Prognostic impact of EGF-receptor in papillary thyroid carcinoma

Transcription

1 Br. J. Caner (1993), 68, (D Mamillan Press Ltd., 1993 Br. J. Caner (1993), 68, Mamillan Press Ltd., 1993 Prognosti impat of EGF-reeptor in papillary thyroid arinoma L.A. Akslen', A.O. Myking', H. Salvesen2 & J.E. Varhaug2 'Department of Pathology, The Gade Institute, 2Department of Surgery, University of Bergen, Bergen, Norway. Summary In this study of papillary thyroid arinomas, immunopositivity for EGF-reeptor was present in a majority of the ases (96%), although different staining patterns were observed. A distint membraneous reation was found in 46%, whereas ytoplasmati positivity of various degrees was present in 90% of the ases. Strong ytoplasmi EGF-reeptor staining was signifiantly assoiated with extra-thyroidal growth of the primary tumour (P = 0.009), and it was furthermore related to dereased reurrene free survival (P = 0.006). Membraneous EGF-reeptor staining was not assoiated with reurrene free survival or patient survival. Multivariate Cox analysis showed that lymph node metastases (P = ) and ytoplasmi EGFreeptor staining (P = ) was independent indiators of tumour reurrenes in this group of surgially treated papillary thyroid arinomas. The prognosis of patients with papillary thyroid arinoma has previously been desribed in several reports (Byar et al., 1979; Tubiana et al., 1985; Carangiu et al., 1985; Akslen et al., 1991), and the influene of sex, age and tumour stage has been disussed. In addition, ertain pathologial features suh as marked nulear atypia, nerosis or vasular invasion are also important (Tennvall et al., 1985; Shindler et al., 1991; Akslen et al., 1993). This information may improve the risk estimation for individual patients. In reent years, alterations in various growth fators and their reeptors have been established as important features of the neoplasti proess (Wynford-Thomas, 1991), among them epidermal growth fator (EGF) and its reeptor (EGF-R). EGF is known to indue ell proliferation in several tissues, and its effet is mediated via a tyrosine kinase type of reeptor (Carpenter & Cohen, 1979). It has previously been shown by different methods that EGF-reeptors are expressed in papillary thyroid arinomas (Duh et al., 1985; Lemoine et al., 1991; Mizukami et al., 1991; Mizukami et al., 1992), but the biologial signifiane of this finding in terms of patient prognosis is not known. The purpose of the present study was therefore to perform a retrospetive immunohistohemial study of EGF-R expression in a series of papillary thyroid arinomas with speial referene to its prognosti importane. Material and methods Patients This material has been desribed previously (Akslen et al., 1993). Briefly, all 263 patients who were surgially treated for thyroid aner at the Department of Surgery, Haukeland Hospital, University of Bergen in the period have been studied retrospetively. After histologial revision and subtyping of the arinomas aording to the WHO riteria (Hedinger, 1988), 173 ases were found to be papillary arinomas with a known primary tumour. One of the main results from our previous study was that very few events were observed in tumours with a diameter of 10 mm or below (miroarinomas aording to the WHO riteria). These were therefore exluded, thereby onentrating on linially signifiant papillary arinomas. Suffiient material was not available in two ases, leaving 125 tumours for further analyses. All patients were surgially treated in our institution, most of them with total/near-total thyroidetomy (93%), and 91% were onsidered to be radially treated, without marosopially remaining tumour tissue. Pathologi lymph nodes were removed mostly by using a 'node-piking' proedure. Correspondene: L.A. Akslen, Department of Pathology, The Gade Institute, Haukeland Hospital, N-5021 Bergen, Norway. Reeived 16 Deember 1992; and in revised form 23 April Variables The following variables were studied: sex; age at diagnosis; primary tumour extension (intra-thyroidal growth, tumour growth in the thyroid gland apsule, major extra-thyroidal extension); lymph node metastases (absent, intra-nodal growth, extra-nodal growth) and EGF-reeptor immunostaining. Membraneous and ytoplasmati staining were reorded separately using a semiquantitative and subjetive grading system, onsidering both the intensity of staining and the proportion of positive ells: Grade 1 = no staining, Grade 2 = weak or moderate positivity, Grade 3 = strong staining in a high proportion of the tumour ells. Immunohistohemistry Immunohistohemial examination was performed on formalin-fixed and paraffin-embedded arhival material using the avidin-biotin omplex method. Setions were inubated overnight at 4 C with the primary antibody MU170-UC against EGF-reeptor (dilution 1:40) (Biogenex, CA). Antigen loalisation was ahieved by the alkaline phosphatase antialkaline phosphatase (APAAP) method. Negative