The first rule of medicine is, "do no harm." Perhaps accepting this precept has produced a logic for disease treatment where our primary purpose is limited to only assisting or helping to correct a malfunctioning system. Can we use drugs that, according to traditional views, would be considered to increase stress on the system in the short term, to actually treat and cure disease in the long term? Is it possible to exacerbate disease for a longer-term gain? Although there are several examples of where this strategy has appeared to work, a systematic testing of the hypothesis has not occurred and, for the majority of diseases, this hypothesis has never been tested.

β-Adrenoceptor antagonists, also called "β-blockers," were originally contraindicated in patients with congestive heart failure (CHF). In the heart, β₁ and β₂-adrenoceptors increase contractility and rate when stimulated by adrenaline or noradrenaline. In CHF, the body's response to the diminished output from the heart is to increase the release of adrenaline and noradrenaline, thereby stimulating cardiac contractility. Therefore, logically, it was thought that to impede contractility further by blocking β-adrenoceptors would exacerbate the disease. However, it became apparent that the body's response of increasing the release of adrenaline and noradrenaline, although beneficial in the short term, led to desensitization of cardiac β-adrenoceptors in the long term and impairment of signal transduction [1]. Thus, β-blockers were used in CHF to possibly "protect" β-adrenoceptors from desensitization. Although the use of β-blockers might have stopped the desensitization of cardiac β-adrenoceptors, it is known that a receptor occupied by an antagonist cannot stimulate cardiac contractility. However, some drugs from this class have, indeed, been shown to increase contractility and decrease mortality in CHF patients (table 1). Importantly, although β-adrenoceptors are desensitized in patients with CHF, these receptors still respond to agonist infusions with an increase in contractility. Therefore, the paradox remains as to why impeding a contractile system results in an increase in contractility. Although various theories exist on why β-blockers are useful in CHF, on the surface these effects appear to be paradoxical.

Other Apparent Paradoxes

Other isolated instances exist where paradoxical pharmacology has become accepted. For example, stimulants such as methylphenidate and amphetamines are used to treat hyperactivity in children [7], and skin irritants such as retinoic acid and benzoyl peroxide are used to treat acne, which is an inflammatory skin condition [8].

There are reports that Alzheimer's disease (AD) and other neurodegenerative dementias can be associated with a hypofunction of the dopamine system [9-11]. There are also reports of a decreased incidence of AD and other neurodegenerative disorders in schizophrenic patients [12,13]. The mainstay of pharmacological treatment of schizophrenia is dopamine receptor antagonists. The authors of one study wrote: "Our findings, as well as those of Purohit et al., prompt the speculation that the neurobiology of schizophrenia (or even its treatment with anti-psychotic medication) may 'protect' against the development of AD pathology" [12]. Schizophrenia is a complex disease that is often misdiagnosed, and it appears more likely that the simpler explanation of the "protective" effect is the common treatment of schizophrenia, rather than the disease itself. Thus, chronic blockade of a system might prevent a hypofunction of the same system.

Ischemic preconditioning can also be viewed as an apparent paradoxical response: exposure of the heart to brief episodes of ischemia protects the heart from cell death produced by a subsequent prolonged ischemic episode. Thus, "stressing" the myocardium results in protection [14].

In the field of immunology, the concept of challenging the immune system for a beneficial long-term response is well established. For example, vaccines, particularly those developed earlier, often produce fever and malaise, but there is a long-term gain.

Psychiatry and psychology also appear to suggest that paradoxical approaches can produce beneficial results in human response. In the disease of addiction, we now suggest not to enable, so that the addict must endure the consequences of their behavior. If the behavior increases the pain, the mind and body might be forced to compensate and change behavior for longer-term gain. Furthermore, branches of these disciplines have employed reverse psychology and paradoxical interventions successfully as treatment modalities.

This strategy has also been inadvertently used through our misuse of antibiotics. The repeated "stressing" of bacteria has produced stronger and more virulent strains. Perhaps it is time to test whether a multicellular organism can do the same.

