The fact that this classification system (ADH through to high grade DCIS) has been accepted internationally reflects the relative consistency of data emerging from the various studies with respect to risk of subsequent development of invasive breast cancer. Perhaps more significantly, it indicates that such evidence supports the compelling hypothesis that some forms of breast carcinoma may arise not only from established forms of DCIS, but that these lesions may arise from epithelial proliferation recognized as ADH and possibly from more common forms of usual hyperplasia. This concept, however appealing, has limitations. Some forms of carcinoma in situ and invasive carcinoma, possibly the majority, do not appear to be associated with such lesions and appear to arise de-novo or from as yet unknown precursor lesions. In addition, ADH was recognized initially based on exclusion rather than positive criteria, i.e. recognition of some but not all features of DCIS and lack of the characteristics of usual type epithelial hyperplasia.⁵ 

Although having found great favor as a concept backed by evidence from long term follow-up studies, the major problem with recognition of these entities has been the difficulty in achieving acceptable levels of concordance or consistency in diagnosis between pathologists. Various strategies have emerged to try and improve reliable recognition, including revision of the criteria, providing a more positive basis for future recognition, education and emphasis of the use of one system by all for diagnostic classification and also the use of lesion size.⁶  Despite the adoption of such principles, reliable classification around the boundaries of ADH remains elusive in most,^{7, 8}  although not all,⁹  studies.

This problem with reproducibility of diagnosis is unlikely to be resolved by the adoption of alternative systems or nomenclature based on the same diagnostic criteria. For example, the mammary intraepithelial neoplasia, ductal type (MIN) system (8) (or ductal intraepithelial neoplasia (DIN)) proposed by Tavassoli ^{10, 11} has some attraction in that it uses a terminology which has been adopted in other organ systems. Unfortunately the MIN system retains the existing diagnostic boundaries and requires the same problematic distinction between intraductal hyperplasia of usual type, ADH, low grade DCIS, intermediate grade DCIS and high grade DCIS. It thus in essence promotes a change in terminology without any change of diagnostic criteria or conceptual basis. In addition this particular proposal also includes a hyperplastic process, ductal epithelial hyperplasia of usual type in a classification of "neoplasia". For these reasons some feel that this classification is unlikely to stand the test of time as it cannot inherently improve on interobserver reproducibility and provides no additional evidence on which to promote a "new" classification system.

One further potential problem exists when using a simple system such as the existing methods or indeed the MIN system in the breast. We currently recognize a wide morphological diversity of breast malignant tumor, both in situ and invasive. This diversity is most probably a consequence of a range of underlying different pathogenic mechanisms and pathways. These may have some common origins, overlap or be entirely distinct with separate pathogenetic origins. Neoplastic processes in other organ systems such as the cervix appear to be far less genetically and morphologically diverse and as a consequence lend themselves to a unifying concept of intraepithelial neoplasia based on morphology. We believe that adaptation of such simplicity, although appealing, is unlikely to be successful due to the morphological and presumably genetic complexity of the breast neoplastic processes. We would suggest that specific clinico-pathological entities should be identified through an understanding of their genetic basis and that these distinct entities could then be further sub-classified using a grading system if they exhibit morphological variation or by alteration of further influential genes.

The classification of DCIS has been a very topical subject in recent years stimulating consensus conferences amongst pathologists, clinicians and radiologists in the USA, Europe and Australia ^12-14  The origin of this interest has been the increased frequency of detection of DCIS, and in particular localized DCIS, due to the uptake of mammography since the 1970's.¹⁵  These lesions have allowed the adoption of conservative surgical management techniques and have led to a search for predictive variables for local recurrence both as in situ and more significantly as invasive disease.^16-20 