ontrols were inubated with PBS, and no positive staining was observed. Follow-up Data onerning loo-regional tumour reurrenes (loal lymph node metastases or soft tissue reurrenes in the thyroid bed), appearane of distant metastases, and patient survival was ahieved through examinations in our institution or by orrespondene to the patients' home physiians. Reurrenes or metastases within 4 months after the primary operation were regarded as part of the primary status and referred to the time of diagnosis. For all patients who died, death ertifiates were examined as well as autopsy reports when available. Last date of follow-up was July 1st, 1989, and the median follow-up time was 7.3 years (maximum 18.3 years). No patient was lost to follow-up. Statistis Analyses were performed by various programmes in the statistial pakage BMDP (Dixon, 1985). Assoiations between different variables were assessed by Pearson's hi-square test. Survival analysis (life-table method) was done by BMDP-1 L using the Mantel-Cox test for differenes between groups, and plots for umulative proportion surviving are given. Reurrene free survival, i.e. the time from diagnosis until the appearane of loo-regional reurrenes or distant spread, and the patient survival (survival time until thyroid aner deaths) were studied. Deaths from interurrent disease, without tumour reurrene, were ensored in the analysis of reurrene-free survival. The influene of o-

2 EGF-RECEPTOR IN PAPILLARY THYROID CARCINOMA 809 variates on survival was analysed by the proportional hazards method (Cox & Oakes, 1984) with BMDP-2L, using a forward stepwise proedure. In these analyses, all variables with a P-value of 0.10 or less in the life-table studies were inluded. Estimated regression oeffiients and P-values are given in the Tables. Results Table I shows the distribution of the ases aording to major liniopathologi variables. Definite extra-thyroidal invasion was present in 19% of the ases, and 45% had lymph node metastases at the time of diagnosis. Tumour reurrenes during the follow-up period ourred in 27 patients (22%). Of the 25 deaths, 14 were due to thyroid aner (56%). EGF-reeptor immunostaining was found in 120 of 125 ases (96%), whereas five ases were ompletely negative. In some ases, a weak positivity was found in folliular ells in adjaent non-neoplasti thyroid tissue. Table II shows that a distint membraneous EGF-reeptor staining (grade 2 and 3) was present in 46% of the ases, and the staining was espeially evident on the apial surfae of the tumour ells (Figure 1). In 90% of the tumours, a ytoplasmi pattern of positive staining was found (Figure 1). This was a diffuse or finely granular staining in most ases, but in some tumours oarse granules were present in the supranulear part of the ytoplasm. Of the ases with ytoplasmati positivity, 12% showed a marked (grade 3) staining reation. Mixed staining patterns were also observed (Figure 1). Table III shows that the degree of ytoplasmi immunopositivity was signifiantly assoiated with the extent of primary tumour infiltration (P=0.009). The frequeny of ases with tumour growth in the thyroid apsule or extrathyroidal invasion was 17%, 63% and 71% in grade 1, 2 and 3 tumours, respetively. Membraneous staining was not assoiated with any of the variables. Univariate analysis of reurrene free survival (life-table method) showed that primary tumour extent and lymph node metastases were signifiant variables (Table IV). Cytoplasmi EGF-reeptor staining was also found to be signifiant, P = using the Mantel Cox' trend test (Table IV, Figure 2). Grade 3 ytoplasmi staining was signifiantly different from Grade 2 positivity (P = 0.05). Patient survival was assoiated with sex, age and primary tumour extension, but EGF-reeptor staining was not signifiant (Table IV). Multivariate survival analysis (Cox' method) of reurrene free survival inluding sex, age, primary tumour extent, lymph node metastases and ytoplasmi EGF-reeptor immunostaining showed that only lymph node metastases (P = ) and EGF-reeptor positivity (P = ) remained as signifiant and independent variables (Table V). EGF-reeptor was found to be the strongest prognosti fator, with a regression oeffiient of 1.28, ompared to 0.81 for lymph node metastases. Multivariate analysis (Cox' Table I Distribution of patients with papillary thyroid arinoma aording to important liniopathologi variables (n = 125) Variable n % Sex males females Age 0-49 years years Primary tumour extent intra-thyroidal tumour thyroid apsular invasion extra-thyroidal invasion Lymph node metastases absent intra-nodal growth extra-nodal growth method) of patient survival was not performed sine EGFreeptor immunostaining turned out to be not signifiant in the univariate life-table study. Disussion The present study of papillary thyroid arinoma indiates that EGF-reeptors are present in a majority of the ases. Corresponding