* From the E.Wolfson Medical Center, Holon, Israel

SUMMARY

Placebo (P) is a widespread phenomenon in medicine and biology but its mechanisms are only partially understood. According to the recent definition, P is any substance devoid of specific activity, which nevertheless, is given to a patient in order to obtain an effect. The latter, named placebo-effect (PE), is not predicted. PE may or may not occur and may be favorable or unfavorable. Negative effects of P are called nocebo. In extreme cases nocebo may lead to severe pathological conditions. The paradox is that P being a neutral inert substance may have unpredictable, and sometimes very powerful effects. Nowadays, the nature of PE is explained by meeting the patient's expectations or by decreasing his anxiety. It has also been suggested that the human response to P is similar to a conditioned response. The author proposes an alternative view of the mechanism of PE, that takes into account the modern theory of self-organization of living systems. The concept of a living organism as a self-regulatory system, and disregulation model of illness and disease could elegantly explain the mechanism of PE. Assessment of self-regulatory capacities of the main functional systems could open new prospects for usage of P for effective treatment of psychosomatic diseases.

Key words: placebo, placebo effect, psychosomatic diseases

INTRODUCTION

P and PE are integral part of every day clinical practice. It is extremely important for health care practitioners to well understand this phenomenon. The present revival of the interest in this topic has been caused by i) intensification of the pharmacological research and clinical trials; ii) new findings in the psychological, sociological, and biological research, iii) a broader view of disease and health, including modern concepts, principles, and philosophies of etiology, diagnosis, and therapy (1).

Nevertheless, we are still far away from a well-documented theory of PE. The approach to the subject being still contradictory, I would like to dwell upon the following subjects:

* Definitions of P and PE.

* Character of PE. Nocebo. "Risk-factors" for a P response.

* Modern concepts of P mechanisms.

* The role of Ps in clinical research: theoretical and practical.

* The different roles of P in medical practice and research.

* Ethical aspects.

* Prospects for the future: author's view.

DEFINITION OF PLACEBO AND PLACEBO EFFECT

P and PE has been known in medicine for more than two hundred years. The word "placebo" first appeared in a medical dictionary in 1785 as a "commonplace method of medicine", and soon after, as "a make-believe medicine". Untill now there has not been an exact definition of P. Nowadays a few definitions of P are accepted.

The Dorland's Medical Dictionary (1988) defined P as "a medical preparation having no specific pharmacological activity against the patient's illness or complaints given solely for the psychophysiological effects of the treatment".

The latest literature defines P as "a dummy treatment administered to the control group in a controlled clinical trial in order that the specific and non-specific effects of the experimental treatment can be distinguished. Ps are putatively inactive substances tested in controlled studies for comparison with presumed active drugs or prescribed for relief of symptoms or to meet a patient's demands" (1).

Nevertheless, many recent articles on the subject include a broader definition like the one given by Shapiro: "P includes all medical procedures no matter how specific - oral and parenteral medication, topical preparations, inhalants, and mechanical, surgical and psychotherapeutic procedures" (2, 3).

P must be differentiated from PE, which may or may not occur and which is bi-directional and may be favorable or unfavorable. PE is defined as the changes produced by Ps. According to Byearly, PE is defined as "any change in a patients symptoms that is the result of the therapeutic intent and not the specific physiochemical nature of a medical procedure". This definition summarizes the attitude of modern medicine to P as a useless and undesired side effect of treatment. The definitions of P and PE are confused. The fact that PE is better than no treatment does not make P an active substance (4). It is paradoxical that P, a neutral inert substance, should be capable of having a very powerful effect, which can sometimes surpass the specific treatment effect (5).

There are three general reasons for clinical improvement in a patient's condition:

1. natural history and regression to the mean;

2. specific effects (verum-specific) of the treatment;

3. non-specific effects of the treatment, attributable to factors other than the specific active components. The latter effect includes PE.

Ps can produce both desirable and adverse reactions. Adverse reactions from Ps are named nocebo effect. This term was introduced by Kennedy in 1961. The most common adverse reactions to P are drowsiness (50%), headache (25%), fatigue (18%), sensation of heaviness (18%), difficulty in concentrating (15%), nausea (10%), sleep disturbance (10%), dry mouth (9%) (3).

The rate of PE is high in psychosomatic diseases. Medical literature is replete with reports on the power of P in helping patients with anxiety, tension, melancholia, shizophrenia, all sorts pain, headaches, cough, insomnia, seasickness, chronic bronchitis, common cold, arthritis, peptic ulcer, hypertension, nausea, senile dementia.

