Interatrial Shunting as a Treatment for Heart Failure: Where Do We Stand?

This article refers to ‘Interatrial shunt therapy in advanced heart failure: Outcomes from the open-label cohort of the RELIEVE-HF trial’ by J. Rodés-Cabau et al., published in this issue on pages 1078–1089. Patients with heart failure (HF) typically suffer from increased left-sided filling pressures causing exertional dyspnoea and pulmonary oedema. The creation of an interatrial shunt as a potential treatment for HF has been proposed a decade ago.1 The idea is that an interatrial shunt permits blood flow directly from the high-pressured, decompensated left atrium to the low-pressured, compliant right atrium. This left-to-right shunt acutely lowers left-sided filling pressures, both at rest and with exercise, which in theory leads to less HF symptoms and better exercise capacity, and presumably also to better outcomes on the long term. In the last 10 years, several interatrial shunt devices have been developed and tested in HF across the entire ejection fraction (EF) spectrum.2-4 The device that has been tested in the most patients is the Corvia atrial shunt, which creates an 8 mm shunt in the fossa ovalis. The specific diameter of 8 mm was chosen because haemodynamic modelling predicted that a shunt diameter of 8–9 mm would allow a clinically meaningful reduction in pulmonary capillary wedge pressure (PCWP), both at rest and with exercise, while right-sided overload and a reduction in systemic cardiac output would be acceptable.1 Two single-arm, open-label studies in HF patients with an EF ≥40% demonstrated safety and adequate shunt patency of the Corvia shunt device and indeed showed a reduction in PCWP at rest and with exercise during a follow-up of up to 6 months.2, 5 This signal for improved haemodynamics following implantation of the device was confirmed, yet in a relatively small clinical trial.6 However, in the much larger REDUCE LAP-HF II, randomized, blinded, sham-controlled, clinical trial in 626 patients, placement of the device was not associated with a reduction in HF rehospitalizations or improvement in functional status in the overall population.7 A post-hoc analysis from this larger trial suggested that the presence of latent pulmonary vascular disease (PVD), defined as an increase in pulmonary vascular resistance (PVR) with exercise, identified a subgroup of patients in which an atrial shunt might even worsen clinical outcomes, whereas in patients without latent PVD, the shunt might be beneficial.8 The RESPONDER-HF randomized trial is currently ongoing in which the Corvia device is tested in HF patients with EF ≥40% and low exercise PVR (ClinicalTrials.gov ID NCT05425459). In this issue of the Journal, Rodés-Cabau et al.9 present the results of the open-label, single-arm, RELIEVE-HF observational study. In this study, the V-Wave Ventura interatrial shunt device with a shunt diameter of 5.1 mm, was tested in 97 patients with HF regardless of EF. All patients had to be on optimal guideline-directed HF therapy, were predominantly male of elder age, 97% were in New York Heart Association class III, more than half had at least one HF hospitalization in the previous 12 months, and around half of the patients had an EF ≥40%. Placement of the device was safe, technical success rate was high, and all shunts were patent at latest follow-up, at least in the patients that were eligible for follow-up echocardiography. During follow-up, quality of life and functional health status, as assessed with the Kansas City Cardiomyopathy Questionnaire (KCCQ), improved at 1 month and remained 12–16 points higher at 12 months than before implantation. No differences in the change of KCCQ scores between patients with EF <40% and those with EF ≥40% were seen. After device implantation, left ventricular size on echocardiography decreased, also right ventricular (RV) size decreased and RV systolic function improved. By the end of follow-up, 13 patients (13.4%) had died and 39 patients experienced at least one HF hospitalization. The latter number was lower than the 56 patients that had at least one HF hospitalization in the year before implantation of the device. The authors compared the actual mortality rate with a predicted mortality rate based on the MAGGIC and BCN Bio-HF scores and reported that their observed mortality rate tended to be lower than the predicted ones. The observational results of a potential beneficial effect of interatrial shunting for HF are in line with the earlier observational studies with the Corvia device.2, 5 The excellent patency of this device is in clear contrast with an earlier generation of the Ventura device, in which a porcine pericardial tri-leaflet valve was sutured that could close when the pressure gradient became right to left. This earlier generation device, with the same shunt size