13th Ljudevit Jurak International Symposium on
Comparative Pathology
Zagreb, Croatia
June 7-8, 2002

juraks@kbsm.hr
conference paper in pdf format

Pathologists are part of the core team of breast units, together with surgeons and radiologists. This means that they do not only play a diagnostic role for breast cancer, as used in the past, since they are requested for pre- and postoperative role in planning the therapeutic approach. Meetings for planning treatment of single cases are of basic importance, and also as an occasion for other specialists to understand our problems of morphological interpretation. Such meetings are a stimulus so that pathologists can better afford the new requests imposed by a rapidly changing scenario. One of the most compelling requests is reproducibility, and to achieve it pathologists must confront with diagnostic guidelines, definitions and landmarks. Guidelines produced by a group established by the EU of pathologists from different European countries involved in breast cancer screening proved effective in favoring reproducibility, as expressed by K statistics. However, in specific areas such as classification and definition of in situ lesions, such as atypical hyperplasia, reproducibility is far from ideal. Participation in a breast unit means highlighting of specific steps, which proved essential in a collaborative work and, ultimately, for the fight against breast cancer. Important steps are: identification of microcalcifications, definition of margins, diagnosis of in situ and microinvasive lesions, detection of metastatic spread. Problems are in fact not only related to morphological interpretations, but to communication as well. Examples are the diagnostic definition of fine needle aspirates and core biopsies, where the use of five standard categories (inadequate, benign; atypic, probably benign; suspect for malignancy, malignant) has proved effective in producing a preoperative diagnosis in conjunction with radiological and clinical criteria. The management, either cytologic or histologic of axillary lymph nodes and especially of sentinel nodes is a new challenge of both scientific and diagnostic impact. Similarly, the evaluation of the prognostic and therapeutic parameters such as HER2 overexpression/ amplification poses new budgetary problems which have to be solved by pathologists not only through diagnostic skill, but on the basis of organization and collaboration as well. 
BACK

Benign and malignant soft tissue tumors of the breast will be presented. Lipomas are the most common benign soft tissue tumors of the breast. When glandular tissue is entrapped within a lipoma, the lesion qualifies as adenolipoma or adenohibernoma according to the type of adipose tissue. The first case of angiomyolipoma HMB 45 positive will be illustrated. A unifying concept concerning solitary fibrous tumors, myofibroblastomas and spindle cell lipomas will be presented. The term of benign spindle stromal cell tumor (BSST) will be proposed for these lesions that are amenable to a common precursor cell of the breast stroma that is vimentin and CD 34 positive. Accordingly, the concept of malignant spindle stromal cell tumors (MSST) will be presented along the line of the benign counterpart. Angiosarcomas will be discussed with emphasis on the Stewart Treves syndrome. 
BACK

Malignant lymphomas of the breast are a rare disease. They may occur as primary or secondary tumors. Morphologically, it is not possible to determine their primary or secondary nature1. The criteria for defining a lymphomatous lesion in the breast as primary were first proposed by Wiseman and Liao in 19722: 1) availability of adequate histologic material; 2) documentation of breast involvement as a primary site; 3) presence of breast tissue in or adjacent to lymphoma infiltrate; 4) no concurrent nodal disease except for the involvement of ipsilateral axillary lymph nodes; and 5) no previous history of lymphoma involving other organs or tissues. Strictly adhering to such criteria, some primary breast lymphomas may be lost since no allowances are made for those primary breast lymphoma cases that may present in higher clinical stages. Obviating this, some authors consider as primary breast lymphomas all those cases in which breast is the first or major site of presentation even though subsequent staging procedures reveal involvement of other sites, such as bone marrow3. 

Clinical Presentation
The lesion most commonly presents as a unilateral breast mass in postmenopausal women (median age 55 to 60 years), although it may occur at any age. In about 10% of cases, it is bilateral. In men, breast lymphoma is exceedingly rare. A subset of patients, characteristically from tropical Africa, are young women, during or immediately after pregnancy, who present with massive bilateral breast swelling. The latter disease is endemic in this part of the world. Histologic examination in these patients reveals Burkitt’s or Burkitt-like lymphoma4. Non-African cases of this type of lymphoma are also on record. The incidence of primary breast lymphoma ranges from 0.04% to 0.5% of breast malignancies in most published series 1,3,5 .The incidence of secondary lymphoma in the breast is difficult to ascertain, since many of those lesions are but one manifestation of disseminated disease and are never biopsied. 

Gross Features
Primary and secondary breast lymphomas usually present grossly as a well defined uni- or multinodular mass of soft or firmer white-gray tissue, sometimes with necrotic and hemorrhagic foci.The tumor varies in size and may attain up to 20 cm in diameter. 

Histopathology 
The vast majority of breast lymphomas are diffuse large cell B lymphomas as defined by recent WHO classification. The latter lymphomas were given different names in older classification schemes, such as reticulum cell sarcoma, histiocytic lymphoma, large cell cleaved or noncleaved lymphomas, centroblastic or immunoblastic lymphoma, etc. In addition to large cell B lymphomas, a variety of other types of lymphoma may also manifest as primary or secondary tumors in the breast. The relation of pre-existing mammary tissue and infiltrating lymphoma varies. In some cases, the bulk of the tumor is located in subcutaneous fatty tissue and breast parenchyma is at its periphery; in other cases, the ducts and lobules of breast tissue are embedded in the infiltrate, while in rare cases pre-existent tissue is overgrown by lymphoma and barely visible. In such cases, remnants of ducts and lobules may only be revealed by using keratin immunostaining. Stroma may be scant or more abundant, sometimes sclerotic and hyalinized. Although most of the tumors are grossly circumscribed, a microscopically different degree of infiltration of the surrounding tissue is always evident. 
Diffuse large cell B lymphoma1,3,5-8: This type of lymphoma is characterized by large lymphoma cells with oval or indented nuclei, with one to three nucleoli and with a narrow rim of basophilic cytoplasm; such cells generally resemble centroblasts. Different number of immunoblasts are frequently admixed, sometimes such cells are predominant. Mitoses are usually numerous. In some cases, cells appear more pleomorphic with wider variation in cell forms and sizes; smaller reactive lymphocytes of B or T types are also present in the infiltrate. Occasionally, reactive histiocytes are numerous, imparting a “starry sky” appearance to the tumor. Adjacent mammary tissue may exhibit lobular atrophy or lymphocytic lobulitis which may be very prominent (lymphocytic mastopathy). Lymphoma cells are immunoreactive for CD20, CD79a, CD45RB, and negative for CD3 and CD45RO. Cases with immunoblastic features may show light chain restriction. Exceptionally, lymphoma cells may express CD30 antigen. 
Follicular lymphoma 1,3,5,7,8: It features neoplastic follicles composed of centrocytes and centroblasts in different proportions and may be graded into 2 or 3 grades, depending on the number of centroblasts inside the neoplastic follicles. Immunohistochemically, the lymphoma cells show positivity for B cell antigens, and for CD10 and bcl- 2, and are negative for CD5 and CD23. Follicular dendritic cells in tight clusters, positive for CD21, delineate neoplastic follicles. Selective infiltration of ductal-lobu- lar units by lymphomas of other types, such as diffuse large cell B lymphoma may mimic neoplastic follicles and could be confounded for true follicular lymphoma. 
Burkitt’s lymphoma4,8: The infiltrate in this lymphoma is composed of medium-sized cells with round nuclei, multiple central nucleoli, coarse chromatin and rather thick nuclear membrane. The cytoplasm is moderate in amount, basophilic with fine vacuoles containing lipids. Mitoses are very numerous. Cells grow in a cohesive pattern, they square off with each other. Numerous tingiblebody macrophages are evenly scattered among lymphoma cells producing characteristic but in no way pathognomonic “starry sky” appearance of the lymphoma. The breast parenchyma is usually hyperplastic and secretory. Ki-67 fraction of viable cells is 100%. Immunohistochemically, pan-B markers are positive, surface immunoglobulins, usually of IgM type, are also positive, while CD5, bcl-2, and TdT are negative. EBV is frequently demonstrated in endemic but not in sporadic cases. IgH and IgL genes are rearranged. Burkitt-like lymphoma is similar in morphological appearance but immunoblast and centroblast-like cells are also admixed. 
Extranodal marginal-zone B-cell lymphoma of MALT type1,3,5 : An undetermined number of breast lymphoma cases belong to the category of MALT lymphoma. The breast is considered to be part of a common mucosal immune system9 and may, during an autoimmune process, acquire lymphoid tissue in lymphoma. Most recent series of breast lymphoma have some cases of MALT lymphoma included. Typically, lymphoma of this type is composed of small lymphocytes, monocytoid (marginal zone type) cells and plasma cells. The latter may dominate the whole microscopic aspect of the lesion. Larger blast type cells may also be present. The infiltrate may be vaguely nodular, reactive follicles may be seen, some of them colonized by monocytoid cells. The lymphoepithelial lesion, i.e. infiltration of the ductal/lobular epithelium by monocytoid (centrocyte-like) cells was originally overestimated as a diagnostic criterion for MALT lymphoma of the breast; its presence is not a prerequisite for diagnosis. Furthermore, it has become increasingly evident that breast epithelium is infiltrated by lymphoma cells of a variety of lymphomas and even more commonly by T reactive cells admixed to lymphomatous infiltrate. Immunohistochemically, MALT lymphoma cells express pan-B markers such as CD20 and CD79a, it is usually bcl-2 positive but negative for CD10, CD5, and CD23. The translocation t(11;18)(q21;q21) has been identified in many MALT lymphomas; analysis did not include breast cases10. The same holds true for recently described trisomy 3 identified in a number of MALT lymphomas 11. Breast may also be involved by secondary MALT lymphoma originating at another MALT site. Some other types of lymphoma may also rarely present in the breast, as primary or secondary lesions, including lymphoblastic lymphoma of either B or T type, extremely rarely peripheral T cell lymphoma, and secondary small lymphocytic lymphoma/CLL or mantle cell lymphoma.

Differential Diagnosis 
Large cell malignant lymphomas may, in certain instances, be misdiagnosed as poorly differentiated duct or lobular carcinoma. Immunohistochemical reactions for keratin resolve any possible dilemma in such cases. Myeloid cell tumors may also be confounded for malignant lymphomas; if basic immunoreactions are inconclusive, myeloperoxidase staining to exclude the former possibility should be used. Inflammatory conditions may mimic MALT lymphomas; in some cases immunohistochemical reactions, flow cytometry and molecular genetic analysis should be employed to determine clonality of the lesion. Reactive follicular hyperplasia can be differentiated from follicular lymphoma by using bcl-2 immunoreactions; reactive follicles are bcl-2 negative. The issue of so-called pseudolymphoma remains unresolved; many authors believe that they really represent MALT type lymphoma. 

Prognosis and Treatment 
Primary breast lymphoma behave in a similar way as lymphomas of corresponding types and stages in other localizations. Localized low-grade lesions, such as MALT type lymphomas, are treated locally by surgery and/or radiation; high grade tumors require systemic chemotherapy with or without irradiation. 

