Br. J. Cancer (1987), 55, 5l3~5l6
`
`The Macmillan Press Ltd, 1987
`
`Epidermal growth factor receptors in non-small cell lung cancer
`
`D. Veale‘, T. Ashcroftz, C. Marshz, G..l. Gibson‘ & A.L. Harris3
`
`Departments of ‘Respiratory Medicine and 2Pathology, Freeman Hospital,‘ Cancer Research Unit and 3I)epartment of Clinical
`Oncology, Newcastle upon Tyne, UK.
`
`Summary The epidermal growth factor receptor is homologous to the oncogene erb~,.’i and is the receptor for
`a class of tumour growth factors (TGF-ct). The clinical correlations with its expression were studied in 77 non-
`small cell lung cancers (NSCLC). They were stained for epidermal growth factor receptor (EGFr) by means of
`an indirect
`immunoperoxidase technique using a monoclonal antibody against
`the receptor. Normal
`lung
`tissue and normal bronchus were stained for comparison. Cancer tissue showed significantly increased staining
`compared to normal lung (P<0.05).
`Staining for EGFr in 40 squamous carcinomas was significantly stronger than in 37 specimens of other
`types of NSCLC (P <0.05), and staining in stage three NSCLC was stronger than in stage 1 and 2 (P<0.05).
`These results suggest that the presence of a high intensity of staining for EGF receptor is associated with
`spread of human non~small cell lung cancer and this receptor may be a suitable target for therapy.
`
`pre-
`comprises
`(NSCLC)
`lung cancer
`cell
`Non-small
`dominantly squamous, adenocarcinomas and large cell un-
`differentiated tumours. Only 5% of such tumours are curable
`by surgery (Crofton & Douglas, 1981), and up to 60% are
`inoperable at
`the time of diagnosis (Crofton & Douglas,
`l98l). Chemotherapy is
`relatively ineffective, with a 17%
`response rate and no long-term survival
`(Simes,
`1985).
`Radiotherapy also is rarely curative (Kjaer, 1982). Survival in
`NSCLC is related to the degree of spread of the tumour at
`diagnosis,
`staged
`by
`the
`tumour
`nodal
`involvement
`metastasis (TNM) system (Mountain et al._. 1974).
`A characteristic of malignant cells in culture is a decreased
`requirement
`for
`serum or
`supplemental growth factors
`(Holley, 1975). This suggests that malignant cells may secrete
`growth factors into their medium. Sporn and Todaro (1980)
`have suggested the term autocrine growth regulation to
`describe the situation where a cell secretes a factor, possesses
`specific receptors for the factor and responds to that factor.
`Epidermal growth factor (EGF) is a growth promoting
`agent found in normal human plasma and tissues which acts
`by binding to receptors for the hormone on cell surfaces.
`When injected into newborn mice, EGF elicits the prolifera-
`tion and differentiation of the epidermis (Cohen, 1983). In
`tissue culture, EGF induces the proliferation of a wide
`variety of cell types, including keratinocytes and transformed
`epithelial cells (Carpenter & Cohen, 1979).
`EGF receptor (EGFr) has three parts, an external domain,
`a transmembrane portion and an internal domain. On
`binding with its receptor, EGF leads to activation of a
`tyrosine lcinase on the internal domain of
`the receptor
`(Carpenter & Cohen, 1979), a property shared with the
`protein product of
`some oncogenes.
`Indeed the erb-B
`oncogene product
`is a truncated EGFr (Downward er al.,
`1984). The receptor is also internalised after stimulation.
`Peptide growth factors produced by transformed cells
`include transforming growth factor
`or
`(TGF-at), which is
`structurally related to EGF and binds to the EGF receptor
`(EGFr) (Marquardt
`er. al., 1983) and transforming growth
`factor [3 (TGFB) which binds to different receptors but
`is
`synergistic with TGFoz (Marquardt et al., 1983).
`In breast cancer, it has been shown that the level of EGFr
`associated with metastatic
`potential, with
`poorly
`is
`differentiated tumours and with a poor prognostic subgroup
`(Sainsbury et al., l985a). The level of EGFr is also associated
`with the degree of
`invasion and poor differentiation of
`bladder cancer (Neal er al., 1985). EGFr have been found on
`NSCLC cell lines grown in cell culture (Sherwin et al., 1981).
`
`Correspondence: A.L. Harris.
