Vetopathy · Pathology Deep Dive · Canine
Most pigmented skin masses on dogs are melanocytomas — benign, and done once they're excised. But cutaneous malignant melanoma is a real diagnosis, it can arise at any haired-skin site, and the two can look deceptively similar on initial impression. Location helps set a prior probability — melanocytic tumors of the nail bed, mucocutaneous junctions, or mouth skew more aggressive than ones on the trunk or limbs — but it isn't diagnostic on its own. Aggressive cutaneous melanomas occur, and relatively indolent tumors occasionally show up in those higher-risk locations. The actual answer comes from the histology, and increasingly, from a proliferation marker that's turned out to be more reliable than almost anything else on the slide.
Why location sets expectations but doesn't decide the case
Median survival time is longest for cutaneous melanocytic tumors generally, shorter for nail bed and mucocutaneous junction tumors, and shortest for oral tumors — a real pattern, and one worth keeping in mind when a mass is first palpated. But it's a population-level pattern, not a rule for an individual case. Breed and age carry a similar amount of information: some breeds skew toward one behavior or the other, but neither is reliable enough to lean on by itself. The histologic features below are what actually separate a tumor's behavior from its location's reputation.
What the slide actually shows
A typical melanocytoma is small, raised, non-ulcerated, heavily pigmented, confined to the dermis, and built from well-differentiated cells with rare mitotic figures. A typical malignant melanoma tends to run the other direction on nearly every one of those features: larger, more often ulcerated, extending beyond the dermis into deeper tissue, frequently less pigmented, with more nuclear atypia and more frequent mitoses. Few individual tumors hit every point on either list cleanly, which is exactly why grading criteria exist — to turn a cluster of impressions into a reproducible threshold.
Mitotic count is one of those thresholds, but the cutoff differs by site: roughly 3 or more mitoses per 2.37 mm² in cutaneous tumors, versus roughly 4 or more in oral and lip tumors, is associated with shorter survival. Nuclear atypia is scored similarly — counting 100 to 200 cells at 40x magnification — with roughly 20% or more atypical nuclei in cutaneous tumors (versus roughly 30% in oral/lip tumors) separating the more aggressive end of the spectrum. Junctional activity, where neoplastic melanocytes are seen at the dermoepidermal junction, is a stronger and more consistent diagnostic feature in oral tumors than in cutaneous ones, where its prognostic value has been inconsistent across studies. Ulceration predicts aggressive behavior specifically in cutaneous tumors — though it's worth remembering a chronically self-traumatized benign mass can also ulcerate, so it's a data point, not a verdict on its own. Vascular invasion, when present, is one of the more unambiguous indicators of malignant behavior regardless of site.
Size and depth carry real weight
Two thickness cutoffs for cutaneous tumors are worth keeping in mind when a submission includes a decent-sized excision: tumor thickness greater than roughly 0.95 cm is associated with shorter overall survival, and thickness greater than roughly 0.75 cm with a shorter disease-free interval. Pigmentation cuts the other direction — tumors with 50% or more pigmented cells tend to carry a better prognosis — though a poorly pigmented (amelanotic) tumor isn't automatically a worse one; it mainly means immunohistochemistry (Melan-A, PNL-2, TRP-1, TRP-2) may be needed to confirm the diagnosis in the first place, since amelanotic melanocytic tumors can closely mimic poorly differentiated carcinomas, soft tissue sarcomas, and round cell tumors on routine stains.
Why Ki-67 has become the most trusted single number
Of all the prognostic markers studied in canine melanocytic tumors, the Ki-67 proliferation index (detected by immunohistochemistry, sometimes under the antibody name MIB-1) has consistently come out as the most reliable — the highest reproducibility and the least interobserver variation of any marker in the comparison, and in several studies a better predictor of survival than mitotic count or nuclear atypia alone. For cutaneous tumors, a threshold around 15% or more Ki-67-positive cells (counted across 500 cells at 40x) is associated with shorter survival; in at least one study, no benign cutaneous tumor crossed that threshold at all. Oral and lip tumors use a different cutoff, scored per grid area rather than percentage. In practice, Ki-67 doesn't replace mitotic count, atypia, and pigmentation — the strongest prognostic read comes from combining them — but when the histologic picture is ambiguous, it's often the number that resolves it.
Quick reference: cutaneous melanocytoma vs. melanoma
| Feature | Favors melanocytoma / lower risk | Favors malignant melanoma / higher risk |
|---|---|---|
| Size / ulceration | Small, non-ulcerated | Larger, ulcerated |
| Depth | Confined to dermis; thickness < ~0.75 cm | Extends beyond dermis; thickness > ~0.95 cm |
| Mitotic count (cutaneous) | < 3 per 2.37 mm² | ≥ 3 per 2.37 mm² |
| Nuclear atypia (cutaneous) | < 20% atypical nuclei | ≥ 20% atypical nuclei |
| Pigmentation | ≥ 50% pigmented cells | Poorly pigmented (may need IHC to confirm diagnosis) |
| Ki-67 index (cutaneous) | < ~15% positive cells | ≥ ~15% positive cells |
| Vascular invasion | Absent | Present — a strong malignancy indicator on its own |
None of these thresholds work in isolation, and location is context rather than a diagnosis. A small, well-pigmented, low-mitotic-count mass on a limb and a similar-looking mass on a digit deserve the same histologic workup — the site changes the prior, not the process.
Eric Snook, DVM, PhD, DACVP — Vetopathy. Have a pigmented mass where the clinical picture and the histology don't quite agree? Happy to talk through whether IHC or a Ki-67 index would help settle it.