Cutting Edge Diagnostics — Companion & Targeted-Therapy Genomic Panels
Vetopathy · Cutting Edge Diagnostics · Canine
A histopathology report has always answered "what is this tumor." Increasingly, a second question is being answered alongside it: "which drugs is this specific tumor likely to respond to." That second question is the job of companion diagnostics — genomic tests that look for a specific mutation or signature known to predict response to a specific therapy. Veterinary oncology doesn't have the deep catalog of targeted drugs human oncology does, but it already has a working example that's been in routine use for over a decade, a urine-based test that can find a tumor's DNA without a biopsy, and a new generation of multi-gene panels trying to scale the whole approach up.
The established case: KIT mutations in mast cell tumors
Mast cell tumors are the best-worked-out example of a true companion diagnostic in veterinary medicine, because the underlying biology lines up so cleanly with an available drug class. Internal tandem duplication (ITD) mutations in exon 11 of the c-Kit gene turn up in roughly 20–30% of canine cutaneous mast cell tumors. These mutations produce a constitutively active KIT receptor — the growth signal stays switched on — which is both why these tumors tend to behave more aggressively (exon 11 ITDs are associated with a higher rate of metastasis and mortality) and why they tend to respond well to the tyrosine kinase inhibitors toceranib (Palladia) and masitinib (Kinavet-CA), drugs built specifically to shut that receptor down.
Exon 8 ITDs show up far less often (roughly 2–5% of cutaneous MCTs) and haven't been shown to carry the same prognostic weight, though tumors carrying them are still expected to respond to the same TKIs. In practice, c-Kit mutation testing is usually run as part of a broader MCT prognostic panel alongside grade, mitotic index, and KIT immunohistochemistry pattern — not as a standalone test — because mutation status answers a treatment-selection question that grade and mitotic count don't.
Finding tumor DNA without a biopsy: BRAF V595E in urine
The BRAF V595E mutation (the canine equivalent of the human BRAF V600E mutation found in melanoma and other cancers) is present in a large majority of canine urothelial and prostatic carcinomas — early work put the figure near 80%, and more recent testing using droplet digital PCR (ddPCR) found it in roughly 75% of urothelial and 85% of prostatic tumor samples. What makes this mutation clinically useful isn't just its prevalence — it's that tumor DNA carrying the mutation sheds into urine in measurable quantities, which means a free-catch urine sample can sometimes answer a question that used to require a cystoscopic or surgical biopsy.
In one study, a ddPCR-based urine assay detected the mutation with roughly 85% sensitivity and 100% specificity against a confirmed-cancer and healthy-dog comparison group — a meaningful improvement over Sanger sequencing of the same urine samples, which caught the mutation only about 61% of the time. The ddPCR method's edge comes from partitioning the sample into thousands of individual reactions, which lets it detect a mutant allele at a frequency of roughly 1 in 10,000 and tolerates the PCR inhibitors that are often present in urine. This doesn't replace cytology or histopathology for a definitive diagnosis, and a negative result doesn't rule out cancer — roughly 1 in 5 of these tumors don't carry the mutation at all — but as a low-invasiveness way to flag a worrying bladder mass or monitor a known case between rechecks, it's a genuinely different diagnostic tool than anything that existed a decade ago.
Scanning the whole panel instead of one gene
The newest layer is the multi-gene oncopanel — commercial tests like SearchLight DNA, run from tumor tissue or cytology samples (FNA slides, FNA liquid, FFPE scrolls) rather than blood, that scan on the order of 100+ cancer-associated genes at once instead of testing for one known mutation. The appeal is obvious: instead of ordering a specific test because a specific tumor type usually carries a specific mutation, the panel looks broadly and reports back whatever predictive biomarkers it finds, matched against known or investigational targeted therapies and the evidence behind each match.
The honest caveat, stated plainly by the labs offering these panels themselves, is that not every tumor sequenced will contain an actionable biomarker. A multi-gene panel is a wider net, not a guarantee of an answer — some tumors come back with nothing a targeted drug exists for yet, and in dogs, many of the matched therapeutics are human oncology drugs being used off-label on the strength of molecular rationale rather than veterinary-specific trial data. That's a meaningfully different evidence bar than a well-validated single-marker test like KIT or BRAF testing, and it's worth communicating that distinction to a client before a panel is ordered on the promise of a targeted-therapy answer.
What this changes about case submission
None of this reduces the importance of good histopathology or cytology — it sits downstream of it. A genomic panel doesn't tell you what a tumor is; it tells you something additional about a tumor whose identity is already established. The main practical implication for submission is sample planning: if targeted-therapy testing is a realistic next step for a given case (a mast cell tumor being considered for TKI therapy, a suspected urothelial carcinoma, or a tumor type increasingly profiled before treatment decisions), it's worth discussing sample needs before the first sample is collected rather than after, since tissue-based panels in particular need material set aside specifically for that purpose.
Quick reference: genomic tests covered here
| Test | Sample / what it tells you |
|---|---|
| c-Kit exon 11 ITD | Tissue (part of MCT prognostic panel); predicts higher metastatic risk and good TKI (toceranib/masitinib) response; present in ~20–30% of canine cutaneous MCTs |
| c-Kit exon 8 ITD | Tissue; less common (~2–5%), expected to respond to the same TKIs, no established prognostic value |
| BRAF V595E (urine) | Free-catch urine via ddPCR; ~85% sensitivity / 100% specificity for urothelial/prostatic carcinoma in studied cohorts; present in ~75–85% of these tumors |
| Multi-gene oncopanel | Tumor tissue or cytology sample; screens 100+ cancer genes for biomarkers matched to targeted/investigational therapies — not all tumors will have an actionable match |
The throughline across all three is the same one running through most of what's genuinely new in veterinary diagnostics right now: the question being asked is shifting from "what is this" toward "what should we do about it specifically" — and that second question is exactly where precision diagnostics earns its name.
Eric Snook, DVM, PhD, DACVP — Vetopathy. Have a case where genomic testing might change the treatment conversation? Happy to talk through sample planning before anything is submitted.