Background
The pathogenesis of various forms of adrenal and low-renin hypertension is attributed primarily to the renal effects of excessive aldosterone on sodium retention and potassium elimination [
1]. Experimental data have shown that aldosterone can cause cardiac, renal and vascular damage independent of its effects on blood pressure [
2]. These experimental data together with growing clinical evidence for the benefits of blocking the actions of aldosterone with mineralocorticoid receptor antagonists (MRAs) in the treatment of heart failure [
3]-[
5] provide the rationale for new therapeutic approaches to inhibit aldosterone production. While MRAs are currently the treatment of choice for pathologic conditions due to aldosterone excess [
6], they induce a counter-regulatory increase in aldosterone production [
7]. This may attenuate the effects of competitive MRAs, and might also contribute to adverse cardiovascular effects through non-genomic mechanisms independent of activation of the mineralocorticoid receptor [
8]-[
10]. An alternative approach to inhibiting the effects of aldosterone is blocking the renin-angiotensin system (RAS), which decreases aldosterone production indirectly. However, the resulting fall in plasma aldosterone does not persist in the long term [
11],[
12]. The therapeutic potential of targeting adrenal steroid synthesis to inhibit aldosterone production has not been fully explored due to the lack of selective aldosterone synthase inhibitors (ASIs).
Aldosterone synthase, which is encoded by the CYP11B2 gene, is highly expressed in adrenal gland glomerulosa cells and is expressed at lower levels in other tissues [
13],[
14]. Aldosterone synthase converts 11-deoxycorticosterone (11-DOC) to aldosterone by sequential 11-β hydroxylation, 18-hydroxylation and 18-oxidation [
15],[
16]. Selective inhibition of aldosterone synthase could provide an effective approach to decrease aldosterone production, thus attenuating both receptor-mediated and non-genomic deleterious consequences of aldosterone excess.
We report the
in vitro and
in vivo effects of aldosterone synthase inhibition with LCI699 (4-[(5R)-dihydro-5H-pyrrolo[1,2-c]imidazol-5-yl]-3-fluorobenzonitrile phosphate) [
17] in rats, non-human primates and humans. We have characterized the enzymatic inhibition and species specificity of LCI699 and have established the relative selectivity of LCI699 for aldosterone synthase over 11β-hydroxylase (encoded by the CYP11B1 gene), which converts 11-deoxycortisol to cortisol and has 93% nucleotide sequence identity with aldosterone synthase [
18]. In order to determine the therapeutic potential of an ASI, the effects of LCI699 on cardiorenal damage and survival were assessed in a double-transgenic (dTG) rat model with ectopic overexpression of human renin and angiotensinogen and the results compared with those of the MRA eplerenone. In healthy human subjects, LCI699 selectively inhibited aldosterone synthase at oral doses ≤ 1 mg daily, but lost specificity above the 1 mg dose. Therefore LCI699 is no longer being developed for essential hypertension, and is currently under development at higher, nonselective doses for the treatment of Cushing’s syndrome [
19].
Discussion
This is the first report of the effects of pharmacologic inhibition of aldosterone synthase in healthy human subjects. Results obtained with the ASI LCI699 indicate that the hormonal and renal effects of blocking the aldosterone pathway in healthy animals translate to humans. In healthy volunteers, once-daily oral dosing with LCI699 0.5 mg selectively reduced plasma and urinary aldosterone, which was associated with natriuresis and an increase in PRA. LCI699 prolonged survival in a rat disease model induced by ectopic overexpression of human renin and angiotensinogen, and was more effective than the MRA eplerenone in preventing cardiac and renal damage. These results support the therapeutic potential of inhibiting aldosterone synthase in diseases characterized by excessive aldosterone production.
