TCIRG1⁺neutrophils: biomarkers of adverse prognosis in tongue squamous cell carcinoma

Abstract

Background

Tongue squamous cell carcinoma (TSCC) is associated with an unfavorable prognosis, and the involvement of systemic inflammation and tumor-associated neutrophils (TANs) remains inadequately elucidated. This study was designed to assess the prognostic significance of inflammatory biomarkers and to delineate functionally distinct TAN subsets in TSCC.

Methods

This retrospective investigation enrolled 851 patients with TSCC, stratified into discovery (n = 596) and validation (n = 255) cohorts. Pre-treatment hematological parameters were utilized to compute systemic inflammatory indices, including neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), platelet-to-lymphocyte ratio (PLR), and systemic immune-inflammation index (SII). Neutrophil infiltration was quantitatively evaluated through hematoxylin and eosin (H&E) staining and multiplex immunohistochemistry (mIHC) in a representative subset of 120 cases. Proteomic profiling facilitated the identification of neutrophil-associated genes, whose prognostic relevance was subsequently corroborated through analysis of The Cancer Genome Atlas (TCGA) dataset and mIHC validation.

Result

Elevated NLR and MLR were associated with diminished overall and disease-free survival, with NLR demonstrating superior predictive capability. Increased intratumoral neutrophil density was significantly correlated with unfavorable clinical outcomes. Proteomic profiling identified T cell immunoregulatory protein 1 (TCIRG1) as a pivotal gene associated with neutrophil activity. mIHC analysis demonstrated that the density of TCIRG1⁺TANs served as a more robust prognostic indicator of poor survival compared to the overall TANs density.

Conclusion

NLR is robust prognostic biomarker in TSCC. The TCIRG1⁺TAN subset emerges as a superior predictor of adverse outcomes, highlighting its potential as a therapeutic target.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-026-07761-5.

Introduction

Inflammation is a hallmark feature of cancer, exerting a profound influence on the initiation and progression of malignant diseases [1]. Within the spectrum of cancer, there is a burgeoning body of evidence highlighting the role of systemic inflammatory responses in dictating patient survival outcomes. These responses are closely linked to alterations in peripheral blood leukocyte profiles, as indicated by the NLR, MLR, PLR, and SII [2–4]. These biomarkers are pivotal for assessing the systemic inflammatory milieu and for prognostication in a spectrum of cancers [5]. Nevertheless, the intrinsic mechanisms that confer prognostic significance to these inflammatory biomarkers in cancer patients are not yet fully elucidated. Oral squamous cell carcinoma (OSCC) represents the most frequent malignancy in the head and neck region, arising from a variety of anatomical sites [6]. A particularly common subtype is TSCC [7]. TSCC poses significant survival challenges and roughly 60% of patients are diagnosed at an advanced local stage during their first clinical presentation, with a 5-year overall survival rate that stands at a mere 50% [6, 8]. This reality underscores an exigent need for the identification of novel therapeutic targets.

Over the past decade, the role and significance of neutrophils in cancer have become increasingly apparent [9]. A wealth of research suggests that neutrophils can contribute to tumor progression through multiple mechanisms, including the promotion of angiogenesis, enhancement of tumor cell mobility, migration, and invasion, as well as modulation of other immune cells, thereby facilitating the ‘immunosuppressive switch’ [10, 11]. Nonetheless, neutrophils are characterized by a short lifespan, with a circulating half-life of approximately 7–10 h in humans [12]. The transient nature of neutrophils, coupled with the technical challenges associated with their study, has meant that much of our understanding of neutrophils in cancer is derived from the analysis of peripheral blood neutrophils [12]. Consequently, the phenotypes and potential functions of TANs in the context of cancer patients remain inadequately explored [12, 13].

In this study, we reveal a significant association between systemic inflammation and patient prognosis in TSCC. Our findings demonstrate that an elevated NLR and heightened neutrophils infiltration within tumors are correlated with diminished survival outcomes. The data additionally highlight the activation of multiple neutrophils signaling pathways in TSCC patients exhibiting a heightened risk of mortality. Notably, TCIRG1⁺neutrophils emerge as potential prognostic indicators of adverse outcomes. These insights underscore the inhibitory influence of neutrophil-associated inflammation on patient prognosis, presenting novel clinical biomarkers and therapeutic targets for TSCC management.