References 
1. MATTIA AR, FERRY JA, HARRIS NL. Breast lymphoma. A Bcell spectrum including low grade B-cell lymphoma of mucosa associated lymphoid tissue. Am J Surg Pathol 1993;17:574-87. 
2. WISEMAN C, LIAO KT . Primary lymphoma of the breast. Cancer 1972;29:1705-12. 
3. HUGH JC, JACKSON FI, HANSON J, POPPEMA S . Primary breast lymphoma. An immunohistologic study of 20 new cases. Cancer 1990;66:2602-11. 
4. SHEPHERD JJ, WRIGHT DH . Burkitt’s lymphoma presenting as bilateral swelling of the breast in women of child-bearing age. Br J Surg 1967;54:776-80. 
5. LAMOVEC J, JANCAR J. Primary malignant lymphoma of the breast. Lymphoma of the mucosa-associated lymphoid tissue. Cancer 1987;60:3033-41. 
6. ABBONDANZO SL, SEIDMAN JD, LEFKOWITZ M, TAVASSOLI FA, KRISHNAN J. Primary diffuse large B-cell lymphoma of the breast. A clinicopathologic study of 31 cases. Pathol Res Pract 1996;192:37-43. 
7. BOBROW LG, RICHARDS MA, HAPPERFIELD LC, ISAACSON PG, LAMMIE GA, MILLIS RR. Breast lymphomas: a clinicopathologic review. Hum Pathol 1993;24:274-8. 
8. LIN Y, GOVINDAN R, HESS JL. Malignant hematopoietic breast tumors. Am J Clin Pathol 1997;107:177-86. 
9. BIENENSTOCK J, BOFUS AD. Review: mucosal immunology. Immunology 1980;41:249-70. 
10. OTT G, KATZENBERGER T, GREINER A, KALLA J, ROSENWALD A, HEINRICH U, OTT MM, MULLER-HERMELINK HK . The t(11;18)(q21;q21) chromosome translocation is a frequent and specific aberration in low-grade but not high-grade malignant non-Hodgkin’s lymphomas of the mucosa-associated lymphoid tissue (MALT) type. Cancer Res 1997;57:3944-8. 11. WOTHERSPOON AC, FINN TM, ISAACSON PG. Trisomy 3 in low-grade B-cell lymphomas of mucosa-associated lymphoid tissue. Blood 1995;85:2000-4. 
BACK

Breast cancer is among the most common tumors affecting women. It is characterized by a number of genetic aberrations. Five to 10% of all cases are estimated to be inherited. The hereditary breast and ovarian cancer syndrome includes genetic alterations of various susceptibility genes, particularly BRCA 1 and BRCA 2. Breast tumors in patients with a germ-line mutations in the BRCA 1 and BRCA 2 gene have an increase in additional genetic defects compared with sporadic breast tumors. Accumulation of somatic genetic changes during tumor progression may follow a specific and more aggressive pathway of chromosome damage in these individuals. Recent advances in genomics and bioinformatics, particularly in DNA-sequencing approaches and DNA-chip technology are revolutionizing target identification of small molecules. Here we review some new findings in the function of BRCA 1 gene function. A major BRCA 1 downstream target gene is the DNA damage-responsive gene GADD 45. Induction of BRCA 1 triggers apoptosis through activation of c-Jun N-terminal kinase/stress-activated protein kinase ( JNK/SAPK). BRCA 1 interacts with the SWI/SNF complex which controls DNA structure. SWI/ SNF is a chromatin remodeling complex important in gene expression. New knowledge about the genetic portrait of breast tumor is coming from differential gene expression profiling using microarrays. Human genome studies as well as development of “DNA chips” provide a window for observing patterns of gene activity in cells, which will revolutionize cancer classification. Knowledge of the molecular characteristics of breast tumor has already made it possible to identify those breast cancer patients who could benefit from therapies that target these features. Progress in basic research in signaling provides the opportunity to attack signal-transduction targets involved in proliferation, survival, invasion, angiogenesis, metastasis and resistance. Exciting knowledge in breast cancer biology is rapidly accumulating in parallel with recent developments in rational selection and validation of relevant targets that provide unique opportunities for development of “intelligent” therapeutics.
BACK

SUMMARY - The completion of the human genome sequence provides unique opportunities to identify new molecular targets for a variety of diseased conditions, especially for neoplastic diseases. Breast cancer is an ideal disease for the implementation of the recently developed, sophisticated genomic technologies, which permit the study of expression of many genes or proteins simultaneously, an approach known as molecular profiling. This approach is considered a major step forward in the development of new drugs that are more effective and less toxic than the current generation of antitumor agents. In this paper, we briefly review the current and future genomics technologies, such as DNA microarrays and proteomics techniques, and their use in the identification of new molecular targets for the treatment of breast cancer. We also discuss the challenge associated with the development of bioinformatics tools to analyze the massive number of data points generated by these technologies. Proof of principle is now emerging, demonstrating that selective agents against abnormal or mutated gene products can indeed be useful in the treatment of cancer. However, despite heavy investment in genomics research by the pharmaceutical industry, the full impact of genomics on drug discovery has yet to be fully demonstrated. 
Key words: Genomics; Microarrays; Proteomics; Molecular target 

Introduction
The completion of the human genome sequence provides a quantum lead towards identifying new molecular targets for a variety of diseased conditions, especially for neoplastic diseases1,2. The repositories of genes and their regulatory sequences represent the starting point of a new challenge, understanding how the 30,000-40,000 genes present in the human genome and their protein products interact and function. In addition of providing the unique opportunity to better understand basic biology and to identify the molecular basis of diseases, the annotation of the human genome offers the promise of an increased rate of drug discovery and development. Breast cancer is a major health problem worldwide and consequently, a large amount of research effort has been focused on the molecular understanding of this disease3. The medical treatment of cancer still has many unmet needs4. The main curative therapies (surgery and radiation) are usually successful only at an early stage, and existing chemotherapeutic treatments are largely palliative. The majority of the current antitumor agents have been unveiled during screening in cytotoxicity assays, although some have also been designed to act on defined molecular targets. However, none of the established cancer drugs were developed in the light of a clear understanding of the molecular differences between neoplastic and normal cells. Breast cancer is an ideal disease for the implementation of the recently developed, sophisticated genomic technologies, which permit to study the expression of many genes or proteins simultaneously, an approach known as molecular profiling. This approach is considered a major step forward in the development of new drugs that are more effective and less toxic than the current generation of antitumor agents. In this paper, we will review the application of genomic technologies for the rational identification of new therapeutic targets for breast cancer. DNA Microarray and Molecular Transcription Profiling Molecular transcription profiling is the large-scale analysis of gene expression using DNA array technologies5,6. DNA microarrays have only been recently introduced to the scientific community7. Microarrays consist of rows and rows of microscopic spots, each of which contains an identical single-stranded polymeric molecule of deoxyribonucleotide (typically oligonucleotides or cDNAs, the probe) attached to a solid support, such as a glass slide or a miniature silicon chip. These arrays can accommodate up to tens of thousands of spots and can be used for high-throughput studies of genomic structure and studies of active gene expression. Figure 1 provides an illustration of these DNA arrays. These arrays use the principle of specific DNA base pairing, i.e. A-T and GC, to allow the large-scale analysis of mRNA abundance as an indicator of gene expression, to detect polymorphisms within a population, or to detect new genes, as unknown DNA sequences can be analyzed8. Over the past few years, a number of different commercially available array products have been introduced. Although most of these products remain relatively expensive, their cost is regularly decreasing, and these products should soon become affordable for most laboratories. Therefore, it is critical that most cancer scientists become familiar with this technology. Recently, customized or in-house microarrays have grown in popularity to help investigators focus on their particular area of interest without being distracted by the huge volume of data generated by some commercial microarrays9. It should be mentioned that, although the microarray technology is the most widely used technique for gene expression analysis, others are available, including serial analysis of gene expression (SAGE) , differential hybridization, differential display or GeneCalling®10. The application of arrays to genomic studies includes the search for single nucleotide polymorphisms (SNPs) and a powerful application of these studies is in the field of pharmacogenomics6. Because each individual has a slightly different genetic makeup, each will have a unique set of SNPs. SNPs are the most frequent form of genetic variation (~3 million/person or approximately 1 SNP/kb), are highly stable, and are relatively easy to identify. Programs to detect and map SNPs in the human genome are well underway with the ultimate aim of establishing a SNP map of the genome11. When SNP analysis is used in conjunction with analytical techniques, such as genetic-linkage mapping or association analysis, a genetic propensity for predisposition to disease, unique metabolism, or adverse events can be identified.Although these SNPs may not be the actual cause of disease, their utility lies in their potential to help predict how an individual may respond to a particular drug. Pharmacogenomics may, therefore, help identify at-risk patients prior to treatment and prevent adverse drug reactions. Furthermore, SNPs may be useful to predict whether a certain drug would be effective in a patient with a specific disease. Although most marketed drugs are efficacious in a vast majority of the patient population, pharmacogenomics offers an opportunity to resurrect drugs that have been discarded because of low efficacy or adverse effects in the entire patient population. In this regard, pharmacogenomics can undoubtedly contribute to a better design of clinical trials in the future. Molecular transcription profiling analyses have already profoundly enhanced our understanding of many diseases, including breast cancer5,12,13. Cell function can be best understood by determining the transcription level of all genes in the genome (the transcriptome). The next step for transcriptional analysis will be the rapid identification and evaluation of potential therapeutic targets. In cancer, the accumulation and combinatorial effects of abnormalities, driving the initiation and progression of cancer, result from mutations and/or changes in expression level of cancercausing genes14. Therefore, therapeutic agents that would target the key molecular abnormalities that lead to malignant progression, have the potential of being more selective than the current non-specific cytotoxic agents, and therefore, more efficacious and less toxic. Proof of principle is now emerging that these selective agents can indeed be useful for the treatment of cancer14. For instance, the HER- 2/neu oncogene is overexpressed in approximately 30 percent of breast cancers, and these tumors are more aggressive and somewhat more resistant to chemotherapy than those not overexpressing the oncogenes15. These observations have led to the development of a monoclonal antibody (Herceptin® or trastuzumab) against the extracellular domain of this receptor tyrosine kinase4. Several clinical trials have demonstrated an improved response rate, a prolongation of the time to disease progression, and an increased overall survival compared to the standard of care, demonstrating the power of this genes-to-drugs paradigm for drug discovery.Herceptin® was approved by the U.S. Federal Drug Administration (FDA) in 1998 for the treatment of HER-2-positive breast cancer. Other receptor tyrosine kinases are frequently overexpressed in cancer. Therefore, several epidermal growth factor (EGF) receptor tyrosine kinase inhibitors, such as ZD-1839 (Iressa®), are currently being developed for the treatment of various cancer types, and results of preclinical studies and preliminary clinical trials indicate that the EGF receptor is indeed a valid target for anticancer therapy16,17. Because microarray technologies examine the expression profile of thousands of genes simultaneously at the mRNA level, it is now possible to study the sum total differences in gene expression between normal and diseased cells. This clearly will lead to the identification of new subtypes of tumors and will not only help the pathologist provide a more refined biological-based diagnosis, but will also enable scientists to identify previously unrecognized therapeutic targets in a rapid and efficient manner9,13,18.

Proteomics 
A major limitation of transcript profiling is that transcriptional activity does not necessarily reflect the activity of the protein product of a particular gene. This is mostly due to variation in cellular location and to complex and versatile protein regulation mechanisms, such as context-dependent post-translational phosphorylation, sulphation and glycosylation5. In addition, assigning a role for a protein based on a gene sequence information is not always feasible, because gene sequence reveals little information about protein function and disease relevance19. Therefore, a recent focus has been shifted towards proteomics, a protein-based approach to provide functional and expression information for proteins on a genome-wide scale (the proteome). 
Currently, the proteomics tools consist mostly of electrophoresis or chromatography coupled with mass spectrometry19. Several new technologies have recently been introduced for high-throughput protein characterization and discovery, such as protein arrays and proteome-scale screens for generic enzyme activities (e.g., protease and phosphatase)19. Applying these technologies to various diseases, and to breast cancer in particular, can work in concert with genomic technologies to identify new potential therapeutic targets. The challenge facing proteomics is, however, enormous, since it is estimated that approximately 75 percent of proteins in multicellular organisms have, as of yet, no known cellular function. Furthermore, human genes are fairly complex, incorporating variable numbers of protein domains into sophisticated functional products, with further protein diversity provided by alternative splicing17. Finally, protein-based technologies are extremely low throughput and more challenging to develop compared to transcription profiling techniques. Laser Capture Microdissection Genomic and proteomic analysis of cells in their native environment can provide the most accurate picture of the alterations that occur in vivo during the disease state. In vitro, cells are not subject to the endocrine and paracrine signals that regulate their overall behavior. However, studying tissues is not an easy task, mostly because tissues are complex three-dimensional structures, composed of large numbers of perpetually interacting cell populations. In the case of breast cancer, neoplastic cells may only account for a small proportion of the tissue analyzed, and the overall genomic and proteomic analyses may be confounded by the presence of large numbers of non-neoplastic cells, such as fibroblasts, endothelial cells or macrophages. 
To overcome these confounding factors, powerful analytical algorithms have been developed to gauge the relative abundance of an unknown cell subpopulation within tissue samples. Such algorithms use known genes associated with particular subpopulations of cells as reference values to estimate the proportion of cancer cells, stromal cells and inflammatory cells5. An alternative to the use of these algorithms is the implementation of microdissection techniques. Laser capture microdissection (LCM) is a technology for rapid and easy procurement of a microscopic and pure cell subpopulation from its complex tissue milieu12. The advantage is that LCM permits the investigator to focus directly on the disease subpopulation or compare several subpopulations of tissue cells from the same patient’s sample. In addition, recent data suggest that LCM would allow quantitative gene expression analysis in formalin-fixed, paraffin-embedded tissues, allowing to take advantage of large numbers of archived pathological tissue specimens20. The disadvantages of LCM are that it is resource intensive and provides only limited amounts of cellular material to study, although the recent development of reliable amplification protocols has partially solved this problem12. LCM has been successfully applied in the study of breast cancer pathogenesis and the identification of potential therapeutic targets, and it is likely that this technology will become a necessary component of the drug discovery process, as well as the cancer biology laboratory. 