Characterization of LCI699 was performed using
in vitro assays and
in vivo models in the rat and monkey. LCI699 showed distinct differences between species; it was at least 200-fold less potent in inhibiting rat recombinant adrenal enzymes (and even less potent using rat adrenal tissue homogenates) and ~20-fold less potent in inhibiting monkey adrenal homogenate enzymes compared with human recombinant enzymes. The
in vivo Ang-II- and ACTH-stimulation models used to characterize the pharmacodynamic effects of LCI699 in rats and monkeys predicted the human oral dose potency and the short duration of action of LCI699 for inhibition of aldosterone synthase. Thus, LCI699 dose-dependently inhibited ACTH-stimulated increases in aldosterone with an EC
50 in the low nanomolar range in monkeys, consistent with the low
in vitro IC
50 (0.7 nmol/L) obtained with recombinant human aldosterone synthase. Both the rat and monkey
in vitro data showed relative selectivity of LCI699 for aldosterone synthase over 11β-hydroxylase (3–5-fold) similar to that observed with the human enzymes. However, both the rat and monkey
in vivo models overestimated the selectivity of LCI699 in humans. In the rat,
in vivo ED
50 values showed 47-fold selectivity of LCI699 for aldosterone synthase over 11β-hydroxylase, while in the monkey there was no attenuation of the ACTH-stimulated cortisol response by LCI699 even at doses above the ED
50, which impaired cortisol synthesis in human subjects. Cai and co-workers evaluated the effects of LCI699 (administered intravenously at doses in the range 0.003–3 mg/kg) on ACTH-stimulated aldosterone and cortisol in rhesus monkeys on a low-salt diet [
25]. They showed no effect of LCI699 on ACTH-stimulated cortisol at doses of 0.003–1 mg/kg, similar to our findings in cynomologus monkeys with LCI699 (administered orally at doses of 0.005–0.15 mg/kg). In rhesus monkeys, only the highest, 3 mg/kg i.v. dose of LCI699 reduced ACTH-stimulated cortisol. Thus, the results of Cai and co-workers [
25] are consistent with our finding of a relatively high
in vivo selectivity of LCI699 for inhibition of synthesis of aldosterone over cortisol in monkeys, which was not observed in humans. Overall, these results demonstrate the utility and limitations of using different species for
in vitro and
in vivo characterization of ASIs and emphasize the importance of early human investigations for selection of ASI drug candidates based on potency and selectivity [
26].
In the dTG rat model, inhibition of aldosterone synthase with LCI699, which dose-dependently blocked increases in aldosterone, prevented development of cardiac and renal functional abnormalities (indicated by changes in LV fractional shortening, IVRT/RR ratio, LV weight, BUN, urinary albumin excretion), and reversed hormonal and electrolyte abnormalities associated with elevated levels of aldosterone. It is noteworthy that the organ protective effects of aldosterone synthase inhibition with LCI699 occurred in the absence of any major reduction in blood pressure. The correlations between plasma aldosterone concentration and observed cardiac and renal structural and functional abnormalities in the present study (Additional file
4) provide additional evidence in support of direct deleterious effects of aldosterone on the heart and kidneys. In addition, dTG rats were polydipsic and polyuric (4-fold higher urine volume flow compared with S-D rats). While Ang-II-induced hypertension with resulting pressure natriuresis has been suggested as the cause for the increase in urine volume [
27], our observation that LCI699 normalized urine volume in dTG rats with little effect on blood pressure indicates that other mechanisms may play a more important role such as impaired renal concentrating ability associated with hyperaldosteronism and hypokalemia [
27]. The finding that aldosterone synthase inhibition exerts organ protective effects independent of blood pressure lowering also suggests that blood pressure should not be used as the sole surrogate marker of cardiorenal risk reduction in early clinical testing of ASIs.
