I am often asked by clients if there is a blood test that can tell if their apparently healthy pet has cancer. Obviously, “cancer” is a scary word, and many of the various diseases given that label are difficult to diagnose and treat. Many people have had personal experiences of a family member, human or non-human, developing cancer, and they are understandably looking for a way to identify such disease as early as possible, with the hope this will make it easy to control or even cure. Unfortunately, like most things in medicine, the reality is far more complex and uncertain than a simple blood test that will identify a hidden cancer in time to treat it.

Cancers differ in their basic biology and behavior, so each requires a different approach to diagnosis. Leukemia may show up as an abnormal number of white blood cells in a complete blood test, but breast cancer will not. High calcium levels have been associated with several cancers, but these range from lymphoma, a common white blood cell cancer spread throughout the body, to anal gland carcinoma, an uncommon tumor that shows up in that particular location only. The variety among cancers make any single test for most or all types very unlikely.

A Bit About Screening Tests

It is also important to understand a bit about the difference between testing for a disease when there is some symptom or other reason to suspect it is present (diagnosis) and testing for the presence of disease in apparently healthy patients (screening). The chances of getting the right answer from a specific test for a particular cancer are much higher when the patient has typical symptoms or other test results that suggest the presence of that cancer. All tests are imperfect, and all give false positive or negative results sometimes. If large numbers of healthy patients are tested, most of whom don’t have the disease, then a lot of the positive tests will turn out to be wrong, but most of the negatives will be true. This is the case even for very, very good tests.

The best way to illustrate this is—brace yourself—with a little math.

Say you have 10,000 people to test, none of whom have any symptoms or problems to suggest they have cancer. However, 200 of them have the cancer you are testing for (2%), while 9,800 do not. You have a test that is about as close to perfect as you can get–98% sensitive and 98% specific. That means that if you have 100 patients with the disease, 98 will test positive and 2 will test negative even though they have it (sensitivity). It also means that if you test 100 patients who don’t have the disease, 98 will test negative and 2 will test positive even though they don’t have it (specificity).

So we test all 10,000. Here are the results we will get:

True positive (test positive and have cancer)- 196

True negative (test negative and don’t have cancer)-9,604

False positive (test positive but don’t have cancer)-196

False negative (test negative but actually have cancer)-4

So we found most of the cases of cancer and we re-assured most of the healthy people. BUT, we also told 4 people with cancer that they were fine. And we told 196 healthy people they had a terrible disease. Did we do more good than harm?

Well, it depends on lots of other things. If early detection leads to more effective treatment and better survival, then at least the true positive patients will benefit. However, those we missed will think all is well and may ignore future symptoms or test results that might actually help them get treatment.

On the other hand, if early detection with this test doesn’t lead to better treatments and outcomes, then all we’ve done is give people bad news which they can’t do anything about even if it’s true. Screening doesn’t help anybody if there is nothing we can do about the disease at the point at which we detect it.

It is also likely that those who got a positive test will have to undergo uncomfortable and expensive testing and maybe even cancer treatment even if they don’t actually have cancer (which, remember, is almost 200 of our patients). We have actively harmed these folks with our test.

There has been a lot of controversy in human medicine recently about screening tests for cancer that were once widely recommended. PSA testing for prostate cancer, for example, turns out to cause a lot of harm. Positive test results often result in expensive and uncomfortable treatment with some serious side effects, and yet it appears not to reliably improve long-term outcomes. In other words, men who get tested will often suffer more harm from treatment without actually living any longer than men who did not get early detection with this test. While the test still has some appropriate uses, widespread screening of men with no symptoms of disease turns out not to be a very good idea. Similar concerns have led to changes in screening for breast cancer and to a growing recognition in human medicine that screening tests of apparently healthy people lead to significant overdiagnosis and harm in many cases.

The situation is a bit different in veterinary medicine. For one thing, there are few blood tests that have been tested at all for the ability to detect cancer before symptoms develop. And for most cancers, there is no research evidence that such early detection would allow us to better treat or even cure such cancers if we could find them. While it is often assumed that the earlier we detect cancer the more we can do about it, this has not been proven in dogs and cats and has actually turned out not to be surprisingly often in humans.

But My Vet Said There Was a Good Screening Test for Cancer

Unfortunately, there are still some veterinarians who will recommend, and sell you, cancer screening blood tests for your pets with a good deal more confidence in their benefits than is justified. A recent article, written by an alternative medicine practitioner I’ve written about before, makes strong claims for his ability to use a couple of blood tests, C-reactive protein, thymidine kinase, and Vitamin D3 levels, to detect cancer or precancerous conditions before your pet is sick and then treat them effectively.

A series of simple, new, inexpensive blood tests bundled into a diagnostic panel now allows doctors to detect cancer and other serious problems before they develop in your dog or cat…These new blood tests can determine your companion’s risk of developing this dreaded disease, and even detect it in its earliest stages.

While these new cancer tests are very helpful for determining the status of your animal’s health before he becomes ill, they are also helpful for monitoring dogs and cats that already have cancer. For those already diagnosed with the disease, monitoring [these] levels helps veterinarians make decisions about changing treatment and to predict when/if the cancer may return and when the animal may come out of remission.

This cancer testing is very cost effective. In our office, we charge under $200 for the entire three-test profile, making it very inexpensive considering the wealth of information we receive from it. Additionally, this profile is the easiest and least expensive way to screen dogs and cats for cancer and other serious inflammatory diseases. There are really no disadvantages to having this testing done.

These claims raise a number of important questions: 1. Can these tests really detect cancer before your pet has symptoms? 2. What should we do if we have a positive test result? 3. Does testing, and whatever we do next, help prevent illness or prolong health and life? 4. Can this testing do any harm?

Can these Tests Really Detect Cancer Before My Pet Gets Sick?

The short answer is, we don’t know. It is important to remember that none of these tests is routinely used or recommended for cancer screening of healthy people or animals by doctors or organizations specializing in cancer detection and treatment. This kind of screening is not a widely accepted practice. Nevertheless, there is research evidence looking at the possible value of these tests for cancer screening. I will offer a brief summary of the evidence and try to answer our four questions.

Thymidine kinase (TK1)- This is a substance which is produced by dividing cells. Since cancer cells typically divide more rapidly than healthy cells, it tends to be produced in greater amounts when there are cancer cells present. This is particularly true for cancers of blood cells, such as lymphoma and leukemia. TK1 it is less consistently elevated in patients with other kinds of tumors. To a lesser extent, elevated TK1 levels have also been associated with viral infections and inflammatory conditions as well as cancer.

There is a lot of evidence to suggest that TK1 levels are higher in dogs with known cancer than in healthy dogs, especially when the cancer is lymphoma or leukemia. Most studies compare health animals with those already diagnosed with cancer by other methods, and they usually find higher TK1 levels in the cancer group. These are not, however, animals with occult cancer, that is cancer that hasn’t caused any symptoms or been diagnosed yet. Whether TK1 levels are consistently higher in dogs who have cancer but not symptoms hasn’t been clearly demonstrated. The TK1 levels of healthy dogs, dogs with cancer, and dogs with other diseases overlap quite a bit. There are even some studies in which as many as half of the dogs with known lymphoma had normal TK1 levels. So there are certainly not clear absolute differences to distinguish these groups.

While it has some potential, TK1 testing is not clearly a useful diagnostic or monitoring tool for dogs with most cancers, much less a screening tool for dogs who might or might not have cancer. As usual, there is even less data related to cats, though there are some suggesting TK1 might behave similarly in this species.

C-Reactive Protein (CRP)-

CRP is a protein associated with inflammation. Inflammation can be caused by many different things, such as infection, autoimmune disease, and of course cancer. CRP is quite sensitive to inflammation, but it does nothing to help distinguish the cause. In dogs, CRP is elevated in at least some kinds of cancer, and the values are generally higher than in dogs without cancer. However, there is significant overlap between these groups, so some dogs with cancer will have normal CRP and some without will have elevated CRP. Interestingly, dogs with lymphoma who are in remission (no illness or detectable tumor but microscopic cancer still present somewhere) generally have lower CRP than dogs with active lymphoma. This means CRP could possibly be used to test for relapse of this cancer. However, dogs in remission are just like dogs without symptoms of cancer who we would be using a test like this to screen, so these data suggest we wouldn’t be able to find lymphomas with this test if there were no symptoms or obvious tumors.

Vitamin D–

Vitamin D, like all vitamins, has a number of essential functions in the body. It has been well known for decades how much vitamin D people (and dogs and cats) require to prevent symptoms of Vitamin D deficiency disease. However, it has been a bit of a fad for a while now to claim that even higher levels would have protective effects against a variety of diseases, including cancer.

This is based largely on a few studies which found that people with certain disease, including some kinds of cancer, tend to have lower Vitamin D levels than healthy people. However, additional research has failed to demonstrate that low levels cause these diseases or that supplements prevent them. People who are sick may have low Vitamin D levels as a symptom of their illness rather than as part of its cause. For the most part, testing for Vitamin D and supplementing to prevent cancer is not recommended in humans unless there are specific reasons to believe an individual patient is at risk for low Vitamin D levels (in other words, routine screening is not recommended).

There is not good evidence to suggest that subclinical Vitamin D deficiency is a real problem in dogs or cats either, or that supplementation beyond the levels currently used in commercial diets is of any benefit.

Veterinary Diagnosic Institute Cancer Tests-

Despite the fact that these tests are not used as general cancer screening aids in humans, and the lack of strong evidence to show they can reliably predict cancer in dogs and cats, there is a company selling these tests as a tool for cancer screening, as well as for cancer diagnosis and monitoring response to therapy. Apart from the general information for each test I’ve already discussed, the company has produced one clinical study to evaluate its screening test.