of 5.1 mm, has been tested in 38 patients with HF with EF >15% and was associated with improvement in functional status, but only 50% of the shunts were patent at the end of the study, probably due to the porcine tri-leaflet valve, that degenerated quickly.4 On a mechanistically point of view, the echocardiographic data in the study of Rodés-Cabau et al. are rather interesting, especially the observed overall improvement in RV function. While such effect can be seen in the setting of recent onset volume load causing a hyperdynamic right ventricle, RV size on average did not increase. This suggests that volume overload was limited and improved RV contraction may be caused in parallel with improvement in left ventricular morphology and function through ventricular interdependence.10 Interestingly however, the effects were seen both in reduced and preserved EF, although the ‘preserved EF’ group also included patients with mildly reduced EF (i.e. 40–49%). It may be that a transformation from post-capillary pulmonary hypertension to volume load from the shunt is beneficial for the right ventricle, at least on the short term. Still, these are average numbers and it is unknown from the data how many patients in fact had increased pulmonary pressures and too much RV loading after the shunt. One matter of concern regarding this is the relative high mean PVR of 2.5 Wood units at baseline, because in patients with actual or latent PVD, increased pulmonary flow causing pulmonary hypertension and RV overload could be detrimental on the longer term. Therefore, especially the results with clinical outcome should be viewed very cautiously in such an observational, non-randomized and small study. Patients enrolled in a clinical trial typically receive more close monitoring that may allow timely adjustment of medication when symptoms occur, that could prevent a hospitalization for HF. Also, a reliable comparison with the MAGGIC and BCN Bio-HF risk scores is hampered by the fact that they were published in 2013 and 2014, respectively, an era with less treatment options for HF.11, 12 The potentially beneficial, clinical effects presented here do not directly guarantee positive results in a subsequent well-powered, randomized controlled trial, as recent experience with the Corvia device has shown. The subsequent RELIEVE-HF trial testing the Venture device in a randomized controlled, double-blind, setting is currently ongoing (ClinicalTrials.gov ID NCT03499236). How do the present results add to the existing data and where do we stand with interatrial shunting as a treatment for HF? The results regarding safety and patency of the new generation Venture V-Wave device are of course very positive for the field of interatrial shunting as a potential treatment for HF. Also the potentially beneficial effects on biventricular function and morphology are hopeful. However, it remains challenging to compare the present results with other studies involving interatrial shunting, mostly due to the heterogeneous study populations, especially regarding differences in the inclusion of patients with reduced, mildly reduced and preserved EF. Second, the use of different shunt sizes also makes a direct comparison between the devices difficult. For instance, the absence of an increase in RV size after the shunt in the present study is different from the effects seen in the studies with the Corvia shunt where RV size increased after creation of the shunt. This is likely due to the smaller shunt diameter of 5.1 mm in the present study compared to the 8 mm shunt of the Corvia device. A smaller shunt will be safer in terms of limited RV volume overload and extra flow through the pulmonary circulation, but may be less effective in reducing PCWP and subsequent HF symptoms. Even in a clinical trial setting with well-defined in- and exclusion criteria, the effects of interatrial shunting in HF thus seem to be highly heterogeneous. The effects of left-to-right shunting may be different in patients with reduced versus preserved EF and in patients with normal versus (mildly) increased PVR. For the latter group of patients with increased PVR, an interatrial shunt may even worsen clinical outcomes. The ideal HF patient receiving an optimal sized, interatrial shunt has yet to be determined (Figure 1). The ongoing RELIEVE-HF and RESPONDER-HF clinical trials will hopefully provide such answers and will further determine the potential place of interatrial shunting as a possible treatment for HF. Conflict of interest: The Department of Cardiology of the University Medical Center Groningen participated in the REDUCE LAP-HF I and REDUCE LAP-HF II trials and currently participates in the RESPONDER-HF trial. T.M.G. reports non-personal consultancy fees from Corvia Medical paid directly to the Institution. All other authors have nothing to disclose.