Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
SHUNT FOR REDISTRIBUTING ATRIAL BLOOD VOLUME
Document Type and Number:
WIPO Patent Application WO/2018/158747
Kind Code:
A1
Abstract:
A shunt for regulating blood pressure between a patient's left atrium and right atrium comprises an anchor comprising a neck region, first and second end regions, and a conduit affixed with the anchor that formed of a biocompatible material that is resistant to transmural and translation tissue ingrowth and that reduces a risk of paradoxical embolism.

Inventors:
EIGLER NEAL (US)
ROSEN LIOR (IL)
HAFELFINGER WERNER (US)
ROZENFELD EREZ (IL)
BEN-DAVID TAMIR (IL)
NAE NIR (IL)
YACOBY MENASHE (IL)
NITZAN YAACOV (IL)
Application Number:
PCT/IB2018/051355
Publication Date:
September 07, 2018
Filing Date:
March 02, 2018
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
V WAVE LTD (IL)
International Classes:
A61B17/02; A61F2/24; A61B17/11; A61F2/01; A61F2/91
Domestic Patent References:
WO2013096965A12013-06-27
WO2016178171A12016-11-10
Foreign References:
US20130030521A12013-01-31
US20050165344A12005-07-28
US6120534A2000-09-19
US6468303B12002-10-22
US20050165344A12005-07-28
US9034034B22015-05-19
US6242762B12001-06-05
US20140350565A12014-11-27
Other References:
HASENFUSS, A TRANSCATHETER INTRACARDIAC SHUNT DEVICE FOR HEART FAILURE WITH PRESERVED EJECTION FRACTION (REDUCE LAP-HF): A MULTICENTRE, OPEN-LABEL, SINGLE-ARM, PHASE 1 TRIAL
KHOSITSTH, TRANSCATHETER AMPLATZER DEVICE CLOSURE OF ATRIAL SEPTAL DEFECT AND PATENT FORAMEN OVALE IN PATIENTS WITH PRESUMED PARADOXICAL EMBOLISM
SCHMITTO, CHRONIC HEART FAILURE INDUCED BY MULTIPLE SEQUENTIAL CORONARY MICROEMBOLIZATION IN SHEEP
Download PDF:
Claims:
WHAT IS CLAIMED IS:

1. A device for regulating blood volume distribution configured for placement in an interatrial septum between a patient's left atrium and patient's right atrium, the device comprising:

an anchor having a first region, a second region, and a neck region joining the first region to the second region and configured to engage the interatrial septum, the anchor having a length in a range of 10 to 15 mm and configured to transition from a contracted delivery state to an expanded deployed state in which the first region extends into the patient's left atrium, and the second region extends at least 5 mm into the patient's right atrium so that an outlet of the second region is located out of a natural circulation flow path of blood entering into the patient's right atrium from an inferior vena cava, whereby emboli entrained in flow from the inferior vena cava is not directed into the outlet of the second region; and

a conduit affixed to the anchor and having a lumen wall defining a lumen, the lumen wall resistant to transmural and translational tissue growth, the lumen having a smallest diameter in the neck region in a range of 5 mm to 6.5 mm, the conduit having a first end that extends from the neck region a first distance of at least 3 mm into the patient's left atrium and a second end that extends from the neck region a second distance of at least 3 mm into the patient's right atrium, thereby preventing pannus formation from narrowing the lumen in the neck region,

wherein the diameter of the lumen and the length of the anchor are selected to reduce a risk of paradoxical embolization.

2. The device of claim 1, wherein the second end is configured so that when implanted the neck region engages a fossa ovalis of the patient's interatrial septum and the second end protrudes a distance of between 5 mm to 15 mm into the patient's right atrium.

3. The device of claim 1, wherein at least the first region is flared.

4. The device of claim 1, wherein the conduit comprises a layer of biocompatible material disposed on the anchor and an exterior surface of the conduit is impermeable to transmural tissue growth and resistant to translational tissue growth.

5. The device of claim 1, wherein the anchor has a diameter in the expanded deployed state larger than a diameter of the conduit, such that when implanted the anchor fills a larger opening in the interatrial septum than is needed to accommodate the conduit.

6. The device of claim 1, wherein a length of the anchor is different than a length of the conduit.

7. The device of claim 1, wherein the anchor, in the expanded deployed state, forms a filter that prevents emboli from entering the second end of the conduit.

8. The device of claim 1, wherein the conduit defines more than one lumen.

9. A device for regulating blood volume distribution between a patient' s left atrium and right atrium, the device comprising:

an anchor having, in a deployed state, a first region configured to be disposed in the patient's left atrium, a second region configured to be disposed in a right atrium, and a neck region disposed between the first and second regions, the neck region configured to engage a fossa ovalis of an atrial septum;

a conduit affixed to the anchor and having a lumen that extends through the first region, the second region and the neck region, the conduit configured so that when implanted an end of the conduit protrudes into the right atrium a distance sufficient to place the end out of a natural circulation flow path in the right atrium, thereby reducing a risk that thrombus entrained in the natural circulation flow path will be directed into the lumen.

10. The device of claim 9, wherein the distance is greater than or equal to 5 mm.

11. The device of claim 9, wherein a wall of the conduit is resistant to transmural and translational tissue growth and the conduit extends into each of the first region and the second region at least 3 mm beyond a location of contact of the anchor to any cardiac structure.

12. The device of claim 9, wherein at least one of the first region and the second region is flared.

13. The device of claim 9, wherein the conduit encapsulates the anchor.

14. The device of claim 9, wherein the anchor has a diameter in the expanded deployed state larger than a diameter of the conduit, such that the anchor is configured to fill a larger opening in the interatrial septum than needed to accommodate the conduit.

15. The device of claim 9, wherein conduit has a length different than a length of the anchor.

16. The device of claim 9, wherein the anchor comprises, in the expanded deployed state, a filter that excludes emboli from entering the lumen.

17. The device of claim 9, wherein the anchor comprises a plurality of longitudinal struts, a subset of the longitudinal struts having eyelets configured to engage a delivery device.

18. The device of claim 17, wherein a subset of the longitudinal struts include radiopaque markers.

19. The device of claim 17, wherein the anchor further comprises a plurality of circumferential sinusoidal struts interconnecting the longitudinal struts, and the conduit encapsulates the anchor except for a portion defining cutouts adjacent to the end of the conduit.

Description:
SHUNT FOR REDISTRIBUTING ATRIAL BLOOD VOLUME

FIELD OF THE INVENTION

[0001] This application generally relates to percutaneously placed implants and methods for redistributing blood from one cardiac chamber to another to address pathologies such as heart failure (HF), myocardial infarction (MI) and pulmonary arterial hypertension (PAH).

BACKGROUND OF THE INVENTION

[0002] Heart failure is the physiological state in which cardiac output is insufficient to meet the needs of the body or to do so only at a higher filing pressure. There are many underlying causes of HF, including myocardial infarction, coronary artery disease, valvular disease, hypertension, and myocarditis. Chronic heart failure is associated with

neurohormonal activation and alterations in autonomic control. Although these

compensatory neurohormonal mechanisms provide valuable support for the heart under normal physiological circumstances, they also play a fundamental role in the development and subsequent progression of HF.

[0003] For example, one of the body's main compensatory mechanisms for reduced blood flow in HF is to increase the amount of salt and water retained by the kidneys.

Retaining salt and water, instead of excreting it via urine, increases the volume of blood in the bloodstream and helps to maintain blood pressure. However, the larger volumes of blood also cause the heart muscle, particularly the ventricles, to become enlarged. As the heart chambers become enlarged, the wall thickness decreases and the heart's contractions weaken, causing a downward spiral in cardiac function. Another compensatory mechanism is vasoconstriction of the arterial system, which raises the blood pressure to help maintain adequate perfusion, thus increasing the load that the heart must pump against.

[0004] In low ejection fraction (EF) heart failure, high pressures in the heart result from the body's attempt to maintain the high pressures needed for adequate peripheral perfusion. However, as the heart weakens as a result of such high pressures, the disorder becomes exacerbated. Pressure in the left atrium may exceed 25 mmHg, at which stage, fluids from the blood flowing through the pulmonary circulatory system transudate or flow out of the pulmonary capillaries into the pulmonary interstitial spaces and into the alveoli, causing lung congestion and if untreated the syndrome of acute pulmonary edema and death.

[0005] Table 1 lists typical ranges of right atrial pressure (RAP), right ventricular pressure (RVP), left atrial pressure (LAP), left ventricular pressure (LVP), cardiac output (CO), and stroke volume (SV) for a normal heart and for a heart suffering from HF. In a normal heart beating at around 70 beats/minute, the stroke volume needed to maintain normal cardiac output is about 60 to 100 milliliters. When the preload, after-load, and contractility of the heart are normal, the pressures required to achieve normal cardiac output are listed in Table 1. In a heart suffering from HF, the hemodynamic parameters change (as shown in Table 1) to maintain peripheral perfusion.

Table 1

[0006] HF is generally classified as either systolic heart failure (SHF) or diastolic heart failure (DHF). In SHF, the pumping action of the heart is reduced or weakened. A common clinical measurement is the ejection fraction, which is a function of the blood ejected out of the left ventricle (stroke volume) divided by the maximum volume in the left ventricle at the end of diastole or relaxation phase. A normal ejection fraction is greater than 50%. Systolic heart failure generally causes a decreased ejection fraction of less than 40%. Such patients have heart failure with reduced ejection fraction (HFrEF). A patient with HFrEF may usually have a larger left ventricle because of a phenomenon called "cardiac remodeling" that occurs secondarily to the higher ventricular pressures.

[0007] In DHF, the heart generally contracts normally, with a normal ejection fraction, but is stiffer, or less compliant, than a healthy heart would be when relaxing and filling with blood. Such patients are said to have heart failure with preserved ejection fraction (HFpEF). This stiffness may impede blood from filling the heart and produce backup into the lungs, which may result in pulmonary venous hypertension and lung edema. HFpEF is more common in patients older than 75 years, especially in women with high blood pressure.

[0008] Both variants of HF have been treated using pharmacological approaches, which typically involve the use of vasodilators for reducing the workload of the heart by reducing systemic vascular resistance, as well as diuretics, which inhibit fluid accumulation and edema formation, and reduce cardiac filling pressure. No pharmacological therapies have been shown to improve morbidity or mortality in HFpEF whereas several classes of drugs have made an important impact on the management of patients with HFrEF, including renin- angiotensin antagonists, beta blockers, and mineralocorticoid antagonists. Nonetheless, in general, FIF remains a progressive disease and most patients have deteriorating cardiac function and symptoms over time. In the U.S., there are over 1 million hospitalizations annually for acutely worsening FIF and mortality is higher than for most forms of cancer.

[0009] In more severe cases of HFrEF, assist devices such as mechanical pumps are used to reduce the load on the heart by performing all or part of the pumping function normally done by the heart. Chronic left ventricular assist devices (LVAD), and cardiac

transplantation, often are used as measures of last resort. However, such assist devices typically are intended to improve the pumping capacity of the heart, to increase cardiac output to levels compatible with normal life, and to sustain the patient until a donor heart for transplantation becomes available. Such mechanical devices enable propulsion of significant volumes of blood (liters/min), but are limited by a need for a power supply, relatively large pumps, and pose a risk of hemolysis, thrombus formation, and infection. Temporary assist devices, intra-aortic balloons, and pacing devices have also been used.

[0010] Various devices have been developed using stents to modify blood pressure and flow within a given vessel, or between chambers of the heart. For example, U.S. Patent No. 6,120,534 to Ruiz is directed to an endoluminal stent for regulating the flow of fluids through a body vessel or organ, for example, for regulating blood flow through the pulmonary artery to treat congenital heart defects. The stent may include an expandable mesh having lobed or conical portions joined by a constricted region, which limits flow through the stent. The mesh may comprise longitudinal struts connected by transverse sinusoidal or serpentine connecting members. Ruiz is silent on the treatment of HF or the reduction of left atrial pressure. [0011] U.S. Patent No. 6,468,303 to Amplatz et al. describes a collapsible medical device and associated method for shunting selected organs and vessels. Amplatz describes that the device may be suitable to shunt a septal defect of a patient's heart, for example, by creating a shunt in the atrial septum of a neonate with hypoplastic left heart syndrome (HLHS). That patent also describes that increasing mixing of pulmonary and systemic venous blood improves oxygen saturation, and that the shunt may later be closed with an occluding device. Amplatz is silent on the treatment of HF or the reduction of left atrial pressure, as well as on means for regulating the rate of blood flow through the device.

[0012] Implantable interatrial shunt devices have been successfully used in patients with severe symptomatic heart failure. By diverting or shunting blood from the left atrium (LA) to the right atrium (RA), the pressure in the left atrium is lowered or prevented from elevating as high as it would otherwise (left atrial decompression). Such an accomplishment would be expected to prevent, relieve, or limit the symptoms, signs, and syndromes associated of pulmonary congestion. These include severe shortness of breath, pulmonary edema, hypoxia, the need for acute hospitalization, mechanical ventilation, and death.

[0013] Shunt flow is generally governed by the pressure gradient between the atria and the fluid mechanical properties of the shunt device. The latter are typically affected by the shunt's geometry and material composition. For example, the general flow properties of similar shunt designs have been shown to be related to the mean interatrial pressure gradient and the effective orifice diameter.

[0014] Percutaneous implantation of interatrial shunts generally requires transseptal catheterization immediately preceding shunt device insertion. The transseptal catheterization system is placed from an entrance site in the femoral vein, across the interatrial septum in the region of fossa ovalis (FO), which is the central and thinnest region of the interatrial septum. The FO in adults is typically 15-20 mm in its major axis dimension and <3 mm in thickness, but in certain circumstances may be up to 10 mm thick. LA chamber access may be achieved using a host of different techniques familiar to those skilled in the art, including but not limited to: needle puncture, stylet puncture, screw needle puncture, and radiofrequency ablation. The passageway between the two atria is dilated to facilitate passage of a shunt device having a desired orifice size. Dilation generally is accomplished by advancing a tapered sheath/dilator catheter system or inflation of an angioplasty type balloon across the FO. This is the same general location where a congenital secundum atrial septal defect (ASD) would be located.

