Abstract
Background: Ductal intraepitelial neoplasia (DIN) represents a spectrum of disease that may progress from usual hyperplasia to ductal carcinoma in situ (DCIS) grade 3. The aim of the study was to asses the correlation between the DIN classification and the surgical treatment including sentinel lymph node biopsy (SLNB). Patients and Methods: In this retrospective study, 229 patients with DIN had undergone conservative or radical surgical treatment and SLNB in cases of DIN1C-DIN3. Results: Breast conservative surgery was the definitive treatment in 80% of the cases. The H&E evaluation of excised sentinel nodes was negative for metastatic disease; neverthless the immunohistochemical (IHC) evaluation revealed the presence of metastatic cells in 6 patients (3.7%). Conclusion: In cases of DIN lesions SLNB is not indicated. The only reason SLNB should be considered is when there is an evidence of invasive foci at definitive histology or when radical mastectomy is proposed.
Ductal intraepithelial neoplasia (DIN) is a growing health problem in western countries where its incidence has showen a 236% increase in the last ten years (1). With the widespread use of screening mammography a dramatic change has occurred in the frequency, management and type of ductal carcinoma in situ (DCIS) detected (2). DCIS is a complex and heterogeneous spectrum of pathological lesions with a widely variable malignant potential and a not yet clearly understood natural history (3). Additionally, breast cancer screening programs have increased the diagnosis of low-grade DCIS in younger women (2). In contrast to the palpable in situ carcinomas of yesteryear, the mammographically detected lesions are invisible to the naked eye. The non palpable stage of DCIS currently represents approximately up to 20% of all new tumors and up to 30% of the screen-detected breast carcinomas (3). At the same time diagnostic and treatment paradigms in breast cancer management have markedly changed with the use of pre-operative core biopsy, the shift from radical mastectomy to breast conserving surgery (BCS), the use of sentinel lymph node biopsy (SLNB) instead of axillary lymph node dissection (ALND) for nodal staging and the routine analysis of prognostic markers (estrogen receptor ER, progesterone receptor PgR, C-erb) (2). While close to 90% of premammography DCIS were of the high-grade comedo type, nearly 60% of mammographically detected lesions are noncomedo, and this percentage is increasing. Relative to the premammographic era (prior to the 1980s) when mastectomy was the treatment of choice, a drastic change has occurred in the management of women with DCIS (2). Since 1997, with increasing support from molecular studies Tavassoli suggested the designation of DIN and offered a translational table for conversion of the currently used terminology of DCIS (2). The term “carcinoma” has been restricted to lesions that have invaded the stroma. “Neoplasia” is defined simply as a proliferation that serves no physiological function. In the DIN system, high-grade DCIS becomes grade 3 DIN, grade two DCIS becomes grade 2 DIN and grade 1 DCIS and atypical intraductal hyperplasia (AIDH) are grouped together as DIN1C. Flat epithelial atypia without intraluminal proliferation is designated as DIN1B, while intraductal hyperplasia (IDH) is designated DIN1A. One important advantage of the DIN terminology is that it incorporates into the system a variety of flat lesions that simply show replacement of the native epithelial cell layer by a single layer of mildly (monomorphous) to severely atypical (polymorphous) cells with or without necrosis. These flat lesions are often missed because of the absence of intraluminal hyperplasia. Although cytologically identical to low-grade DCIS and unequivocally proven to be very similar to adjacent well-differentiated in situ or invasive carcinomas at the molecular level (4), the flat (monomorphous) epithelial atypia remains in the DIN1B category. By definition DCIS is a noninvasive disease and lymphatic metastasis should not occur, but there is no obvious solution to this discrepancy (5). The status of the axillary lymph nodes is the most powerful prognostic indicator of invasive breast cancer (6) but the value of lymph node evaluation in the management of DCIS has been much debated. Routine axillary dissection, in this low-risk group, would be associated with an unacceptably high level of morbidity. The advent of mammographic breast screening with stereotactic/ultrasound (US)-guided core biopsy has provided for preoperative diagnosis and quality-control criteria in screening programs demands a low rate of open biopsy. However, in several studies the histological underestimation of invasive disease has been described in 16-35% of patients with a preoperative diagnosis of DCIS (6-8). This poses a significant problem regarding surgical management because lymphatic staging is the standard of care for breast carcinoma. Some patients with DCIS on diagnostic biopsy may show clinical or radiological suspicion of invasive disease. Nodal staging should be considered in these patients with DCIS who have an increased risk of harboring invasive carcinoma and of axillary metastases (7). A cohort of in situ lesions of the breast was retrospectivally reviewed according to the DIN classification in order to define optimal surgical management.
