Skip to main content
Premium Trial:

Request an Annual Quote

Classic Samples, New Analysis


Formalin-fixed, paraffin-embedded tissue is routinely used for histopathological diagnosis, and by some accounts FFPE tissue remains the best tissue for microscopic examination in a clinical environment. However, FFPE tissue has had its issues — such as cross-linking of proteins and yield — that make its use in 'omics-related research and non-clinical settings challenging. It also contains RNA that can be fragmented during fixation and storage, making FFPE unsuitable for certain types of analysis. Genome Technology's Tony Fong spoke with Stephen Hewitt from the National Cancer Institute's Laboratory of Pathology about the current state of FFPE technology. What follows is an excerpt of that conversation, edited for space.

Genome Technology: What classic problems associated with using FFPE tissue for 'omics research have not been addressed adequately?

Stephen Hewitt: I think that the use of formalin-fixed, embedded tissue in 'omics research too frequently remains the use of convenience samples, meaning samples that were not collected with the intent to be used for 'omics research. They may be old clinical samples, they may have been research specimens, but no one at the time of their collection was anticipating they would be used for the isolation of biomolecules. As a result, the standard operating procedures … are frequently either unknown or non-existent.

GT: What standard operating procedures does a researcher have to follow handling this type of tissue for 'omics experiments that they may not have to follow when it is used for other types of research?

SH: That's been one of the great challenges — that there has been a real diversity of standard operating procedures for formalin-fixed, paraffin embedding, and it's common that one will read a manuscript and it will say, 'per standard protocol.' And what you discover is that that protocol varies hospital to hospital, clinic to clinic.

Recently, [NCI's] Office of Biospecimens and Biorepository Research [now known as the Biorepositories and Biospecimen Research Branch] has generated a set of guidelines for SOPs for handling tissue for research purposes. More directly, in a clinical setting ... the Clinical and Laboratory Standards Institute has issued a set of guidelines for immunohistochemistry that provide SOPs for handling formalin-fixed, embedded tissue in a clinical setting. These guidelines still have a fair amount of diversity, but the difference is that they finally do set boundary parameters and are based on data.

GT: Is FFPE suitable for research use or is it really still for use in clinical studies?

SH: The way I'd describe it would be I'm not sure I would use formalin-fixed, paraffin-embedded tissue for discovery, but I certainly would use it for validation.

GT: Why?

SH: There is no doubt [that] there are going to be differences between fresh and fixed tissue. Fresh tissue, if of high quality, will have limited degradation, while formalin-fixed tissue will always demonstrate degradation. Fresh and frozen tissue may be degraded if poorly handled or previously defrosted, but can be of high quality if collected and handled well. Formalin-fixed material is cross linked and will generally show features similar to prolonged hypoxia.

When you're trying to make something that you're going to use as a test in a clinical setting, you want something that's robust, reproducible. And at that point, you're looking at the selection process of 'Well, what's going to be something that I can measure in the real-world?' And that's why I always try to make that jump to FFPE earlier rather than later. What I don't like to see is somebody who discovered something and then did a bunch of experiments on frozen samples for the next two, three, four, five years. Then it's like 'I'm ready to put it in a clinical trial,' and then all of a sudden [they] run smack into the wall. Let's try to find out where the wall is earlier.

GT: Is yield an issue with DNA or proteins?

SH: For DNA generally, unless you are really dependent on a fragment length assay, yield is no longer a substantial issue. PCR is just so robust, and, unless you really need longer fragments, it's just really not an issue. For proteins, yield remains an issue, but it's being overcome by more sensitive proteomic technologies. The mass specs are pretty sensitive now. … The difference is RNA. In RNA, only 30 percent of the molar weight of the RNA is present in FFPE as compared to the same starting material in frozen tissue, whereas for DNA and protein, that number is substantially higher and are probably in the order of 70 [percent] to 90 [percent] and possibly higher.

GT: What are some of the more common missteps that researchers make when they use FFPE tissue?

SH: Probably the most common is not using the right tissue, not having the consultant pathologist or collaborative pathologist confirm that that's the right tissue and that it's got the right lesion and everything else. That still, unfortunately, undermines a lot of work. I think that's the biggest problem. I think the second one is that … FFPE tissue does degrade, and people will think 'I've got an FFPE block, I can still get X out of it.' And it is degrading over time. Slides degrade much faster than blocks do, but blocks degrade for RNA content, and we've been able to demonstrate that proteins will degrade in them as well. And so people need to make sure that their FFPE blocks are obtained from the highest quality sources. Not every FFPE block is equal. And that's an area where there's still a lot of active research. It's a real challenge.

The Scan

Study Links Genetic Risk for ADHD With Alzheimer's Disease

A higher polygenic risk score for attention-deficit/hyperactivity disorder is also linked to cognitive decline and Alzheimer's disease, a new study in Molecular Psychiatry finds.

Study Offers Insights Into Role of Structural Variants in Cancer

A new study in Nature using cell lines shows that structural variants can enable oncogene activation.

Computer Model Uses Genetics, Health Data to Predict Mental Disorders

A new model in JAMA Psychiatry finds combining genetic and health record data can predict a mental disorder diagnosis before one is made clinically.

Study Tracks Off-Target Gene Edits Linked to Epigenetic Features

Using machine learning, researchers characterize in BMC Genomics the potential off-target effects of 19 computed or experimentally determined epigenetic features during CRISPR-Cas9 editing.