Bone Cancer in Sighthounds: What the Research Actually Says

Osteosarcoma is a cancer of the cells whose job is to build bone. They keep building, badly and without a stop signal, while at the same time the bone around them is eaten away. Building and destroying in the same place at once. That is what makes the X-ray look so chaotic, and it is what causes the pain.

It is about 85 per cent of all primary bone tumours in dogs.

Where it starts

Almost always at the end of a long bone, just below the growth plate. Vets say it lands away from the elbow and towards the knee. In practice that means the wrist end of the front leg, the shoulder, and either side of the stifle behind.

Those are the growth plates that build most of a tall dog’s height and stay active longest. The more bone-building a dog has to do, the more chances there are for something to go wrong.

Why it hurts so much

• The bone loses strength, so it develops tiny fractures under normal use.
• The tumour pushes out against the sheath around the bone, which is one of the most nerve-rich tissues in the body.
• The inflammation makes local nerves oversensitive, so ordinary pressure starts registering as pain.

How it kills, and it is not the leg

Cancer cells get into the bloodstream and lodge in the first place they reach, which is the lungs. This happens early and silently.

So amputation removes the pain and the tumour you can see, but in about nine cases out of ten the disease has already left the leg. Chemotherapy exists to go after that invisible lung seeding. That is the whole of what it is for.

↑ On this page

The best UK figures come from a study that searched the records of over 900,000 dogs. Across the whole dog population this is uncommon, roughly one dog in 3,000 a year. That number is close to useless for a sighthound owner, because the spread between breeds is enormous.

Odds against a crossbreed

• Scottish deerhound: 118 times. The highest of any UK breed.
• Leonberger: 56 times.
• Great dane: 34 times.
• Rottweiler: 27 times.
• Greyhound: 12 times.
• Lurcher: 2.7 times. Crossing dilutes the risk a lot. It does not remove it.

Now forget breed names and look at shape

• Long-muzzled build: 2.7 times the odds of a medium-muzzled dog.
• Short-legged build: one tenth the odds.
• Heavier than average for the breed: adds about 65 per cent.

That is the real pattern. It is not a disease of big dogs, it is a disease of long bones, and a sighthound is built out of long bones.

How much of this is decided at birth

A great deal of it, and this matters if you are lying awake wondering what you did wrong.

• Across greyhounds, rottweilers and wolfhounds, 33 inherited genes explain 55 to 85 per cent of the risk.
• In Irish wolfhounds the heritability is 0.65, and 93 per cent of osteosarcoma deaths happen before the dog turns 10.
• In Scottish deerhounds a single high-risk gene copy carries over 75 per cent risk, against under 5 per cent without it. More than 15 in every 100 deerhounds are affected in their lifetime.

Neutering age: what is fair to say

The main study followed 683 rottweilers. Bone cancer appeared in about 12 in every 100 overall, but in roughly 25 in every 100 of the dogs neutered before their first birthday.

The 2025 international consensus is careful about this. It says it is sensible to delay neutering until a dog is fully grown in at-risk breeds, while being clear the link is not settled.

My honest reading: waiting until the growth plates close in a big sighthound is a cheap, sensible precaution. It is not prevention, and no dog ever got this because of when you had her spayed.

↑ On this page

Every owner lands in the same place afterwards. If only I had caught it a month sooner.

Here is the shape of the disease, and it is worth sitting with:

• The tumour starts in the leg and is completely silent.
• Cells reach the lungs early, while there is still nothing to see.
• The dog starts quietly doing less. This is the first outward sign, and it looks like ageing.
• A limp appears, or a leg breaks under normal load.
• Diagnosis. About 9 in 10 dogs already have lung disease, but it is far too small to appear on an X-ray. Only about 1 in 20 shows a visible lung shadow.

Moving the diagnosis a month earlier moves one line on that list. It does not move the lungs.

What earlier detection does buy

• More time before a break forces an emergency decision at ten o’clock at night.
• More options while the bone is still structurally sound.
• Pain relief sooner, which is worth a great deal on its own.
• A calmer decision instead of a rushed one.

What it does not buy

A different ending. This is also why no screening programme for high-risk breeds has ever been shown to help. If you missed the early signs, you did not cost your dog her life. This disease is built to be missed.

