INTERNATIONAL JOURNAL OF ONCOLOGY 26: 1177-1185, 2005

Expression of EGFR, HER2, HER3, and HER4 in metastatic squamous cell carcinomas of the oral cavity and base of tongue T. EKBERG1, M. NESTOR1, M. ENGSTRÖM1, H. NORDGREN2, K. WESTER3, J. CARLSSON4 and M. ANNIKO1 1Otorhinolaryngology 4Biomedical

and Head & Neck Surgery, Akademiska Sjukhuset, 2Genetics and Pathology, 3Experimental Urology, Radiation Sciences, Rudbeck Laboratory, Uppsala University, SE-751 85 Uppsala, Sweden Received October 21, 2004; Accepted December 15, 2004

Abstract. The expressions of all four receptors in the epidermal growth factor receptor family, EGFR. HER2, HER3, and HER4 were evaluated by immunohistochemistry in 19 cases of metastatic squamous cell carcinoma of the oral cavity and base of tongue. EGFR had a similar and high expression in both primary tumours and the corresponding metastases, while the expression in normal epithelium was lower in most cases. HER2 was not expressed to the same extent as EGFR. However, when HER2 was well expressed, it was in most cases expressed to the same extent and intensity in the primary tumours, metastases, and normal epithelium. The expression of HER3 and HER4 varied and was mainly cytoplasmic in all cases studied. No overexpression of HER3 and HER4 in tumours was seen as compared to normal epithelium. In order to further investigate the distribution of HER3, two HER3 expressing cell lines originating from tongue cancer were analysed in vitro, using radiolabelled anti-HER3 antibodies directed to the extracellular domains of the receptor. The results indicated that HER3 was not present in measurable amounts in the cellular membrane. There is a need for improved diagnostics and therapy for the studied type of tumours, e.g. using radiolabelled antibodies or ligands, and EGFR seemed suitable as target since the expression was high, membrane associated and similar in the primary tumours and the corresponding metastases. Introduction Squamous cell carcinomas of the head and neck region (HNSCC) initially spread locally in the near epithelium. The macroscopic appearance of the tumours may be well defined, but the seemingly normal adjacent epithelium might be invaded with single tumour cells and small islands of microscopic tumours. In a later phase the tumours form lymph node

_________________________________________ Correspondence to: Dr Tomas Ekberg, Otorhinolaryngology and Head & Neck Surgery, Akademiska Sjukhuset, Uppsala University, SE-751 85 Uppsala, Sweden E-mail: [email protected]

Key words: EGFR, HER, HNSCC, metastasis, receptors, squamous carcinoma

metastases. The incidence of nodal metastases varies with size, histological grade of differentiation, mode of invasion and primary site of the carcinoma (1). Therapy usually consists of surgery, external radiotherapy, or a combination of these. Chemotherapy is included when dissemination is suspected. Radiation and chemotherapy have substantial toxicity to normal tissue, which also can be additive, and sometimes prevent optimal therapeutic doses to be given. Tumours can also become resistant to chemotherapy. There is a need for new approaches to both diagnostics and treatment, possibly combined with existing forms. One such approach is receptor mediated tumour targeting using radiolabelled antibodies or ligands (2). Receptor targeting is becoming a clinical reality for tumours expressing epidermal growth factor receptor (EGFR) (3-6) and HER2 (7), both members of the erbB/HER type 1 tyrosine kinase receptor family, which includes EGFR (erbB-1/ HER1), HER2 (neu/erbB-2), HER3 (erbB-3), and HER4 (erbB-4). They are important for mediating signals to control proliferation and differentiation (8). The signal transduction is mediated through binding of a ligand to the receptor (except for HER2 that has no known ligand) (9), which then forms a hetero- or homodimer with another receptor in the family. The receptors initiate intracellular signalling via various phosphorylation steps. The HER3 receptor has no intrinsic tyrosine kinase activity, but can be transphosphorylated by both EGFR and HER2 (10). Since there are four known members of this receptor family, and several ligands, there are multiple possibilities of hetero- and homodimers and signalling patterns (11). Overexpression of EGFR and HER2 has been associated with malignant transformation. HER3 has not been shown to transform cells to a malignant phenotype on its own, but if overexpressed in the presence of HER2 it may transform cells (10). Overexpression of EGFR is common in HNSCC (12,13). Increased levels of HER2 have been suggested in several studies of HNSCC (14), with possible prognostic value in oral SCC (15-18). HER3 has been shown to be overexpressed in HNSCC cell lines and associated with clinical malignant progression (18-20) and to be overexpressed also in other types of malignancy (21). However, HER3 can also be expressed in the normal surface squamous epithelium of the tongue, oropharynx and oesophagus (22). A number of human tissues and some human mammary carcinoma cell lines have HER4 transcripts (23) but the role of HER4 in cancer is less clear

