The “bute” in horsemeat

The horsemeat scandal in Europe continues to surprise and shock.  Henry Gee has written about the controversy for this site and for Occam’s Corner but I wanted to comment on one of the safety issues.

It seems that a drug, colloquially known as “bute”, has entered the human food chain via some of the horsemeat produced in the UK.  This revelation piqued my pharmacological interest.  I wasn’t familiar with “bute” or to give it its proper chemical name, phenylbutazone, so I decided to dig around a bit.

Quite quickly I found that phenylbutazone had been marketed for human use under the trade name of butazolidine. For me, hearing this name was Proustian, and I was transported back more than 40 years to my father’s pharmacy.  As a teenager, I would occasionally help him dispense drugs and I well remember counting out butazolidine tablets all those years ago.  In those days, blister packs did not exist; there were large bottles of tablets in the dispensary and it was the job of the pharmacist to transfer the correct number to a small, carefully labelled glass bottle.

This anecdote is part of my history but it does tell us one thing: phenylbutazone is an old drug.  In fact it was first synthesised by Geigy Pharmaceuticals in Basel and marketed as butazolidine in the late 1940’s.   The drug was used to treat arthritis and other painful inflammatory conditions and for patients that tolerated the drug it was an effective remedy.  It is classified as a Non Steroidal Anti-inflammatory Drug (NSAID), a class of drugs widely used to treat pain and inflammation.   Aspirin was the first NSAID and many people will be familiar with ibuprofen, another member of this group of drugs.  NSAIDs relieve pain and inflammation by inhibiting an enzyme that synthesises chemical mediators called prostaglandins.

Whereas aspirin and ibuprofen are quite well tolerated by patients, the use of phenylbutazone was associated with a wide range of side effects.  The most serious of these were blood dyscrasias where bone marrow production of blood cells was affected.  This disorder was fatal for about 1 in 45000 of patients taking the drug.  More than a quarter of patients taking the drug reported other side effects including gastrointestinal bleeding, oedema, and toxic effects on the liver or kidney.

By the mid 1980’s, phenylbutazone was withdrawn from general use, partly because of the side effects and partly because better drugs such as ibuprofen had appeared on the market.  Phenylbutazone still features in the British National Formulary as a treatment for the severe rheumatic disorder ankylosing spondylitis, but for use only when other drugs are unsuitable.

Despite being withdrawn for general human use, the drug has found a niche in veterinary medicine where it is widely used to treat pain and inflammation in horses.  Although horses treated with phenylbutazone are not supposed to enter the human food chain, it seems that this rule has been flouted in some cases.  The drug has been found recently in horses slaughtered in the UK but destined for human consumption in continental Europe.

The question then is whether this poses a risk to human health so we need to know the levels of drug found in the horsemeat.  The most straightforward way to express the levels of phenylbutazone is to work out how much there is in a kilo of the meat and in the recent tests, the highest amount found was 1.9 mg.  The Chief Medical Officer, Dame Sally Davies has said that this level of phenylbutazone poses no risk to human health and I thought it might be useful to think about how this conclusion was reached.  In the past, when phenylbutazone was used in patients, a typical dose would have been 400 mg a day, spaced out in four equal amounts.  To reach this level you would need to eat as much as 200 kilos of contaminated horsemeat a day!   On that basis there would appear to be little or no risk.

I wonder, however, if there is a bit more complexity to the story and here are some issues that give me pause for thought:

1.  The argument about the levels of phenylbutazone in horsemeat assumes that it would be necessary to consume up to 400 mg a day of the drug to see effects.  The relation between drug dose and drug response is, however, a complex one, especially for phenylbutazone. I don’t know whether 400 mg of drug is just enough to give a response or whether it is well over that level.  If the latter is true, then lower amounts of drug might have small effects.
2. Some drugs are said to interact so that the effects of one drug on a patient may be modified if they also take a second drug.  There is a very well known interaction between phenylbutazone and the anticoagulant drug warfarin and patients taking warfarin can suffer severe gastrointestinal bleeding if they also take phenylbutazone.  The mechanism of this interaction is complex and, as far as I can tell, it is not fully understood.  This complicates arguments about safety of phenylbutazone in horsemeat.
3. Phenylbutazone is broken down in animals and in humans to another drug called oxyphenbutazone.  This has many of the same actions as phenylbutazone and I wonder if oxyphenbutazone has been tested for in the horsemeat.
4. Drugs developed nowadays are very well characterised for potential side effects and off-target interactions.  Because phenylbutazone was developed so long ago, it will not have been characterised in the same way.  I doubt if we know all the potential actions of the drug.  If any of these occurs at lower concentrations than the effects on pain and inflammation, this could complicate arguments about safety.