For example, in cardiovascular diseases the rate of improvement of symptoms both subjective and objective has been assessed to be 30% to 80% (3).

Some people seem to be more prone to influence of P than others. These differences may relate to the recipient's personality. Probably, the most important factors here are those relating to specific attitudes toward the illness, medication, and physician.

Investigations documented, that 50% of healthy volunteers with competitive and aggressive behavior pattern had had a subjective nocebo response to P (6, 7). Other authors, have reported such predictors of positive P response to antidepressant treatment, as matrimony, mild severity of illness, ages of 60 and above (8).

Still others emphasized the importance of PEs in pain treatment and research. They also indicate that individuals are not consistent in their P responses. Yet, there was an attempt to point out the role of anxiety, expectations, and learning in PE (9).

MODERN CONCEPTS OF MECHANISMS OF PLACEBO EFFECT

Why inert substances can be effective is unknown. One theory to explain PE is the classical conditioning. After an unconditioned stimulus has been continually accompanied with a conditioned one, the conditioned stimulus alone can elicit the same response as the unconditioned stimulus. In this way the environment affects behavior (10).

Another theory to explain PE is that P acts through meeting the expectations of the patient. It is thought that a person's beliefs and hopes about the treatment, as well as his/her sensory experience and thoughts can affect the body's neurochemical and other biochemical systems, including the hormonal and immune systems.

Some explain P-response by neurohormone mechanisms. Endogenous opioid peptides play a variety of roles in the CNS, from development to immune modulation. These functions are mediated mostly via specific opioid receptors localized in different brain regions and cells. Endogenous neuroendocrine polypeptides, including beta-endorphins, enkephalins, and antiopioids, are activated by many factors. These factors include Ps, vigorous exercise, and other stressors.

Endogenous opioid system (EOS) is both related to psychological events and closely connected with other neurotransmitter systems in the brain (11). Therefore, interaction between EOS and different neurotransmitter systems in the brain mediates PE on mood and behavior of both healthy and sick people.

A third model of PE is based on the fact that mental imagery could produce specific and measurable physiologic effects. This model explains the relation between psychological and physiological components of PE. In the brain, psychological P related imagery into a physiologic P response, i.e. there my be "a placebo effect without the placebo" (12).

In fact, the newly identified neuroimmunomodulatory (NIM) pathways may be one of the "non-specific mechanisms" (NSM) by which Ps exert their effects. Activation of these pathways can produce dramatic changes in immunity and in some circumstances, may be as powerful as a direct pharmacological action, or even more than that. It is supposed (6) that PE can be delivered through the expectation of benefit via higher centers, activating NIM pathways. The therapies, including alternative therapies, may work through this higher center P-expectancy effect, but they may also activate NIM pathways directly, without necessarily involving the patient's expectation (6). The activation of NSM including NIM pathways is likely to differ with individuals. This may account for variability of PE and the capability of P to sometimes excel the therapy in a double-blind trial.

As for nocebo-effect, it was believed to be similarly produced by conditioned responses but activated by negative expectations. Nocebo-stimuli such as anxiety, fear, doubt, etc. may both reduce PE and induce negative side-effects.

Nevertheless, the nocebo effect is not the dark side of PE. In recent years, some authors shown that neuropeptide cholecystokinin (CCK) has an antiopioid effect and increases anxiety (13, 14, 15). It is also documented that CCK antagonist proglumide blocks nocebo through a mechanism not involving opioids (13). They also point out, that P analgesic response can be modulated in two opposite directions: it can be partially abolished by opioid antagonist naloxone and potentated by proglumide. P activation by proglumide occurred only in P responders, but not in non-responders. This fact suggests that activation of an EOS is, nevertheless, a necessary condition for proglumide activity (15).

In my opinion, it is very important to investigate the role of nitric oxide (NO) in the placebo-nocebo effects mechanism. As it has been documented, NO may serve as an important regulator of basal CCK release (16). On the other hand, NO is implicated in the action of opioids, mediating the morphine effects (17).

THE ROLE OF PLACEBOS IN CLINICAL RESEARCH:

Theoretical aspects

To assess the effectiveness of a new medication, both positive and negative PE must be ruled out as a negative "noise". This noise includes all non-specific factors that influence the effectiveness of the treatment.

