Introduction

The prevalence of heart failure (HF) is approximately 1–2% in adults, and the overall incidence of HF is increasing, most notably in the context of an ageing population [1,2,3]. HF with preserved ejection fraction (HFpEF) already accounts for more than half of all HF hospital admissions [4, 5]. In the outpatient setting, the European Society of Cardiology (ESC) Long-Term Registry reports that 60% have HF with reduced ejection fraction (HFrEF), 24% have HF with mildly reduced ejection fraction (HFmrEF), and 16% have HFpEF [1, 6]. Patients with HFpEF present functional limitation, poor quality of life (QoL), and a higher mortality compared to healthy age-matched population [7,8,9,10]. Several studies also reported similar outcomes in HF patients irrespective of left ventricular ejection fraction (LVEF), with an annual mortality rate ranging from 1.3 to 24% [11,12,13].

Medical treatment has demonstrated to improve cardiac function, symptoms, and prognosis in patients with HFrEF, but failed to demonstrate benefits in HFpEF in large-scale clinical trials [1, 14,15,16,17,18,19,20,21,22,23]. To date, only diuretics are recommended in patients with HFpEF and signs of congestion in order to alleviate symptoms of HF. Recently, empagliflozin was found to be effective in reducing the primary composite endpoint of cardiovascular (CV) death or hospitalization for HF in the Empagliflozin in Heart Failure with a Preserved Ejection Fraction (EMPEROR-Preserved) trial [3, 24]. Also, dapagliflozin improved patient-reported symptoms compared to placebo in the PRESERVED-HF trial and reduced by 18% the combined risk of worsening HF or CV death among patients with HF and an LVEF > 40% in the Dapagliflozin Evaluation to Improve the LIVEs of Patients With PReserved Ejection Fraction Heart Failure (DELIVER) trial [25,26,27].

However, the management and treatment of HFpEF still remains a major unmet clinical need [3]. Treatments alternative to medical therapies, including devices, may have a major role [28].

The aim of this review is to summarize evidence regarding the use of percutaneous interatrial shunting devices in patients with HFpEF.

The role of left atrial pressure and the rationale for the development of interatrial shunt devices

Impaired left ventricular (LV) relaxation and compliance are the hallmark features of HFpEF. They lead to elevated LV filling pressure and left atrial pressure (LAP) at rest or during exercise and may cause pulmonary hypertension (PH), with consequent exertional dyspnoea and exercise limitation or pulmonary oedema (Fig. 1) [29,30,31,32,33,34].

Fig. 1
figure 1

Pathophysiologic mechanisms leading to dyspnoea in patients with HFpEF and possible targets for therapies, including the interatrial shunt devices. 6MWD, 6-min walking distance; HFpEF, heart failure with preserved ejection fraction; LA, left atrium; LV, left ventricle; QoL, quality of life; PCWP, pulmonary capillary wedge pressure; SGLT2, sodium/glucose cotransporter 2

Full size image

A rise in LAP results in atrial remodelling and failure. Left atrial (LA) disease has been recently introduced in the ESC guidelines for the management of HF and defined as a complex of subclinical structural, electrophysiological, and functional changes that affect the atria with the potential to produce clinical consequences [1]. LA disease causes HF symptoms, an increase risk of atrial fibrillation (AF), right ventricular (RV) dysfunction, impaired exercise capacity, and adverse outcomes [35]. Higher LAP, measured as pulmonary capillary wedge pressure (PCWP) at rest or during exercise, has been associated with higher mortality and morbidity [36], whereas decongestion therapies guided by a real-time indirect monitoring of LAP resulted in a reduction of HF hospitalizations [37,38,39].

Unloading the LA by shunting blood to the lower pressure reservoir of the right atrium (RA) and systemic veins may reduce pulmonary venous pressure and improve symptoms and outcomes in these patients (Fig. 1). The hypothesis of a potential benefit of an iatrogenic interatrial shunt in HF was based on the observation that in the setting of mitral stenosis, a condition also associated with elevated LAP and LA dysfunction, the coexistence of a congenital atrial septal defect (named Lutembacher syndrome) was associated with fewer symptoms and a more favourable clinical course, without exceeding the risk of right HF or stroke [40]. On the other hand, the closure of atrial septal defects in patients with unrecognized LV dysfunction may lead to abrupt elevation of LAP and pulmonary oedema [41].

