Conflict of Interest Declaration: I like animals (often more than I like humans) and so I am biased towards keeping Geronimo alive… but I will try and be objective in the following blog.
TB in cattle is a major economic problem, but here we’re really talking about Mycobacterium bovis (bovine TB) rather than the Mycobacterium tuberculosis that mainly affects humans. Cattle that test positive cannot be sold, moved, bred from, or used for milk production, as they can infect other cattle. Importantly M. bovis can infect humans (and even cats but we don't cull Tabby cats!). Cattle are culled, and the farm is locked down until other animals are able to produce negative tests. It’s a strictly governed system and the annual testing is dreaded by farmers; I can only imagine how distressing and financially debilitating this is for the farmers, it must be truly awful.
Now alpacas are slightly different as they are not usually farmed for anything other than their wool. Apparently you can milk them, and you probably could eat them… but their wool is much more valuable and hence the attraction for farming. In fact, there are about 35,000 alpacas in the UK and the numbers are increasing. But the rules are if an alpaca has TB, then like cattle, they are culled so that they don’t spread the infection to other animals, including the cattle.
Alpacas are known as New World Camelids along with Llamas and Vicunas, being related to the Old World Camelids like Dromedary and Bactrian camels.
So, what’s the fuss with Geronimo?
Geronimo was imported to the UK 5 years ago from… no not Peru… but New Zealand! His new owners wanted him as a stud animal to breed their herd of alpacas. Before being sold and travelling to the UK, Geronimo had 4 negative tests for TB (he couldn’t be imported without the negative tests). After arriving in the UK, he tested positive for TB! He was been tested again, and again the test was positive! As a result of the tests, he has been kept away from the herd and sentenced to be culled. His owner is upset about this and has been trying to find a way to keep Geronimo alive but this week she lost her legal battle through the High Court and Geronimo is now scheduled to be killed.
So, this got me thinking; how would I go about arguing why Geronimo should be spared; are the tests valid and accurate?
Questions…
- What test was negative in New Zealand?
- What test is positive in the UK?
- Is the test validated for use in alpacas?
- What is the risk of the test cross reacting to another organism?
- What are the positive and negative predictive values?
- Is the animal sick? Is the animal a risk to other animals or humans?
So, let’s begin…
What test was negative in New Zealand?
We don’t know for sure what test was used in New Zealand, but the internationally accepted test is a tuberculin skin test (a bit like a TB Mantoux test in humans), and is the same as the test used for cattle. A small amount of the TB antigen mix is injected intradermally into the alpaca’s axilla, and then 2-3 days later the size of the lump is measured to look for a localised skin reaction suggesting the alpaca is launching an immune response to the injection. A lump indicates a pre-existing cellular response and hence past or current infection.
The reactive skin test would be good evidence of infection; however the test is usually negative, as there is a high false negative rate. In fact, the false negative rate can be as high as 50-80%.
This information doesn’t help our case for Geronimo as the test he had before moving to the UK may well have been falsely negative.
What test is positive in the UK?
The test used in the UK is the Enferplex serology blood test. This is an antibody test developed for detecting TB in cattle. Basically, the test uses 7 “TB antigens” as targets for antibodies produced by the test animal (antibodies react with antigens). If the animal has antibody against 2 antigens then the test is positive.
On the face of it the Enferplex is a pretty good test. We use antibody tests all the time in human medicine for testing if someone is immune or has been exposed to a disease. However, these antibody tests are known to give false positives as sometimes there are cross reactions between antibody against one microorganism and antigens produced by another microorganism.
Is the test validated for use in alpacas?
Looking at the literature, the Enferplex test has been used a few times to look for TB in alpacas. A joint study between the Department for Environment, Food and Rural Affairs (DEFRA) and the British Alpaca Society in 2012, reported a sensitivity of 66.7% and specificity of 96.9% for detecting M. bovis in alpacas. However, another perfectly good study from Poland in 2020 showed the Enferplex test had a sensitivity of 0% (yep, it detected no cases at all!) and a specificity of 95%. Given the discrepancies in sensitivity and specificity in the DEFRA and Polish studies, when used in the “real world”, what might be the cause of these differences?
