Design: Diagnostic accuracy study against RNA-NGS and FISH
Population: 79 pigmented spindle cell naevi of Reed
Primary outcome: Agreement of pan-TRK IHC with molecular NTRK3 detection
Effect: Specificity and PPV 92–100%; sensitivity 33–36%

Italian investigators reclassified 79 pigmented spindle cell naevi of Reed under 2025 WHO criteria and stained them with an optimised pan-TRK protocol using extra amplification and a red chromogen. RNA-NGS was available in 59 and NTRK3 FISH in 51.

RNA-NGS found MYO5A::NTRK3 fusion in 56% and FISH found NTRK3 rearrangement in 65%. Pan-TRK was positive in only 20%, always with diffuse granular cytoplasmic staining. Against the molecular tests, specificity and positive predictive value were high (92–100%), but sensitivity was about 33–36%.

A positive pan-TRK stain in a spitzoid pigmented lesion is useful support for an NTRK3-driven Reed naevus. A negative stain is not reassurance that there is no fusion, and should not be used to close the question where the diagnosis matters.

Treat diffuse granular cytoplasmic pan-TRK staining as strong support for an NTRK3 fusion
Do not use a negative pan-TRK stain to exclude NTRK3 rearrangement
Send for RNA-NGS or FISH when the distinction from melanoma depends on it
Check BRAF V600E and NRAS Q61R stains are negative before invoking a Reed naevus

Why it matters

Reframes pan-TRK in pigmented spitzoid lesions as a rule-in tool, not a screen.

The statistics, in plain English

High specificity means a positive stain is rarely wrong. Sensitivity of about a third means the stain misses roughly two of every three fusions the molecular tests find, so a negative result carries little weight.