Methods

Patient cohort characteristics

This investigation enrolled a total of 851 patients with histologically confirmed TSCC, comprising a discovery cohort of 596 cases and a validation cohort of 255 cases, who were treated at our tertiary care center between January 2010 and December 2020. All participants were recruited from our single-center institution and met strict inclusion criteria: (i) aged 18 years or older, with primary TSCC confirmed by two experienced pathologists; (ii) clinical stages I-IV, classified according to the 8th Edition of the American Joint Committee on Cancer (AJCC) guidelines; (iii) an Eastern Cooperative Oncology Group performance status ranging from 0 to 1; (iv) no history of prior anti-cancer therapy; and (v) tumors deemed locally resectable by a head and neck surgical oncologist. Exclusion criteria included other pathological types, absence of surgery, incomplete pre-treatment data, or loss to follow-up. The study protocol was approved by the ethics committee (SYSKY-2024-694−01), and the requirement for informed consent was waived in light of the study’s retrospective design.

Clinical characteristics and laboratory testing

A retrospective analysis was performed utilizing comprehensive clinicopathological data extracted from medical records, including patient age, TNM classification, histological grade, and routine inflammatory laboratory parameters (neutrophil, lymphocyte, monocyte, and platelet counts) assessed within one week preceding surgical intervention. From these parameters, we calculated established inflammatory indices: NLR, MLR, PLR, SII, defined as the product of platelet count and NLR. DFS was defined as the time interval from surgical intervention to either disease recurrence or mortality from any cause. OS was calculated from the date of surgery to the date of death from any cause. For patients without disease progression or mortality by the January 2024 data cut-off, the last follow-up date was designated as the censoring point.

H&E staining

For pathological assessment, tissue specimens obtained from 120 TSCC patients meeting the predefined inclusion criteria were processed using hematoxylin and eosin (H&E) staining and subsequently evaluated independently by two board-certified pathologists under blinded conditions, with no access to clinical data or each other’s interpretations. The H&E-stained TSCC sections were systematically analyzed for neutrophil and monocyte infiltration using standardized quantification protocols. Neutrophil density was determined by enumerating cells at 400× magnification and classified according to a four-tier scoring system: 0 (absent), 1 (1–19 cells), 2 (20–100 cells), and 3 (> 100 cells). Monocyte quantification followed an identical methodological approach, with cell counts categorized using the same scoring criteria: 0 (absent), 1 (1–19 cells), 2 (20–100 cells), and 3 (> 100 cells). This standardized scoring system ensured consistent and reproducible quantification of immune cell infiltration in the tumor microenvironment.

Proteomic analysis

Protein extraction and tryptic digestion

Twelve fresh tongue cancer tissue samples were collected from 6 patients with poor survival outcomes and 6 patients with favorable survival outcomes for proteomic analysis. Patients were categorized into the poor survival group based on mortality, while those who remained alive were classified into the favorable survival group. The samples were initially washed with phosphate-buffered saline (PBS) to remove blood and cellular debris. Subsequently, the tissues were homogenized in lysis buffer (8 M urea, pH 8.0) supplemented with protease inhibitors (Roche) and phosphatase inhibitors, followed by incubation on ice for 20 min. The lysates were sonicated on ice for 2 min (3 s on, 3 s off) and centrifuged at 21,000 × g for 10 min to pellet tissue debris. The supernatants were collected, and protein concentrations were quantified using the bicinchoninic acid (BCA) assay. Proteins were reduced with 2 mM dithiothreitol at 56 °C for 30 min and alkylated with 10 mM iodoacetamide at room temperature for 30 min. Digestion was performed using trypsin at an enzyme-to-substrate ratio of 1:50 at 37 °C for 16 h. The resulting peptides were desalted using SepPak C18 cartridges (Waters, MA) and dried under vacuum using a SpeedVac concentrator.

Nano-LC-MS/MS

Peptide samples were analyzed using an EASY-nLC 1200 LC system (Thermo Fisher Scientific) coupled to an Orbitrap Fusion mass spectrometry (Thermo Fisher Scientific). Peptides were reconstituted in mobile phase A (0.1% formic acid) and directly loaded onto a 75 μm × 150 mm capillary column packed with 1.9 m particle size C18 AQ stationary phase, using an 8–38% ACN gradient containing 0.1% FA over 120 min at a flow rate of 300 nL/min. An MS1 scan was performed over a range of 350–1800 m/z (at 60000 resolution, 100% AGC), followed by MS/MS data-dependent acquisition using HCD and detection in the Iontrap (with a 35 ms injection time and 100% AGC).

