Epidermal Growth Factor Receptor 1 Expression Is Upregulated in Undifferentiated Thyroid Carcinomas in Humans

Abstract

Background:

Epidermal growth factor receptor 1 (EGFR1) signaling is involved in human cancer cell progression and is responsible for aggressive biological behavior and poor clinical outcome in several human malignancies. Activation of the EGFR1 pathway has been proposed, among others, as being involved in the progression of thyroid cancer toward a thyroid-stimulating hormone (TSH)-independent phenotype. We have previously observed that undifferentiated thyroid carcinoma cells are hyper-sensitive to EGF signaling of downstream intracellular pathways, and this correlated both with the loss of TSH-dependency and increase in EGF-dependent proliferation and migration. Thus, we hypothesized that the upregulation of EGFR1 protein expression may be enhanced in parallel with transition toward a poorly differentiated phenotype in human thyroid carcinomas.

Methods:

The expression of EGFR1 was evaluated, by immunohistochemistry, in a series of 49 human thyroid carcinomas at different degrees of tumor differentiation.

Results:

The expression of EGFR1 protein was significantly upregulated in poorly differentiated and anaplastic thyroid carcinomas, whereas it was absent or faint in normal thyroid gland tissue and in differentiated thyroid papillary carcinomas. Of note, selected thyroid tumors characterized by a mixed population of differentiated and undifferentiated tumor cells, likely progressing from well to poorly differentiated and anaplastic phenotypes, exhibited EGFR1-negative differentiated fields together with EGFR1-positive poorly differentiated and anaplastic areas.

Conclusions:

Upregulation of EGFR1 expression may be a molecular marker of dedifferentiation in thyroid epithelial carcinomas, likely being responsible for the activation of EGF signaling observed in tumor cells and favoring progression toward an angiogenic, poorly differentiated, TSH-independent phenotype.

Introduction

T he majority of thyroid tumors are well-differentiated malignancies, with favorable prognosis and high sensitivity to standard treatments, whereas a minority of thyroid epithelial cancers are poorly differentiated or anaplastic, and are characterized by massive lymphatic invasion and high metastatic potential, which accounts for their poor outcome and aggressive clinical behavior. While some of these malignancies exhibit an aggressive undifferentiated phenotype from the outset, others eventually become aggressive as a result of a progressive dedifferentiation process from initially well-differentiated tumors (1,2).

In thyroid epithelial carcinomas, differentiation refers to the maintenance of cellular functions that are characteristic of normal thyroid follicular cells, and, among others, thyroid-stimulating hormone (TSH) dependency, which is responsible for the likelihood of a beneficial response to radioiodine therapy. By contrast, the major trait of poorly differentiated thyroid carcinomas is the loss of TSH dependency, and, as a consequence, the inability to express sodium-iodide symporter upon TSH stimulation and to respond to radioiodine therapy (3,4). It has been proposed that the loss of TSH receptor signaling correlates with the parallel activation of other growth factor signaling pathways with known function in epithelial-to-mesenchymal transition (EMT), and likely involved in driving tumor progression toward a TSH-independent phenotype (5). In such a perspective, we recently reported that the full activation of epidermal growth factor receptor 1 (EGFR1) signaling in human thyroid carcinoma cells may represent, among others, a molecular event responsible for the transition toward a poorly differentiated and aggressive phenotype and for the loss of TSH-dependency (6). However, unlike other human solid malignancies that are characterized by the presence of activating mutations of the tyrosine kinase domain of EGFR1 (7), these have been detected only in a minority of thyroid carcinomas (8 –10), suggesting that the activation of EGF signaling may be driven by the upregulation of its components, rather than by genomic mutation events. Based on this hypothesis, we studied the expression of EGFR1 protein in human thyroid carcinomas, to determine the likelihood of it being a mechanism responsible for the enhanced EGF signaling that we previously noted in TSH independent thyroid carcinoma FRO cells (6).

Patients and Methods

Study population and diagnostic criteria

The study group comprised 49 patients with thyroid cancers at different degrees of tumor grading and 10 control cases of lung and oral epithelial cancers; all the controls were previously characterized for EGFR1 status in our laboratories. The histopathological diagnoses were performed at the Histopathology Sections at Foggia and Siena Universities. The study population of thyroid tumors was subdivided into three subgroups: (i) well-differentiated (30 cases), (ii) poorly differentiated (9 cases), and (iii) anaplastic (10 cases) carcinomas, according to the World Health Organization (WHO) classification of tumors (11). Further, tumor staging was established according to the TNM system (12). Patients' characteristics are reported in Table 1. The study was performed in double blind with respect to histology and tumor grading. Express written informed consent to use biological specimens for investigational procedures was obtained from all patients or from their relatives.

