Abstract 748

NCL, AA, AK & AS contributed equally to the study

A large number of acquired mutations have been identified in haematological malignancies in recent years. An increasing number of targets present a new problem for diagnostic molecular pathology laboratories. It is not practical for a laboratory to design increasing numbers of gene specific assays for these targets. For these reasons a single assay capable of detecting a large number of gene mutations is desirable. Here we describe a next generation sequencing approach using the Roche 454 platform which is capable of the simultaneous mutation analysis of 17 genes at an appropriate sensitivity level in multiple patient samples. This methodology allows mutation scanning of all coding exons of 5 complete genes (TP53, EZH2, DNMT3a, RUNX1 and CEBPa) scanning of the mutational hotspots of 5 genes (ASXL1 exon 12, JAK2 exon 12, FLT3, MPL and CBL) and the analysis of specific mutation hotspots in a further 8 genes (NRAS, KRAS, NPM1, IDH1, IDH2, KIT, BRAF, JAK2 exon 14). In order to cover all these regions, a total of 99 PCR amplicons (average length 300bp range 250–400bp) were amplified for each patient sample. To avoid time consuming quantification and normalisation of individual amplicons downstream of PCR, we developed a simple method to first designate the amplicons into 4 groups based on the efficiency of PCR amplification. Following amplification the 1st round PCR products are diluted according to the grouping assigned to the particular amplicon prior to a second round of PCR in which MIDs or “barcodes” are added in a patient specific manner. The products of the 2nd round PCRs are then pooled for each patient and normalised before pooling the entire library for subsequent sequencing. The sequencing is carried out using the Roche GS FLX titanium reagents.

To validate this approach we have prepared libraries from 80 AML patients with normal cytogenetics. Sequencing of these patients was split over 3 runs of the FLX instrument using standard conditions recommended by the manufacturer. The average number of reads per amplicon was 300 with 95% of amplicons having a minimum coverage of 200 reads. The approach allows the detection of mutations at a sensitivity of approximately 5%. Mutations were detected in 14 /17 genes in at least one patient sample. No mutations of BRAF, JAK2 or MPL were detected. In line with previously published data, mutations were found frequently in NPM1 (67%), DNMT3a (55%) FLT3 ITD (44%) and less frequently in the remaining genes. These data correlate well with published data for normal karyotype AML. In 6/17 genes the mutation status of the patient samples had been previously analysed using: Fragment analysis (NPM1 and FLT3 ITD), RFLP (FLT3 TKD), Sanger Sequencing (IDH1, IDH2, ASXL1, TP53). Significantly, 12 additional mutations were detected (NPM1 n=2, FLT3 ITD n=2, FLT3 TKD n=4, IDH1 n=2, IDH2 n=2, ASXL1 n=4) using the new methodology demonstrating an improvement in sensitivity over and above individual gene specific assays. A single run of the FLX instrument enables the analysis of 30 patients in a batch. In our hands library preparation, sequencing and analysis of 30 patients using the workflow described takes a skilled operator 10 full days. With an estimated test price of \P650 ($1000), this compares extremely favourably with the price of providing these analyses separately estimated to be in excess of \P4000 ($6500).

To conclude we have developed a next generation amplicon sequencing approach to assay 17 individual target genes including whole gene coding sequences and more focused mutational hotspots. In addition we have validated a strategy which dramatically reduces the amount of operator time required compared with most amplicon sequencing approaches. This approach offers a highly flexible platform for analysing multiple gene targets in multiple samples. Changes to the targets analysed or the effective sensitivity can easily be incorporated without the need to make major modifications to the procedure.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution