Separate quality-control measures are necessary for estimation of RNA and methylated DNA from formalin-fixed, paraffin-embedded specimens by quantitative PCR

Abstract

Estimations of RNA abundance and DNA methylation by quantitative PCR (qPCR) from formalin-fixed, paraffin-embedded (FFPE) tissue specimens are not yet routine in clinical laboratory practice. Excluding specimens with poorly preserved nucleic acids is an important quality-control step for avoiding unreliable results. Because the assays for RNA abundance and DNA methylation have different critical limiting factors, we examined the extent of overlap of excluded specimens for RNA abundance versus methylated DNA. The transcript abundance of three reference genes and of the test gene, estrogen receptor 1 (ESR1), was estimated by SYBR Green qPCR in 250 breast cancer specimens. The estrogen receptor (ER) protein was identified by IHC, and concordance between ESR1 and ER was estimated by Cohen's κ. TaqMan PCR for the ALU-C4 sequence was performed with bisulfite-treated DNA to determine usability in the MethyLight assay. Excluding specimens with mean reference gene CT values exceeding the group mean by >1 SD led to significant improvement of the concordance of ESR1 and ER. Specimens with usable DNA after bisulfite treatment likewise had ALU-C4 CT values of less than the group mean + 1 SD. Samples with low-quality RNA and DNA were partly nonoverlapping. RNA and DNA extracted from the same FFPE block need separate exclusion criteria for qPCR assays of transcript abundance and methylated DNA.