Advanced Human Papillomavirus–Negative Head and Neck Squamous Cell Carcinoma: Unmet Need and Emerging Therapies

Abstract

Despite notable progress in the treatment of advanced head and neck squamous cell carcinoma (HNSCC), survival remains poor in patients with recurrent and/or metastatic (R/M) human papillomavirus (HPV)–negative HNSCC. Worse outcomes in patients who are HPV-negative may be partly related to loss of cell-cycle regulators and tumor suppressors as well as a noninflamed and hypoxic tumor microenvironment, both of which contribute to treatment resistance and disease progression. Anti–programmed cell death protein 1–based regimens as current standard-of-care treatment for R/M HNSCC are associated with durable responses in a limited number of patients. The anti-EGFR mAb, cetuximab, has antitumor activity in this treatment setting, but responses are short-lived and inevitably curtailed due to treatment resistance. Crosstalk between the EGFR and hepatocyte growth factor–dependent mesenchymal–epithelial transition (c-MET) receptor tyrosine kinase pathway is a known mechanism of resistance to cetuximab. Dual targeting of EGFR and c-MET pathways may overcome resistance to cetuximab in patients with HPV-negative HNSCC. Here, we review clinical data of treatments evaluated in patients with R/M HPV-negative HNSCC and highlight the potential role of combining hepatocyte growth factor/c-MET and EGFR pathway inhibitors to overcome cetuximab resistance in this population.

Introduction

Head and neck squamous cell carcinomas (HNSCC) are a heterogeneous group of tumors that develop from the mucosal epithelium and arise predominantly in the oral cavity, oropharynx, hypopharynx, and larynx (1, 2). In the United States, HNSCC is the ninth most common cancer and the 12th leading cause of cancer-related deaths (3). Globally, the incidence of HNSCC has increased by 37% over the past decade (4), with an expected 30% increase annually by 2030 across developed and developing countries (2, 5). HNSCC is more prevalent in men than that in women and is generally diagnosed in patients 50 to 70 years of age (6).

The combination of tobacco and alcohol consumption accounts for 72% of HNSCC cases (5). In addition, infection with an oncogenic strain of human papillomavirus (HPV), primarily HPV-16, has been identified as a risk factor for tumors originating in the oropharynx (2). As a result, the increase in HNSCC incidence is largely due to the increase in HPV-positive oropharynx HNSCC (7). Although HPV-negative HNSCC incidence has not increased, prognosis is worse, and overall outcomes have not significantly changed despite recent advances (7, 8). Other populations at a higher risk of developing HNSCC include firefighters, veterans, and individuals repeatedly exposed to sawdust, burn pits, and leather dust (9–11).

Current treatments for patients with HNSCC are associated with morbidity and resistance, especially in the recurrent and/or metastatic (R/M) setting (12, 13). The standard-of-care first-line therapy for R/M HNSCC is the programmed cell death protein 1 (PD-1) inhibitor pembrolizumab with or without platinum-based chemotherapy (14, 15). In addition, cetuximab, an antibody targeting EGFR, combined with platinum-based chemotherapy, known as the EXTREME regimen, has been part of the treatment protocol for more than a decade (14–16). Despite progress in the development of novel treatments for patients with HNSCC, the median overall survival (mOS) of patients with HPV-negative R/M HNSCC is approximately 1 year (8). For patients who progress on or become refractory to a PD-1 inhibitor and chemotherapy, there are few effective treatment options, and the expected OS is approximately 6 months (15, 17).

Herein, we briefly summarize the genomic and immunologic perturbations that characterize HPV-negative R/M HNSCC, review selected prior prospective trials evaluating targeted therapies and combination regimens in this population, and introduce potential novel therapeutic targets of relevance.

Molecular Characterization of HPV-Negative HNSCC

HPV-negative HNSCC is partially driven by the acquisition of mutations related to carcinogens contained in tobacco and alcohol, most common mutations involving tumor suppressor genes (18, 19), which dysregulate critical cellular processes and ultimately lead to tumorigenesis (20). In addition, tobacco use is associated with an immunosuppressive tumor immune microenvironment characterized by lower numbers of CD8⁺ cytotoxic T cells, PD-L1–positive cells, and suppression of IFN response pathways (21). The worse outcomes in patients with HPV-negative HNSCC may partly be related to loss of cell-cycle regulators and tumor suppressors such as CDKN2A and TP53 (1, 19). The CDKN2A and TP53 gene alterations are almost exclusively seen in HPV-negative HNSCC (58% and 84%, respectively) versus HPV-positive HNSCC (0% and 3%; ref. 22). CDKN2A copy number loss was associated with poor clinical prognosis in patients with HPV-negative HNSCC (23). In addition, CCND1 copy number gain is more frequent in patients with HPV-negative HNSCC, with CCND1 gene alterations reported in 31% versus 3% in patients with HPV-positive HNSCC (22). Furthermore, HPV-negative HNSCC is associated with frequent focal deletions in other tumor suppressor genes, including NSD1, FAT1, NOTCH1, and SMAD4 (2).

Activation of nuclear erythroid 2–related factor 2 (NRF2), kelch-like ECH associated protein 1, and cullin 3 complex was also associated with oncogenesis and poor survival in HPV-negative HNSCC (24). NRF2 has emerged as a regulator of oxidative stress. Under normal conditions, NRF2 is suppressed through kelch-like ECH associated protein 1 –dependent ubiquitination–proteasomal degradation (25). In contrast, activation of NRF2 in the tumor environment promotes tumor growth and resistance to oxidants and anticancer drugs (25).

Tumor growth, proliferation, and metastasis require angiogenesis, which is governed by several different growth factor receptors, including VEGFR and EGFR (26). VEGF, an angiogenesis stimulator, is overexpressed in up to 90% of HNSCC cases (26, 27). VEGF activates VEGFR, which leads to the phosphorylation of the intracellular tyrosine kinase domain that activates the rat sarcoma virus (RAS)/PI3K/MAPK pathways essential for cancer cell development and growth (28, 29). The pathologic VEGF-receptor pathway triggers signaling processes that promote vascular damage and the growth of abnormal blood vessels (30). Overexpression of VEGF is associated with tumor growth, cell migration, and metastases in HNSCC.

EGFR is overexpressed in up to 90% of HNSCC cases and is associated with VEGF expression in HPV-negative HNSCC (31, 32). Alterations in the EGFR pathway are frequent in HPV-negative (15%) versus HPV-positive (6%) HNSCC (22). In HNSCC, binding of EGF induces a conformational change of EGFR, which stimulates protein tyrosine kinase activity, leading to autophosphorylation and receptor activation (33). This effect triggers activation of the RAS/rapidly accelerated fibrosarcoma/MAPK/PI3K-AKT pathway. Phosphorylated MAPK translocates into the nucleus and phosphorylates transcription factors that activate target genes, which promote angiogenesis, proliferation, metastasis, and invasion.