In clinical research are accepted two main models of efficacy (6): pragmatic efficacy model (PEM) and fastidious efficacy model (FEM).

PEM compares two treatments under clinical conditions. In terms of P this model of randomized clinical trial usually assumes that various non-specific effects interact with themselves and with any specific effect. Therefore, it is difficult to make a clear distinction between verum effect and P. This model provides less scientifically useful, but more clinically relevant information.

FEM is accepted in research as a tool for randomization and blinding. This model is a foundation stone of the modern classical pharmacological research. FEM assumes that verum-specific effects and non-specific effects are separable. The total outcome effect is equal to its active effect plus its P effect. The concept of FEM is based on application of laboratory experiments to ill human beings. This model is based on several unproven assumptions (6):

- a part of the medicine response (MR) results from PE,

- the same amount of MR results from a PE as in a P group or in the P phase of trial;

- in patients receiving both medicine and P, the baselines and conditions were the same;

- PE and MR are linear, and PE is relatively constant in magnitude throughout the trial.

However, the interaction of the trial medicine and P is complex and some of these assumptions are not valid. According to this model, efficacy beyond the PE becomes an inherently unanswerable question.

Practical aspects

First of all, it must be emphasized that subjective benefit is not equivalent to a PE. P triggers the body's natural non-specific response. In this sense P as a biological trigger is formally an active substance. From this standpoint, all the non-specific factors may be triggers of non-specific effects and may be defined as natural healing force.

Secondly, many factors that are classified as non-specific may have a pathogenic effect that has not yet been estimated. For example, a stable vasodilative effect after cardiac catheterization could be explained by mechanical stimulation of the endothelium function induced by the procedure. Another example might be "non-specific" vasodilative action of a meal also as a result of specific interaction between gastro-intestinal hormones, endorphins, and nitric oxide.

In clinical trials aimed to assess the efficacy of new pharmacological therapies, several types of controls may be used, namely (18):

- P concurrent control;

- dose comparison concurrent control;

- no-treatment concurrent control;

- active-treatment concurrent control;

- historical control.

DIFFERENT ROLES OF PLACEBO IN MEDICAL PRACTICE AND RESEARCH

Currently, the main application of P is in clinical research. In clinical trials Ps are given with a different goal than in medical practice. Here P is given with the expectation that, in addition to the specific treatment effect, the patients will benefit from non-specific effects of P. In clinical practice PE should be maximized, whereas in clinical trials one is interested in increasing specific effects only. Minimizing PE in clinical trials leaves more "therapeutic space" for specific effect.

The term PE refers to diverse non-specific, desired or non-desired, effects of substances or procedures and interactions between the patient and the therapist. The truth is that PE is counterfeited by a variety of factors, including the natural history of the disease, regression to the mean, concomitant treatments, obliging reports, experimental subordination, methodological defects in the studies (18).

It is possible to exclude all of the mentioned factors using

- an untreated control group (3 arm trials: untreated-placebo-verum)

- 4-arm trials for testing interaction of specific and non-specific factors named "balanced placebo design". The model is a factorial design consisting of a pharmacological and non-pharmacological intervention.

In a lot of clinical trials it is more preferable to use a standard treatment as a reference factor.

ETHICAL ASPECTS

In clinical practice and research, P as a control intervention is ethically acceptable in the following cases (2, 3):

1. No "gold standard" intervention is available which causally interferes with the underlying condition.

2. There is no adequate therapy for the disease and/or active therapy could presumably have serious side-effects.

3. P treatment should not last too long.

4. P treatment should not inflict unacceptable risks.

5. The experimental subject should be adequately informed and an informed consent given.

Therapeutic and diagnostic use of P is ethically acceptable only in the same well-defined cases. While "therapeutic" application of P almost inevitably implies deception, this is not the case for its use in research. Unpredictability of PE necessitates placebo-controlled designs for most trials.

According to FDA rules, using P in clinical trials is desirable but not obligatory.

PROSPECTS FOR THE FUTURE: AUTHOR'S VIEW

Healing process is based on phylogenetic non-specific mechanisms of recovery using homeostatic pathways common for mammals. PE is an integral part of this natural process.

In my opinion, it may be useful to analyze the mechanism of PE using a new conception of a living organism as a self-regulating system. This conception suggests, that the principle of self-organization toward a pathological state is discrepancy between the virtual process and the actual one. The adaptive systems of the organism tend to evolve so that this discrepancy may be minimized (19).

