Skip to main content

Gene expression profiles from formalin fixed paraffin embedded breast cancer tissue are largely comparable to fresh frozen matched tissue.

Mittempergher L, de Ronde JJ, Nieuwland M, Kerkhoven RM, Simon I, Rutgers EJ, Wessels LF, Van't Veer LJ

Publication: PLoS One (2011) • Vol 6 • Page e17163PMID: 21347257View on PubMed

Purpose

The purpose of this paper was to determine the effects of fixation on cDNA-mediated annealing, selection, extension, and ligation (DASL) analysis of breast tumors.

Conclusion

DASL hybridization and average signal intensity was lower in formalin-fixed paraffin-embedded (FFPE) specimens than frozen tissues, but 94% of probes that hybridized in frozen specimens also did so in FFPE specimens. The hybridized probes that were enriched in the FFPE specimens were more likely to be GC rich and represent cell-cycle genes and the underrepresented hybridized probes had lower GC content and represented RNA processing, DNA repair and transport genes. FFPE technical replicates showed a correlation of 0.95 compared to 0.99 for frozen specimens. In clustering analysis, FFPE and the matched frozen specimens grouped together with a correlation of 0.65-0.89. In FFPE tissue, estrogen receptor alpha (ER-alpha), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) expression correlated well with immunohistochemical (IHC) staining. Generally, when expression was matched to signatures used for diagnosis or prognosis, FFPE specimens correlated well with frozen specimens.

Studies (1)

Purpose

The purpose of this study was to compare gene expression in FFPE and frozen breast specimens by DASL.

Summary of Findings

Hybridization and average signal intensity was lower in FFPE specimens than frozen tissues (p<0.001). There were 11883 probes hybridized in all of the FFPE samples. 94% of these overlapped with the 13370 probes hybridized in all frozen specimens. Cell cycle gene probes that hybridized were enriched in FFPE tissue and RNA processing, DNA repair and transport probes were underrepresented. FFPE technical replicates showed a correlation of 0.95, and the correlation between different specimens was 0.91. In contrast, the correlation for technical replicates of frozen specimens was 0.99. After median recentering, FFPE and the matched frozen specimens grouped together, along with their technical replicates, with a correlation of 0.65-0.89 considering all probes. In contrast, the correlation between non-matched pair FFPE and frozen specimens was 0.002-0.83. In FFPE tissue, the correlations of ER-alpha, PR, and HER2 expression with IHC staining results were 0.71 (p<0.01), 0.89 (p<0.01) and 0.57-0.82 (p<0.01), respectively. When FFPE gene expression data was classified by the single sample predictor or genomic grade index, 19 of 20 FFPE specimens agreed with the frozen counterpart. Using a 60 gene index derivative of the 70 gene prognosis signature, there was a correlation of 0.94 between matched frozen and FFPE specimens, and both specimen types correlated highly with the MammaPrint Indices (>0.88). The probes for which there was the highest correlation between FFPE and frozen specimens were GC rich.

  • Biospecimens

    TissueBreast
  • Preservation Types

    FormalinFrozen
  • Diagnoses

    NeoplasticCarcinoma
  • Platforms

    RNADASL
    ProteinImmunohistochemistry
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen PreservationType of fixation/preservation
    Formalin (buffered)
    Frozen
    Immunohistochemistry SpecificTargeted peptide/protein
    Estrogen receptor alpha
    HER2
    Progesterone receptor
    DASL SpecificTechnology platform
    Immunohistochemistry