`Received 25 July 1986; and in revised form, 5 January 1987.
`
`We have therefore examined lung tumours for the presence
`of EGFr and investigated the relationship between receptor
`density, histological subtype and tumour staging. We have
`examined the relationship of the presence of EGFr on these
`tumous with prognostic variables such as
`the degree of
`differentiation of the tumours.
`
`We used a monoclonal antibody to the receptor in an
`immunoperoxidase technique (DeLellis et al., 1979).
`
`Patients and methods
`
`We studied tumour material from 77 patients (53 male and
`24 female) undergoing surgery or diagnostic bronchoscopy.
`The mean age of the patients was 60.4 years. Tumours were
`staged by the tumour nodal involvement metastasis (TNM)
`system on examination of the resected material. Tumours
`with distant metastases or mediastinal nodes are graded
`stage three, as also are tumours extending into the parietal
`pleura or
`involving a main bronchus
`less
`than two
`centimetres from the carina, or any tumour associated with
`atalectasis or obstructive pneumonitis of an entire lung or
`pleural effusion,
`.
`Forty patients had squamous tumours, of whom 17 had
`stage three tumours. Twenty patients had adenocarcinomas
`(six stage three), 13 had large cell undifferentiated carcinomas
`(live stage three) and four tumours had features of both
`adeno and squamous carcinomas,
`two of which were stage
`three.
`
`Lung tumour samples from operation were frozen in liquid
`nitrogen in 71 of the patients within 30 minutes of resection.
`Blocks of tissue were taken from different areas of the
`
`tumour and 13 dual samples tested for the presence of BGFr.
`Six
`tumour
`specimens were
`similarly
`obtained
`at
`bronchoscopy.
`The presence of EGFr was also assessed in samples of
`uninvolved lung in 17 cases and in one sample from a
`patient without malignancy. Samples of normal bronchus
`were tested for EGFr in 8 patients.
`The EGF receptor was identified by means of an indirect
`immunoperoxidase technique (DeLeHis et al., 1979) with a
`murine monoclonal antibody (EGFR1)
`(gift of Dr M.
`Waterlield). The antibody was raised from an epidermoid
`carcinoma cell line (A431) which expresses a high concentra-
`tion of EGF receptors (Waterlield er al., 1982).
`Five pm cryostat sections were cut and picked up on
`lysine-coated slides. After drying with a fan for 30 min, they
`were fixed in acid for 20 min and washed twice for 5min in
`saline buffered to pH 7.6 with tris-l-1C1. The sections were
`then covered with normal serum (diluted 1:4 with Ms-
`bulfered saline) as a blocking agent for 10min. The sections
`
`‘
`
`APOTEX EX. 1026-001
`
`

`
`514
`
`D. VEALE et al.
`
`were incubated at room temperature with primary antibody
`for 30 min. After
`two washes in tris—buffered saline,
`the
`blocking step was repeated. The sections were then incubated
`with
`peroxidase-conjugated
`rabbit
`anti-mouse . imrnuno«
`globulin (Dakopatts) for 30 min. After
`two further 5-min
`washes, peroxidase activity was developed by means of a
`solution of diaminobenzidine. The sections were washed in
`
`water, counterstained with haetnaoxylin, dehydrated and
`mounted. The positive control was human placenta, which
`contains large numbers of EGFr. For negative controls we
`omitted the primary antibody. Controls used in each run
`showed similar intensity of staining.
`The intensity of staining was assessed by two observers
`reading the sections independently without knowledge of the
`tumour stage or degree of differentiation. The sections were
`graded on a scale from 0 to + + + according to the intensity
`of staining of malignant cells relative to the positive control.
`The placental control stained strongly positively on all
`occasions
`and showed no staining when the primary
`antibody was omitted. The histological typing of the tumours
`was carried out on parafiin sections of the resected material.
`Statistical analysis was performed using the chi—square test
`on a two by four table of results.
`
`Results
`
`Normal lung and bronchus EGFr
`
`Results of staining for EGFr in 17 specimens of normal lung
`showed no staining in 5 specimens and mild staining (1+) in
`9. Only two specimens of normal
`lung tissue showed
`moderate staining graded + +, while one specimen had
`strong staining for EGFr. In normal bronchial epithelium
`there was strong positive staining for EGFr in a thin band in
`the basal layer (Figure l) and weaker staining in bronchial
`glands.
