Adrenaline and hypertension

Introduction Several studies based on the assessment of adrenaline kinetics, and performed both in experimental animal models of hypertension and in human hypertension, have prepared the ground for Esler and coworkers, in this issue of the Journal [1], to revisit the so-called ‘adrenaline hypothesis of hypertension pathogenesis’ [2]. Convincing evidence for cardiac adrenaline release has been provided in the aging process, heart failure states and panic disorders [3–5]. This new article by Rumantir et al. [1] clearly demonstrates that the hypertensive state is characterized by an augmented cardiac adrenaline release, which is coupled to an increased release of noradrenaline from the heart, confiming the previous demonstration by the same group of investigators that noradrenaline release is augmented in hypertension [6]. Background The hypothesis that adrenaline is involved in the pathogenesis of the essential hypertensive state was first proposed approximately 20 years ago by Brown and Macquin [7]. Originally, the hypothesis included three major assumptions. First, that adrenaline promotes noradrenaline release from sympathetic nerve terminals via stimulation of prejunctional β-adrenoceptors. Second, that adrenaline release from the adrenal medulla (and/or from other cardiovascular regional districts) is enhanced during the development of the hypertensive state. Finally, that the augmented secretion of adrenaline with time contributes to the development of the hypertensive condition. Regarding the first assumption, it has been claimed that endogenous adrenaline, after being taken up by sympathetic nerve endings, is co-released with noradrenaline from the adrenergic vescicles and stimulates presynaptic β-adrenoceptors to facilitate noradrenaline release. This process, which has been repeatedly documented in isolated tissues and in experimental animals [7–9], has also been extensively studied in man to determine whether and to what extent it plays a functional role in physiological or pathophysiological conditions. The human investigations, however, have provided conflicting results, with some studies showing that exogenous adrenaline administration may augment noradrenaline release and potentiate vasoconstrictor responses to a variety of neurogenic stimuli [10–12] (thus confirming, although indirectly, the facilitatory effects of adrenaline on sympathetic vasoconstrictor tone), while others have produced negative results [13–15]. As far as the second assumption of the adrenaline hypothesis is concerned, it would be expected that circulating levels of adrenaline are increased in the early hypertensive phases, i.e. in borderline hypertension or in the mild hypertensive state of young patients. The results of studies and meta-analyses published to date lend some support to the hypothesis by showing that indeed plasma adrenaline levels are often augmented in hypertensive patients compared with age-matched controls, but a lack of intra-individual concomitant increase of both catecholamines casts doubt on the validity of the first assumption, i.e. that co-release of adrenaline facilitates noradrenaline release [16]. Finally, the third assumption of the adrenaline hypothesis implies that the increased plasma levels of this adrenergic neurotransitter would trigger with time a stable blood pressure elevation. Although difficult to be reconciled with the classical haemodynamic effects of adrenaline, i.e. a modest increase in systolic blood pressure, a slight reduction in diastolic blood pressure coupled with muscle vasodilatation, the validity of the hypothesis has been suggested by some studies performed during the past few years [17–19]. Again, however, the evidence collected cannot be considered unequivocal because some reports failed to demonstrate a delayed pressor action of adrenaline when infused in the systemic circulation (the so-called ‘epinephrine after-effects’) [20]. The new evidence In their article, Rumantir and coworkers [1], by employing the noradrenaline and adrenaline radiolabelled technique, investigate whole body and cardiac clearance and spillover rate of catecholamines in a group of middle-age essential hypertensive patients, and compare the results with those obtained in age-matched normotensive controls. The main findings of the study can be summarized as follows. First, the hypertensive state, at variance from what is found in the normotensive condition, is characterized by a consistent release of adrenaline from the cardiac sympathetic nerves. Second, in the hypertensive patients, this is accompanied by an increased spillover of noradrenaline from the same cardiovascular regional districts. Third, systemic adrenaline spillover is somewhat greater (although not significantly) in hypertensive subjects compared with the normotensive patients. Fourth, in the hypertensive group, cardiac release of adrenaline is significantly and directly related to noradrenaline release. Can these results be interpreted as in favour of the ‘adrenaline hypothesis’? The answer is probably negative because, as properly emphasized by the authors, the data obtained in their study do not allow a definite ‘verdict’ in favour of (or against) the adrenaline hypothesis. They do, however, add a new piece of evidence to the current knowledge, by documenting that, in the hypertensive state, a cardiovascular regional district of crucial importance for cardiovascular homeostasic control (i.e. cardiac) is indeed characterized by an enhanced release of both adrenaline and noradrenaline, thus confirming the validity of the second assumption of the adrenaline hypothesis. The unsolved issues As noted by Rumantir and coworkers [1], one concern regarding the adrenaline hypothesis is the variable relationship between adrenaline and noradrenaline spillover. Although, in their article, the range of cardiac adrenaline release was larger in normotensives compared with hypertensives, there was no correlation with cardiac noradrenaline spillover in normotensives but a significantly positive correlation in hypertensives. Furthermore, a recent study from the same laboratory on patients with panic disorder [5] showed that this condition is associated with normal levels of cardiac noradrenaline spillover despite an augmented cardiac adrenaline release at rest. Finally, panic attacks coinciding with this investigation induced dramatic increases in total body and cardiac adrenaline spillover, but only a minor increase in noradrenaline spillover. Thus, while it is tempting to speculate that the adrenaline mechanism indicated by Rumantir et al. may contribute to the differences in development of hypertension demonstrated between secluded nuns and laywomen [21], as well as between Kuna Indians remaining in their traditional environment and those moving to a Westernized society [22], the maintenance of normal cardiac noradrenaline release and normal blood pressure levels in patients with panic disorder of long duration raises the possibility that the response to neural adrenaline release, rather than the adrenaline release ‘per se’, is the factor distinguishing those that develop hypertension from those that remain normotensive in a stressful environment. A second question regarding the ‘adrenaline hypothesis’ concerns regional differentiation. A recent study by Johansson et al. [23] demonstrated that hypertension is associated with increased renal adrenaline release at rest, but no correlation between renal adrenaline and noradrenaline spillover was found. Furthermore, exercise induced a parallel increase in cardiac adrenaline and noradrenaline spillover in healthy subjects (cardiac catecholamine kinetics not investigated in hypertensives), whereas renal adrenaline spillover decreased and noradrenaline spillover increased in both hypertensives and normotensives during exercise. Although transmitter handling apparently differs between the heart and the kidneys, these observations raise the question of whether the coupling between adrenaline and noradrenaline release, demonstrated by Rumantir and coworkers, is specific for the heart. If this is the case, its importance for the development of hypertension must be tempered accordingly. While intricate and invasive methods such as the isotope dilution technique yield a limited number of observations, they enable a mechanistic insight that is not accessible with less invasive techniques. Although the cautious interpretation advocated by Rumantir and coworkers is clearly warranted, their findings pave the way for ‘the adrenaline hypothesis of hypertension’ to join the ranks of possible mechanisms contributing to the multi-factorial entity of human essential hypertension. Future studies must address the regional differentiation of this adrenaline-noradrenaline relationship at rest, as well as in response to various stimuli.