BACKGROUND OF THE INVENTIONThe present invention relates to pumps usable as implantable ventricular assist devices, to components useful in such pumps, and to methods of using the same.
In certain disease states, the heart lacks sufficient pumping capacity to meet the needs of the body. This inadequacy can be alleviated by providing a mechanical pump referred to as a ventricular assist device to supplement the pumping action of the heart. It would be desirable to provide a ventricular assist device which can be implanted and which can remain in operation for months or years to keep the patient alive while the heart heals, or which can remain in operation permanently during the patient's lifetime if the heart does not heal, or which can keep the patient alive until a suitable donor heart becomes available.
Design of a ventricular assist device presents a daunting engineering challenge. The device must function reliably for the desired period of implantation. Moreover, blood is not a simple fluid, but instead is a complex system containing cells. Severe mechanical action can lead to hemolysis, or rupture of the red blood cells, with serious consequences to the patient. Also, blood in contact with an artificial surface, such as the surfaces of a pump, tends to clot. While this tendency can be suppressed to some extent by proper choice of materials, surface finishes and by administration of anticoagulants, it is still important to design the pump so that there are no regions within the device where blood can be trapped or flow is interrupted for relatively prolonged periods. To provide clinically useful assistance to the heart, the device must be capable of delivering a substantial blood flow at a pressure corresponding to normal blood pressure. For example, a ventricular assist device for an adult human patient of normal size should deliver about 1-10 liters per minute of blood at a pressure of about 70-110 mm Hg depending on the needs of the patient.
One type of ventricular assist device or pump uses a balloon. The balloon is placed within the aorta. The balloon is connected to an external pump adapted to repeatedly inflate and deflate the balloon in synchronism with the contractions of the heart muscle to assist the pumping action. Balloon assist devices of this nature have numerous limitations including limited durability and limited capacity.
As described, for example, in U.S. Pat. No. 6,688,861, a miniature electrically-powered rotary pump can be implanted surgically within the patient. Such a pump has a housing with an inlet and an outlet, and a rotor which is suspended within the housing and driven by a rotating magnetic field provided by a stator or winding disposed outside of the housing. During operation, the rotor is suspended within the housing by hydrodynamic and magnetic forces. In such a pump, the rotor may be the only moving part. Because the rotor does not contact the housing during operation, such a pump can operate without wear. Pumps according to the preferred embodiments taught in the '861 patent and related patents have sufficient pumping capacity to provide clinically useful assistance to the heart and can be small enough that they may be implanted within the heart and extend within the patient's thoracic cavity. Pumps of this nature provide numerous advantages including reliability and substantial freedom from hemolysis and thrombogenesis. However, implantation of such a pump involves a majorly invasive surgical procedure.
As described, for example, in Nash, U.S. Pat. No. 4,919,647; Siess, U.S. Pat. No. 7,011,620; and Siess et al., U.S. Pat. No. 7,027,875; as well as in International Patent Publication No. WO 2006/051023, it has been proposed to provide a ventricular assist device in the form of a rotary pump which can be implanted within the vascular system, such as within the aorta during use. Aboul-hosn et al., U.S. Pat. No. 7,022,100, proposes a rotary pump which can be placed within the aorta so that the inlet end of the pump extends through the aortic valve into the left ventricle of the heart.
A ventricular assist device implanted into the vascular system must be extraordinarily compact. For example, such a device typically should have an elongated housing or other element with a diameter or maximum dimension transverse to the direction of elongation less than about 13 mm, and most preferably about 12 mm or less. To meet this constraint, the vascularly-placed ventricular assist devices proposed heretofore resort to mechanically complex arrangements. For example, the device described in U.S. Pat. No. 7,011,620 incorporates an electric motor in an elongated housing. The motor drive shaft extends out of the housing and a seal surrounds the shaft. An impeller is mounted at the distal end of the drive shaft outside of the motor housing and within a separate tubular housing. The pump taught in U.S. Pat. No. 7,022,100 consists of a separate motor using a flexible drive shaft extending through the patient's vascular system to the impeller, with an extraordinarily complex arrangement of seals, bearings, and a circulating pressurized fluid to prevent entry of blood into the flexible shaft. The arrangement taught in WO 2006/051023 and in U.S. Pat. No. 4,919,647 also utilizes flexible shaft drives and external drive motors. These complex systems are susceptible to failure.
