Breast Imaging
Calcifications of Lobular Carcinoma In Situ of the Breast
Radiologic—Pathologic Correlation
Audio Available | 
Share
OBJECTIVE. Because lobular carcinoma in situ is thought to be an incidental finding at breast pathology, the finding of lobular carcinoma in situ at core biopsy poses a diagnostic dilemma of radiologic-pathologic concordance. The purpose of this article is to describe the radiologic-pathologic correlation of calcifications associated with lobular carcinoma in situ of the breast.
MATERIALS AND METHODS. Between July 1999 and July 2000, seven excisional biopsies resulted in findings of lobular carcinoma in situ of mammographic calcifications. The radiographic features of the calcifications were characterized by the Breast Imaging Reporting and Data System lexicon, and pathologic features were reviewed.
RESULTS. Two forms of lobular carcinoma in situ were associated with calcifications: the classic form with small, uniform cells, and the pleomorphic form with larger cells frequently associated with central necrosis. On mammography, all calcifications were clustered, punctate, high density, and smaller than or equal to 0.5 mm, although mammographically visible calcifications found in the pleomorphic type tended to be larger and more dense. Additionally, infiltrating lobular carcinoma was found after surgical excision in two (40%) of five patients with pleomorphic lobular carcinoma in situ.
CONCLUSION. Calcifications can be associated with lobular carcinoma in situ and therefore concordant at stereotactic core biopsy. The classic form may be incidental and clinically innocuous. The pleomorphic form is morphologically similar to ductal carcinoma in situ and may have a greater tendency for invasion.
| Introduction |   |
|---|
The clinical significance of lobular carcinoma in situ has been debated since it was first described in 1941 [1]. It is now considered a high-risk marker for the development of infiltrating ductal carcinoma and not a premalignant lesion [2]. Moreover, lobular carcinoma in situ is thought to be an incidental finding at breast pathology [2], and therefore without corollary signs of mass or calcification on mammography. Hence, the finding of lobular carcinoma in situ at stereotactic biopsy poses a diagnostic dilemma. Are pathology results of lobular carcinoma in situ concordant with the radiographic findings of calcifications when the targeted calcifications are removed at core biopsy? Calcifications associated with lobular carcinoma in situ have been noted in the pathology literature [3] but have been only rarely emphasized in radiology [4]. This article describes the radiologic-pathologic correlation of calcifications associated with lobular carcinoma in situ.
| Materials and Methods | |
|---|
Between July 1999 and July 2000, six women (age range, 35-76 years; mean age, 51 years) underwent biopsies in which lobular carcinoma in situ was associated with mammographic calcifications. Of 174 stereotactic biopsies performed in a 1-year period, the biopsies in three patients revealed lobular carcinoma in situ. Three cases were determined after needle localization and excisional biopsies, although the biopsy specimens for two patients were from outside institutions and had been referred for second opinions. There were 128 needle localizations during the study. The department of pathology reported 18 patients with pure lobular carcinoma in situ by needle localization or surgical biopsy, which included these six patients. Additionally, one patient had a biopsy 2 years earlier of a different sample from the same breast quadrant for identical mammographic findings. A total of seven excisional biopsy specimens were reviewed for their mammographic and pathologic features.
The radiographic features of the calcifications were characterized using the Breast Imaging Reporting and Data System (BI-RADS) lexicon by a board-certified radiologist specializing in breast imaging [5]. The calcifications were measured on the nonmagnified mammographic views. The pathologic features were reviewed by a board-certified pathologist with specialization in breast pathology.
| Results | |
|---|
All calcifications were in the lobules that contained lobular carcinoma in situ. Two forms of distinctly sized calcifications were noted. In two patients, a small calcification was identical in morphology to calcifications also present in surrounding benign tissue (Fig. 1A). Pathogenesis of the calcifications was unknown. This form of lobular carcinoma in situ is the classic form of lobular carcinoma in situ with small, uniform cells [6]. In contrast, large calcifications that were formed in central necrosis (Fig. 2A) were noted in five of the biopsy specimens; the cells in this type of lobular carcinoma in situ were larger and more pleomorphic.
