Introduction: A previously developed NGS-based assay (ENGAUGE™-Cancer-DLBCL), which consists of the genes BRAF, BTK, CARD11, CD79A, CD79B, EZH2, KRAS, MYD88, NOTCH1, PIM1, SYK, TNFAIP3 and TP53 was analytically validated for clinical use in patients with Diffuse large B-cell lymphoma. We analyzed the performance of this assay with respect to the reproducibility, accuracy, precision, linearity, as well as limits of detection and quantification in a clinical reference laboratory environment in tumor sections taken from formalin-fixed paraffin-embedded (FFPE) tissue blocks.
Methods: Illumina TrueSeq® Custom Amplicon on MiSeq was used for this study. The basic design of the experiments comprised the evaluation of 26 samples in replicate starting from genomic DNA isolated from a pooled sample in which each replicate underwent independent library preparation and was analyzed in a multiplexed run in the same flow cell. We repeated the analysis of the same 26 samples three times on different days (run-to-run reproducibility). All reactions were performed in triplicate with both positive and negative controls, and detection thresholds for each gene were chosen by receiver-operator curve analysis. The sequence data analysis was performed with NextGene, Variant Studio and Diagnovus’s proprietary bioinformatics analysis software. Unsupervised clustering and component analysis were performed to examine the variance and correlation of gene profiles amongst patients and sections within FFPE blocks. Sensitivity and Specificity were also calculated.
Results: In all the sequencing runs performed in the validation studies, the average library coverage from clinical FFPE samples was 1847. The allele frequency for 22 out of 23 (96%) previously identified mutations was concordant with those observed by Sanger sequencing. With respect to limits of detection, the NGS assay can detect mutations present at 5-6% allele frequency with at least 95% confidence. Extraction of total DNA yielded high-quality DNA ranging from 250 to 500 ng per sample. Unsupervised clustering and principal component analysis revealed a strong correlation between blocks from the same patient, as well as between tumor samples from the same block. Using only 250ng total dsDNA input, the ENGAUGE™-Cancer-DLBCL NGS assay is capable of detecting all mutations including single base substitutions and small indels, with 5% LOD, ≥95% intra‐run reproducibility (precision), 100% inter‐run reproducibility, and with ≥95% analytical sensitivity and specificity. 17 critical variants with frequencies ranging from 5% to 100% (10 SBS, 4 deletions, and 3 deletions with insertions) were detected in 26 tumor FFPE samples. The most frequently observed mutations were NOTCH1 (27%), CARD11 (19%) and TP53 (19%).
Conclusions: Extensive reagent and analytical performance studies were conducted to evaluate the reliability and reproducibility of a gene test panel (ENGAUGE™-Cancer-DLBCL). Our validation study demonstrates that this Illumina TrueSeq® Amplicon panel combined with Diagnovus’s proprietary bioinformatics analysis software provides a highly sensitive and sensitive test with a reportable analytical range of 5% to 100% mutation allele frequency.
This assay is suitable for screening diffuse large B-cell patient FFPE tumor specimens for a spectrum of clinically relevant somatic mutations.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.