Introduction
Junctional ectopic tachycardia (JET) is a postoperative supraventricular arrhythmia characterised by atrioventricular dissociation and increased automaticity at the atrioventricular node or bundle of His. It typically occurs after repair of congenital heart defects with incidence rates ranging from 1.8 to 22% depending on surgical complexity.Reference Yoneyama, Tokunaga and Kato1–Reference Kylat and Samson3 In JET, atrial contraction occurs against closed AV valves, limiting preload and potentially causing hypotension.Reference Kylat and Samson3 While JET-related mortality has become rare, the arrhythmia remains clinically relevant due to its potential to cause haemodynamic instability.Reference Alasti, Mirzaee and Machado4–Reference Paluszek, Brenner and Pichlmaier7
In unstable patients, antiarrhythmic therapy is initiated promptly. In stable cases, initial non-pharmacologic strategies—such as electrolyte correction, optimised sedation (including prophylactic dexmedetomidineReference Ghimire and Chou8), targeted hypothermia, atrial overpacing, and catecholamine reduction—are preferred. If these fail, pharmacologic treatment is initiated.Reference Sasikumar, Kumar and Balaji9 Amiodarone remains widely used due to its broad antiarrhythmic profile. As a class III agent with a long half-life (14–47 days), it blocks potassium, sodium, and calcium channels, prolonging action potentials and refractory periods.Reference Singh10,Reference Punnam, Goyal and Kotaru11 It also suppresses automaticity and conduction but may cause bradycardia, hypotension, QT prolongation, and organ toxicity.Reference Singh10–Reference Florek, Lucas and Girzadas12
Other antiarrhythmics such as esmolol, propranolol, sotalol, procainamide, and ivabradine have been used, but none combine rapid onset, titratability, and minimal inotropic effects. Landiolol, an ultra-short-acting, highly β1-selective intravenous beta-blocker (half-life 3.5 min), offers precise dose control and fast reversibility. Developed in Japan and approved in Europe in 2016 for supraventricular tachycardias, it has the highest β1-selectivity ratio (277:1) among beta-blockers, which contributes to its favourable haemodynamic profile.Reference Iguchi, Iwamura and Nishizaki13–Reference Syed15 Importantly, landiolol exerts only minimal negative inotropic effects, making it particularly suitable for use in vulnerable postoperative paediatric patients. Its mechanisms—heart rate reduction, slowed conduction, and prolonged AV nodal refractoriness—support its use in JET, though paediatric data remain limited.Reference Syed15
While isolated case reports have described the use of landiolol in paediatric JET—either in combination with amiodaroneReference Hasegawa, Oshima and Maruo16 or following amiodarone monotherapyReference Saiki, Nakagawa and Ishido17—systematic data on this combination remain lacking. To our knowledge, this is the first comparative study evaluating landiolol plus amiodarone versus amiodarone monotherapy for postoperative JET in children.
Material and methods
Between January 2022 and June 2024, 962 children with congenital heart defects were admitted to our paediatric cardiac ICU at the German Paediatric Heart Centre Bonn following heart surgery. Of these, 31 developed postoperative JET and were treated with either amiodarone alone or in combination with landiolol. These patients were included in this retrospective single-centre cohort study.
Inclusion criteria were (1) age <18 years, (2) postoperative JET onset within 7 days, (3) treatment with amiodarone ± landiolol, and (4) documented JET onset, treatment protocol, and response. Patient selection is summarised in Figure 1a.
Figure 1. (a) Flowchart illustrating patient inclusion and exclusion. Of 962 children undergoing cardiac surgery, 31 developed postoperative JET and received either amiodarone monotherapy or a combination of amiodarone and landiolol. Seven patients were excluded due to hypotension (n = 3) or ECMO support (n = 4), resulting in a final analysis cohort of 24 patients. (b) Time to frequency control (FQC) and sinus rhythm (SR) in patients treated with amiodarone alone (A) or in combination with landiolol (A & L). *p = 0.02, ns = not significant. A: amiodarone, FQC: frequency control, L: landiolol, ns: not significant, SR: sinus rhythm.