EKBERG et al: SQUAMOUS CELL CARCINOMAS OF THE ORAL CAVITY

1178

Table I. Antibody characteristics and immunostaining procedures used for immunohistochemistry. ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Antibody Clone Source Staining and dilution Epitope retrieval Chromogen Counterstaining ––––––––––––––––– Manual Ventana ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– EGFR 31G7 Zymed Labs., South San 1:40 Enzymatic DAB Harris Francisco, CA, USA hematoxylin EGFR

31G7

Zymed Labs., South San Francisco, CA, USA

HER2

Polyclonal

Dako, Glostrup, Denmark

HER2

Polyclonal

Dako, Glostrup, Denmark

HER3

RTJ.1

BD Biosciences PharMingen, San Diego, CA, USA

HER4

HFR.1

1:20 1:300

1:100

1:100

Enzymatic

DAB

Harris hematoxylin

HIER in prediluted CCl buffer

DAB

Harris hematoxylin

HIER in 0.01 M citrate buffer pH 6.0

DAB

Harris hematoxylin

None

DAB

Harris hematoxylin

NeoMarkers, Fremont, 1:40 None DAB Harris CA, USA hematoxylin –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

(24). It has been suggested that coexpression patterns including at least pairs of the four receptors may be associated with malignant phenotypes (18,25,26). Immunohistochemical stainings of EGFR and HER2 have shown pronounced membraneous staining. In contrast, HER3 and HER4 staining has been mainly cytoplasmic (23,24). Furthermore, EGFR and HER2 have been reported to express at high levels in both tumours and metastases. Thus, it seems as EGFR and HER2 are potential targets for macromolecular therapy while this might not be the case with HER3 and HER4. For targeting EGFR several monoclonal antibodies have undergone or are under clinical testing, and for HER2 there is a humanised monoclonal antibody (Trastuzumab) used for a subset of mammary malignancies. Some earlier studies have shown low expression levels of the EGFR-family receptors in HNSCC. These studies have often included tumours with a wide variation regarding site, stage and treatment (27), and it is known that the properties of tumours from different sites may differ significantly (28). In our study we first compared the expression of the EGFR-family receptors in HNSCC of the oral cavity and base of tongue with their corresponding metastases and normal epithelium. This was done to give further information on which of these receptors that are of interest for diagnostic and/or therapeutic procedures. Attempts were also made to determine the cellular localization of HER3, i.e. membraneous or cytoplasmic. Materials and methods Tissue specimens. Patients with HNSCC of the oral cavity or base of tongue and metastatic disease were included. They presented at the Department of Otorhinolaryngology, Akademiska Sjukhuset, Uppsala, Sweden, between 1996

and 2002. Carcinomas of the tonsils and nasopharynx were excluded, as were patients without viable cancer cells in the tissue specimen. Paraffin sections from both the primary tumour and the corresponding lymph node metastases were required for inclusion. This was the case for 19 patients. All tissues were fixed in 10% buffered-formalin and paraffinembedded according to standard procedures at Akademiska Sjukhuset, Uppsala, Sweden. Immunohistochemistry. Paraffin sections of 4 µm thickness were placed on Superfrost/Plus slides (Mentzel, Germany), deparaffinized in xylene and rehydrated in graded alcohol concentrations. Antibody characteristics and immunostaining procedures are specified in Table I. Endogenous peroxidase was blocked in 0.3% H2O2 in phosphate-buffered saline (PBS) for 20 min. To reduce non-specific binding of the primary antibody, sections were pre-incubated in 0.5% BSA-c (Aurion, Wageningen, The Netherlands) in PBS. Manual immunostaining was performed by incubation of primary antibodies for 16 h at 4˚C, followed by incubation for 45 min at 20˚C in EnVision (Dako). Developing was done using diaminobenzidine (DAB) (Sigma, St. Louis, MO, USA) for 6 min. Counterstaining was done in Harris hematoxylin (Sigma). Finally, sections were dehydrated using graded alcohols, cleared in xylene and mounted in organic mounting medium (Pertex, Histolab, Gothenburg, Sweden). Immunostainings in an automated stainer (Ventana Medical Systems, Tucson, USA) were performed according to recommendations of the manufacturer using a standard DAB detection-kit. For manual immunohistochemistry, heat induced epitope retrieval (HIER) was done by boiling the slides for 7 min at full pressure (16 psi) in a decloacing chamber (Biocare

INTERNATIONAL JOURNAL OF ONCOLOGY 26: 1177-1185, 2005

1179

in an atmosphere containing humified air with 5% CO2. Cells were trypsinized and grown in dishes two days prior to the experiments. Antibodies for cell culture experiments. Primary monoclonal antibodies commercially available from NeoMarkers and Sigma were used for experiments on cell lines SCC-9 and SCC-25. The antibodies from NeoMarkers were c-erbB-3/ HER3 Ab 5, c-erbB-3/HER3 Ab 8, c-erbB-4/HER4 Ab 1 and EGFR Ab 2. The antibody from Sigma was anti-c-erbB-4 (HER4) Ab 4-36. All epitopes targeted were within the extracellular domains. As secondary antibody rabbit anti-mouse immunoglobulin (code no. Z0412) from Dako was used.

