Ketotifen and cardiovascular effects of xamoterol following single and chronic dosing in healthy volunteers

Abstract

Aims

To study whether desensitization occurs after long-term administration of the β₁-adrenoceptor partial agonist xamoterol and, if so, whether this can be influenced by ketotifen.

Methods

In a double-blind, randomized design 10 young, healthy males received ketotifen (2×1 mg day⁻¹ p.o.) or placebo for 3 weeks with xamoterol (2×200 mg day⁻¹ p.o.) administered concomitantly during the last 2 weeks. β₁-adrenoceptor mediated responses were assessed as exercise-induced tachycardia and isoprenaline-induced shortening of heart rate corrected electromechanical systole (QS₂c); isoprenaline-induced tachycardia was measured as a mixed β₁-/β₂-adrenoceptor-mediated effect.

Results

The first dose of xamoterol significantly increased resting heart rate and systolic blood pressure and significantly shortened QS₂c. The last dose of xamoterol after 2 weeks of treatment still produced the same responses. Ketotifen did not influence these effects of xamoterol on resting haemodynamics. The first dose of xamoterol caused a rightward shift of the exercise- and isoprenaline-induced tachycardia (mean dose ratios±s.e.mean: 1.20±0.05 and 2.46±0.23) and the isoprenaline-evoked shortening of QS₂c (dose ratio 3.59±0.68). This rightward shift was even more pronounced after 2 weeks xamoterol treatment. This additional rightward shift after 2 weeks of xamoterol was not affected by ketotifen (mean difference (95% CI) of log transformed dose ratios between placebo and ketotifen: exercise tachycardia 0.001 (−0.03; 0.04); isoprenaline tachycardia 0.03 (−0.15; 0.21); isoprenaline induced shortening of QS₂c 0.13 (−0.22; 0.48)).

Conclusions

In humans xamoterol is a partial β₁-adrenoceptor agonist with positive chrono- and inotropic effects at rest and antagonistic properties under conditions of β-adrenoceptor stimulation. These effects were well maintained after chronic dosing with no signs of β₁-adrenoceptor desensitization. Ketotifen does not change the β-adrenoceptor mediated responses of xamoterol after chronic dosing.

Introduction

It is now generally accepted that following long-term exposure to β-adrenoceptor agonists β-adrenoceptors are desensitized. This holds true for exogenously applied agonists (e.g. in the treatment of asthma) but also for chronically increased endogenous catecholamines (e.g. in phaeochromocytoma or chronic heart failure). Such a desensitization may limit the therapeutic efficacy of β-adrenoceptor agonists in the therapy of asthma [1] and in chronic heart failure may contribute to progression of the disease [2].

It would, therefore, be desirable to find drugs that are capable of preventing agonist-induced desensitization. The antiallergic agent ketotifen might be such a drug. We have previously shown that, in humans, ketotifen can prevent (or at least attenuate) agonist-induced desensitization of β₂-adrenoceptor mediated cardiovascular effects (isoprenaline-infusion induced decrease in diastolic blood pressure or increase in plasma catecholamines) [3]; moreover, very recently we found, that ketotifen can also attenuate agonist-induced desensitization of β₂-adrenoceptor-mediated positive ino- and chronotropic effects [4]. However, it is not known whether ketotifen can also blunt agonist-induced desensitization of β₁-adrenoceptors, for example, in chronic heart failure [2]. If this were the case it could represent an alternative way to maintain β₁-adrenoceptor function in heart failure. Unfortunately the only β₁-adrenoceptor selective agonist currently available for oral use in humans is xamoterol which is a partial β₁-adrenoceptor agonist [5]. We have previously shown, in healthy volunteers, that following 2 weeks treatment with xamoterol cardiovascular β₁-adrenoceptor-mediated effects (exercise-induced increase in heart rate and isoprenaline induced increase in systolic blood pressure) are attenuated [6]. This could be due to desensitization; it could also be due to an antagonist effect of the partial agonist xamoterol. The aim of the present study was, therefore, to find out whether, in healthy volunteers, 2 weeks treatment with xamoterol might desensitize β₁-adrenoceptor mediated cardiovascular effects and if so, whether ketotifen might attenuate these effects.

