*Article* **Assessment of Epinephrine and Norepinephrine in Gastric Carcinoma**

**Alina Maria Mehedint,eanu 1,†, Veronica Sfredel 2,†, Puiu Olivian Stovicek <sup>3</sup> , Michael Schenker <sup>1</sup> , Georgică Costinel Târtea 2,\* , Octavian Istrătoaie 4,\*, Ana-Maria Ciurea <sup>1</sup> and Cristin Constantin Vere <sup>5</sup>**


**Abstract:** The aim of our study was to assess the sympathetic nervous system's involvement in the evolution of gastric carcinoma in patients by analyzing the mediators of this system (epinephrine and norepinephrine), as well as by analyzing the histological expression of the norepinephrine transporter (NET). We conducted an observational study including 91 patients diagnosed with gastric carcinoma and an additional 200 patients without cancer between November 2017 and October 2018. We set the primary endpoint as mortality from any cause in the first two years after enrolment in the study. The patients were monitored by a 24-h Holter electrocardiogram (ECG) to assess sympathetic or parasympathetic predominance. Blood was also collected from the patients to measure plasma free metanephrine (Meta) and normetanephrine (N-Meta), and tumor histological samples were collected for the analysis of NET expression. All of this was performed prior to the application of any antineoplastic therapy. Each patient was monitored for two years. We found higher heart rates in patients with gastric carcinoma than those without cancer. Regarding Meta and N-Meta, elevated levels were recorded in the patients with gastric carcinoma, correlating with the degree of tumor differentiation and other negative prognostic factors such as tumor invasion, lymph node metastasis, and distant metastases. Elevated Meta and N-Meta was also associated with a poor survival rate. All these data suggest that the predominance of the sympathetic nervous system's activity predicts increased gastric carcinoma severity.

**Keywords:** gastric carcinoma; norepinephrine transporter; plasma free metanephrines and normetanephrines

## **1. Introduction**

Gastric cancer is a malignant disease with a high degree of lethality; according to Globocan 2018, it ranks fifth in terms of the incidence of malignancies, with 1,033,701 cases annually (5.7%). It is also the third leading cause of cancer mortality, with an annual death toll of approximately 782,685 (8.2%) [1]. This condition is much more common among men, occupying third place for the total number of neoplasms, whereas this condition is ranked in fifth place regarding incidence for women [1]. The factors leading to this disease have not been identified exactly, but strong correlations have been found between its occurrence and diet (i.e., a diet rich in salty and smoked foods), *Helicobacter pylori* infection, vitamin deficiency, a low consumption of fruits and vegetables, smoking, a family history of gastric cancer, stress, and long-term stomach inflammation [2].

**Citation:** Mehedint,eanu, A.M.; Sfredel, V.; Stovicek, P.O.; Schenker, M.; Târtea, G.C.; Istr˘atoaie, O.; Ciurea, A.-M.; Vere, C.C. Assessment of Epinephrine and Norepinephrine in Gastric Carcinoma. *Int. J. Mol. Sci.* **2021**, *22*, 2042. https://doi.org/ 10.3390/ijms22042042

Academic Editor: Hidekazu Suzuki

Received: 1 February 2021 Accepted: 14 February 2021 Published: 18 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

The evolution of gastric cancer is unpredictable, and because of the extensive nature of the diagnostic methods, as well as diagnosis often occurring in advanced stages of the disease, the disease only becomes symptomatic in the advanced stages for the vast majority of patients [3]. The involvement of the autonomic nervous system in the development and evolution of gastric cancer has not been fully elucidated, but numerous studies have shown that there is a close relationship. Not only does the autonomic nervous system innervate the digestive tract but the tumor cells secrete growth factors and exhibit elevated levels of catecholamines and various receptors [4–6]. The involvement of the vegetative nervous system was first demonstrated by Batsakis approximately 30 years ago when he described the presence of nerves located in the vicinity of human epithelial carcinomas, such as gastric, head and neck, or prostate cancers [7,8]. These nerves have been described as directly involved in metastatic dissemination through a process called perineural invasion (PNI) in which neoplastic cancer cells are able to invade and migrate into, around, and through the nerves, with PNI frequently being associated with poor clinical results [9].

In this study, we wanted to evaluate the involvement of the sympathetic nervous system in the evolution of patients with gastric carcinoma by analyzing the mediators of this system (epinephrine and norepinephrine) as well as the histological expression of the norepinephrine transporter (NET). The norepinephrine transporter is a monoamine transporter responsible for capturing extracellular norepinephrine (N-Meta), also known as noradrenaline. The latter has an inhibitory role in the gastrointestinal tract. This is also true for epinephrine, which mainly enters the gastrointestinal tract through the bloodstream after being secreted by the adrenal medulla directly into the circulation. It is known that this transporter is also involved in the uptake of extracellular dopamine; the reuptake of the two neurotransmitters plays an important role in regulating their concentrations in the synaptic terminals [10–13].

In order to obtain homogeneous data, we chose to analyze the influence of epinephrine and norepinephrine only on gastric tumors because of the physiological particularities of the stomach's innervation. It is necessary to mention that the stomach is much more dependent on extrinsic neural inputs, represented by nuclei located in the caudal brainstem, from which sympathetic and parasympathetic pathways start or are controlled. In contrast to the stomach, the small and large intestines have a high degree of independent neuronal control and can function even if they lack extrinsic neural inputs [14].

#### **2. Results**

#### *2.1. Assessment of the Heart Rates of the Patients Included in the Study*

To assess the predominance of the sympathetic or parasympathetic autonomic nervous system, we evaluated 91 patients suffering from gastric cancer using a Holter electrocardiogram (ECG) for approximately 24 h, both at the time of diagnosis and before starting any antineoplastic therapy, by calculating the average heart rate (HR) during the day, during the night, and for 24 h. For the controls, 200 patients without gastric carcinoma who belonged to the same age group as the gastric carcinoma patients were evaluated using a Holter ECG. We observed (Figure 1A,B) that during the day, higher heart rates predominated in the group of patients with gastric carcinoma than in the control group (HR during the day = 90.76 ± 13.64 beats per minute (bpm) in the group of patients with gastric carcinoma versus 82.29 ± 7.86 bpm in the control group; *p* = 0.0012). These differences were maintained overnight (HR at night = 65.74 ± 16.44 bpm in the gastric carcinoma patients versus 58.21 ± 5.11 bpm in the control group; *p* = 0.0015) and for the entire monitoring period of approximately 24 h (HR for the 24 h period = 78.25 ± 14.01 bpm in the gastric carcinoma patients versus 68.96 ± 6.80 bpm in the control group; *p* < 0.0000). Another observation was that the patients with gastric carcinoma had no significant difference in heart rate during the day compared to that during the night, as shown in patients without cancer (Figure 1C,D). All these data suggest an increased predominance of sympathetic nervous system influences in the patients with gastric carcinoma versus patients without cancer.

without cancer.

**Figure 1.** Evaluation of heart rates (HRs) by Holter electrocardiogram (ECG) of the patients included in the study. (**A**) Minimum, maximum, and average HRs in patients with gastric carcinoma and patients without cancer, with (**B**) a volcano association plot. (**C**) Representative histogram of HR for a 24 h period of a patient with gastric carcinoma; approximately the same HRs can be observed during the day and night (yellow arrows). (**D**) Representative histogram of HR for a 24 h period of a cancer-free patient; it was higher during the day (as normal) and lower at night (blue arrows). Av HR, atrioventricular heart rate; bpm, beats per minute. Student's *t*-test, \*\*\* *p* < 0.001 and \*\*\*\* *p* < 0.0001. **Figure 1.** Evaluation of heart rates (HRs) by Holter electrocardiogram (ECG) of the patients included in the study. (**A**) Minimum, maximum, and average HRs in patients with gastric carcinoma and patients without cancer, with (**B**) a volcano association plot. (**C**) Representative histogram of HR for a 24 h period of a patient with gastric carcinoma; approximately the same HRs can be observed during the day and night (yellow arrows). (**D**) Representative histogram of HR for a 24 h period of a cancer-free patient; it was higher during the day (as normal) and lower at night (blue arrows). Av HR, atrioventricular heart rate; bpm, beats per minute. Student's *t*-test, \*\*\* *p* < 0.001 and \*\*\*\* *p* < 0.0001.

#### *2.2. Relationship between Norepinephrine Transporter Expression and Clinicopathological 2.2. Relationship between Norepinephrine Transporter Expression and Clinicopathological Features*

other observation was that the patients with gastric carcinoma had no significant difference in heart rate during the day compared to that during the night, as shown in patients without cancer (Figure 1C,D). All these data suggest an increased predominance of sympathetic nervous system influences in the patients with gastric carcinoma versus patients

*Features* The expression of the norepinephrine transporter was analyzed in samples from patients with gastric carcinoma (N = 91) as well as from the 200 patients (controls) who required gastric resection for benign reasons (Figure 2A–D). Using multispectral microscopy (Figure 3A–D), we analyzed the expression of the norepinephrine transporter, calculating the integrated optical density (IOD) only for the target color signal. Depending on the tumor grading, we observed an increase in IOD from well-differentiated (G1) to moderately differentiated (G2) and poorly differentiated (G3) tumors (Figure 4A and Supplementary Table S1). The norepinephrine transporter expression was higher in the patients with gastric carcinoma in those aged < 60 years (*p* = 0.0115) and in those with localization of the tumor in the gastric body or pyloric area (*p* = 0.0033), with tumor invasion The expression of the norepinephrine transporter was analyzed in samples from patients with gastric carcinoma (N = 91) as well as from the 200 patients (controls) who required gastric resection for benign reasons (Figure 2A–D). Using multispectral microscopy (Figure 3A–D), we analyzed the expression of the norepinephrine transporter, calculating the integrated optical density (IOD) only for the target color signal. Depending on the tumor grading, we observed an increase in IOD from well-differentiated (G1) to moderately differentiated (G2) and poorly differentiated (G3) tumors (Figure 4A and Supplementary Table S1). The norepinephrine transporter expression was higher in the patients with gastric carcinoma in those aged < 60 years (*p* = 0.0115) and in those with localization of the tumor in the gastric body or pyloric area (*p* = 0.0033), with tumor invasion T3–4 (*p* = 0.0093), with lymph node metastasis N≥<sup>2</sup> (*p* = 0.0371), and with TNM classification of malignant tumors (TNM) stages TIII–IV (*p* = 0.003) (Figure 5a and Supplementary Table S2).