Patients and Methods
Patient selection and diagnosis. Between April 1997 and January 2007, 229 consecutive patients with mammographic, cyto/microhistological investigations suspicious of preinvasive breast lesion were referred to our institution (The Breast Unit of the Department of Surgery, Tor Vergata University Hospital in Rome, Italy) for surgical characterization and treatment. Data from these patients had been collected in a retrospective breast disease database and were reviewed for this study. The Demographic and assessed risk factors for breast cancer were also collected for statistical analysis. Preoperative evaluation had included mammography with magnification and ultrasonography, and magnetic resonance imaging (MRI) in selected patients. Mammographically detected microcalcification was assigned a score (M1-M5) according to the BIRADS® classification (Breast Imaging Reporting and Data System, it sets up a classification for Level of Suspicion (LOS) for the possibility of breast cancer). The cytology assessed by fine needle ago biopsy (FNAB) and reported using the standard reporting categories: acellular/inadequate, benign, atypical, suspicious and malignant (C1-C5) (26).
Diagnosis and treatment protocols. From April 1997 to 2003 patients with mammographic (M4-M5 microcalcifications, parenchimal distortion, opacities) or cytologyc suspicious of early breast cancer had undergone open biopsy with intraoperative specimen mammography in the case of microcalcifications, and histological diagnosis. From 2003 the patients underwent preoperative 11G vacuum assisted core biopsy (VACB) (Mammotome, Ethicon Endo-Surgery, 11-gauge) of radiologically suspicious sites. The patients with DIN1C or higher grade at VACB, or a negative VACB with a M5 pattern underwent open biopsy. Hook wires or the radio occult lesion localization technique had been used in order to localize non palpable lesions. The patients with a histological diagnosis of DCIS/DIN1C-DIN3 had undergone SLNB. After routine 11G VACB was adopted, open biopsy and SLNB were combined in the same operation and SLNB was performed according to standard procedures. Immediately prior to the 24 hour period before surgery a radioactive tracer, namely 80 to 100Mbq of 99m TC-labelled human albumin microcolloid (Nanocoll; Sorin Biomedica Diagnostics, Saluggia, Italy) in a volume of 0.2 ml, was injected intra and peritumorally sterotatically or under US control. Lymphoscinitigraphy (LSG) was then performed and the estimated location of the SLN was skin marked.
Histopathological examination. The pathological assessment included an evaluation of the primary tumor size, grade and histological type, as well as of the lymph node status after SLNB. The breast resection and mastectomy specimens from all the patients were re-assessed by two pathologists with specific expertise in breast cancer. All the lesions were divided into three classes both according to the European Pathologist Working Group (EPWG) classification (G1, G2, G3), adopted by the European Organization for the Research and Treatment of Cancer (EORTC) and according to Tavassoli et al. classification (DIN1C, DIN2, DIN3), adopted by the World Health Oganization (WHO) (3). The cell size was evaluated as either small or large. Four different architectural patterns were considered: comedo, cribriform, papillary and solid (Table III). Necrosis and microcalcifications were also noted. The extension of cancer cells beyond the basement membrane into the adjacent tissues, with no single focus larger than 1 mm in greatest dimension, was considered microinvasion. To further evaluate the microinvasion, immunohistochemical (IHC) Calponin staining was performed in selected patients, including all those with tumor-positive SLN findings. When possible ER and PgR status, Ki-67 labeling index determined with the MIB1 monoclonal antibody and HER2/neu overexpression were assessed. The HER2/neu overexpressions were evaluated using a 1:100 dilution of a polyclonal antiserum (Dakko, Glostrup Denmark) and considering only moderate to strong complete membrane staining in at least 10% of the neoplastic cells as evidence of overexpression. For the evaluation of ER and PgR status and Ki-67 labeling index, the percentage of cells exhibiting definite nuclear staining in at least 2,000 neoplastic cells examined at ×400 magnification was recorded. Only nuclear immunoreactivity was evaluated for ER, PgR and MIB1. The threshold for HER2/neu overexpression was 10%, for ER and PgR the positivity was 1%, and for MIB1 the positivity was 20% .