Table 1.

Baseline Characteristics of the Patients

Patients	n=49
Age
Median (range)	62 (22–96)
Sex
Female	36
Male	13
Histotypes
Papillary carcinomas	30
Classic type	20
Follicular variant	5
Oncocytic variant	1
Columnar variant	1
Fasciitis-like stroma variant	1
Diffuse sclerosing variant	2
Poorly differentiated carcinoma	9
Insular variant	5
Papillary variant	4
Anaplastic Carcinoma	10
Unifocal/multifocal	36/13
TNM
pTx	1
pT1	18
pT2	1
pT3	12
pT4	17
pNx	6
pN0	29
pN1	14
cMx	19
cM0	26
cM1	4

Immunohistochemistry

Four-micrometer serial sections from formalin-fixed and paraffin-embedded blocks were cut and mounted onto poly-L-lysine-coated glass slides. Immunostaining was performed using the linked streptavidin-biotin horseradish peroxidase technique. After sequential deparaffinization and hydration, the slides were treated with 0.3% (v/v) H₂O₂ for 15 minutes to quench endogenous peroxidase. Antigen retrieval was performed by microwave heating—the first time for 3 minutes at 650 W, and the second and third times for 3 minutes at 350 W—the slides immersed in 10 mM citrate buffer, pH 6. After heating, the sections were blocked for 60 minutes with 1.5% (v/v) horse serum (Santa Cruz Biotechnology, Segrate, Italy) diluted in phosphate-buffered saline buffer and incubated for 32 minutes at 24°C with the primary prediluted anti-EGFR1 antibodies (clone 31G7; Ventana, San Diego, CA), following protease pretreatment. All the slides were treated with biotinylated species-specific secondary antibodies and streptavidin-biotin enzyme reagent, and the color was developed by 3,3′-diaminobenzidine tetrahydrochloride using the standard automatic Ventana Benchmark System^® (Ventana). Sections were counterstained with Mayer's hematoxylin and mounted using xylene-based mounting medium.

The results of the immunohistochemical staining were evaluated separately by two independent observers. Immunostained cells were counted in at least 10 high-power fields analyzed by a light microscope. Staining patterns were also evaluated and recorded as membranous, cytoplasmic, or mixed membranous-cytoplasmic. For each case, the cumulative percentage of positive cells in all sections examined was determined. Some cases showed in the same slides a progression from well to poorly differentiated and/or anaplastic carcinomas. In the latter cases, EGFR1 evaluation was performed separately in at least four fields corresponding to different cancer differentiation areas. After quantitative analysis of cells stained for EGFR1, all the cases were scored semiquantitatively, as follows: 0, negative; +1, weak reactivity that was membranous, cytoplasmic, or both; +2, circumferential membrane staining of intermediate intensity and frequent cytoplasmic reactivity that was less intense than the membrane reactivity; +3, complete strong circumferential staining, usually associated with weaker cytoplasmic staining (13).

Statistical analysis

Chi-square test was used to establish the statistical significance between the difference in EGFR1 levels, comparing the frequency of EGFR1-negative (0 or +1) and EGFR1-positive tumors (+2 or +3), in differentiated versus poorly differentiated and/or anaplastic carcinomas. Statistically significant values (p<0.05) are reported in the Results section.

