Prognostic Impact and Clinicopathological Significance of TIM‐3 and LAG‐3 Expressions in Laryngeal Squamous Cell Carcinoma

ABSTRACT

Introduction

Research on immune checkpoint molecules is increasing, reflecting their critical role in cancer pathogenesis. While PD‐1, PD‐L1, and CTLA‐4 have been widely studied in head and neck cancers, outcomes remain unsatisfactory, highlighting the need for novel prognostic markers and therapeutic targets. This study evaluated the prognostic significance of TIM‐3 and LAG‐3 expressions in laryngeal squamous cell carcinoma.

Methods

Fifty patients who underwent open partial or total laryngectomy for laryngeal squamous cell carcinoma (2016–2018) were retrospectively analyzed. Tumor specimens were assessed by immunohistochemistry for TIM‐3 and LAG‐3 expressions, as well as overall immune infiltration. Survival analyzes were performed using Kaplan–Meier and multivariable Cox regression models.

Results

TIM‐3 expression showed no significant prognostic impact in the overall cohort but was associated with worse overall survival in the low‐to‐moderate immune infiltration subgroup (p = 0.045). LAG‐3 positivity correlated with reduced recurrence‐free survival in the full cohort (p = 0.015) and with both overall and recurrence‐free survival in the low/moderate infiltration subgroup (p = 0.029 and p = 0.016, respectively). Multivariable Cox regression suggested LAG‐3 as an independent predictor of recurrence, while TIM‐3 showed an effect on overall survival only in the low/moderate immune infiltration subgroup.

Conclusion

TIM‐3 and LAG‐3 expressions are associated with adverse outcomes in laryngeal cancer, particularly in tumors with limited immune infiltration. These findings are exploratory and support further multicenter studies to validate their prognostic and potential immunotherapeutic relevance.

Level of Evidence

3.

1. Introduction

Laryngeal carcinoma stands as one of the most prevalent head and neck cancers, with a 5‐year survival rate of 67% [1]. Despite advances in diagnostic and therapeutic approaches, survival outcomes have remained largely unchanged over the last decades. This stagnation highlights the need for a better understanding of the biological mechanisms driving tumor progression and treatment resistance [2]. In this context, the tumor immune microenvironment has gained increasing attention because of its pivotal role in cancer development, progression, and treatment resistance. Although established immune checkpoint pathways have been extensively investigated in head and neck squamous cell carcinoma, clinical responses remain limited in many patients, supporting the need to explore alternative immune checkpoint molecules as prognostic markers and potential therapeutic targets [3, 4, 5, 6, 7].

TIM‐3 (T‐Cell Immunoglobulin and Mucin Domain‐3) and LAG‐3 (CD223) (Lymphocyte Activation Gene‐3) are emerging immune checkpoint molecules that are expressed primarily on tumor‐infiltrating lymphocytes (TILs) and, in some cases, on tumor cells themselves. Both are closely associated with T‐cell exhaustion and tumor immune escape. TIM‐3 is expressed on activated CD4+ and CD8+ T cells, where its interaction with ligands such as galectin‐9 and CEACAM‐1 (Carcinoembryonic Antigen Cell Adhesion Molecule 1) inhibits T‐cell proliferation and cytokine production [8, 9, 10]. LAG‐3, as an indicator of T cell exhaustion, similarly binds to MHC class II (Major Histocompatibility Complex II) molecules and L‐selectin, suppressing T‐cell effector functions and enhancing Treg‐mediated immunosuppression through modulation of cytokine secretion [11, 12, 13, 14, 15]. Both markers, together with PD‐1, indicate CD8+ T‐cell exhaustion and are highly expressed on regulatory T cells (Tregs), where they enhance immunosuppression through upregulation of CTLA‐4 and PD‐1 [16, 17, 18].

