Tuesday, April 22, 2008

His master's voice - GammaIFN

We have deliberately not posted much about the gamma interferon blood test, pending the result of a court ruling. The case sought a Judicial Review into some quite extraordinary results of its use experienced by some farms, unlucky enough to have it inflicted on their cattle herds.

One such was the Somerset based Higher Burrow Organic Farmering Partnership, where over 430 animals were tested late last year. Of those, 14 had a positive skin test result. But in accordance with Defra policy of using the newly approved blood test as a herd test in such cases, the cattle underwent parallel GammaIFN testing. The results showed that 86 were positive, while eight results were indeterminate.

Concerned farmers in several herds so affected, requested the primary skin test as backup, as the disparity between the blood test results and intradermal skin test was so great. Defra said no and ordered slaughter. So for months several farms have been snarled up with hundreds of cattle awaiting the results of last week's request to the High Court for a Judicial Review into that apparent disparity.

Newspaper reports which covered the story quote:

Hugh Mercer QC, for High Burrow, who said the two sets of tests showed "a massive disparity". Slaughter without re-testing would be "unlawful, irrational and disproportionate".
The case has been adjourned once already, so that reams of "evidence" could be cogitated and digested, but in one fell swoop, last week, sitting judge, Mr. Justice Mittings dismissed the request saying "the policy is lawful".

He said: "The policy is lawful. The outcome – unhappy and potentially disastrous though it may be – flows from it."
So what of this policy? Where did it have its roots and how was it trialled in the UK?

Defra will say that the blood test was "developed in Australia in the 1980's", but what they are not so keen to explore is that Australia achieved Tb free status ahead of its licensing, using the intradermal skin test - and a clearance of feral cattle and water buffalo, acting as a maintenance reservoir of tuberculosis. Australia also has a hugely different environment from the UK, with many less contaminants to affect the test results.

But in 2006, Defra began a pilot study known as SB4021. They snuck under the radar of 24 farms in 3/4 year testing areas, using bloods from other testing regimes, and tested for bovine Tb. In these areas they would not have expected to find anything at all, but they did. In fact seven percent of samples proved positive. As this test was not a recognised diagnostic test at the time, they could not confirm their findings with slaughter. Specificity and sensitivety of any diagnostic test are interlinked values, but the authors of the report had a much better idea. Chuck out omit any results which didn't fit the expected result. These included cattle from areas of high Tb incidence which may have encountered a sniff of m.bovis , and also postives - up to 25 per cent in a few herds - to another micobacterium which was subsequently identified as M. kansassii (5.4).

The pilot study indicated vastly differing results for length of time in transit of bloods, possible co contamination with recent skin tests, and several other known contaminants including skin granulomas, the precence of antibodies or vaccinates to Johnnes disease (M. avium paratuberculosis) and certain mosses - as well as M. kansassii.

No post mortems were carried out to confirm their results, but Defra presented this ragbag of selectively culled information to the the Standing Veterinary Committee of the EU for approval as a secondary diagnostic test for bTb.

As a secondary test, a gamma negative animal still requires a primary intradermal skin test to regain its status, yet in the bizarre world inhabited by Defra, a postive result means the chop. And this is at the root of what the farmers challenging official Defra policy last week in the High Court were asking for. Parity for both animals, particularly as the bloods had shown such, err - disparity, of results.

Mr. Justice Mittings described Defra's decision to slaughter the animals as
"not only lawful but mandatory".
Mandatory under EU law, based on a small "pilot study" which excluded or ignored results which it did not expect? Rock solid science then.

The blood test was sold to an unsuspecting industry as "flexible in interpretation", but we see little sign of that, with a fixed cut off point now applied and no allowances made for other known contaminants. It was also described as picking up "very early cases". But that too turns out to be somewhat of an exaggeration. The difference between the latency of the intradermal skin test and gammaIFN is about two weeks, with the skin test averaging 42 days and gamma 28 in experimentally infected animals.

But a delicious twist has given one farmer in the queue for a re-test using the primary intradermal skin test, exactly what he wanted. While the cattle at Higher Burrow await their fate, 31 cattle at Tom Maidment's farm in Wiltshire have been retested. Mr. Maidment and his vets had engaged in long correspondence with Defra (London) and Defra (Veterinary Laboratories Agency) but not the clerks at the local AHO who, very obligingly, when the 60 day re-test was due, ordered Mr. Maidment to - err, retest his cattle.

Only too keen to oblige, Mr. Maidment's cattle were skin tested and the results interpreted under severe interpretation. All were clear. All passed. No lumps. Anywhere. Not one.

Which makes Mr. Justice Mittings' statement:

there was no evidence that those blood tests might, if retaken, prove negative...
rather timely. (Or had the learned judge not realised that the last thing the farmers were asking for was another blood test, but the primary skin test?)

And which is probably why, when assessing the relative accuracy of the gammaIFN blood test, Defra are not so keen to reveal that while about half (48 - 51 percent)skin test reactors slaughtered show VL (visible lesions) from the cattle's exposure to M. bovis, at post mortem, 81.4 percent of cattle slaughtered as gamma reactors show no sign of disease whatsoever.


Anonymous said...

