—  SYMPOSIUM #14  —

Pathology and Molecular Biology of Pre-invasive Breast Lesions
Moderator: Werner Boecker

Section 3 - Pathology Work-up of DCIS Specimens

Thomas Decker, Muenster , Germany
Daniel Faverly, Brussels , Belgium


Click here to download handout in pdf format for the current section (644 KB)



Introduction
The aim of this presentation is to highlight the importance of pathological examination of surgical specimens in all types of DCIS. With respect to pathological work-up, DCIS is a challenge for the surgical pathologist. Firstly, identifying the lesions is difficult. They are macroscopically nonpalpable and invisible and, therefore, as a rule, detected due to microcalcifications at mammography. Secondly, breast-conserving therapy (BCT) is a serious option in DCIS. However, the therapeutic decision to proceed with BCT depends on a number of pathologic features. There is no doubt that all the information contained in the pathology report primarily depends on the correct sampling of specimens. This requires optimal macroscopic work-up and sampling of surgical specimens. Microscopic identification of the radiological target lesion is a prerequisite. The data obtained is of decisive importance for planning further clinical management. Inadequate evaluation of surgical specimens can lead to both over and under-treatment. Only by providing all clinically relevant data can the pathologist meet his commitments as a member of a multidisciplinary breast team.

Principles of Distribution of DCIS in the Breast
Knowledge of the lobar architecture of the mammary gland is a prerequisite for understanding DCIS. In 1975 Wellings et al. were able to show by means of extensive submacroscopic examinations of 196 breasts with DCIS [1] that, with few exceptions, DCIS takes its origin from the terminal duct lobular unit (TDLU), starting with distension of the ductular structures and unfolding of lobules by the proliferating tumor cells. Further expansion then leads to involvement of the extralobular ductal system. Thus, the neoplasia spreads within the pre-existing ductal lobular system. It follows that the very architecture of the mammary ductal system is one of the key factors that determine the varied distribution patterns observed in these tumors [1, 2]. The ductal-lobular system forms segments from the nipple to the periphery, which appear to be pyramid-shaped: the base lying in the periphery of the breast with its peak pointing towards the nipple [3].

Using computer-assisted three-dimensional reconstruction, Ohtake and coworkers were able to show that the various segments are of different size and usually overlapping in their parenchymal structures [4]. Therefore, these segments do not follow the geometry of the artificial system of quadrants. Furthermore, an analysis of complete galactograms of 85breasts [5] disclosed that 37% of segments are aligned in an outer subcutaneous layer, 41% in an inner layer beneath this, and 22% are aligned centrally. This explains why segmentectomy usually contains several independent segments [6].

Size and intraglandular extension of DCIS
Due to the complex architecture of breast tissue only three-dimensional analysis will yield realistic DCIS extension patterns [3, 4, 6, 7]. Based on the technique of Wellings mentioned above, the Nijmegen group developed an original stereomicroscopic method for investigating the mammary gland. It enabled reconstruction of the glandular tree with its lobes, segments and lobules and, in addition, correlation of 3D-findings with conventional 2D-histology as well as detailed radiograms of specimen slices (Fig. 1).

Thus, data from two such studies has provided evidence that DCIS most frequently occurs within the same segment [8, 9]. In contrast, DCIS foci are only rarely fo und in clearly separate segments [10] (Fig. 2).

Findings by Ohtake and coworkers who examined the intraductal growth of DCIS components in invasive carcinomas [6] also indicate that a total of 81.3% of DCIS showed some type of intraductal expansion towards the nipple; and that tumors obviously also spread in a retrograde fashion into the periphery of segments. In addition, Ohtake et al. established so-called network models and determined the maximum angle of the segment of DCIS growth from the mamillary pole towards the periphery [4]. In all cases cone or pyramid-like shapes with a broad base towards the periphery were found. Furthermore, overlap of ductal lobular systems from different segments could be detected, as well as connecting anastomoses at different levels of the ductal system.