Thus, despite very considerable effort devoted in the art heretofore to development of ventricular assist devices, further improvement would be desirable.
SUMMARY OF THE INVENTIONOne aspect of the invention is an implantable blood pump. The pump according to this aspect of the invention includes a housing defining a bore having an axis, one or more rotors disposed within the bore, each rotor including a plurality of magnet poles, and one or more stators surrounding the bore for providing a rotating magnetic field within the bore to induce rotation of each of the one or more rotors. The one or more rotors may be constructed and arranged so that during operation of the pump the one or more rotors are suspended within the bore of the housing and out of contact with the housing solely by forces selected from the group consisting of magnetic and hydrodynamic forces. In this embodiment the pump has a maximum lateral dimension in any direction perpendicular to the axis of the bore, or a diameter of up to about 20 mm. In one embodiment the diameter of the bore is about 14 mm. In another embodiment the diameter of the bore is between 9 and 11 mm. The pump of the present invention can impel from about 1-3 liters of blood per minute. In one embodiment the pump is adapted to impel about 2 liters of blood per minute. Blood pressure can be maintained within the range of from 70-120 mm Hg between the inlet and outlet. The pump is adapted for positioning within an artery, and may include a gripper adapted to engage the wall of an artery.
Another aspect of the invention includes methods of providing cardiac assistance to a mammalian subject as, for example, a human. Methods according to this aspect of the invention include advancing a pump including a housing having a bore, one or more rotors disposed within the bore and one or more motor stators disposed outside of the housing through the vascular system of the subject until the pump is disposed at an operative position at least partially within an artery of the subject, and securing the pump at the operative position. The method includes the step of actuating the pump to spin the one or more rotors and pump blood distally within the artery solely by applying electrical currents to the one or more motor stators and to suspend the one or more rotors within the bore solely by forces selected from the group consisting of magnetic and hydrodynamic forces applied to the one or more rotors.
Still further aspects of the present invention include rotor bodies having helical channels formed longitudinally along the length of the body of the rotor. Each helical channel is formed between peripheral support surface areas facing substantially radially outwardly and extending generally in circumferential directions around the rotational axis of the rotor. Each channel has a generally axial downstream portion. The helical and axial portions of each of the channels cooperatively define one or more continuous flow paths extending between the upstream and downstream ends of the rotor. In one embodiment, the axial regions of the channels have greater aggregate cross-sectional area than the one or more passages. The support surfaces of the rotor body are formed on a plurality of lobes. Each lobe has a circumferential extent which increases in a radially outward direction away from the rotational axis of the rotor. The support surfaces face generally radially outwardly away from the rotor axis and define hydrodynamic bearing surfaces. The circumferential extent of the support surfaces is greater than the circumferential extent of peripheral surface areas.
BRIEF DESCRIPTION OF THE DRAWINGSFIG. 1 is a diagrammatic perspective view of a pump in accordance with one embodiment of the present invention, with components omitted for clarity of illustration.
FIG. 2 is a diagrammatic view of components used in the pump ofFIG. 1, with certain components depicted as transparent for clarity of illustration.
FIG. 3 is a partial cut-away view of the pump depicted inFIGS. 1 and 2.
FIG. 4 is a fragmentary sectional view depicting a portion of the pump shown inFIGS. 1-3.
FIG. 5 is a further fragmentary sectional view depicting another portion of the pump shown inFIGS. 1-3.
FIGS. 6 and 7 are perspective views depicting a rotor used in the pump ofFIGS. 1-5.
FIGS. 8 and 9 are elevational views of the rotor shown inFIGS. 6 and 7.
FIGS. 10-14 are sectional views taken along frames10-14 respectively inFIG. 9.
