparisons between the two groups are confounded by baseline risk differences, and these findings should be interpreted as exploratory and hypothesis‐generating.

FIGURE 3.

Event‐free survival (A) and overall survival (B) in the CCRT and NCCRT arms. CCRT, concurrent chemoradiotherapy; NCCRT, nimotuzumab and concurrent chemoradiotherapy.

In the propensity score‐matched cohort, the 2‐year EFS rates were 73.8% in the CCRT group and 58.3% in the NCCRT group (log‐rank p = 0.087), while the 2‐year OS rates were 86.2% and 82.5%, respectively (log‐rank p = 0.342). These results align with those observed in the unmatched cohort, suggesting that the potential survival benefit associated with nimotuzumab in the extremely high‐risk group was not influenced by baseline imbalances.

We further evaluated the association between clinicopathological characteristics and survival outcomes in patients with LA‐HNSCC. In the NCCRT group, the univariate analysis showed that patients with perineural invasion tended to have worse EFS, although the difference was not statistically significant (p = 0.086). Additionally, patients with a tumor infiltration depth of ≥ 0.95 cm also exhibited a trend toward poorer OS, but no significant statistical difference was observed (p = 0.068). In the CCRT group, an infiltration depth of ≥ 0.95 cm also showed a non‐significant trend toward worse OS (p = 0.093). Multivariate analysis did not identify a clear correlation between clinicopathological characteristics and prognosis (Figure 4). Given the limited sample size and exploratory nature of these analyses, these non‐significant trends should be interpreted with caution and are insufficient to establish definitive prognostic associations.

FIGURE 4.

Univariate analyses of EFS and OS in patients with locally advanced head and neck squamous cell carcinoma (A, EFS of NCCRT; B, OS of NCCRT; C, EFS of CCRT; D, OS of CCRT). CCRT, concurrent chemoradiotherapy; CI, confidence interval; dNLR, derived neutrophil‐to‐lymphocyte ratio; EFS, event‐free survival; HALP, hemoglobin‐albumin‐lymphocyte‐platelet score; HR, hazard ratio; LYM, lymphocyte; NCCRT, nimotuzumab and concurrent chemoradiotherapy; NLR, neutrophil‐to‐lymphocyte ratio; OS, overall survival; PLR, platelet‐to‐lymphocyte ratio; PNI, prognostic nutritional index; SII, systemic immune‐inflammation index; WBC, white blood cell.

Given the potential prognostic impact of HPV status in HNSCC, we conducted an exploratory subgroup analysis based on HPV status. In the overall cohort, HPV‐positive patients (n = 6) exhibited numerically better survival outcomes than HPV‐negative patients, although the small number of HPV‐positive cases precluded formal statistical comparisons. Due to the limited sample size and the predominance of HPV‐negative disease in both groups (90.0% in the CCRT group and 92.0% in the NCCRT group), further stratified analysis by HPV status was not feasible. This limitation is acknowledged, and future studies with larger HPV‐positive cohorts are needed to clarify the role of HPV status in this treatment setting.

Regarding inflammatory and nutritional biomarkers, in the NCCRT group, univariate analysis revealed that elevated dNLR showed a trend toward poorer EFS (p = 0.071), while patients with PLR ≥ 149 or LYM < 1.265 exhibited worse OS, without statistical significance (p = 0.081 and p = 0.081). In the CCRT group, patients with PLR ≥ 149 and HALP < 44.17 showed worse OS (p = 0.019 and p = 0.009). Other inflammatory and nutritional indicators were not significantly associated with survival outcomes (Figure 4). Similarly, the multivariate analysis revealed that inflammatory and nutritional biomarkers were not associated with patient prognosis.

In summary, our study investigated the association between clinicopathological features and inflammatory and nutritional biomarkers in relation to survival outcomes among patients with LA‐HNSCC. In the CCRT group, PLR and HALP levels had a more pronounced impact on OS. However, given the exploratory nature of these analyses and the limitations inherent to the study design, these findings require confirmation in independent, adequately powered cohorts.