T3–4 (*p* = 0.0093), with lymph node metastasis N≥2 (*p* = 0.0371), and with TNM classification of malignant tumors (TNM) stages TIII–IV (*p* = 0.003) (Figure 5a and Supplementary Table

S2).

*Int. J. Mol. Sci.* **2021**, *22*, 2042 4 of 14

*Int. J. Mol. Sci.* **2021**, *22*, 2042 4 of 14

**Figure 2.** Examples of histological samples collected from the patients included in the study. (**A**) Normal gastric mucosa. (**B**) Gastric carcinoma with a high degree of cell differentiation—G1. (**C**) Gastric carcinoma with a moderate degree of cell differentiation—G2. (**D**) Gastric carcinoma with a poor degree of cell differentiation—G3. Hematoxylin and eosin staining. Magnification, 400×. Scale bars represent 20 µm. **Figure 2.** Examples of histological samples collected from the patients included in the study. (**A**) Normal gastric mucosa. (**B**) Gastric carcinoma with a high degree of cell differentiation—G1. (**C**) Gastric carcinoma with a moderate degree of cell differentiation—G2. (**D**) Gastric carcinoma with a poor degree of cell differentiation—G3. Hematoxylin and eosin staining. Magnification, 400×. Scale bars represent 20 µm. **Figure 2.** Examples of histological samples collected from the patients included in the study. (**A**) Normal gastric mucosa. (**B**) Gastric carcinoma with a high degree of cell differentiation—G1. (**C**) Gastric carcinoma with a moderate degree of cell differentiation—G2. (**D**) Gastric carcinoma with a poor degree of cell differentiation—G3. Hematoxylin and eosin staining. Magnification, 400×. Scale bars represent 20 µm.

unmixing microscopy: (**B**) well-differentiated (G1), (**C**) moderately differentiated (G2), and **(D)** poorly differentiated (G3). (1) Images from the optical microscopy; (2) slides immuno-stained for norepinephrine transporters with 3,3′-diaminobenzidines (DAB) and counterstained with hematoxylin (signals are shown overlapping); (3) images with pure hematoxylin only (nuclei and cell membranes); (4) images with pure DAB only (only the signal for norepinephrine transporters). Magnification, 400×. Scale bars represent 20 µm. **Figure 3.** Assessment of the expression of norepinephrine transporters in normal (**A**) and tumor gastric tissue (**B**–**D**) by spectral unmixing microscopy: (**B**) well-differentiated (G1), (**C**) moderately differentiated (G2), and **(D)** poorly differentiated (G3). (1) Images from the optical microscopy; (2) slides immuno-stained for norepinephrine transporters with 3,3′-diaminobenzidines (DAB) and counterstained with hematoxylin (signals are shown overlapping); (3) images with pure hematoxylin only (nuclei and cell membranes); (4) images with pure DAB only (only the signal for norepinephrine transporters). Magnification, 400×. Scale bars represent 20 µm. **Figure 3.** Assessment of the expression of norepinephrine transporters in normal (**A**) and tumor gastric tissue (**B**–**D**) by spectral unmixing microscopy: (**B**) well-differentiated (G1), (**C**) moderately differentiated (G2), and **(D)** poorly differentiated (G3). (1) Images from the optical microscopy; (2) slides immuno-stained for norepinephrine transporters with 3,30 diaminobenzidines (DAB) and counterstained with hematoxylin (signals are shown overlapping); (3) images with pure hematoxylin only (nuclei and cell membranes); (4) images with pure DAB only (only the signal for norepinephrine transporters). Magnification, 400×. Scale bars represent 20 µm.

**Figure 3.** Assessment of the expression of norepinephrine transporters in normal (**A**) and tumor gastric tissue (**B**–**D**) by spectral

**Figure 4.** (**A**) Norepinephrine transporter expression quantified according to the integrated optical density (IOD) in cancer-free patients and in different tumor differentiation gradings in patients with gastric carcinoma. Plasma levels of free metanephrine (**B**) and normetanephrine (**C**) in cancer-free patients and in different stages of tumor differentiation in patients with gastric carcinoma. Correlations between metanephrine and norepinephrine transporter (**D**), normetanephrine and normetanephrine transporter (**E**), and normetanephrine and metanephrine (**F**). Kaplan–Meier curves depending on the norepinephrine transporter IOD (**G**), on plasma free metanephrine (**H**), and on plasma free normetanephrine (**I**). One-way ANOVA; \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001, and \*\*\*\* *p* < 0.0001. **Figure 4.** (**A**) Norepinephrine transporter expression quantified according to the integrated optical density (IOD) in cancer-free patients and in different tumor differentiation gradings in patients with gastric carcinoma. Plasma levels of free metanephrine (**B**) and normetanephrine (**C**) in cancer-free patients and in different stages of tumor differentiation in patients with gastric carcinoma. Correlations between metanephrine and norepinephrine transporter (**D**), normetanephrine and normetanephrine transporter (**E**), and normetanephrine and metanephrine (**F**). Kaplan–Meier curves depending on the norepinephrine transporter IOD (**G**), on plasma free metanephrine (**H**), and on plasma free normetanephrine (**I**). One-way ANOVA; \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001, and \*\*\*\* *p* < 0.0001.

#### *2.3. Relationship between Plasma Free Metanephrine (Meta) and Normetanephrine and Clinicopathological Features* Another analysis performed in our study was measuring the free metanephrine and nor-*2.3. Relationship between Plasma Free Metanephrine (Meta) and Normetanephrine and Clinicopathological Features*

metanephrine in the plasma for all the patients included before starting any therapy. Both the plasma free metanephrine and plasma free normetanephrine were higher in the patients with gastric carcinoma than those without. The cancer-free patients had plasma free metanephrine (Meta) values of 34.05 ± 13.23 pg/mL and plasma free normetanephrine (N-Meta) values of 128.3 ± 46.70 pg/mL. In the patients with gastric carcinoma, increased plasmatic levels of metanephrine and normetanephrine were correlated with tumor grading, increasing from welldifferentiated (Meta = 48.09 ± 16.45 pg/mL and N-Meta = 152.1 ± 57.05 pg/mL) to moderately differentiated (Meta = 54.14 ± 19.59 pg/mL and N-Meta = 178.1 ± 65.53 pg/mL) and poorly differentiated (Meta = 59.13 ± 21.88 pg/mL and N-Meta = 225.4 ± 91.22 pg/mL) tumors (Figure 4B,C). In terms of the clinicopathological features concerned, we observed that higher free metanephrine could be found in patients with gastric carcinoma according to the histological type, i.e., adenocarcinoma (Ad.c.) as opposed to mixed carcinoma/signet-ring-cell carcinoma (M.c./S.r.c.c.) (*p* = 0.0004); the location of the tumor in the gastric body or pyloric area (*p* = 0.0047); tumor invasion T3–4 (*p* = 0.0165); lymph node metastasis N≥2 (*p* = 0.0473); and TNM stage TIII–IV (*p* = 0.0148) (Figure 5B and Supplementary Table S3). We also observed higher plasma free normetanephrine in patients with gastric carcinoma with a tumor size ≥5 cm (*p* = 0.0217), histological type Ad.c as opposed to M.c./S.r.c.c. (*p* = 0.0253), a location of the tumor in the gastric body or pyloric area (*p* = 0.0132), a tumor invasion T3–4 (*p* = 0.0177), a lymph node metastasis N≥2 (*p* = 0.0127), and a TNM stage TIII–IV (*p* = 0.0275) (Figure 5C and Supplementary Table S4). It should be noted that patients who had the signet-ring-cell gastric cancer type, Another analysis performed in our study was measuring the free metanephrine and normetanephrine in the plasma for all the patients included before starting any therapy. Both the plasma free metanephrine and plasma free normetanephrine were higher in the patients with gastric carcinoma than those without. The cancer-free patients had plasma free metanephrine (Meta) values of 34.05 ± 13.23 pg/mL and plasma free normetanephrine (N-Meta) values of 128.3 ± 46.70 pg/mL. In the patients with gastric carcinoma, increased plasmatic levels of metanephrine and normetanephrine were correlated with tumor grading, increasing from well-differentiated (Meta = 48.09 ± 16.45 pg/mL and N-Meta = 152.1 ± 57.05 pg/mL) to moderately differentiated (Meta = 54.14 ± 19.59 pg/mL and N-Meta = 178.1 ± 65.53 pg/mL) and poorly differentiated (Meta = 59.13 ± 21.88 pg/mL and N-Meta = 225.4 ± 91.22 pg/mL) tumors (Figure 4B,C). In terms of the clinicopathological features concerned, we observed that higher free metanephrine could be found in patients with gastric carcinoma according to the histological type, i.e., adenocarcinoma (Ad.c.) as opposed to mixed carcinoma/signet-ring-cell carcinoma (M.c./S.r.c.c.) (*p* = 0.0004); the location of the tumor in the gastric body or pyloric area (*p* = 0.0047); tumor invasion T3–4 (*p* = 0.0165); lymph node metastasis N≥<sup>2</sup> (*p* = 0.0473); and TNM stage TIII–IV (*p* = 0.0148) (Figure 5B and Supplementary Table S3). We also observed higher plasma free normetanephrine in patients with gastric carcinoma with a tumor size ≥5 cm (*p* = 0.0217), histological type Ad.c as opposed to M.c./S.r.c.c. (*p* = 0.0253), a location of the tumor in the gastric body or pyloric area (*p* = 0.0132), a tumor invasion T3–4 (*p* = 0.0177), a lymph node metastasis N≥<sup>2</sup> (*p* = 0.0127), and a TNM stage TIII–IV (*p* = 0.0275) (Figure 5C and Supplementary Table S4). It should be noted that patients who had the signet-ring-cell gastric cancer type, despite

having a low degree of differentiation (G3), had low plasma levels of free metanephrine and normetanephrine. despite having a low degree of differentiation (G3), had low plasma levels of free metanephrine and normetanephrine.