FIG. 15 is a fragmentary, diagrammatic sectional view depicting a portion of the rotor depicted inFIGS. 6-14 in conjunction with another component of the pump.
FIG. 16 is a diagrammatic perspective view depicting another rotor utilized in the pump ofFIG. 1.
FIG. 17 is a diagrammatic elevational view of the rotor shown inFIG. 16.
FIG. 18 is a diagrammatic perspective view of the pump depicted inFIGS. 1-17 in conjunction with a further component.
FIG. 19 is a partially blocked diagrammatic view depicting the pump ofFIGS. 1-17 in operating position in the cardiovascular system of a subject.
FIGS. 20 and 21 are diagrammatic perspective views depicting components used in further embodiments of the invention.
FIG. 22 is a diagrammatic perspective view of a rotor according to a further embodiment of the invention.
FIG. 23 is a diagrammatic sectional view taking along line23-23 inFIG. 22.
FIG. 24 is a diagrammatic elevational view depicting a rotor according to yet another embodiment of the invention.
FIG. 25 is a diagrammatic elevational view depicting a component of a stator used in a pump ofFIGS. 1-17.
FIG. 26 is an electrical schematic diagram of the stator used in the pump ofFIGS. 1-17.
DETAILED DESCRIPTIONApump10 in accordance with one embodiment of the invention includes a housing12 (FIGS. 1,2 and3).Housing12 is a ceramic tube defining acentral bore14 having anaxis16.Bore14 is cylindrical and has a constant diameter over the major portion of its length. Theinterior surface13 of the housing defining bore14 is smooth, and desirably has a surface roughness on the order of 4 micro inches rms or less. Merely by way of example, the inside diameter ofbore14 in this constant diameter region may be about 0.178 inches (“in”), and the wall thickness of the housing may be about 0.010 in.Housing12 defines aninlet18 at anend20 of the housing, referred to herein as the inlet or upstream end, and anoutlet22 communicating withbore14 at an output ordownstream end24 of the housing. The inside diameter ofinlet18 is slightly less than the inside diameter ofbore14. The housing includes aninlet transition section26 having an inside diameter which increases progressively in the downstream direction at the juncture betweeninlet18 and bore14. The inside diameter of the housing increases progressively at anoutlet transition section23 immediately upstream fromoutlet22.
As best seen inFIG. 4, a thin-walledmetallic tube28 is fitted over the inlet or upstream end of the housing so that the interior oftube28 communicates withinlet opening18.Tube28 may be formed from a metal such as titanium, titanium alloy, or platinum, and may be brazed to the ceramic housing. An upstream end fitting30 surroundstube28, and also surrounds theupstream end20 of theceramic housing12. Aflexible intake tube32 surrounds the upstream end of fitting30 andtube28, and is held in place by acrimp metal band34. As best seen inFIG. 1,intake tube32 extends upstream from fitting30 and fromhousing12, and terminates at acastellated opening36 at its upstream end. As best seen inFIG. 4, the interior oftube32 communicates withinlet18 ofhousing12, and thus with thebore14 of the housing, throughtube28. In one embodiment,intake tube32 is formed from a non-thrombogenic flexible polymer such as, for example, a fluoropolymer, polydimethylsiloxane, silicone polycarbonate urethanes, thermoplastic polyurethanes, polycarbonate urethanes, segmented polyurethanes, poly(styrene-b-isobutylene-b-styrene, or sulfonated styrene containing copolymers.
A metallic outflow tube40 (FIG. 5) surrounds thedownstream end24 ofhousing12 and communicates with thebore14 of the housing.Outflow tube40 may be formed from materials as discussed above with respect totube28. The downstream end ofoutflow tube40 defines the outlet41 of thepump10.
A downstream end fitting42 surroundsoutflow tube40 and thedownstream end24 ofhousing12. An elongatedelectrical cable44, of which only a portion is shown inFIGS. 1 and 5, is secured to downstream end fitting42. As best seen inFIG. 1, the downstream end fitting42 carries several miniatureelectrical feedthroughs46, which are electrically isolated from one another and which are connected to the individual conductors ofcable44.