3.3. Safety

All adverse events were graded according to the Radiation Therapy Oncology Group (RTOG) acute radiation injury criteria and Common Terminology Criteria for Adverse Events (CTCAE) v5.0 guidelines. The addition of nimotuzumab (NCCRT group) did not result in an increase in toxicity compared to standard CCRT. In the CCRT group, radiodermatitis graded 1–2 occurred in 28 (56.0%) patients, while grades 3–4 occurred in 4 (8.0%) patients. In the NCCRT group, radiodermatitis graded 1–2 occurred in 7 (28%) patients, with grades 3–4 occurring in 1 (4.0%) patient. Renal dysfunction was mild (grade 1–2) in 19 (38.0%) patients in the CCRT group and in 3 (12.0%) patients in the NCCRT group. Additionally, 29 (58.0%) patients experienced grade 1–2 leukopenia and 3 (6.0%) patients experienced grade 3–4 leukopenia in the CCRT group, while 20 (80.0%) and 2 (8.0%) patients experienced the same in the NCCRT group, respectively. Both regimens demonstrated manageable acute safety profiles. Notably, our analysis was limited to acute adverse events, and late toxicities, which are clinically relevant in postoperative HNSCC, were not systematically assessed. Therefore, the safety conclusions should be limited to the acute treatment phase. Adverse events related to treatment are shown in Table 2.

TABLE 2.

Treatment‐related adverse events

	N (%)
	CCRT (N = 50)		NCCRT (N = 25)
Adverse events	Grade 1–2	Grade 3–4	Grade 1–2	Grade 3–4
Leukocytopenia	29 (58.0)	3 (6.0)	20 (80.0)	2 (8.0)
Neutropenia	27 (54.0)	5 (10.0)	12 (48.0)	2 (8.0)
Anemia	26 (52.0)	0 (0.0)	16 (64.0)	0 (0.0)
Thrombocytopenia	15 (30.0)	0 (0.0)	6 (24.0)	0 (0.0)
Hypokalemia	20 (40.0)	0 (0.0)	9 (36.0)	0 (0.0)
Renal dysfunction	19 (38.0)	0 (0.0)	3 (12.0)	0 (0.0)
Oral mucositis	31 (62.0)	2 (4.0)	15 (60.0)	1 (4.0)
Radiodermatitis	28 (56.0)	4 (8.0)	7 (28.0)	1 (4.0)
Limited mouth opening	26 (52.0)	1 (2.0)	10 (40.0)	0 (0.0)
Infusion reaction	1 (2.0)	0 (0.0)	1 (4.0)	0 (0.0)
Rash	1 (2.0)	0 (0.0)	2 (12.0)	0 (0.0)

Abbreviations: CCRT, concurrent chemoradiotherapy; NCCRT, nimotuzumab and concurrent chemoradiotherapy.

4. DISCUSSION

Locally advanced HNSCC presents a significant therapeutic challenge, particularly in patients with high‐risk pathological features. ¹⁷ While clinical trials have established that PF (cisplatin + 5‐fluorouracil) and TPF (docetaxel + cisplatin + 5‐fluorouracil) regimens achieve substantial tumor regression in locally advanced oral squamous cell carcinoma, these neoadjuvant approaches unfortunately failed to demonstrate improvements in disease‐free survival (DFS) or OS. ¹⁸ , ¹⁹ This therapeutic limitation emphasizes the urgent need for new treatment strategies to enhance survival outcomes in patients with LA‐HNSCC. Recent advances in immunotherapy have shown promising results. Growing evidence suggests that immune checkpoint inhibitors (ICIs) can leverage the abundant tumor antigens and immune cell infiltration present in the preoperative tumor microenvironment to potentiate the efficacy of neoadjuvant therapy. The landmark KEYNOTE‐689 trial demonstrated that pembrolizumab combined with neoadjuvant and adjuvant therapies significantly improved EFS in patients with LA‐HNSCC. The pembrolizumab group achieved a 3‐year EFS of 57.6% compared to 46.4% in the control group (p = 0.008). Importantly, perioperative pembrolizumab administration did not compromise surgical feasibility and maintained a favorable safety profile. ²⁰ Adjuvant treatment paradigms for LA‐HNSCC have also been extensively investigated. The RTOG‐9501 study established CCRT as the standard treatment for patients with postoperative HNSCC with high‐risk features, including positive margins or ENE. Subsequent clinical trials have consistently demonstrated that concurrent adjuvant high‐dose cisplatin with radiotherapy achieves better DFS and OS outcomes than radiotherapy alone. ²¹ , ²² Nevertheless, the persistently high recurrence rates among patients with high‐risk LA‐HNSCC highlight the urgent need for more effective therapeutic approaches in this challenging patient population.