TIM‐3 and LAG‐3 have been increasingly investigated in several solid tumors [15, 19, 20, 21]. Both markers are involved in T‐cell exhaustion and tumor immune escape, suggesting that their expression may have prognostic relevance in the tumor immune microenvironment [11, 22, 23, 24]. However, their relationship with clinicopathological features, survival outcomes, and the degree of intratumoral immune infiltration in laryngeal cancer has not been clearly established.

In this study, we aimed to investigate the expression of TIM‐3 and LAG‐3 in LSCC, assess their associations with clinicopathological features and immune infiltration levels, and evaluate their impact on survival parameters. We further examined whether their prognostic relevance differs according to the density of tumor‐infiltrating lymphocytes, with the goal of identifying potential biomarkers and future targets for immunotherapy in laryngeal cancer.

2. Materials and Methods

In this retrospective cohort study, we adhered to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines in the design, execution, and reporting.

2.1. Patient Selection and Clinical Evaluation

Fifty patients diagnosed with laryngeal squamous cell carcinoma between 2016 and 2018 who underwent partial or total laryngectomy by open surgery were included in the study. Lateral neck dissection was also performed in all patients, either bilateral or unilateral. A single team with expertise in treating head and neck cancers conducted all surgeries. Patients with a history of another malignancy, those who had undergone chemotherapy or radiation treatment previously, and those with hepatitis or HIV infection were excluded from the study. The 8th edition of AJCC (American Joint Committee on Cancer) TNM (Tumor Node Metastasis) staging was used for both pathological and clinical staging. Pathology specimens were examined by a single pathologist specialized in head and neck cancer pathology.

Patients' demographics, histopathological characteristics of the tumors, such as the tumor's location, TNM stage, histological grade, presence of vascular/lymphatic/perineural/cartilage invasion, and extranodal extension, were assessed. The type and the date of the surgery, presence of recurrence and/or metastases and the date of event were noted. Surveillance parameters as overall survival (OS), disease‐specific survival (DSS), disease‐free survival (DFS), and recurrence‐free survival (RFS) were calculated. Immune infiltration, expression of TIM‐3 and LAG‐3 were evaluated utilizing immunohistochemistry and light microscopy.

2.2. Immunohistochemical Staining

After evaluating H&E stained slides of the cases under a light microscope (Olympus, Tokyo, Japan), one paraffin block was selected for each case, containing sufficient tumor tissue and suitable for immunohistochemical study. To determine the immunohistochemical staining properties, seven positively charged slides (Epiglass, Istanbul, Turkey) were obtained from the selected paraffin blocks in 2‐μm sections, using a microtome knife Patho‐Cutter II (Erma, Tokyo, Japan). The 2‐μm sections on a positively charged slide were kept in an oven (Hybaid, UK) set at 56° for 2 h.

An automated staining device (Ventana Benchmark XT, Tuscon, Arizona, USA) was used for immunohistochemical staining. For the procedure, a ready kit, UltraView Universal DAB Detection Kit (Ventana, Basel, Switzerland), containing biotin‐free HRP multimer‐based hydrogen peroxide and 3,3‐diaminobenzidine tetrachloride (DAB) chromogen suitable for the device, was used. Monoclonal antibodies including LAG‐3 (CD223) [D2G4O] Conc. 0.1 mL (1:100–200) (Epitomics, Nanterre, France) and TIM‐3 [D5D5R] Conc. 0.1 mL (1:400) (CellSignaling, Massachusetts, USA) were applied to the slides. Additionally, CD4 (T4) [4B12] RTU Bond 7 mL (LEICA/Novocastra, Illinois, USA) and CD8 (T8) [4B11] RTU Bond 7 mL (LEICA/Novocastra, Illinois, USA) immunostaining was performed to assess tumor‐infiltrating lymphocytes. Counterstaining was performed with hematoxylin solution (Biognost, Zagreb, Croatia) in the dyeing device, and the process was completed after dehydration and xylene stages. The slide was sealed with entellan and made ready for evaluation under the light microscope.

