Radiosensitizing effect of galunisertib, a TGF-ß receptor I inhibitor, on head and neck squamous cell carcinoma in vitro

Head and Neck Squamous Cell Carcinoma (HNSCC) is the seventh most common type of cancer worldwide, affecting more than 5.5 million people, causing over 380,000 deaths every year [1‐3]. HNSCCs represents a group of tumors that can arise from the mucosa of the nasal or oral cavity, the pharynx or larynx. Despite multidisciplinary treatment approaches, advanced HNSCC has a 5-year overall survival rate of only ~ 50% [4]. Radiotherapy is an important cornerstone in the treatment of HNSCC patients, either as adjuvant therapy after surgery or as definitive radiochemotherapy in patients unfit for surgery or with locally advanced disease. Resistance to radiation is therefore particularly associated with a poor prognosis and represents a major clinical problem [5]. Strikingly, only the epidermal growth factor receptor antibody cetuximab has been approved as a radiosensitizing agent in the last 50 years [6].

Radioresistance is defined as either no or only partial response of the tumor to radiation therapy or early recurrence within a few weeks after initial response [5]. The biological mechanisms are complex and involve, among others, the Transforming Growth Factor (TGF) pathway with TGF-ß as one of its key members [7]. Although TGF-ß has tumor-suppressing functions in healthy cells and most early-stage cancers, its activation in late-stage disease can promote tumorigenesis [8]. Notably, increased TGF-ß expression has been found in 80% of HNSCC and TGF-ß expression levels correlate with more advanced disease and worse patient survival [9].

The experimental cancer drug galunisertib (LY2157299) is a small molecule inhibitor of the TGF-ß receptor I and has shown antineoplastic effects in various cancer entities in vitro and in vivo [10]. Furthermore, galunisertib is currently being tested in cancer clinical trials for several different cancers, highlighting the expected potential of this novel drug [11]. To date, however, no data exist on the effect of galunisertib on HNSCC. The aim of this study was, therefore, to preclinically evaluate the effect of galunisertib in single- and combination treatment with radiation on head and neck squamous cell carcinoma in vitro.

This is the first study to describe the in vitro effect of galunisertib in head and neck squamous cell carcinoma. Analysing three head and neck squamous cell carcinoma cell lines using standard preclinical assays, we observed varying antineoplastic effects in all tested cell lines at clinically relevant drug concentrations.

Radiation therapy is a cornerstone of treatment in a majority of patients suffering from head and neck squamous cell carcinoma. Radioresistance, therefore, poses a major problem, which could be potentially abrogated using radiosensitizing agents. Research into the molecular mechanisms of radioresistance has led to the identification of many potential targets [17]. Transforming growth factor-beta (TGF-ß) has emerged as such a target [18] and the availability of the small molecule TGF-ß receptor I inhibitor galunisertib, which is currently under clinical investigation as an antineoplastic agent, has led us to investigate its radiosensitizing potential in head and neck squamous cell carcinoma.

Since its first demonstration of antitumor effects in xenograft models of non-small lung cancer and breast cancer [19], galunisertib has progressed through phase I and II clinical trials and is one of the most advanced candidates among small molecule TGF-ß inhibitors [20]. Clinical studies investigating galunisertib in hepatocellular carcinoma [21] and pancreatic cancer [22] have shown improved outcome of the treatment group. Most importantly, various clinical trials reported a safe toxicity profile with no dose-limiting events [20]. Clinical trials in phase I and I/II investigating the combination of galunisertib with radiation treatment in hepatocellular carcinoma (NCT02906397) and malignant glioma (NCT01220271), respectively, are underway. Interestingly, we could not find any in vitro studies that investigated the combination of galunisertib with radiation. However, a study by Hardee et al. has shown increased sensitivity to radiation therapy in glioblastoma treated with the TGF-ß receptor I inhibitor LY364947 [23]. Furthermore, Yang et al. have shown that another TGF-ß receptor inhibitor (LY2109761) increased radiosensitivity in gastric cancer in vitro and in vivo [24], highlighting the expected potential of this treatment combination.