Target Validation
A critical step after the identification of a putative therapeutic target is to validate the target’s relevance to the disease process21. An important part of this validation step can be achieved through the development of appropriate preclinical animal disease models. In particular, scientists now have the ability to genetically manipulate the mouse through transgenesis and gene targeting to test hypotheses regarding gene function and their role in disease. These knockout or transgenic mouse models (an important part of functional genomics) provide a powerful tool to the gene-to-drug paradigm for drug discovery22. The literature contains an enormous number of examples where a genetically engineered mouse model has helped better define the relevance of a specific gene product in a disease model. For instance, p53 knockout mice rapidly develop neoplasms in various tissues, as seen in patients with the Li-Fraumeni syndrome who have germline mutations of p5323-25. Other molecular tools are also available for target validation in animal models, such as antisense oligonucleotides, ribozymes and neutralizing antibodies4,26. 
Another method used to validate a target for breast cancer is analysis of its expression in a large population of breast tumors by immunohistochemistry and/or in situ hybridization10. While such a validation step can still be achieved analyzing one slide at a time, this can be very tedious and laborious. Hence, the recent development of tissue microarrays to increase the throughput of the process27. Tissue microarrays consist of hundreds to thousands cylindrical tissue biopsies, ranging from 0.6 up to 4 mm in diameter, each from a different patient, all distributed on a single glass slide. The tissue microarray technology has been tested and validated in several cancer types, including breast cancer28. The data confirmed many of the clinicopathological correlation of gene amplifications or immunostaining reactions reported with conventional techniques on the basis of whole tumor analysis5. Full automation of tissue array creation and screening is being developed to expeditiously validate the large numbers of newly identified potential therapeutic targets for cancer treatment. 
Bioinformatics 
While the cancer biologist will definitely benefit from these new technologies, it is clear that one of the major current challenges is to develop informatics techniques to facilitate the processing and analysis of this large amount of data. Interpretation of data and the development of models that facilitate the understanding of specific biological phenomena have become priorities. A common characteristic of contemporary drug discovery projects is their increasing complexity compared to the past, where discovery efforts were largely dominated by chemistry and pharmacology. Genomics techniques have led to the creation of a new research discipline, called bioinformatics. 
Initially, the focus of bioinformatics was on the analysis, processing, and archiving of genomic sequence data. Because of the rapid progress of the large-scale genome sequencing projects culminating in a “first draft” of the human genome, bioinformatics is now moving from the genome to the transcriptome and proteome level, with the focus shifting from the evaluation and annotation of genomic sequence data to the analysis of actual gene products29. The use of DNA microarrays generates a massive number of individual data points, which must then be analyzed by using data mining tools sufficiently sophisticated to categorize all these data group them in a meaningful manner8. These analytical tools are aimed toward the hunting of potential drug targets, the deciphering of possible cellular pathways, and the generation of hypotheses regarding the potential roles of certain genes. Protein-focused bioinformatics efforts aim to better understand the cellular expression, post-translational modifications, family relationships, structures and functions of proteins, as well as to evaluate their potential as suitable drug targets29. 
Several private and public databases exist for genome mapping, nucleic acid and protein sequences, and protein structures. In particular, Internet resources provide invaluable information and make available databases with relevance to drug discovery and genomic technologies21. For instance, the DNA sequence information available in these public databases can be used to identify transcripts differentially expressed in normal breast epithelial cells and breast tumor cells30. Similar approaches can easily be adapted and applied to other tumor types with sufficient transcript sequences available in the public databases. 

Conclusion
In drug discovery, initial expectations of new technologies are often too high. Despite large-magnitude efforts and resource commitments to new technologies (such as high-throughput screening and combinatorial chemistry) in the last decade, our ability to produce high-quality drug candidates has not become significantly enhanced. Obviously, the long period of time it takes for novel drugs to reach the market makes it very difficult to assess the immediate impact of a technology. In the 1990s, the primary goal of genomics research in the pharmaceutical industry was to increase the number of identified molecular targets and to gain proprietary rights to use those targets31. This goal has partly been successful: many more targets have been identified, intellectual property abounds, and proof of principle that these targets are valid has been demonstrated. However, an increased productivity of the pharmaceutical industry has yet to be demonstrated. Will genomics have its expected impact on drug discovery? First, it is important to remember that these technologies add a substantial new level of complexity to the drug discovery process, which will require the formation of large multidisciplinary teams to properly integrate these tools in the existing drug discovery and development process. In addition, in the drug discovery and development process, a long road with many hurdles separate the identification of a relevant gene target and the regulatory approval of an innovative drug: is the new gene a drugable target? Can adequate pharmacokinetic properties be achieved? Does the drug modulate the target in patients? Are there side effects and if so, are they manageable? Is there significant therapeutic benefit? What biomarker can be used to quickly predict efficacy? Are combinatorial genomebased therapies necessary for efficacy? What is the appropriate patient population? While drug development has not become much easier or faster, the promise of a major impact of genomics on drug discovery is still largely intact: it is likely that the genomics technologies will provide scientists with a large new collection of molecular targets for the treatment of various diseases, including breast cancer. However, success will require the proper and effective use of complementary technologies, expertise and innovative design of clinical trials, and proper integration of these various aspects is likely to become the definitive competitive advantage to pharmaceutical companies. 

References 
1. International Human Genome Sequencing Consortium. Initial sequencing and analysis of the human genome. Nature 2001;409: 860-921. 
2. VENTER JC, ADAMS MD, MYERS EW, LI PW, MURAL RJ, SUTTON GG, et al. The sequence of the human genome. Science 2001;291:1304-51. 
3. HORTOBAGYI GN. Treatment of breast cancer. N Engl J Med 1998;339:974-84. 
4. GIBBS J. Mechanism-based target identification and drug discovery in cancer research. Science 2002;287:1969-73. 
5. LIOTTA L, PETRICOIN E. Molecular profiling of human cancer. Nat Rev Genet 2000;1:48-56.
6. GRAVES DL. Powerful tools for genetic analysis come of age. Trends Biotechnol 1999;17:127-34. 
7. SCHENA M, SHALON D, DAVIS RW, BROWN PO. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 1995;270:368-9. 
8. MAUGHAN NJ, LEWIS FA, SMITH V. An introduction to arrays. J Pathol 2001;195:3-6. 
9. ALIZADEH AA, ROSS DT, PEROU CM, Van de RIJN M. Towards a novel classification of human malignancies based on gene expression patterns. J Pathol 2001;195:41-52. 
10. PEALE FVJr, GERRITSEN ME. Gene profiling techniques and their application in angiogenesis and vascular development. J Pathol 2001;195:7-19. 11. BENTLEY DR. The human genome project: an overview. Med Res Rev 2000;20:189-96. 
12. SGROI DC, TENG S, ROBINSON G, LeVANGIE R, HUDSON JR, ELKAHLOUN AG. In vivo gene expression profile analysis of human breast cancer progression. Cancer Res 1999;59: 5656-61. 
13. LAKHANI SR, ASHWORTH A. Microarray and histopathological analysis of tumours: the future and the past? Nat Rev 2001;1:151-7. 
14. CLARKE PA, te POELE R, WOOSTER R, WORKMAN P. Gene expression microarray analysis in cancer biology, pharmacology, and drug development: progress and potential. Biochem Pharmacol 2001;62:1311-36. 
15. SLAMON DJ, CLARK GM, WONG SG, LEVIN WJ, ULLRICH A, McGUIRE WL. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987;235:177-82. 
16. ADJEI AA. Epidermal growth factor receptor tyrosine kinase inhibitors in cancer therapy. Drugs Fut 2001;26:1087-92. 
17. WORKMAN P, CLARKE PA. Innovative cancer drug targets: genomics, transcriptomics and clinomics. Expert Opin Pharmacother 2001;2:911-5. 
18. ALIZADEH AA, EISEN MB, DAVIS RE, MA C, LOSSOS IS, ROSENWALD A, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 2000;403 (6769):503-11. 
19. EDWARDS AM, ARROWSMITH CH, des PALLIERES B. Proteomics: new tools for a new era. Modern Drug Discov 2000;3: 36-44. 
20. SPECHT K, RICHTER T, MUELLER U, WALCH A, WERNER M, HOEFLER H. Quantitative gene expression analysis in microdissected archival formalin-fixed and paraffin-embedded tumor tissue. Am J Pathol 2001;158:419-29. 
21. SAWYER TK. Deciphering therapeutic targets. Biotechniques 2001;30:1086-90. 
2. WEST DB, IAKOUGOVA O, OLSSON C, ROSS D, OHMEN J, CHATTERJEE A. Mouse genetics/genomics: an effective approach for drug discovery and validation. Med Res Rev 2000;20: 216-30. 
23. DONEHOWER LA, HARVEY M, SLAGLE BL, McARTHUR MJ, MONTGOMERY CAJR, BUTEL JS et al. Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumors. Nature 1992;356:215-21. 
24. HARVEY M, McARTHUR MJ, MONTGOMERY CAJr, BUTEL JS, BRADLEY A, DONEHOWER LA. Spontaneous and carcinogen-induced tumorigenesis in p53-deficient mice. Nat Genet 1993;5:225-9. 
25. VARLEY JM, EVANS DG, BIRCH JM. Li-Fraumeni syndrome: a molecular and clinical review. Br J Cancer 1997;76:1-14. 
26. TAYLOR MF. Target validation and functional analyses using antisense oligonucleotides. Expert Opin Ther Targets 2001;5:297-301. 
27. KONONEN J, BUBENDORF L, KALLIONIEMI A, BARLUND M, SCHRAML P, LEIGHTON S, et al. Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med 1998;4:844-7. 
28. HEISKANEN M, KONONEN J, BARLUND M, TORHORST J, SAUTER G, KALLIONIEMI A, et al. CGH, cDNA and tissue microarray analyses implicate FGFR2 amplification in a small subset of breast tumors. Anal Cell Pathol 2001;22:229-34. 
29. BAJORATH J. Rational drug discovery revisited: interfacing experimental programs with bio- and chemo-informatics. Drug Discovery Today 2001;6:989-95. 
30. LEERKES MR, CABALLERO OL, MACKAY A, TORLONI H, O’HARE MJ, SIMPSON AJG, et al. In silico comparison of the transcriptome derived from purified normal breast cells and breast tumor cell lines reveals candidate upregulated genes in breast tumor cells. Genomics 2002;79:257-65. 
31. WARD SJ. Impact of genomics in drug discovery. Biotechniques 2001;31:626-34.
BACK