Results

EGFR1 expression was evaluated by immunohistochemistry in 49 thyroid carcinomas at different degrees of tumor differentiation and in the surrounding peritumoral noninfiltrated thyroid gland. The majority of the study population was characterized by well-differentiated papillary carcinomas, 20 classic types, and 10 more rare variants. Nine poorly differentiated and 10 anaplastic thyroid carcinomas were also included in the analysis (Table 1). EGFR1 expression was assessed semiquantitatively (i) according to criteria adopted for colon cancer (13), (ii) as percentage of stained cells, and (iii) according to cellular localization of positive staining. Results are reported as case-by-case detailed descriptions of EGFR1 expression in thyroid carcinomas (Table 2), and as levels of EGFR1 expression according to histotype variants and degree of tumor differentiation (Table 3). EGFR1 immunostaining was negative in the vast majority of normal thyroid noninfiltrated specimens (Table 2 and Fig. 1a), only 7/42 normal thyroid glands showing focal, weak cytoplasmic staining. Complete circumferential staining in the majority of cells was observed in just two normal thyroid glands (Table 2). Consistently, 18/20 classic thyroid papillary carcinomas exhibited negative or faint EGFR1 staining. Low or negative EGFR1 expression was also observed in 8/10 variants of papillary carcinomas (4/5 follicular, 1/1 oncocytic, 1/1 columnar, 1/1 fasciitis-like stroma, and 1/2 diffuse sclerosing variants) (Table 2 and Fig. 1b–d). EGFR1 was highly expressed in 2/20 classic papillary carcinomas, in 1/5 follicular, and in 1/2 diffuse sclerosing variants of papillary carcinoma (Table 2). The quantification of EGFR1 expression as a percentage of positive papillary carcinoma cells revealed a broad range (5%–100%) of EGFR1-positive cells with a prevalent cytoplasmic or a mixed cytoplasmic/membranous pattern. The presence of EGFR1-positive staining in a minority of papillary carcinoma (4/30 cases) was not associated with higher primary tumor size and/or tumor stage or traits of more aggressive clinical behavior (lymph node or distant metastases, shorter progression-free, or overall survival). Of note, the only 2 EGFR1-positive papillary classic carcinomas were both microcarcinomas, whereas several other microcarcinomas showed low or absent ERGF1 expression (Table 2).

FIG. 1.

Epidermal growth factor receptor 1 (EGFR1) expression in normal thyroid gland and in 7 thyroid carcinomas at different degrees of differentiation representative of our study population. (a) Normal thyroid gland, showing negative staining for EGFR1; (b) papillary carcinoma, columnar variant (case no. 27), with moderate cytoplasmic staining; (c) papillary carcinoma, fasciitis-like stroma variant (case no. 28), showing membranous staining; (d) papillary carcinoma, classic type (case no. 10), with faint cytoplasmic staining; (e) poorly differentiated carcinoma, with a solid-insular histological pattern, showing negative or focal faint staining (case no. 33); (f) poorly differentiated carcinoma (case no. 35), with strong membrane-cytoplasmic staining; (g) poorly differentiated carcinoma (case no. 39), showing strong membrane and moderate cytoplasmic positivity; (h) anaplastic carcinoma (case no. 42), with intense membrane staining especially in acantholytic aggressive cells; (i) anaplastic carcinoma (case no. 44), with strong complete circumferential membrane staining in all the cancerous cells. Note the negative or faint cytoplasmic staining in normal thyroid infiltrated by the tumor in (F). Linked streptavidin biotin-horseradish peroxidase (LSAB-HRP) technique, nuclear counterstaining with hematoxylin; all original magnifications are 200×. Color images available online at www.liebertonline.com/thy

Table 2.

Clinical and Histopathological Features and Epidermal Growth Factor Receptor 1 Expression in Patients with Thyroid Carcinoma