Processing of proteomic data and published dataset

Data structure tidying was performed using the tidyr package. Proteins with missing values in < 50% of samples were imputed with the KNN algorithm (k = 10) from the impute package. Differential gene expression analysis was conducted using the limma package. All differentially expressed genes (DEGs) were identified based on normalized expression values, applying a logarithmic fold change (LogFC) threshold of 0.584. We sourced the neutrophil dataset from Wang et al. (PMC10086491) and obtained the neutrophil gene set labeled GSEA3982. By inftegrating data from three distinct cohorts, we aimed to explore the genes significantly associated with neutrophil function.

mIHC

In accordance with the manufacturer’s guidelines, we conducted mIHC staining on formalin-fixed, paraffin-embedded (FFPE) tumor specimens from 120 patients with TSCC. Utilizing the Opal 7-Color Kit (Catalog Number 0004100100, Panovue, Beijing, China), we applied the following primary antibodies: TCIRG1 at a dilution of 1:100 (Catalog Number 12649-1-AP, Proteintech) and Myeloperoxidase (MPO) at a dilution of 1:1000 (Catalog Number 22225-1-AP, Proteintech). Each staining round included the application of a secondary horseradish peroxidase-conjugated antibody from Panovue (Beijing, China), followed by incubation with the respective tyramide-coupled fluorophores, Opal 650 and Opal 520. The mIHC-stained images were captured using the Akoya Vectra Polaris imaging system (Akoya Biosciences, Marlborough, MA, USA). Subsequently, the InForm software (Akoya Biosciences, Marlborough, MA, USA) was employed to quantify the density (number of cells per mm2) of MPO-positive and TCIRG1⁺/MPO⁺ dual-positive cells across the entire tumor area of each sample.

Statistical analysis

Categorical data were presented as proportions and analyzed using Chi-square tests, supplemented by Yates’ correction where necessary. Continuous variables were summarized with medians and interquartile ranges (IQRs) and compared using Student’s t-tests or Mann-Whitney U tests, depending on the data distribution. The median follow-up duration was ascertained via the reverse Kaplan-Meier technique. To identify the optimal thresholds for blood test and mIHC outcomes, Receiver Operating Characteristic (ROC) curves were utilized, applying the Youden index for determination. These values were then dichotomized for subsequent analysis. Univariate and multivariate Cox regression analyses were conducted to uncover the correlation of NLR, MLR, PLR, and SII with overall and disease-free survival. Survival outcomes were evaluated using both the Cox proportional hazards model and the Kaplan-Meier method. In the context of regression analysis, multi-category independent variables were reformatted using dummy coding. All reported P-values were two-tailed, with statistical significance defined as a value of 0.05 or less. The statistical analyses were executed using R software for Windows, version 4.2.1.

Results

MLR and NLR as key independent prognostic indicators for TSCC patients

To determine the most reliable biomarkers for prognostic assessment in TSCC patients, a systematic evaluation of hematological parameters and histopathological specimens was performed. Tumor tissue samples from the cohort were subjected to comprehensive proteomic profiling, with the resultant data subsequently validated through mIHC analysis, as detailed in the methodological workflow diagram (Fig. 1A). This study encompassed a cohort of 1,138 patients diagnosed with TSCC. The study population was rigorously selected through the application of predefined inclusion and exclusion criteria. Specifically, patients were excluded based on the following parameters: prior history of anti-cancer therapy (n = 84), incomplete clinical data (n = 15), non-TSCC pathological subtypes (n = 112), and loss to follow-up (n = 76). Following this systematic exclusion process, the final analytical cohort comprised 851 TSCC patients with comprehensive clinicopathological profiles and randomized into discovery (n = 596) and validation (n = 255) cohorts (Fig. 1B). As presented in Table 1, the proportion of individuals aged over 60 years was 34.4% in the discovery cohort and 32.2% in the validation cohort. Within the discovery cohort, 305 patients (51.2%) were classified at clinical stages III + IV, 194 (32.5%) exhibited T3 + T4 tumor classification, 244 (40.9%) demonstrated lymph node metastasis, and 270 (45.3%) presented with well-differentiated tumors. Similarly, in the validation cohort, 124 patients (48.6%) were at clinical stages III + IV, 64 (25.1%) had T3 + T4 classification, 105 (41.2%) showed lymph node metastasis, and 125 (49%) displayed well-differentiated tumors. The median follow-up duration in the discovery cohort was 55.3 months (95% CI: 39.1–107.7 months), with 31.5% (188/596) of patients experiencing disease recurrence and 35.6% (212/596) succumbing to the disease. In the validation cohort, the median follow-up was 55.4 months (95% CI: 45.7–67.5 months), with 20% (51/255) experiencing recurrence and 32.9% (84/255) dying from the disease.

Fig. 1.

MLR and NLR as prognostic factors in TSCC patients. A Schematic representation of the experimental design and analytical approach. We integrated clinicopathological data and laboratory parameters of TSCC patients with proteomic sequencing and mIHC staining of tumor samples to identify TCIRG1+TANs linked to prognosis. B Patient Selection Flowchart. This study initially enrolled a cohort of 1138 TSCC patients. However, 287 patients were subsequently excluded due to incomplete clinical data. The remaining 851 patients with comprehensive clinical profiles were randomly allocated into a discovery group and a validation group in a 2:1 ratio. C The Venn diagram analysis revealed that both the MLR and NLR are significantly correlated with OS in TSCC. D The Venn diagram analysis revealed that both the MLR and NLR are significantly correlated with DFS in TSCC

Table 1.