JET diagnosis and monitoring
JET was diagnosed based on surface ECG showing heart rates >95th percentile for age or >170 bpm, with or without AV dissociation and narrow QRS complexes.Reference Collins, Van Hare and Kertesz18,Reference Lue, Wu and Wang19 Atrial ECGs via temporary wires were used to confirm the diagnosis. If there was doubt regarding the diagnosis of JET without atrioventricular dissociation, adenosine was given to confirm the diagnosis. Sustained sinus rhythm was defined as an intrinsic atrial rhythm observed on a surface electrocardiogram (ECG) in the absence of atrial or ventricular pacing, persisting for at least 12 consecutive hours. When atrial pacing was required solely due to a slow sinus rate, but with preserved sinus node activity and atrioventricular conduction, this was classified as sinus rhythm restoration. However, atrial pacing was not considered equivalent to sinus rhythm in patients with absent sinus activity or junctional escape rhythm.
Treatment protocol
Initial non-pharmacologic management included correction of electrolytes, deepened sedation (incl. dexmedetomidine), muscle relaxation in unstable cases, targeted hypothermia (35.0–35.9 °C), reduction of catecholamines—especially epinephrine—to the lowest effective dose based on serial echocardiographic assessment of ventricular function, and atrial overdrive pacing (AOO or VAT mode, i.e., atrial pacing without sensing or synchronisation to intrinsic ventricular activity).Reference Janousek, Vojtovic and Gebauer20
If ineffective, pharmacologic rate control was initiated using amiodarone (20 mg/kg/day, with a bolus of 5 mg/kg over 30–60 min) alone or with landiolol. The antiarrhythmic approach and landiolol addition were at the physician’s discretion. Titration followed no standardised protocol but was adjusted stepwise to achieve adequate heart rate control while maintaining haemodynamic stability and avoiding hypotension or perfusion deficits. Landiolol was started at 10 µg/kg/min and titrated up to 80 µg/kg/min as needed.Reference Domanovits, Wolzt and Stix21
In patients with impaired ventricular function, echocardiography was performed at least every 8 h. Data on central venous saturation and lactate levels, used as surrogate markers for heart function, were measured before starting landiolol and 3 h (for lactate levels) or 4–6 h (for central venous saturation) after initiation. Landiolol was tapered or stopped if arterial hypotension or increased norepinephrine requirement occurred. Thyroid function was monitored during amiodarone therapy, with substitution as needed.
Outcome measures
Primary endpoints:
-
a) Time to frequency control: Defined as a ≥20% reduction in heart rate from peak JET rate.
-
b) Time to sinus rhythm: Defined as return to sinus rhythm confirmed by electrocardiogram.
-
c) Absence of recurrence within 12 h of achieving frequency control or sinus rhythm.
Landiolol-specific evaluations:
-
• Central venous saturation (at baseline and 4–6 h post-initiation)
-
• Lactate levels (at baseline and 3 h post-initiation)
-
• Incidence of arterial hypotension requiring treatment modification
Statistical analysis
Continuous variables were reported as means ± SD or medians (IQR), as appropriate. Normality was tested using the Shapiro–Wilk test. Group comparisons were performed using the t-test or Mann–Whitney U test for continuous variables and the chi-square or Fisher’s exact test for categorical variables.
For both populations, categorical outcomes (e.g., frequency control rates, amiodarone bolus administration, and vasoactive support) were analysed using Fisher’s exact test. Time-to-event outcomes (e.g., time to frequency control or sinus rhythm) were compared using the Mann–Whitney U test. Associations between landiolol dose and clinical endpoints were evaluated using Spearman’s correlation coefficient.
Both intention-to-treat (ITT, n = 31) and per-protocol (PP, n = 24) analyses were performed. The ITT population included all treated patients, while the PP analysis excluded those who required ECMO or discontinued landiolol due to arterial hypotension.
Effect sizes (Cohen’s d or r) were reported where significant. A p-value <0.05 was considered statistically significant. Statistical analyses were conducted using GraphPad Prism 10 (GraphPad Software, San Diego, CA, USA) and SPSS v29.0 (IBM Corp., Armonk, NY, USA).
