Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Robustness of RNA sequencing on older formalin-fixed paraffin-embedded tissue from high-grade ovarian serous adenocarcinomas.

Author(s): Zhao Y, Mehta M, Walton A, Talsania K, Levin Y, Shetty J, Gillanders EM, Tran B, Carrick DM

Publication: PLoS One, 2019, Vol. 14, Page e0216050

PubMed ID: 31059554 PubMed Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   This study investigated the suitability of archival serous ovarian adenocarcinoma FFPE blocks for NGS and investigated the effects of block storage duration on RNA quality and sequencing metrics. The authors also compared different RNA quality assessment methods for evaluating specimen suitability for NGS. This study included 67 FFPE high-grade serous ovarian adenocarcinomas that were retrospectively collected from three sites in the Surveillance, Epidemiology and End Results (SEER) registry. As specimens were collected retrospectively, details surrounding the formulation of formalin used, processing regime, storage temperature and conditions were not provided and are potentially confounding factors. Storage duration calculated from collection to RNA extraction ranged from 7-32 years. RNA was extracted from five 10 µm thick sections using the AllPrep DNA/RNA FFPE Kit following xylene deparaffinization. RNA quality was assessed using an Agilent Bioanalyzer and the RT² RNA QC PCR Array which is based on ACTB (174 bp) and HPRT1 (57 bp) amplicons. NGS libraries were prepared from 21-1000 ng RNA using NEBNext Ultra II RNA Library Prep Kit without fragmentation and sequenced using the Illumina NextSeq 500 using NextSeq High Output v2 kit using 1x151 bps single-read run for 64 specimens and 2x76 bps for the remaining 3 samples. Twenty-five specimens with low sequencing yields were resequenced using 1x151 bps runs to get a higher sequencing depth. Bacterial contamination of the RNA was evaluated using FastScreen. The sections before and after those used for RNA extraction underwent pathology review to confirm that ≥50% of cells had nuclei consistent with malignant cells, and ≤50% of cells were necrotic.

   Summary of Findings:

   RNA integrity was successfully evaluated in 61 of the 67 specimens evaluated, yielding a median RIN  

   of 2.4 (range 1.0-2.8). RINS were weakly correlated with storage duration (R=0.15). The RIN values for 5 of 7 pairs with replicate curls were comparable but the remaining two pairs had dissimilar RINs, as one replicate had a RIN close to 1 with the other 2.3-2.4. The median percentage of fragments >200 (DV200) was 13% (range 1-50%), but the median DV100 and DV50 values were 56% (11-86%) and 98% (76-100%), respectively, indicating highly fragmented RNA. Importantly, DV estimates could not be calculated for 14 of the 67 specimens.  The median DV200 and DV100 values were 20% and 53%, respectively, for specimens stored 7-12 years, 14% and 46%, respectively, for specimens stored 13-22 years, and 8% and 34%, respectively, for specimens stored 23-32 years. The DV100 and DV200 values were modestly negatively correlated with storage duration (R=-0.39 and R=-0.44, respectively) indicating an adverse effect of storage on RNA integrity. Storage duration was also weakly negatively correlated with library yield (R=-0.21), with an average yield of 500 ng, 87 ng and 41 ng for specimens stored 7-12, 13-22 and 23-32 years, respectively.  Overall 69-91% of bases had a quality score of >30, the percentage of reads mapped ranged from 7.4-94.9%, the percent of unique mapping ranged from 2-86%, and the percent of non-duplicated reads ranged from 1-76%, indicating a wide range of quality in sequencing results. Further indicating inconsistent effects of block storage, intronic bases varied from 17-72%, intragenic bases ranged from 11-81%, and mRNA bases ranged from 1-34%. The percentage of UTR, intronic and intragenic bases were very weakly correlated with FFPE block storage (R=-0.13, R=-0.16 R=0.21).  Further, the GC content in some specimens was much lower (5%) than the expected 40-60% while in others it was much higher (85%). Interestingly, both 5’ and 3’ biases were observed in the data set, but this bias was less prevalent in specimens with an mRNA percentage >10. The percentage of reads mapped and non-duplicate reads were most strongly correlated with DV100 (R=0.58 and R=0.53, respectively) followed by DV200 (R=0.43 and R=0.50, respectively), with lower correlations observed with RIN (R=0.14 and R=0.19, respectively). Based on the superior resolution of DV100 compared to DV200 the authors suggest a DV100 >40 be used as a cut-off for sequencing and using a higher input volume when DV100<60 as opposed to Illumina’s recommendation of DV200. The real-time PCR-based RT RNA QC PCR Array was unable to amplify ACTB or HPRT1 in any of the FFPE specimens evaluated although no PCR inhibition was detected. More than 1% bacterial contamination was detected in 56 of the 67 specimens, with 41 specimens having >5% bacterial contamination and 28 specimens having >5% contamination. Mouse contamination was detected in 6 specimens obtained from two different sites. Importantly, the collection/storage site did not significantly affect the DV100 value or the percentage of mRNA. An analysis of 189 genes that were highly expressed across specimens included the P13K-Akt signaling pathway, indicating that valuable NGS data is still obtainable from FFPE specimens regardless of quality; however, while  correlations in sequencing data between curls of the same specimen were strong for 4 specimen pairs (R=0.7-0.8) they were weak (R<0.22) for the remaining 3 pairs. Clustering based on expression of the 193 genes in the TCGA dataset revealed that 26 of the FFPE samples had a gene expression pattern similar to specimens in TCGA fresh-frozen ovary tumor and the human body normal ovary data sets.

   Biospecimens

   • Tissue - Ovary

   Preservative Types

   • Formalin

   Diagnoses:

   • Neoplastic - Carcinoma

   Platform:

   Analyte	Technology Platform
   RNA	Automated electrophoresis/Bioanalyzer
   RNA	Real-time qRT-PCR
   RNA	Next generation sequencing

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Storage	Storage duration	7-12 years 13-22 years 23-32 years
   Automated electrophoresis/Bioanalyzer Specific	Technology platform	RIN DV200 DV100

   View more