Methods

Subjects

Ten young, healthy, male volunteers (mean age±s.e.mean: 28.7 1.3 years, mean weight±s.e.mean: 84.5±3.4 kg) who on the basis of medical history, physical examination, ECG and routine laboratory screening were judged to be healthy participated in this study. The study protocol had been approved by the Ethics Committee of the University Clinics of Essen and was conducted in agreement with the Declaration of Helsinki. The subjects confirmed their written consent after detailed information had been provided.

Study design and general study procedures

The study was performed in a double-blind, randomized, placebo-controlled, cross-over design. We compared the effects of either placebo or ketotifen on β-adrenoceptor mediated cardiovascular functions before and after 2 weeks treatment with the partial β₁-adrenoceptor selective agonist xamoterol. At day 1 and 2, i.e. prior to the commencement of any drug therapy, subjects underwent a baseline exercise and isoprenaline infusion test. Thereafter they were treated with ketotifen 2×1 mg day⁻¹ orally at 07.00 h and 19.00 h or matching placebo respectively for 3 weeks, i.e. until the end of the study period. At days 8 and 9, i.e. after 1 week therapy with ketotifen or placebo, respectively, the exercise and isoprenaline infusion test were repeated 90 min after the first application of 200 mg xamoterol orally. From day 9 to day 23 subjects received xamoterol, 2×200 mg day⁻¹ orally, at 07.00 h and 19.00 h. At day 22 and 23, with the volunteers still on treatment, the exercise and isoprenaline infusion test were again repeated 90 min after administration of the morning dose of xamoterol. During the test days the morning dose of xamoterol was always taken in the study room. The order in which the exercise and isoprenaline infusion test were performed (exercise on day 1, 8, 22; isoprenaline infusion on day 2, 9, 23 or vice versa) was randomly allocated between subjects but kept constant for each subject during the entire study. All tests were performed after an overnight fast between 08.00–12.00 h with the subjects in the supine position in a quiet room at a constant ambient temperature after subjects had been resting for at least 30 min. There was a washout period of 5 weeks duration between the two treatment periods (placebo, ketotifen).

Cardiovascular measurements

Systolic and diastolic (phase V) blood pressure (SBP, DBP) were measured with a standard mercury sphygmomanometer. Baseline recordings of blood pressure (5 measurements) and systolic time intervals (STI) (five cardiac cycles) were performed after 30 min of complete supine rest prior to administration of xamoterol and again 90 min after administration of xamoterol. The 90 min recordings served as baseline for the subsequent isoprenaline infusion or exercise test.

STI were obtained noninvasively, according to standard techniques [7] and as previously described by our group [8, 9]. The following parameters were measured: RR interval (ms) of the ECG from which heart rate (beats min⁻¹) was calculated; duration of the electromechanical systole (QS₂) (ms); duration of left ventricular ejection time (LVET) (ms). The duration of the pre-ejection period (PEP) (ms) was derived mathematically by subtracting LVET from QS₂. QS₂ was corrected for changes in heart rate as previously described [8, 10] and is herein referred to as QS₂c.

Dynamic bicycle exercise was performed in the supine position on a Bosch ergometer (Bosch, Berlin, Germany). Subjects started with an initial workload of 50 watt which was increased by 25 watt every 3 min until a maximum workload of 150 watt was attained.

The non-selective β-adrenoceptor agonist isoprenaline (Aleudrina®, Boehringer Ingelheim) was administered as a continuous intravenous (i.v.) infusion at five incremental doses of 1.75, 3.5, 7.0, 17.5, and 35 ng kg⁻¹ min⁻¹ by means of a perfusion pump (Braun, Melsungen, Germany). Each dose was infused for 10 min. Blood pressure was measured at 1 min intervals during the last 5 min of each dose step and STI were recorded following the last blood pressure measurement. The mean of these five blood pressure measurements and of five cardiac cycles (STI) respectively was chosen for analysis of the dose-response curve.