2.3. Determination of the Immune Infiltration

There are currently no defined criteria for evaluating immune infiltration and tumor‐infiltrating lymphocyte density within a tumor. In this study, we utilized a previously described method involving light microscopic evaluation of hematoxylin and eosin (H&E) (Biognost, Zagreb, Croatia)—stained slides of patients. Subsequently, patients were classified based on the presence of TILs as high (diffuse infiltration of lymphocytes in > 80% of tumor and stroma), moderate (infiltration in 20%–80% of tumor and stroma), or low (sparse or absent infiltration in < 20% of tumor and stroma) [25]. For survival analysis, low and moderate groups were combined because preliminary evaluation showed overlapping survival outcomes, and separate modeling of three groups was limited by small subgroup sizes. Therefore, a two‐tier classification (low/moderate vs. high) was applied for Kaplan–Meier and Cox regression analyzes to increase statistical reliability.

2.4. Immunohistochemical Evaluation

Initially, five areas with over 70% cellularity, minimal necrosis, and more than 50% tumor parenchyma were chosen from H&E‐stained tumor sections for each patient. These same areas were identified in slides subjected to immunohistochemistry with TIM‐3 and LAG‐3 monoclonal antibodies. Evaluations were conducted under a light microscope (Olympus, Tokyo, Japan) at 200× magnification (20× objective lens), with TILs and tumor cells stained with mentioned dyes manually counted by a head and neck cancer pathologist.

2.5. Evaluation of TIM‐3 and LAG‐3 Expressions

For both markers, there are no defined and universally accepted evaluation criteria in the literature yet. The combined positive score (CPS) algorithm, which was previously found to be effective, was used in this study to classify the slides [26, 27, 28]. In this calculation, the total number of TIM‐3 or LAG‐3 expressing tumor cells and TILs in a tumor section (0.05 mm dia under 200× magnification) is divided by the total number of tumor cells and multiplied by 100. It is considered positive if CPS is equal to or higher than 1. Two groups were established for both TIM‐3 and LAG‐3 as positive or negative (Figure 1). Parameters were then compared between these two groups.

FIGURE 1.

Demonstration of TIM‐3 and LAG‐3 expressions by immunohistochemistry (200×). (A) TIM‐3 positive, (B) TIM‐3 negative, (C) LAG‐3 positive, (D) LAG‐3 negative.

2.6. Statistical Analysis

A sample size estimation was performed during the ethics approval process using G*Power 3.1 (Heinrich Heine University, Düsseldorf, Germany) to provide a planning framework rather than a formal determination of statistical power for the present analyzes. Effect size (h = 1.41) assumptions were derived from comparable published data [11], targeting a clinically relevant hazard ratio of approximately 2.0 and a 60/40 allocation between marker‐positive and marker‐negative groups, corresponding to an estimated sample size of 48 patients under a log‐rank approximation (two‐sided α = 0.05, 1−β = 0.95). However, given the retrospective design and the event‐driven nature of survival outcomes, this calculation should be interpreted as a theoretical approximation. Therefore, the precision of multivariable estimates was primarily evaluated based on confidence interval width, and adjusted model findings were considered exploratory.

SPSS 23.0 software (IBM Corp., Armonk, New York, USA) was used for statistical analysis. Descriptive statistics were presented as numbers and percentages for categorical data and as means, standard deviations, median values, minimum, and maximum values for numerical data. The comparison of categorical variables between groups was performed using Pearson chi‐squared Analysis. For numerical variables not satisfying the normal distribution, the Mann–Whitney U Test was used for two independent groups, and the Kruskal–Wallis Test for more than two independent groups. The statistical alpha significance level was accepted as p < 0.05. As the parametric test requirement for correlation analysis between numerical variables could not be satisfied, Spearman Correlation Analysis was applied.