[0015] U.S. Patent Publication No. 2005/0165344 to Dobak, III describes apparatus for treating heart failure that includes a tubular conduit having a emboli filter or valve, the device configured to be positioned in an opening in the atrial septum of the heart to allow flow from the left atrium into the right atrium. Dobak discloses that shunting of blood may reduce left atrial pressures, thereby preventing pulmonary edema and progressive left ventricular dysfunction, and reducing LVEDP. Dobak describes that the device may include deployable retention struts, such as metallic arms that exert a slight force on the atrial septum on both sides and pinch or clamp the device to the septum.

[0016] Two types of percutaneously implantable shunts have been described in the medical and patent literature. In short-term, small-size clinical trials, both types have been shown to be associated with improvements in symptoms, quality of life measurements, and exercise capacity. Both shunts also have observed and theoretical drawbacks, which may limit their effectiveness and use.

[0017] The first type of shunt is henceforth referred to as an orifice-plate mesh shunt. Orifice-plate mesh shunts comprise a metallic mesh that wraps around both sides of the septum with a hole in the center and anatomically mimics the location and geometrical characteristics of a small congenital secundum ASD. The shunt geometry generally resembles a thin plate with a hole in it. In most embodiments, the "plate" comprises both mesh material and atrial septal tissue encased by the mesh. One example of such devices, designed by Corvia Medical, Inc., Tewksbury MA, consists of a self-expanding nitinol mesh that forms a pair of disc-like flanges with an open orifice in the center. The maximal diameter of the discs is 19.4 mm and the orifice diameter is 8 mm. Each disc flange has multiple truss-like legs that deploy into a preset configuration that wraps around the LA and RA sides of the interatrial septum and applies a clamping force to the tissue.

[0018] Another example of such a mesh type device, developed by Occlutech

International AB, Helsingborg, Sweden, resembles a dual-disc occluder used for closing congenital secundum ASDs, which additionally includes a short open barrel orifice in the center that connects the two discs. [0019] A major benefit of the foregoing orifice-plate mesh shunts over other shunt designs is simplicity of manufacture. Although relatively simple in theory and construction, orifice-plate mesh type shunts have several important drawbacks that are expected to reduce their overall potential for clinical safety and effectiveness.

[0020] A first drawback of orifice-plate devices is the susceptibility to narrow or close during the post-implantation healing period. For example, neoendocardial tissue ingrowth, referred to as pannus, grows from the underlining tissue to cover the mesh and narrow or partially occlude the shunt orifice. During the period following implantation, local trauma caused by crossing and dilating the FO, plus the chronic effects of continuous pressure applied by the mesh material on the septal tissue, provoke a localized healing response. This response entails activation of an inflammatory process, attracting lymphocytes and macrophages to the area of tissue injury. These inflammatory cells in turn release a variety of cytokines that signal fibroblasts and smooth-muscle cells from the wound margins to dedifferentiate, migrate, proliferate and encapsulate affected portions of the implanted device. The fibroblasts and smooth muscle cells then secrete extracellular matrix material composed of collagen and proteoglycans, which extracellular matrix forms the bulk of the pannus. The duration of this healing phase in humans is typically up to 6-9 months, but may be longer if there is a chronic source for tissue injury such as device compression or erosion of adjacent tissue. Eventually this pannus is covered with neoendothelial cells, causing the pannus growth to stop or stabilize. In the long term, the collagen of the pannus remodels, but generally retains its space occupying properties. Such tissue ingrowth typically spreads over the surfaces of the implant's struts, mesh, or discs, and may substantially narrow the orifice lumen or even entirely occlude the shunt. Narrowing or occlusion of the shunt prevents LA decompression and limits any positive effect for the patient.

[0021] The degree of luminal narrowing may be quite variable between patients due to differences in the severity of local injury— the more injury, the more exaggerated the pannus formation. Also, variability results from differences in host wound healing responses. For example, the amount and character of extracellular matrix may affect the duration of healing and amount of material deposited. Thus, for an orifice-plate mesh shunt, the eventual orifice lumen size will be highly variable. These processes will be familiar to one skill in the art as it is generally analogous to the type of late lumen loss that occurs in arteries when bare metal stents are used to treat atherosclerotic stenosis. [0022] In a trial described in the publication, "A Transcatheter Intracardiac Shunt Device for Heart Failure with Preserved Ejection Fraction (REDUCE LAP-HF): A Multicentre, Open-label, Single-arm, Phase 1 Trial" by Hasenfuss, et al., 14 of 64 patients implanted with an orifice-plate mesh shunt device had no demonstrable flow across the shunt on

transthoracic echocardiographic Doppler imaging at 6 months after implantation. It has not reported whether the shunts were occluded or if the imaging study was simply too technically difficult to tell for certain. Although additional interventional cardiology procedures may be undertaken to restore lost luminal patency, such procedures may pose unacceptable risks, including death and stroke from embolization of the orifice-clogging material.

[0023] A second drawback of an orifice-plate mesh shunt is the potential for paradoxical embolization. Paradoxical embolization refers to thromboembolism originating in the venous vasculature (venous thromboembolism or VTE), such that an embolus traverses right-to-left through a cardiac shunt into the systemic arterial circulation. The most severe complication of paradoxical embolization occurs when an embolus lodges in the cerebral circulation with resulting cerebral infarction (stroke). Similarly, if a paradoxical embolus enters the coronary arterial circulation, myocardial infarction (MI) may ensue. Other embolic syndromes result from embolization to the mesenteric, renal, and peripheral arteries supplying the limbs.

These may cause respectively, ischemic bowel syndrome, hematuria with worsening renal function, and gangrene requiring amputation.

[0024] Most frequently, VTE in adults is the consequence of in situ thrombosis in the deep veins (deep venous thrombosis or DVT) of the lower extremities or pelvis. For the most part, clinically relevant venous emboli develop in the popliteal veins or more proximally in larger veins of the upper thigh or pelvis. In patients with DVT involving the popliteal vein, the venous diameter averaged 11.4 mm (range from 6.2 mm to 20.1 mm). Often, emboli are described as having the form of a cast of the vein's lumen with a width equal to the diameter of the vein of origin. These thrombi also tend to be elongated, corresponding to the length of the occluded venous segment.

[0025] The risk factors associated with thromboembolic disease include a variety of anatomic, physiological, rheological variables and disease states. Heart failure is a well- recognized risk factor for DVT and VTE, especially in patients with reduced left ventricular systolic function. About 3% of deaths in heart failure patients are due to VTE, usually associated with pulmonary embolism. Patients with transvenous endocardial pacing leads and an intracardiac shunt have a 3-fold increased risk of systemic thromboembolism, suggesting that paradoxical embolism is a contributing underlying cause. There is evidence that the risk of paradoxical embolism is directly related to the orifice size of naturally occurring atrial level shunts such as ASD and patent foramen ovale (PFO). The presence of an atrial septal aneurysm is an additional risk factor. For example, as described in the publication "Transcatheter Amplatzer Device Closure of Atrial Septal Defect and Patent Foramen Ovale in Patients with Presumed Paradoxical Embolism" by Khositsth, et al., in a series of 103 adult patients with paradoxical embolization, an ASD was present in 12%, whereas PFO was present in 81%. In patients with clinically significant ASD referred for closure, the incidence of paradoxical embolus has been reported to be up to 14%.

[0026] It has been asserted that in order for VTE to enter the systemic circulation, the prevailing LA to RA pressure gradient must be temporarily reduced, eliminated or reversed so that blood will either flow slowly across the shunt, cease to flow across the shunt or flow retrograde across the shunt. Echo/Doppler imaging studies often reveal some amount of shunting in both directions (bi-directional shunting) in patients with congenital ASD, even when LA to RA flow predominates. Bidirectional shunting may be best demonstrated when a subject performs a Valsalva maneuver (straining caused by exhalation against a closed glottis). Valsalva increases intrathoracic pressure, which causes the RA and LA pressures to equalize after several seconds and then for the RA pressure to transiently exceed LA pressure on exhalation. Intermittent bidirectional flow also may be observed at rest when the interatrial pressure gradient is low, or intermittently during the cardiac cycle when LA contraction is delayed compared to RA contraction (interatrial conduction delay). This is seen especially when the atria are enlarged or diseased, such as in heart failure. In this setting, interatrial electrical conduction delay results in retardation of LA contraction.

Bidirectional shunting can also be seen transiently during inspiration, when venous return to the RA is increased, during coughing, with abdominal compression, during forced exhalation, or in the presence of severe tricuspid valve regurgitation. Chronically increased pulmonary arterial pressure, as seen in severe pulmonary hypertension, whether primary or secondary to chronic lung disease, recurrent pulmonary embolism, or due to chronic right ventricular volume overload, has been associated with chronic and more severe RA to LA shunting.

[0027] Additional phenomena associated with RA to LA shunting are diminished pulmonary blood flow and decreased arterial oxygen saturation due to systemic venous admixing. When these findings are also transient, they are generally well tolerated. Thus, prevention of significant or larger paradoxical emboli is the primary concern rather than preventing reverse shunting per se. As the consequences of paradoxical embolization can be catastrophic, it is desirable particularly in high-risk patients, that an implantable shunt be equipped with mechanism(s) that limit or minimize the chances of paradoxical embolization or minimize the chances of transporting large emboli.

[0028] From these data, it seems reasonable to expect that an orifice-plate mesh shunt, by virtue of its anatomic similarities with congenital secundum ASD, would have a similar risk of paradoxical embolization. It is easily understandable that a thin plate-orifice mesh type of artificial shunt might be more susceptible to paradoxical embolization than other types of shunts with longer orifice geometries, e.g., a nozzle. For any given quanta of RA volume (blood or thrombus), the statistical likelihood of traversing retrograde across the shunt and into the LA would be expected to be a complex function of the duration of pressure gradient reversal, flow patterns in the RA, shunt tunnel distance affecting the length of the flow velocity streamlines, and flow velocity and orifice or lumen size.

[0029] A third drawback of an orifice-plate mesh shunt is that percutaneous removal from the shunt body is only possible at the time of implantation. Should the shunt become a nidus for infection, develop fatigue or corrosion fractures of its metallic framework, or erode or otherwise impinge on other vital cardiac structures, it cannot be removed by percutaneous retrieval/removal techniques. This is because the shunt, with its large "footprint" on the interatrial septum, is encased in pannus tissue. Attempts at percutaneous removal may result in tearing of the septum, pericardial tamponade, and device embolization into the systemic circulation, resulting in death or the need for emergency surgery. Safe removal would require performing open heart surgery. This entails that the heart be bypassed using an

extracorporeal membrane pump oxygenator (cardiopulmonary bypass), so the heart can be opened, the shunt removed, and the septum repaired. Performing such surgical procedures in patients with already established severe heart failure, including its frequently associated co- morbid conditions such as peripheral, cerebrovascular, and coronary artery disease, renal dysfunction, and diabetes, would be expected to have substantial risks for mortality or severe morbidity.

[0030] A fourth drawback of an orifice-plate mesh type of shunt is that its geometry renders it relatively inefficient in supporting high flow. For any given pressure gradient across the shunt, the geometry of an orifice plate requires a larger orifice because it has a reduced effective orifice size compared with other geometries, such as a venturi-shaped lumen, or a conical shaped nozzle. This is because with an office-plate, there are more energy losses associated with eddy currents at the edges of the plate. Orifice-plate geometries may be categorized as having a relatively low discharge coefficient, which is a dimensionless fluid-mechanical parameter that relates flow to actual orifice size. For practical purposes, the discharge coefficient is the ratio of areas of the exiting jet vena contracta, which is the narrowest portion of the jet, compared to the shunt orifice. For example, the coefficient of discharge for orifice plates placed in pipes tends to be

approximately 0.6, but rarely exceeds 0.65. The discharge coefficient is affected by the orifice and chamber dimensions, the pressure gradient, and the viscosity of blood and/or the Reynolds number of the specific flow condition. This differs from the more efficient passage of flow through a classic venturi type of narrowing, where the discharge coefficient usually exceeds 0.9 and is typically in the range of 0.94 to 0.98. The result is that, in comparison with more efficient shunt lumen geometries, an orifice-plate mesh shunt requires a larger orifice diameter to accommodate the same amount of flow for any given pressure differential across the shunt.

[0031] A fifth drawback of an orifice-plate mesh shunt is that it occupies a large area or footprint on the interatrial septum. The flanges of the device that anchor the shunt typically occupy the entire area of the fossa ovalis and may overlap adjoining muscular portions of the interatrial septum. These flanges exert persistent pressure on the septum, causing injuring and stimulating an exaggerated healing response as described above. Also, the rigidity of the mesh may interfere with the normal motion of the muscular septum. The flanges additionally may impinge on adjacent cardiac structures such as the roof of the left atrium, the ostia of the pulmonary veins, and the aorta root and sinuses of Valsalva, where due to chronic rubbing contact or sandwiching compressive forces, they may erode into these vital structures. Such erosion has been associated with severe complications including cardiac tamponade and death. For example, the similarly sized Amplatzer ASD disc occlusion device described above has been occasionally associated with erosion into adjoining tissues with resulting catastrophic outcomes.

[0032] Additional issues associated with placing relatively large devices with complex three-dimensional geometries are potential difficulties in positioning the shunts accurately in the FO, obtaining sufficient tissue anchoring to prevent migration, and having devices conform to irregularities of the cardiac anatomy. For example, in a report of attempted implantation of orifice-plate mesh shunts in 66 patients in the above cited publication authored by Hasenfuss, et al., device placement was not possible in two patients. And of the 64 implanted patients, the device had to be removed and re-implanted in another 3 patients due to misplacement, migration, or embolization of the first attempted implant.