Given such growth patterns, even DCIS sizes of up to 100mm could be explained in terms of unicentric lesions with continuous segmental growth via connecting channels between different segments. Although the literature on DCIS sizes (including those treated with mastectomy) is limited, large lesions seem to account for a considerable number of cases. Holland and coworkers found that 51% (42/82) of DCIS lesions were more than 50mm in diameter 11. In the course of the CMP-CHIREC study undertaken by Faverly et al. a full histological revision of 80 consecutive surgical specimens with DCIS was performed (unpublished data). All patients had previously been subjected to a vacuum-assisted biopsy, and none of them had any radiological contraindication for BCT. The mean size of DCIS was 24.4 mm, 37 % of the lesions did not show any residual tumor after the vacuum-assisted biopsy, whereas 63 % did and were larger than 10 mm. In 36 % of all patients the diameter of DCIS was larger than 30 mm. This data was confirmed by the Berlin-Buch experience: The following data on the sizes of 411DCIS is derived from excision and re-excision specimens as well as from (primary or secondary) mastectomy specimens. About 27.2% of DCIS were smaller than 15mm in diameter, while 16.3% were 16–40mm and 56.4% were 40mm or larger (Table 1).
Table 1: Pathological assessment of size in 411 patients*
0-15 mm 15-40 mm >40 mm Total
112 (27.2%) 67 (16.3%) 232 (56.4%) 411 (100%)

*1993–2002 Berlin-Buch Breast Center, including all DCIS patients irrespective of therapy.

In more than half of the DCIS cases (56%: 232/411), the lesions were fo und to be larger than 40mm. For results of both groups see Fig. 3.

Multifocality and multicentricity
In DCIS cases treated by breast-conserving therapy discontinuous growth may cause serious clinical problems. DCIS foci discontinuous to the main lesion may remain in the breast as residual tumors and cause local recurrence. In order to estimate the risk of such recurrence, it is essential to consider DCIS expansion patterns and the likelihood of distant foci elsewhere in the breast.

Unfortunately, due to the use of inconsistent definitions, the data in the literature concerning the incidence of multicentricity of DCIS ranges from 0to 78% [12, 13, 14, 15]. We define multicentricity as the presence of two or more separate foci in the breast more than 40mm apart. Holland et al. detected only a single case in a total of 60 mastectomy specimens containing two foci of completely isolated DCIS that fulfilled the above criteria (Fig. 4). In 232mastectomy specimens of DCIS patients of the Berlin-Buch Breast Unit, we found two cases with multicentric foci that were more than 45mm apart. Multicentric DCIS as defined above, therefore, seems to be the exception. Furthermore, this conclusion is supported by data showing that local recurrence after breast-conserving therapy usually occurs within the area of previous surgery and that bilateral DCIS is extremely rare [14, 16].

In another study 60cases of DCIS were analyzed with a combination of radiological and pathological techniques in a three-dimensional approach [7, 9]. In 50% of cases only continuous growth expansion was fo und, while the other half displayed discontinuities (gaps) in their intraductal growth. When the distances between discontinuous DCIS areas were measured, their length was less than 5mm in 63%, and less than 10mm in 83% of cases. In total, only 8% of DCIS cases showed discontinuities between DCIS foci of more than 10mm. Furthermore, Faverly and coworkers fo und a correlation between the respective grade of a given DCIS and the frequency and size of gaps between foci. While 90% of poorly differentiated DCIS (high-grade) showed no gaps at all, 70% of well-differentiated (low-grade) DCIS did. In addition, the gaps were shorter in poorly differentiated lesions, while intermediate and well-differentiated DCIS displayed larger gaps. All gaps of more than 10mm were found in the latter group. The only case of multicentric DCIS with a distance of more than 40mm between foci was a well-differentiated DCIS.

