Skip to main content
NCBI home page
ESMO Open logoLink to ESMO Open
. 2025 Aug 12;10(8):105538. doi: 10.1016/j.esmoop.2025.105538

The immunologic landscape of HRAS-mutant head and neck squamous-cell carcinoma

P Economopoulou 1, T Rampias 2, A Spathis 3, I Kotsantis 1, A Kyriazoglou 1, M Anastasiou 1, A Pantazopoulos 1, G Gomatou 1, A Margeli 1, M Papageorgiou 1, P Foukas 3, A Psyrri 1,∗
PMCID: PMC12361755  PMID: 40803018

Abstract

Background

HRAS mutations define a distinct biologic subset of head and neck squamous-cell carcinoma (HNSCC). There are limited data regarding HRAS-mutant (mut) tumors’ sensitivity to immunotherapy. We sought to evaluate the mutational landscape and transcriptional profile, as well as analyze the tumor microenvironment (TME) of HRAS-mut tumors to provide the conceptual framework for combinatorial treatment approaches.

Materials and methods

We analyzed mutational and transcriptome data from The Cancer Genome Atlas (TCGA). In addition, genomic DNA from baseline tumor biopsies was targeted for sequencing. Our study included 10 patients with HRAS-mut and 40 with HRAS-wild-type (WT) HNSCC. Programmed death-ligand 1 (PD-L1) expression in formalin-fixed paraffin-embedded tumor samples was assessed using the PD-L1 IHC 22C3 pharmDx assay. We characterized subpopulations of exhausted CD8(+) T cells by measuring the expression of T-cell factor-1 (TCF1) and programmed cell death protein 1 (PD-1) in both the center and the periphery of the tumors using multiplex immunohistochemistry, followed by analysis using a manually trained algorithm in QuPath software.

Results

The analysis of TCGA HNSCC mutation and mRNA expression data demonstrated that 6% of HNSCCs harbor mutant HRAS. Transcriptome analysis showed that HRAS-mut HNSCCs are infiltrated by immune cells (CD8A, CD8B, CD2) and have higher expression levels of CXCL11, CXCL10, CXCL9 and CCL4 chemokines. Moreover, the percentage of HRAS-mut samples increased in higher PD-L1 score groups (11% versus 20% versus 100% in tumor positive scores <1%, 1%-49% and ≥50%, respectively, P = 0.006). The analysis of TME showed that HRAS-mut tumors have a statistically significant higher number of total immune cells (5123.17/mm2 versus 3527.93/mm2, P = 0.002) and a higher percentage of pre-exhausted CD8(+) PD-1(+) TCF1(+) T cells in the periphery (384.67/mm2 versus 51.18/mm2, P = 0.040) than HRAS-WT tumors.

Conclusions

HRAS-mut HNSCCs are characterized by a significantly increased number of pre-exhausted PD-1(+) TCF1(+) T cells and PD-L1 expression, suggesting a potential sensitivity to immunotherapy.

Key words: HRAS mutations, head and neck squamous-cell carcinoma, PD-L1 expression, pre-exhausted CD8 (+) T cells, TCF1

Highlights

  • •

    Based on TCGA HNSCC mutational and mRNA expression data, we found that 6% of HNSCC cases harbor HRAS mutations.

  • •

    Transcriptome analysis showed that HRAS-mut cases are infiltrated by immune cells and express high levels of chemokines.

  • •

    The percentage of HRAS-mut samples increased in higher PD-L1 score groups.

  • •

    Analysis of TME revealed that HRAS-mut HNSCC cases have a higher number of pre-exhausted T cells.

Introduction

Despite the development of novel drugs and the implementation of immune checkpoint inhibitors in the treatment algorithm of recurrent/metastatic (R/M) disease, survival of patients with advanced head and neck squamous-cell carcinoma (HNSCC) remains dismal.1 Moreover, although the molecular landscape of HNSCC has been thoroughly studied,2, 3, 4 most gene alterations that occur during carcinogenesis are not druggable. Indeed, both human papillomavirus (HPV)-related and HPV-negative HNSCC tumors are characterized by PIK3CA/RTK/RAS pathway activation through a wide range of contributing mutations and amplifications, including EGFR, HER2, FGFR, activating PI3K mutations and RAS.5 However, despite similarities in the overall mutational burden in HPV-negative and HPV-related tumors, it has been shown that HPV-negative tumors mainly harbor mutations in TP53, CDKN2A, MLL2, CUL3, NSD1, PIK3CA and NOTCH genes.6 On the contrary, HPV-positive tumors display unique mutations in DDX3X and FGFR2/3 and aberrations in PIK3CA, MLL2/3 and NOTCH1. In addition, HPV-positive tumors are further characterized by mutations in KRAS and alterations in DNA-repair genes.6 On the other hand, ∼4%-8% of HNSCCs harbor activating mutations in HRAS, a member of the RAS family of oncogenes that is important in regulating cell growth and survival.7 Our group has previously demonstrated that HRAS mutations are associated with an unfavorable prognosis and cetuximab resistance in HNSCC.8

Tipifarnib is a highly selective farnesyl transferase inhibitor (FTI) that has shown promising activity in HRAS-mutant (HRAS-mut) R/M HNSCC.9 FTIs are biologically active anticancer agents that have demonstrated exceptional potency against tumor cells in preclinical models.10 These compounds inhibit the activation of various target proteins, including HRAS, by inhibiting the enzyme farnesyl transferase, ultimately resulting in cell growth arrest.11 Nevertheless, despite encouraging data on clinical activity in phase II clinical trials, molecular mechanisms and related biomarkers of response and resistance to tipifarnib are largely unknown.