Blood for determination of plasma xamoterol levels was drawn from an antecubital vein 90 min after intake of the first and the last dose of xamoterol on day 8 and day 22 respectively. Xamoterol plasma levels were analysed by high performance liquid chromatography with fluorometric detection as previously described [11] using a LiChrospher column (Knauer, Berlin, Germany).

Study drugs

Both ketotifen capsules containing 1.38 mg ketotifen fumarate equivalent to 1 mg ketotifen and matching placebos were produced by our hospital pharmacy. Xamoterol (Corwin®, Tablets containing 200 mg) were obtained from Stuart Pharmaceuticals Ltd, Wilmslow, Cheshire, England and isoprenaline (Aleudrina®) from Boehringer Ingelheim.

Statistical evaluation

The effects of xamoterol on resting haemodynamics were investigated by comparing the measurements performed 90 min after administration of xamoterol on day 8 and day 22 with corresponding measurements during the control day 1 by two-way-ANOVA analysing the effects of treatment (ketotifen, placebo), time (study day 1, 8, 22) and treatment-time-interaction. When ANOVA indicated a significant effect post-hoc paired t-tests with Bonferroni multiple comparisons procedure were used to assess where the significant differences lay.

The influence of xamoterol on the haemodynamic effects induced by i.v. isoprenaline and bicycle exercise were examined by construction of cumulative dose-response curves for the changes in cardiovascular responses from baseline for each isoprenaline infusion and bicycle exercise test. For each individual dose-response curve the doses of isoprenaline required to increase heart rate by 15 beats min⁻¹, to reduce diastolic blood pressure by 10 mmHg and to shorten QS₂c by 30 ms were calculated from a fitted linear (increase in heart rate during i.v. isoprenaline and during exercise) and log linear (fall of diastolic blood pressure and shortening of QS₂c during isoprenaline) regression equation. These doses are referred to as effective doses (ED; e.g. ED₁₅ for heart rate, ED₃₀ for QS₂c etc.) and were expressed in ng kg⁻¹ min⁻¹ throughout. Accordingly for the bicycle exercise test the workload (Watt) required to increase heart rate by 50 beats min⁻¹ (ED₅₀ for heart rate) was calculated.

The effect of the first dose of xamoterol administered on day 8 on the haemodynamic changes induced by i.v. isoprenaline and exercise was described by comparing the effective doses on day 8 with the effective doses on day 1, i.e. on the control day when no xamoterol had been given by calculation of an effective dose ratio (ED-ratio). Similarly the effect of chronic administration of xamoterol on cardiovascular responsiveness was described by calculating the ED-ratios of the effective dose of day 22, i.e. after 2 weeks administration of xamoterol, over the effective dose of day 8, i.e. after the first dose of xamoterol.

In order to test whether ketotifen was able to modulate the influence of xamoterol on cardiovascular function the ED-ratios during placebo were compared with the ED-ratios during ketotifen treatment by paired t-test. For statistical analysis ED-ratios were generally log-transformed (log₁₀). All quoted P-values for the comparison of ED-ratios between placebo and ketotifen as well as the corresponding 95% confidence intervals refer to log-transformed ED-ratios.

In some isoprenaline dose-response curves it was not always possible to define phase V diastolic blood pressure at the two highest doses of isoprenaline when Korotkoff sounds did not disappear or the fall in diastolic blood pressure reached a plateau distorting the log linear dose response relationship which was evident for the lower dose levels. When this occured these doses were not included into log linear regression analysis for calculation of ED values. For one subject (number 4) during day 8 of the placebo treatment phase, it was not possible to calculate a meaningful dose response curve for diastolic blood pressure. Completely atypically, this subject showed an increase in diastolic blood pressure from baseline during the first two doses to a maximum of 8.4 mmHg at 3.5 ng kg⁻¹ min⁻¹ before a fall occurred. Log linear regression resulted in a fit with a slope not significantly different from zero (P =0.206). Subject number 6 injured his knee during the placebo treatment phase prior to day 22. Therefore he could not perform the exercise test on day 22. For these reasons these subjects were excluded from statistical analysis, i.e. n =9 for isoprenaline induced change in diastolic blood pressure and for exercise evoked tachycardia.