In summary, based on the architecture of the ductal lobular tree, the expansion of DCIS can be seen to start within the terminal ductal lobular unit (TDLU) and extend in different directions, preferentially towards the nipple. Therefore, DCIS growth occurs within one segment due to the anatomic lobar units of the mamillary gland. However, DCIS can involve several lobes or segments via intersegmental anastomoses and thus affect a larger area of breast tissue.

Examination of Excision Specimens
As a rule, DCIS grade, the type of calcification associated, and its correlation with mammography are nowadays determined preoperatively using minimally invasive biopsy techniques (MIB). Therefore the vast majority of excision specimens in modern breast medicine are performed for therapeutic reasons as segmental resections after MIB, rather than for diagnostic reasons.

These specimens are generally larger than diagnostic excisions or the "lumpectomies" of earlier times. Complete processing of such specimens may be ideal but seems to us impossible in daily practice. Anyway, in DCIS cases the pathologist is f undamentally confronted with three tasks: the lesion must be fo und and classified, the size of the lesion must be determined and, finally, resection margins must be assessed. There is no doubt that all the information contained in the pathology report primarily depends on the correct sampling of specimens. The following protocol for "systematic and guided" work-up of DCIS operational specimens is based on the principles of distribution of DCIS in the breast as demonstrated in the first part of our presentation. This systematic pathological examination of the excision specimen considers the segmental architecture of the ductal-lobular system, and it is immediately obvious that the ducts of the segment contained in a given specimen are generally oriented along virtual lines between the nipple and the periphery. In addition the associated microcalcifications are used as guidance for sampling [17].

Interdisciplinary Prerequisites
In order to examine DCIS lesions appropriately, the pathologist needs the support of radiologists and surgeons: Before he can determine the histological type of DCIS, the lesion has to be found within the specimen. As mammographic microcalcifications are the primary indication of DCIS, the mammographic report and images as well as the specimen radiograph are needed to identify the target lesion. For determining the size of DCIS lesions and for margin assessment an intact surgical specimen is required. Cuts and tears in the specimen can impede reliable diagnosis, while fragmentation renders it virtually impossible. Immediately after removing the surgical specimen the surgeon should mark the mamillary margin of resection. The pectoral fascia should be removed in such a way that the dorsal resection margin is formed by the fascia. The external surface of this dorsal (fascial) resection margin should be marked with blue dye. Ideally, the surgeon then places the excision specimen on a foil with a schematic drawing of the breast outlines. Alternatively the surgeon can mark the surface of the six margins with sutures of different lengths, colors or materials to ensure proper orientation.

After performing specimen radiography the microcalcification area is marked with several pins. This marking for the pathologist is done independently of the preoperative marking for the surgeon.

Systematic and guided examination
The systematic pathological examination of the excision specimen has to consider the ductal-lobular extension and segmental growth of the DCIS with special emphasis on resection margins. Generally, the ducts run from the nipple to the peripheral resection margin. If the specimen is sliced perpendicularly to the mamillary-peripheral axis, it is quite obvious that most ducts will be cross-sectioned, which increases the chance of recognizing as many ducts as possible within the histological slide (Fig. 5). To be able to identify non-visible DCIS detected as a result of mammographic calcifications, the pathologist must be guided, for example, by pins marking the microcalcification area observed in the specimen radiograph.

Measurement of Specimens
The measurement of the specimen's dimensions – the lengths of the mamillary-peripheral axis and the dorsal-ventral extension – is essential for later reconstruction of the size of a DCIS and should therefore be the first step in examination.

Marking of margins
Marking the surface of excision tissue is essential for microscopic identification of the original resection margins. This can be done by painting the specimen's surface with marker substances that adhere to the tissue during fixation, dehydration, embedding, cutting and staining and must also be visible at microscopy. Various substances such as ink, gelatin, dyes, latex (correction fluid) or silver nitrate can be used. If marking of the surface is done by the surgeon before specimen radiography, radio-opaque substances (latex, silver nitrate) should be avoided. As a rule, all these materials adhere better to fresh than to fixed tissue. To avoid seeping of the marking substance into crevices of the surface, the specimen should not be dipped. Rather, the surface marker should be applied carefully with a small brush or a cotton applicator. It is possible to mark different margins by using different colors for recognition on the microscopic slide.