The development and spread of malignant tumors is a multi-step process, involving a variety of alterations in the mechanisms controlling cell proliferation, differentiation and genetic alterations. Understanding of the biological process involved in tumorigenesis has practical application in the clinical areas of diagnosis, prognosis and treatment1,2. The goal of clinicians managing patients with malignancy is to create therapy to give maximum benefit to each individual patient. The decisions are usually based in part on predictors of the likely biological behavior of a given tumor. The major tumor characteristics in breast cancer known to be of value in prognosis include tumor size, tumor histologic type and grade, axillary lymph node status, steroid hormone receptor status, ploidy, and cell kinetics3,4. Establishing prognosis based on these parameters is successful to a large extent, but they still fail to accurately predict the clinical course of all patients. Therefore, the search for better means of integrating prognostic data and for new prognostic markers in breast cancer patients still remains a major goal. 
The human epidermal growth factor receptor-2 (HER-2) protooncogene encodes 185 transmembrane glycoprotein, often simply calledHER-2/neu or c-erbB- 2 protein receptor. In vitro and animal studies have indicated that HER-2/neu gene amplification and protein overexpression play a role in oncogenic transformation, tumorigenesis and metastasis. Furthermore, the growth of tumors and human breast carcinoma cell lines overexpressing HER-2/neu receptor is inhibited by anti- HER-2/neu monoclonal antibody, opening a new avenue for targeted cancer therapy. In 1987, Slamon et al.6 first reported a significant relationship between amplification of the HER-2/neo oncogene and adverse clinical outcome in patients with breast cancer. Although subsequent studies have largely confirmed this association in patients with node positive disease, whether or not HER-2/neo gene amplification or overexpression is an independent prognostic factor in patients with node-negative breast cancer remains a matter of controversy5-10. Although still somewhat controversial, the majority of clinical studies suggest that HER-2/neu is amplified and overexpressed in approximately 20%-30% of breast carcinomas, and that among the new biological indicators of tumor aggressiveness it is potentially useful in predicting the outcome of patients with breast carcinoma and can be used effectively to improve the identification of high-risk patients6,7,11,12.It is generally accepted that HER-2 overexpression is associated with shorter overall survival, low level of ER, and higher tumor grade.However, considerable variation in the incidence of amplification/overexpression and prognostic significance of HER-2/neu has been reported. Some investigators found amplification in only 10% of patients and no correlation to clinical outcome13, whereas others found overexpression in up to 50% of patients and a strong association with outcome4-14. Indeed, in our study the incidence of HER-2/neu overexpression was 42.7% with some marginal association with outcome in univariate analysis (p=0.059) and no significant influence on survival in multivariate analysis15. Conflicting results of numerous studies, including our study, highlight some of the persisting controversies surrounding the use of HER-2/neu as a prognostic marker. Also, these results emphasize the importance of considering HER2/neu status in the light of information provided by other prognostic variables. For this reason, we tried to assess the prognostic significance of HER-2/neu overexpression in association with other known prognostic factors, and showed association with tumor size (p=0.041) and grade (p=0.037), DNA ploidy (p=0.046) and cathepsin D expression in stromal macrophages (p=0.024). These findings pointed to HER2/neu overexpression as an indicator of prognosis in grade II breast carcinoma, suggesting that determination of both tumor size and DNA ploidy in combination with HER- 2/neu overexpression appear to enhance the ability to recognize the patients at different risk15. More recently, there has been considerable interest in the potential role of HER-2/neu gene amplification and overexpression as a predictor of response to various therapeutic modalities in patients with breast cancer. In particular the results of recent clinical trials have indicated that treatment with monoclonal antibody to HER-2/neu protein (Herceptin®) may be useful in prolonging the survival of patients with metastatic disease. In our small group of 17 advanced breast carcinoma patients who were enrolled in the study from the beginning of 1999 until July 2000 at the Zagreb University Hospital Center, all were treated with Herceptin® and Taxol® in combination. All patients had tumors positive for HER2/neu by HercepTest. Partial response to therapy was observed in 47%, stable disease in 29% and progression of disease in 24% of patients. 
However, complete response was not observed in the investigated group of patients. In this way, the overall therapy benefit (stable disease and partial response) was found in 76% of patients. Complications of therapy include neutropenia, thrombocytopenia, diarrhea and onycholysis, but there were no signs of congestive heart failure16. Furthermore, some studies have indicated that tumors with HER-2/neu overexpression may show resistance to certain forms of cytotoxic therapy and sensitivity to others. Finally, some recent experimental and clinical studies have suggested that HER-2/neu overexpression is associated with resistance to tamoxifen even when tumors were ER positive, and therefore the success of Herceptin® therapy depends upon the selection of the most appropriate patients for treatment. Candidates for Herceptin® therapy can be identified by the evaluation of tumor cells for the presence of altered HER-2/neu. As a result of this information, there is a growing clinical demand for HER-2/neu analysis of current and archived breast cancer specimens. There are a variety of methods available to determine the HER-2/neu status of breast cancers. These include assays to evaluate: 1) gene amplification as Southern blot, slot blot, dot blot, polymerase chain reaction (PCR), in situ hybridization and fluorescence in situ hybridization (FISH), 2) assays to determine mRNA overexpression such as Northern blot analysis, slot blot, and in situ hybridization, and 3) methods to assess protein overexpression (Western blot analysis, immunoassay, and immunohistochemistry (IHC). Many of these methods are beyond the scope of most pathology laboratories for technical reasons. Furthermore, most of these assays require prospective collection of fresh tissue and are not applicable to archival material. The IHC method performs well and gives a clear picture of the heterogeneity ofprotein expression in tumor cells; it distinguishes tumor cells from normal cells and is easy to perform on routine paraffin-embedded material. HercepTest is a semi-quantitative IHC assay to determine HER-2/ neu protein overexpression in breast cancer tissues routinely processed for histologic evaluation. For the determination of HER-2/neu protein overexpression, only the membrane staining intensity and pattern should be evaluated using scale 0-3+. Even if the standardized HER-2/ neu protocol is followed, the subjective nature of histologic scoring can lead to spurious results. The need to distinguish between “faint” and “weak” staining is subject to bias even when positive control tissues are included in each stain run.Based on the problems associated with IHC (i.e. lack of standardization and subjective bias), some experts argue that FISH is a better alternative. FISH measures HER-2/neu gene amplification but tends to be more expensive, complicated, and time-consuming than IHC, and HER-2/neu must be scored on the invasive component of breast cancer (because a high % of breast in situ lesions have altered the HER-2/neu status). 
Although each of these methods has its advantages and disadvantages, direct comparisons of these two assays have been few and are limited by small numbers of cases. Utility of IHC versus FISH for the selection of breast cancer therapy requires thorough analysis of the results of clinical trials now underway that address this issue. Any correlative study comparing IHC with FISH without statistically significant outcome data of patients treated with Herceptin® is of limited value for resolving the IHC- FISH controversy. Wang et al.17correlated IHC results with FISH over the past 18 months. These results shoved excellent correlation in over 98 percent of HercepTestnegative cases (0, 1+) lacking HER-2/neu gene amplification, and in tumors with high expression (3+) demonstrating amplification. A poor correlation was found between cases considered to be weakly positive (2+) with the HercepTest and amplification with FISH. However, only 11 percent of all breast cancer cases demonstrated 2+ IHC staining. Based on these results, the authors suggest that clinically and economically, significant value is testing for HER-2/neu by HercepTest (IHC) with reflex to FISH only in cases of weakly positive results (2+). The cost of reflex testing of all 0 or 1+ cases is significant, since this group comprises up to 88 percent of all breast cancer cases. Controversy still exists as to whether the cases that show 0 or 1+ staining but demonstrate amplification (two percent of all cases examined) will respond to Herceptin. 
In our hands also HercepTest provides excellent reproducibility and standardization, and close to 90% of all breast cancers can be adequately and reliably determined by IHC, with molecular evaluation reserved for borderline cases. Our own findings in 45 patients tested by IHC and FISH for HER-2/neu detection showed excellent correlation between IHC HER-2/neu analysis and the molecular technique for HER-2/neu amplification (FISH), except for the weakly positive (2+) IHC results as determined by the FDA-approved HercepTest. By IHC 12/45 (26.6%) were HER-2/neu positive. Six out of seven IHC high positive specimens (3+) showed gene amplification by FISH (85.7%), and 3/5 IHC medium positive specimens (2+) showed no gene amplification (60%). None of the cases negative by IHC showed expression of HER-2/neu by FISH. Concordances between FISH and IHC results were seen in 42 out of 45 cases (93.5%). We conclude that all 2+ IHC cases can in turn be subject to FISH analysis to confirm the presence of an altered HER-2/neu gene. This combination of assays would help ensure that patients who are most likely to benefit from Herceptin are identified. FISH testing is used to determine HER-2/neu status in equivocal circumstances, but the use of HercepTest as a screening assay allows for improved cost control and turnaround time without detriment to the patient. 
It is known that the FISH procedure required more technician time and more interpretation time per case for the pathologist than IHC. Reagent costs were subsequently higher for FISH than for IHC. There is a high level of correlation between FISH and IHC in the evaluation of HER-2/neu status of breast cancers using formalin- fixed, paraffin-embedded specimens, although the choice of which assay to use should be left for individual laboratories to make based on technical and economic considerations. The results published to date may make it difficult to justify the routine use of FISH for detection of HER2 status in breast cancers18-20. It is generally accepted, at this level of knowledge that the best approach is to combine both IHC and FISH assays, and to use the IHC assay as a triage step, followed by FISH to analyze the IHC medium and high positive cases18-20. 

References 
1. FEARON ER, VOGELSTEIN B. A genetic model for colorectal tumorigenesis. Cell 1990;61:759-67. 
2. THOMPSON AM, STEEL CM, CHETTY U, CARTER DC. Evidence for the multistep theory of carcinogenesis in human breast cancer. Breast 1992;1:29-35. 
3. ELLEDGE RM, MCGUIRE WK, OSBORNE CK. Prognostic factors in breast cancer. Semin Oncol 1992;19:244-53. 
4. MILLER WR, ELLIS IO, SAINSBURG JRC, DIXON JM. Prognostic factors - ABC of breast diseases. BMJ 1994;309:1573- 6. 
5. CLARK GM, WONG SG. Human breast cancer: correlation of relapse and survival with amplification of the HER2/neu oncogene. Science 1987;235:177-82. 
6. SLAMON D, CLARK GM, WONH SH, LEVIN WJ, ULLRICH A, MCGUIRE WL. Human breast cancer: Correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987;235:177-82. 
7. BORG A, TANDON AK, SIGURDSSON H, et al. HER2/neu amplification predicts poor survival in node-positive breast cancer. Cancer Res 1990;50:4332-7. 
8. LOVEKIN C, ELLIS IO, LOCKER A, et al. c-erbB-2 Oncoprotein expression in primary and advanced breast cancer. Br J Cancer 1991;63:439-43. 
9. QUENEL N, WAFFLART J, DONICHON F, et al. The prognostic value of c-erbB-2 in primary breast carcinomas: a study on 942 cases. Breast Cancer Res Treat 1995;35:283-91. 
10. ROSEN PP, LESSER ML, ARROYO CD, et al. Immunohistochemical detection of HER2/neu in patients with axillary lymph node negative breast caracinoma: a study of epidemiologic risk factors, histologic features, and prognosis. Cancer 1995;75:1320-6. 
11. REVILLION F, BONNETERRE J, PEYRAT JP. ERBB2 oncoprotein in human breast cancer and its clinical significance. Eur J Cancer 1998;34:791-808. 
12. PRESS MF, BERNSTEIN L, THOMAS PA, et al. HER2/neu gene amplification characterized by fluorescence in situ hybridization: poor prognosis in node-negative breast carcinoma. J Clin Oncol 1997;15:2894-904. 
13. ZHOU DJ, AHUJA H, CLINA MJ. Proto-oncogene abnormalities in human breast cancer: c-erbB-2 amplification does not correlate with recurrence of disease. Oncogene 1989;4:105-8. 
14. NOGUCHI M, KAWASAKI N, NAGAYOSHI O, et al. C-erbB- 2 oncoprotein expression versus internal mammary lymph node metastases as additional prognostic factors in patients with axillary lymph node-positive breast cancer. Cancer 1992;69:2953-60. Acta clin Croat 2002; 41:135-186 Conference papers 148 Acta clin Croat, Vol. 41, No. 2, 2002 
15. JAKI.-RAZUMOVI. J, PETROVE»KI M, UAAREVI. B, GAMULIN S. Mutual predictive value of c-erbB-2 overexpression and various prognostic factors in ductal invasive breast carcinoma. Tumori 2000;86:30-6. 
16. MRSI. M, GRAGI. M, BUDI©I. Z, PODOLSKI P, BOGDANI. V, LABAR B, JAKI.-RAZUMOVI. J, RESTEKSAMARAIJA N, GO©EV M. Trastuzumab in the treatment of advanced breast cancer: single-center experience. Ann Oncol 2001;12 (Suppl):95-6. 
17. WANG S, SABOORIAN MT, FRENKEL E, HYNAN L, GOKASLAN ST, ASHFAQ R. Laboratory assessment of the status of HER2/neu protein and oncogene in breast cancer specimens: comparison of immunohistochemistry assay with fluorescence in situ hybridization assays. J Clin Pathol 2000;53:374-81. 
18. JACOBS TW, GOWN AM, YAZIJI H, BARNES MJ, ASHNITT SJ. Comparison of fluorescence in situ hybridization and immunohistochemistry for the evaluation of HER2/neu in breast cancer. J Clin Oncol 1999;17:1974-82. 
19. DOWSETT M, COOKE T, ELLIS IO, GULLICK WJ, GUSTERSON B, MALLON E, WALKER R. Assessment of HER2 status in breast cancer: why, when and how? Eur J Cancer 2000; 36:170-6. 
20. HENDRICKS JB. Histopathology at the trailing edge (editorial). J Histotechnol 2000;23:297.
BACK