Afirst stator48 surroundshousing12 adjacent the upstream end thereof, shown inFIG. 25. Thestator48 includes a magneticallypermeable frame700. The frame is configured as a cylindrical,tubular ring702 having a plurality ofpoles704 projecting inwardly from thering702 to the exterior ofhousing12; six poles are used in this particular embodiment depicted.Ring portion702 is concentric withhousing12 and boreaxis16. Each pole includes a widened portion at the tip of the pole, where the pole confronts the exterior surface ofhousing12. The poles define six slots,706-1 through706-6, between them. The dimensions of the stator are substantially uniform along the axial length of the stator. In this embodiment the stator is formed from numerous uniform laminations stacked on one another. The laminations are formed from a magnetically permeable material selected to minimize power losses due to magnetic hysteresis. For example, the laminations may be formed from 29-gauge silicon steel of the type sold under the designation M15 electrical steel.
In the particular embodiment depicted, the exterior diameter OD ofring portion702 is about 0.395 inches, and the interior diameter ID of the ring portion is about 0.304 inches. The width or circumferential extent PW of each pole is about 0.035 inches at its juncture with thering portion702. The interior diameter PD between opposed pole tips may be about 0.221 inches. The axial length of the frame is selected according to desired output power, and may be, for example, about 0.35 inches for about 1 watt output to about 0.85 inches for about 3 watts output.
Stator48 further includescoils710,712, and714 shown in electrical schematic inFIG. 26. In the particular embodiment depicted, each coil includes about 11 to about 14 turns of 33- or 34-gauge insulated wire, and may be impregnated with a material such as a varnish after winding. The coils are connected in a WYE configuration to a common neutral718.Coils710,712, and714 are disposed respectively in slots706-1 through706-6 of frame700 (FIG. 25).Coil710 is wound through slots706-1 and706-4, whereascoil712 is wound through slots706-3 and706-6, andcoil714 is wound through slots706-5 and706-2. This arrangement is commonly referred to as a balanced three-phase integral slot winding. The ends of the coils remote from theneutral point718 are connected to inputs P1, P2, and P3. When these inputs are energized with three sinusoidal voltages offset from one another in phase by 60°, the coils provide a magnetic field within the bore which is directed transverse to thebore axis16 and which rotates around the axis, within a first region50 (FIGS. 1 and 2) of the bore. Asecond stator52 is disposed downstream fromfirst stator48, and is arranged to apply a rotating magnetic field within a second region54 ofbore14 downstream from the first region. The second stator may be similar to the first stator.Pump10 further includes acasing47, depicted in broken lines inFIG. 1, extending between theoutflow fitting42 and the inflow fitting30, and covering thestators48 and52. A potting material (not shown) fills space withincasing47 around the stators.
A first orupstream rotor56 is disposed withinbore14 adjacent the upstream end of bore, within region thefirst region50. Asecond rotor58 is disposed within the bore downstream fromfirst rotor56, within region54.
Thefirst rotor56, shown in FIGS.2 and6-15, is formed as a solid, unitary body of a ferromagnetic, biocompatible material, such as an alloy including platinum and cobalt, as, for example, an alloy consisting essentially of platinum and cobalt such as 77.3% Pt and 22.7% Co. The rotor has an upstream orinlet end60, a downstream oroutlet end62, and arotational axis64 depicted in dotted line inFIG. 6. The rotor includes a unitarycentral shaft portion66 immediately surrounding the axis and coaxial therewith, extending throughout the length of the rotor. The central shaft portion has a generally spherical dome67 at its upstream end and a conical, tapered region65 at thedownstream end62.
The body ofrotor56 is described herein with reference toaxis64. As used herein with reference to a structure such as rotor having upstream and downstream ends and an axis, the upstream direction is the direction parallel to the axis toward the upstream end, whereas the downstream direction is the opposite direction. A “radial” direction is a direction outwardly, away from the axis. A “circumferential” direction is a direction around an arc in a plane perpendicular toaxis64. The “forward” circumferential direction indicated by one end F of the arrow FR inFIG. 6 corresponds to the direction of rotation ofrotor56 aboutaxis64 in service. The opposite circumferential direction indicated by one end R of the arrow FR (FIG. 7) is referred to herein as the reverse circumferential direction.
