SEARCH

SEARCH BY CITATION

INTRODUCTION

  1. Top of page
  2. INTRODUCTION
  3. CLASSIFICATION
  4. SIZE (EXTENT, DISTRIBUTION) AND MARGINS
  5. OPTIMAL TISSUE PROCESSING
  6. INVASION
  7. BIOLOGIC MARKERS
  8. FUTURE DIRECTIONS
  9. Acknowledgements

Ductal carcinoma in situ (DCIS) constitutes a spectrum of non-invasive proliferative epithelial lesions with a predilection for the terminal duct-lobular units of the breast. Influenced by the ubiquitous use of screening mammography and emerging from relative obscurity only a generation ago, DCIS now represents up to one-fourth of breast cancer diagnoses in some institutions. Since it is non-invasive, it lacks the ability to metastasize, but it has been considered an initial step in a process that has the potential to progress to invasive cancer if not treated. For this reason, mastectomy, until recently, was the standard treatment for this disease.

Because it has been documented that only a minority of patients develop invasive cancer after excision of DCIS, some physicians have been treating selected patients with DCIS by local excision only, or local excision and radiation therapy, rather than by mastectomy. Differences in treatment recommendations abound because physicians have been unable to determine precisely which patients with DCIS fit best into each of these treatment categories, and if any women are well served by a recommendation for treatment that does not include the whole breast. Furthermore, pathologists do not universally agree upon the features that define DCIS or those that might predict the subsequent risk of local recurrence or invasive cancer. Therefore, it is not surprising that optimal treatment is equally controversial.

To address these various issues, an international consensus conference was convened in Philadelphia, sponsored by The Breast Health Institute and The Fashion Group International, Philadelphia, and held at Thomas Jefferson University on April 25-28, 1997. Co-sponsor was the Department of Pathology of Jefferson Medical College/Thomas Jefferson University Hospital. The above named panelists, representing the disciplines of surgical pathology, surgery, breast imaging (radiology), radiation oncology, and biostatistics constituted the panel.

The panelists attempted to reach consensus on the following issues:

  • 1
    The pathologic classification of DCIS
  • 2
    The identification of specific features that may convey prognostic significance
  • 3
    Methods for determining size or extent of DCIS
  • 4
    Margin assessment and the confirmation of complete local excision
  • 5
    The appropriate manner in which surgical specimens from patients with suspected DCIS (mammographic abnormalities) should be processed from the time they are removed from the patient until the final microscopic slides are reviewed by the pathologist and a report rendered.

The conference participants recognized that there are situations in which the distinction between DCIS and atypical ductal hyperplasia, on one end of this spectrum, and between DCIS and DCIS with "microinvasion," on the other end, is difficult. Nevertheless, for the purposes of this conference, the panel limited its focus to DCIS. Atypical duct hyperplasia was excluded from consideration, as was microinvasive carcinoma. Any lesion that has any evidence of microinvasion, however limited, is classified as invasive carcinoma (pT1mic).

CLASSIFICATION

  1. Top of page
  2. INTRODUCTION
  3. CLASSIFICATION
  4. SIZE (EXTENT, DISTRIBUTION) AND MARGINS
  5. OPTIMAL TISSUE PROCESSING
  6. INVASION
  7. BIOLOGIC MARKERS
  8. FUTURE DIRECTIONS
  9. Acknowledgements

In the past several years, a number of new classification systems for DCIS have been proposed. In contrast to the traditional system of classification, primarily based upon architecture, newer classification systems stratify DCIS primarily on the basis of nuclear grade, and secondarily, on the presence and amount of necrosis and cell polarization. Any classification system for DCIS should reflect the biologic potential of these lesions for local recurrence and/or progression to invasive carcinoma. While the panel did not endorse any one system of classification, the consensus reached by the panel did recommend that the following features be documented in a pathology report that confirms the presence of DCIS:

  • 1
    Nuclear grade
  • 2
    Necrosis
  • 3
    Polarization
  • 4
    Architectural pattern(s)

The panel also recommended that the following information associated with the DCIS also be documented in the report:

  • 5
    Margins (distance from any margin to the nearest focus of DCIS; focal or diffuse involvement of margins)
  • 6
    Size (extent and distribution of DCIS)
  • 7
    Microcalcifications associated with DCIS and calcifications outside the area of DCIS
  • 8
    Correlation of the tissue specimen with specimen x-ray and mammographic findings
  • 9
    Consideration of a synoptic report for DCIS was suggested

The classification of DCIS should reflect its biology by establishing a grading system with definable clinical outcomes. Endpoints to consider in the assessment of the biology of DCIS include, but are not limited to, the following:

  • 1
    Probability of local recurrence (LR)
  • 2
    Risk of subsequent mastectomy
  • 3
    Probability of subsequent invasive carcinoma
  • 4
    Breast carcinoma specific mortality

Such a classification would be a standard for studying (1) the natural history of the disease, including average time intervals to the various outcomes, and (2) treatment options, including the need for mastectomy following failed breast conservation.

  • Definitions

    • Nuclear grade

      The panel recommended that DCIS should be stratified primarily by nuclear grade.

      • Low grade nuclei (NG 1)

        Appearance:

        Monotonous (monomorphic).

        Size:

        1.5-2.0 normal RBC or duct epithelial cell nucleus dimensions.

        Features:

        Usually exhibit diffuse, finely dispersed chromatin, only occasional nucleoli and mitotic figures. Usually associated with polarization of constituent cells.

        Caveat:

        The presence of nuclei that are of similar size but are pleomorphic precludes a low grade classification.

      • High grade nuclei (NG 3)

        Appearance:

        Markedly pleomorphic.

        Size:

        Nuclei usually >2.5 RBC or duct epithelial cell nuclear dimensions.

        Features:

        Usually vesicular and exhibit irregular chromatin distribution and prominent, often multiple nucleoli. Mitoses may be conspicuous.

      • Intermediate grade nuclei (NG 2)

        Nuclei that are neither NG 1 nor NG 3.

    • Necrosis

      Necrosis has been shown in a number of studies to modify the risk associated with nuclear grade. The participants recommend that it be included in the features cited in the pathology report.

      Definition of necrosis

      Presence of ghost cells and karyorrhectic debris. The panel agreed that these are important features distinguishing necrotic debris from secretory material.

      Necrosis quantification

      Comedonecrosis:

      Any central zone necrosis within a duct, usually exhibiting a linear pattern within ducts if sectioned longitudinally.

      Punctate:

      Non-zonal type necrosis (foci of necrosis that do not exhibit a linear pattern if longitudinally sectioned).

    • Cell polarization

      Polarization reflects the radial orientation of the apical portion of tumor cells towards intercellular (lumen-like) spaces, either larger lumina or minute "microacinar" spaces which produce a rosette-like appearance. Such polarization is characteristic of lower grade DCIS with cribriform and solid architecture, but can also be recognized in epithelial protuberances, bridges, arcades and micropapillae of DCIS of lower grades with micropapillary architecture.

      The group recognized that these features, A through C, would apply to the majority of recognized DCIS, but that a small proportion of cases, such as apocrine and signet ring cell types are not so easily classified. For these subtypes stratification is not yet established.

    • Architectural pattern

      The consensus panel recognized that architectural pattern alone does not stratify DCIS satisfactorily with regard to outcome. Moreover, there is not a consistent association with nuclear grade, so that any pattern may contain any nuclear grade. However, there is some evidence that micropapillary architecture, when present in its pure form, is more commonly associated with more extensive, multifocal, and multicentric disease. Therefore, the panel agreed that architectural patterns present within a DCIS lesion should be cited in the report. In cases of a multiplicity of patterns within the same lesion, they should be listed in order of decreasing amounts, and the notation made that there are several patterns.

