Gene Fusion Testing: FISH, Immunohistochemistry, DNA and RNA Sequencing

Fluorescence In-Situ Hybridisation

Break-apart FISH uses two coloured probes flanking a gene. When the gene is intact the colours overlap; when it is rearranged they separate. It detects that a break has occurred without identifying the partner.

FISH can miss small intrachromosomal rearrangements where the probes barely separate, and it requires interpretation of a set number of nuclei against a cut-off.

Immunohistochemistry

Immunohistochemistry detects abnormal expression of the fusion protein, for example ALK protein that is normally absent in lung tissue. It is fast, cheap and works on small samples, making it a good screen.

Specificity varies by target: it is high for ALK but lower for ROS1 and NTRK, so positive results for those often need confirmation.

DNA-Based Sequencing

DNA next-generation sequencing can detect fusions if it covers the intronic breakpoint regions. The difficulty is that breakpoints often fall in very large or repetitive introns that are hard to tile with probes, so some fusions are missed.

DNA panels are good at point mutations and copy number, so a combined approach is common.

RNA-Based Sequencing

RNA sequencing reads the spliced transcript, where the fusion junction is a single clean exon-to-exon boundary regardless of where in the introns the DNA break occurred. This makes it well suited to detecting fusions and identifying the partner and variant.

It requires good-quality RNA, which can be limited in older or poorly preserved samples, and proficiency-testing data show performance varies between laboratories.

Interpretation Notes

A negative fusion result should be read against the method used: a DNA panel with limited intronic coverage, or immunohistochemistry alone, does not exclude a fusion as confidently as RNA sequencing.

When a targetable fusion is strongly expected but not found, testing by an orthogonal method is reasonable.

Key Takeaways

• ·FISH detects that a break occurred but not the partner; small rearrangements can be missed.
• ·Immunohistochemistry is a good screen, with variable specificity by target.
• ·DNA sequencing can miss fusions with breakpoints in large introns.
• ·RNA sequencing reads the clean fusion junction and is often the preferred confirmatory method.