Comparison of Landiolol and Esmolol on Haemodynamic Responses During Weaning of Intensive Care Unit Patients with Reduced Ejection Fraction after Vascular Surgery

Postoperative weaning off mechanical ventilation and tracheal extubation are frequently performed procedures in the intensive care unit (ICU) that can be associated with complications in patients with cardiomyopathy. Sympathetic nervous system activation is a well-known contributor to the deterioration of cardiac performance, especially in patients with coronary disease or left ventricular dysfunction.^1,2 Weaning and tracheal extubation are stressful procedures that may be associated with adverse haemodynamic changes, such as tachycardia or hypertension; these are poorly tolerated in patients with coronary disease or impaired left ventricular function, who can develop MI or acute pulmonary oedema as a result.^3,4

Prophylactic management to prevent such adverse events has been explored, including the use of opiates, β-blockers or dexmedetomidine.^5,6 The use of opiates to treat pain and blunt stress is limited by their dose-dependent respiratory depressant effect.⁶ Blunting the adrenergic response is an alternative option, and β-blockers or α₂-adrenoceptor agonists are potential candidates, but the latter may be associated with an unwanted sedative effect.⁵

Dexmedetomidine has been used in difficult-to-wean ICU patients because it can reduce agitation and delirium. Dexmedetomidine reduces sympathetic tone by its central action, which contributes to its sedative effect, but can also be associated with undesired haemodynamic effects such as hypotension or bradycardia.^6,7

Short-acting β-blockers are currently used in many different clinical scenarios in the perioperative and critical care settings to treat tachyarrhythmia or control heart rate to reduce ischaemia or cardiac dysfunction.⁸

Esmolol bolus has been used successfully to treat tachycardia or hypertension during extubation or weaning procedures.^9–14 However, its routine use as a continuous infusion to prevent such haemodynamic adverse events is limited by the potential for the development of hypotension.¹¹ Landiolol is a novel ultra-short-acting β-blocker with a β_{1-adrenoceptor selectivity eightfold greater than that of esmolol.^8,15 Comparisons of esmolol and landiolol in animal models have shown that landiolol results in a more profound reduction in heart rate associated with minimal impact on blood pressure, whereas esmolol has a marked dose-dependent hypotensive effect.¹⁵}

Preliminary observations in surgical patients tend to confirm these results, with esmolol being associated with a significant blood pressure-lowering effect during extubation but landiolol showing no significant difference compared with saline.¹⁶

The capacity of landiolol to decrease heart rate while maintaining coronary perfusion and preserving cardiac contractility should theoretically be of benefit to patients with coronary disease and/or cardiac dysfunction.

The aim of the present study was to evaluate the cardiovascular response to and tolerability of landiolol in postoperative patients with cardiac dysfunction in the ICU during weaning and extubation compared with esmolol.

Methods

This study was a prospective single-centre randomised open-label study conducted in the ICU of the General Hospital of Athens ‘Georgios Gennimatas’ during the period from 2018 to 2020. This study was conducted as an open-label, unblinded trial because blinding was not possible due to differences in the rate of infusion of the two drugs.

All participants provided written informed consent and the study protocol was approved by the Ethics Committee of the General Hospital of Athens ‘Georgios Gennimatas’ (approval number: 3527/27-01-2018).

The primary endpoint was incidence of hypotensive episodes (defined as a mean arterial pressure [MAP] <60 mmHg for at least 5 minutes in a consecutive 10-minute period or systolic arterial pressure [SAP] <100 mmHg). Secondary endpoints were the incidence of serious adverse events such as bradycardia, pulmonary oedema, cardiac shock, bronchospasm, ischaemia, stroke and death.

Patient Population

Patients were included in the study if they were aged >18 years, had signed informed consent forms, were undergoing elective vascular surgery and had cardiac dysfunction with New York Heart Association (NYHA) functional classification >II.

