Melanoma in dogs, explained honestly and in full.

Melanoma is a cancer of melanocytes, the pigment-producing cells that colour skin, hair, and mucous membranes. When they turn cancerous they form tumors that are often, but not always, darkly pigmented. In dogs melanoma is fairly common, and unlike in people, sun exposure is not thought to be the main driver. Certain breeds are over-represented, including Scottish Terriers, Cocker Spaniels, Poodles, Chow Chows, Golden Retrievers, and Dachshunds, and most affected dogs are middle-aged or older.

The most important thing to understand is that a melanoma's behaviour depends almost entirely on where it grows. The word "melanoma" alone tells you very little. Veterinary oncologists group it mainly by site, because site is the single best predictor of how it will behave.

Oral melanoma

The mouth is the most common site for malignant melanoma in dogs, and oral melanoma is the most common malignant oral tumor overall. It appears on the gums, lips, palate, or tongue as a pigmented (or sometimes non-pigmented) mass that may bleed, smell, or interfere with eating. Oral melanoma is locally invasive and spreads readily, most often to the regional lymph nodes and the lungs. This is the form that demands aggressive, well-coordinated treatment.

Digital (nail-bed) melanoma

Melanoma of the toe, also called subungual or digital melanoma, usually shows up as a swollen toe, a lost or deformed nail, or a non-healing sore that is sometimes first mistaken for an infection. It is generally malignant and carries an intermediate-to-guarded prognosis, with spread to lymph nodes and lungs possible.

Cutaneous (haired skin) melanoma

Melanomas on the haired skin are usually well-behaved. The majority are benign or low-grade, and complete surgical removal is often curative. The mitotic index, meaning how many tumor cells are actively dividing, is an important clue: low-mitotic-index skin melanomas tend to behave benignly, while the minority with high mitotic activity can act more like their oral counterparts.

Ocular and mucocutaneous melanoma

Melanoma can also arise in the eye (often treated by removing the eye when it threatens vision or comfort) and at mucocutaneous junctions such as the lips and eyelids. Behaviour varies, and your oncologist will weigh location, size, and biopsy findings together.

Diagnosis often begins with a fine needle aspirate (FNA), where a small needle samples cells from the mass for examination under a microscope. When the tumor is pigmented, the dark melanin granules can make the diagnosis relatively straightforward.

The complication is amelanotic melanoma, meaning tumors that produce little or no visible pigment. These can be difficult to identify on cytology alone and are sometimes mistaken for other cancers. To confirm the diagnosis, a pathologist examines a tissue biopsy and, when needed, runs immunohistochemistry using melanocytic markers such as Melan-A and PNL2. This confirms the cells are truly melanocytic even when pigment is absent.

Because oral and digital melanoma spread, staging is an important part of the work-up:

• Regional lymph node sampling (the mandibular nodes for oral tumors), since melanoma can spread to nodes that feel normal.
• Chest imaging, either radiographs or, increasingly, a CT scan, to look for spread to the lungs.
• Abdominal ultrasound in some cases, plus bloodwork and urinalysis before chemotherapy.

For oral melanoma, oncologists commonly use a World Health Organization staging scheme based largely on tumor size: Stage I is under 2 cm, Stage II is 2 to 4 cm, Stage III is over 4 cm or with lymph node involvement, and Stage IV is distant spread. Larger tumors, lymph node spread, and a high mitotic index all worsen the outlook.

Prognosis tracks closely with stage. To see how steeply, according to PubMed, one prospective study of 67 dogs with oral melanoma treated with electrochemotherapy reported median survival of about 16.5 months at stage I, 9 months at stage II, 7.5 months at stage III, and 4.5 months at stage IV (Tellado et al., Radiology and Oncology, 2020). Those figures come from one specific treatment in one cohort and are medians, not predictions for any single dog, but the stage gradient they show holds across the disease: catching oral melanoma early changes the outlook more than almost any single treatment choice. Cutaneous melanomas with a low mitotic index, by contrast, are frequently cured by complete removal. For why a median is a midpoint rather than a verdict, see how to read a canine cancer study.

Surgery

Surgery is the foundation of treatment for most melanomas, and the goal is wide, complete removal. For cutaneous tumors this is often straightforward and may be curative on its own. For oral melanoma, achieving clean margins frequently requires removing part of the jaw, either a mandibulectomy (lower jaw) or a maxillectomy (upper jaw). These operations sound daunting, but dogs generally tolerate them remarkably well and most eat and drink comfortably within days. For digital melanoma, the standard is amputation of the affected toe, which gives good local control with little effect on mobility. Surgery controls the local tumor; what it cannot do is address cancer cells that have already travelled elsewhere, which is why oral and digital melanoma are usually treated with surgery plus something else.