The reproducibility of the isoprenaline and exercise induced changes in cardiovascular parameters was assessed by calculating ED-ratios for the effective doses (ED) on day 1 of treatment phase 2 over the effective dose (ED) on day 1 of treatment phase 1. If there is no systematic change in cardiovascular responsiveness to isoprenaline or exercise with time this ED-ratio should not significantly differ from 1. This was tested by paired t-test.

Throughout a P value of <0.05 (two-tailed) was considered statistically significant. All values are provided as mean±s.e.mean unless otherwise indicated.

Results

1 Reproducibility of dose-response curves

Effective doses for isoprenaline induced increases in heart rate, shortening of QS₂c and fall in diastolic blood pressure and for the exercise induced tachycardia did not differ significantly between day 1 of the two study periods and the corresponding ED-ratios were not significantly different from 1 (Table 1).

Table 1.

Reproducibility of dose-response curves.

2 Cardiovascular effects of xamoterol at rest

Compared with the control day 1 (when no xamoterol was given) resting supine heart rate and systolic blood pressure were significantly higher and QS₂c and PEP were significantly shorter 90 min after administration of xamoterol on day 8 and day 22 (Table 2). There were no differences for any of these parameters between day 8 and day 22. On the other hand, diastolic blood pressure 90 min following the first dose of xamoterol (day 8) was not significantly different from diastolic blood pressure at the control day (day 1) but there was a significant difference in resting diastolic blood pressure between day 8 and day 22 (Table 2). There were no significant differences between placebo or ketotifen for resting systolic and diastolic blood pressure, QS₂c and PEP. Resting supine heart rate on day 8 and day 22 tended to be slightly slower with ketotifen, but this trend was not statistically significant (P =0.096 ANOVA treatment time interaction).

Table 2.

Baseline values for heart rate, blood pressure and systolic time intervals immediately prior to start of the isoprenaline infusion in the absence of xamoterol (day 1) and 90 min following the first (day 8) and the last (day 22) dose of xamoterol, respectively.

3 Effect of the first dose of xamoterol on isoprenaline induced haemodynamic changes and exercise induced increase in heart rate

Isoprenaline dose-dependently increased heart rate (Figure 1) and systolic blood pressure, lowered diastolic blood pressure and shortenend QS₂c. Following 1 week pretreatment with placebo the first dose of xamoterol shifted the dose-response curves for all isoprenaline induced cardiovascular responses and for exercise evoked tachycardia to the right (Table 3a). After 1 week treatment with ketotifen the first dose of xamoterol given on day 8 caused a significantly smaller rightward shift in the isoprenaline induced increase in heart rate compared with the rightward shift after 1 week treatment with placebo (P <0.05; 95% CI: 0.04, 0.31 for log ED₁₅-ratio) (Table 3). All other parameters were not influenced by 1 week’s pretreatment with ketotifen as evidenced by ED-ratios (1/day 1) which were not different between placebo and ketotifen (Table 3).

Figure 1.

Isoprenaline induced change in heart rate during the control day (day 1: ○, •) and 90 min after the first (day 8: ▵, ▴) and last (day 22: □, ▪) dose of xamoterol. The effects of xamoterol were investigated in the absence (a) white, open symbols and presence (b) black, filled symbols of ketotifen. Doses of isoprenaline are presented on a logarithmic scale.

Table 3.

a) Effect of ketotifen on the cardiovascular responses to i.v. isoprenaline and exercise after the first dose of xamoterol. Numbers indicate mean dose ratios (DR) (±s.e.mean) for effective doses (ED) of day 8 over effective doses of day 1. 95% confidence intervals (CI) were calculated on the basis of log-transformed dose ratios.

