Sunday, May 25, 2008

The 'Early detection' potential of gammaIFN

Following our postings below on Defra's newly discovered love affair with gammaIFN, and its subsequent carnage , the twitterings spirited defence of the test from London continues to concentrate on its ability to facilitate 'early detection' of Tb. In written parliamentary questions - the answering incorrectly of which is a hanging offence - an epidemiological roadmap of bTb has emerged, which has proved most useful.
Dec 2003: Column 218W

Mr. Paterson: To ask the Secretary of State for Environment, Food and Rural Affairs what plans she has to replace the TB skin test used on cattle as a preliminary diagnostic aid with a more accurate and sensitive test. [141969]

Mr. Bradshaw: Improved diagnosis of TB in cattle is a major objective of Defra's wide-ranging research programme. The current most promising candidate is the gamma interferon test, a laboratory-based blood test that measures the immune response to M. bovis (the causative agent for bovine TB) of T-cells in cattle blood.

This test was officially recognised by the EU in July 2002, but only for use as a supplement to the Single Intradermal Comparative Cervical Test (SICCT) in TB affected herds. The test is considered more sensitive than the SICCT, but less specific, meaning that it results in a higher probability of false positives. For this reason, the gamma interferon test cannot be used on its own as a screening test for TB for the time being.

So, a supplementary test with a higher probability of delivering false positives, but are Defra correct in informing their ministers and the public that gammaIFN will detect Tb at an 'earlier stage'. No. From answers to written parliamentary questions, they are not.
23rd March 2004: Column 689W
Mr. Paterson: To ask the Secretary of State for Environment, Food and Rural Affairs how long after exposure to an infective dose of M.Bovis bacilli a bovine animal becomes infectious, and what the typical duration of infectivity is [159074]

Mr. Bradshaw: Published figures for the time from exposure to being infectious include 87 - 226 days after natural exposure in a United Kingdom study.[ ] The typical duration of infectivity is not known.

Now we are sure you don't need reminding that Tony Yewdall's cattle waited in limbo for at least six months before slaughter, as did those of the Somerset based Burrough Farm Partnership. And we have discussed before the latency of the skin test as being a published 30 - 50days from exposure to flagging up an immune response to a bovine tuberculin antigen jab. (Gamma has a latency of approx 28 days) The maximum exposure window for these cattle, could have been that length of time before the previous skin test which may have been at 60 days, six months, a year or several years, depending on herd and parish disease status.

Only 7 of 177 cattle slaughtered after gamma bloods had visible signs of Tb. And those were not necessarily at an infectious stage. How long had they been waiting? The thick end of eight months - or 240 days from the latest possible exposure. And how long did PQs tell us was the maximum time of an animal becoming infectious after natural exposure? Yup, you read it correctly. A range " 87 - 226 days ".

So we repeat - as does our big sister site, what an appalling, absolutely avoidable waste.


Matthew said...

This message has come in from a veterinary colleague with vast experience of Tb control and testing, and we are grateful to post it.

Bradshaw’s PQ answer of Dec 2003 contains a howler:

"The test is considered more sensitive than the SICCT, but less specific, meaning that it results in a higher probability of false positives. For this reason, the gamma interferon test cannot be used on its own as a screening test for TB for the time being."

Clearly this answer was written by a civil servant who had no conception of the basic principles of Veterinary Medicine.

A test with high sensitivity and low specificity is ideal for use as a screening test but totally unsuitable for a follow up to a test of higher specificity. In the highly successful eradication of Bovine Brucellosis a screening test of high sensitivity and low specificity was used routinely. Every positive was then tested with two other tests of higher specificity (and lower sensitivity) to reduce the false positives to a very low level.

This is the exact opposite of the way that the gamma interferon test is being used.

The traditional single intradermal comparative test is, actually, two tests carried out at the same time. The injection of bovine tuberculin is a test of relatively low specificity: if all animals with a reaction were slaughtered there would be many more false positives; the injection of avian tuberculin increases the specificity since all animals with an avian reaction equal to or greater than the bovine are deemed to have passed the test. In effect the test is a high sensitivity, low specificity screening test carried out at the same time as a supplementary test to increase the specificity.

No sane person would advocate retesting a herd with reactors by means of a skin test involving only the bovine injection, yet that is perilously close to the use that is now being made of the gamma interferon test.

The use of a test of lower specificity to refine the results of a screening test could only make sense if the second test had a very much greater sensitivity.

The real reason why the gamma interferon test is unsuitable as a screening test is that it is expensive; screening tests need to be cheap, as for example, the ELISA test for brucellosis carried out on bulk milk samples.

Another very obvious fact has been missed by the gamma interferon enthusiasts. Immunological tests detect evidence of an immune response. Eradication programmes relying on immunological tests have to take serious account of the fact that animals which have mounted an entirely successful immune response will react to the test while animals whose immune defences are entirely defective will pass. That, I think, explains the absence of any evidence of disease in so many reactors to the tuberculin test and the great majority of gamma interferon positives. It also explains the very rare finding of animals with progressive disease which pass tests, as well as the absence of any sign of progressive disease in the overwhelming majority of those who fail.

Anonymous said...

Anyone concerned have a look


Matthew said...

Anon 12.35
Thanks for that. Our own vets had mentioned the dictat. They were not amused. We've done a posting and linked to the story the FG covered.