Baseline clinical characteristics of TSCC patients in discovery cohort and validation cohort

Characteristics	Discovery cohort	Validation cohort
Age (year), n (%)
≤ 60	391(65.6%)	173(67.8%)
> 60	205(34.4%)	82(32.2%)
Tumor differentiation, n (%)
Well	270 (45.3%)	125 (49%)
Moderate	255 (42.8%)	106 (41.6%)
Poor	71 (11.9%)	24 (9.4%)
T classification, n (%)
T1 + T2	402 (67.5%)	191 (74.9%)
T3 + T4	194 (32.5%)	64 (25.1%)
N classification, n (%)
N0	352 (59.1%)	150 (58.8%)
N+	244 (40.9%)	105(41.2%)
AJCC8, n (%)
Ⅰ+Ⅱ	291(48.8%)	131(51.4%)
Ⅲ+Ⅳ	305(51.2%)	124 (48.6%)
Recurrence, n (%)12
No	408 (68.5%)	204 (80%)
Yes	188 (31.5%)	51 (20%)
Death, n (%)
No	384 (64.4%)	171 (67.1%)
Yes	212 (35.6%)	84 (32.9%)
Follow-up Time, (month)a	55.3 (39.1, 107.7)	55.4 (45.7, 67.5)

^amedian (P25, P75)

A comprehensive analysis was performed to investigate the association between inflammatory biomarkers and clinical outcomes in TSCC patients. Optimal cutoff values for hematological parameters were established using the Youden index and subsequently dichotomized, as presented in Tables S1 and S2. The predictive power of the NLR, MLR, PLR and SII was assessed in both the discovery and validation cohorts using ROC curves, as depicted in Figure S1. The analysis revealed that the NLR index possesses the highest predictive value. Furthermore, both univariate and multivariate Cox regression analyses underscored the correlation of NLR, MLR, PLR, and SII with OS in the discovery cohort, as detailed in Table S3. In the validation cohort, the association with OS was significant only for NLR and MLR, as shown in Table S4. Concurrently, NLR and MLR were found to be significantly linked to DFS, as illustrated in Supplementary Table S5 and Table S6. These findings underscore the prognostic significance of MLR and NLR, highlighting their substantial predictive utility for TSCC patient outcomes (Fig. 1C-D).

Elevated NLR and MLR are significantly correlated with adverse outcomes and clinicopathological features in patients with TSCC

The prognostic significance of the NLR and MLR was assessed through Kaplan-Meier survival analysis. The OS curves revealed statistically significant disparities between subgroups stratified by MLR levels, with the high MLR cohort demonstrating markedly inferior clinical outcomes compared to the low MLR group (Fig. 2A and Figure S2A). In alignment with our hypothesis, patients exhibiting elevated MLR levels demonstrated significantly reduced DFS compared to those with lower MLR levels (Fig. 2B and Figure S2B). Additionally, an increased NLR was significantly associated with worse OS (Fig. 2C and Figure S2C) and DFS (Fig. 2D and Figure S2C) across both the discovery and validation cohorts. These findings collectively establish NLR and MLR as independent prognostic biomarkers for patients with TSCC.

Fig. 2.

Association of prognostic indictors with patient outcomes and clinicopathological parameters in TSCC within discovery cohorts. A Kaplan–Meier analysis of OS in patients with high and low MLR levels. B Kaplan–Meier analysis of DFS in patients with high and low MLR levels. C Kaplan–Meier analysis of OS in patients with high and low NLR levels. D Kaplan–Meier analysis of DFS in patients with high and low NLR levels. E Analysis of correlations between MLR and clinicopathological parameters including clinical T category, N category, disease stage, and histological differentiation. F Analysis of correlations between NLR and clinicopathological parameters including clinical T category, N category, disease stage, and histological differentiation. ***, P<0.001, ****, P<0.0001, ns, non-significant

Through comprehensive analysis of the association between clinicopathological features and inflammatory biomarkers, we identified that elevated MLR levels were significantly associated with advanced tumor classification (T3-T4), lymph node metastasis, and advanced clinical stages (III-IV) in the discovery cohort (p < 0.05), though no statistically significant correlation was observed between MLR and tumor differentiation grade (Fig. 2E). Similarly, increased NLR levels demonstrated significant associations with T3-T4 tumor classification, lymph node metastasis, advanced clinical stages (III-IV), and poor tumor differentiation (Fig. 2F). These findings were subsequently validated in an independent cohort, where both elevated NLR and MLR levels consistently correlated with more advanced clinicopathological parameters (Figure S2). These results collectively indicate that NLR and MLR serve as robust prognostic indicators for clinical outcomes in patients with TSCC.

