SNP-LOH Analysis of Lobular Carcinoma In Situ and Associated Invasive Lobular Cancer.

Abstract

Abstract Background: Lobular carcinoma in situ (LCIS) is a risk factor for the development of subsequent invasive breast carcinoma in either breast. Approximately 50-70% of these subsequent cancers are invasive lobular carcinoma (ILC). Recent molecular evidence suggests that LCIS may be a precursor of ILC. However, not all LCIS progresses to invasive disease and at present there are no biomarkers that predict which cases are going to develop invasive disease. SNP-LOH analysis may be a useful method to analyse genome wide loss of heterozygosity (LOH) and copy number variation in LCIS and associated invasive disease in order to identify such biomarkers. The study of LCIS poses specific technical challenges as available samples are almost exclusively formalin fixed paraffin embedded (FFPE) tissue and LCIS tends to occur in small scattered foci which require micro-dissection.Aims: To assess the feasibility of performing SNP-LOH on micro-dissected FFPE tissue To compare patterns of LOH in pure LCIS and LCIS associated with ILCTo examine the genetic relationship between LCIS and associated ILCMethods: Tumour blocks were collected as part of the GLACIER Study (a study to investigate the Genetics of LobulAr Carcinoma In situ in EuRope). Tissue was micro-dissected from FFPE specimens. DNA was extracted with the DNEasy kit (Qiagen) if it was possible to micro-dissect by hand under a light microscope. For cases extracted using the laser capture microscope (LCM) the Picopure Kit (MDS Analytical Tech)) was used to extract DNA. SNP-LOH was performed using the GoldenGate Assay(Illumina) on 28 samples (20 LCIS, 8 ILC). A minimum of 250ng of DNA was required. For samples with low DNA yields, DNA was amplified using the Genomeplex Amplification Kit(Sigma).Results: Results were obtained on 24 samples that were micro-dissected under the light microscope and extracted using the DNEasy kit(Qiagen). Amplification of DNA was also possible from these samples and the SNP-LOH results for amplified DNA correlated with those for unamplified DNA(1 case).SNP-LOH was performed on 13 cases of classical LCIS: 6 pure LCIS, 7 with associated ILC. LOH events were more common in LCIS associated with invasive disease (range 2-5, median 3) than pure LCIS (range 0-5, median 2). The commonest change in both groups was 16q LOH (6/6 LCIS with invasive disease, 4/7 pure LCIS, p=0.12, Fishers Exact Test). 1q LOH was also present in all samples associated with invasive disease, but only identified in 3/7 samples of pure LCIS (p=0.049, Fishers Exact Test). There were also other regions of LOH that were more common in LCIS associated with ILC, but did not reach significance.Comparing LOH patterns in LCIS and associated ILC revealed that in 3/6 cases of ILC genetic changes were identical to those in the LCIS. In 2/6 cases the ILC had the same genetic changes but had also acquired new genetic changes. Interestingly in 1 case the LCIS had more genetic changes than the ILC.Conclusions: This pilot study has shown that it is possible to perform SNP-LOH on small amounts of micro-dissected FFPE tissue. The results confirm the findings of others that LCIS is a likely precursor lesion of ILC and suggest that even with this limited number of samples pure LCIS and LCIS associated with ILC may have different genetic profiles. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 5169.