Aregion68 of the rotor adjacent theupstream end60, referred to herein as the “helix” region,rotor56 hashelical channels74 and76 defining a pair of raisedperipheral surface areas70 and72 radially outwardly from thecentral shaft66. Thechannel74 defines asurface area78 facing in the forward circumferential direction, referred to herein as the pressure surface or, alternatively referred to as the leading surface.Pressure surface78 is a helical surface of variable pitch along the axial length. The pressure surface has a pitch angle A (FIG. 8) of about 60°. As used in this disclosure with reference to a helical surface, the term “pitch angle” refers to the angle between the axis of revolution,axis64, and a line tangent to the surface. Thechannel76 defines asurface area80, facing in the rearward circumferential direction, referred to herein as the suction surface or alternatively referred to as the trailing surface.Suction surface80 is also helical, but has a slightly larger pitch angle thanpressure surface78, so that the thickness or circumferential extent of raisedsurface area70 increases progressively in the downstream direction (to the right as seen inFIG. 9). Thesuction surface80 is truncated by a smallflat surface83 in a plane perpendicular to theaxis64 at the upstream end of the helix region, thereby defining a sharp, radially-extensive edge85, having a radius of about 0.003 inches, at the upstream end of the helix region. The raisedperipheral surface area70 is arcuate and of constant radius aboutaxis64. Thesurface area70 extends through about 130° of arc aboutaxis64 from its upstream edge to its downstream edge.
Thesurface area72 is identical tosurface area70, and is offset fromsurface area70 by 180° aboutaxis64. As best appreciated with reference to the cross-sectional view ofFIG. 10, thesurface areas70 and72 are thin. The circumferential extent ofsurface area70 is about 15° of arc aroundaxis64, and thus the aggregate circumferential extent ofsurface areas70 and72 amounts to about 30° of arc.
As used in this disclosure, the term “major diameter” of a body having an axis refers to the dimension which is twice the greatest radius from the axis to any point on the body in a particular plane perpendicular to the axis. Forrotor56, the major diameter is simply the length of a line extending between theperipheral surfaces70 and72 throughaxis64. As used in this disclosure, the term “solidity” refers to the ratio between the cross-sectional area of the solid features of the body to the area of a circle having a diameter equal to the major diameter of the body. The solidity of thehelix portion68 is in the range of about 10-20% at the upstream or inlet end of the body. In one embodiment the solidity is in the range of about 10-15%, and about 14%, and increases progressively to about 15-25% at the downstream end of the helix region, to about 18-23%. In one embodiment the solidity is about 20%. Stated another way, the helix region is largely open for entry of blood at its upstream end.
Therotor56 further includes a support region88 (FIG. 6) disposed downstream from thehelix region68. The rotor has afirst lobe90 andsecond lobe92 projecting outwardly fromcentral shaft66 in the support region.First lobe90 has a pressure surface94 (FIG. 9), also referred to as a leading surface, facing in the forward circumferential direction.Pressure surface94 is continuous with thepressure surface78.First lobe90 also has a suction surface96 (FIG. 6), also referred to as a trailing surface.Suction surface96 is continuous with thesuction surface80. Thus, the periphery of thefirst lobe90 constitutes a continuation of theperipheral surface area70 in the downstream direction. The first lobe also has a surface98 (FIG. 13) facing generally radially outwardly away fromaxis64, this surface being referred to herein as a “support” surface. Theopposite lobe92 has asimilar pressure surface99 continuous with the pressure surface atarea72, and suction surface100 (FIG. 9) continuous with the suction surface atarea72.Lobes90 and92 are diametrically opposite to one another and definepassages106 and108 between them.Passage106 is continuous withchannel74, whereaspassage108 is continuous withchannel76. The passages extend to the downstream end of the body and are open at the downstream end, so that the channels of the helix region and the passages of the support region cooperatively define continuous flow paths extending between the upstream and downstream ends of the body. The pressure and suction surfaces of the lobes have substantially constant pitch angle, and the pitch angles of the pressure and suction surfaces are substantially equal to one another, so that the circumferential extent of each lobe remains substantially constant throughout thesupport region88. The pitch angle of the pressure and suction surfaces of the lobes are substantially smaller than the pitch angle of the pressure and suction surfaces in the helix regions. For example, the pitch angles of the lobe pressure and suction surfaces may be on the order of about 10°.
