Detection of ALK and ROS1 fusion transcripts in FFPE samples of non-small cell lung cancer patients using a novel RT-PCR based assay and targeted RNA sequencing.

e21695 Background: Detection of genomic rearrangements like ALK and ROS1 fusions are mandatory in non-small cell lung cancer (NSCLC) as those alterations can be targeted by an increasing number of drugs. Fluorescence in-situ hybridization (FISH) has been so far the gold standard but multiplexed technologies would allow analysis of many potential targets simultaneously. We have evaluated a novel RT-PCR based assay and a targeted RNA sequencing solution for the detection of ALK and ROS1 rearrangement in NSCLC patients. Methods: 41 patients with late stage NSCLC were included in the study. ALK and ROS1 status were screened on FFPE tissue sections using immunohistochemistry and confirmed with FISH. Total nucleic acids were extracted from tissue sections using the Maxwell RSC RNA FFPE kit (Promega). Detection of rearrangements was performed by RT-PCR using the prototype ALK/RET/ROS1 Fusion Panel assay (Roche), which detects 7 ALK fusion variants, 13 ROS1 variants and 6 RET variants. For RNA sequencing, the libraries were prepared with the QIAseq Targeted RNAscan Lung Cancer Panel (Qiagen). Sequencing was performed on a MiSeq instrument (Illumina). Results: We tested 16 ALK FISH+ tumors, 4 ROS1 FISH+ tumors, and 21 samples that were either IHC- or IHC+/FISH-. The RT-PCR assay detected 15 of the 16 ALK FISH+ tumors, and 3 of the 4 ROS1 FISH+ tumors. In the 21 IHC- or IHC+/FISH- samples, 1 additional ROS1 fusion and 1 RET fusion were detected. The remaining 19 assays samples were negative. 2 fusions were missed by RT-PCR because the assay design does not cover these rearrangements (KLC1-ALK and CTNND1-ROS1). With targeted RNA sequencing, we were able to detect 14 of the 16 ALK FISH+ tumors and all 4 of the ROS1 FISH+ tumors. We also identified 1 additional ROS1 fusion and 1 RET fusion in the 21 IHC- or IHC+/FISH- samples. 17 patients were negative and 2 tests were not contributive. Overall, using FISH as a reference, the sensitivity of both assays was 90% (18/20). Conclusions: Using limited amount of biological material, both RT-PCR and targeted RNA sequencing were able in one assay to detect the overwhelming majority of ALK and ROS1 fusions identified with FISH. These approaches also allowed us to detect gene rearrangements that were missed by the initial IHC/FISH screening. The RT-PCR assay requires less RNA (as little as 3 ng) and has a much faster turn-around time. The RNA sequencing approach allowed us to identify the fusion partners, but it requires much more RNA and is rather difficult to implement in routine practice.