[0033] Finally, the large footprint on the atrial septum may hinder or render impossible performing other interventional procedures that require transseptal access. The large flange diameter and small mesh pore sizes generally make catheter crossing of the atrial septum possible only through the central shunt orifice itself. Transseptal procedures using small diameter catheters, such as atrial fibrillation RF ablation, may be conducted through the orifice-plate lumen only if it is not obstructed by pannus and the orifice location permits entry into all four pulmonary veins. Other structural heart disease procedures that have large diameter delivery systems and/or require crossing the FO in specific locations may encounter difficulties or simply not be possible. These procedures include left atrial appendage occlusion, mitral valve edge-to-edge ("MitraClip") repair, and transvascular mitral valve replacement. For example, placing of a MitraClip optimally requires crossing the FO in its superior-posterior quadrant. The guiding catheter has a tip inner diameter of 7.7 mm (23 Fr). Similar transseptal access is needed to perform reconstructive mitral annuloplasty with the Cardioband device marketed by Valtech. In these cases, the only alternatives might be higher risk therapeutic approaches involving trans-left ventricular apical access or open heart surgery.

[0034] The second type of shunt is referred to as a valved unidirectional shunt. These shunts attempt to overcome some of the drawbacks of orifice-plate devices. For example, valved unidirectional shunts have embodiments containing a one-way or check-valve to limit reverse shunting and paradoxical embolization. Some of the valve configurations are designed to open when the LA-RA pressure gradient exceeds a predefined threshold. Other valve configurations close only when the RA pressure exceeds LA pressure (reversed gradient).

[0035] U.S. Patent No. 9,034,034 to Nitzan, the entire contents of which are incorporated by reference herein, solves many of the drawbacks of plate-like orifice mesh shunts describe above. An embodiment of the Nitzan-type shunt comprises an hourglass or diabolo outer shape, having a small FO footprint minimizing septal injury, which is expected to minimize pannus growth and obliteration of the shunt lumen. Its one-way valve also is designed to reduce the potential for reverse shunting and paradoxical embolization. The relatively small footprint of the shunt in contact with the septum and encapsulated collapsible nitinol frame is designed to facilitate percutaneous extraction from the septum and retrieval from the body using a standard goose-neck snare and large-bore sheath, thus making the device more easily retrieved. The venturi tube-like inner lumen of the diabolo shape provides better bulk flow characteristics, permitting a smaller orifice for the same amount of flow compared to orifice plate shunts. And finally, the small footprint on the FO and the hourglass shape are designed to facilitate accurate placement and retention during implantation. This geometry also minimizes interference with normal motion of the interatrial septum, and the small footprint provides space surrounding the shunt for other potential interventional procedures that require transseptal catheterization.

[0036] One embodiment of the Nitzan design, manufactured by V-Wave, Ltd (Caesarea, Israel), designed to support unidirectional left-to-right flow, comprises a self-expanding frame constructed from a laser-cut nitinol tube. The frame includes five sinusoidal circumferential struts interconnected by six longitudinal bars. The frame is heat-set so that it has an asymmetrical hourglass shape or a diabolo shape. The shunt is deployed so that the neck (5.3 mm outer diameter) is placed across the FO and secured in place by its external surface geometry. The shunt's widest portion has a conical shape with an approximately 14.3 mm outer diameter at the LA end of the shunt, which serves as an "entry" port on the distal end of the entry funnel. The entry funnel is deployed in the left atrium, and registers the neck of the shunt to the region of the FO. A second, slightly narrower bell-shaped portion forms the exit portion of the shunt, which expands to a maximum outer diameter of 11.4 mm at the RA end of the shunt. The shunt does not require flanges, discs, or tissue anchors to secure it in place. Septal retention is achieved without applying persistent pressure, tension or rubbing contact on the tissue adjoining the device neck.

[0037] The V-Wave shunt has a single inner lumen where flow is entrained into the entry funnel in the LA and passes through the constricted neck having a 5.1 mm inner diameter, which resembles a venturi-type orifice, and then exits through a bioprosthetic valve positioned near the RA end of the shunt. The entry funnel and the central neck region are encapsulated with expanded polytetrafluoroethylene ("ePTFE") to form a skirt or cover over the frame. The skirt is designed to facilitate laminar flow and limit pannus ingrowth during device healing. The exit bell-shaped portion contains three, glutaraldehyde-fixed, porcine pericardial leaflets sutured to the frame at the right atrial extent of the ePTFE encapsulation. The leaflets are designed to create a smooth exit channel and remain in the open position, closing only when the RA pressure exceeds LA pressure by 1-2 mmHg, thus preventing reverse right-to-left shunting.

[0038] For deployment, the V-Wave shunt is compressed in a loading tube where it is attached to a triple-latch cable delivery catheter. The loading tube is inserted into a 14F delivery sheath that has been previously placed after a transseptal catheterization from the right femoral vein across the FO. The shunt then is advanced through the sheath until the entry funnel has been deployed in the LA. The entire system is withdrawn as a unit until the LA funnel is in contact with the left side of the FO. The delivery catheter latches are unhooked from the shunt, the delivery catheter withdrawn so the right atrial side of the shunt is held only by its radial force against the delivery sheath. Then the delivery sheath is withdrawn, thereby deploying the exit bell-shaped portion of the shunt on the RA side of the FO. Device placement may be guided and confirmed by fluoroscopy and echocardiography, e.g., intracardiac echo or transesophageal echo.

[0039] Pre-clinical testing on the V-Wave shunt was performed in an established juvenile ovine (sheep) model that created an ischemic cardiomyopathy form of heart failure. The sheep were pre-treated with sequential coronary artery microembolization as described in the publication, "Chronic Heart Failure Induced by Multiple Sequential Coronary

Microembolization in Sheep" by Schmitto et al. After several weeks, the sheep manifested evidence of severe left ventricular systolic dysfunction and develop elevated LV, LA, and pulmonary artery pressures. In a 12-week survival study, this V-Wave shunt was associated with significant improvements in LA pressure and left ventricular ejection fraction. All manifestations of worsening heart failure were improved and in some cases reversed with interatrial shunting. Concurrent control animals with established heart failure, but were not implanted with the V-Wave shunt, demonstrated progressive worsening of LV ejection fraction, and intracardiac/pulmonary pressure during 3-month follow-up. The physiological improvements in shunted animals were substantial even though the shunt volume was assessed to be small. The pulmonary blood flow/systemic blood flow ratio (Qp/Qs) was between 1.1 to 1.2, as measured by oximetry, which is consistent with a very small shunt. Naturally occurring ASDs, with a Qp/Qs less than 1.5, are generally left untreated as they are well tolerated for decades by the compliant right heart and pulmonary vasculature, without evidence of worsening right ventricular failure despite mild chronic volume overload. This was confirmed in the sheep model where RA and pulmonary artery pressures decreased to baseline levels with shunting, but progressively worsened in the control animals.

[0040] A total of 38 patients were implanted with the V-Wave hourglass-shaped shunt having valve leaflets in two similar feasibility studies. The baseline characteristics of the combined study populations are summarized in Table 1 below.

Table 1. Baseline characteristics of 38 patients implanted with valved hourglass-shaped shunt device

NYHA=New York Heart Association heart failure classification;

DM=diabetes mellitus; HTN=hypertension;

AFIB=atrial fibrillation;

ACEi-ARB=receiving angiotensin converting enzyme inhibitor or angiotensin receptor blocker;

BB=receiving beta blocker;

MRA=receiving mineralocorticoid antagonist;

DIUR=receiving loop diuretic;

CRT-D=implanted with combination cardiac resynchronization therapy pacemaker with ICD; ICD=implantable cardioverter/defibrillator;

CRT-P=implanted with cardiac resynchronization therapy pacemaker without combination ICD;

NT-proBNP=N-terminal pro b-type natriuretic peptide;

eGFR=estimated glomerular filtration rate; 6MWT=6 minute walk test distance;

PCWP=pulmonary capillary wedge pressure;

RAP=right atrial pressure;

PAP=pulmonary artery pressure;

CI=cardiac index;

PVR=pulmonary vascular resistance;

LVEF=left ventricular ejection fraction;

HFrEF=heart failure with reduced ejection fraction;

HFpEF=heart failure with preserved ejection fraction.

These parameters and abbreviations are well known to one skilled in the art.

[0041] All patients had New York Heart Association (NYHA) Class III or ambulatory Class IV heart failure symptoms at the time of study enrollment. Patients with either reduced or preserved left ventricular ejection fraction were included. There was a high frequency of co-morbidities known to be associated with a poorer prognosis including coronary artery disease, diabetes mellitus, atrial fibrillation, and chronic kidney dysfunction. All patients received appropriate guideline-driven medical and device therapies prior to study enrollment. Patients had evidence of elevated levels of natriuretic peptides, reduced exercise capacity, elevated intra-cardiac and pulmonary artery pressures, increased pulmonary vascular resistance, and reduced cardiac output. These factors have also been associated with poor outcomes. Patients were excluded if they had severe right ventricular dysfunction or severe pulmonary hypertension. [0042] Implantation of the V-Wave shunt was successful in all 38 patients and no device replacements were performed. Shunts remained implanted in the atrial septum without dislodgements, migrations or apparent interference with normal septal motion on

fluoroscopic or echocardiographic imaging. No shunts have required removal or replacement for infection or strut fracture. Follow-up imaging studies show that there are adjacent locations on the FO, that are available and amenable for performing transseptal procedures to treat other cardiac conditions, including, for example, atrial fibrillation ablation, left atrial appendage occlusion, and mitral valve repair. The valve apparatus, when functioning normally, has been shown to effectively prevent reverse (right-to-left) shunting.

Echocardiographic contrast and Doppler studies during rest or Valsalva maneuver show that there is no reverse shunting in the early months after human implantation. Furthermore, no thromboembolic clinical events, including paradoxical embolization, have been observed during the first year of follow-up.

[0043] Shunt patency is defined as LA to RA flow through the shunt as observed during transesophageal echo/Doppler study. At 3-months after implantation of the V-Wave shunts, patency was confirmed in all patients. The pulmonary to systemic flow ratio (Qp/Qs), as measured by echocardiography, increased from 1.04±0.22 at baseline to 1.18±0.16 shortly after implantation (p<0.03). In the subgroup of 30 patients with HFrEF presented by Dr. William Abraham, MD at TCT 2016 in Washington DC, there were statistically significant (p<0.05) improvements in clinician-assessed symptoms, patient assessed quality-of-life scores, and exercise capacity as measured by a 6-minute hall walk testing at 3, 6, and 12 months following implantation. There was no deterioration in natriuretic hormone levels, echocardiographic, or hemodynamic parameters. Most importantly, the annualized (Poisson) heart failure hospitalization rate with shunting (0.17 heart failure hospitalization per patient year), was substantially reduced in comparison to a well matched historical control groups (CHAMPION trial Control and Treatment groups, 0.90 and 0.67 heart failure hospitalization per patient year, respectively). These data provide adequate proof-of-concept that interatrial shunting is of benefit to patients with severe symptomatic heart failure. Moreover, these data strongly support moving forward with larger-scale clinical trials including randomized clinical trials. [0044] Notwithstanding the initial success observed in the foregoing trial, device occlusion, e.g., shunts having undetectable LA to RA flow, was observed in some valved interatrial shunt devices after long-term implantation, e.g., by 1 year. Further, shunts may develop bidirectional shunting that was not present early on. Bidirectional shunting is indicative of an incompetent valve, e.g., a valve where one or more leaflets do not fully coapt during closure, resulting in an open channel for reversed flow, and depending on the severity of the incompetence, may create a potential path for paradoxical embolus to traverse from the RA to LA.

[0045] To assess the effective orifice size of valved shunts over time, transesophageal echo/Doppler measurements of the diameter of the vena contracta, measured on the left-to- right flow jets through the shunt, were found to be consistent with progressive shunt narrowing. The vena contracta diameter monotonically decreased with time after

implantation from 4.0±1.1 mm just after implantation, to 3.6±1.0 mm at 3 months, and 2.7±1.4 mm at 6-12 months (p<0.01). This equates, on average, to shunts losing more than half of their orifice area by 12 months. Moreover, some of the left-to-right jets appeared to be exiting the shunt at an angle substantially different from the long axis of the shunt body. This skewing of the jet is consistent with material inside the shunt such as a valve leaflet with impaired mobility, which diverts the direction of the jet. This observation gives rise to concern about a decrease in the clinical effectiveness of the shunts over time

[0046] Clinical effectiveness also may be measured by the rate of hospitalization for worsening heart failure. In the 38 patients, during the first 6 months after implantation of the V-Wave shunt, the hospitalization rate was 0.16 per patient year, which increased to 0.40 per patient year between months 6-12. These data suggest there may be a loss of shunting benefit consistent with the time course associated with shunt narrowing or occlusion.

[0047] There are several possible mechanisms working alone or in combination that could explain these observations.

[0048] The least likely cause of shunt occlusion is collapse of the shunt due to external forces applied by the septum. For example, it is possible that contraction of pannus tissue formed during the later stages of healing (remodeling) could result in extrinsic compression of the shunt. However, there is no evidence to support this scenario based on multiple observations of frame geometry seen during pre-clinical studies and during follow-up transesophageal echocardiography (TEE), CT, or fluoroscopic imaging in humans. In all cases, the observed shunt frame has not been observed to be extrinsically compressed or in any other way narrowed, deformed, or fractured.

[0049] Another possible mechanism is in situ thrombosis of the shunt. However, all patients were treated with monitored anticoagulation for the first three months, or indefinitely if there were other indications for chronic anticoagulation, which was most commonly required in patients with a history of atrial fibrillation. Subjects were also treated

simultaneously with low-dose aspirin, which was continued indefinitely. Having experience with prosthetic cardiac valves as a predicate, valve thrombosis would have been expected to be seen earlier, typically within 30-45 days after implantation, especially in patients with a history of subtherapeutic anticoagulation therapy.

[0050] In the 38 patients implanted with the V-Wave valved hourglass-shaped shunt described above, no thrombi were detected on 121 consecutive post-implantation

echocardiograms. These studies systematically looked for intracardiac or device thrombus by an independent Echocardiographic Core Laboratory at time points including one day after implantation, and at 1, 3, 6, and 12 months after implantation. None of the patients presented with stroke or other clinical manifestations of thromboembolic events. Of 9 patients with suspected shunt occlusion or incompetent valves, most were taking therapeutic doses of anticoagulants (warfarin or New Oral Anticoagulant agents) at the time the shunt anomaly was discovered. Another reason that thrombosis is unlikely is the observation of progressive vena contracta narrowing over a time course of 6 months or more. Thrombosis would be expected to result in sudden lumen loss, and not progress slowly over a period of months.