A series of devices, creating a left-to-right controlled interatrial shunt, have been developed to decompress the LA in HF patients [32, 42,43,44]. The shunts work as on-demand, self-regulating LAP lowering systems, according to the pressure gradient between the LA and RA [45]. As LAP increases in response to any conditions (exercise, rise in systolic blood pressure, etc.), a small amount of LA blood is shunted to the RA, leading to a reduction in the LAP. The interatrial shunt creates a mismatch between pulmonary and systemic flow (Qp:Qs), possibly determining a risk of worsening right-sided HF. However, if the shunted volume is limited, the reduction in LAP overcomes the increase in right-sided volume, and pulmonary artery and right heart pressures remain unchanged or may be reduced as well. Traditionally, Qp:Qs ratios < 1.5 are known to be well-tolerated and without deleterious right-sided changes. Thus, the therapeutic interatrial shunt ratio (Qp:Qs) goal should be around 1.2 or slightly greater.

The following sections will describe devices which are currently under investigation in patients with HFpEF. Some of them already received CE approval (Tables 1 and 2, Fig. 2).

Table 1 Overview of trials on the use of device-based percutaneous treatments to decompress the left atrium in heart failure with preserved ejection fraction
Full size table
Table 2 Ongoing trials on the use of device-based percutaneous treatments to decompress the left atrium in heart failure with preserved ejection fraction
Full size table
Fig. 2
figure 2

The main characteristics of devices used to reduce left atrial pressure in patients with HFpEF. AF, atrial fibrillation; AFR, atrial flow regulator; CS, coronary sinus; DOACs, direct oral anticoagulants; LA, left atrium; IASD, interatrial shunt device; RA, right atrium

Full size image

Interatrial shunt device

Interatrial shunt device (IASD, Corvia Medical, Tewksbury, MA, USA) is a bare-metal nitinol frame device, creating a permanent 8-mm communication between the atria, allowing a physiological pressure-dependent left-to-right flow [46]. IASD is implanted using percutaneous trans-septal access via the femoral vein. The LA disc is exposed and retracted to the septum, and the RA disc is unsheathed to secure the device in place. The design of the device is based on predictive haemodynamic modelling which evaluated the relationship between shunt size and LAP reduction, based on human data. Particularly, the IASD targets a Qp:Qs of 1.3 [47].

IASD was firstly evaluated in a pilot study, assessing safety and efficacy in 11 patients with symptomatic HFpEF. The study showed that an IASD could be safely implanted. The legs of the device are flat on the LA side to minimize the risk of thrombus formation. After the procedure, patients were treated with aspirin, lifelong, and clopidogrel, generally for 6 months [48]. Patients with a history of AF were maintained on oral anticoagulants and clopidogrel at the discretion of the physicians. At 30-day follow-up, the devices remained patent, and LV filling pressures were reduced by 5.5 mmHg (19.7 ± 3.4 vs. 14.2 ± 2.7; p = 0.005) with evidence of early clinical benefit, improvement in New York Heart Association (NYHA) functional class, 6-min walking distance (6MWD), and QoL (Table 1) [49, 50].

These preliminary results were confirmed in the Reduce Elevated Left Atrial Pressure in Patients With Heart Failure (REDUCE LAP-HF) trial, a larger, non-randomized study, including 68 patients with LVEF > 40%. There were no peri-procedural complications, or major adverse cardiovascular events (MACE) or need for cardiac surgical intervention for device-related complications at 6-month follow-up. A significant reduction in exercise PCWP was observed at 6 months. Moreover, there was an improvement in functional and exercise capacity, whereas a modest increase in right heart cardiac output and RA pressure (RAP) was observed. Sustained device patency at 6 months was confirmed by left-to-right shunting (pulmonary/systemic flow ratio: 1.06 ± 0.32 at baseline vs. 1.27 ± 0.20 at 6 months, p = 0.0004) [51]. Follow-up of these patients was subsequently extended to 12 months, providing evidence of a sustained and meaningful clinical benefit, evaluated through QoL score, 6MWD, and NYHA functional class. Invasive hemodynamic studies performed in a subset of patients demonstrated a sustained reduction in the workload-corrected exercise PCWP (p < 0.01). Echocardiographic parameters at 12 months showed a modest but stable reduction in the LV end-diastolic volumes, without changes in LA and RA volumes. By contrast, a small but significant increase in the RV end-diastolic volumes raised some concerns [52].