Results
Unless stated otherwise, results refer to the per-protocol cohort (n = 24): 15 patients received amiodarone plus landiolol (A & L), and 9 patients received amiodarone monotherapy (A) (see Figure 1a). The patient populations in both treatment groups were similar with respect to surgical complexity. The Aristotle score did not differ significantly between the A & L group and the A group (median 9, IQR 2 in both groups; p = 0.49). Likewise, within the A & L group, patients who received landiolol first had a median score of 8.5 (IQR 3), compared to 10 (IQR 0.9) in those who received amiodarone first (p = 0.11). Median cardiopulmonary bypass times were similar between groups: median 167 min (IQR 40.5) in the A & L group versus 159.5 min (IQR 46.0) in the A group (p = 0.57). Aortic cross-clamp times were likewise comparable: 104 min (IQR 46.0) versus 105 min (IQR 25.5), respectively (p = 0.60). Initial postoperative arterial lactate levels showed no significant difference: a mean of 2.35 mmol/L (SD 1.02) in the A & L group versus 2.41 mmol/L (SD 1.00) in the A group (p = 0.89).
No landiolol-associated complications were observed in the final study cohort. However, three patients excluded from analysis developed arterial hypotension, most likely attributable to landiolol administration, at a maximum dose of 40 µg/kg/min.
Among the 15 patients receiving combination therapy, 9 initially received amiodarone before landiolol was added. Both groups had comparable underlying congenital heart disease diagnoses.
Patient characteristics and treatment response
Demographics were similar except for a higher proportion of female patients in the A & L group (60 vs. 0%, p = 0.002). The three most common congenital heart diseases were atrioventricular septal defect, tetralogy of Fallot, and ventricular septal defect.
JET characteristics and treatment
As shown in the table,
The timing of JET onset after surgery was similar between groups:
-
• Combination therapy group: median 11.2 h (range 0–133.2)
-
• Amiodarone monotherapy group: median 11.3 h (range 0.3–178.7) (p > 0.99).
Dexmedetomidine was administered postoperatively in 86.7% of A & L patients and 88.9% of A-only patients; maximum doses did not differ significantly (median 1.0 vs. 1.1 µg/kg/h, p = 0.31).
A total of 47.8% of all patients maintained a core temperature within the target range (35.0–35.9°C). Time to frequency control (median 6.5 vs. 10.8 h, p = 0.30, r = 0.19) and time to sinus rhythm restoration (median 41 vs. 45.2 h, p = 0.81) were not significantly different between those with or without target temperature achievement.
Time to frequency control, sinus rhythm, and heart rate development
Time to frequency control was significantly shorter in patients who received amiodarone and landiolol compared to those who received amiodarone alone (median 6.7 vs. 14.7 h, p = 0.02, Cohen’s d = 1.05; large effect), as shown in the table and Figure 1b.
Among patients who started with landiolol, heart rate decreased significantly within 40–120 min (mean difference: −23.7 beats per min, 95% CI: −45.4 to −1.9, p = 0.04, r = 0.45; medium effect), while no significant reduction occurred in patients who started with amiodarone during this period (−3.5 beats per min, 95% CI: −12.7 to 5.9, p = 0.43, r = 0.10; small effect), as illustrated in Figure 2. Time to sinus rhythm restoration remained comparable between groups (median 51 vs. 40 h, p = 0.93), as shown in Figure 1b and the table. Within patients who received amiodarone and landiolol, initiating landiolol first was associated with a trend toward a shorter time to frequency control (median 2.4 vs. 8 h, p = 0.05, Cohen’s d = 1.49; very large effect), as shown in Figure 2 and the table; however, this did not result in significant differences in sinus rhythm restoration (p = 0.55).
Figure 2. Heart rate reduction after 40–120 min: comparison between landiolol and amiodarone. A: amiodarone, bpm: beats per minute, HR: heart rate, L: landiolol.
Per-protocol analysis (n = 24) showed a 100% frequency control rate in both groups. In the intention-to-treat analysis (n = 31)—which included four ECMO cases and three therapy discontinuations due to hypotension—frequency control was achieved in 78.9% of A & L patients (15/19) and 75.0% of A-only patients (9/12) (p = 1.00).
