Rat Poison

I

In the small hours of 24 September 1955, in Denver, the president of the United States woke with a pain in his chest that he put down to dinner.

He was sixty-four, he had played golf the day before [and had eaten a hamburger with onion, a thing that entered the official account] and he led a reasonably healthy life for the period, except for the cigarettes, which he had given up some years earlier after smoking four packets a day.

It was a heart attack. He was moved to a military hospital, spent seven weeks admitted and walked out.

Among the things he was given was an anticoagulant then known above all for something else. It had been sold, since 1948, as rat poison.

That the president received it and recovered changed the product's reputation within months [and without a single new trial being published]. The doctors who doubted stopped doubting, and a substance that was in hardware shops to kill rats became the most prescribed oral anticoagulant in the world for fifty years.

II

The origin is a Wisconsin farmer who in the winter of 1933 drove three hundred kilometres with a milk churn full of blood that would not clot, a dead cow and a bale of mouldy sweet clover.

His cows were bleeding to death internally after being castrated or dehorned; they were dying of haemorrhages from things that had not killed before. The disease was known and was called “sweet clover disease”; the mouldy hay produced a substance that prevented clotting, and stockmen lost animals every year.

A university biochemist, Karl Link, spent six years isolating it —six years with the bale of clover in the laboratory—. In 1940 he identified the compound responsible and afterwards made more potent versions.

The obvious application was not medical: if that made cows bleed, it would kill rats. One of those versions was put on the market in 1948 with a name formed from the initials of the university's foundation and the ending of the chemical family [the initials are those of the office that handled its patents].

That name is the one on the pharmacy box today.

There is an intermediate episode that closes the circle. In 1951 a navy recruit drank an enormous dose of the rat poison to kill himself; he was admitted, given vitamin K and recovered completely. That attempt demonstrated two things at once. That the poisoning was reversible and that a cheap antidote existed. Four years later they were prescribing it to the president.

III

I have bought it at a hardware shop, in the garden section, for four euros ninety.

It is a blue bait block, the size of a bar of soap, with grooves and a hole in the middle to spike it on the trap holder. It smells faintly of cereal. The box carries a small skull, a toxicology telephone number and the sentence, in small print, that it contains an anticoagulant.

I bought it for a country house with mice in the false ceiling and I used it twice.

What interested me was the label [the small print, not the skull]. The active principle is a direct relative of the one my uncle has taken for eleven years, with his orange card [he carries it in his wallet, behind his identity document], his monthly blood test and his list of foods he cannot eat in varying quantities.

The difference between the blue block from the hardware shop and my uncle's tablet is not the molecule. It is the dose and it is who decides when to stop.

Both boxes carry warnings; the rat poison's is clearer.

IV

Inside, this is a sabotage of recycling and it has to be told slowly because it explains everything else.

To make the factors that clot the blood, the liver needs vitamin K. It uses it as a tool. Every time it finishes a factor, it spends a unit of the vitamin, which is left oxidised and useless. Since the vitamin is scarce, the liver has an enzyme dedicated to repairing it and making it usable again, over and over.

Warfarin blocks the one that repairs.

It does not touch the vitamin, does not touch the factors, does not destroy anything already circulating. It simply prevents recycling, and the liver is left with no tool to make new factors.

From there come the three oddities of this drug.

The first: it takes two or three days to take effect, because you have to wait for the factors already in the blood to be used up. It is of no use in an emergency.

The second: anything that changes the amount of vitamin K in the diet changes the necessary dose. A week eating spinach, broccoli or cabbage [the green leaf is that vitamin's larder] alters the result, and that is why whoever takes it should not stop eating vegetables but always eat the same amount.

The third: there is a perfect and cheap antidote, which is vitamin K itself.

That is why it has to be measured. Every few weeks, a drop of blood, a number, and an adjustment of a quarter of a tablet up or down.

In a rat, the same molecule produces an internal haemorrhage within a few days, because nobody measures anything for it or gives it vitamin.

The manufacturers know this and have used it. When resistant rats appeared, the industry brought out much more potent, long-acting versions, designed so that a single meal is enough. Those are the ones that poison the dogs that eat them by accident, and the ones that oblige veterinary clinics to keep vitamin K.

V

What this story teaches is not that poisons can cure, which is the easy sentence and anybody has known it since Paracelsus.

It is another thing, duller and harder to accept. The difference between a medicine and a poison is not in the substance; it is in the infrastructure around it. A laboratory number every three weeks, a card, a list of interactions, an antidote at the hospital and somebody who adjusts.

Take that infrastructure away and the pharmacy box and the blue block from the hardware shop are the same thing.

Add that infrastructure and a substance that kills rodents keeps millions of people alive with a mechanical valve or with a heart in fibrillation.

VI

The objection has to do with the very story I have just told.

The president's episode worked as proof, and it was not proof. It was a case, with a single patient, with no control group and with all the world's press watching; the drug spread by “borrowed prestige”, not by evidence.

That it afterwards turned out to be a good drug does not fix the procedure. It fixes the result, which is a different thing and is a matter of luck.

It is the same mechanism that appears with insulin and with chlorpromazine, and it repeats because it is what really drives a drug's adoption. A visible case weighs more than a table. What changes from one story to another is whether that case turned out to be representative.

And by now I have spent four or five pieces saying the same thing from different angles, so I apply it to myself: I too tend to tell these stories by the result, because the result is what makes them stories.

VII

Eisenhower went back to work, stood for re-election the following year and won. He died in 1969, at seventy-eight, of his heart.

Warfarin went on being the standard anticoagulant until a few years ago, when it was gradually displaced by others that do not require measuring so often and that do not have so good an antidote.

At the hardware shop the blue block is still there, in the same aisle as the hoses.

Winter of 1933, a milk churn with blood that would not clot.

1948, marketed as rat poison.

24 September 1955, the heart attack, seven weeks admitted.

Fifty years as the most used oral anticoagulant in the world.

Bait block at the hardware shop: four euros ninety.

Box of tablets at the pharmacy: less than two euros.