Intratumoral neutrophils as potent prognostic indicators for TSCC patients

Our analysis has identified NLR and MLR as potential biomarkers for TSCC patients, suggesting that neutrophils and monocytes may exert a pivotal influence on patient outcomes [15, 16]. To validate these findings, we analyzed H&E-stained tissue sections from 120 TSCC patients, establishing a quantitative scoring system to evaluate neutrophil and monocyte infiltration densities, as detailed in Table 2. The infiltration scores were subsequently correlated with corresponding peripheral blood NLR and MLR values. Statistical analysis revealed that patients with elevated NLR exhibited significantly higher neutrophil infiltration scores compared to the low NLR cohort (p < 0.05). In contrast, MLR values demonstrated an inverse relationship with monocyte infiltration, though this association did not reach statistical significance. These results suggest a potential association between peripheral blood NLR levels and intratumoral neutrophil infiltration (Figure S3). For analytical purposes, we stratified patients into high (scores 2–3) and low (scores 0–1) neutrophil and monocyte infiltration groups. The prognostic significance of these immune cell infiltration patterns was evaluated through comprehensive clinicopathological correlation analysis, with individual patient outcomes visualized in a waterfall plot Fig. 3A. Patients were categorized into favorable (alive) and poor (deceased) prognostic groups based on survival status. Subsequent survival analysis revealed a significant correlation between neutrophil infiltration levels and TSCC patient outcomes, as quantified in Table S7. Kaplan-Meier survival curves demonstrated superior DFS and OS in patients with low neutrophil infiltration compared to those with high infiltration (Fig. 3B). In contrast to neutrophils, and despite the prognostic value of the systemic MLR, the density of intratumoral monocytes, as assessed by H&E staining, showed no significant association with DFS or OS in our TSCC cohort (Figure S4). This suggests that the prognostic signal from the MLR may be driven more by its lymphocyte component or systemic monocyte function rather than local monocyte infiltration in the tongue TME.Additionally, we showcase the expression of neutrophils and tumor recurrence status in two representative TSCC cases—Case 1 with high neutrophil density and Case 2 with low neutrophil density (Fig. 3C). These observations reinforce the notion that the intratumoral presence of neutrophils is significantly associated with adverse outcomes in TSCC patients.

Table 2.

Baseline clinical characteristics of 120 patients with pathological scores

Characteristics	N = 120
Monocytes(Infiltration score)*, n (%)
0	38 (31.7%)
1	52 (43.3%)
2	26 (21.7%)
3	4 (3.3%)
Neutrophils(Infiltration score)*, n (%)
0	33 (27.5%)
1	30 (25%)
2	29 (24.2%)
3	28 (23.3%)
Recurrence, n (%)
No	74 (61.7%)
Yes	46 (38.3%)
Death, n (%)
No	54 (45%)
Yes	66 (55%)
Follow-up Time, (month)a	51.2 (42.6, 62.2)

^amedian (P25, P75)

^*Histological scoring of microscopic cell infiltration in pathological specimens was categorized as: 0 (absent), 1 (1–19 cells), 2 (20–100 cells), and 3 (> 100 cells)

Fig. 3.

Intratumoral neutrophils as prognostic indicators in TSCC patients. A The waterfall plot illustrates the association between intratumoral neutrophil density and clinicopathological characteristics. B Kaplan–Meier analysis comparing OS among patients with high versus low levels of intratumoral neutrophil expression. C Kaplan–Meier analysis comparing DFS among patients with high versus low levels of intratumoral neutrophil expression. D Surgical therapeutic responses in two representative TSCC patients with high or low intratumoral neutrophils expression

Elevated TCIRG1 expression foreshadows adverse outcomes in TSCC patients

To elucidate the functional significance of neutrophils in the tumor microenvironment, we systematically analyzed the biological profiles of patients with TSCC through Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses. Proteomic profiling was conducted on tumor specimens obtained from 12 advanced TSCC patients, categorized into distinct cohorts based on survival outcomes—favorable versus unfavorable. Our results demonstrated a significant upregulation of neutrophil-related biological processes, including “neutrophil migration” and “neutrophil-mediated immunity,” in patients with poor prognostic outcomes (Fig. 4A). These findings underscore the pivotal role of neutrophils in influencing TSCC progression and highlight their potential as prognostic biomarkers, particularly in cases associated with adverse clinical outcomes.

Fig. 4.

