Comparison of Molecular Testing Methodologies for CIC-Rearranged Sarcomas.

Abstract

Context.—: Molecular detection of a capicua transcriptional repressor (CIC) rearrangement is critical for diagnosing CIC-rearranged sarcoma (CIC-RS) but is analytically challenging.

Objective.—: To compare the technical performance of fluorescence in situ hybridization (FISH), whole-transcriptome sequencing (RNA-seq), and DNA methylation profiling for CIC-rearrangement detection in a large, mainly pediatric cohort.

Design.—: The study cohort consisted of 44 distinct patient tumors that were positive, equivocal, or suggestive for CIC rearrangement, including 18 central nervous system and 26 extra-central nervous system solid tumors. Forty tumors underwent FISH to detect CIC rearrangement, 31 underwent transcriptome sequencing, and 34 underwent methylation array analysis. Results for tumors tested by multiple testing modalities were compared.

Results.—: Fusions were detected in 27 cases: CIC::double homeobox 4 (DUX4) (n = 15), CIC::NUT midline carcinoma family member 1 (NUTM1) (n = 4), CIC::leucine twenty homeobox (LEUTX) (n = 3), CIC::NUT family member 2B (NUTM2B) (n = 1), ataxin 1 (ATXN1)::NUTM1 (n = 1), ATXN1::NUT family member 2A/B (NUTM2A/B) (n = 1), CIC::DUX4 proximity effect (n = 1), and dedicator of cytokinesis 1 (DOCK1)::DUX4 (n = 1). Twenty-five tumors were tested by all 3 testing modalities. Apparent false-negative rates were 20% (3 of 15) for CIC FISH, 14% (2 of 14) for transcriptome sequencing, and 14% (2 of 14) for methylation array analysis. Both false-negative methylation array results had CIC::LEUTX fusion.

Conclusions.—: Awareness of molecular testing pitfalls in the appropriate detection of CIC rearrangement is critical. Any CIC FISH result may need to be further confirmed, either with unequivocal immunohistochemical support or by another molecular method. A positive RNA-seq or methylation array analysis result may be sufficient evidence for a diagnosis of CIC-RS in the appropriate histologic context. A negative or inconclusive/unclassified RNA-seq or methylation array analysis result in a tumor with high initial suspicion for CIC-RS likely requires careful reevaluation.