Patients were excluded from the study if they had a contraindication to β-blockers (e.g. severe hypotension), cardiogenic shock, heart conduction block (second-degree or higher atrioventricular block), sinus node disease, severe bradycardia (heart rate <50 BPM), phaeochromocytoma, were experiencing an asthma crisis, had pulmonary hypertension (mean pulmonary artery pressure >15 mmHg) or pulmonary oedema, were pregnant or were not able to provide signed informed consent.

Patient Management and Procedures

Patients were managed identically during the surgical procedure, with a standard anaesthetic protocol used routinely at the General Hospital of Athens ‘Georgios Gennimatas’. Baseline cardiac function was evaluated by experienced clinicians using ultrasound machines (iE33 Philips Healthcare), including both left ventricular systolic (modified Simpson’s method) and diastolic function.

Patient heart rate, SAP, diastolic arterial pressure (DAP) and MAP were monitored using a Philips MX 550 monitor, and cardiac rhythm was recorded every minute up to 30 minutes after extubation of the trachea. Continuous oxygen saturation and blood gas measurement (pCO₂, pO₂) data were also collected. Collected data (heart rate, SAP, DAP and rhythm status) were exported from the hospital database and analysed in Microsoft Excel.

Description of the Intervention

After randomisation, immediately before the weaning procedure was started, patients were administered a continuous infusion of β-blockers (esmolol or landiolol), without a loading dose. Drug doses were determined according to the drug regimen recommended in the product information summary and based on our clinical experience with the drugs to reflect differences in drug potency. Landiolol (RAPIBLOC AOP Orphan) was initiated at a rate of 1 µg/kg/min and esmolol (ESMOCARD AOP Orphan) was initiated at a rate of 50 µg/kg/min.

The drugs could be uptitrated up to 10 µg/kg/min for landiolol and 200 µg/kg/min for esmolol based on patient haemodynamic responses. In the case of hypotension (as MAP <60 mmHg or SAP <100 mmHg) or bradycardia (heart rate <50 BPM), the dose of landiolol and esmolol could be adjusted (stepwise reductions of 1 or 50 µg/kg/min, respectively) or they could be stopped if there was no recovery in the haemodynamic response after a 10-minute period.

Statistical Analysis

Sample size was calculated based on the primary endpoint as a binomial dichotomous variable, and considering an expected incidence of hypotension of 8% and 35% in the landiolol and esmolol groups, respectively, for patients with cardiac dysfunction.^17,18 The online EasyMedStat tool (https://www.easymedstat.com/sample-size-calculator) was used to calculate sample size.

Assuming a 35% incidence of hypotension with esmolol as the reference group and targeting an incidence of 8% hypotension in landiolol group, after applying continuity correction, the sample size for each group was determined to be 19 to achieve a power of 80% to detect a difference in proportions of −0.27 between the two groups, at a two-sided p-value of 0.05. No major losses to follow-up were anticipated due to the short time frame of the study and the routine collection of data related to the study endpoint; thus, we considered a minimal 5% loss to follow-up and planned to recruit 20 patients per group.

Continuous variables (heart rate, SAP, DAP, MAP) are expressed as the mean ± SD at each time point for each group. The significance of differences between the two groups in mean heart rate, SAP, DAP and MAP at each time point was evaluated using Student’s t-test. Calculations were performed in Excel.

Results

Thirty-nine postoperative patients with heart failure were included in the study and randomly assigned to either landiolol (n=19) or esmolol (n=20). There were no statistically significant differences between the 2 groups. Mean age, LVEF, NYHA functional class, and the Acute Physiology and Chronic Health Evaluation II and Sequential Organ Failure Assessment scores were similar in the two groups. Demographic data for the landiolol and esmolol groups are presented in Table 1.

Heart failure was characterised by a mean left ventricular ejection fraction of 36.6 ± 7.6%, with 76.9% (n=30/39) of patients having NYHA Class II and only 23% (9/39) having NYHA Class III. The mean infusion rate was 2 ± 2.1 μg/kg/min for landiolol and 150 ± 50 μg/kg/min for esmolol.

