Immunophenotypic and Gene Expression Analyses of the Inflammatory Microenvironment in High-Grade Oral Epithelial Dysplasia and Oral Lichen Planus

Oral lichen planus (OLP) is a chronic inflammatory condition frequently affecting the oral cavity. Although multiple agents have been previously associated with OLP, no specific etiology is currently described as the cause. The persistent issue in the diagnostic separation between OLP and oral epithelial dysplasia (OED) is the considerable clinical and histologic overlap between the two entities in the oral cavity [1]. There are still no widely accepted diagnostic criteria for OLP. Microscopic differentiation is based upon subjective interpretation of the microscopic presentation by the examining pathologist and the clinical presentation. Although direct immunofluorescence (DIF) studies can help diagnose OLP, DIF has to be correlated with the appropriate clinical presentation and permanent slides, as basement membrane fibrinogen deposition is not pathognomonic of OLP and has been described in OED specimens and other mimickers [2, 3]. Nevertheless, distinguishing between OED with chronic interface mucositis and OLP with reactive cellular atypia can be challenging, necessitating a subjective evaluation of the present morphologic and cytologic features [4].

Oral squamous cell carcinoma (OSCC) is believed to be preceded by a precancerous lesion, mainly defined by clinical visual and histopathologic examination. Based on the presence of OED, these lesions have been grouped under the umbrella term “Oral Potentially Malignant Disorders” (OPMDs). These represent a heterogeneous group of clinical mucosal lesions that may have an increased likelihood of transforming into OSCC [5]. On microscopic examination, the presence and severity of epithelial dysplasia within such lesions are graded from low to high depending on the extent of intraepithelial abnormalities and carry a reported overall risk of malignant transformation of up to 36% [6‐8]. Although the presence of dysplasia in biopsied material is a reliable predictor of transformation into OSCC, multiple studies have reported between 0 and 28% of malignant transformations arising from mucosal lesions that do not previously show OED during microscopic examination [7, 9‐12].

Other studies have previously shown a direct link between inflammation and malignant transformation [13‐15]. The progressive and increasing inflammatory infiltrate density in the microenvironment, alongside qualitative changes in distinct cell populations in OPMD and OSCC, suggests that reversing the environment in OPMD could unveil anti-lesion immunity, leading to immune regression before the onset of OSCC [13, 16].

Evidence indicates that the initial phases of carcinogenesis are linked to alterations in the immune response within the microenvironment [14, 17]. We previously reported a direct link between the presence and distribution of CD4 + and CD8 + lymphocytes and high-grade OED [18]. However, tumor-associated macrophages (TAM) have also been reported to play a role in early carcinogenesis, tumor progression, and later metastasis [19].

Several recent studies have shown increased PD-1/PD-L1 cells in OSCC and OPMD, particularly lesions that progressed to oral cancer [15, 20]. However, accurate interpretation of the immunohistochemical stains can be challenging for the practicing pathologist [21, 22].

Our study hypothesizes that OPMD-induced inflammation can be a surrogate marker to diagnostically differentiate between OED and OLP and to predict malignant transformation. Utilizing immunofluorescence/immunohistochemistry (IF/IHC), we analyzed the presence and distribution of CD4 + and CD8 + infiltrating lymphocytes, CD163/STAT1 + macrophages, and expression of PD-1/PDL-1 in the epithelial compartment of biopsy samples previously diagnosed with OLP and OED. We also performed bulk-RNA sequencing on all the samples and compared the immune genetic expression between OED and OLP.

The Institutional Review Boards of the University of North Carolina at Chapel Hill (UNC-CH) and East Carolina University (ECU) registered, reviewed, and approved the study protocol, with the numbers 16-5988 and UMCIRB-21-001256, respectively. All study procedures followed the Declaration of Helsinki and Research Committee Regulations. Oral tissue samples received between July 1, 2005, and January 31, 2022, at the UNC-CH and ECU Oral & Maxillofacial Pathology biopsy tissue archives that were diagnosed by board certified Oral and Maxillofacial Pathologists with oral lichen planus (OLP) and moderate to severe (high-grade) epithelial dysplasia (OED), and met the inclusion criteria, were collected [23]. Biopsy accession files, dental and medical records, and available sample follow-up information were reviewed and collected for each sample. For our study, 10 cases of OED and 10 cases of OLP were identified and collected. The specimens included cases with sufficient tissue for additional analysis from the tongue, buccal mucosa, or floor of the mouth. Tissue samples from patients with a previous history of cancer, prior treatment with steroids and/or immunosuppressed status, subjects under 21 years of age, and patients with a prior history of radiation to the head and neck were excluded. The hematoxylin and eosin (H&E) slides were retrieved and reviewed, and the diagnosis was confirmed by multiple observer consensus. All cases selected for the OLP group were diagnosed according to the diagnostic guidelines published by Cheng et al., currently the American Academy of Oral and Maxillofacial Pathology position paper regarding the diagnosis of OLP, which include patients with intraoral widespread and symmetrical oral lesions, clinical correlation with medication, allergy and chronic use of cinnamon-containing products was ruled out, and incisional biopsies were performed with consistent microscopic and direct immunofluorescence findings [24]. The follow-up information was recorded with the available surgical pathology records with microscopic diagnoses in the same location and respective diagnosis (e.g., invasive carcinoma) or clinical information seen in patient records.