Figure 1. Examples of immunostaining in tumours. (A), EGFR; extent 3, intensity 3. (B), HER2; extent 3, intensity 1. (C), HER3; extent 3, intensity 3. (D), HER4; extent 3, intensity 2.

Medical, Walnut Creek, CA, USA). Enzymatic retrieval was done in 0.05% protease type VIII (Sigma, St. Louis, MO, USA) in PBS for 10 min at room temperature. In the automated immunohistochemistry in Ventana Benchmark, both heat induced antigen retrieval and proteolytic treatment were done using prediluted solution from the manufacturer. Evaluation of immunohistochemistry. Evaluation of the immunostainings was performed independently by two of the authors (T.E., J.C.). The extent and intensity were semiquantitatively judged as proportion of tumour cells stained (0, negative; 1, less than one third; 2, between one and two thirds; and 3, more than two thirds) and intensity of immunostaining (0, negative; 1, weak; 2, moderate; 3, intense). If disagreements occurred, evaluation was repeated in a multiheaded microscope until agreement was achieved. Using the above criteria it was evaluated if the tumour expressed the receptors at an adequate level and had a distribution that might allow for successful tumour targeting. Examples of staining classifications are given in Fig. 1. Besides the internal negative controls (inflammatory cells and fibroblasts), parallel sections, where the primary antibody was replaced for PBS, served as negative controls. Breast cancer tissue sections (HER2, HER3 and HER4) and tissue sections from placenta (EGFR), with known expression of the respective receptor, were used as positive controls. In addition, for HER2 a commercially available (Dako) control slide, containing cultured cells (SK-BR-3, MDA 175 and MDA 231) with unique expression levels of HER-2, was used as positive control. When comparing the receptor expression in apparently normal epithelium to tumour, an overall judgement of extent and intensity was used. Cell lines. The HNSCC cell lines SCC-9 and SCC-25 (obtained from American Type Culture Collection) have cancer of the tongue as their origin. They were cultured in a 1:1 mixture of Ham's F12 and Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% foetal calf serum, 0.4 mM hydrocortisone, 2 ml L-glutamine and antibiotics (100 IU penicillin and 100 µg/ml streptomycin). Cells were incubated in 37˚C

Labelling of the secondary antibody for cell culture experiments. The 125I labelling of the secondary rabbit antimouse antibody was performed according to standard procedure with chloramine-T (CAT). Ten MBq 125I were diluted in 120 µl PBS and then mixed with 10 µl antibody (4 mg/ml PBS) and 20 µl CAT (4 mg/ml water). The reaction proceeded on ice for 5 min, after which the process was stopped by adding 40 µl Na2SO4 (4 mg/ml water). The sample was separated on a NAP-5 column and the fractions containing the 125I labelled antibody were pooled. Cell culture experiments. Cell lines SCC-9 and SCC-25 were grown in separate 3-cm dishes up to approximately 50,000 cells in each dish. The dishes were incubated in triplets for each antibody to be tested, and then washed once in cold serum-free medium (SF). To minimise internalisation of receptor-antibody complexes the incubations were made on ice. The cell lines had previously shown to function well for several hours under such conditions. Each cell line was divided in three groups and treated as follows: a) the purpose was first to evaluate the binding of the primary antibody to receptor in the cellular membrane. 0.125 µg primary antibody diluted in medium was added and the dishes were incubated for 2 h on ice. The dishes were then washed twice in cold SF, and normal medium was added. Once again the dishes were incubated for 2 h on ice; b) the purpose was to verify the specificity of the secondary antibody. This was done by preventing the radiolabelled secondary antibody from binding to the primary antibody by adding an excess of non-labelled secondary antibody to the cells. 0.125 µg primary antibody diluted in medium was added and the dishes were incubated for 2 h on ice. The dishes were then washed twice in cold SF, and 1.25 µg of secondary rabbit anti-mouse antibody was added. The dishes were then incubated for another 2 h on ice; c) the purpose was finally to evaluate a possible non-specific binding of the secondary antibody. Medium was added and the dishes were incubated for a total of 4 h with the secondary antibody on ice (all the time without primary antibody). The dishes were then (in all 3 cases) washed in SF, and 1.25 µg 125I labelled secondary rabbit anti-mouse antibody was added, and the dishes were once more incubated for 2 h on ice. Finally they were washed six times in SF, trypsinized, and resuspended in media. The cells were counted and radiation measured in a gamma counter. The gamma counting was made on the trypsin solution containing the cells. Since the cell lines had previously shown to express EGFR, which is