[0051] A third potential cause of occlusion is neoendocardial tissue overgrowth or pannus formation that narrows the lumen at the neck of the hourglass-shaped shunt. Applicants' earlier ovine studies suggest otherwise. Specifically, the shunt lumen surface at the neck of the hourglass contained only microscopic amounts of cellular material. On gross

pathological examination, there was no visible loss of the lumen area in neck region. A human shunt specimen has been examined in an explanted heart from a patient that underwent cardiac transplantation 2.5 years after shunt implantation. The ePTFE surfaces of the shunt including the lumen at the neck contained no pannus formation or narrowing of any kind. [0052] In another example, a left atrial pressure sensor implanted across the FO by transseptal catheterization and used for guiding the medical therapeutic dosing in

symptomatic patients with severe heart failure was observed to experience pannus formation. In the original embodiment of the sensor, the sensing diaphragm, located at the distal end of the sensor module body, protruded into the left atrium by 1-mm beyond its three anchoring legs that rested on the left atrial side of the septum. In a later, improved geometry version, the legs were placed more proximal on the sensor module body so that sensing diaphragm protruded into the LA by an additional 1.5 mm.

[0053] In a comparative inter-species pathology study, neoendocardial tissue (pannus) formation was observed over the sensing diaphragm in 20 of 31 original sensors compared with only 3 of 40 specimens with the improved geometry sensor. Of the 20 original sensors with tissue coverage, 7 had demonstrable artifacts in the LA pressure waveform. In each case with artifacts, pannus formation over the sensing diaphragm had a thickness >0.3 mm. These data indicate that when tissue coverage exceeds this thickness, the tissue interferes with fluid pressure measurement. None of the improved sensors had waveform artifacts or tissue thickness >0.3 mm.

[0054] In addition to producing waveform artifacts, the time course of tissue

encapsulation of the sensing diaphragm could be estimated by assessing LA pressure waveforms for baseline drift with or without the development of artifacts. It was

hypothesized that as neoendocardial tissue grows over the sensing diaphragm, measured LA pressure increased due to a drifting baseline caused by tension applied from the tissue capsule covering the diaphragm through its contiguous connection with the atrial wall. This healing phenomenon may be initiated as early as several weeks' post implant in animals and starts around 3-4 months in humans. Using the timing of drift to indicate tissue coverage in humans, it was shown that in a group of 46 heart failure patients with the original sensor design geometry, about 25% developed the characteristic drift pattern associated with tissue coverage of the sensing diaphragm during the first year after implantation. Of 41 similar patients implanted with the improved geometry sensor, none developed drift.

[0055] Pannus formation on devices that traverse the interatrial septum has been observed to start at the portions of the device in contact with the septum in the region of local tissue injury. Tissue growth progresses contiguously, extending translationally along the external surfaces of the device that protrude into each atrial chamber. This pannus growth thins as a function of distance from the sites of cardiac contact until it becomes essentially a monolayer of neoendothelial cells. The process naturally stops after about 6-12 months in humans. Thereafter, the remaining tissue may remodel but active growth of pannus is completed. From these data, Applicants observed that tissue coverage typically grows a distance of about 3 mm from its starting place on the septal wall before stopping or becoming thin enough so as not to impede device function.

[0056] Thus, for pannus to cause narrowing of the lumen at the shunt neck, it would have to extend contiguously from the site of injury on the septum for some distance to reach the neck. Applicants have determined that translational tissue growth over a distance of 3 or more millimeters becomes much less likely.

[0057] Pannus formation affecting the valve leaflets is the most likely stand-alone mechanism that explains all of the untoward observations seen in human subjects implanted with V-Wave shunts, including progressive shunt narrowing, incompetence of the valve with bidirectional flow, and eventual loss of shunt flow with associated loss of clinical efficacy.

[0058] Tissue overgrowth affecting the valve leaflets bases and commissures was the predominant histopathological finding in the ovine pre-clinical study described above. Gross pathological examination of shunts implanted for 3 months showed pannus infiltration extending from the adjacent FO into the valve leaflet bases with thickening of the leaflet bodies in 5 out of 6 shunts. In 4 shunts, there was fusion of at least 2 of the 3 valve commissures where the leaflet edges were sutured to the shunt frame. Fusion of all 3 commissures was observed in 3 shunts. One case showed severe narrowing at the

commissures with a luminal area of 4 mm 2 or a 75% area stenosis in comparison to the normal 19.6 mm 2 lumen at the device neck. The leaflets were described as semi-pliable or stiffened in 4 out of 6 shunts. In two of the devices, commissural fusion and leaflet thickening were so pronounced that complete leaflet coaptation could not likely occur during valve closure. In none of these cases has pannus formation been seen to narrow the shunt neck.

[0059] On examination of microscopic sections, pannus thickness tends to be greater on the side of the leaflets facing the atrial septum where the ePTFE/leaflet junction was infiltrated with pannus that was contiguous with the adjoining atrial tissue. Pannus extended from the atrial septum on and around the right atrial edge of the ePTFE skirt and into the base and commissures of the valve leaflets. At 3 months, the pericardial leaflets showed varying degrees of pannus coverage ranging from mild to marked. In general, pannus is thickest at the leaflet bases and commissures, and tapers toward the free edges. In 2 sheep, the pannus on the leaflets measured 2 to 3 times the original thickness of the leaflets.

[0060] The pannus was generally well healed or organized by 3 months. It was composed of collagen and proteoglycan matrix surrounding smooth muscle cells, fibroblasts and rare focal areas of inflammation with lymphocytes, macrophages, and occasional multinucleated (foreign body type) giant cells. The pannus tissue was mostly covered with neoendothelium consistent with near complete healing. No leaflet calcification or thrombi were observed.

[0061] Although animal models of cardiovascular devices are limited in their ability to represent human tissue healing responses, the major differences are characteristically limited to the temporal duration of the response. For example, in a comparative pathology study described in the publication, "Comparative Pathology of an Implantable Left Atrial Pressure Sensor" by Roberts, et al., of a percutaneously implantable titanium/nitinol-enclosed LA pressure sensor, implanted on the interatrial septum, it was found that sheep at 1.5 to 8 months and canines implanted for 1 to 25 months, closely approximated the pathological findings seen in humans implants of 3 to 56 months duration. Histology had a similar appearance in humans and animals, and confirmed that the tissue covering the device was composed of a neoendocardium lined with a neoendothelium. The appearance of the neoendocardial tissue covering the sensor described above was similar to what is observed with ASD closure devices.

[0062] This mechanism of pannus formation preferentially affecting the bioprosthetic valve material compared to the ePTFE encapsulated portions of the shunt was observed in the human explanted specimen referred to earlier. After 2.5 year of implantation in heart, the 3 pericardial leaflets were severely thickened, immobile, infiltrated at their bases and commissures with pannus resulting in valvular stenosis with a reduction in outflow area of 52% relative to the non-obstructed shunt neck. Although this shunt was patent, it would have been incompetent, allowing bidirectional flow, and would have shunted less than half of the flow expected for any given pressure gradient. [0063] To further evaluate the tendency of this bioprosthetic valve to become infiltrated with pannus, valved and valveless designs of the V-Wave shunt were implanted by applicants in a non-diseased juvenile ovine (n=9) model. Specifically, this study was designed to highlight the resistance of a valveless, ePTFE encapsulated shunt (n=6) to pannus formation, narrowing and occlusion, relative to the legacy valved version previously used in humans (n=3), by creating a highly proliferative model expected in healthy juvenile sheep where the left-to-right interatrial pressure gradient was expected to be small. In the valveless design, the bioprosthetic valve material and its attaching polypropylene suture were removed and the ePTFE encapsulation was extended to cover the entire nitinol frame of the shunt except for the last 1.5 mm on the RA side where the shunt was coupled to its delivery system for deployment. The ePTFE used had an internodal distance of up to 30 microns. At 12 weeks the sheep where euthanized. The gross pathology findings showed that the 3 valved shunts were heavily infiltrated with pannus formation, extending from the septum into the regions containing the bioprosthetic leaflets. The leaflets were fused, immobile and highly stenotic leaving only a pinhole opening. The degree of pannus formation was much exaggeration versus prior experience in the ovine heart failure model. Thick pannus extended retrograde contiguously from the leaflet bases toward the hourglass neck of the shunts. The pannus growth from the original septal site of injury to the tips of the valve leaflets exceeded 3 mm in distance. Pannus appeared to grow through the valve commissures and through the suture holes attaching the porcine pericardial leaflets to the frame and the ePTFE skirt. Pannus formation was associated with mononuclear inflammatory cell infiltrates and multinucleated giant cells.

[0064] All 6 of the valveless, ePTFE encapsulated shunts were widely patent with only minimal pannus formation attaching the FO tissue to the external surface of the device.

Applicants observed no pannus growing translationally more than 3 mm along the external surface of the ePTFE from the septum. No visible pannus reached from the septum all the way into the lumen portion of either the left atrial entry cone or right atrial exit cone of the device. The lumina at the necks of all of the shunts were widely patent on gross and microscopic examination. There was no evidence of pannus formation permeating through the ePTFE encapsulation into the shunt lumen. [0065] From these combined observations, applicants have determined that length of translational pannus growth from the site of healing may be dependent on the type of biomaterial surface. In the case of the ePTFE encapsulated shunt, pannus formation severe enough to interfere with device function tends to translate a maximum of about 3 mm from the site of injury, whereas in the case of the bioprosthetic valve material tested, the amount of pannus formation and translational length of pannus tissue growth were exaggerated.

[0066] Also, from these data, it is reasonable to expect that the near complete shunt healing seen after 3 months in the juvenile ovine model will be predictive of the

histopathological findings in humans at 9-12 months. Moreover, these gross and microscopic observations, with their anticipated species-to-species conservation of findings, leads to the conclusion that the healing response in sheep is likely indicative of the mechanism causing shunt closure, valvular incompetence, and progressive stenosis in humans. Thus, there exists a need for a more durable shunt configuration that maintains luminal patency for extended periods of time.

[0067] It further would be desirable to provide a shunt for redistributing atrial blood volumes and reducing interatrial pressure imbalances that reduces the risk of paradoxical embolism caused by emboli transfer from the right to left atria.

[0068] It also would be desirable to provide an interatrial shunt configuration that reduces the risk of pannus formation after a prolonged period of implantation, where the degree of pannus formation and tissue ingrowth is not strongly dependent on the manner or location in which the shunt is implanted in the FO.

SUMMARY OF THE INVENTION

[0069] In view of the foregoing drawbacks of previously-known interatrial shunts, a shunt constructed in accordance with the principles of the present invention provides a more durable configuration that maintains luminal patency for extended periods of time. The inventive shunt further enables redistribution of interatrial blood volumes and pressure imbalances while reducing a risk of paradoxical embolism caused by emboli moving through the shunt from the right to left atria. [0070] Shunts constructed in accordance with the principles of the present invention also provide greater safety by enhancing long-term patency and reducing the risk of pannus formation after a prolonged period of implantation by reducing the impact of the manner in which the shunt is implanted in the interatrial septum.

[0071] In accordance with the principles of the present invention, shunts having an anchor and conduct are provided for redistributing atrial blood volumes, in which the shunt dimensions, contours and materials maintain long-term patency while reducing the risk of paradoxical embolism. It is hypothesized that such shunt designs will provide reductions in left atrial pressure, relieve pulmonary congestion, and lower pulmonary artery pressure, among other benefits. The inventive devices are configured for implantation through the atrial septum, and preferably through the fossa ovalis.

[0072] In particular, shunts designed in accordance with the principles of the present invention are designed to control LAP by transferring a small portion of the blood normally flowing from the left atrium to the left ventricle and diverting it instead to the right atrium, thereby modestly reducing LV end-diastolic filling volume. When the LAP is elevated, the LV operates on a steeper portion of its diastolic compliance curve. Accordingly, even a modest reduction in LV end-diastolic volume leads to a substantial fall in LV end-diastolic pressure. That reduction causes a commensurate reduction in upstream filling pressures including LAP, pulmonary venous pressure, and pulmonary artery pressure. The anticipated clinical result of these pressure reductions is expected to relieve or even prevent pulmonary congestive symptoms. At smaller interatrial gradients with less shunting, the effect on LV volume and filling pressures becomes progressively smaller until it is negligible. As interatrial shunting primarily affects LV filling and not afterload, beneficial effects on lowering end-diastolic pressure are expected, regardless of LV systolic function, for patients with heart failure associated with reduced ejection fraction (HFrEF) and patients with heart failure and preserved ejection fraction (HFpEF).

[0073] In accordance with one aspect of the present invention, the inventive devices include an anchor configured to be implanted in the interatrial septum, preferably the FO, and a conduit affixed to the anchor. The anchor includes a first region, a second region, and a neck region joining the first region to the second region and sized and shaped to engage the interatrial septum. The anchor has a length in a range of 10 to 15 mm and may transition from a contracted delivery state to an expanded deployed state in which the first region extends into the patient's left atrium, and the second region extends at least 5 mm into the patient's right atrium so that an outlet of the second region is located out of a natural circulation flow path of blood entering into the patient's right atrium from an inferior vena cava, whereby emboli entrained in flow from the inferior vena cava is not directed into the outlet of the second region.

[0074] The conduit includes a luminal wall defining a lumen, such that the luminal wall comprises a biocompatible material that is resistant to transmural tissue growth, and that limits translational tissue growth to 3 mm or less from the site of contact to the nearest cardiac structure. For example, a first end of the conduit extends from the neck region a first distance of at least 3 mm into the patient's left atrium and a second end of the conduit extends from the neck region a second distance of at least 3 mm into the patient's right atrium, thereby preventing pannus formation from narrowing the lumen in the neck region.

[0075] In one preferred embodiment, that anchor may have an hourglass or "diabolo" shaped frame with a neck region adjoining flared end regions, and the conduit may comprise a biocompatible material that encapsulates the frame. For example, the smallest diameter in the neck region may be in a range of 5 mm to 6.5 mm. Thus, the diameter of the lumen and the length of the anchor are selected to reduce a risk of paradoxical embolization. The frame, which may be formed of a biocompatible elastically or plastically deformable material, or shape memory material. The device may be implanted by forming a puncture through the atrial septum, particularly through the FO, and then percutaneously inserting the device therethrough, such that the neck region lodges in the puncture, the first end region extends into the left atrium, and the second end region extends into the right atrium.