All ECMO cases had initially received combination therapy. In two patients, landiolol had been discontinued ≥120 min before cannulation, making a causal role in the subsequent haemodynamic decompensation unlikely given its ultra-short half-life. One of the remaining patients experienced sudden haemodynamic decompensation following a 10 mg/kg amiodarone bolus—higher than the dose recommended in our protocol—which necessitated ECMO implantation.
All patients requiring ECMO support recovered fully during the subsequent clinical course without any residual deficits.
The three patients who discontinued landiolol early due to hypotension also received combination therapy. All seven were classified as ITT non-responders, as rate control could not be assessed due to early termination of treatment.
Landiolol dosing and outcomes
Landiolol dosing did not significantly impact time to frequency control or sinus rhythm. The median landiolol dose administered was 30 µg/kg/min (range 10–80; mean 38 ± 25), as shown in the table. The correlation analysis confirmed no meaningful relationship between the maximum landiolol dose and these outcomes (Spearman’s r = 0.095, p = 0.74 for frequency control, 95% confidence interval: −0.464 to 0.580, p = 0.78 for sinus rhythm).
The overall duration of landiolol therapy was 29.3 h (range 6–82). Amiodarone therapy lasted for a median of 85.7 h (range: 15.7–260) in patients who received amiodarone and landiolol and 68 h (range: 8.33–200) in patients who received amiodarone alone (p = 0.59).
The interval between initiation of the first and second antiarrhythmic agents was similar in both combination subgroups (median 3.3 vs. 3.7 h, p = 0.39; see table)
As shown in the table, catecholamine use was similar between groups (all p > 0.05). The use of epinephrine (67 vs. 44%), norepinephrine (73 vs. 44%), and milrinone (73 vs. 67%) at JET onset did not differ significantly (all p ≥ 0.21). An amiodarone bolus (5 mg/kg) was administered in 47% of A & L patients and 56% of A-only patients (p = 1.00).
Lactate levels and central venous saturation were measured before and—as described in the Methods—3–6 h after initiation of landiolol therapy, allowing intra-group comparisons. Although data for the amiodarone group were available, these were not analysed and therefore no inter-group comparison was performed for these variables.
Discussion
Combination therapy with landiolol and amiodarone resulted in faster frequency control than amiodarone alone (median 6.7 vs. 14.7 h, Cohen’s d = 1.05), especially when landiolol was initiated first (median 2.4 vs. 8 h, d = 1.49). Despite relatively high doses, time to frequency control was slower than in previous studies—likely due to differences in patient selection or study design.Reference Saiki, Nakagawa and Ishido17,Reference Sagawa, Suzuki and Takei22,Reference Miyake, Fujita and Yoshizawa23
In addition, animal data suggest that locally released norepinephrine can accelerate otherwise slow AV junctional pacemaker activity, supporting a role of heightened sympathetic tone in sustaining automaticity-driven JET and limiting antiarrhythmic efficacy.Reference Sugiyama, Satoh and Ishida24
Sinus rhythm restoration times were similar (51 vs. 40 h, p = 0.93), consistent with amiodarone monotherapyReference Kovacikova, Hakacova and Dobos25,Reference Raja, Hawker and Chaikitpinyo26 but longer than in landiolol-only cohorts.Reference Yoneyama, Tokunaga and Kato1,Reference Tokunaga, Hiramatsu and Kanemoto14 The lower conversion rate to sinus rhythm in our cohort may partly reflect our stricter definition (≥12 h sustained), which may have excluded transient or borderline cases. Longer amiodarone use in the combination group likely reflects clinical routine, where landiolol is tapered after control is achieved, while amiodarone was maintained for continued stability.
As outlined earlier, impaired AV synchrony in JET reduces preload. Rapid heart rate reduction with landiolol may restore haemodynamic stability by enabling effective overdrive pacing and AV synchrony—even before sinus rhythm is achieved. Thus, the initial therapeutic goal is rate control rather than immediate rhythm conversion. Earlier rate control has been linked to shorter ventilation and ICU times.Reference Lim, Mok and Loh27 In our cohort, early landiolol use had the greatest effect (Cohen’s d = 1.49), and a significant heart rate drop within 40–120 min (r = 0.45) suggests rapid onset of action.