LCI699 prolonged survival in both young dTG rats and older animals with established hypertension and cardiorenal disease. The MRA eplerenone at a dose of 30 mg/kg body weight also improved survival in a young dTG rat model, but increased plasma and urinary aldosterone and had little effect on several key markers of cardiorenal damage that were normalized by LCI699. Our results confirm and extend those of Fiebeler
et al., who showed that the aldosterone synthase (CYP11B2) inhibitor FAD286 or adrenalectomy attenuated cardiorenal damage and reduced mortality in young dTG rats [
28]. Collectively, these data suggest that aldosterone, in addition to Ang II and hypertension, is an important mediator of cardiac and renal damage in the dTG rat model. Synergistic genomic and non-genomic interactions between aldosterone and Ang II have been described in both
in vitro and
in vivo studies, and have been implicated in vascular inflammation, fibrosis and remodeling [
29],[
30]. Although corticosterone, the physiologic glucocorticoid in rodents, and cortisol in humans have the same affinity for the mineralocorticoid receptor as aldosterone and are present in higher concentrations than aldosterone [
31], plasma corticosterone levels in dTG rats were not higher than those observed in age-matched S-D rats and so it is unlikely that corticosterone contributed significantly to the cardiorenal disease observed in the dTG rat model. It is notable that enalapril pre-treatment extended median survival by approximately 4 months in the dTG rat. This may be due to the cardiorenal protective effects of reducing Ang II levels by inhibiting ACE that may influence survival beyond what can be achieved by inhibition of aldosterone synthesis. Regardless of possible synergy between Ang II and aldosterone, our results indicate that inhibiting aldosterone synthase alone is sufficient to prevent development and slow progression of cardiac and renal disease in the dTG rat model.
LCI699 was synthesized as the first orally active aldosterone synthase inhibitor for human use based on the chemical structure of FAD286A, the dextroenantiomer of the aromatase inhibitor fadrozole. Studies in recombinant hamster fibroblasts expressing human CYP11B1 and CYP11B2 showed that LCI699 was a more potent inhibitor of aldosterone synthesis than FAD286A (IC
50 0.2 vs 0.8 nmol/L) and more selective for inhibition of aldosterone vs cortisol synthesis (selectivity factor 15-fold vs 7.9-fold) [
32]. However, selectivity of LCI699 for aldosterone synthase over 11β-hydroxylase was comparable to that for FAD286 based on recombinant human CYP11B1 and CYP11B2 enzymatic assays [
20]. LCI699 and FAD286A also exhibited broadly similar selectivity for inhibition of aldosterone synthase over 11β-hydroxylase in the
in vivo rat ACTH stimulation model (selectivity factor by exposure 47-fold for LCI699 in the present study, vs 48-fold for FAD286A in a previous study) [
21]. Preclinical studies in dogs showed that LCI699 exhibited greater oral bioavailability than FAD286A (96% vs 33% for a 1 mg/kg oral dose), a longer half-life (0.6–3.6 h vs 0.9–1.2 h), and more potent inhibition of Ang II-stimulated aldosterone release (EC
50 41 nmol/L vs 136 nmol/L) (D. Rigel, personal communication). FAD286A was tested in a similar dTG rat model to that used in the present study [
28]. The effects of FAD286A (4 mg/kg in the diet from Weeks 4–7) on albuminuria and cardiac hypertrophy appeared less pronounced than those that we observed with LCI699; however, subsequent studies using higher doses of FAD286A found comparable pharmacologic benefits of LCI699 and FAD286A (D. Rigel, personal communication).
