Comparative Analysis of Molecular Testing Methods for FGFR2 Fusion/Rearrangement in Intrahepatic Cholangiocarcinoma

Abstract

Abstract Backgrounds: Targeted therapies for intrahepatic cholangiocarcinoma (ICC) patients with FGFR2 fusion/rearrangement have shown significant effectiveness. Reliable and efficient testing strategies for detecting FGFR2 fusion/rearrangement are crucial following FDA and NMPA approval. Methods: A cohort of 226 cases underwent RNA-based NGS, DNA-based NGS, and break-apart FISH analysis to evaluate FGFR2 fusion/rearrangement. The fusion detection rates, fusion partner identification, and concordance between the methods were analyzed. Results: The positive rates for each method were as follows: RNA-based NGS − 9.8% (22/225), DNA-based NGS − 7.1% (16/226), and break-apart FISH − 10.2% (23/226). The integrated approach increased the prevalence rate of FGFR2 fusion/rearrangement to 11.5% (26/226). Among the 26 positive cases detected by any of the three methods, only 15 cases (57.7%) showed positive results for all three techniques, resulting in an overall concordance rate of 95.1% (215/226). RNA-based NGS played a critical role in confirming the presence of oncogenic FGFR2 fusion in 77% (20/26) of the fusion-positive cases and uncovered five novel putative oncogenic fusions involving FGFR2 (A1CF, ANKS1B, FOXP1, TFEC, and DAAM2). The combination of RNA- and DNA-based NGS detected a total of 29 FGFR2 fusion/rearrangements in 23 out of 226 cases (10.2%), involving 23 partner genes. The most frequent fusion partner was BICC1 (21.7%), while others varied. Relative to the oncogenic FGFR2 fusions identified by the RNA-based NGS assay, DNA-based NGS exhibited a sensitivity of 70.0% and specificity of 99.0%, while FISH showed a sensitivity of 99.0% and specificity of 98.1%. Among the 23 cases with positive FGFR2-FISH results, 18 cases (78%) demonstrated the typical split 3'/5' pattern with a signal distance of ≥ 2, while 5 cases (22%) displayed atypical patterns, including 1 case with short split signals (signal distance < 2) and 4 cases with a single 5' signal. Conclusion: FGFR2 fusion/rearrangement exhibits genetic heterogeneity. Although DNA-based NGS is highly specific and FISH is highly sensitive, the use of RNA-based NGS is crucial for detecting FGFR2 fusion/rearrangement. A combined approach utilizing RNA-based NGS, DNA-based NGS, and FISH assay is recommended to enhance the accuracy of identifying FGFR2 fusions.