The Transcatheter Interatrial Shunt Device for the Treatment of Heart Failure With Preserved Ejection Fraction (REDUCE LAP-HF I) trial was the first randomized, sham-controlled trial designed to determine the effectiveness of the IASD in patients with HFmrEF or HFpEF. The trial met its primary endpoint of effectiveness, with statistically significant lowering of PCWP during exercise at 1-month follow-up (p = 0.028). No major complications, including death, myocardial infarctions (MI), IASD occlusions or removals after the procedure, or strokes/transient ischemic attacks (TIA), were reported in either of the study arms. The rate of HF-related hospitalizations or emergency department/acute care facility visits requiring intravenous treatment was < 1% in the treatment arm compared with 9.1% in the control arm. However, the small sample size prevented to reach statistical significance [32].

The impact of IASD on CV mortality and HF events remained to be assessed. In patients treated with this device in the open-label REDUCE LAP-HF cohort, Kaye et al. observed a 33% lower mortality rate than that predicted by the Meta-analysis Global Group in Chronic Heart Failure (MAGGIC) prognostic model over the entire observation period [53].

The atrial shunt device for heart failure with preserved and mildly reduced ejection fraction (REDUCE-LAP II) was a prospective, randomized, multicentre, blinded, sham-controlled trial, enrolling 626 patients with symptomatic HF, a LVEF of at least 40%, PCWP during exercise ≥ 25 mmHg while exceeding RAP by at least 5 mmHg. Patients were randomly assigned (1:1) to receive either a shunt device or sham procedure. The primary endpoint was a hierarchical composite of CV death or non-fatal ischemic stroke at 12 months, rate of total HF events up to 24 months, and change in Kansas City Cardiomyopathy Questionnaire (KCCQ) overall summary score at 12 months. The trial failed to demonstrate difference between groups for the primary endpoint (win ratio 1.0 [95% CI 0.8–1.2]; p = 0.85) and in the individual components of the primary endpoint. There were no differences in the composite safety endpoint between the two groups (n = 116 [38%] for shunt device vs. n = 97 [31%] for sham procedure; p = 0.11), but MACEs were more common in the atrial shunt group compared with the sham control (p = 0.025). Patency was 100% at 12 months [54].

Ongoing studies investigating IASD (NCT03191656, NCT04632160) are reported in Table 2. Since the device has received CE approval in the Europe, the REDUCE LAP III (NCT03191656) will collect post-market data in consecutive patients with HF treated with the IASD System II, to further evaluate efficacy, safety, and QoL outcomes in a real-world setting.

V-Wave Shunt

The V-Wave Shunt device (V-Wave Ltd., Caesarea, Israel) consists of an hourglass-shaped nitinol frame encapsulated with a partially expanded polytetrafluoroethylene cover serving as an anchor for 3 porcine pericardial leaflets held together using a Prolene suture [46, 55]. The device is delivered via femoral vein and interatrial septal puncture. In the first design, a one-way bioprosthetic valve ensured only left-to-right shunting. A first-in-man experience with this device demonstrated initial safety and early beneficial clinical and haemodynamic outcomes in patients with HFrEF [56]. In a single-arm open-label study, enrolling HFrEF and HFpEF patients with NYHA functional class III or IV, interatrial shunting with the V-Wave system was safe and feasible (Table 1). However, a high rate of shunt stenosis (36%) or occlusion (14%) at 12 months, likely due to early valve degeneration, resulted in loss of efficacy [57].

These observations prompted the creation of a second-generation device, which preserved the hourglass shape but eliminated the valve component; instead, a 5.1-mm central opening was created. This device was tested in a pilot study including 10 patients with chronic HF and NYHA class ≥ III, despite optimal tolerated drug and device therapies [58].

The ongoing RELIEVE-HF (Reducing Lung Congestion Symptoms using the V-Wave Shunt in Advanced Heart Failure) (NCT03499236) trial will evaluate safety and efficacy of the second-generation V-Wave Shunt in patients with advanced HF (NYHA functional class III or IV), regardless of LVEF. The co-primary endpoints are the frequency of major device-related adverse events (time frame: 30 days after randomization) and a hierarchical composite of death, heart transplant or left ventricular assist device (LVAD) implantation, HF events, and change KCCQ (time frame: follow-up duration at endpoint analysis ranges from a minimum of 12 to a maximum of 24 months) (Table 2). The trial aims to randomize 500 patients. First results on 97 patients showed high implantation success rates and safety, as well as sustained improvements in QoL from the first month. Moreover, shunt patency through 12 months was 100% [57].

