Real-time quantitative RT-PCR shows variable, assay-dependent sensitivity to formalin fixation: implications for direct comparison of transcript levels in paraffin-embedded tissues.

Real-time quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) is a versatile tool for precise quantification of gene expression. Formalin-fixed and paraffin-embedded (FFPE) tissue is well suited for qRT-PCR, if RNA extraction is optimized and small amplicon sizes are used. However, little is known whether individual assays may show variable sensitivity to fixation. This is of great importance, if a direct comparison of different transcripts is performed within the same sample, such as for mRNA splice variants. We established a cell culture model to test for and quantify differences in performance of individual qRT-PCR assays on FFPE as compared with fresh material, using TaqMan methodology. RNA was isolated from 7 different cell lines either directly or after preparation of a FFPE cell block. RNA from both sources was reverse transcribed and gene expression quantified using 13 different TaqMan assays. All assays allowed highly reproducible target quantification, using both fresh and FFPE-derived cDNA. The 13 assays showed an average Ct difference of 3.2 between fresh and FFPE cells, if identical amounts of cDNA were used as template. However, the Ct shifts varied from 1.8 to 5.1 for individual assays, indicating variable resistance to fixation. These Ct shift differences were statistically highly significant in 27/78 (35%) of all possible combinations of assays. Because the Ct shift remained constant for each assay, they could be used for calculation of correction factors which rendered FFPE-derived expression data highly comparable to those obtained from fresh material, and as a consequence among each other. Thus, a standardized assessment of qRT-PCR assay efficiencies in FFPE allows for precise intraindividual comparison of mRNA species, such as splice variants with different biologic functions, in archival tissues.