It is important to consider the potential differences in clinical benefit between inhibiting the synthesis of aldosterone with an ASI and inhibiting the mineralocorticoid receptor with an MRA. These have been reviewed in detail, with specific reference to LCI699, by Azizi and colleagues [
33]. One issue of particular clinical relevance is hyperkalemia, which is a well-known adverse effect of MRAs and other inhibitors of RAS activity. It is, however, uncertain if the different mechanism of action of ASIs compared with MRAs will alter the risk of hyperkalemia. Some insights may be gained by comparing the phenotypes of mineralocorticoid receptor- and aldosterone synthase-deficient mice generated by gene targeting. Mineralocorticoid receptor-deficient mice show a partial defect of the amiloride-sensitive Na
+ channel (ENaC) in the colon and kidney [
34],[
35]. The phenotype is characterized by strongly enhanced fractional excretion of sodium resulting in hyponatremia and severe hyperkalemia with reduced fractional excretion of potassium compared with wild type mice, reflecting a transport defect in the principal cells of the collecting duct. In contrast, mice with genetic deficiency of aldosterone synthase demonstrate modest hyperkalemia with normal urinary potassium excretion [
36]. Interestingly, an increase in potassium excretion on Day 7 was observed in our study of healthy subjects following administration of LCI699, but not eplerenone. One possible explanation for this difference is that 11-DOC (increased by LCI699 but not eplerenone) which been shown to stimulate potassium excretion in the renal collecting tubules, may have a relative potassium-sparing effect without promoting sodium retention [
37],[
38]. The clinical relevance of this remains uncertain, however, as a study conducted in patients with mild-to-moderate hypertension showed a significant dose-response relationship for LCI699 (total daily dose range 0.25 mg to 1 mg) and serum potassium levels; moreover, changes in serum potassium concentration with LCI699 were similar in magnitude to those observed with the MRA eplerenone (50 mg twice daily) [
39].
The clinical benefits of MRAs have been demonstrated previously [
3]-[
5]. However, the efficacy of MRAs may be limited by the fact that the renal loss of sodium and retention of potassium induced by MRAs stimulate a dose-dependent increase in renin, Ang II and aldosterone [
7]. This may counteract the beneficial actions of MRA treatment both by competition of increased aldosterone at the level of the mineralocorticoid receptor and by stimulating non-genomic (mineralocorticoid receptor-independent) effects of aldosterone [
40]. On the other hand, cardiomyocyte mineralocorticoid receptors can be occupied by physiologic glucocorticoids, and under conditions of tissue damage cortisol can become a mineralocorticoid receptor agonist [
41], mimicking the experimental effects of aldosterone through a mechanism that would not be affected by an ASI. Our results that cardiomyocyte cell size was normalized by LCI699 in the dTG rat suggest that such a mechanism plays little, if any, role in this experimental model.
The majority of untreated dTG rats in our studies experienced unexpected and sudden death, whereas the majority of LCI699-treated animals eventually became moribund, necessitating euthanasia. Although the mode of death is uncertain, it is possible that inhibiting aldosterone synthase may reduce mortality in part by preventing sustained ventricular arrhythmias and subsequent hemodynamic deterioration [
42]. The clinical relevance of such findings is suggested by the Randomized ALdactone Evaluation Study (RALES), which demonstrated that in patients with congestive heart failure, the addition of spironolactone to standard medical therapy significantly reduced both cardiac mortality and sudden cardiac death [
3].
This study provides the first comprehensive characterization of the hormonal effects of inhibiting aldosterone synthase in humans. In the clinical study, moderate sodium restriction was used to activate the renin-angiotensin pathway and stimulate baseline aldosterone levels, and thereby facilitate detection of dose-dependent decreases in aldosterone and counter-regulatory increases in renin. Sodium and potassium intake were strictly controlled while subjects were housed in the clinic and verified by measuring urinary sodium and potassium excretion. The potassium content of the diet was increased by 20 mEq/day after 7 days to evaluate the persistence of LCI699 effects in response to further stimulation of aldosterone release. Single doses of LCI699 caused dose-dependent reductions in plasma and urinary aldosterone levels; the maximal inhibition of plasma aldosterone reached a plateau of approximately 80% at the 10 mg single dose, with no further reduction at higher doses up to 200 mg. These results were consistent with those from the Ang-II- and ACTH-stimulation models in rats and monkeys, in which the upper dose of LCI699 also did not completely inhibit aldosterone production (maximal inhibition of 80–90%). Incomplete pharmacologic inhibition of aldosterone synthase by LCI699 at the doses used is the most likely explanation for this observation, because in monkeys a 3000 μg/kg dose of LCI699 decreased plasma aldosterone concentrations to below the detection limit for 24 h (data not shown). Indeed, the steroid profile of patients with genetic aldosterone synthase deficiency is characterized by low to undetectable plasma aldosterone levels and elevated levels of mineralocorticoid precursors (corticosterone or 11-DOC) [
16].