results have been reported by others using biohemial (Duh et al., 1985; Makinen et al., 1988; Masuda et al., 1988; Di Carlo et al., 1990) or immunohistohemial methods (Lemoine et al., 1991; Song et al., 1991; Mizukami et al., 1992). Our results add further evidene to the role of onogenes in development and progression of thyroid tumours. In a reent study, the presene of transforming growth fator ox Figure 1 Patterns of EGF-reeptor immunostaining in papillary thyroid arinomas. a, membraneous staining, x 428; b, mixed membraneous and ytoplasmi staining, x 428;, strong (grade 3) ytoplasmi staining, x 428. (TGF-a), a known EGF-R ligand, as well as TGF-a mrna was found in the epithelial omponent of papillary arinomas, indiating an autorine growth fator prodution (Haugen et al., in press). In addition to alteraa b

3 810 L.A. AKSLEN et al. tions in the EGF-R system, we have previously shown that -erbb-2 expression is also inreased (Haugen et al., 1992). The pattern and intensity of EGF-reeptor immunostaining varied onsiderably between different tumours. Inreased expression in the tumour ell ytoplasm was signifiantly assoiated wtih extra-thyroidal growth of the primary tumours. However, no relationship to o-existing lymph node metastases was found, in line with studies on breast aner (Bolla et al., 1992). Previous reports on arinomas of the urinary bladder, stomah and large bowel have also suggested an assoiation between EGF-R positivity and invasive growth (Neal et al., 1985; Yasui et al., 1988; Smith et al., 1989; Yonemura et al., 1991), but the mehanism is not lear. Some in vitro studies indiate that EGF may stimulate the Table II Patterns and intensity of EGF-reeptor immunostaining in papillary thyroid arinoma (n = 125) Membraneous Cytoplasmi staining staining Staininga n % n % Grade Grade Grade agrade 1 = no staining; Grade 2 = slight or moderate staining; Grade 3 = marked staining. Table III Assoiations between pattern of EGF-reeptor immunostaining and important liniopathologi variables in patients with papillary thyroid arinoma (n = 125) P-value" Membraneous Cytoplasmi Variables staining" stainingb Sex Age Primary tumour extentd 0.2 O.009e Lymph node metastasesf apearson's hi-square. bgrades 1, 2, 3 (see Material and methods). C0-49 years, 50 + years. dintra-thyroidal, thyroid apsular invasion, extra-thyroidal invasion. efor details, see Results. fabsent, intra-nodal growth, extra-nodal growth. seretion of proteolyti enzymes (Lee & Weinstein, 1978; Boyd, 1989), but in a study of EGF-R and athepsin D in endometrial and ervial tumours, no signifiant relationship was found (Sambia et al., 1991). In addition, ellular migration may be stimulated by EGF (Westermark et al., 1982). Positivity for EGF-reeptor was signifiantly related to inreased risk of reurrent disease in papillary thyroid arinomas, in ontrast to a reent study where no prognosti importane ould be found (Mizukami et al., 1992). Our present results are thus in general agreement with reports on breast aner, where the presene of EGF-reeptors has been doumented to be an important preditor of tumour reurrenes as well as patient survival (Sainsbury et al., 1987; Grimaux et al., 1989; Niholson et al., 1991; Toi et al., 1991). Further studies using quantitative methods should now be performed. The mehanism of EGF-reeptor influene on patient prognosis has not yet been larified. In breast aner, inreased expression of EGF-R has been related to dedifferentiation of tumour ells, with inreased ellular atypia, m' 1.0._.._ en a a.> 0.4 E Grade3(n=14) Grade 1 (n = 12) Grade 2 (n = 99) EGF-RECEPTOR (ytoplasmi staining) Mantel-Cox test, trend: p = lo 15 Years after diagnosis Figure 2 Reurrene-free survival aording to ytoplasmi EGF-reeptor immunostaining in papillary thyroid arinomas. Table IV Univariate survival analysis (life-table method) of patients with papillary thyroid arinoma aording to liniopathologi variables and EGF-reeptor immunostaining, n = 125 (the figures in parenthesis give the number of patients who were alive after 10 years and eligible for estimation of 10-year survival; the P-values orrespond with a standard life-table analysis, based on all patients and observed events during the whole follow-up period) 10-year reurrene 10-year patient Variables n free survival (%)a pb survival (%) pb Sex females 89 (28) males 36 (10) Age 0.05 < years 71 (25) years 54 (13) Primary tumour extent intra-thyroidal 51 (22) thyroid apsule invasion 50 (12) extra-thyroidal invasion 24 (4) Lymph node metastases < absent 67 (22) intra-nodal growth 20 (6) extra-nodal growth 36 (9) EGF-reeptor immunostaining membraneous staining grade 1 67 (16) grade 2 46 (17) grade 3 12 (5) ytoplasmi staining grade 1 12 (7) grade 2 99 (25) grade 3 14 (6) aof 125 ases inluded, nine were not radially treated and exluded in the analyses of reurrene free survival. bmantel-cox' test. 'Trend version of the Mantel-Cox' test.

