Head and neck squamous cell carcinoma (HNSCC) is a heterogeneous disease that may develop in large carcinogen-induced premalignant fields of genetically altered cells. Field cancerization was first observed and described by Slaughter et al. in 1953 (1). At that time, the investigators concluded that the mucosa of the upper aerodigestive tract had undergone changes which were likely due to continued carcinogen exposure and could therefore become more susceptible to the development of multiple foci of malignant transformation (2). At its core, HNSCC evolution involves a multifaceted interplay of genetic mutations and interactions with the local microenvironment. It is well established that the process of tumor evolution progresses through stages of premalignancy wherein cells acquire abnormal geno- and phenotypes that increase their propensity for malignant transformation. Clonally unrelated premalignant cells in a cancerized field can often be found distant from each other and contribute to local disease recurrence or the development of a second primary tumor in the upper aerodigestive tract (3–9). However, the exact order and timing of these genetic changes remains to be determined for HNSCC pathogenesis. Frequent alterations in gene copy number, for example, are a hallmark of advanced cancers (10). Normal epithelium on the other hand displays very rarely copy number alterations (11) but may instead harbor numerous somatic mutations (12–14). Much less is known about the process of transformation from normal epithelium to premalignant lesions and subsequent evolution into malignant tumors. Evidence suggests that copy number alterations may play a key role in driving early malignancy (11, 15, 16). This concept adds another layer of complexity to the understanding of how malignancies develop in the upper aerodigestive tract and is particularly significant in the context of HNSCC, as it underscores the importance of not only treating the primary tumor but also monitoring and addressing potential premalignant changes in the surrounding tissue. Understanding the biology of premalignancy is therefore a critical enabler to developing the right approaches to intercepting the disease. Yet knowing which individuals with premalignancies will progress and develop clinically significant cancers during their lifetimes remains one of the most important challenges of reducing HNSCC morbidity and mortality. Simultaneously, it has become increasingly evident that the progression of tumors and the immune responses are profoundly interconnected, extending beyond genetic mutations and cellular transformations (17–19). Recent advancements in immune checkpoint therapy have demonstrated improved overall survival in a proportion of patients with recurrent/metastatic HNSCC (20–24). Given the impact of immunotherapy in advanced disease, modulation of the immune microenvironment might prove beneficial as a prevention strategy in patients with high-risk premalignant mucosal lesions. Available data indicates that immune surveillance may play an important role in preventing the transition of premalignancy to cancer (25–27). However, the extent to which an active immune system influences the genome evolution of a tumor (or vice versa) remains unknown, particularly with respect to the extensive genomic and transcriptomic heterogeneity within a tumor and its surrounding microenvironment (28, 29).

Here, we retrospectively analyzed archived FFPE tissue samples from primary and recurrent tumors as well as matched premalignant and normal squamous mucosa from patients with p16-negative HNSCC. We used targeted panel Next Generation Sequencing (NGS) and shallow whole genome sequencing (WGS) on a selected number of tissues to profile the mutational landscape of multi-regionally and sequentially collected samples. Additional patients’ samples were analyzed with the Nanostring PanCancer Immune Panel to explore the mechanisms underlying immune evasion. The aims of this study were to provide a characterization of the preconditioned squamous mucosa in a cancerized field and their corresponding primary and, if available, recurrent tumors to shed light into their genomic, immune microenvironmental and evolutionary relationship. We confirm that genomic alterations, along with dysregulation in immune-related gene expression, are detectable at an early stage in the development of upper aerodigestive tract tumors. These observations underscore the potential for targeting the mechanisms involved in the early stages of disease progression, offering new avenues for therapeutic intervention.

In this study, we have conducted a focused analysis of the genome and immune microenvironment from multiple, matched normal squamous tissues, premalignant lesions, primary, and recurrent tumors from seven patients with p16-negative HNSCC with the aim to identify novel molecular mechanisms that could drive tumor evolution in the upper aerodigestive tract. The unique aspect of this study is the comprehensive profiling of multi-regionally and sequentially collected and well annotated samples at different disease stages, which also included a focus on the role of the immune microenvironment.