1180

EKBERG et al: SQUAMOUS CELL CARCINOMAS OF THE ORAL CAVITY

Table II. Characteristics of included patients. ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Pts. Tumour Tumour Metastasis TNM at Staging at Histological Pts. Sex no. site irradiated irradiated diagnosis diagnosis differentiation age ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 1 Gingiva No No T4N1M0 IVA Moderate 82 M 2 Bucca No No T4N2M0 IVA Moderate 79 F 3 Tongue Yes Yes T2N0M0 II Moderate 72 F 4 Bucca No No T4N1M0 IVA Poor 75 F 5 Gingiva No No T4N1M0 IVA Poor 72 F 6 Soft palate No Yes T4N2M0 IVA Poor 68 M 7 Base of tongue Yes Yes T4N0M0 IVA Poor-moderate 63 M 8 Gingiva Yes Yes T1N1M0 III High 63 F 9 Tongue Yes Yes T4N2M0 IVA High 58 M 10 Gingiva No Yes T1N2M0 IVA Poor 53 M 11 Floor of mouth No Yes T4N0M0 IVA Moderate 53 M 12 Base of tongue Yes Yes T2N1M0 III Moderate 49 F 13 Tongue No No pT1N1M0 III Moderate 75 F 14 Bucca Yes Yes T3N0M0 III Moderate 68 F 15 Tongue No No T2N2M0 IVA Moderate 57 M 16 Floor of mouth No Yes T2N1M0 III High 59 F 17 Gingiva No No T4N0M0 IVA Poor-moderate 55 F 18 Tongue No No T3N0M0 III Moderate 80 F 19 Floor of mouth No No T4N1M0 IVA Poor-moderate 62 M –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

well known to be located in the membrane, this served as a positive control. When receptors are present in the membrane, the detected radiation should be considerably higher in the cells treated according to a) as compared to b). The cells treated according to c) served as negative control. Immunohistochemical calibration of the cell lines. The SCC-9 and SCC-25 cells were trypsinized and grown on Superfrost/ Plus slides (Mentzel, Germany) in medium for 48 h, then washed in SF. They were fixated in methanol at -20˚C for 20 min, in acetone at 4˚C for 1 min, then washed in PBS. The antibodies used were EGFR Ab 2, c-erbB-3/HER3 Ab 5, c-erbB-3/HER3 Ab 8 and c-erbB-4/HER4 Ab 1, also used in the cell culture experiments. Antibodies were diluted in PBS to 7.5 µg/ml. Controls using Ab 4 (diluted in PBS to 5 µg/ml), which has the intracellular domain of HER4 as epitope, were also made, since it had proven to function well in IHC of the tissue samples. Immunostaining was done as described for paraffin sections. Evaluation of the immunostaining was performed independently by three of the authors (T.E., J.C., M.N.). The intensity of the staining was semi-quantitatively judged as: negative, 0; weak, 1; weak-moderate, 2; moderate, 3; intense, 4; or very intense, 5. The intensity of immunostaining for EGFR in each cell line was set to 5. If disagreements occurred, evaluation was repeated in a multiheaded microscope until agreement was achieved.

normal epithelium from 19 patients were analysed (8 male and 11 female). The carcinomas were located in tongue (n=5), gingiva (n=5), bucca (n=3), floor of mouth (n=3), base of tongue (n=2) and soft palate (n=1). The tumours and corresponding metastases were classified according to histological grade of differentiation, and staged according to the TNMsystem (UICC, TNM Classification of Malignant Tumours, 6th edition). It was noted if the tumours or metastases had received radiation therapy prior to biopsy or operation. The patient data are specified in Table II. Immunohistochemistry of tumours and metastases. The staining patterns were membraneous for EGFR and HER2, with or without concomitant cytoplasmic staining, and the staining patterns for HER3 and HER4 were mainly cytoplasmic. A summary of the evaluations of extent and intensity of the receptor immunostainings are shown in Table III. Some examples of immunostainings are shown in Fig. 1.

Results

EGFR. All primary tumours expressed EGFR and only one metastasis was classified negative (Table III). Using a cut-off score set to higher than 1 for EGFR, there was a good agreement between primary tumours and the corresponding metastases in all except 3 cases. This generally good agreement is seen in the scoring patterns shown in Fig. 2. In the subgroup with HNSCC of the tongue there was a tendency for lower scores in the metastases compared to the corresponding primary tumours.

Patient characteristics. Squamous cell carcinomas of the oral cavity and base of tongue, lymph node metastases and

HER2. HER2 had generally lower expression as compared to EGFR. Five of the 19 cases had no HER2 immuno-

INTERNATIONAL JOURNAL OF ONCOLOGY 26: 1177-1185, 2005

1181

Table III. Extent and intensity of immunostaining in primary tumours and corresponding metastases for the four receptors of the EGFR-family. ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Pts. no. EGFR EGFR HER2 HER2 HER3 HER3 HER4 HER4 primary metastasis primary metastasis primary metastasis primary metastasis tumour tumour tumour tumour ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 1 Extent 3 3 0 0 3 2 3 2 Intensity 2 2 0 0 3 2 2 2 2 Extent Intensity

