Skip to main content

Analyses and interpretation of whole-genome gene expression from formalin-fixed paraffin-embedded tissue: an illustration with breast cancer tissues.

Kibriya MG, Jasmine F, Roy S, Paul-Brutus RM, Argos M, Ahsan H

Publication: BMC Genomics (2010) • Vol 11 • Page 622PMID: 21059268View on PubMed

Purpose

The purpose of this paper was to determine the effects of using formalin-fixed paraffin embedded (FFPE) specimens rather than frozen specimens on the identification of genes differentially expressed between breast tumor and normal adjacent tissue by cDNA-mediated Annealing, Selection, extension and Ligation (DASL) assay.

Conclusion

Similar signal intensities and a similar number of probes were detected in frozen and FFPE specimens. However, the largest variation in expression between the two type of specimens was attributed to specimen preservation method (30.32%) followed by interindividual variability (25.43%) and disease state (4.94%). 1864 genes showed a more than 2 fold difference in expression between matched frozen and FFPE specimens. Analysis of variance (ANOVA) identified an interaction between diagnosis and formalin fixation for 772 genes, resulting in improper classification or exclusion as tumor markers. Using Gene Set Enrichment Analysis (GSEA), 100 genes were identified as being differentially expressed between frozen tumor and normal adjacent tissue, but 440 genes were identified as being differentially expressed between FFPE tumor and normal adjacent tissue, including only 38% (38 genes) of those identified using frozen specimens.

Studies (1)

Purpose

The purpose of this study was to determine the effects of using FFPE rather than frozen specimens on the identification of genes differentially expressed between breast tumor and normal adjacent tissue by DASL. After xylene deparaffinization of the FFPE specimens and proteinase K digestion of all specimens, RNA was extracted using the HighPure kit.

Summary of Findings

Similar signal intensities and a similar number of probes were detected in frozen and FFPE and tumor and normal adjacent specimens. Further, the correlation in expression between normal adjacent and tumor specimens was r2=0.85 for frozen and r2=0.72 for FFPE specimens. The largest variation was attributed to specimen preservation method (30.32%) followed by interindividual variability (25.43%) and disease state (4.94%). 1864 genes showed a more than 2 fold difference in expression between matched frozen and FFPE specimens. Importantly, the transcripts affected by formalin fixation included 5 of the 61 identified breast cancer related genes in the Breast Cancer Gene Database (BCGD) (APC, CDKN2A, IGF1R, TGFA and TSG101) and estrogen receptor alpha (ESR1, ER alpha). Cluster analysis effectively separated frozen and FFPE specimens. There was also a significant effect of block age and batch effect, particularly for FFPE specimens (F=2.56 and 2.34 for FFPE, respectively; and F=0.48 and 1.03 for frozen, respectively). Interestingly, ANOVA analysis identified an interaction between diagnosis and formalin fixation for 772 genes. Consequently, when FFPE tissues were used, genes were identified as being differentially expressed between tumor and normal adjacent tissue when no such effect was found in frozen specimens, or vice versa. Finally, using Gene Set Enrichment Analysis, 100 genes were identified as being differentially expressed between frozen tumor and normal adjacent tissue, but 440 genes were identified as being differentially expressed between FFPE tumor and normal adjacent tissue, including only 38% (38 genes) of those found in frozen specimens.

  • Biospecimens

    TissueBreast
  • Preservation Types

    FrozenFormalin
  • Diagnoses

    NeoplasticNormal Adjacent
    NeoplasticCarcinoma
  • Platforms

    RNADASL
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen PreservationType of fixation/preservation
    Formalin (buffered)
    Frozen
    StorageStorage duration
    3-6 years