							EGFR1 protein		EGFR1 gene
Case no.	Age	Sex	Histotype	Tumor size (cm)	Unifocal/multifocal	TNM	Normal thyroid	Tumor standard score ^a	Tumor quantitative evaluation (%) ^b	Tumor cellular localization of staining
1	45	F	Papillary, classic type	0.6	U	pT1pN0cM0	Neg.	0	Neg.	—
2	68	F	Papillary, classic type	0.8	U	pT1pN0cM0	Neg.	0	Neg.	—
3	62	F	Papillary, classic type	0.1	U	pT1pN0cM0	Focal C	+1	100	C/M
4	55	M	Papillary, classic type	0.3	U	pT1pN0cM0	Neg.	+1	30	C
5	71	F	Papillary, classic type	0.3	U	pT1pN0cM0	Focal C	0	0	—
6	22	F	Papillary, classic type	0.4	U	pT1pN0cM0	Positive C	+1	100	C
7	36	F	Papillary, classic type	0.6	U	pT1pN0cM0	Neg.	+1	100	C
8	61	F	Papillary, classic type	0.2	U	pT1pN0cM0	Neg.	0	Neg.	—
9	73	F	Papillary, classic type	0.7	U	pT1pN0cM0	Positive C	+3	40	M/C
10	43	F	Papillary, classic type	0.4	U	pT1pN0cM0	Neg.	+1	30	C
11	62	F	Papillary, classic type	1.5	U	pT1pN0cM0	Neg.	+1	25	C/M
12	33	F	Papillary, classic type	1.5	U	pT1pN0cM0	Neg.	0	Neg.	—
13	51	F	Papillary, classic type	1.8	U	pT3pN0cM0	Neg.	+1	30	C/M
14	55	F	Papillary, classic type	1.8	U	pT1pN0cM0	Neg.	+1	40	C/M
15	41	F	Papillary, classic type	1.7	U	pT1pN0cM0	Neg.	0	Neg.	—
16	48	M	Papillary, classic type	2.0	U	pT1p0cM0	Neg.	+1	100	C/M
17	38	F	Papillary, classic type	1.3	U	pT3pN1cMx	Neg.	+1	100	C
18	50	F	Papillary, classic type	1.5	U	pT1pN0cMx	Neg.	+1	30	C/M
19	33	F	Papillary, classic type	1.2	U	pT1pNxcMx	Neg.	+2	100	C/M
20	50	F	Papillary, classic type	2.0	U	pT2pN0cM0	Neg.	+1	50	C
21	58	F	Papillary, follicular variant	1.7	U	pT3pN0cM0	Neg.	+2	60	M/C
22	73	M	Papillary, follicular variant	3.0	U	pT3pN1cM0	Neg.	+1	30	C/M
23	25	M	Papillary, follicular variant	5.0, 3.0, 8.0	MF	pT4pN1cM0	Focal C	+1	100	C/M
24	65	F	Papillary, follicular variant	1.5	U	pT4pN0cMx	Focal C	+1	40	C
25	76	F	Papillary, follicular variant	3.0	U	pT4pN1cM0	Neg.	0	Neg.	—
26	58	M	Papillary, oncocytic variant	0.8	U	pT1pN0cM0	Neg.	0	Neg.	—
27	78	M	Papillary, columnar variant	2.5	U	pT3pN1cMx	n.v.	0	<5	C
28	46	F	Papillary, fasciitis-like stroma variant	4.0	MF	pT3pN0cMx	Neg.	+1	20	M
29	41	M	Papillary, diffuse sclerosing variant	Diffuse	MF	pT4pNxcM0	Focal C	+2	30	M/C
30	42	F	Papillary, diffuse sclerosing variant	Diffuse	MF	pT3pN1cMx	Focal C	0	<5	M
31	26	F	Poorly differentiated, insular variant	Diffuse	MF	pT4pN1cM1	Neg.	+1 (AF)	100	C
32	69	M	Poorly differentiated, insular variant	3.0	U	pT3pNxcMx	Neg.	0	Neg.	—
33	79	M	Poorly differentiated, insular variant	Diffuse	MF	pT3pN1cMx	Neg.	0	8	C
34	69	F	Poorly differentiated, insular variant	Diffuse	MF	pTxpNxcMx	Neg.	+3 (AF)	30	M
35	24	F	Poorly differentiated, papillary variant	2.0	U	pT3pN1cMx	Neg.	+3 (AF)	80	M/C
36	69	M	Poorly differentiated, papillary variant	Diffuse	MF	pT4pN1cM1	Neg.	+3 (AF)	70	M
37	54	F	Poorly differentiated, papillary variant	4.0	U	pT4pN1cM1	n.v.	+1 (AF)	<10	M/C
38	81	F	Poorly differentiated, insular vaiant	Diffuse	MF	pT4pN1cMx	n.v.	0	<5	M
39	74	M	Poorly differentiated, papillary variant	Diffuse	MF	pT4pNxcMx	Neg.	+2 (AF)	40	M/C
40	96	F	Anaplastic	1.5	MF	pT4pNxcMx	Neg.	+3	100	M
41	79	F	Anaplastic	Diffuse	MF	pT3pN0cM0	Neg.	+1	20	C
42	78	F	Anaplastic	10.0	U	pT4pN1cM0	n.v.	+3	100	M
43	81	M	Anaplastic	7.0	U	pT4pN0cM0	Neg.	+1	40	C/M
44	70	M	Anaplastic	4.0	U	pT3pN0cM0	Neg.	+3	100	M
45	75	F	Anaplastic	8.0	U	pT4pN0cMx	Focal C	+3	60	M
46	81	F	Anaplastic	5.0	U	pT4pN1cMx	n.v.	+3	60	M
47	83	F	Anaplastic	Diffuse	MF	pT4pN0cM1	Neg.	+2	75	C/M
48	79	F	Anaplastic	6.0	U	pT4pN0cMx	n.v.	+2	30	M/C
49	78	F	Anaplastic	8.0	U	pT4pNxcMx	n.v.	+1	60	C/M

Standard score employed according to criteria adopted for colon cancer.