Moreover, the activation of alternative signaling pathways, such as the hepatocyte growth factor (HGF)/mesenchymal–epithelial transition (c-MET) receptor tyrosine kinase (RTK) signaling axis, has been implicated as a mediator of EGFR resistance (31). More than 20% of patients with HNSCC harbor either a copy number gain or an amplification of c-MET, and more than 80% show c-MET protein overexpression. c-MET is an RTK that, upon binding to its ligand HGF, leads to dimerization of two c-MET receptors followed by autophosphorylation of tyrosine residues, which serve as docking sites for downstream signaling molecules that mediate activation of the RAS/rapidly accelerated fibrosarcoma/ERK/MAPK/PI3K-AKT/PKB/mTOR, Crk-Rap, Rac-Pak, and STAT3 pathways (31, 34). c-MET activation promotes the epithelial–mesenchymal transition, a process characterized by proliferation, migration, invasion, and metastasis in HNSCC (35). In addition, c-MET and the RTK AXL are overexpressed in radiotherapy- and cisplatin-resistant HNSCC, and their expression correlates with aggressive tumors (35). Downstream of RTK, the PI3K oncogenic pathway, including PIK3CA, PTEN, and PIK3R1, is often mutated in HNSCC (36). PIK3CA is among the most frequently mutated and/or copy number altered genes in both HPV-negative (34%) and HPV-positive (56%) HNSCC tumors (22).

Epigenetic changes also play a role in driving oncogenesis in HPV-negative HNSCC (6). In a study evaluating genome-wide methylation levels in primary HNSCC tumor samples and lymph node metastases from individuals with different HPV statuses, HPV-negative tumor tissues were more frequently hypomethylated than HPV-positive tumors (37). Hypermethylation in HPV-positive HNSCC is driven by the viral oncoproteins, E6 and E7, which regulate DNA methylation in the host genome. Hypermethylated genes include those regulating cell cycle, apoptosis, cell adhesion, cell migration, differentiation, and G-coupled receptor signaling (38, 39). DNA methylation also regulates the tumor microenvironment (TME); therefore, methylation patterns may be a biomarker for response to immunotherapy in HNSCC (38).

HNSCC also has an immunosuppressive TME (40). The TME in HNSCC is a complex and heterogeneous mix of tumor cells and stromal cells that includes endothelial cells, cancer-associated fibroblasts (CAF), and immune cells (Fig. 1A). Several signaling pathways have been implicated in driving immune suppression in the TME. c-MET/HGF signaling via HGF-mediated upregulation of PD-L1 expression induces immune suppression through interaction with PD-1 on cytotoxic T cells resulting in decreased activation and immune-mediated antitumor activity (31, 41, 42). In a preclinical study, AXL and PI3K/AKT expression was associated with PD-L1 expression in tumors from patients with locally advanced HPV-negative HNSCC treated with surgery and postoperative radiotherapy (43). Additionally, AXL/PI3K signaling was associated with PD-L1 expression in three separate HPV-negative HNSCC cohorts (RPPA, mRNA, and TCGA mRNA cohorts). c-MET/HGF signaling may also impair cytotoxic T-cell activation through increased glycolytic metabolism and subsequent efflux of lactate from cancer cells into the TME (31). Hypoxia is also intertwined with immune suppression in the TME, and HNSCC is considered to have one of the most hypoxic and immunosuppressive TMEs among other solid tumors (40). Decreased T cells, antigen presenting cells, and proinflammatory M1 macrophages and increased immunosuppressive cells, as exemplified by a decreased ratio of cytotoxic T cells/FOXP3⁺ Tregs, are found in hypoxic tumors compared with nonhypoxic tumors (40). Additionally, recurrent tumors are more likely to be hypoxic than newly diagnosed primary tumors (40). Hypoxia-driven immunosuppression can occur through multiple mechanisms, some of which are driven by hypoxia-inducible factor 1α. Both tumor cells and CAFs produce growth factors such as VEGF that recruit endothelial cells, stimulating neovascularization and supplying oxygen and nutrients to the tumor (2). Hypoxia-inducible factor 1α upregulates VEGF, suppresses tumor immunity by inhibiting the mature dendritic cells, and induces immunosuppressive cells including regulatory T cells, tumor-associated macrophages, and myeloid-derived suppressor cells (MDSC; refs. 44, 45). Furthermore, AXL also plays numerous immunosuppressive roles in the TME, including recruiting MDSCs and regulatory T cells, mediating the secretion of immunosuppressive cytokines and chemoattractants, upregulating the expression of PD-L1 and PD-L2, and promoting the immunosuppressive protumor M2-like phenotype (46, 47). Additionally, hypoxic tumors display increased expression of EGFR and TGFβ pathway genes and targeting of EGFR with cetuximab decreased expression of hypoxia signature genes (40).

Figure 1.

Crosstalk between the EGFR and c-MET pathways in the HNSCC tumor microenvironment. A, Depiction of the tumor microenvironment in patients with HNSCC. B, Overexpression of EGFR and c-MET RTKs in HNSCC leads to the activation of downstream signaling pathways including PI3K/AKT, MAPK, and CDK4/6. Through dysregulation of cell death pathways, activation of these signaling cascades stimulates the survival of cancer cells. Cetuximab’s inhibition of EGFR is insufficient to inhibit downstream signaling mediated by HGF/c-MET activation due to crosstalk between RTKs. Activation of the receptor tyrosine kinase c-MET via its ligand, HGF, mediates proliferation, motility, and differentiation in several different cell types. C, Targeting the HGF/c-MET pathway and EGFR simultaneously may prevent binding of the EGF and HGF ligands to their respective receptors, EGFR, and c-MET, and downstream activation of PI3K and CDK4/6 signaling in the tumor microenvironment. AKT, AKT serine/threonine kinase; GRB-2, growth factor receptor bound protein 2; MOA, mechanism of action; MOD, mode of disease; RAF, rapidly accelerated fibrosarcoma. “Created with BioRender.com.”

An increase in the incidence of young patients with HPV-negative HNSCC who present without traditional risk factors (e.g., smoking) has been observed (5). Most of these young patients are female nonsmokers; the most frequent subsite of disease is the oral tongue (48, 49). Lower socioeconomic status, poor diet, physical inactivity, and poor oral hygiene have also been associated with HPV-negative HNSCC (4, 5). Young patients developing HPV-negative HNSCC, especially in the tongue, tend to have aggressive disease and disease-free survival comparable to that of older patients (50). The exact reason is unknown; however, it is hypothesized to be due partially to aberrant expression of the transmembrane protein discoidin, CUB, and LCCL domain containing 1 (DCBLD1). When coactivated with EGFR, DCBLD2, a neuropilin-like transmembrane protein that is upregulated during arterial remodeling (51) and a paralog of DCBLD1, has been shown to be a marker of poor prognosis in patients with HNSCC. Additionally, the SNP rs6942067, which causes an upregulation of the DCBLD1 gene, has been associated with HNSCC in young (age <40 years) nonsmokers with HPV-negative HNSCC (48). Moreover, patients with DCBDL1-high disease have an increased occurrence of oral tongue cancer and are more often female.

