With Great FFPE Challenges, Comes Novel Research Opportunities
Have you tried working with poor-quality samples before in your molecular analysis? Were cDNA synthesis yields low? Are sequencing depths in Next-Generation Sequencing (NGS) continuing to increase? Handling poor-quality samples, like formalin-fixed, paraffin-embedded (FFPE) specimens, is notoriously difficult in molecular analysis following the FFPE processing which induces chemical modifications that have plagued researchers for decades. FFPE samples have been around for over 100 years and continue to be important sources for many pathological and clinical studies, with samples frequently being paired with detailed clinical documentation. Poor-quality samples can be challenging to work with, but they don’t have to be, and it has been suggested that NGS is the ideal tool for analyzing FFPE tissues. In this blog post, we will briefly discuss what FFPE samples are, highlight Admera Health’s optimized FFPE workflows, and look to the horizon to see what’s next for these critical scientific resources.
What Are FFPE Samples?
Estimates suggest that over a billion FFPE samples exist in hospitals and tissue banks worldwide (Cazzato G. 2021), indicating their importance for many retrospective studies; but also in prospective investigations, where researchers may be interested in omitting the use of fresh-frozen (FF) tissue (Hedegaard, J 2014). FFPE processing provides an affordable, practical means for preserving highly important tissues (Spencer DH 2013). Formalin preserves tissue morphology and protects cellular protein epitopes that can be used by clinicians and researchers alike. Optimally though, fixation should occur rapidly following tissue harvest to ensure the integrity of nucleic acids for use downstream. So, what does FFPE processing ‘do’ to the sample? In the 1970s, it was discovered that formalin formed crosslinks between DNA and protein (Feldman MY 1973), where now, we know that FFPE samples have a variety of alterations, from individual base modifications to covalent cross-linkage between various other macromolecules (DNA-DNA, DNA-RNA); to base excision and (DNA backbone breakage)—all of which can impair polymerase efficiency during amplification, whereby reducing the library complexity (Steiert TA 2023) and amount of amplifiable template (Sikorsky JA 2007). The most frequent chemical alteration is the spontaneous deamination of cytosine and 5’methyl-cytosine to form uracil and thiamine, respectively, which can cause sequencing artifacts and increase the rate of false positives in the data. This is especially of concern for mutations at lower allelic frequencies (Steiert TA 2023). During the deparaffinization process that removes the wax surrounding the tissue, temperature can also impact the integrity of nucleic acids, where one study suggested an optimal paraffin melting temperature of 75°C for 5 minutes to preserve tissue DNA (Kerick M 2011). There are a variety of approaches to mitigate the chemical modifications following FFPE processing, including pre-analytical sample quality control measures, wet lab DNA repair treatments, analytical sample preparation, and bioinformatic analysis (Steiert TA 2023). Although challenges can be expected when working with highly fragmented nucleic acids isolated from FFPE tissues, NGS may be the ideal platform for analysis due to its ability to assemble information from millions of short DNA fragments (Hedegaard J 2014).
How Admera Health Can Help
Before starting any molecular analysis, it is necessary to evaluate sample concentration and integrity. Admera Health’s expert team has extensive experience in handling FFPE samples from a variety of input types: slides, curls, as well as blocks for cutting. Our extraction procedure uses a magnetic bead-based purification step that combines heat and oil for a gentle overnight deparaffinization process and makes use of specialized buffers during extraction, to reverse macromolecule cross-linkage and yield higher quantity/quality extracts. We provide clients with detailed quality control reports showing the results of extraction, highlighting any potential problem samples of very low amount or quality that can impact successful sequencing and analysis. In this way, we empower our clients to ‘move forward’ with samples that have the best chances for success. Our extensive experience in handling FFPE tissues provides clients with several end-to-end NGS workflows for genomics, bulk RNA sequencing (RNA-seq), and single-cell RNA-seq (scRNA-seq).