`
`
`
`in normal
`lmmunoperoxidase staining for EGFr
`Figure 1
`tissue showing strong positive staining in the basal
`bronchial
`layer. (x 240).
`
`Carcinoma EGFr ~ Relation to histological type and stage
`
`Staining in cancer specimens was significantly stronger in
`tumour than in normal lung tissue (x2=8.47, P<0.01). The
`results for staining for EGFr in tumours are snown in Table
`l, and illustrated in Figure 2. Staining in sections from
`different parts of the tumour showed a good correlation of
`intensity of EGFr staining in 10 cancers examined. Eight of
`10 tumours examined had identical grading from different
`areas. No tumour graded EGFr negative was then graded
`positive but one
`tumour which
`graded positive was
`subsequently negative.
`stained
`tumours
`squamous
`forty
`Twenty—seven
`of
`moderately strongly or strongly (+ + or + + +), while only
`nine of twenty adenocarcinomas stained similarly. Overall,
`the staining in squamous tumours was significantly stronger
`
`lrnrnunoperoxidase staining for EGFr in a squamous
`Figure 2
`lung
`cancer.
`(x 57).
`(a) primary
`antibody
`omitted;
`(b)
`monoclonal primary antibody.
`
`Table I EGFr in non-small cell lung cancer (normal lung
`and bronchus)
`
`Staining grade
`
`NSCLC
`
`Squamous
`Adeno
`Large cell
`Adenosquamous
`Lung
`Bronchus
`
`0
`
`5
`6
`4
`0
`5
`2
`
`1
`
`8
`5
`3
`l
`9
`S
`
`2
`
`3
`
`Total
`
`l3
`8
`5
`2
`2
`0
`
`l4
`l
`l
`l
`l
`O
`
`40
`20
`13
`4
`17
`7
`
`Table II EGFr staining re tumour stage
`
`Staining grade
`
`0
`
`14
`l
`
`I
`
`7
`10
`
`2
`
`17
`11
`
`3
`
`9
`8
`
`Total
`
`47
`30
`
`Stages 1 and 2
`Stage 3
`
`the
`than for other NSCLC (x2=8.88, P<0.05). When all
`tumours were grouped according to stage there was signifi-
`cantly stronger
`staining in the 30 stage three tumours
`compared to 47 stage one and stage two tumours (x2=9.87,
`P<0.05) (Table II). There was no significant difference in the
`number of stage three squamous tumours (17/40) compared
`to the number of stage three adeno and large cell carcinomas
`(13/37) (12 = 0.44) (Table III).
`When all the tumours were divided according to degree of
`differentiation (well and moderately differentiated versus
`
`APOTEX EX. 1026-002
`
`

`
`EGF RECEPTORS IN NON-SMALL CELL LUNG CANCER
`
`515
`
`Table Ill Relation of stage to histological
`subtype
`
`Squamous Nomsquamous
`
`23
`l7
`40
`
`24
`13
`37
`
`Stages I and II
`Stage Ill
`
`NS (x2 = 0.44).
`
`there was no significant difference in
`poorly differentiated),
`intensity of staining for EGFr between the groups. Likewise
`squamous tumours alone showed no significant relationship
`between EGFr score and degree of differentiation, although
`there were only two well differentiated squamous tumours.
`There was no statistically significant difference in the degree
`of differentiation in different stages of tumour spread.
`
`Discussion
`
`We have demonstrated the presence of epidermal growth
`factor receptors in non-small cell
`lung cancer and the fact
`that squamous carcinomas show more staining than non—
`squamous tumours. We have also shown that the intensity of
`staining for EGFr is related to the stage of spread of the
`tumours.
`
`The histology of NSCLC can vary throughout the tumour
`(Mountain et al., 1974) and thus we have taken samples from
`different areas of the tumours and found no significant
`variation of staining within tumours.
`Hendler and Ozanne (1984) demonstrated EGFr in 11
`squamous lung cancers. They showed that 7 of 8 adeno-
`carcinomas failed to stain for EGFr, using a semiquantitative
`autoradiography technique. The greater intensity of staining
`for EGFr
`in squamous
`tumours is consistent with the
`findings that squamous tumours at other sites, such as head
`and neck or cervical cancer, and squamous carcinoma cell
`lines express more EGF receptors
`than other cancers
`(Cowley et al., I984). Ozanne er al. have shown a 2 to 10-
`fold excess of receptor sites in squamous cancers of various
`sites, including lung, compared to normal skin (Gusterson et
`al., 1984).