Targeting the immunosuppressive interaction of the T-cell activation-induced programmed cell death protein 1 (PD-1) with its ligand programmed death-ligand 1 (PD-L1) has also shown activity in a wide variety of cancer types,12,13 including HNSCC.14 Based on landmark phase III trials, PD-1 checkpoint inhibitors nivolumab and pembrolizumab have been approved in R/M HNSCC.14, 15, 16 Despite the clinical activity of immunotherapeutic drugs, only a minority of patients with HNSCC respond, and resistance mechanisms remain undefined.

Nevertheless, it has been recently shown that responses to immunotherapy predominantly occur in tumors with a pre-existing antitumor T-cell response, which can most accurately be represented through the presence of dendritic cells, a crucial component of the antigen-presenting machinery, and cytotoxic CD8+ T cells that function as eradicators of cancer cells.17 Therefore, a so-called T-cell-‘inflamed’ phenotype defines a tumor with high expression of an immune signature that includes T-cell-related genes.18 On the other hand, T-cell exhaustion, defined as dysfunctional T cells stimulated by continuous antigen exposure, strongly affects the response to immune checkpoint blockade (ICB).

We have previously shown that HNSCCs harboring HRAS mutations are associated with higher PD-L1 expression scores and co-existence of TP53 mutations.19 However, there are limited data regarding the sensitivity of HRAS-mut tumors to ICB. In this translational study, we sought to (i) analyze the genomic landscape of HRAS-mut HNSCC to assess the prognostic impact of concurrent mutations in HRAS-mut tumors and their potential impact on modulating the T-cell-inflamed phenotype, and (ii) assess the immune microenvironment of HRAS-mut tumors by exploring the subpopulations of pre-exhausted and exhausted T cells to provide the conceptual framework for combinatorial strategies.

Materials and methods

HNSCC TCGA data

All meta-analyses carried out in this manuscript used data generated by The Cancer Genome Atlas (TCGA) research network which were retrieved from cBioPortal (http://cbioportal.org). Our analyses relied exclusively upon patient data which are publicly available. More specifically, RNA sequencing (RNA-seq) and mutation data were retrieved from 515 head and neck tumor samples (Head and Neck Squamous Cell Carcinoma TCGA-PanCancer Atlas cohort). To further characterize the molecular and immune profile of HRAS-mut tumors in HNSCC, the TCGA mutational and transcriptome data were analyzed using the TIMER 2.0 platform for comprehensive analysis of tumor-infiltrating immune cells.20, 21, 22

Identification of HRAS mutations by targeted sequencing

Genomic DNA from baseline biopsies of tumors was purified and subjected to targeted sequencing. Extracted DNA used for sequencing had a purity ratio of A260/A280 and a concentration of 20 ng/μl, consistent with common requirements.

Briefly, DNA samples were quantified using the Qubit double-stranded DNA high sensitivity fluorometric method (ThermoFisher Scientific, Waltham, MA). Forty nanograms from each sample was fragmented using the Covaris S220 focused ultrasonicator, with a setting of 100 W peak incident power, 30% as duty factor, 1000 cycles per burst and treatment time of 2.5 min per sample at 4°C. Fragmented DNAs were used as input for the Illumina TSO500 library prep protocol, according to manufacturer’s instructions (TruSight Oncology Reference Guide, Illumina Inc., San Diego, CA). Sequencing was carried out using Illumina NextSeq 500 sequencer, according to manufacturer’s instructions (NextSeq System Denature and Dilute Libraries Guide, Illumina Inc.). We identified 10 patients with HRAS-mut tumors. To compare the genomic and immune profiles between HRAS-mut and -wild-type (WT) tumors, we included 40 patients with HRAS-WT tumors in the analysis. All participants provided informed consent in compliance with the Declaration of Helsinki, which had been previously approved by the Ethics Committee of Attiko Hospital, Athens, Greece.

PD-L1 expression in formalin-fixed tumor samples was assessed using the PD-L1 IHC 22C3 pharmDx assay (Agilent Technologies, Carpinteria, CA) and characterized by the combined positive score (CPS). Assessment of CPS in the center and the periphery of the tumor was carried out.

Tissue microarrays—multiplex immunohistochemistry

Paraffin-embedded tissue blocks were selected in order for the tumor cells to represent ∼40%-60% of the total number of cells in the section. The selected areas had no necrotic or heavily inflammatory regions. Based on hematoxylin–eosin-stained slides after examining all the available sections from each case, representative areas were selected by an experienced pathologist (P.F.) to identify tumor regions reflecting the characteristics of the whole tumor. Regions from areas of interest (i.e. tumor center, tumor periphery and non-neoplastic tissue) were annotated and 1-mm diameter tissue cores were obtained from each one (tumor center: 0-3 cores; tumor periphery: 0-3 cores; non-neoplastic tissue: 0-2 cores). A hollow needle of 2-mm diameter was used to obtain the cores. Cores were subsequently inserted into paraffin recipient blocks forming the tissue microarrays (TMAs).

Multiplex (triple) immunohistochemistry (mIHC) staining was carried out on 3-μm-thick sections from each TMA block for the simultaneous detection of PD-1, CD8 and T-cell factor-1 (TCF1) in order to characterize subpopulations of CD8(+) T cells by measuring the co-expression of TCF1, a marker of T cells with stem-like properties, and PD-1 in both the C and P of the tumor. Furthermore, as tumor-draining lymph nodes (TDLNs) are important immune sites for the priming and development of tumor-specific TCF1(+) T cells,23 we sought to characterize the distribution of dendritic cells, cytotoxic CD8(+) T cells and regulatory T cells in lymph node parenchyma by using mIHC staining with anti-CD11c, anti-CD8 and anti-Foxp3, respectively.