Genomics
Our lab has been successful in using FFPE tissues for both whole exome sequencing (WES) and whole genome sequencing (WGS), although increased rates of duplication, smaller library insert sizes, and less uniform coverage are common during analysis and have been described in the literature (Steiret TA 2023). To improve the chances of success, we find using all available material as input during extraction and eliminating fragmentation entirely for highly fragmented samples during library preparation, can be helpful. There are examples of WES & WGS that have compared matched FF & FFPE tissues showing the modifications to FFPE tissues can be managed successfully. For example, one study saw small, but detectable differences in sixteen FF/FFPE-matched lung adenocarcinoma samples when using WES, although there continued to be high concordance (>99.99%) between base calls suggesting the differences were negligible (Spencer DN 2013). Similarly, another group saw strong concordance for 70-80% of the variants between thirty-eight paired FF/FFPE samples using WES, (Hedegaard J 2014). Regarding the use of FFPE tissues for WGS one investigation evaluated fifty-two FF/FFPE matched samples and saw agreement in 71% of single-nucleotide variants (SNVs) that were detected, but also 98% of clinically actionable somatic copy number variations (CNVs) following optimized extraction procedures with reductions to crosslinking temperatures (Robbe P 2018). Thus in general, the above studies confirm the use of FFPE tissues in WES (Spencer DN 2013, Hedegaard J 2014) and WGS (Robbe P 2018), wherein individual success is dependent on each sample's concentration and integrity.
Bulk RNA-seq
We have extensive experience handling FFPE tissues for bulk transcriptomics. In our experience, eliminating fragmentation entirely during library preparation can improve the chances of success downstream. Unlike DNA, RNA presents challenges during library preparation. RNA is less stable compared to DNA due to its single-stranded composition but is also prone to heat degradation and susceptible to RNases ubiquitous to the environment—all of which can reduce the quantity of cDNA generated, limit gene detection, and insert artifacts into the data (Pennock ND 2019). Yet, FFPE-derived RNA has shown robust and reproducible whole transcriptome analysis (Newton Y 2020). For example, the Hedegaard et al. study from above also used RNA-seq to compare the paired FF/FFPE samples and saw expression profiles that were highly correlated between the tissue pairs (r=0.9±0.05 Hedegaard J 2014). A high correlation (r>0.89) was also seen in another study of six FF/FFPE paired ovarian tumor samples (Graw S 2015). It is worth noting that both these studies saw consistently higher gene expression profiles and a lower percentage of reads mapping to intronic regions in the FF compared to FFPE samples (Hedegaard J 2014 & Graw S 2015). A more recent study of twenty-five FF/FFPE replicates also saw a high median correlation (r=0.95) between tissue pairs over the whole transcriptome and showed FF samples were typically associated with templates of greater length and transcript integrity, and had lower GC content (Newton Y 2020). Thus in general, FFPE tissues can be used in bulk RNA-seq, although the use of FF tissue is still preferred in most situations (Hedegaard J 2014, Graw S 2015 & Newton Y 2020).
scRNA-seq
Our lab has seen interest in the Chromium Single Cell Gene Expression Flex workflow since 10x Genomics launched the assay last August, for its ability to perform scRNA-seq on fixed tissues—even FFPE—where recently scRNA-seq has been constrained to either fresh or cryopreserved tissues. An early pre-print showed the ability to do single-nucleus RNA sequencing (snRNA-seq) in FFPE samples by combining an optimized nuclei extraction protocol with the Flex kit, which uses a probe-based system to target the whole transcriptome (Vallejo AF 2022). Vallejo and colleagues compared matched fresh/fixed samples using a prostate cancer cell line and showed the Flex kit detected ~2x more transcripts compared to the “traditional approach” which used the Chromium NextGEM Single-cell 3’ Reagent kit (v3.1, 10x Genomics) that reverse transcribes oligo(dT) captured transcripts; likely a consequence of poor cDNA generation during library preparation from the heavily degraded post-mortem samples (Vallejo AF 2022). Janesick and colleagues also saw a higher median gene sensitivity when using the Flex assay compared to both the Chromium 5’ or 3’ Gene Expression kits, when sequencing depth was kept constant across platforms (~10k reads/cell) (Janesick A 2023). Here, the authors used FFPE sections from human breast tissue and leveraged the shared probe sets between the Flex and Visium Spatial assays (18,536 genes targeted by 54,018 probes) to accurately deconvolute the cell type information in the Visium spatial data for investigation of the tumor microenvironment; highlighting innovative research applications from the integration of single-cell and spatial gene expression datasets (Janesick A 2023). To benchmark the use of FFPE tissue in scRNA-seq, Trinks and colleagues evaluated three patient-matched fresh, cryopreserved, and archival FFPE cancer tissues, to show robust preservation of clinically relevant cell type information in the FFPE tissue vs. fresh tissue, and high correlations between the matched tissues in several clinically relevant signaling pathways (Trinks A 2024). This data suggests that FFPE tissues can be used in scRNA-seq studies for both retrospective and prospective analysis (Trinks A 2024). In our lab, we also find interest in the Flex assay for its ability to batch samples collected at various time points/facilities and minimize any technical effects from analysis over several runs. All things considered, the studies shared here support the use of FFPE samples for scRNA-seq/snRNA-seq, although are currently limited to investigations using human and mouse samples (Janesick A 2023 & Trinks A 2024).