The Kaplan–Meier method was used to estimate survival curves, and group comparisons were performed using the log‐rank test. Multivariable Cox proportional hazards regression analyzes were conducted to assess independent prognostic effects. To reduce the risk of model overfitting in relation to the number of events, clinicopathological variables included in multivariable models were limited to established prognostic factors (lymphatic invasion, perineural invasion, and extranodal extension) and immunological markers (TIM‐3 and LAG‐3 expression) were evaluated within the same framework. The overall significance of the multivariable Cox regression models was evaluated using the Omnibus test of model coefficients, with p‐values < 0.05 considered indicative of a statistically significant model fit.

2.7. Ethical Considerations

All research is performed in accordance with the Declaration of Helsinki. A Clinical Research Ethics Committee Approval Form was obtained prior to the commencement of the study. Informed consent forms were obtained from all patients before surgery. All methods were performed in accordance with relevant guidelines and regulations.

3. Results

Fifty patients (48 male, 2 female) who underwent laryngectomy for LSCC were included in this study. Mean age was 62.46 ± 8.79 years (from 45 to 84). The mean follow‐up period was 44.40 ± 16.20 months (min: 4 months, max: 64 months). During the follow‐up period, 18 patients died, constituting the overall survival events, and 10 deaths were disease‐specific. Ten patients developed locoregional recurrence, which was evaluated as an event for recurrence‐free survival. In this study, local, regional, and combined locoregional recurrences were recorded and analyzed together under the category of locoregional recurrence. Distant metastasis was observed in four patients, 2 of whom also had locoregional recurrence. Overall, 12 patients were classified as having disease‐free survival events, including 10 patients with locoregional recurrence and two patients with distant metastasis alone. Data considering the clinical and histopathological characteristics are summarized in Table 1.

TABLE 1.

Summary of clinical and histopathological data.

Clinical and histopathological parameters		Number of patients (n)	Percentage (%)
Tumor location	Glottic	12	24
	Supraglottic	19	38
	Transglottic	19	38
Type of operation	Partial laryngectomy	32	64
	Total laryngectomy	18	36
Neck dissection	Unilateral	6	12
	Bilateral	44	88
Adjuvant radiotherapy	+	46	92
	−	4	8
Adjuvant chemotherapy	+	15	30
	−	35	70
pT	T1	1	2
	T2	8	16
	T3	15	30
	T4	26	52
Grade	1	2	4
	2	30	60
	3	18	36
Vascular invasion	−	26	52
	+	24	48
Lymphatic invasion	−	8	16
	+	42	84
Perineural invasion	−	29	58
	+	21	42
Thyroid cartilage invasion	−	27	54
	+	23	46
pN	N0	27	54
	N1	8	16
	N2	5	10
	N3	10	20
Extranodal extension	−	40	80
	+	10	20
Inflammatory response	Mild	13	26
	Moderate	20	40
	Dense	17	34
Recurrence	−	40	80
	+	10	20
Metastasis	−	46	92
	+	4	8
Disease‐specific death	−	40	80
	+	10	20

The associations between TIM‐3 and LAG‐3 expression and clinicopathological parameters are summarized in Table 2. No significant association was observed between TIM‐3 expression and clinicopathological parameters in the overall cohort. LAG‐3 positivity was significantly associated with recurrence (p = 0.031; r = 0.336). In the low‐to‐moderate immune infiltration subgroup (n = 33), the associations between TIM‐3 and LAG‐3 expression and clinicopathological parameters are presented in Table 3. In this subgroup, TIM‐3 positivity was significantly associated with thyroid cartilage invasion (p = 0.011; r = 0.454) and advanced T stage (p = 0.013; r = 0.448). Similarly, LAG‐3 positivity was significantly associated with thyroid cartilage invasion (p = 0.050; r = 0.343) and advanced T stage (p = 0.017; r = 0.434).

TABLE 2.

Relationship of TIM‐3 and LAG‐3 expressions with histopathological and clinical parameters.

