Oral squamous cell carcinoma (OSCC) is one of the most common types of malignant head and neck cancer (1). Generally, patients with head and neck squamous cell carcinoma (HNSCC), including OSCC, have advanced disease at the time of the initial diagnosis (2). Despite the advent of chemotherapy, radiotherapy, and multimodal therapy, the 5-year overall survival (OS) rate of HNSCC has remained approximately 50% for the past 30 years (3). Currently, there are no reliable clinical biomarkers that can be used to stratify HNSCC patients according to their prognosis (3).

Treatment planning for OSCC is generally based on disease staging; i.e., clinical assessment of the tumor, lymph nodes, and distant metastases (TNM staging) (4). Histopathological prognostic markers are mainly associated with the properties of cancer cells (e.g., the degree of differentiation, nuclear pleomorphism, and mitotic activity). However, tumor growth also depends on the stroma surrounding the tumor and the tumor microenvironment (TME), which plays an important role in cancer progression (5). The tumor stroma basically consists of the non-malignant cells in the TME, including cancer-associated fibroblasts (CAF), innate and adaptive immune cells, microvessels, and extracellular matrix (ECM) cells (6). The tumor stroma is formed by a complex process induced by tumor-host interactions, and studies have shown that the histological features of the tumor stroma may serve as independent prognostic factors in various solid tumors (7).

Although several studies have reported the prognostic value of stromal features in OSCC, such as the tumor-stromal ratio (8, 9) and the number of CAF (alpha-smooth muscle actin-positive fibroblasts) (10) no previous studies have categorized stromal pattern (SP) by analyzing both stromal desmoplasia and immune/inflammatory cell reactions in an integrated manner.

Here, we propose a new method for categorizing SP in OSCC. This categorization system is largely based on Ueno's criteria (11), but we modified it by adding a fourth category; i.e., the immune/inflammatory type. This modification was made because we observed that a subset of OSCC showed characteristic immune/inflammatory cell stromal reactions in the absence of marked desmoplasia. We show here that the categories in this new system are closely associated with various clinicopathological parameters, such as the histological grade, lymphovascular invasion, neural invasion, and a diffuse invasion pattern, and represent an independent prognostic factor for OSCC.

A desmoplastic reaction (DR) is defined by the growth of fibrous or connective tissue at sites of stromal invasion by cancer (15). The presence of a DR has been reported to be associated with poor outcomes in several types of cancer, such as those of the colon, pancreas, neuroendocrine organs, and skin (11, 16–18). On the basis of the distinctive histological features of hematoxylin and eosin (HE)-stained desmoplastic stromal tissue, Ueno et al. divided the DR of colon cancer into three groups; i.e., into mature, intermediate, and immature types (19). This DR categorization system provided prognostic information about colorectal cancer patients; i.e., the 5-year survival rate was 80% for the mature type, 55% for the intermediate type, and 27% for the immature type (19).

In addition to DR, immune/inflammatory cells, such as granulocytes, lymphocytes, and macrophages, also play a key role in the TME (20). These cells are involved in various immune responses and defensive activities against tumor cells. Therefore, the immune system in the TME may influence patient prognosis and the treatment response (20). Previous studies have shown that tumor-infiltrating immune cells are closely associated with the prognoses of colorectal, renal, and oral cancers (21–23).

In the present study, we classified the SP of OSCC into immune/inflammatory, mature, intermediate, and immature types, based on the findings of HE-stained slides. We examined the utility of this new SP classification by analyzing its clinicopathological correlations and prognostic value.

One main finding of this study was that the immature type was associated with aggressive tumor features, such as poor differentiation (grade 3), lymphovascular or neural invasion, and an infiltrative invasive pattern (YK-4C or YK-4D), and patient survival.

A myxoid stroma has also been reported to be associated with aggressive tumor histological features and a poor prognosis in colorectal cancer (11, 24, 25). Ueno et al. classified the DR of colorectal cancer into 3 types (mature, intermediate, and immature). The immature type was found to be associated with the dedifferentiation of tumor cells; hypovascularity; a restricted distribution of immune cells; and abundant tumor-supporting extracellular components, including fibronectin and tenacin (24, 25). The prognostic power of this DR categorization system for stratifying 5-year RFS was greater than that of any other conventional prognostic factor, such as the TNM stage, venous invasion, or tumor grade (11, 24).