OPMDs encompass a heterogeneous group of lesions and conditions that exhibit an increased risk of transforming into oral cancer. The global incidence of OPMDs varies across populations, with factors such as age, gender, tobacco and alcohol consumption, and socioeconomic status contributing to regional disparities. Studies have consistently shown a higher prevalence of OPMDs in older individuals, especially those engaging in tobacco and alcohol use [34]. Various disorders may exhibit clinical and histopathological traits akin to OED and OLP [35, 36]. While OED typically manifests as a solitary lesion with varying proportions of white and red color changes and ulceration, a multifocal presentation is well-recognized, particularly in patients with advanced tobacco-related mucosal injury and proliferative verrucous leukoplakia (PVL) [37].

A band-like chronic inflammatory cell infiltrate in the superficial lamina propria commonly accompanies OED. When viewed under low-magnification microscopy, this can closely resemble the histopathological features of OLP [38]. Substantial clinical and histologic overlap has been observed between OLP and OED of the oral cavity, particularly in cases of PVL. This raises questions about whether purported transformations of oral lichen planus to squamous cell carcinoma represent undetected premalignant lesions with a nonspecific inflammatory response mimicking OLP [9, 39].

The infiltration of various immune cells, including lymphocytes, macrophages, and dendritic cells, mark the immune microenvironment of OED. These cells play a pivotal role in the surveillance and response to dysplastic changes. Studies have demonstrated an increased density of CD8 + cytotoxic T cells in OED lesions, suggesting their involvement in antitumor immunity [40]. Additionally, regulatory Tregs have been noted, indicating a complex balance between effector and suppressor immune responses [16]. The immune microenvironment of OED reflects a dynamic struggle between immunosurveillance and immune evasion mechanisms. While an influx of immune cells indicates an attempt to eliminate dysplastic cells, OED lesions often employ immune evasion strategies, including upregulation of programmed death-ligand 1 (PD-L1) on dysplastic epithelial cells, dampening the antitumor immune response [41].

Very few studies have focused on identifying molecular alterations in the immune microenvironment surrounding OPMD. The abundance and composition of intraepithelial and underlying immune cells in OPMD may play a role in predicting cancer progression, prognosis, and possible immunotherapy as a preventive measure. Hence, the genomic and molecular profile of said immune intraepithelial cells is still needed. Regarding the epithelial compartment, cytogenetic alterations such as copy number gain of 16q, 8q, and loss of 3p, 8p, 9p, 4q, 5q, and 13q are biomarkers for premalignant oral lesions [42]. Also, protein overexpression and mutation of the p53 gene have been explored in OPMD [43, 44]. However, these molecular alterations have yet to be translated into clinical utility because their effectiveness as a biomarker is limited [45, 46], as chemoprevention may prove ineffective when the target cell population carries a significantly high risk or has already undergone preneoplastic irreversible cellular changes [46, 47]. Hence, exploring the relationship between epithelial dysplastic changes and the immune response profile in the microenvironment setting for planning a targeted treatment can be useful for diagnostic and, later, management purposes.

Our study only analyzed immune signatures in both study groups. Gene Ontology studies provide a systematic and standardized framework for annotating and categorizing genes, facilitating a deeper understanding of gene function, biological processes, and their implications in health and disease. According to our GO analysis, aberrant expression in OED was mainly associated with immune signatures associated with immunosurveillance, lymphocyte infiltration, cytotoxic response, and surrogate markers of tumor-associated macrophages. These results are in accordance with the existing bulk RNA-sequencing studies that have facilitated the identification of specific immune cell populations infiltrating OED lesions. Notably, increased expression of genes associated with cytotoxic T lymphocytes, such as CD8A and granzyme B (GZMB), has been observed in previous studies [41, 48].