Carcinoma of the breast is the most common malignant tumor in women in the western world. Its incidence has been rising steadily over the past decades. Despite considerable progress in early diagnosis and treatment, mortality remains relatively high and approximately every third or second woman with breast cancer will ultimately die of the disease. 
TNM stage is generally accepted as the most important determinant of outcome in breast cancer. In addition to hormonal status of the patient (pre- or postmenopausal) and the presence of steroid receptors in the tumor, TNM stage has long been the only factor used by clinicians to make therapeutic decisions. 
During the past two decades, there has been a gradual change in the treatment of breast cancer with a tendency toward less radical surgery and more adjuvant systemic therapy. In the early era of chemotherapy, this type of treatment was mainly used in patients with advanced disease; however, meta-analyses of large clinical trials have clearly shown that systemic chemotherapy and hormonal therapy also reduce the risk of cancer recurrence and mortality in patients with early breast cancer. Accordingly, the question of who should be treated by systemic therapy has eventually changed to the question of who should not be so treated. 
Due both to more widespread public education and to early diagnosis by mammography screening programs, the percentage ofpatients with node negative breast cancer (N0) is increasing. The majority of these patients (about 70%) do not experience disease recurrence after surgery and/or radiotherapy alone (local therapy); therefore, it seems inappropriate to suggest systemic adjuvant therapy for all node negative breast cancer patients. Since the relative risk reduction is constant across different tumor stages and risk groups, it is obvious that the absolute benefit from adjuvant therapy may be quite small in some specific low-risk group of patients. It is reasonable to attempt to avoid excessive treatment morbidity and cost by using selective prognostic markers to identify prognostically relevant subsets of patients. However, an “ideal” prognostic indicator or a widely accepted combination of markers able to identify patients at low versus high risk has yet to be clearly defined. Numerous studies have reported that several clinicopathologic features have prognostic importance in nodenegative breast carcinoma patients. These features include tumor size, histologic grade, and histologic type of tumor, vascular invasion and some new biological markers. Only a few studies have simultaneously evaluated the relative prognostic weight of various newer biological markers compared with all conventional clinicopathological markers by performing a multivariate analysis. Also, the majority of these studies included heterogeneous group of patients regarding tumor size. 
Recently, an International Consensus Panel proposed a 3-tiered risk classification for patients with negative axillary lymph nodes (Table 1) and defined the low-risk group as those with tumor size of 1 cm or less, positive ER or PR status and histologic grade 1; they suggested that this is the only group in which adjuvant systemic therapy could be omitted. Since this proposal is not based on multivariate analyses of large data sets, its true prognostic value remains uncertain. 
Although the patients with stage I (T1N0M0) breast carcinoma (i.e. tumor measuring 2 cm or less without regional lymph node and distant metastasis) have an excellent short term prognosis, approximately 20% will eventually develop distant metastases and die of the disease. However, the remaining majority would be cured by surgery alone and gain no benefit from adjuvant systemic therapy. Unfortunately, there is no general agreement how to best identify the latter group. Less than 15 published studies evaluated the prognostic value of different factors in T1N0M0 breast carcinoma and only few of them included the newer biologic indicators, such as c-erbB-2, p53, bcl2, Ki67 (MIB-1), flowcytometric DNA ploidy or S-phase fraction determination. Although the results of these studies are somewhat controversial, histologic or nuclear grade, proliferative activity (as assessed by either mitosis counting, SPF determination or Ki67 expression), vascular invasion and tumor size most often emerged as significant prognostic factors. We recently investigated a group of 270 patients with T1N0M0 breast carcinoma who were treated at the Institute of Oncology Ljubljana and were followed for a median of 12.5 years. All original slides were reviewed and examined for histologic type, mitotic index (MI), Nottingham histologic grade (NHG) and its components (extent of tubule formation, pleomorphism, mitotic counts) and presence of vascular invasion. Representative tumor slides were stained immunohistochemically for p53, bcl-2, c-erbB-2, MIB-1(Ki67), CEA, ER and PR using LSAB method and Dako TechMate 500 automatic immunostainer. The prognostic value of investigated features was evaluated using univariate and multivariate survival analysis. Survival of our patients (84.4% cancer-specific survival, CSS, and 77.4% metastasis-free survival, MFS, at 10 years) was similar to that reported in other series of patients with T1N0M0 tumors. In keeping with other reports, late recurrences were not uncommon. The prognostic value of tumor size was not confirmed in our study: although survival was somewhat worse in patients with tumors larger than 1 cm (T1c) than in those with tumors measuring 1 cm or less (T1ab), the difference was not significant. We confirmed the prognostic significance of NHG: both MFS and CSS were significantly better in patients with grade 1 than in those with grade 2 or 3 tumors. Apart from NHG, in univariate analysis, MI, vascular invasion and c-erbB-2 expression were significant predictors of MFS and CSS. In addition, CEA expression and MIB-1 reactivity were significantly related to MFS, and histologic type to CSS. Age, menopausal status, type of treatment, PR or ER status and expression of bcl-2 or p53 were not significantly associated with survival. The relative importance of prognostic variables was tested in Cox’s proportional hazard model. When all variables were entered in the model, MI, histologic type, vascular invasion and CEA expression emerged as significant independent prognostic factors for both MFS and CSS. 
MI was the single most important prognostic factor for MFS and CSS; however, our findings suggest that optimal cutoff values for different prognostic groups may be lower than those proposed in NHG. When testing various multivariate models to predict CSS, NHG retained its independent prognostic value only in the model that did not include MI and histologic type, whereas it was replaced by MIB-1 reactivity in multivariate analysis of MFS. 
By combining four independent factors (MI, histologic type, vascular invasion and CEA expression) into a prognostic index, the patients could be allocated into three prognostic groups. Patients in the first group (15%) developed metastatic disease in almost one half of cases and those in the second (60%) in one third of cases. In the third group (25%), prognosis was excellent, with more than 90% MFS at 15 years after surgery. In the latter group, the use of adjuvant chemotherapy may be unnecessary. By applying the aforementioned Consensus Panel criteria, the group of patients in whom adjuvant systemic therapy could be omitted would be considerably smaller. 

References
1. EIFEL P, AXELSON JA, COSTA J, et al. National Institutes of Health Consensus Development Conference Statement: adjuvant therapy for breast cancer, November 1-3, 2000. J Natl Cancer Inst 2001; 93:979-89. 
2. Polychemotherapy for early breast cancer: an overview of the randomised trials. Early Breast Cancer Trialists’ Collaborative Group. Lancet 1998; 352:930-42. 
3. Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy. 133 randomised trials involving 31,000 recurrences and 24,000 deaths among 75,000 women. Early Breast Cancer Trialists’ Collaborative Group. Lancet 1992; 339:1-15. 
4. Early stage breast cancer: consensus statement. NIH Consensus Development Conference, June 18-21, 1990. Cancer Treat Res 1992; 60:383-93. 
5. GOLDHIRSCH A, WOOD WC, SENN HJ, et al. Meeting highlights: international consensus panel on the treatment of primary breast cancer. J Natl Cancer Inst 1995; 87:1441-5. 
6. YARBRO JW, PAGE DL, FIELDING LP, et al. American Joint Committee on Cancer Prognostic Factors Consensus Conference. Cancer 1999; 86:2436-46. 
7. FITZGIBBONS PL, PAGE DL, WEAVER D, et al. Prognostic factors in breast cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 2000; 124:966-78. Acta clin Croat 2002; 41:135-186 Conference papers 150 Acta clin Croat, Vol. 41, No. 2, 2002 
8. SAUERBREI W, HÜBNER K, SCHMOOR C, et al. Validation of existing and development of new prognostic classification schemes in node negative breast cancer. German Breast Cancer Study Group. Breast Cancer Res Treat 1997; 42:149-63. 
9. MOON TE, JONES SE, BONADONNA G, et al. Development and use of a natural history data base of breast cancer studies. Am J Clin Oncol 1987; 10:396-403. 
10. QUIET CA, FERGUSON DJ, WEICHSELBAUM RR, et al. Natural history of node-negative breast cancer: a study of 826 patients with long-term follow-up. J Clin Oncol 1995; 13:1144-51. 
11. ROSEN PP, GROSHEN S, KINNE DW. Survival and prognostic factors in node-negative breast cancer: results of long-term follow- up studies. J Natl Cancer Inst Monogr 1992; 11:159-62. 
12. ROSEN PP, GROSHEN S. Factors influencing survival and prognosis in early breast carcinoma (T1N0M0-T1N1M0). Assessment of 644 patients with median follow-up of 18 years. Surg Clin North Am 1990; 70:937-62. 
13. RAUSCHECKER HF, SAUERBREI W, GATZEMEIER W, et al. Eight-year results of a prospective non-randomised study on therapy of small breast cancer. The German Breast Cancer Study Group (GBSG). Eur J Cancer 1998; 34:315-23. 
14. JOENSUU H, PYLKKÄNEN L, TOIKKANEN S. Late mortality from pT1N0M0 breast carcinoma. Cancer 1999; 85:2183-9. 15. LEE AK, LODA M, MACKAREM G, et al. Lymph node negative invasive breast carcinoma 1 centimeter or less in size (T1a,b NOMO): clinicopathologic features and outcome. Cancer 1997; 79:761-71. 
16. LEITNER SP, SWERN AS, WEINBERGER D, et al. Predictors of recurrence for patients with small (one centimeter or less) localized breast cancer (T1a,b N0 M0). Cancer 1995; 76:2266-74. 
17. STIERER M, ROSEN H, WEBER R. Nuclear pleomorphism, a strong prognostic factor in axillary node-negative small invasive breast cancer. Breast Cancer Res Treat 1992; 20:109-16. 
18. STAEL O, DUFMATS M, HATSCHEK T, et al. S-phase fraction is a prognostic factor in stage I breast carcinoma. J Clin Oncol 1993; 11:1717-22. 
19. STENMARK-ASKMALM M, STAEL O, OLSEN K, et al. p53 as a prognostic factor in stage I breast cancer. South-East Sweden Breast Cancer Group. Br J Cancer 1995; 72:715-9. 
20. RAILO M, LUNDIN J, HAGLUND C, et al. Ki-67, p53, Er-receptors, ploidy and S-phase as prognostic factors in T1 node negative breast cancer. Acta Oncol 1997; 36:369-74. 
21. FRKOVI. GRAZIO S, BRA»KO M. Long term prognostic value of Nottingham histological grade and its components in early (pT1N0M0) breast carcinoma. J Clin Pathol 2002; 55:88-92. 
 Table 1. Risk categories for women with node-negative breast cancer 
Low-risk Intermediate-risk High-risk 
(has all (between the other (has at least listed factors) 2 categories) 1 listed factor) 
Tumor size ??1cm 1-2 cm > 2cm 
ER or PR status positive positive negative 
Tumor grade grade 1 grade 1-2 grade 2-3
BACK

