Cardiac natriuretic peptides during exercise and training after heart transplantation

Time for primary review 30 days. The cardiac natriuretic system, composed of atrial and brain natriuretic peptides (ANP and BNP), plays a major role in blood pressure and fluid homeostasis, protecting the organism from volume and pressure overloads. Accordingly, the cardiac hormones have been shown to delay the occurrence of overt heart failure through their diuretic, natriuretric and vasodilatory properties [1–3]. Adrenomedullin, mainly produced by vascular smooth muscle cells and by vascular endothelial cells, is also secreted by the failing human heart. This potent vasorelaxing and natriuretic peptide can thus be considered as a third cardiac hormone, involved in circulation control [4,5]. Short term survival is no longer the pivotal issue for most heart-transplant recipients (Htx) because of enhancement in organ preservation, surgical and medical therapies. Consequently, improving quality of life after heart transplantation arises as an important goal, which might be reached through exercise and training programs [6–8]. Since cardiac natriuretic peptides greatly participate in cardiovascular adaptations, it appeared interesting to focus this review on ANP, BNP and adrenomedullin (ADM) responses to exercise and training after heart transplantation. Relative rather than absolute work load will be used to compare groups since it is generally considered to be a better indicator of the magnitude of fluid-regulating hormone changes, both in normal subjects and in Htx [9]. After discussing why circulating ANP, BNP and ADM are elevated in Htx, we will investigate the stimuli of their secretion during exercise and present the few data available concerning their response to exercise training after heart transplantation. After successful cardiac transplantation, normalization of filling pressures, as well as normalization of the renin–angiotensin–aldosterone and the sympathetic systems, generally occur, explaining why persistent elevation of circulating ANP and BNP was unexpected [10–13]. In fact, replacing the failing heart by a so-called ‘normal’ heart does not totally restore the cardiovascular system and we yet know that the denervated transplanted heart is characterized by diastolic dysfunction. Diastolic cardiac dysfunction, together with vascular dysfunction (preexistant before transplantation and enhanced by the immunosuppressive therapy), associated with volume and/or pressure overload, appear thus to play a key role in the cardiac natriuretic system stimulation observed in Htx. Endothelin participates also likely in such increase after heart transplantation, either directly or through its hypertensive and renal deleterious effects [10–16]. Circulating adrenomedullin has also been recently shown to be increased after heart transplantation, likely in relation with endothelin, hypertension and increased left ventricular mass index [17,18]. In healthy humans, ANP has been shown to increase during and immediately following exercise [19,20]. Rather than an increase in atrial pressure, the predominant stimulus for the cardiac hormone release is an increase in atrial stretch secondary to increased venous return [21,22]. The first report on ANP response during exercise after heart transplantation supported an ANP hypersecretion in Htx, as compared to controls [23]. Controversial data have nevertheless been reported thereafter (Tables 1 and 2), challenging the proposed factor (altered atrial anatomy), whose effects are modulated by humoral (catecholamines, arginine vasopressin and endothelin) or nervous factors (cardiac denervation), responsible for an eventual ANP hypersecretion during exercise after heart transplantation [9,24–34]. Selected data from studies investigating the ANP response to submaximal exercise in controls and heart transplant recipientsd MAP was calculated from mean systolic and diastolic blood pressures. Values were estimated from graphs. Heart rate was recalculated from author's data. Values are mean±S.E.M. Delay is time elapsed since transplantation. Δ, relative increase between peak and rest values; MAP, mean arterial blood pressure; Htx, heart transplant patients; Ctrl, health control subjects; Submax, submaximal exercise. Selected data from studies investigating the ANP response to submaximal exercise in controls and heart transplant recipientsd MAP was calculated from mean systolic and diastolic blood pressures. Values were estimated from graphs. Heart rate was recalculated from author's data. Values are mean±S.E.M. Delay is time elapsed since transplantation. Δ, relative increase between peak and rest values; MAP, mean arterial blood pressure; Htx, heart transplant patients; Ctrl, health control subjects; Submax, submaximal exercise. Selected data from studies investigating