The major diameter of the support section defined by the lobes is equal to the major diameter of helix section. However, as best appreciated by comparison ofFIG. 13 withFIG. 10, the circumferential extent of support surfaces98 and104 of the lobes is much greater than the circumferential extent of the peripheral edge surfaces82 and84 of the helix regions. For example, each support surface may have a circumferential extent of about 90-110° of arc aboutaxis64, so that the aggregate circumferential extent of the support surfaces is about 180-220°. Moreover, the solidity of the support region including the lobes is substantially greater than the solidity of the helix region. The solidity of the support region may be about 30-40%.
As also apparent fromFIGS. 13 and 14, thepressure surface94 andsuction surface96 oflobe90 diverge from one another in the radially outward direction, away fromaxis64. Similarly, thepressure surface99 andsuction surface100 oflobe92 diverge from one another in the radially outward direction. Stated another way, the circumferential extent of each lobe increases progressively in the radially outward direction, so that the mass of each lobe is concentrated in the region of the lobe remote fromaxis64.
Support surface104 oflobe92 has a trailing land area110 (FIGS. 8 and 9) disposed at the same radius fromaxis64 as the peripheral surfaces of the helix regions. The trailingland area110 extends along the trailing or suction edge ofsupport surface104, i.e., the edge of the support surface at its juncture with thesuction surface100 oflobe92.Land area110 merges into theperipheral surface84 ofhelix area72, as best seen inFIG. 8. Thesupport surface104 further includes a first or upstream hydrodynamic-bearingsurface112 and a separatingland114 extending in the forward circumferential direction from trailingedge land110 to the forward edge of the support surface, at its juncture withpressure surface99.Land114 lies at the same radius fromaxis64 as the trailingedge land110. Thus, bearingsurface112 is bounded on its trailing and upstream sides by trailingedge land surface110 andperipheral surface84, and on its downstream side by separatingland114. As best seen inFIG. 15, bearingsurface112 is disposed radially inwardly from the land surfaces and slopes radially outwardly towards its trailing edge, i.e., in the reverse circumferential direction toward trailingland surface110. Stated another way, the land surfaces define a generally cylindrical surface at the major diameter, and bearingsurface112 defines a depression in this generally cylindrical surface which tapers to a decreasing depth in the reverse circumferential direction. As also shown inFIG. 15, the major diameter of the rotor defined byland surfaces110 is just slightly less than the internal diameter ofbore14 in the housing. For example, the internal diameter of the bore may be about 0.002 inches larger than the major diameter of the rotor.
Support surface104 further includes a second ordownstream bearing surface116 immediately downstream from separatingland114, and a downstreamend land surface118 immediately downstream of the bearingsurface116.Bearing surface116 is configured in the same way as bearingsurface112, and forms a similar depression in the cylindrical outer surface tapering to a progressively decreasing depth in the reverse circumferential direction, toward the trailingedge land surface110.
All of the surfaces ofrotor56 are smooth, desirably to a surface roughness of about 4 micro inches or less.Rotor56 may be formed, for example, by machining from a solid rod and polishing using techniques such as electropolishing and drag polishing.Rotor56 has a permanent magnetization with a flux direction transverse toaxis64, so thatlobe92 forms one pole of a permanent magnet, wherelobe90 forms the opposite pole.
Rotor56 may have an axial length, from the upstream edges of the helix areas to the downstream end of the lobes of about 0.5-0.95 inches, preferably 0.6 inches long. The helix region may be about 0.15-0.25 inches long, preferably about 0.2 inches long whereas the support region may be about 0.35-0.45 inches long in the axial direction, preferably about 0.4 inches long. The ratio between the length of support region and the length of the helix region is about 1:1 to 3:1, preferably about 2:1.