Lymphatic analysis. The sentinel nodes were sent to the pathology laboratory as separate samples. The fresh specimens were cleaned of all extracapsular fat tissue, measured and sliced into 1-1.5 mm thick sections perpendicular to their long axis. The tissue was fixed in formalin and embedded in paraffin. Two sections were stained with H&E. When a metastasis 2 mm or larger was found in a frozen section, only H&E sections were made from the paraffin-embedded tissue. If no metastatic tissue was detected or ITC or a micrometastasis was found, a Cam 5.2 immunostain (Becton Dickinson Immunocytometry Systems, San Jose, CA, USA) was performed on the paraffin-embedded tissue in addition to the regular H&E sections. Metastases of 0.2 mm or less were considered as micrometastases and as ITC (9). The nodal status was determined according to the revised tumor-node-metastasis (TNM) staging system for breast cancer as presented in the sixth edition of the American Joint Committee's Cancer Staging Manual (10).
Patient characteristics at the time of presentation.
Tumor characteristics at the time of presentation.
Post surgical treatment. All the DIN1C-DIN3 patients had been referred to an oncologist for post-surgical treatment. External radiotherapy (60Gy) and Tamoxifen for 5 years was proposed to all the patients with DIN1C-DIN3 final histology who had undergone BCS with ER/PgR responsive cancer when no specific contraindication was present and after pelvic-trans-vaginal US evaluation. The patients who had undergone mastectomy were proposed Tamoxifen only for 5 years. Alternatively, Anastrozole or Letrozole or Examestane treatment was used. The patients were followed-up every three months for 2 years, then every 6 months for the following two years and yearly starting from the fifth year. Yearly Rx-mammography, breast US and total body bone scintigraphy were performed.
Statistical analysis. Fisher's exact and Chi-square tests were used to compare the proportional data. The means and medians were compared using the Mann-Whitney U-test. Two-tailed p-values <0.05 were considered statistically significant.
Results
Patients characteristics. In total the 229 patients were included in the study of whom 168 had been diagnosed as DIN1C-DIN3 and had undergone SLNB. The patient characteristic at the time are shown in Table I. Previous pregnancies were reported in 77.5% of the patients recruited (the mean age at first pregnancy: 25, S.D. 7.9; the mean number 1.71, range 0-7). Substitutive hormonal therapy was reported in 6.7% of the postmenopausal women and 7.6% had taken oral contraception for more than 6 months in the 10 years before diagnosis. The age at diagnosis, the age at menarche and menopause, the age at first pregnancy, oral contraception, substitutive hormonal therapy and the dimension of the lesion did not show any association with the DIN grade (Pearson's Chi-square >0.05%).
Main histopathologic characteristics of 229 DIN cases.
Main histopathologic characteristics of the DIN1C-DIN3 lesions.
Histopathological features. The average size of the DIN was 9.4 mm. The rate of multifocality was 35.1% (Table II). The histological aspects identified are listed in Table III. The receptor profile and other markers analysed in the DIN1C-DIN3 lesions are listed in Table IV: the ER/PgR status was available for 92 patients (54.76%), while other markers such as Ki-67 and HER2/neu were analysed in a smaller patient number (n=43, 24%). Histologically, cribriform proliferation and comedonecrosis were significantly more composed to DIN1C and DIN3 (Pearson's Chi-square <0.001). All the SLNs sampled showed an absence of metastatic or micrometastatic nodal involvement by H&E stain while the IHC assessment for cytokeratin showed 3.7% (6/168) positivity. Figure 1 shows the localization of the DIN lesions.