the ANP response to maximal exercise in controls and heart transplant recipientsd MAP was calculated from mean systolic and diastolic blood pressures. Values were estimated from graphs. Values are mean±S.E.M. Delay is time elapsed since transplantation. Δ, relative increase between peak and rest values. MAP, mean arterial blood pressure; Htx, heart transplant patients; Ctrl, health control subjects; Max, exercise was performed until exhaustion. Selected data from studies investigating the ANP response to maximal exercise in controls and heart transplant recipientsd MAP was calculated from mean systolic and diastolic blood pressures. Values were estimated from graphs. Values are mean±S.E.M. Delay is time elapsed since transplantation. Δ, relative increase between peak and rest values. MAP, mean arterial blood pressure; Htx, heart transplant patients; Ctrl, health control subjects; Max, exercise was performed until exhaustion. The first factor proposed to explain ANP hypersecretion during exercise in Htx was increased atrial volume and mass resulting from the surgical procedure [23]. Indeed, according to Laplace's law, higher atrial volume induces higher diastolic and systolic wall stress, which is known to be an important determinant of ANP secretion [1]. Moreover, an increased atrial mass may augment the heart's ability to release ANP. Nevertheless, maximal atrial ejection force is similar in Htx and controls [35] and ANP hypersecretion was also observed in Htx with total excision of recipient atria [32,36]. Stretch-induced ANP secretion has been shown to be enhanced by humoral factors such as epinephrine, norepinephrine, arginine vasopressin (AVP) and endothelin (ET) [37–40]. Exercise-induced epinephrine increase is reduced in Htx mainly because of their lower peak exercise capacity [30,34]. Higher or similar norepinehrine levels, together with AVP hypersecretion were observed during exercise in Htx [29,30]. Together with the fact that basal plasma endothelin is generally increased [41,42], it suggests that these hormones might participate in an exaggerated ANP secretion in Htx, directly or through volume- and/or pressure overload [1]. To date, however, evidence of a stimulating effect of norepinephrine, AVP or ET on ANP release remain to be demonstrated during exercise after heart transplantation. As previously reported, antihypertensive treatment could decrease ANP secretion during exercise if associated with a decrease in atrial wall stress [28]. On the other hand, β-blockers resulting in attenuated heart rate increase induce ANP hypersecretion during exercise [43,44]. Cyclosporine might also enhance ANP secretion during exercise in Htx through both an increase in cardiac afterload and mass resulting in diastolic dysfunction and through its potential stimulatory effects on the sympathetic system and endothelin [45]. Similarly, the corticotherapy could result in saline retention, leading to volume overload. It is also known to directly stimulate the transcription of the gene encoding for ANP [46]. Thus, therapy will either enhance or reduce exercise-induced ANP secretion in Htx. Cardiac innervation is not necessary for ANP secretion, but the cardiac hormone release may be under neural control. Although controversial, an inhibitory role of cardiac innervation on ANP release has been proposed, and heart transplantation, by means of cardiac denervation, could correspond to a breakdown of this inhibition [47,48]. Cardiac denervation may also indirectly stimulate ANP release during exercise through both the need for elevated catecholamine concentrations for chronotropic incompetence limitation and elevated AVP resulting from loss of afferent information from cardiac mechanoreceptors [29,49]. However, atrial stretch being the most likely explanation for exercise-induced increase in ANP, cardiac denervation might act predominantly on ANP release by modulating atrial stretch. This effect may further be amplified in Htx, because of lessened pericardial restriction and impairment in ventricular compliance [28,50–52]. ANP hypersecretion in Htx may thus reflect greater atrial stretching [30,53]. In this view, the cardiac denervation-induced chronotropic limitation [6,54–57] may result in heart rate and venous return mismatch, favoring distention of the cardiac chambers through a compensatory increase in stroke volume. Such a mechanism occurs early during exercise and has been nicely demonstrated after heart transplantation [58]. Accordingly, an inverse relationship between ANP and heart rate increase has been recently reported during early exercise in Htx [34] (Fig. 1). Interestingly however, and consistently with the analysis of all data reported in the literature (Fig. 2), peak exercise ANP level was similar in Htx and controls, probably because of the lower power output reached