The second rotor58 (FIGS. 16 and 17) is similar to thefirst rotor56 discussed above. Thus, the second rotor includes a helix region defined bychannels174 and176 adjacent the upstream or inlet end160 of the rotor. The second rotor haslobes190 and192 defining passages between them, the passages between the lobes being continuous with the channels in the helix region.Lobes190 and192 are configured in substantially the same way as the lobes of the first rotor discussed above with reference toFIG. 13. Thus, in this rotor as well, each lobe has a circumferential extent which increases in the radially outward direction so as to provide a support surface having a substantial circumferential extent. Here again, the solidity of the rotor in the support region occupied by the lobes is substantially greater than the solidity of the rotor in the helix region.
Second rotor58 has a pitch opposite to the pitch of the first rotor. The forward circumferential direction F′ ofsecond rotor58 is the clockwise direction of rotation aboutaxis164 as seen from the upstream end160 of the rotor, whereas the forward circumferential direction of the first rotor56 (FIG. 6) is the counterclockwise direction as seen from theupstream end60 of the rotor. Also, thesecond rotor58 is substantially shorter in the axial direction than the first rotor. The axial length of the second rotor may be about 0.3-0.5 inches. In one embodiment the axial length is 0.4 inches. Of this length, approximately 0.15 inches is occupied by the helix region, and approximately 0.25 is occupied by the support section consisting of thelobes190,192. The pitch angle A′ (FIG. 17) of the channel surfaces defining the helix region is substantially greater than the pitch angle A (FIG. 8) of the corresponding surfaces in the first rotor. Here again, each channel surface extends helically aroundaxis64 by about 130° from the upstream end to its juncture with the associated lobe.
As best seen inFIG. 18, pump10 desirably is provided with an expansible gripper adapted to engage the interior surface of an artery. Unless otherwise stated, dimensions of the pump referred to herein exclude the gripper. The gripper depicted is astent200 which includes a thin-walledcylindrical shell204 having numerous perforations extending through it.Stent200 also includes acentral collar206 and threelegs208 which extend from the collar to the downstream edge ofshell204, i.e., the edge of the shell facing upwardly inFIG. 18.Collar206 is mounted on the outflow end fitting42 ofpump10 at the downstream or outlet end of the pump.Power cable44 projects downstream through the interior ofshell204.
In the expanded condition depicted inFIG. 18,stent204 is spaced radially outwardly from the pump.Stent204 has a collapsed condition (not shown) in which the stent is disposed downstream from the outlet fitting42, withlegs208 extending generally parallel to the axis of the pump and downstream from the pump. In this collapsed condition,stent204 has an exterior diameter approximately equal to or smaller than the exterior diameter ofpump10, i.e., about 13 mm or less. In one embodiment the stent diameter is about 12 mm.
In oneembodiment stent200 is formed from a shape-memory alloy such as the alloy sold under the registered trademark Nitinol™. The stent is initially provided in the collapsed condition, and is arranged to return spontaneously to the expanded condition when the stent is left unconstrained and heated to body temperature.
In operation, pump10 andstent200 are inserted into the patient's vascular system as, for example, into the femoral artery or another artery having good access to the desired placement site, and advanced through the vascular system, with theintake tube32 leading, until the pump is in the desired location. As shown inFIG. 19, the pump may be placed within the patient's aorta, withintake tube32 extending through theaortic valve220 of the subject's heart, and with theupstream end36 of the inflow tube protruding into the left ventricle. In this position,power cable44 extends through the aorta.
The step of advancing the pump through the vascular system may be performed using generally conventional techniques for placement of intra-arterial devices. For example, an introducer catheter may be placed using a guidewire; the guidewire may be removed, and then the pump may be advanced through the introducer catheter, whereupon the introducer catheter is removed. Alternatively, the pump may be provided with fittings suitable for engaging the guidewire. For example, the stent itself may serve as one such fitting at the downstream or output end of the pump, whereas theinflow tube32 may be provided with a hole (not shown) extending through its wall adjacent itsupstream end36, so that the guidewire is threaded through the interior of the stent and through the hole in the intake tube. In this case, the pump is advanced over the guidewire without using an introducer catheter.