Previous studies have established that surgical margin status and ENE are independent prognostic factors that significantly influence DFS and OS in patients with HNSCC. ²³ For patients with LA‐HNSCC, combining high‐dose cisplatin with postoperative radiotherapy (RT) is significantly more effective than RT alone, without causing excessive late complications. Specifically, the 5‐year PFS rates were 47% and 36% in the combined treatment and RT groups, respectively (p = 0.04). Similarly, the 5‐year OS rates were 53% and 40%, respectively (p = 0.02). According to the Kaplan–Meier analysis, the 2‐year PFS and OS rates were less than 70% and 80%, respectively, in the combined treatment group. Furthermore, the cumulative incidence of local or regional recurrence was significantly reduced in the combined treatment group (p = 0.007). ²⁴ Meta‐analyses further confirmed that positive margins and ENE are key determinants of poor outcomes in LA‐HNSCC. Notably, adjuvant cisplatin‐based CCRT benefits only those with these specific high‐risk features who are chemotherapy‐eligible, whereas radiotherapy alone remains standard for other intermediate‐risk patients. ²⁵ In our study, intermediate‐risk patients receiving CCRT (radiotherapy plus cisplatin) exhibited 2‐year EFS and OS rates of 74.1% and 85.6%, respectively. As all enrolled patients had negative margins, those with ENE were stratified into the very high‐risk group and treated with nimotuzumab‐enhanced CCRT, achieving 2‐year EFS and OS rates of 60.6% and 83.6%, respectively. The Phase III NIVOPOSTOP trial demonstrated that adjuvant nivolumab plus CCRT significantly improved DFS (hazard ratio [HR] 0.76) in high‐risk patients, establishing a new benchmark. ²⁶ Conversely, the IHN01 trial, which investigated adjuvant nimotuzumab, did not meet its primary efficacy endpoints in the overall intention‐to‐treat population. ²⁷ However, discrepancies between clinical trials and real‐world practice often arise from patient selection. Our study focused exclusively on an ENE‐positive cohort for nimotuzumab intensification. ENE implies a biologically aggressive phenotype with high metastatic potential. The “negative” result of IHN01 may reflect the dilution of effect in a broad high‐risk population that included patients without ENE. Our data are consistent with the hypothesis that for the highest‐risk subset, EGFR blockade may offer meaningful disease control by enhancing radiosensitivity in hypoxic bulky nodal disease remnants. However, due to the absence of an ENE‐positive control group receiving CCRT alone, these observations remain hypothesis‐generating and do not establish causal treatment benefits. Furthermore, the safety profile of nimotuzumab, confirmed in the IHN01 trial to be comparable to that of a placebo, contrasts favorably with the immune‐related adverse events (irAEs) of checkpoint inhibitors, making it a viable option for patients ineligible for immunotherapy. Evidence suggests that for patients with oral cancer with isolated adverse pathological features, neither CCRT nor accelerated radiotherapy provides additional benefits over standard radiotherapy. However, patients with multiple high‐risk factors (e.g., T3–T4 disease, N2–N3 nodal involvement, and ENE) may derive PFS and OS improvements from intensified therapy. ²⁸ Contrasting findings emerged from a study on intermediate‐risk HNSCC (defined as ≥ 1 postoperative risk factor), in which radiotherapy combined with cetuximab showed 2‐year OS from the Kaplan–Meier curve as approximately 85.0%. ²⁹ In a PSM study, patients with intermediate‐risk LA‐HNSCC did not benefit from postoperative CCRT compared to RT alone, although postoperative CCRT did not improve OS and EFS in patients with multiple moderate risk factors who had significantly poorer survival. ³⁰ However, in another PSM study, pN2–3 and pT3–4 were independent adverse prognostic factors for intermediate‐risk HNSCC patients without margin‐negative or ENE, and adjuvant CCRT improved survival in intermediate‐risk patients with pN2–3 or pT3–4. ³¹ Therefore, for patients with intermediate‐risk LA‐HNSCC, CCRT seems to improve the prognosis of patients, but further stratification of risk factors is needed to better identify the population that can benefit from CCRT. However, in the extremely high‐risk LA‐HNSCC patient population, the evidence for additional targeted therapy remains inconclusive A trial evaluating afatinib maintenance therapy post‐CCRT in resected HNSCC patients with macroscopically complete resection reported no DFS benefit (2‐year DFS: 61% vs. 64%; HR 1.12, 95% CI 0.70–1.80), discouraging its use in this setting. ³² Conversely, a study in high‐risk, resected HPV‐negative HNSCC (margin < 1 mm, ENE, perineural/vascular invasion, or ≥ 2 positive lymph nodes) demonstrated that panitumumab‐enhanced CCRT improved outcomes, with 2‐year PFS and OS rates of 70% and 72%, respectively, alongside favorable tolerability, and showed improved clinical outcomes in high‐risk, resected, HPV‐negative HNSCC patients. ³³ Another trial in stage II–IV HNSCC (margin ≤ 5 mm and/or ENE) compared CCRT plus lapatinib maintenance versus placebo, finding no therapeutic advantage but increased toxicity with lapatinib. ³⁴ However, in a separate analysis of pathological stage III–IV HNSCC (positive margins, ENE, or ≥ 2 metastatic lymph nodes), patients receiving cisplatin‐ or docetaxel‐based CCRT with cetuximab achieved 2‐year DFS rates of 57% and 66%, and 2‐year OS rates of 69% and 79%, respectively. ¹⁰ Thus, combining anti‐EGFR antibodies with CCRT may benefit extremely high‐risk LA‐HNSCC, though optimal combination partners and patient selection require further study. In conclusion, our real‐world study suggests that cisplatin‐based CCRT improves the prognosis of intermediate‐risk patients with LA‐HNSCC, whereas nimotuzumab combined regimens may benefit very high‐risk patients with ENE‐positive. Notably, in the very high‐risk group, adding nimotuzumab to CCRT yielded survival outcomes comparable to those of the moderate‐risk group receiving CCRT alone. However, these findings remain exploratory and require prospective validation to refine risk‐stratified personalized treatment for LA‐HNSCC.