Selting KA, Sharp CR, Ringold R, Knouse J. Serum thymidine kinase 1 and C-reactive protein as biomarkers for screening clinically healthy dogs for occult disease. Vet Comp Oncol. 2013 Jul 16. doi: 10.1111/vco.12052. [Epub ahead of print]

This study involved collecting blood from about 360 dogs (German shepherds and golden retrievers) to test for CRP and TK1 and then following them for 6 months or more to see which developed cancer or died of some other disease. The dogs were not evaluated by a veterinarian as part of the study, and most health information was obtained from questionnaires completed by owners. This raises the possibility that some dogs may not have been properly classified or diagnosed, which might change the results and conclusions significantly.

As expected, over the short course of the study the incidence of cancer was pretty low, about 3%. This means 11 dogs developed some kind of cancer. Because CRP is not very specific for cancer but simply detects any inflammation, it was not used by itself to predict cancer development. However, the TK1 test tends to be more specific for cancer but not as sensitive as the CRP, so the authors looked at both the TK1 by itself and the combination of the two tests (called the neoplasia index or NI) to predict cancer development.

Sensitivity and specificity of a diagnostic test are always in opposition to one another. The more sensitive a test is (that is, the more actual cases of disease it finds), the less specific it is (that is, the more healthy patients will test positive for a disease they don’t have). This is true for any test, and the balance between the two numbers is related to what value of the test one uses as a cutoff. In this study, the authors followed the common practice of  analyzing their data with a variety of cutoffs to try and find the best balance between sensitivity and specificity for the test.

In the case of cancer screening for apparently healthy patients, there are dangers to both a low sensitivity and a low specificity. If the test is very sensitive but not very specific, then it will indicate a lot of healthy animals have cancer when they don’t. Apart from the worry this creates for their owners, there will be a lot of expensive, and sometimes uncomfortable or even dangerous, follow-up testing and possibly even unnecessary treatment for a disease that isn’t there. This is the danger of overdiagnosis.

On the other hand, if a test is very specific but not very sensitive, a lot of animals with cancer will show up as healthy on the test. This is a problem if early detection and treatment could lead to a better outcome, though as I mentioned earlier this has not generally been demonstrated for dogs and cats with cancer.

For TK1 alone, the best sensitivity achieved was 100%, meaning all dogs that developed cancer later had an abnormal test at that particular cutoff. However, when that level was used, the specificity was 0, meaning that everybody would test positive whether they had cancer or not. At the other end, the best specificity achieved was 98%, but at this level sensitivity was only 27%. Between these extremes, it was possible to get one or the other value around 80% with the opposite value about 70%. This is not a very good compromise for a test intended to be used in a population with a very low rate of cancer because there will be a lot of patients misclassified with regard to a very serious disease.

The authors expected, of course, the TK1 to be pretty specific (good at ruling out cases that don’t have cancer) but not very sensitive. This is why they combined it with the more sensitive CRP to try and improve the overall sensitivity without losing too much specificity. This combined neoplasia index had an optimal balance of sensitivity and specificity at 82% sensitivity and 91% specificity. In the case of a cancer screening test, it is probably better to favor specificity so you don’t do unnecessary tests and treatments on healthy animals. A lower sensitivity will miss some true cases of disease, but it isn’t clear yet that finding these cases earlier improves their outcome, and of course they will eventually develop symptoms that will lead to a diagnosis.

In the case of a screening test, which is intended to be used on a population which is mostly healthy and disease is rare (about 3% in this study, for example), there are a couple of numbers more useful than specificity and sensitivity. These are the positive predictive value (PPV) and the negative predictive value (NPV). The PPV is, essentially, the percentage of positive tests that are correct, that is the patient actually has the cancer the test says they have. The PPV for this test in this study was 22.5%. That means that when a patient has a positive test, only about 22% of the time is there actually a cancer there. That leads us to  the second question….

What Should I Do if my Pet Has a Positive Test?

The veterinarian promoting this test offers some hint of how he responds to positive results.

In my practice, animals with increased TK levels are supplemented with specific herbal remedies (Healthy Qi, CA Support, etc.) to support their immune systems. The blood is retested one month later. If TK levels are back to normal, no further testing is needed for four months. If TK levels continue to increase, then we must search further for the reason why it’s increasing. A persistent elevation of TK levels predicts the likelihood of cancer developing over the next six months.

So this doctor recommends responding to a positive test with the use of unproven, and untested, herbal products from the pre-scientific system of myths and metaphors known as Traditional Chinese Veterinary Medicine (TCVM). Given the implausible and unvalidated methods used in this system, the risks of untested and unregulated herbal supplements, and the fact that the concept of “boosting the immune system” is itself nonsense, this is certainly not a useful response. Of course, the plan proposed here also ignores the fact that values such as these vary with time for many reasons, and the fact that they might decrease when such remedies are used does not, of course actually mean the remedies are responsible, though that is undoubtedly the implication here.

I would like to be able to say that this approach is only that of this individual “holistic” veterinarian, and it does not necessarily imply that VDI has a similarly dubious and evidence-poor approach to the use of their products. However, among the informational materials VDI offers vets to help understand how to use this test is an article, entitled The Holistic View, by this same doctor. Clearly, the company is willing to endorse such dubious practices in the interest of marketing its products.

Simply following TK1 levels over time, if one chooses to measure them, may be a more reasonable strategy. Persistent elevations or increases over time are more likely to be meaningful than individual values. But instituting dubious treatment in response to an elevation is not rational. Neither is an extensive diagnostic workup looking for cancer. While the risks of diagnostic imaging, such as chest x-rays and abdominal ultrasounds, are low, these tests do have costs, they can find irrelevant abnormalities which would lead to unnecessary further testing and treatment, and they can generate false reassurance when they fail to find an abnormality. Effective use of such tests requires establishing a reasonable clinical index of suspicion which can help guide interpretation of the results and further actions.

This author makes similar recommendations for responding to elevated CRP levels, including the use of supplements of unproven or questionable value without a specific diagnosis or clear clinical purpose for making use of them.

In animals with elevated CRP or Hpt levels, specific nutritional supplements (antioxidants, fatty acids, etc.) are administered to reduce harmful inflammatory proteins. Diseases know to be associated with inflammation, such as dental disease, arthritis, and allergies, are diagnosed and treated appropriately to reduce inflammation in the body. The blood is retested in one month and CRP/Hpt levels should have returned to normal. If they haven’t, further investigation is undertaken.

And finally, the author makes a number of strong claims about the prevalence of Vitamin D deficiency and the need for supplementation that are not supported by real evidence or accepted by veterinary nutrition specialists.

While pet food is formulated with enough vitamin D3 to prevent deficiency disease (rickets), the levels are too low for maximum health. Testing shows most dogs have blood levels considered insufficient for optimal health, and would benefit from supplementation. Once the vitamin D3 test results are available, supplementation is given with the goal of shooting for a blood level of 100 mg/ml (in studies, animals with cancers tended to have vitamin D3 blood levels lower than 100.) Additional testing is done to confirm if the prescribed amount of vitamin D3, typically given once daily with food, is enough to reach the recommended blood level.

While there is certainly evidence that dogs with cancers, and other diseases, often have lower levels of Vitamin D than healthy dogs, as in humans it has not yet been clearly shown that the low levels increase the risk of developing disease or that supplementation prevents it. The specific number this author recommends as a target comes from a single study of dogs with a particular kind of cancer. And Vitamin D supplementation has risks as well, so recommending it based on inadequate data is not appropriate or helpful.

When screening tests of unproven value is recommended, and all abnormal results are taken as an indication to prescribe (i.e. sell) supplements that also have unclear value, it is difficult not to see profit as a significant motive in the marketing of this test. And while I don’t doubt this individual believes he is finding, and then treating or preventing cancer with this approach, this does not eliminate the potential influence of the obvious profit potential in this approach, nor excuse the lack of concern for the very weak evidence behind the practice.

A more evidence-based answer to the question of how to respond to a positive test result simply isn’t available, which is one of the major reasons why this sort of testing is not widely recommended. Until it is clear that the test consistently identifies real disease at a stage where it is both otherwise undetectable and can be more effectively treated, the testing and the diagnostic or therapeutic followup is based on little more than speculation or intuition, not sound, science-based medicine. This is also the answer to our third question….

Will Testing & the Followup Prevent Illness or Prolong Life?

The answer, unfortunately, is no one knows. The current evidence suggests some of these tests could lead to more effective detection and treatment of some cancers. However, as I mentioned earlier, this has not yet been proven to actually be true for humans or veterinary patients. The majority of promising ideas in medicine fail to fulfil their promise, which is why thorough scientific research is vital. Leaping on the bandwagon too early might occasionally work out well, but the odds are far better that it will lead to taking useless or even harmful actions that later evidence clearly shows we would be better off not having taken. If this particular testing strategy does turn out to work well for at least some patients, the folks promoting it will naturally look like prescient geniuses. If, as is more likely, it turns out to be less useful than hoped or even worthless or harmful, most people will probably not notice as we will likely move right on to the next promising idea. But a careful review of medical history suggests that we could make much better decisions, for ourselves and our pets, if we waited until the evidence was a lot stronger. After all, we are talking about testing healthy animals with no signs of disease! How much should we really put them through in an effort to prevent diseases they may not get using a test and therapies we don’t know actually work? Let’s not forget our last question…

Can These Tests Do Any Harm?

As I’ve already mentioned, overdiagnosis is a real, and serious, risk. The data shows that the PPV for this test, the proportion of positive results that are actually true, is quite low. So we are going to identify a lot of animals as having a disease they don’t have. If we put these animals through a lot of expensive diagnostic tests or, even worse, subject them to untested treatments as suggested by some proponents of this screening test, then we will generate a great deal of unnecessary stress for pets and their owners and unnecessary expense for some clients, and we may even harm some of the pets we are trying to help.