Surgical outcomes for 168 DIN1C-DIN3 lesions.
Surgical treatment. From April 1997 to January 2007 229 patients underwent surgical treatment for DIN lesions. As a first treatment, 118 excisional biopsies, 75 wide resections, 32 quadrantectomies and 4 total mastectomies had been performed. The involved margin rate was 6% for the patients undergoing BCS, giving a clear margin rate of 94% . When the free margins of resection were “close” or less than 5mm further resections had been performed, i.e. 16 wide resections, 12 quadrantectomies and 8 mastectomies were performed. In a single case of DIN1C where the “close” was 2mm from the margin the patient was sent directly to radiotherapy after appropriate counselling. Total, subcutaneous or skin sparing mastectomy were performed in the cases of small sized breast or when there was evidence of multifocality. In six further patients mastectomy was performed as a third treatment. According to the criteria of post mastectomy reconstructive surgery, selected patients underwent prosthetic implant immediately or after skin expansion. The surgical treatments of the patients with DIN1C-DIN3 are shown in Table V. LSG identified the SLN in 166 patients (72.5%), 161 patients showed a typical location of the SLN in the ipsilateral axilla, in 5 cases the SLN was located in the ipsilateral internal mammary chain. No SLN was identified in two patients who had previously undergone a wide resection. The intraoperative use of a gamma detecting probe allowed the identification and sampling of a median number of 1.3 SLN (210 axillary SLN, 6 internal mammary chain SLNs). The overall success rate of the LSG method was 97.7% .
Discussion
In the mammographic period about 98% of DIN have shown the presence of microcalcifications (11-16). Furthermore, microcalcifications were associated with malignancy in up to 35% of the cases (15). In the present series clinical presentations included microcalcifications, but also in a minority of cases architectural distorsion, ductal asymmetry, circumscribed nodules and ill defined masses. The evaluation of microcalcifications strongly depends on the radiologist's expertise, the availability of digital imaging and the comparison with previous examinations. Eighty seven % of the patients who were referred to our breast surgery service were recruited after mammographic screening in our institution. External patients were often required to have a Rx-mammography re-assesment or to undergo a full field digital mammography. Taken together these factors may explain how in the present study the mean maximum diameter was about 10mm which was comparable to a recent single institution series in Italy (17), but considerably smaller if compared to other series. The screening policy provided by the national health service could also have contributed to these results. The role of other imaging modalities including breast MRI and breast US in diagnosing and staging the extent of DCIS within the breast has yet to be established (18, 19). MRI is increasingly used to evaluate the breast and may identify lesions that are not seen mammographically, but its use in DCIS must be assessed in order to avoid overestimation. At present it cannot be considered as a mass screening tool for early lesion of the breast and it should always be followed by an US second look. In our experience FNAC has little place in the diagnosis of pure DCIS, because this diagnosis implies accurate exclusion of stromal invasion, which can be accomplished only by microscopic examination of a histologically intact specimen to analyze the overall histological architecture of the diseased tissue. False negatives can be difficult to distinguish from truly benign lesions. In this setting the patient may go on to a surgical biopsy to confirm the benign nature of the lesion. In fact, the routine use of 11C VACB, FNAC offers a reliable tool for the detection of malignant lesions but it does not allow a correct preoperative diagnosis in terms of grade, invasion and histological subtype. VCAB using specifically designed devices under stereotactic or US guidance allows a better, even if more expensive, preoperative diagnosis with a low rate of underestimates. Preoperative assessment of the lesion allowed us to schedule open biopsy and SLNB in a single day surgery with a considerable return in terms of the effectiveness and overall cost of the treatment. It might be questioned if the finding of AIDH at VACB still requires further characterization by open biopsy. In the present series open biopsy was