Virtual Library
Start Your Search
A. Van Bokoven
Author of
-
+
MO01 - Lung Cancer Biology - Techniques and Platforms (ID 90)
- Event: WCLC 2013
- Type: Mini Oral Abstract Session
- Track: Biology
- Presentations: 1
- Moderators:S. Lu, P. Waring
- Coordinates: 10/28/2013, 10:30 - 12:00, Bayside 204 A+B, Level 2
-
+
MO01.03 - Transitioning To Next Gen Testing Of Lung Carcinoma (ID 2891)
10:35 - 10:40 | Author(s): A. Van Bokoven
- Abstract
Background
The feasibility of multigene testing in a clinical setting has been demonstrated by the Lung Cancer Mutation Consortium (LCMC) which has evaluated over 1000 cases from multiple institutions in a CLIA environment. The initial platforms used by the LCMC were SNaPshot and Ion Torrent, allele specific tests. More recently the sequencing by synthesis method (Illumina) used for whole genome sequencing has been scaled for sequencing of a limited number of targeted genes. In this study we compare the performance characteristics of Next Generation testing on the MiSeq platform with the older allele specific SNaPshot platform and evaluate the applicability of Miseq-based testing to a clinical, CLIA regulated setting.Methods
Two Illumina kits, the TruSeq and TruSight evaluating 221 hotspots in 48 gene and 175 exons in 26 genes, respectively, were compared. To assess analytical sensitivity, cell lines with known mutations and SNPs were titered into liver DNA known to be wild-type for the selected mutations, at tumor cell concentrations ranging from 3% to 50%. In addition, 24 formalin-fixed paraffin-embedded lung tumors that had previously been evaluated by SNaPshot or direct sequencing were tested to compare sensitivities and specificities of methods. Paraffin embedded human tumor tissue samples were enriched for tumor cells by coring of paraffin block or macrodissection using a pneumatic cell collector. DNA was extracted by proteinase K digestion and column chromatography, end repaired and phosphorylated. Libraries were prepared from each sample by ligating index adapters that allow for mixing of samples and binding adapters that link DNA fragments to flow cell. Combined libraries were added to flow cells at an appropriate concentration, clusters generated, and sequencing reaction commenced. Results were evaluated by software developed by Illumina or locally at the University of Colorado.Results
Spiking studies indicated that analytic sensitivity for Miseq at loading quantities of 100 to 300 ng (TruSeq) was ~5% for known KRAS and TP53 mutations and several synonymous polymorphisms in other covered genes, comparable to SNaPshot. For clinical samples, average depth of coverage was 5700 (+/- 2267). Unfiltered results using Illumina software supplied with the Miseq instrument showed an average of 88 heterozygous SNPs, 12 insertions and 17 deletions (uncurated for relevance). All of the mutations that were previously found by SNaPshot were also detected by Miseq TruSeq and TruSight protocols (100% concordance). Variants representing known polymorphisms, synonymous changes and variants identified in the context of low coverage were excluded. Data analysis using locally developed software indicated the presence of 1-9 SNPs in each sample that were not predicted by SNaPshot testing, attributable to the wider coverage of the Miseq platforms. None of the additional mutations represented treatable targets with currently available drugs.Conclusion
Next-generation testing is feasible in a CLIA environment using the Miseq platform. However, rigorous software validation is necessary before this platform can be adopted by a busy clinical laboratory. Software limitations currently being addressed include long turnaround time, inadequate vetting of new and recurrent SNPs for clinical significance and limited software development resources.
