- Email: [email protected]
Metallothionein expression in duct carcinoma in situ of the breast
Metallothionein Expression in Duct Carcinoma In Situ of the Breast A. G. DOUGLAS-JONES, PHD, FRCPA, K. W. SCHMID, MD, B. BIER, MD, K. HORGAN, FRCSI, K. LYONS, FRCR, N. D. DALLIMORE, MRCPATH, I. J. MONEYPENNY, FRCS, AND B. JASANI, PHD, MRCPATH In a previous study immunocytochemically detectable metallothionein (MT) expression in tumor cells of invasive duct carcinoma of the breast was shown to be associated with a more aggressive behavior and these findings have been subsequently confirmed by others. The aim of this study was to examine the prevalence and significance of MT positivity in preinvasive duct carcinoma i n situ (DCIS). Fifty-five specimens of pure screen-detected DCIS were stained immunocytochemically for MT using the antibody E9. The intensity and distribution of MT staining were assessed using a semiquantitative method resulting in intensity distribution (ID) scores allowing duct by duct analysis in relation to architectural and cytological features of the DCIS. In general, myoepithelial cells around benign and malignant structures stained uniformly strongly for MT. Staining in DCIS was analyzed by architecture irrespective of cytology
and by nuclear grade irrespective of architectnre. The results showed that MT staining was significantly greater in comedo (ducts with necrosis) DCIS (ID = 97) compared with n o n c o m e d o (ducts without necrosis) DCIS (ID = 56) (P = .05 by Mann Whitney Ustatistic)and that low cytological grade (ID = 50) was associated with less MT staining than was high cytological grade (ID = 92) (P = .05 by Mann Whitney Ustatistic). These observations thus are consistent with the previously observed association between MT positivity and more aggressive behavior in invasive duct carcinoma of the breast. HUM PATHOL 26:217--222. Copyright © 1995 by W.B. Saunders Company Key words: methallothionein, duct carcinoma of breast, duct carcinoma in sitn, immunohistochemistry, breast. Abbreviations: DCIS, duct carcinoma in sitn; ID, intensity distribution°
I n a r e c e n t study invasive d u c t c a r c i n o m a s of the breast exhibiting immunocytochemically detectable m e t a l l o t h i o n e i n (MT) e x p r e s s i o n were s h o w n to have a significantly p o o r clinical o u t c o m e i n t e r m s o f t u m o r free interval. 1 I n two f u r t h e r i n d e p e n d e n t studies cond u c t e d o n 99 a n d 79 u n s e l e c t e d s p e c i m e n s of b r e a s t c a r c i n o m a , a significant association was o b s e r v e d between i n c r e a s e d levels o f M T e x p r e s s i o n a n d tum o r type, local r e c u r r e n c e , a n d p a t i e n t l e n g t h o f survival. 2,3 T h e aim of this study was to e x a m i n e preinvasive d u c t a l m a l i g n a n t e p i t h e l i a l p r o l i f e r a t i o n s ( d u c t carcin o m a i n situ or DCIS) of the b r e a s t to d e t e r m i n e w h e t h e r or n o t M T is e x p r e s s e d i n these early lesions a n d w h e t h e r this e x p r e s s i o n a n d its level are r e l a t e d to previously d e s c r i b e d m o r p h o l o g i c a l p r e d i c t o r s o f clinical behavior, i n c l u d i n g a r c h i t e c t u r a l type, grade, a n d necrosis.
tween 50 and 65 years. A representative block from each specimen containing DCIS with background breast tissue was selected for immunocytochemical staining of MT and its correlation with morphological characteristics as assessed by observation of the material on an adjacent serial section stained with hematoxylin-eosin.
MATERIALS AND METHODS
Case Material Fifty-five specimens of screen-detected pure DCIS of breast were retrieved from the files of the Department of Pathology at the Llandough Hospital and University Hospital of Wales, Cardiff. The ages of the patients were within the range of the National Breast Screening Programme, ie, be-
From the Department of Pathology, University of Wales College of Medicine, and the Breast Test Wales Group, Cardiff, Wales; and the Department of Pathology, Universityof Munster/Westphalia, Germany. Accepted for publicationJune 28, 1994. Address correspondence and reprint requests to A. G. DouglasJones, PhD, FRCPA, Department of Pathology, University of Wales College of Medicine, Cardiff, South Wales, UK CF4 4XN. Copyright © 1995 by W.B. Saunders Company 0046-8177/95/2602-001355.00/0
217
Morphological Classification of Lesions Classification according to duct architecture. The architectural pattern of the DCIS was determined in the following way. If more than 70% of the structures present were of a single architectural type, the case specimen was considered to be in this category and three ducts of this architecture were assessed. If there was a mixture of two or more architectural types (<70% of any one type), ie, mixed type, then three structures of each type were scored separately. Using these criteria any one specimen could contain one or more types of DCIS. For the purpose of this duct by duct analysis, cytology was ignored and ducts were classified as "with necrosis" or "comedo" if they contained any degree of central necrosis (Table 1). Classification according to cytological grade. The tumor cytology was subjectively assessed using the same criteria as applied in routine grading of infiltrating breast carcinoma irrespective of duct architecture or necrosis. 4 Briefly, grade 1 lesions showed uniform, rounded, regular nuclei with uniform chromatin staining. Grade 2 lesions showed moderate nuclear pleomorphism and grade 3 lesions showed marked nuclear pleomorphism, coarse chromatin, and hyperchromasia. Nuclear grade of DCIS was generally uniform within each individual specimen, although one specimen contained structures containing epithelium of different grades; these were assessed for MT staining separately (Table 1). Classification according to extent of intraduct necrosis. A third type of analysis based on a case by case comparison was performed on the basis of categorizing the lesion or case as a whole as either pure comedo (>90% of ducts containing necrosis), DCIS with necrosis (DN+), or DCIS without necrosis (<10% of ducts containing necrosis) ( D N - ) (Table 1).
HUMAN PATHOLOGY
Volume 26, No, 2 (February 1995)
I m m u n o c y t o c h e m i c a l Staining of MTs Paraffin sections 5-#m thick were cut from each block, dewaxed, and incubated with 1:20,000 dilution of the monoclonal anti-MT (E9) applied overnight at 4°C. The monoclonal anti-MT antibody, E9, was used in its unfractionated ascites form. This was obtained as a generous gift from Professor John Kay, Department of Biochemistry, University College of Cardiff, Cardiff, Wales. This antibody was found by Jasani and Elmes 5 to be immunocytochemically reactive against a conserved epitope shared by the 1 and 2 isoforms of human, rat, and horse MT, as tested on formalin-fixed, paraffin-embedded tissues obtained from the species, s The antibody is now commercially available in its purified form from Dako (Dako Ltd, High Wycombe, UK). The details of the preparation of the original antibody are as previously described. 1 The tissue binding of the antibody was disclosed using a standard alkaline phosphatase antialkaline phosphatase technique employing new fuchsin as the substrate. Sections were counterstained with hematoxylin.