Of course, we will certainly detect at least some animals who actually do have cancer. Isn’t this a good thing, and doesn’t this balance the possible expense and harm we might do? Unfortunately, we don’t really know the answer to this either. Since we can’t be sure that detecting these cases earlier actually helps them, there is no way to balance that against the costs and risks of the testing.

Bottom Line

Screening tests, which try to identify hidden disease in apparently healthy animals, can sometimes be beneficial. If these tests are sufficiently accurate, and if we have truly useful treatments to offer the cases we find, then we can use such screening to reduce disease and suffering. However, at the moment there is no such screening blood test for cancer in dogs and cats.

The specific tests suggested here, thymidine kinase, C-reactive protein, and Vitamin D, do have some potential value in testing for cancer and monitoring cancer treatment. The evidence, however, is very preliminary and does not clearly tell us that these tests, alone or in combination, are sufficiently accurate to find most hidden cancers and not misidentify too many healthy animals as having cancer. Much more research needs to be done before these tests can be recommended as part of soundly scientific and evidence-based medicine.

And there is little evidence that if we had such a test we could actually help the pets with cancer we identified live longer or healthier lives. In addition to working on developing good screening tests for cancer in dogs and cats, we will need to find out which treatments, if any, can help these pets to live longer and stay healthier once we find them. Given unproven herbal remedies and supplements to these animals, even if the tests showing they have hidden disease are correct, does not benefit them and may do harm.

It is natural to want a clear, simply way to reassure ourselves our pets are healthy, and to want effective treatments to give them if we find they have cancer even though they appear healthy. However, taking the best care possible of our animal companions does not mean subjecting them to testing which is unreliable and therapies which aren’t proven to be safe and effective. Good preventative health care means scientific, evidence-based healthcare, not chasing after comforting promises without real evidence.

The best strategy for preventing disease in our pets is to feed them nutritionally complete diets, maintain them at a healthy, lean body condition, keep up with appropriate preventative care for infectious diseases and parasites, and provide them with appropriate socialization and training, physical and mental exercise, and love. While less dramatic in some ways that blood tests, supplements, and other kinds of constant tinkering with their physiology, this is still the best preventative health care we can give our pets.

 Bibliography

There are many basic science articles investigating TK!, CRP, and Vitamin D in dogs and cats (mostly dogs). This is intended as a sample of the literature to illustrate the state of the existing evidence. There have not yet been systematic reviews of these tests and compounds for companion animals, but plenty of other studies are indexed on PubMed.

1. Vet J. 2013 Sep;197(3):854-60. doi: 10.1016/j.tvjl.2013.05.036. Epub 2013 Jul 4.

High levels of inactive thymidine kinase 1 polypeptide detected in sera from dogs
with solid tumours by immunoaffinity methods: implications for in vitro
diagnostics.

Kiran Kumar J(1), Sharif H, Westberg S, von Euler H, Eriksson S.

Determination of serum thymidine kinase 1 (STK1) activity has been used as a
proliferation marker for neoplastic diseases in both human and veterinary
medicine. The purpose of this study was to determine STK1 activity and enzyme
levels in different dog tumours. Serum samples from three dogs with leukaemia,
five with lymphoma, 21 with solid tumours and 18 healthy dogs were analyzed for
STK1 activity, using an optimized [(3)H]-deoxythymidine (dThd) phosphorylation
assay, and for STK1 protein levels using an immunoaffinity/western blot assay.
STK1 activity in dogs with haematological tumours was significantly higher than
in the solid tumour and healthy dog groups (mean ± standard deviation [SD] = 65 ±
79, 1.1 ± 0.5, and 1.0 ± 0.4 pmol/min/mL, respectively). Serum samples were
analyzed after immunoaffinity isolation by western blot and the TK1 26 kDa band
intensities quantified revealing that concentrations were significantly higher in
dogs with haematological tumours and solid tumours compared to healthy dogs (mean
± SD=33 ± 12, 30 ± 13, and 10 ± 5 ng/mL, respectively). Pre-incubation with the
reducing agent dithioerythritol (DTE) showed a decrease in STK1 activity and
protein levels in most samples, but an increase of about 20% in sera from healthy
dogs and from those with haematological malignancies. Compared to animals with
solid tumours, the specific STK1 activity (nmol [(3)H]-deoxythymidine
monophosphate (dTMP)/min/mg of TK1 protein of 26 kDa) was 30-fold higher in
haematological malignancies and 2.5-fold higher in healthy dogs, respectively.
The results demonstrate that there is a large fraction of inactive TK1 protein,
particularly in sera from dogs with solid tumours. The findings are important in
the use of STK1 as a biomarker.
2. Vet Comp Oncol. 2013 Jul 16. doi: 10.1111/vco.12052. [Epub ahead of print]

Serum thymidine kinase 1 and C-reactive protein as biomarkers for screening
clinically healthy dogs for occult disease.

Selting KA(1), Sharp CR, Ringold R, Knouse J.

Thymidine kinase (TK1) is a biomarker that correlates well with diagnosis and
prognosis in certain canine cancers. Canine C-reactive protein (cCRP) is a widely
accepted marker of inflammation correlated with increased risk and severity of
various diseases. We evaluated serum TK1 and cCRP concentrations in apparently
healthy dogs (n?=?360). All dogs were followed up for a minimum of 6?months by
health questionnaire. All dogs with cancer were identified using a proprietary
dual-biomarker algorithm [termed Neoplasia Index (NI)]. Specificity of positive
NI is 0.91 and high positive is 0.98. All-cause mortality was 20% in dogs with
elevated cCRP and 3% in dogs with low cCRP. The performance of serum TK1 and cCRP
as tools for screening for occult cancer is improved when evaluated together.
Serum TK1 and cCRP (unified in the NI) are useful in the screening of occult
canine cancer. cCRP is useful in screening for other serious diseases.

3. Vet Comp Oncol. 2013 Mar;11(1):1-13. doi: 10.1111/j.1476-5829.2011.00296.x. Epub
2011 Dec 8.

Thymidine kinase assay in canine lymphoma.

Elliott JW(1), Cripps P, Blackwood L.

The aim of the study was to evaluate if thymidine kinase (TK) correlated with
duration of first remission (DFR) or survival in dogs with lymphoma and if
initial TK levels correlated with stage and substage; and also to assess if TK
level at diagnosis correlated with immunophenotype. TK was assayed in 73 dogs
with treatment naïve lymphoma, then again after treatment; 47% had an initial TK
above the reference interval. Dogs with B-cell lymphoma had higher initial TK
activities than dogs with T-cell lymphoma. TK levels were not higher in dogs with
higher stage disease and TK activity prior to treatment was not associated with
DFR or survival. Where TK was elevated at diagnosis, it fell into the reference
range during remission. TK was normal in 53% dogs at diagnosis, which is higher
than previously reported. Further studies are warranted to assess the utility of
TK in dogs with lymphoma.

4. J Feline Med Surg. 2013 Feb;15(2):142-7. doi: 10.1177/1098612X12463928. Epub 2012
Oct 17.

Serum thymidine kinase activity in clinically healthy and diseased cats: a
potential biomarker for lymphoma.

Taylor SS(1), Dodkin S, Papasouliotis K, Evans H, Graham PA, Belshaw Z, Westberg
S, von Euler HP.

The thymidine kinases are enzymes that convert deoxythymidine to deoxythymidine
monophosphate and have a function in DNA synthesis. Rapidly proliferating cells
will have higher levels of thymidine kinase. Serum thymidine kinase activity
(sTK) is a useful tumour marker in humans and dogs, with utility as a prognostic
indicator in lymphoma. In the current study serum samples were collected from 49
clinically healthy cats, 33 with lymphoma, 55 with inflammatory disease and 34
with non-haematopoietic neoplasia (NHPN). sTK was measured using a radioenzyme
assay and a reference interval (1.96 × SD) was established from the clinically
healthy cats (<5.5 U/l). Mean sTK activity for healthy cats was 2.2 U/l (range
0.8-8.4, ± SD 1.7). Mean sTK activity for cats with lymphoma was 17.5 U/l (range
1.0-100.0 SD ± 27.4). Mean sTK activity for cats with NHPN was 4.2 U/l (range
1.0-45.0, SD ± 8.6). Mean sTK activity for the inflammatory group was 3.4 U/l
(range 1.0-19.6, SD 3.9). Cats with lymphoma had significantly higher sTK
activity than healthy cats or cats with inflammatory disease (P <0.0001) and cats
with NHPN (P <0.0002). sTK activity is a potentially useful biomarker for feline
lymphoma and further study is required to assess its utility as a prognostic
indicator.

5. Vet Comp Oncol. 2012 Dec;10(4):292-302. doi: 10.1111/j.1476-5829.2011.00298.x.
Epub 2011 Oct 20.

Elevated serum thymidine kinase activity in canine splenic hemangiosarcoma*.

Thamm DH(1), Kamstock DA, Sharp CR, Johnson SI, Mazzaferro E, Herold LV, Barnes
SM, Winkler K, Selting KA.

Thymidine kinase 1 (TK1) is a soluble biomarker associated with DNA synthesis.
This prospective study evaluated serum TK1 activity in dogs presenting with
hemoabdomen and a splenic mass. An ELISA using azidothymidine as a substrate was
used to evaluate TK1 activity. Sixty-two dogs with hemoabdomen and 15 normal
controls were studied. Serum TK1 activity was significantly higher in dogs with
hemangiosarcoma (HSA) than in normal dogs (mean ± SEM = 17.0 ± 5.0 and 2.01 ±
0.6, respectively), but not dogs with benign disease (mean ± SEM = 10.0 ± 3.3).
Using a cut-off of 6.55 U/L, TK activity demonstrated a sensitivity of 0.52,
specificity of 0.93, positive predictive value of 0.94 and negative predictive
value of 0.48 for distinguishing HSA versus normal. When interval thresholds of
<1.55 and >7.95 U/L were used together, diagnostic utility was increased. Serum
TK1 evaluation may help to discriminate between benign disease and HSA in dogs
with hemoabdomen and a splenic mass.
6. Int J Oncol. 2009 Feb;34(2):505-10.