3 3

3 3

0 0

0 0

3 3

1 1

3 2

1 1

3 Extent Intensity

3 3

3 3

1 1

1 1

2 3

3 2

1 1

3 2

4 Extent Intensity

2 2

2 2

1 1

0 0

0 0

0 0

2 1

0 0

5 Extent Intensity

3 3

3 2

0 0

0 0

2 2

1 1

2 2

0 0

6 Extent Intensity

3 3

3 3

1 1

1 1

1 1

1 2

3 2

0 0

7 Extent Intensity

2 3

1 1

1 1

1 1

1 1

2 1

1 1

2 1

8 Extent Intensity

3 2

2 3

1 1

1 2

3 3

2 3

3 2

2 2

9 Extent Intensity

1 1

0 0

0 0

0 0

2 1

0 0

2 1

0 0

10 Extent Intensity

2 3

3 3

2 1

2 1

0 0

1 1

1 1

1 1

11 Extent Intensity

3 3

3 2

2 1

0 0

1 2

2 2

1 1

2 2

12 Extent Intensity

2 3

3 3

2 1

2 1

2 2

3 2

3 2

3 2

13 Extent Intensity

3 3

2 2

1 1

0 0

1 2

2 2

1 1

3 1

14 Extent Intensity

3 3

3 3

1 1

1 1

2 2

3 3

3 2

2 1

15 Extent Intensity

2 3

1 1

3 2

2 1

2 3

1 2

3 2

2 1

16 Extent Intensity

3 2

3 3

2 1

0 0

1 2

2 2

2 1

0 0

17 Extent Intensity

3 2

2 1

3 1

2 2

2 1

0 0

3 2

1 1

18 Extent Intensity

3 3

2 2

2 2

2 1

2 2

2 2

2 2

2 1

19 Extent 3 3 0 0 3 2 3 3 Intensity 3 3 0 0 3 1 2 2 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

1182

EKBERG et al: SQUAMOUS CELL CARCINOMAS OF THE ORAL CAVITY

Figure 2. Immunostaining for EGFR. Distribution of scores for extent and intensity in (A) primary tumours and (B) corresponding metastases.

Figure 4. Immunostaining for HER3. Distribution of scores for extent and intensity in (A) primary tumours and (B) corresponding metastases.

Figure 3. Immunostaining for HER2. Distribution of scores for extent and intensity in (A) primary tumours and (B) corresponding metastases.

Figure 5. Immunostaining for HER4. Distribution of scores for extent and intensity in (A) primary tumours and (B) corresponding metastases.

staining neither in primary tumours nor in the corresponding metastases. Furthermore, 4 other metastases were negative. However, no cases with HER2 expression in the metastases had a negative primary tumour, and only 1 metastasis had a higher total score (extent and intensity summarised) compared to the corresponding primary tumour. None of the 3 cases with HNSCC of the floor of mouth had any expression of HER2 in the metastases, and in the tongue cancers 4/5 had either no expression or lower scores in the metastases compared to the corresponding primary tumours. The general impression from the results in Table III and Fig. 3 is that the HER2 expression in the metastases was somewhat lower compared to the corresponding primary tumours.

HER3. Only 1/19 cases was negative for HER3 in both the primary tumour and the corresponding metastasis. In addition, 1 primary tumour and 2 metastases were negative, while there was some expression in the paired metastases or tumours, respectively. However, the differences between primary tumours and corresponding metastases were small and the general impression is that there is a similarity between them (Table III, Fig. 4). HER4. All primary tumours had some expression of HER4, but 5/19 metastases were negative. The general impression from Table III and Fig. 5 is that the HER4 expression was somewhat lower in the metastases than in the primary tumours.

INTERNATIONAL JOURNAL OF ONCOLOGY 26: 1177-1185, 2005

1183

Figure 7. Counts per minute (CPM) per 100,000 cells for antibodies used in experiments on EGFR and HER3 expressing cell lines. Error bars represent maximal errors.

Figure 6. Examples of immunostaining in normal epithelium for (A) EGFR, (B) HER2, (C) HER3 and (D) HER4.

Table IV. Expression of receptors in the tumours compared to the expression in corresponding normal epithelium. ––––––––––––––––––––––––––––––––––––––––––––––––– Pts. no. EGFR HER2 HER3 HER4 ––––––––––––––––––––––––––––––––––––––––––––––––– 1 + = + 2 / / / / 3 = + = 4 / / / / 5 + + + 6 + + + 7 / / / / 8 + = / 9 = = = = 10 + + + 11 + + = 12 / / / / 13 + = 14 + = + = 15 + + + = 16 + = = 17 = + + 18 + + + + 19 + = + + ––––––––––––––––––––––––––––––––––––––––––––––––– +, higher expression in the tumour; -, lower expression in tumour; =, similar expression in tumour and normal epithelium; /, no normal epithelium in biopsy.

–––––––––––––––––––––––––––––––––––––––––––––––––

Influence of previous therapy. There seemed to be no clear differences between the tissues irradiated prior to biopsy or operation and those, which were not irradiated. However, the number of cases was too low to allow analysis of statistical significance.

