In the course of patient care, formalin-fixation and paraffin-embedding (FFPE) of biopsy tissue samples are routinely performed, where these samples can be analyzed by histology and archived to link the sample with clinical long-term follow-up. With the development of advanced NGS-based oncology gene panels, it is becoming increasingly important to consider pre-analytic variables when extracting nucleic acids from FFPE-treated samples. This post covers frequently asked questions (FAQs) around the extraction of nucleic acids from FFPE samples for downstream NGS analysis.
During the typical FFPE embedding process, a sample is first exposed to formaldehyde, which reacts with proteins and other primary and secondary amine-containing compounds. This reaction leads to extensive crosslinking, which stabilizes the structure of the sample for subsequent processing and histological analyses. Longer fixation times tend to lead to better stabilization for such analyses. However, longer fixation times also make the nucleic acids in the sample more difficult to access, and – ideally – fixation should not be carried out for more than 24 hours if the goal is to analyze the nucleic acids in the sample.
After fixation, formalin-fixed samples are first dehydrated with ethanol in order to remove water, which is not miscible with xylene. Then, xylene is used in order to remove ethanol, which is not miscible with paraffin. Finally, xylene is replaced with molten paraffin, which is later allowed to harden in order to generate FFPE blocks.
These can subsequently be cut into sections that are placed onto slides, or into tubes as curls. For histological analyses, 5 micron sections are typical because they allow for improved assessment of tissue and cell morphology compared to thicker sections. For next-generation sequencing (NGS) analyses, 10 micron sections are becoming more common because they contain twice the amount of sample – especially, when any microdissection is to be performed.
FFPE extraction reverses this process, although the reagents and steps differ between manufacturers. In all cases, it is essential to first remove paraffin in order to make the nucleic acids and proteins accessible to aqueous solutions. It is also essential to use proteinase K, protease, or similar in order to degrade the crosslinked structures surrounding the nucleic acids. If this reaction is not carried out to completion, then the ability to recover and analyze nucleic acids is compromised.
The following suggestions may improve the ability to extract nucleic acids reproducibly from FFPE curls.
- The QIAGEN QIAamp® DNA FFPE Tissue Kit and QIAamp® DSP DNA FFPE Tissue Kit and Ambion RecoverAll™ Total Nucleic Acid Isolation Kit for FFPE dissolve paraffin with xylene. The “pellet” that may result from the subsequent centrifugation step can be very difficult to see and can detach easily – especially, in the presence of any static electricity.
- For this reason, xylene is best removed with a very narrow pipette tip that is unlikely to aspirate a significant amount of the “pellet.” The pellet in the subsequent ethanol wash is more visible and better attached to the wall of the tube, but careful pipetting is still necessary. Evaporating all ethanol is important for optimal proteinase K digestion.
Regarding Proteinase K digestion:
- With the QIAGEN kits, Proteinase K digestion is carried out at 56 °C for 1 hour “or until the sample has been completely lysed.” Carrying out the digestion until the sample has been completely lysed is important or DNA yields may be low or biased to regions that are more accessible. According to the related DNeasy® Blood & Tissue Kit that includes similar reagents, “if it is more convenient, samples can be lysed overnight; this will not affect them adversely.” Similarly, with the Ambion kit, digestion is carried out for 16 hours at 50 °C if the goal is to isolate DNA and “increasing the incubation time at 50°C to up to 48 hr usually results in recovery of DNA with slightly increased functionality.”
- However, it should be noted that for the therascreen® EGFR and KRAS RGQ PCR Kit companion diagnostics that make use of the DSP version of the kit, “proteinase K digestions … must be performed for 1 hour” because the assays were validated this way.
Roche Companion Diagnostics kits for EGFR and KRAS mutation testing:
- The Roche cobas® EGFR and KRAS Mutation Test companion diagnostics use a similar protocol to the QIAGEN kits, but with two important differences.
- First, during deparaffinization, an equal volume of ethanol is added directly into the sample after the paraffin has been dissolved in xylene. This limits the amount of FFPE material that can be processed because paraffin is only sparingly soluble in 50% xylene 50% ethanol. However, the resulting pellet appears and behaves more like a pellet that is in ethanol, so the removal of the supernatant is easier than the removal of xylene, but careful pipetting is still necessary.
- Second, the amount of added proteinase K is about 4-fold higher, so samples may get digested more in the prescribed 1 hour at 56 °C. However, the buffers used for digestion differ with Roche including Triton X-100 and QIAGEN including SDS.
- A modification of the Roche deparaffinization procedure can improve the handling of the sample pellet. This involves adding an equal volume of PBS relative to the xylene after the addition of ethanol. The mixture of 500 µl xylene, 500 µl ethanol, and 500 µl PBS leads to a separation of phases. After centrifuging, the pellet is in ~1000 µl of 50% ethanol and under ~500 µl of xylene/paraffin. The upper organic and most of the lower aqueous phase can be removed and the pellet washed in ethanol and dried. The subsequent proteinase K digestion should be carried out for longer than usual.
Alternative deparaffinization reagents:
- The QIAGEN QIAamp® DNA FFPE Tissue Kit has a supplementary protocol that uses their Deparaffinization Solution. Based on the applicable SDS, this solution is based on the hydrocarbon cetane, which is at the lower end of mineral oils and slightly larger than naphtha, which is used as a substitute for xylene in histology (e.g. Thermo Scientific’s Shandon™ Xylene Substitute). Using this reagent makes sample processing easier because no pellets are generated. However, it may be helpful to increase the recommended incubation times for deparaffinization and digestion because DNA recovery can be lower than with other methods and biased.
- The same recommendation applies to methods that use mineral oil, hot soapy water, etc. in order to remove paraffin. If the entire sample is not accessible to proteinase K or protease, then digestion will be incomplete. If digestion is not carried out long enough, then only a subset of released DNA – which may be biased – can be analyzed.