Abstract
Comprehensive genomic profiling (CGP) is increasingly used as a clinical laboratory test and being applied to cancer treatment; however, standardization and external quality assessments (EQA) have not been fully developed. This study performed cost-effective EQA and proficiency tests (PT) for CGP testing among multiple institutions those belong to the EQA working group of Japan Association for Clinical Laboratory Science (JACLS). This study revealed that preanalytical processes, such as derived nucleic acids (NA) extraction from formalin fixed paraffine embedded (FFPE) samples, are critical. First, EQA with extracted DNA from cell lines showed a detection rate of 100% (9 out of 9) in KRAS (c.38G > A; p.G13D), PIK3CA (p.H1047R), and B-Raf proto-oncogene, serine/threonine kinase (BRAF) (c.1799 T > A; p.V600E) in cases of > 10% variant allele frequency (VAF). However, BRAF (c.1799 T > A; p.V600E) detection decreased to 67% (6 out of 9) for a VAF of 4.9%. Second, when DNA was extracted from FFPE samples, pathogenic variants and variants with companion diagnostic indications were detected in all 10 participating laboratories. Each variant had < 20% VAFs on average (8.1-19.1%) and wide variability among laboratories was observed (relative standard deviation, 13-60%). Nonetheless, BRAF (c.1798_1799delinsAA; p.V600K) of 8.1% VAF, EGFR (c.2235_2249del; p.E746_A750del) of 9.7% VAF, and EGFR (c.2254_2277del; p.S752_I759del) of 9.8% VAF were detected with 70% (7/10), 70% (7/10), and 60% (6/10) frequency, respectively. Therefore, 10% VAF in pre-analytic processing for DNA extraction from FFPE was critical for variant detection in CGP analysis. Further, incorrect results were reported in case independent variant calling of BRAF; c.1798_1799delinsAA (p.V600K) was mistakenly interpreted as c.1798G > A, and c.1799 T > A was on the other strand. In conclusion, the EQA/PT among 10 institutes with common samples revealed the importance of VAF in pre-analysis and helped us understand the significance of the pipeline and common pitfalls usually ignored by the internal quality control in a single institute.