Percentage of cell stained irrespective of staining intensity and cellular localization (see Patients and Methods section).

U, unifocal; MF, multifocal; Neg., negative; n.v., not valuable (lack of normal thyroid tissue); C, cytoplasmic; M, membranous; C/M, mixed cytoplasmic and membranous staining with prevalent cytoplasmic localization; M/C, mixed staining with prevalent membranous localization; AF, anaplastic field; EGFR1, epidermal growth factor receptor 1.

Table 3.

Epidermal Growth Factor Receptor 1 Expression in Thyroid Carcinomas According to Histotype Variants

	EGFR1 quantification ^a
Histotypes	0	+1	+2	+3
Papillary carcinomas^b,c	10	16	3	1
Classic type	6	12	1	1
Follicular variant	1	3	1	—
Oncocytic variant	1	—	—	—
Columnar variant	1	—	—	—
Fasciitis-like stroma variant	—	1	—	—
Diffuse sclerosing variant	1	—	1	—
Poorly differentiated Carcinomas	3	2	1	3
Insular variant	3	1	—	1
Papillary variant	—	1	1	2
Anaplastic carcinomas	—	3	2	5

According to colon cancer criteria (13).

Papillary vs. anaplastic: p=0.002.

Papillary vs. all: p=0.003.

Poorly differentiated thyroid tumors showed a mixed EGFR1 expression pattern: 5/9 tumors revealed negative or weak EGFR1 immunoreactivity, whereas 4/9 cases exhibited high EGFR1 protein levels (Table 3 and Fig. 1e–g). Interestingly, among the poorly differentiated carcinomas, insular variants more frequently (4/5 tumors) exhibited negative or low EGFR1 expression, whereas papillary variants more frequently (3/4 cases) showed positive staining with EGFR1 (Table 3 and Fig. 1e–g). By contrast, in 7/10 cases, anaplastic thyroid carcinomas showed high levels of membranous-cytoplasmic EGFR1 expression (Table 3 and Fig. 1h, i). Consistently, the percentage of EGFR1-positive cells was generally higher in anaplastic thyroid carcinomas compared to differentiated tumors (Table 2). Noteworthily, the difference in EGFR1 levels was statistically significant when we compared EGFR1-negative (0 or +1) and EGFR1-positive tumors (+2 or +3) in papillary versus anaplastic carcinomas (chi-square test, p=0.002) and in papillary versus poorly differentiated and anaplastic carcinomas (chi-square test, p=0.003).

Since these results suggest that the expression of EGFR1 is progressively upregulated from differentiated to poorly differentiated and anaplastic thyroid carcinomas, we questioned whether EGFR1 expression is induced in thyroid tumor carcinoma cells undergoing progression toward a poorly differentiated phenotype. To this purpose, six poorly differentiated thyroid tumors with concomitant presence in the same slides as fields with well-differentiated and poorly differentiated cells were selected from our study population and stained for EGFR1 expression. Figure 2 reports EGFR1 staining in poorly differentiated carcinoma no. 39 as a representative case of this subgroup of tumors, showing papillary (Fig. 2a) and insular tumor areas (Fig. 2b) as well as regions of undifferentiated carcinoma (Fig. 2c, d). Interestingly, in these tumors a correlation between loss of differentiation and increase in EGFR1 expression was observed. Indeed, tumor cells in more differentiated tumor regions were always characterized by negative or low EGFR1 immunoreactivity, whereas EGFR1 expression was higher in contiguous areas with poorly differentiated tumor cells. Accordingly, the EGFR1 quantification reported in Table 2 for these 6 tumors refers to undifferentiated fields. This observation suggests that the upregulation of EGFR1 expression may be a marker of loss of differentiation in thyroid carcinomas undergoing progression toward a poorly differentiated phenotype.

FIG. 2.