Slicing
The specimen should be serially sectioned from the mamillary pole to the periphery of the segment. Care should be taken to slice the tissue thinly (about 4mm). Slice radiographs (in addition to the radiograph of the intact specimen) provide an even more precise orientation by identifying individual slices that contain microcalcifications.

Sampling
It is obvious that tissue samples with microcalcifications (as incidentally macroscopically suspicious findings) must be examined. In our protocol, all slices with microcalcifications identified by slice radiography must be embedded completely for histological examination. Due to the discrepancy between the size of areas of microcalcification and of the histologically detected DCIS, we also embed the corresponding tissue of the neighboring slices. Furthermore, assessment of the resection margins of all of these slices is mandatory. Finally, the two first mamillary as well as the two last peripheral slices are embedded for assessment of 10mm margins in these directions (Fig.5).This means that slices of the nipple and peripheral margins are therefore embedded tangentially for processing. Two cassettes for the mamillary slice and four to six cassettes for the peripheral slice are usually required. The other resection margins are contained in the slices cut perpendicular to the long axis of the resection specimen. We also include the dorsal and ventral margins in the examination when the pectoral fascia is removed. This is indispensable for three-dimensional size determination and for quality control of the surgical techniques. Both three-dimensional reconstruction of the lesion and exact information about the direction of an involved margin require recognition of the original location of every tissue block. Since the color of the surface marking substance is not sufficient for this purpose, we use a coding system to designate the original slice location of the excision specimen examined: the serially sectioned slices are labeled by ascending numbers from mamillary to peripheral slice. Abbreviations are used to designate the direction of the margin (ma = mamillar, p = peripheral, cra = cranial, cau = caudal, d = dorsal, v = ventral, m = medial and l = lateral). For example, the code 39004/06-3-cau would indicate patient number 39004 from year 2006, with tissue of the third slice containing the caudal resection margin.

Analysis of pathological data of 299 DCIS excision specimens
We analyzed 299excision specimens of patients with DCIS worked up pathologically according to the protocol described above between 1993 and 2003. Of the primary excision specimens, 254of299 showed DCIS exactly at the resection line or within the 10mm margin. In each of these cases a second operation was carried out, if possible as a re-excision. In cases with an unfavorable relation of the lesion to breast size we advised patients to undergo mastectomy. In more than 85% of cases (254/299) residual DCIS could be detected in the re-operation specimen. (Table 1).

Table 2: Residual tumor after involved margins* in primary excisions in 299 DCIS patients

Re-excision Mastectomy (%) Total
Residual tumor 142 67 209 (82.3 %)
Tumor-free 45 0 45 (17.7 %)
Total 187 67 254 (100.0 %)
*DCIS within10mm

If one compares the results the difference between re-excisions and mastectomies is rather striking: whereas there were 17.7 % tumor-free re-excisions, secondary mastectomies were never tumor-free. This can be attributed to the fact that whenever a decision in favor of mastectomy was made, it did not solely depend on the margin status, but other factors as well (see below).