Formalin-fixed paraffin-embedded tissue sections were hatched at 60°C for 1 h for deparaffinization and rehydrated in xylene solution and ethanol series. Epitope retrieval was carried out with target retrieval solution (pH: 9, 30 min, 97°C) using a commercial kit [Envision Flex, High pH (Link), Dako, Glostrup, Denmark] and then washed with Tris-Wash Buffer B, TBS (20×) for 5 min. Subsequently, a 2-day standardized IHC protocol was carried out for tumor sections: on day 1 anti-PD-1 [rabbit monoclonal, clone EPR4877(2) (abcam), 1 : 50 dilution, DAB chromogen (DAKO, DM827)] staining followed by anti-CD8 antibody [mouse monoclonal, clone C8/144B (DAKO), 1 : 50 dilution, magenta chromogen (DAKO, DM857)] staining and on day 2, after sample boiling in target retrieval solution, anti-TCF1 [rabbit monoclonal, clone C63D9 (Cell Signaling), 1 : 50 dilution, green chromogen (Bio SB, BSB 0128)] staining. For TDLN sections, the staining sequence was as follows: anti-CD8 [mouse monoclonal, clone C8/144B (DAKO), 1 : 50 dilution, DAB chromogen (DAKO, DM827)] and anti-CD11c [rabbit monoclonal, clone EP157 (Bio SB), 1 : 100 dilution, magenta chromogen (DAKO, DM857)], both on day 1 and anti-Foxp3 [rabbit monoclonal, clone SP97 (ZYTOMED), 1 : 50 dilution, green chromogen (Bio SB, BSB 0128)] on day 2. Following triple IHC staining completion, the slides were counterstained with hematoxylin for 25 s and washed under tap water. The sections were immersed in ethanol series 2× (50%, 70%, 90%) and were rinsed two times in xylene. Then, the slides were dried until xylene evaporation and were covered for scanning and QuPath software analysis.

Digitized slides and image analysis

Following IHC staining, TMA slides were scanned using a whole slide imaging scanner (D-Sight 200 Fluo by Menarini Diagnostics s.r.l., Florence, Italy) using ×20 magnification objective. Digitized images were subsequently ported to the open-source software QuPath v.0.4.3. (Queen’s University, Belfast, Northern Ireland). The image type settings in QuPath were set to ‘Brightfield (other)’ and the stain vectors were automatically evaluated by QuPath algorithms. Images of cores containing whole tissue were included in the study and analysis, while images of cores with no analyzable tissue were excluded from the study; such cases had focus issues or damaged and not measurable tissue.

Total counted cells were estimated using the positive cell count capability of QuPath (functionality analyze-cell detection). Then, machine learning (ML) algorithms were trained using numerous tissue sections to classify cells as (i) negative, (ii) single stain positive for each IHC staining (i.e. three ML classifiers), (iii) double stain positive for the different pairs of IHC staining (i.e. three ML classifiers) or (iv) triple stain positive if a cell was positive for all three IHC stains. Eventually, eight different ML classifiers were trained to identify negative cells or the various types of positivity in IHC staining. The in-built random tree ML classifiers available through the QuPath software were used for this purpose. The performance of each classifier on cell identification according to IHC staining type was evaluated by calculating the false-positive and false-negative indices; these were less than 5% and 0.7%, respectively, for the worst case of the classifiers. The outcome of this step was a set of various ML classifiers capable of identifying the IHC staining status of cells in digitized TMA sections.

The next step involved tissue annotation as (i) tumor center, (ii) tumor periphery and (iii) non-neoplastic tissue, and the cell/nuclei detection algorithms and object classifiers were applied on each image to detect cells and classify them as (i) negative for IHC staining, (ii) DAB positive [PD-1(+)], (iii) magenta positive [CD8(+)], (iv) green positive [TCF1(+)], (v) magenta-DAB positive [CD8(+) PD-1(+)], (vi) magenta-green positive [CD8(+) TCF1(+)], (vii) DAB-green [PD-1(+) TCF1(+)] and (viii) magenta-DAB-green positive [CD8(+) PD-1(+) TCF1(+)]. CD8 and PD-1 staining patterns are membranous, while TCF1 has a nuclear staining pattern. For each core, the total number of positive cells was normalized for 1 mm2 tissue area, yielding positive cell density (number of positive cells per mm2). Moreover, percentages of the different [CD8(+) PD-1(+) CD8(+) PD-1(+) TCF1(+)] T-cell subpopulations/total CD8(+) T cells were calculated.

Statistical analysis

The measurements obtained from QuPath were exported into Microsoft Excel spreadsheet files (Microsoft Corporation, Redmond, WA) and then were ported to SPSS for Windows version 24.0 software platform (IBM Corp., Armonk, NY) for statistical analysis. Descriptive values were expressed as median and standard error for the arithmetic data or frequency and the relevant percentages for the categorical data. Comparisons between groups for the qualitative parameters were made using the chi-square test. For the arithmetic variables, normality was not always possible to be ensured via the Shapiro–Wilk test; thus, nonparametric tests were applied, specifically the Kruskal–Wallis test for three or more groups and the Mann–Whitney U test for two groups. Furthermore, correlations of arithmetic characteristics were evaluated via the Spearman correlation coefficient. The significance level cut-off for the study was set to P = 0.05.

