The implications of abnormal signal patterns of break-apart FISH probes used in the diagnosis of bone and soft tissue tumours

Abstract

Many subtypes of bone and soft tissue tumours harbour specific chromosome translocations leading to chimeric fusion genes. The identification of these specific fusion genes is the basis of molecular diagnoses in such tumours. Break-apart FISH is a robust method that is commonly used to identify these translocations and provide diagnostic support to histological interpretations. The signal patterns of the break-apart probes are usually easily interpreted. However, some cases show abnormal signal patterns leading to equivocal and challenging interpretation. The incidence of these abnormal patterns is largely unknown. Using a retrospective cohort we explored the incidence of abnormal signal patterns across common bone and soft tissue tumour types to raise awareness of this occurrence and to aid in the interpretation. In total, 1,087 bone and soft tissue tumours tested by break-apart probes were examined. The abnormal signal patterns were classified as deletion, additional copy and amplification, which were found at highest frequency in low-grade fibromyxoid sarcoma (32%, 6/19), and at moderate frequencies in those from alveolar rhabdomyosarcoma (10%, 9/94), nodular fasciitis (9%, 18/209), synovial sarcoma (8%, 17/207) and Ewing sarcoma/round cell sarcoma with EWSR1-non-ETS fusions (6%, 29/497). The lowest frequency was found in clear cell sarcoma (1%, 1/61). Despite the equivocal results from the abnormal signal patterns, the specific fusion genes were confirmed by orthogonal molecular techniques such as FISH with fusion probes, RT-PCR or next-generation sequencing.

Keywords: FISH; abnormal signal pattern; bone and soft tissue tumour; break-apart probe; cancer; gene rearrangement; genomics; translocation.

FIGURE 1

Usual signal pattern of break-apart probe. (A) Negative signal pattern of two yellow signals. (B) Positive signal pattern of one yellow and a break-apart (one green separated from one red). (C) Negative: Three to 4 copies of yellow signals. (D) Negative: Multi-copies of yellow signals.

FIGURE 2

An alveolar rhabdomyosarcoma with FOXO1 gene amplification. (A) Round cells with pseudoalveolar pattern. (B) FOXO1 break-apart probe shows positive signal pattern with amplification of 3′ end of the FOXO1 locus (Green). (C) Multi fusion yellow signals of PAX7::FOXO1 by FISH of PAX7::FOXO1 fusion probe. (D) No fusion signal found by FISH of PAX3::FOXO1 fusion probe.

FIGURE 3

A NFATC2-rearranged sarcoma with EWSR1 gene amplification. (A) Undifferentiated blue small round cells composed of cords of cells in a fibrous stroma. (B) EWSR1 break-apart probe shows atypical signal pattern: amplification of 5′ end of the EWSR1 locus. (C) NFATC2 break-apart probe confirmed NFATC2 gene rearranged with amplification of 3′ end of the NFATC2 locus. EWSR1::NFATC2 fusion gene detected by NGS.

FIGURE 4

A synovial sarcoma with deletion of 3′ end of SS18 locus. (A) Monophasic-type SS. (B) SS18 break-apart probe shows atypical signal pattern of one yellow and one red signals: deletion of 3′ end of the SS18 locus. (C) SS18::SSX2 fusion gene detected by RT-PCR and NGS.

FIGURE 5

An Ewing sarcoma with no break-apart and an extra copy of EWSR1 gene locus. (A) Monomorphic small blue round cell tumour. (B) EWSR1 break-apart probe shows atypical signal pattern of two yellow and one red signals: one extra copy of the 5′ end of the EWSR1 locus. (C) EWSR1::FLI1 type I fusion gene detected by RT-PCR and NGS.

FIGURE 6

A clear cell sarcoma with extra copies of 5′ end of EWSR1 locus. (A) Polygonal cells with vesicular nuclei and cytoplasmic clearing forming sheets and nests. (B) EWSR1 break-apart probe shows atypical signal pattern of two yellow and two red signals: extra copy of 5′ end of the EWSR1 locus. (C) Circos plot of EWSR1::CREM fusion gene detected by WGS.