Head and neck squamous cell carcinomas (HNSCC) are among the most common solid cancers worldwide with an annual incidence of over 600,000 cases [1
]. Besides the two well-established risk factors, tobacco and alcohol consumption, which account for more than 70% of all HNSCC [2
], infection by high-risk types of human papillomavirus (HPV), in particular, HPV16, is a causative factor for an increasing incidence of oropharyngeal squamous cell carcinoma (OPSCC) [3
]. Recent studies showed HPV-related OPSCC to be a distinct clinical and molecular entity [5
]. Patients affected by HPV-related OPSCC tend to be younger and have better outcomes. This has led to the introduction of de-escalation strategies to reduce long-term toxicity [7
]. However, the global 5-year-survival rate for all HNSCC sites is only 40–50% [10
], although there are substantial differences among countries. Global survival rates are comparable with the survival in Germany where our study was based [11
]. Current therapeutic paradigms do not adequately address the distinct clinical and biological heterogeneity of HNSCC with varying treatment responses [12
]. Diagnosis of most HNSCC at more advanced tumor stages also contributes to the unfavorable prognosis. Since therapeutic resistance frequently develops, the successful treatment of HNSCC is significantly impaired [14
]. Therefore, there is an urgent demand in valuable biomarkers and therapeutic targets for HNSCC patients.
Mitogen-activated protein kinase (MAPK) cascade is a critical pathway for transduction extracellular signals to cellular responses, which plays a significant role in tumor cell survival, proliferation, and resistance to current therapies [15
]. Amongst those are the MEK/ERK1/2 signaling components. They are the final effectors of the MAPK signaling pathway. Previous studies addressing radioresistance in HNSCC provided compelling experimental evidence that activation of MEK-ERK1/2 signaling mediates the inadequate response to therapy [16
]. Moreover, inhibition of irradiation-induced MEK-ERK1/2 activation, significantly suppressed colony forming ability, and enhanced apoptosis in epithelial cancer cells [17
]. Activation of the phosphatidylinositol-3-kinase (PI3-K)/protein kinase B (AKT) pathway plays a central role in numerous cellular processes, including metabolism, cell growth, apoptosis, survival, and differentiation [20
]. The cascade is also known for impacting on radioresistance development in HNSCC [21
]. HPV-positive oropharyngeal SCC showed activating mutations of PIK3CA in 56% [24
]. In a recent study, we found pAKT (Ser473) expression serves as an independent prognostic marker for progression-free survival in OPSCC [25
]. So far, only inhibitors of the epidermal growth factor receptor (EGFR) upstream of MAPK/ERK signaling have gained approval for HNSCC treatment [26
]. Reliable molecular biomarkers for early detection as well as for the response to treatment are lacking for HNSCC. Activation of MAPK/ERK signaling is detrimental to response characteristics in HNSCC and MEK/ERK inhibitors are actually present in clinical trials. Hence, it is obvious to examine the impact of activated MAPK/ERK as a prognostic biomarker. To the best of our knowledge, there are no data yet available on how activation of ERK1/2 by phosphorylation is associated with clinical parameters and particularly with the HPV status in OPSCC. In the present study, we aim to shed new light on the role of ERK1/2 activation in relation to the HPV status as a potential predictive biomarker.
In the present study, we found that HPV-related OPSCCs express lower pERK1/2 staining patterns than non-HPV-related counterparts, which suggested that well-documented differences in the survival outcome might be in part attributed to lower activation of MAPK/ERK signaling as a cancer cell intrinsic feature. MAPK/ERK signaling is critical for survival, dissemination, and resistance to therapy in numerous human cancer cells, and ERK1/2 phosphorylation results in the activation of multiple downstream substrates that are critically implicated in these processes [32
]. Increased pERK1/2 and total ERK1/2 expression have been associated with radioresistance and poor prognosis in nasopharyngeal carcinoma [33
] and clival chordomas [34
]. This is in line with our previous results demonstrating that ERK1/2 activation impairs radiosensitivity of HNSCC cells [18
]. There is evidence for pERK1/2 being a prognostic biomarker in several malignancies, including breast cancer [36
], malignant melanoma [40
], and hepatocellular carcinoma [41
]. High expression of pERK1/2 enhances tumorigenicity and metastasis and has been linked to poor prognosis in esophageal cancer [42
], hepatocellular carcinoma [43
], and ovarian cancer [44
]. However, it is not known how ERK signaling is associated with survival outcome in HNSCC or OPSCC, especially in relation to their HPV status.