Results

Analysis of TCGA HNSCC mutational data

The analysis of TCGA HNSCC mutation and mRNA expression data (HNSCC Pan Cancer Atlas cohort) revealed that 6% of HNSCCs harbor mutant HRAS (HRAS-mut group), whereas 4.7% of HNSCCs overexpress WT HRAS (HRAS-WTOV group). The rest of the cohort (89.3%) consists of tumors with no HRAS alterations (mutations or overexpression) and, therefore, represents the non-altered HRAS group (Figure 1A).

Figure 1.

Figure 1

Open in a new tab

HRAS alterations in head and neck cancer. (A) TCGA HRAS mutation and transcriptomic data. Tumors with HRAS expression that corresponds to z-score ≥1.5 were assigned to the HRAS-overexpressing group (HRAS-WTOV). (B) Chart showing the relative size of HRAS-mut and HRAS-WTOV groups in TCGA HNSCC cohort (PanCancer Atlas). (C) GO analysis of differentially up-regulated genes in the HRAS-mut group. (D) Heatmap indicating the mRNA expression levels of HRAS and immune genes based on the TCGA HNSCC mutation and transcriptomic data (PanCancer Atlas). (E) Fold change expression of specific immune genes in the HRAS-mut group of tumors compared with the non-altered HRAS group of tumors based on TCGA transcriptomic data (PanCancer Atlas). GO, gene ontology; HNSCC, head and neck squamous-cell carcinoma; mut, mutant; NF-κB, nuclear factor-κB; NK, natural killer; TCGA, The Cancer Genome Atlas; WT, wild type.

Comparison of transcriptomics between HRAS-mut and non-altered HRAS tumors

Comparing the transcriptome between HRAS-mut and non-altered HRAS groups, we identified 547 differentially expressed genes (DEGs) up-regulated in HRAS-mut tumors. Gene ontology (GO) analysis for this set of genes indicated pathways that are mainly related to immune responses and cytokine/chemokine-mediated signaling pathways (Figure 1B). As shown in Figure 1D and E, this transcriptome analysis revealed that HRAS-mut tumors display a higher degree of infiltration by immune cells (CD8A, CD8B, CD2) and higher expression levels for CXCL11, CXCL10, CXCL9 and CCL4 chemokines and cytokines.

Our findings were further validated by the analysis of TCGA RNA-seq data using the TIMER2.0 platform (Supplementary Figure S1, available at https://doi.org/10.1016/j.esmoop.2025.105538).

Comparing the transcriptome between HRAS-mut and HRAS-WTOV groups, we failed to identify a similar transcriptomic signature for these immune genes in HRAS-WTOV tumors, suggesting a different immunological profile compared with the HRAS-mut group (Figure 2A and B).

Figure 2.

Figure 2

Open in a new tab

Immune profile of tumors with HRAS alterations in head and neck cancer. (A) Heatmap indicating the mRNA expression levels of HRAS and immune genes based on the TCGA HNSCC mutation and transcriptomic data (PanCancer Atlas). (B, C) Boxplot indicating the CD8A and CD274 mRNA expression levels among the HRAS-mut, HRAS-WTOV and non-altered HRAS group of head and neck tumors. ANOVA was used for the statistical analysis of expression levels among the three groups. ANOVA, analysis of variance; HNSCC, head and neck squamous-cell carcinoma; mut, mutant; TCGA, The Cancer Genome Atlas; WT, wild type.

The comparative analysis of RNA-seq data from the HRAS-WTOV and non-altered HRAS groups revealed 2955 up-regulated genes in HRAS-overexpressing tumors. The GO analysis for these DEGs did not indicate pathways related to immune response, further indicating that the HRAS-WTOV group displays a quite different immunological profile compared with the HRAS-mut group. As shown in Figure 3A, the above GO analysis revealed that overexpressed genes in the HRAS-WTOV group are mostly involved in pathways related to enhanced translation and mitochondrial electron transport chain (ETC) activity. Among the differentially expressed components of the electron transfer pathway, UQCC3 and ATP5F1D were the most up-regulated in HRAS-WTOV tumors (Figure 3B and C). A significantly lower expression of ETC components was detected in HRAS-mut tumors (Supplementary Figure S2A, available at https://doi.org/10.1016/j.esmoop.2025.105538), further supporting a different gene expression profile between HRAS-mut and HRAS-WTOV tumors.

Figure 3.

Figure 3

Open in a new tab

Gene expression profile of the HRAS-WTOV group of tumors. (A) Gene ontology analysis of differentially up-regulated genes in the HRAS-WTOV group. (B) Heatmap indicating the mRNA expression levels of HRAS and mitochondrial DNA genes based on the TCGA HNSCC transcriptomic data (PanCancer Atlas). (C) Fold change expression of mitochondrial DNA genes in the HRAS-WTOV group of tumors compared with the non-altered HRAS group of tumors based on TCGA transcriptomic data (PanCancer Atlas). HNSCC, head and neck squamous-cell carcinoma; TCGA, The Cancer Genome Atlas; WT, wild type.

It is well accepted that oncogenic HRAS can regulate the expression of genes involved in the ETC and that overexpression of both HRAS and components of the ETC has been observed in different cancer types.24 In our analysis, we have identified a strong correlation between HRAS expression and the expression of many different ETC components, including UQCC3 and ATP5F1D, independent of HRAS mutation status (Supplementary Figure S2B, available at https://doi.org/10.1016/j.esmoop.2025.105538).

The presence of HRAS mutations is associated with higher PD-L1 expression

PD-L1 expression levels were assessed by IHC in 6 HRAS-mut HNSCCs and 28 HNSCCs harboring WT HRAS. For all cases, PD-L1+ cells inside and at the periphery of the tumor were assessed independently. The CPS and tumor proportion score (TPS) were both evaluated.