Correlation of TCIRG1 expression with patient outcomes in TSCC. A GO and KEGG analyses of proteomically derived gene sets from TSCC patients. B Venn diagram analysis identified seven pivotal genes related to neutrophil function. C The volcano plot depicted the expression levels of the seven key genes. D The heatmap illustrated the expression profiles of these seven genes across a cohort of individual TSCC patient samples. E Kaplan–Meier analysis disclosed a correlation between the upregulation of key genes and overall survival in OSCC patients of TCGA databases

To clarify the molecular mechanisms underlying neutrophil function in tumor biology, we systematically identified a critical neutrophil-associated gene signature by integrating data from three independent cohorts. Through comparative analysis, seven consensus genes were identified at the intersection of these datasets, as depicted in the Venn diagram (Fig. 4B). Subsequent volcano plot analysis revealed a distinct differential gene expression profile, with six genes demonstrating significant upregulation and one gene showing downregulation in patients with poor prognosis compared to those with favorable outcomes, and TCIRG1 had the highest expression variance (Fig. 4C and Table S8). A corresponding heatmap further delineated the expression profiles of these genes across individual TSCC patient samples (Fig. 4D). To validate the clinical significance of these genes, we conducted a comprehensive analysis of the correlation between their expression profiles and overall survival in TSCC patients, utilizing data extracted from TCGA database. The optimal cutoff values for gene expression were determined using the Youden index, and the expression levels were subsequently dichotomized into binary variables for further analysis. Notably, TCIRG1 exhibited a statistically significant association with patient prognosis, as illustrated in Fig. 4E and Figure S5. These findings collectively suggest that TCIRG1, which is implicated in neutrophil activity, is positively associated with adverse survival outcomes in TSCC patients.

TCIRG1⁺ neutrophils are associated with poor prognosis in TSCC patients

Our findings indicate that intratumoral neutrophils, also classified as TANs, are prevalent in TSCC patients exhibiting poor clinical outcomes. Nonetheless, the prognostic significance of TAN infiltration remains a contentious issue, largely attributed to their functional heterogeneity [17]. This includes the antitumor phenotype (termed N1-TANs) and the tumor-promoting phenotype (designated N2-TANs) [12]. To investigate the impact of specific TAN subsets on TSCC patient outcomes, we conducted mIHC staining on tumor samples from 120 TSCC patients, who were retrospectively classified into favorable (alive, n = 54) and unfavorable (deceased, n = 66) prognostic groups based on their survival status.

TANs were identified using the MPO marker, which has been extensively associated with TANs across various cancers [18, 19]. The cellular density of MPO-positive cells was assessed using the Inform software. Patients were stratified by the optimal density cut-off (determined by ROC analysis). Our results demonstrate a significant enrichment of MPO⁺Cells in TSCC patients with unfavorable prognoses (Fig. 5A). Furthermore, elevated levels of TANs were robustly correlated with diminished OS (P < 0.0001, HR = 5.172, 95% CI: 3.171–8.438) and DFS (P < 0.0001, HR = 3.400, 95% CI: 2.118–5.457) (Fig. 5B).

Fig. 5.

Prognostic significance of TCIRG1⁺TANs density in TSCC patients. A mIHC staining of DAPI and MPO in TSCC tissues from patients with favorable and unfavorable prognoses. Scale bar represents 100 μm.;assessment of TANs density in tumor samples between the well outcome group (n = 54) and the poor outcome group (n = 66). B Kaplan–Meier survival analysis comparing OS and DFS for patients with high versus low TANs expression levels. C mIHC staining of DAPI, MPO, and TCIRG1 in TSCC tissues from patients with favorable and unfavorable prognoses. Scale bar represents 100 μm;assessment of TCIRG1⁺TANs density in tumor samples between the well outcome group (n = 54) and the poor outcome group (n = 66). D Kaplan–Meier survival analysis comparing OS and DFS for patients with high versus low TCIRG1⁺TANs expression levels. E ROC analysis of the predictive value of TANs and TCIRG1⁺TANs for OS and DFS

