麻豆传媒

IN the summer of 1986, a team of British researchers stumbled upon a remarkable new treatment for the painful swelling that is the fate of people who have their wisdom teeth removed. This treatment involved massaging patients鈥� cheeks with an ultrasound probe, a cousin of the device used to monitor pregnancies. And the results were nothing short of astonishing. The swollen jaws shrank by 35 per cent compared with a control group.

Not, however, because of any healing powers of ultrasonic waves. Unknown to both the patients and the dental surgeon the researchers had rigged the machine to look like it was switched on when it was in fact unplugged. Charlotte Feinmann, Malcolm Harris and their colleagues at the University College and Middlesex Medical School in London concluded that they had witnessed an especially potent example of the power of medical persuasion -the ability of sugar pills, sham surgery, and, in this case, unplugged probes to speed the recovery of even the sickest patients.

But the British team鈥檚 undeniably effective treatment had little chance of catching on because of one seemingly insoluble problem. Doctors are bound by duty not to deceive their patients. Yet, if patients know they are getting a bogus, 鈥減lacebo鈥� treatment, wouldn鈥檛 any beneficial effect vanish? On the face of it, there seems a simple choice: placebo medicine or open and honest medicine.

Now, an Australian psychologist called Nicholas Voudouris of La Trobe University in Melbourne, Australia, says that he has a practical solution to this ethical dilemma which will allow medicine to harness the power of the placebo without deception.

Voudouris has shown that student volunteers can be spared much of the pain of a series of small electric shocks if they are also treated by a counterfeit 鈥渁naesthetising鈥� skin cream. Remarkably, it works even when the students know the cream is phoney. The trick, says Voudouris, is to first train the students to associate the cream with pain relief just as Pavlov鈥檚 dogs were conditioned to associate the bell ringing with food.

If all goes well, sometime around next Christmas Voudouris and Milton Cohen, a rheumatologist at the Scottish Hospital in Sydney, plan to test a similar conditioning strategy in real patients, suffering from all sorts of pain, from crippling arthritis, to nerve damage, to the kind of constant agony that is simply inexplicable. These patients will receive placebo pills mixed into their frequent doses of powerful opiate-like painkillers. 鈥淭he beauty of this strategy is that you鈥檒l have less drugs,鈥� says Voudouris. 鈥淲hen you have less drugs, you鈥檒l have less of all the bad side effects that go with drugs.鈥�

Doctors should also end up with cheaper treatments, a potential benefit that will warm the hearts of governments struggling to control escalating health costs. However, the day when the placebo effect becomes a respectable medical tool is still some way off. Doctors are officially forbidden to use placebos other than in private practice and in clinical trials. And while many medical practitioners will privately admit to using sham injections or fake pills, this is not because of any great faith in their healing powers. More often it is because medical practitioners wrongly see placebos as a way to expose a patient鈥檚 complaints as imaginary.

Indeed, according to Patrick Wall, a neurologist at St Thomas鈥檚 Hospital in London, some doctors still give credence to a long disproved notion that only patients with a 鈥渟pecial mentality鈥� 鈥� read crazy 鈥� respond to placebos. 鈥淥penly give a patient a placebo, and people say 鈥� here, here, you鈥檙e encouraging irrational behaviour.鈥�

Confusion about placebos even extends to the scientific literature. Edzard Ernst, a doctor and researcher at the University of Exeter, and his colleagues have looked at hundreds of articles published over the past 20 years and found that only a quarter that claimed to study the placebo effect used the word properly. According to Ernst, the studies often make no distinction between subjects who have been given fake medicines and subjects who have received nothing at all.

In fact, scientists only seem to invoke the placebo effect 鈥渨hen they don鈥檛 quite know what鈥檚 happening and really don鈥檛 care鈥�, says Phil Richardson, a psychologist at Guy鈥檚 Hospital in London. One reason for such apathy is a lack of financial incentive, says Robert Ader, a psychologist at the University of Rochester, New York. When you approach scientists employed by pharmaceuticals companies, they do get excited about the clinical possibilities, says Ader. 鈥淏ut then they put on their company hats and say 鈥� nobody is going to give you a dime to do this work.鈥�

Nuts and bolts

The problem isn鈥檛 simply that nobody is going to award patents for sugar pills. Outside and inside companies, medical researchers invariably feel more comfortable with therapies that can be understood at the level of biology鈥檚 chemical nuts and bolts 鈥� drugs receptors, genes, cell-signalling molecules and the like.