Given the variability in timing and overlap with disease-related haemodynamic instability, we did not quantify hypotension in the amiodarone group. Known amiodarone-associated hypotension—typically occurring within 1–12 hReference Kovacikova, Hakacova and Dobos25, Reference Saul, Scott and Brown28, Reference Maghrabi, Uzun and Kirsh29 —likely contributed to the decision to avoid bolus administration in 44% of patients. This risk was underscored by one case of acute haemodynamic decompensation requiring ECMO shortly after a 10 mg/kg amiodarone bolus. In contrast, arterial hypotension likely attributable to landiolol required therapy discontinuation in 11.1% of patients. However, consistent with prior reports, these events were milder and more rapidly reversible, reflecting landiolol’s ultra-short half-life. These findings—together with ECMO requirements in select cases—highlight the importance of tailoring antiarrhythmic strategies to individual haemodynamic profiles, particularly when combining agents with overlapping depressant effects.
Landiolol dosing was adjusted individually, as no standardised paediatric protocol exists. Its short half-life permits precise titration, though close monitoring remains essential. Compared to esmolol—another β1-selective agent used in paediatric JET—landiolol offers higher β1-selectivity and faster offset (3.5 vs. 9 min), which may improve tolerability. While direct paediatric comparisons are lacking, data from adult ICU settings suggest superior rate control with better haemodynamic stability,Reference Si, Yuan and Shi30 supporting its use in vulnerable populations such as neonates and postoperative cardiac patients.
While amiodarone remains a cornerstone in JET management, our findings suggest that early initiation of landiolol may facilitate more rapid rate control without compromising haemodynamic stability. Lactate levels and central venous oxygen saturation—used as surrogate markers of cardiac function—remained stable during and after landiolol initiation. Its complementary mechanism—β-blockade reducing heart rate and myocardial oxygen demand, combined with amiodarone’s suppression of ectopic automaticity—may provide a synergistic effect that promotes recovery. This hypothesis warrants validation in prospective trials.
Limitations
This study’s retrospective design, small sample size (n = 24), and potential confounders (e.g., inflammation, surgical complexity, baseline function) limit generalisability. As an observational analysis, findings reflect associations rather than causation. Inconsistent adverse event reporting and varying sinus rhythm definitions hinder comparability with other studies. The lack of long-term follow-up precludes evaluation of late recurrences. Diagnostic and procedural heterogeneity may limit subgroup applicability. The non-protocolled choice of initial therapy introduces selection bias. Haemodynamic stability was assessed with invasive monitoring, catecholamine requirements, lactate levels, and central venous saturation; however, cardiac output and echocardiographic data were not systematically analysed.
Conclusions
The combination of landiolol and amiodarone—particularly with early landiolol use—was associated with faster frequency control and generally preserved haemodynamic stability in paediatric postoperative JET. This complementary strategy may offer clinical benefit and warrants further evaluation in prospective studies.
Table 1. Patient characteristics, treatment details, catecholamine use, and outcomes in postoperative junctional ectopic tachycardia treated with landiolol plus amiodarone versus amiodarone monotherapy
Data are presented as the number of patients (%) or median (range) unless otherwise specified.
JET: junctional ectopic tachycardia, FQR: frequency control, SR: sinus rhythm, A: amiodarone, L: landiolol, HR: heart rate, CVS: central venous saturation.
Acknowledgements
The authors would like to express their gratitude to the staff at the German Paediatric Heart Centre for their support in data collection and patient care. We also thank our colleagues for their valuable insights and contributions during the development of this study.
Financial support
This publication was supported by the Open Access Publication Fund of the University of Bonn.
Competing interests
The authors declare none.
Ethical standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the responsible institutional committee on human experimentation and with the principles of the Helsinki Declaration of 1975, as revised in 2008. The study protocol was approved by the Ethics Board of the University Hospital of Bonn (Approval Number 2024-386-BO, granted on September 26, 2024).
Open access