Fine-needle aspiration cytology (FNA) is a simple, rapid, accurate and cost-effective method that has become a standard of care in the evaluation of breast lesions. Over recent years, FNA has become the method to assess palpable and non-palpable breast lesions, and contributes to management decisions at surgical and medical levels, being the source of primary diagnosis in several cases. In addition, cytology is also largely used to diagnose lymph node metastasis and to evaluate pleural effusions in patients with breast cancer. As the treatment planning is frequently made preoperatively based on cytological material, the diagnosis should be as precise as possible, and as much prognostic information should be gained from the cytologic specimens as possible. 
FNA has largely replaced frozen sections and, in cases candidates for primary chemotherapy it can provide hormonal assessment as well as other useful parameters relevant for the prognosis and prediction of therapeutic response. Furthermore, FNA material can also be used for some special studies such as immunohistochemistry, cytometry, in situ hybridization and molecular biology techniques. 
The aim of this review is to report the major application of the new technologies in cytologic material obtained from breast FNA biopsies. 

Identification of Myoepithelial Cells
The presence of myoepithelial cells is one of the most important criteria to support a diagnosis of benign lesion in FNA of the breast. However, on cytologic examination, the correct identification of myoepithelial cells is sometimes difficult, as they might be confused with apoptotic cells, stromal cells, and even epithelioid histiocytes. Since some years ago, several markers have been used in cytologic material in an attempt to identify myoepithelial cells. However, most of these markers (such as smooth-muscle actin, calponin, H-caldesmon, cytokeratins 5/6/14) have shown cross-reaction with other cells (myofibroblasts, luminal cells, stromal cells, pericytes) and are expressed at the cytoplasm of myoepithelial cells that can be lost in smears. 
Recently, two novel markers have been used to identify myoepithelial cells at the histologic and cytologic levels. P63, a p53-homologue nuclear transcription factor, is a protein that is necessary for the maintenance of the basal compartment of several multilayered epithelia and is selectively expressed in basal cells of stratified epithelia, in the basal cells of prostate and myoepithelial cells of the breast, salivary and lacrimal cells. Recently, Barbareschi et al. and our group have shown that p63 is a reliable myoepithelial cell marker in histologic sections. Moreover, some preliminary studies have pointed out that p63 could be better than other conventional myoepithelial cell markers because, as it is localized in the nuclei of myoepithelial cells, it overcomes the cytoplasmic fragility of myoepithelial cells in FNA. Studying more than 90 cases of breast FNA smears with immunocytochemistry for p63 antibody, we demonstrated that this marker highlighted the nuclei of two distinct cell populations: 1) all of the oval-tospindle- shaped cells with dark nuclei on epithelial cell clusters, and 2) all of the naked nuclei observed in the background of the smears. No immunoreactivity for p63 was observed in the cells admixed with fibrillary matrix of fibromyxoid stroma or in those isolated cells with oval nuclei and sparse-to-moderate cytoplasm. Based on these results, we strongly suggest that p63 is a reliable marker for myoepithelial cells in breast FNAs and that the majority of naked nuclei, defined as oval-to-spindle-shaped cells without any discernible cytoplasm, show a myoepithelial origin and thus, they might be included in the major criteria to diagnose benign breast lesions. Moreover, we observed that p63 helps identify myoepithelial cells overlying malignant cell clusters that we found consistently in cases of ductal carcinoma in situ. However, further studies with large series of patients using similar methodology are required in order to define how specific and sensitive this finding is in ruling out invasion in FNA of the breast. 
The other myoepithelial cell marker used by our group is maspin. Maspin is a member of the serpin family of serine protease inhibitors, and it has been claimed to be a tumor and metastasis suppressor and to have antiangiogenic properties. Maspin is consistently expressed by myoepithelialcells and our group has shown that myoepithelial tumors of the breast are positive for maspin. Although initial studies demonstrated cytoplasmic expression only, we showed for the first time that maspin is also expressed in the nuclei, a finding recently confirmed by other groups. 

Estrogen and Progesterone(ER/PR) Assessment
Besides giving prognostic information, hormonal receptor analysis is a useful tool to predict hormonal response in human breast cancer. According to the College of American Pathologists (CAP) Consensus Statement 1999, concerning the prognostic factors in breast cancer, hormone receptor analysis should be performed routinely in all primary breast carcinomas. Although hormone receptor analysis has been traditionally performed on surgically removed specimens, FNA offers a suitable alternative for this determination in a number of situations: a) inoperable cases and metastatic or recurrent tumors in which the size and accessibility to surgical biopsy presents a problem; b) cases in which preoperative irradiation or presurgical therapy is the initial treatment option; and c) advanced tumors in which serial hormone receptor studies may provide information regarding response to therapy. 
The assessment of hormone receptors on cytologic material has been performed by immunocytochemistry with good correlation with histologic and biochemistry determinations. Some years ago, we described a method of immunocytochemical assessment of estrogen receptor status on alcohol-fixed smears obtained by FNA from breast cancer patients, using a commercially available monoclonal antibody with antigen retrieval, and the results were compared with the assessment by ER immunocytochemical assay using the same procedure on formalin- fixed tissue and with assessment by ER-ICA assay on frozen sections. The results were scored semiquantitatively using a five grade scoring system. Although we have been critical in the use of score system in cytologic specimens, especially in relation to the extension ofimmunoreactivity, we found a good correlation between the results obtained on the cytologic specimens and on the histologic material. The heterogeneity of ER expression within the tumor should be taken in consideration whenever using FNA material for semi-quantification of ER because it might cause discrepant results. The putative usefulness of quantification in cytochemical hormone receptor assays remains controversial, and no consensus about the use of semiquantitative scoring system or mere division of tumors into positive and negative ones has been attained so far. Although it has not been shown that quantitative values beyond a defined level are helpful in selecting treatment options, some authors using FNA material showed better response to treatment with tamoxifen in patients with more than 50% ER-positive tumor cells. HER2/neu Assessment 
The proto-oncogene c-erbB2 is localized on chromosome 17 and encodes a 185-kD transmembrane glycoprotein with tyrosine kinase activity, which possesses a close sequence homology with the epidermal growth factor receptor. This oncogene is overexpressed in about one third of breast cancers and its overexpression is associated with high histologic grade, reduced survival, lower responsiveness to methotrexate-based treatment regimens and hormone receptor modulators such as tamoxifen, as well as higher responsiveness to doxorubicin-based regimens. Recently, a humanized antibody against HER2/neu was developed (Herceptin - Genentech, Inc, South San Francisco, CA), and might be used as a novel neoadjuvant primary therapy for HER2/neu positive breast cancer patients. Some authors showed that it is possible to determine the immunocytochemical expression of HER2/neu in previously Papanicolaou-stained aspirates of breast carcinomas. They found a strong correlation between HER2/ neu immunocytochemistry of aspirates and their corresponding tissue biopsies. In fact, the determination of HER2/neu in smears and cytoblock preparations may be as sensitive as, or even more sensitive than that of formalin- fixed, paraffin-embedded tissue. In cytology, the pattern of expression of HER2/neu is relatively uniform and is evidenced by membrane staining. NCL-CB11, a monoclonal antibody against the internal domain of the HER2/ neu protein, gave better results in cell smears with strongest reaction and least background staining. The clinical use of Herceptin requires the evaluation of HER2/neu amplification from every potentially eligible patient. Fluorescence in situ hybridization (FISH) is currently regarded by the FDA as the gold standard method for detecting HER-2/neu amplification. A big deal of discussion has emerged recently concerning the accordance between the immunohistochemical assessment of HER2/ neu overexpression and the real amplification of the gene assessed by FISH. For tissue sections, the semiquantitative immunohistochemical approach accepted by FDA is defined as positive membranous staining in more than 10% of the neoplastic cells. Partial or incomplete, weak to moderate, and moderate to strong membranous staining in more than 10% of the tumor cells must be scored as 1+ (negative), 2+ (weak positive), and 3+ (strong positive), respectively. Several reports showed a good correlation between a 3+ immunoexpression of HER2/neu and the amplification of the gene using formalin-fixed paraffin embedded tissue; however, in 1+ and 2+ cases, there is no correlation between these parameters and these patients may have benefits assessing c-erb-B2 amplification by FISH analysis. In the last years, some authors have successfully assessed the amplification of HER2/neu by FISH analysis in archival cytologic fine needle aspirates and showed a good concordance with paraffin-embedded tissue. 
The findings reported so far support that FNA cytologic samples might constitute the most cost-effective and easiest way to assess the HER2/neu amplification and overexpression, however, further studies are required to characterize the method for semiquantitative analysis of its immunocytochemical expression. Besides, FISH analysis requires microscopes with special filters and complex image analysis system to interpret the fluorescent signal. With the aim to supervene this drawback, the chromogenic in situ hybridization (CISH), a new modification of FISH, that enables detection of HER-2/neu gene copies with conventional peroxidase reaction in breast cancer specimens using regular microscopes, has been tested with outstanding results. Further studies in this front are needed to verify the applicability of this new method in FNA cytologic samples. 

P53 Expression
P53 encodes a 53 kD nuclear phosphoprotein with tumor suppressor activity. The wild type P53 protein, which is a transcription regulator, is present in the nuclei of all mammalian cells where it appears to be involved in the regulation of cell proliferation and apoptosis. Recent clinical evidence supports a critical role of P53 status in providing prognostic information, mainly in node-negative breast cancer patients. In fact, there is increasing evidence that tumors lacking normal P53 function are clinically more aggressive as they acquire a selective growth advantage becoming more resistant to ionizing radiation and some anticancer drugs. P53 immunostaining is nuclear and could be determined in previously Papanicolaou-stained aspirates of breast carcinomas. When we compare P53 gene mutations versus overexpression, the data obtained by molecular biology methods for assessment of mutations give better prognostic information than immunohistochemistry performed with PAb 1801 monoclonal antibody. In histologic material using a semi-quantitative approach we found an association between the presence of mutation in SSCP analysis and strong P53 staining and absence of mutation in cases with scarce and weakly positive neoplastic cells, however, we do not test this system in FNA smears. Since it is possible to detect P53 mutations and deletion in FNA material using polymerase chain reaction single strand conformation polymorphism (PCR-SSCP) and DNA sequencing analyses in DNA extracted from cell suspensions or archived stained smears and that specific mutations in the P53 gene are associated to primary resistance to chemotherapy, the assessment of P53 mutations on FNA material could be very useful to predict clinical behavior and responsiveness to therapy in breast cancer. This is of particular value in a primary chemotherapy setting, as complete tumor regression may occur and FNA-based pre-chemotherapy information may represent the only available information unaffected by therapy. The application of molecular biology techniques to the existing archival smears may become a valuable tool to detect genetic changes in samples from breast cancer aspirates, making FNA a reliable and helpful tool for the diagnosis, prognostic assessment and therapeutic management of breast cancer patients. 