Analysis of Staining: Intensity Distribution Scoring of MT Staining Within an individual ductal structure containing malignant epithelium, the total percentage of positive cells was assessed. Then the percentages of weakly, moderate, and strongly staining cells were assessed, such that the sum of these categories equalled the overall percentage of positivity. A staining score was then calculated as follows: score (out of a maximum of 300) = 1 × percentage of weak (+), 2 X percentage of moderate (+ +), or 3 × percentage of strong (+ + +) positivity. The assessment was performed on a consensus basis by two pathologists (AGD-J and BJ) using a double-headed microscope. Initially, an overall decision on whether or not an individual specimen showed homogeneous staining was made. If all the ducts present showed similar levels of staining (homogeneous), three ducts were chosen at random for assessment. If staining appeared to be different in different ducts (heterogeneous), the three strongest staining ducts and the three weakest staining ducts were assessed and the corresponding mean values were calculated. An average value was calculated from these means to represent the overall score for the lesion. Statistical analysis was performed using nonparametric methods (Mann Whitney Ustatistic and NemenyiDunn analysis of variance).
RESULTS
FIGURE 1. Nonneoplastic background breast Iobule showing
no staining for MT in acinar or ductal epithelium but strong staining in myoepithelial cells. (Monoclonal antibody E9: alkaline phosphatase antialkaline phosphatase technique, Hematoxylin counterstain; original magnification x500.)
strongly positive. In s o m e specimens within an individual ductal structure adjacent individual cells varied f r o m b e i n g strongly positive to b e i n g entirely negative in a " m o s a i c " p a t t e r n (Fig 2). In o t h e r specimens there was variation in staining f r o m d u c t to duct, s o m e showing u n i f o r m l y strong a n d s o m e u n i f o r m l y weak staining a n d s o m e showing n o staining at all except for the p e r i p h e r a l myoepithelial cells (Fig 3). T h e semiquantitative intensity distribution (ID) scoring system thus was a d o p t e d as a m e a n s o f quantifying the staining a n d indicating its variability t h r o u g h s t a n d a r d e r r o r calculations. T h e overall results o f ID scoring are s u m m a r i z e d in Table 1.
I m m u n o s t a i n i n g Patterns of MTs
D u c t C a r c i n o m a In Situ With Different Types of A r c h i t e c t u r e
Generally, in the b a c k g r o u n d n o n n e o p l a s t i c breast tissue d u c t u l a r a n d acinar epithelium showed n o detectable MT staining. Metallothionein-positive staining c o u l d be seen in the nucleus, cytoplasm, or both. For the p u r p o s e s o f analysis in this study n o distinction was m a d e between n u c l e a r a n d cytoplasmic staining. Myoepithelial cells associated with b o t h b e n i g n epithelium a n d DCIS were consistently strongly positive for MT (Fig 1). For the p u r p o s e o f the analysis o f DCIS myoepithelial cell staining thus was ignored. S o m e d u c t structures showing epithelial proliferation were seen a n d the luminal cells showed either low levels o f MT expression or were negative, b u t n o f o r m a l assessment o f these structures was made. Malignant epithelium (DCIS) showed variable intensity o f staining f r o m negative to
T h e results o f analysis o f MT by architecture o f DCIS in individual ducts (ignoring cytology) are shown in Fig 4. T h e m e a n ID score (out o f a possible total o f 300) for ducts with necrosis ( c o m e d o ) was 94 +_ 58 (n = 35), solid 78 _+ 63 (n = 9), cribriform 57 + 57 (n = 19), a n d micropapillary 30 _+ 38 (n = 9). Differences between c o m e d o (ducts with necrosis) a n d micropapillary scores were significant (P = .05) a n d between come d o a n d cribriform scores were significant (P = .01) by N e m e n y i - D u n n analysis o f variance. W h e n data for solid, micropapillary, a n d cribriform were c o m b i n e d ( m e a n ID = 56), there was a significant difference between c o m e d o (ducts with necrosis) a n d n o n c o m e d o (ducts without necrosis) DCIS ( m e a n ID, 94; P = .05 by M a n n W h i t n e y U statistic).
218
METALLOTHIONEIN IN DCIS OF THE BREAST (Douglas-Jones et al)
J
FIGURE 3. Cribriform DCIS of low cytological grade in which there is no epithelial positivity, but surrounding myoepithelial cells show strong positivity. (Monoclonal antibody E9: alkaline phosphatase antialkaline phosphase technique; hematoxylin counterstain; original magnification ×250.)
form r o u n d e d nuclei (grade 1) had a mean ID score of 50 (+55; n = 21), DCIS with moderate nuclear pleomorphism (grade 2) had a mean ID score of 106 (+67; n = 15), and marked DCIS with nuclear pleomorphism
Fifty-FiveSpecimens of Pure Screen-Detected DCIS
TABLE 1.
FIGURE 2. Comedo type DCIS showing marked variation in MT staining in individual ceils. Some stain very strongly, whereas adjacent cells are unstained, (Monoclonal antibody E9: alkaline phosphatase antialkaline phosphatase technique, hematoxylin counterstain; original magnification ×250,)
Duct Carcinoma In Situ With and Without Necrosis The analysis based on the presence and extent of intraduct necrosis is shown in Fig 5. The mean ID score for pure c o m e d o ( > 9 0 % of ducts showing necrosis) was 107 (_+65; n = 18), for DCIS with necrosis (DN+; 10% to 90% of ducts showing necrosis) it was 77 (_+49; n = 18), and for DCIS without necrosis (DN; < 1 0 % of ducts showing necrosis) it was 49 (+59; n = 19). Scores of pure c o m e d o DCIS were significantly different from scores for DCIS without necrosis (P = .05 by Mann Whitney U statistic).
Mean ID Score (Maximum of 300) (SD) Architecture Comedo (with necrosis) Micropapillary Cribriform Solid Total assessments* Grade 1 2 3 Total assessmentst Necrosis DCIS (pure comedo >90%) DCIS + (necrosis (10% to 90%)) DCIS (no necrosis <10%) Total specimens
94 30 57 78
No. of Specimen/ Assessments
(58) (38) (57) (63)
35 9 19 9 73
50 (55) 106 (67) 8O (54)
21 15 20 56
107 (65)
18
77 (47) 49 (59)
18 19 55
* If there was a mixture of patterns within one lesion, three examples of each architecture were selected and scored separately. J- One specimen showed high (3) and low (1) grades in different ducts within a single lesion.
