Immunomodulatory role of tumor microenvironment on oncological outcomes in advanced laryngeal cancer

Abstract

Background

The study evaluated the prognostic impact of the immune microenvironment in LSCC with markers of major immune cells to identify the key determinants of short-term disease-free survival (ST DFS) and reveal factors related to disease progression.

Methods

The study cohort included 61 patients who underwent total laryngectomy, 83.6% of whom were male with a mean age of 64.3 years at the time of surgery. Twenty-five patients had long term DFS (over 5 years), 8 – had moderate DFS (between 2 and 5 years), and 28 had short-term DFS (less than 2 years). Immunohistochemical staining and evaluation were performed on samples collected after the laryngectomy.

Results

The samples’ assessment revealed that the mean expression of all analysed markers was the highest both in stroma and the tumor compartment for short term DFS (ST DFS) patients. Analysis confirmed that a high stromal density of CD8 cells (p = 0.038) significantly correlated with DFS, and that the increased presence of CD57 cells (p = 0.021) was significantly associated with ST DFS. Moreover, the high density of CD68 cells in the tumor epithelial compartment had a negative prognostic impact on DFS (p = 0.032). Analysis of overall survival in the studied cohort with Kaplan-Meyer curves revealed that a high stromal density of CD68 cells was a significant negative predictor of OS (p = 0.008).

Conclusions

The observed associations of CD68 cells infiltration with progression and prognosis in patients with LSCC provide potential screening and therapeutic opportunities for patients with unfavourable outcomes.

Introduction

Head and neck squamous cell carcinoma (HNSCC) is still ranked worldwide among the cancers with poor prognoses. Laryngeal squamous cell cancer (LSCC) constitutes nearly one third of HNSCCs. For many years, LSCC therapy of LSCC has been consistently based on surgery or chemoradiotherapy. Both modalities in advanced cancer stages severely impair voice, respiratory and swallowing functions with serious psychophysical consequences that handicap patients’ quality of life. Over the past three decades there has been no improvement in the 5-year survival of patients with laryngeal cancer. The global data collected by the International Agency for Research on Cancer reported over 184 000 cases of laryngeal cancer in 2020 and predict a growing trend in the future [1]. In the same year, the mortality rate was 99 800 [1].

Therefore, comprehensive research continues to identify specific histological and molecular patterns of LSCC. This knowledge would enable a better understanding of disease biology and the implementation of targeted strategies to maximize functional and outcome results.

A growing amount of evidence confirms that the immune system plays a vital role in cancer development and progression. The immune reaction associated with the presence of malignant tumors involves numerous cells and factors that act as immunological markers. Tumor-infiltrating immune cells have been identified as promising prognostic factors in solid tumors including lung [2, 3], colorectal [4] and breast cancer [5]. The specific interaction between the immune environment and the development of laryngeal cancer has not yet been clearly elucidated, and very few studies have explored this issue [6,7,8,9,10,11]. Therefore, there is an ongoing need to search for further components to understand mechanisms and find pathways for potential therapeutic targets in LSCC.

For research on tumor immunology, distinguishing findings in the tumor stroma and tumor cells is essential. The stroma is called a nourishing microenvironment, where growth factors, cytokines and enzymes modify the extracellular matrix to favour the proliferation and survival of tumor cells, angiogenesis, cell migration and metastasis, but may also enhance the host immune reaction [12]. Analysis of both these compartments provides varied results for the same factors, confirming the need for separate assessment. High expression of CD45, CD11b, CD3 and COX-2 in the peritumoral stroma is correlated with an increased risk of metastasis, but the findings within the tumor are different [9]. The host’s immune response to tumor cells involves various types of immune cells including lymphocytes - both T (cytotoxic and helper) and B, NK cells, dendritic cells, macrophages, mast cells and neutrophils. Of special importance is the presence of so-called tumor infiltrating lymphocytes - TILs. The primary presentation of cancer antigens to T cells involves dendritic cells that initiate subsequent antitumor responses (humoral and cellular).

