SOFT Case 4 (Jan 1st, 2005)

A 58-year old female underwent a routine mammography screening. Mammography showed three separate clusters of "pleomorphic calcification". The lesion was excised with three guide wires. 9 images of the lesion presented.

Presented by Dr Sanjay Navani
Associate Histopathologist and Immunohistochemist,
Breach Candy Hospital, Mumbai.
snavani@vsnl.com

DISCUSSION OF SOFT CASE 4

Total Number of Pathologists Replies Received 14

Most pathologists entertained a differential diagnosis that included part or all of the following:
Atypical Ductal Hyperplasia, Intraductal Carcinoma (DCIS), Intraductal Carcinoma (DCIS) with Microinvasion.

Four pathologists thought it was a frankly invasive carcinoma ranging from invasive cribriform to adenoid cystic carcinoma. Two pathologists were bothered by what appeared to be "streaming" of nuclei in picture 6 Two pathologists felt immunohistochemistry with either E-cadherin or a double-immunostaining with E-cadherin and HMW Cytokeratin should be carried out to assess invasion. One pathologist thought it was an Adenocarcinoma of alveolar epithelium. One pathologist thought the lesion was completely benign - Collagenous Spherulosis with Blunt Duct Adenosis.

The advantage that I had before looking at the gross specimen for frozen section on this case was a mammography report saying there were three areas of "pleomorphic calcification". The areas were marked with guidewires so I also knew where to look for the lesion. After locating the three lesions I made the mistake of letting my focus slip and missed another grossly abnormal area that was also missed on the mammo. That area was subsequently sampled on paraffin and showed features similar to that viewed by all of you. Moral of the Gross examination: Even radiologists can't see everything. Pathologists can, but not if they are lax.

Pictures 1,2,3 showed marked distension of acini that still retain their lobular architecture. When one looks around the group of distended acini it still is possible to say that it looks like one lobular unit only much bigger. The same is also true of adjacent lobular units- this creates a problem because then one becomes unsure whether it is another lobule with similar in situ changes or whether it all actually respresents invasive tumor. A closer look (high power - Images 8 & 9) show peripherally compressed spindled cells that appear to be myoepithelial (Congratulations to Sachin Kale for this astute observation).

The "cribriform" spaces are definitely "cribriform". The trick is to concentrate on the orientation of the cells surrounding the space rather than the space itself. If one sees an orientation of cells towards the central space forming a well appreciated lumen - it favors a carcinoma. Additionally the nuclei go towards the base of the cell giving rise to apical cytoplasm bordering the space (Picture 8). The nuclei of course are at least mildly atypical in most structures further supporting the diagnosis of Intraductal carcinoma as does the comedonecrosis.

The problems appear in Picture 6. Are those islands invasive or in situ? Further Pictures 5 & 7 show a Sclerosing adenosis that in Picture 7 is partially involved by the lesion. This gives rise to odd shaped nests in a desmoplastic, reactive stroma and hence the question - Is that invasion?

Here the observation of compressed spindled cells at the base (as Sachin detected) is invaluable because they prompt the question - Can I confirm the myoepithelial cells with immunohistochemistry? Yes, we can - smooth muscle actin and calponin should do the trick. Both these were carried out and showed well preserved myoepithelial cells. (Good paper on myoepithelial cell detection in AJSP August 2004 by Melinda Lerwill - it's a MUST READ). (IHC pictures have now been put up)

Invasive cribriform islands are recognised by their angulated and jagged contours - the cribriform areas in the current case all have well rounded outlines. Myopethelial cells, of course will never be present in conventional invasive cancer but may be seen at the base in Intraductal carcinoma (DCIS). Collagenous spherulosis must have collagenous rounded nodules surrounded by cells that show no orientation to the collagen - these were absent.

Yes, it does appear there is nuclear streaming in picture 6 - I thought it was artefactual and seen only in that section. However the other pictures show clear Intraductal Carcinoma.

Invasive Carcinoma with an Intraductal pattern has been reported (Breast Pathology, Rosen). Apparently, not everything appearing in situ is insitu. I appreciate all those who considered this lesion to be extensively invasive versus Intraductal. It is a distinct differential to be ruled out.

My diagnosis on this case was Extensive Intraductal Carcinoma, Intermediate Grade with Comedonecrosis in Sclerosing Adenosis (Cribriform, Solid, Micropapillary Patterns). She still underwent node dissection (the fallout of second opinion). All were negative for metastasis. After 26 months of follow up she is free of disease.

Thank you all for an opportunity to present this case.

Sincerely
Dr Sanjay Navani
AssociateSurgical Pathologist and Immunohistochemist
Breach Candy Hospital Trust
Mumbai.