`lung showed far less staining for
`In our series, normal
`EGFr, and only 1/17 normal lungs were strongly positive.
`This may reflect
`individual variation or possibly EGFr
`involvement in other chronic inflammatory processes.
`Our results suggest that epidermal growth factor receptor
`may play a part
`in the genesis or spread of lung cancers.
`This is the third common primary carcinoma in which we
`have found a correlation of the presence of EGFr with some
`of the prognostic features (Neal et al., 1985; Sainsbury er al.,
`1985b). This may suggest
`that,
`rather
`than EGFr being
`involved in tumour initiation, subclones of tumours that
`express more EGFr or have had rearrangement of EGFr
`genomic DNA may be selected for growth,
`invasion and
`
`References
`
`metastasis. This may be a relatively late change in tumour
`development.
`(1986) examined 63 lung tumours, mainly
`Cerny er al.
`bronchial biopsies, and found 86% of squamous cancers
`stained positively for EGFr, but none of 15 small cell
`tumours stained. However, only 5 adenocarcinomas or large
`cell
`tumours were examined and the relation to stage was
`not assessed.
`
`(1985) showed that staining for EGFr was
`Neal et al
`related to poor differentiation of bladder cancers, and
`Sainsbury et al. (19851)) found a similar relationship in breast
`cancers, but we did not observe this in our specimens. in our
`series of bronchial
`tumours,
`there was no significant
`association of differentiation with stage and this may be a
`reflection of the large proportion of tumours that were
`moderately or poorly differentiated.
`lndeed in adeno-
`carcinoma of the bronchus there is no significant survival
`difference as
`regards degree of differentiation (Katlic &
`Carter, 1979).
`in the bronchial
`The detection of normal basal cells
`mucosa staining for EGFr, first reported by Gusterson et al.
`(I984), suggests that
`these stem cells of the mucosa may
`require EGF for proliferation and differentiation to normal
`mucosal cells. If these cells are the site of transformation for
`lung cancers,
`it would be expected that
`some adeno-
`carcinomas and large cell tumours would also express EGFr,
`as we have found, in contrast to Ozanne’s group (Hendler &
`Ozanne, I984).
`The demonstration of a high concentration of EGFr in the
`worse prognosis NSCLC suggests a novel potential target for
`therapy. To quantitate the binding characteristics, we have
`assayed membranes from these primary tumours and have
`found high affinity sites (Kd 1x10”9 molar) for EGF and
`much higher numbers of
`receptors
`than breast cancers
`(190fmolmg"1 membrane protein versus
`4—47 fmolmg“
`membrane protein) (Sainsbury et aI.,
`l985a). EGF bound
`drug would be expected to react specifically at these sites,
`thus leading to higher concentration of drug in tumour than
`in other
`tissue. Drugs linked to EGF should be more
`selective than current
`therapies, particularly as
`receptor
`stimulation is
`linked to internalisation, and squamous
`tumours express several-fold more EGFr than reported for
`other normal tissues, including normal keratinocytes (Ozanne
`er al., 1985).
`The monoclonal antibody used here recognises the external
`domain of the EGFr, so this could also be used for targeting.
`The relationship of EGFr to other prognostic features in
`three common primary tumours shows that this type of drug
`therapy might be generally applicable in patients in whom
`there is little alternative therapy at present. We are currently
`studying this approach in human non-small cell cancer lines.
`
`We thank Mr G.N. Morritt, Mr C. Hilton, Mr A. Hedley~Brown
`and Mr A. Blesovsky for help in collection of tumour specimens and
`Dr S. Nariman for permission to collect bronchoscopy specimens.
`We thank Dr M. Wateriield for kindly donating the antibodies to
`EGF receptor.
`
`CARPENTER, G. & COHEN, S. (l979). Epidermal growth factor. Ann.
`Rev. Biochem., 48, I93.
`CERNY, T., BARNES, D.M., HASLETON, P.S. & 4 others (i986).
`Expression of epidermal growth factor
`receptor (EGF-R) in
`human lung tumours. Br. J. Cancer, 54, 265.
`COHEN, S. (1983). The epidermal growth factor. Cancer, 51, 1787.