Traditional histological classification of DCIS based on architecture appears to have limited value in this role. Most of the improved histological classification systems are based on a three-tier grading or differentiation system and use nuclear grade,^{21, 22}  differentiation and polarity ²³ and/or the presence or absence of necrosis.^{24, 25}  Mel Silverstein and colleagues have been particularly innovative in promoting use of histological grade coupled with lesion size and excision margin distance in the form of a prognostic index,²⁶  and have shown significant predictive power for local recurrence. However, although many of the various histological classification systems described appear predictive in many of the studies published to date, questions still remain about diagnostic reproducibility between pathologists and their long-term clinical relevance. At present there is limited data on reproducibility ^{27, 28, 19, 22}  but these early studies do appear to show moderate to good levels of concordance for many of the three-tier grading systems. It is however possible that the histological classification of established DCIS may be irrelevant if viable surgical excision is achieved. Total ablation of DCIS by mastectomy is generally accepted to be curative. Conservative surgical management, particularly with "close" margins, is complicated by a significant risk of local recurrence which can be reduced (but not negated) by adjuvant radiotherapy.²⁹  Use of complete local excision aiming to achieve excision by over 10 mm, coupled with diligent pathology examination of margins, can reduce the risk of local recurrence,³⁰  but to date no randomized control trials examining such surgical and pathological techniques and comparing them with a less diligent approach combined with radiotherapy have been conducted.

The existing information on DCIS indicates that it is typically a unicentric lesion confined to a single duct system which presumably could lend itself to complete surgical excision without the need for complete mastectomy.^31-33  The main problem however is our present lack of ability to map disease extent pre-operative and thus plan the surgical procedure.^{33, 34}  Current imaging techniques do not provide sufficient accuracy to allow such tailored conservative surgery and most centres offer a standardized technique resecting an area of calcification with a circumferential extended margin of "normal" tissue of varying extent.

In many three-tier classification systems in pathology, pathologists have little difficulty in recognizing the entities at either end of the spectra. Problems of concordance of classification are generally found in the middle group and its boundaries. Two-tier classification systems could potentially improve on reproducibility, but at present there is no clear agreement where such a boundary should be and thus in two-tiered systems at present reproducibility remains poor.²⁸  We feel that most authorities would choose the boundary between low and intermediate grade DCIS rather than a boundary between intermediate and high grade DCIS. In practice, however, the distinction around either boundary is difficult. Most of the newly promoted classification systems use nuclear grade, some with necrosis as an additional morphological feature, for this distinction. Nuclear grading is assisted in the latest grading systems through reference to red cell size; low grade nuclei are identified as being less than or equal to two red blood cells in size, intermediate and high grade nuclei being over two red cells in area. In practice this size criteria is relatively restrictive and certainly in current UK practice there are only a small proportion of cases of DCIS which fulfil these criteria for low grade designation.

It is for the above reasons that currently we retain the use of the standard terminology of usual epithelial hyperplasia, ADH and DCIS, the latter sub-classified into three grades. However we feel that recently generated data may offer a more appropriate and pragmatic approach to future classification which could resolve the diagnostic dilemmas for pathologists and also provide a more rational basis for clinical management.

There is a need for an improved and innovative classification system for epithelial proliferative lesions and in situ malignancy of the breast. We feel strongly that there is little value in revising the existing systems in view of their faults, as described above, unless or until it is possible to devise a classification method which has biological and clinical relevance and which can be applied consistently by pathologists. We suspect that such a system based on recognition of genetic lesions is now emerging and it is tempting to draw comparisons with the developments in leukemia and lymphoma classification over the last 10 years which recognize distinct clinico-pathological entities based on fundamental genetic changes. Similar specific and diagnostically helpful genetic changes are being identified in many soft tissue tumors.