The presence of HRAS mutations was more common in higher TPS PD-L1 scores (11% versus 20% versus 100% in TPS <1%, 1%-49% and ≥50%, respectively, P = 0.006). Similar results were obtained for CPS (11% versus 43% in CPS <10 and ≥10, P = 0.050).

Immune cell densities in the center and periphery of HRAS-mut versus HRAS-WT tumors

In order to interrogate the spatial (center, periphery) distribution and proportions of various T-cell subsets, we carried out mIHC on both HRAS-mut and HRAS-WT tumors. The cells of interest in our study were: overall CD8(+) T cells, CD8(+) PD-1(+) TCF1(−) T cells (indicative of exhausted cells), CD8(+) PD-1(−) TCF1(+) (indicative of naive/memory stem-like cells) and CD8(+) PD-1(+) TCF1(+) (indicative of activated/precursor-exhausted cells).

Regarding the whole cohort of patients, significantly higher densities of CD8(+) PD-1(+) TCF1(+) naive/memory stem-like T cells and CD8(+) PD-1(+) TCF1(−) exhausted T cells were found in the center of the tumors, compared with the infiltrative margin (P = 0.001 for both).

To interrogate deeper into the distribution of the various T-cell subsets, we compared their densities in HRAS-mut and HRAS-WT patients both in the center and periphery regions. HRAS-mut tumors present higher densities of CD8+ T cells in the periphery (851.10/mm2 versus 333.30/mm2, P = 0.066) compared with their WT counterparts. Importantly, the densities of CD8(+) PD-1(+) TCF1(+) activated/precursor-exhausted T cells and of CD8(+) PD-1(+) TCF1(−) exhausted T cells were higher in the periphery of HRAS-mut tumors (384.67/mm2 versus 51.18/mm2, P = 0.040, Figure 4; 152.01/mm2 versus 9.76/mm2, P = 0.030, respectively), whereas in the center of the tumors only the density of CD8(+) PD-1(+) TCF1(−) exhausted T cells was significantly higher (13.77% versus 2.67% of total CD8+ cells, P = 0.022) in HRAS-mut patients (Figure 5).

Figure 4.

Figure 4

Open in a new tab

Bar graph showing that the percentage of pre-exhausted CD8+ T cells, defined as PD-1+ TCF1+, is elevated in the P ofHRAS-mut tumors (384.67/mm2versus 51.18/mm2,P= 0.040). CI, confidence interval; mut, mutant; P, periphery; PD-1, programmed cell death protein 1; TCF1, T-cell factor-1; WT, wild type.

Figure 5.

Figure 5

Open in a new tab

Bar graph showing that exhausted T cells, defined asPD-1+TCF1− are more abundant in the C ofHRAS-mut tumors compared with WT (13.77% versus 2.67% of total CD8+ cells,P= 0.022). C, center; CI, confidence interval; mut, mutant; P, periphery; PD-1, programmed cell death protein 1; TCF1, T-cell factor-1; WT, wild type.

Moreover, the pixel area occupied by CD11c+ dendritic cells and the density of CD8(+) T cells were higher in regional, metastasis-free, lymph nodes from HRAS-mut patients (79.25% versus 38.80%, P = 0.036, and 10 160.43/mm2 versus 4438.28/mm2, P = 0.036, respectively).

Discussion

In the era of precision medicine, there is a continuous effort to identify new targeted therapies for solid tumors with rare genetic alterations. Constitutive activating mutations in RAS genes are found in ∼20% of human cancers.25 Among them, HRAS mutations are mostly prevalent in skin cancer, head and neck cancer and bladder cancer, whereas KRAS is highly mutated in colorectal cancer, pancreatic cancer and lung adenocarcinoma. Moreover, several in vitro and in vivo studies have demonstrated that WT HRAS overexpression can promote fibroblast transformation, tumor initiation and progression.26, 27, 28, 29 Mutations in oncogenic HRAS occur relatively rarely (4%-8%) in patients with HNSCC and are more frequently present in patients with HPV-negative tumors.7 Tipifarnib is the only investigational drug for patients with R/M HNSCC harboring HRAS mutations to have been granted the Breakthrough Therapy Designation by the Food and Drug Administration. To improve the understanding of the underlying biology and leverage the findings to optimize the management of HRAS-mut tumors, we sought to uncover the underpinnings of sensitivity to immunotherapy by analyzing the genomic landscape and tumor microenvironment (TME) of HRAS-mut tumors. We found that these tumors display a T-cell-inflamed phenotype, characterized by higher levels of CD8(+) T-cell infiltration, total immune cells, high expression of chemokines and cytokines, high PD-L1 expression and pre-exhausted T cells, suggesting a potential sensitivity of these tumors to ICB.

It has long been known that T cells’ presence in cancers is correlated with improved prognostic outcomes in a variety of solid tumors. Although cytotoxic T cells are consistently regarded as major drivers of antitumor immunity, the diversity in T-cell reactivity, activation and dysfunctional states raises important questions for tumor immunologists30: firstly, how tumor-specific T-cell populations are maintained and secondly, which T cells mediate ICB antitumor responses. Indeed, exhausted CD8(+) T cells within the TME are reinvigorated by ICB to produce clinical remission in a subgroup of cancer patients.23,31,32 Nevertheless, exhausted CD8(+) T cells display excessive heterogeneity, encompassing subsets of pre-exhausted and terminally exhausted T cells.33,34 Among these, pre-exhausted T cells expressing TCF1 likely represent predominate responders to PD-1/PD-L1 ICB.23,35,36