This attitude irks Ader. 鈥淚t鈥檚 not a question of whether psychology is better than chemistry,鈥� he says. 鈥淏oth are operating, so the best treatment will optimise both.鈥� Besides, he argues, the day will come when placebo effect can be explained by the actions of molecules and cells. There鈥檚 no magic about it, he insists.

Ader and like-minded researchers are busy looking for clues in the way the brain 鈥渢alks鈥� to the immune system (see 鈥淟et mind talk unto body鈥�, 麻豆传媒, 23 July 1994). The list of brain chemicals that can interact directly with immune cells or their molecular minions is growing all the time. Each one of these chemicals may play a small role in the placebo effect, suggests Ader.

But while the precise mechanisms by which a bogus medicine can trick the human body to heal itself are still unknown, that鈥檚 no reason to hold back clinical research into placebos, say Voudouris and Cohen. Far from it. All you need to start developing placebos as clinical tools, they say, is a lesson from Pavlov and his salivating dogs.

Pavlov not only showed that dogs could be conditioned to drool at the tinkling of a bell. He also got them to respond to a bell as if it were the drug morphine, sliding into drowsy oblivion. And in the decades since Pavlov, researchers have steadily refined their conditioning skills. Well-trained lab rats and dogs will respond to injections of salty water or sugar pills as if they had received all manner of potent or toxic medicines. In the latter case, placebo-fed animals have even been known to keel over and die.

In the normal scheme of things, humans don鈥檛 respond to placebos because of any deliberate conditioning. They respond, explains Voudouris, because of the incidental conditioning of a lifetime. Every time you pop an aspirin you are, in effect, strengthening an association between 鈥渨hite pill鈥� and 鈥渇eeling better鈥�. Since no two medical histories are ever identical, we are all conditioned in slightly different ways.

Green or pink?

Which could explain the one observation above all others that has persistently made medical researchers wary of any attempt to harness the placebo effect: the fact that the response is notoriously unreliable, occurring in between 0 and 100 per cent of patients depending on the drug, placebo and disease.

To make matters worse, nobody has succeeded in predicting who will be most susceptible to imitation therapy, or identifying the type of placebo that is most likely to do the job. A person who responds to a green sugar pill may not be affected by a pink one, for instance. Also, placebo injections are generally more powerful than placebo capsules, which are in turn more powerful than placebo pills 鈥� at least, some of the time.

Voudouris is confident his strategy will provide a way around the problems. 鈥淕iven the right conditioning, we can make nearly anyone respond to a placebo,鈥� he says. He suspects that the placebo鈥檚 variable effects are most troublesome with 鈥渇irst-time鈥� doses because patients鈥� previous experiences of pills and injections are also highly variable. And he aims to iron out these differences 鈥� not by switching pill colours, or sizes, but by changing how or when the mock medicine is administered.

Timing is crucial. Give a patient placebo injections one week and jabs of real morphine the next and the conditioning will work to eliminate part or all of the real drug鈥檚 effect. But reverse the order and the placebo jabs will produce some of the effects of morphine. 鈥淵ou can easily wash out the drug,鈥� notes Voudouris, 鈥渂ut you don鈥檛 wash out the conditioning effects of the drug.鈥� And this, he says, raises the prospect of giving patients suffering from chronic illnesses mixed doses of drug and placebo, while easing their pain to the same extent.

First, however, there is the problem of deception to solve. There can be no doubt that what doctors tell their patients can make all the difference between success and failure. Three decades ago, researchers led by Richard Sternbach, then at the Massachusetts Mental Health Center in Boston, laid the groundwork for placebo studies when they looked at stomach contractions. The team found that they could manipulate the response of patients to inactive pills with just a few suggestive words: the speed of contractions increased, decreased or stayed put depending on what patients were told the drug was designed to do.

How, then, can honest doctors make placebos work? Once again, Voudouris believes the answer lies in conditioning. Doctors can be open about using placebos, he argues, provided they also take the trouble to condition patients to respond to the treatments. His evidence comes from some student guinea pigs to whom he gave mild electric shocks with electrodes strapped to their forearms.