Monitoring therapy in canine malignant lymphoma and leukemia with serum thymidine
kinase 1 activity–evaluation of a new, fully automated non-radiometric assay.

Von Euler HP(1), Rivera P, Aronsson AC, Bengtsson C, Hansson LO, Eriksson SK.
Thymidine kinase 1 (TK), which is involved in the synthesis of DNA precursors, is
only expressed in S-G2 cells. Serum TK levels correlate to the proliferative
activity of tumor disease. Determinations of TK levels have so far relied on
radio enzyme assay (REA) and experimental ELISA methods, which have limited the
clinical use of this biomarker, although recent studies in dogs with malignant
lymphoma (ML) demonstrate its wide potential. A non-radiometric method based on a
competitive immunoassay with specific anti-3′-azido-deoxythymidine monophosphate
(AZTMP) antibodies has been further developed into the fully automated Liaison TK
assay (DiaSorin). Sera from healthy dogs (n=30), and dogs with leukemia (LEUK)
(n=35), ML (n=84), non-hematological tumors (n=50), and inflammatory disease
(n=14) were tested using both methods. Lymphoma and LEUK samples were available
before and during chemotherapy. The coefficients of variation for the Liaison TK
assay in this study were 6.3 and 3.4% (low/high TK, respectively), and the
correlation between TK REA (X) and the Liaison TK assay (Y) was y=0.9203x+1.3854
(R2=0.9501). The TK1 levels measured during chemotherapy gave very clear
differences between dogs in complete remission and dogs out of remission. A
Tukey-Kramer analysis showed that all LEUKs and MLs out of remission differed
significantly from the other groups. The Liaison TK assay showed high precision,
high sensitivity and a good correlation to the TK REA. The Liaison TK assay
provides valuable clinical information in the treatment and management of canine
LEUK and ML, with a potential to be further validated in human trials.

7. J Vet Intern Med. 2004 Sep-Oct;18(5):696-702.

Serum thymidine kinase activity in dogs with malignant lymphoma: a potent marker
for prognosis and monitoring the disease.

von Euler H(1), Einarsson R, Olsson U, Lagerstedt AS, Eriksson S.

Serum thymidine kinase (sTK) activity was evaluated as a tumor marker for canine
malignant lymphoma (ML). The objective was to investigate if sTK, as in humans,
could be used as a prognostic marker for survival time in dogs with ML and if sTK
could identify early signs of progression of disease in treated dogs. Serum
samples from 52 dogs with ML were tested for initial TK activity. Samples from 21
normal dogs and 25 dogs with nonhematologic neoplasms were used for comparison.
Forty-four dogs with ML were treated. Serum TK activity was measured in treated
dogs before each treatment and every 4 weeks thereafter until relapse. Dogs with
ML had 2-180 times higher TK activity (TK 5-900 U/L) than normal dogs (TK <7 U/L)
based on the mean + 2 standard deviations. In the group of other neoplasms, only
2 dogs had a moderate increase (6.4 and 7.5 U/L) compared with the controls. Mean
sTK activities in the dogs with ML that had gone into complete remission (CR)
were not significantly different from activities in healthy controls (P = .68).
Mean sTK at least 3 weeks before and at the time of relapse was significantly
higher than activity measured at CR (P < .0001). Dogs with ML that initially had
sTK >30 U/L had significantly shorter survival times (P < .0001). Furthermore,
sTK activity reflected the clinical staging of ML. Measuring sTK can be used as a
powerful objective tumor marker for prognosis and for predicting relapse before
recurrence of clinically detectable disease in dogs with ML undergoing
chemotherapy.

8. J Vet Intern Med. 2004 Sep-Oct;18(5):595-6.

Serum thymidine kinase activity: an alternative to histologic markers of cellular
proliferation in canine lymphoma.

Madewell BR.

Thymidine kinase (TK) is a cellular enzyme which is involved in a ‘salvage
pathway’ of DNA synthesis. It is activated in the G1/S phase of the cell cycle,
and its activity has been shown to correlate with the proliferative activity of
tumor cells…. Clinical studies have reported high serum TK concentrations in a
variety of neoplasias. The majority of these studies concerned hematological
malignancies. TK seems to be a useful marker in non-Hodgkin’s lymphoma, where it
correlates with clinical staging and provides marked prognostic information on
(progression-free) survival.

PMID: 15515571  [PubMed – indexed for MEDLINE]
9. J Vet Med Sci. 1997 Oct;59(10):957-60.

Plasma thymidine kinase activity in dogs with lymphoma and leukemia.

Nakamura N(1), Momoi Y, Watari T, Yoshino T, Tsujimoto H, Hasegawa A.

Plasma thymidine kinase (TK) activity was evaluated as a plasma marker for canine
lymphoma and leukemia. A tentative “cut-off” value was set at 6.0 U/l as the
upper level of plasma TK based on the mean + 2SD of plasma TK activity in 13
clinically healthy dogs. The levels of plasma TK activity in all of the 20 dogs
with lymphoma and leukemia were higher than the cut-off value, whereas those in
dogs with lymphoma decreased in parallel with the reduction of the tumor mass
after chemotherapy. These findings suggested that estimation of plasma TK
activity can be used as a plasma marker for lymphoma and leukemia in the dog.

10. Vet Comp Oncol. 2014 Jul 8. doi: 10.1111/vco.12101. [Epub ahead of print]

Serum 25-hydroxyvitamin D concentrations in dogs – correlation with health and
cancer risk.

Selting KA(1), Sharp CR, Ringold R, Thamm DH, Backus R.

25-hydroxyvitamin D (25(OH)D) is important in bone health as well as many
diseases including cancer. Supplementation may increase responsiveness of cancer
cells to chemotherapy. Serum 25(OH)D, intact parathyroid hormone (iPTH) and
canine C-reactive protein (c-CRP) were measured in healthy dogs and dogs with
haemoabdomen. Regression analysis determined optimal 25(OH)D concentrations. In
healthy dogs (n = 282), mean iPTH concentrations correlated inversely (r(2) =
0.88, P < 0.001) to 25(OH)D concentrations. Variation in both iPTH and c-CRP
plateaued at 25(OH)D concentrations of 100-120 ng mL(-1) . Haemoabdomen dogs (n =
63, 43 malignant and 20 benign) had 25(OH)D concentrations ranging from 19.4 to
>150 ng mL(-1) . Relative risk of cancer increased with decreasing 25(OH)D
concentrations [RR = 3.9 for 25(OH)D below 40 ng mL(-1) (P = 0.0001)]. Serum
25(OH)D concentrations in dogs vary widely, and are influenced by dietary VitD
content. Serum vitD measurement can identify dogs for which supplementation may
improve health and response to cancer therapy.

11. Vet Parasitol. 2014 Jun 16;203(1-2):153-9. doi: 10.1016/j.vetpar.2014.02.001.
Epub 2014 Feb 25.

Serum acute phase protein concentrations in dogs with spirocercosis and their
association with esophageal neoplasia – a prospective cohort study.

Nivy R(1), Caldin M(2), Lavy E(1), Shaabon K(1), Segev G(1), Aroch I(3).

Spirocerca lupi, the dog esophageal worm, typically induces formation of
esophageal nodules, which may transform to sarcoma. Ante mortem discrimination
between benign and malignant esophageal masses is challenging. Serum acute phase
proteins (APPs) are utilized in diagnosis and prognosis of various canine
diseases as markers of inflammation. This study characterized serum APPs
concentrations in dogs with benign and malignant esophageal spirocercosis and
evaluated their accuracy in differentiating benign from malignant lesions.
Seventy-eight client-owned dogs with esophageal spirocercosis were included.
Serum C-reactive protein (CRP), haptoglobin, serum-amyloid A (SAA) and albumin
concentrations were measured upon diagnosis and follow-up visits, and compared
with healthy dogs, and between malignant and benign cases. Haptoglobin, CRP and
SAA concentrations were higher, and albumin concentration was lower (P<0.001 for
all) in infected dogs compared to healthy controls. Dogs with suspected neoplasia
had significantly higher CRP (P=0.011), haptoglobin (P=0.008) and SAA (P=0.05),
and lower albumin (P=0.012) concentrations compared to dogs with benign
esophageal nodules. APPs moderately discriminated between suspected malignant and
benign esophageal disease. None of the dogs with suspected neoplasia had
concurrent normal concentrations of all APPs. The present results indicate that
canine spirocercosis is characterized by an acute phase reaction, both at
presentation and during treatment. When concentrations of all four APPs are
within reference range, esophageal malignancy is highly unlikely. Although
concentrations of all positive APPs were significantly higher in suspected
neoplastic cases compared to benign ones, moderate discriminatory power limits
their clinical use. Neither APP was useful to monitor response to treatment.
12. Vet Rec. 2012 Jun 23;170(25):648. doi: 10.1136/vr.100401. Epub 2012 Jun 1.

Acute phase protein levels in dogs with mast cell tumours and sarcomas.

Chase D(1), McLauchlan G, Eckersall PD, Pratschke J, Parkin T, Pratschke K.