Sialyl-Tn Expression
Sialyl-Tn (STn) is a core region carbohydrate antigen formed by the premature 2-6 sialylation of N-acetylgalactosamine whose expression is associated with some human malignancies. In fact, neoplastic transformation is almost invariably associated with marked changes in cell membrane glycoconjugates due to abnormal expression or depression of DNA encoding glycosyl transferases and the expression of simple mucin-type antigens, including STn is highly restricted in normal adult tissues. In our experience, the expression of this marker in breast cancer is associated with the presence of axillary metastases, lack of hormonal receptor, and high histologic grade. Moreover, some authors have demonstrated that STn positivity appears to be a marker of resistance to adjuvant chemotherapy. Using immunocytochemistry, we documented STn expression in mammographically detected breast lesions diagnosed by FNA cytology. Therefore, the determination of expression of STn in breast aspirates could be a useful marker to assess resistance to chemotherapy as well as to identify cases with high risk of axillary metastases.

Assessment of Telomerase Activity
Telomeres are repetitive sequences at the ends of chromosomes that protect chromosomes from incomplete replication, nuclease degradation, and end-to-end fusion during replication, playing a major role during DNA replication. In most somatic cells, after each cell division, the telomeres are eroded, leading to a progressive shortening of their length. When one telomere reaches the critical point, the cell stops dividing, and senesces. The maintenance of telomeres depends on the telomerase activity. Telomerase is a ribonucleoprotein complex responsible for de novo telomere synthesis and addition of telomeric repeats to existing telomeres. Telomerase activity (TA) is almost restricted to embryonic cells, germ cells, and malignant neoplastic cells; very low levels of this enzyme have been detected in somatic tissues, mainly restricted to the basal cell layer ofseveral epithelia and cells in the terminal ducts of the breast. Telomerase activity can be measured in vitro by using the telomeric repeat amplification protocol (TRAP), including cells obtained by FNA specimens. The presence of TA was demonstrated in 80% to 90% of breast carcinoma FNA samples; however, some benign lesions presented some level of telomerase activity, including fibroadenomas. Further evaluation of the sensitivity and specificity of TA for malignant cells is required before this technique could be accepted as a new marker in routine cytology. 

References
1. BARBARESCHI M, PECCIARINI L, CANGI MG, MACRI E, RIZZO A, VIALE G, DOGLIONI C. P63, a p53 homologue, is a selective nuclear marker of myoepithelial cells of the human breast. Am J Surg Pathol 2001; 25:1054-60. 
2. FITZGIBBONS PL, PAGE DL, WEAVER D, THOR AD, ALLRED DC, CLARK GM, RUBY SG, O’MALLEY F, SIMPSON JF, CONNOLLY JL, HAYES DF, EDGE SB, LICHTER A, SCHNITT SJ. Prognostic factors in breast cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 2000;124:966-78. 
3. HIYAMA E, SAEKI T, HIYAMA K, TAKASHIMA S, SHAY JW, MATSUURA Y, YOKOYAMA T. Telomerase activity as a marker of breast carcinoma in fine-needle aspirated samples. Cancer 2000;90:235-8. 
4. KLIJANIENKO J, COUTURIER J, GALUT M, EL-NAGGAR AK, MACIOROWSKI Z, PADOY E, MOSSERI V, VIELH P. Detection and quantitation by fluorescence in situ hybridization (FISH) and image analysis of HER-2/neu gene amplification in breast cancer fine-needle samples. Cancer 1999;87:312-8. 
5. LAVARINO C, CORLETTO V, MEZZELANI A, DELLA TORRE G, BARTOLI C, RIVA C, PIEROTTI MA, RILKE F, PILOTTI S. Detection of TP53 mutation, loss of heterozygosity and DNA content in fine-needle aspirates of breast carcinoma. Br J Cancer 1998; 77:125-30. 
6. MOORE JG, TO V, PATEL SJ, SNEIGE N. HER-2/NEU gene amplification in breast imprint cytology analyzed by fluorescence in situ hybridization: direct comparison with companion tissue sections. Diagn Cytopathol 2000;23:299-302. 
7. NIZZOLI R, BOZZETTI C, NALDI N, GUAZZI A, GABRIELLI M, MICHIARA M, CAMISA R, BARILLI A, COCCONI G. Comparison of the results of immunocytochemical assays for biologic variables on preoperative fine-needle aspirates and on surgical specimens of primary breast carcinomas. Cancer 2000;90:61-6. 
8. POREMBA C, SHROYER KR, FROST M, DIALLO R, FOGT F, SCHAFER KL, BURGER H, SHROYER AL, DOCKHORN- DWORNICZAK B, BOECKER W. Telomerase is a highly sensitive and specific molecular marker in fine-needle aspirates of breast lesions. J Clin Oncol 1999;17:2020-6. 
9. REIS-FILHO JS, SCHMITT FC. Taking advantage of basic research: p63 is a reliable myoepithelial and stem cell marker. Adv Anat Pathol 2002; in press. Acta clin Croat 2002; 41:135-186 Conference papers 154 Acta clin Croat, Vol. 41, No. 2, 2002 
10. REIS-FILHO JS, MILANEZI F, SILVA P, SCHMITT FC. Maspin expression in myoepithelial tumors of thebreast. Pathol Res Pract 2001;197:817-22. 
11. SCHMITT FC, BENTO MJ, AMENDOEIRA I. Estimation of estrogen receptor content in fine-needle aspirates from breast cancer using the monoclonal antibody 1D5 and microwave oven processing: correlation with paraffin embedded and frozen sections determinations. Diagn Cytopathol 1995;13:347-51. 
12. SCHMITT FC. Comments on p53 protein expression, cell proliferation and steroid hormone receptors in ductal and lobular in situ carcinomas of the breast. Eur J Cancer 1997; 33:1903. 
13. SCHMITT FC, SOARES R, CIRNES L, SERUCA R. P53 in breast carcinomas: association between presence of mutation and immunohistochemical expression using a semiquantitative approach. Pathol Res Pract 1998;194:815-19. 
14. SCHMITT FC, MARINHO A, AMENDOEIRA I. Expression of sialyl-Tn in fine-needle aspirates from mammographically detected breast lesions: a marker of malignancy? Diagn Cytopathol 1998; 18:325-9. 
15. SOLOMIDES CC, ZIMMERMAN R, BIBBO M. Semiquantitative assessment of c-erbB-2 (HER-2) status in cytology specimens and tissue sections from breast carcinoma. Anal Quant Cytol Histol 1999;21:121-5. 
16. TANNER M, GANCBERG D, DI LEO A, LARSIMONT D, ROUAS G, PICCART MJ, ISOLA J. Chromogenic in situ hybridization: a practical alternative for fluorescence in situ hybridization to detect HER-2/neu oncogene amplification in archival breast cancer samples. Am J Pathol 2000;157:1467-72. 
BACK

Mammary neoplasms are very prevalent in dogs and cats, but are rare in other domesticated species. Because of their prevalence in these companion animals, and because they are a model for human breast cancer, they are deeply investigated clinically, histologically, immunohistochemically and with other molecular pathology techniques. Mammary neoplasms are the third most frequently occurring tumor in female cats, following hematopoietic neoplasms and skin tumors. The incidence of mammary tumors in this species is less than half that of humans and dogs, nevertheless, these tumors account for 17% of neoplasms in female cats. The great majority (about 80%) of feline mammary tumors are malignant (adenocarcinomas). Breed predisposition (Siamese) has been speculated but not proven. Mammary tumors occur primarily in intact cats from 9 months to 23 years of age, with a mean age of 10 to 12 years. Several reports have documented an association between the prior use of progesterone-like molecules and the development of benign or malignant feline mammary neoplasms. While low concentrations of progesterone receptors have been found in the cytoplasm of some feline mammary tumors, dihydrotestosterone receptors have not been identified. Only 10% of the feline tumors examined were positive for estrogen receptors: a much higher percentage of estrogen receptor positive tumors are seen in dogs and humans. Tubular, papillary, and solid adenocarcinomas are the most common malignant tumors and the majority present a combination of tissue types in each tumor. Sarcomas, mucinous carcinomas, duct papillomas, adenosquamous carcinomas, and adenomas are rarely seen. Mammary gland dysplasia, while infrequent, needs to be differentiated from malignant neoplasms. Lobular hyperplasia and fibroepithelial hyperplasia are types of non-inflammatory hyperplasia identified in the mammary gland of the cat. These hyperplastic conditions are relatively common, and thought to be associated with hormonal stimulation of the glandular tissue, which decreases following ovariohysterectomy. Benign tumors include simple and complex adenomas, lowand high-cellularity fibroadenomas, benign mixed tumors and duct papillomas. Feline mammary gland neoplasm can involve any or all of the glands and is distributed equally between the left and right sides. Multiple gland involvement occurs in more than half of affected cats. Metastatic lung and thorax involvement may be extensive. Tumor size is the single most important prognostic factor for malignant feline mammary tumors. Other significant prognostic factors affecting recurrence in and survival of feline malignant mammary tumors are the extent of surgery and histologic grading of the tumor. Cats with a tumor size of greater than 3 cm in diameter will have a median survival time of 4 to 6 months. Cats with a tumor size of 2 to 3 cm in diameter will have a significantly increased survival time with a median of about 2 years. Cats with tumor less than 2 cm in diameter have a median survival time of more than 3 years. Thus, early diagnosis and treatment is a very important prognostic factor for malignant feline mammary tumors. Mammary tumors are extremely common in intact female dogs, and account for at least 50% of all reported neoplasms. At least 70% of intact bitches will develop a clinically detectable mammary tumor if they live to 15 years of age, and almost all of them will have microscopic tumor foci. Ovariohysterectomy prior to the first heat cycle greatly reduces the risk (more than 80%) of developing mammary neoplasia later in life, with the benefits reduced with time and no benefits of spaying after the second heat. Some report that 50% of canine mammary tumors aremalignant and 50% of the malignant neoplasms metastasize, others reports that metastasis is much less frequent. The most reliable predictor of true behavioral malignancy in canine mammary tumors is local invasion. While local invasion is readily detected by histologic examination of the excised tumors, it can also be predicted with accuracy by clinical examination. Tumors that seem to be fixed to the underlying tissue, cross the midline, or are otherwise obviously infiltrative are likely to be true malignancies. A bitch with one detected mammary tumor very often will have numerous microscopic foci in the same or other glands. Malignant tumors of the dog include several types of carcinomas: noninfiltrating (in situ), complex, simple, tubulopapillary, solid, anaplastic and some special types of carcinomas including spindle, squamous, mucinous, and lipid rich carcinomas. Sarcomas, which are much less frequent, include fibrosarcomas, osteosarcomas, and other sarcomas. In addition, carcinosarcoma and carcinoma or sarcoma within a benign neoplasm have been described. Benign tumors include simple, complex, basaloid adenomas, fibroadenomas, benign mixed tumors, duct papillomas. Other described changes in the canine mammary gland include ductal, lobular, and epithelial hyperplasia, adenosis, cysts, duct ectasia, fibrosclerosis and gynecomastia.
BACK

Diagnostic imaging (particularly echocardiography and MRI) has increased the frequency of identification of cardiac tumors. Although still uncommon, cardiac neoplasia is more than an academic curiosity. In human patients, benign tumors are more frequent than malignant, with myxomas being most prevalent. A sex predisposition has been observed, with a higher prevalence in females. Cardiac tumors are diagnosed in patients of all ages. The most frequent cardiac tumors in dogs are hemangiosarcomas, chemodectomas, ectopic thyroid carcinomas, and pericardial mesothelioma. For most of these tumor types there is a demonstrated breed predisposition; for example chemodectomas are most commonly found in brachiocephalic breeds. Diagnostic imaging is needed to guide surgical treatment: surgical removal of these tumors has grown following the introduction of echocardiographic examination. Diagnostic imaging is useful even in prenatal diagnosis of cardiac neoplasias such as rhabdomyoma and rhabdomyosarcoma. Almost all cardiac related tumors could be identified by echocardiography, particularly intracardiac ones. When transthoracic echocardiography is unable to fully evaluate the edges of cardiac masses, transesophageal echocardiography can be used to provide more complete diagnostic information. MRI synchronized with cardiac cycle can achieve additional information. Echocardiography and MRI are complementary diagnostic tools. In the dog, cardiac tumors are commonly typified by echocardiography based on their peculiar morphological features. Even in human patients, echocardiography and MRI give enough information to identify the type of cardiac tumor.
BACK