In our cohort of patients with HRAS-mut HNSCC, we found a statistically significant higher density of CD8(+) T cells in both the periphery and the center of the tumors as compared with their HRAS-WT counterparts by IHC. We confirmed this result using TCGA transcriptomic data that showed a higher degree of infiltration by immune cells (CD8A, CD8B, CD2) and higher expression levels for CXCL11, CXCL10, CXCL9 and CCL4 chemokines and cytokines, compared with HRAS non-altered tumors. Most importantly, we showed for the first time that HRAS-mut tumors have a higher percentage of pre-exhausted CD8(+) PD-1(+) TCF1(+) T cells in the periphery, suggesting a potential sensitivity to ICB. In a recent study by Kareff et al., the authors analyzed the molecular and immune profile of 524 HRAS-mut solid tumors including HNSCC.37 They calculated immune cell fractions through the deconvolution of whole transcriptome sequencing data using an immune deconvolution algorithm and similarly found increased CD8+ T cells in HRAS-mut tumors. However, these tumors were treated with immunotherapy before participation in the study. Interestingly, HRAS-mut triple-negative breast cancer and urothelial cancer cases that were also included in this study exhibited decreased CD8(+) and CD4(+) T cells.37 In addition, Lyu et al. used single-sample gene-set enrichment analysis scores to compare enrichment levels of 20 immune signatures between HRAS-mut and HRAS-WT HNSCC tumors (three cancer multi-omics datasets, including TCGA HNSCC dataset, an HNSCC gene expression profiling dataset and an HNSCC cohort treated with immunotherapy).38 In accordance with our finding, they found that HRAS mutations were characterized by enhanced immune signatures, indicating an increased sensitivity to ICB.

Furthermore, we evaluated PD-L1 expression by IHC in HRAS-mut and -WT HNSCCs. We found higher CPS and TPS in HRAS-mut tumors as compared with their WT counterparts. More recently, Coleman et al. identified 249 patients with HRAS-mut HNSCC from four patient datasets and reported their demographics, clinical outcomes and mutational profile.7 In this study, PD-L1 TPS was measured by IHC. Consistent with our findings, HRAS-mut tumors were less frequently PD-L1 negative (<1%) than HRAS-WT tumors (7.7% versus 53.4%, P < 0.044).7 In addition, in the study by Lyu et al., several immune checkpoint genes including PD-1, PD-L1 and PD-L2 were more highly expressed in HRAS-mut than in HRAS-WT HNSCCs.38

Accumulating evidence supports the notion that residency of T cells (either tumor, periphery or TDLNs) plays a substantial role in their capacity to generate antitumor response to ICB. It is well known that CD8(+) T cells that reside either in the blood circulation or lymph nodes are traditionally substratified based on their differentiation/functional status into naive T cells, effector T cells and memory T cells.30 However, modern molecular profiling techniques such as single-cell RNA-seq have enabled detailed profiling of T-cell subsets in human cancers and have revealed the immense heterogeneity of intertumoral T-cell states. Recently, Huang and fellow researchers used multiple preclinical models, including mice and cell lines, and demonstrated different functional states of tumor-reactive CD8+ T cells in the TDLNs. More specifically, CD8(+) T cells expressing the phenotype TCF1(+) TOX(−) were found to be bona fide memory T cells and function as the genuine responders to PD-1/PD-L1 ICB.23 In our cohort, we found an increased percentage of CD11c+ dendritic cells and CD8(+) T cells in draining TDLNs from HRAS-mut as compared with HRAS-WT tumors, consistent with data showing that maintenance of TCF1 stem cell marker by intratumoral T cells requires continuous migration from draining lymph nodes.

An interesting observation derived by transcriptomics in our study is that tumors that overexpress WT HRAS (named as HRAS WTOV) display a different transcriptomic signature for studied immune genes, suggesting a distinct, more immunologically cold phenotype compared with HRAS-mut tumors. Notably, GO analysis revealed that overexpressed genes in the HRAS-WTOV group are mostly involved in pathways related to enhanced translation and mitochondrial ETC activity. Based on transcriptomic data, HRAS-WTOV tumors represent a distinctive entity presenting a different gene expression profile compared with HRAS-mut tumors. However, the finding that HRAS-WTOV tumors display high sensitivity to tipifarnib treatment suggests a high degree of dependency on HRAS activity by these tumors.39 In addition, HRAS inhibition by tipifarnib in this group of tumors has been associated with re-sensitization to cisplatin.40 Mitochondrial dysfunction and increased oxidative stress correlate with oncogenic transformation induced by mutant KRAS overexpression,41 but HRAS overexpression has not been studied in that context. Our study is the first to report a different immunological and genomic profile of HRAS-WT tumors displaying overexpression compared with HRAS non-altered tumors.

A major limitation of our study is the small number of participants, which is attributed to the rarity of HRAS mutations in HNSCC. This limitation may be ameliorated if we take into consideration that the focus of the study is not the description of demographics and clinical outcomes of patients with HRAS-mut HNSCC, but a translational-based analysis of the TME of HRAS-mut tumors. Validation of our results in future studies with larger cohorts is needed.

In conclusion, our study represents a pioneering effort to evaluate the immune context of HRAS-mut HNSCCs. We show for the first time that HRAS-mut tumors are characterized by a significantly higher number of total immune cells, pre-exhausted PD-1(+) TCF1(+) T cells and PD-L1 expression, suggesting a potential sensitivity of these tumors to immunotherapy alone or in combo with tipifarnib.

Acknowledgments

Funding

This work was supported by Kura Oncology, Inc., with translational research grants (no grant number) to the National Kapodistrian University of Athens (AP) and the Bioacademy of Athens (TR).