Despite the known mechanisms underlying HPV-negative HNSCC, the discovery of effective predictive biomarkers that can optimize patient selection remains an unmet need. In that regard, multi-omics analyses represent useful tools to identify potential biomarkers with strong biological rationale and to identify novel molecular targets for drug development. For example, using multi-omics analysis, Haung and colleagues identified three molecular subtypes amongst HPV-negative HNSCC, as exemplified by EGFR ligand abundance, Rb phosphorylation in combination with CCND1/CDKN2A aberrations, and concordant multiple immune checkpoint upregulation (52). Huang and colleagues further suggest that based on these three molecular subtypes, HPV-negative HNSCC can be stratified into those that will respond to EGFR targeted therapy, cyclin-dependent kinase (CDK) inhibitors, and immune checkpoint inhibitors (ICI), respectively (52). Furthermore, several additional targets, including KIT, FCER1G, PLAU, SERPINE1, TOP2A, matrix metalloproteinases, cell cycle and DNA damage–related kinases, and over expressed cancer/testis antigens and neoantigens, have been identified with these analyses (52).

Selected Clinical Trials in Patients with HPV-Negative HNSCC

Several trials have evaluated the efficacy and safety of targeted agents in patients with R/M HPV-negative HNSCC (Tables 1 and 2).

Table 1.

Summary of clinical data investigating anticancer therapies in HPV-negative recurrent or metastatic HNSCC.

	Pathway involved	Study design/phase	Numbers of patients	Line of therapy	Key efficacy endpoint(s)b	Study
Anticancer therapya
Pembrolizumab	Immunotherapy (PD-1 inhibitor)	Single arm, phase Ib	60 (HPV-negative, 37)	1L+	ORR: 18%; 14% (HPV-negative disease)	Seiwert and colleagues (53); KEYNOTE-012
Pembrolizumab vs. pembrolizumab plus chemotherapy vs. cetuximab plus chemotherapy	Immunotherapy (PD-1 inhibitor) EGFR	Randomized, phase III	882 (HPV-negative, 692)	1L	mOS: All patients: 11.5 months (pembrolizumab) vs. 10.7 months (cetuximab plus chemotherapy); 13.0 months (pembrolizumab plus chemotherapy) vs. 10.7 months (cetuximab plus chemotherapy) PD-L1 CPS <1: 7.9 months (pembrolizumab) vs. 11.3 months (cetuximab plus chemotherapy) PD-L1 CPS ≥1: 12.3 months (pembrolizumab) vs. 10.4 months (cetuximab plus chemotherapy); 13.6 months (pembrolizumab plus chemotherapy) vs. 10.6 months (cetuximab plus chemotherapy) PD-L1 CPS ≥20: 14.9 months (pembrolizumab) vs. 10.8 months (cetuximab plus chemotherapy); 14.7 months (pembrolizumab plus chemotherapy) vs. 11.1 months (cetuximab plus chemotherapy)	Burtness and colleagues (8); Burtness and colleagues (54); Harrington and colleagues (55); KEYNOTE-048
Pembrolizumab plus cabozantinib	Immunotherapy (PD-1 inhibitor)	Randomized single arm, phase II	36 (HPV-negative, 12)	1L+	ORR: 52% mPFS: 14.6 months mOS: 22.3 months	Saba and colleagues (56)
Pembrolizumab vs. pembrolizumab plus lenvatinib	Immunotherapy (PD-1 inhibitor)	Randomized, phase III	511 (HPV-negative, 396)	1L	ORR: 25.4% (pembrolizumab); 46.1% (pembrolizumab plus lenvatinib) mPFS: 2.8 months (pembrolizumab); 7.0 months (pembrolizumab plus lenvatinib) mOS: 17.9 months (pembrolizumab); 15.0 months (pembrolizumab plus lenvatinib)c	Licitra and colleagues (57)
Nivolumab vs. standard therapy (methotrexate, docetaxel, or cetuximab)	Immunotherapy (PD-1 inhibitor)	Randomized, phase III	361 (HPV-negative, 86)	2L+	RR: 13.3% (nivolumab) and 5.8% (standard therapy) mPFS: 2.0 months (nivolumab) and 2.3 months (standard therapy) mOS: 7.7 months (nivolumab) and 5.1 months (standard therapy)	Ferris and colleagues (58); Ferris and colleagues (59); CheckMate 141
Afatinib vs. methotrexate	ERBB family blocker	Randomized, phase III	483 (HPV-negative, 208)	2L	mPFS: 2.6 months (afatinib) and 1.7 months (methotrexate)	Machiels and colleagues (60); Lux-Head & Neck 1
Duligotuzumab vs. cetuximab	EGFR and HER3	Randomized, phase II	121 (HPV-negative, 85)	2L+	ORR: 12% (duligotuzumab) and 14.5% (cetuximab) mPFS: 4.2 months (duligotuzumab) and 4 months (cetuximab) mOS: 7.2 months (duligotuzumab) and 8.7 months (cetuximab)	Fayette and colleagues (61)
Cetuximab combination therapy
Cetuximab plus nivolumab	EGFR and immunotherapy (PD-1 inhibitor)	Single arm, phase II	95 (HPV-negative, 48)	1L and 2L+	mPFS: 3.4 months (2L+) and 6.15 months (1L) mOS: 11.5 months (2L+) and NR (1L)	Chung and colleagues (62); Chung and colleagues (63)
Cetuximab plus pembrolizumab	EGFR and immunotherapy (PD-1 inhibitor)	Nonrandomized, phase II	33 (HPV-negative, 21)	1L+	ORR: 45%	Sacco and colleagues (64)
Cetuximab plus durvalumab	EGFR and immunotherapy (PD-L1 blocker)	Single arm, phase II	35 (HPV-negative, 7)	1L+	ORR: 39% mPFS: 5.8 months mOS: 9.6 months	Gulati and colleagues (65)
Cetuximab plus ficlatuzumab	EGFR and HGF	Single arm, phase I	13 (HPV-negative, 12)	2L+	ORR: 17% mPFS: 5.4 months mOS: 8.9 months	Bauman and colleagues (17)
Cetuximab plus ficlatuzumab vs. ficlatuzumab	EGFR and HGF	Randomized, phase II	58 (HPV-negative, 32)	2L+	ORR: 19% [cetuximab plus ficlatuzumab (all patients)]; 38% [cetuximab plus ficlatuzumab (HPV-negative disease)] mPFS: 3.7 months [cetuximab plus ficlatuzumab (all patients)]; 4.1 months [cetuximab plus ficlatuzumab (HPV-negative disease)]	Bauman and colleagues (66)
Cetuximab plus tivantinib vs. cetuximab	EGFR and c-MET	Randomized, phase II	78 (HPV-negative, 47)	1L+	ORR: 7.5% (all patients); 12.5% (HPV-negative disease) mPFS: 3.5 months (cetuximab plus tivantinib) mOS: 7.4 months (cetuximab plus tivantinib)	Kochanny and colleagues (67)
Cetuximab plus palbociclib vs. cetuximab	EGFR and CDK4/6	Randomized, phase II	125 (HPV-negative, 125)	1L+	mPFS: 3.9 months (cetuximab plus palbociclib) and 4.6 months (cetuximab) mOS: 9.7 months (cetuximab plus palbociclib) and 7.8 months (cetuximab)	Adkins and colleagues (68)
Cetuximab plus CDX-3379	EGFR and HER3	Single arm, phase II	30 (HPV-negative, 30)	3L+	ORR: 6.7%	Bauman and colleagues (69)
Cetuximab plus PX-866 vs. cetuximab	EGFR and PI3K	Randomized, phase II	83 (HPV-negative, 20)	2L+	mPFS: 80 days (both arms) mOS: 211 days (cetuximab plus PX-866) and 256 days (cetuximab)	Jimeno and colleagues (70)

