Skip to main content

Effective quality management practices in routine clinical next-generation sequencing.

de Abreu FB, Peterson JD, Amos CI, Wells WA, Tsongalis GJ

Publication: Clin Chem Lab Med (2016)PMID: 26872315View on PubMed

Purpose

This paper investigated whether quality metrics can predict next generation sequencing success using formalin-fixed paraffin embedded (FFPE) cytology, biopsy and resection specimens.  

Conclusion

Of the 879 specimens for which library construction was successful, only 2 failed to meet sequencing quality metrics despite inclusion of specimens with low DNA yield (<1.7 ng/ μL) or integrity (Kappa Q129/Q41 ratio of < 0.4). Cytology and biopsy specimens accounted for 75% of the 126 specimens that failed library construction, and cytology specimens accounted for 82% of the specimens with low DNA yield. It is unclear what percentage of specimens that failed library construction were also of low yield. Of the 39 specimens that failed the DNA integrity assay, only 11 (28%) were unsuccessful in library construction; therefore the authors conclude that Q129/Q41 is not a reliable predictor of library success for FFPE specimens. 

Studies (1)

Purpose

This study investigated whether quality metrics can predict next generation sequencing success using FFPE cytology, biopsy and resection specimens. This study included FFPE blocks of 408 non-small-cell lung carcinomas (NSCLCs), 304 colon adenocarcinomas, 120 glioma/glioblastomas, 95 melanomas, 27 breast carcinomas, and 74 non-specified tumors of other types. Specimens included 397 biopsy, 195 cytology (FNA, pleural fluid or alveolar lavage) and 421 resection specimens. The tumor area was determined by H&E microscopy. DNA was extracted from microdissected areas from at least 8 unstained sections of specimens containing at least 10% tumor content using the Gentra Puregene Kit. DNA was quantified using PicoGreen and DNA integrity was assayed in 243 specimens using the real-time PCR based KAPA hgDNA system. NGS library construction was performed using Ion AmpliSeq Cancer Hotspot Panel v2. 

Summary of Findings

Only 23 of the 1028 specimens did not contain sufficient tumor content to permit DNA extraction. The remaining 1005 specimens underwent library construction, of which 126 (12.5%) failed to produce sufficient yield (>100 pM). Ninety-two of these 126 (75%) were biopsy or cytology specimens.  Similarly, 82% of the specimens that had low DNA yield (<1.7 ng/μL) also were cytology specimens, but it is not clear if these specimens also failed library production. Of the 39 specimens that did not have a Q129/Q41 ratio greater or equal to 0.4, only 11 (28%) failed library construction, indicating that Q129/Q41 is not a reliable predictor of library success. Of the 879 specimens for which a library was constructed, only 2 (0.02%) failed to meet sequencing quality criteria (For the run: >70% chip loading, >55% usable sequences, <35% polyclonality and <20% low quality reads; For the specimen : >90% on-target reads, >90% coverage uniformity, and ≥95% amplicons with 500x coverage; For the variants: coverage ≥ 500x, ≥ 5% allelic frequency). Of the 877 remaining specimens, 127 (14.5%)  were found to be wildtype, leaving the remaining 750 specimens (85.5%) positive for one (40%) or more (60%) of the variants.

  • Biospecimens

    FluidPleural Fluid
    FluidBronchial Lavage
    TissueLung
    TissueColorectal
    TissueSkin
    TissueBreast
    TissueNerve
  • Preservation Types

    Formalin
  • Diagnoses

    NeoplasticCarcinoma
    NeoplasticMelanoma
    NeoplasticNot specified
    NeoplasticOther
  • Platforms

    DNAFluorometry
    DNAReal-time qPCR
    DNANext generation sequencing
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen AcquisitionMethod of tissue acquisition
    Biopsy
    Fine needle aspiration
    Surgical resection
    Next generation sequencing SpecificTechnology platform
    Kappa ratio
    PicoGreen