      The following architectural patterns were accepted by the committee:

      • a
        Comedo
      • b
        Cribriform
      • c
        Papillary
      • d
        Micropapillary
      • e
        Solid

Necrosis may occur in any of these patterns. The term comedo refers specifically to solid intraepithelial growth within the basement membrane with central (zonal) necrosis. Such lesions are often but not invariably of high nuclear grade. This is a major departure from prior systems of classification that used the term comedo to indicate both central necrosis and high grade nuclear features. This current system permits the separation of nuclear grade from architecture, when appropriate.

  • 2
    Heterogeneity of Nuclear Grade The nomenclature employed should reflect the highest nuclear grade, but the panel recognized that some DCIS exhibits heterogeneity of nuclear grade. It is presently unknown if the relative proportion of each nuclear grade affects outcome. The panel agreed that a report could cite additional nuclear grades present in the lesion and include the apparent proportion of the highest grade component in such circumstances.

SIZE (EXTENT, DISTRIBUTION) AND MARGINS

  1. Top of page
  2. INTRODUCTION
  3. CLASSIFICATION
  4. SIZE (EXTENT, DISTRIBUTION) AND MARGINS
  5. OPTIMAL TISSUE PROCESSING
  6. INVASION
  7. BIOLOGIC MARKERS
  8. FUTURE DIRECTIONS
  9. Acknowledgements

The panel unanimously concurred that the size or extent of DCIS and the status of margins should be addressed in the pathology report regardless of the methods of assessment. Both, but particularly margin status, have been shown to be related significantly to outcome.

However, the panel also recognized that several methods to assess size and margins exist. These depend upon surgical orientation and various processing techniques in individual surgical pathology laboratories.

OPTIMAL TISSUE PROCESSING

  1. Top of page
  2. INTRODUCTION
  3. CLASSIFICATION
  4. SIZE (EXTENT, DISTRIBUTION) AND MARGINS
  5. OPTIMAL TISSUE PROCESSING
  6. INVASION
  7. BIOLOGIC MARKERS
  8. FUTURE DIRECTIONS
  9. Acknowledgements

The traditional surgical procedure has been a needle-guided (wire directed) localization and excision of an area of calcifications or mass with calcifications, with a specimen radiograph confirming the successful excision of the suspicious area. The following steps are considered requisite for the processing of that specimen.

  • 1
    Whenever feasible, the specimen should be oriented by the surgeon for the pathologist. Addressing the segmental anatomy of the ductal system of the breast, some panelists recommended identifying the radial axis with a suture sewn into the nipple edge of the specimen.
  • 2
    Whenever feasible, correlation of the specimen radiograph with the preoperative mammograms should be performed, usually by the radiologist, and this information should be communicated to the pathologist. The specimen radiograph must be available to the pathologist for radiographic-pathologic correlation. It is often helpful to place a metallic clip on the specimen near the site of the calcifications if they are faint or only within a very small area of the entire specimen. The additional specimen radiograph with the clip in place is then sent to the pathologist to aid the orientation of the specimen.
  • 3
    If the surgical margins are not submitted as separate specimens (vide infra), the surfaces of the specimen should be inked to permit orientation of the histologic sections. Some pathologists find that different colored inks are preferable to a single color to facilitate orientation when the tissue is examined microscopically. However employed, the principle is the careful identification of the margins of the specimen to be able to measure distance from the surgical margins to any focus of DCIS.
  • 4
    There are several ways in which the surgical specimen may be processed. When feasible and practical the entire specimen should be processed in sequence in separate cassettes. This is the best way to assess the size of the area of DCIS encountered. Using this technique, the entire specimen is sectioned transversely (breadloafed) by the pathologist into uniform sections, preferably 2 to 3 mm. Such an exhaustive examination may not be possible in all cases. Alternatively, the specimen should be sectioned, the slices radiographed, and only the slices demonstrating the mammographic abnormality and the immediately adjacent slices are submitted for the initial histologic examination. The remaining tissue may then be processed only in cases in which ductal carcinoma in situ is identified in these initial sections. This will then determine the size and extent of the lesion, evaluate the microscopic margins, and rule out invasion.
  • 5
    Determination of size or extent of DCIS is usually an estimate. This may be approximated by the number of segments, each approximately 2-3 mm thick, that contain DCIS microscopically. If DCIS is present only on a single slide, the greatest dimension of the lesion should be measured and reported. However, if DCIS is present on more than one slide, determination of size and/or extent of DCIS depends upon the manner in which the specimen has been processed.
    If DCIS is associated with characteristic linear, branching, casting type microcalcifications and margins are involved, the preoperative mammograms may provide a minimum estimate of DCIS size or extent. When the mammograms are used to estimate the size of DCIS, the largest area of calcifications may be measured. A single measurement of the maximum extent of calcifications should be recorded. The method of determining tumor size should be noted in the report, i.e. clinical palpation, mammogram, tissue specimen macroscopic or microscopic.
  • 6
    When the margins are not separately submitted by the surgeon (see number 7, below), the margins of the single specimen are determined by measurements of the distance between the DCIS and an inked resection margin. This narrowest (closest) margin will determine the margin status for a particular resection.
  • 7
    Some breast surgeons among the panelists believe that the current technique of inking the margins of a single specimen was not the optimal method to examine margins. An alternative technique of margin assessment was presented.

