ABSTRACT
KIF5B-RET fusions have recently been reported to occur in pulmonary adenocarcinomas, thereby being proposed as a novel genetic alteration in adenocarcinoma of the lung. However, clinically useful methods to detect RET-rearrangement in pulmonary adenocarcinoma have not been well established. 53 cases of lung adenocarcinomas harbored "triple (EGFR, KRAS and ALK)-negative" were tested for KIF5B-RET fusions using whole-transcriptome sequencing, fluorescence in situ hybridization (FISH), immunohistochemistry (IHC), and long-range PCR. Dual color break-apart probes and KIF5B-RET fusion probes were used for FISH. Three different commercial antibodies against C-terminal RET protein were tested for IHC. Primers designed for 3 different variants of KIF5B-RET fusions were used for long-range PCR. Three patients (5.6%) showed RET rearrangement in whole-transcriptome sequencing, which were used as a gold standard. All those three patients were also positive in FISH for both KIF5B-RET fusion and RET break-apart probes. None of remaining patients showed positive result, resulting in 100% concordance rate of FISH and transcriptome sequencing methods. However, fused RET proteins were not detected by IHC in none of true positive patients. Moreover, 6 patients without RET fusions showed gain of gene copy number of both KIF5B and RET. All those three true positive cases were detected by long-range PCR methods and none with true negative cases were positive. Both FISH and PCR may be useful methods to detect novel KIF5B-RET rearrangements in pulmonary adenocarcinomas rather than IHC. However, as there may be additional variant of fusion mutation, FISH may be better than PCR method in terms of sensitivity.