↑ On this page

Knowing the order makes the appointment less frightening. The 2025 consensus is also refreshingly blunt that in a typical case you do not need every test.

1. The X-ray does most of the work

A plain radiograph of the sore leg is the main tool and is often conclusive on its own. Your vet is looking for a specific combination:

• The hard outer shell of the bone visibly eaten away.
• The internal scaffolding dissolving into a moth-eaten grey.
• New bone being thrown down at the same time, sometimes in a radiating sunburst pattern.
• A wedge of new bone where the tumour has lifted the sheath off the shaft.
• It does not cross the joint. That one feature separates it from most infections.

2. A needle sample, usually not a biopsy

A fine needle sample is now first choice and gets the answer 70 to 97 times out of 100. A full core biopsy is more thorough but weakens an already fragile bone, so where the picture is typical most oncologists skip it and confirm afterwards.

3. Staging the chest

Three chest X-rays, and a feel of the local lymph node. A CT scan picks up smaller spots if you can afford it.

4. Bloods

One number matters: alkaline phosphatase. It does not diagnose anything, but a dog whose level is normal before treatment lives significantly longer than one whose level is raised. Ask for it.

Things that make the outlook worse

• Raised alkaline phosphatase. The most consistent bad sign in the whole literature.
• A tumour in the shoulder. Consistently the worst location.
• A bodyweight of 40kg or more.
• Spread to a lymph node. Rare, under 5 in 100, but serious when it happens.

↑ On this page

These are the figures the 2025 international consensus puts its name to. They are medians, so half of dogs do better and half do worse, and they come from populations rather than from your dog.

• Pain relief only: a few weeks to a few months.
• Amputation alone: 3 to 4 months. Around 1 in 10 alive at a year.
• Amputation plus carboplatin: 8 to 14 months. About 4 in 10 alive at one year, 1 in 5 at two.
• Targeted radiotherapy instead of amputation: about 8 months, but 4 in 10 dogs fracture the treated bone afterwards.
• Limb-sparing surgery: about 9 months, with high complication rates. Not better than amputation plus chemotherapy.

If chemotherapy is the plan, it should start within 5 days of the amputation. Dogs who start later do worse. Book the oncologist before the surgery, not after it.

Amputation, honestly

It is the treatment of choice, and it is worth being clear what it is for. Amputation is the pain treatment. It removes the source completely and immediately. It adds relatively little on its own to how long a dog lives, because the lungs are already seeded.

That is not an argument against it. Turning a dog in constant severe pain into a comfortable dog is worth doing for its own sake.

Dogs cope better than owners expect. The honest caveats are very heavy dogs, dogs with arthritis in the other legs, and front-leg amputees, because a dog carries about 60 per cent of its weight on the front.

Chemotherapy, honestly

Carboplatin is the drug of choice. Dogs tolerate it far better than people do, because the doses are set for quality of life rather than maximum tumour kill. Most dogs get a few off days and a dip in white cells around day seven. Hair loss is uncommon outside continuously coated breeds.

Two findings that save money and car journeys: two drugs are no better than one, and giving chemotherapy before surgery rather than after did slightly worse in dogs, which is the opposite of the human picture.

If surgery is not right for your dog

Palliative care is a real choice, not a lesser one, and there is more available than owners are often told.

• Pain-relief radiotherapy, typically four weekly sessions.
• Layered pain relief: an anti-inflammatory first, then opioids, nerve-pain drugs and others alongside.
• Bisphosphonates, which slow the cells eating the bone and help with pain.

↑ On this page

This is one of the most heavily researched cancers in veterinary medicine, because it is nearly identical to the childhood human disease and dogs get it about ten times more often. That produces a lot of headlines. Here is what is behind them.

Promising
Personalised immunotherapy

The dog’s own tumour is used to make a vaccine, then its immune cells are taken out, activated and put back. It became the first treatment of its kind approved by the US Department of Agriculture, in March 2025, specifically for this disease.

The result: 71 in 100 alive at one year, against 21 in 100 of the comparison dogs.

The catch: only 14 dogs in each group, the comparison dogs were matched rather than randomly assigned, and those comparison dogs did worse than chemotherapy normally achieves, which makes the gap look bigger than it may be. Full results are due late 2026. It costs around 20,000 US dollars, and it is not available in the UK, partly because it has to be made from tissue taken at your dog’s own surgery.