Duct Carcinoma In Situ of Different Grades Analysis of the data by grade is shown in Fig 6. Ductal carcinoma in situ composed of cells with uni219
HUMAN PATHOLOGY
Volume 26, No. 2 (February 1995)
200-
2000.05
0.1
NS
•••-
150t--
o
100-
t.. m
100n
50MP
CR
S
50-
C
FIGURE 4. Staining score for MT out of a maximum of 300 related to architectural type of DCIS (ignoring cytology) MP, micropapillary; CR, cribriform; S, solid; C, c o m e d o (ducts with necrosis).
1
I
!
2
3
FIGURE 6. Staining score for MT out of a maximum of 300
(grade 3) had a m e a n ID score of 80 (_+54; n = 20). In practice it was often difficult to assign a particular lesion to grade 2 or 3, whereas grade 1 lesions were clearly identifiable because of the regularity of their nuclei. If data on grade 2 and 3 lesions are c o m b i n e d (ID = 92) and c o m p a r e d with those of grade 1 lesions (ID = 50), the difference is significant (P = .05 by Mann Whitney U statistic). Overall the results indicate that there is m o r e MT expression in ducts with necrosis than in those without and in ducts with cytologically high grade DCIS. DISCUSSION
T h e r e are two electrophoretically separable subgroups or is.forms of MT in m a m m a l s referred to as 200,
150,
100
50 ¸
DN-
DN+
PC
FIGURE 5. Staining score for MT out of a maximum of 300 related to presence of necrosis. DN-, DCIS without necrosis (less than 10% of ducts show necrosis); DN+, DCIS with necrosis ( b e t w e e n 10% and 90% of ducts show necrosis); PC, pure c o m e d o DCIS (more than 90% of ducts show necrosis). 220
related to cytological g r a d e (irrespective of architecture).
MT 1 and MT 2. These i s . f o r m s show small differences in their binding affinity for metal ions but a p p e a r to have similar biochemical properties. T h e antibody used in this study recognizes a d o m i n a n t immunoreactive epitope shared by MT 1 and MT 2; at present it is not possible to distinguish between the i s . f o r m s immunocytochemically. 5 T h e significance of differential i s . f o r m expression is not known. Metallothioneins are not normally expressed constitutively in tissues in amounts detectable by immunocytochemistry, except for a few specific cell types, such as my.epithelial, renal, and thyroid epithelial cells. Metallothionein expression is readily and rapidly inducible in m a n y tissues by a wide variety of stress factors, such as exposure to high levels of heavy metal ions, lipopolysaccharide, interleukin-1, interferon, glucocorticoids, and ultraviolet light. ~s Metallothionein overexpression has b e e n detected immunocytochemically in a n u m b e r of h u m a n tumors and is associated with a worse prognosis in breast carcinoma 1-~ and in malignant m e l a n o m a . 9 T h e biological basis for this overexpression in tumors and the m e c h a n i s m of its influence on prognosis is not currently understood. T h e results of this study and a separate qualitative study of benign and proliferative lesions of breast epithelium show that immunocytochemically detectable MT expression is not an innate feature of background, presumably normal, ductal or glandular breast epithelium. Consistently strong staining was observed in m y . epithelial cells, but no lumenal cell positivity was seen in aden.sis, sclerosing aden.sis, or papillomas. In a small n u m b e r of specimens included in the qualitative study, the MT staining of the infiltrating duct carcin o m a was similar to that of the duct carcinoma in situ associated with it. 1° T h e findings of the present study show that the
METALLOTHIONEIN IN DCIS OF THE BREAST (DougLas-Jones et al)
degree of MT positivity in preinvasive breast duct carcin o m a varies with histological architecture, cytological grade, and presence and extent of lumenal necrosis. Metallothionein positivity in tumor cells may be the result of expression of either a metabolically more stable form of MT or increased synthesis over degradation leading to net accumulation of MT in cytoplasm a n d / or nuclei of the t u m o r cells. The normally present housekeeping MT is most probably b o u n d to zinc ions and has a high turnover disallowing any detectable level of accumulation. Metallothionein b o u n d to copper or cadmium ions is known to be more stable, could account for MT positivity, and may play a role in the generation of malignancy. Histochemically detectable copper, zinc, and iron have been f o u n d in malignant melanoma but not in benign intradermal nevi, reflecting the MT findings in melanoma, n Studies quantifying trace metal concentrations in benign and malignant tumors, including breast carcinomas, show increased levels of copper or zinc in malignant tumors when compared with corresponding benign tissue. 12-16 An in vitro model has been proposed implicating copper in the generation of hydroxyl free radicals that can damage DNA. Copper complexes are reduced by superoxide radicals or other reducing agents to the cuprous state, and then they react with hydrogen peroxide to form hydroxyl radicals. These radicals may cause double-stranded DNA breakage, which is not repairable by cellular mechanisms. ~7 T h e r e is evidence to suggest that cells u n d e r oxidative stress are likely to overexpress MT as a protective response to the DNA-damaging influence of hydroxyl radicals, ls'19 Thus, observational data and theoretical evidence link raised heavy metal levels with carcinogenesis, and the raised MT levels observed in breast cancer may reflect this involvement because MT is acting as the main binding protein for these metals. Increased MT synthesis in response to inflammatory stress mediated by glucocorticoids or cytokines is another explanation. T h e r e is at present no evidence to support this possibility. It is possible, on the other hand, that the tumor cells are responding to intracellular factors or that the MT synthesis is switched on nonspecifically because of a general loss of transcription control in neoplastic cells. At least two varieties of tumors have been shown to have a consistent gene defect associated with increased MT expression. 2°'2a Thus, it is clear that further studies are n e e d e d to examine the mechanism of MT overexpression in breast tumors before the basis for the increased malignant potential of MT-positive breast duct carcinoma can be understood. Previous studies on the subtypes of DCIS have shown that histological pattern and grade have clinical relevance. 22 The n e e d for strict definitions of criteria for subtype classification has been stressed. Invasive recurrence only followed high grade DCIS and most commonly followed c o m e d o DCIS. It was suggested that the term " c o m e d o " be applied to the high grade DCIS solid pattern irrespective of the a m o u n t of necrosis. 22 Other investigators have accepted some degree of intraduct necrosis within a given architectural type of 221
DCIS. 23 We attempted to overcome the problems of classification associated w i t h " m i x e d type" DCIS by performing a duct by duct analysis rather than by forcing a particular lesion into one or other diagnostic category. Overexpression of c-erb B2 oncoprotein has been demonstrated immunocytochemically in DCIS associated with large cell size, high proliferative index, and histological c o m e d o subtype. 24 Mixed type DCIS with central necrosis showed higher S-phase fraction and proliferative index compared with lesions of mixed type DCIS without necrosis. It has b e e n proposed that the presence of focal necrosis in DCIS represents an intermediate position with regard to cell proliferation rate between pure c o m e d o and small cell cribriform and micropapillary types. 24'25 This hypothesis is reflected in the results reported here in which expression of MT in DCIS with necrosis is intermediate between DCIS without necrosis and pure c o m e d o DCIS. These correlations indicate that MT expression may be linked to cell proliferation rate. The results presented here show that although DCIS of low grade and more "benign types" (micropapillary or cribriform) can express MT, it is at a lower level than that expressed in high grade or c o m e d o type DCIS, reflecting the previous findings in infiltrating duct carcinoma. 1 In the intraductal preinvasive lesions studied here, there is a wide variety of expression of MT among malignant cells in the same duct (Fig 2), giving ample scope for a clonal selection process to occur during progression of the tumor, if MT expression confers some biological advantage to those malignant cells containing it in abnormally large amounts. In the cases described here there has been no therapeutic intervention and so the variation in MT overexpression observed cannot be the result of drug-related induction or therapy-related selection pressure. Although immunocytochemistry does not allow distinction between wild type and mutant p53 expression, accumulation of immunoreactive p53 protein has been demonstrated in approximately 25% of pure DCIS and is associated with large cell size and increased morphological degree of necrosis. 26 These findings reflect the overexpression of MT reported here and imply a relationship between MT and p53 expression. T h e r e are other interesting parallels between p53 and MT because increased expression of both is seen after exposure of skin to ultraviolet light s'27 and both molecules appear to have a role in protection against certain forms of DNA damage. 2s'29 Interestingly, the pattern of p53 expression in large cell DCIS illustrated by Poller et a126 is remarkably similar to that seen in high grade, c o m e d o DCIS illustrated here with strongly positive cells and adjacent negative ones ("mosaic" pattern, Fig 2). Double immunolabeling will allow the investigation of the relationship between p53 and MT expression in individual cells. The pivotal position of p53 in the protection of the g e n o m e after DNA damage has emerged, 99 and it is possible that MT has a related role in adaptation of cells to genomic instability. Studies are in progress to assess the relationship between p53 mutation and MT status in individual tumors.