The acute phase proteins (APP) form part of a non-specific host response to
inflammation. They may be induced by a range of different causes, including
infection, inflammation, cancer and trauma. As they form part of the earliest
response to such insults, they have potential for early identification of
disease. In people, APP levels have been shown to correlate both with the extent
of disease and also the prognosis in several forms of neoplasia, including
prostate, oesophageal and colorectal cancer. As such, they can be used as
prognostic and monitoring tools. To date, similar studies in veterinary patients
have been limited, largely retrospective in nature and many are non-specific for
tumour type. The purpose of this study was to evaluate a panel of four APPs in
dogs with naturally occurring mast cell tumours (MCTs) and sarcomas to identify
in the first instance whether increased levels of individual APPs, or
identifiable combinations of APPs, was linked with the presence of disease. In
the patients with MCTs, C-reactive protein (CRP) and a-1 acid glycoprotein levels
increased, with a concurrent drop in serum amyloid A levels. In the sarcoma
patients, CRP, a-1 acid glycoprotein and haptoglobin were increased. These
findings suggest that specific solid tumour types in dogs may be associated with
specific changes in APP profiles.
13. Vet Clin Pathol. 2009 Sep;38(3):348-52. doi: 10.1111/j.1939-165X.2009.00139.x.
Epub 2009 Apr 16.

Evaluation of serum haptoglobin and C-reactive protein in dogs with mammary
tumors.

Planellas M(1), Bassols A, Siracusa C, Saco Y, Giménez M, Pato R, Pastor J.

Author information:
(1)Department of Animal Medicine and Surgery, Universitat Autonoma de Barcelona,
Barcelona, Spain. marta.planellas@uab.cat

BACKGROUND: In veterinary medicine, there is increasing interest in measuring
acute phase proteins as a tool in the diagnosis and monitoring of neoplastic
diseases. Although mammary neoplasms are the most common type of cancer in dogs,
acute phase proteins have not been extensively evaluated in dogs with mammary
tumors.
OBJECTIVES: The aim of this study was to evaluate serum haptoglobin (Hp) and
C-reactive protein (CRP) concentrations in the dogs with mammary tumors and
assess their potential association with malignancy.
METHODS: A retrospective study of dogs with mammary tumors was performed. Serum
concentrations of CRP and Hp were determined in healthy control dogs (n=20) and
dogs with mammary tumors before surgery (n=41). Mammary tumors were grouped as
carcinomas (n=24), fibrosarcoma (n=1), malignant mixed tumors (n=7), benign mixed
tumors (n=6), and adenomas (n=3). CRP and Hp concentrations were compared in dogs
with different tumor types and were also compared based on tumor size, lymph node
infiltration, skin ulceration, fixation to underlying tissue, and time between
tumor identification and removal.
RESULTS: Hp concentration was significantly (P<.043) higher in dogs with mammary
tumors (median 2.03 g/L, range 0.09-2.94 g/L) compared with controls (1.38 g/L,
range 0.08-3.00 g/L), but the range of values overlapped considerably. CRP
concentration was higher in dogs with carcinomas (4.70 mg/L, range 0.63-128.96
mg/L) vs controls (2.11 mg/L, range 0.25-6.57 mg/L) (P=.0008) and in dogs with
ulcerated skin (14.8 mg/L, range 5.7-128.9 mg/L, n=3) compared with those without
ulceration (2.4 mg/L, range 0.11-30.3 mg/L, n=38) (P=.048).
CONCLUSIONS: Serum Hp and CRP do not appear to have value in diagnosing or
predicting malignancy of mammary tumors in dogs. Higher CRP concentrations in
dogs with mammary carcinoma suggest a role for inflammation in this tumor type.

PMID: 19392756  [PubMed – indexed for MEDLINE]
14. J Vet Intern Med. 2007 Nov-Dec;21(6):1231-6.

Serum C-reactive protein concentration as an indicator of remission status in
dogs with multicentric lymphoma.

Nielsen L(1), Toft N, Eckersall PD, Mellor DJ, Morris JS.

BACKGROUND: The acute-phase protein C-reactive protein (CRP) is used as a
diagnostic and prognostic marker in humans with various neoplasias, including
non-Hodgkin’s lymphoma.
OBJECTIVE: To evaluate if CRP could be used to detect different remission states
in dogs with lymphoma.
ANIMALS: Twenty-two dogs with untreated multicentric lymphoma.
METHODS: Prospective observational study. Blood samples were collected at the
time of diagnosis, before each chemotherapy session, and at follow-up visits,
resulting in 287 serum samples.
RESULTS: Before therapy, a statistically significant majority of the dogs (P =
.0019) had CRP concentrations above the reference range (68%, 15/22). After
achieving complete remission 90% (18/20) of the dogs had CRP concentrations
within the reference range, and the difference in values before and after
treatment was statistically significant (P < .001). CRP concentrations of dogs in
complete remission (median, 1.91; range, 0.2-103) were significantly different (P
= .031) from those of dogs with partial remission (median, 2.48; range, 0-89),
stable disease (median, 1.77; range, 1.03-42.65), or progressive disease (median,
8.7; range, 0-82.5). There was profound variation of CRP measurements within each
dog.
CONCLUSIONS: CRP is useful in determining complete remission status after
treatment with cytotoxic drugs. However, the individual variation between dogs
means CRP concentration is not sufficiently different in other remission states
to permit its use in monitoring progression of the disease. Greater reliability
in determining remission status might be achieved by combining CRP concentration
with other serum markers.

15. Vet J. 2007 Jul;174(1):188-92. Epub 2006 Aug 9.

Changes in C-reactive protein and haptoglobin in dogs with lymphatic neoplasia.

Mischke R(1), Waterston M, Eckersall PD.

Acute phase proteins (APP) are regarded as a useful diagnostic tool in humans
with lymphomas, leukaemias and multiple myeloma. C-reactive protein (CRP) and
haptoglobin concentrations were measured in dogs with malignant multicentric
(high grade) lymphoma (n=16), acute lymphoblastic leukaemia (ALL) (n=11), chronic
lymphocytic leukaemia (CLL) (n=7) and multiple myeloma (n=8). Twenty-five healthy
dogs served as controls. Measurements of the CRP plasma concentration were
performed using a commercial ELISA and haptoglobin was measured with an assay
based on its haemoglobin binding capacity. Global group comparisons using
Kruskal-Wallis-test revealed significant group differences for both APPs
(P<0.0001). Median CRP concentrations were increased in all groups with
neoplastic lymphatic disorders (lymphoma: 37.2mg/L, ALL: 47.8mg/L, CLL: 35.5mg/L,
myeloma: 17.6mg/L) compared to controls (1.67mg/L; P<0.001). Compared to the
healthy controls (median=0.59g/L), haptoglobin was especially increased in dogs
with ALL (6.8g/L, P<0.0001) followed by dogs with malignant lymphoma (3.8g/L,
P<0.0001), CLL (3.2g/L, P=0.0008), and multiple myeloma (3.0g/L, P=0.0163). For
both APPs, a wide range of values was found in all patient groups. The results
indicate that particularly severe and acute lymphatic neoplasia, such as high
grade lymphoma and ALL, cause significant acute phase reactions in dogs and must
be included in the differential diagnoses of increased blood levels of these
APPs.

16. J Am Vet Med Assoc. 2007 Feb 15;230(4):522-6.

Serum C-reactive protein concentrations in dogs with multicentric lymphoma
undergoing chemotherapy.

Merlo A(1), Rezende BC, Franchini ML, Simões DM, Lucas SR.

OBJECTIVE: To determine whether serum C-reactive protein (CRP) concentration is
high in dogs with multicentric lymphoma, whether CRP concentration changes in
response to chemotherapy, and whether CRP concentration can be used as a marker
for relapse in dogs with multicentric lymphoma.
DESIGN: Cohort study.
ANIMALS: 20 dogs with multicentric lymphoma and 8 healthy control dogs undergoing
chemotherapy with cyclophosphamide, vincristine, and prednisone (CVP) or with
vincristine, cyclophosphamide, methotrexate, and L-asparaginase (VCMA) and 20
other healthy dogs.
PROCEDURES: Serum CRP concentration was measured weekly during the first month of
chemotherapy and then at 3-week intervals until relapse in dogs with multicentric
lymphoma, weekly for 16 weeks in healthy dogs undergoing chemotherapy, and once
in the healthy dogs not undergoing chemotherapy.
RESULTS: For both groups of dogs with lymphoma, mean serum CRP concentration
during week 1 (prior to treatment) was significantly higher than mean
concentrations following induction of chemotherapy and at the time of relapse.
Mean serum CRP concentration in the healthy dogs undergoing chemotherapy was not
significantly different at any time from mean concentration for the healthy dogs
not undergoing chemotherapy. No significant differences were observed between
dogs treated with CVP and dogs treated with VCMA.
CONCLUSIONS AND CLINICAL RELEVANCE: Results suggest that serum CRP concentration
is high in dogs with multicentric lymphoma but that serum CRP concentration is
not a useful marker for relapse and that chemotherapy itself does not affect
serum CRP concentration.

19. J Vet Intern Med. 2005 Nov-Dec;19(6):865-70.

Preliminary studies of serum acute-phase protein concentrations in hematologic
and neoplastic diseases of the dog.

Tecles F(1), Spiranelli E, Bonfanti U, Cerón JJ, Paltrinieri S.