Abbreviations: 1L, first line; 2L, second line; 3L, third line; CDK, cyclin-dependent kinase; NR, not reached.

^aAgents/combinations with positive clinical trial efficacy outcomes reported in this table are highlighted in bold text.

^bResults are reported for all patients unless specified otherwise.

^cORR and mPFS were significantly improved with pembrolizumab + lenvatinib, but mOS was not.

Table 2.

Summary of toxicity profiles of anticancer and cetuximab combination therapies in HPV-negative recurrent or metastatic HNSCC.

	Pathways involved	Notable toxicities (≥5% grade three AEs)	Study
Anticancer therapy
Pembrolizumab	Immunotherapy (PD-1 inhibitor)	No single grade three TRAE was found in ≥5% of the population	Seiwert and colleagues (53); KEYNOTE-012
Pembrolizumab vs. pembrolizumab plus chemotherapy vs. cetuximab plus chemotherapya	Immunotherapy (PD-1 inhibitor) EGFR	Pembrolizumab Anemia (5%)	Burtness and colleagues (8); KEYNOTE-048
		Pembrolizumab + chemotherapy Anemia (25%) Neutropenia (18%) Neutrophil count decreased (11%) Mucosal inflammation (10%) Thrombocytopenia (9%) Stomatitis (8%) Fatigue (7%) Hypokalemia (7%) Nausea (6%) Platelet count decreased (5%) WBC count decreased (5%) Decreased appetite (5%)
		Cetuximab + chemotherapy Neutropenia (21%) Anemia (17%) Neutrophil count decreased (13%) Thrombocytopenia (9%) WBC count decreased (9%) Nausea (6%) Hypokalemia (6%) Rash (6%) Fatigue (5%) Mucosal inflammation (5%) Hypomagnesemia (5%)
Pembrolizumab plus cabozantinibb	Immunotherapy (PD-1 inhibitor)	Dysphagia (8.3%) Hypertension (8.3%) Increased AST level (5.6%) Back pain (5.6%) Hypotension (5.6%) Oral mucositis (5.6%)	Saba and colleagues (56)
Pembrolizumab vs. pembrolizumab plus lenvatinibb	Immunotherapy (PD-1 inhibitor)	Pembrolizumab No single grade 3–4 TRAE was found in ≥5% of the population	Licitra and colleagues (57)
		Pembrolizumab + lenvatinib Hypertension Pneumonia Anemia Dysphagia Decreased weight Decreased appetite
Nivolumab vs. standard therapy (methotrexate, docetaxel, or cetuximab)c	Immunotherapy (PD-1 inhibitor)	Nivolumab No single grade three TRAE was found in ≥5% of the population	Ferris and colleagues (58) CheckMate 141
		Standard therapy Neutropenia (7.2%) Anemia (4.5%)
Afatinib vs. methotrexated	ERBB family blocker	Afatinib Rash or acne (10%) Diarrhea (9%) Stomatitis (6%) Fatigue (6%)	Machiels and colleagues (60); Lux-Head & Neck 1
		Methotrexate Stomatitis (8%) Neutropenia (6%) Anemia (5%) Leukopenia (5%)
Duligotuzumab vs. cetuximabe	EGFR and HER3	Duligotuzumab Infection and infestations MedDRA SOCf (22%)	Fayette and colleagues (61)
		Cetuximab Rash and related MedDRA termsg (7%) Infection and infestations MedDRa SOCf (12%) Hypomagnesemia (5%) Dyspnea (7%)
Cetuximab combination therapy
Cetuximab plus nivolumabd	EGFR and immunotherapy (PD-1 inhibitor)	1L: Rash acneiform (9%) Hypophosphatemia (5%) Immune-related AE: fatigue (5%)	Chung and colleagues (62); Chung and colleagues (63)
		2L+: Fatigue (13.3%) Immune-related AE: fatigue (6.7%)
Cetuximab plus pembrolizumabd	EGFR and immunotherapy (PD-1 inhibitor)	Oral mucositis (9%) Colitis (6%) Rash acneiform (6%) Hypomagnesemia (6%)	Sacco and colleagues (64)
Cetuximab plus ficlatuzumabc	EGFR and HGF	Infection (20%) Thromboembolism (10%) Hypophosphatemia (10%) Peripheral edema (10%)	Bauman and colleagues (17)
Cetuximab plus ficlatuzumabe	EGFR and HGF	Rash, acneiform (19%) Hypoalbuminemia (16%) Hypokalemia (13%) Lung infection (13%) Bacteremia (9%) Headache (9%) Abdominal pain (6%) Dysphagia (6%) Death NOS (6%) Anemia (6%) Skin and soft tissue infection (6%) Musculoskeletal pain (6%) Dyspnea (6%) Thromboembolic event (6%)	Bauman and colleagues (66)
Cetuximab plus tivantinibe	EGFR and c-MET	Neutrophil count decreased (12.5%) WBC decreased (12.5%) Lymphocyte count decreased (10%) Fatigue (7.5%) Stomatitis/pharyngitis (7.5%) Anemia (5%) Rash, acneiform (5%)	Kochanny and colleagues (67)
Cetuximab plus palbociclibd	EGFR and CDK4/6	Neutropenia (26.6%) Leukopenia (18.8%) Rash (9.4%) Thrombocytopenia (6.3%)	Adkins and colleagues (68)
Cetuximab plus CDX-3379h	EGFR and HER3	AEs related to CDX-3379: Diarrhea (20%) Dermatitis acneiform (20%) Hypomagnesemia (10%) Hypokalemia (10%) Mucosal inflammation (7%)	Bauman and colleagues (69)
		AEs related to cetuximab: Dermatitis acneiform (27%) Magnesium metabolism disorders (17%) Hypokalemia (10%) Mucosal inflammation (7%)
Cetuximab plus PX-866e	EGFR and PI3K	Dysphagia (13%) Fatigue (5%) Hypokalemia (5%)	Jimeno and colleagues (70)

Abbreviations: 1L, first line; 2L, second line; MedDRA, Medical Dictionary for Regulatory Activities; NOS, not otherwise specified; NR, not reported; SOC, standard of care; TEAE, treatment-emergent adverse event; WBC, white blood cell.