Radiation

Melanoma is relatively radioresponsive, and radiation therapy plays a major role when surgery cannot fully remove the tumor or the location makes aggressive surgery impractical. Hypofractionated protocols, which use a handful of larger doses rather than many small ones, are commonly used and can give good local control with fewer anaesthetic events. Radiation is frequently combined with other treatments: it handles the local and regional disease while systemic therapy targets the risk of spread.

Chemotherapy

Melanoma is, frustratingly, relatively resistant to conventional chemotherapy. There is no chemotherapy protocol that reliably extends survival in canine oral melanoma the way it does for some other cancers. Carboplatin is the agent most often reached for in measurable or metastatic disease, and it produces responses in a minority of dogs, but these tend not to be durable. Because of that limited benefit, the field has put much of its energy into immunotherapy instead.

Here is a piece of history many owners do not know: the first therapeutic cancer vaccine licensed for use in any species, human or animal, was a canine melanoma vaccine. ONCEPT, originally developed by Merial (now Boehringer Ingelheim), received a USDA conditional licence in 2007 and full licensure in 2010.

ONCEPT is a xenogeneic DNA vaccine: it delivers the gene for human tyrosinase, an enzyme melanocytes use to make pigment. Because the human version is just different enough from the canine version, it nudges the dog's immune system into responding to tyrosinase, and that response can cross-react with the dog's own melanoma cells. It is given by a transdermal device as a series of injections, for dogs with stage II or III oral melanoma in which local disease has already been controlled by surgery or radiation.

The evidence is worth describing precisely, because neither side of the debate rests on a randomised controlled trial. The study behind its licensure followed 58 dogs with stage II or III oral melanoma and compared them with 53 historical controls, reporting significantly longer survival in the vaccinated group; its first author worked for the manufacturer (Grosenbaugh et al., American Journal of Veterinary Research, 2011). A later single-centre retrospective review of 45 dogs found no improvement in progression-free survival, disease-free interval, or median survival (Ottnod et al., Veterinary and Comparative Oncology, 2013). Historical controls and retrospective records are both weaker designs than a randomised trial, which is the honest reason the question remains open (reviewed in Pellin, Veterinary Sciences, 2022). The reasonable summary in 2026: ONCEPT is a low-toxicity option whose magnitude of benefit is genuinely debated, and that is a question for your oncologist.

ONCEPT targets one shared self-antigen, tyrosinase, that every dog's melanoma has in common. The next generation of cancer vaccines flips that idea around: instead of one antigen for every patient, they target the mutations unique to one patient's tumor. Melanoma is, interestingly, the cancer where this personalized neoantigen approach has its clearest human proof. In people, adding a personalized vaccine (mRNA-4157) to the immunotherapy drug Keytruda cut the risk of melanoma recurrence or death by about 44 percent versus Keytruda alone in a randomised phase 2b trial (Weber et al., Lancet, 2024, the KEYNOTE-942 trial, 157 patients); that was a phase 2b result, with a larger phase 3 trial now underway.

The same pipeline works in dogs: the species reference is the canine genome, the immune-presentation molecules are DLA rather than human HLA, and the codon table differs, but the structure is identical. We should be equally clear about our own position, though. That 44 percent is human data, in one of the most heavily mutated cancers there is, and a result in people does not transfer automatically to dogs. No personalized cancer vaccine, including the ones we design, has published canine efficacy data in melanoma or any other dog cancer. The human precedent is a reason to pursue the work responsibly, not a claim about what is available for dogs today. Our plain-language explainer covers how the approach works.

Two practical steps protect your options. First, if surgery is planned and any personalized approach might interest you, ask that the tumor tissue be preserved properly, following the receiving laboratory's own instructions (our short guide explains why). Second, come to the oncology appointment with the questions that matter:

1. Where is the tumor, and what does that location mean for how it is likely to behave?
2. What did the biopsy show, and what is the mitotic index? Was immunohistochemistry needed to confirm it?
3. Did staging (lymph nodes, chest imaging) show any sign of spread?
4. What stage is it, and what is the realistic prognosis with versus without the treatment you recommend?
5. Is wide surgical removal achievable? If it means a mandibulectomy, maxillectomy, or toe amputation, what is recovery like?
6. Would radiation add meaningful local control in our case?
7. Is the ONCEPT vaccine appropriate for my dog, and how do you weigh its mixed evidence?
8. Are there clinical trials or experimental options worth knowing about, and is our case a reasonable fit?
9. What does follow-up monitoring look like, and what are we watching for?

If you do not yet have a veterinary oncologist, your primary veterinarian can refer you, or see our find-help directory.