Serum concentrations of acute-phase proteins (APPs): haptoglobin (Hp),
ceruloplasmin (Cp), serum amyloid A (SAA), and C-reactive protein (CRP) were
determined in healthy dogs (n = 15) and dogs with different diseases grouped as
acute inflammation (I, n = 12), hematologic neoplasias (HT, including leukemia
and lymphoma, n = 16), nonhematologic neoplasias (NHT, including epithelial,
mesenchymal, and mixed, n = 20), and autoimmune hemolytic anemia (AIHA, n = 8).
SAA and CRP were analyzed using commercially available enzyme-linked
immunosorbent assay (ELISA) kits, and Hp and Cp were measured using colorimetric
methods, all previously validated for use in dogs. Increased concentrations of
all APPs were observed in all groups of diseased dogs, but statistical
significance only was observed with Hp (I, P < .001; HT, P < .05), Cp (I, P <
.05; AIHA, P < .01), and CRP (I, P < .001; HT, P < .001; AIHA, CRP P < .05). High
variability in individual APPs within each group of diseases was found with no
significant differences between leukemia and lymphoma as well as among different
types of neoplasia. The AIHA group had smaller increases in Hp, SAA, and CRP but
higher concentrations of Cp. When follow-up of individual cases was possible, a
decrease in APPs generally was found in cases with favorable outcome. The results
of this study suggest that neoplasia and hematologic diseases such as AIHA should
be considered as possible causes of mild increases in APPs in dogs. Measurement
of APPs may be helpful to assess clinical evolution and monitor treatment of
these processes.

20. Biochem Biophys Res Commun. 2003 Jan 31;301(1):212-7.

The canine mast cell activation via CRP.

Fujimoto T(1), Sato Y, Sasaki N, Teshima R, Hanaoka K, Kitani S.

We report here canine mastocytoma-derived cell (CMMC) activation via two
pentraxin, limulus- and human-CRP. Mast cell chemotaxis was measured by Boyden’s
blindwell chamber. To confirm that the cell migration was chemotactic,
“checkerboard” analysis was performed. We used Fura-2 to investigate CRP-mediated
cytosolic calcium elevation. To examine whether CRP-induced stimulation is
mediated through G-proteins, CMMC were incubated with pertussis toxin (PTx)
before use in chemotaxis assay and Ca(2+) mobilization. CMMC migration in
response to CRP was both chemokinetic and chemotactic. Limulus-CRP induced a
transient Ca(2+)-mobilization dose-dependently. Preincubation of the cells with
PTx inhibited CRP chemotaxis and Ca(2+)-mobilization, suggesting that G-proteins
of the Gi-class are involved in the chemotaxis. We suggest that CRP may
participate in the migration of mast cells to inflamed tissues during an
acute-phase response. CRP-mediated recruitment of mast cells might play an
important role in hypersensitivity and inflammatory processes.

 

 

An interesting article appeared recently in a widely-read veterinary journal that discusses so-called “integrative medicine” in veterinary cancer care in terms of evidence-based medicine (EBM). This article raises a number of interesting questions and concerns.

Raditic, DM. Bartges, JW. Evidence-based integrative medicine in clinical veterinary oncology. Vet Clin Small Animal 2014;44:831-853.

I’ve written about the notion of “integrative veterinary medicine” (IVM) before, and it is a highly problematic concept. These authors define it as “the use of complementary and alternative medicine (CAM) with conventional Western medicine systems.” This already contains some fallacies in that it labels science-based or “conventional” medicine as “Western,” which ignores the fact that the scientific approach to health and disease that underlies modern medicine has been developed and very successfully employed in all cultures around the world. While much of the development of modern science initially took place in Europe and North America, critical contributions to scientific medicine have come from many other places. And there is nothing intrinsically “Western” about scientific medicine. It has come to dominate medical practice around the world because of its demonstrable success, crossing cultural barriers rapidly and continuously.

This label also implies that the alternatives to conventional scientific medicine are fundamentally culturally distinct in origin and character. This may be true for some, such as Traditional Chinese Medicine (itself somewhat misleadingly named), but it is certainly not true for chiropractic, homeopathy, and many herbal and nutritional therapies which are include under the CAM umbrella and yet which were invented and practiced predominantly in Europe and North America.

In any case, defining IVM and the use of CAM along with conventional medicine begs the question of what CAM itself really is. As I’ve discussed before, it is largely an ideological category with little scientific meaning. It exists in part because of fundamental philosophical differences between CAM practitioners and proponents of science-based medicine, and partly to protect specific therapies in which proponents have personal faith from criticism and rejection regardless of the state of the scientific evidence concerning them. To call something CAM is not to define it, since many of the therapies included under the label are mutually incompatible in fundamental ways and have very different evidentiary profiles. The label signifies a particular kind of attitude and belief, not a real difference from the perspective of how or if one can demonstrate these therapies to be safe and effective using scientific means.

The authors further expand on their definition of IVM by defining CAM in terms of specific therapies they would include in the category and, tellingly, as anything “that are rational and supported by evidence to alleviate physical and emotional symptoms, improve quality of life (QOL), and possibly improve adherence to oncology treatment regimens.” This definition contains within it the assumption that these therapies are rational and supported by evidence. However, this often turns out not to be true for many of the practices included in the CAM fold, so this cannot be justly blithely assumed.

The authors then go on to justify the importance of knowledge about CAM therapies by pointing out some CAM therapies are widely employed. I agree, and I think it is vital that clinicians understand what the category itself is about and what the evidence says about specific therapies, though I suspect my conclusions would be quite different from those of the authors. They also acknowledge that the major drive for use of these therapies is psychological rather than based on any objective reason to expect they are safe and effective:

…human oncology patients use natural products to empower themselves, attempt to take control of their health… logically pet owners would apply these same emotions.

Once again, I agree, and I think it is important that veterinarians understand the roots of CAM use among our clients. This doesn’t mean, as the authors appear to imply, that we must support or encourage these practices, only that we must be knowledgeable about them and effective in understanding and communicating with our clients about their use. Often, if we are practicing truly evidence-based medicine, this communication is likely to involve discouraging the use of therapies for which there is no good reason to expect they are safe and effective.

The authors go on to acknowledge that the research evidence, particularly the most useful types of research, clinical trials, is scarce for most CAM practices. However, they imply that the reason for this is not the lack of a belief on the part of CAM practitioners that their therapies need scientific evaluation but simply a lack of funding for therapies that are not likely to be profitable:

Evidence-based research on CAM in an IM plan in veterinary clinical oncology is scarce, which is expected because large-scale research funding is typically provided for projects with potential for profits, such as with new, patented drugs.

This would seem to be contradicted by their own admission that, “In 2010, US herbal supplement sales exceeded $5.2 billion.” The reality is that CAM therapies are often very profitable. The reason herb and supplement manufacturers invest far less in scientifically evaluating the safety and efficacy of their products than to pharmaceutical companies is not because there is no profit to be made in these products but simply because the government does not require them to. Pharmaceutical products must demonstrate safety and efficacy to a certain, albeit imperfect, standard set by the FDA. Politics and ideology have led, however, to a different and far more lax standard for herbs and supplements, which are often assumed to be safe despite the abundant evidence that they can cause as much harm as conventional medicines.

In the absence of much clinical trial evidence, the authors place greater emphasis on less reliable kinds of evidence, such as in vitro and animal model studies and extrapolation across species. They hold out particular hope that metabolomics, the study of how the compounds and chemicals involved in metabolism are affected by particular diseases and interventions, such as herbs and supplements. While a great deal of useful knowledge can be gained by this kind of research, the study of such proxy markers for clinically relevant outcomes cannot substitute for clinical trials. The response of cancer cells in a petri dish or in a mouse to a purported cancer therapy have never proven to be very reliable indicators of the actual clinical response of cancer patients to the same therapy.

The rest of the article is largely a review of selected clinical and pre-clinical research on a few herbs, dietary supplements, and acupuncture. This is fine as far as it goes, but like any narrative review the only studies mentioned are the ones the authors, who are avowed proponents of such therapies, choose to discuss, so one has to bear in mind they may not be representative of the real state of the scientific evidence. In any case, what strikes me about this part of the paper is that the authors are describing the process of scientific evaluation of potential medical therapies from in vitro and animal model studies up through clinical trials. How, exactly, is there anything complementary or alternative about this?

What they describe is essentially the process of scientific, “Western” medicine, which routinely investigates the therapeutic and preventative value of such interventions. The only reason to call any of this alternative or distinguish it from conventional scientific medicine is to suggest that the use of these therapies should be somehow viewed differently from other medicines such as pharmaceuticals. The usual next step, which the authors never make explicit here, is to claim that these therapies should be open to use regardless of the state of the scientific evidence. They do, however, suggest that rather than evaluating the science for themselves as they would ordinarily do, veterinary oncologists should rely on the expertise and personal experience of veterinarians already using such therapies and already convinced of their value despite the lack of scientific evidence.

Collaboration between conventional oncologists and practitioners of IM, who have knowledge, experience, and training to use HDS and acupuncture, is needed to explore the possibilities of integrative veterinary…Although individual herbs are discussed here, integrative practitioners more often use herbal formulas and have identified safe, reliable sources of HDS products with known content; they also know possible interactions and understand dosing to prevent adverse effects. Veterinarians with this special training, education, and experience can be found through the organizations in Box 1.

The suggestion here is that these remedies can be viewed as “safe and reliable” and that adverse effects can be predicted and avoided despite the lack of scientific research simply because these veterinarians have personal experience in their use. This illustrates the underlying theme that occurs again and again in apologia for CAM. Proponents use language like “evidence-based” and promote supportive scientific research because these things have marketing value; they generate acceptance and the appearance of scientific validity. Yet in their hearts, the lack of scientific evidence for safety and efficacy does not diminish a confident belief in the safety and efficacy of these therapies. Science is a nice extra, something to convince the skeptics, but it isn’t seen as fundamentally necessary to the evaluation and safe use of these therapies. This attitude is incompatible with evidence-based medicine despite the appropriation of EBM terminology.

