Letter: Affirming Cardiac Safety While Highlighting Areas for Further Research

We read with interest the article by Bajaj et al. [1] Their pooled analysis of three Phase III trials provides reassuring evidence on terlipressin's cardiac safety in HRS-AKI, showing bradycardia is more common but rarely serious (none reported as such) and infrequently requires intervention (dose interruption in 4.5%, reduction in 9.1%). This supports terlipressin's role in decompensated cirrhosis, consistent with AASLD and EASL guidelines [2, 3]. While the study benefits from the robust pooling of three well-conducted Phase III trials (OT-0401, REVERSE and CONFIRM), several methodological aspects merit further discussion to improve the interpretability and generalisability of the results. First, the retrospective cardiac adverse event analysis, with bradycardia reported at investigators' discretion without predefined criteria, risks reporting bias. Lack of standardised continuous ECG monitoring may under-detect subclinical arrhythmias [4]; QT prolongation was only systematically assessed in one trial, affecting just two patients. Future studies should incorporate protocol-driven ECG to enhance event capture. Second, patient exclusions for significant cardiovascular disease, while appropriate, limit real-world applicability where comorbidities prevail. In this advanced cohort (mean MELD ~33), subgroup analyses by age, sex or ACLF grade could reveal vulnerabilities. The near-significant link between higher baseline serum albumin and bradycardia (OR 1.74, p = 0.057) suggests volume status or albumin co-administration as moderators. Exploring these, especially in elderly or alcohol-associated hepatitis subgroups (well represented here), would refine risk profiles. Third, the generalisability of these findings beyond North American populations deserves consideration, given the trials' predominant enrolment in the United States and Canada. Aetiologies of cirrhosis and standards of care, including albumin use and ICU availability, vary globally, potentially influencing AE profiles [5]. Additionally, the relatively small number of events (e.g., only 22 bradycardia cases in the terlipressin group) limits statistical power to detect differences in rare outcomes such as serious atrial fibrillation. Adjusting for confounding variables—such as prior use of midodrine or octreotide (reported in ~40% of patients)—through multivariable models could strengthen causal interpretations, as univariate analyses showed no significant associations. This study paves the way for several important future research directions. Prospective trials should integrate advanced cardiac monitoring—such as Holter ECG or telemetry—to better capture transient arrhythmias, including torsades de pointes, especially given the high background prevalence of QT prolongation in cirrhosis. Incorporating cardiac biomarkers (e.g., troponin and BNP) and echocardiographic parameters could improve risk stratification by identifying patients with underlying cirrhotic cardiomyopathy who may be more susceptible to haemodynamic alterations. Direct comparative studies with norepinephrine—assessing not only arrhythmia rates but also dynamic haemodynamic responses—should prioritise non-ICU settings to evaluate real-world feasibility, given terlipressin's bolus administration advantage. Furthermore, personalised dosing strategies based on baseline heart rate or baroreceptor sensitivity may help minimise AEs and encourage interdisciplinary collaboration between hepatologists and cardiologists. In summary, Bajaj et al. [1] contribute valuably by affirming terlipressin's favourable cardiac profile in HRS-AKI. These suggestions aim to strengthen future methodological rigour and clinical relevance. Shanshan Huang: conceptualization, investigation, funding acquisition, visualization, writing – review and editing, writing – original draft. Jie Zhou: writing – original draft, formal analysis, software, writing – review and editing, supervision, project administration. The authors have nothing to report. The authors declare no conflicts of interest. This article is linked to Bajaj et al. papers. To view these articles, visit https://doi.org/10.1111/apt.70297. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.