Table V. Calibration immunostaining of cell lines SCC-9 and SCC-25.a ––––––––––––––––––––––––––––––––––––––––––––––––– Ab 225 Ab 8 Ab 5 Ab 1 Ab 4 EGFR HER3 HER3 HER4 HER4 ––––––––––––––––––––––––––––––––––––––––––––––––– SCC-9 5 2 2 0 0 SCC-25 5 2 2 0 0 ––––––––––––––––––––––––––––––––––––––––––––––––– aScale:

0-5 (see text).

–––––––––––––––––––––––––––––––––––––––––––––––––

Normal epithelium. The immunostaining in apparently normal adjacent epithelium was for EGFR and HER2 generally located near or at the basal membrane and decreased as the epithelium differentiated. For HER3 and HER4 more pronounced staining was seen in the middle and in the superficial layers, and there was no staining near and at the basal membrane (Fig. 6). The EGFR staining in the primary tumours and the corresponding metastases was in most cases stronger than in the normal epithelium. Actually, EGFR did not show weaker staining in tumours compared to normal epithelium in any case, while the results varied largely for the other receptors (Table IV). Cell culture experiments. The 125I labelled secondary antibody reacted specifically with the primary antibodies and had no non-specific binding directly to the cells. The radiation detected when using Ab 2 (EGFR) was high, as expected (Fig. 7). The number of counts per minute (CPM) per 105 cells was about two orders of magnitude higher than the background. However, in the experiments with the antibodies directed to extracellular domains of HER3 (Ab 5, Ab 8) and HER4 (Ab 1, Ab 4-36) the amount of radiation detected was neglectable, i.e. in the same range as the radiation detected in the negative controls. Nevertheless, the cells contained HER3 as shown in the immunohistochemical stainings (Table V). Thus, it seems from these tests that there were no, or only very small, amounts of HER3 (and HER4) receptors in the cellular membranes. The cytoplasmic appearance of HER3 (and HER4) is also in accordance with the results obtained in the immunohistochemical stainings of tissue sections.

1184

EKBERG et al: SQUAMOUS CELL CARCINOMAS OF THE ORAL CAVITY

Discussion We analysed in detail a small group of patients with aggressive disease in a well-defined area of the head and neck. This group is the most difficult to give curative treatment, and would therefore benefit most from a possible complementary method of diagnosis and treatment in terms of targeting the tumour cells. The receptors in the EGFR family were analysed for that purpose. When determining the expression or overexpression of receptors with IHC, it is of course important to analyse also the expression of the same receptors in near normal epithelium from the same patient. Earlier studies have shown similar patterns of expression in normal oral mucosa as in the epithelium located near the carcinoma (29), which justify that it is enough to analyse the normal epithelium near the tumour cells in our tissue specimens. The expression of EGFR was high in, and concordant between, primary tumours and their corresponding metastases which is in accordance with recent reports (13). We had no case where the expression of EGFR was more pronounced in normal epithelium compared to the tumours. Considering scores 2 and 3 versus scores 0 and 1 (both extent and intensity) there was a good agreement between primary tumours and the corresponding metastases in all except 3 cases. Thus, EGFR seems generally stable when comparing primary tumours with the corresponding metastases. The HER2 receptor is overexpressed and subject to targeting therapies in a subset of mammary malignancies (30). In HNSCC the results from earlier studies on HER2 expression have been diverse (14,17). Furthermore, there are reports on membrane staining (31) and on only cytoplasmic staining (32). The expression of HER2 in our material was not as strong as the expression of EGFR, and in some tumours the expression of HER2 was equal to or lower than the normal epithelium (Table V). Note that we did not find any expression of HER2 in the metastases from HNSCC of the floor of mouth. The HER3 receptor is less well documented in HNSCC with metastatic disease, but may have prognostic significance in oral cancer (20). There are reports on HER3 expression in HNSCC and coexpression of HER3 with other family members (18,25,29). Our results vary somewhat but there seems to be a general similarity between the primary tumours and the corresponding metastases. However, the HER3 expression in normal epithelium was in some cases as high as in the tumours. The role of HER4 in tumour development is not clear. Like HER3 it might have prognostic significance in combination with other receptors, but when investigated on its own it does not seem to be overexpressed in HNSCC as previously indicated (33). The results of our immunostainings neither supported nor contradicted overexpression. Many cases had immunostaining to a high degree, but others had expression levels equal to or lower than normal epithelium. Furthermore, tissues such as kidney, skin and adrenal cortex seem to have high levels of HER4 (33) and adding that information indicates that HER4 might not be suitable for macromolecular targeting therapies. There are some earlier studies of expression of the receptors in the EGFR-family for metastatic HNSCC of other sites. In