A relatively high concentration of TILs in the tumor microenvironment (TME), especially cytotoxic T cells (CD8) induces an inflammatory reaction and is a good marker of the host immune response with an expected increase in overall survival [7, 10]. Assessment of the degree of peritumoral lymphocyte infiltration has been proved to be a reliable supplementary component of the TNM classification in patients with colon cancer, and is helpful in predicting recurrence and mortality [13, 14]. On the other hand, tumor cells induce mechanisms to evade the immune system surveillance. One of the mechanisms involves immunological checkpoints and is related to the upregulated expression of programmed death ligand 1 (PD-L1) on the surface of tumor cells surface, which binds to the programmed death 1 (PD-1) receptor on TILs inducing tumor tolerance [15]. Another mechanism favouring tumor development and progression promotes a state of chronic inflammation in the TME with depletion of T cells and increased inflammatory infiltration by neutrophils [16]. To identify LSCC patients with a poor prognosis effectively and to expand treatment options in this group we need to search for molecular biomarkers related to the outcomes of patients with in laryngeal cancer.

In the present study, by visualizing major immune cells: B cells (CD20), general T cells (CD3), helper T cells (CD4), cytotoxic T cells (CD8), NK cells (CD57) and macrophages (CD68), we evaluated LSCC immune microenvironment to identify the key determinants of short-term disease-free survival and reveal factors related to disease progression.

Materials and methods

Clinicopathological data and study subjects

The study was approved by the Bioethics Committee at the Medical University of Warsaw (No. AKBE/173/2021).

Patients with a primary diagnosis of LSCC staged T3 or T4a and treated with total laryngectomy at the Department of Otorhinolaryngology Head and Neck Surgery, Medical University of Warsaw between 2013 and 2016 were identified in medical database. The exclusion criteria included: recurrent disease with salvage laryngectomy, a history of chemotherapy or neck radiotherapy due to other cancers, immunosuppression therapy, no access to follow up data, and incomplete adjuvant treatment with radio/chemotherapy according to the recommendations (adjuvant radiotherapy/systemic treatment for confirmed adverse features on histopathology including cartilage infiltration, perineural and perivascular infiltration, extranodal extension, nonradical resection or close margins). For the included patients the availability of paraffin blocks in the repository of the Department of Pathology was checked and tissue material for 61consequtive patients was deposited for further processing. Clinical, pathological, additional therapy and outcome data were verified for all patients. Our cohort was divided into three groups according to the period from surgery to disease recurrence (disease free survival-DFS): long-term disease-free survival (LT DFS) (> 5 years), moderate long-term disease-free survival (MT DFS) (2–5 years) and short-term disease-free survival (ST DFS) (< 2 years). Overall survival (OS) was estimated from the time of laryngectomy to the date of death t or last follow-up visit in 2023.

Immunohistochemical staining

Post laryngectomy haematoxylin and eosin (H&E)-stained slides and formalin-fixed, paraffin-embedded tissue blocks were retrieved from the archives of the Department of Pathology, Medical University of Warsaw. H&E- stained slides were examined by an experienced pathologist to choose the most representative samples containing LSCC. The samples were stained with antibodies against CD3 (Polyclonal Rabbit Anti-Human, Ready-to-Use, Agilent Dako), CD4 (Monoclonal Mouse Anti-Human, Clone 4B12, Ready-to-Use, Agilent Dako), CD8 (Monoclonal Mouse Anti-Human, Clone C8/144B, Ready-to-Use, Agilent Dako), CD20 (Monoclonal Mouse Anti-Human, Clone L26, Ready-to-Use, Agilent Dako), CD57 (Monoclonal Mouse Anti-Human, Clone TB01, Ready-to-Use, Agilent Dako) and CD68 (Monoclonal Mouse Anti-Human, Clone KP1, Ready-to-Use, Agilent Dako) and visualized using Dako Omnis EnVision FLEX Visualization System in accordance with the manufacturers’ recommendations.