performed on 38 cases of AIDH diagnosed after VACB revealing underestimation in only three patients (7.9%) who had DIN1C at final histology. Similarly Liberman et al. (20) and Bedei et al. (21) reported an underdiagnosis rate of 10% and 11.7% respectively. On the other hand a huge variability in the rate of underdiagnosis exists between different authors, a series analysed by Winchester DJ (22) showed that only 83% had an accurate diagnosis when AIDH was present while Joshi (23) reported 0% of AIDH underestimates. It is probable that, in the case of AIDH histology after VACB, an open excisional biopsy is still needed only in selected cases (i.e. highly suspicious microcalcifications, the absence of microcalcifications at specimen radiography, lesions larger than 20 mm, a previous history of in situ or invasive breast cancer). A six month follow-up of the lesions should be suggested to patients with AIDH at VACB by conventional imaging or if necessary a second VACB. The underestimation of in situ lesions at VACB occurred in 6 DIN3 cases (2 microinvasive and 4 infiltrating carcinomas). As all in situ lesions should undergo excisional biopsy, underestimation would not affect the management of these patients. Underestimation seems more likely to occur for high grade lesions and when the peripheral edge is affected. It should be noted that the use of VACB drastically reduced the number of open biopsies for AIDH compared to the period before the routine use of VACB (p<0.0001). In situ ablation of the cancer with 11G VACB does not ensure complete removal of a malignancy and should be completed by open biopsy (20). Surgical treatment has drastically changed in the mammographic era to favour BCS (1). Furthermore, clinical presentation such as, large palpable masses, bloody nipple discharge or Paget's disease of the nipple are rare (24). Consistently BCS has been the first choice for 97.5% of the patients with DIN1C-DIN3; and for about 80% it proved to be the definitive treatment. BCS is indicated for localized DCIS of less than 3 cm in diameter, without evidence of gross multicentricity or diffuse malignant-appearing calcifications, with acceptable cosmetic results (25-27). Mastectomy is still indicated in cases of large tumors (>3 cm), small breast, multiple tumors or with diffuse malignant-appearing microcalcification, persistent positive margins after two or three excisions, a history of previous therapeutic irradiation of the breast region or other conditions that contraindicate breast irradiation (28-31). Furthermore in two young patients (43 and 46 years old) with monolateral multifocal DIN3 bilateral nipple sparing mastectomy with immediate prosthetic reconstruction was performed in consideration of the life expectation and the long term cosmetic results. The importance of margin assessment of in situ lesions of the breast has been widely discussed. Complete tumor excision, confirmed by specimen radiography, evaluation of resection margins and post-excision mammogram are important determinants of local control. Involved margins have been identified as the most important independent prognostic variable for predicting local relapse (3). A small percentage of patients (6%) in the present study had margins which were less than 5mm (Figure 2). The definition of “positive” and “negative” margins varies between institutions (32). Decisions regarding re-excision or mastectomy depend on the volume of tissue resection required to achieve margin clearance and the anticipated cosmetic outcome. Such decisions require thorough counselling with the patient (32). The higher rate of involved margins found during second re-excision in the presents study showed the difficulty of achieving clear margin when multifocality was present (p<0.0001, OR=1.085, 95% CI 1.038-1.134). The present relapse rate (10.1 %) at a mean follow-up of 76 months (range 4-121 months) indicated that 5 mm represented a safe clear margin for DIN and suggested that a shorter margin such as 2 mm could be considered acceptable as supported by other papers (2-33). Trials assessing adjuvant radiotherapy after excision with involved margins have shown ‘acceptable local control’, a finding that questions the importance of obtaining clear margins if adjuvant radiotherapy is already planned (34).
(Patient No.5) A 43 year old with a no palpable mass in the left breast and group 2 DCIS of the Van Nuys classification. The mammography shows focal microcalcifications in the lower mid portion of left breast.
Immunostained lymph node positive for cytokeratins.