Conclusion
Working with poor-quality samples like FFPE specimens presents both a challenge and opportunity for any researcher. The sheer quantity of FFPE tissues available, particularly when combined with additional clinical patient data, permits novel exploration for many important and unresolved biological questions. The power of NGS to generate information from millions of short DNA fragments, like those isolated from FFPE samples, is why it has been suggested as the ideal tool for working with such samples, although FF tissues are still preferred in most situations (Hedegaard J 2014). The use of FFPE tissues in scRNA-seq with the Flex assay is particularly exciting, overcoming cross-linkage and strand cleavage obstacles that have plagued researchers for decades (Janesick A 2023) and hints at the potential for single-cell profiling of both prospective and retrospective analysis (Trinks A 2024). Stay tuned for upcoming information regarding the use of FFPE tissues for spatial transcriptomic assays, powered by 10x Genomics and STOmics. Working with poor-quality samples can be challenging, but they don’t have to be. Admera Health’s expert team of scientists and optimized workflows provide researchers with a range of solutions for handling FFPE samples, including genomics, bulk RNA-seq, and single-cell transcriptomics.
References
1. Cazzato, G. et al. Formalin-fixed and paraffin-embedded samples for next-generation sequencing: problems and solutions. Genes (2021). https://doi.org/10.3390/genes12101472
2. Feldman, M.Y. Reactions of nucleic acids and nucleoproteins with formaldehyde. Prog Nucleic Acid Res Mol Biol (1973). https:/doi.org/10.1016/s0079-6603(08)60099-9
3. Hedegaard, J. et al. Next-generation sequencing of RNA and DNA isolated from paired fresh-frozen and formalin-fixed paraffin-embedded samples of human cancer and normal tissue. PLoS ONE (2014). https://doi.org/10.1371/journal.pone.0098187
4. Kerick, M. et al. Targeted high throughput sequencing in clinical cancer Settings: formaldehyde fixed-paraffin embedded (FFPE) tumor tissues, input amount and tumor heterogeneity. BMC Medical Genomics (2011). https://doi.org/10.1186/1755-8794-4-68
5. Marshall, C.R. et al. Best practices for the analytical validation of clinical whole-genome sequencing intended for the diagnosis of germline disease. NPJ Genom Med (2020). https://doi.org/10.1038/s41525-020-00154-9
6. 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). https://doi.org/10.1093/nar/27.22.4436
7. Newton, Y. et al. Large scale, robust, and accurate whole transcriptome profiling from clinical formalin-fixed paraffin-embedded samples. Scientific Reports (2020). https://doi.org/10.1038/s41598-020-74483-1
8. Pennock, N.D. et al. RNA-seq from archival FFPE breast cancer samples: molecular pathway fidelity and novel discovery. BMC Med Genom (2019). https://doi.org/10.1186/s12920-019-0643-z
9. Robbe, P. et al. Clinical whole-genome sequencing from routine formalin-fixed, paraffin-embedded specimens: pilot study for the 100,000 Genomes Project. Genetics in Medicine (2018). https://doi.org/10.1038/gim.2017.241
10. Sikorsky, J.A. et al. DNA damage reduces Taq DNA polymerase fidelity and PCR amplification efficiency. Biochemical and Biophysical Research Communications (2007). https://doi.org/10.1016/j.bbrc.2007.01.169
11. Spencer, D.H. et al. Comparison of clinical targeted next-generation sequence data from formalin-fixed and fresh-frozen tissue specimens. Journal of Molecular Diagnostics (2013). https://doi.org/10.1016/j.jmoldx.2013.05.004
12. Steiert, T.A. et al. A critical spotlight on the paradigms of FFPE-DNA sequencing. Nucleic Acids Research (2023). https://doi.org/10.1093/nar/gkad519
13. Trinks, A. et al. Robust detection of clinically relevant features in single-cell RNA profiles of patient-matched fresh and formalin-fixed paraffin-embedded (FFPE) lung cancer tissue. Cellular Oncology (2024). https://doi.org/10.1007/s13402-024-00922-0.
14. Vallejo, A.F. et al. snPATHO-seq: unlocking the FFPE archives for single nucleus RNA profiling. bioRxiv (2022). https://doi.org/10.1101/2022.08.23.505054