Immunohistochemistry evaluation

All H&E and IHC slides were digitalized with a Hamamatsu NanoZoomer 2.0-HT scanner and viewed via NDP.view2 software. The expression of selected biomarkers: CD3, CD4, CD8, CD20, CD57 and CD68 was independently evaluated by two pathologists blinded to the patients’ follow-up data. The method of evaluation was based on the Guidelines for the Assessment of TILs in Solid Tumors: Recommendations by an International Immuno-Onoclogy Biomarker Working Group [16, 17]. The presence of immunocompetent cells was examined at 200x magnification in two compartments: the intratumoral epithelial compartment (tumor cell nests) and the tumor stromal compartments (tissue between cancer cell nests within tumor). The average density of specific cells was expressed as a continuous variable by estimation of the ratio of the area occupied by IHC-positive cell infiltrates in selected compartment (tumor cell nests or tumor stroma) to the entire intratumoral or tumor stromal area (e.g. stromal % CD3 = area occupied by CD3 cells in the tumor stroma/total tumor stromal area). For each slide, a full assessment of the tumor area was performed, however areas of ulceration and necrosis were excluded from the analysis. Representative examples of H&E and IHC stains are presented in Fig. 1.

Fig. 1

Examples of immunohistochemistry expression of selected biomarkers in LSCC. (A) Microphotograph of an LSCC with a visible intratumoral epithelial compartment (tumor cell nests marked with T) and intratumoral stromal compartment (marked with asterix), routine H&E staining. (B) CD68 expression in moderate number of cells in tumor nests (marked with T) and tumor stroma (marked with asterix). (C) CD57 expression in a small number of scattered cells in the tumor stroma and only few cells in tumor nests. (D) Diffuse infiltration of CD20 cells in the stromal compartment, with only a few cells in the epithelial compartment. (E) CD3 cells visible in the stromal compartment around tumor cell nests with moderate amounts of CD3 cells in the epithelial compartment. (F) CD8 expression in TILs shows a similar pattern to that of CD3 expression, with the majority of CD8 cells being visible in the tumor stroma surrounding the cell nest. Original magnification in photomicrographs: 100×

Statistical analysis

Statistical analysis included comparisons of average cell density expression levels with selected antibodies between the LT DSF, MT DSF and ST DSF study groups. Kruskal-Wallis one-way analyses of variance with a Bonferroni correction for multiple hypothesis testing were used. The Mann-Whitney U tests was used to compare variables between two study groups. For dichotomous variables chi-square tests and Fisher’s exact tests were used. Kaplan-Meier estimators and log rank tests were used for survival analysis. P-values of less than 0.05 were considered statistically significant. The analysis was performed using R (version 4.1.2) and IBM SPSS Statistics (version 28.0.1.0) programs.

Results

Clinicopathological and immunohistochemical characteristics

The study cohort included 61 patients who underwent total laryngectomy, mostly males (83.6%) with a mean age of 64.3 years at the time of surgery. In most cases the extent of the tumor overlapped with a few other laryngeal locations (60.6%). Pathologically confirmed advancement was mostly the T3 stage (77.04%). Over 70% of patients had no nodal metastases. The median disease-free survival (DFS) and overall survival (OS) rates were 4.05 ± 3.33 and 4.7 ± 2.99 years, respectively. Twenty-five patients were identified as having long-term disease-free survival (LT DFS) without disease recurrence for 5 years following surgery. The moderate long-term disease-free survivals (MT DFS) group consisted of 8 patients who experienced the disease recurrence between 2 and 5 years after surgery, and 28 experienced patients short-term disease-free survivals (ST DFS) with relapse within 2 years. There was a significant difference (p = 0.003) in the mean age among the three groups, with the highest average for MT DFS (72.6 years), followed by ST DFS (66.6 years) and LT DFS (59.1 years). The other features including sex distribution, tumor localization and advancement were not significantly different between groups. Table 1. presents the clinical, histological and outcome characteristics of the patients.

Table 1 Clinical, histological and outcome characteristics

Associations between immune markers expression and disease-free survival

The samples assessment revealed the highest mean percentage of CD3 cells in the stromal compartment (24.3%), but for the tumor epithelial compartment CD68 cells predominated (9.3%). For the whole study cohort, the evaluated immunological markers had higher expression in the stromal compartment than in the tumor compartment; however, the tumor/stromal ratio differed for specific cell types ranging from 0.16 for CD20 to 0.57 for CD57. The mean expression of all analysed markers was highest in both the stroma and tumor cell nests of patients with ST DFS. The tumor/stromal ratios for CD8 (0.5) and CD68 (0.53) were greater in ST DFS patients than in MT DFS patients (0.44 and 0.38, respectively) and LT DFS (0.45 and 0.45, respectively). Table 2. comprehensively summarizes the immunohistochemistry results for the tumor epithelial and stromal compartments of selected markers in relation to the outcomes: LT DFS, MT DFS and ST DFS.