after heart transplantation [34]. Further supporting this assumption, an attenuated increase in ANP was observed in Htx showing a nearly normal heart rate response to exercise, corresponding likely to partial cardiac reinnervation or exercise training [9,59]. Maximal ANP, BNP and ADM changes, from rest to peak exercise, in controls subjects (Ctrl) and heart transplant recipients (Htx). For ANP changes, data were obtained from Refs. [23–25,29,30,34]. For BNP changes, data were obtained from Ref. [34]. For ADM changes, data were obtained from Ref. [75]; means±S.E.M. Relative hormones and heart rate (HR) changes between rest and 70% of peak values obtained during a maximal exercise test in control subjects (white bars) and heart transplant recipients (black bars). Data were obtained from Refs. [30] and [34] for atrial natriuretic peptide (ANP), from Ref. [34] for brain natriuretic peptide (BNP) and unpublished data for adrenomedullin (ADM). Means±S.E.M. *, Difference between Ctrl and Htx groups, P<0.05. In healthy humans, all but one report showed a lack of significant BNP change in response to exercise [34,60–63]. This is consistent with the idea that BNP is released mainly from the ventricles in a constitutive manner and that short term exercise cannot stimulate synthesis and/or secretion of the cardiac hormone. However, BNP has also been shown to be acutely released during exercise in patients presenting with congestive heart failure, ischemia and/or hypertension [64–66], challenging such hypothesis. To the best of our knowledge, there are only two reports on the BNP response to exercise after heart transplantation [34,67]. Although similar and non-significant increases were observed during a graded 50-W, 10-min submaximal exercise [67], enhanced BNP increase has been observed in Htx as compared to normal subjects, in response to a graded 10-min maximal bicycle exercise performed until exhaustion [34]. Atrial stretch, ventricular hypertrophy and cardiac diastolic dysfunction have been proposed as potential stimuli for the BNP release during exercise. Both ANP and BNP have been colocated in atrial granules and might be cosecreted as shown during supraventricular tachyarrhythmias [68,69]. Since atrial stretch is the main stimulus for ANP release during exercise, it might be hypothesized that atrial stretch may participate in BNP release during exercise through ANP and BNP cosecretion. Accordingly, a positive correlation was observed during exercise between ANP and BNP after heart transplantation [34]. Such correlation was nevertheless weak and the specific enhancement of BNP release in Htx suggests that BNP might originate from other tissues such as the ventricles rather then from atrial granules [70]. Interestingly, exercise-induced increase in BNP is greater in hypertensive patients with left ventricular hypertrophy as compared to patients without cardiac hypertrophy [63,66]. This supports a role for increased cardiac mass in BNP release. Accordingly, BNP increment from rest to peak exercise was positively correlated with left ventricular mass index after heart transplantation [34]. Although not demonstrating a causal relationship, it further supports that BNP release is related to cardiac mass and that the enhanced BNP secretion observed in Htx might be due to their cardiac hypertrophy, directly and/or through cardiac diastolic dysfunction. Indeed, the relationship between left ventricular mass and BNP may be partly explained by diastolic dysfunction. Thus, plasma BNP level is increased in proportion with the degree of cardiac diastolic dysfunction [71]. Although cardiac diastolic dysfunction was not investigated, it might have participated in Htx's enhanced BNP response to exercise since both altered late-diastolic passive LV properties and blunted acceleration of LV relaxation during exercise contribute to the exaggerated exercise-induced elevation of LV end-diastolic pressure (LVEDP) after heart transplantation [34,50,54]. Accordingly, enhanced BNP is correlated to LVEDP at rest and throughout exercise in cardiovascular patients [61,64]. Such approach might be particularly interesting in Htx in view of the direct positive lusitropic effect of BNP [72,73], since it has been recently demonstrated that BNP infusion causes beneficial hemodynamic and neurohormonal effects during exercise in patients with isolated diastolic heart failure [72]. In healthy humans, in contrast to submaximal exercise, maximal exercise increased significantly circulating ADM [61,63,74]. Similarly, maximal exercise-induced increase in ADM was significant after heart transplantation but it tended to be lower in Htx as compared to normal controls [75]. Heart rate and exercise intensity or duration have been discussed in the literature as potential stimuli for ADM release during exercise. Exercise physiology