Intracystic Papillary Carcinoma of Breast: Interrelationship With in Situ and Invasive Carcinoma and a Proposal of Pathogenesis

Array Comparative Genomic Hybridization Study of 14 Cases

Thaer Khoury; Qiang Hu; Song Liu; Jianmin Wang


Mod Pathol. 2014;27(2):194-203. 

In This Article

Materials and Methods


Clinical and pathologic databases were searched for 'intracystic', 'encapsulated' or 'papillary' carcinoma of the breast at Roswell Park Cancer Institute between 1992 and 2010. Only cases that had available slides and tissue blocks are included in the study.

The original slides of these cases were reviewed. The pathologic features including tumor size (invasive ductal carcinoma and intracystic papillary carcinoma), lymph node status, nuclear grade and margin status were recorded. Myoepithelial markers including p63 (clone 4A4 (Dako), with 1:50 dilution, and EDTA antigen retrieval), smooth muscle myosin (clone SMMS-1 (Cell Marque), prediluted and EDTA antigen retrieval) and Calponin (clone CALP (Dako), with dilution 1:200, and Vector antigen retrieval) were performed on all cases. Complete or partial loss of myoepithelial cell layer was required for the diagnosis of intracystic papillary carcinoma.[1] We adopted the WHO definition of invasion in intracystic papillary carcinoma as a tumor that assumes a pattern of invasive ductal carcinoma and lacks the papillary architecture.[1] Invasive ductal carcinoma was graded using modified Bloom Richardson grading system.[10] For a lesion to qualify to be ductal carcinoma in situ, the involved ducts had to be completely surrounded by the myoepithelial cell layer with non-papillary growth pattern. Ductal carcinoma in situ was graded based on European Pathologists Working Group classification system.[11] All cases were stained with estrogen receptor (clone 1D5 (Dako), with 1:100 dilution, and TRS/Vector antigen retrieval) and progesterone receptor (PgR636 (Dako), with dilution 1:100, and TRS/Steamer antigen retrieval).

Laser Capture Microdissection

Paraffin-embedded breast tissue was cut at 8 μm and mounted on polyethylene naphthalate-membrane-coated slides (Leica, Wetzlar, Germany; cat. no. 11505158) and dried overnight at room temperature. The next day slides were deparaffinized in three changes of xylene, rehydrated using graded alcohols and stained manually with hematoxylin and eosin. All the solutions were prepared with double-distilled water and changed for the each sample. The laser capture microdissection was carried out using Leica Laser Microdissection systems. DNA was extracted from the laser capture microdissection tissues by QIAamp DNA Micro Kit (Qiagen; cat. no. 56304) and eluted in 20 μl nuclease-free water, and the concentration of DNA was measured by NanoDrop, Spectrophotometer, ND-1000.

Array Comparative Genomic Hybridization Analysis

One microgram of laser capture microdissection captured formalin-fixed, paraffin-embedded tumor genomic DNA was Cy5 labeled (test) and 1 μg of pooled control DNA from 20 healthy male donors was Cy3 labeled (reference) using the Bioarray Kit (Enzo Life Sciences). Following purification, equal molar aliquots of the test and reference probes were combined, denatured and hybridized to an RPCI 21K BAC array (Roswell Park Cancer Institute).[12] After a 20-h hybridization, the arrayed slides were washed and then scanned on a GenePix 4200AL Scanner (Molecular Devices) to generate high-resolution (5 μm) images for both Cy5 (test) and Cy3 (reference) channels. Image analysis was performed using ImaGene (version 8.0.0) software from BioDiscovery. The log 2 test/reference ratios were normalized using a subgrid loess correction. Mapping information was added to the resulting log 2 test/reference values. The mapping data for each BAC was found by querying the human genome sequence at

Statistical Analysis

A loess-corrected log 2 ratio of the background subtracted test/control was calculated for each clone. The data are further normalized using median-based normalization and segmented using DNAcopy.[13] Instead of using a fixed cutoff value, copy number gain/loss was called by CGHcall with 75% assumed tumor purity.[14] The test of difference in copy number changes among intracystic papillary carcinoma, ductal carcinoma in situ and invasive ductal carcinoma was carried out using CGHMultiArray.[15] We performed two analyses for the genomic variation in intracystic papillary carcinoma cases, one based on the presence or absence of concurrent ductal carcinoma in situ (intracystic papillary carcinoma with ductal carcinoma in situ (n=6) vs intracystic papillary carcinoma without ductal carcinoma in situ (n=8)) and one based on the presence or absence of concurrent invasive ductal carcinoma (intracystic papillary carcinoma with invasive ductal carcinoma (n=6) vs intracystic papillary carcinoma without invasive ductal carcinoma (n=8)). Only three cases had concurrent ductal carcinoma in situ and invasive ductal carcinoma, precluding statistical analysis. Genomic variation analysis in intracystic papillary carcinoma group with relation to patient age, tumor grade and tumor recurrence was also performed. Finally, we compared the genomic variation between intracystic papillary carcinoma and concurrent ductal carcinoma in situ and invasive ductal carcinoma. An Online Mendelian Inheritance in Man (OMIM) search was performed on all the involved genes in the genomic variation.