4 Effect of chronic dosing of xamoterol on isoprenaline induced haemodynamic changes and exercise induced increase in heart rate

After 2 weeks treatment with xamoterol ED-ratios (1/day 8) for all isoprenaline effects and exercise induced tachycardia were greater than 1 indicating a further rightward shift with chronic dosing. The magnitude of this additional rightward shift between day 8 and day 22 was however not affected by treatment with ketotifen (Table 3b).

Compared with the effects of the first dose of xamoterol on day 8 one subject (number 10) was extraordinarily sensitive to the antagonistic effects of xamoterol on isoprenaline induced shortening of QS₂1/day 8 of 38.1 (placebo) and 18.1 (ketotifen), respectively. The goodness of fit of these subject’s dose-response curves for day 8 and 22 was however satisfactory with coefficients of determinations (r²) ranging from 0.80 to 0.97 and the slopes of the regression were significantly different from zero. Mean (±s.e.mean) dose ratios (1/day 8) including this subject were 5.74±3.60 (placebo) and 4.53±1.99 (ketotifen); if this subject was excluded they were 2.15±0.35 and 3.02±1.44 respectively. The result of significance testing by paired t-test was not affected if this subject was excluded from comparison of the day 22/day 8 QS₂c dose ratios.

5 Xamoterol plasma concentrations

Xamoterol plasma concentrations were higher on day 22 90 min following intake of the last dose of xamoterol compared with xamoterol levels on day 8 90 min following intake of the first dose. There was, however, no significant difference between plasma xamoterol levels during placebo or ketotifen treatment (Table 4).

Table 4.

Mean plasma xamoterol concentrations (ng ml⁻¹) 90 min after administration of 200 mg p.o.

Discussion

We have previously shown that chronic administration of the β₂-adrenoceptor agonist terbutaline desensitizes β₂-adrenoceptor mediated cardiovascular functions in young, healthy volunteers and that this desensitization can be attenuated by ketotifen [3, 4]. In this study we aimed to investigate whether chronic stimulation of β₁-adrenoceptors by xamoterol would desensitize β₁-adrenoceptor mediated cardiovascular responses and if so whether this could likewise be influenced by ketotifen.

1 Effect of xamoterol on cardiovascular function in healthy volunteers

Our study confirms the β₁-adrenoceptor partial agonist activity of xamoterol in healthy humans. Under resting conditions xamoterol slightly increased heart rate and systolic blood pressure and shortened QS₂c and PEP indicating its positive chronotropic and inotropic action. However, during bicycle exercise or i.v. infusion of isoprenaline, i.e. under the conditions of increased adrenergic drive xamoterol behaved as an antagonist causing a rightward shift of all cardiovascular responses. During isoprenaline infusion the most pronounced rightward shift was seen for the shortening of QS₂c. Since this is a response mainly mediated by β₁-adrenoceptors [8, 12] this observation is consistent with a predominant action of xamoterol at β₁-adrenoceptors as shown previously [5, 13, 14].