Prior data have established a strong correlation between TCIRG1 and neutrophil activity, with the Kaplan-Meier survival curves indicating that elevated TCIRG1 expression is a harbinger of poor prognosis. These insights have led us to hypothesize that TCIRG1 may serve as a pivotal marker for distinguishing TAN subsets. Consequently, we assessed the cellular density of TCIRG1 and MPO double-positive Cells (TCIRG1⁺MPO⁺Cells) as a predictive indicator of patient outcomes. We found that infiltration of TCIRG1⁺MPO⁺Cells were significantly associated with advanced tumor classification (T3-T4), lymph node metastasis, and advanced clinical stages (III-IV) (all p < 0.05), but not with tumor differentiation grade. (Figure S6). Furthermore, a higher density of these cells was observed in TSCC patients with poor clinical outcomes (Fig. 5C). Critically, increased infiltration of TCIRG1⁺MPO⁺Cells was strongly associated with worse OS (P < 0.0001, HR = 4.566, 95% CI: 2.739–7.613) and DFS (P < 0.0001, HR = 3.985, 95% CI: 2.430–6.537) (Fig. 5D). To further investigate the prognostic value of TCIRG1⁺MPO⁺Cells, we calculated the optimal cutoff values for their density, as depicted in Table S9. Statistical evaluations demonstrated that the density of TCIRG1⁺MPO⁺Cells offer enhanced predictive accuracy for the OS of TSCC patients, surpassing that of MPO⁺Cells alone (AUC = 0.850 for TCIRG1⁺MPO⁺Cells versus AUC = 0.749 for MPO⁺Cells; Fig. 5E). This enhanced predictive value was also evident for DFS (AUC = 0.841 for TCIRG1⁺MPO⁺Cells versus AUC = 0.717 for MPO⁺Cells; Fig. 5E). Further Receiver Operating Characteristic (DCA)was conducted to validate the clinical utility of the net benefit-based model. The results indicated that the DCA for the TCIRG1⁺MPO⁺ Cells significantly deviated from both extremes (all positive or all negative), demonstrating better clinical utility compared to MPO⁺Cells (Figure S7).These findings underscore the significant improvement in predictive power when TCIRG1⁺MPO⁺Cells density is considered over MPO⁺Cells alone, suggesting that TCIRG1⁺MPO⁺Cells may represent a promising biomarker for the diagnostic stratification of TSCC patients regarding their prognosis.

Discussion

Previous research has firmly established the NLR as an independent prognostic factor for various cancers [20]. Our findings corroborate this association, revealing a consistent link between elevated NLR and diminished survival in TSCC patients. Despite this, the precise mechanisms connecting high NLR to the adverse prognosis of TSCC patients are not fully understood. A growing body of evidence implicates a high NLR as a marker of increased neutrophil presence in peripheral blood [16]. Neutrophils, as the front-line defenders against pathogens, migrate to inflammatory sites [15], where their rapid recruitment is a defining feature of the inflammatory response [17]. Notably, neutrophil-driven inflammation is central to tumor progression [17]. Furthermore, neutrophils possess a remarkable capacity to transmigrate from the bloodstream into the tumor microenvironment (TME) [21]. Recent studies have also indicated that elevated circulating neutrophil counts are correlated with increased tumor-infiltrating neutrophils in patients [22, 23]. Consequently, we evaluated the density of tumor-infiltrating neutrophils using a neutrophil scoring system and observed that TSCC patients with higher levels of these neutrophils tend to have poorer outcomes. These results align with previous reports, including a meta-analysis showing that neutrophils are prevalent in most solid tumor samples [20]. Notably, a high ratio of tumor-infiltrating neutrophils has emerged as a robust predictor of poor prognosis across various cancer types [20]. Overall, these insights shed new light on the critical role of neutrophils in the prognosis of TSCC patients.

In recent years, there has been a significant upsurge in research interest in the complex roles and notable heterogeneity of neutrophils in the cancer context [10, 13]. Studies have highlighted the dualistic nature of tumor-infiltrating neutrophils, possessing both pro- and anti-tumor capabilities [11, 17]. The TME can harness neutrophil functions, including recruitment and production, to manifest pro-tumor characteristics. These include the promotion of genetic instability, tumor cell proliferation, angiogenesis, and the suppression of anti-tumor or inflammatory responses [9]. Conversely, neutrophils can exert anti-tumor effects through direct cytotoxicity against tumor cells or by enhancing anti-tumor functions via interactions with T cells [12]. Thus, we tried to recognized a subset of TANs that contribute to poor prognosis in TSCC patients and substantial endeavors have been devoted to elucidating their potential as therapeutic targets. Our proteomics analysis of tumor samples from advanced TSCC patients revealed notable activation of neutrophil-related signaling pathways in those with worse outcomes, underscoring the significant role of TANs in the TME. To delineate TAN subtypes, we conducted an intersectional analysis of three independent cohorts, including differentially expressed proteins from our proteomics data, neutrophil gene sets from GSEA3982, and a granule gene set derived from published single-cell RNA sequencing studies. We further found that the high TCIRG1 expression led to a poor survival by analyzing TCGA databases. Further analysis of TCGA databases indicated that high expression of TCIRG1 is linked to poor survival. Despite its historical nomenclature linked to T-cell studies [24], contemporary single-cell transcriptomic atlases from Human Protein Atlas (https://www.proteinatlas.org) reveal that TCIRG1 mRNA is predominantly expressed in myeloid cells, with the highest enrichment observed in neutrophils. This expression pattern aligns with its documented oncogenic roles; TCIRG1 dysregulation promotes malignant progression in several cancers [25]. Specifically, it functions as a metastasis enhancer in hepatocellular carcinoma by modulating cell proliferation and death pathways and serves as a prognostic biomarker in glioblastoma multiforme [26]. Furthermore, TCIRG1 expression strongly correlates with immune cell infiltration, including neutrophils [25, 27, 28]. Integrating these insights, we hypothesized that TCIRG1-positive TANs (TCIRG1⁺TANs) might be associated with adverse outcomes in TSCC patients, as detailed in Fig. 6. Image analysis of tissue sections immunostaining for TCIRG1 and MPO demonstrated that a high density of TCIRG1⁺TANs was significantly correlated with poor prognosis in TSCC patients. Notably, the predictive value of TCIRG1⁺TANs was found to surpass that of TANs alone. While there was a correlation between the NLR cutoff and the hazard ratio for TSCC patient outcomes, its predictive accuracy at approximately 75% was considerably lower than that of TCIRG1⁺TANs. These findings collectively suggest that TCIRG1⁺TANs may serve as a promising prognostic indicator for TSCC patients.