Width of Tumor-free Margins
The tumor-free margin of the surgical specimen in BCT correlates with the risk of residual tumor and the frequency of recurrence [18, 19, 20, 21, 22]. The concept of local recurrence developing from remaining DCIS in situ is based on this correlation, which has been demonstrated in numerous studies. Nevertheless, opinions differ as to the optimal minimal distance, and there is no generally accepted standard for determining this parameter. Not until the 1980s when breast-conserving therapy was introduced for DCIS was special attention paid to resection margins. The NSABP B-17 protocol only stipulated that the DCIS must not be transected [23]. This requirement was certainly insufficient: patients whose DCIS had been treated by excision without radiotherapy suffered a local recurrence rate of 43% [24]. Tumor-free margins suggested by other experts were equally small: Lagios [25] und Silverstein [26] originally chose 1mm, Solin 2mm [27]. However, as early as 1994 Silverstein documented that a margin of 1mm was insufficient [28]. Residual tumor was found in 45% of mastectomy and re-excision specimens in these cases. Holland and coworkers were able to prove in 1985 that 1mm wide resection margins involved a high risk in view of the large amounts of residual tumor tissue [29]. Therefore they proposed leaving a rim of normal glandular tissue between tumor and resection line as a criterion for a safer excision. Subsequent studies by Faverly and co-workers from the same working group provided an explanation for this finding: discontinuities in the intraductal neoplastic growth [9]. The most important finding relating to the debate on histological minimal distances is that only in 8% of all DCIS cases gaps of more than 10mm occur. Theoretically, excision with a tumor-free margin of 10mm should remove DCIS with a probability of 90%. This pathomorphological data corresponds with the clinical results presented by the Van Nuys group in 1996 [30], the Nottingham group in 1997 [31] and the combined Children's Hospital and Van Nuys study in 1997 [(32)]. In DCIS excisions with larger than 10mm-wide, tumor-free margins the rate of local recurrence ranged from 8 to 6%, or was even as low as 5%. Correspondingly, local recurrence rates are higher in cases with shorter minimal distances to the resection margins. Nevertheless, several studies have yielded results that are in glaring contrast, among them the prospectively randomized NSABP B-17 study. This study revealed hardly any, or no correlation between the width of the histological tumor-free margins and residual tumor or the risk of recurrence [33]. Recent data published by the Boston group [34] shows a 5-year local recurrence rate of 12 % in cases of non-high grade DCIS excised with 10 mm free margins. Discrepancies among this data may be chiefly attributed to three reasons:
  1. Different techniques used to examine resection margins;

  2. Variations in the extent of margin assessment; and

  3. Varying minimum requirements regarding the tumor-free margin.
Our experience with excisions of 299 DCIS is that all cases with complete primary excision as well as all re-excision specimens with 5mm wide tumor-free margins showed even 10mm wide DCIS-free zones along the resection margins.

We were not able to establish any differences in the rate of residual tumor between the group with direct involvement of margins and those with tumor-free margins of 1mm or 1–5mm.

Most of the 142DCIS cases (85%) that contained tumor within the 10mm margins of the segmental excisions and residual tumor in re-excision specimens showed involvement of more than one of the margins (in several directions). In cases with residual tumor the mamillary resection margin was far more often involved (87%) than all the other five. We consider findings of tumor tissue within a 10mm margin along the resection line (under standardized conditions of systematic and specimen radiograph-oriented slicing) to be an indicator of residual tumor in the breast. The probability of residual tumor increases with the number of affected resection lines as well as with involvement of the margin facing the nipple.

Size of DCIS
As DCIS is considered an obligate precursor lesion of invasive breast carcinoma, the aim of breast-conserving therapy should be complete removal of the lesions. Naturally, the size of the DCIS largely determines the success of such an approach. The greater the extent of the DCIS the larger the excision has to be. Many studies have been performed regarding the relationship between DCIS and specimen size, resection margin findings and recurrence risk [33, 34, 35, 36, 37, 38]. Nevertheless, the available data on the influence DCIS size on treatment outcome is poor, mainly because of methodological problems, inconsistencies in study design, and the influence of patient selection.