4 EGF-RECEPTOR IN PAPILLARY THYROID CARCINOMA 811 Table V Multivariate analysis of reurrene free survival (Cox' proportional hazards method) of patients with papillary thyroid arinoma (n = 11 3)a Regression Standard Variables oeffiient error P-valueb Lymph node metastases EGF-reeptor" (ytoplasmi staining) aof 125 ases, nine were not radially treated and were therefore exluded, and three ases were exluded due to lak of information. bl ratio test of signifiane. Categories: absent, intra-nodal growth, extra-nodal growth. dcategories: grade 1, grade 2, grade 3. redued onentration of oestrogen reeptors and a higher proliferative fration as measured by Ki-67 immunopositivity (Fitzpatrik et al., 1984; Bolla et al., 1992). An inverse orrelation between EGF-R ontent and TSH-response has also been noted in anaplasti thyroid arinomas (Di Carlo et al., 1990). These findings indiate that inreased EGF-R expression is assoiated with proliferative ativity in tumour ells and redued dependeny of normal growth regulators. Interestingly, two speifi staining patterns were observed in the present study. A distint membraneous positivity was found in 46% of the ases, whereas a diffuse or granular ytoplasmi staining was observed in 90% of all tumours. These patterns have been briefly noted by others (Kashima et al., 1991; Mizukami et al., 1992). Only the ytoplasmati form of staining, however, showed a signifiant assoiation with patient prognosis. Earlier studies (Lemoine et al., 1990; Aasland et al., 1988; Aasland et al., 1990) indiate that inreased expression of EGF-R in thyroid tumours is not due to gene amplifiation or gross rearrangements. Therefore, epigeneti hanges may be involved, and reent Western blot studies of fresh material indiate that modified proteins are present in some ases (Haugen et al., in preparation). In onlusion, the present study using semiquantitative immunohistohemial assessment indiates that expression of EGF-reeptors may be an important feature of papillary thyroid arinomas, and various staining patterns seem to be of different biologial signifiane. Strong ytoplasmati immunopositivity was assoiated with extra-thyroidal tumour invasion and found by multivariate analysis to be the strongest independent preditor of reurrent disease. However, quantitative methods should be used in further studies to establish this assoiation. This study has been supported by the Norwegian Caner Soiety. We thank Tor Christensen, Inger Hernar, Anne-Marie Larsen and Bendik Nordanger for exellent tehnial assistane. Referenes AASLAND, R., LILLEHAUG, J.R., MALE, R., JOSENDAL, O., VAR- HAUG, J.E. & KLEPPE, K. (1988). Expression of onogenes in thyroid tumours: o-expression of -erbb-2/neu and -erbb. Br. J. Caner, 57, AASLAND, R., AKSLEN, L.A., VARHAUG, J.E. & LILLEHAUG, J.R. (1990). Co-expression of the genes enoding transforming growth fator-alpha and its reeptor in papillary arinomas of the thyroid. Int. J. Caner, 46, AKSLEN, L.A., HALDORSEN, T., THORESEN, S.O. & GLATTRE, E. (1991). Survival and auses of death in thyroid aner: a population-based study of 2479 ases from Norway. Caner Res., 51, AKSLEN, L.A., MYKING, A.O., SALVESEN, H. & VARHAUG, J.E. (1993). Prognosti importane of various liniopathologial features in papillary thyroid arinoma. Eur. J. Caner, 