**Figure 5.** (**A**) Relationship between norepinephrine transporter expression and clinicopathological features. Relationship between plasma free metanephrine (**B**) and normetanephrine (**C**) and clinicopathological features. Ad.c., adenocarcinoma; M.c./S.r.c.c., mixed carcinoma/signet-ring-cell carcinoma; P&B, gastric body or pyloric area. **Figure 5.** (**A**) Relationship between norepinephrine transporter expression and clinicopathological features. Relationship between plasma free metanephrine (**B**) and normetanephrine (**C**) and clinicopathological features. Ad.c., adenocarcinoma; M.c./S.r.c.c., mixed carcinoma/signet-ring-cell carcinoma; P&B, gastric body or pyloric area.

#### *2.4. Correlation between Norepinephrine Transporter Expression and Plasma Free Metanephrine and Normetanephrine*

We observed a moderate positive correlation between the IOD for norepinephrine transporter expression and plasma free metanephrine (r = 0.3929; 95% confidence interval = 0.2034–0.5540; R-squared = 0.1544), and a strong positive correlation between the IOD for norepinephrine transporter expression and plasma free normetanephrine (r = 0.5151; 95% confidence interval = 0.3459–0.6519; R-squared = 0.2654). There was also a strong positive correlation between the plasma free metanephrine and plasma free normetanephrine (r = 0.5901; 95% confidence interval = 0.4373–0.7098; R-squared = 0.3842). These data are summarized in Figure 4D–F.

#### *2.5. Univariate Analysis of Prognostic Factors*

The 91 patients with gastric carcinoma were divided into a low-NET group (N = 36/91) and a high-NET group (N = 55/91) based on the median IOD for the norepinephrine transporter. Depending on cut-off levels for plasma free metanephrine (65 pg/mL) and normetanephrine (196 pg/mL), the patients with gastric carcinoma were divided into low- (N = 39/91) and high-Meta groups (N = 52/91) and into low-(N = 33/91) and high-N-Meta groups (N = 58/91). Univariate analysis with a log-rank test indicated that the high-NET patients had a significantly poorer survival rate at two years after inclusion in the study than the low-NET patients (44.23% vs. 63.88%; *p* = 0.0358; hazard ratio and its reciprocal = 1.956 and 0.5140, respectively; Figure 4G). Lower survival rates were also observed in the patients with higher metanephrine but without statistical significance (48.07% vs. 58.33%; *p* = 0.1487; hazard ratio and its reciprocal = 1.462 and 0.6838, respectively; Figure 4H), as well as in those with high free normetanephrine with statistical significance (42.30% vs. 66.67%; *p* = 0.0104; hazard ratio and its reciprocal = 2.289 and 0.4369, respectively; Figure 4I).

#### **3. Discussion**

In this study, we analyzed the correlations between the clinicopathological aspects of patients diagnosed with gastric cancer (gastric adenocarcinoma) and the involvement of the autonomic nervous system in the carcinogenesis process by identifying certain features of the sympathetic nervous system and the norepinephrine transporter, identified in neoplastic cells. We attempted to support the claim that the vegetative nervous system can influence the development and evolution of gastric cancer.

The sympathetic nervous system is part of the vegetative nervous system and responsible for the fight reaction, also known as the sympathetic–adrenal response. It secretes adrenaline (epinephrine) and noradrenaline (norepinephrine), catecholamines that are subsequently released into the blood [15]. The action exerted by the sympathetic nervous system causes a series of reactions in various organs of the body, such as an increased heart rate, decreased motility in the large intestine, reduced secretions by salivary glands, and vasoconstriction [16]. Free plasma metanephrine and normetanephrine are metabolites of catecholamines, the latter being considered a hormone that is released into the blood, especially during periods of physical or emotional stress, depression, or anxiety, causing both psychological and endocrine changes [17]. These catecholamines produced from the precursor tyrosine can, on the one hand, alter the immune response and, on the other hand, promote several biological signaling pathways involved in tumor initiation, growth, and metastasis [16–19].

Regarding the involvement of the sympathetic nervous system in the development and evolution of gastric cancer, the primary pathway is mediated by the action of neurotransmitters on β2-adrenergic receptors, activating an intracellular signaling cascade via adenylyl cyclase [18,19]. Studies have highlighted the impact of the main neurotransmitter of the sympathetic nervous system, norepinephrine, on vascular endothelial growth factor (VEGF) and matrix metalloproteinase 2/9 (MMP-2 and MMP-9) [20]. Other studies have also shown an important role of norepinephrine in epithelial–mesenchymal transition

(EMT). For example, Shan et al. demonstrated that norepinephrine causes, in gastric carcinoma, a decrease in E-cadherin expression and an increase in vimentin expression; both changes increase cell motility and confer the ability of tumor invasion [21]. This mechanism can occur through the β2-adrenergic receptor (AR)–hypoxia-inducible factor-1-alpha axis, which is also involved in the promotion of tumor progression by chronic stress in animal cancer models [22]. EMT can also be initiated in gastric cancer by the β2-AR–metalloproteinase (MMP)-7 pathway through the activation of AP-1 and signal transducer and transcriptional activator 3 (STAT3) [23,24].

We found that the highest plasma values of serum metanephrine and normetanephrine were increased in patients with poorly differentiated gastric adenocarcinoma. They varied according to the degree of differentiation, and increased values were found among those with localization in the gastric body or pyloric area and with histopathological aspects of adenocarcinoma, as well as among patients who had metastases in regional lymph nodes or distant metastasis. In this regard, a recent study that evaluated the activity of periostin, which mediates the critical steps in gastric carcinoma, showed that it is expressed in the stroma of gastric carcinoma but not in normal gastric tissue, and this is strongly correlated with the expression of alpha-smooth muscle actin (SMA) [25]. Isoprenaline causes an increase in periostin expression in gastric cancer, with the activation of the previously mentioned axis, but it can also promote angiogenesis by stimulating VEGF secretion and the upregulation of VEGFR2 and plexin-A1 [26,27].

Another aspect highlighted by our study is the implications of psychological stress for the initiation and progression of gastric carcinoma. Psychological stress initiates a response of the hypothalamic–pituitary–adrenal axis, which raises catecholamine levels; catecholamines interact with certain biological components of tumor cells through certain signaling pathways. This can lead to the progression of certain cancers, such as those of the ovaries, nasopharynx, or pancreas [27,28].

Regarding the influence of catecholamines on therapy for severe gastric cancer, it has been observed that the stimulation of gastric cancer cells with catecholamines in vitro increases trastuzumab resistance by not only activating STAT3 and extracellular signalregulated kinases (ERKs) but also by upregulating mucin 4 (MUC4) expression [29]. However, these cellular signaling mechanisms induced by catecholamines may become possible therapeutic targets. For example, propranolol, a non-selective adrenergic blocker, can cause cell cycle arrest and induce apoptosis in gastric carcinoma cells by blocking nuclear factor-kB (NF-kB), MMP2/9, VEGF, and cyclooxygenase-2 (COX-2) [30,31].

It is well known that the sympathetic nervous system influences cardiac activity, causing an increase in heart rate, as was shown in our patients. Most of the patients with gastric cancer in our study had increased heart rates directly proportional to the plasma levels of free metanephrine and normetanephrine. Shi et al. recently reported that the severity of gastric cancer in diagnosed patients can also be predicted by perturbations in the nonlinear dynamic patterns of heart rate variability (HRV) [32].

In other primary tumors, the activity of the sympathetic nervous system has been evaluated in both preclinical and clinical studies. The plasma norepinephrine and epinephrine concentrations are significantly higher in patients with oral and oropharyngeal squamous cell carcinoma (SCC) than in non-cancer patients [33]. In epithelial ovarian cancer, norepinephrine reduces cisplatin's efficacy and can affect DNA integrity [34]. Epinephrine increases the phosphorylation of p38 MAPK in breast cancer cells and, thereby, enhances the malignancy of this type of cancer [35]. A chemical sympathectomy markedly reduces the incidence of fibrosarcoma and significantly prolongs survival in rats [36]. Regarding other tumors of the gastrointestinal tract, it has been found that norepinephrine facilitates tumor growth in pancreatic cancer [37], induces hepatocellular carcinoma invasion and anoikis resistance through β2-AR-mediated epidermal growth factor receptor transactivation [38], and facilitates cell proliferation in esophageal squamous cell carcinoma [39]. Catecholamines have also been shown to promote metastasis and tumor progression in prostate and lung cancers and melanoma [40].

#### **4. Materials and Methods 4. Materials and Methods** *4.1. Patients*

#### *4.1. Patients*

This was an observational study, in which 91 patients diagnosed with gastric carcinoma with different degrees of tumor differentiation were consecutively included, following surgery or upper digestive endoscopy, at the Emergency County Hospital of Craiova, Romania, between November 2017 and October 2018. For the controls, we delimited a group of 200 patients without cancer of the same age group and gender as the cancer patients. We set the primary endpoint as mortality from any cause in the first two years after enrolling in the study. The design of the study is shown in Figure 6. This was an observational study, in which 91 patients diagnosed with gastric carcinoma with different degrees of tumor differentiation were consecutively included, following surgery or upper digestive endoscopy, at the Emergency County Hospital of Craiova, Romania, between November 2017 and October 2018. For the controls, we delimited a group of 200 patients without cancer of the same age group and gender as the cancer patients. We set the primary endpoint as mortality from any cause in the first two years after enrolling in the study. The design of the study is shown in Figure 6.

transactivation [38], and facilitates cell proliferation in esophageal squamous cell carcinoma [39]. Catecholamines have also been shown to promote metastasis and tumor pro-

*Int. J. Mol. Sci.* **2021**, *22*, 2042 9 of 14

gression in prostate and lung cancers and melanoma [40].