Many clinical and pathological factors affect the prognosis of patients with HNSCC, such as tumor size, depth of invasion, lymphatic and perineural infiltration, and marginal status. ³⁵ Recently, many studies have also found that tertiary lymphoid structure is one of the prognostic factors for HNSCC. ³⁶ Perineural invasion and tumor infiltration depth ≥ 0.95 cm showed non‐significant trends toward poorer prognosis (p > 0.05), possibly due to limited sample size. We also analyzed the relationship between peripheral blood inflammatory factor indicators, nutritional status indicators, and the prognosis of patients with LA‐HNSCC. Elevated dNLR was associated with a trend toward poorer EFS, and elevated PLR or reduced LYM was associated with worse OS in the NCCRT group, although these trends were not statistically significant (p > 0.05). However, PLR and HALP levels were significantly associated with OS in the CCRT group (p = 0.019 and p = 0.009, respectively). These biomarkers are derived from routine blood tests and offer a cost‐effective and accessible means of refining risk stratification in patients with intermediate‐risk LA‐HNSCC. Their integration into postoperative assessments could help identify patients who may benefit from closer surveillance or treatment modifications; however, prospective validation is needed. Previous studies suggested that a high dNLR was an independent prognostic factor in patients with esophageal squamous cell carcinoma, and that patients with a high dNLR had a worse prognosis. ³⁷ Additionally, in an investigation involving patients with oral cancer, it was revealed that elevated dNLR levels were correlated with OS. ³⁸ Therefore, high dNLR may be associated with the survival of patients with LA‐HNSCC and can well predict patient prognosis. However, the clinical value of HALP in patients with HNSCC remains unclear. The study showed that low HALP predicted poor OS and DFS (p < 0.001) and was significantly associated with oral cavity squamous cell carcinoma. ³⁹ Further research corroborated that diminished HALP was linked to extended DFS among patients suffering from locally advanced esophageal squamous cell carcinoma undergoing neoadjuvant chemoimmunotherapy. ⁴⁰ Additionally, a meta‐analysis found that oral squamous cell carcinoma patients with elevated PLR had worse OS and DFS. ⁴¹ Another study also demonstrated that patients with HNSCC with higher PLR had a poorer prognosis. ⁴² The HALP score captures key tumor‐host interactions. Low hemoglobin reflects hypoxia, which promotes radioresistance and metastasis via HIF‐1α. Hypoalbuminemia indicates poor nutrition, which can potentially reduce tolerance to chemotherapy and immune function. Lymphocytes are crucial for anti‐tumor immunity, while elevated platelets may shield tumor cells from immune attack and secrete immunosuppressive factors like TGF‐β. Thus, a low HALP score may indicate the convergence of hypoxia, malnutrition, and immune evasion. Similarly, an elevated PLR signals a shift toward a pro‐tumor inflammatory milieu. These readily available biomarkers enable refined risk stratification beyond conventional TNM staging. ⁴³ In summary, our study and previous research suggest that perineural invasion, tumor infiltration depth, elevated dNLR, and elevated PLR are associated with poorer prognosis in patients with LA‐HNSCC. Furthermore, HALP and PLR may serve as candidate prognostic biomarkers to guide postoperative adjuvant chemoradiotherapy in intermediate‐risk patients with LA‐HNSCC. However, prospective validation is required.