The top upregulated genes were NKG7, MS4A4A, CD8A, MST1R, and NCR3LG1. CD8A encodes the CD8 alpha chain, a cell surface glycoprotein expressed on cytotoxic T cells. It plays a key role in antigen recognition and the immune response against infected or abnormal cells [49]. MST1R encodes the receptor for macrophage-stimulating protein, and it is involved in various cellular processes, including cell survival, migration, and immune responses [50]. Similar findings have been reported upregulated in similar studies in high-grade dysplasia or early OSCC [48, 51, 52]. These findings were also corroborated by the increased intraepithelial infiltration of CD8 + lymphocytes and CD163/STAT1 + macrophages by IHC/DF in all cases of high-grade OED versus OLP (Fig. 1). This suggests an active involvement of cytotoxic T cells in the immune response against dysplastic cells in OED. Interestingly, the infiltration of said immune cells follows the dysplastic changes within the intraepithelial tissue. STAT is an IFN-inducible product, but specifically, the activation of STAT1 in TAM might lead to the up-regulation of inducible nitric oxide synthase and Arginase I activity. Nitric oxide synthase and Arginase could play a role in suppressing T cells. This is why STAT1 was chosen as an antibody for identifying TAM in our study [53, 54].

Upregulation of NKG7 and MST1R also indicates a cytotoxic nature of the immune microenvironment elicited by the premalignant changes occurring in the epithelium. Hence, we can consider that the RNA-sequencing data may have originated from T- and NK-enriched parts rather than all of the OED tissues. This makes sense, as cytotoxic cells will eliminate keratinocytes undergoing malignant changes in the epithelial compartment. Moreover, the coexpression of CD163 and STAT1 in macrophages in OED suggests a pro-inflammatory phenotype; hence, it is likely that infiltrated CD4 + cells modulate the phenotype of associated macrophages in oral premalignant lesions.

It is also important to highlight that in a small portion of the OED cases when taken together, immune signatures reported as being involved in immune suppressive mechanisms, such as Treg cells and PD1/PDL1, were enriched. However, no significant overexpression of PD1/PDL1 was identified by IHC/DIF, nor clinical correlation with malignant transformation, when compared with available follow-up information.

The most common downregulated genes in OED were TRADD, CX3CL1, and ILI24. While there might not be specific studies on the downregulation of these genes in OED, the general understanding of their roles in cancer suggests that their downregulation could contribute to dysplastic changes in oral epithelial cells. Dysregulation of apoptosis and inflammation are key features in the progression of precancerous lesions like OED. The TNFRSF1A-associated death domain or TRADD gene mediates cellular responses to tumor necrosis factor receptor 1 (TNFR1) signaling. While direct studies on the downregulation of TRADD in carcinogenesis may be limited, alterations in TNFR1 signaling pathways, including TRADD, have been implicated in various malignant neoplasms. Some studies suggest that dysregulation of TNFR1 signaling, which includes TRADD, can contribute to cancer development and progression [55]. CX3CL1 also plays a role in the recruitment and activation of immune cells within the tumor microenvironment. Its downregulation can contribute to an immunosuppressive milieu, allowing tumors to evade immune surveillance. Reduced CX3CL1 expression has been associated with enhanced tumor invasion and metastasis. The interaction between CX3CL1 and CX3CR1 regulates the adhesion of immune cells to tumor cells, influencing the metastatic potential of cancer [56, 57]. Interleukin 24 (IL24) has been extensively implicated in carcinogenesis, affecting various aspects of cancer development. Downregulation of IL24 has been associated with multiple cancer types, and its tumor-suppressive properties are attributed to its ability to induce apoptosis and inhibit tumor cell growth [58].

In contrast for OLP, our data revealed differential gene expression mainly related to lymphocyte infiltration, T-cell regulation, and cytotoxicity. Reported findings consistently reveal gene alterations associated with immune responses, inflammation, and epithelial integrity [59]. Similar studies have also found dysregulation in signaling pathways related to inflammation, immune responses, and cell adhesion [60]. The top upregulated genes in the OLP group were TCIRG1, DCG1, and NRIP1.

All these genes regulate immune responses, particularly in T-cell and dendritic cell function. While direct evidence of upregulation in OLP may not be readily available, the involvement of immune regulatory genes like TCIRG1 and DCG1 in their function in immune regulation suggests a possible role in the inflammatory processes associated with OLP. However, specific studies on its upregulation in OLP are limited [60].

Our limitations in this pilot study included the retrospective nature of the project and limited research funds, which prevented the inclusion of more cases in both study groups, as well as low-grade dysplasia and invasive OSCC groups. Another limitation was the lack of follow-up information or longer follow-up time since the initial microscopic diagnosis to assess malignant transformation in many cases. Many of the patients in the OED group were referred to outside institutions for definitive treatment after the incisional biopsy and diagnosis were made; hence, we did not have access to this information.

The immune microenvironment of high-grade OED is a dynamic and intricate network involving immune cells, cytokines, and inflammatory mediators. Elucidating these interactions is crucial for unraveling OED pathogenesis, predicting malignant transformation, and developing targeted therapeutic strategies. Our preliminary data suggests that the inflammatory microenvironments associated with OED and OLP differ; as such, they could be used for objective and accurate diagnostic separation during microscopic examination in the future. Further research is necessary, including analyzing more cases (comprising hyperkeratosis/reactive mucosal lesions, low-grade OED, and invasive OSCC) and spatial transcriptomics.