2 Does xamoterol induce desensitization of β-adrenoceptor mediated function in man in vivo?

The positive chronotropic and inotropic effects of xamoterol at rest were not attenuated following 2 weeks administration. This suggests that chronic xamoterol treatment does not desensitize cardiac β₁-adrenoceptors. On the other hand, after 2 weeks administration xamoterol produced a further rightward shift in the exercise and isoprenaline induced tachycardia and the isoprenaline induced shortening of QS₂c. We cannot completely exclude the possibility that this additional shift is due to β₁-adrenoceptor desensitization. A more likely explanation, however, is the higher plasma levels of xamoterol observed after 2 weeks dosing. Taken together, our results do therefore not provide conclusive evidence for a desensitization of β₁-adrenoceptor mediated effects after chronic dosing of xamoterol. In keeping with our results, no evidence for the development of ‘tolerance’ to the haemodynamic effects of xamoterol has been obtained in clinical trials in patients with heart failure [15–17]. On the other hand, these data appear to contradict our previous conclusion [6] that a 2 weeks treatment of healthy volunteers with xamoterol desensitized β₁-adrenoceptor mediated effects as demonstrated by a reduction in exercise induced tachycardia and isoprenaline induced increase in systolic blood pressure. However, in contrast to the present study, in our previous study cardiovascular responses after 2 weeks treatment with xamoterol were compared with control data that had been determined before the first dose of xamoterol had been administered. Accordingly, it is well possible, that the β₁-adrenoceptor antagonistic effect of the partial agonist xamoterol was responsible for the ‘desensitizing’ effect of xamoterol observed in our previous study [6]. It has been shown both in patients with heart failure [18, 19] and in healthy volunteers [20] that chronotropic responses and parameters of cardiac systolic performance and inotropy can desensitize following long-term exposure to a full β₁-adrenoceptor agonist. We therefore suggest that the failure of xamoterol to desensitize human cardiac β₁-adrenoceptors is due to the fact that, in humans it is only a weak partial β₁-adrenoceptor agonist.

3 Is ketotifen able to modulate β-adrenoceptor mediated cardiovascular function in humans in vivo?

Ketotifen did not change the cardiovascular responses to xamoterol after chronic dosing both at rest and during adrenergic stimulation. Since xamoterol per se did not desensitize β₁-adrenoceptor mediated responses after chronic dosing the lack of an effect of ketotifen is not surprising.

However, pretreatment with ketotifen significantly attenuated the antagonistic effect of the first dose of xamoterol on isoprenaline induced tachycardia. Plasma xamoterol levels were not significantly different following ketotifen and placebo pretreatment. Therefore this difference cannot be explained by a pharmacokinetic interaction between ketotifen and xamoterol whereby ketotifen could potentially alter the disposition of xamoterol. In addition this effect was specific for the isoprenaline induced tachycardia; had it been the consequence of a pharmacokinetic interaction with different plasma drug levels for placebo and ketotifen pretreatment one would have expected a systematic difference in the responsiveness of all cardiovascular parameters. Neither the isoprenaline induced shortening of QS₂c nor the exercise induced tachycardia were affected by 1 week pretreatment with ketotifen. The power of our study to detect differences in these two parameters was superior to the power to detect differences in isoprenaline induced tachycardia. Therefore we are confident that we have not overlooked any systematic change in QS₂c and exercise induced tachycardia by ketotifen. Isoprenaline induced tachycardia is mediated by both β₁- and β₂-adrenoceptors [21, 22] whereas isoprenaline mediated shortening of QS₂c [8, 12] and exercise induced tachycardia [21, 22] are predominantly mediated by β₁-adrenoceptor stimulation. Consequently the differential impact of ketotifen on isoprenaline and exercise induced tachycardia (in the presence of xamoterol) would suggest that 1 week pretreatment with ketotifen specifically sensitized the β₂-adrenoceptor mediated component of the increase in heart rate caused by isoprenaline. This result is entirely consistent with our previous finding that, in healthy volunteers, ketotifen administration shifted the ratio high-to-low affinity state of lymphocyte β₂-adrenoceptors towards high affinity state [23]. Since the high affinity state of the β₂-adrenoceptor is essential for coupling to adenylyl cyclase [24]1/adenylyl cyclase coupling resulting in an enhanced β₂-adrenoceptor responsiveness.

Our results demonstrate, that, in humans, xamoterol is a partial β₁-adrenoceptor agonist with positive inotropic and positive chronotropic properties at rest and β₁-adrenoceptor blocking activity under the conditions of increased adrenergic drive. We found no evidence for a desensitization of β₁-adrenoceptor mediated cardiovascular functions following 2 weeks administration of xamoterol and, hence, no influence of ketotifen on xamoterol effects. In the presence of xamoterol ketotifen sensitized the β₂-adrenoceptor mediated component of the isoprenaline induced tachycardia so that the antagonistic effect of xamoterol on isoprenaline induced tachycardia was significantly reduced compared with placebo.