Targeted analysis of genetic alterations revealed a characteristic mutational pattern in HNSCC, akin to findings reported in previous studies (39–41). The most prevalent genetic mutations observed in our cohort were in the TP53 and NOTCH1 genes, followed by mutations in CDKN2A, CREBBP, NOTCH3, CTNNB1, FGFR3, and FGFR4, among others. A notable proportion of mutations were exclusive to dysplasia and carcinoma, respectively, although approximately 20% were shared between these two stages. Gene mutations rated as (likely) pathogenic that were already identified in low-grade and moderate dysplastic samples included inactivating mutations in the tumor suppressor genes TP53, CDKN2A, and NOTCH1, implying a crucial role in stepwise carcinogenesis towards HNSCC. These results corroborate the empirical progression model of Califano et al., where TP53 is disrupted early (5, 42). In that model, the earliest event of disease progression includes loss of chromosome arm 9p, which harbors CDKN2A (43), progressing to loss of chromosome arms 17p, where TP53 resides, and 3p. In our cohort, in earlier grades of dysplasia, individual genes were present at low prevalence (<15% VAF). This may support an early role in tumorigenesis as suggested by Leshchiner et al. who inferred genetic progression from exome sequencing of primary HNSCC tumors (44). While the role of TP53 as tumor suppressor gene in HNSCC is well established and different mutations could be linked to increased malignancy (45) and therapy-resistance (46), the functional significance of NOTCH1 mutations is less clear (47). NOTCH1 and NOTCH3 belong to the family of cell surface receptors that transduce juxtacrine signals of delta-like canonical Notch ligands and jagged canonical Notch ligands from adjacent cells, with significant roles in directing tissue commitment and cell differentiation (48). Loss of function of NOTCH1 has been linked to the acquisition of stem-like properties and a more aggressive phenotype. However, tumor-suppressing mechanisms of NOTCH1 signaling are not entirely understood (49). High expression of Notch3 on the other hand has been linked to poorer prognosis in SCC (50). Observed loss of 9p and therefore CDKN2A with subsequent cell cycle dysregulation as an early event in the carcinogenesis are in line with previous findings (51). The function of tumor suppressors is inherently difficult to restore. However, a recent study could show meaningful response rates to selective CDK4/6 inhibitors in CDKN2A-altered HNSCC, proving the clinical significance of CDKN2A mutations (52). Whether this approach could also target premalignant lesions, needs to be established. Recent work has demonstrated a role for chromosomal instability driving selection in cancer evolution with early somatic copy number alterations potentially being implicated in tumorigenesis (53). Our results also support an increasing genomic instability over time. Case 5 showed a high rate of chromosomal alterations even in moderate dysplasia, including loss of 3p and 9p, that was followed by multiple disease recurrences. Similarly, case 6 demonstrated progressive loss of 3p with subsequent local tumor recurrences. Previous studies have shown that dysplastic lesions that harbor 3p loss are 33 times more likely to progress to invasive carcinomas than those without (54) and that 3p deletion may be associated with poor disease outcomes (55). Furthermore, high-grade dysplasia showed a similarly rearranged genome as was seen in HNSCC, confirming its behavior like an invasive carcinoma. The existence of these multilocular distinct premalignant conditions therefore presents the need and opportunity to define which are at greatest risk for progression, including if, how, and when to intervene.

A recent comprehensive study on early-stage untreated lung cancers from the TRACERx cohort demonstrated an interplay between the immune and genomic landscapes through copy number loss as a mechanism of subclonal immunoediting. Immunoediting mechanisms may impact on tumor evolution, affecting either antigen presentation or neo-antigenic mutations both at DNA and RNA level with chromosomal instability driving loss of neo-antigens (28). Other studies were able to establish a link between somatic copy number alterations and markers of immune evasion, with highly aneuploid tumors being less responsive to immunotherapy (56). In HNSCC, cumulative loss of chromosome arm 9p has been shown to be the strongest driver of immune evasion. Simultaneously, copy number-defined high-risk oral preinvasive and early invasive lesions were immunogenic, suggesting a possible clinical benefit of augmenting immune surveillance (57). Nevertheless, the exact role of the immune microenvironment in the early stages of cancer development remains largely unknown, and particularly how survival and clonal expansion of premalignant cells may be fostered by an immunosuppressive microenvironment that hinders immune surveillance remains elusive. In our study, pathways associated with the interferon alpha and gamma response exhibited heightened activity, even in moderate dysplastic lesions, with a progressively amplified expression in higher dysplasia grades and in carcinoma. Interferon gamma has been linked to antitumor-immune response and increased interferon gamma signaling has been proposed as a prediction marker for response to immune checkpoint therapy in HNSCC (58). In contrast, low-dose interferon gamma might lead to tumor progression (59). Early upregulation of this signaling pathway in dysplasia might indicate an increased, yet insufficient, immune response. Likewise, interferon alpha signaling contributes to antitumor activity in early stages and has been shown to promote tumor growth in equilibrium and escape phase of cancer development (60). It can therefore be speculated, whether immunomodulatory therapy targeting these pathways in dysplasia (in combination with immune checkpoint therapy) could have beneficial effects. Interestingly, a recent study could show that immune checkpoint inhibition can achieve regression in high-risk oral leukoplakia (61). Moreover, progression to invasive carcinoma could be linked to 9p21.3 loss, encompassing the type I interferon gene cluster and CDKN2A, pointing towards a crucial role in therapy resistance and carcinogenesis.