A DR may represent the histological consequences of ECM remodeling generated by CAF (15). In the TME, CAF play a critical role in mediating the epithelial-mesenchymal transition (EMT) of carcinoma cells (6). EMT is a complex and reversible biological process through which epithelial tumors change from a polar, adhesive phenotype to a mesenchymal phenotype characterized by increased cell migration, invasive potential, cytoskeletal remodeling, and resistance to apoptosis (26). Ueno et al. speculated that keloid-like collagen and a myxoid stroma may represent histological features that are associated with the induction of EMT in tumor cells (27).

In this study, we modified the DR classification described by Ueno et al. by adding the immune/inflammatory type, which is characterized by diffuse immune/inflammatory cell infiltration and inconspicuous desmoplasia at the invasive front, as a new category. We showed that the immune/inflammatory type was associated with shallow invasion, early stage disease, the absence of lymphovascular or neural invasion, and expansive-type invasion. These results are in line with the findings obtained for cutaneous SCC (18). In cutaneous SCC, the presence of a myxoid stroma was found to be associated with a lack of inflammatory and immune reactions, perineural invasion, tumor depth, and metastatic risk (18). The authors speculated that the absence of an immune response in an immature myxoid stroma facilitates the metastatic process (18). In our study, immune/inflammatory cells were largely lymphoid cells, as far as we could identify on HE-stained sections. As infiltration of immune cells may represent host reaction to combat malignancy, it would make sense that the immune/inflammatory type was associated with favorable histologic features and better prognosis. Results supporting this notion have recently been reported by Troniano et al. These authors demonstrated that paucity of immune reaction predicted poor prognosis in tongue squamous cell carcinoma (28).

The existence of an immune/inflammatory subtype is supported by recent studies of carcinoma of the lung, breast, or bladder based on RNA sequencing or proteomic data (21–23, 29–31). In breast cancer, an immunohistochemistry-based classification exhibited substantial agreement with an mRNA expression-based classification (30). The immunomodulatory type, which is characterized by the infiltration of CD8-positive T cells into the tumor parenchyma, showed a better prognosis than the other subtypes and relatively high expression of immune checkpoint molecules, such as cytotoxic T-lymphocyte-associated protein 4 (CTLA4), programmed death ligand 1 (PD-L1), indoleamine 2, 3-dioxygenase 1 (IDO1), and programmed cell death protein 1 (PD-1) (30). In bladder cancer, the inflamed type demonstrated the highest levels of many immune cell types, including cytotoxic lymphocytes, T cells, and the immune regulators nuclear factor of activated T-cells, cytoplasmic 2 (NFATC2) and signal transducer and activator of transcription 4 (STAT4), while the MYC-driven subtypes, including the Myc type and early type, exhibited repressed immune responses and worse prognoses than the inflamed type (31).

This study had several limitations. First, the number of cases was smaller than those in previous studies of colon cancer (11). Second, immature type was strongly correlated with advanced cases, neural invasion, and YK 4C, 4D. Although our multivariate analysis indicated an independent predictive value of stromal pattern in our cohort, the close associations among multiple factors may make it difficult to firmly establish that SP is a truely independent prognostic factor. A larger number of cases from multiple facilities might be required to confirm the current findings. Third, the precise molecular mechanisms by which the SP contributes to the TME in OSCC remain unknown.

The molecules involved in the interactions between tumor cells and the adjacent fibroblasts may include Wnt/beta-catenin, transforming growth factor beta, and C-X-C chemokine receptor type 4 (CXCR4) (24, 27, 32–34). Recently, Ueno et al. showed that both the mRNA and protein expression levels of periostin were high in the immature type of colon cancer (35). Periostin has been shown to be an important regulator of bone and tooth formation and maintenance and EMT (36). Ogawa et al. identified three distinct stromal types that influence survival, immunity, and molecular features in pancreatic cancer (37). Further studies are warranted to elucidate the molecular mechanisms underlying each type of SP and their effects on the TME.

In conclusion, in OSCC the SP is associated with prognosis and may predict tumor aggressiveness. Our SP categorization method may be an excellent tool for aiding decision-making regarding treatment options in precision medicine and provide a histological basis for further clinical and molecular research.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Jichi Medical University. The Ethics Committee waived the requirement of written informed consent for participation. Written informed consent was not obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article. There are no potentially identifiable data or images are presented in the article.

YA: conceptualization, investigation, and writing original draft. AK and TM: prepared the tissue samples and validated the data. MH: prepared the tissue samples, validated the data, and performed histological evaluation. DM: validated the data. NF, HN, and YM: collected the clinical records. TN: resources, supervision, writing review, and editing. All authors critically reviewed and approved the manuscript.

This work was supported in part by Japan Society for the Promotion of Science KAKENHI Grant Number 19K10094 (to YA).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.