Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Analysis of quality metrics in comprehensive cancer genomic profiling using a dual DNA-RNA panel.

Author(s): Watanabe K, Kohsaka S, Tatsuno K, Shinozaki-Ushiku A, Isago H, Kage H, Ushiku T, Aburatani H, Mano H, Oda K

Publication: Pract Lab Med, 2024, Vol. 39, Page e00368

PubMed ID: 38404525 PubMed Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   This study compared two DNA integrity markers (Q-value and ddCq) and one RNA integrity metric (DV200) using DNA and RNA extracted from 585 formalin-fixed, paraffin-embedded (FFPE) blocks that had been stored for <1, 1-2, 2-3, 3-5 y or approximately 5 years before extraction. The authors investigated the predictive performance of these metrics for four next-generation sequencing quality metrics. Potential effects associated with the hospital of collection/processing and cancer type on the DNA/RNA quality metrics were also examined. A total of 585 tissues that included 114 lung, 90 bowel, 58 ovarian/fallopian tube, 51 uterus, 38 soft tissue, 31 cervix, and 203 specimens of types representing <5% of the total were collected (further details were not provided) and fixed in 10-20% neutral buffered formalin for 24-72 h before paraffin embedding. The specimens were contributed from multiple hospitals, with six hospitals providing >10 specimens. DNA (582 specimens) and RNA (572 specimens) were extracted using the GeneRead DNA FFPE Kit and the RNeasy FFPE Kit, respectively. Double stranded DNA was quantified by Qubit and amplifiable DNA was quantified using the TaqMan Copy Number Reference Assay Human RNase P.  The Q value was then calculated as the percentage of dsDNA that was amplifiable.  NA quantity was defined as the amount of double-stranded DNA when the Q-value was >1 and as the amount of amplifiable DNA when the Q-Value was ≥1. The ddCq value was calculated based on the ratio of two real-time PCR amplicons in the FFPE DNA QC Assay version 2 Kit. The percentage of RNA fragments >200 bp (DV200) was assessed using a 2200 TapeStation. DNA and the RNA libraries were prepared using a SureSelectXT Custom Kit and the SureSelect RNA Capture Kit, respectively, and sequenced on a Next-seq Instrument.

   Summary of Findings:

   The duration of FFPE block storage had a deleterious effect on DNA integrity, with significantly higher ddCq values and lower Q-values noted with progressive storage. ddCq values were higher in specimens stored for >1 year compared to those stored for approximately 1 year (P<0.001, all), stored >3 years compared to 1-2 years (P<0.001, both), or stored approximately 5 years compared to those stored 2-3 or 3-5 years (P<0.001, both). Q-values were lower in specimens that were stored for 5 years than 3-5 years (P<0.01), 2-3 years (P<0.001), 1-2 years (P<0.001), or approximately 1 year (P<0.001), or stored for 3-5 years compared to those stored 1-2 years (P<0.01) or approximately 1 year (P<0.01) but were not significantly different in specimens stored <3 years when compared to those stored for approximately 1 year. ddCq was negatively correlated with the percentage on-target reads (r=-0.47, P<0.001), mean depth (r=-0664, P<0.001), and target exon coverage (r=-0.362, P<0.001); Q-value was correlated with the percentage on-target reads (r=0.353, P<0.001) and coverage uniformity (r=0.411, P<0.001). In a generalized linear model, the percentage of on-target reads was associated with ddCq, but mean depth was associated with ddCq and Q-value. The predictive power of a ddCq ≤5.36 for a mean depth >500 was 91.6%, but the same cut-off was less predictive of coverage uniformity ≥99% (55.1%) or target exon coverage ≥99% (60.6%).  The predictive power of a Q-value ≥0.928 for a mean depth >500, coverage uniformity ≥99%, and target exon coverage ≥99% was 79.9%, 81.5%, and 83.1% respectively.  For specimens with a low ddCq (≤5.36) but unacceptable Q-values, the on-target read percentage and mean depth were still generally acceptable but uniformity and target exon coverage were <99%. ddCq and Q-values displayed significant differences among hospitals even after correcting for tumor type. Q-values were lowest in lung tumors.

    Similarly, DNA and RNA integrity were adversely affected by block storage. The percentage of RNA fragments >200 nt (DV200) decreased progressively with FFPE block storage duration, with significantly lower DV200 values from specimens that were stored >2 years than those stored approximately 1 year (P<0.001 all) or those stored for 3-5 years or approximately 5 years compared to those stored approximately 1 year (P<0.01, both). DV200 was weakly positively correlated with the coverage of housekeeping genes (r=0.369, P<0.001).  The predictive power of a DV200 >41 for a coverage of housekeeping genes ≥70% was 92.1%. DV200 was not significantly different among hospitals but was significantly higher in lung cancer specimens than in bowel or ovarian/fallopian cancer specimens.

   Biospecimens

   • Tissue - Colorectal
   • Tissue - Lung
   • Tissue - Other
   • Tissue - Uterus
   • Tissue - Cervix
   • Tissue - Ovary

   Preservative Types

   • Formalin

   Diagnoses:

   • Neoplastic - Carcinoma
   • Not specified
   • Neoplastic - Not specified

   Platform:

   Analyte	Technology Platform
   DNA	Fluorometry
   DNA	Real-time qPCR
   RNA	Next generation sequencing
   RNA	Automated electrophoresis/Bioanalyzer
   DNA	Next generation sequencing

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Biospecimen Acquisition	Biospecimen location	Lung Bowel Ovarian Uterus Cervix Soft tissue Other
   Biospecimen Acquisition	Locale of biospecimen collection	Different hospitals compared
   Storage	Storage duration	Approximately 1 year 1-2 years 2-3 years 3-5 years Approximately 5 years
   Real-time qPCR Specific	Technology platform	ddCq Q-value

   View more