It is unlikely that the basic pathological process of hyperplasia and neoplasia in the breast are different from other organ systems. We feel that a fundamental distinction between hyperplasia and neoplasia will remain in breast disease classification and serve as one distinct boundary. At present this separation is based mainly on morphological criteria but in recent years additional features have become recognized which could assist in achieving diagnostic reproducibility. We would propose that the boundary between usual type ductal hyperplasia and neoplastic proliferation of the breast should be based on recognition of a clonal cell process but not necessarily based on molecular investigations of clonality. A clonal process, particularly of low grade cellular morphology, potentially can also broadly be recognized by its uniformity of morphology and of phenotype such as cytokeratin expression or hormone receptor expression. Although not widely recognized, some groups have demonstrated that usual epithelial hyperplasia is morphologically and also phenotypically heterogeneous, being composed of a mixed cell population arranged in a haphazard architectural structure and showing mixed / heterogeneous cytokeratin and estrogen receptor (ER) expression.³⁵  The finding of ER heterogeneity in this hyperplastic condition is not surprising when one considers the focal expression seen in normal epithelium. In contrast ADH and established low grade DCIS have not only a uniformity of cell type but also show uniformity of cytokeratin, particularly cytokeratin 5 / 6, and estrogen receptor expression. High grade DCIS is, however, frequently ER negative.³⁶ 

The concept of in situ epithelial neoplasia exists in other organs such as cervix and prostate. The concept of a progression from a benign to malignant neoplastic processes is well established in some sites particularly in the colo-rectum, as the adenoma-carcinoma sequence. At present, in the breast one could argue that the conceptual distinction between benign neoplasia and in situ malignancy has been arbitrarily drawn at the boundary between ADH and low grade DCIS. Is this the appropriate boundary and should we pursue the need for distinction between benign and malignant neoplasia? Indeed, is it appropriate to distinguish between benign and malignant in situ neoplasia in the breast? Would it be more appropriate to identify established neoplastic processes and then sub-classify these according to their grade and extent; the former being related to risk of progression to invasive carcinoma, the latter being used to direct the appropriate technique of ablative removal?

The molecular genetic studies of low grade DCIS and ADH using loss of heterozygosity (LOH) techniques have demonstrated similar genetic lesions and, in informative cases, confirmatory evidence that these are clonal processes and therefore fulfil the basic concept of a neoplastic process.³⁷  The frequency of LOH in cases of usual hyperplasia is much lower in most series and, although the finding of LOH in this process is intriguing, suggests that it is appropriate to draw the boundary line between hyperplasia and neoplasia between usual epithelial hyperplasia and ADH. The presence of LOH in some cases of usual hyperplasia ^{38, 39} most probably indicates that clonal lesions may emerge from this hyperplastic process. This is a well-recognized concept at other sites. These are not, however, morphological distinct and have not reached autonomy and dominance over the otherwise hyperplastic process.

Equally exciting studies, in our view, are the very recently published work using comparative genomic hybridization (CGH) to investigate DCIS of the breast. These have prompted the proposal of a hypothetical model for the pathogenesis of DCIS, which recognizes genetic lesions associated with particular morphological sub-types. Well-differentiated DCIS is associated with 16q and 17p loss whilst tumors of intermediate and high grades often have allelic losses of significantly more chromosomal arms, frequently including 1p, 1q, 6q, 9p, 11p, 11q, 13q, and 17q.⁴⁰  High grade DCIS particularly has gains at 17q but also 11q and 13q.⁴¹  Intermediate grade DCIS appears to include a combination of lesions which show 16q loss but gains at other chromosomes particularly 1q and also cases which show gain at 11q13q, but lack the gain at 17q12 which is a feature of high grade DCIS.⁴²  As in intraductal proliferations (ADH and DCIS), atypical lobular hyperplasia and lobular carcinoma in situ similarly show the same genetic mutations as each other with loss of material from 16p, 16q, 17p, and 22q and gain of material from 6q.⁴³ 

We believe that these studies may herald a new era in classification of breast disease. To date these series have themselves analyzed groups based on nuclear grading criteria, but we would expect that further refinement of genetic abnormalities will derive from analysis of distinct morphological types of DCIS.

