Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Alternative tissue fixation for combined histopathological and molecular analysis in a clinically representative setting.

Author(s): Meecham A, Miranda E, Morris HT, Hair J, Oien KA, Gerrard G, Guppy N, Mooney D, Shaw EC, Ashton-Key M, Lees R, Flanagan A, Rodriguez-Justo M

Publication: Histochem Cell Biol, 2021, Vol. 156, Page 595-607

PubMed ID: 34905068 PubMed Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   This study compared H&E, IHC, and FISH staining scores in matched FFPE and PFPE sections from 203 tissue specimens. The yield, purity, integrity and amplifiability of DNA and RNA and NGS metrics of DNA extracted from PFPE sections were compared with that from FFPE sections. Matched ~5mm³ segments of 203 tissue resection specimens (61 colorectal, 43 prostate, 32 lung, 19 lymph node, 12 breast, 9 bladder, 7 kidney, 6 sarcoma, 3 esophagus, 3 stomach, 2 pleura, 2 ovary, 1 thymus, 1 pancreas, 1 skin, and 1 spleen, diagnosis not specified) were placed in buffered formalin and PAXgene Tissue fixative and stored at 4°C. Within 24 h, PAXgene specimens were transferred to PAXgene stabilizer solution, and all specimens were then processed overnight in an Excelsior AS Processor and embedded in paraffin. FFPE and PFPE blocks were stored at -20°C until use. To evaluate morphological preservation, 4µm sections were H&E stained, and nuclear, cytoplasmic, and cell membrane features were scored (0–4) by two pathologists. Protein expression was visualized by immunohistochemistry staining. As part of IHC optimization, staining was compared in eight PFPE specimens with or without the addition of formalin fixation for 24 h prior to staining with four antibodies. To visualize RNA localization, PFPE sections were incubated in formalin for 24 h after which FISH was performed on both FFPE and PFPE sections. DNA and RNA were extracted from four FFPE sections for each of the 65 specimens using the AllPrep DNA/RNA FFPE Kit and from PFPE sections using the PAXgene Tissue DNA Kit and PAXgene Tissue RNA Kit, respectively. DNA and RNA were quantified using a Qubit3 Fluorometer and purity was assessed by spectrophotometry. DNA and RNA fragmentation were analyzed on an Agilent 2200 Tapestation. DNA quality was further assessed using the real-time PCR-based Infinium FFPE QC Kit, and RNA quality was assessed by real-time PCR amplification of Glucuronidase Beta (GUSB). Sequencing libraries were constructed from matched DNA isolated from 4 specimens using the 50-gene Ampliseq Cancer Hotspot Panel v2 and Oncomine Solid Tumor Library Prep Kit and sequenced using an Ion Torrent Personal Genome Machine.

   Summary of Findings:

   Membrane and nuclear staining scores were comparable between FFPE and PFPE sections, but cytoplasmic staining scores were lower in PFPE than FFPE sections (3.508 versus 3.633, P=0.0323). However, when data from each of the three collection sites was analyzed independently, significantly higher cytoplasmic staining scores were observed in PFPE than matched FFPE specimens at one site (site 3, 16 specimens, P=0.0061), lower scores in PFPE than FFPE specimens were observed at another site (site 2, 94 specimens, P=0.0001), and comparable cytoplasmic staining scores were observed in FFPE and PFPE specimens at the third site (Site 1, 70 specimens). Further comparison at the site level revealed no significant differences in morphological staining overall or specifically for nuclear, cytoplasmic, or cell membrane staining between FFPE and PFPE specimens from site 1, but significant differences (in opposing directions) were observed between FFPE and PFPE specimens for each parameter at the other two sites. The authors state the pathologists reported increased eosin staining, tissue fragility, poorer lymph node preservation, and swelling of the erythrocytes in PFPE specimens. However, the preference for FFPE versus PFPE was both tissue- and site-dependent, with each site having a different preference for each tissue type. Similar to H&E staining, significant differences in IHC staining were observed between FFPE and PFPE specimens, but the differences depended on the tissue type and preservation method of the specimen analyzed. Higher mean staining was observed in FFPE than PFPE for 4 of the 7 antibodies evaluated in colon specimens, 2 of 3 antibodies in lung specimens, and 1 of 6 antibodies in prostate specimens. Higher mean staining in PFPE relative to FFPE was only observed for the MNF116 antibody in lung specimens (P=0.0005). Pretreatment of PFPE sections with formalin improved staining of 3 of the 4 antibodies evaluated.  The authors report that the observers generally preferred staining in FFPE to PFPE sections based on blind scoring. FISH staining was adequate in 100% of FFPE and 92.5% (37 of 40) of PFPE sections, but EGFR staining was significantly higher in FFPE than PFPE sections (P< 0.0001). While not significant, a trend toward higher cyclin-dependent kinase inhibitor 2A (CDKN2A) staining intensity was observed in PFPE than FFPE.