The liver is the central organ of drug metabolism and also a major target of drug toxicity. Many drugs and chemicals are potentially hepatotoxic, including drugs used for treatment of liver diseases. It is known from several studies that about 2% to 10%, in older age groups even more, of hospitalized icteric patients are drug victims and about 25% of acute necrotizing liver diseases are related to drug intake. Moreover, drugs are also causes of chronic liver diseases. Practically all types of acute and chronic liver disorders can be drug-induced. Therefore, diagnosis depends on close cooperation between the pathologist and clinician, and particularly on the information about the patient. The diagnosis of adverse drug effects is important since further administration of a drug potentiates liver injury and may even lead to a chronic course. Hepatotoxic drugs can be divided into intrinsic and idiosyncratic types. Intrinsic hepatotoxins mostly have cytotoxic and less often cholestatic effects, which are dosedependent, fairly well predictable with a short latent period. The lesions are reproducible in animal experiments. Liver disease due to this type of drugs is rather rare and mostly associated with overdose. A typical example is acetaminophen (paracetamol) intoxication. 
Liver injury caused by idiosyncratic liver toxins is more often seen. Only a small percentage of susceptible patients are affected. These drug effects are unpredictable and usually not experimentally reproducible. The latent period between the drug administration and onset of liver disease is often long and variable, but shorter latent periods are seen upon readministration. The idiosyncratic type of drug-induced liver injury is either caused by an immune response against the drug or a metabolite in association with cellular proteins, or by genetically determined variations in drug metabolism and resulting toxic metabolites. The morphologically detectable consequences of adverse drug reactions are variable and in principle nonspecific, although eosinophilia, pigment-containing macrophages, steatosis, and ductular reaction are suggestive of drug etiology. Morphologic alterations comprise adaptive responses of hepatocytes, hepatitic and non-hepatitic lesions, granuloma formation, cholestasis, finally fibrosis or even cirrhosis, and vascular and neoplastic changes. Cellular adaptation with increased amounts of smooth endoplasmic reticulum is morphologically characterized by ground glass-like appearance of hepatocytes, predominantly in centrilobular areas. 
The most frequent manifestations of intrinsic hepatotoxins are steatosis (macro- or microvesicular) and necrosis, usually in acinar zone 3. Macrovesicular steatosis for example may result from the administration of corticosteroids, the microvesicular form may be caused e.g. by tetracyclins, anticonvulsants or anti-inflammatory drugs. Microvesicular steatosis often results from the inhibition of mitochondrial ?-oxidation. Parenchymal necrosis may be zonal (perivenous, periportal, intermediate), non-zonal or massive. Intrinsic hepatotoxins tend to the induction of zonal, idiosyncratic or non-zonal necroses. Typically, perivenous zonal necrosis is caused by an overdose of paracetamol. Acute hepatitis may follow the administration of diverse drugs, including anesthetics but also herbal preparations. It resembles an idiosyncratic reaction with latent periods between exposure and manifestation of liver disease varying between days and months. The morphology is similar or even identical to acute viral hepatitis with variable degrees of necrosis, steatosis, cholestasis, lobular activity and portal inflammation. The presence of epithelioid cell granulomas suggests drug etiology. A classical example of this type of drug reaction is halothane hepatitis. In its early stage, centrilobular necrosis caused by toxic metabolites prevails, whereas in the later stage the hepatitic picture predominates. This adverse reaction to halothane occurs in about 1 to 10 of 30000 exposed individuals but the frequency increases to about 1 in 100 following multiple exposures. 
Multiple or prolonged exposures to certain drugs may lead to chronic liver disease. Drug-induced chronic hepatitis is indistinguishable from chronic viral or autoimmune hepatitis both with respect to morphology and serology (e.g., occurrence of autoantibodies). Drug-induced chronic hepatitis is often accompanied by pronounced lobular necrosis, lymphocytic, plasmacytic and eosinophil-granulocytic inflammation in variable combinations. Chronic hepatitis may be induced by acetaminophen (paracetamol), tetracyclins, sulfonamides, anti-inflammatory, anticonvulsive, antihypertensive and antiarrhythmic drugs. Steatohepatitis, characterized by steatosis, hepatocellular ballooning, Mallory body formation, inflammation and pericellular fibrosis, morphologically resembling alcoholic hepatitis, can be caused by a variety of drugs, including synthetic estrogens, amiodarone, methotrexate and tamoxifen. The risk of development of steatohepatitis is enhanced by obesity. 
Cholestasis is a common manifestation of drug-induced liver disease. It may be classified as canalicular, hepatocanalicular or ductular. Pure canalicular cholestasis without morphologic signs of liver cell damage and inflammation occurs in a small percentage of genetically susceptible individuals after administration of 17-?-alkylated anabolic or contraceptive steroids. Jaundice usually develops 1-6 months after commencement of treatment. In contraceptive pills, the estrogen component is the major mediator of cholestasis. Hepatocanalicular cholestasis develops, e.g., in 1%-2% of patients treated with chlorpromazine and is associated with portal and lobular inflammation of varying degrees, enlargement of hepatocytes, multinucleation and mitoses. The lesion may vary from almost pure cholestasis to pronounced hepatitis. Chronic drug-induced cholestatic conditions are occasionally observed and may be accompanied by bile duct lesions finally leading to a vanishing bile duct syndrome. About 30% of granulomas in the liver are drug-induced. Hepatic granulomas should, therefore, always rise suspicion of an adverse drug reaction. Many drugs can be incriminated, with non-steroidal anti-inflammatory drugs playing a major role in this respect. 
The vascular system of the liver may also be the target of adverse drug reactions (anabolic and contraceptive steroids, cytostatic drugs, herbal preparations), which in- clude sinusoidal dilatation (anabolics, contraceptives) and endothelial lesions with resulting veno-occlusion disease (cytostatic drugs). 
The association of hepatocellular adenomas with longterm administration of contraceptives and anabolics is a well known although a rather rare event. The risk associated with contraceptives increases with prolonged intake, higher estrogen content and in individuals of older (above 30 years) age groups. 
In conclusion, drug-induced lesions of the liver show a variable morphology on liver biopsy. The diagnosis is important since discontinuation usually results in recovery. Drug-induced liver injury should always be considered in patients with hepatobiliary symptoms, particularly in older age groups and in females who are more susceptible. Close cooperation between the pathologist and clinician is an imperative to establish a correct diagnosis. 

General Reading
1. SCHEUER PJ, LEFKOWITCH JH. Liver biopsy interpretation. London, Edinburgh, New York, Philadelphia, St.Louis, Sydney, Toronto: W.B. Saunders, 2000. 
2. SHERLOCK S. Diseases of the liver and biliary system. 8th Ed. Oxford, London, Edinburgh, Boston, Melbourne: Blackwell Scientific Publications, 1989. 
3. ZIMMERMAN HJ. Hepatotoxicity. The adverse effects of drugs and other chemicals on the liver. 2nd Ed. Philadelphia: Lipincott Williams & Wilkins, 1999. 
4. ZIMMERMAN HJ, ISHAK K. Hepatic injury due to drugs and toxins. In: Mac SWEEN RNM, BURT AD, PORTMANN BC, ISHAK KG, SCHEUER PJ, ANTHONY PP, eds. Pathology of the liver, London, Edinburgh, New York, Philadelphia, St. Louis, Sydney, Toronto: Churchill Livingstone, 2002: 621-709.
BACK

The School of Veterinary Medicine at the University of Pennsylvania was founded in 1884. The aims of the School, as stated in the Announcement of its formation, were “to give instruction, both theoretical and practical, in all branches pertaining to the scientific study of the elements of medicine, and the practical application of these elements to the domestic animals, in the preservation of their health, in their employment as useful aids to man, and in the diseases to which they are subject.” Thus, from the outset, the School was identified as a branch of medicine. Of note, ten of the School’s original faculty had M.D. degrees. The philosophy of there being only “one medicine”, involving both humans and animals, has guided the school over the last 117 years. Veterinary medicine has certainly benefited from advances in human medicine in a number of specialty areas, including veterinary clinical toxicology. Conversely, veterinary medicine has improved the health and well-being of people in many significant ways. More specifically, veterinarians and veterinary toxicologists have contributed to the advancement of the specialty of human clinical toxicology by 1) improving the diagnosis and treatment of human and animal toxicantinduced disease, 2) improving our understanding of toxicologic mechanisms and comparative toxicology and pathology, 3) enhancing the recognition of new and emerging toxicologic threats, and 4) preventing human and animal toxicant-induced disease. Veterinary toxicologists and analytical chemists have been at the forefront in the utilization of a variety of analytical techniques for the detection of toxicants in a number of biological and environmental matrices. The day-to-day use of inductively coupled plasma argon emission spectroscopy (ICPAES) for the diagnosis of metal intoxications was first instituted in veterinary diagnostic toxicology laboratories. Adaptation of sophisticated analytical techniques for broad-based toxicant screening continues with the use of liquid chromatography . mass spectroscopy and ICP-mass spectroscopy systems. Veterinarians were early investigators of the safety and efficacy of fomepizole (4-methylpyrazole) for the treatment of ethylene glycol intoxication of dogs. From a comparative toxicology standpoint, recent investigations in cats have demonstrated species differences with regard to the effectiveness of fomepizole to inhibit alcohol dehydrogenase. The impact on human health of persistent organic pollutants such as PCBs has been difficult to assess. Investigation of seal mortalities in the Baltic has demonstrated that ingestion of PCB-contaminated fish can impair immune function and potentially exacerbate infectious disease outbreaks. Such studies have direct application to the potential risks to humans of eating PCB-contaminated foodstuffs. Aflatoxins were first isolated from peanut meal during investigations of an epizootic of “turkey X” disease in England; histopathologic examination of affected turkeys demonstrated the hepatotoxicity of these mycotoxins. Subsequent studies revealed their ani- mal and human carcinogenic potential. Veterinary toxicologists were the first to isolate and chemically characterize the fumonisin mycotoxins, common contaminants of corn and the cause of several animal diseases including equine leucoencephalomalacia (ELEM). Isolation and chemical characterization of these mycotoxins has led to considerable investigation into their mechanism(s) of toxic action and their potential impact on human health. Veterinary toxicologists play an essential role in preventing toxicant-induced human disease. They are involved in assuring that feedstuffs intended for livestock consumption are free of chemical contaminants that might result in violative residues in meat, milk and eggs destined for human consumption. Investigation into unexplained foodanimal deaths by veterinary toxicologists often prevents chemically-contaminated animal products from being used as human food. Veterinary toxicologists play an essential role in the use of animals as biomonitors of environmental quality. The diagnosis of lead exposure and/or intoxication in a household is often initially made in a pet such as a dog or cat. Such a diagnosis can prevent significant exposure of young children to lead or early recognition of an adverse effect. In the U.S., the establishment of the National Animal Poison Control Center (NAPCC) nearly 15 years ago has provided a valuable database of animal exposures to toxicants. The NAPCC database assists in the recognition of potential toxicant hazards to humans. For example, information on the intoxication of animals by some human dietary supplements has been useful in characterizing their potential hazard to people. Additional animal poison information services are now available in England and France. In summary, veterinary clinical toxicologists interface directly with their human colleagues in a variety of ways. It is important that human and veterinary clinical toxicologists recognize that there is only “one clinical toxicology” and seek ways to promote and exploit such a concept for the mutual benefit of our human and animal patients. 
BACK