		TIM‐3			LAG‐3
Histopathological parameters		TIM‐3 negative	TIM‐3 positive	p, correlation (r)	LAG‐3 negative	LAG‐3 positive	p, correlation (r)
pT	T1–3	12	12	p = 0.136 r = 0.196	12	12	p = 0.136 r = 0.196
	T4	8	18		8	18
Grade	1–2	11	21	p = 0.217 r = −0.153	11	21	p = 0.217 r = −0.153
	3	9	9		9	9
Vascular invasion	−	11	15	p = 0.477 r = 0.049	11	15	p = 0.477 r = 0.049
	+	9	15		9	15
Lymphatic invasion	−	4	4	p = 0.401 r = 0.089	2	6	p = 0.297 r = −0.134
	+	16	26		18	24
Perineural invasion	−	11	18	p = 0.475 r = −0.050	12	17	p = 0.525 r = 0.033
	+	9	12		8	13
Thyroid cartilage invasion	−	14	13	p = 0.058 r = 0.262	13	14	p = 0.163 r = 0.180
	+	6	17		7	16
pN	N0	12	15	p = 0.343 r = 0.098	13	14	p = 0.163 r = 0.180
	N+	8	15		7	16
Extranodal extension	−	16	24	p = 0.635 r = 0.000	17	23	p = 0.365 r = 0.102
	+	4	6		3	7
Recurrence	−	17	23	p = 0.365 r = 0.102	19	21	p = 0.031 r = 0.306
	+	3	7		1	9
Metastasis	−	18	28	p = 0.528 r = −0.060	18	28	p = 0.528 r = −0.060
	+	2	2		2	2
Disease‐specific death	−	16	24	p = 0.635 r = 0.000	18	22	p = 0.139 r = 0.204

Note: Statistically significant values are presented in bold.

TABLE 3.

Relationship of TIM‐3 and LAG‐3 expressions with histopathological and clinical parameters in low‐to‐moderate immune infiltration subgroup (n = 33).

		TIM‐3			LAG‐3
Histopathological parameters		TIM‐3 negative	TIM‐3 positive	p, correlation (r)	LAG‐3 negative	LAG‐3 positive	p, correlation (r)
pT	T1–3	10	4	p = 0.013 r = 0.448	8	6	p = 0.017 r = 0.434
	T4	5	14		3	16
Grade	1–2	8	13	p = 0.224 r = −0.196	5	15	p = 0.125 r = −0.267
	3	7	5		6	6
Vascular invasion	−	8	6	p = 0.211 r = 0.201	5	9	p = 0.547 r = 0.043
	+	7	12		6	13
Lymphatic invasion	−	4	1	p = 0.116 r = 0.293	2	3	p = 0.550 r = 0.060
	+	11	17		9	19
Perineural invasion	−	6	9	p = 0.412 r = −0.100	6	9	p = 0.355 r = 0.129
	+	9	9		5	13
Thyroid cartilage invasion	−	11	5	p = 0.011 r = 0.454	8	8	p = 0.050 r = 0.343
	+	4	13		3	14
pN	N0	8	8	p = 0.437 r = 0.089	7	9	p = 0.195 r = 0.214
	N+	7	10		4	13
Extranodal extension	−	12	13	p = 0.459 r = 0.090	10	15	p = 0.158 r = 0.250
	+	3	5		1	7
Recurrence	−	13	11	p = 0.105 r = 0.286	11	13	p = 0.013 r = 0.433
	+	2	7		0	9
Metastasis	−	13	16	p = 0.626 r = −0.034	9	20	p = 0.407 r = −0.131
	+	2	2		2	2
Disease‐specific death	−	12	12	p = 0.324 r = 0.149	10	14	p = 0.104 r = 0.289
	+	3	6		1	8

Note: Statistically significant values are presented in bold.

