This article is reproduced with the permission of New Scientist for exclusive use by Nova users.

Seeing the seeds of cancer
24 March 2001
From New Scientist Print Edition.
Eugenie Samuel

When my grandmother found a lump in her breast at the age of sixty-four, her doctor told her it was nothing to worry about. By the time he realised his mistake, it was too late. "You've killed me," she told him on his final visit before her death.

She had already lost her mother and grandmother to breast cancer. Today, the hereditary nature of some breast cancers is better understood, and cancer researchers are seeking reliable ways to screen women with this genetic predisposition. The available techniques are not perfect: they can still miss a developing cancer, and with it the opportunity for lifesaving intervention. But a recent discovery involving the weird quantum laws that govern the microscopic world of atoms could provide a better solution.

Researchers are now testing a magnetic scanner that creates a strange quantum link between atomic nuclei in different parts of the body. The two linked nuclei act as one, synchronising their movements even if they are centimetres apart. But this strange behaviour only occurs when the tissue under scrutiny does not contain any abnormalities. This means that the quantum link can deliver vital information about any developing health problems. Clinical trials of the technique are already under way, promising to detect tumours earlier than ever before, and to tell malignant and benign tumours apart at a very early stage. Quantum effects used to be an odd abstraction, with little relevance to our daily existence. But one day soon, the weird nature of quantum physics might just save your life.

The potential of this quantum imaging technique first came to light nearly a decade ago when a group of Princeton University researchers, led by chemist Warren Warren, uncovered a flaw in the 50-year-old theory of magnetic resonance imaging (MRI). This technology is currently considered the most sensitive way to screen for breast cancers, and is used as a follow-up when suspiciously dark areas of breast tissue show up on mammograms (see "Screen test").

Warren and his collaborators were investigating the structure of proteins by hitting small samples of protein molecules in water with complicated sequences of radio pulses. They hoped that the response of the molecules to this radiation would cast light on the structure. But they didn't understand the results they were getting.

The nuclei of certain atoms behave like tiny magnetic spinning tops. In the strong magnetic field of an MRI machine the spinning nuclei all line up with their axes aligned with the magnetic field. When a radio-wave pulse hits the aligned nuclei, it knocks them out of alignment. They carry on spinning, but at an angle—just as a knocked top will continue to spin even when it is leaning over. Having all these spins at an angle to the main field creates a measurable magnetic signal, which fades as the nuclei gradually return to their aligned, equilibrium position.

Warren and his colleagues thought they would be able to learn about the protein's structure from the different ways in which this magnetic signal faded when they varied the characteristics of the radio-wave pulse. But the signal in Warren's samples didn't fade in the way that MRI theory suggested it should. "It was even happening in control samples containing pure water," says Warren. "We thought something must be wrong with the instruments."

When the instruments checked out, he thought he might have made a mistake in designing his pulse sequences. So he pared down the experiment until he was using the simplest possible sequence of just two pulses: a radio-wave pulse followed by a short-lived magnetic field pulse called a crusher. The crusher kicks each of the spins by a different amount, and so should leave them in disarray, destroying any measurable magnetic signal in the sample.

But there was a signal. It was only about a tenth as strong as the usual magnetic resonance signals, but it was still more than a thousand times stronger than the background noise. Warren performed the experiment time and again. There was no doubt that the signal was there.

The researchers found the results of these simple sequences so bizarre that they called the sequences "CRAZED". The term has stuck: it is now the accepted name for the pulses that produce these unexpected weak signals. Warren suspects that he wasn't the first researcher to see signals from CRAZED pulse sequences. "When it happened before, people probably thought 'I made a mistake on my sequence'," he says. But Warren was sure of what he was doing and what he was seeing. All he needed now was an explanation.

Having traced the way the unexpected magnetisation gradually faded, Warren used the mathematical technique known as Fourier transform to reveal the different frequency components of the signal. This provided him with the mysterious signal's "fingerprint"—and it looked strangely familiar.

Strange connection

When a pulse of radio waves is produced under certain conditions, its photons acquire a quantum mechanical connection, called "coherence", which keeps them in step with each other. When two coherent photons hit a pair of spinning nuclei they transfer their quantum coherence to the two nuclei. This link binds the nuclei together into one quantum state. One result of this is that the spins will return to equilibrium at exactly the same time, and this simultaneous response gives rise to the characteristic frequency spectrum that Warren recognised.