The benefits in routine clinical practice from following a simplified classification system such as nuclear grade are a greater chance of achieving reproducibility of recognition between pathologists and clinical relevance, particularly prediction of risk of local recurrence. The problem in terms of classification development in the modern era is that they may over simplify. In the past pathologists have recognized a range of morphologically distinct types of DCIS based on their architectural and cytomorphological features; these include comedo, micropapillary, cribriform, small and large cell solid, papillary, apocrine, signet ring, clear cell, hypersecretory and neuroendocrine DCIS and as well as lobular carcinoma in situ (LCIS). Should we abandon recognition of these morphologically distinct tumors? We believe that despite problems in achieving diagnostic reproducibility these morphological types probably hold the key to understanding the underlying pathogenic mechanisms. We would not be surprised to find distinct genetic lesions associated with each morphological sub-type and feel that histopathologists may re-visit morphological sub-typing of DCIS in the near future.

We would make the following proposal with regard to future classifications of the these lesions:

1. The distinction of the boundary between hyperplasia and neoplasia in the breast should be recognized on the basis of a dominant, uniform, clonal cell population using morphological, immunophenotypical and molecular genetic information, if feasible. This is practicable at present and pathologists can utilize immunohistochemistry facilities to detect heterogeneity in staining, for example in ER reactivity.

2. Established clonal, neoplastic breast lesions (including lesions presently classified as ADH, ALH, LCIS and DCIS) should be sub-categorized according to their specific molecular genetic basis. This sub-categorization could routinely be based on morphological characteristics, but we would anticipate that this would be supported by the expression or loss of phenotypic markers related to specific genetic lesions. These genetically distinct in situ neoplasms could be further sub-categorized according to additional genetic changes and the extent of the lesion.