EGFR1 expression in different areas of poorly differentiated thyroid carcinoma no. 39. (a) Differentiated papillary carcinoma area, showing very faint cytoplasmic EGFR1 expression; (b) insular carcinoma area, with negative immune-staining; (c and d) poorly differentiated carcinoma area, showing strong membrane staining for EGFR1 especially in the fusiform fibroblastoid dedifferentiated cells. LSAB-HRP technique, nuclear counterstaining with hematoxylin. Color images available online at www.liebertonline.com/thy

Discussion

EGFR1 is a cell-membrane receptor that plays a key role in cancer development and progression (14). Several lines of evidence suggest that the overexpression of EGFR1 gene correlates with advanced tumor stage, aggressive biological behavior, and poor clinical outcome in several human malignancies, including lung, colorectal, breast, bladder, head, and neck carcinomas (15). Indeed, the induction of EGFR1 signaling by either stimulation with its ligands or activating mutations of the receptor's tyrosine kinase (TK) domain leads to the switching of a number of intracellular signaling pathways, and this in turn triggers a cascade of cell responses ranging from cell growth, proliferation, and invasion to cell survival and angiogenesis (14).

The role of EGFR1 signaling in the transition toward a poorly differentiated phenotype in thyroid cancer is the object of investigation, with contradictory results having so far been achieved. While previous studies demonstrated increased EGFR1 expression in human anaplastic thyroid carcinoma (16,17), recently, it has been suggested that EGFR1 expression is critical in the fate of well-differentiated thyroid carcinomas, since the loss of its expression correlates with lymphatic invasion and lymph node metastases in papillary microcarcinomas (18). In the present study, we report that EGFR1 protein expression is progressively upregulated from well-differentiated thyroid papillary carcinomas to poorly differentiated and anaplastic thyroid tumors, suggesting that the overexpression of EGFR1 may be a molecular event involved in the progression of thyroid carcinomas. Indeed, EGFR1 expression is absent or faint both in normal thyroid gland and in the vast majority of well-differentiated thyroid carcinomas and this is consistent with the maintenance of TSH dependency in these tissues (1,3). The results of this study are in line with our previous observation that differentiated thyroid carcinoma cells are characterized by low levels of EGFR1 expression and by lack of activation of AKT signaling, cell migration, and proliferation in response to EGF (6). It is important to note that we found definite, intense membrane circumferential EGFR1 expression just in cancer cells showing a spindle fibroblastoid or large anisocytotic shape and/or an acantholytic pattern of growth (see Figs. 1 and 2), whereas the majority of well-differentiated thyroid tumor cells showed a faint cytoplasmic pattern of EGFR1 expression. As regards its intracellular localization, it should be borne in mind that this intracellular EGFR1 content may be mainly made up of ligand-bound receptor multimers, which are likely inactive and destined to be internalized in clathrin-coated pits and targeted for ubiquitylation and destruction, or recycled to the cell membrane (19). It is also noteworthy that selected poorly differentiated thyroid tumors undergoing progression from differentiated to anaplastic phenotypes are characterized by EGFR1-negative tumor cells in more differentiated tumor areas and EGFR1-positive cells in poorly differentiated and anaplastic areas. These results support the hypothesis that EGFR1 is progressively upregulated in parallel with the loss of differentiation, likely playing a role in driving tumor progression. Thus, we suggest that EGRF1 upregulation may be interpreted as a feature associated with the poorly differentiated phenotype in thyroid tumors. The hypothesis is further supported by the low frequency of activating mutations in the TK domain of EGFR1 reported in thyroid malignancies (8 –10), which strongly suggests that the activation of EGF signaling is driven by receptor upregulation rather than by genomic mutation events.

In such a perspective, recent in vitro evidence reported by our group suggests that increased sensitivity to EGF stimulation may be involved in the loss of TSH-dependency and, possibly, in the EMT of thyroid carcinomas (6). Indeed, prolonged EGF stimulation of initially TSH-dependent thyroid carcinoma cells resulted in a progressive loss of TSH dependency. Accordingly, ligand-dependent EGFR1 activation induced a selective upregulation of AKT phosphorylation in human thyroid TSH-independent carcinoma cells, concomitantly with increased rates of cell proliferation and migration (6). Based on this evidence, it is intriguing to hypothesize that, while TSH is the major regulator of cell fate in well-differentiated thyroid tumors (1), the acquired sensitivity to TK-dependent growth factors, and among others EGF, may favor the transition toward a more aggressive mesenchymal-like undifferentiated condition, as provided by the loss of TSH-dependency, the activation of a more invasive and metastatic phenotype and the induction of angiogenic potential (5). In such a perspective, it is likely that EGFR1 signaling cooperate with other signaling pathways activated by BRAF V600E, the most common somatic mutation in papillary thyroid carcinoma. Indeed, it has been recently proposed that the effects of BRAF V600E on cell surface receptors and extracellular matrix noncellular components trigger different pathologic biological responses, such as the activation of angiogenesis, proliferation, adhesion, and motility as well as resistance to apoptosis, and are thus involved in thyroid tumor progression (20).