There was no significant relationship between TIM‐3 expression and OS (p = 0.437), DSS (p = 0.601), DFS (p = 0.586), and RFS (p = 0.242) parameters in the total patient cohort (Figure 2). LAG‐3 expression was found as a significantly negative prognostic factor for RFS (p = 0.015). Although LAG‐3 positivity seemed to be a negative factor for OS (p = 0.082), DSS (p = 0.082), and DFS (p = 0.086), the difference was not statistically significant (Figure 3).

FIGURE 2.

Kaplan–Meier diagrams depicting the relationship between TIM‐3 expression and survival parameters in the entire patient population. (A) overall survival (OS), (B) disease‐specific survival (DSS), (C) disease‐free survival (DFS), (D) recurrence‐free survival (RFS). (Tick marks on the curves represent the censored data).

FIGURE 3.

Kaplan–Meier diagrams depicting the relationship between LAG‐3 expression and survival parameters in the entire patient population. (A) overall survival (OS), (B) disease‐specific survival (DSS), (C) disease‐free survival (DFS), (D) recurrence‐free survival (RFS). (Tick marks on the curves represent the censored data).

Among the 33 patients with low to moderate immune infiltration in the tumor, TIM‐3 positivity had a significant negative impact on OS (p = 0.045). The TIM‐3 positive group showed a trend toward poorer outcomes in terms of DSS (p = 0.265), DFS (p = 0.168), and RFS (p = 0.059), without statistical significance (Figure 4). For the subset of patients with low to moderate immune infiltration, the presence of LAG‐3 expression had a statistically significant negative effect on OS (p = 0.029) and RFS (p = 0.016). While the LAG‐3 positive group did show a trend toward worse outcomes in terms of DSS (p = 0.098) and DFS (p = 0.101), this difference did not reach statistical significance (Figure 5).

FIGURE 4.

Kaplan–Meier diagrams depicting the relationship between TIM‐3 expression and survival parameters in the patient population with low immune response against the tumor. (A) overall survival (OS), (B) disease‐specific survival (DSS), (C) disease‐free survival (DFS), (D) recurrence‐free survival (RFS). (Tick marks on the curves represent the censored data).

FIGURE 5.

Kaplan–Meier diagrams depicting the relationship between LAG‐3 expression and survival parameters in the patient population with low immune response against the tumor. (A) overall survival (OS), (B) disease‐specific survival (DSS), (C) disease‐free survival (DFS), (D) recurrence‐free survival (RFS). (Tick marks on the curves represent the censored data).

In multivariable Cox regression analysis for RFS, including LAG‐3 expression, lymphatic invasion, perineural invasion, and extranodal extension, LAG‐3 positivity remained independently associated with an increased risk of recurrence (HR = 8.98, 95% CI: 1.12–72.10; p = 0.039). In the subgroup with low to moderate immune infiltration, multivariable Cox analysis for OS showed that TIM‐3 expression remained associated with mortality after adjustment for lymphatic invasion, perineural invasion, and extranodal extension (HR = 5.49, 95% CI: 1.10–27.44; p = 0.038). LAG‐3 demonstrated a similar direction of effect but did not reach statistical significance in the adjusted model (HR = 6.95, 95% CI: 0.87–55.68; p = 0.068). In the subgroup with low to moderate immune infiltration, no recurrence events occurred in LAG‐3–negative tumors, resulting in unstable effect estimates and wide confidence intervals in the Cox model for RFS.

4. Discussion

Immune checkpoint pathways have been extensively investigated in head and neck cancers; however, their clinical benefit remains limited to a subset of patients [29, 30, 31, 32]. In this context, TIM‐3 and LAG‐3 have attracted increasing attention as alternative immune checkpoint molecules that may contribute to tumor immune escape and treatment resistance [3, 33].