Philosophical Arguments

We stress the cardiovascular system with exercise for longer-term gain. Why can a similar concept not apply in pharmacology? If we stress muscle we get better function. Such stress can result in beneficial compensatory effects such as angiogenesis. Is it possible to "pharmacologically stress" to gain long-term benefit? Can exacerbating a disease make use of the body's compensatory and redundant mechanisms to achieve a beneficial long-term response?

The practice of blood-letting, which Hippocrates wrote about, remained in use for over 2000 years. While fully acknowledging that blood-letting was borne out of ignorance, perhaps we threw the baby out with the bath water when we banished the practice of stressing systems as a possible cure. Perhaps all we need to do is to design more selective "leeches." Pharmacology has done that, but these drugs have not been systematically and rigorously tested as "leeches." Indeed, because for over 2,000 years the principal medical treatment was blood-letting, it is not surprising that the first precept of medicine became "do no harm." However, as scientists, we know the correct conclusion from the experiment was simply do not blood-let for most diseases. As a reviewer might say, the rest was speculation.

We exercise, we diet, we save money and we discipline, all to a degree of short-term discomfort for a longer-term gain. Perhaps medicine in general, and pharmacology specifically, should examine a strategy that works in so many other areas of our lives.

Possible Orthodox Explanations for Paradoxical Pharmacology

It is important to remember that paradoxical pharmacology will not be paradoxical if it works. Eventually the logical explanation for the apparent paradox will be determined. As previously stated, there are now numerous theories on how β-blockers exert their beneficial effects in heart failure and also how stimulants work in hyperactive children. One possible general hypothesis for why paradoxical pharmacology might work is the difference between the chronic versus the acute effect of drugs. In the case of the use of β-blockers in heart failure, initially the patient often reports feeling worse, but this is followed by an improvement of function and a decrease in mortality. Clearly, the chronic response differs from the acute response. There are numerous other examples of differences between acute and chronic responses. For example, classical antidepressants often take weeks of treatment to produce their effects. In addition, acute and chronic effects of drugs often produce opposite effects. This is particularly true for receptor-mediated events. For example, acute agonist exposure can produce activation of receptors and increased signaling, whereas chronic exposure can produce desensitization and decreased signaling. Indeed, this is precisely what happens in the progression of heart failure (table 2). Even the infusion of exogenous β-adrenoceptor agonists, although good in the short term (i.e. patient's symptoms are improved), increases mortality with chronic use. Finally, to again draw the parallel with exercise, acutely, exercise raises blood pressure, whereas chronically there is a decrease in blood pressure.

Although it is known that chronic and acute responses often differ substantially and, indeed, are often opposite in nature, most contraindications for drugs are predicted on the basis of the assumption that the chronic effect will be simply a more prolonged version of the acute effect. Experiments that analyze chronic effects are not performed, in part, because we cannot see how to get by the initial contraindication. Perhaps the effects of β-blockers in heart failure can teach us a lesson. When trials with β-blockers in heart failure were first started, the doses were too high and the drugs were poorly tolerated. However, it was determined that treatment should begin at a low dose and gradually the dose should be increased over a period of weeks. Therefore, perhaps the chronic experiment can be performed, if we start at a low dose and gradually increase the dose.

"Look For What Seems Out of Place" (Peter Murphy from "Cuts You Up")

Table 2 shows some of the similarities in outcomes when using β-adrenoceptor ligands in CHF and asthma. Admittedly, one difference between asthma and CHF is the hyperactivity of the sympathetic nervous system in CHF, which is not believed to be present in asthma. Therefore, following the determination of the consequences of chronic hyperactivity, it became logical to run the experiments with the chronic use of antagonists in CHF. In asthma, running the experiment would have to be more intuitive than logical. However, if one explained to a non-scientist the following: (1) we have two classes of drugs that have opposing effects, stimulants and blockers; (2) in this disease we know that short-term use of stimulants is beneficial, whereas long-term use is detrimental; and (3) we know that short-term use of blockers is detrimental, one would surmise that the non-scientist would say, "well, what does long-term blocker treatment do?" Table 3 shows the missing chronic experiment of intermittent agonist treatment in hypertension. Although it is known that chronic infusions of noradrenaline and adrenaline are deleterious and produce effects such as heart failure, we have not attempted to mimic exercise by performing chronic intermittent infusions. It is likely that a set of experiments has been omitted in the study of many diseases, based on the assumption that the chronic effect would be the same as the acute effect. But is this the logical interpretation? Have we become too bright to run the experiment?