Promising
Losartan with toceranib

A blood pressure tablet at high dose, paired with a licensed canine cancer drug. It blocks the immune cells the tumour recruits to help it settle in the lungs.

The result: half the dogs with established lung spread got a meaningful benefit, at a stage where almost nothing else works.

Why it matters here: both drugs are cheap, oral and already available in the UK. Of everything on this page, this is the one most worth asking your oncologist about.

Early days
Vaccines, checkpoint drugs and viruses

• A vaccine against a protein called CSPG4 produced longer survival than untreated dogs.
• Immune checkpoint drugs, the class that transformed human melanoma, are real and the target exists in dogs. No survival benefit has been shown in this disease yet.
• Viruses engineered to attack tumour cells have shown safety and some benefit. Nothing near clinical availability.

Did not work
The graveyard, and why it is the useful part

Every one of these was being described the way the promising ones are described now.

• Sirolimus. 324 dogs, properly randomised, the best-designed trial ever run in this disease. No benefit at all.
• Inhaled interleukin-15. Encouraging early results, then stopped early because dogs were doing worse than the comparison group.
• A bacterial vaccine. Withdrawn after it caused infections that could pass to people in the household, and later studies showed no survival benefit.
• Low-dose continuous chemotherapy and several add-on drugs. No survival advantage.

The lesson I take from that list: a small study with no control group is a hypothesis, not a result. Ask how many dogs, and ask what the dogs who did not get it were doing.

↑ On this page

Start with the fact that governs the rest of this section. No diet or supplement has ever been tested against bone cancer in dogs. Every study below is a different cancer, or healthy dogs. Anyone telling you a food fights bone cancer is guessing, and usually selling.

Proven
Keeping her lean

The strongest diet evidence in all of dog medicine. 48 labradors in littermate pairs, one of each fed 25 per cent less from eight weeks old until death, followed for fourteen years.

It is the only dietary thing on this page with proper long-term proof, and it is free.

Proven
Enough food and enough protein during treatment

The real nutritional enemy is muscle wasting, and it is badly under-recognised. Around 37 in every 100 dogs had already lost more than 5 per cent of their bodyweight by the time they were diagnosed. Body condition predicts survival directly.

The goal during treatment is not fighting the tumour. It is keeping the dog eating and holding her muscle. Appetite is the thing to watch.

Cornell’s position, and I agree with it: do not change the diet without a medical reason. A complete balanced food she will actually eat beats an exotic one she will not.

Reasonable
Omega-3 oils

This is the most quoted diet evidence there is, and it is worth reading properly. The famous study put 32 dogs with lymphoma on a fish oil diet against an identical soybean oil diet.

Overall, survival was no different between the two diets. The benefit only appeared when the dogs were split by stage, in one subgroup. That is a very different thing from “fish oil extends survival”, which is how it is nearly always quoted.

Still worth using. Low risk, genuinely anti-inflammatory, helps with appetite and muscle. Expect that, not months.

Not shown
Keto diets, turmeric and CBD

• Sugar feeds cancer. The mechanism behind this is established in human cancers but has not been confirmed in dogs. In one of the few low-carbohydrate studies in dogs, there was no effect on remission or survival.
• Turmeric and curcumin. Real activity against canine cancer cells in a dish, but it is poorly absorbed in a living dog and there is no clinical trial in any canine cancer.
• CBD. A 2026 review of the whole field concluded the mechanisms are not understood and the products are not standardised. There is no trial of CBD against any canine cancer.

Be careful
Two things worth pausing

• High-dose antioxidant supplements during chemotherapy or radiotherapy. A 2021 paper in the Veterinary Record argues these should be actively discouraged, because they may protect the tumour along with everything else. Antioxidants in ordinary food are not the concern. Supplement doses are. Tell your oncologist everything your dog is getting, so interactions are known.
• Raw food while on chemotherapy. Not a verdict on raw feeding generally. Chemotherapy knocks the white cells down, and during those windows a dog cannot fight off a bacterial infection normally. UK data shows raw-fed dogs carry much higher rates of salmonella and resistant E. coli. Cornell, the AVMA and WSAVA all advise against raw while a dog is immunosuppressed.