The company who produces the Enferplex test say it is “developed for use in alpacas” but I have tried looking for literature to support the statement and I can find little evidence for this. In fact, the company’s website also says:
“The Camelid societies recognise the development and understanding of serological testing for bTB in camelids is still in its infancy. There is a lot more to learn. As the Enferplex is based on quantitative, rather than qualitative outputs, and relies on seven individual antigens that are monitored separately, the potential to refine the test, to both maximise sensitivity and specificity, is far higher than with the other tests.”
To me this reads as “we are developing a test for TB in camelids but we still have work to do…”, the test may well be “developed” for alpacas but it appears it may not quite yet be “validated” as an alpaca test.
What is the risk of cross reaction to another organism?
Both the DEFRA and Polish studies showed that the Enferplex test cross reacts with another bacterium M. microti. Camelids are susceptible to M. bovis as well as M. microti (which also causes them a TB like disease yet I’ve never seen this as a cause of TB in humans and I presume it does not cause TB in cattle).
Another study tried to find a distinguishing test for M. bovis in alpacas. They had two herds that tested positive with the Enferplex test, but further investigation showed that one herd had M. bovis and the other herd had M. microti. This study showed that the Enferplex test cannot distinguish between these two species of mycobacteria. In fact the Enferplex website does confirm that the test can cross react! So, Geronimo has been tested using Enferplex twice, and condemned, based on a test known to cross react with another bacterium that infects alpacas, but M. microti is not an indication for culling. Only M. bovis requires culling!
If Geronimo was human and had tested positive for a disease with a test that was known to cross react with something else, he would have been tested with a second test BUT one that was different to the first test. A repeat Enferplex test should not be used as a confirmatory test; it’s bad practice. Isn’t it a definition of madness to do the same thing again but expect a different outcome? Using the same test again doesn’t exclude a false positive or cross-reacting test. I would argue that a single positive test, or repeat positives of the same test, DOES NOT PROVE THAT GERONIMO HAS TB!
What are the positive and negative predictive values?
Any reader of the Bug Blog will know that positive and negative predictive values (PPV and NPV) are a pet subject of mine; although they can be dull and confusing! Ultimately though it is the positive and negative predictive values of a test that matters and these are dependent on both the test and the prevalence of the disease within the population studied.
If we use the DEFRA values and assume a prevalence of 5% (which is a recognised statistical assumption, and may in this case even be an overestimate) then the PPV for the Enferplex test would be 54% and the NPV would be 98%. Another way of saying that is there is a 46% chance that the test is a false positive (whereas the test is a great test to show negative results - 98% accurate). If the disease prevalence is actually lower, then the risk of a false positive increases (e.g. a prevalence of 1% means an 88% chance of a false positive); that’s really high! 88% chance of getting a wrong result, that’s like relying on a Covid-19 lateral flow test!!
So, what multi-test regimen would I use to test Geronimo?
If I wanted to prove whether Geronimo did or did not have TB, then I would not rely on a single test as there is no single test that is good enough.
I would use a multistage testing regimen using a very sensitive screening test and confirm with a different more specific test. In fact, this was the conclusion of DEFRA’s own study in 2012; they discuss using a combination of blood tests. The problem is that there is still no really specific test widely available.
In human medicine we would use a PCR test for the specific test which would distinguish M. bovis from M. microti. These PCR tests are available, but they are expensive and (unsurprisingly) unvalidated for alpacas. However, PCR on tissue is only dependent on the bacterium not “the host” so validation in the host animal is less of a problem. The problem is that the sensitivity in “the host animal” is unknown and may be low, therefore there would need to be quite a bit of bacterium in the sample to be tested.
However, using the widely accepted sensitivity of 70% and specificity of 99.8% for TB PCR in humans we would have a PPV of 95% and NPV of 98% and I suspect that would be enough to convince everyone on balance of probability whether Geronimo could have TB.