**Figure 6.** Design of the study. **Figure 6.** Design of the study.

All of the stages of the study were explained to the patients before commencement, and participation was possible only after providing written consent, with the patients being informed about the confidentiality of personal data and the procedures being performed in accordance with current regulations, without negatively influencing normal All of the stages of the study were explained to the patients before commencement, and participation was possible only after providing written consent, with the patients being informed about the confidentiality of personal data and the procedures being performed in accordance with current regulations, without negatively influencing normal diagnostic or treatment procedures.

diagnostic or treatment procedures. This study was approved by the Ethics Committee of the University of Medicine and Pharmacy of Craiova (No. 71/02.04.2017), respecting the ethical principles underlying the Declaration of Helsinki and the University Code of Ethics on Good Conduct of Research This study was approved by the Ethics Committee of the University of Medicine and Pharmacy of Craiova (No. 71/02.04.2017), respecting the ethical principles underlying the Declaration of Helsinki and the University Code of Ethics on Good Conduct of Research and the codes of practice established by the Code of Medical Ethics.

and the codes of practice established by the Code of Medical Ethics. Patients diagnosed with gastric cancer, confirmed by postoperative histopathological examination or biopsy, had blood samples taken for the measurement of serum metanephrine and normetanephrine, after which they underwent Holter ECGs, before starting Patients diagnosed with gastric cancer, confirmed by postoperative histopathological examination or biopsy, had blood samples taken for the measurement of serum metanephrine and normetanephrine, after which they underwent Holter ECGs, before starting chemotherapeutic treatment, to avoid possible post-drug interactions.

chemotherapeutic treatment, to avoid possible post-drug interactions.

#### *4.2. Assessment of Heart Rate*

To assess the predominance of the sympathetic or parasympathetic nervous system, the patients were also monitored using a Holter ECG TLC5000 (Contec Medical Systems, Qinhuangdao, Hebei Province, China), together with a former analysis of heart rate variability in terms of both frequency and the field of time. Ten patch electrodes were applied to the patients' chests, through which 12 ECG leads were connected. The important parameters in our study were the minimum, average, and maximum heart rate, not only over 24 h but also during the night and day. The ECG Holter parameters were interpreted according to the recommendations of the American College of Cardiology/American Heart Association (ACC/AHA) ambulatory electrocardiography guide [41]. It should be noted that the patients were examined by this method after the diagnosis of gastric adenocarcinoma was confirmed. Another important criterion was a lack of previous medication that could have influenced the heart rhythm (especially betablockers, calcium channel blockers, current blockers, or other antiarrhythmics). All the patients in our study performed at least 150 min of moderate-intensity aerobic physical activity per week. There were no differences between the groups included in the study regarding the degree of physical activity. Since all the patients underwent gastrectomy and since this intervention involved damage to local nerve plexuses, we unfortunately did not consider it appropriate to analyze the heart rate variability postoperatively. Moreover, postoperative stress is another factor that inevitably changes heart rate variability.

#### *4.3. Histopathological Examination*

A histopathological examination was performed following surgery or tumor biopsy via upper digestive endoscopy. The biological samples were introduced into 10% formic aldehyde solution, in which a neutral pH was created by adding calcium bicarbonate to neutralize the formic acid for fixation. After fixation, the biological samples were washed with water and then paraffin embedded as follows. The samples were completely dehydrated by passing them through ethyl alcohol of different concentrations. They were then clarified by removing the alcohol from the tissue, and the samples were passed through successive paraffin baths, incorporated into paraffin blocks, and solidified. Finally, the paraffin blocks were sectioned, then the sectioned tissues were glued to slides and stained with hematoxylin–eosin, after which histopathological diagnoses were established with certainty.

#### *4.4. Immunohistochemistry*

The paraffin-embedded tissues were sectioned into 3 µm serial sections using the HM350 rotary microtome (Thermo Fisher Scientific, Waltham, MA, USA) equipped with a section transfer system in a cold-water bath with a Peltier cooling mode. These slides were then transferred to a water bath heated to 40 ◦C to be stretched and evened out. The slides were recovered from the bath on blades covered with poly-lysine with a positively charged amino acid residue to increase the adhesion of the sections on the blades. The poly-lysine-coated slides were dried in a thermostat at 37 ◦C for 24 h.

The next day, the slides were first deparaffined by passing them through three successive xylene baths, for 15 min each, and then rehydrated in alcohol solutions with decreasing concentrations; any final traces of alcohol were removed by washing with distilled water. For antigenic recovery, the slides were boiled for 21 min in successive cycles of 3 min each, in a microwave oven, at a power of 600 W, in a solution of sodium citrate at a pH of 6. This was followed by cooling the slides for 30 min and then washing them with tap water and distilled water for 15 min. To block endogenous peroxidase activity, the slides were incubated in 1% hydrogen peroxide and distilled water for 30 min at room temperature, stored for another 30 min in 3% skimmed-milk powder in phosphate-buffered saline (PBS), and then incubated with primary antibodies at 4 ◦C for 18 h.

The primary antibodies used in this work were norepinephrine transporter monoclonal antibody (CL3063)/NBP2-62704 (dilution 1:20; Novus Biological, Abingdon, UK).

Finally, the signal was identified via 3,30 -diaminobenzidines (DAB) (Dako, Glostrup, Denmark). Subsequently, the slides were cover-slipped in DPX (Sigma-Aldrich, St. Louis, MO, USA) after hematoxylin and eosin staining. All the slides stained for each primary antibody were processed at the same time to observe the protocol correlation, along with the control sections, which were stained with either DAB or hematoxylin–eosin in order to obtain the pure spectrum for those colors. Negative controls were obtained by omitting primary antibodies.

#### *4.5. Acquisition and Image Processing*

For the quantification of the target immunohistochemical signal, and taking into account the histopathological aspect, light microscopy images were obtained using a Nikon Eclipse 90i motorized microscope (Apidrag, Bucharest, Romania). This microscope was equipped with a Nuance FX multispectral camera as well as the Nuance imaging analysis software (Perkin Elmer, Hopkinton, MA, USA). An optical microscopy image was initially obtained, followed by a mixed image (the color spectra for hematoxylin and DAB were separately superimposed on this image). In another step, separate images were obtained for hematoxylin and DAB (Supplementary Figure S1). The unmixed DAB signal was quantified by randomly obtaining 10 images captured with a 20× lens. The color signal was quantitatively analyzed based on the integrated optical density, using the Image-Pro Plus AMS 7 image analysis software (Media Cybernetics, Bethesda, MD, USA). With the help of this software, regions of interest were defined where we evaluated the color signals and calculated the IOD, while the stroma was manually excluded from the obtained images.

#### *4.6. Dosage of Plasma Free Metanephrine and Normetanephrine*

Blood samples were taken from fasted patients who had avoided alcohol and caffeine for 24 h before sampling. They were also informed to avoid certain drugs that may influence serum metanephrine and normetanephrine, such as acetaminophen, tricyclic antidepressants, phenoxybenzamine, alpha-agonists, or monoaminoxidase inhibitors. For patients undergoing treatment with these drugs, their medication was discontinued for at least five days before sampling.

Venous blood was collected in a pre-cooled ethylenediaminetetraacetic acid (EDTA) K3 vacutainer, which was then gently shaken, overturned, and placed on ice; it was transported to the laboratory within 2 h. The samples were processed no later than 2 h after sampling.

The serum metanephrine and normetanephrine were quantified via the competitive enzyme-linked immunosorbent assay (ELISA), and the values were determined after a precipitation stage. Their reference values were the following: metanephrine, <65 pg/mL, and normetanephrine, <196 pg/mL, with detection limits of 5 and 10 pg/mL, respectively [42].

#### *4.7. Statistical Analysis*

The data obtained with the Image-Pro Plus AMS 7 image analysis software were exported to Microsoft Excel 2010 (Microsoft Corporation, Redmond, WA, USA) and analyzed using GraphPad Prism 8 (San Diego, CA, USA). All the results are reported as the means and standard deviations. To compare the means of two groups, we used Student's *t*-test. To compare the means of more than two groups, we used an ANOVA. To examine the correlations between the different categories of data, we used Pearson's correlation test. To analyze whether there was a link between a variable and survival time, we used the log-rank test. *p* < 0.05 was considered to indicate a statistically significant difference between the compared means from the various groups.

#### **5. Conclusions**

The predominance of the sympathetic nervous system's activity in patients with gastric cancer, through increased heart rates, elevated plasma free metanephrine and normetanephrine, and increased expression of the norepinephrine transporter in tumor cells, is a negative prognostic factor for these patients. These observations may highlight future therapeutic or prognostic targets.

**Supplementary Materials:** The supplementary materials can be found at https://www.mdpi.com/ 1422-0067/22/4/2042/s1: Figure S1, example of spectral unmixing for the series of slides; Table S1, norepinephrine transporter expression quantified according to the integrated optical density in cancer-free patients and in different tumor differentiation gradings in patients with gastric carcinoma; Table S2, norepinephrine transporter expression quantified depending on clinicopathological features; Table S3, plasma free metanephrine level (pg/mL) depending on the clinicopathological features; Table S4, plasma free normetanephrine level (pg/mL) depending on the clinicopathological features.