Here, we aimed at investigating the effect of galunisertib at clinically relevant doses. We, therefore, searched the literature for studies investigating the plasma concentration of galunisertib in clinical trials first and found that patients who received a dose of 150 mg twice daily reached a plasma concentration of ~ 10,000 ng/ml (corresponding to ~ 27 µM) [25]. We thus subsequently prespecified to investigate galunisertib at concentrations of 20 and 40 µM, since higher serum concentrations might not be achievable in the clinical setting. Notably, at comparable concentrations of up to 10 µM, galunisertib showed potent and selective inhibition of SMAD 2/3 phosphorylation in hepatocellular carcinoma cells [26] and anaplastic carcinoma cells [27].

We observed the strongest effects of galunisertib in SCC-25. Cell migration and clonogenic survival were significantly inhibited by single treatment and showed additive effects in combination with radiation, suggesting a radiosensitizing antineoplastic effect. Surprisingly, galunisertib treatment had a stimulating effect on metabolic activity, a surrogate marker for cell proliferation, in SCC-25, which also antagonized the inhibitory effect of radiation treatment. We furthermore observed a cell cycle shift with a significant reduction of cells in G₁ phase. This could potentially be explained by an inhibition of TGF-ß mediated G₁ phase arrest, which is often lost in malignant cells [28], but appears to be still intact in SCC-25. Indeed, it has been shown that TGF-ß signaling intermediates can remain intact and activate TGF-ß responsive promoters in some HNSCC cell lines [29]. It would appear that the increase in metabolic activity at 72 h could be associated with this effect, emphasizing the dichotomous role of TGF-ß in tumor progression [30]. Interestingly, a recent publication by Oshimori et al. identified TGF-ß-responding SCCs that displayed hallmarks of malignancy during a slow-cycling proliferation. Interestingly, those cells showed better protection against DNA damaging agents [31]. An inhibition of this mechanism might therefore explain more efficient radiation-induced DNA damage. In the other two tested cell lines, drug response was not as distinct. In FaDu, we could observe a reduction in metabolic activity and cell migration after galunisertib treatment. In CAL27, an inhibitory effect could be observed for cell migration after galunisertib treatment with an additive effect with radiation. Consistent with our results, galunisertib showed inhibitory effects on cell migration in primary cholangiocarcinoma cells [32] and hepatocellular carcinoma [26] at comparable concentrations of 50 and 10 µM, respectively. Furthermore, a study by Zhang et al. showed a reduction in cell viability in ovarian cancer cells in vitro, however, at concentrations above 100 µM [33].

Mechanistically, the observed differences between cell lines could be explained by different degrees of TGF-ß escape. Martin et al. have shown that CAL27 harbour an inactivating SMAD4 mutation, which is strictly required for downstream signaling of TGF-ß receptor I [34]. FaDu has been shown to be SMAD4-deficient [35]. Notably, galunisertib has shown high selectivity for the TGF-ß pathway and potent inhibition of TGF-ß mediated SMAD2/3 phosphorylation [36]. This poses the question of the underlying mechanism of observed differences between CAL27 and FaDu. Interestingly, while CAL27 remained insensitive to TGF-ß stimulation after SMAD4 re-expression in a study by Martin et al. [34], suggesting TGF-ß receptor alterations, FaDu showed partial restoration of TGF-ß responsiveness in a study by Hummer et al. [35]. In a subsequent analysis of tissue microarrays for SMAD4 expression, Martin et al. found detectable SMAD4 expression in ~ 82% of HNSCC patients by immunohistochemistry, suggesting that a substantial patient population could potentially benefit from a TGF-ß receptor inhibitor therapy.

Here, we explored the effect of TGF-ß receptor I inhibition in HNSCC monoculture, neglecting the paracrine signaling observed in the tumor microenvironment. However, autocrine TGF-ß potentiation appears to play an important role in tumor progression and could explain the observed effects. [37]. Therefore, our in vitro results cannot be directly applied to predict the response of HNSCC to Galunisertib treatment in vivo. The evident importance of the tumor microenvironment in TGF-ß signaling calls for more advanced cancer models, including cells of the tumor stroma, in future studies [38]. Notably, similar studies investigating the effect of Galunisertib both in vitro and in vivo, found consistent antineoplastic effects [33, 39, 40]. It can therefore reasonably be assumed that the observed effect of Galunisertib on HNSCC cell lines might also be relevant in vivo.

In conclusion, we demonstrated that galunisertib shows significant radiosensitizing and antineoplastic effects in one out of three tested HNSCC cell lines in vitro. This observation correlates with evidence for TGF-ß signaling alterations in the two less responsive cell lines, therefore providing a rationale for further investigation of this new drug.