When the surgical specimen is excised in the operating room and removed from the operative field, a cavity remains. Rather than ink the excised tissue to see what has been removed, arcs of additional tissue following the contour of the cavity are excised by shaving the edges of the cavity as if one were peeling an onion from the inside out. These are separately excised, labeled and submitted, usually from medial, lateral, superior, inferior, and deep margins (base of wound). These are processed by the pathologist separately. Any involvement within the tissue submitted as a "marginal biopsy" is considered a positive margin. As many separate specimens may be submitted as desired, although most of the group accepted the five additional specimens as sufficient. Moreover, if a single margin is positive, it can be addressed at a subsequent procedure if breast conservation is to be employed, since the exact location of the positive margin is unequivocal.

INVASION

  1. Top of page
  2. INTRODUCTION
  3. CLASSIFICATION
  4. SIZE (EXTENT, DISTRIBUTION) AND MARGINS
  5. OPTIMAL TISSUE PROCESSING
  6. INVASION
  7. BIOLOGIC MARKERS
  8. FUTURE DIRECTIONS
  9. Acknowledgements

This consensus statement applies to DCIS only and assumes the absence of any invasion, including that which is called "microinvasion." Criteria, definitions and significance of microinvasion were not considered. As noted above, the presence of any invasion, however, "micro," confers a different degree of risk that has not been addressed by these proceedings. If present, it should be identified and quantified in the diagnosis.

BIOLOGIC MARKERS

  1. Top of page
  2. INTRODUCTION
  3. CLASSIFICATION
  4. SIZE (EXTENT, DISTRIBUTION) AND MARGINS
  5. OPTIMAL TISSUE PROCESSING
  6. INVASION
  7. BIOLOGIC MARKERS
  8. FUTURE DIRECTIONS
  9. Acknowledgements

Biologic markers currently have no uniquely definitive role in the management (i.e. the evaluation of therapeutic options or prognosis) of a patient with DCIS. However, the group recognized that this statement may change as data accumulate concerning correlations between the value of these different markers and patient outcome. Several participants in the consensus shared their experience to date, using estrogen and progesterone receptors, nuclear antigen Ki-67, gene products p53 and C-erbB2 as examples.

Now that virtually all markers can be determined using slides made from the paraffin blocks of formalin fixed tissue, irrespective of time between biopsy and these determinations, discarding the paraffin blocks after any length of time was strongly discouraged.

FUTURE DIRECTIONS

  1. Top of page
  2. INTRODUCTION
  3. CLASSIFICATION
  4. SIZE (EXTENT, DISTRIBUTION) AND MARGINS
  5. OPTIMAL TISSUE PROCESSING
  6. INVASION
  7. BIOLOGIC MARKERS
  8. FUTURE DIRECTIONS
  9. Acknowledgements
  • 1
    Further studies are essential to determine how best to combine the features of DCIS, as described above, into a universally acceptable, reproducible, and clinically useful system of classification.
  • 2
    The identification of biologic and molecular markers that predict recurrence, progression to invasion, and response to therapy is an important goal and merits continued study.
  • 3
    Although the lesions discussed here traditionally have been considered "ductal carcinoma in situ," a few of the participants questioned the categorization of these lesions as "carcinoma in situ," recognizing the uncertain biologic potential of these lesions, the difficulty in distinguishing low grade DCIS from atypical hyperplasia in some cases, and the trend away from the use of this term in other organ systems. One suggestion was to use the term, "mammary intraepithelial neoplasia" to encompass the entire spectrum of proliferative breast lesions, including those lesions currently designated as DCIS.