Complete excision of the lesion was almost never achieved in patients with DCIS of more than 40mm in diameter. With very few exceptions, these large DCIS showed involvement of another quadrant or of the central portion of the gland. Due to the poor anatomical definition of the central portion of the mammary gland this analysis was restricted to a retroareolar area up to 25mm from the nipple. In cases where a lesion was mammographically detected in an overlapping area of two quadrants, only one quadrant was evaluated. Ninety-eight percent of the DCIS with diameters larger than 40mm were also larger than 60mm. In these cases, even in women with larger breasts, re-excisions were not successful and mastectomies had to be performed. Therefore, we consider a maximum diameter of 40mm to be the upper limit for BCT,independent of breast size.

Residual Tumor Risk Evaluation
DCIS size, resection margin findings and evidence of residual tumor are closely related. The frequency of DCIS within 10mm margins increases with the total size of the DCIS: 100% of the DCIS of less than 15mm in diameter showed 10mm free margins. This could only be achieved after re-excision of nearly half of DCIS with diameters of 16–40mm. In contrast, we were unable to excise DCIS larger than 40mm in diameter with 10mm tumor-free margins. Moreover, in DCIS of 40mm or more, detection of tumor within the 10mm margin with involvement of the mamillary resection margin indicates that the lesion cannot be removed completely by BCT. We could not find any correlation between grade (or comedo necrosis) and residual tumor risk.

By combining margin findings (distance and number of margins involved) and size the pathologist is able to estimate different risks of residual tumor. Figs. 6A-C show three examples for residual tumor risk estimation: A )Low residual tumor risk: Margin >10mm, lesion size <16mm. B) Intermediate residual tumor risk: Margin <5mm, only one margin involved, mamillary margin not involved, lesion size 16–40mm. This combination indicates a reasonable risk of residual tumor; nevertheless there is a good chance that breast conservation by completing re-excision will be successful. C) Highest residual tumor risk. Margin <5mm, multiple margins involved (mamillary margin included), lesion size >40mm. This combination indicates a very high risk of residual tumor; there is no chance of completely excising the DCIS. The patient should be advised to undergo mastectomy with reconstruction.

Conclusions
Biologically, DCIS is considered to be an 'early' lesion without invasion and metastasizing potential. Nevertheless, DCIS is not necessarily a 'small' lesion – it can involve a considerable area of the glandular body. Size is the most important limiting factor for BCT, because there is a close relation between DCIS size and resection margin findings: the larger the DCIS, the greater the risk for involved resection margins. DCIS larger than 40mm in diameter in the primary excision specimen almost always show involved resection margins. Here the resection margin findings are more than just an indication for re-excision. DCIS of more than 40mm with involved resection margins under 5mm always extend into the retromamillary central glandular body and/or other quadrants of the gland and are therefore in total much larger. Starting at 40mm, size becomes the essential risk marker for residual tumor. The complete removal of such lesions can only be achieved by (subcutaneous or skin-sparing) mastectomy. In contrast, in our experience, small DCIS with diameters of less than 40mm can be excised completely with 10mm-wide free resection margins if preoperative planning is performed according to our protocol [17] (Fig. 7).

One of the most interesting questions of modern practical management of DCIS is the quantitative analysis of vacuum-assisted biopsies to estimate risk factors such as size and correlation with microcalcification before the surgical intervention.

Finally, questions of individualized therapy within BCT are currently a topic of discussion. Which patients do not need additional radiotherapy? Which patients really benefit from additional tamoxifen therapy? Any attempt to answer these questions must be based on prospective pathological data. In addition to the histopathological classification and biological markers, data on size and quantity of the lesions is und oubtedly of importance, especially with reference to resection margins.

References
  1. Wellings SR, Jensen HM, Marcum RG. An atlas of subgross pathology of the human breast with special reference to possible precancerous lesions. J Natl Cancer Inst 1975;55:231-73.

  2. Moffat DF, Going JJ. Three dimensional anatomy of complete duct systems in human breast: pathological and developmental implications. J Clin Pathol 1996;49:48-52.

  3. Mai KT, Yazdi HM, Burns BF, Perkins DG. Pattern of distribution of intraductal and infiltrating ductal carcinoma: a three-dimensional study using serial coronal giant sections of the breast. Hum Pathol 2000;31:464-74.

