Analytical Accuracy of RET Fusion Detection by Break-Apart Fluorescence In Situ Hybridization

Abstract

Context.— RET gene fusions are oncogenic drivers in nonsmall cell lung cancer and nonmedullary thyroid cancer. Selpercatinib (RETEVMO), a targeted inhibitor of RET, was approved by the US Food and Drug Administration for the treatment of RET fusion–positive nonsmall cell lung cancer and nonmedullary thyroid cancer emphasizing the need for rapid and accurate diagnosis of RET fusions. Fluorescence in situ hybridization (FISH) has been used to detect gene rearrangements, but its performance detecting RET rearrangements is understudied. Objective.— To validate and describe the performance of Abbott Molecular RET break-apart FISH probes for detecting RET rearrangements. Design.— A training set with RET fusion–positive (13) and RET fusion–negative nonsmall cell lung cancer and nonmedullary thyroid cancer samples (12) was used to establish criteria for FISH scoring. The scoring criteria was then applied to a larger validation set of samples (96). Results.— A cutoff of 19% or more positive nuclei by FISH was established in the training set and determined by the mean ±3 SD. The validation set was tested using Abbott Molecular RET break-apart FISH compared with sequencing. With this cutoff, a sensitivity of 86% (12 of 14) and specificity of 99% (81 of 82) was achieved. Bootstrapping showed sensitivity could be optimized by using a greater than 13% cutoff with indeterminate samples of 13% to 18% abnormal nuclei requiring confirmation by an orthogonal method. Using this 3-tier scoring system sensitivity increased to 100% (14 of 14) and specificity was 96% (79 of 82). Conclusions.— Abbott Molecular break-apart FISH probes can be used to detect RET fusions. Laboratories can optimize cutoffs and/or testing algorithms to maximize sensitivity and specificity to ensure appropriate patients receive effective, timely therapy.

Clinicopathologic characteristics and diagnostic methods of RET rearrangement in Chinese non-small cell lung cancer patients

“…RT-PCR is specific, but it is limited to known fusion partners and thus may underestimate prevalence. RET FISH is highly sensitive (100%) but has suboptimal specificity (45-60%) (17,18). Moreover, break-apart displays low sensitivity in detecting non-canonical RET fusions (19).…”

Case Report: Complete pathologic response to neoadjuvant selpercatinib in a patient with resectable early-stage RET fusion-positive non-small cell lung cancer

“…To mitigate this challenge, molecular adjudication via FISH was utilized ( 7 ). A negative RET FISH result provided evidence of the lack of residual tumor (at least related to the original clone) and helped determine that the initially perceived “viable tumor cells” by two of three pathologists represented a benign reactive proliferation ( 8 ). The absence of driver alterations suggests a lack of residual tumor since driver alterations are typically retained during the course of cancer evolution ( 9 ).…”

Targeted therapy of RET fusion-positive non-small cell lung cancer

“…FISH works by labeling DNA or RNA probes with specific nucleotide molecules that can find, qualitatively, and quantitatively assess DNA or RNA sequence at the probe’s site ( 22 , 32 ). FISH has the highest sensitivity for the detection of classical RET fusions (KIF5B-RET, CCDC6-RET) at 86%-91.7% and 95%, respectively, while other non-classical fusions such as NCOA4-RET have a sensitivity of only 66.7% ( 33 , 34 ). The FISH specificity has been reported to reach 50.6%-99% ( 33 , 34 ).…”

“…FISH has the highest sensitivity for the detection of classical RET fusions (KIF5B-RET, CCDC6-RET) at 86%-91.7% and 95%, respectively, while other non-classical fusions such as NCOA4-RET have a sensitivity of only 66.7% ( 33 , 34 ). The FISH specificity has been reported to reach 50.6%-99% ( 33 , 34 ). But FISH is associated with demerits like high detection cost, extended detection time, need for professional supervision, and high subjectivity ( 35 ).…”