Formalin-fixed, Paraffin-embedded (FFPE) tissues

Extracting Gene Expression and miRNA Data from FFPE Samples for Clinical Trials


The Challenge and Potential of FFPE Samples


Formalin-fixed, paraffin-embedded (FFPE) tissue samples can be a rich resource for retrospective discovery studies because large numbers of samples with clinical outcome data can be rapidly acquired and analyzed. Millions of samples have been archived and many thousands are added each year from biopsies and resections. However, because of the RNA degradation and variability between samples, it is generally thought that they are not suitable for gene expression studies. Asuragen scientists have now optimized the methods required to achieve high quality microarray and qRT-PCR data from this important resource, and routinely obtain valuable data for our own in-house research and development and for external service customers.


Example Biomarker Data


With the appropriate RNA isolation methods, quality control procedures, and maximal assay sensitivity, analysis of archived FFPE samples can yield practical biomarker signatures. As an example, Figure 1 illustrates preliminary qRT-PCR validation of microRNAs (miRNA) targets identified by array analysis of FFPE samples in a diagnostic biomarker discovery project. This example demonstrates the use of FFPE samples to identify a panel of differentially expressed targets using microarrays that were subsequently verified by qRT-PCR.



Figure 1: DNA microarray and qRT-PCR data from fixed tissues. Note the remarkably small p-values within each set, illustrating the reproducibility of the markers and procedures relative to the fold-change of the markers (Y axis). The overall Pearson’s correlation coefficient between array and qRT-PCR data is 0.958.


This study demonstrated excellent overall correlation between DNA microarray and qRT-PCR data from FFPE tissues. While this example uses miRNA, Asuragen scientists have performed similar analyses for mRNA targets by microarray and qRT-PCR (data not shown).


RNA Isolation and Quality Control


The major barriers to the use of FFPE tissues are the extreme degradation of the RNA contained in these samples. Figure 2 demonstrates this extreme degradation of RNA in fixed tissues, which renders the most common quality control methods for RNA inadequate.


Figure 2:Comparison of RNA electropherogram from high-quality “intact” RNA (A) to highly degraded RNA typical of FFPE samples (B). The median size of the degraded RNA is approximately 100 nucleotides – significantly larger than miRNA. (~20 nt).


Asuragen scientists have evaluated a number of methods for RNA extraction to develop optimized protocols for preparation and QC. Asuragen has developed and validated these methods for the reliable isolation of total RNA, including miRNA, from fixed archived tissues. In addition, we implemented rigorous quality control to assess each sample prior to final analysis. The first check relies on a spiked-in Armored RNA® process control. The second and third checks use highly specific detection assays on several carefully chosen endogenous mRNAs and miRNAs. Together, these QC assays provide the basis for a “no-call” determination during the development of a biomarker or in clinical practice, as shown in Figure 3. As a result of these efforts, it is now possible to isolate sufficient quantities of RNA with adequate quality for array or qRT-PCR gene expression studies (data not shown).


Figure 3: Example QC data from a process control, miRNA, and mRNA references on paired tumor (TUM) and normal adjacent (NAT) FFPE samples fixed for 1, 4, and 11 years. All data is from the TaqMan system (Applied Biosystems). Circle: process failure based on high CT value of QCref (Armored RNA process control), Star: sample failure based on high CT values of endogenous mRNA and miRNA.


Enhancing qRT-PCR Sensitivity


Another practical concern with fixed tissues can be the amount of tissue available. For some archived specimens or clinical samples very small amounts of FFPE tissue will be obtainable. While we have successfully isolated RNA from sections as thin as 3 µm affixed to slides, the diversity of surgical and fixation procedures and sample-specific effects such as tissue type and tumor or necrotic content often limit the yield of RNA to only a few nanograms. In this context, it is critical to maximize the sensitivity of qRT-PCR-based methods. As shown in Figure 4, novel assay configurations can overcome the amplicon size limit posed by conventional RT-PCR designs to probe even smaller amplicons and allow greater sensitivity.


Figure 4: Relative sensitivity of gene specific qRT-PCR assays as a function of amplicon length. Conventional TaqMan assays are restricted to ~50 nt or larger.


Summary


Despite the highly cross linked and degraded nature of RNA in FFPE tissue blocks, high quality mRNA gene array data can be routinely extracted by the use of optimized procedures. Most clinical applications rely on detection of small regions of select genes using methods such as qRT-PCR. Furthermore, miRNAs, a newly discovered class of small (~20 nt) regulatory RNA, are potentially more robust analytes for both diagnostic and prognostic applications and are very amenable to extraction and analysis from FFPE tissue. Thus, fixed tissue, whether from archives or prospective studies, is now a viable sample for mRNA and miRNA profiling and biomarker signatures. Unlocking the wealth of gene expression information in archived FFPE samples will allow biomarker validation, patient stratification, and even mechanism of action studies from this valuable clinical sample type.


References

  1. Lewis, F., et al., Unlocking the archive--gene expression in paraffin-embedded tissue. J Pathol, 2001. 195(1): p. 66-71.
  2. Masuda, N., et al., Analysis of chemical modification of RNA from formalin-fixed samples and optimization of molecular biology applications for such samples. Nucleic Acids Res, 1999. 27(22): p. 4436-43.
  3. Coombs, N.J., A.C. Gough, and J.N. Primrose, Optimisation of DNA and RNA extraction from archival formalin-fixed tissue. Nucleic Acids Res, 1999. 27(16): p. e12.
  4. Chung, J.Y., T. Braunschweig, and S.M. Hewitt, Optimization of recovery of RNA from formalin-fixed, paraffin-embedded tissue. Diagn Mol Pathol, 2006. 15(4): p. 229-36.