Table 2 Summary of immunohistochemistry results for the tumor epithelial and stromal compartments of selected markers in relation to the outcomes: LT DFS, MT DFS and ST DFS

The analysis confirmed that a high stromal density of CD8 cells (p = 0.038) and CD57 cells (p = 0.021) was significantly associated with DFS, whereas a high CD68 cells density in the tumor compartment had a negative prognostic impact on DFS (p = 0.032). The comparison of the immunohistochemical data between LT DFS, MT DFS and ST DFS in the tumor and stroma compartments is presented in Table 3.

Table 3 Comparison of the immunohistochemical data between LT DFS, MT DFS and ST DFS in the tumor epithelial and stromal compartments (p-values)

Analysis of overall survival in the studied cohort with Kaplan-Meyer curves revealed that a high stromal density of CD68 cells was a significant negative predictor of OS (p = 0.008). Figure 2. presents the Kaplan-Meier curve of overall survival (OS) in patients with laryngeal cancer stratified by CD68 density in the stromal compartment (blue line – low density of CD68 in the stroma; green line – high density of CD68 in stroma). The mean survival of laryngeal cancer patients with respect to the density of stromal CD68 cells is presented in Table 4. Analysis of the expression of the other antibodies on OS yielded no statistically significant differences.

Fig. 2

Kaplan-Meier curve of overall survival (OS) in patients with laryngeal cancer stratified by CD68 density in the stroma (blue line – low density of CD68 in the stroma; green line – high density of CD68 in the stroma)

Table 4 The mean survival of laryngeal cancer patients with respect to the density of stromal expression of CD68 cells

Association of immune markers with clinicopathological features

Comparison of the IHC expression of selected immune markers in both the stroma and the tumor epithelial compartment revealed no correlation with patient age, tumor grade or nodal metastases. However, we found a significantly greater stromal density of CD68 cells in tumors with T4a advancement than in those with T3 advancement (p = 0.038). Also, the comparison between III and IV stage advancement was close to significance level 0.05. Moreover, the density of CD68 cells in tumor cell nests was significantly greater in patients with perineural invasion (11.77 +/- 5.76) than in other patients (7.67 +/- 4.2) (p = 0.03). The density of CD68 cells in the stromal compartment was significantly greater for patients with cartilage infiltration (22.76 +/- 6.84 vs. 17.37 +/- 6.48 respectively) (p = 0.012). A summary of the immunohistochemical data comparing different clinical variables is presented in Table 5.

Table 5 Summary of the immunohistochemistry data comparing different clinical variables

Discussion

Laryngeal cancer TNM staging inadequately addresses the tumor features both for reliable prognosis and treatment matching. For many years, the pathological confirmation of perineural and/or vascular tumor infiltration has been recognized as an additional negative prognostic factor and has clinical implications for the implementation of adjuvant radiotherapy. Increasing evidence confirms that tumor microenvironment factors provide of essential information regarding tumor behavior, prognosis, and response to treatment. Tumor cells are antigenic and therefore induce an immune response. The composition and extent of the host immune response to tumor differ among solid malignancies, but in many cases, it has prognostic and predictive significance. Consequently, evaluating immune infiltrates in laryngeal cancer is important for identifying optimal biomarkers and further selecting patients who may benefit from specific immunotherapeutic options.