**Author Contributions:** Conceptualization, A.M.M., V.S., G.C.T., and C.C.V.; methodology, V.S. and O.I.; software, G.C.T.; validation, V.S., O.I., M.S., and A.-M.C.; formal analysis, A.-M.C. and P.O.S.; investigation, A.M.M., P.O.S., M.S., and A.-M.C.; writing—original draft preparation, G.C.T. and O.I.; writing—review and editing, A.M.M. and V.S.; visualization, P.O.S., M.S., and A.-M.C.; supervision, C.C.V. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of UNIVERSITY OF MEDICINE AND PHARMACY OF CRAIOVA, ROMANIA (No. 71/02.04.2017).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Acknowledgments:** The authors would like to especially thank Pirici Daniel from the Department of Histology, University of Medicine and Pharmacy of Craiova, Romania, for the processing of the histological samples and for the help given in the multispectral evaluation of the microscopic images.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Abbreviations**


#### **References**


## *Review* **Updated Insights on EGFR Signaling Pathways in Glioma**

**Alexandru Oprita 1,†, Stefania-Carina Baloi 1,†, Georgiana-Adeline Staicu <sup>1</sup> , Oana Alexandru <sup>2</sup> , Daniela Elise Tache <sup>1</sup> , Suzana Danoiu <sup>3</sup> , Elena Simona Micu 4,\* and Ani-Simona Sevastre 5,\***


**Abstract:** Nowadays, due to recent advances in molecular biology, the pathogenesis of glioblastoma is better understood. For the newly diagnosed, the current standard of care is represented by resection followed by radiotherapy and temozolomide administration, but because median overall survival remains poor, new diagnosis and treatment strategies are needed. Due to the quick progression, even with aggressive multimodal treatment, glioblastoma remains almost incurable. It is known that epidermal growth factor receptor (EGFR) amplification is a characteristic of the classical subtype of glioma. However, targeted therapies against this type of receptor have not yet shown a clear clinical benefit. Many factors contribute to resistance, such as ineffective blood–brain barrier penetration, heterogeneity, mutations, as well as compensatory signaling pathways. A better understanding of the EGFR signaling network, and its interrelations with other pathways, are essential to clarify the mechanisms of resistance and create better therapeutic agents.

**Keywords:** glioma; pathways; EGFR; clinical trials

## **1. Introduction**

With an overall survival of less than 35% in five years [1], malignant primary brain tumors are the most difficult to treat cancers. Of those, the most common type is represented by gliomas. Based on the expression patterns' differences, glioblastomas are divided into three subtypes as follows: classical, proneural, and mesenchymal [2]. Because glioblastoma multiforme (GBM), a grade IV glioma [3], is one of the most aggressive primary brain tumors, recent studies and reviews have focused on deepening our understanding of the disease [4–9].

At present, GBM's pathogenesis is better understood due to recent advances in molecular biology. For newly diagnosed glioblastoma, the current standard of care is represented by resection, followed by radiotherapy and temozolomide (TMZ) administration [10], but the median overall survival (OS) is not fully improved; therefore, new diagnosis and treatment strategies are needed [11,12].

Glioblastoma is the most common and the most deleterious glioma [13]. The 2011– 2015 Statistical Report of the Central Brain Tumor Registry of the United States (CBTRUS) showed that glioblastoma represents 48% of the malignant brain and central nervous system tumors, with an incidence rate in the United States 1.58 times higher in males

**Citation:** Oprita, A.; Baloi, S.-C.; Staicu, G.-A.; Alexandru, O.; Tache, D.E.; Danoiu, S.; Micu, E.S.; Sevastre, A.-S. Updated Insights on EGFR Signaling Pathways in Glioma. *Int. J. Mol. Sci.* **2021**, *22*, 587. https:// doi.org/10.3390/ijms22020587

Received: 15 November 2020 Accepted: 5 January 2021 Published: 8 January 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

compared to females [14]. Due to the quick progression, even with aggressive multimodal treatment, glioblastoma remains almost incurable [15,16].

Nowadays, chemotherapy has a significant role in glioblastoma's treatment strategies, with numerous research studies aimed to develop more efficient chemotherapeutic drugs [17]. Understanding the disease's pathogenesis has a key role in identifying disease biomarkers and developing new potential chemotherapeutic drugs. We present some of the most promising signaling pathways involved in pathogenesis, with their specific targeting components.

GBM is characterized by nuclear atypia, cellular pleomorphism, mitotic activity, anaplasia, and rapid proliferation alternated with an aggressive invasion of the surrounding brain tissue. In its microenvironment, glioma cells are faced with many challenges such as acidity, hypoxia, and low nutrient availability. To maintain rapid growth, they need to modulate metabolic activity [18,19].

In multicellular organisms, tyrosine phosphorylation is involved in signal transduction, leading to differentiation, proliferation, migration, and survival [20,21].

Receptor tyrosine kinases (RTKs) are activated by binding their extracellular domain to corresponding ligands determining their oligomerization. This process activates the intracellular domain, facilitating the recruitment of proteins that start a signaling cascade, integrating numerous signaling pathways that lead to specific cellular responses [22]. Among all RTKs, epidermal growth factor receptor (EGFR) is the most amplified in GBM [23]. EGFR amplification is observed in the classical subtype of glioma [2]. EGFR gene amplification is detected in 57.4% of primary GBM patients, leading to high levels of EGFR protein, contributing to tumorigenesis and progression [24].

However, targeted therapies against this type of receptor have not yet shown a clear clinical benefit. Many factors contribute to resistance, such as ineffective blood–brain barrier penetration, heterogeneity, mutations, and compensatory signaling pathways. A better understanding of the EGFR signaling network and its interrelations with other pathways are essential to improve drug activity, clarify the mechanisms of resistance, and develop better therapeutic agents.

#### **2. Understanding EGFR Features**

The transmembrane receptor of tyrosine kinase epidermal growth factor (EGFR), also known as HER (human EGFR related) 1 or ErbB1, along with HER2/neu (ErbB-2), HER3 (ErbB-3), and HER4 (ErbB-4), is a member of the ErbB family and it is located on chromosome band 7p12 [25].

Like all RTKs, EGFR has an extracellular region, a single transmembrane domain, an intracellular juxtamembrane domain, a tyrosine kinase, and a C-terminal region. The ligands of ErbB receptors are divided into two main groups: EGFR activators called EGF agonists, and neuregulins that bind to ErbB3 and ErbB4 [26,27].

The extracellular region of EGFR has two homologous domains (I and III) that bind ligands and two cysteine rich domains (II and IV) [28].

The juxtamembrane region tethers inactive EGFRs to the plasma membrane cytosolic surface, which contributes to EGFR activation [29]. Structural studies highlight the functional importance for certain regions, such as the structure of the first 30 amino acids from the intracellular juxtamembrane region of EGFR and the C-terminal 190 amino acids [27].

There are more than 40 EGFR ligands that control its signaling. They can be divided into high-affinity ligands, such as epithelial growth factor (EGF), heparin-binding EGF-like growth factor (HB-EGF), Transforming growth factor alpha (TGF-α), betacellulin (BTC), and low-affinity ligands, such as epiregulin (EREG), amphiregulin (AREG), and epigen (EPGN).

Expression of EGF-family proteins and activation of EGFR are features of cardiac disease [30,31]. Moreover, molecular alterations of EGFR include overexpression, deletion, or amplification, in different types of cancer. In GBM, EGFR amplification promotes invasion, proliferation, and drug resistance to radio- and chemotherapy [32].

Several trials on EGFR targeted therapy have failed to produce conclusive evidence, maybe because of the EGFR molecular heterogeneity in GBM, of the low specificity of the designed drugs, as well as because of low brain penetration [33]. Despite all this, the detection of EGFR alterations is still used as a prognostic marker for GBM because 24–67% of GBMs are characterized by a mutated gene, 40% by amplification, and 60% by EGFR overexpression [34].

In recent years, studies have proved that EGFR has pro-survival kinase-independent functions in malignant cells. This fact has offered a different perspective of understanding EGFR implications in cancer, with new ideas of EGFR targeted cancer therapy [35–37].

#### **3. Mechanisms of EGFR Pathway Activation**

There are several different mechanisms of EGFR pathway activation, such as increased ligand production or overexpression/defective inactivation/mutation of the receptor. Many studies focused on the EGFR signaling mechanism in recent years, trying to conclude how the extracellular EGFR-ligand binding propagates through the single transmembrane helix (TM) to trigger intracellular kinase activation [38–40].

#### *3.1. EGFR Activation Mechanisms in Normal Physiologic Status*

The expression of EGFR in normal cells is about 4 <sup>×</sup> <sup>10</sup>4–10 <sup>×</sup> <sup>10</sup><sup>4</sup> receptors/cell [41], whereas, in cancer cells, more than 10<sup>6</sup> receptors/cell are observed [42]

The EGFR RNA expression is increased by stimulating the EGFR-specific transcription factor (ETF). The receptor expression is regulated by epidermal growth factor (EGF) itself and other proteins such as E1A, Sp1, and AP2 [36].

Like all RTKs, EGFR is activated by ligands featuring receptor-specificity. Briefly, ligand binding leads to a dimeric active conformation of EGFR by homodimerization (complexed with another EGFR) or heterodimerization (complex with another ErbB member). The tyrosine residues from other RTKs are autophosphorylated after ligand stimulation, and phenylalanine substitutions significantly impair the kinase signaling and the downstream signaling. Differently, EGFR Tyr-845 phosphorylation is not a required mechanism for ligand-induced EGFR activation, but it may represent the main mechanism for EGFR transactivation [43,44].

Proteins that express a proto-oncogene tyrosine-protein kinase (Src) homology domain 2 (SH2) region bind to the activated receptor, areactivated, and forward the signal to the downstream effectors, propagating critical cellular signaling pathways [45]. EGFR can simultaneously activate several signal transduction pathways such as phosphatidylinositide 3 kinase (PI3K) and serine–threonine kinase (AKT) and RAS/MAPK pathways [46].