I appreciate the authors’ acknowledgement of the lack of convincing evidence to support the use of most CAM therapies in veterinary cancer care. And I agree that despite this, there is interest in such therapies on the part of clients. Veterinarians absolutely should be knowledgeable about these practices so they can guide and advise their clients. However, the appropriate response to the lack of good evidence for these practices is not to suggest that it cannot be obtained or that we can comfortably rely on individual experience and expertise instead.

Ultimately, for potential therapies with reasonable plausibility, the usual practice of science should be followed. Pre-clinical testing followed, if the evidence warrants, by clinical trials. Any use prior to the development of strong scientific evidence should be viewed as experimental, as likely to harm and to help the patient, and this should be clearly disclosed to clients. There is no reason to treat these therapies differently from any others or to give them special exemption from the usual standards of evidence.

 

 

By far the most common response to those posts I write questioning the claims for medical therapies advertised on the internet is a flood of testimonials from people who believe the scientific evidence is wrong or irrelevant because the product seemed to work for them and because so many people report it works. There are many reasons why anecdotes aren’t trustworthy, and I’ve written about this at length before. As psychologically compelling as our own experience are, and as hard-wired as we are to appreciate and believe stories more than data, the truth is that our uncontrolled observations are deeply unreliable.

Cognitive biases are a particularly significant part of the problem. These are little mental quirks and shortcuts that lead us to the wrong conclusions and to bad decisions. I’ve written about the effect of these biases on veterinary clinical decision making, and many, many books have been written about the specific details of these biases, how they work, and what effects they have on our reasoning**

A recent study has produced some specific and compelling evidence for the degree to which internet reviews of medical products grossly overestimate the actual value of the treatments being reviewed. The bottom line is that people who have positive experiences are far more likely to write testimonials than people who don’t, which creates a false impression of how well a therapy works, if it does at all.

Mícheál de Barra; Kimmo Eriksson; Pontus Strimling. How Feedback Biases Give Ineffective Medical Treatments a Good Reputation. J Med Internet Res 2014;16(8):e193)

The study first compared reviews of a diet plan on Amazon with results of clinical research on the same diet.

After 6 months on the diet, 93% (64/69) of online reviewers reported a weight loss of 10 kg or more while just 27% (19/71) of clinical trial participants experienced this level of weight change.

This figure shows the results of three clinical studies compared with those reported on Amazon. Clearly, the studies found consist results far less dramatic than those indicated by the online testimonials.

The same phenomenon was found when evaluating a dietary supplement/herbal product supposed to improve fertility.

And when the investigators evaluated the influence of reviews on decision-making, they found that positive reviews did affect people’s choice of a weight loss plan.

The author’s conclusions were:

We found that the reputed benefit of weight loss diets and fertility treatments is larger than the real benefit, apparently because people with typical or poorer outcomes are less inclined to tell others about their experiences. Thus, the real-world reputation of medical treatments seems to be subject to a reporting bias akin to the publication bias toward positive results that is seen in scientific research. Moreover, we found the resultant reputation distortion to be large enough to influence people’s decisions about which diet to begin…

Researchers have pointed out that several processes make it very difficult to identify benefits and harms of medical treatments when data are not systematically collected. In particular, treatments with no direct effect will sometimes appear effective because of the statistical phenomenon known as regression to the mean and the physiological phenomenon known as the placebo effect. It has also been suggested that treatments that prolong illness may, perversely, spread better because they are “demonstrated” for a longer period than effective treatments. Here, we have explored an additional mechanism, reporting bias, and its logical consequence: when people with poor outcomes remain silent, the reputed benefit of a treatment will exceed its real effect.

Though there are already more than enough nails in the coffin of the idea that anecdotes and testimonials can be trusted when making decisions about medical therapies, this study provides yet another solid reason not to rely on this kind of evidence. That will not, of course, stop people from responding to substantive, evidence-based critiques of specific products with pointless testimonials, but it will help remind all of us why these shouldn’t be taken very seriously.

 

**Brief Bibliography

Burton, R. (2008). On Being Certain: Believing You’re Right Even When You’re Not. New York: St. Martin’s Press

Carroll, RT. (2000) Becoming a Critical Thinker – A Guide for the New Millennium. Boston: Pearson Custom Publishing.

Gilovich, T. (1993). How We Know What Isn’t’ So: The Fallibility of Human Reason in Everyday Life. New York: The Free Press.

Kahneman, D. (2011). Thinking, Fast and Slow. New York: Farrar, Straus and Giroux.

Kida, T. (2006). Don’t Believe Everything You Think: The 6 Basic Mistakes We Make in Thinking. New York: Prometheus Books.

McKenzie, BA. Veterinary clinical decision-making: cognitive biases, external constraints,and strategies for improvement. Journal Amer Vet Med Assoc. 2014;244(3):271-276.

Park, RL. (2001) Voodoo Science: The Road from Foolishness to Fraud. Boston: Oxford University Press.

Sagan, C. (1995). The Demon-Haunted World: Science as a Candle in the Dark. New York: Random House.

Shermer, M. (1997). Why People Believe Weird Things: Pseudoscience, Superstition, and Other Confusions of Our Time. New York: Holt, Holt & Company.

Tavris C. Aronson, E. (2008) Mistakes Were Made (But Not by Me):Why we Justify Foolish Beliefs, Bad Decisions, and Hurtful Acts. Boston: Mariner Books.

Burch, D. (2009). Taking the Medicine: A Short History of Medicine’s Beautiful Idea and our Difficulty Swallowing It. London: Chatto & Windus

The Skeptic’s Dictionary

Veterinary Medicine and the Philosophy of Science

 

A recent editorial from Trends in Molecular Medicine does an excellent job explaining why clinical trials testing the effects of highly implausible therapies, like homeopathy and energy medicine, are not an example of evidence-based medicine but of science tricked into serving as advertising.

Gorski DH. Novella SP. Clinical trials if integrative medicine: testing whether magic works?
Trends in Molecular Medicine 2014; In Press

Clinical trials are the last in a series of steps to evaluate potential therapies, and those that fail at earlier levels, or that have not been subjected to appropriate pre-clinical testing, should not be inflicted on patients. It is unethical and dangerously misleading to conduct such clinical trials.

When poorly designed and conducted, such trials generate spurious positive findings that mislead. When properly designed and conducted, they most often generate negative findings (read Snake Oil Science for numerous examples). However these negative studies do nothing to discourage the use of these therapies among proponents. Conducting such studies in hopes of finding a rare exception in which they prove effective and long-standing scientific principles are overturned is not an efficient use of limited resources for testing novel therapies. We should focus our energy on therapies that don’t conflict with well-established principles of basic science and which are promising on the basis of pre-clinical research. Most of these will fail at the clinical trial level as well, of course, but at least the odds are somewhat better than for trials on what amounts to magic.

This is relevant to veterinary medicine as well as human medicine, though as usual the scope of the problem is smaller. For example, conducting trials of homeopathic remedies for cancer (Banerji Protocol as Adjunct Nanotherapy for Treating Canine Appendicular Osteosarcoma: A Double Blind, Randomized Study Design), despite the overwhelming evidence that homeopathy is inconsistent with fundamental and well-established laws of physics and chemistry and has consistently failed pre-clinical and clinical testing (1, 2), is simply unjustifiable ethically and scientifically. It is an irrational use of limited resources for investigation of novel treatments.

Here is part of Dr. Gorski and Dr. Novella’s conclusion from the editorial:

In RCTs [randomized controlled clinical trials] testing modalities with low pre-test probability (i.e., low plausibility), confounding effects are vastly magnified, easily producing false-positives. In these days of extreme scarcity of research funding, it is difficult to justify spending precious research dollars carrying out RCTs of treatments where the likelihood of producing a true positive trial is so low and that have real risks that can lead to outcomes like the TACT or pancreatic cancer trials. All clinical trials, not just RCTs, should be based on scientifically well-supported preclinical observations that justify them, preferably with biomarkers to guide patient selection and follow-up. Until specific CAM and IM modalities achieve that level of preclinical evidence, RCTs testing them cannot be scientifically or ethically justified. That is science-based, rather than evidence-based, medicine.

Sadly, many of the vets I talk to, even those who are sincerely committed to science and evidence-based medicine, appear to still believe that conducting clinical trials on even the most egregiously pseudoscientific therapies is appropriate and will make a difference in whether these therapies are chosen by clinicians. There was a time when I believed this. Unfortunately, this view doesn’t reflect the reality of alternative medicine research.

The problem is that the clinical trial literature, which is often questionable in mainstream human medicine and frequently poor quality in veterinary medicine overall, is consistently abysmal when it comes to alternative medicine research. I have previously reviewed the positive literature presented by the Academy of Veterinary Homeopathy last year in defense of homeopathy, for example. It took many hours of locating and reading dozens of papers in detail to determine that this evidence was thoroughly unconvincing. And yet these poor quality studies are constantly put forward as supportive of continued clinical use for these modalities, and the majority of clinicians simply don’t have the time or interest to go beyond the abstracts and appraise them critically.

Such studies create the appearance of legitimacy, which is used to make such modalities more widely acceptable even when the literature doesn’t actually support efficacy. In some cases, for example in China, negative trials of alternative therapies are virtually never published, which illustrates the uniquely unreliable nature of this literature. It is effectively marketing rather than science.

So in light of this, clinical trials of highly implausible treatments don’t accomplish their ostensible purpose. They don’t accurately evaluate safety and efficacy, and they don’t discourage the use of modalities that are pretty clearly ineffective. The work of the National Center for Complementary and Alternative Medicine (NCCAM) is another great example. This agency has spent billions of dollars testing therapies that have not met basic preclinical plausibility criteria. The vast majority of these trials have found no evidence of efficacy. And these results have had little effect on the popularity or use of the modalities studied. (3, 4, 5).

