Whole exome sequencing (WES) on formalin-fixed, paraffin-embedded (FFPE) tumor tissue in gastrointestinal stromal tumors (GIST).
Author(s): Astolfi A, Urbini M, Indio V, Nannini M, Genovese CG, Santini D, Saponara M, Mandrioli A, Ercolani G, Brandi G, Biasco G, Pantaleo MA
Publication: BMC Genomics, 2015, Vol. 16, Page 892
PubMed ID: 26531060 PubMed Review Paper? No
Purpose of Paper
This paper compared DNA integrity and next generation sequencing results from formalin-fixed paraffin embedded (FFPE) and frozen gastrointestinal stromal tumors (GIST).
Conclusion of Paper
DNA isolated from FFPE specimens had lower success rates for amplification of large PCR amplicons (500 bp) and lower Kappa ratios, indicative of DNA fragmentation. For the three FFPE specimens that yielded 500 bp amplicons, the average insert size in the library was slightly shorter than frozen specimens or blood specimens, but the number of reads and exome coverage were comparable. Further, 94-96% of variants detected in frozen specimens were found in the matched FFPE specimen. In contrast when compared to the case-matched frozen specimen, the FFPE specimen with a maximum amplicon size of 300-400 bp library construction required two reactions, had a shorter insert size, exome coverage was lower, and fewer reads were generated. Perhaps most importantly, this low quality FFPE specimen did not allow detection of 47% of the variants identified in the frozen specimen, and 20 of the 33 variants identified in the FFPE specimen were not identified in the matched frozen specimen.
Studies
-
Study Purpose
This study compared DNA integrity and next generation sequencing results from FFPE and frozen GIST. For some specimens a matched blood specimen was used for genotyping. A portion of each tumor was snap-frozen in liquid nitrogen and the remainder was fixed in formalin for 6 to 72 h and paraffin embedded. Using an H&E stained slide as a guide, tumor regions were macrodissected from 2-3 10 µM sections. DNA was extracted from macrodissected FFPE and frozen specimens after an overnight Proteinase K digestion using the QIAamp DNA micro kit. DNA was extracted from blood using the QIAamp DNA mini kit. DNA yield was determined by spectrophotometer and Picogreen. DNA integrity was evaluated by RAPD PCR and the Kapa HgDNA quantification and quality kit. Sequencing libraries were prepared with the Nextera Rapid Capture Exome Enrichment protocol and sequenced at 2 x 100 bp using Illumina Sequencing by synthesis.
Summary of Findings:
DNA quantification methods differed, as Picogreen estimated DNA yield to be approximately half that reported by spectrophotometry.
While it was possible to amplify multiple fragments of 500 bp or longer from frozen specimens, the longest fragment generated from FFPE specimens was 500 bp for 3 of the 4 FFPE specimens and 300-400 bp for the remaining FFPE specimen. While 129 and 41 bp fragments were equally amplifiable in frozen specimens using the Kapa assay, amplification of the 129 bp fragment was 20-50% that observed for the 41 bp fragment in the FFPE specimen with the best quality. Similarly, the ratio of 305 bp to 41 bp fragments was 0.75 in frozen specimens, but 0-0.20 in FFPE specimens.
While libraries of 1.8-2.7 µg of DNA were successfully produced for most specimens, the library generated for the FFPE specimen that had a maximum amplicon length of 300-400 bp was limited to 0.5 µg after pooling two reactions. The average fragment size of the NGS library from frozen tumor and blood specimens was 295 bp while the size range from FFPE specimens was dependent on DNA quality. High quality FFPE specimens (that had a maximum amplicon length of 500 bp) produced libraries with an average fragment size of 230-240 bp while the average fragment size was 190 bp for the low quality FFPE specimen. Frozen, FFPE and blood specimens displayed comparable average Q-scores, percentages of bases with a Q-score greater than or equal to 30, and percentages of PCR duplicates. While an average of 55 and 56 million reads were generated using frozen and high quality FFPE specimens, respectively, only 14 million reads were generated from the low quality FFPE specimen. Further, the average exome coverage was 55-71X and 58-77X for frozen and FFPE tumor specimens, respectively, but only 17X for the low quality FFPE specimen. The majority of variants identified in matched frozen specimens were also identified in high quality FFPE specimens (94-96%), while only 53% of these variants were identified in low quality FFPE specimens. While only 0-1 false positive variant calls (variants found in FFPE but not frozen specimens) were found in the three high quality FFPE specimens, 20 of the 33 variants found in low quality FFPE specimens were not identified in the matched frozen specimen. When investigated by nucleotide type it was found the low quality FFPE specimen had more C>T/G>A, C>A/G>T and A>G/T>C substitutions. Importantly, high quality FFPE specimens did not differ from frozen specimens in the frequency of individual nucleotide substitutions.
Biospecimens
Preservative Types
- Formalin
- Frozen
Diagnoses:
- Neoplastic - Sarcoma
Platform:
Analyte Technology Platform DNA Fluorometry DNA Next generation sequencing DNA PCR DNA Spectrophotometry Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Biospecimen Preservation Type of fixation/preservation Formalin (buffered)
Frozen
PCR Specific Length of gene fragment 41 bp
129 bp
Spectrophotometry Specific Technology platform Picogreen