  4. Ohtake T, Kimijima I, Fukushima T, et al. Computer-assisted complete three-dimensional reconstruction of the mammary ductal/lobular systems: implications of ductal anastomoses for breast-conserving surgery. Cancer 2001;91:2263-72.

  5. Lanyi M. Die gesunde Brust. In: Brustkrankheiten im Mammogramm: Diagnostik und pathomorphologische Bildanalyse. Berlin, Heidelberg , New York: Springer; 2003. p.9-13.

  6. Ohtake T, Abe R, Kimijima I, et al. Intraductal extension of primary invasive breast carcinoma treated by breast-conservative surgery. Computer graphic three-dimensional reconstruction of the mammary duct-lobular systems [see comments]. Cancer 1995;76:32-45.

  7. Faverly DR, Holland R, Burgers L. An original stereomicroscopic analysis of the mammary glandular tree. Virchows Arch A Pathol Anat Histopathol 1992;421:115-9.

  8. Holland R, Hendriks JH, Vebeek AL, Mravunac M, Schuurmans Stekhoven JH. Extent, distribution, and mammographic/histological correlations of breast ductal carcinoma in situ. Lancet 1990;335:519-22.

  9. Faverly DR, Burgers L, Bult P, Holland R. Three dimensional imaging of mammary ductal carcinoma in situ: clinical implications. Semin Diagn Pathol 1994;11:193-8.

  10. Holland R, Faverly D. The local distribution of DCIS of the breast. Whole organ studies.In: Silverstein M (ed.), DCIS of the breast. Philadelphia , Baltimore, New York: Lippincott Williams & Wilkins; 2002. Ch. 19 p.240-248

  11. Holland R, Hendriks JH. Microcalcifications associated with ductal carcinoma in situ: mammographic-pathologic correlation. Semin Diagn Pathol 1994;11:181-92.

  12. Carter DJ. Intraductal Papillary Tumors of the Breast. A Study of 78 Cases. Cancer 1977;39:1689-92.

  13. Lagios MD, Westdahl PR, Margolin FR, Rose MR. Duct carcinoma in situ. Relationship of extent of noninvasive disease to the frequency of occult invasion, multicentricity, lymph node metastases, and short-term treatment failures. Cancer 1982;50:1309-14.

  14. Rosen PP, Braun DW, Jr., Kinne DE. The clinical significance of pre-invasive breast carcinoma. Cancer 1980;46:919-25.

  15. Schwartz GF, Patchefsky AS, Feig SA, Shaber GS, Schwartz AB. Clinically occult breast cancer. Multicentricity and implications for treatment. Ann Surg 1980;191:8-12.

  16. Page DL, Dupont WD, Rogers LW, Landenberger M. Intraductal carcinoma of the breast: follow-up after biopsy only. Cancer 1982;49:751-8.

  17. Decker T, Boecker W, Ruhnke M. et al. Pathological diagnosis in surgical specimens.In: Boecker W. (ed.), Preneoplasia of the breast. Philadelphia: Saunders; 2006. Ch. 6.

  18. Silverstein MJ, Lagios MD, Groshen S, et al. The influence of margin width on local control of ductal carcinoma in situ of the breast. N Engl J Med 1999;340:1455-61.

  19. Van Zee KJ, Liberman L, Samli B, et al. Long term follow-up of women with ductal carcinoma in situ treated with breast-conserving surgery: the effect of age. Cancer 1999;86:1757-67.

  20. Vicini FA, Kestin LL, Goldstein NS, et al. Relationship between excision volume, margin status, and tumor size with the development of local recurrence in patients with ductal carcinoma-in-situ treated with breast-conserving therapy. J Surg Oncol 2001;76:245-54.

  21. Chan KC, Knox WF, Sinha G, et al. Extent of excision margin width required in breast conserving surgery for ductal carcinoma in situ. Cancer 2001;91:9-16.