29A, BOLLA, M., CHEDIN, M., COLONNA, M., MARRON, J., ROSTAING- PUISSANT, B. & CHAMBAZ, E. (1992). Prognosti value of epidermal growth fator reeptor in a series of 303 breast aners. Eur. J. Caner, 28, BOYD, D. (1989). Examination of the effets of epidermal growth fator on the prodution of urokinase and the expression of the plasminogen ativator reeptor in a human olon aner ell line. Caner Res., 49, BYAR, D.P., GREEN, S.B., DOR, P., WILLIAMS, E.D., COLON, J., VAN GILSE, H., MAYER, A., SYLVESTER, M. & GLABBEKE, M.V. (1979). A prognosti index for thyroid arinoma. A study of the E.O.R.T.C. thyroid aner ooperative group. Eur. J. Caner, 15, CARCANGIU, M.L., ZAMPI, G., PUPI, A., CASTAGNOLI, A. & ROSAI, J. (1985). Papillary arinoma of the thyroid. A liniopathologi study of 241 ases treated at the University of Florene, Italy. Caner, 55, CARPENTER, G. & COHEN, S. (1979). Epidermal growth fator. Ann. Rev. Biohem., 48, COX, D.R. & OAKES, D. (1984). Analysis of Survival Data. Chapman and Hall: London. DI CARLO, A., MARIANO, A., PISANO, G., PARMEGGIANI, U., BEGUINOT, L. & MACCHIA, V. (1990). Epidermal growth fator reeptor and thyrotropin response in human thyroid tissues. J. Endorinol. Invest., 13, DIXON, W.J. (1985) (hief ed.). BMDP Statistial Software. University of California Press. DUH, Q.-Y., GUM, E.T., GEREND, P.I., RAPER, S.E. & CLARK, O.H. (1985). Epidermal growth fator reeptors in normal and neoplasti thyroid tissue. Surgery, 98, FITZPATRICK, S.L., BRIGHYWELL, J., WITTLIFF, J.L., BARROWS, G.M. & SCHULTZ, G.S. (1984). Epidermal growth fator binding by breast tumor biopsies and relationship to estrogen reeptor and progestin reeptor. Caner Res., 44, GRIMAUX, M., ROMAIN, S., REMVIKOS, Y., MARTIN, P.M. & MAGDELENAT, H. (1989). Prognosti value of epidermal growth fator reeptor in node positive breast aner. Br. Caner Res. Treat., 14, HAUGEN, D.F., AKSLEN, L.A., VARHAUG, J.E. & LILLEHAUG, J. (1992). Expression of -erbb-2 protein in papillary thyroid arinomas. Br. J. Caner, 65, HAUGEN, D.F., AKSLEN, L.A., VARHAUG, J.E. & LILLEHAUG, J. (1993). Demonstration of a TGF-a - EGF reeptor autorine loop and -my protein overexpression in papillary thyroid arinomas. Int. J. Caner (in press). HEDINGER, C. (1988). (ed). Histologial Typing of Thyroid Tumours, WHO, Springer-Verlag: Berlin, Heidelberg. KASHIMA, K., YOKOYAMA, S., NAKAYAMA, I. & NOGUCHI, S. (1991). Immunohistohemial study on expression of -my, p53, -erbb-2 and epidermal growth fator reeptor in human thyroid tumors. Ata Histohem. Cytohem., 24, LEE, L.S. & WEINSTEIN, I.B. (1978). Epidermal growth fator, like phorbol esters, indues plasminogen ativator in HeLa ells. Nature, 274, LEMOINE, N.R., WYLLIE, F.S., LILLEHAUG, J.R. & 8 others (1990). Absene of abnormalities of the -erbb-1 and -erbb-2 protoonogenes in human thyroid neoplasia. Eur. J. Caner, 26, LEMOINE, N.R., HUGHES, C.M., GULLICK, W.J., BROWN, C.L. & WYNFORD-THOMAS, D. (1991). Abnormalities of the EGF reeptor system in human thyroid neoplasia. Int. J. Caner, 49, MAKINEN, T., PEKONEN, F., FRANSSILA, K. & LAMBERG, B.A. (1988). Reeptors for epidermal growth fator and thyrotropin in thyroid arinoma. Ata Endorinol., 117, MASUDA, H., SUGENOYA, A., KOBAYASHI, S., KASUGA, Y. & IIDA, F. (1988). Epidermal growth fator reeptor on human thyroid neoplasms. World J. Surg., 