These observations may open up new perspectives in the clinical management of thyroid carcinomas. A diagnostic issue in thyroid cancers is the early identification of tumors undergoing rapid loss of differentiation to provide more aggressive and effective treatments (21). The evidence that a minority of well-differentiated tumors are characterized by high expression of EGFR1 suggests that its upregulation may represent an early event in tumor progression and deserves to be further evaluated as a marker for the early detection of those thyroid tumors undergoing rapid dedifferentiation. This issue may be clinically relevant in the early prognostic characterization of the small subgroup of papillary microcarcinomas that, while still maintaining the morphological features of well-differentiated tumors, are characterized by more aggressive clinical behavior (22). In such a perspective, the size of our study population does not allow for any conclusions about the clinical significance of high EGFR1 expression in well-differentiated thyroid tumors. Further, EGFR1 signaling may be a molecular target for the treatment of poorly differentiated thyroid malignancies, since in the last few years several pharmacological tools have been developed to inhibit the EGFR1 pathway (23). However, a phase II study demonstrated that the treatment of human iodine-refractory or anaplastic thyroid carcinomas with the EGFR1 TK inhibitor gefitinib achieves only 48% disease stability, raising concerns about the clinical activity of this agent in thyroid malignancies (24). By contrast, some studies reported significant inhibition of EGF-dependent growth in anaplastic thyroid carcinoma cell lines (25), and vandetanib, a dual inhibitor of EGFR1 and VEGFR2, showed a significant antitumor activity in mice xenografts of human anaplastic thyroid carcinoma cells (26). In line with this observation, our group recently demonstrated that poorly differentiated thyroid carcinoma cells are sensitive to anti-EGFR1 inhibitors, used either as single agents or in combination with chemotherapeutics (27). Thus, further studies are needed to evaluate whether, among several molecular modifications that characterize thyroid tumor cells, EGFR1 upregulation may represent a predictive marker of sensitivity to anti-EGFR1 therapy.

Footnotes

Acknowledgments

This work was supported by Grant 26/08 from Lega Italiana per la Lotta ai Tumori (LILT) to M.C., and Grants IG8780 from the Associazione Italiana per la Ricerca sul Cancro (AIRC), and Ministero dell'Istruzione dell'Università e della Ricerca (PRIN 2008) to M.L. We thank Anthony Green for proofreading the article and suggesting stylistic improvements.

Disclosure Statement

The authors declare that no competing financial interests exist.

References

Riesco-Eizaguirre

, Santisteban

. 2007. New insights in thyroid follicular cell biology and its impact in thyroid cancer therapy. Endocr Relat Cancer, 14:957–977.

Patel

, Shaha

. 2006. Poorly differentiated and anaplastic thyroid cancer. Cancer Control, 13:119–128.

Kinder

. 2003. Well differentiated thyroid cancer. Curr Opin Oncol, 15:71–77.

Are

, Shaha

. 2006. Anaplastic thyroid carcinoma: biology, pathogenesis, prognostic factors, and treatment approaches. Ann Surg Oncol, 13:453–464.

Vasko

, Saji

. 2007. Molecular mechanisms involved in differentiated thyroid cancer invasion and metastasis. Curr Opin Oncol, 19:11–17.

Landriscina

, Piscazzi

, Fabiano

, Maddalena

, Costantino

, Farese

, Bufo

, Cignarelli

. 2009. Targeting epidermal growth factor receptor 1 signaling in human thyroid-stimulating hormone-independent thyroid carcinoma FRO cells results in a more chemosensitive and less angiogenic phenotype. Thyroid, 19:629–637.

Ferguson

. 2008. Structure-based view of epidermal growth factor receptor regulation. Annu Rev Biophys, 37:353–373.

Mitsiades

, Kotoula

, Poulaki

, Sozopoulos

, Negri

, Charalambous

, Fanourakis

, Voutsinas

, Tseleni-Balafouta

, Mitsiades

. 2006. Epidermal growth factor receptor as a therapeutic target in human thyroid carcinoma: mutational and functional analysis. J Clin Endocrinol Metab, 91:3662–3666.

Lee

, Lee

, Kong

, Kook

, Yang

, Park

, Keam

, Park do

, Cho

, Kim

, Chung

, Lee

, Kim

. 2007. Epidermal growth factor receptor status in anaplastic thyroid carcinoma. J Clin Pathol, 60:881–884.