HUMAN PATHOLOGY
Volume 26, No. 2 (February 1995)
Acknowledgment. We a c k n o w l e d g e Dr H. Navabi for technical assistance, Dr M. M o r g a n for assistance in the statistical analysis o f the data, a n d M. S t a n t o n a n d N. G a r r a h a n for p r o d u c i n g the p h o t o m i c r o g r a p h s . REFERENCES 1. Schmid KW, Ellis IO, Gee JMW, et al: Presence and possible significance ofimmunocytochemicallydemonstrable metallothionein over-expression in primary invasive ductal carcinoma of the breast. Virchows Arch A. Pathol Anat Histopathol 422:153-159, 1993 2. Reid A, Pereira H, Galea M, et al: Metallothionein in human breast cancer. J Pathol 167:29, 1992 (abstr) 3. Fresno M, Wu W, RodriguezJM, et al: Localisation of metallothionein in breast carcinomas. An immunohistochemical study. Virchows Arch A Pathol Anat Histopathol 423:215-219, 1993 4. Elston CW: Grading of invasive carcinoma of the breast, in Page DL, Anderson TJ (eds) : Diagnostic Histopathology of the Breast. Edinburgh, Scotland, Churchill Livingstone, 1987, pp 300-311 5. Jasani B, Elmes ME: Immunohistochemical detection of metallothionein. Methods Enzymol 205:95-107, 1991 6. Kagi JHR: Overview of metallothionein. Methods Enzymol 205:613-626, 1991 7. Evering WE, Haywood S, Elmes ME, et al: Histochemical and immunohistochemical evaluation of copper and metallothionein in the liver and kidney of copper-loaded rats.J Pathol 160:305-312, 1990 8. Jasani B, Anstey A, Marks R, et al: Wild type p53 and metallothionein are expressed simultaneously in UV irradiated skin: A possible link to photocarcinogenesis. J Invest Dermatol 101:422, 1993
(abstr) 9. Zelger B, Schir M, Hittmair A, et al: hnmunohistochemically demonstrated metallothionein expression in malignant melanoma. Histopathology 23:25%263, 1993 10. Bier B, Douglas-Jones AG, Totsch M, et al: Immunofiistochemical demonstration of metallothionein in normal human breast tissue and malignant breast lesions. Breast Cancer Res Treat 30:213221, 1993 11. Bedrick AE, Ramasamy G, TchertkoffV: Histochemical determinations of copper, zinc, and iron in pigmented nevi and melanoma. AmJ Dermatopathol 13:575-578, 1991 12. Mulay IL, Roy R, Knox BE, et al: Trace metal analysis of cancerous and non-cancerous human tissues. J Natl Cancer Inst 47:111, 1971 13. DeJorge FB, GoesJS, Guedes JL, et al: Biochemical studies
222
on copper, copper oxidase, magnesium, sulphur, calcium and phosphorus in cancer of the breast. Clin Chim Acta 12:403406, 1965 14. Schwartz AE, Leddicote GW, Fink RW, et al: Trace elements in normal and malignant human breast tissue. Surgery 76:325-329, 1974 15. Margalioth EJ, Schenker JG, Chevion M: Copper and zinc levels in normal and malignant tissues. Cancer 52:868-872, 1983 16. Santoliquido PM, Southwick HW, Olwin JH: Trace metal levels in cancer of the breast. Surg Gynecol Obstet 142:65-70, 1976 17. Samuni A, Chevion M, Czapski G: Unusual copper-induced sensitization of the biological damage due to superoxide radicals. J Biol Chem 256:12632-12635, 1981 18. Thornalley PJ, Vasak M: Possible role of metallothionein in protection against radiation-induced oxidative stress. Kinetics and mechanism of its reaction with superoxide and hydroxyl radicals. Biochem Biophys Acta 827:36-44, 1985 19. Chubatsu S, Meneghini R: Metallothionein protects DNA from oxidative damage. BiochemJ 291:193-198, 1993 20. Le Beau MM, Diaz MO, Karin M, et al: Metallothionein gene cluster is split by chromosome 16 rearrangements in myelomonocytic leukaemia. Nature 313:709-711, 1985 21. Schmidt CJ, Hamer DH: Cell specificity and an effect of RAS on human metallothionein gene expression. Proc Natl Acad Sci U S A 83:3346-3350, 1986 22. Bellamy COC, McDonald C, Salter DM, et al: Noninvasive ductal carcinoma of the breast: The relevance of histological categorization. HUM PATHOL 24:16-23, 1993 23. Lagios MD, Margolin FR, Westdahl PR, et al: Mammographically detected duct carcinoma in situ. Frequency of local recurrence following tylectomy and prognostic effect of nuclear grade on local recurrence. Cancer 63:618-624, 1989 24. Poller DN, Galea MH, Locker AP, et al: Ductal carcinoma in situ of the breast: A new simplified histological classification. Association between cellular proliferation and c-erb B2 protein expression. Breast 1:164, 1992 (abstr) 25. Poller DN, Galea M, Locker A, et al: DCIS of the breast-A histological, cytometric and molecular classification. J Pathol 164:359A, 1991 (abstr) 26. Poller DN, Roberts EC, BellJA, et al: p53 Protein expression in mammary ductal carcinoma in situ. HUM PATHOL24:463-468, 1993 27. Hall PA, McKee PH, Menage H du P, et al: High levels of p53 protein in U-V irradiated normal human skin. Oncogene 8:203207, 1993 28. Kastan MB, Onyekwere O, Sidransky D, et al: Participation of p53 protein in the cellular response to DNA damage. Cancer Res 51:6304-6311, 1991 29. Lane DP: p53, Guardian of the genome. Nature 358:15-16, 1992
and by nuclear grade irrespective of architectnre. The results showed that MT staining was significantly greater in comedo (ducts with necrosis) DCIS (ID = 97) compared with n o n c o m e d o (ducts without necrosis) DCIS (ID = 56) (P = .05 by Mann Whitney Ustatistic)and that low cytological grade (ID = 50) was associated with less MT staining than was high cytological grade (ID = 92) (P = .05 by Mann Whitney Ustatistic). These observations thus are consistent with the previously observed association between MT positivity and more aggressive behavior in invasive duct carcinoma of the breast. HUM PATHOL 26:217--222. Copyright © 1995 by W.B. Saunders Company Key words: methallothionein, duct carcinoma of breast, duct carcinoma in sitn, immunohistochemistry, breast. Abbreviations: DCIS, duct carcinoma in sitn; ID, intensity distribution°