We have been told on two separate occasions by Defra officials that the longer the interval between taking blood samples and those samples receiving laboratory treatment the greater the risk of a positive result. I understand that the maximum recommended interval is 12 hours ( 8 hours if you are in South Africa!) Here in Cornwall samples have to be taken back to Truro by mid-afternoon for delivery to Bristol the same day providing of course that there are no hold-ups on the motorway, car breakdowns etc. With the best will in the world and keeping to the speed limit they are not likely to arrive during office hours.I am not aware of Defra working overnight so the chances are the samples will not be seen to before the next day, well outside of guidelines. Even if samples are treated the same day the animals tested first would appear to have a greater chance of testing positive than those tested later in the day especially with a large herd taking several hours to complete a test.If I were faced with the huge discrepancy between the skin test and blood test outlined in the recent court case I would be inclined to challenge the actual blood testing procedures.They appear to be a complete lottery.

Matthew said...

Anon 12.36.

Yes, it would appear from recent challenges to test results that judges are inclined to decide on procedural failures, rather than tackle head on the fundamental right in common law of an individual to challenge results of statute, which may have the opportunity for error further down the line.


Anonymous said...

We all know that the skin test is effective as herd test.

A herd test - it identifies infected herds.

The safest route would be to 'remove' infected herds.

We don't do that, so anything that finds infected animals that the skin test misses should help.

We're talking about bacteria here - not something that can be seen. Just 'cos there's no VL's doesn't mean no infection

Matthew said...

Anon 10.12

Yes the herd test is recognised as the primary diagnostic tool worldwide.

* "A herd test - it identifies infected herds."

No. It identifies individual cattle within a herd - and the more cattle tested the better - which have had exposure to m.bovis bacteria, in enough quantity (just 70 cfu [colony forming units]) to provoke an immune response in that animal.(see last point)

* "The safest route would be to 'remove' infected herds."

Not necessary and not effective if the problem is not in the cattle.

Herd clearances have been done in the past, (some by default)and restocks several months later have experienced Tb breakdowns. Two farmers known to us personally, had cattle taken with clinical FMD. Farms were disinfected within an inch of their lives, and left fallow for months. In each case the restock was a complete herd, from non Tb areas, and preMT before entry.

One had a breakdown within 6 months and the other lasted just a year. The spoligotype identified in the reactor cattle was indigenous to the farm of breakdown, not the area of origin of the cattle, which in each case was several hundred miles away.

Another farmer, after several years of restriction, got clear and the farm was sold. After a period of time, the buildings were demolished and a complete new dairy set up built. Within a couple of years the new owner - and his herd were in trouble.

* "... anything that finds infected animals that the skin test misses should help."

That assumes that the skin test used as a whole herd test is missing anything, and that gammaIFN is specific to m.bovis. Neither assumption is backed by fact.

* "We're talking about bacteria here - not something that can be seen. Just 'cos there's no VL's doesn't mean no infection."

Agree, which is why all post mortem samples are cultured. Open lesions are necessary to facilitate onwards transmission of bacteria and that has proved to be difficult in laboratory trials of confined cattle, and leads everyone except the ISG to the conclusion that in field situations, and with regular testing, cattle to cattle transmission is over stated.

We are aware of the argument re 'infectious badgers' i.e those with open lesions, transmitting bacteria and 'infected'- those whose immune system has walled off the lesion ... for a while. Shedding is thus intermittant. A reactor cow is shot. So, with regular testing, little opportunity to get to the 'open lesion' stage.

So much cfu bacteria are present in badger samples - 300,000 cfu in 1ml of urine - that a microscope is sometimes not needed, we're told. In fact "jumping off the slide" was the expression used. And just 70 cfu needed to provoke the immune respose to the skin test in a cow.

George said...

Even if a herd is infected that doesn’t mean all the cattle are infected. In fact, the commonest number of reactors is one. Removing that one animal will often solve the local problem - after a 60 day test or two with no further infection the herd can be released from restrictions. So shooting the rest of the herd will achieve nothing. If there are further reactors to be found, the skin test is good at finding them at the subsequent tests – if it wasn’t then the successful eradication of TB in cattle from a number of countries could never have been carried out. There is no evidence that significant numbers of infected cattle are missed by the skin test. Continuing cattle problems happen where TB is found in wild species too. Here the disease has to be tackled in all the major species involved in order to make progress or a lot of money will be wasted. Killing more cattle will achieve nothing, as the disease will constantly trickle back into the cattle from an increasing reservoir in the wildlife.

Anonymous said...

Well George it all depends what yiou consider significant doesn't it?

You suggest that "There is no evidence that significant numbers of infected cattle are missed by the skin test".

Ever heard of gamma interferon George?

The results of the trial of this published on 10 August 2006, showed that in problem TB herds about 12% of animals passing the skin test are testing positive to the blood test and of the blood test positives about 18% are shown to be grossly infected or culture positive in almost half of the herds blood tested.

Combined use of skin testing and blood testing resulted in more infected animals being detected than using skin testing alone.


George said...

The skin test is repeated every 60 days and it is very likely that the reactors identified slightly earlier by the gamma interferon test would be found at subsequent skin tests. As I said, if a significant number of infected cattle were missed by the skin test then the successful eradication of TB in cattle could never have been carried out anywhere using the skin test.
The blood test certainly identifies more reactors, but a high proportion of these are never confirmed as having TB and they probably haven’t got it. Of those identified that are infected, most have early disease and are not spreading it, so shooting them earlier does not achieve anything.