Detection of rare somatic mutational profiles in metastatic colorectal cancer (mCRC) during routine RAS sequencing using next generation sequencing (NGS)
Résumé
In most patients with mCRC who are being considered for anti-EGFR antibody therapy, RAS mutation testing i.e. KRAS and NRAS exon 2, 3 and 4, is routinely assessed using PCR-based assays only detecting major hotspot mutations of exon (ex) 2 (codon 12 and 13), 3 (codon 59 and 61) and 4 (codon 117 and 146). We performed deep sequencing of the entire exons using NGS as an alternative to detect additional rare mutations profiles with significant frequency of mutated allele (FMA). Methods: 188 formalin-fixed paraffin-embedded tumor samples from primary or metastatic lesions of patients (M/F sex ratio 1.27, mean age 69 years, range 32-90) with mCRC (150 colon, 38 rectum) were analyzed. DNA was extracted from macrodissected slides (mean tumor cell content 43.3%, range 5-80). Results: RAS mutation testing was routinely assessed using NGS in 177 mCRC samples. NGS could not be performed in 11 cases (6.2%) due to the insufficient quantity or quality of DNA. NGS sensitivity was 1% at X1000 depth. RAS mutations were found in 103 samples (62%) and relatively distributed as 69.9% KRAS ex2, 3.9% KRAS ex3, 14.6% KRAS ex4, 4.8% NRAS ex2, 1.0% NRAS ex3, 1.0% NRAS ex4 and 4.8% multiple mutations. Uncommon mutational profiles were detected in 10 cases (9.7%): 2 KRAS ex2 c.37G > T p.G13C single mutation with FMA > 30%, 5 silent mutations (4 with FMA > 25%), alone (n = 2) or combined with other rare mutations (n = 3) with lower but significant FMA ( > 1%), and 6 multiple mutation profiles among which 2 double hotspot mutation (KRAS ex2 c.34G > A p.G12S and NRAS ex3 c.181C > A p.Q61K, KRAS ex2 c.34G > A p.G12S and NRAS ex2 c.38G > T p.G13V), 1 secondary rare mutation associated with a KRAS ex2 c.35G > A p.G12D hotspot mutation, and 3 multiple mutations only with rare but potentially deleterious mutations located around the loops responsible for nucleotide (GTP) binding. In only 1 case, the FMA of the secondary mutations was < 1%. As a whole, 7 cases (6.8%) had RAS mutations out of hotspots. Conclusions: NGS proved accurate, sensitive and suitable for routine RAS genotyping in mCRC. It can detect uncommon RAS mutation profiles with significant FMA that can potentially impair the patient response to anti-EGFR antibody.