[0076] The biocompatible material that may be a polymer, such as expanded

polytetrafluoroethylene (ePTFE), polyurethane, DACRON (polyethylene terephthalate), silicone, polycarbonate urethane, Ultra High Molecular Weight Polyethylene (UHMWPE) or PTFE. The biocompatible material may also be a metal, ceramic, carbon nanotube array or any other suitable material known to those familiar with the art that provides the shunt with the following properties. One purpose of the biocompatible covering is to form a conduit, with the biocompatible material serving as a barrier to isolate the shunt lumen from the exterior of the conduit. Additionally, the biocompatible material isolates the lumen from penetration by cellular proliferation (pannus formation) occurring on the exterior surface of the conduit, where it contacts the septum or FO, which result from the processes associated with device healing. The biocompatible material should also impede translational growth of pannus along the luminal wall of the conduit for more than about 3 mm from the site of contact with any cardiac structure.

[0077] The concept of having a separate anchor to provide shape and a conduit to provide isolation, which when combined comprise the shunt, is solely for the general convenience of developing practical device embodiments. It will be apparent to one skilled in the art that a shunt device with the requisite shape, expansion and covering characteristics could be constructed from a single unitary material that serves as both anchor and conduit. For example, one such embodiment may comprise injection molded silicone rubber that forms a single piece self-expanding shunt. Also, superelastic polymers are under development that have mechanical and biocompatible properties comparable to nitinol alloys. Thus, the anchor or frame used interchangeably throughout this specification should be considered in the general sense to refer to any composition of linked physical members that contribute in substantial part to the shunt device's shape and other physical properties that govern the shunt's transition from pre-deployment constrainment to the expanded and deployed state where it is in contact with tissue. All of the shunt device embodiments described in this patent application can be understood in terms of component parts (anchor and conduit) or as a unitary device with certain specified physical properties including shape geometry in pre-and post-deployment states and biocompatible surface properties.

[0078] The cross-sectional profile of shunt lumen perpendicular to its axis of flow may be round, oval, rectangular, or any other regular or irregular polygonal shape. The cross- sectional profile may vary from one shape to another along the axis of flow, which may be a straight line or may be curvilinear. The cross-sectional profile may rotate along the axis of flow. The shunt may have a single lumen or there may be a plurality of lumina.

[0079] In one aspect of the present invention, a device for regulating blood distribution between a patient's left atrium and right atrium comprises an anchor having a neck region joining first and second end regions, the neck region configured to engage the fossa ovalis of the patient's atrial septum; and a conduit affixed to the anchor so that the conduit extends into the right atrium by a distance selected to reduce the risk of paradoxical embolism. The conduit preferably comprises a biocompatible material that limits (or inhibits excessive) tissue ingrowth into the lumen of the conduit. The anchor and conduit are configured to accommodate endothelial or neointima layer growth up to a thickness of about 0.6 mm or less, so as to render such material inert, inhibit hyperplasia, and substantially inhibit obstruction of the flow path through the device.

[0080] In one preferred embodiment, the anchor comprises hourglass-shaped frame having a plurality of circumferential struts interconnected by longitudinal struts that, when deployed, form first and second flared end regions connected by a neck. In some

embodiments, when the shunt is deployed across the patient's atrial septum, the first flared end region protrudes 3 to 10 mm into the left atrium beyond the surface of the left septal wall. The second flared end region may protrude 5 to 10 mm into the right atrium beyond the surface of the right septal wall. The neck has an inner diameter of 4 to 8 mm, where preferably the inner diameter is in a range of 5 to 6.5 mm. The first flared end region preferably has a diameter selected in the range of between 10 and 20 mm, and the second flared end region preferably has a diameter selected in the range of between 9 and 15 mm. The first and second flared end regions each preferably flare outward from the longitudinal axis of the shunt by an amount selected from between about 25 to 60 degrees, although such angles may be different for each of the first and second flared regions. For example, in one embodiment, the steepest part of the outer surface of the first flared end region is at an angle of approximately 40 degrees relative to the longitudinal axis of the device, while the steepest part of the outer surface of the second flared end region may be at an angle of approximately 37.5 degrees relative to the longitudinal axis of the device.

[0081] In preferred embodiments, the shunt is configured to transition between a collapsed state suitable for percutaneous delivery and an expanded state when deployed across the patient's fossa ovalis, such that the shunt assumes an hourglass configuration in the expanded state. The hourglass configuration may be asymmetric. The shunt may be configured for implantation through a portion of the fossa ovalis, away from the surrounding limbus, inferior vena cava, and atrial wall.

[0082] Methods of treating a subject with heart pathology also are provided, including providing a shunt having first and second end regions and a neck region disposed

therebetween; deploying the shunt across a puncture through the subject's interatrial septum, preferably through the FO, such that the neck region is positioned in the puncture with the first end region disposed in the left atrium, and the second end region disposed in the right atrium, such that flow through the device redistributes blood between the left atrium and the right atrium through the device when the left atrial pressure exceeds the right atrial pressure. [0083] Subjects with a variety of heart pathologies may be treated with, and may benefit from, the inventive device. For example, subjects with heart failure and pulmonary congestion, reducing the left atrial pressure and left ventricular end diastolic pressure may provide a variety of benefits, including but not limited to decreasing pulmonary congestion; decreasing pulmonary artery pressure; increasing ejection fraction; increasing fractional shortening; and decreasing left ventricle internal diameter in systole. Other heart pathologies that may be treated include myocardial infarction, which may be treated by deploying the device during a period immediately following the myocardial infarction, e.g., within six months after the myocardial infarction, or within two weeks following the myocardial infarction, to reduce myocardial remodeling.

[0084] Patients with pulmonary arterial hypertension (PAH) due to idiopathic cause or associated with other disorders such as connective tissue diseases, drugs or toxins, HIV infection, portal hypertension, or congenital heart disease have been shown to benefit from atrial septostomy procedures that cause interatrial shunting from the right to the left atrium (right to left shunt). These procedures include blade or balloon septostomy or placement of devices such as uncovered diabolo stents or fenestrated atrial septal occlusion devices. It will be apparent to persons of skill in the art that the embodiments already described and other preferred embodiments described in this patent specification are applicable to treating patients with PAH.

BRIEF DESCRIPTION OF DRAWINGS

[0085] FIGS. 1 A to 1C are, respectively, perspective, end and side views of a preferred embodiment of a shunt constructed in accordance with the principles of the present invention.

[0086] FIG. 2 is a side view of an alternative embodiment of a shunt of the present invention having a cutout in its polymeric encapsulation to secure the shunt to a delivery system.

[0087] FIG. 3 is a perspective view of another alternative embodiment of a shunt of the present invention having an alternative cutout in its encapsulation.

[0088] FIGS 4A and 4B are, respectively, end and side views of a further alternative embodiment of a shunt constructed in accordance with the principles of the present invention having eyelets that engage a delivery system. [0089] FIGS. 5 A and 5B are plan views of further alternative embodiments of anchors suitable for use in the inventive shunt, cut along line 5A— 5A and 5B— 5B, and unrolled to a flat configuration.

[0090] FIG. 6 is a graph comparing theoretical flows through a shunt design having a Venturi contour with 5 mm and 6 mm diameter orifices compared to theoretical flows obtained using orifice plate-type devices.

[0091] FIGS. 7A and 7B are, respectively, a plan view of the right atrial side of the atrial septum, illustrating implantation of a shunt through a portion of the fossa ovalis, and a perspective view of an embodiment of the shunt of FIGS. 1 A-IC positioned in the fossa ovalis of the atrial septum.

[0092] FIGS. 8 A and 8B schematically depict pannus formation on an hourglass-shaped embodiment of the shunt of the present invention positioned in the fossa ovalis orthogonal to the atrial septum wall, immediately after implantation and after pannus formation.

[0093] FIGS. 9 A and 9B schematically depict pannus formation on an hourglass-shaped embodiment of the shunt of the present invention positioned in the fossa ovalis non- orthogonal to the atrial septum wall, invention immediately after implantation and after pannus formation.

[0094] FIGS. 10 through 15 depict various alternative embodiments of shunts constructed in accordance with the principles of the present invention.

[0095] FIGS. 16A and 16B are, respectively, side and end views of anchor suitable for a further alternative shunt embodiment having self-expanding flexible arms that form a filter over the right atrial side of the conduit.

[0096] FIG. 17 is a graph comparing theoretical flows through shunt designs constructed in accordance with the principles of the present invention compared to a previously known valved shunt design. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0097] Interatrial shunts are provided for redistributing interatrial blood volumes and reducing left atrial pressure, which may be advantageous in treating subjects suffering from heart failure (HF) or other disorders associated with elevated left atrial pressure. A preferred embodiment of the inventive device includes an anchor, which may be an hourglass or "diabolo" shaped stent or frame, and a conduit, formed by encapsulating the frame in a synthetic biocompatible material. The shunt is configured to be lodged securely within a passage formed in the atrial septum, preferably the fossa ovalis, and provides one-way blood flow from the left atrium to the right atrium, when blood pressure in the left atrium exceeds that on the right.

[0098] Referring now to FIGS. 1 A to 1C, an illustrative embodiment of shunt 10 of the present invention is described. Shunt 10 generally comprises anchor 12 having three regions: flared or funnel-shaped end region 14, flared or funnel-shaped end region 18, and neck region 16 disposed between end regions 14 and 18. Neck region 16 is configured to lodge in a puncture formed in the atrial septum, preferably in the fossa ovalis. Flared end regions 14 and 18 are configured to partially engage and protrude beyond the right and left sides, respectively, of the atrial septum when implanted. Shunt 10 further comprises a conduit, illustratively formed by encapsulating anchor 12 with biocompatible material 20 that covers all or substantially all of anchor 12 to form a conduit defining a lumen or interior passageway 22.

[0099] Flared region 14 is configured to be disposed in the right atrium, while flared region 18 is configured to be disposed in the left atrium. In one embodiment, anchor 12 includes six longitudinal struts 24 interconnected by five circumferential struts 26a-26e. Longitudinal struts 24 prevent foreshortening of the anchor during expansion, while the sinusoidal or serpentine bends in circumferential struts 26a-26e permit the anchor to transition between a radially collapsed substantially cylindrical delivery state to an expanded, flared, deployed state as illustrated in FIGS. 1 A to 1C. As depicted in the figures, a conduit is formed by biocompatible material 20 that encapsulates the entirety of neck 16, flared end region 18, and flared end region 14. Biocompatible material 20 preferably is affixed to anchor 12 using a suitable biocompatible adhesive or by sandwiching the anchor between inner and outer layers of biocompatible material using sintering techniques. [00100] In a preferred embodiment, anchor 12 comprises a self-expanding material, such as a shape memory alloy, and circumferential struts 26a-26e are treated to expand a predetermined amount when deployed, so that together with encapsulation 20, lumen 22 has a contour that permits substantially laminar flow between flared end section 18 (in the left atrium) and flared end section 14 (in the right atrium). Sinusoidal or serpentine bends 28 in circumferential struts on flared end region 14 preferably are 180 degrees out of phase with the sinusoidal or serpentine bends 28 in neck region 16 and flared end region 18, so that the sinusoidal or serpentine bends do not extend beyond the ends of longitudinal struts 24 in either the collapsed delivery state or deployed state.

[00101] Anchor 12 may comprise a biocompatible metal framework or laser-cut solid metallic tube made from nitinol, titanium alloy, cobalt chromium alloy, MP35n, 316 stainless steel, L605, Phynox/Elgiloy, platinum chromium or other biocompatible metal such as are known to persons of skill in the art. While a preferred embodiment employs a shape memory self-expanding alloy, anchor 12 alternatively may comprise an elastically or plastically deformable material, e.g., balloon expandable, or may be a shape memory alloy that responds to temperature changes to transition between contracted delivery and expanded deployed states. The surface finish applied to the material of the anchor may be selected to control the distance, thickness, composition and/or growth pattern of pannus formation, e.g., the external surfaces of anchor 12 may be electro-polished.

[00102] In accordance with the principles of the present invention, the radial dimensions, axial lengths and contours of neck region 16 and flared end regions 14 and 18 preferably are selected to provide laminar flow through the interior of the shunt, to reduce the formation of eddy currents when implanted, and thus inhibit thrombus formation; to inhibit pannus formation that could obstruct the neck region; to promote tissue ingrowth around the exterior of the neck region to secure the shunt against migration; to provide a desired rate of blood flow between the left and right atria at physiological pressure differentials; and to prevent retrograde paradoxical embolization.

[00103] Biocompatible material 20 forming the conduit preferably is resistant to the transmural and translational ingrowth of pannus material having a tissue thickness greater than 0.6 mm. For example, in experimental ePTFE vascular grafts, those with a 60-micron internodal distance showed rapid, transmural infiltration with proliferating smooth muscle cells and granulation tissue, whereas ePTFE grafts with a 30-micron internodal distance were observed to develop only a slow growing, thin sheet of endothelium that advanced only a few millimeters into the graft lumen from the adjacent artery. Porous polyester fabric coverings employed on some atrial septal defect ("ASD") occlusion devices would be poor choices for use in the shunt of the present invention, because such materials become completely enmeshed with penetrating fibrotic tissue. It is expected that when shunt 10 comprises anchor 12 made of, for example, electro polished nitinol, and biocompatible material 20 may be an inert polymer such as ePTFE with an internodal distance of 30 microns or less, or is PTFE, such that pannus will grow to a thickness no greater than about 0.6 mm after extending translationally a distance of 3 mm from the site of contact with the Foramen Ovalis ("FO") tissue. In such cases, interior lumen of the conduit is not expected to narrow beyond a total of 1.2 mm from its original diameter and the neck. For the purposes of this patent the term "luminal narrowing" shall be defined as a loss of minimal shunt lumen diameter of greater than 25% and the term "luminal obstruction" is defined as total (100% loss of lumen diameter) blockage of the lumen to the flow of blood.