laryngeal cancer no positivity for EGFR or HER2 for patients with or without metastases was found in one study (34), while another study found EGFR expression in relapsed laryngeal node metastases (35). As mentioned earlier, different sites may have different expressions, possibly with higher expression of EGFR (28) and HER2 (17) in oral cancer. There is one published report claiming that HNSCC from the larynx, pharynx and oral cavity mainly had negative stainings for both EGFR and HER2 in both primary tumours and metastases (36). As indicated by others (20,22,33), both primary tumours and corresponding metastases showed a cytoplasmic immunostaining pattern for HER3 and HER4, rather than the expected membranous distribution. This does not seem to be an artefact caused by histo-processing, e.g. formalin fixation, since frozen sections show the same pattern (22,33). In the present study preparatory testings indicated that the age of paraffin sections only marginally affected staining intensity (data not shown). When comparing formalin-fixed paraffin-embedded sections to frozen sections, the latter generally showed weaker staining intensity but also a poor morphology, which partly hampered the evaluation of intracellular distribution of the receptors. Included in these tests were several tissue types and fixatives representing different fixation mechanisms. Despite the cytoplasmic distribution of the immunostaining, it remains to be determined if small subsets of the receptors are localized in the cellular membrane. The primary antibodies used in our experiments on the cell lines all recognize the extracellular domain of the respective receptor. The immunostainings showed that the cells SCC-9 and SCC-25 expressed EGFR and HER3 but not HER4. A system where we did not alter the binding properties of the primary antibodies in the process of radioactive labelling was then created. We received high radioactivity uptake for the anti-EGFR antibodies, but for the anti-HER3 antibodies, the uptake did not differ from the uptake using anti-HER4 antibodies or the negative controls. The results indicate that the HER3 receptor is not present at a measurable level in the cellular membrane of the living cells. Since HER4 was not detectable with IHC in the cell lines we cannot draw conclusions of its distribution from the in vitro experiments. Targeted radionuclide diagnostics, maybe in combination with targeted therapy, might be possible if a tumour and its possible metastases have a strong receptor expression. Our findings with an EGFR expression that is high, membrane associated and similar in primary tumours and corresponding lymph node metastases indicate that EGFR is a candidate for such targeting. From the diagnostic point of view it is important to gain as much information as possible on the tumour margins and the presence of possible metastases. From the therapeutical point of view, even if external beam radiation has been given, or surgery including a radical lymph node dissection of the neck has been performed, there is always a risk of ‘minimal residual disease’ in the neck or residual tumour at the primary site, and then there is a need for targeted therapy as a follow-up. Acknowledgements This study was supported by the Swedish Cancer Society.

INTERNATIONAL JOURNAL OF ONCOLOGY 26: 1177-1185, 2005

References 1. Byers RM, Wolf PF and Ballantyne AJ: Rationale for elective modified neck dissection. Head Neck Surg 10: 160-167, 1988. 2. Carlsson J, Forssell Aronsson E, Hietala SO, Stigbrand T and Tennvall J: Tumour therapy with radionuclides: assessment of progress and problems. Radiother Oncol 66: 107-117, 2003. 3. Baselga J: Targeting the epidermal growth factor receptor with tyrosine kinase inhibitors: small molecules, big hopes. J Clin Oncol 20: 2217-2219, 2002. 4. Herbst RS and Shin DM: Monoclonal antibodies to target epidermal growth factor receptor-positive tumors: a new paradigm for cancer therapy. Cancer 94: 1593-1611, 2002. 5. Mendelsohn J and Baselga J: The EGF receptor family as targets for cancer therapy. Oncogene 19: 6550-6565, 2002. 6. Baselga J: Targeting the epidermal growth factor receptor: a clinical reality. J Clin Oncol 19 (Suppl 18): 41S-44S, 2001. 7. Arteaga CL, Moulder SL and Yakes FM: HER (erbB) tyrosine kinase inhibitors in the treatment of breast cancer. Semin Oncol 29 (Suppl 11): 4-10, 2002. 8. Yarden Y and Sliwkowski MX: Untangling the ErbB signalling network. Nat Rev Mol Cell Biol 2: 127-137, 2001. 9. Peles E, Ben-Levy R, Tzahar E, Liu N, Wen D and Yarden Y: Cell-type specific interaction of Neu differentiation factor (NDF/ heregulin) with Neu/HER-2 suggests complex ligand-receptor relationships. EMBO J 12: 961-971, 1993. 10. Alimandi M, Romano A, Curia MC, et al: Cooperative signaling of ErbB3 and ErbB2 in neoplastic transformation and human mammary carcinomas. Oncogene 10: 1813-1821, 1995. 11. Olayioye MA, Neve RM, Lane HA and Hynes NE: The ErbB signaling network: receptor heterodimerization in development and cancer. EMBO J 19: 3159-3167, 2000. 12. Rikimaru K, Tadokoro K, Yamamoto T, Enomoto S and Tsuchida N: Gene amplification and overexpression of epidermal growth factor receptor in squamous cell carcinoma of the head and neck. Head Neck 14: 8-13, 1992. 13. Santini J, Formento JL, Francoual M, et al: Characterization, quantification and potential clinical value of the epidermal growth factor receptor in head and neck squamous cell carcinomas. Head Neck 13: 132-139, 1991. 14. Craven JM, Pavelic ZP, Stambrook PJ, et al: Expression of cerbB-2 gene in human head and neck carcinoma. Anticancer Res 12: 2273-2276, 1992. 15. Werkmeister R, Brandt B and Joos U: Clinical relevance of erbB-1 and -2 oncogenes in oral carcinomas. Oral Oncol 36: 100-105, 2000. 16. Werkmeister R, Brandt B and Joos U: The erbB oncogenes as prognostic markers in oral squamous cell carcinomas. Am J Surg 172: 681-683, 1996. 17. Xia W, Lau YK, Zhang HZ, et al: Strong correlation between c-erbB-2 overexpression and overall survival of patients with oral squamous cell carcinoma. Clin Cancer Res 3: 3-9, 1997. 18. Xia W, Lau YK, Zhang HZ, et al: Combination of EGFR, HER-2/ neu and HER-3 is a stronger predictor for the outcome of oral squamous cell carcinoma than any individual family members. Clin Cancer Res 5: 4164-4174, 1999. 19. O-charoenrat P, Rhys-Evans PH, Archer DJ and Eccles SA: CerbB receptors in squamous cell carcinomas of the head and neck: clinical significance and correlation with matrix metalloproteinases and vascular endothelial growth factors. Oral Oncol 38: 73-80, 2002.