Fig. 6.

Schematic illustration of the regulatory role of TCIRG1⁺TANs in TSCC patient outcome

Despite the robustness of our findings, several limitations must be acknowledged. The retrospective design of the study introduces potential biases, including selection bias and incomplete data collection. Additionally, the generalizability of our results may be constrained by the single-center study design. Finally, while TCIRG1 has been identified as a promising biomarker, its precise biological function and the signaling pathways underlying its effects remain to be fully elucidated.

Conclusion

This study highlights the critical significance of neutrophil-associated biomarkers, specifically the NLR and TCIRG1⁺TANs, in forecasting adverse clinical outcomes in patients with TSCC. By employing a comprehensive analysis of a substantial patient cohort and integrating multi-omics methodologies, we identified NLR and TANs as reliable predictors of reduced survival rates. Notably, our findings demonstrate that the density of TCIRG1⁺TANs exhibits superior prognostic accuracy compared to overall TAN density, establishing this subset as a more precise indicator of unfavorable survival outcomes. This discovery not only enhances our comprehension of neutrophil heterogeneity and their functional roles within the tumor microenvironment but also identifies potential therapeutic targets. Although our study provides a robust foundation for future clinical applications, additional prospective and mechanistic investigations are essential to validate the clinical efficacy of these biomarkers. Collectively, this research contributes novel insights into neutrophil-mediated inflammation in TSCC and offers promising avenues for advancing patient management strategies.

Abbreviations

OSCC

Oral Squamous Cell Carcinoma

TSCC

Tongue Squamous Cell Carcinoma

NLR

Neutrophil-to-lymphocyte ratio

MLR

Monocyte-to-lymphocyte ratio

PLR

Platelet-to-lymphocyte ratio

SII

Systemic inflammation index

ROC

Receiver Operating Characteristic

AUC

Area under the curve

DCA

Decision Curve Analysis

IQRs

Interquartile ranges

AJCC

American Joint Committee on Cancer

OS

Overall Survival

DFS

Disease Free Survival

H&E

Hematoxylin and eosin

mIHC

multiplex immunohistochemistry

GO/KEGG

Gene Ontology/Kyoto Encyclopedia of Genes and Genomes

TCGA

The Cancer Genome Atlas

TME

Tumour immune microenvironment

MPO

Myeloperoxidase

TANs

Tumor-associated neutrophils

TCIRG1

T cell immunoregulatory protein 1

Authors’ contributions

Bowen Li designed and supervised research, conducted experiments, analyzed data, and wrote the manuscript. Zhaoyu Lin and Qunxing Li designed research, analyzed data, and revised the manuscript. Guoxin Huang, Chenchang Li, Jingjing Han and Zheng Cheng conducted experiments and analyzed data. Jingbin Huo, Yupeng Wu, Lianxi, Mai, Guangxin Rao, Yingnan Ma, Wenhao Sun, Xiaojuan Wang conducted experiments and provided reagents.

Funding

This work was supported by the National Natural Science Foundation of China (Grant Nos. 82403268, 82472823, 82303758, 82272788, 82072990 and 82072988); Guangdong Science and Technology Development Fund (2024A1515012876, 2023A1515110834); CSCO-Meck Cancer Research Fund (Y-MSDZD2022-0459); CSCO-HengRui Cancer Research Fund (Y-HR2022QN-0751).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki.The protocol of all study cohorts was approved by the institutional review board and ethics committee of Sun Yat-Sen Memorial Hospital (SYSKY-2024-694-01), and informed consent was waived due to the retrospective nature of the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.