 View larger version (182K) | Fig. 1A. —Classic lobular carcinoma in situ in 43-year-old woman. Photomicrograph of biopsy specimen shows calcifications (double arrows) in lobular carcinoma in situ and similar-appearing calcifications (single arrow) in focus of adenosis. (H and E, ×100) |
 View larger version (195K) | Fig. 2A. —Pleomorphic lobular carcinoma in situ in 34-year-old woman. Photomicrograph of biopsy specimen shows central necrosis and calcification (arrow) in lobule with lobular carcinoma in situ (arrowheads). (H and E, ×400) |
These calcifications were sometimes fragmented within the lobule. The case featured in Figure 2A,2B,2C had the largest amount of necrotic debris in the lobule: the lobule was distended and distorted to form the shape of the calcifications noted in the mammogram (Fig. 2B), which were not as round as those seen in Figures 1A,1B and 3. No necrosis was present in ducts.
 View larger version (149K) | Fig. 2B. —Pleomorphic lobular carcinoma in situ in 34-year-old woman. Photomicrograph of biopsy specimen shows membranous staining of cells (solid arrows) in benign lobules with complete absence of staining of cells in lobular carcinoma in situ (open arrows). (E-cadherin immunostain, ×100) |
 View larger version (161K) | Fig. 2C. —Pleomorphic lobular carcinoma in situ in 34-year-old woman. Mammogram shows cluster of calcifications with three large (1 mm) pleomorphic calcifications (arrow) surrounded by punctate calcifications. Although no calcifications were branching in morphology, calcifications were thought to represent ductalcomedocarcinoma in situ. E-cadherin staining in (B) was significant in distinguishing lobular carcinoma in situ from ductal carcinoma in situ. (Spot magnification, ×1.7) |
 View larger version (156K) | Fig. 1B. —Classic lobular carcinoma in situ in 43-year-old woman. Mammogram shows cluster of punctate calcifications in lobular carcinoma in situ (arrow). Benign calcifications were in adjacent tissues (arrowheads). (Spot magnification, ×1.7) |
 View larger version (163K) | Fig. 3. —Mammogram of 58-year-old woman shows cluster of punctate calcifications (arrows) smaller than or equal to 0.5 mm. Appearance of calcifications was similar to that seen in all patients. Calcifications were in pleomorphic lobular carcinoma in situ. |
Biopsy specimens in two of the four patients who underwent stereotactic core biopsy yielded pleomorphic lobular carcinoma in situ. Both patients subsequently underwent surgical excision. The findings of an additional three patients with pleomorphic lobular carcinoma in situ were diagnosed at surgical excision. In all five patients, the calcifications were clustered in distribution; in two patients with pleomorphic lobular carcinoma in situ, there was more than one cluster in a pattern, suggestive of a ductal distribution. All calcifications were high in density and punctate or pleomorphic, although they tended to be round. No mammographic features distinguished the classic lobular carcinoma in situ from the pleomorphic lobular carcinoma in situ calcifications other than that the latter tended to be larger. Almost all calcifications were equal to or smaller than 0.5 mm but they varied in size (Fig. 3). One case of pleomorphic lobular carcinoma in situ had additional calcifications as large as 1 mm and oval to borderline linear (Fig. 2B), with slightly angular margins. None of the calcifications was branching, amorphous, or indistinct [5].
In two patients, the clustered punctate calcifications were surrounded by scattered punctate and amorphous calcifications (Fig. 1B), which were in classic lobular carcinoma in situ. The clustered calcifications had been considered suspicious for malignancy because they were focal in distribution and new. The surrounding calcifications were interpreted as benign because of mammographic stability.
No masses or architectural distortion were associated with the lobular carcinoma in situ calcifications.
| Discussion | |
|---|
The issue of mammographic—pathologic concordance is the pivotal point in the treatment of patients who undergo core biopsy, because women who are thought to have discordant results are recommended for a second core biopsy or surgical excision. Calcifications can be a finding of lobular carcinoma in situ and not just a serendipitous finding at pathology as is currently considered [3]. We acknowledge that differences in our study design and selection criteria for biopsy or histologic technique in contrast to previous studies may be contributing to this impression. The most significant finding in our study was that were two forms of calcifications found at pathologic analysis (Figs. 1A,1B,2A,2B,2C,3): one had a common appearance to surrounding calcifications in benign proliferative change, and the other was reminiscent of comedocarcinoma calcifications of ductal carcinoma in situ.