Disclosure

AP has received honoraria and research funding from BMS, MSD, Roche, Pfizer, Merck Serono, AstraZeneca, Ipsen. She has also received research support from Kura Oncology. All other authors have declared no conflicts of interest.

Supplementary data

Supplementary Figures
mmc1.docx (551.5KB, docx)

References

  • 1.Mody M.D., Rocco J.W., Yom S.S., Haddad R.I., Saba N.F. Head and neck cancer. Lancet. 2021;398(10318):2289–2299. doi: 10.1016/S0140-6736(21)01550-6. [DOI] [PubMed] [Google Scholar]
  • 2.Mountzios G., Rampias T., Psyrri A. The mutational spectrum of squamous-cell carcinoma of the head and neck: targetable genetic events and clinical impact. Ann Oncol. 2014;25(10):1889–1900. doi: 10.1093/annonc/mdu143. [DOI] [PubMed] [Google Scholar]
  • 3.Stransky N., Egloff A.M., Tward A.D., et al. The mutational landscape of head and neck squamous cell carcinoma. Science. 2011;333(6046):1157–1160. doi: 10.1126/science.1208130. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Chung C.H., Guthrie V.B., Masica D.L., et al. Genomic alterations in head and neck squamous cell carcinoma determined by cancer gene-targeted sequencing. Ann Oncol. 2015;26(6):1216–1223. doi: 10.1093/annonc/mdv109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Pearson A.T., Vokes E.E. Is this the dawn of precision oncology in head and neck cancer? J Clin Oncol. 2021;39(17):1839–1841. doi: 10.1200/JCO.21.00569. [DOI] [PubMed] [Google Scholar]
  • 6.Seiwert T.Y., Zuo Z., Keck M.K., et al. Integrative and comparative genomic analysis of HPV-positive and HPV-negative head and neck squamous cell carcinomas. Clin Cancer Res. 2015;21(3):632–641. doi: 10.1158/1078-0432.CCR-13-3310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Coleman N., Marcelo K.L., Hopkins J.F., et al. HRAS mutations define a distinct subgroup in head and neck squamous cell carcinoma. JCO Precis Oncol. 2023;7 doi: 10.1200/PO.22.00211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Rampias T., Giagini A., Siolos S., et al. RAS/PI3K crosstalk and cetuximab resistance in head and neck squamous cell carcinoma. Clin Cancer Res. 2014;20(11):2933–2946. doi: 10.1158/1078-0432.CCR-13-2721. [DOI] [PubMed] [Google Scholar]
  • 9.Ho A.L., Brana I., Haddad R., et al. Tipifarnib in head and neck squamous cell carcinoma with HRAS mutations. J Clin Oncol. 2021;39(17):1856–1864. doi: 10.1200/JCO.20.02903. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Gilardi M., Wang Z., Proietto M., et al. Tipifarnib as a precision therapy for HRAS-mutant head and neck squamous cell carcinomas. Mol Cancer Ther. 2020;19(9):1784–1796. doi: 10.1158/1535-7163.MCT-19-0958. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Simanshu D.K., Nissley D.V., McCormick F. RAS proteins and their regulators in human disease. Cell. 2017;170(1):17–33. doi: 10.1016/j.cell.2017.06.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Borghaei H., Paz-Ares L., Horn L., et al. Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer. N Engl J Med. 2015;373(17):1627–1639. doi: 10.1056/NEJMoa1507643. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Wolchok J.D., Kluger H., Callahan M.K., et al. Nivolumab plus ipilimumab in advanced melanoma. N Engl J Med. 2013;369(2):122–133. doi: 10.1056/NEJMoa1302369. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Ferris R.L., Blumenschein G., Jr., Fayette J., et al. Nivolumab for recurrent squamous-cell carcinoma of the head and neck. N Engl J Med. 2016;375(19):1856–1867. doi: 10.1056/NEJMoa1602252. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Burtness B., Harrington K.J., Greil R., et al. Pembrolizumab alone or with chemotherapy versus cetuximab with chemotherapy for recurrent or metastatic squamous cell carcinoma of the head and neck (KEYNOTE-048): a randomised, open-label, phase 3 study. Lancet. 2019;394(10212):1915–1928. doi: 10.1016/S0140-6736(19)32591-7. [DOI] [PubMed] [Google Scholar]
  • 16.Harrington K.J., Burtness B., Greil R., et al. Pembrolizumab with or without chemotherapy in recurrent or metastatic head and neck squamous cell carcinoma: updated results of the phase III KEYNOTE-048 study. J Clin Oncol. 2023;41(4):790–802. doi: 10.1200/JCO.21.02508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Trujillo J.A., Sweis R.F., Bao R., Luke J.J. T cell-inflamed versus non-T cell-inflamed tumors: a conceptual framework for cancer immunotherapy drug development and combination therapy selection. Cancer Immunol Res. 2018;6(9):990–1000. doi: 10.1158/2326-6066.CIR-18-0277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Ayers M., Lunceford J., Nebozhyn M., et al. IFN-γ-related mRNA profile predicts clinical response to PD-1 blockade. J Clin Invest. 2017;127(8):2930–2940. doi: 10.1172/JCI91190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Psyrri A., Anastasiou M., Spathis A., et al. Abstract 3882: Translating genotype to immunophenotype in HRAS mutated head and neck squamous cell carcinoma (HNSCC) to identify effective tipifarnib partners for optimal patient outcomes. Cancer Res. 2022;82(suppl 12):3882. [Google Scholar]