The incidence of hypotension was 60% (12/20) in the esmolol group, compared with 0% (0/19) in the landiolol group (p=0.0001; 95% CI [32.8–78.1]). Although esmolol produced a more pronounced decrease in blood pressure than landiolol, both esmolol and landiolol decreased MAP relative to baseline at 15 minutes (−37 ± 3.4 mmHg, 95% CI [−30.7, −43.1 mmHg], p<0.0001 and −22 ± 3.1 mmHg, 95% CI [–15.7, –28.3 mmHg], p<0.0001, respectively) and 30 minutes (−50 ± 3.5 mmHg, 95% CI [−42.8, −57.1 mmHg], p<0.0001 and −26 ± 2.9 mmHg; CI [–20.0, –31.9 mmHg], p<0.0001, respectively). Esmolol-induced hypotension was managed by reducing the dose, with prompt blood pressure recovery. No patient required vasopressor support or fluid administration.

Both esmolol and landiolol decreased heart rate significantly compared with baseline at 15 minutes (−27 ± 3.6 BPM, 95% CI [−19.7, −34.29 BPM], p<0.0001 and −42 ± 3.3 BPM [95% CI –35.3, –48.7 BPM], p<0.0001, respectively) and 30 minutes (−34 ± 4.1 BPM, 95% CI [−25.6, −42.3 BPM], p<0.0001 and −48±3.3 BPM, 95% CI [–41.3, –54.7 BPM], p<0.0001, respectively). Landiolol produced a more rapid and substantial decrease in heart rate than esmolol, as shown in Figure 1. Mean heart rate measured 30 minutes after tracheal extubation was −40 ± 20 BPM in the landiolol group, compared with −30 ± 16 BPM in the esmolol group. However, this −10-BPM difference between the two groups did not reach statistical significance (p=0.09; 95% CI [−21.7, 1.72 BPM]). The more profound decrease in heart rate was not associated with haemodynamic deterioration in the landiolol group, although a significant reduction in MAP was recorded in the esmolol group. Changes in mean blood pressure over time are shown in Figure 2.

There were no differences in serious adverse events, except for hypotension, between the two groups (Table 2).

Discussion

To the best of our knowledge, this study is the first to compare the haemodynamic effects of esmolol and landiolol in a population of postoperative ICU patients with cardiac dysfunction. Data comparing landiolol and esmolol are scarce, and there are no data on patients with cardiac dysfunction. The principal reason for this may be due the lack of a validated indication for esmolol in this patient population. Furthermore, the use of esmolol in Japan is limited to intraoperatively, whereas outside Japan esmolol has been widely used both intra- and postoperatively in the cardiac surgery setting.^19,20 These differences have led to an absence of clinical studies comparing esmolol and landiolol in the postoperative period or other non-surgical settings. The only studies available comparing both drugs include a bolus regimen or short infusion, intraoperatively, making it difficult to detect any differences in haemodynamic responses. Indeed, one study reported no differences in terms of heart rate or blood pressure when landiolol or esmolol was used to control the haemodynamic response at intubation.²¹ However, another study was able to show a difference in terms of blood pressure response between the two drugs.¹⁶

There is no available study in white patients, but pharmacological studies evaluating the efficacy and safety of landiolol in White healthy volunteers showed a more pronounced bradycardic effect with a minimal impact on blood pressure compared with esmolol.^22–24 These results in healthy volunteers, along with our results, are in line with data from animal studies, which demonstrated that landiolol had less of a negative inotropic effect and less of a hypotensive effect than esmolol while maintaining a marked dose-dependent bradycardic effect.^12,25,26

The use of landiolol use to manage haemodynamic responses during extubation has been described in Japan, but only in patients without cardiac dysfunction.^27–29 The degree to which landiolol reduced heart rate and blood pressure tended to be higher among normotensive than hypertensive patients.^27–29 In one study using landiolol during emergency from anaesthesia and tracheal extubation, the degree to which landiolol reduced heart rate and blood pressure tended to be higher among normotensive than hypertensive patients.²⁹

Given the profile of patients who are at risk of developing complications at weaning and extubation, and the contribution of catecholamine stress to weaning failure, we believe landiolol can be valuable and provide safe heart rate control in this setting. Indeed, recovery from anaesthesia or a reduction in sedation to facilitate the withdrawal of ventilatory support can provoke a stress response, catecholamine release and agitation, resulting in tachypnoea, tachycardia and hypertension.² If not adequately controlled, it often requires increasing or resuming sedation, potentially prolonging mechanical ventilation. These haemodynamic changes are usually transitory, variable and unpredictable, and may lead to complications in patients with coronary or cardiac dysfunction.