[00104] In the preferred embodiment depicted in FIGS. 1 A to 1C, anchor 12 has an hourglass shape formed of a shape memory metal, e.g., nitinol, or any other suitable material known in the art. Circumferential struts 26a-26e and longitudinal struts 24 preferably comprise a unitary construction, that is, entire anchor 12 is laser cut from a tube of shape memory metal. Biocompatible material 20 may comprise, for example, a sheet of a polymer such as expanded polytetrafluoroethylene ("ePTFE"), polytetrafluoroethylene ("PTFE",) silicone, polycarbonate urethane, DACRON (polyethylene terephthalate), Ultra High

Molecular Weight Polyethylene (UHMWPE), or polyurethane. The biocompatible material may also be a metal, ceramic, carbon nanotube array or any other suitable biocompatible material. For example, biocompatible material 20 may comprise ePTFE with an up to 30- micron internodal distance, and may be applied as inner and outer layers sintered together to form a unitary conduit. Alternatively, biocompatible material 20 may be applied to the inner lumen and the outside of the anchor using electrospinning techniques. Other methods of encapsulation and other suitable polymers that prevent transmural ingrowth of pannus tissue may alternatively be used, as will be understood by one skilled in the art. Bare metal regions of anchor 12, and any other regions of the anchor, optionally may be electropolished or otherwise treated to inhibit thrombus formation using known methods. [00105] As noted above, neck 16 of shunt 10 preferably is configured for implantation through the fossa ovalis of the atrial septum, and more preferably near or at the central portion of the fossa ovalis. As known to those skilled in the art, the fossa ovalis is a thinned portion of the atrial septum formed during fetal development of the heart, which appears as an indent in the right side of the atrial septum and is surrounded by a thicker portion of the atrial septum. While the atrial septum itself may be several millimeters thick and muscular, the fossa ovalis may be only approximately one millimeter thick, and is formed primarily of fibrous tissue.

[00106] In some embodiments of the present invention, shunt 10 may be asymmetrically shaped to take advantage of the natural features of the atrial septum near the fossa ovalis, and to provide suitable flow characteristics. For example, in a preferred embodiment, the anchor comprises an hourglass or diabolo shape where a LA entry funnel resembles a conical-shaped nozzle and a RA exit funnel is "bell" shaped, with the wide mouth lumen of the bell at the RA exit port in the RA. The narrow entrance to the bell-shaped exit funnel connected to the orifice of the neck region may be configured to approximate the curved surface of a parabola. This type of convergent-divergent nozzle resembles the shape of a classical de Laval nozzle used in rocket engines. Left to right flow is largely governed by the smooth convergence of streamlines in the entry cone and the divergence of streamlines exiting the bell. Such a nozzle configuration is very efficient in the forward flow direction having a discharge coefficient resembling a classic venturi tube, e.g., 0.95-0.98. In addition, in the event there is retrograde flow, e.g., when right atrial pressure exceeds left atrial pressure, streamlines entering the bell must make an abrupt directional change to converge in the orifice which create eddies at the entrance to the orifice portion of the neck region. This reduces the likelihood of paradoxical embolization because bulk retrograde flow is reduced relative to antegrade flow for the same absolute value of the pressure differential across the shunt.

[00107] Referring now to FIG. 1C, points B and C are located on the leftmost

circumferential strut 26e, which defines the LA entry port. Points A and D are located on circumferential strut 26d along the LA entry funnel proximal to strut 26e. Points H and E are located on circumferential strut 26b along the RA exit funnel, and points G and F are located on circumferential strut 26a, which defines the RA exit port. In preferred embodiments, the diameter of lumen 22 in the neck region of the shunt orifice ranges from 5 to 6.5 mm. The portion of the shunt crossing the FO, bounded by points ADEH may be 3 mm in axial length but may be extended up to 10 mm in patients with a thicker FO. The diagonal length between points AB, CD, EF, and/or GH is preferably >3 mm so that pannus cannot grow translationally inward from the ends of the shunt and thus obstruct neck region 16. In addition, the horizontal component length between points AB, CD, EF, and/or GH is preferably <15 mm, to avoid interference with existing cardiac structures when implanted. In accordance with another aspect of the invention, it has been determined that providing a length of segments EF and GH generally greater than 5 mm is expected to ensure that the end region that extends into the right atrium is disposed generally out of the flow path of blood returning from the inferior vena cava, which is most likely to have entrained emboli that could cause paradoxical embolization. Truncated funnel cones bounded by ABCD and/or EFGH may have volumes <2 ml.

[00108] Other embodiments of the shunt of the present invention may include anchors with different combinations and configurations of circumferential ring and axial strut elements. Specifically, such embodiments, may have more or less than 6 longitudinal struts 24 and more or less than five circumferential struts 26a-26e. These configurations may yield other shunt lumen geometries. In another embodiment, anchor 12 may be made of a self- expanding polymer. Alternatively, the anchor need not be self-expanding, and may be made from a plastically deformable biocompatible metal such as 316 L stainless steel, cobalt chromium alloys, or any other such suitable materials known to those skilled in the art. Such a deformable shunt anchor may be delivered by an expanding member, such as a balloon, that is configured to achieve the desired luminal geometry. The deformable anchor may be designed to expand prismatically or at certain localized sites where ductile hinges are configured for more selected expansion as taught by U.S. Patent No. 6,242,762 to Shanley, the contents of which are incorporated by reference herein.

[00109] Referring now to FIG. 2, an alternative embodiment of a shunt constructed in accordance with the principles of the present invention is described. Shunt 30 includes anchor 31 is similar in construction to that described for the embodiment of FIGS 1 A-IC, and has flared end regions 32 and 33 and neck region 34. When implanted in a patient's interatrial septum, flared end region 32 is disposed in the patient's right atrium, while flared end region 33 is disposed in the patient's left atrium, with neck region 34 situated in a passage formed in the interatrial septum. Anchor 31 includes longitudinal struts 35 and circumferential struts 36a-36e, and is encapsulated by biocompatible material 37. Anchor 31 may comprise a self-expanding or plastically deformable material as described herein above.

[00110] Shunt 30 of FIG. 2 differs from the previous embodiment in that biocompatible material 37, for example ePTFE, includes cutout 38 adjacent to circumferential strut 36a. Cutout 38 may extend proximally from circumferential strut 36a for a distance of 0.5 mm to 2 mm, and more preferably about 1 mm, to permit circumferential strut 36e to be releasably engaged with a delivery system during deployment, for example, hooks, as described by Yacoby in US Patent Publication 2014/0350565. Biocompatible material 37 may be trimmed manually or mechanically from circumferential strut 36a to create cutout 38 or by laser- cutting. In this manner, shunt 30 may be positioned and repositioned in a passage formed in the interatrial septum until the clinician is satisfied with the device placement, before being released. In a preferred embodiment, the conduit formed by biocompatible material 37 extends a distance of at least 3 mm beyond neck region 34 into flared end region 32, to ensure that pannus cannot grow translationally along luminal wall far enough to partially occlude the flow area of neck region 34. Additionally, flared end region 32 extends a distance of at least 5 mm into the right atrium when implanted in the interatrial septum to ensure that the entry of flared end region 34 is generally not aligned with natural circulation flow paths generated by blood entering the right atrium from the inferior vena cava, thereby reducing the risk that emboli carried from the lower extremities into the right atrium will cause paradoxical embolism by passing through shunt 30. Thus, emboli entrained in flow from the inferior vena cava is not directed into the entry of flared end region 34.

[00111] With respect to FIG. 3, another alternative embodiment of inventive shunt is described. Shunt 40 includes anchor 41 having flared end regions 42 and 43 joined by neck region 44, as described for the preceding embodiments. Anchor 41 includes longitudinal struts 45 joined by circumferential struts 46a-46e and biocompatible material 47, for example a thin layer of ePTFE or other suitable material as described above. Shunt 40 differs from the embodiment of FIGS. 1 A to 1C in that the polymeric encapsulation includes cutouts 48 on alternating peaks of the sinusoidal bends formed by circumferential strut 46a that permit a delivery device to releasably engage shunt 40. Shunt 40 also includes skirt 49 of

biocompatible material that extends beyond circumferential strut 46e. In a preferred embodiment, cutouts 48 include circular sectors having angles in the range of 60° to 180°, more preferably 120°, such that largest distance between the edge of the polymeric encapsulation and circumferential strut 46a is in the range of 0.5 to 2 mm, and more preferably 1 mm. The configuration of cutouts 48 of shunt 40, which may be laser cut, advantageously maximize the encapsulated area of the shunt while still enabling proper engagement to the delivery system hooking mechanism. As will be apparent to those skilled in the art, other possible cutting patterns or methods may be employed.

[00112] Referring now to FIGS. 4A and 4B, another embodiment of a fully encapsulated hourglass shunt constructed in accordance with the principles of the present invention is described. Shunt 50 includes anchor 51 having end regions 52 and 53 joined by neck region 54. Anchor 51 has longitudinal struts 55 coupled to circumferential struts 56a-56e as described for preceding embodiments, and includes a conduit formed of biocompatible material 57 as also described hereinabove. Shunt 50 differs from the embodiment of FIGS. 1 A to 1C in that alternating longitudinal struts 55 include elongated portions 58 having eyelets 59 for engagement with a delivery system extending from right atrial end region 52. Shunt 50 may have between 2 to 6, and preferably 3 elongated portions 58 and eyelets 59 left as bare-metal, i.e., without polymeric encapsulation. Elongated portions 58 preferably are short, protruding a minimum additional distance into the right atrium or alternatively are constructed to bend into the right atrium RA exit port on release from the delivery system to serve as filter to block paradoxical emboli from passing into the lumen of the conduit at end region 52. An alternative approach that also filters the size of emboli is to construct the shunt with a plurality of passageways or lumina that transport blood in parallel such that the total cross-sectional area of all the of the passageways conserves the flow characteristics needed for adequate shunting to achieve the redistribution of blood between the atria as desired.

[00113] With respect to FIGS. 5A and 5B, further alternative embodiments of an anchor suitable for constructing a shunt in accordance with one aspect of the present invention are described. Anchor 60 is similar in design to anchor 51 of the embodiment of FIGS. 4A and 4B, and includes longitudinal struts 61 joined to circumferential struts 62a-62e, which include sinusoidal bends. Accordingly, anchor 60 when expanded includes flared end regions joined by a neck region to form a generally hourglass shape, while longitudinal struts 61 prevent foreshortening, i.e., axial shrinkage, during deployment. For purposes of illustration, anchor 60 as depicted in FIG. 5 A is shown cut along one of longitudinal struts 61 (along line 5 A— 5A) and flattened, although the anchor preferably is cut from a tubular material. As for preceding embodiments, anchor 61 includes a polymeric encapsulation that forms a conduit, omitted for clarity from FIG. 5 A, that covers the anchor between circumferential struts 62a and 62e. Anchor 60 includes elongated portions 63 and eyelets 64 that extend into the right atrium when the shunt is employed. In accordance with one aspect of the invention, alternating eyelets 64 include radiopaque markers 65, for example made of platinum iridium, gold, tantalum, or any other similar suitable material, which enhance visualization of the shunt under fluoroscopy. Eyelets 64 that do not accommodate radiopaque markers 65 permit the shunt to be releasable engaged by a delivery system for percutaneous transluminal delivery.

[00114] In FIG. 5B, anchor 66 is similar in design to anchor 60 of the embodiment of FIG. 5A, except that in this embodiment circumferential struts 67a-67e having sinusoidal bends that extend between longitudinal struts 68 all face in the same direction. Anchor 66 additionally includes eyelets 69 that extend from alternating longitudinal struts 68 for use in releasably coupling the shunt to a percutaneous transluminal delivery system. One advantage of this design is that retrieval of a self-expanding shunt using anchor 66, by its delivery system when halfway deployed or fully deployed, requires less pull-back force to collapse the shunt than the embodiment of FIG. 5 A. As in preceding embodiments, anchor 66 when expanded includes flared end regions joined by a neck region to form a generally hourglass shape, while longitudinal struts 68 prevent foreshortening during deployment. For purposes of illustration, anchor 66 as depicted in FIG. 5B is shown cut along one of longitudinal struts 68 (along line 5B— 5B) and flattened, although the anchor preferably is cut from a tubular material. Anchor 66 further includes a conduit formed by encapsulating the anchor with a biocompatible material, omitted for clarity from FIG. 5B, that covers the anchor between struts 67a and 67e.

[00115] In accordance with one aspect of the present invention, an interatrial hourglass- shaped shunt with flow characteristics resembling a venturi tube and a discharge coefficient of approximately 0.96-0.97 may have a minimal neck orifice inner diameter ranging from 5 mm to approximately 6.5 mm. Having a somewhat larger orifice diameter, within this range, e.g. 6.0 mm, will support approximately 35% more flow for any given pressure gradient compared with a 5.1 mm shunt, as shown in FIG. 6. This may not only create improved hemodynamic conditions but provide additional benefit in maintaining shunt flow should some shunt narrowing due to pannus ingrowth occur during device healing. [00116] In accordance with another aspect of the invention, various nozzle geometries with high discharge coefficients relative to an orifice-plate geometry advantageously may be used to provide laminar flow through the shunt. These include but are not limited to various variations of venturi tubes, conical convergent nozzles (with convergence angles from 20 to 80 degrees), cylindrical convergent nozzles, and the Addy type nozzle with a convergent curved entrance wall leading to a length of cylindrical tubing having a diameter equivalent to the orifice diameter. The latter two appear similar in appearance to the horn of a trumpet. In another preferred embodiment, the shunt lumen may be a cylindrical tube with no or minimal dilation at the entry or exit ports.

[00117] The cross-section of lumen 22 (see FIG. IB) need not be circular and/or the lumen need not be coaxial with a straight horizontal line axis when viewed longitudinally. Although these latter geometries may be difficult to deliver through catheters with circular luminal cross-sections, they may be constrained to such catheter lumens and expand into non-circular cross-sectional or curved longitudinal geometries upon deployment. Other preferred embodiments include any combination of entry, orifice, and exit geometries where the exiting jet vena contracta cross-sectional area is 70% or greater compared with the minimal orifice area, over the range of physiological interatrial pressure gradients, thereby having a higher discharge coefficient than an orifice-plate.