The necrotic debris in the lobules of pleomorphic lobular carcinoma in situ was fragmented, which accounted for the slightly angular margins in the calcifications that tended to be round. It is the fragmented necrotic debris in ductalcome-docarcinoma in situ that distinguishes this sign of malignancy from the classic thick, rod-shaped calcification of secretory disease [7]. Analogously, similar fragments in lobular carcinoma in situ necrosis distinguishes these calcifications from the benign round acinar calcifications of the terminal duct lobular unit [7]. The necrosis may explain why the calcifications of pleomorphic lobular carcinoma in situ were radiographically dense, again similar to that of ductal comedocarcinoma in situ. We are uncertain of the pathophysiology of the classic lobular carcinoma in situ and cannot explain why these were also radiographically dense. Because ductal carcinoma in situ and lobular carcinoma in situ share the pathophysiology of necrosis in some mammographically similar features, radiologists may not be able to distinguish between the two types prospectively.
Just as the radiographic features of necrotic lobular carcinoma in situ and ductal carcinoma in situ may be similar, the pathologic distinction of these calcifications may be difficult. This differentiation appears simple: lobular carcinoma in situ shows loss of cohesion, intracytoplasmic vacuoles, and pagetoid ductal involvement, three elements that are absent in ductal carcinoma in situ [3]. In contrast, microacini, absent in lobular carcinoma in situ, are present in ductal carcinoma in situ. In reality, the distinction is not always clear. Regarding this controversy, Rosen [2] has noted that there is no common agreement regarding the terminology and suggests “examples of ductal and lobular carcinoma in situ.” Tavassoli [8] has stated that “necrosis does `rarely occur in lobular neoplasia' and that does not `necessarily imply ductal growth pattern.”' We agree with this statement—the calcifications in our patients were in lobular carcinoma in situ and not ductal carcinoma in situ.
A recent development in pathology that has made the differentiation between lobular carcinoma in situ and ductal carcinoma in situ more precise is that of the stain for e-cadherin [9]. E-cadherin is a transmembrane glycoprotein responsible for calcium dependent cell—cell adhesion [10]. This protein is lost in lobular but not in ductal carcinomas (Fig. 2B). Jacobs et al. [11] had no false-negative findings in their 17 cases of lobular carcinoma in situ and in nine cases of indeterminate in-situ carcinoma. These two groups would be congruent with our group of patients with necrotic lobular carcinoma in situ as noted by their architectural definitions. Similarly, Acs et al. [12] showed that 26 of 27 lobular carcinomas showed complete loss of membranous staining, and that all cases of ductal carcinoma showed e-cadherin reactivity. Therefore, the e-cadherin stain shows significant specificity in distinguishing lobular from ductal malignancies.
Because this necrotic form of pleomorphic lobular carcinoma in situ has been infrequently noted, the clinical prognosis is unknown. In our series, this finding was found in most patients (5/7, 71%). This high frequency is likely a result of the increased use and specificity of e-cadherin in distinguishing ductal from lobular neoplasms. Cases that were previously considered to be ductal carcinoma in situ because of the necrosis are now being determined to be lobular carcinoma in situ (pleomorphic type). The question remains whether the physiology that leads to the necrotic deposition of calcification in the lobule is clinically similar to that of ductal carcinoma in situ and, therefore, more likely associated with the development of invasive carcinoma than the high-risk marker of classic lobular carcinoma in situ. This clinical significance is critical to determine, particularly for patients undergoing stereotactic core biopsy, because correct estimation of malignancy is key when interpreting stereotactic results. Therefore, in the setting of necrotic calcifications with lobular carcinoma in situ, one may need to suggest further surgical excision because this form of lobular carcinoma in situ may be an underestimation of invasive malignancy, analogous to ductal carcinoma in situ. In this series, there were five such cases. In two (40%), there was invasive disease at excision, both lobular within the area of lobular carcinoma in situ. One patient underwent preoperative MR imaging, and the extent of pleomorphic lobular carcinoma in situ was determined to be greater in extent on MR imaging than that expected by the extent of calcifications shown on mammography. This underestimation of extent of disease is similar to the diagnostic dilemma that plagues ductal carcinoma in situ [13].