I n a r e c e n t study invasive d u c t c a r c i n o m a s of the breast exhibiting immunocytochemically detectable m e t a l l o t h i o n e i n (MT) e x p r e s s i o n were s h o w n to have a significantly p o o r clinical o u t c o m e i n t e r m s o f t u m o r free interval. 1 I n two f u r t h e r i n d e p e n d e n t studies cond u c t e d o n 99 a n d 79 u n s e l e c t e d s p e c i m e n s of b r e a s t c a r c i n o m a , a significant association was o b s e r v e d between i n c r e a s e d levels o f M T e x p r e s s i o n a n d tum o r type, local r e c u r r e n c e , a n d p a t i e n t l e n g t h o f survival. 2,3 T h e aim of this study was to e x a m i n e preinvasive d u c t a l m a l i g n a n t e p i t h e l i a l p r o l i f e r a t i o n s ( d u c t carcin o m a i n situ or DCIS) of the b r e a s t to d e t e r m i n e w h e t h e r or n o t M T is e x p r e s s e d i n these early lesions a n d w h e t h e r this e x p r e s s i o n a n d its level are r e l a t e d to previously d e s c r i b e d m o r p h o l o g i c a l p r e d i c t o r s o f clinical behavior, i n c l u d i n g a r c h i t e c t u r a l type, grade, a n d necrosis.
tween 50 and 65 years. A representative block from each specimen containing DCIS with background breast tissue was selected for immunocytochemical staining of MT and its correlation with morphological characteristics as assessed by observation of the material on an adjacent serial section stained with hematoxylin-eosin.
MATERIALS AND METHODS
Case Material Fifty-five specimens of screen-detected pure DCIS of breast were retrieved from the files of the Department of Pathology at the Llandough Hospital and University Hospital of Wales, Cardiff. The ages of the patients were within the range of the National Breast Screening Programme, ie, be-
From the Department of Pathology, University of Wales College of Medicine, and the Breast Test Wales Group, Cardiff, Wales; and the Department of Pathology, Universityof Munster/Westphalia, Germany. Accepted for publicationJune 28, 1994. Address correspondence and reprint requests to A. G. DouglasJones, PhD, FRCPA, Department of Pathology, University of Wales College of Medicine, Cardiff, South Wales, UK CF4 4XN. Copyright © 1995 by W.B. Saunders Company 0046-8177/95/2602-001355.00/0
217
Morphological Classification of Lesions Classification according to duct architecture. The architectural pattern of the DCIS was determined in the following way. If more than 70% of the structures present were of a single architectural type, the case specimen was considered to be in this category and three ducts of this architecture were assessed. If there was a mixture of two or more architectural types (<70% of any one type), ie, mixed type, then three structures of each type were scored separately. Using these criteria any one specimen could contain one or more types of DCIS. For the purpose of this duct by duct analysis, cytology was ignored and ducts were classified as "with necrosis" or "comedo" if they contained any degree of central necrosis (Table 1). Classification according to cytological grade. The tumor cytology was subjectively assessed using the same criteria as applied in routine grading of infiltrating breast carcinoma irrespective of duct architecture or necrosis. 4 Briefly, grade 1 lesions showed uniform, rounded, regular nuclei with uniform chromatin staining. Grade 2 lesions showed moderate nuclear pleomorphism and grade 3 lesions showed marked nuclear pleomorphism, coarse chromatin, and hyperchromasia. Nuclear grade of DCIS was generally uniform within each individual specimen, although one specimen contained structures containing epithelium of different grades; these were assessed for MT staining separately (Table 1). Classification according to extent of intraduct necrosis. A third type of analysis based on a case by case comparison was performed on the basis of categorizing the lesion or case as a whole as either pure comedo (>90% of ducts containing necrosis), DCIS with necrosis (DN+), or DCIS without necrosis (<10% of ducts containing necrosis) ( D N - ) (Table 1).
HUMAN PATHOLOGY
Volume 26, No, 2 (February 1995)
I m m u n o c y t o c h e m i c a l Staining of MTs Paraffin sections 5-#m thick were cut from each block, dewaxed, and incubated with 1:20,000 dilution of the monoclonal anti-MT (E9) applied overnight at 4°C. The monoclonal anti-MT antibody, E9, was used in its unfractionated ascites form. This was obtained as a generous gift from Professor John Kay, Department of Biochemistry, University College of Cardiff, Cardiff, Wales. This antibody was found by Jasani and Elmes 5 to be immunocytochemically reactive against a conserved epitope shared by the 1 and 2 isoforms of human, rat, and horse MT, as tested on formalin-fixed, paraffin-embedded tissues obtained from the species, s The antibody is now commercially available in its purified form from Dako (Dako Ltd, High Wycombe, UK). The details of the preparation of the original antibody are as previously described. 1 The tissue binding of the antibody was disclosed using a standard alkaline phosphatase antialkaline phosphatase technique employing new fuchsin as the substrate. Sections were counterstained with hematoxylin.