Atrial flow regulator

The atrial flow regulator (AFR, Occlutech, Istanbul, Turkey) is a self-expandable nitinol mesh braided into 2 flat discs creating a 1- to 2-mm fenestrated neck. As previous devices, it is delivered via femoral venous access following a trans-septal puncture. The central opening can have various diameters (6, 8, and 10 mm). However, only the 8-mm and 10-mm diameters obtained the CE mark for HF patients. The device is designed to allow interatrial bidirectional flow [42, 46, 59]. Due to the possibility of creating a bidirectional shunt, AFR was initially successfully tested in patients with severe, irreversible pulmonary arterial hypertension (PAH) [59, 60].

The Pilot Study to Assess Safety and Efficacy of a Novel Atrial Flow Regulator in Heart Failure Patients (PRELIEVE) was an open-label, prospective, non-randomized, first-in-man study investigating the feasibility up to 1-year follow-up of AFR implantation (8 mm or 10 mm diameters), in patients with HFrEF (n = 24) or HFpEF (n = 29) [61]. Among inclusion criteria, PCWP was ≥ 15 mmHg at rest or ≥ 25 mmHg during exercise. At 3 months, rest PCWP decreased by 5 mmHg (p = 0.0003) in the whole cohort. When analysed separately, the PCWP change was significant for HFpEF patients as compared with HFrEF patients. RAP remained unchanged after 3 months. Echocardiographic data at 12-month follow-up showed that LA/LV diameter and LVEF remained unchanged, with significant improvement of the E/E′ ratio, a parameter that reflects PCWP [62]. A mild significant dilatation of RV diameter was observed in the HFpEF cohort due to increased volume, although without deterioration of right heart function. The authors also observed an improvement of NYHA functional class, QoL, and 6MWD at 1 year. Shunt patency with unidirectional left–right shunting was proven in all patients with sufficient echocardiography readout at both 3 and 12 months [63]. After the procedure, aspirin (75–325 mg daily) indefinitely associated with a P2Y12 inhibitor or anticoagulant therapy (warfarin or a direct-acting oral anticoagulant) for 6 months has been recommended (empirically) [48].

Currently, several studies are ongoing to further test the safety and efficacy of the AFR (Table 2). The “Flow Regulation by Opening the Septum in Patients With Heart Failure Trial (FROST-HF)” trial (NCT03751748) will enrol 230 patients with LVEF > 45%. Furthermore, receiving CE approval, a large observational registry (Follow-up Study to Monitor the Efficacy and Safety of the Occlutech AFR in Heart Failure Patients [AFteR] Registry) (NCT04405583) will include patients undergoing AFR implantation for the monitoring up to 3 years after the procedure.

Novel perspectives

Devices capable of creating interatrial shunts have proven to be feasible, safe, and effective. In a recent meta-analysis including 6 studies (5 single-arm open-label studies, 1 sham-controlled trial) and 226 patients with chronic HF, the pre-defined primary outcome of change in 6MWD from baseline to 12 months was improved by 28 m (95% confidence interval (CI) 10.9–45.3), without significant interaction between devices (p = 0.66) and LVEF subgroups (p = 0.21) [64].

All the aforementioned devices use the interatrial septum as the site of shunt placement. However, patients may develop a certain amount of right heart overload and enabling right-to-left shunting may lead to hypoxemia and systemic embolization. In addition, placement in the interatrial septum limits subsequent trans-septal punctures for percutaneous procedures [65].

In order to overcome these limitations, a novel approach has been proposed. The Transcatheter Atrial Shunt System (Edwards Lifesciences) was created to reduce PCWP, through the creation of a shunt from the LA to the coronary sinus (CS) (Fig. 2). It is a bare-nitinol implant with 4 arms and an internal shunting diameter of 7 mm. The device is deployed between the LA and the CS through a percutaneous atriotomy, a procedure involving CS cannulation from the right internal jugular vein, CS-to-LA puncture, and balloon dilation of the LA wall. Intraprocedural CS angiography, fluoroscopy, transoesophageal echocardiography, and hemodynamic assessment are performed to document appropriate device seating and LA-to-CS shunting. The shunt device is fully recapturable up until the point of final arm deployment, while deployed shunts can be closed using the commercially available Amplatzer Duct Occluder or Amplatzer Septal Occluder in the event of excessive shunting or evidence of RV compromise [65].