10.

Masago

, Asato

, Fujita

, Hirano

, Tamura

, Kanda

, Mio

, Katakami

, Mishima

, Ito

. 2009. Epidermal growth factor receptor gene mutations in papillary thyroid carcinoma. Int J Cancer, 124:2744–2749.

11.

DeLellis

, Lloyd

, Heitz

, Eng

. 2004. Pathology, Genetics of Tumours of Endocrine Organs. IARC Press: Lyon.

12.

Greene

, Page

, Fleming

, Fritz

, Balch

, Haller

, Morrow

. 2002. American Joint Committee on Cancer. Cancer Staging Manual, 6th. Springer: New York.

13.

Goldstein

, Armin

. 2001. Epidermal growth factor receptor immunohistochemical reactivity in patients with American Joint Committee on Cancer Stage IV colon adenocarcinoma: implications for a standardized scoring system. Cancer, 92:1331–1346.

14.

Hynes

, MacDonald

. 2009. ErbB receptors and signaling pathways in cancer. Curr Opin Cell Biol, 21:177–184.

15.

Dei Tos

. 2007. The biology of epidermal growth factor receptor and its value as a prognostic/predictive factor. Int J Biol Markers, 22,1 Suppl 4:S3–S9.

16.

Ensinger

, Spizzo

, Moser

, Tschoerner

, Prommegger

, Gabriel

, Mikuz

, Schmid

. 2004. Epidermal growth factor receptor as a novel therapeutic target in anaplastic thyroid carcinomas. Ann N Y Acad Sci, 1030:69–77.

17.

Murakawa

, Tsuda

, Tanimoto

, Tanabe

, Kitahara

, Matsubara

. 2005. Expression of KIT, EGFR, HER-2 and tyrosine phosphorylation in undifferentiated thyroid carcinoma: implication for new therapeutic approach. Pathol Int, 55:757–765.

18.

Lim

, Baek

, Lee

, Park

, Kim

, Kang

, Jeon

, Lee

, Yun-Cha

, Lee

, Son

, Kang

. 2007. Clinical, histopathological, and molecular characteristics of papillary thyroid microcarcinoma. Thyroid, 17:883–888.

19.

Ratushny

, Astsaturov

, Burtness

, Golemis

, Silverman

. 2009. Targeting EGFR resistance networks in head and neck cancer. Cell Signal, 21:1255–1268.

20.

Nucera

, Lawler

, Parangi

. 2011. BRAF(V600E) and microenvironment in thyroid cancer: a functional link to drive cancer progression. Cancer Res, 71:2417–2422.

21.

Liu

, Wu

, van Hasselt

, Chen

. 2010. Carcinogenesis and therapeutic strategies in thyroid cancer. Curr Drug Targets, 11:716–732.

22.

Grodski

, Delbridge

. 2009. An update on papillary microcarcinoma. Curr Opin Oncol, 21:1–4.

23.

Ciardiello

, Tortora

. 2008. EGFR antagonists in cancer treatment. N Engl J Med, 358:1160–1174.

24.

Pennell

, Daniels

, Haddad

, Ross

, Evans

, Wirth

, Fidias

, Temel

, Gurubhagavatula

, Heist

, Clark

, Lynch

. 2008. A phase II study of gefitinib in patients with advanced thyroid cancer. Thyroid, 18:317–323.

25.

Nobuhara

, Onoda

, Yamashita

, Yamasaki

, Ogisawa

, Takashima

, Ishikawa

, Hirakawa

. 2005. Efficacy of epidermal growth factor receptor-targeted molecular therapy in anaplastic thyroid cancer cell lines. Br J Cancer, 92:1110–1116.

26.

Gule

, Chen

, Sano

, Frederick

, Zhou

, Zhao

, Milas

, Galer

, Henderson

, Jasser

, Schwartz

, Bankson

, Myers

, Lai

. 2011. Targeted therapy of VEGFR2 and EGFR significantly inhibits growth of anaplastic thyroid cancer in an orthotopic murine model. Clin Cancer Res, 17:2281–2291.

27.

Landriscina

, Maddalena

, Fabiano

, Piscazzi

, La Macchia

, Cignarelli

. 2010. Erlotinib enhances the proapoptotic activity of cytotoxic agents and synergizes with paclitaxel in poorly-differentiated thyroid carcinoma cells. Anticancer Res, 30:473–480.