Analysis of Staining: Intensity Distribution Scoring of MT Staining Within an individual ductal structure containing malignant epithelium, the total percentage of positive cells was assessed. Then the percentages of weakly, moderate, and strongly staining cells were assessed, such that the sum of these categories equalled the overall percentage of positivity. A staining score was then calculated as follows: score (out of a maximum of 300) = 1 × percentage of weak (+), 2 X percentage of moderate (+ +), or 3 × percentage of strong (+ + +) positivity. The assessment was performed on a consensus basis by two pathologists (AGD-J and BJ) using a double-headed microscope. Initially, an overall decision on whether or not an individual specimen showed homogeneous staining was made. If all the ducts present showed similar levels of staining (homogeneous), three ducts were chosen at random for assessment. If staining appeared to be different in different ducts (heterogeneous), the three strongest staining ducts and the three weakest staining ducts were assessed and the corresponding mean values were calculated. An average value was calculated from these means to represent the overall score for the lesion. Statistical analysis was performed using nonparametric methods (Mann Whitney Ustatistic and NemenyiDunn analysis of variance).
RESULTS
FIGURE 1. Nonneoplastic background breast Iobule showing
no staining for MT in acinar or ductal epithelium but strong staining in myoepithelial cells. (Monoclonal antibody E9: alkaline phosphatase antialkaline phosphatase technique, Hematoxylin counterstain; original magnification x500.)
strongly positive. In s o m e specimens within an individual ductal structure adjacent individual cells varied f r o m b e i n g strongly positive to b e i n g entirely negative in a " m o s a i c " p a t t e r n (Fig 2). In o t h e r specimens there was variation in staining f r o m d u c t to duct, s o m e showing u n i f o r m l y strong a n d s o m e u n i f o r m l y weak staining a n d s o m e showing n o staining at all except for the p e r i p h e r a l myoepithelial cells (Fig 3). T h e semiquantitative intensity distribution (ID) scoring system thus was a d o p t e d as a m e a n s o f quantifying the staining a n d indicating its variability t h r o u g h s t a n d a r d e r r o r calculations. T h e overall results o f ID scoring are s u m m a r i z e d in Table 1.
I m m u n o s t a i n i n g Patterns of MTs
D u c t C a r c i n o m a In Situ With Different Types of A r c h i t e c t u r e
Generally, in the b a c k g r o u n d n o n n e o p l a s t i c breast tissue d u c t u l a r a n d acinar epithelium showed n o detectable MT staining. Metallothionein-positive staining c o u l d be seen in the nucleus, cytoplasm, or both. For the p u r p o s e s o f analysis in this study n o distinction was m a d e between n u c l e a r a n d cytoplasmic staining. Myoepithelial cells associated with b o t h b e n i g n epithelium a n d DCIS were consistently strongly positive for MT (Fig 1). For the p u r p o s e o f the analysis o f DCIS myoepithelial cell staining thus was ignored. S o m e d u c t structures showing epithelial proliferation were seen a n d the luminal cells showed either low levels o f MT expression or were negative, b u t n o f o r m a l assessment o f these structures was made. Malignant epithelium (DCIS) showed variable intensity o f staining f r o m negative to
T h e results o f analysis o f MT by architecture o f DCIS in individual ducts (ignoring cytology) are shown in Fig 4. T h e m e a n ID score (out o f a possible total o f 300) for ducts with necrosis ( c o m e d o ) was 94 +_ 58 (n = 35), solid 78 _+ 63 (n = 9), cribriform 57 + 57 (n = 19), a n d micropapillary 30 _+ 38 (n = 9). Differences between c o m e d o (ducts with necrosis) a n d micropapillary scores were significant (P = .05) a n d between come d o a n d cribriform scores were significant (P = .01) by N e m e n y i - D u n n analysis o f variance. W h e n data for solid, micropapillary, a n d cribriform were c o m b i n e d ( m e a n ID = 56), there was a significant difference between c o m e d o (ducts with necrosis) a n d n o n c o m e d o (ducts without necrosis) DCIS ( m e a n ID, 94; P = .05 by M a n n W h i t n e y U statistic).
218
METALLOTHIONEIN IN DCIS OF THE BREAST (Douglas-Jones et al)
J
FIGURE 3. Cribriform DCIS of low cytological grade in which there is no epithelial positivity, but surrounding myoepithelial cells show strong positivity. (Monoclonal antibody E9: alkaline phosphatase antialkaline phosphase technique; hematoxylin counterstain; original magnification ×250.)
form r o u n d e d nuclei (grade 1) had a mean ID score of 50 (+55; n = 21), DCIS with moderate nuclear pleomorphism (grade 2) had a mean ID score of 106 (+67; n = 15), and marked DCIS with nuclear pleomorphism
Fifty-FiveSpecimens of Pure Screen-Detected DCIS
TABLE 1.
FIGURE 2. Comedo type DCIS showing marked variation in MT staining in individual ceils. Some stain very strongly, whereas adjacent cells are unstained, (Monoclonal antibody E9: alkaline phosphatase antialkaline phosphatase technique, hematoxylin counterstain; original magnification ×250,)
Duct Carcinoma In Situ With and Without Necrosis The analysis based on the presence and extent of intraduct necrosis is shown in Fig 5. The mean ID score for pure c o m e d o ( > 9 0 % of ducts showing necrosis) was 107 (_+65; n = 18), for DCIS with necrosis (DN+; 10% to 90% of ducts showing necrosis) it was 77 (_+49; n = 18), and for DCIS without necrosis (DN; < 1 0 % of ducts showing necrosis) it was 49 (+59; n = 19). Scores of pure c o m e d o DCIS were significantly different from scores for DCIS without necrosis (P = .05 by Mann Whitney U statistic).
Mean ID Score (Maximum of 300) (SD) Architecture Comedo (with necrosis) Micropapillary Cribriform Solid Total assessments* Grade 1 2 3 Total assessmentst Necrosis DCIS (pure comedo >90%) DCIS + (necrosis (10% to 90%)) DCIS (no necrosis <10%) Total specimens
94 30 57 78
No. of Specimen/ Assessments
(58) (38) (57) (63)
35 9 19 9 73
50 (55) 106 (67) 8O (54)
21 15 20 56
107 (65)
18
77 (47) 49 (59)
18 19 55
* If there was a mixture of patterns within one lesion, three examples of each architecture were selected and scored separately. J- One specimen showed high (3) and low (1) grades in different ducts within a single lesion.
Duct Carcinoma In Situ of Different Grades Analysis of the data by grade is shown in Fig 6. Ductal carcinoma in situ composed of cells with uni219
HUMAN PATHOLOGY
Volume 26, No. 2 (February 1995)
200-
2000.05
0.1
NS
•••-
150t--
o
100-
t.. m
100n
50MP
CR
S
50-
C
FIGURE 4. Staining score for MT out of a maximum of 300 related to architectural type of DCIS (ignoring cytology) MP, micropapillary; CR, cribriform; S, solid; C, c o m e d o (ducts with necrosis).
1
I
!
