Editorials
Atrial Fibrillation Does Not Matter in Japanese Ventricular Assist Device Patients? ― Half-Way Up the Hill ―
2019 Volume 83 Issue 6 Pages 1202-1203
Details
2019 Volume 83 Issue 6 Pages 1202-1203
Atrial fibrillation (AF) is a common comorbidity worldwide and an important determinant of worse clinical outcomes in patients with heart failure (HF) regardless of its phenotypes.1–3 Based on the hypothesis that AF prevention would improve survival in HF patients with impaired left ventricular (LV) function, many clinical trials have tested it but no pharmacological interventions has been convincingly shown to improve outcomes. For instance, the AF-CHF (Atrial Fibrillation-Congestive Heart Failure)4 and the ANDROMEDA (Antiarrhythmic Trial with Dronedarone in Moderate to Severe CHF Evaluating Morbidity Decrease)5 trials failed to find an advantage of a rhythm control strategy in patients who had moderate-severe LV systolic dysfunction. In addition, recently published impressive trials of a catheter-based rhythm control strategy for HF patients concomitant with AF could not conclude its superiority, as discussed focusing especially on the study design (e.g., large number of exclusions, less effect in more advanced stage of HF, limited number of events).6–9 Accordingly, clinicians should be cautious when interpreting these findings regarding rhythm control strategy compared with established anticoagulation strategies.10
Article p 1286
Notably, we may not need to pay too much attention to AF in patients who receive a LV assist device (LVAD). Because all LVAD patients receive anticoagulation therapy, and the importance of atrial function appears insignificant compared with sufficient hemodynamic support from the LVAD to provide enough perfusion to tissues and organs. Thus, compared with the need to balance the risks of hemorrhage and thrombus in patients with LVADs, the priority of AF management in these patients is much lower. Although the prevalence of preoperative and the incidence of postoperative AF in HF patients who undergo LVAD implant procedures, and the prognostic effects of hemocompatibility-related adverse events (HRAEs), have been demonstrated to date,11 little evidence in Japanese patients has been published. Actually, important regional differences exist not only in terms of health economy, medical infrastructure and management but also patients’ characteristics and their adherence to medical advice and lifestyle, which affect outcome prediction.3,12 Therefore, findings from different regions need to be confirmed in each region, especially in these specific HF populations with LVADs.
In this issue of the Journal, the J-MACS (Japanese Mechanically Assisted Circulatory Support) registry provides new evidence regarding the association between the presence of preoperative AF and subsequent HRAEs in Japanese HF patients who underwent LVAD.13 In 190 patients who were registered to the nationwide J-MACS, 23 (%) had preoperative AF. Over 70% of cases of AF persisted following LVAD implantation. Consistent with previous findings, predominantly from populations in the USA, 1-year survival rates free from HRAEs were comparable regardless of the presence of AF, even after propensity score-matched analysis. Moreover, they found a relatively higher rate, but not significantly, of surgically managed pump thrombosis in the AF group (0.16 vs. 0.04 events/year, incidence rate ratio 3.75, 95% confidence interval (CI) 0.87–16.1, P=0.075) when compared with the non-AF group among HRAEs. The lack of association between preoperative AF and HRAEs supports our clinical practice of deciding whether aggressive management of AF, including rhythm control, is necessary during the perioperative period of LVAD implantation, despite a minor effect that could not be ultimately excluded in their limited population. Their finding indicates that any hemodynamic effect of preoperative AF in the Japanese HeartMate II population might be minimal and that conventional anticoagulation is appropriate for this patient population. However, they have not shown the prognostic effects of persistent AF following LVAD implantation, maybe because of the limited sample size as they mention in the Study Limitations section. Enriquez et al reported that persistent AF (i.e., AF both pre- and post-LVAD implantation) was an independent determinant of death or HF hospitalization in patients who received a LVAD.14 In that report, the prognostic impact was significantly higher for HF hospitalization including right ventricular failure (hazard ratio 7.37, 95% CI 2.12–25.64) than for death (hazard ratio 2.65, 95% CI 0.96–7.35).14 Importantly, in Japan we also need to consider the development of right ventricular failure, as well as HRAEs, due to persistent AF after LVAD implantation, because of the longer waiting period to heart transplantation. Further investigations with larger sample sizes in Japanese LVAD populations to confirm the association between persistent AF and long-term post-LVAD outcomes are warranted (Figure).
Effect of atrial fibrillation (AF) on left ventricular assist device (LVAD) patients in the USA and Japan. HF, heart failure; PT-INR, prothrombin time-international normalized ratio.
In summary, Imamura et al should be congratulated for confirming the findings from US LVAD populations regarding the minor prognostic effect of preoperative AF on subsequent HRAEs in the largest Japanese LVAD cohort. Although this study has the big statistical limitation that negative results could not be confirmed with such a small sample size and insufficient power, and although special attention should be paid when interpreting the findings, their data were derived from the current maximum LVAD population in Japan, as they emphasize, and the study provides a preliminary guide for clinicians who manage Japanese advanced HF patients.
None.