12, MIZUKAMI, Y., NONOMURA, A., HASHIMOTO, T., MICHIGISHI, T., NOGUCHI, M., MATSUBARA, F. & YANAIHARA, N. (1991). Immunohistohemial demonstration of epidermal growth fator and -my onogene produt in normal, benign and malignant thyroid tissues. Histopathology, 18, MIZUKAMI, Y., NONOMURA, A., MICHIGISHI, T., YOKOYAMA, K., NOGUCHI, M., HASHIMOTO, T., NAKAMURA, S. & MAT- SUBARA, F. (1992). Immunohistohemial demonstration of epidermal growth fator reeptors in normal, benign and malignant thyroid tissues. Int. J. Onol., 1, NEAL, D.E., MARSH, C., BENNET, M.K., ABELL, P.D., HALL, R.R. & SAINSBURY, J.R.C. (1985). Epidermal growth fator reeptors in human bladder aner: omparison of invasiveness and superfiial tumors. Lanet, i,

5 812 L.A. AKSLEN et al. NICHOLSON, S., RICHARD, J., SAINSBURY, C. & 7 others (1991). Epidermal growth fator reeptor (EGFR) results of a 6 year follow-up study in operable breast aner with emphasis on the node negative subgroup. Br. J. Caner, 63, SAINSBURY, J.R.C., FARNDON, J.R., NEEDHAM, G.K., MALCOLM, A.J. & HARRIS, A.L. (1987). Epidermal growth fator reeptor status as preditor of early reurrene of and death from breast aner. Lanet, i, SCAMBIA, G., PANICI, P.B., FERRANDINA, G., BAIOCCHI, G., DISTEFANO, M. & MANCUSO, S. (1991). Cathepsin D in primary endometrial and ervial tumors: relationship with histopathologial parameters and with estrogen, progesterone, and epidermal growth fator reeptor. Caner J., 4, SCHINDLER, A.-M., VANMELLE, G., EVEQUOZ, B. & SCAZZIGA, B. (1991). Prognosti fators in papillary arinoma of the thyroid. Caner, 68, SMITH, K., FENNELLY, J.A., NEAL, D.E., HALL, R.R. & HARRIS, A.L. (1989). Charaterization and quantitation of the epidermal growth fator reeptor in invasive and superfiial bladder tumors. Caner Res., 49, SONG, B. (1991). Immunohistohemial demonstration of epidermal growth fator reeptor and eruloplasmin in thyroid diseases. Ata Pathol. Jpn., 41, TENNVALL, J., BI0RKLUND, A., M0LLER, T., RANSTAM, J. & AKERMAN, M. (1985). Prognosti fators of papillary, folliular and medullary arinomas of the thyroid gland. Retrospetive multivariate analysis of 216 patients with a median follow-up of 11 years. Ata Radiol. Onol., 24, TOI, M., OSAKI, A., YAMADA, H. & TOGE, T. (1991). Epidermal growth fator reeptor expression as a prognosti indiator in breast aner. Eur. J. Caner, 27, TUBIANA, M., SCHLUMBERGER, M., ROUGIER, P., LAPLANCHE, A., BENHAMOU, E., GARDET, P., CAILLOU, B., TRAVAGLI, J. & PARMENTIER, C. (1985). Long-term results and prognosti fators in patients with differentiated thyroid arinoma. Caner, 55, WESTERMARK, B., MAGNUSSON, A. & HELDIN, C.-H. (1982). Effet of epidermal growth fator on membrane motility and ell loomotion in ultures of human lonal glioma ells. J. Neurosi. Res., 8, WYNFORD-THOMAS, D. (1991). Onogenes and anti-onogenes: the moleular basis of tumour behaviour. J. Pathol., 165, YASUI, W., SUMIYOSHI, H., HATA, J., KAMEDA, T., OCHIAI, A., ITO, H. & TAHARA, E. (1988). Expression of epidermal growth fator reeptor in human gastri and oloni arinomas. Caner Res., 48, 137. YONEMURA, Y., SUGIYAMA, K., FUSHIDA, S., KAMATA, T., OHOYAMA, S., KIMURA, H., YAMAGUCHI, A. & MIYAZAKI, I. (1991). Tissue status of epidermal growth fator and its reeptor as an indiator of poor prognosis in patients with gastri aner. Anal. Cell. Pathol., 3,