The status of the axillary lymph nodes remains the most powerful prognostic indicator of invasive breast cancer (6), but the role of axillary staging with SLNB for DIN is controversial (35). By definition DIN does not have the potential to spread to regional lymph nodes unless there is an invasive component. Nonetheless, surgeons have been removing lymph nodes in patients with a primary diagnosis of DIN for a variety of reasons and with variable results. Furthermore, the SLNB is generally a well tolerated and the associated morbidity with this minimally invasive procedure is significantly less than that of an axillary dissection (8). In several studies the complications associated with the SLNB were all transient (8). Positive sentinel nodes have been reported in 0% to 13% (35) of DIN patients. In the present series positive nodes were not identified with H&E and positive sentinel nodes detected by IHC were uncommon (3.7%). The information from axillary dissection or SLNB was of little value in the treatment of the patients with DIN in whom no invasive cancer was found. IHC was revealed to be a more sensitive tool for the identification of positive nodes, nevertheless, it should not be used to upstage patients with DCIS (35). We suggest that when DIN, and no invasive component, is present on final pathology, no form of nodal staging is indicated. The only reason to perform SLNB in patients with DIN diagnosed by core needle biopsy examination is the small number of cases (3.5%) in which complete excision would show an invasive cancer. Whenever the rate of upstaging should be higher (36), and if an invasive cancer is found the patient can safely undergo SLNB a later time (37, 38). Therefore in patients diagnosed with DIN on core biopsy examination SLNB should be reserved for those at high risk of invasive disease including patients with palpable lesions, DIN larger than 40 mm, or high nuclear grade or patients undergoing mastectomy where SLNB could not be postponed (39). Post-surgical treatment generally included radiotherapy when conservative therapy was performed plus endocrine therapy in hormono-responsive tumors (Table VI). The recommendations place a high value on postoperative radiotherapy, stating that radiotherapy following BCS reduces the risk of subsequent invasive and in situ recurrences for all women with DCIS, regardless of the grade or pathological subgroup (40). Subsets of patients with a high risk of recurrence (e.g. ‘women with a high-grade DCIS with necrosis, close margins and larger lesions’) particularly benefit from adjuvant radiotherapy (40). Indeed 8 out of 17 patients in the present study who experienced ipsilateral breast tumor recurrence (IBTR) had refused any adjuvant therapy, confirming even in this small non randomized setting the significant benefit of adjuvant treatment in DIN lesions (Chisquare= 4.936; p=0.026; d.f.=1).
Rate of ipsilateral breast tumor recurrence (IBTR) after different adjuvant therapies at mean follow up of 76 months (Range 4-121 months).
Conclusion
DIN of the breast is an increasingly common finding requiring an appropriate surgical management. BCS is mandatory in most DIN cases and should follow an accurate preoperative assessment including imaging and VACB histology. SLNB, with few exceptions, is not recommended when a pure DIN lesion is suspected and then confirmed at final histology with few exceptions. Nevertheless, the development of sentinel lymph node biopsy (SLNB) has signaled a paradigm shift in the surgery of axilla for breast cancer management (6). Although the incidence of metastases in patients with an initial diagnosis of DCIS is considered to be low, SLNB remains an attractive option when considering DCIS (6). In patients at high risk, occult invasion does occur and without lymphatic analysis, undertreatment is possible. DCIS represents a spectrum of disease that may progress to invasive carcinoma so a consideration of such pathological features as invasion predictors to optimize surgical management (3) and the practice of diagnostic procedures such as SLNB are suggested. This is a minimally invasive procedure that gives important information for treatments possibilities on the bases of disease staging. Further, as shown in this study, the wide range in reported positive SLNs may also depend on the different techniques utilized to stain axillary nodes and the subsequent kind of nodal metastasis detected (8). Adjuvant RT and endocrine treatment should be considered helpful in preventing long term IBTR.
- Received July 31, 2008.
- Revision received December 19, 2008.
- Accepted February 13, 2009.
- Copyright© 2009 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.