Single gene expression analysis revealed a few markers differentially expressed in different grades of dysplasia compared to HNSCC. One candidate gene that was upregulated in moderate dysplastic compared to low-grade dysplastic samples with highest expression in HNSCC was secreted phosphoprotein 1 (SPP1), also known as osteopontin (OPN). SPP1, an integrin-binding glycophosphoprotein, has been shown to be associated with immune cell infiltration and found to be upregulated in multiple cancers, including HNSCC (62). High SPP1 expression has previously been shown to be associated with poor overall survival in colorectal cancer, with increasing expression from normal tissue to primary tumors and ultimately to metastases (63). Indoleamine 2,3-dioxygenase 1 (IDO1), another significantly differentially expressed gene that we found to be upregulated from benign to malignant phenotype, is a rate-limiting enzyme in the conversion of the essential amino acid tryptophan to kynurenine. IDO1 is overexpressed in various tumor types and associated with worse overall survival (64). It has been shown to be involved in immunosuppression in the tumor microenvironment through increased tryptophan metabolism, thereby leading to suppression of antitumoral T-cells (65–68). IDO1 inhibitors that could restore the anti-tumoral function of T-cells and shift the tumor microenvironment from immunosuppressive to immunogenic are currently under investigation in various clinical trial settings (69, 70). Interleukin 6 (IL6), a pleiotropic cytokine, which is an important player in several cellular processes such as proliferation, immune regulation, inflammation, and invasion, has been shown to be upregulated in various cancer types and also predicts poor clinical outcomes in patients with HNSCC (71–74). In our study, we could already show an increase in IL6 expression in moderate and high-grade dysplastic samples. In accordance with the literature, we also found lymphocyte activation gene-3 (LAG3) overexpressed in HNSCC samples. LAG3 is an immune checkpoint control protein that negatively regulates T-cell and immune response (75). Here, our results demonstrated that LAG3 upregulation could already be observed in moderate dysplasia. Previously, Choi et al. have demonstrated a stepwise alteration of T-cell repertoires during HNSCC progression, implying changes in the immune microenvironment before malignant invasion (76). In the case of serine peptidase inhibitor Kazal type 5 (SPINK5), its downregulation was even evident in low-grade dysplastic lesions, signifying an early deactivation event in the disease’s evolutionary trajectory. SPINK5 encodes for a serine protease inhibitor, which plays a crucial role in regulating various cellular processes, including tissue remodeling and inflammation. In HNSCC, SPINK5 has emerged as a critical player, often serving as a tumor suppressor gene (77). This suggests that SPINK5’s reduced expression may occur early in the development of HNSCC, potentially contributing to the initiation and progression of the disease.

Our study has several limitations, primarily stemming from the relatively small sample size. We acknowledge that drawing comprehensive conclusions from a cohort of just seven patients is inherently limited. However, we believe that our study adds valuable insights to the limited body of research that has explored the relationship of the genetic landscape with immune microenvironmental changes in p16-negative HNSCC and matching premalignant and normal squamous mucosa. Additionally, the absence of sequencing data from matched peripheral blood lymphocytes is a constraint, as it hampers the ability to filter out variants originating from clonal hematopoiesis of indeterminate potential (78). Nonetheless, we made substantial efforts to address this issue through rigorous filtering methods, including cross-referencing with the dbSNP database. One notable strength of our study lies in the detailed clinical characterization of the enrolled patients, with extensive follow-up periods and a diverse array of available samples, encompassing, primary and recurrent tumor tissues, as well as different grades of dysplastic lesions and matched normal squamous mucosa. The utilization of archived FFPE tissue aligns with common clinical practice, making our approach more pragmatic for potential clinical implementation.

In summary, findings from this study show that genetic changes and aberrations in immune gene expression become detectable in the early stages of upper aerodigestive tumor development, underscoring the prospect of intervening in the initial disease progression mechanisms. A better understanding of the concept that premalignant lesions may transform presumably with acquired mutations and copy number alterations to escape immune surveillance is crucial. It could provide scientists with new targets for intervention in those patients who are at risk of progression to HNSCC and clinicians with biomarkers for risk stratification and therapy decision-making in patients with high-risk premalignant lesions. In future, a deeper and more comprehensive assessment of epithelial, stromal, and immune heterogeneity in morphologically comparable grades of dysplasia in a cancerized field could potentially help predicting clinical outcomes and guide personalized therapies in patients with high-risk premalignancy. Ultimately, personalized diagnostics to guide treatment decisions, thereby avoiding unnecessary surgical procedures, their associated morbidities as well as reducing health economic costs may only be achieved by larger studies to confirm promising molecular biomarkers.

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

The studies involving humans were approved by Local ethics committee of the Ludwig-Maximilians-Universität (LMU) in Munich (ref. no. 18-446). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

AL: Writing – original draft, Writing – review & editing. JK: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. CW: Writing – original draft, Writing – review & editing. AJ: Writing – original draft, Writing – review & editing. PB: Writing – original draft, Writing – review & editing. SF: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.