Taylor (20) has proposed a disregulation model of illness and disease. This model attributes somatic disorders to a variety of regulatory disturbances, including deficits in affective regulation and in the ability to maintain a stable and cohesive self. To repair deficits in a patient's self-organization, chaotic impulses and various emotions could be evaluated for their signal function and modulated through imagination activity and communication with others.

Analysis of the patient's pathologic internal object relations could modify his self-organization and enhance his self-regulatory capacities (20).

Investigation of PE on the circadian rhythm of temperature and hand-grip strength rhythms indicated that desynchronized circadian temperature rhythm was obliterated by Ps. Desynchronized circadian rhythms of left and right hand-grip strength were all obliterated by P in men only (21).

In my opinion, endothelial cell function may be the final common pathway for organ self-organization and integration. Morphologic and functional changes of the endothelial cells can be directly responsible for triggering the metabolic process.

Assessment of self-regulatory capacities of the main functional systems could open new prospects for usage of PE in treatment of psychosomatic diseases.

1. Straus JL. Placebo effects. Issues for clinical practice in psychiatry and medicine. Psychosomatics, 1996, 37, 4: 315-26

2. DeDeyn PP, D'Hooge RR. Placebo in clinical practice and research. J Med Ethics, 1996, 22: 140-6

3. Bienenfeld L, Frishman W, Glasser S. The placebo effect in cardiovascular disease. Am Heart J, 1996, 132, 6: 1207-21

4. Zajicek G. The placebo effect is the healing force of nature (Editorial). The cancer J, 1997, 8:1-3

5. Watkuns AD. The role of alternative therapies in the treatment of allergic disease. Clin Exper Allergy, 1994, 24: 813-25

6. Kaptchuk TJ, Edwards R, Eisenberg DM. Complementary medicine: efficacy beyond the placebo effect. Complementary medicine. An objective appraisal. Ernst Edzard (ed.), Butterwort Heinemann, Oxford, 1997, p 42-70

7. Drici M-D, Raybaud F, De Lunardo C, et al. Influence of the behavior pattern on the nocebo response of healthy volunteers. Br J Clin Pharmac, 1995, 39: 204-6

8. Wilcox CS, Cohn JB, Linden RD, et al. Predictors of placebo response: a retrospective analysis. Psychopharmacology Bulletin, 1992, 28: 157-62

9. Turner JA, Deyo RA, Loeser JD, et al. The importance of placebo effects in pain treatment and research. JAMA, 1994, 271, 20: 1609-14

10. Voudouris NJ, Peck CL, Coleman G. The role of conditioning and verbal expectancy in the placebo response. Pain, 1990, 43: 121-8

11. Sher L. The placebo effect on mood and behavior: the role of the endogenous opioid system. Med Hypotheses 1997, 48: 347-9

12. Egbert ID, Battit GE, Welch CE, Bartlett MK. Reduction of post-operative pain by encouragement and instruction of patients. N Engl J Med, 1964, 270: 825-7

13. Beneditti F, Amanzio M, Casadio C, et al. Blockade of nocebo hyperalgesia by the cholecystokinin antagonist proglumide. Pain, 1997, 71: 135-40

14. Beneditti F, Amazio M. The neurobiology of placebo analgesia: from endogenous opioids to cholecystokinin. Prog Neurobiol, 1997, 52: 109-25

15. Beneditti F. The opposite effects of the opiate antagonist naloxone and the cholecystokinin antagonist proglumide on placebo analgesia. Pain, 1996, 64: 535-43

16. Mangel AW, Scott L, Prpic V, Liddie RA. Regulation of cholecystokinin secretion in STC-1 cells by nitric oxide. Am J

Physiol, 1996, 271(4Pt1): G650-4

17. Kivastik T, Rutkauskaite J, Zharkovsky A. Nitric oxide synthesis inhibition attenuates morphine-induced place preference. Pharmacol Biochem Behav, 1996, 53, 4: 1013-5

18. Kienle GS, Kiene H. Placebo effect and placebo concept: a critical methodological and conceptual analysis of reports of the placebo effect. Altern Ther Health Med, 1996 2, 6: 39-54

19. Ito K, Gunji YP. Self-organization of living systems towards critically at the edge of chaos. Biosystems, 1994, 33: 17-24

20. Taylor GJ. Clinical application of a dysregulation model of illness and disease: a case of spasmodic torticolis. Int J Psychoanal, 1993, 74, (Pt 3): 581-95