TIM‐3 serves as an exhaustion indicator on CD8+ T lymphocytes, alongside PD‐1 [16], with FoxP3+ cells exhibiting the highest TIM‐3 expression among CD4+ cells [34]. TIM‐3‐expressing Tregs are notably more concentrated in cancer tissue than in peripheral blood [17], and those isolated from the tumor demonstrate immunosuppressive effects [35]. TIM‐3 expression has been associated with poor prognosis across various solid tumors [27, 36, 37, 38, 39]. In esophageal cancer, which shares features with head and neck malignancies, TIM‐3 expression correlated with blood vessel invasion, and patients with TIM‐3(+)PD‐1(+)CD8‐low tumors had the worst outcomes [40]. Shan et al. [41] reported similar findings, associating TIM‐3 with higher TNM stage, lymph node metastasis, and depth of invasion.

Preclinical studies showed that TIM‐3 blockade in animal models of head and neck cancer enhances the immune response against tumors by increasing T lymphocyte activity [22]. In clinical settings, TIM‐3 expression increases after cetuximab therapy and correlates with poor response [24, 42], while high TIM‐3 expression is observed in treatment‐refractory cases [43]. In oral cavity cancer, nimotuzumab resistance demonstrated a relation with higher TIM‐3 expression, but no significant correlation was found with overall survival or histopathologic characteristics [44]. However, another study demonstrated a relation between TIM‐3 expression and lymph node metastasis in oral cavity cancer [33]. To the best of our current knowledge, there is not yet a study evaluating TIM‐3 expression specifically in laryngeal cancers. In our study, TIM‐3 expression did not significantly affect survival in the entire patient population. Yet, in the patient population with low to moderate immune response against the tumor (n = 33), OS was significantly reduced with TIM‐3 positivity (p = 0.045). Moreover, in this subgroup, a correlation was observed between TIM‐3 positivity and thyroid cartilage invasion and T stage (p = 0.011; r = 0.454 and p = 0.013; r = 0.448, respectively). In multivariable Cox analysis, TIM‐3 retains independent prognostic significance in the low to moderate immune response subgroup after adjustment for clinicopathological factors (HR = 5.49, 95% CI: 1.10–27.44; p = 0.038). Nevertheless, given the limited number of events and the wide confidence intervals, this association should be interpreted cautiously. These findings suggest that TIM‐3 expression may have greater prognostic relevance in tumors with limited immune infiltration (54.55%), whereas its presence in highly immune‐infiltrated tumors (70.59%) may represent an adaptive immune regulatory phenomenon rather than a direct indicator of adverse outcome.

LAG‐3, expressed on activated CD4+ and CD8+ T lymphocytes, functions as a co‐inhibitor immune checkpoint molecule [11]. It is highly expressed on Tregs, and its blockade in animal models enhances antitumor immunity by attenuating Treg‐mediated suppression [17, 45]. The expression of LAG‐3 was associated with poor prognostic factors in various malignancies [15, 21, 38].

LAG‐3 expression in head and neck cancer tissue is higher when compared to normal mucosa, and elevated expression has been found to be associated with poor differentiation, lymph node metastases, larger tumor size, and recurrence [11]. LAG‐3 has also been identified as an independent prognostic factor in oral cavity cancers, particularly in patients with nodal metastases, and may contribute to resistance to anti‐PD‐1 therapy [43, 44]. Our study is the first to evaluate LAG‐3 expression in laryngeal cancers. RFS was significantly lower in the LAG‐3 positive patient group (p = 0.015). Although other survival outcomes (OS, DSS, DFS) did not reach statistical significance, they trended worse in LAG‐3‐positive patients. However, in the low to moderate immune infiltration subgroup (n = 33), both OS and RFS were significantly lower with LAG‐3 positivity (p = 0.029 and p = 0.016, respectively), and no recurrences occurred in LAG‐3‐negative cases. LAG‐3 positivity in this subgroup also correlated with thyroid cartilage invasion and T stage (p = 0.050; r = 0.343 and p = 0.017; r = 0.434). LAG‐3 expression was associated with adverse outcomes overall, with its effect most pronounced in tumors with limited immune infiltration. In the multivariable Cox model, LAG‐3 expression remained independently associated with RFS (HR = 8.98, 95% CI: 1.12–72.10; p = 0.039). However, within the low to moderate immune infiltration subgroup it did not retain its impact on OS (HR = 6.95, 95% CI: 0.87–55.68; p = 0.068). In this subgroup, no recurrence events occurred in LAG‐3–negative tumors, resulting in unstable effect estimates and wide confidence intervals in the Cox model for RFS. Therefore, these results should be interpreted cautiously and considered exploratory. Biologically, LAG‐3–mediated inhibitory signaling may contribute to reduced antitumor immune activity, which could partly explain its higher expression in tumors with lower immune infiltration observed in our cohort (66.67% vs. 47.06%, respectively).