[00118] A shunt with a single LA conical entry funnel, with an hourglass-shaped lumen, or with a tubular lumen, having a discharge coefficient of 0.70 or larger, generally has a longer tunnel of entrained flow by nature of its longer length, typically 6 to 30 mm long, versus an orifice-plate mesh type shunt, which may be defined by the thickness of the FO itself and is typically shorter than 6 mm, e.g., 3 mm or less. For paradoxical embolization to occur, i.e., for a paradoxical embolus to embolize from the heart into the systemic arterial circulation, the paradoxical embolus must pass completely or nearly completely through the shunt.

Emboli may be propagated by their residual kinetic energy against a left-to right gradient or when there is no gradient, or may be carried along when a reversed pressure gradient creates right to left bulk flow. Depending on the relative magnitude of the kinetic energy of the embolus and the bulk flow directional status, a longer lumen shunt will tend to pass fewer emboli compared to an orifice-plate shunt with a shorter lumen. This is likely to be the case in the presence of normal left to right bulk flow or when there is zero net flow. This is also likely to be true during very transient pressure gradient reversals, such as during a cough. Therefore, in another preferred embodiment, a shunt with a flow lumen length of 6 to 30 mm, or more typically 10 to 15 mm, by virtue of its increased lumen length, will have less tendency for paradoxical embolization than an orifice-plate mesh shunt. Thus, the length of the flow lumen of the anchor is selected to reduce a risk of paradoxical embolization.

[00119] Referring now to FIG. 7 A, a preferred location for implanting shunt 10 of FIGS. 1 A-1C of the present invention is described. FIG. 7A is a plan view of the right atrial side of atrial septum 70, including implantation site 71 located at a central position of fossa ovalis 72. Preferably, implantation site 71 is selected so that the shunt may be implanted spaced apart from the surrounding limbus 73, inferior vena cava (IVC) 74, and atrial septum 75. For example, as shown in FIG. 7B, flared end region 14 is configured to be implanted in right atrium 76 and may be tapered so as to have a more cylindrical shape than does flared end region 18, which is configured to be implanted in left atrium 77. The more cylindrical shape of flared end region 14 may reduce or inhibit contact between flared end region 14 and limbus 73 of fossa ovalis 72, that is, between flared end region 14 and the prominent margin of the fossa ovalis, while still anchoring device 10 across atrial septum 75. The more cylindrical shape of flared end region 14 further may reduce or inhibit contact between flared end region 14, and the right side of atrial septum 70, as well as ridge 77 separating the coronary sinus from the IVC 74 (shown in FIG. 7 A but not FIG. 7B).

[00120] Still with respect to FIG. 7A, a preferred location for shunt implantation may be slightly anterior to the centerline of the long axis of the fossa ovalis, i.e., located on the right hand side of the ovale. This location leaves potential space in the upper left quadrant (posterior-superior) of the fossa, which has been found to be optimal for crossing the fossa to perform structural heart disease procedures on the mitral valve, including edge-to-edge repair with MitraClip® transcatheter mitral valve repair system offered by Abbott, Abbott Park, IL and mitral annuloplasty with Cardioband, offered by Valtech Cardio, Or Yehuda, Israel. This preferred location also leaves potential space in the lower left quadrant (posterior-inferior) of the fossa, which has been found to be optimal for crossing the fossa to perform structural heart disease procedures to occlude the left atrial appendage. A shunt with an hourglass shape that occupies the smallest possible location on the fossa, as described herein, facilitates these other procedures. [00121] Again, referring to FIG. 7B, shunt 10 preferably is configured so as to avoid imposing significant mechanical forces on atrial septum 75, thus allowing the septum to naturally deform as the heart beats. For example, the thicknesses of muscular areas of septum 75 may change by over 20% between systole and diastole. It is believed that any significant mechanical constraints on the motion of atrial septum 75 in such areas would lead to the development of relatively large forces acting on the septum and/or on atrial tissue that contacts shunt 10. Such forces could invoke an inflammatory response and/or hyperplasia in the atrial septum tissue, and possibly cause shunt 10 to eventually lose patency. However, by configuring shunt 10 so that neck region 16 may be implanted entirely or predominantly in the fibrous tissue of the fossa ovalis 72 with a small footprint, the hourglass shape of shunt 10 is expected to be sufficiently stable so as to be retained in the septum, while reducing mechanical loads on the surrounding atrial septum 75. Tissue ingrowth from atrial septum 75 in regions 78 may further enhance binding of shunt 10 to the septum. Preferably, there should be a substantial rim of fossa around the shunt after implantation, e.g., for a thickness of 1-2 mm, as depicted in FIG. 7B.

[00122] Also, because neck region 16 of shunt 10 is significantly narrower than flared end regions 14 and 18, shunt 10 will "self-locate" in a puncture through atrial septum 75, particularly when implanted through the fossa ovalis, with a tendency to assume an orientation where its longitudinal axis is substantially orthogonal to the FO. In some embodiments, neck region 16 may have a diameter suitable for implantation in the fossa ovalis, e.g., that is smaller than the fossa ovalis, and that also is selected to inhibit blood flow rates exceeding a predetermined threshold. Neck region 16 preferably provides a passage having a diameter between about 4 and about 7 mm, and more preferably between about 5 mm and about 6.5 mm. For example, diameters of less than about 4 mm may in some circumstances not allow sufficient blood flow through the shunt to decompress the left atrium, and may reduce long-term patency of the shunt. Conversely, diameters of greater than about 7 mm may allow too much blood flow, resulting in right ventricular volume overload and pulmonary hypertension. Preferably, the effective diameter at the narrowest point in shunt 10 is about 5 mm to 6.5 mm. In addition, by virtue of its smaller orifice diameter at the neck region, the shunt will have less tendency for paradoxical embolization than an orifice-plate mesh shunt. Thus, the orifice diameter at the neck region of the anchor is selected to reduce a risk of paradoxical embolization. [00123] The diameters of flared end regions 14 and 18 further may be selected to stabilize shunt 10 in the puncture through atrial septum 45, e.g., in the puncture through fossa ovalis 72. For example, flared end region 18 may have a diameter of 10 to 20 mm at its widest point, e.g., about 13 to 15 mm; and flared end region 14 may have a diameter of 9 to 15 mm at its widest point, e.g., about 9 to 13 mm. The largest diameter of flared end region 14 may be selected so as to avoid mechanically loading the limbus of the fossa ovalis 72, which might otherwise cause inflammation. The largest diameter of flared end region 18 may be selected so as to provide a sufficient angle between flared end regions 14 and 18 to stabilize shunt 10 in the atrial septum, while limiting the extent to which flared end region 18 protrudes into the left atrium (e.g., inhibiting interference with flow from the pulmonary veins), and providing sufficient blood flow from the left atrium through neck region 16.

[00124] In accordance with the principles of the present invention, the length of end region 14 is selected to protrude into the right atrium by a distance sufficient to inhibit tissue ingrowth that may otherwise interfere with the operation of shunt 10. Applicants have observed that tissue ingrowth inwards along an impermeably membranes of specified biomaterials from the end that contacts tissue generally stops after about 3 mm. Accordingly, to ensure that tissue ingrowth from the ends of the conduit does not extend into and partially occlude the flow area of neck region 16, the distance R between the narrowest portion of neck region 16 and the end of region 14 should be at least 3 mm plus half of the thickness of the septal region, i.e., fossa ovalis, contacting the exterior of shunt 10. Assuming that the fossa ovalis has a thickness of about 3.0 mm, then the minimum distance R should be about 4.5 mm, based on applicants' observations. Likewise, end region 18 preferably does not significantly engage the left side of atrial septum 75, so that distance L also preferably is at least 4.5 mm. Due to patient-to-patient variability in the thickness of the FO, e.g., due to the patient's general health and age, and because neck region 16 may not be precisely aligned with the mid-point of the FO, each distances R and L preferably fall within a range of 3 to 6 mm. Accordingly, for some embodiments, the overall dimensions of shunt 10 may be about 9-12 mm long (L+R, in FIG. 7B) to prevent tissue ingrowth from the ends of the conduit, i.e., end regions 14 and 18, from partially occluding neck region 16.

[00125] In another preferred embodiment, regardless of the geometrical shape of the conduit, there should be a minimum of 3 mm of material resistant to translational tissue growth, i.e., extending inward from the ends of the end regions to accommodate neoendocardial tissue growth over the shunt surfaces starting from a location in contact with the atrial septum, such that tissue growth cannot reach the orifice (site of minimal diameter of the shunt lumen or cross-sectional area of lumen 22 shown in FIG. IB). With this preferred embodiment, the minimal orifice diameter of an interatrial shunt device will be rendered largely unaffected by pannus formation. In another preferred embodiment, there should be a minimum of 3 mm of conduit length for neoendocardial tissue to grow over the shunt luminal surfaces starting from a location in contact with the atrial septum, before reaching the entrance or exit port sites of the shunt lumen. With such an embodiment, there is even less potential for pannus to encroach the shunt lumen.

[00126] Referring now to FIGS. 8A and 8B, the expected healing response invoked by implanting shunt 10 of FIGS. 1A-1C orthogonally across the FO is described, while FIGS. 9A and 9B correspond to implantation of the shunt non-orthogonally so that an outer surface of the LA entry cone contacts the atrial septal tissue. FIGS. 8A and 9A depict positioning of the shunts immediately post implantation, while FIGS. 8B and 9B depict shunt positioning after the completion of the healing phase.

[00127] In each of FIGS. 8 A and 8B, the FO is shown as bowed towards the RA and concave towards the LA. In patients with dilated cardiomyopathy or restrictive physiology, including most patients with left ventricular failure, regardless of etiology, the FO portion of the interatrial septum generally is bowed toward the right atrium. This gives the LA a generally concave or near hemispherical shape in the region centered on the FO. Conversely, the RA side of the FO is generally convex in shape. This orientation of the FO was confirmed by echocardiography (n=178 examinations) in the 38 patients implanted with the V-Wave Nitzan-type valved shunt discussed in the Background of the Invention portion of this specification. In measurements of more than 100 patients exhibiting heart failure with preserved ejection fraction (HFpEF), the LA volume generally averaged 85 ml with a minimum volume of 54 ml, while for a like number of patients exhibiting heart failure with reduced ejection fraction (HFrEF), the LA volume generally averaged 104 ml with a minimum volume of 71 ml. Although the LA is often approximated by a sphere or an ellipsoid, there are frequently exceptions to this, for example, where the LA appears squashed when viewed in its anterior-posterior dimension. Although not specifically quantified, the RA appeared to be similar in size to the LA. [00128] Although exceptions to RA bowing of septal anatomy occur, they generally do so in the presence of isolated right ventricular failure or severe pulmonary hypertension in the absence of left ventricular dysfunction or mitral valve disease, e.g. as occurs in pulmonary arterial hypertension (PAH). In those instances, RA pressure tends to exceed LA pressure causing the FO to bow in the opposite direction toward the LA. Such patients generally would derive no clinical benefit from left-to-right interatrial shunting. However, patients with severe pulmonary hypertension in the absence of left-sided heart failure may benefit from right-to-left shunting as a means to improve low systemic cardiac output. Several of the embodiments described in this disclosure would be provide improved performance compared to right-to-left shunts currently available to that population of patients.

[00129] Another geometrical constraint is the frequent presence or need to place transvenous endocardial electrical pacing or defibrillation leads in or through the RA of heart failure patients. In the 38-patient feasibility study conducted with the V-Wave Nitzan-type shunt, 74% of patients had already been implanted with cardiac rhythm management devices prior to interatrial shunting. Most of these patients had 2 or 3 such electrical leads placed. Leads most often enter the RA from the superior vena cava (SVC). Right atrial pacing leads usually loop up and terminate anterio-laterally in the RA appendage, but in some

circumstances, they are attached to a muscular portion of the interatrial septum. RV pacing and defibrillation leads usually course along the lateral wall of the RA, then cross the tricuspid valve, and terminate in the interventricular septum, RV apex, or pulmonary outflow tract. LV leads enter the coronary sinus, which is just below and anterior to the FO.

Occasionally, leads must be placed from unusual sites of origin and may enter the RA from the inferior vena cava ("IVC"). Leads are usually left with enough slack so that they do not put tension on their terminal ends when the heart moves or changes position. Much of this slack results in a web of excess lead body material that is often concentrated in the RA.

[00130] The observations of septal bowing, the range of chamber dimensions observed and the consequences of multiple transvenous endocardial lead placement have important implications for interatrial shunt device design. If a shunt protrudes into the LA chamber, it preferably is placed so that it generally projects orthogonally with respect to the FO as shown in FIG. 8 A. Orthogonal placement is expected to minimize impingement on other adjacent or nearby critical cardiac structures, such as the aortic root, the mitral valve annulus, the roof and the posterior wall of the LA, and the pulmonary veins. Alternatively, if not placed substantially orthogonally, as shown in FIG. 9A, the shunt geometry should be selected to prevent the shunt from interacting with these structures. Proper accounting for such design considerations will prevent erosion of the shunt into critical cardiac structures, and prevent blockage of flow through the shunt by luminal impingement by adjacent cardiac structures. Ideally, the shunt should also occupy minimal space within the LA and only minimally disturb its normal flow pattern. The LA fills from the pulmonary veins during ventricular systole and drains into the left ventricle when the mitral valve opens during diastole. Blood coming from the right superior pulmonary veins tends to course along and hug the interatrial septum preventing stasis near the FO.

[00131] In a preferred embodiment of shunt 10, the volume of blood displaced by the portion of the shunt protruding into the LA, i.e., the volume of blood in the portion of the shunt lumen protruding into the LA, should be less than or equal to 5% of the LA diastolic volume expected in the patient population. This is typically 2.0 ml or less in adult patients with heart failure. Moreover, the shunt should not protrude into the LA by more than 15 mm, or more typically 3 to 10 mm. These dimensional considerations may also be accomplished in conjunction with other shunt features that facilitate a substantially orthogonal orientation, such as an LA entry funnel.