^aIncludes grade 3–5 events.

^bReported as all AEs with grade 3 or 4 toxicity.

^cAEs were reported as TRAEs and included grade 3 or 4 events.

^dAEs were reported as TRAEs or drug-related AEs; no grade 4 or 5 events were reported in ≥5% of patients.

^eAEs were presented only as grade ≥3.

^fMedDRA System Order Class Infections and Infestations terms: paronychia (22% in duligotuzumab arm vs. 10% in cetuximab arm; no grade 3 AEs), conjunctivitis, pneumonia, respiratory tract infection, bronchitis, lower respiratory tract infection, oral candidiasis, Candida infection, folliculitis, oral fungal infection, upper respiratory tract infection, cellulitis, sepsis, urinary tract infection, device-related infection, infection, nasopharyngitis, lung infection, nail infection, pharyngitis, oral herpes, abscess, abscess neck, bacteremia, Clostridium difficile infection, ear infection, fungal infection, groin abscess, infected bites, influenza, meningitis, mucosal infection, onychomycosis, osteomyelitis, rash pustular, rhinitis, skin bacterial infection, skin infection, staphylococcal infection, viral upper respiratory tract infection, abscess limb, acute sinusitis, bacteriuria, furuncle, herpes zoster, infected fistula, oral infection, pneumonia necrotizing, sinusitis, stoma site infection, and tracheostomy infection.

^gRash and related MedDRA terms: dermatitis acneiform, rash, rash maculopapular, rash macular, rash erythematous, rash pruritic, rash generalized, rash papular, rash pustular, genital rash, and mucocutaneous rash.

^hAEs were reported as TEAEs; one grade 4 AE related to cetuximab (magnesium metabolism disorders) was reported in 7% of patients.

Anti–PD-1 immune checkpoint inhibitors

In a phase Ib trial evaluating pembrolizumab in patients with HPV-negative (62%) and HPV-positive (38%) HNSCC, the overall objective response rate (ORR) was 18%. ORR was numerically lower (14%) in patients with HPV-negative HNSCC than in patients with HPV-positive (25%) HNSCC (53). Grade 3 treatment-related adverse events (TRAE) were reported in 17% of all patients and included increased alanine aminotransferase (3%), increased aspartate aminotransferase (AST; 3%), hyponatremia (3%), fatigue (2%), rash (2%), atrial fibrillation (2%), congestive heart failure (2%), diarrhea (2%), lymphopenia (2%), musculoskeletal pain (2%), and neck abscess (2%). Fourteen percent of patients with HPV-negative HNSCC had grade 3 TRAEs.

In the phase III, randomized KEYNOTE-048 trial, patients with HPV-negative (78.4%) and HPV-positive (21.6%) R/M HNSCC were treated with pembrolizumab alone or with chemotherapy (platinum and 5-fluorouracil) rather than the prior standard of care of cetuximab plus chemotherapy (8). Of note, patients were excluded if they had progressive disease within 6 months of prior systemic treatment. Patients were stratified by PD-L1 expression, p16 status, and performance status and were allocated 1:1:1 to treatment arms (8). With a median follow-up of 45 months, in the overall population, pembrolizumab alone had noninferior OS compared with cetuximab plus chemotherapy [mOS, 11.5 vs. 10.7 months; HR, 0.81; 95% confidence interval (CI), 0.68–0.97; P = 0.00994; ref. 55]. In patients with PD-L1 expression ≥1, pembrolizumab alone improved mOS compared with cetuximab plus chemotherapy [PD-L1 combined positive score (CPS) ≥20: 14.9 vs. 10.8 months; HR, 0.61; 95% CI, 0.46–0.81; P = 0.00034; PD-L1 CPS ≥1: 12.3 vs. 10.4 months; HR, 0.74; 95% CI, 0.61–0.89; P = 0.00080]. However, in patients with PD-L1 CPS <1, pembrolizumab failed to demonstrate OS benefit over cetuximab plus chemotherapy (mOS, 7.9 vs. 11.3 months; HR, 1.51; 95% CI, 0.96–2.37), highlighting that patients with HNSCC and low PD-L1 expression may not benefit from pembrolizumab alone (54). Grade ≥3 adverse events (AE) occurred in 55% of 300 treated patients in the pembrolizumab-alone arm, 85% of 276 in the pembrolizumab plus chemotherapy arm, and 83% of 287 in the cetuximab plus chemotherapy arm. In patients treated with pembrolizumab alone, the most common grade ≥3 AE was anemia (5%). The most common grade ≥3 AEs with the combination of pembrolizumab and chemotherapy were anemia (25%), neutropenia (18%), neutrophil count decreased (11%), and mucosal inflammation (10%) with pembrolizumab plus chemotherapy. With cetuximab plus chemotherapy, the most common grade ≥3 AEs were neutropenia (21%), anemia (17%), and neutrophil count decreased (13%).

In the phase III CheckMate 141 trial, patients received nivolumab or standard, single-agent systemic therapy (methotrexate, docetaxel, or cetuximab; refs. 58, 59). This trial included patients who experienced disease progression or recurrence within 6 months after the last dose of prior systemic therapy. Treatment effect on survival by p16 status was assessed in a post hoc analysis. With a minimum long-term follow up of 24.2 months, the mOS was 7.7 months with nivolumab versus 5.1 months with standard therapy (HR, 0.68; 95% CI, 0.54–0.86), with a slightly improved OS observed in patients with HPV-positive (mOS, 9.1 vs. 4.4 months; HR, 0.60, 95% CI, 0.37–0.97) versus HPV-negative HNSCC (mOS, 7.7 vs. 6.5 months; HR, 0.59, 95% CI, 0.38–0.92). Rates of any-grade and grade 3/4 TRAEs were higher in the standard therapy group than those in the nivolumab group (any grade: 79.3% vs. 61.9%; grade 3/4: 36.9% vs. 15.3%). The most common grade ≥3 TRAEs with standard therapy were neutropenia (7.2%) and anemia (4.5%); no grade 3 to 4 TRAEs occurred in ≥5% of the population treated with nivolumab.