The scoring systems of tumor immune infiltrates used in research and proposed for different tumors vary widely in detail, scope, and resource requirements. Therefore, the International Immunooncology Biomarkers Working Group has proposed protocol to standardize the method of assessing tumor-infiltrating [16]. These guidelines recommend rating TILs separately in the tumor epithelial compartment and intratumoral stroma as a percentage of the compartment area infiltrated by TILs to the total compartment area. Although this semi-quantitative method was originally designed for the evaluation of routine H&E slides, it may also be used as a fast, accessible, and affordable way of assessing the presence and extent of specific IHC-stained immunocompetent cells involved in the tumor microenvironment immune response. As suggested, we reported our data in the form of continuous variable, without setting any thresholds to provide more biologically relevant information and allow more accurate statistical analyses. Our study presents immune markers data from advanced laryngeal cancer specimens according to these recommendations for a better understanding of the role of specific markers in determining outcomes and identifying factors relevant for prognosis and treatment options in the future.

Head and neck cancers differ not only in the etiology but also in molecular and immune profiles. HPV-positive oropharyngeal carcinoma is very distinctive from tobacco-induced cancers in the same location, but even more different from the other sites head and neck malignancies. However, the available literature on immune infiltrates in patients with LSCC is still limited. Therefore, in the following discussion we concentrated mostly on the data concerning laryngeal squamous cell carcinoma but also analysed known data about HNSCC in other locations.

The main finding of our study was the association of greater density of all analysed immune markers in resected laryngeal specimens from patients with a DFS shorter than 2 years. This finding contradicts the majority of published results thus far; however, this is not an entirely new observation. An analyses of tumor microenvironment by Sun et al. confirmed that the infiltration of a large number of regulatory T cells, dysplastic plasma cells, and macrophages in the LSCC tissue indicates an immunosuppressive state [18]. A recent study by Missale et al. also confirmed an association between an immune-rich TME and a poorer disease specific survival in LSCC patients [19]. De Virgilio et al. revealed the correlation between high density of specific TILs (CD3, CD4) and CD68 cells and an increased probability of lymph node metastasis in salivary gland cancers [20]. The negative prognostic impact of immune infiltrates was also confirmed in renal clear cell carcinoma [21]. Further data will be discussed with a focus on the role of selected immune cells in cancer progression.

The predominant component of the tumor microenvironment in solid tumors are tumor infiltrating lymphocytes, including CD3, CD4, and CD8 cells. Macrophages (CD68), natural killer cells (CD57), and dendritic cells constitute a relatively small part of the infiltrate [22]. The subtype of CD4 cells – T helpers, CD8 cells, NK cells and M1 macrophages protect the host against tumor development and progression [23]. Conversely, the subtypes of CD4 cells – the regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs) promote tumor invasion [23].

CD3 is present on T lymphocytes in all stages of their development, and as the receptor complex, it transduces the activating signal to the cytoplasm. Survival analysis of HNSCC patients treated with definitive chemoradiation by Balermpas et al. revealed that high CD3 and CD8 expression significantly improved the overall survival, progression free survival and distant metastasis free survival according to univariate analysis, but multivariate analysis revealed that the benefit was limited only to OS [24]. This study however included a predominant group of oropharyngeal cancers with only 5 laryngeal malignancies, four of which were scored as having low CD3. A study by Hoing et al. on LSCC confirmed the association of a high density of CD3 cells in the tumor core with a reduced metastasis rate [9]. Similar observations were provided by Karpathiou et al., who reported that a higher density of CD3, CD8, and CD57 cells was a positive prognostic factor in laryngeal and pharyngeal cancer patients (p = 0.02, 0.01, and 0.02, respectively) [10]. Accordingly, Zhang et al. reported improved OS and DFS in LSCC patients with a higher density of TILs (CD3, CD4, and CD8), specifically, a high density of CD3 cells was associated with better OS [25]. Tevetoğlu et al. confirmed that a high infiltration of CD3, CD20, and CD4 had a positive impact on disease-specific survival, disease-free survival, and recurrence-free survival in LSCC patient and overall survival was positively affected by high CD3 and CD4 infiltrations [26]. In contrast, higher CD3 cell density represented a risk factor for locoregional recurrence-free survival and was the only covariate independently associated with this outcome in the study by Missale et al. on LSCC [19]. Our present analysis provides no evidence for the prognostic accuracy of survival and outcomes for CD3 infiltrates in either the stroma or the tumor compartment. Notably, however, there was an increased density of CD3 cells in both compartments in the specimens from patients with early recurrence compared with those from patients without the relapse.