  • 20.Li T., Fu J., Zeng Z., et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020;48(W1):W509–W514. doi: 10.1093/nar/gkaa407. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Liu C.C., Steen C.B., Newman A.M. Computational approaches for characterizing the tumor immune microenvironment. Immunology. 2019;158(2):70–84. doi: 10.1111/imm.13101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Li B., Severson E., Pignon J.C., et al. Comprehensive analyses of tumor immunity: implications for cancer immunotherapy. Genome Biol. 2016;17(1):174. doi: 10.1186/s13059-016-1028-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Huang Q., Wu X., Wang Z., et al. The primordial differentiation of tumor-specific memory CD8+ T cells as bona fide responders to PD-1/PD-L1 blockade in draining lymph nodes. Cell. 2022;185(22):4049–4066.e25. doi: 10.1016/j.cell.2022.09.020. [DOI] [PubMed] [Google Scholar]
  • 24.Weinberg F., Hamanaka R., Wheaton W.W., et al. Mitochondrial metabolism and ROS generation are essential for Kras-mediated tumorigenicity. Proc Natl Acad Sci U S A. 2010;107(19):8788–8793. doi: 10.1073/pnas.1003428107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Prior I.A., Hood F.E., Hartley J.L. The frequency of Ras mutations in cancer. Cancer Res. 2020;80(14):2969–2974. doi: 10.1158/0008-5472.CAN-19-3682. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Theodorescu D., Cornil I., Fernandez B.J., Kerbel R.S. Overexpression of normal and mutated forms of HRAS induces orthotopic bladder invasion in a human transitional cell carcinoma. Proc Natl Acad Sci U S A. 1990;87(22):9047–9051. doi: 10.1073/pnas.87.22.9047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Bai S., Feng Q., Pan X.Y., et al. Overexpression of wild-type p21Ras plays a prominent role in colorectal cancer. Int J Mol Med. 2017;39(4):861–868. doi: 10.3892/ijmm.2017.2903. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Maruyama C., Tomisawa M., Wakana S., et al. Overexpression of human H-ras transgene is responsible for tumors induced by chemical carcinogens in mice. Oncol Rep. 2001;8(2):233–237. doi: 10.3892/or.8.2.233. [DOI] [PubMed] [Google Scholar]
  • 29.Diaz R., Lopez-Barcons L., Ahn D., et al. Complex effects of Ras proto-oncogenes in tumorigenesis. Carcinogenesis. 2004;25(4):535–539. doi: 10.1093/carcin/bgh026. [DOI] [PubMed] [Google Scholar]
  • 30.van der Leun A.M., Thommen D.S., Schumacher T.N. CD8+ T cell states in human cancer: insights from single-cell analysis. Nat Rev Cancer. 2020;20(4):218–232. doi: 10.1038/s41568-019-0235-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Brahmer J.R., Tykodi S.S., Chow L.Q., et al. Safety and activity of anti-PD-L1 antibody in patients with advanced cancer. N Engl J Med. 2012;366(26):2455–2465. doi: 10.1056/NEJMoa1200694. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Hashimoto M., Kamphorst A.O., Im S.J., et al. CD8 T cell exhaustion in chronic infection and cancer: opportunities for interventions. Annu Rev Med. 2018;69:301–318. doi: 10.1146/annurev-med-012017-043208. [DOI] [PubMed] [Google Scholar]
  • 33.Im S.J., Hashimoto M., Gerner M.Y., et al. Defining CD8+ T cells that provide the proliferative burst after PD-1 therapy. Nature. 2016;537(7620):417–421. doi: 10.1038/nature19330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Kurtulus S., Madi A., Escobar G., et al. Checkpoint blockade immunotherapy induces dynamic changes in PD-1-CD8+ tumor-infiltrating T cells. Immunity. 2019;50(1):181–194.e6. doi: 10.1016/j.immuni.2018.11.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Siddiqui I., Schaeuble K., Chennupati V., et al. Intratumoral Tcf1+PD-1+CD8+ T cells with stem-like properties promote tumor control in response to vaccination and checkpoint blockade immunotherapy. Immunity. 2019;50(1):195–211.e10. doi: 10.1016/j.immuni.2018.12.021. [DOI] [PubMed] [Google Scholar]
  • 36.Miller B.C., Sen D.R., Al Abosy R., et al. Subsets of exhausted CD8+ T cells differentially mediate tumor control and respond to checkpoint blockade. Nat Immunol. 2019;20(3):326–336. doi: 10.1038/s41590-019-0312-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Kareff S.A., Trabolsi A., Krause H.B., et al. The genomic, transcriptomic, and immunologic landscape of HRAS mutations in solid tumors. Cancers (Basel) 2024;16(8):1572. doi: 10.3390/cancers16081572. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Lyu H., Li M., Jiang Z., Liu Z., Wang X. Correlate the TP53 mutation and the HRAS mutation with immune signatures in head and neck squamous cell cancer. Comput Struct Biotechnol J. 2019;17:1020–1030. doi: 10.1016/j.csbj.2019.07.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Kessler L., Malik S., Leoni M., Burrows F. Potential of farnesyl transferase inhibitors in combination regimens in squamous cell carcinomas. Cancers (Basel) 2021;13(21):5310. doi: 10.3390/cancers13215310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Rampias T., Hoxhallari L., Avgeris M., et al. Sensitizing HRAS overexpressing head and neck squamous cell carcinoma (HNSCC) to chemotherapy. Ann Oncol. 2019;30:v462–v463. [Google Scholar]
  • 41.Luo Y., Ma J., Lu W. The significance of mitochondrial dysfunction in cancer. Int J Mol Sci. 2020;21(16):5598. doi: 10.3390/ijms21165598. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Figures
mmc1.docx (551.5KB, docx)

RESOURCES