#### 3.1.1. Extracellular Domain Activation

For EGFR, the dimerization is completely receptor-mediated, with no physical interaction between two activating ligands. In normal physiologic status, the receptors are in a dynamic monomer–dimer equilibrium. In the absence of ligands, the extracellular domain presents a tethered configuration (intra-molecular links entirely block the dimerization arm), and the intracellular tyrosine kinase domain (TKD) is inactive. Ligand binding leads to a conformational change that exposes the buried dimerization arm, and the extracellular domain dimerizes, inducing conformational changes of the intracellular domain and enabling kinase activation [45].

A recent study by Chung et al. described physiological EGFR activation as being due to a ligand-mediated extracellular domain dimerization that stabilizes the *N*-terminal transmembrane dimer and disrupts autoinhibition, allowing the C-terminal juxtamembrane (JM-B) segment to stabilize the asymmetric kinase domain (KD) dimer, resulting in activation of EGFR signaling. They also concluded that the stimulus stabilizes the active KD conformation in pathological states and further the asymmetric KD dimerization. The inside–out coupling is weaker than the physiological outside–in coupling, suggesting

that the extracellular (EC) dimer is linked through the *N*-terminal TM dimer with the asymmetric oncogenic KD dimer [38].

#### 3.1.2. Intracellular Domains Activation

By ligand-induced dimerization, the cis-autoinhibition is released, and through a unique allosteric mechanism, the kinase activity of EGFR isactivated. It is well known that this mechanism consists of physical interaction between the C-terminal tail of the activator kinase and the other kinase *N*-terminal tail (receiver kinase) of the dimer pair, inducing conformational changes of the *N*-lobe of receiver kinase and trans-phosphorylation Cterminal tail of the activator [47].

#### 3.1.3. Downstream Signaling of EGFR

EGFR activation and autophosphorylation result in the recruitment of downstream signaling proteins. Almost all autophosphorylation sites are binding sites for Src Homology 2-(SH2) or Phosphotyrosine binding-(PTB) signaling proteins. The SH2- proteins may be bound directly to the receptor, or indirectly through docking proteins using PTB domains [48]. EGFR can recruit and regulate many signaling pathways such as PI-3 K/AKT, RAS/MAPK, and JAK2/STAT. Therefore, EGFR functions as a hub involved in regulating various cellular processes [21,23], as shown in Figure 1.

**Figure 1.** EGFR signaling pathway (EGFR—epithelial growth factor receptor, EGFRvIII—Epidermal growth factor receptor variant III, Pi3K—Phosphoinositide 3-kinase, RAS—family of genes involving cellular signal transduction, PTEN— Phosphatase and tensin homolog, NF1—Neurofibromatosis type 1, RAF—serine/threonine-specific protein kinases, MEK— Mitogen-activated protein kinase, ERK—extracellular signal-regulated kinase, AkT—Protein kinase B, mTOR—mammalian target of rapamycin, Src—Proto-oncogene tyrosine-protein kinase, cMyc—c proto-oncogene, NFKB—nuclear factor kappalight-chain-enhancer of activated B cells, Block arrow—inhibition activity, Point arrow—pathway flow).

The PI-3K/AKT signaling pathway involves PI3K, an enzyme with SH2-signal transducer and its downstream effector AKT, regulating apoptosis and cell survival. Once EGFR is activated and phosphorylated, PI3K is brought to the cell membrane, and it phosphorylates phosphatidylinositol 4,5-bisphosphate (PIP2), forming phosphatidylinositol (3,4,5)-trisphosphate (PIP3). AKT reacts with PIP3, and it is phosphorylated at Threonin308 by phosphoinosite-dependent protein kinase-1 (PDK1) and at Serine 473 by the mammalian target of rapamycin complex 2 (mTORC2), reaching full activity. The phosphatase and tensin homolog (PTEN) negatively regulate the PI3K/AKT pathway by dephosphorylating and delocalizing PIP3 from the cellular membrane, resulting in the relocalization of AKT in the cytoplasm, where it is unable to be reactivated [49,50].

Class IA is one of the three different classes of PI3Ks featuring subunits with regulatory activity such as p85. Active EGFR achieves association with regulatory p85 through dimerization with human HER3, or via the docking protein GRB2-associated binder 1 (GAB1), relieving the inhibitory effect of p85 [51]. GAB1 is a scaffolding protein involved in recruiting additional signaling proteins such as PI3K, SHP2, and p120RasGap. It is involved in many EGFR signaling outputs, and is the predominant mechanism linking EGFR to PI3K/Akt signaling [52,53].

Due to its increasing importance in different human cancers, GAB1 may represent an emerging potential therapeutic target.

The RAS/MAPK signaling pathway involves the growth-factor-receptor bound-2 (GRB2), which forms a complex with Son of Sevenless (SOS), a guanine-nucleotide exchange factor (GEF) and activates the RAS G-protein by exchanging guanosine diphosphate (GDP) with guanosine triphosphate (GTP) [54]. Consequently, RAS and mitogen-activated protein kinases (MAPKs) initiate a downstream signaling cascade to phosphorylate the nuclear protein Jun. Jun creates complexes with different nuclear proteins leading to the key transcription factor activator protein 1 (AP-1), responsible for translation and transcription of proteins involved in the growth and division of cells. Activated RAS is negatively regulated by GTPase activating proteins (GAPs), such as the tumor suppressor neurofibromin 1 (NF1) [55].

Signal transduction and activator of transcription 3 (STAT3) is tyrosine-phosphorylated or activated as pSTAT3 due to EGFR-regulation of interleukin-6 (IL-6) expression. This mechanism leads to a feed-forward in the IL-6/Janus kinase (JAK)/STAT3 loop [21,56–58].

#### *3.2. Oncogenic Status and EGFR Activation*

The EGFR is one of the most frequently altered oncogenes in brain cancers. Except for hematopoietic cells, the majority of cell types express ErbB family members [35].

In glioblastoma cells, the EGFR tyrosine kinase activity may be dysregulated by multiple oncogenic mechanisms, such as gene mutation, overexpression of EGFR protein, increased gene copy number, rearrangements of chromosomes, and activation by autocrine function [59].

#### 3.2.1. Mutations of Cell Signaling Regulators

The EGFR gene is located on chromosome 7p11.2 and consists of 28 exons encoding a transmembrane protein receptor composed of 464 amino acids. Exons 5–7 and 13–16 encode the ligand binding domain, and exons 18–24 encode the tyrosine kinase domain. The region encoded by exons 25–28 is the site of autophosphorylation.

Although EGFR is one of the most important drug targets in cancer therapies, its mutations present an organ–site asymmetry, depending on the cancer's organ of origin [60]. Although mutations occur in the kinase domain (KD) in other tumors, in gliomas, heterogeneous mutations and deletions are focused on the ligand-binding ectodomain (ECD). This tissue-specific feature leads to type-II tyrosine kinase inhibitors (TKIs) with high sensitivity for the inactive symmetric KD dimer (sKD), when administered in GBM mutations [61]. However, both intra- and extracellular GBM mutations result in ligand-independent oncogenic activation.

Almost 50% of the tumors characterized by EGFR amplification are positive for the mutant EGFRvIII and EGFR single nucleotide variants (SNVs). Due to this tumorspecific feature, novel therapeutic agents are currently under development to target the overexpressed EGFR or EGFRvIII proteins. An in-frame deletion of exons 2–7 characterizes the EGFRvIII, which results in overexpression of a truncated receptor that lacks some significant parts of the ECD. This prototypic oncoprotein is unable to bind ligands, and it is constitutively active. Several studies examined the effect of the EGFRvIII constitutive activity on the wtEGFR and ErbB2 protein levels. For example, one study evaluated the effect of Tyrphostin AG1478 on the protein levels and demonstrated that its administration increased protein levels of wtEGFR and erbB2 in vIIIA1 cells, due to the catalytic activity of EGFRvIII, while in its absence, the levels were reduced [62]. Furthermore, the unique peptide sequence of EGFRvIII generated by the fusion of exons 1 and 8 may serve as a tumor-specific target in immunotherapy [63], although subsequent phase III trial results are not as promising as initially anticipated [64].

A meta-analysis performed in 2017 by Felsberg et al. proved that EGFRvIII and EGFR SNVs do not represent prognostic keys in EGFR-amplified glioma patients. However, the amplification of EGFR is retained in recurrent glioma [63], although improved long-term survival by EGFRvIII therapy has been reported in glioblastoma patients [65].

Nevertheless, the research on EGFRvIII continues, producing inconclusive results. For example, Struve et al. just published in early 2020 the results of a study focused on the effect of EGFRvIII in regulating DNA mismatch repair. They tested if EGFRvIII influences temozolomide's sensitivity and demonstrated that, under standard treatment with temozolomide, EGFRvIII expression leads to prolonged survival only in patients with tumors with O6-methylguanine-DNA methyltransferase (MGMT) methylated promoter. Their results showed that EGFRvIII sensitizes a type of GBM to the current standard of care treatment with temozolomide through the upregulation of DNA mismatch repair (MMR) [65]. However, patients with tumors that have both EGFRvIII and MGMT methylation are very uncommon, and the conclusion that EGFRvIII status was associated with increased survival had a *p* = 0.06. This level would not normally be considered significant, especially not in this sort of multivariate analysis [66].

#### 3.2.2. Overexpression and Gene Amplification

The EGFR gene is amplified in approximately 40% of glioblastomas. The primary and secondary GBM differ in genetic profiles and primary GBMs have a higher prevalence of EGFR gene amplification and overexpression than secondary GBMs [67]. In a study performed by Watanabe et al., EGFR gene amplification was associated with protein overexpression in most tumor cells, but 10% of GBM with overexpression of EGFR protein lacked EGFR gene amplification [68]. However, previous studies have stated that EGFR overexpression or activation does not necessarily cause a simple amplification of its downstream signals, but dose-dependent changes in oncogene-induced downstream signaling and biological responses have been reported [69].