Preventative versus Curative

The strategy outlined in this article might also help prevent diseases. Indeed, the strategy might work better prophylactically. The example given above with schizophrenic patients provides circumstantial evidence for this, because it is likely that many of the patients would have been started on anti-psychotic drugs before exhibiting signs of dementias. The theory for prophylactic use is that if we begin "stressing" the system before it "breaks," perhaps we can begin preparing the body for a period when the system will cease to work properly. Also, before disease sets in, there might be more unaffected compensatory or redundant systems that can be employed to offset the disease. The near completion of the Human Genome Project might provide us with very good clues for research on prophylactic treatments using this strategy.

Implications

First, this strategy will not work in many situations. For example, it is unlikely that increasing tumor size would be useful in treating cancer. Second, dosing will be a serious concern in certain diseases. For example, there are numerous single-dose studies that report severe reactions and even deaths following the administration of β-blockers to asthmatics. As stated, the lesson from β-blockers in heart failure is to start at a very low dose and increase the dose over a period of weeks; this might serve as a good general rule. There might also be dosing problems in the treatment of symptoms even if the strategy works. For example, if, theoretically, chronic use of β-blockers improved asthma symptoms in a subset of asthmatics, we would still be left with what to do when the acute attack occurs. Currently, the treatment of acute asthma attacks involves the use of β₂-adrenoceptor agonists and their effectiveness would be severely impeded by the presence of an antagonist. Although such concerns exist, there are also likely to be answers to these dosing problems. For example, in asthma, if the β-blocker was being taken systemically, the use of higher concentrations of the inhaled β₂-adrenoceptor agonist might be reasonable for an acute attack because cardiac β-adrenoceptors would also be blocked, and tachycardia from the absorbed agonist might not be a problem. Alternatively, phosphodiesterase inhibitors could be used to bypass the receptor altogether.

But, why pursue the strategy if it is risky? One reason would be for those diseases for which pharmacological intervention via the traditional "correcting" or "assisting" approach has been a relative failure. In 1999, the National Center for Health Statistics reported that in the U.S. death rates from asthma increased by 56% from 1970 to 1997. Clearly, current pharmacological interventions, although useful for treating symptoms, are not sufficiently effective against disease progression. Interestingly, until recently, pharmacological intervention in heart failure had also been a relative failure with respect to mortality.

In addition, this paradoxical strategy is worth pursuing for another reason: it might help accelerate acceptance of radical, yet established and proved, treatment modalities. There are now various published studies that show a definitive beneficial role for some β-blockers in heart failure (table 1). The first was published in 1996, yet it is estimated that despite the five years of dissemination and intense marketing by pharmaceutical companies, <30% of eligible patients with heart failure are taking a β-blocker [16]. The probable explanation is that it is very difficult to re-train physicians who where told for 25 years that β-blockers are absolutely contraindicated in patients with heart failure. With some studies reporting a 30-40% decrease in mortality, this lack of dissemination is probably costing human lives [16]. Increasing the number of examples of paradoxical results that are reported would probably lead to a faster acceptance of the results.

Where to Begin?

If acute versus chronic responses are often opposite in nature, and if the contraindications have been made based on the acute effects, there is a suggested list of where basic research can begin to look for clues to investigate paradoxical pharmacology. It is the list under "Contraindications" because the opposite of contraindicated is indicated. This is the list where one would have found β-blockers in CHF just a short time ago. I suggest we test the first precept of medicine, "do no harm," and determine its validity by performing basic research with paradoxical pharmacology. If medicine and pharmacology behave as other areas where short-term discomfort produces longer-term benefit, it might well be that we have paid a high price for accepting a presumption.

All wisdom is plagiarism; only stupidity is original.
                            - Hugh Kerr (professor, editor, preacher) [22]