Extubation is associated with tachycardia and weaning increases cardiopulmonary demand, which may trigger coronary ischaemia and MI.^2,3 The higher β₁-adrenoceptor selectivity of landiolol allows β₂-adrenoceptor-mediated relaxation of the coronary vessels, and its profound bradycardic effect should contribute to improving the oxygen demand/supply balance.

Similarly, stopping mechanical ventilation reduces intrathoracic pressure and increases the autonomic response, increasing left ventricular preload and afterload, which, in turn, increases left ventricular end-diastolic pressure.⁴ Hence, patients with hypertrophic cardiomyopathy or diastolic left ventricular dysfunction should benefit from a reduction in heart rate provided it is not offset by excessive negative inotropic effects.

Landiolol has been safely used to treat supraventricular tachyarrhythmia in patients with relatively severe ventricular cardiac dysfunction.^30–35 Landiolol is indicated in decompensated heart failure associated with tachyarrhythmia, but remains contraindicated in decompensated heart failure when the condition is considered not to be related to arrhythmia.

Our results confirm that landiolol can be safely used in patients with cardiac dysfunction and that it has a better profile than esmolol in this patient population. Our findings are also in line with the pharmacodynamic differences between and landiolol esmolol seen in White healthy volunteers, which were awaiting confirmation in patients.^22–24,36

Study Limitations

This study has several limitations. First, there was low incidence of short-term serious adverse events and we did not plan to collect data on serious adverse events in the long term. Our sample size was calculated based on the short-term incidence of hypotension and not tailored to identify any differences in long-term serious adverse events, especially those with such a low incidence. One could question the clinical relevance of the incidence of hypotension as our primary endpoint, but we believe that combined with a decrease in heart rate it reflects the better risk–benefit balance of landiolol in this patient population. The perioperative use of beta-blocker to prevent ischaemia or arrhythmia has been associated with hypotension, stroke and increased long-term mortality.^31,37 Thus far, the use of landiolol to prevent arrhythmia in cardiac surgery patients with cardiac dysfunction has not been associated with hypotension.³⁸ The only study providing long-term safety data did not find any negative effect of landiolol on mortality up to 5 years after landiolol had been used to prevent arrhythmia in patients undergoing thoracic oesophageal cancer surgery.³⁹

Second, our study design included a regimen that may have been optimised with esmolol, either by reducing the infusion dose at initiation (25 or 50 µg/kg/min) to minimise hypotension or increasing the dose up to 200 µg/kg/min to maximise heart rate control. These adjustments would have led to a greater gap in the heart rate-lowering effect between landiolol and esmolol (former option) or increased the incidence of hypotension in the esmolol group (latter option).

Third, our results only apply to the population of patients with cardiac dysfunction, in whom landiolol has already been used for several years in Japan but is only just starting to be used in white patients.^35,40,41 Fourth, extubation in other surgical settings is associated with tachycardia and hypertension, where the hypotensive effect of esmolol may be a benefit. Esmolol has been proven to be a useful option in this setting, which differs from the setting in the present study.^42–48

Conclusion

This is the first study comparing landiolol and esmolol for controlling heart rate in surgical patients after extubation. Landiolol was associated with a faster and more profound reduction in heart rate compared with esmolol. In addition to exhibiting a more pronounced bradycardic effect, landiolol had no effect on blood pressure, unlike esmolol, which was associated with a significant reduction in MAP. Further trials are required to compare landiolol and esmolol in different settings and in patients without heart failure.