#### 3.2.3. Rearrangements of Chromosomes

Breakpoint sequence analyses proved different types of chromosomal rearrangements and mechanisms of DNA repair. Analyses of single nucleotide polymorphisms suggested that different deletions may appear from amplified non-vIII EGFR precursor [70].

In a study performed in 2018 on glioma tumor samples by Tomoyuki et al., complex chromosomal rearrangements involving chromosome 7 were observed [70].

A study performed by Lopez-Gines et al. showed that trisomy/polysomy 7 and monosomy 10 were frequently associated with glioma. The combination of these anomalies is important in glioblastoma's tumorigenesis. Moreover, the association seems to be independent of EGFR gene amplification [71].

#### 3.2.4. Activation by Autocrine Function

It is well known that wild-type EGFR ligands such as transforming growth factoralpha (TGF-alpha) and heparin-binding EGF (HB-EGF) are often increased in glioblastoma leading to an autocrine loop resulting in the autonomy growth of glioma cells [72]. GBM expresses an EGFR mutant (EGFRvIII) that signals constitutively, does not bind ligand, and is considered to have more tumorigenicity than wild-type EGFR. In a U251-MG glioma cell line, the expression of EGFRvIII may result in specific up-regulation of some genes (TGF-α, EPHA2, HB-EGF, IL8, FOSL1, MAP4K4, DUSP6, and EMP1) influencing signaling pathways involved in oncogenesis. TGF-α and HB-EGF (EGFR ligands) induce the expression of EGFRvIII, suggesting that EGFRvIII has a role in creating an autocrine loop with wild-type EGFR. By inhibiting HB-EGF activity with neutralizing antibodies, EGFRvIII-induced cell proliferation may be reduced, suggesting that EGFRvIII-HB-EGFwild-type EGFR autocrine loop has a major role in signal transduction in glioblastoma cells [73]. Furthermore, studies have demonstrated that the expression of the EGFR alone has a poor transformation effect on cells. Though, coexpression of TGF-α ligand leads to a significant increase in transformation and therapies based on neutralizing the ligands have demonstrated the decreased growth of cells that harbor such loops [74,75].

#### **4. Applied Theory—Therapies Targeting EGFR**

The distribution of EGFR in cancer cells is the basic pillar of many targeted strategies pursued to inhibit its signaling pathway [76,77].

EGFR activity may be controlled by binding to the tyrosine kinase domain or binding to the extracellular component. There are three generations of tyrosine kinase inhibitors approved for clinical use. The first mechanism targets signal transduction and is characteristic of the tyrosine kinase inhibitors (RTKIs, TKIs), which bind to the tyrosine kinase domain of EGFR and inhibit its activity. First-generation TKIs, inhibit the receptor by competitive binding with ATP. Subsequent generations of TKIs were created to overcome drug resistance. Second-generation TKIs irreversibly inhibit all four ERBB (originally named because of the homology with the erythroblastoma viral gene product, v-erbB) receptors, whereas the third-generation TKI are specifically designed to target the T790M resistance mutation [78]. As first-generation inhibitors, active drugs include: erlotinib, gefitinib, lapatinib and vandetanib. Afatinib, dacomitinib, and tesevatinib are examples of second-generation small molecule EGFR inhibitors. Osimertinib is the first third generation RTKI. The monoclonal antibodies act as receptor blockers by binding to the extracellular component of the EGFR and block it from binding to its ligands. Cetuximab, necitumumab, and panitumumab are examples of biological therapy targeting EGFR [21]. An overview of clinical trials focused on anti-EGFR strategies used in GBM is provided in Table 1.

**Table 1.** Epidermal growth factor receptor (EGFR)-targeted therapies for adult high-grade gliomas currently in investigational and/or clinical use.



**Table 1.** *Cont.*


**Table 1.** *Cont.*

Other strategies consist of radio-immunotherapy, docking molecule conjugate toxins, chimeric antigen receptor T cells (CAR-T cells), RNA-based therapies, oncolytic viruses, exosomes, and nanoparticles [100]. EGFR-targeted nanoparticles may be combined with focused ultrasound to achieve local drug delivery [101]. Studies have shown that magnetic nanoparticles' superparamagnetic properties allow them to be guided by an external magnet. However, their therapeutic use is limited in treating in vivo brain pathologies due to insufficient local ability to cross the blood–brain barrier. So, focused ultrasound combined with magnetic targeting synergistically delivers drug-loaded magnetic nanoparticles at the target tissue [102]. Boronated EGFR binding compounds are under investigation in so-called boron neutron capture therapy (BNCT). To improve the unsatisfactory bioavailability of large molecules or viruses due to low blood–brain barrier permeability, the convection technique (CED) is also being investigated [103]. Studies have shown that magnetic nanoparticles' superparamagnetic properties allow them to be guided by an external magnet. However, their therapeutic use is limited in treating in vivo brain pathologies due to insufficient local ability to cross the blood–brain barrier. So, focused ultrasound combined with magnetic targeting synergistically delivers drug-loaded magnetic nanoparticles at the target tissue [101,102].

For example, a study using Cetuximab conjugated magnetic iron oxide nanoparticles showed a significantly enhanced anti-tumor activity compared tocetuximab alone. This was due to improved cellular targeting and uptake, EGFR internalization, EGFR signaling alterations, and apoptosis induction in glioma stem-like cells and tumor non-stem cells that expressed EGFR [104]. In Figure 2, the EGFR-based therapies used in glioblastoma are mentioned.

#### *4.1. Small Molecule Receptor Tyrosine Kinase Inhibitors*

Although several compounds are approved for various diseases, none are approved for glioblastoma due to numerous negative clinical trials. Trials have not shown efficacy either alone or in combination for the oldest small molecule kinase inhibitors: gefitinib, erlotinib, lapatinib, and afatinib [105].

Gefitinib (Iressa, ZD1839) is the first approved EGFR-targeted small-molecule. Initial results from the clinical studies proved that gefitinib was safe when administered for lung carcinoma. However, responses were observed only in a subset of patients featuring chemotherapy–refractory advanced NSCLC (Nonsmall-cell lung carcinoma). The specific mutations of the EGFR gene explained this. It was suggested that these mutations stabilize the interaction of ATP and gefitinib with EGFR. Nevertheless, in the Phase II trial for recurrent GBM, gefitinib did not show improved overall survival [106], neither in the Phase I/II trial when combined with radiation in newly diagnosed GBM [107].

**Figure 2.** EGFR-based therapies in glioblastoma. (BBB—blood brain barrier, mABs—monoclonal antibodies, CAR-T—Chimeric anti-gen receptor T cell therapy)

Erlotinib (Tarceva, OSI-774) proved to prolong NSCLC patients' survival rate upon chemotherapy [108]. These results cannot be achieved in GBM, because EGFR mutations occur in the extracellular domain in GBM, whereas in lung cancers, they are typically observed in the kinase domain. Therefore, unlike NSCLC, GBMs are not sensitive to first-generation EGFR inhibitors [103]. As a single agent, it showed no efficacy in newly diagnosed GBM [109] and later studies that co-administered temsirolimus or bevacizumab were also unsuccessful [110,111].

Lapatinib (Tykerb, GSK 572016) had minimal efficacy alone or in combination with pazopanib in recurrent glioblastoma [112].

Afatinib (Tovok, BIBW2992) had limited efficacy as a single agent in one clinical trial in recurrent glioblastoma [79].

One of the drawbacks of the small molecule inhibitors is their brain penetrance. A study performed by Liu et al. showed that erlotinib could be distributed inside an intracranial U87 xenograft [113]. In another clinical trial, Gefitinib tissue concentration was two- to three-fold plasma concentrations, which was not the cause of insufficient efficacy [114].

Tesevatinib is another second-generation RTKI that is currently under evaluation in patients with recurrent GBM [92]. The first results should be published this year. This trial investigates the drug activity in EGFRvIII positive and negative GBM, with or without EGFR amplification.

Dacomitinib (Vizimpro, PF299804) is a second-generation EGFR inhibitor. Despite its poor global results in a phase II trial in recurrent GBM, Dacomitinib significantly benefited some patients [115].

A recent study investigating Osimertinib (AZD9291), a third-generation EGFR inhibitor, showed that it inhibits with high potency (<100 nM) the constitutive activity of EGFRvIII tyrosine kinase while also inhibiting its downstream signaling. Furthermore, Chagoya et al. proved that osimertinib inhibited the in vitro growth of the D317 cell line and heterotopic and orthotopic xenograft models [89].

To date, there have been eight completed clinical studies involving glioma and Vandetanib (Caprelsa, ZD6474), a second-generation EGFR inhibitor. They all investigated vandetanib's effect together with other therapies (radiotherapy or therapeutic agents), but the results were not satisfactory [116].

#### *4.2. Monoclonal Antibodies*

Cetuximab (Erbitux, DTXSID70142901) is a monoclonal antibody (mAb) targeting the L2 domain of EGFR, preventing dimerization and subsequent cross-activation, thus interrupting downstream signal transduction. It has been approved for the treatment of colorectal, head, and neck cancers [117]. In the progressive high-grade glioma (HGG) patient population, the drug was well tolerated but had limited activity and failed to demonstrate benefit [118]. In new research, the photo-immunoconjugate nanoparticle (PIC-NP) significantly enhanced the photosensitizers in cancer cells and increased the light-activated cytotoxicity in U87 cells overexpressing EGFR [119,120].

Nimotuzumab (OSAG101) is another antibody targeting the L2 domain of EGFR. It was tested in clinical trials for its efficiency in adults with glioblastoma, but the results were not satisfactory. Currently, there is an ongoing clinical trial investigating the effect of nimotuzumab co-administered with temozolomide and radiotherapy. Preliminary results show that nimotuzumab was well tolerated, with an increased survival rate in newly diagnosed GBM patients [86].