Previous preclinical and clinical studies have suggested that TIM‐3 and LAG‐3 may contribute to tumor immune escape by promoting T‐cell exhaustion and Treg‐mediated immunosuppression. Blockade of these pathways has been shown to restore antitumor immune responses in experimental models, and increased expression of TIM‐3 and LAG‐3 has also been implicated in resistance to anti‐PD‐1–based treatment [11, 22, 23, 24]. In this context, the observed association of these markers with adverse survival outcomes in our cohort may reflect their potential role in immune evasion in laryngeal squamous cell carcinoma. Consistent with many studies, we also observed a significant correlation between TIM‐3 and LAG‐3 expression (p < 0.001). Co‐expression of these markers, both linked to poor prognosis, highlights their potential as future immunotherapy targets in laryngeal cancer.

This study has several limitations that should be acknowledged. First, the cohort size was modest (n = 50), with a relatively limited number of survival events, which may limit the statistical strength of subgroup analyzes. The wide confidence intervals reflect statistical imprecision and indicate that these findings are exploratory. Second, the study design was retrospective, which introduces potential selection and information biases. Assessment of immune infiltration was based on single‐observer evaluation and may be subject to observer‐related variability. Lastly, there are currently no standardized, universally accepted cut‐off criteria for TIM‐3 and LAG‐3 positivity in laryngeal cancer, which may affect the reproducibility and generalizability of our findings. In this study, TIM‐3 and LAG‐3 expressions were evaluated using a binary CPS‐based classification, and further stratification according to higher CPS thresholds, such as CPS > 10 or CPS > 20, could not be performed due to the limited cohort size and number of survival events. Future studies with larger, multiinstitutional cohorts and standardized evaluation methods should investigate whether different CPS cut‐off levels provide additional prognostic value in laryngeal squamous cell carcinoma.

5. Conclusion

Our study evaluated the prognostic significance of TIM‐3 and LAG‐3 expressions in laryngeal squamous cell carcinoma. TIM‐3 expression showed an adverse prognostic impact mainly in tumors with low to moderate immune infiltration, particularly for overall survival, while LAG‐3 was associated with worse outcomes in the overall cohort and especially in tumors with limited immune infiltration, notably for recurrence‐free survival. The novel observation that the prognostic effect of these markers was more pronounced in tumors with low intratumoral immune infiltration highlights the need for further investigation. These findings suggest that TIM‐3 and LAG‐3 may serve as potential prognostic biomarkers in laryngeal cancer and could inform future immunotherapy research. However, given the modest sample size and statistical imprecision, these results should be considered exploratory and require validation in larger, multiinstitutional studies.

Funding

This study was funded by the Istanbul University—Cerrahpasa Scientific Research Projects Coordination Unit [Project Number: 2020‐34821].

Ethics Statement

Ethical approval for this clinical study was obtained from the Clinical Research Ethics Committee of the Cerrahpaşa Faculty of Medicine, Istanbul University‐Cerrahpaşa [Document No: 2019‐122461].

Consent

Written informed consent was obtained from all participants. No personal data or patient images are disclosed in this manuscript. All patients and authors provided consent for the publication of the study data.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

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