  22. Douglas-Jones AG, Logan J, Morgan JM, Johnson R, Williams R. Effect of margins of excision on recurrence after local excision of ductal carcinoma in situ of the breast. J Clin Pathol 2002;55:581-6.

  23. Fisher ER, Costantino J, Fisher B, et al. Pathologic findings from the National Surgical Adjuvant Breast Project (NSABP) Protocol B-17. Intraductal carcinoma (ductal carcinoma in situ). The National Surgical Adjuvant Breast and Bowel Project Collaborating Investigators. Cancer 1995;75:1310-9.

  24. Fisher B, Costantino J, Redmond C, et al. Lumpectomy compared with lumpectomy and radiation therapy for the treatment of intraductal breast cancer. N Engl J Med 1993;328:1581-6.

  25. Lagios MD. Duct carcinoma in situ. Pathology and treatment. Surg Clin North Am 1990;70:853-71.

  26. Silverstein MJ, Waisman JR, Gamagami P, et al. Intraductal carcinoma of the breast (208 cases). Clinical factors influencing treatment choice. Cancer 1990;66:102-8.

  27. Solin LJ, Fowble BL, Schultz DJ, Goodman RL. The significance of the pathology margins of the tumor excision on the outcome of patients treated with definitive irradiation for early stage breast cancer. Int J Radiat Oncol Biol Phys 1991;21:279-87.

  28. Silverstein MJ, Gierson ED, Colburn WJ, et al. Can intraductal breast carcinoma be excised completely by local excision? Clinical and pathologic predictors. Cancer 1994;73:2985-9.

  29. Holland R, Veling SH, Mravunac M, Hendriks JH. Histologic multifocality of Tis, T1-2 breast carcinomas. Implications for clinical trials of breast-conserving surgery. Cancer 1985;56:979-90.

  30. Silverstein MJ, Lagios MD, Craig PH, et al. A prognostic index for ductal carcinoma in situ of the breast. Cancer 1996;77:2267-74.

  31. Sibbering DM, Blamey RW. Nottingham experience. In: Silverstein MJ, editor. Carcinoma in situ of the breast. Baltimore : Williams & Wilkins; 1997. p.271-84.

  32. Lagios MD, Silverstein MJ. Ductal carcinoma in situ. The success of breast conservation therapy: a shared experience of two single institutional nonrandomized prospective studies. Surg Oncol Clin N Am 1997;6:385-92.

  33. Fisher ER, Dignam J, Tan-Chiu E, et al. Pathologic findings from the National Surgical Adjuvant Breast Project (NSABP) eight-year update of Protocol B-17: intraductal carcinoma. Cancer 1999;86:429-38.

  34. Wong JS, Kaelin C M, Troyan SL, et al. Prospective Study of Wide Excision Alone for Ductal Carcinoma in Situ of the Breast. J Clin Oncol 2006;24:1031-36

  35. Fisher ER. Pathobiological considerations relating to the treatment of intraductal carcinoma (ductal carcinoma in situ) of the breast. CA Cancer J Clin 1997;47:52-64.

  36. Osteen RT. Limits of resection for ductal carcinoma in situ. J Surg Oncol 1998;69:63-5.

  37. Silverstein MJ. Ductal carcinoma in situ of the breast. Br J Surg 1997;84:145-6.

  38. Silverstein MJ. Predicting residual disease and local recurrence in patients with ductal carcinoma in situ. J Natl Cancer Inst 1997;89:1330-1.

  39. Schnitt SJ, Connolly JL. Classification of ductal carcinoma in situ: striving for clinical relevance in the era of breast conserving therapy. Hum Pathol 1997;28:877-80.

  40. Tornos C, O'Hea B. Ductal carcinoma in situ (DCIS) of the breast: pathologic features predictive of residual disease after initial excisional biopsy. Mod Pathol 2000;13:48A