1185

20. Shintani S, Funayama T, Yoshihama Y, Alcalde RE and Matsumura T: Prognostic significance of ERBB3 overexpression in oral squamous cell carcinoma. Cancer Lett 95: 79-83, 1995. 21. Gullick WJ: The c-erbB3/HER3 receptor in human cancer. Cancer Surv 27: 339-349, 1996. 22. Prigent SA, Lemoine NR, Hughes CM, Plowman GD, Selden C and Gullick WJ: Expression of the c-erbB-3 protein in normal human adult and fetal tissues. Oncogene 7: 1273-1278, 1992. 23. Plowman GD, Culouscou JM, Whitney GS, et al: Ligandspecific activation of HER4/p180erbB4, a fourth member of the epidermal growth factor receptor family. Proc Natl Acad Sci USA 90: 1746-1750, 1993. 24. Srinivasan R, Gillett CE, Barnes DM and Gullick WJ: Nuclear expression of the c-erbB-4/HER-4 growth factor receptor in invasive breast cancers. Cancer Res 60: 1483-1487, 2000. 25. Bei R, Pompa G, Vitolo D, et al: Co-localization of multiple ErbB receptors in stratified epithelium of oral squamous cell carcinoma. J Pathol 195: 343-348, 2001. 26. Krahn G, Leiter U, Kaskel P, et al: Coexpression patterns of EGFR, HER2, HER3 and HER4 in non-melanoma skin cancer. Eur J Cancer 37: 251-259, 2001. 27. O-charoenrat P, Rhys-Evans P, Modjtahedi H and Eccles S: The role of c-erbB receptors and ligands in head and neck squamous cell carcinoma. Oral Oncol 38: 627, 2002. 28. Takes RP, Baatenburg de Jong RJ, Schuuring E, Litvinov SV, Hermans J and van Krieken JH: Differences in expression of oncogenes and tumor suppressor genes in different sites of head and neck squamous cell. Anticancer Res 18: 4793-4800, 1998. 29. Ibrahim SO, Vasstrand EN, Liavaag PG, Johannessen AC and Lillehaug JR: Expression of c-erbB proto-oncogene family members in squamous cell carcinoma of the head and neck. Anticancer Res 17: 4539-4546, 1997. 30. Yarden Y: Biology of HER2 and its importance in breast cancer. Oncology 61 (Suppl 2): 1-13, 2001. 31. Kearsley JH, Leonard JH, Walsh MD and Wright GR: A comparison of epidermal growth factor receptor (EGFR) and c-erbB-2 oncogene expression in head and neck squamous cell carcinomas. Pathology 23: 189-194, 1991. 32. Field JK, Spandidos DA, Yiagnisis M, Gosney JR, Papadimitriou K and Stell PM: C-erbB-2 expression in squamous cell carcinoma of the head and neck. Anticancer Res 12: 613-619, 1992. 33. Srinivasan R, Poulsom R, Hurst HC and Gullick WJ: Expression of the c-erbB-4/HER4 protein and mRNA in normal human fetal and adult tissues and in a survey of nine solid tumour types. J Pathol 185: 236-245, 1998. 34. Takes RP, Baatenburg de Jong RJ, Schuuring E, et al: Markers for assessment of nodal metastasis in laryngeal carcinoma. Arch Otolaryngol Head Neck Surg 123: 412-419, 1997. 35. Almadori G, Cadoni G, Galli J, et al: Epidermal growth factor receptor expression in primary laryngeal cancer: an independent prognostic factor of neck node relapse. Int J Cancer 84: 188-191, 1999. 36. Takes RP, Baatenburg de Jong RJ, Wijffels K, et al: Expression of genetic markers in lymph node metastases compared with their primary tumours in head and neck cancer. J Pathol 194: 298-302, 2001.