Acknowledgements

  1. Top of page
  2. INTRODUCTION
  3. CLASSIFICATION
  4. SIZE (EXTENT, DISTRIBUTION) AND MARGINS
  5. OPTIMAL TISSUE PROCESSING
  6. INVASION
  7. BIOLOGIC MARKERS
  8. FUTURE DIRECTIONS
  9. Acknowledgements

The conference gratefully acknowledges additional grant support from the following companies:

  Biopsys Medical, Inc.

  Oxford Health Plans

  Zeneca

The members of the Consensus Conference Committee were as follows: Conference Chairman: Dr. Gordon F. Schwartz, Jefferson Medical College, Philadelphia, PA. Conference Co-Chairman: Dr. Michael D. Lagios, St. Mary's Medical Center, San Francisco, CA. Pathologists: Dr. Darryl Carter, Yale University School of Medicine, New Haven, CT; Dr. James Connolly, Beth Israel Deaconess Medical Center, Boston, MA; Dr. Ian O. Ellis, City Hospital, NHS Trust, Nottingham, United Kingdom; Dr. Vincenzo Eusebi, Universita di Bologna, Bologna, Italy; Dr. Gerald C. Finkel, Thomas Jefferson University Hospital, Philadelphia, PA; Dr. Fred Gorstein, Thomas Jefferson University Hospital, Philadelphia, PA; Dr. Roland Holland, University Hospital Nijmegen, Nijmegen, The Netherlands; Dr. Robert V. P. Hutter, Saint Barnabas Medical Center, Livingston, NJ; Dr. Michael D. Lagios, St. Mary's Medical Center, San Francisco, CA; Dr. Shahla Masood, University of Florida Health Science Center, Jacksonville, FL; Dr. Rosemary R. Millis, Guy's Hospital, London, United Kingdom; Dr. Frances P. O'Malley, London Health Science Centre, London, Ontario, Canada; Dr. Juan Thomas Palazzo, Jefferson University Hospital, Philadelphia, PA; Dr. Arthur S. Patchefsky, Fox Chase Cancer Center, Philadelphia, PA; Dr. Juan Rosai, Memorial Sloan-Kettering Cancer Center, New York, NY; Dr. Stuart J. Schnitt, Beth Israel Deaconess Medical Center, Boston, MA; Dr. Roland Schwarting, Thomas Jefferson University Hospital, Philadelphia, PA; Dr. John P. Sloane, Royal Liverpool University Hospital, Liverpool, United Kingdom; Dr. Fattaneh A. Tavassoli, Armed Forces Institute of Pathology, Washington, DC. Mammographers: Dr. Stephen A. Feig, Thomas Jefferson University Hospital, Philadelphia, PA; Dr. Daniel B. Kopans, Massachusetts General Hospital, Boston, MA. Radiation Oncologist: Dr. Beryl McCormick, Memorial Sloan-Kettering Cancer Center, New York, NY. Surgeons: Dr. Edward M. Copeland III, University of Florida College of Medicine, Gainesville, FL; Dr. Armando E. Giuliano, John Wayne Cancer Institute, Santa Monica, CA; Dr. Gordon F. Schwartz, Jefferson Medical College, Philadelphia, PA; Dr. Melvin J. Silverstein, The Breast Center, Van Nuys, CA; Dr. Joop A. van Dongen, The Netherlands Cancer Institute, Amsterdam, The Netherlands. Biostatistician: Dr. Carol Bodian, the Mount Sinai Medical Center, New York, NY.