Now of course by saying “alternative medicine,” I am lumping together many very different approaches, most of which are mutually incompatible theoretically as well as sometimes inconsistent with the mainstream scientific understandings of health and disease. But I am not suggesting none of these deserve to be evaluated in clinical trials. Many herbal remedies, manual therapies, dietary manipulations, nutraceuticals, etc. are quite plausible and can and should be evaluated scientifically like any other novel therapy.

But when talking about methods which could only work if fundamental scientific principles upon which much of our technology and medicine is based is wrong (e.g. energy medicine, homeopathy, the theoretical structures of TCM, etc.), or which have effectively been disproven already (e.g. homeopathy, magnet therapy) I can see nothing to gain from clinical trials. They will almost certainly be of low quality and evidentiary value, and even if negative they will not discourage the use of such therapies.

And yes, there are plenty of different evidentiary problems with conventional medicine. Pharmaceutical companies often shamelessly manipulate data, treatments like hyperbaric oxygen and stem cell therapy are rushed into use well before sufficient evidence is available to justify this, and so on. But clinical trials examining therapies which either can’t work, unless much of established science is wrong, or which have already been extensively tested and failed to show evidence of efficacy, are simply wasteful and misleading.

And from an ethical perspective, it is truly disturbing to see veterinary cancer patients treated with homeopathy, especially when otherwise respected individuals and institutions are involved.. I think this is unfair to the patients even in the context of a clinical trial when there is so much evidence in human medicine that there have been systematic reviews of the systematic reviews of research on homeopathy (which, as I have discussed previously, have shown that the positive results come only from inadequately controlled studies).

It seems fundamentally unethical to do work such as this in the name of science when it is clear that nothing short of a miraculous overturning of established science could allow for a positive result, and it is equally clear that no one who believes in homeopathy is going to give it regardless of a negative outcome in such a trial. Such research wastes resources, creates no clarity, does not allow for real falsification, and exposes patients to therapies for which there is really no reasonable expectation of a benefit.

A number of interesting and important research studies have been published in the last 5 years on the subject of the risks and benefits of neutering in dogs, and these have provided important information to help guide veterinarians and owners in making decisions about neutering. I published a narrative review of the literature on neutering in 2010, before most of these papers appeared, and I have been following this area of study with interest and evaluating individual papers as they are published. Though each study has its limitations, as is always true in science, and we must be careful not to overgeneralize the results of particular studies or lurch wildly from one set of recommendations to another on the basis of limited evidence, these studies are influencing traditional dogma about neutering. I have made several significant changes in both the information I give to owners and the recommendations I make about neutering based on the emerging evidence.

Another important study has recently been published which adds to the evidence in this area.

Hart BL, Hart LA, Thigpen AP, Willits NH. Long-Term Health Effects of Neutering Dogs: Comparison of Labrador Retrievers with Golden Retrievers. PLoS ONE 2014;9(7): e102241. doi:10.1371/journal.pone.0102241

This study was produced by the same group which published an earlier retrospective study of Golden retrievers, and it is quite similar in its methods and some of its limitations. The authors reviewed medical records from the veterinary hospital at the University of California Davis veterinary college. The study looked for Golden retrievers and Labrador retrievers from 1 to 8 years of age and evaluated the occurrence of 6 diseases relative to neutering status and age at neutering.

The findings are summarized in Table 1 for Labradors and Table 2 for Golden retrievers. For comparison, I have also extracted similar data from the earlier paper by on Golden retrievers by the same group in Table 3. This paper classified age at neutering differently (early being <12 months of age and late being >12 months of age), so the studies are not directly comparable, but it is useful to scan for any possible patterns that might help us make sense of these complex data.

First off, what does the study appear to show? Well, perhaps the most important finding is that there are significant differences in the pattern of possible effects of neutering between males and females, and also between the breeds. This latter point is extremely important. People tend to jump on the findings from a study in a single breed and want to generalize those findings to all dogs and make broad recommendations based on that single study. However, if the possible effects of neutering differ significantly between two breeds as closely related and similar as Golden retrievers and Labrador retrievers, likely these effects will be even more dissimilar in breeds less closely related. A study of Rottweilers or Viszlas, for example, may tell us little or nothing about what to expect in Beagles or Poodles. This is worth keeping in mind before we rush into making sweeping recommendations for all dogs.

The findings in this study suggest some increase in the risk of some diseases with neutering, but the patterns were not very consistent. Neutering before 6 months of age appeared to be associated with higher risk of cruciate ligament disease and elbow dysplasia in male Labradors but not with the risk of hip dysplasia. In females, however, the pattern was the opposite, with neutering at less than 6 months being associated with greater risk of hip dysplasia but not the other two orthopedic diseases. But in Golden retrievers, neutering before 6 months of age was associated with greater risk of cruciate ligament disease in both sexes, but with greater risk of hip dysplasia only in males, not females.

Comparing the current study with the previous study of Golden retrievers also adds to the confusion. In the current study, risk of lymphoma seems to be higher in dogs neutered between 6-11 months than in intact dogs, but not in dogs neutered before 6 months or after 11 months. In the previous study, however, lymphoma risk appeared higher in males neutered before 12 months but not in females neutered at this age. In another example, late neutering in the first paper seemed to increase the risk of hemangiosarcoma in female Goldens but not in males. This was contrary to the hypothesis that earlier neutering might increase the risk of this cancer. However, in the current study, no differences in hemangiosarcoma risk were found for any neutered group of males or females compared to intact dogs.

There are, as always, a number of limitations that we must bear in mind when evaluating this study. As with the previous study, dogs were those seen at a university veterinary hospital. This is a very different population than usually seen at regular general veterinary practices. The veterinary hospital at UC Davis typically sees seen sicker dogs, dogs with more affluent owners, and dogs that live in different areas than those seen by regular practices in the rest of California. So we have to be careful about assuming than any relationships seen between health and neutering in this population will hold for the very different population of primary care practice patients.

Dogs younger than 12 months were excluded from this study. While the disease looked at are less common in dogs less than 1 year of age than in older dogs, they do occur. If dogs who get these diseases under 12 months are different, in terms of neuter status, from the dogs looked at in this study, that might change the apparent relationship. For example, if dogs who are intact are more likely to get one of these diseases earlier than neutered dogs, then excluding younger dogs would make neutering seem like a risk when it might actually be protective.

Similarly, dogs over 9 years of age were excluded as well. These dogs are particularly likely to get some of the cancers looked at in the study, and again if the ones who do differ in neuter status from the ones in this study, the apparent relationships identified in the study might be very different from the real relationships between these diseases and neutering.

A number of potential benefits of neutering, such as preventing mammary tumors and uterine infections in females and possibly being associated with longer overall lifespans, were not included in this study. While these are not directly related to the questions the authors sought to investigate, it is important that we consider overall risks and benefits when deciding when or if to neuter our dogs. If neutering prevents some cancers and increases the risk of others, for example, we need to make a decision about which effect is greater or more important in a particular dog. This study only looks at a small piece of that equation, namely the risks but not the benefits.

Finally, there are many comparisons made in this study, and each one is evaluated statistically. This can create a problem known as the multiple comparisons problems, in which the usual threshold for statistical significance is not appropriate. It is common, in studies which make multiple comparisons between subgroups and don’t correct for this problem to find a smattering of statistically significant results which are really the result of chance. In this study, I am not certain if a correction for multiple comparisons was made. If not, this could explain some of the significant differences found, particularly when the overall pattern is not consistent between studies or does not fit expectations based on biology.

Bottom Line
Overall, this study does suggest that neutering under 6-12 months of age might be associated with an increased risk of some orthopedic diseases in Golden retrivers and Labrador retrievers. In Labs, neutering does not appear to increase cancer risk. In Goldens, there may be some effect on cancer risk, but it differs between males and females. And in females, it may even be the case that earlier neutering is associated with a lower cancer risk than neutering later in life. But there are a number of inconsistencies and limitations which make these findings tentative at best.

The study also appears to show a strong difference in the effects of neutering on disease risk between even very closely related breeds. So it is important to recognize that the effects seen in one breed cannot reliably predict those that might be seen in another.

No single, universal recommendation for neutering or not neutering dogs of any sex, breed, or age is justified by the scientific evidence. There is some suggestion that females may experience significant health benefits, particularly in prevention of pyometra and mammary tumors, though the evidence concerning mammary tumors is weaker than generally believed. For males, there does not seem to be a compelling, consistent health benefit to neutering, and in large breeds there may be some benefit to waiting until after 1 year of age. In any case, as I always stress, it is important to balance potential risks and benefits carefully.

And not all benefits of neutering are related to the health of the individual dog. Millions of unwanted dogs are born and euthanized in the U.S. every year, and while every client I talk to dismisses the possibility that their dog could contribute to that problem if left intact, most of these dogs are the product of breeding by owned pets. So if you choose to keep your dog intact for health reasons, it is important to accept that this creates an additional responsibility to make sure he or she does not contribute to the serious problem of overpopulation and the suffering and euthanasia of unwanted dogs.

With my own clients, I discuss the complexity and inconsistency of the evidence, the general trends that seem to be apparent, and the important of making individualized decisions based on all the relevant factors, including age, breed, sex, intended purpose, and the circumstances of the owner. For the most part, I recommend neutering of females before their first heat, though the evidence is still not ironclad and more work needs to be done to elucidate the details of potential risks and benefits in different breeds and with different neutering ages. For males, I no longer see a compelling reason to neuter unless there is clear evidence of aggression towards other dogs, however I emphasize to owners that they must be aware of the added responsibility to prevent roaming and unwanted breeding.

These recommendations are different from those I gave when I first graduated from veterinary school, and they have changed as a result of changing and improving evidence. Likely they will continue to evolve as more research is done and we have a better understanding of the issue.

Table 1

Table 2

Table 3