Rapid and Cost-Efficient Detection of <i>RET</i> Rearrangements in a Large Consecutive Series of Lung Carcinomas

Abstract

RET-kinase-activating gene rearrangements occur in approximately 1–2% of non-small-cell lung carcinomas (NSCLCs). Their reliable detection requires next-generation sequencing (NGS), while conventional methods, such as immunohistochemistry (IHC), fluorescence in situ hybridization (FISH) or variant-specific PCR, have significant limitations. We developed an assay that compares the level of RNA transcripts corresponding to 5′- and 3′-end portions of the RET gene; this test relies on the fact that RET translocations result in the upregulation of the kinase domain of the gene and, therefore, the 5′/3′-end expression imbalance. The present study included 16,106 consecutive NSCLC patients, 14,449 (89.7%) of whom passed cDNA quality control. The 5′/3′-end unbalanced RET expression was observed in 184 (1.3%) tumors, 169 of which had a sufficient amount of material for the identification of translocation variants. Variant-specific PCR revealed RET rearrangements in 155/169 (91.7%) tumors. RNA quality was sufficient for RNA-based NGS in 10 cases, 8 of which carried exceptionally rare or novel (HOOK1::RET and ZC3H7A::RET) RET translocations. We also applied variant-specific PCR for eight common RET rearrangements in 4680 tumors, which emerged negative upon the 5′/3′-end unbalanced expression test; 33 (0.7%) of these NSCLCs showed RET fusion. While the combination of the analysis of 5′/3′-end RET expression imbalance and variant-specific PCR allowed identification of RET translocations in approximately 2% of consecutive NSCLCs, this estimate approached 120/2361 (5.1%) in EGFR/KRAS/ALK/ROS1/BRAF/MET-negative carcinomas. RET-rearranged tumors obtained from females, but not males, had a decreased level of expression of thymidylate synthase (p RET testing in facilities that do not have access to NGS due to cost or technical limitations.

Keywords

• NSCLC
• fusion
• tyrosine kinases
• <i>RET</i>
• PCR
• NGS