Another antibody targeting the L2 domain of EGFR is Panitumumab (Vectibix, ABX-EGF). In combination with irinotecan, it was not very effective in solid tumors [121]. Panitumumab -IRDye800 is currently under investigation in Phase II trials as a GBM diagnostic agent [122].

bscEGFRvIIIxCD3 is a bispecific T-cell engager antibody (BiTEs) that binds to the CD3 T-cell coreceptor and recruits cytotoxic T cells. It was designed to redirect the T-cells towards tumors expressing EGFRvIII. Used in vitro and in vivo on mice, bscEGFRvIIIxCD3 showed the potent killing of GBM expressing EGFRvIII [123].

mAB806 is an antibody targeting the EGFRvIII-specific sequence. The antibody mAb806 is under investigation for glioblastoma treatment, although its mechanism of action remains unknown [124]. It was shown to potentiate the sensitivity of glioma xenotransplants to radiotherapy [125].

#### *4.3. Targeted Isotopes*

The isotope 125I mAB425 is a radioactive isotope conjugated with a specific antibody. Emrich et al. demonstrated that the administration of 125IMAb425 and intensive medical management led to a significant increase in median survival in patients with high-grade gliomas [126]. However, the results of subsequent studies failed expectations.

#### *4.4. Immunotherapy*

#### 4.4.1. CAR-T Cells Targeting EGFRvIII

A new technology developed in cancer therapy is the engineering of T cells to recognize their target by expressing a chimeric antigen receptor (CAR). Glioblastomas express the EGFRvIII, with its unique site of antigenicity. Therefore, these chimeric antigen receptor (CAR)-T cells were engineered to recognize the vIII-receptor mutation through a humanized single-chain antibody fragment (scFv) fused with some key constituents of T-cell receptor intra-cytoplasmic signaling domains [127].

This strategy is currently in early clinical trials. Sahin et al. developed a thirdgeneration chimeric antigen receptor (CAR), specific for EGFRvIII (G3-EGFRvIII), that expresses CD28 (Cluster of Differentiation 28) and CD134 (Cluster of Differentiation 134). Their findings suggest that G3-EGFRvIII CAR represents a potential anti-glioblastoma strategy [128]. A Phase 1 pilot study that investigated the safety and feasibility of CAR-T-EGFRvIII in treating patients with EGFRvIII+ glioblastoma just terminated in 2019, and results are expected to be published [129]—currently, 15 trials are still recruiting.

#### 4.4.2. EGFR as an Immunologic Target—Vaccination

EGFRvIII represents the most common mutation of EGFR. It creates a tumor-specific antigen detectable in almost 30% of human GBM. Deleting the EGFR exons 2–7 results in EGFRvIII with a truncated extracellular domain, resulting in a unique, GBM cell-specific, antibody-reactive antigen. An EGFRvIII-specific peptide conjugated to a keyhole limpet hemocyanin represents the structure of a vaccine called Rindopepimut (CDX110). The latest Phase III clinical trial showed that rindopepimut did not increase the survival rate in newly diagnosed glioblastoma patients [130–132].

#### *4.5. Targeting the Regulation of EGFR Gene Expression*

This strategy consists of using antisense oligonucleotides, siRNA, ribozymes, and miRNA. In glioblastoma, the microRNAs control the post-transcriptional gene expression of receptor tyrosine kinase (RTK) signaling pathways by blocking or accelerating the mRNA. Recent work demonstrates that extracellular vesicles (EVs) can carry and transfer EGFR [133] and that cell communication through EVs enhances glioblastoma's intratumoral heterogeneity [134]. Bronisz et al. showed that miR-1 could interact with a major EV protein Annexin A2 (ANAXA2) to reduce glioblastoma tumorigenicity [135]. Furthermore, one recent study of Liao et al. showed that the extracellular EC domain methylation using protein arginine methyltransferase 1 (PRMT1) increased EGF binding and dimerization, with enhanced receptor activation counteracting the effect of cetuximab in a mouse model of colon cancer [136]. The expression of the EGFRvIII is also under investigation. Unfortunately, none of the strategies targeting EGFR gene expression regulation have yet had any preclinical development.

#### *4.6. Nanoparticles*

In order to efficiently deliver the therapeutic agent, novel pharmaceutical formulations are currently used. It is well known that the bioavailability of drugs may be low because of the blood-brain barrier (BBB) permeability. Nanoparticles are vesicular carriers able to increase the bioavailability due to targeted drug release while protecting their content. EGFR is an ideal molecule for glioblastoma tumor targeting and numerous agents have been entrapped in a variety of nanoparticles [137].

There is only limited experience in early clinical trials for cetuximab conjugated liposomes [138]. A Phase II study just terminated in March 2020 investigated the effect of combining Temozolomide and a nanocomplex called SGT-53 (systemic gene therapy—53) to treat recurrent glioblastoma [139]. Previous results showed prolonged survival rates in glioblastoma mouse models [140].

Regardless of strategy, all therapeutic agents face the main problem of delivery across the blood–brain barrier, often cited as the explanation for EGFR targeting failure in glioblastoma [103].

#### **5. Facing a Real Challenge—Drug Resistance**

There is evidence that targeted therapy towards mutations responsible for cancer growth and progression is effective in different tumor types. For GBM, the responses to EGFR-pathway inhibitors were not as expected, and they are mainly explained by drug resistance [141]. Two major mechanisms could explain the EGFR therapy resistance.

The first resistance mechanism involves target independence. In this situation, glioma cells without EGFR protein expression experience no negative impact from EGFR inhibitors. For this type of resistance, the loss of extra-chromosomally encoded EGFR is a frequent mechanism. Target independence may occur after small molecule therapy. The dynamic EGFRvIII expression regulation by small circular fragments of extra-chromosomal DNA is involved in the resistance to EGFR inhibition. Some studies demonstrated that GBM cells treated with erlotinib reversibly suppressed mutant EGFR by producing extrachromosomal DNA, making the GBM cells resistant to EGFR inhibition. After withdrawing

erlotinib, the mutations re-emerged on extra-chromosomal DNA, leading to the upregulation of EGFRvIII with consequent re-sensitization of GBM cells [142].

The second mechanism regards target compensation. In this situation, glioma cells fight back against EGFR inhibition by activating compensatory pathways independent of EGFR signaling. Insulin-like growth factor 1 (IGFR1), platelet-derived growth factor β (PDGFβ), mesenchymal-epidermal transition (cMET), and their downstream targets are involved in these compensatory pathways [77].

Given these two mechanisms, rational strategies should include multi-target therapies targeting truncation mutations for the first mechanism and multi-target therapies targeting compensatory proteins for the second mechanism. Resistance may be overcome by dosing/epigenetic therapy, targeting truncation mutations, or through multitarget therapy.

For the first strategy, glioma cells' re-sensitization may be achieved by using drug scheduling [143]. Pulsatile intermittent drug therapy with EGFR-inhibitors in high doses can lead to better inhibition of the target, delay of therapy resistance, and toxicity reduction compared to continuous dosing [144].

Targeting truncated mutations is a suitable strategy for glioma treatment due to their high frequency in this pathology. As candidates for co-targeting, the following deserve discussion: PTEN with PI3K as a molecular target and cyclin-dependent kinase inhibitor 2 A (CDKN2A) with cyclin-dependent kinase (CDK)4/6 as a molecular target. Some PI3K inhibitors are currently undergoing clinical trials, such as GDC-0084, PX-866, pilaralisib, buparlisib, and XL765 [145]. Abemaciclib, palbociclib, and ribocyclib are examples of FDA (Food and Drug Administration)-approved oral drugs with good BBB permeability [146,147] that may be investigated for targeting CDK4/6.

In many types of cancer, multitarget therapy is a preferred option. Therapies may become more efficient by combining EGFR inhibitors with other downstream blocking agents. Several glioma specific epitopes such as IL13RA2 and EphA2 are under investigation for poly-target therapy with antibody drug conjugates (ADCs) and CAR-T cells [148–151].

In a recent 2020 study, Meng et al. proved that the cross-activation of EGFR and MET signaling pathways contributes to temozolomide resistance in glioblastoma patients. To simultaneously diminish both EGFR and MET activation, they developed a nanoinhibitor with double functionalized brain-target (BIP-MPC-NP) by conjugating cMBP and Inherbin3 modified poli-2-methacryloyloxyethyl phosphorylcholine (MPC)-nanoparticles. The study shows that DNA damage repair is reduced, and sensitivity is augmented by downregulating the E2F1 transcription factor in temozolomide resistant glioma in mice. These results demonstrate that the abovementioned nanoinhibitor is a suitable candidate for overcoming drug resistance in glioma [152].

Furthermore, patient mutations affecting the trafficking of therapeutic antibodies is another potential mechanism contributing to therapeutic resistance [153,154].

#### **6. Conclusions**

The global understanding of EGFR signaling has dramatically advanced in the last ten years. However, extensive work is still required in order to understand all signaling pathways and their implications fully. The application of EGFR-targeted therapy for glioma treatment has been less successful than expected. Gliomas require a complex signaling network that dictates the tumor sensitivity of EGFR-targeted therapies. Low BBB penetration, as well as tumor heterogeneity, secondary mutations, and compensatory signaling pathways, contribute to resistance. The development of new combinatorial therapies may improve patient quality of life through personalized, tailored choices of appropriate therapeutic strategies. An integrated approach is required to offer a complete view of this critical receptor by combining cellular, biochemical, structural, and genetic modeling techniques.

**Author Contributions:** Conceptualization, A.-S.S.; methodology, A.O., S.-C.B., D.E.T., S.D.; investigation, A.O., S.-C.B.; resources, G.-A.S.; writing—original draft preparation, O.A.; writing—review and editing, A.O., E.S.M., A.-S.S.; visualization, A.-S.S.; supervision, A.-S.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are openly available.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