CD8 positive T lymphocytes, also called cytotoxic T cells, destroy targeted neoplastic cells via direct lysis. A low CD8 cells count was correlated with more aggressive disease with an advanced nodal stage of HNSCC; however the study by Balermpas et al. did not include laryngeal cancer patients [27]. Other studies on oropharyngeal and hypopharyngeal cancer confirmed that high infiltration of CD8 cells was associated with significantly better outcomes [28,29,30,31]. In line with the abovementioned reports, the study by Sánchez-Canteli et at. found high CD8 density infiltration in laryngeal and hypopharyngeal cancer favourable on DFS and OS [6]. Moreover, the multivariate analysis by Chatzopoulos et al. revealed that higher TIL density was correlated with a lower risk for relapse and death in the entire cohort, but high CD8 infiltration was not a significant independent prognostic marker in any analysis setting [32]. In contrast, Wolf et al.reported that a higher CD8 infiltrate density in oral cancer patients was associated with tumor recurrence [33]. The present evaluation of our cohort also revealed that a higher stromal density of CD8 cells is a significant marker of DFS in advanced laryngeal SCC patients. In summary, the current data concerning the infiltration of CD8 cells and their prognostic utility are contradictory and warrant further exploration. Both the anatomic subsite and the compartments of the tumor microenvironment could be responsible for differences in findings.

CD4 lymphocytes play crucial roles in initiating and maintaining anticancer immune responses. However, the impact of CD4 cells infiltration in tumors is even more controversial, because of the heterogeneity of these T cell populations, which currently include activated CD4 CD69 cells, regulatory CD4 Foxp3 cells, and mixed CD4 CD25 cells. A study by Badoual et al. in HNSCC patients confirmed a positive correlation between elevated levels of tumor infiltrating CD4 CD69 T cells and OS and locoregional control [34]. In the population of persistent or recurrent laryngeal squamous cell carcinoma patients treated with salvage surgery, Hoeseli et al. also reported that elevated CD4 and CD8 levels are significant factors associated with improved disease free survival [35]. In contrast, a study by Moreira et al. on oral SCC revealed that CD4 is a predictor of poor outcome, with high counts of CD4 cells associated with significantly decreased survival compared with patients with low cell counts [36]. Our findings did not reveal significant differences in CD4 cells density in either the stroma or the tumor with respect to the disease free survival period. A further analysis of the distribution of CD4 cells including subsets in LSCC is therefore necessary.

CD20 antigen is expressed on the surface of lymphocyte B cells throughout all stages of their development, and enables optimal B-cell immune response, specifically against lymphocyte T antigens. CD20 cells are infrequently included in TME studies. A dense profile of CD20 infiltration seems to be more relevant for the immune response in metastatic tissue than in primary tumors according to a study by Pretscher et al. on oropharyngeal and hypopharyngeal cancer [37]. The authors noted a statistically significant improvement in DFS in patients with large numbers of CD20 cells in the peritumoral lymphoid tissue in metastatic lymph nodes, but no effect was observed at the primary cancer site. Therefore, CD20 intratumoral cells may be more relevant in the interaction between the tumor and the humoral immune system in metastatic deposits in lymph nodes than at the primary tumor site. However, an investigation by Missale et al. on LSCC revealed that a population of CD20 cells in the tumoral area was associated with an increased risk of distant recurrence [19]. Our evaluation revealed no significant differences in CD20 density with respect to disease recurrence or the OS.

CD57 has been identified as a phenotypically stable marker of differentiated and highly cytotoxic NK cells responsible for the lysis of target cells without sufficient MHC class 1 expression on their surface. Therefore, CD57 cells have been identified as antitumour infiltrating lymphocytes. A meta-analysis by Bisheshar et al. investigated the correlation between CD57 and the known clinicopathologic features and their impact on prognosis [38]. They found one study confirming that high infiltration of CD57 cells was correlated with advanced tumor stage and neck metastasis, which are predictors of worse prognosis [39]. In contrast, three other included studies revealed a correlation between high CD57 cells counts and favourable clinical features, including lower locoregional advancement and a decreased incidence of disease recurrence and death [10, 40, 41]. A recent study by de Ruiter et al. reported that the presence of intratumoral CD57 infiltration did not correlate with prognosis in advanced stage, HPV-negative HNSCC patients treated with chemoradiotherapy [42]. These contradictory findings are even more questionable in the presence of our results confirming the significant association of a high stroma density of CD57 cells with disease recurrence.