Is the animal sick? Is the animal a risk to other animals?
This is tricky. Okay, some get a consumptive illness and waste away, yet very few get respiratory symptoms. In fact TB in alpacas can be completely asymptomatic. The theory behind this is that they have physiological adaptations to living at high altitude in their native countries that mean they are able to cope with respiratory pathogens.
So, therefore it is possible that Geronimo could be a risk to other animals even though he isn’t sick. He could still be contaminating his environment with infective organisms despite showing no symptoms. However, no one really knows the level of this risk, but I think it is likely to be low. In humans it is the coughing out of bacteria that exposes others to the bacterium, and Geronimo doesn’t have a cough. BUT don’t camelids spit?!? Do camelids spit at each other or is it just at tourists? Let’s not go there, we’re trying to save Geronimo!!
Closing arguments
So, in summary Your Honour, I believe Geronimo should be given a stay of sentence because the bad science used to condemn him indicates there is sufficient doubt about the diagnosis and therefore this alpaca may not actually have bovine TB:
- The test used to diagnose TB has not been properly validated in alpacas
- The test used to confirm the original diagnosis was the same test as the first test; a false positive reaction cannot be ruled out if using the same test
- The best published data for the use of the test suggests a false positive rate of 46% or more, casting reasonable doubt on the validity of the positive result
- The test used to diagnose TB is well described to cross react with other Mycobacterium spp., particularly M. microti, a bacterium known to infect alpacas and which does not require culling
- This alpaca has tested negative 4 times with an alternative, internationally accepted test even though this test is also imperfect
- This alpaca is not sick; there is no clinical evidence to support a diagnosis of TB. He remains isolated from other animals in the interim, and so does not pose a major risk to other animals
I ask Your Honour to suspend the decision to cull Geronimo whilst a multi-test regimen can be used to ascertain a more reliable result and better inform the decision to cull or not to cull.
The Defence rests….
BUT none of the science may matter!
What I hear you cry?
Well as at the end of the day the UK has a law called The Animal Health Act 1981 and unfortunately this allows the Secretary of State for DEFRA to order any animal where TB is suspected to be put down… they seemingly don’t really need any proof! To reverse this decision Geronimo’s owner either needs to convince the Secretary of State that there is reasonable doubt (I think the 46% false positive shown above might give them cause to claim doubt exists) or argue the law itself is wrong, and that is extremely unlikely to be successful.
As we have seen with Covid-19, we do not always follow the science; politics always seems to get in the way!
Unproven FACTS from Wikipedia: Geronimo was a prominent leader, medicine man and brave defender of the Apache people; he apparently shouted his own name as he jumped from a height to attack the US Cavalry. The exclamation is now commonly associated with bravely jumping from things because (allegedly) a US private, Aubrey Eberhardt, was the first person to say it whilst being one of the first parachutists jumping from an aeroplane.
Good luck Geronimo! Be brave… fight on!
References
- Tuberculosis in camelids. Wernery U, Kinne J. Rev Sci tech OFF int Epiz 2012 31 (3) 899-906
- Tuberculosis in camelids – present situation and tests. Hoen M. Vet Times March 2017
- Evaluation of Gamma Interferon and Antibody Tuberculosis Tests in Alpacas. Rhodes S, Holder T, Clifford D, et al. Clin Vaccine Immunology. Oct 2012, Vol 19, No. 10, 1677-1683
- Development and Evaluation of a Serological Assay for the Diagnosis of Tuberculosis in alpacas and Llamas. Infantes-Lorenzo J, Whitehead C, Moreno I, et al. Frontiers in Veterinary Science 2018, Vol. 5 Article 189
- Transboundary tuberculosis: importation of alpacas infected with Mycobacterium bovis from the United Kingdom to Poland and potential for serodiagnostic assays in detecting tuberculin skin test false-negative animals. Krajewska-Wedzina M, Didkowska A, Sridhara A, et al. Transboundary Emerg Dis 2020; 67: 1306-1314