B68
Using formalin fixed paraffin embedded (FFPE) core biopsy samples for molecular marker studies of breast cancer: Feasibility study of copy number and gene expression analysis
Mahesh Iddawela1, Oscar Rueda1, Yuker Wang3, Malek Faham3, Mohamed El-Sheemy2, Ged Cowley2, Jennifer Eremin2, Oleg Eremin2, Helena Earl4, Carlos Caldas1
1CRUK Cambridge Research Institute, Cambridge, UK, 2United Linconshire Hospitals NHS Trust, Lincoln, UK, 3Affymetrix Inc., San Francisco, USA, 4University of Cambridge, UK
Background
FFPE tissues are a valuable source of tumour material for translational research that has thus far been under-used as only a limited amount of partially degraded DNA or RNA can be extracted from these samples. But progress in technology has allowed the analysis of genome wide copy number and expression changes using small amounts of degraded DNA/RNA. We used two platforms: molecular inversion probes assay (MIP, Affymetrix) and c-DNA mediated Annealing Selection extension and Ligation assay (DASL, Illumina) for analysis of RNA/DNA extracted from routine core biopsies.
Method
FFPE diagnostic core biopsies of 63 patients treated with neoadjuvant adriamycin/cyclophosphamide followed by taxotere in a clinical study were used. MIP assay with 50K SNP panel was used for copy number analysis (CNA). DASL assay assessed expression of 502 genes using three probes per gene.
Results
RNA and DNA of good quality was extracted from 52/63 (82%) and 46/63 (73%) core biopsy samples, respectively. CNA showed well described aberrations such as high probability of gains in 1q, 8q, 11q and 20q and loss in 4q, 12q, 13q, 18q and 22q (probability of alteration ≥0.35). Several replicates were analysed and they clustered together. Samples with pathological complete response had high probability of gains in 17q, 19q, 20q and 11q. 5/46 samples were HER-2 3+ with IHC and all showed amplification around the 17q locus. Gene expression profiling was carried out using the DASL assay and there was good correlation between IHC for ER with area under the curve of 0.72-0.94 following ROC analysis for the three transcripts assessed .There was statistically significant correlation between ER IHC and gene expression (p<0.05) but this was dependent on the location of the probe.
Conclusion
Reliable copy number and expression data can be generated using these assays and FFPE samples are useful for translational research.