[00132] Similar considerations exist for the RA side of the FO. The shunt should occupy a minimal volume and have only a small effect on normal flow patterns. In a preferred embodiment, the same occupying volume and protrusion distance considerations, apply to the RA side of the shunt, that is, the device and its lumen should occupy less than or equal to 5% of the RA diastolic volume, e.g., 2.0 ml or less in adult patients with heart failure, and protrude into the RA by no more than, for example, 15 mm, or more typically 3 to 10 mm. These dimensional considerations can also be accomplished in conjunction with other shunt features that facilitate a substantially orthogonal orientation, such as RA exit funnel. These same criteria apply when the shunt is used in an application where RA to LA shunting is desirable, e.g., pulmonary arterial hypertension (PAH). The shunt should protrude in the RA the least amount necessary so that it does not foul pacing leads or abrade their electrical insulation.

[00133] As described earlier, the propensity for venous thromboembolism ("VTE") to cross in the retrograde direction through a shunt is expected to be a function of not only the amount and duration of retrograde shunt flow from the RA to the LA, but also a result of the flow patterns in the RA. The path of flow in the adult RA is complex because blood enters the chamber from multiple sources which include the inferior vena cava (IVC), the superior vena cava (SVC), the coronary sinus and from the LA through the shunt. These flow paths include directional changes and asymmetries whose topology has been assessed by color flow Doppler imaging and more recently from magnetic resonance velocity mapping.

[00134] Since the overwhelming majority of VTE in adult patients originate from the lower extremities and pelvic veins, the path traveled by paradoxical emboli are most likely similar to the flow vectors for blood coming from the IVC. Flow from the inferior vena cava courses along the posterior wall of the RA chamber before looping around the roof, where it is directed toward the tricuspid valve by coursing along the interatrial septum. The rest of the cavity generally contains pooled blood. Thus, blood entering the RA from the IVC forms a clockwise vortex descending along the RA side of the interatrial septum in most patients with normal anatomy. Advantageously, this flow pattern of blood downwards from the roof of the RA and along the interatrial septum reduces the risk of blood pooling in the vicinity of neck region 16 of the inventive shunt 10, thus reducing the risk of local thrombus formation due to blood stasis. Further, these flow pathway observations suggest that a thrombus originating from inferior vena cava will a have a trajectory that passes very close to the RA orifice of a naturally occurring secundum type atrial septal defect or an orifice-plate mesh type shunt. Because in this case thrombus is essentially presented by the flow path within the RA to the orifice, even a small reversal of shunt flow could embolize the thrombus across the orifice into the LA.

[00135] In accordance with another aspect of the present invention, a preferred

embodiment of an inventive shunt includes an exit port (end region 14) that extends a distance into the RA, e.g., 3 to 15 mm, or more typically 5 to 10 mm, sufficient to place the orifice of the exit port out of the naturally occurring flow paths in the RA. In particular, the exit port projects partially or completely through the stream of blood originating from the IVC that loops down across the interatrial septum. Such a shunt geometry thus will be expected to have a lower risk of paradoxical embolization compared with an orifice-plate mesh type shunt where the exit port is directed at the passing looped IVC flow stream.

[00136] Referring now to FIGS. 10 and 11, additional alternative embodiments are described, where a conduit is registered with respect to the fossa ovalis of the interatrial septum by an external, unencapsulated bare metal anchor similar to anchor 12 of the embodiment of FIGS. 1A-1C. Specifically, shunt 80 of FIG. 10 includes anchor 81, which may be employed to register conduit 82 within the interatrial septum. Conduit 82 may include a separate encapsulated tubular frame or may comprise a tube of solid material, and may include a variety of geometries to achieve specific characteristics as previously described. Anchor 81 and conduit 82 may be physically affixed to each other prior to insertion in the body by mechanical interference, welding, adhesives, or other well-known means, and preferably includes a skirt that prevents bypass flow between anchor 81 and conduit 82. Alternatively, anchor 81 may be delivered across the septum deployed, and then conduit 82 may be inserted through and deployed within anchor 81 and held in place by mechanical interference or expansion with a balloon. The advantages of such a two-part design are two-fold. First, pannus will grow thick only on the outside surface of anchor 81 because the LA and RA ends of conduit 82 are offset from, and thus do not contact, adjacent cardiac structures. Second, the design creates a longest straight channel for high velocity flow, but limits the ability of paradoxical emboli to transit conduit 82 during a transient pressure gradient reversal. The dimensional aspects noted above with respect to the description of shunt 10 of FIG. 1C above may be applied to shunt 80.

[00137] FIG. 11 illustrates another preferred embodiment with benefits similar to that of the shunt of FIG. 10. More specifically, shunt 90 may include anchor 91 as described above with the respect to frame 12 of the embodiment of FIGS. 1 A-1C. Conduit 92 may include flared end regions as described above, e.g., to form an hourglass shape in the deployed state. One of ordinary skill in the art will appreciate that the specific shape of the flared end regions may be conical, parabolic, or horned shaped, and may be present at either or both ends of the shunt device depending on the desired hydraulic properties. The dimensional aspects noted above with respect to the description of shunt 10 of FIG. 1C above may be applied to shunt 90.

[00138] The shunt types depicted in FIG 10 and FIG 11, or shunts with similar

characteristics that would be apparent to one of ordinary skill in the art, may be particularly applicable to the clinical situation where too large an aperture defect has been created in the FO and where interatrial shunting to treat heart failure is required. Consider the case of a patient with severe mitral regurgitation and poor left ventricular function, where it would be clinically desirable to first perform a repair procedure on the mitral valve, e.g. MitraClip® of mitral annuloplasty by the percutaneous transseptal approach, followed by interatrial shunt placement. These mitral valve procedures currently use a 23Fr ID. (~8 mm O.D) guiding catheter to cross the FO. After mitral repair, an anchor with an outer minimal diameter matching the larger aperture defect caused by the prior procedure may be implanted, wherein the conduit as a smaller diameter desirable for shunting (e.g. 5.0 to 6.5 mm). Likewise, such shunts advantageously may be used where, during the transseptal procedure, the fossa ovalis has been torn, thus creating a larger aperture defect than required for various shunt embodiments described with respect to FIGS. 1 to 5. Again, a shunt of the kind described with respect to FIGS. 10 or 11 could be used to address such a situation.

[00139] FIGS. 12-15 show further alternative shunt embodiments 95, 100, 110 and 120, respectively that use different shunt geometries in combination with anchors and anchoring tabs. The conduits of these shunts may be cylindrical, conical or have other lumen geometries as previously described herein. More specifically, in FIG. 12 anchor 95 suitable for use in an inventive shunt includes flared region 96 configured for deployment in the left atrium and substantially cylindrical region 97 that extends through the atrial septum and into the right atrium. Flexible struts 98 bend distally, i.e., towards the septum when the anchor is released from its delivery sheath, and preferably include U-shaped inverted ends that contact, but do not penetrate, the right atrial wall in the fully deployed position, as depicted in FIG. 12. Preferably, anchor 95, other than flexible struts 98 includes a conduit formed by encapsulating the anchor with polymeric material that prevents tissue ingrowth from obstructing the lumen of cylindrical region 97, and may be made of a biocompatible shape memory alloy, as described for preceding embodiments.

[00140] Shunt 100 of FIG. 13 may include a plurality of collapsible tab-like retention elements 101 disposed on the RA region of a cylindrical shunt. Retention elements 101 are designed to engage the FO to prevent migration/embolization of shunt 100 into the LA or beyond. With a much-thickened FO, retention elements 101 may become buried within the FO wall itself. In addition, shunt 100 may include conical anchor 102 extending at an angle into the LA from the LA side 103 of shunt 100, similar in construction to flared end region 18 of frame 12 of the embodiment of FIGS. 1A-1C. The advantage of this configuration is that it may be deployed in an FO that has any wall thickness (typically up to 10 mm). The other dimensional aspects noted above with respect to the description of shunt 10 of FIG. 1C above may be applied to shunt 100. [00141] In FIG. 14, shunt 110 is similar in construction to shunt 100 and includes retention elements 111 on the RA side, but omits conical anchor 102 on the LA side. Instead, shunt 110 may include plurality of collapsible tabs 112 on LA side 113 of the shunt designed to offset cylindrical shunt 110 from the FO or other cardiac structures. An advantage of this configuration is that there is less structure occupying the free space in the LA. The other dimensional aspects of shunt 10 of FIG. 1C above may be applied to shunt 110.

[00142] In FIG. 15, shunt 120 comprises an encapsulated expanded LA side 121, and a simple cylinder on RA side 122 that includes a plurality of retention elements 123. An advantage of this configuration is that shunt 120 may be constructed from a singular tubular frame. The other dimensional aspects of shunt 10 of FIG. 1C above may be applied to shunt 120.

[00143] Referring now to FIGS. 16A and 16B, anchor 130 of an alternative embodiment of a shunt constructed in accordance with the principles of the present invention is described. Anchor 100 is similar to anchor 12 of the embodiment of FIGS. 1 A-IC, but further includes a plurality of flexible arms 131 attached to the circumferential strut nearest the exit port in the right atrium. Flexible arms 131 self-expand when the shunt is deployed to form a meshwork or filter that partially obstructs the exit port of the shunt. In particular, upon deployment, flexible arms 131 unfold to extend across the lumen in the vicinity of the lumen of the RA exit port, ideally near the location of its widest opening, to form a filter that prevents larger paradoxical emboli from passing into the left atrium. Flexible arms 131 permit blood to pass in either direction with minimal resistance while excluding the passage of paradoxical emboli that are generally larger than the mesh size, e.g., venous thromboemboli above a certain size, which may be on a paradoxical trajectory. In this case, the size of the emboli excluded is determined by the geometry of mesh. Prior to deployment, these arms may also serve as locations of attachment of the shunt to its delivery system. While in the embodiment depicted in FIGS. 16A and 16B, flexible arms 131 comprise struts that fold across the exit port of anchor 130 upon deployment, in alternative embodiments, flexible arms 130 may take any of a number of configurations, including a plurality or multiplicity of bars or arches that fold across the exit port to create a filter. In an alternative embodiment, as already described, larger paradoxical emboli could be excluded by having a plurality of passageways or lumina through the shunt device. [00144] FIG. 17 is a graph depicting the effects of orifice size on the flow characteristics, e.g., bench testing quantified flow vs. pressure relationships, of two types of V-Wave Nitzan- type shunts as described in the above-incorporated application. Measurements were made in saline at 37 degrees Celsius, under constant pressure gradient conditions over the expected range of left-to-right pressure gradients. Flow was measured for the V-Wave 5.1 mm inner diameter orifice Nitzan-type hourglass-shaped valveless shunt and for a 6-mm inner diameter orifice valveless version of the shunt built upon the same nitinol frame. As depicted in FIG. 17, the 6-mm shunt has about 35% more flow than the 5 mm valved shunt. Also shown in FIG. 17, is the simulated flow for venturi tubes with orifice inner diameters of 5.1 and 6 mm with discharge coefficients of 0.97 and 0.96 respectively. These data suggest that the performance of the valveless hourglass shunts is closely approximated by a classical venturi. Simulations of a conical convergent nozzle (not shown) with a convergence angle of 37 and 36 degrees and a discharge coefficient of 0.92 for the 5.1 and 6 mm orifice inner diameters, respectively, showed similar predictive accuracy with actual shunts.

[00145] Referring again to FIG. 6, that figure depicts theoretical flows for a 5.1 mm and

6.0 mm venturi tube (discharge coefficient 0.97 and 0.96, respectively), as described above, along with flows through 6.4 mm and 7.5 mm orifice plates (discharge coefficient 0.61), respectively. As shown in FIG. 6, an orifice plate device requires an inner diameter of 7.5 mm to have flow characteristics similar to a 6 mm venturi tube. Similarly, an orifice plate device requires an inner diameter of about 6.4 mm to have flow characteristics similar to a

5.1 mm venturi tube. These measured data and simulations show that the valveless lumen of the hourglass-shaped V-Wave Nitzan-type shunt is more efficient at supporting bulk flow over the expected physiological range of pressure gradients than an orifice-plate shunt.

[00146] In particular, an hourglass-shaped shunt permits a smaller orifice than an orifice- plate shunt with similar bulk flow capacity (7-8 mm in diameter). The smaller orifice, in turn, prevents proportionally larger thrombi from passing retrograde through the shunt and into the systemic circulation. Since ischemic damage from the lodging of embolus is limited to the watershed organ territory supplied by the occluded vessel, larger emboli tend to cause more damage and have more associated dangerous consequences, especially when the occluding vessel supplies the brain. Thus, with a smaller orifice size, paradoxical embolic strokes, if they occur, are likely to be smaller than with an orifice-plate mesh type shunt. Accordingly, in a preferred embodiment, a shunt having a discharge coefficient of 0.70 or greater will, by virtue of its smaller diameter or area orifice, have less tendency for paradoxical embolization than an orifice-plate mesh shunt with similar flow characteristics.

[00147] Clinical studies conducted using a orifice-plate mesh shunt offered by Corvia Medical, Inc., Tewksbury, MA, indicate that a 8-mm Corvia orifice-plate mesh shunt had a Qp/Qs =1.27±0.20 at 6 months compared to 1.06±0.32 just prior to implantation. This ratio was likely higher just after implantation due to some degree of shunt narrowing as a result of pannus formation that would be expected by 6 months. By comparison, for the V-Wave Nitzan-type valved shunt with a 5 mm orifice inner diameter, Qp/Qs derived from

echo/Doppler analysis in the aforementioned patient cohort was relatively small at 1.18±0.16 shortly after implant compared to 1.04±0.22 at baseline (p<0.03). Qp/Qs decreased slightly to 1.12±0.14 by 6-12 months (p=0.10), consistent with the observed narrowing of the shunts over this same time period. These data suggest that the V-Wave Nitzan-type valved shunt, that was shown to have substantial early clinical benefit, was associated with a very small Qp/Qs ratio, and no evidence of worsening right heart failure or pulmonary hypertension. The data also suggest that a shunt of similar geometry can be made with a larger inner diameter, e.g., 6.5 mm inner diameter, without exceeding a Qp/Qs ratio of 1.5: 1.

[00148] It is appreciated that certain features of the invention, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the invention, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable subcombination.

[00149] While various illustrative embodiments of the invention are described above, it will be apparent to one skilled in the art that various changes and modifications may be made herein without departing from the invention.