Development of a semi-conductor sequencing-based panel for genotyping of colon and lung cancer by the Onconetwork consortium

Abstract

Background: The number of predictive biomarkers that will be necessary to assess in clinical practice will increase with the availability of drugs that target specific molecular alterations. Therefore, diagnostic laboratories are confronted with new challenges: costs, turn-around-time and the amount of material required for testing will increase with the number of tests performed on a sample. Our consortium of European clinical research laboratories set out to test if semi-conductor sequencing provides a solution for these challenges.

Methods: We designed a multiplex PCR targeting 87 hotspot regions in 22 genes that are of clinical interest for lung and/or colorectal cancer. The gene-panel was tested by 7 different labs in their own clinical setting using ion-semiconductor sequencing.

Results: We analyzed 155 samples containing 112 previously identified mutations in the KRAS, EGFR en BRAF genes. Only 1 sample failed analysis due to poor quality of the DNA. All other samples were correctly genotyped for the known mutations, even as low as 2%, but also revealed other mutations. Optimization of the primers used in the multiplex PCR resulted in a uniform coverage distribution over the amplicons that allows for efficient pooling of samples in a sequencing run.

Conclusions: We show that a semi-conductor based sequencing approach to stratify colon and lung cancer patients is feasible in a clinical setting.

Figure 1

Schematic scheme of the 3 phases to assess the performance of the gene-panel.

Figure 2

Schematic representation of the results of phase 3 of the panel validation. Indicated are the variants per sample identified in the 28 CRC (A) and 59 NSCLC (B) samples. Variant frequencies were extrapolated to 100% neoplastic cells and/or related to other variants present in the sample (intra-sample comparison). Variants in black correspond to ‘driver’ mutations (present in >35% of neoplastic alleles), while variants in grey are present in only a minority of the neoplastic cells (<35% of neoplastic alleles or an allele frequency 0.5x that of the driver mutation present in the sample). Genes containing >1 variant in different frequencies are indicated in grey/black boxes.

Figure 3

Average coverage distribution over the individual amplicons. The coverage distribution plot over the individual amplicons in gene-panel v1 (A) and v2 (B). For panel v1 5 samples are pooled per 316 chip; with panel v2 8 samples are pooled per 316 chip (B). Note the logarithmic scale of the y-axis.