Combination of anti–PD-1 and VEGFR inhibitors

Given the frequency of overexpression of VEGF and its receptors (VEGFR1–3) in HNSCC, combination approaches with immunotherapy (i.e., anti–PD-1) are being explored. Cabozantinib, a VEGFR tyrosine kinase inhibitor (TKI) with inhibitory activity against c-MET and AXL, decreased the activity of immunosuppressive regulatory T cells and MDSCs and increased infiltration of CD8⁺ T cells into the tumor in preclinical studies (35, 71). Therefore, in addition to blocking VEGFR, AXL, and c-MET, cabozantinib may improve the efficacy of anti–PD-1 therapy by increasing the number of intratumoral CD8⁺ T cells and decreasing the number of MDSCs (35).

The combination of cabozantinib plus pembrolizumab was evaluated in a phase II trial of patients with R/M HNSCC (NCT03468218; ref. 56). Of the 36 patients enrolled, 33% had HPV-negative and 47% had HPV-positive HNSCC. The ORR was 52%, median progression-free survival (mPFS) was 14.6 months, and mOS was 22.3 months. No association was found between OS and tumor HPV status. Interestingly, baseline CD8⁺ T-cell tumor infiltration positively correlated with clinical response. Furthermore, patients with PD-L1 CPS ≥20 had a considerable OS benefit compared with patients with PD-L1 CPS <20 (mOS, 32.9 vs. 14.6 months). The reported OS exceeded the historical control with pembrolizumab alone in this population. In contrast, in the LEAP-010 phase III trial of patients with R/M HNSCC with PD-L1–expressing tumors, a combination of pembrolizumab plus lenvatinib failed to improve OS compared with pembrolizumab plus placebo (NCT04199104; ref. 57). The contradictory outcomes observed with the PD-1 inhibitor and VEGFR TKI combination must be evaluated further and suggest validity of targeted therapy combination approaches that exclude immuno-oncology agents.

Results from ongoing trials evaluating these combinations, including the phase II BiCaZO trial of nivolumab plus cabozantinib in patients with advanced melanoma or R/M HNSCC (NCT05136196), will help inform their role in the HNSCC treatment paradigm.

HER family inhibitors

Afatinib, an irreversible ERBB family blocker, was evaluated in a phase III trial as second-line treatment in patients with R/M HNSCC (60). Patients were randomized 2:1 to receive afatinib or methotrexate. Evaluation of HPV status in 285 evaluable patients (59% of 483 total) showed that 208 (73%) had HPV-negative HNSCC and 49 (17%) had HPV-positive HNSCC. In the overall population, mPFS was longer in the afatinib group than that in the methotrexate group (2.6 vs. 1.7 months; HR, 0.80; 95% CI, 0.65–0.98; P = 0.030). A post hoc analysis by HPV status showed that the benefit of afatinib compared with methotrexate was more pronounced in patients with HPV-negative HNSCC (HR, 0.69; 95% CI, 0.50–0.96) than that in those with HPV-positive HNSCC (HR, 0.95; 95% CI, 0.51–1.75; interaction test P = 0.32). In the overall population, the most frequent grade 3/4 TRAEs in the afatinib and methotrexate groups were rash (10% and 0%), diarrhea (9% and 2%), stomatitis (6% and 8%), fatigue (6% and 3%), and neutropenia (<1% and 7%). Serious AEs occurred in 14% of patients with afatinib and 11% with methotrexate (60).

The efficacy and safety of duligotuzumab, a dual-action humanized IgG1 antibody that binds to EGFR and HER3, were compared with cetuximab in a phase II trial. ORR was similar with duligotuzumab and cetuximab (12% vs. 14.5%, respectively). Interestingly, responses in both arms were confined to patients with HPV-negative HNSCC. Duligotuzumab contributed to higher rates of serious AEs (41% vs. 29.5%) versus cetuximab (61).

Despite the near-ubiquitous aberrant EGFR signaling in HNSCC, cetuximab is limited by modest and short-lived efficacy and acquired resistance (17). Cetuximab resistance can result from compensatory signaling through the activation of alternate RTKs, including other HER family members and c-MET (66, 72). Evidence suggests that acquired cetuximab resistance is also mediated by increased angiogenesis, EGFR internalization and degradation, altered EGFR subcellular localization, depletion of epithelial-like cells and enrichment of mesenchymal-like cells, activation of effectors downstream of EGFR, and increased expression of HER family growth factors (72).

Additionally, evidence of cetuximab-mediated immunogenicity through antibody-dependent cell cytotoxicity (ADCC) has informed approaches evaluating cetuximab combination regimens (73). Besides blocking EGFR signaling through direct binding to its receptor, cetuximab has an IgG1 backbone that can bind CD16 fragment crystallizable receptors located on NK cells, inducing immunologic antitumor effects (Fig. 1B). The binding of the IgG1-fragment crystallizable part of cetuximab to CD16 on NK cells triggers ADCC cytolytic activity, which is predominantly mediated by perforin and granzymes.

Combination of cetuximab and anti–PD-L1 inhibitors

Prospective trials have evaluated cetuximab in combination with immunotherapeutic agents in R/M HNSCC (Tables 1 and 2). To evaluate the antitumor activity of PD-1 blockade with EGFR inhibition in R/M HNSCC, cetuximab in combination with nivolumab was studied in a phase II trial of 88 patients for response and survival outcomes (62, 63). Patients were stratified by receipt of prior therapy. Of the patients whose tumor tissues were assessed for p16, 55% were HPV-negative and 45% were HPV-positive. PFS and OS based on p16 IHC status did not show a difference. In patients who had received prior systemic therapy in the R/M setting (n = 45), mOS was 11.5 months, mPFS was 3.4 months, and the ORR was 22%. In patients who had not received prior therapy in the R/M setting, mOS was not reached, mPFS was 6.2 months, and the ORR was 37%. Grade ≥3 TRAEs that occurred in ≥2 patients with prior therapy were fatigue (n = 6; 13.3%) and rash (n = 2; 4.4%). Grade ≥3 TRAEs occurring in ≥2 patients without prior therapy were rash acneiform (n = 4, 9%), hypophosphatemia (n = 2; 5%), and hypomagnesemia (n = 2; 5%). Hypomagnesemia was the only grade 4 TRAE (2%). The only grade 3 immune-related AE occurring in ≥2 patients was fatigue (n = 3; 6.7% vs. n = 2; 5%) with and without prior therapy, respectively.

A phase II trial evaluated cetuximab in combination with pembrolizumab in patients with platinum-resistant, R/M HNSCC (64). HPV status was evaluated as an exploratory endpoint; 64% of patients were HPV-negative and 36% were HPV-positive. By 6 months, the ORR was 45% (95% CI, 28%–62%). More patients with HPV-negative HNSCC had a response by 6 months than those with HPV-positive HNSCC (57% vs. 25%, respectively), although the difference was not significant (P = 0.18). The most common grade 1/2 TRAEs were rash acneiform (76%), hypomagnesemia (33%), dry skin (33%), increased AST (27%), nausea (24%), and fatigue (21%). The most common grade ≥3 TRAEs were oral mucositis (9%), rash acneiform (6%), hypomagnesemia (6%), and colitis (6%); serious TRAEs occurred in 15% of the patients.

Cetuximab has also been evaluated in combination with durvalumab, a PD-L1 inhibitor, in a phase II trial in patients who had received prior systemic therapy for R/M HNSCC. Patients who had received both ICI therapy and cetuximab were excluded (65). In a post hoc analysis, evaluation of HPV status showed that 20% of patients had p16-negative tumors, 11.4% had p16-positive tumors, 2.9% had oropharyngeal tumors, and 65.7% non-oropharyngeal tumors with unavailable p16 status. mOS was 9.6 months, mPFS was 5.8 months, and the ORR was 39%. Of the 13 responders, six had HPV-negative HNSCC, one had HPV-positive HNSCC, and six had unknown HPV status. There were 16 grade 3 TRAEs and 1 grade 4 TRAE. The most common TRAEs included maculopapular rash (77%), acneiform rash (57%), fatigue (46%), nausea (31%), and hypomagnesemia (31%). Grade ≥3 TRAEs reported in ≥1 patient included blurred vision, colitis, localized edema, maculopapular rash, hypomagnesemia, syncope, alanine aminotransferase increased, alkaline phosphatase increased, AST increased, bilirubin increased, diabetic acidosis, hyperglycemia, acute kidney injury, aspiration pneumonia, pleural effusion, and thromboembolic event. NK cell cytotoxic activity increased with cetuximab and further increased in responders with the addition of durvalumab.

Cetuximab has also been studied in prospective trials in combination with agents targeting the HGF/c-MET signaling pathway, CDK proteins, other members of the HER family [e.g., Erb3 (HER3)], and PI3K (Tables 1 and 2).

Combination of cetuximab and HGF/c-MET inhibitors

Given that c-MET and its sole ligand, HGF, are overexpressed in 70% to 80% of HPV-negative HNSCC cases, and that compensatory signaling of the c-MET pathway may act as a resistance mechanism to EGFR inhibition, studies have evaluated the possible synergy of combining a c-MET inhibitor with an EGFR inhibitor in R/M HNSCC, including patients with HPV-negative HNSCC (Fig. 1C). A phase II trial evaluating the combination of cetuximab with the c-MET inhibitor tivantinib reported similar efficacy outcomes between groups, yielding an ORR of 7.5% with the combination regimen compared with 7.9% with cetuximab monotherapy (67). Interestingly, the ORR was higher in patients with HPV-negative HNSCC but still comparable between the treatment groups (12.5% vs. 13%, respectively). In all patients, survival was not significantly improved by the combination, with mOS of 7.4 months (95% CI, 4.7–10.3 months) for the tivantinib and cetuximab combination and 8.6 months (95% CI, 5.7–11.5 months) for the cetuximab arm (P = 0.99). In addition, compared with cetuximab alone, the combination treatment was associated with an increased rate of grade ≥3 hematologic toxicities, including decreased white blood cell count in 12.5% of patients, decreased neutrophil count in 12.5%, and decreased lymphocyte count in 10% (67).

Another study tested whether dual EGFR and HGF/c-MET targeting could overcome cetuximab resistance. In this phase I trial, ficlatuzumab, a specific anti-HGF IgG1 mAb, was evaluated in combination with cetuximab in platinum-refractory and cetuximab-resistant HNSCC (17). Ninety-two percent of patients had HPV-negative HNSCC. The ORR was 17%, mPFS was 5.4 months, and mOS was 8.9 months (17). Interestingly, responders demonstrated an 83% increase in effector CD8⁺T cells (17). Ficlatuzumab is mechanistically distinct from c-MET inhibitors such as tivantinib as it binds the sole ligand of c-MET rather than the c-MET receptor, sequestering HGF and preventing residual c-MET signaling (66, 67). Additionally, the combination of two IgG1 mAbs, cetuximab and ficlatuzumab, could trigger ADCC in the TME followed by T-cell cross priming.

These findings led to a phase II trial evaluating ficlatuzumab combined with cetuximab in patients with pan-refractory (i.e., resistant to cetuximab, platinum-based chemotherapy, and an anti–PD-1 mAb treatment) R/M HNSCC. In this trial, PFS in the combination arm was prolonged relative to a historical control of 2 months. The prespecified significance criterion was met in the combination arm, with mPFS of 3.7 months (lower-bound 90% CI, 2.3 months; P = 0.04). The ORR in all patients was 19% (95% CI, 7%–36%; ref. 66). When patients treated with the combination were stratified by HPV status, those with HPV-negative HNSCC experienced superior ORR (38% vs. 0%) and mPFS (4.1 vs. 2.3 months) compared with patients with HPV-positive HNSCC (66). The most common grade ≥3 toxicities in the combination arm were acneiform rash (19%), hypoalbuminemia (16%), and hypokalemia (13%). Lung infection was observed in 13% of patients.

Dual targeting of the HGF/c-MET pathway has demonstrated promising results in patients with pan-refractory, HPV-negative metastatic HNSCC (66). The tolerability and efficacy of ficlatuzumab in combination with cetuximab versus cetuximab are being evaluated in a phase III trial in patients with HPV-negative HNSCC who experienced relapse on systemic therapies including ICIs and chemotherapy (NCT06064877; ref. 74).

Combination of cetuximab and cell-cycle inhibitors

In a phase II trial evaluating the combination of the CDK4/6 inhibitor palbociclib with cetuximab, there was no significant difference in OS compared with cetuximab monotherapy in patients with relapsed HPV-negative R/M HNSCC (68). mOS with the palbociclib-plus-cetuximab combination versus cetuximab monotherapy was 9.7 versus 7.8 months (HR, 0.82; 95% CI, 0.54–1.3; P = 0.18). The most common grade 3 toxicities with combination treatment were neutropenia (26.6%), leukopenia (18.8%), and treatment-related rash (9.4%). Another CDK4/6 inhibitor, dalpiciclib, is being evaluated in combination with cetuximab in a phase II trial in patients with HPV-negative R/M HNSCC relapsed/refractory to PD-1 (NCT05721443); results have not yet been reported.

Combination of cetuximab and HER3 inhibitors

CDX-3379, a human IgG1 mAb that binds HER3 to prevent it from binding to its ligand, neuregulin-1, or from dimerizing with EGFR, was tested in HNSCC studies. In a phase II trial, CDX-3379 plus cetuximab combination therapy was evaluated in patients with R/M HPV-negative HNSCC whose disease had relapsed on standard therapies, including cetuximab and pembrolizumab (69). ORR in the combination arm was low (6.7%), with mPFS of 2.2 months and mOS of 6.6 months. The most common AEs were diarrhea (83%), hypomagnesemia (30%), hypokalemia (23%), and acneiform dermatitis (53%), with dose modification required in 70% of patients. Grade ≥3 TRAEs occurred in 53% of patients; the most common grade ≥3 AEs related to CDX-3379 were diarrhea (20%), dermatitis acneiform (20%), hypomagnesemia (13%), and hypokalemia (10%); dermatitis acneiform (27%), magnesium metabolism disorders (23%), and hypokalemia (10%) were cetuximab-related. The modest ORR, coupled with the excessive, dose-limiting toxicity of the combination, precluded further clinical development (69).

Combination of cetuximab and PI3K inhibitors

In a phase II trial including patients with HPV-negative (43%) and HPV-positive (57%) relapsed HNSCC, a combination of PX-866, a PI3K inhibitor, with cetuximab yielded similar efficacy outcomes but higher overall toxicities compared with the cetuximab monotherapy arm (70). Across all patients, the ORR was 10% versus 7% in the combination and cetuximab monotherapy arms, respectively. Overall, the most common AEs reported in the combination versus monotherapy arms included nausea (53% vs. 23%), rash (45% vs. 31%), vomiting (45% vs. 15%), fatigue (43% vs. 23%), diarrhea (40% vs. 21%), and hypokalemia (25% vs. 10%; ref. 70). The most common grade ≥3 AEs with the combination were dysphagia (13%), fatigue (5%), and hypokalemia (5%).

Emerging targets in HPV-negative HNSCC

Other emerging signaling pathways are being investigated as potential therapeutic targets in patients with HPV-negative HNSCC.

TGFβ inhibitors

TGFβ signaling is a key regulator of normal epithelial cell proliferation and development (75). In HNSCC, TGFβ signaling is commonly dysregulated and promotes cell invasion, metastasis, proliferation, and drug resistance (Fig. 1B). A preclinical study in HNSCC cell lines supported a protumorigenic role for the TGFβ pathway, suggesting that it serves as an auxiliary growth factor pathway (76). TGFβ2 was transcriptionally upregulated in response to EGFR- and FGFR-specific TKIs. Moreover, a triple combination of EGFR, FGFR, and TGFβ receptor 1 (TGFβR1) inhibitors blocked HNSCC cell growth more effectively than the TKIs alone. In preclinical head and neck cancer models, blockade of EGFR signaling with cetuximab led to increased TGFβ secretion and CAF formation, contributing to cetuximab resistance (77). BCA101, a bifunctional EGFR and TGFβ inhibitor, demonstrated an advantage over cetuximab in activating the immune system, which was known to be suppressed by the presence of TGFβ (78). In an ongoing phase IIb trial, a combination of BCA101 plus pembrolizumab was evaluated in a first-line setting in patients with R/M HNSCC with PD-L1 CPS ≥1 (79). The combination showed antitumor activity with additive potential among patients with HPV-negative HNSCC, with an ORR of 58% in this subgroup compared with 44% in all patients (NCT04429542).

Furthermore, the bifunctional checkpoint inhibitor bintrapfusp alfa (M7824), consisting of human TGFβRII (TGFβ trap) linked to anti–PD-L1, has been shown to decrease TGFβ-induced signaling in the TME (80). Based on these preclinical results, a clinical trial evaluated the antitumor activity of bintrapfusp alfa (M7824) in 14 patients with HPV-negative resectable HNSCC (NCT04247282; ref. 81). The 1-year recurrence-free rate was 86%. The most common TRAEs were oral hemorrhage (35.7%), fatigue (35.7%), epistaxis (35.7%), rash (28.6%), and pruritus (21.4%); grade ≥3 TRAE was vasculitis that occurred in one patient (7.1%).

HRAS

HRAS mutations occur in 4% to 8% of patients with R/M HNSCC and define a predominantly HPV-negative biologic subset characterized by enrichment for wild-type TP53 and caspase-8 mutations (82). Activating mutations in HRAS with inactivating CASP8 mutations and wild-type TP53 may drive an alternative tumorigenesis in HPV-negative HNSCC via RAS and alterations in cell death and NF-κB (22). The most common HRAS mutation in patients with HNSCC is G12S (83). Tipifarnib is a farnesyltransferase inhibitor that disrupts HRAS function. A phase II study enrolled 30 patients with R/M HNSCC (82). Of the 20 evaluable patients with high HRAS mutation variant allele frequency, the ORR was 55%, mPFS was 5.6 months, and mOS was 15.4 months. Grade ≥3 TEAEs in ≥10% of patients were hematologic-related events [anemia (37%), lymphopenia (13%), neutropenia (10%), and leukopenia (10)] and gastrointestinal disturbances [nausea (10%)].

WEE1

TP53 is one of the most frequently altered genes in HPV-negative HNSCC (22). Recently, the WEE1 inhibitor adavosertib has attracted attention because of its selective cytotoxicity against TP53-mutated cells and its promising activity in early-phase clinical trials (84). WEE1 is a kinase that regulates replication stress and cell-cycle inhibition at the G2/M checkpoint. DNA damage during the S-phase activates WEE1, phosphorylating CDK1 at Tyr15 to inactivate its kinase activity. Because TP53 regulates the G1/S checkpoint, TP53-mutated cells rely on the G2/M checkpoint to arrest the cell cycle to repair DNA damage. Thus, treatment of TP53-mutated cells with adavosertib causes the cells to enter the M-phase without DNA repair, which in turn causes cell death. Combining adavosertib with alisertib, an aurora kinase inhibitor, leads to synergistic antitumor effects in both in vitro and in vivo HNSCC models (85). These findings indicate a novel rational combination, providing a promising therapeutic pathway for TP53-mutated cancers.

Summary

HPV-negative R/M HNSCC is a group of heterogeneous malignancies with poor outcomes and limited efficacious therapies. To improve upon the status quo, a concerted effort at developing novel efficacious therapies with biologically rational targets and treatment combinations as well as discovering effective predictive biomarkers will be crucial. In this regard, a continued refining of the genomic and immunologic framework of HPV-negative HNSCC and the consequent development of practical disease classification systems aimed at informing treatment selection will be key in unlocking a breakthrough. Additionally, lessons learned from clinical and translational studies of other difficult-to-treat tumor types with shared molecular and immunologic characteristics such as those with inherent immune suppressive or hypoxic TMEs will be important in achieving success and avoiding repeating failures.

Authors’ Disclosures

C.H. Chung reports personal fees from AVEO Oncology, Seagen, Regeneron, Exelixis, Genmab, and Fulgent outside the submitted work. No disclosures were reported by the other author.