This device was initially tested in 11 patients with symptomatic HF (7 HFpEF, 4 HFrEF) despite maximally tolerated guideline-directed medical and device therapy and PWCP ≥ 15 mmHg with a gradient from PCWP to RAP > 5 mmHg. Among 8 patients undergoing successful implantation, a significant improvement in symptoms and haemodynamic parameters was observed [65, 66]. A prospective early feasibility clinical trial (ALt FLOW US) is ongoing to evaluate safety and efficacy of this device in a larger patient population (NCT03523416) (Table 2).

More recently, a radiofrequency ablation–based interatrial shunt (RAIAS) therapy with a novel non-implanted device was developed. This device, named the NoYA system (NoYA MedTech, Hangzhou, China), consisted of a self-expanded flowerlike nitinol stent fixed connected to the radiofrequency generator. The flowerlike stent was configured with a diameter adjustable from 4- to 12-mm waist in the middle, on which the electric poles were located. Under the radial force of the stent and the power of radiofrequency energy, an interatrial communication was made. After ablation, the device was removed from the body, leaving nothing but an artificial ASD [59]. This novel approach was evaluated first in 11 domestic pigs and then in 9 HFpEF patients and 1 with HFmrEF with NYHA functional classes II–IV [67]. RAIAS therapy was successfully administered to all patients, and an evidence interatrial left-to-right shunt flow was detected with a mean Qp:Qs of 1.3 ± 0.3. No major safety event was observed during the whole study. However, there was a progressive decreased in diameter of the defect from 5.0 mm post-procedure to 3.0 at 6 months, and two patients showed complete closure of the defect (1 of 3-month and 1 of 6-month follow-up), confirmed by transoesophageal echocardiography. Patients showed improvements of NYHA functional class (p = 0.003) at 6 months, a reduction of median NT-proBNP from 3533 to 1347 pg/mL (p = 0.028), and an improvement in 6MWD from 349 to 440 m (p = 0.008). In addition, echocardiographic parameters showed a mean decrease of 3.1 ± 3.9 mm in the LA diameter at 6 months compared with baseline (p = 0.042), reflecting the reduced overloading of the LA after RAIAS [67].

In conclusion, this first-in-man trial suggested that this new approach was a safe and feasible strategy for patients with HFpEF. Figure 3 summarizes the main criteria for the selection of patients that may benefit from interatrial shunt device implantation. However, future prospective randomized clinical trials are needed to clarify the efficacy and long-term safety.

Fig. 3
figure 3

A practical guide for the selection of patients to be implanted with devices creating a left-to-right atrial shunt. Single asterisk indicates that right ventricular dysfunction is defined as TAPSE < 14 mm, RV volume ≥ LV volume, and PASP > 60 mmHg for AFR implantation; TAPSE < 12 mm or RVFAC ≤ 25% for V-Wave Shunt; and PVC > 20 mmHg or RAD > 45 mm, or TAPSE < 14 mm for radiofrequency ablation–based. ASD, atrial septal defect; AVR, aortic valve replacement; CABG, coronary artery bypass surgery; ESC, European Society of Cardiology; HFpEF, heart failure with preserved ejection fraction; LA, left atrium; MI, myocardial infarction; PCI, percutaneous coronary intervention; PCWP, pulmonary capillary wedge pressure; RAP, right atrial pressure; TIA, transient ischaemic attack

Full size image

Two further devices are being evaluated in HFpEF patients: one is able to create a personalized atrial septostomy through the combined use of radiofrequency ablation and balloon dilation (CURB) (NCT04573166); the other, the Alleviant System, a therapeutic interatrial shunt without a permanent heart implant (NCT04583527, NCT04838353).

Conclusion

The management of HFpEF remains a challenge due to limited data about effectiveness of medical treatments. The use of devices capable of creating interatrial shunts, to reduce LAP, represents a promising therapeutic option. Early trials demonstrate feasibility, safety, and effectiveness in reducing PCWP, improving patients’ symptoms and QoL. Data on mortality and HF re-hospitalizations are still limited.