LCI699 treatment induced a natriuresis on Day 1 that was similar with all doses and equal to that achieved by eplerenone 100 mg (the approved maximum once-daily dose for use in patients with essential hypertension). The rise in PRA observed after repeated administration of LCI699 reflects the magnitude of RAS counter-regulation following aldosterone synthase inhibition and consequent aldosterone suppression and sodium depletion. LCI699 0.5 mg daily resulted in similar increases in PRA to eplerenone 100 mg, while LCI699 3 mg stimulated markedly larger increases in PRA. The larger dose-dependent counter-regulatory activation of the RAS by LCI699 relative to eplerenone presumably reflects greater suppression of the aldosterone pathway by LCI699 at the higher doses. The consequent increase in levels of Ang II may limit effects of aldosterone synthase inhibition, but this was not tested in the present study.
Repeated once-daily dosing with LCI699 was associated with a resetting of plasma aldosterone levels during the second half of the dosing interval (12–24 h post-dose) followed by a more gradual, chronic increase in baseline aldosterone levels. This pattern probably reflects a combination of the short half-life of LCI699 (approximately 4 h in humans), counter-regulatory stimulation of the RAS, a gradual increase in adrenal aldosterone synthase enzyme, and slow clearance of precursors such as 11-DOC. In the first 12 h after dosing, LCI699 effectively inhibited the activity of aldosterone synthase and plasma aldosterone was decreased. However, during the next 12 h, when LCI699 plasma concentrations had decreased, aldosterone synthase was presumably no longer inhibited. Moreover, the adrenal gland was exposed to elevated levels of renin and Ang II, which stimulate aldosterone release [
43]. Thus, with once-daily administration of LCI699 in the morning, the pharmacologic effect of the drug to inhibit aldosterone synthase will essentially have dissipated during the nocturnal rise in ACTH, cortisol and aldosterone. This was the pharmacokinetic and dynamic basis for twice-daily dosing of LCI699 for subsequent studies in patients.
The selectivity of LCI699 for aldosterone synthase over 11β-hydroxylase was evaluated by an ACTH stimulation test in experimental models and in humans. The ACTH test provided safety monitoring for clinically relevant inhibition of glucocorticoid synthesis, which occurred with repeated administration of LCI699 at the 10 mg dose [
44]. Measurements of plasma levels of precursors of aldosterone synthesis (11-DOC) and cortisol synthesis (11-deoxycortisol) also provided useful markers of the relative selectivity of aldosterone synthase inhibition. Both precursors are synthesized in the adrenal gland and their production is dependent on normal ACTH levels [
45],[
46]. LCI699 0.5 mg was shown to be selective for inhibition of aldosterone synthase over 11β-hydroxylase throughout the dosing period, as evidenced by reductions in plasma and urinary aldosterone and increases in the aldosterone precursor 11-DOC without effect on basal or ACTH-stimulated cortisol or on the cortisol precursor 11-deoxycortisol. LCI699 1 mg was selective for aldosterone synthase after a single dose on Day 1 and on Day 6 of multiple dose administration, but showed a time-dependent loss of selectivity with continued dosing. Although there was no effect on basal cortisol or 11-deoxycortisol, LCI699 1 mg did result in attenuation of the ACTH-stimulated cortisol response on Day 13. This was confirmed in approximately 20% of hypertensive patients treated with LCI699 1 mg once daily [
47]. This lag in pharmacodynamic effect in the absence of a change in plasma pharmacokinetics, may reflect a longer half-life of LCI699 within the adrenals. The 3 mg dose of LCI699 inhibited 11β-hydroxylase on Day 6 (as demonstrated by both an attenuated cortisol response to ACTH and increase in 11-deoxycortisol) and also resulted in a small increase in plasma potassium which, in conjunction with the clinical observations, suggested development of mild hypoaldosteronism.