References

1. Dupont WD, Page DL. Relative Risk of Breast-Cancer Varies With Time Since Diagnosis of Atypical Hyperplasia. Human Pathology 1989;20:723-5.
2. Page DL, Dupont WD. Anatomic Indicators (Histologic and Cytologic) of Increased Breast- Cancer Risk. Breast Cancer Research and Treatment 1993;28:157-66.
3. Connolly JL, Schnitt SJ. Clinical and Histologic Aspects of Proliferative and Nonproliferative Benign Breast Disease. Journal of Cellular Biochemistry 1993:45-8.
4. Marshall LM, Hunter DJ, Connolly JL, Schnitt SJ, Byrne C, London SJ, et al. Risk of breast cancer associated with atypical hyperplasia of lobular and ductal types. Cancer Epidemiology Biomarkers & Prevention 1997;6:297-301.
5. Page DL, Dupont WD, Rogers LW, Rados MS. Atypical Hyperplastic Lesions of the Female Breast - a Long-Term Follow-Up-Study. Cancer 1985;55:2698-708.
6. Page DL, Rogers LW. Combined Histologic and Cytologic Criteria For the Diagnosis of Mammary Atypical Ductal Hyperplasia. Human Pathology 1992;23:1095-7.
7. Sloane JP, Amendoeira I, Apostolikas N, Bellocq JP, Bianchi S, Boecker W, et al. Consistency achieved by 23 European pathologists from 12 countries in diagnosing breast disease and reporting prognostic features of carcinomas. Virchows Archiv-an International Journal of Pathology 1999;434:3-10.
8. Rosai J. Borderline Epithelial Lesions of the Breast. American Journal of Surgical Pathology 1991;15:209-21.
9. Schnitt SJ, Connolly JL, Tavassoli FA, Fechner RE, Kempson RL, Gelman R, et al. Interobserver Reproducibility in the Diagnosis of Ductal Proliferative Breast-Lesions Using Standardized Criteria. American Journal of Surgical Pathology 1992;16:1133-43.
10. Tavassoli FA. Ductal carcinoma in situ: Introduction of the concept of ductal intraepithelial neoplasia. Modern Pathology 1998;11:140-54.
11. Tavassoli FA. Mammary intraepithelial neoplasia: A translational classification system for the intraductal proliferations. The Breast Journal 1997;3:48-58.
12. Schwartz GF, Lagios MD, Carter D, Connolly J, Ellis IO, Eusebi V, et al. Consensus conference on the classification of ductal carcinoma in situ. Cancer 1997;80:1798-802.
13. Recht A, Vandongen JA, Fentiman IS, Holland R, Peterse JL. 3rd Meeting of the DCIS Working Party of the EORTC (Fondazione-Cini, Isola-S-Giorgio, Venezia, 28-February-1994) - Conference Report. European Journal of Cancer 1994;30A:1895-900.
14. Australian - New Zealand Breast Cancer Trials Group. Recommendations from the Pathology and Multidisciplinary DCIS Workshops. In: Diagnosis of DCIS; 1996; Hunter Valley; 1996.
15. Pinder SE, Ellis IO. Atypical ductal hyperplasia, ductal carcinoma in situ and in situ atypical apocrine proliferations of the breast. Current Diagnostic Pathology 1996;3:235-42.
16. Solin LJ, Kurtz J, Fourquet A, Amalric R, Recht A, Bornstein BA, et al. Fifteen-year results of breast-conserving surgery and definitive breast irradiation for the treatment of ductal carcinoma in situ of the breast. Journal of Clinical Oncology 1996;14:754-63.
17. Silverstein MJ. Predicting residual disease and local recurrence in patients with ductal carcinoma in situ. Journal of the National Cancer Institute 1997;89:1330-1.
18. Silverstein MJ, Lagios MD, Martino S, Lewinsky BS, Craig PH, Beron PJ, et al. Outcome after invasive local recurrence in patients with ductal carcinoma in situ of the breast. Journal of Clinical Oncology 1998;16:1367-73.
19. Badve S, A'Hern RP, Ward AM, Millis RR, Pinder SE, Ellis IO, et al. Prediction of local recurrence of ductal carcinoma in situ of the breast using five histological classifications: A comparative study with long follow-up. Human Pathology 1998;29:915-23.
20. Bijker N, Bijker JL, Peterse N, vandeVijver M, Julien JP. Classification of ductal carcinoma in situ of the breast in the EORTC 10853 trial: Do subtypes predict risk of recurrence? Results of the central pathology review of the EORTC 10853 DCIS trial. European Journal of Cancer 1998;34:21.
21. National Coordinating Group for Breast Screening Pathology. Classifying epithelial proliferation. In: Pathology reporting in breast cancer screening. No.3 ed. Sheffield: NHS BSP Publications; 1995. p. 18-28.
22. Sneige N, Lagios M, Schwarting R, Colburn W, Atkinson E, Weber D, et al. Interobserver reproducibility (IR) of the Lagios nuclear grading system for ductal carcinoma in situ (DCIS). Modern Pathology 1998;11:A28.
