The Last on the List

I

In the summer of 1992, a forty-eight-year-old woman who had been told to make her preparations sent a sample of her tumour by post to a laboratory in Los Angeles.

Her name was Barbara Bradfield. She had a breast cancer that had come back, had spread to her neck and had already been through surgery, radiation and chemotherapy. Her oncologist was proposing another round of the same and she said no.

At the laboratory, a researcher called Dennis Slamon looked at that tissue and found one of the highest quantities he had ever seen of a particular protein on the surface of the tumour cells.

That made her the perfect candidate for an experimental antibody that had spent years looking for patients.

She began the treatment with fifteen other women. Of those sixteen, she was the one left.

Thirty-odd years later she is still without cancer.

II

The word that changes category here is “candidate”.

Up to here the scheme was always the same: there is a disease, there is a drug and it is given to whoever has the disease. With more or less accuracy, the criterion is the diagnosis.

Not here. Here the drug is of use only if the tumour has that protein in abnormal quantities, and that happens in a proportion of cases and not in the rest. Giving the antibody to a woman whose tumour does not have it is not a mild treatment: it is a useless treatment, with its effects and its bill.

So before the drug a test has to be done, and that test decides.

It is the first time a medicine reached the market tied to a test that says who it can be given to [the drug and its test were approved on the same day]. Today that has a name, “precision medicine”, and it is half of oncology.

The test costs, at a pathology laboratory, considerably less than a single dose of the drug it gives access to. The cheap test governs the expensive treatment, and that hierarchy is new in medicine.

III

I waited nine days for the result of a biopsy in 2022.

It was a mole on my back that my partner could see and I could not, with an edge the dermatologist did not like. She took it off then and there with local anaesthetic, two stitches, eight minutes; she showed me the little piece in the pot before sending it, something I have never asked anybody for.

It cost a hundred and sixty euros with the consultation.

The following nine days were a curious thing. I was not afraid continuously. I was afraid in patches, almost always at seven in the evening, and the rest of the time I was normal, working, eating and watching series. The body does not know how to keep an alarm on that long, so it switches it on and off.

On day nine I rang them, because they were not ringing.

It was benign. I hung up, stood a moment in the hallway and then went to the kitchen to make myself a coffee, and I remember that and not the sentence the woman on the phone said.

Nine days to know. Barbara Bradfield was given, in 1992, a prognosis with no term and no appeal, and the difference between her wait and mine is that hers did not end with a telephone call.

IV

Inside, this is not a classical drug and it has to be said from the start: what is administered is a manufactured defence.

An antibody is a Y-shaped piece the body produces to grab hold of something specific and mark it. What was done here was to choose the target first and commission the antibody afterwards.

The target is an aerial. On the surface of many cells there are aerials that receive the order to grow and divide; in a proportion of breast tumours, the cell makes that aerial in absurd quantities, hundreds of times more than normal, so that it receives the order to grow all the time even though nobody is giving it.

The antibody sticks to that aerial from outside. It does three things at once: it hinders the signal, it marks the cell so that the body's own defences recognise it and attack it, and it drags the aerial inwards so that it is destroyed.

Notice the difference from chemotherapy —which is the whole point—. Chemotherapy kills what divides fast, and so it kills tumour cells, hair, mucosa and marrow [hence the hair loss, the ulcers and the transfusions]. This does not kill what divides fast; it kills what wears that aerial.

The price is also specific and it is not small. That same aerial is in the heart muscle, in normal quantities, and that is why this treatment obliges cardiac function to be monitored while it lasts [an echocardiogram every three months].

Targets are not exclusive. They are only more selective.

V

What this case inaugurates is the idea that the patient is chosen.

Before, a clinical trial took in everybody who had the diagnosis and measured the average. If the drug worked marvellously in a quarter and not at all in the rest, the average came out mediocre and the drug was discarded.

That antibody came close to dying that way. In the first trials with unselected patients, the results were so unremarkable that the company came close to abandoning the programme several times.

What saved it was selecting: measuring the protein, treating only the women whose level was high, and then the effect appeared.

There is the uncomfortable lesson. For years we were measuring averages of mixed populations and calling ineffective drugs that were highly effective in a subgroup we did not know how to identify.

VI

The objection is supplied by the case itself and it is statistical.

Barbara Bradfield is the only survivor of that group of sixteen. Telling her story as proof that the drug works is the error this work tries not to commit; the other fifteen also received the treatment.

The drug works, and we know it from the later trials with thousands of women, not from her.

Her case is what kept the programme from being closed, and that is another thing and is important in another way. A single body can save a drug from cancellation without being proof of anything.

It is a mechanism that appears more often than the industry admits. The decisions to continue or stop a programme are taken in meetings, with budgets and with deadlines, and in those meetings a spectacular case weighs what a table of medians does not [even though the table is the only thing that demonstrates anything].

VII

Slamon stayed in the same laboratory and spent years looking for funding while the company hesitated. The approval came in 1998 [six years after that posting].

Bradfield has told her story a good few times and always insists on naming the other fifteen.

Of them there are no published names, no ages, nothing of where they were from. They are in the trial file as subject numbers, protected by the same anonymity that makes the result reliable, and it is the third time in this work that that anonymity appears on the right side and hurts all the same.

Summer of 1992, a sample sent by post.

Sixteen women, one long-term survivor.

1998, the approval.

Nine days of waiting for a biopsy and a hundred and sixty euros.

A prior test that decides who can receive the treatment and who cannot.