Skip to main content

Quantitative expression profiling in formalin-fixed paraffin-embedded samples by affymetrix microarrays.

Abdueva D, Wing M, Schaub B, Triche T, Davicioni E

Publication: J Mol Diagn (2010) • Vol 12 • Page 409-17PMID: 20522636View on PubMed

Purpose

The purpose of this paper was to determine the suitability of formalin-fixed paraffin-embedded (FFPE) specimens for microarray analysis of gene expression.

Conclusion

RNA yield and quality were significantly reduced in FFPE compared to frozen specimens. Amplification yield was significantly reduced in FFPE tissue compared to frozen specimens and further declined when the specimen had been stored for > 1.5 years. When RNA was extracted from frozen tissues, there were 1.4 fold more present calls, less variance and a larger signal distribution then when the RNA was extracted from FFPE tissue. Further, the signal intensity was reduced in probes detecting the 5' end compared to the 3' end of transcripts from FFPE, but not frozen tissues. The concordance of genes differentially expressed in tumor and normal kidney between FFPE and frozen tissues increased when only genes with a fold change of >5 were compared or when a more stringent p value was used.

Studies (1)

Purpose

The purpose of this study was to compare RNA quality and suitability for microarray analysis of FFPE and frozen kidney, colon and lung tissues and to determine the effects of storage. Different RNA extraction protocols were used for frozen and FFPE specimens.

Summary of Findings

The yield of RNA was significantly lower from FFPE tissue than from frozen specimens (p<0.02). A non-significant decrease in RNA yield was observed with storage both in frozen (2 fold) and FFPE tissue (4 fold). RNA from FFPE tissue was highly degraded (<500 bp) when compared to RNA from frozen specimens. Despite five times more starting RNA, less amplification product was produced from FFPE than frozen tissues (p=0.0004). The amplification yield from FFPE tissue was also 2 fold greater in tissues that had been stored for <1.5 years compared to tissues stored for 7.5 years (p=0.009). When RNA came from frozen tissues, there were 1.4 fold more present calls on the microarray, then when the RNA came from FFPE tissues. Further, the variance was higher from FFPE tissues than from frozen tissues and the raw signal intensity had a smaller range. The signal intensity was reduced in probes detecting the 5' end compared to the 3' end of transcripts from FFPE, but not frozen tissues. Although there was >50% concordance of genes differentially expressed (>2 fold) between tumor and normal tissue for FFPE and frozen specimens, the magnitude of the fold change differed significantly. When only genes with changes of greater than 5 fold were examined, the concordance between FFPE and frozen specimens increased to 90%. Further, the use of more stringent p value cut-offs increased the concordance of differentially expressed genes between tumor and normal tissue in FFPE and frozen specimens.

  • Biospecimens

    TissueKidney
    TissueColorectal
    TissueLung
  • Preservation Types

    FormalinFrozen
  • Diagnoses

    Neoplastic
    NeoplasticNormal Adjacent
  • Platforms

    RNASpectrophotometry
    RNAAutomated electrophoresis/Bioanalyzer
    RNADNA microarray
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen PreservationType of fixation/preservation
    Formalin (buffered)
    Frozen
    StorageStorage duration
    0.5 years
    2.5 years
    7.5 years