Finally, the CD68 immune marker is widely expressed in tumor associated macrophages (TAMs). Currently we divide CD68 TAMs into two subtypes: classically activated antitumour/proinflammatory M1macrophages and alternatively activated protumour/anti-inflammatory M2 macrophages,- both of which play distinct roles in TME immune response. Interestingly, elevated levels of CD68 cells in tumor samples are correlated with an adverse prognosis in many solid tumors including glioblastoma, renal clear cell carcinoma, liver hepatocellular carcinoma, lung squamous cell carcinoma, and thyroid carcinoma [43]. A meta-analysis by Bisheshar et al. which included ten studies that investigated the relationship between CD68 TAMs and OS, revealed that a small number of CD68 cells were correlated with better OS [44]. Only four of the included studies separately evaluated the tumor and stromal compartments and both Kikuchi et al. and Lin et al. reported that a trend toward a low density of intratumoral CD68 TAMs was correlated with better OS [45, 46]. Compared with these data, our results confirmed high tumor CD68 cell density as a significant predictor of disease relapse and ST DFS and identified high stromal CD68 cell density as an individual negative predictor of OS. These data seem very promising for further exploration. Moreover, increased CD68 cell infiltration was also predictive of LSCC progression, revealing a significant correlation with tumor advancement and cartilage infiltration in the stromal compartment, and with perineural invasion of the tumor zone.

Research on the prognostic value of immune infiltrates in the TME of LSCC and other subsites of HNSCC has resulted in many discrepant findings [47]. For the same cancer primary site, the outcomes can be influenced by the type of selected treatment option, because surgery, radiotherapy and chemotherapy influence the immune response and the TME composition may differ by modality type. HNSCC should be investigated according to the primary tumor site. HPV negative and positive cancers should also be evaluated separately. Some of the results is obtained from the analysis of the material collected during tumor biopsy and the consistency with findings from specimens with complete tumor resection has not been confirmed thus far. It is important to standardize assessments performed by pathologists according to proposed recommendations with an emphasis on analysing both intratumoral compartments. The differences in findings can be explained by differences in tumor sites and the advancement-specific composition of the environment; however the phenomenon of cancer-associated immune exhaustion has not been thoroughly explored [48].

The main limitations of the present study are the small sample size and the retrospective nature of the analysis. In our study, we included only patients with advanced stage LSCC treated with primary surgery to avoid confounding results based on pretreatment chemo/radiotherapy implications for the composition of the tumor microenvironment, and to collect and analyse the uniform specimens from resected tumors. Moreover, our cohort was selected from distant years because a long observation time was necessary to reliably assess the outcomes, especially the LT DFS. This monocentric study had to be conducted at a feasible level for a wide assessment of immune markers with the primary aim of defining their predictive role in advanced LSCC. Furthermore, age is a morbidity factor for most cancers and was confirmed to be a significant factor for DFS in our study. The problem, however, is unsolved, and we still do not have evidence as to whether it is due to accumulated mutations in cells, a decreased efficiency of DNA repair mechanisms, or even more importantly, the impaired function of the immunological system.

Conclusions

The analysis of advanced LSCC samples revealed an association between greater density of all analysed immune markers in both the tumor and stromal compartment in ST DFS patients (less than 2 years). We found no considerable evidence or prognostic accuracy on survival and outcomes for patients with CD3, CD4, or CD20 lymphocyte infiltrates in either the stroma or the tumor compartment. A greater stromal density of CD8 cells and CD 57 cells was identified as a significant markers of DFS in advanced LSCC patients. A high tumor CD68 cells density was confirmed to be a significant predictor of disease relapse and ST DFS. Higher stromal CD68 density was identified as an individual negative predictor of OS. Moreover, high CD68 infiltration was shown to be predictive of LSCC progression, revealing a significant correlation with tumor advancement and cartilage infiltration in the stromal compartment, and with perineural invasion in the tumor epithelial compartment.