after heart transplantation allows an unique opportunity to study the effect of heart rate on ADM release during exercise. Indeed, the comparison of ADM and heart rate changes in controls and Htx (heart rate increase being both delayed and blunted in Htx), suggested clearly that heart rate does not play a key role in ADM release during exercise [75]. From previous studies performed in both normal and cardiovascular subjects, it appeared that exercise duration is not critical for ADM secretion during exercise. Thus, submaximal exercises, consisting of two fixed work loads (40 and 80 W) for 4 min each, failed to increase ADM in controls and in patients presenting with hypertension or myocardial infarction [61,63]. Similarly, our preliminary data suggest that a 45-min endurance exercise test, mainly performed below 50% of peak VO2, does not significantly modify plasma ADM in either controls or Htx (unpublished). On the other hand, short duration (10 min) graded maximal exercise, performed until exhaustion, increased circulating ADM both in controls and in Htx [75]. Thus, exercise intensity, rather than duration, predominantly stimulates ADM secretion, the main factor modulating such ADM response still needs to be determined. Endurance and resistance exercise training are well tolerated in Htx and are now considered an essential adjunct therapy after heart transplantation [6–8,59,76,77]. Indeed, long term endurance training allows Htx to achieve peak heart rate and VO2 values that approach age-matched norms [59]. Similarly, resistance exercise training counteracts corticosteroid-induced osteoporosis and skeletal muscle myopathy [77]. Although neuroendocrine activation is well known to be associated with poor long-term prognosis in heart failure, there are relatively few data concerning cardiac natriuretic peptides and training in cardiovascular diseased patients [78–80]. Interestingly however, it has been recently shown, in the first randomized controlled study, that a 16-week endurance training significantly reduces rest vasoconstrictive (angiotensin, aldosterone, vasopressin) and ANP plasma levels in heart failure patients [80]. Controversially, after heart transplantation, no significant changes were observed in circulating ANP and vasopressin before and after a 6 week modified interval training program [81]. Nevertheless, the fact that plasma renin activity and aldosterone tended to decrease after similar training when considering both exercise and recuperation values, suggests that longer-term exercise training might also be beneficial by reducing neurohormonal activation after heart transplantation [82]. Indeed, although resulting in a significant increase in maximal tolerated power, it might be hypothesized that the 6-week training program performed by the patients was too short to greatly reduce their neurohormonal activation [82]. In summary, despite an early ANP hypersecretion, maximal exercise-induced increase in the cardiac hormone appears similar in Htx and controls. The lower peak power reached after heart transplantation might therefore counterbalance the early greater atrial stretching secondary to heart rate and venous return mismatch, resulting in a smaller ANP increase during late exercise. A key factor modulating the ADM release during exercise is unknown, but BNP appears to be hypersecreted during exercise in Htx, likely because of ventricular hypertrophy associated with diastolic dysfunction. The physiological role of the cardiac natriuretic peptides during exercise remains to be investigated. However, in view of the beneficial effect of BNP infusion during exercise in patients with diastolic heart failure [72], one may suggest that BNP might favor ventricular relaxation in cardiac hypertrophic Htx. Furthermore, BNP acting in synergy with ANP and probably with ADM, could exert modulatory influences on the vasoconstrictive and fluid retention systems thus allowing a better oxygen supply to the working muscles. Similarly, although the percentage fall in renal blood flow at peak exercise appeared greater in Htx than in controls [83], our preliminary data showed similar natriuresis after maximal exercise in normal subjects and after heart transplantation. This is further supported by previous data showing that exogenous and endogenous ANP significantly participate in sodium and water renal excretion after heart transplantation [84–87]. Finally, long-term exercise training are warranted after heart transplantation, but not only because of their well known beneficial effects on physical status and lowering effect on the therapy-induced complication rate. Indeed, such a training program may reduce the neurohormonal activation, thus improving the long-term survival of heart-transplant recipients. Further studies will be useful to investigate this important issue.