Regarding safety, our study demonstrated that incorporating nimotuzumab into CCRT did not exacerbate serious adverse reactions such as radiation dermatitis or hematological toxicity in patients with high‐risk LA‐HNSCC. A study evaluating the efficacy of nimotuzumab combined with intensity‐modulated radiotherapy (IMRT) in elderly patients with locally advanced nasopharyngeal carcinoma showed that 10 patients (33%) developed grade 3 mucositis, and 3 patients (10%) developed grade 3 pneumonia. ⁴⁴ Another study also verified that NCCRT in the treatment of LA‐HNSCC did not show significant adverse reactions. ⁴⁵ Thus, NCCRT was associated with a manageable acute safety profile in this patient population. However, the absence of a systematic late toxicity assessment represents a notable gap, and safety conclusions should only be interpreted within the context of acute treatment.

Our clinical investigation had several limitations. First, this was a single‐center retrospective study with a relatively small sample size, which may limit the generalizability and introduce selection bias. Although PSM was used, residual confounding from the unmeasured variables could not be excluded. Moreover, ROC‐derived cut‐offs were generated and evaluated within the same cohort, introducing an optimism bias, which requires external validation. Second, the follow‐up duration was insufficient to assess long‐term outcomes. Third, the lack of a direct control group for ENE+ patients (CCRT alone) is inherent to real‐world data. Thus, the outcomes in the NCCRT group cannot be directly attributed to nimotuzumab, and the findings should be considered exploratory and hypothesis‐generating. Fourth, the statistical power was limited by the modest sample size, particularly in the NCCRT group (n = 25), affecting subgroup analyses and the detection of modest but clinically relevant differences. Non‐significant trends should not be overinterpreted, and negative findings should not exclude meaningful associations. Fifth, safety assessments were confined to acute toxicities, and late adverse events were not systematically assessed. Sixth, the HPV status was predominantly negative, with only six HPV‐positive cases precluding meaningful subgroup analysis, and the findings are primarily applicable to HPV‐negative disease. Seventh, several important variables (smoking, alcohol consumption, and surgical approach) were not included in the multivariate analysis due to incomplete documentation, potentially introducing residual confounding factors. To address these limitations, we plan to conduct prospective randomized trials to validate the clinical value of nimotuzumab in high‐risk patients and CCRT in intermediate‐risk LA‐HNSCC.

5. CONCLUSION

In this real‐world study, risk‐stratified adjuvant CCRT, with or without nimotuzumab, was associated with promising outcomes and acceptable acute safety in patients with high‐risk postoperative LA‐HNSCC. The addition of nimotuzumab to CCRT was associated with improved survival outcomes in ENE‐positive patients, comparable to those observed in intermediate‐risk patients who received CCRT alone. PLR and HALP have emerged as candidate prognostic biomarkers for intermediate‐risk patients and warrant further validation. These findings are hypothesis‐generating and require confirmation in prospective randomized studies with dedicated control groups and systematic long‐term follow‐ups.

AUTHOR CONTRIBUTIONS

Gangling Tong conceived the study. Gangling Tong, Yuejia Zhou, and Linting Zhang drafted the manuscript. Wenjuan Lai, Jingyun Tan, and Jing Gao are local investigators at the participating centers. Yajie Liu, Shubin Wang, and Feng Peng are responsible for data collection. The study is supervised and coordinated by Gangling Tong, Yuejia Zhou, and Linting Zhang.

CONFLICT OF INTEREST STATEMENT

The authors declare that this study was conducted in the absence of any commercial or financial relationships that could be construed as potential conflicts of interest.

ETHICS APPROVAL

This study was approved by the Ethics Committee of the Peking University Shenzhen Hospital (2025‐018). Due to the retrospective nature of this study, the requirement for individual informed consent was waived by the Institutional Review Committee.

GENERATIVE AI STATEMENT

The authors declare that no Generative AI was used in the creation of this manuscript.

DATA AVAILABILITY STATEMENT

The datasets used and/or analyzed in the current study are available from the corresponding author upon reasonable request.