2
3
FIGURE 6. Staining score for MT out of a maximum of 300
(grade 3) had a m e a n ID score of 80 (_+54; n = 20). In practice it was often difficult to assign a particular lesion to grade 2 or 3, whereas grade 1 lesions were clearly identifiable because of the regularity of their nuclei. If data on grade 2 and 3 lesions are c o m b i n e d (ID = 92) and c o m p a r e d with those of grade 1 lesions (ID = 50), the difference is significant (P = .05 by Mann Whitney U statistic). Overall the results indicate that there is m o r e MT expression in ducts with necrosis than in those without and in ducts with cytologically high grade DCIS. DISCUSSION
T h e r e are two electrophoretically separable subgroups or is.forms of MT in m a m m a l s referred to as 200,
150,
100
50 ¸
DN-
DN+
PC
FIGURE 5. Staining score for MT out of a maximum of 300 related to presence of necrosis. DN-, DCIS without necrosis (less than 10% of ducts show necrosis); DN+, DCIS with necrosis ( b e t w e e n 10% and 90% of ducts show necrosis); PC, pure c o m e d o DCIS (more than 90% of ducts show necrosis). 220
related to cytological g r a d e (irrespective of architecture).
MT 1 and MT 2. These i s . f o r m s show small differences in their binding affinity for metal ions but a p p e a r to have similar biochemical properties. T h e antibody used in this study recognizes a d o m i n a n t immunoreactive epitope shared by MT 1 and MT 2; at present it is not possible to distinguish between the i s . f o r m s immunocytochemically. 5 T h e significance of differential i s . f o r m expression is not known. Metallothioneins are not normally expressed constitutively in tissues in amounts detectable by immunocytochemistry, except for a few specific cell types, such as my.epithelial, renal, and thyroid epithelial cells. Metallothionein expression is readily and rapidly inducible in m a n y tissues by a wide variety of stress factors, such as exposure to high levels of heavy metal ions, lipopolysaccharide, interleukin-1, interferon, glucocorticoids, and ultraviolet light. ~s Metallothionein overexpression has b e e n detected immunocytochemically in a n u m b e r of h u m a n tumors and is associated with a worse prognosis in breast carcinoma 1-~ and in malignant m e l a n o m a . 9 T h e biological basis for this overexpression in tumors and the m e c h a n i s m of its influence on prognosis is not currently understood. T h e results of this study and a separate qualitative study of benign and proliferative lesions of breast epithelium show that immunocytochemically detectable MT expression is not an innate feature of background, presumably normal, ductal or glandular breast epithelium. Consistently strong staining was observed in m y . epithelial cells, but no lumenal cell positivity was seen in aden.sis, sclerosing aden.sis, or papillomas. In a small n u m b e r of specimens included in the qualitative study, the MT staining of the infiltrating duct carcin o m a was similar to that of the duct carcinoma in situ associated with it. 1° T h e findings of the present study show that the
METALLOTHIONEIN IN DCIS OF THE BREAST (DougLas-Jones et al)
degree of MT positivity in preinvasive breast duct carcin o m a varies with histological architecture, cytological grade, and presence and extent of lumenal necrosis. Metallothionein positivity in tumor cells may be the result of expression of either a metabolically more stable form of MT or increased synthesis over degradation leading to net accumulation of MT in cytoplasm a n d / or nuclei of the t u m o r cells. The normally present housekeeping MT is most probably b o u n d to zinc ions and has a high turnover disallowing any detectable level of accumulation. Metallothionein b o u n d to copper or cadmium ions is known to be more stable, could account for MT positivity, and may play a role in the generation of malignancy. Histochemically detectable copper, zinc, and iron have been f o u n d in malignant melanoma but not in benign intradermal nevi, reflecting the MT findings in melanoma, n Studies quantifying trace metal concentrations in benign and malignant tumors, including breast carcinomas, show increased levels of copper or zinc in malignant tumors when compared with corresponding benign tissue. 12-16 An in vitro model has been proposed implicating copper in the generation of hydroxyl free radicals that can damage DNA. Copper complexes are reduced by superoxide radicals or other reducing agents to the cuprous state, and then they react with hydrogen peroxide to form hydroxyl radicals. These radicals may cause double-stranded DNA breakage, which is not repairable by cellular mechanisms. ~7 T h e r e is evidence to suggest that cells u n d e r oxidative stress are likely to overexpress MT as a protective response to the DNA-damaging influence of hydroxyl radicals, ls'19 Thus, observational data and theoretical evidence link raised heavy metal levels with carcinogenesis, and the raised MT levels observed in breast cancer may reflect this involvement because MT is acting as the main binding protein for these metals. Increased MT synthesis in response to inflammatory stress mediated by glucocorticoids or cytokines is another explanation. T h e r e is at present no evidence to support this possibility. It is possible, on the other hand, that the tumor cells are responding to intracellular factors or that the MT synthesis is switched on nonspecifically because of a general loss of transcription control in neoplastic cells. At least two varieties of tumors have been shown to have a consistent gene defect associated with increased MT expression. 2°'2a Thus, it is clear that further studies are n e e d e d to examine the mechanism of MT overexpression in breast tumors before the basis for the increased malignant potential of MT-positive breast duct carcinoma can be understood. Previous studies on the subtypes of DCIS have shown that histological pattern and grade have clinical relevance. 22 The n e e d for strict definitions of criteria for subtype classification has been stressed. Invasive recurrence only followed high grade DCIS and most commonly followed c o m e d o DCIS. It was suggested that the term " c o m e d o " be applied to the high grade DCIS solid pattern irrespective of the a m o u n t of necrosis. 22 Other investigators have accepted some degree of intraduct necrosis within a given architectural type of 221
DCIS. 23 We attempted to overcome the problems of classification associated w i t h " m i x e d type" DCIS by performing a duct by duct analysis rather than by forcing a particular lesion into one or other diagnostic category. Overexpression of c-erb B2 oncoprotein has been demonstrated immunocytochemically in DCIS associated with large cell size, high proliferative index, and histological c o m e d o subtype. 24 Mixed type DCIS with central necrosis showed higher S-phase fraction and proliferative index compared with lesions of mixed type DCIS without necrosis. It has b e e n proposed that the presence of focal necrosis in DCIS represents an intermediate position with regard to cell proliferation rate between pure c o m e d o and small cell cribriform and micropapillary types. 24'25 This hypothesis is reflected in the results reported here in which expression of MT in DCIS with necrosis is intermediate between DCIS without necrosis and pure c o m e d o DCIS. These correlations indicate that MT expression may be linked to cell proliferation rate. The results presented here show that although DCIS of low grade and more "benign types" (micropapillary or cribriform) can express MT, it is at a lower level than that expressed in high grade or c o m e d o type DCIS, reflecting the previous findings in infiltrating duct carcinoma. 1 In the intraductal preinvasive lesions studied here, there is a wide variety of expression of MT among malignant cells in the same duct (Fig 2), giving ample scope for a clonal selection process to occur during progression of the tumor, if MT expression confers some biological advantage to those malignant cells containing it in abnormally large amounts. In the cases described here there has been no therapeutic intervention and so the variation in MT overexpression observed cannot be the result of drug-related induction or therapy-related selection pressure. Although immunocytochemistry does not allow distinction between wild type and mutant p53 expression, accumulation of immunoreactive p53 protein has been demonstrated in approximately 25% of pure DCIS and is associated with large cell size and increased morphological degree of necrosis. 26 These findings reflect the overexpression of MT reported here and imply a relationship between MT and p53 expression. T h e r e are other interesting parallels between p53 and MT because increased expression of both is seen after exposure of skin to ultraviolet light s'27 and both molecules appear to have a role in protection against certain forms of DNA damage. 2s'29 Interestingly, the pattern of p53 expression in large cell DCIS illustrated by Poller et a126 is remarkably similar to that seen in high grade, c o m e d o DCIS illustrated here with strongly positive cells and adjacent negative ones ("mosaic" pattern, Fig 2). Double immunolabeling will allow the investigation of the relationship between p53 and MT expression in individual cells. The pivotal position of p53 in the protection of the g e n o m e after DNA damage has emerged, 99 and it is possible that MT has a related role in adaptation of cells to genomic instability. Studies are in progress to assess the relationship between p53 mutation and MT status in individual tumors.