To condition the students, Voudouris and his colleagues applied medicinal-smelling, but otherwise ordinary pink cold cream to the students鈥� arms before they turned the current on. One group was told that the cream was a powerful painkiller, another that it was just a placebo. What none of the subjects knew, though, was that during the conditioning session, the researchers were turning the current down to half its former level whenever the cream was applied.

The point of the exercise was to 鈥渢each鈥� the subjects by experience that the cream curbed the pain. And it worked. When the current was cranked up to its previous level, subjects reported up to 80 per cent less pain when the cream was applied 鈥� even those students who had been told the cream would be useless.

This experiment still required a degree of deceit. It would probably have flopped if the researchers hadn鈥檛 concealed the fact that the electric current was halved during the conditioning phase. But it establishes an important principle, says Voudouris 鈥� namely, that you can admit you are using a placebo without destroying its potency.

The next step, testing the power of conditioning on real patients, will involve no deceit at all. Quite the opposite. Voudouris and his colleagues will be scrupulously honest with the 60 or so patients with arthritis, back pain and other ills who agree to take part in their clinical trial. They will all sign a treatment contract explaining that patients will sometimes be given placebos instead of their usual painkiller, which in most cases will be morphine and its derivatives. But before any switch to placebo takes place, all patients will spend weeks receiving the genuine article.

According to previous studies by Voudouris and his team, patients conditioned this way should still be able to respond to a fake drug even though they know placebos are being used. The only proviso is that nobody must know when the substitution takes place.

Side effects

This last condition, however, may be harder to fulfil than it sounds. 鈥淚t鈥檚 going to be hard to keep patients and doctors blinded long term, particularly if the drugs have strong, recognisable side effects,鈥� warns Charles Weijer, a doctor who examines the ethics and science of placebos at McGill University in Montreal. And Voudouris isn鈥檛 making it easy for himself. Chronic pain sufferers are, by definition, those that medicine hasn鈥檛 been able to permanently cure. So their incidental conditioning, before the trial starts, tells the patients that all medicine ultimately fails.

Voudouris admits that no one really knows whether such obstacles can be overcome with his therapy. 鈥淲e鈥檙e in completely uncharted territory here,鈥� he says. 鈥淭here are lots of problems we鈥檒l encounter.鈥�

There is also little doubt that critics will question whether it鈥檚 right to give patients inactive pills in place of tried and tested drugs. At present, the only sanctioned public use in science is in conventional clinical research. Drugs companies like to demonstrate the effectiveness of experimental drugs by showing that they can outperform placebos in clinical trials. Even here, there are ethical worries, especially over studies where patients are assigned a placebo instead of a therapy that is already known to be effective. Wouldn鈥檛 doctors who doled out placebos in everyday clinics invite even sterner rebuke?

Moral quagmire

Not necessarily, argues Weijer. While he is rather critical of the unethical use of placebos in trials, Weijer believes that Voudouris has managed to sidestep much of the moral quagmire. For a start, the Australian鈥檚 aim is not to usurp effective treatments with placebos but to mix the two. 鈥淭hese patients will not be deprived of medicine, they will not be lied to, and they may get some benefit,鈥� says Weijer. 鈥淭he ethics shouldn鈥檛 be problematic.鈥�

Voudouris would go further and turn the ethical argument on its head. If placebos really do work and patients receive the same benefits for less money and with smaller doses of a noxious drug, it may one day be considered unethical not to use them. Indeed, Voudouris foresees a time when nearly every prescription includes a premixed packet of sugar pills 鈥� and no doctor is embarrassed to prescribe it. He does believe, however, that there would have to be strict limits on the kinds of diseases that could be treated this way.

鈥淭here are definitely cases where we know that a placebo isn鈥檛 going to work,鈥� says Ader. It is impossible, for example, to condition a placebo response for something with no effective treatment. That means no placebo conditioning for intractable cancers.

If Voudouris鈥檚 clinical trial is successful, expect to hear a lot more talk about placebo therapies. But don鈥檛 expect a sudden medical revolution. The researchers will still have to persuade clinicians and patients that the benefits they offer aren鈥檛 the result of magic or mental illness but of biomolecular machinery 鈥� however poorly understood. Only then will diehard sceptics refrain from drawing disparaging parallels between placebo users and healers waving chicken鈥檚 innards over prostrate believers. And for that final step, those pulling for the placebo will need more than good science on their side: they鈥檒l need some good PR.