Several limitations of the human study should be acknowledged. First, although dietary sodium and potassium intake were controlled, there was no assessment of total metabolic sodium and potassium balance during repeated dose administration of LCI699. Second, urinary collections were not fractionated. This likely gave rise to an artifactual apparent attenuation in the effect of LCI699 to inhibit aldosterone synthase when 24-h urinary aldosterone was compared across Days 1, 7 and 14 of administration because of the short half-life of LCI699 and consequent rebound increase in aldosterone 12–24 h after dosing.
There are two potential therapeutic uses of LCI699, as noted by Azizi and colleagues in their review of the therapeutic potential of aldosterone synthase inhibition [
33]. First, LCI699 at low doses (below 1 mg) can be used as a selective inhibitor of aldosterone production to treat subsets of patients with hypertension and cardiorenal disease associated with elevated aldosterone levels. In a small, short-term study in 14 patients with primary aldosteronism, LCI699 (0.5 mg and 1 mg) administered twice daily effectively reduced plasma and urinary aldosterone levels and corrected hypokalemia after 4 weeks of treatment. In these low-renin hypertensive patients, reductions in blood pressure were very modest (mean 4 mmHg decrease in ambulatory systolic blood pressure) [
48]. In another study of 524 patients with essential hypertension, LCI699 administered at doses between 0.25 and 1 mg resulted in placebo-corrected reductions in ambulatory systolic and diastolic blood pressures of 6–9 mmHg and 3–6 mmHg, respectively after 8 weeks of treatment [
39]; ACTH-stimulated cortisol release was suppressed in approximately 20% of subjects receiving LCI699 at a total daily dose of 1 mg in that study. Second, the potency of LCI699 in inhibiting 11β-hydroxylase at daily doses above 3 mg led to further development of LCI699 as a drug candidate for indications in which metyrapone is currently used, both for diagnosis and treatment. Indeed, LCI699 has low pharmacokinetic variability, and may avoid the limitations of metyrapone and other steroidogenesis inhibitors such as ketoconazole and mitotane [
49]. In a study in 12 patients with Cushing’s disease, LCI699 2–50 mg twice daily was well tolerated and highly effective in reducing urinary free cortisol (below upper limit of normal in 11 patients) and in lowering blood pressure after 70 days of treatment [
19].
Competing interests
JM has received fees as a consultant to Novartis, Actelion, Roche and NicOx, and has received no stock or stock options. DFR, CW, FF, MB, JL, WC, DL, GL, SR, YZ and WPD are employees of Novartis and are therefore eligible for Novartis stock and stock options. AYJ was formerly an employee of Novartis.
Authors’ contributions
All authors made substantial contributions to the conception or design of the work and/or the acquisition, analysis, or interpretation of data for the work. JM and WPD were responsible for the design of the overall study program, preclinical and clinical data analysis and interpretation. DFR was responsible for the design and oversight of preclinical rat and monkey in vivo studies, and data analysis and interpretation. AYJ was responsible for the design and oversight of in vitro assays and preclinical rat and monkey ex vivo sample analyses, and data analysis and interpretation. FF, MB, JL and WC were responsible for the conduct and data analysis of preclinical rat and monkey in vivo studies. CH and JL were responsible for analysis of preclinical rat and monkey ex vivo samples and data. DL was responsible for conduct of in vitro assays and data analysis. GL was responsible for the analysis of preclinical rat and monkey pharmacokinetic samples and data. SR was responsible for the analysis of clinical pharmacokinetic data and its interpretation. YZ was responsible for the analysis of clinical study data. CW was responsible for the design and oversight of the clinical study conduct and summary and interpretation of its data. All authors were involved with drafting the manuscript or revising it critically for important intellectual content, and approved the final version of the manuscript. In addition, all authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.