23. Holland R, Peterse JL, Millis RR, Eusebi V, Faverly D, Vandevijver MJ, et al. Ductal Carcinoma in-Situ - a Proposal For a New Classification. Seminars in Diagnostic Pathology 1994;11:167-80.
24. Silverstein MJ, Poller DN, Waisman JR, Colburn WJ, Barth A, Gierson ED, et al. Prognostic Classification of Breast Ductal Carcinoma-in-Situ. Lancet 1995;345:1154-7.
25. Poller DN, Silverstein MJ, Galea M, Locker AP, Elston CW, Blamey RW, et al. Ductal carcinoma in situ of the breast: A proposal for a new simplified histological classification association between cellular proliferation and C-erb b-2 protein expression. Modern Pathology 1994;7:257-62.
26. Silverstein MJ, Lagios MD, Craig PH, Waisman JR, Lewinsky BS, Colburn WJ, et al. A prognostic index for ductal carcinoma in situ of the breast. Cancer 1996;77:2267-74.
27. Douglas-Jones AG, Gupta SK, et al. A critical appraisal of six modern classifications of ductal carcinoma in situ of the breast (DCIS): correlation with grade of associated invasive carcinoma. Histopathology 1996;29:397-409.
28. Scott MA, Lagios MD, Axelsson K, Rogers LW, Anderson TJ, Page DL. Ductal carcinoma in situ of the breast: Reproducibility of histological subtype analysis. Human Pathology 1997;28:967-73.
29. Fisher B, Costantino J, Redmond C, Fisher E, Margolese R, Dimitrov N, et al. Lumpectomy Compared With Lumpectomy and Radiation-Therapy For the Treatment of Intraductal Breast-Cancer. New England Journal of Medicine 1993;328:1581-6.
30. Sibbering DM, Blamey RW. Ch. 36. Nottingham experience. In: Silverstein MJ, editor. Ductal carcinoma in situ of the breast. Baltimore: Williams and Wilkins; 1997. p. 367-72.
31. Faverly D, Holland R, Burgers L. An Original Stereomicroscopic Analysis of the Mammary Glandular Tree. Virchows Archiv a-Pathological Anatomy and Histopathology 1992;421:115-9.
32. Faverly DRG, Burgers L, Bult P, Holland R. 3-Dimensional Imaging of Mammary Ductal Carcinoma in-Situ - Clinical Implications. Seminars in Diagnostic Pathology 1994;11:193-8.
33. Holland R, Hendriks JHCL, Verbeek ALM, Mrauvunac M, Schuurmans Stekhoven JH. Extent, distribution, and mammographic/histological correlations of breast ductal carcinoma in situ. Lancet 1990;335:519-22.
34. Holland R, Hendriks JHC. Microcalcifications Associated With Ductal Carcinoma in-Situ - Mammographic-Pathological Correlation. Seminars in Diagnostic Pathology 1994;11:181-92.
35. Bocker W, Decker T, Ruhnke M, Schneider W. Ductal hyperplasia and ductal carcinoma in situ. Pathologe 1997;18:3-18.
36. Lagios MD. Classification of duct carcinoma in situ (DCIS) with a characterization of high grade lesions: Defining cohorts for chemoprevention trials. Journal of Cellular Biochemistry 1996:108-11.
37. Lakhani SR, Collins N, Stratton MR, Sloane JP. Atypical Ductal Hyperplasia of the Breast - Clonal Proliferation With Loss of Heterozygosity On Chromosomes 16q and 17p. Journal of Clinical Pathology 1995;48:611-5.
38. O'Connell P, Pekkel V, Fuqua S, Osborne CK, Allred DC. Molecular-Genetic Studies of Early Breast-Cancer Evolution. Breast Cancer Research and Treatment 1994;32:5-12.
39. Kasami M, VnencakJones CL, Manning S, Dupont WD, Page DL. Loss of heterozygosity and microsatellite instability in breast hyperplasia - No obligate correlation of these genetic alterations with subsequent malignancy. American Journal of Pathology 1997;150:1925-32.
40. Fujii H, Szumel R, Marsh C, Zhou WB, Gabrielson E. Genetic progression, histological grade, and allelic loss in ductal carcinoma in situ of the breast. Cancer Research 1996;56:5260-5.
41. Chuaqui RF, Zhuang ZP, EmmertBuck MR, Liotta LA, Merino MJ. Analysis of loss of heterozygosity on chromosome 11q13 in atypical ductal hyperplasia and in situ carcinoma of the breast. American Journal of Pathology 1997;150:297-303.
42. Buerger H, Otterbach F, Simon R, Poremba C, Diallo R, Decker T, et al. Comparative genomic hybridization of ductal carcinoma in situ of the breast - evidence of multiple genetic pathways. J Pathol 1999;187:396-402.
43. Lu YJ, Osin P, Lakhani SR, DiPalma S, Gusterson BA, Shipley JM. Comparative genomic hybridization analysis of lobular carcinoma in situ and atypical lobular hyperplasia and potential roles for gains and losses of genetic material in breast neoplasia. Cancer Research 1998;58:4721-7.