HUMAN PATHOLOGY
Volume 26, No. 2 (February 1995)
Acknowledgment. We a c k n o w l e d g e Dr H. Navabi for technical assistance, Dr M. M o r g a n for assistance in the statistical analysis o f the data, a n d M. S t a n t o n a n d N. G a r r a h a n for p r o d u c i n g the p h o t o m i c r o g r a p h s . REFERENCES 1. Schmid KW, Ellis IO, Gee JMW, et al: Presence and possible significance ofimmunocytochemicallydemonstrable metallothionein over-expression in primary invasive ductal carcinoma of the breast. Virchows Arch A. Pathol Anat Histopathol 422:153-159, 1993 2. Reid A, Pereira H, Galea M, et al: Metallothionein in human breast cancer. J Pathol 167:29, 1992 (abstr) 3. Fresno M, Wu W, RodriguezJM, et al: Localisation of metallothionein in breast carcinomas. An immunohistochemical study. Virchows Arch A Pathol Anat Histopathol 423:215-219, 1993 4. Elston CW: Grading of invasive carcinoma of the breast, in Page DL, Anderson TJ (eds) : Diagnostic Histopathology of the Breast. Edinburgh, Scotland, Churchill Livingstone, 1987, pp 300-311 5. Jasani B, Elmes ME: Immunohistochemical detection of metallothionein. Methods Enzymol 205:95-107, 1991 6. Kagi JHR: Overview of metallothionein. Methods Enzymol 205:613-626, 1991 7. Evering WE, Haywood S, Elmes ME, et al: Histochemical and immunohistochemical evaluation of copper and metallothionein in the liver and kidney of copper-loaded rats.J Pathol 160:305-312, 1990 8. Jasani B, Anstey A, Marks R, et al: Wild type p53 and metallothionein are expressed simultaneously in UV irradiated skin: A possible link to photocarcinogenesis. J Invest Dermatol 101:422, 1993
(abstr) 9. Zelger B, Schir M, Hittmair A, et al: hnmunohistochemically demonstrated metallothionein expression in malignant melanoma. Histopathology 23:25%263, 1993 10. Bier B, Douglas-Jones AG, Totsch M, et al: Immunofiistochemical demonstration of metallothionein in normal human breast tissue and malignant breast lesions. Breast Cancer Res Treat 30:213221, 1993 11. Bedrick AE, Ramasamy G, TchertkoffV: Histochemical determinations of copper, zinc, and iron in pigmented nevi and melanoma. AmJ Dermatopathol 13:575-578, 1991 12. Mulay IL, Roy R, Knox BE, et al: Trace metal analysis of cancerous and non-cancerous human tissues. J Natl Cancer Inst 47:111, 1971 13. DeJorge FB, GoesJS, Guedes JL, et al: Biochemical studies
222
on copper, copper oxidase, magnesium, sulphur, calcium and phosphorus in cancer of the breast. Clin Chim Acta 12:403406, 1965 14. Schwartz AE, Leddicote GW, Fink RW, et al: Trace elements in normal and malignant human breast tissue. Surgery 76:325-329, 1974 15. Margalioth EJ, Schenker JG, Chevion M: Copper and zinc levels in normal and malignant tissues. Cancer 52:868-872, 1983 16. Santoliquido PM, Southwick HW, Olwin JH: Trace metal levels in cancer of the breast. Surg Gynecol Obstet 142:65-70, 1976 17. Samuni A, Chevion M, Czapski G: Unusual copper-induced sensitization of the biological damage due to superoxide radicals. J Biol Chem 256:12632-12635, 1981 18. Thornalley PJ, Vasak M: Possible role of metallothionein in protection against radiation-induced oxidative stress. Kinetics and mechanism of its reaction with superoxide and hydroxyl radicals. Biochem Biophys Acta 827:36-44, 1985 19. Chubatsu S, Meneghini R: Metallothionein protects DNA from oxidative damage. BiochemJ 291:193-198, 1993 20. Le Beau MM, Diaz MO, Karin M, et al: Metallothionein gene cluster is split by chromosome 16 rearrangements in myelomonocytic leukaemia. Nature 313:709-711, 1985 21. Schmidt CJ, Hamer DH: Cell specificity and an effect of RAS on human metallothionein gene expression. Proc Natl Acad Sci U S A 83:3346-3350, 1986 22. Bellamy COC, McDonald C, Salter DM, et al: Noninvasive ductal carcinoma of the breast: The relevance of histological categorization. HUM PATHOL 24:16-23, 1993 23. Lagios MD, Margolin FR, Westdahl PR, et al: Mammographically detected duct carcinoma in situ. Frequency of local recurrence following tylectomy and prognostic effect of nuclear grade on local recurrence. Cancer 63:618-624, 1989 24. Poller DN, Galea MH, Locker AP, et al: Ductal carcinoma in situ of the breast: A new simplified histological classification. Association between cellular proliferation and c-erb B2 protein expression. Breast 1:164, 1992 (abstr) 25. Poller DN, Galea M, Locker A, et al: DCIS of the breast-A histological, cytometric and molecular classification. J Pathol 164:359A, 1991 (abstr) 26. Poller DN, Roberts EC, BellJA, et al: p53 Protein expression in mammary ductal carcinoma in situ. HUM PATHOL24:463-468, 1993 27. Hall PA, McKee PH, Menage H du P, et al: High levels of p53 protein in U-V irradiated normal human skin. Oncogene 8:203207, 1993 28. Kastan MB, Onyekwere O, Sidransky D, et al: Participation of p53 protein in the cellular response to DNA damage. Cancer Res 51:6304-6311, 1991 29. Lane DP: p53, Guardian of the genome. Nature 358:15-16, 1992