Warren had only ever seen this link occur between nuclei within the same molecule. But when he examined the quantum coherence created by the CRAZED pulse he was surprised to find that the linked nuclei were micrometres apart, far more than the width of a molecule. Instead of wreaking its normal havoc and destroying all correlation between the spins as it was expected to, the crusher had left the delicate quantum link between two hugely distant spins intact.

Warren eventually worked out what was happening. Although the crusher rotates the axis of each of the spinning nuclei by a different amount, that rotation doesn't destroy every one of the quantum coherences. If it rotated one spin by 10 degrees, say, and another by 370 degrees—a full circle plus 10 degrees—their original relationship would remain the same. Far from knocking all the spins out of alignment, the crusher pulse would allow certain nuclei to remain in step and so maintain their ghostly link.

Since this discovery, Warren and his collaborators have published a string of papers in the journal Science, explaining their data and its implications. The most far-reaching of these, Warren has realised, is that it is possible to change the characteristics of the crusher pulse and alter the distance over which the link occurs. The researchers have since established a quantum link between two nuclei that are centimetres apart. "We've even done this between molecules in different test tubes," Warren says.

Although these long-distance correlations are impressive and strange, it is the possibility of going down the scale, and producing quantum coherences between nuclei just one-tenth of a millimetre apart that has got imaging researchers excited. Warren's quantum coherences only form between pairs of nuclei that are in exactly the same environment—which means they must be in tissue that's in the same state of health. So by imaging neighbouring molecules, and finding where the coherences don't form, Warren can pinpoint exactly where the health of the tissue changes. This should allow him to trace out, say, the border of a tumour. The technique will provide a resolution about 50 times finer than the limit of conventional MRI.

And this is just the start. Because Warren's technique provides an accurate measure of the oxygen concentration in body tissues, it can also diagnose the exact state of a tumour. Tumours use oxygen in a very particular way. The outside of a malignant tumour co-opts its own blood supply from the body to provide a stream of oxygen for growth. But the inside of the tumour has stopped growing and is largely dead and deoxygenated. Most normal tissue has an oxygenation level somewhere between these two extremes.

The oxygenation level of the tissue determines the rate at which its spinning nuclei, disturbed by the radio-wave pulse, will return to their original orientation. Only two spins that return at the same rate will form a quantum coherence and give out the fingerprint signal. At the borders of a tumour, where there is a sharp change in the level of oxygenation, the quantum coherence can't form, and thus there is no signal.

With Mitch Schnall at the Hospital of the University of Pennsylvania, in Philadelphia, Warren has now begun clinical trials of breast tissue imaging using the new quantum states. They have two aims: to pick up breast tumours too small to see with current techniques, and to use the better resolution to pick out more detail in the blood supply. If they can see exactly how the tumour is growing they should be able to determine the level of malignancy without the need to remove a sample of tissue for biopsy—a painful and distressing procedure. The clinical trials will run for a year, and until they are over, Schnall doesn't want to make any concrete claims. "We're not yet in a position to say it's better, but it has the potential for substantial impact," he says.

Researchers at the Institute of Cancer Research at the Royal Marsden NHS Trust hospital in Surrey are also generating images using the new technique. Angelo Bifone and Martin Leach have produced the first quantum coherence images of the human brain and of brain tumours, although they are still learning how to interpret the pictures. "We know there's a lot of new information in there," says Leach, "but we don't know how to use it yet." Nevertheless the team can see attractive possibilities ahead. "If you want to study smaller blood vessels you can just use a pulse that's twice as long or twice as strong—there's nothing like that in conventional MRI," says Bifone.

In his latest study, Bifone has used the new method on the bone of people suspected of having osteoporosis, where the bone becomes porous, or "trabecular". "You can look at holes on different length scales in the trabecular bone to see if and how they're thinning," he says.

It's still too soon to tell exactly how good the technique will prove in all these situations, but Warren believes that quantum imaging could eventually save lives. His own mother died of breast cancer the same year he discovered the quantum links. She was diagnosed in 1986, and fought the tumour through surgery and chemotherapy. But nine years later, the cancer recurred and killed her. At the time Warren was some way from his first images. "Today," he says, "she would have been a prime candidate for what we hope we can achieve: aggressive, very early detection."

I can't help wondering whether this quantum imaging would also have saved my grandmother's life. Or whether, one day, the strange nature of quantum physics will come between me and my genetic fate.

From issue 2283 of New Scientist magazine, 24 March 2001, page 42

For the latest from New Scientiist visit www.newscientist.com