Things that are actively dangerous, and all of it circulates

• Garlic and onion. Promoted as an immune booster. Damages red blood cells.
• Apricot kernels, laetrile, so-called vitamin B17. They release cyanide, and there is no anti-cancer evidence in any species.
• Human supplements dosed for a 70kg adult, and anything containing xylitol.
• Grapes and raisins, and essential oils given by mouth.

↑ On this page

Take this list in with you. You will not remember it in the room.

• Is the X-ray typical, or is there another explanation worth ruling out?
• What is her alkaline phosphatase, and is it normal or raised?
• Which bone and which end, and does that spot carry a worse outlook?
• Chest X-rays or a chest CT, and would the CT change anything?
• If we amputate, can chemotherapy start within 5 days? Is the oncologist already booked?
• Is she a realistic candidate for amputation, given her weight and her other three legs?
• If we do not amputate, what is the full pain plan? Radiotherapy, bisphosphonates, which drugs?
• Would high-dose losartan with toceranib be reasonable for her, now or later?
• Is there any trial she could join?
• What exactly will tell me the pain is no longer controlled?
• What does each option cost, and what will the insurance actually pay?
• If she were your dog, what would you do?

↑ On this page

Every figure on this page traces to one of these. I have put the guidelines and the randomised trials first, because those are the ones that count.

Guidelines and trials

• Osteosarcoma of the appendicular skeleton in dogs: consensus and guidelines. Frontiers in Veterinary Science, 2025. The source for diagnosis, staging, survival figures and treatment protocols.
• Adjuvant sirolimus does not improve outcome: a randomised trial of 324 dogs. Clinical Cancer Research, 2021.
• Inhaled interleukin-15 phase 2 trial, halted for futility.
• Losartan with toceranib in canine metastatic osteosarcoma. Clinical Cancer Research, 2022.
• Stereotactic radiotherapy in 123 dogs. Veterinary and Comparative Oncology, 2021.

Breed risk and genetics

• Dog breeds and conformations predisposed to osteosarcoma in the UK: a VetCompass study. 905,552 dogs. The source of every breed and body-shape figure above.
• Genome-wide analyses implicate 33 loci in heritable dog osteosarcoma. Genome Biology, 2013.
• Heritability and segregation analysis of osteosarcoma in the Scottish deerhound. Genomics, 2007.
• Heritability of osteosarcoma in Irish wolfhounds. 2021.
• Endogenous gonadal hormone exposure and bone sarcoma risk. Cooley and others, 2002. The 683-rottweiler neutering study.

Immunotherapy

• American Veterinary Medical Association on the 2025 approval.
• Owner overview of the immunotherapy, including cost and availability. VCA Animal Hospitals.

Diet

• Nutrition and oncology in dogs and cats: a comprehensive review. Frontiers in Veterinary Science, 2024.
• Effects of diet restriction on life span and age-related changes in dogs. The 14-year Purina life span study.
• Fish oil, arginine and chemotherapy in dogs with lymphoma. The omega-3 study, in full.
• Should we be discouraging use of anticancer supplements in dogs and cats with cancer? Veterinary Record, 2021.
• Diet decisions during cancer treatment. Cornell University College of Veterinary Medicine.

↑ On this page

I am a sighthound trainer and boarder, not a vet, and nothing here is a diagnosis or a treatment plan. This is what I read after losing my own dog, gathered in one place because it was not gathered anywhere I could find it.

Where I have quoted a number I have linked the study it came from, so you can check it or take it to your vet. Median survival figures describe groups of dogs, not yours. Every decision here belongs with the vet who can actually see your hound.

If something on this page worries you about your own dog, that is the phone call to make, and make it sooner than feels necessary.

I am Don Jordan, The Sighthound Guy. I look after greyhounds, whippets and lurchers across London, with collection around Hampstead, North Finchley and Highgate, a 350ft secure private garden, a climate-controlled home and daily WhatsApp photo and video updates so you always know how your dog is getting on.

The Sighthound Guy is fully insured and bonded, Canine First Aid certified, and holds a 5-star home boarding licence (no. 2023-2989). We hold a 5-star Google rating, are a Google Selected Business and have been featured in Time Out. Sighthounds only.

To talk through your hound, see what I offer, read how I work, or get in touch to arrange a meet and greet.

For Storm, who never let on.