Lobular carcinoma in situ is most commonly considered to be a risk marker for the development of invasive ductal carcinoma bilaterally. Therefore, lobular carcinoma in situ on surgical specimens usually leads to surveillance only, rather than further surgery or other treatment. However, some invasive lobular carcinoma is thought to arise from lobular carcinoma in situ on the basis of anecdotal data [14]. As a result, pathologists are becoming more accepting of the idea that infiltrating lobular carcinoma may arise from lobular carcinoma in situ [15]. Our findings support this opinion, as noted in Figure 4. However, epidemiologic data is lacking. Moreover, if this were true, a more elusive question analogous to that of ductal carcinoma in situ is, what are the factors that might predict this progression?
 View larger version (191K) | Fig. 4. —Photomicrograph of biopsy specimen shows pleomorphic lobular carcinoma in situ with focus of infiltrating lobular carcinoma (double arrows) adjacent to lobular carcinoma in situ (single arrows). Cytologic features of cells in infiltrating lobular carcinoma were identical to cells in lobular carcinoma in situ. (H and E, ×400) |
The classic form of nonnecrotic lobular carcinoma in situ calcifications (Fig. 1A,1B) may be more innocuous clinically because this form of calcification is also found in surrounding benign tissue. When these calcifications are removed by core biopsy and correlate with lobular carcinoma in situ at pathology, they are also concordant findings. These may represent the cases of lobular carcinoma in situ that have been considered incidental at surgical pathology and, hence, not clinically worrisome other than as a risk marker. Patients with these findings may not need surgical follow-up after stereotaxis; however, our two patients, with only 6-month follow-up, do not provide proof of this hypothesis.
Liberman et al. [4] suggest that surgical excision is warranted for lobular carcinoma in situ if there is discordance with mammographic findings, coexistent atypia, or radial scar, or if the pathologist cannot distinguish between lobular carcinoma in situ and ductal carcinoma in situ in the core biopsy specimens. This latter scenario may reflect cases of pleomorphic lobular carcinoma in situ that can now be distinguished using e-cadherin stains [9]. Reynolds [16] similarly recommends surgery, because after an exhaustive review of the literature, there were two cases of nearby ductal disease at surgical followup of four cases of lobular carcinoma in situ at core biopsy. He notes that incidental lobular carcinoma in situ coexistent with ductal malignancy at surgical pathology is not rare. Whether these cases were the pleomorphic lobular carcinoma in situ with necrotic calcifications or the classic nonnecrotic form is not known.
In summary, these findings show that calcifications can be associated with lobular carcinoma in situ but that there are two forms: the pleomorphic necrotic and the classic nonnecrotic. Our findings show that the necrotic calcifications of lobular carcinoma in situ have particular imaging findings that are similar to ductalcomedocarcinoma in situ, which can be pathologically distinguished using e-cadherin staining. Because it shares similar morphology with ductalcomedocarcinoma in situ, this form of lobular carcinoma in situ may be clinically similar to ductal carcinoma in situ and not just a marker of high risk.
Therefore, we advocate further surgical excision after core biopsy for the pleomorphic type of lobular carcinoma in situ in particular. Confirmation of our observations by other investigators is needed before this recommendation is universally adopted by other radiologists. Classic lobular carcinoma in situ with nonnecrotic calcifications may not warrant further tissue excision if similar-appearing calcifications are noted in adjacent benign tissue without other atypical or malignant findings. Long-term follow-up with the recognition of these distinct types of calcifications is necessary to understand their clinical significance.
Address correspondence to D. Georgian-Smith.
| References | |
|---|