`COWLEY, G., SMITH, 1., GUSTERSON, B., HENDLER, F. & OZANNE,
`B. (1984). The amount of EGF receptor is elevated on squamous
`cell carcinomas. In The Cancer Cell, p. 5. Cold Spring Harbor:
`New York.
`CROFTON, J. & DOUGLAS, A. (l98l).
`ed, p. 653. Blackwell.
`
`ln Respiratory DiSea.s'('.\'. 3rd
`
`DELELLIS, R.A., STERNBERGER, L.A., MANN, R.B., BANKS, l’.M. &
`NAKANE, P.l(. (i979). Immunoperoxidase teehnics in diagnostic
`pathology. Am. J. Clin. Path., 71, 483.
`DOWNWARD, J., YARDEN, Y., MAYER, L. & 6 others (1984). Close
`similarity ofepidermal growth factor receptor and v—erb-b oncogene
`protein sequences. Nature, 307, 521.
`GUSTERSON, M., COWLEY, G., SMITH, LA. & OZANNE, A. (l984).
`Cellular localisation of human epidermal growth factor receptor.
`Cell Biol. Internat. Rep., 8, 649.
`HENDLER, F.J. & OZANNE, B.W. (1984). Human squamous cell lung
`cancers express increased epidermal growth factor receptors. J.
`Clin. Invest., 74, 647.
`
`APOTEX EX. 1026-003
`
`

`
`516
`
`D. VEALE er al.
`
`HOLLEY, R. (1975). Control of growth of mammalian cells in cell
`culture. Nature, 258, 487.
`(1979). Prognostic implications of
`I).
`KATLIC, M. & CARTER,
`histology, size and location of primary tumours. Prog. Canc.
`Res. Then, 11, 143.
`KJAER, M. (l982). Radiotherapy of squamous, adeno~ and large cell
`carcinoma of the lung. Cane. Treat. Rev., 9, l.
`MARQUARDT, H., HUNKAPILLER. M.W., HOOD, LF. & 4 others
`(1983). Transforming growth factors produced by retrovirus—trans-
`formed rodent fibroblasts and human melanoma cells: Amino
`acid sequence homology with epidermal growth factor. Proc.
`Natl Acad. Sci. USA, 80, 4684.
`(1974). A
`MOUNTAIN, C.F.._ CARR, D.T. & ANDERSON, W.A.D.
`system for the clinical staging of lung cancer. Am. J. Roentgenol,
`120. I30.
`NEAL. D.E.. MARSH, C, BENNETT. M.K. & 4 others U985).
`Epidermal—growth~factor
`receptors
`in human bladder
`cancer:
`Comparison of invasive and superficial tumours. Lamset, i, 366.
`OZANNE, 8.. SHUM, A., RICHARDS. C.S. & 5 others U985). Evidence
`for an increase of EGF receptors in epidermoicl malignancies. In
`Cancer Cells 3. Growth Factors and Transformation. Feramisco, J.
`er al. (eds) pp. 4l. Cold Spring Harbor: New York.
`
`SAINSBURY, J.R.C., FARNDON, J.R., SHERBET, G.V. & HARRlS, A.L.
`(l985a).
`Epidermal~growth-factor
`receptors
`and
`oestrogen
`receptors in human breast cancer. Lancet, i, 364.
`SAINSBURY,
`.l.R.C.. MALCOLM, A.J., APPLETON, D.R., FARNDON.
`LR. & HARRIS, A.L. (198512). Presence of epidermal growth factor
`receptor as an indicator of poor prognosis in patients with breast
`cancer. J. Clin. Pathol., 38, I225.
`SHERWIN, S.A., MINNA, J.D., GAZDAR, A.F. & TODARO, GJ. (l98l).
`Expression of epidermal and nerve growth factor receptors and
`soft agar growth factor production by human lung cancer cells.
`Cancer Res., 41, 3538.
`SIMES, RJ. (1985). Risk-benefit relationships in cancer clinical trials:
`The ECOG experience in non—small cell
`lung cancer. J. Clin.
`Oncol., 3, 462.
`(1980). Autoerine secretion and
`SPORN, M.B.. & TODARO, GJ.
`malignant transformation of cells. N. Engl. J. Meal, 303, 878.
`WATERFIELD. M.D., MAYERS, E.L.V., STROOBANT. P. & 5‘oIhcrs
`(I982). A monoclonal antibody to the human epidermal growth
`factor receptor. Cell Bin:-hem., 20, I49.
`
`APOTEX EX. 1026-004