Our study describes activated ERK1/2 as a strong independent risk factor for poor prognosis in OPSCC. In detail, 68 of the 124 OPSCC samples showed high pERK1/2 IRS which had a significantly reduced progression-free and overall survival. The observed HR of 2.042 (95% CI 1.229–3.392) reflects an increase in the hazard for death by factor 2.042 for each level of the pERK1/2 high scores in the PFS interval and HR of 1.844 (95% CI 1.068–3.185) in the DSS interval. Specifically, tumor size, smoking, HPV association, and high pERK1/2 expression were determined to be risk factors in our cohort, calculated by univariate Cox regression analysis for PFS and OS. Not surprisingly, we found nicotine consumption to serve as an independent risk factor for survival outcome of OPSCC patients. So far, HPV status has been demonstrated as the most suitable independent predictor of survival in OPSCC patients. HPV-positive OPSCC have a significant survival advantage over HPV-negative tumors, with a 58% reduction in mortality risk [45
]. HPV-related HNSCC tumors have a higher response to treatment as ionizing radiation that might explain their favorable outcomes [46
]. High pERK1/2 IRS were strongly associated with HPV-negative OPSCC (61 out of 68 samples, p
< 0.001). Vice versa, in HPV-related samples we found low pERK1/2 expression in 22/29 cases. Proliferation of HPV-positive cancer cells is mainly triggered by viral oncoproteins, in particular E7 [47
]. As a consequence, activation of signaling cascades, including MAPK/ERK driving cell cycle progression seems to be less relevant in HPV-positive OPSCC which is supported by the fact that there are less genetic alterations in regulators of MAPK signaling, e.g., EGFR [48
]. EGFR alterations (high gene copy numbers, overexpression) have been even found to be inversely correlated to HPV status in OPSCC [7
]. Furthermore, ERK phosphorylation has been shown to be impaired by viral oncoproteins. The binding partner of E6, the ubiqutin ligase UBE3A was reported to dampen down basal level ERK activation through removing the p53 tumor suppressor protein in cervical cancer cells [49
]. Our assumption concerning the relationship between ERK activation and HPV is supported by our finding regarding pERK1/2 expression and histopathological grading. High pERK1/2 levels were significantly associated with better cellular differentiation (G1/2) while poorly differentiated tumors were linked to low pERK1/2 expression. It has been shown that HPV-driven HNSCC frequently have a poorly differentiated histopathology [50
] which is in line with our data.
The EGFR as upstream of ERK1/2 and AKT is expressed in more than 90% of all HNSCC [52
]. Few data exist on the relationship between EGFR and HPV-induced oropharyngeal cancers. Several studies indicate that there is an inverse correlation between HPV infection and EGFR protein expression [54
]. The relationship of HPV status with EGFR protein expression and survival outcome has been addressed in a few studies. These suggest that the best outcomes are observed in patients with HPV-related OPSCC with low EGFR expression and the worst in HPV-negative with high EGFR expression or high EGFR gene copy number, respectively [56
]. This is in accordance with our findings. We assume that in our cohort MEK-ERK1/2 and PI3K/AKT are most likely activated by upstream factors such as EGFR in HPV-negative OPSCC showing worse survival outcome. Co-activation of both pathways MEK-ERK1/2 and PI3K/AKT was common in HPV-negative OPSCC, however, additional AKT (Ser473) phosphorylation did not increase the predictive power of pERK1/2. As OPSCC with co-activation of both pathways MEK-ERK1/2 and PI3K/AKT did not show worse outcome and were not significantly correlated with clinical and histopathological features we propose that these patients might not benefit from targeted therapy against MAPK/ERK. In summary, we demonstrated that phosphorylation of ERK1/2 is a rare event in HPV-positive OPSCC but quite frequent in HPV-negative tumors. Our data support the hypothesis that low levels of intrinsic ERK1/2 activation contribute partially to the positive prognosis of HPV-associated OPSCC.
Our study was limited by the small sample size and the fact that not all of the patients received intensity-modulated radiation therapy causing better tumor targeting. However, we were able to identify activated ERK1/2 as a strong prognostic indicator in OPSCC for the first time. As we already suggested from previous in vitro and ex vivo data, the utilization of small molecule inhibitors against MEK-ERK1/2 kinases might prove itself as a promising specific therapeutic agent. The subgroup of patients which is supposed to benefit most is HPV-negative and show high pERK1/2 expression according to our findings. We assume that activated ERK1/2 decreases radiosensitivity, especially in HPV-negative OPSCC. OPSCC patients with HPV-negative but high pERK expression may be candidates for MEK-targeted therapy because such tumors show a more aggressive clinical course and are likely to respond to ERK inhibition. Further studies in a larger cohort of clinical samples are necessary to address the clinical significance of varying degrees of ERK1/2 pathway activation.