Skip to main content

Practical guide for the comparison of two next-generation sequencing systems for solid tumour analysis in a universal healthcare system.

Maxwell P, Hynes SO, Fuchs M, Craig S, McGready C, McLean F, McQuaid S, James J, Salto-Tellez M

Publication: J Clin Pathol (2019) • Vol 72 • Page 225-231PMID: 29386326View on PubMed

Purpose

This paper compared variant detection using different gene panels and investigated the reproducibility and analytical sensitivity of the Oncomine panel. The minimum DNA load of the Oncomine panel was also investigated.

Conclusion

The percentage of amplicon targets with ≥500 X reads was slightly higher for the Actionable panel than the Oncomine panel. Of the 44 variants identified using the Actionable panel, 30 were also identified using the Oncomine and Hotspot panels, 11 were identified only with the Oncomine panel, and three were unique to the Actionable panel. Of the 146 variants identified using the Oncomine panel, 102 were also identified using the Cancer hotspot panel and three were unique to the Oncomine panel. Results were strongly correlated between all of the panels. The Oncomine panel showed good reproducibility between separate chips and runs. The variant frequency for the Oncomine panel declined and number of background variants decreased with decreasing input below 4 ng and the overall sensitivity was 5%.

Studies (1)

Purpose

This study compared variant detection using different gene panels and investigated the reproducibility and analytical sensitivity of the Oncomine panel. The minimum DNA load of the Oncomine panel was also investigated. This study used residual FFPE lower gastrointestinal adenocarcinoma (n=17), lung adenocarcinoma (n=2), and malignant melanoma (n=9) specimens and a cell line. No details of specimen procurement, fixation, processing, or storage were provided. DNA was extracted using Maxwell 16, Roche Sample Prep Kit, or DNEasy kits following the manufacturer’s instructions (no information as to which kit for which specimen). NGS libraries were constructed using the Ampliseq Cancer Hotspot Panel v2, the Human Tumour Actionable Mutations Panel, and the Oncomine panel using 21 (Ampliseq and Actionable) or 19 (Oncomine) PCR cycles for initial target enrichment and seven (Ampliseq and Oncomine) or four PCR cycles for library amplification. Libraries were quantified by bioanalyzer. Templates for the Ampliseq and Oncomine panels were prepared using the OT2 200 Kit or the Hi-Q OT2 Kit and for the Actionable panel using MiSeq Reagent Kit v2. All data was processed using the same bioinformatics pipeline. Results were compared with reference data produced by quantitative PCR; Sanger sequencing of variants in IDH1 and 2, BRAF, KRAS, NRAS, and PIK3CA; and NGS analysis using the Ampliseq Cancer Hotspot Panel v2. The analytical sensitivity of the Oncomine panel was determined by dilution of variant samples into wildtype. The reproducibility of the Oncomine panel was tested by repeating the runs on the same chip and on separate chips. The effect of input on variant detection was evaluated by using various starting amounts of DNA starting with (between 0.2 ng and 34 ng).

Summary of Findings

The percentage of amplicon targets with ≥500 X reads was slightly higher for the Actionable panel than the Oncomine panel (>99% versus 98.2). A total of 44, 146, and 199 variants were identified using the Actionable, Oncomine, and Hotspot panels, respectively. Of the 44 variants identified using the Actionable panel, 30 were also identified using the Oncomine and Hotshot panels, 11 were identified only with the Oncomine panel, and three were unique to the Actionable panel. Of the 146 variants identified using the Oncomine panel, 102 were also identified using the Cancer Hotspot panel and three were unique to the Oncomine panel. Results were strongly correlated between the Oncomine and the Hotspot panels (R=0.975, P<0.001), the Oncomine and the Actionable panels (R=0.977, P<0.001), and between the Actionable and Hotspot panels (R=0.974, P<0.001). Importantly, the Oncomine panel showed good reproducibility between separate chips and runs. The variant frequency declined and number of background variants decreased with decreasing input below 4 ng. The overall sensitivity of the Oncomine panel was 5%.

  • Biospecimens

    TissueSmall Bowel
    TissueLung
    TissueSkin
  • Preservation Types

    Formalin
  • Diagnoses

    NeoplasticMelanoma
    NeoplasticCarcinoma
  • Platforms

    DNANext generation sequencing
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Next generation sequencing SpecificTechnology platform
    Ampliseq Cancer Hotspot Panel v2
    Human Tumour Actionable Mutations Panel
    Oncomine Panel
    Next generation sequencing SpecificTemplate/input amount
    0 ng
    1 ng
    17 ng
    2 ng
    34 ng
    4 ng
    9 ng
    Diluted to contain 5-30% mutant DNA