Expression Level of Estrogen Receptor Determined by PCR Quantification of mRNA from Whole Slide Formalin-Fixed, Paraffin-Embedded Breast Cancer Tissue Sections Correlates with Semiquantitative Immunohistochemical Level and Biochemical Measurements.
Trine Tramm, Guido Hennig, Jan Alsner, Flemming B Soerensen, Jens Overgaard. Aarhus University Hospital, Denmark; Siemens Healthcare Diagnostics Holding GmbH, Eschborn, Germany; Vejle Hospital, Denmark
Background: Formalin fixed, paraffin embedded tissue (FFPE) often contains an admixture of normal tissue, premalignant changes and invasive cancer. This has called into question the specificity of the results from gene expression analysis of the total amount of mRNA extracted from a whole slide tissue section. Often a threshold for tumor cell content in the FFPE blocks is being used or macrodissection is performed to avoid contamination from non-neoplastic tissue. The use of FFPE for gene expression analysis has been considered impractical, time consuming and characterized by a low grade of automation. The need for macrodissection further ads to these issues.
Design: Two FFPE blocks from each of 21 breast carcinomas were chosen. From each block a whole section and a manually trimmed, tumor enriched section (discarding surrounding non-invasive tissue) were prepared. mRNA was isolated with a fully automated technique currently under development at Siemens (Siemens Healthcare Diagnostics, Deerfield, IL). Tumor content defined as invasive carcinoma with interposed stroma was estimated stereologically. Eluates were analyzed with qRT-PCR for housekeeping gene RPL37A and 3 target genes (ESR1, PGR and ERBB2). Raw data (CT values) for target genes were normalized to RPL37A, and relative expression levels calculated and compared to immunohistochemical and biochemical data.
Results: RNA was successfully extracted from all sections, and gene expression reliably quantified. Agreement between whole slide and trimmed sections were optimal, indicating that expression levels for ESR1, PGR and HER2 are not strongly influenced by contamination from surrounding tissue. Agreement between RNA- and protein expression determined by immunohistochemistry as well as by Enzyme Immuno Assay was excellent for ESR1, making it possible to define a mRNA treshold, distinguishing between ESR1-positive and negative samples.
Conclusions: Isolation and quantification of ESR1, PGR and HER2 mRNA from FFPE with qRT-PCR are feasible without prior trimming of tissue. High level of agreement between quantitative mRNA expression level, semiquantitative IHC level and biochemically measured quantitative protein level for ESR1. Quantitative expression analysis using qRT-PCR in routinely processed FFPE is feasible and could be adapted in diagnostic testing.
Tuesday, March 1, 2011 9:30 AM
Poster Session III # 51, Tuesday Morning