   DNA yield was slightly but not significantly higher from PFPE than FFPE sections. DNA fragment length and DNA integrity numbers were both significantly higher for DNA extracted from PFPE than FFPE specimens (P<0.001, both). qPCR analysis using the Infinium FFPE QC Kit was successful using DNA from 33 of 34 FFPE specimens and 32 of 34 PFPE specimens, but mean dCT values were lower (better) using DNA from PFPE than FFPE specimens (-0.58 versus 7.98, P<0.01). NGS quality control metrics for the 4 matched specimens, found a significantly longer mean read length for DNA from PFPE than FFPE specimens (P<0.05), but the number of on-target reads, uniformity, and the number of variants detected were comparable. RNA yield was slightly higher from FFPE than PFPE specimens (P=0.03), but RIN and RNA purity were significantly higher for RNA isolated from PFPE than FFPE specimens (2.85 versus 2.03, P<0.01 and 3.66 versus 2.98, P<0.01, respectively). Interestingly mean CT values for real-time PCR amplification of GUSB were significantly lower for RNA from FFPE than PFPE specimens (18.72 versus 27.5, P<0.01).

   Biospecimens

   • Tissue - Breast
   • Tissue - Colorectal
   • Tissue - Lymph Node
   • Tissue - Prostate
   • Tissue - Spleen
   • Tissue - Kidney
   • Tissue - Stomach
   • Tissue - Ovary
   • Tissue - Pancreas
   • Tissue - Thymus Gland
   • Tissue - Skin
   • Tissue - Lung
   • Tissue - Esophagus
   • Tissue - Bladder

   Preservative Types

   • PAXgene
   • Formalin

   Diagnoses:

   • Not specified
   • Neoplastic - Carcinoma
   • Neoplastic - Sarcoma

   Platform:

   Analyte	Technology Platform
   Morphology	H-and-E microscopy
   Protein	Immunohistochemistry
   DNA	Real-time qPCR
   DNA	Spectrophotometry
   DNA	Next generation sequencing
   RNA	Spectrophotometry
   RNA	In situ hybridization
   RNA	Automated electrophoresis/Bioanalyzer
   RNA	Real-time qRT-PCR
   DNA	Automated electrophoresis/Bioanalyzer
   RNA	Fluorometry
   DNA	Fluorometry

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Biospecimen Acquisition	Biospecimen location	Lung Colon Prostate Lymph Kidney Stomach Breast
   Biospecimen Acquisition	Locale of biospecimen collection	3 different collection/analysis sites compared
   Real-time qRT-PCR Specific	Targeted nucleic acid	GUSB
   In situ hybridization Specific	Targeted nucleic acid	CDKN2A EGFR
   Immunohistochemistry Specific	Targeted peptide/protein	MLH1 MSH2 PMS2 MSH6 p53 p16 p63 EGFR TTF1 MNF116 34BE12 CK5
   Biospecimen Preservation	Type of fixation/preservation	PAXgene followed by Formalin Formalin (buffered) PAXgene

   View more