Before or after placement of the pump, the end ofpower cable44 remote from the pump is connected to acontrol unit222. Thecontrol unit222, in turn, is connected to astorage battery224. The control unit and battery may be provided as a unitary device in a common implantable housing.Control unit222 is electrically connected throughcable44 to thestators48 and52 of pump10 (FIGS. 1 and 3), and is arranged to apply appropriate excitation currents to these stators to provide rotating magnetic fields as discussed below. The control unit senses the voltage on the stators and thus detect back EMF generated by rotation of the rotors. The control unit controls the excitations of the stators so as to maintain the rotors at a predetermined speed of rotation.Battery224 is a rechargeable battery, and is connected to atransdermal power connection226Control unit222. Thebattery224 may be implanted in a location within the subject's body outside of the vascular system. The transdermal power connection may include a coil adapted to draw energy from a magnetic field applied through the skin, or may include a fitting extending through the patient's skin and carrying connectors adapted for conductive connection to a power source.
With the pump in place and secured,controller222 actuates the first or upstream stator to apply magnetic flux withinregion50 of housing bore14 (FIG. 2). The flux direction is transverse to theaxis16 of the housing, and hence transverse to the axis of the first orupstream stator48. The controller varies the direction of the flux progressively, so that the magnetic field direction rotates aboutaxis16 in the forward direction offirst rotor56. Because of its permanent magnetization,first rotor56 aligns itself with the flux direction, and thus spins about its axis at the same rate as the flux. Thelobes90 and92 provide a strong permanent magnet. Moreover, the magnetic material of the lobes is concentrated near the outside of the stator, in close proximity to the wall ofbore14, and thus in close proximity tostator48. This provides a strong magnetic interaction between the stator and the rotor. In one embodiment the field and rotor rotate within the range of about 20,000-60,000 revolutions per minute (rpm), most typically about 50,000 rpm.
As the rotor spins about its axis, the bearing surfaces on the lobes advance with the rotor in the forward circumferential direction. As best appreciated with reference toFIG. 15, there is a relatively large clearance (about 0.004 inches) between the interior surface oftube14 andbearing surface112 at the forward edge, where the surface intersects forward surface99 of the lobe. There is a smaller clearance at the rearward edge of the bearing surface, nearland110, where the bearing surface transitions smoothly into theland110. There is an even smaller clearance between theland110 and the wall of the housing. Thus, as the bearing surface advances in the forward direction, a high hydrodynamic pressure is created at the rearward portion of the bearing surface. The same is true for the downstream bearing surface116 (FIGS. 8 and 9), and for the bearing surfaces of theopposite lobe90. The hydrodynamic pressures keep the rotor centered in the bore and out of contact with the wall of the bore. Theland portion114 between bearing surfaces forms a barrier to axial flow of blood between theupstream bearing surface112 and thedownstream bearing surface116. The same is true for the bearing surfaces of the opposite lobe. This helps to assure that the bearing surfaces provide independent separating forces at axially spaced locations along the support region of the rotor, so that the rotor resists pitch or yaw of the rotor axis relative to the central axis of the bore.
The magnetic field applied bystator48 maintains the rotor in axial alignment with the stator, and prevents the rotor from moving axially within the bore. Thus, during operation, the rotor is suspended within the bore by the hydrodynamic and magnetic forces applied to it, and is entirely out of contact with any solid element of the pump. The rotor thus operates with no wear on the rotor or the housing.
As the first rotor spins, the leading and suction surfaces of thechannels74 and76 (FIG. 6) of the first rotor impinge on blood present within thebore14 of the housing, and impel the blood downstream. The relatively low solidity provided by the helix region promotes inflow of blood into the flow channels and thus helps to provide effective pumping action. Thelobes90 and92 provide relatively little pumping action. However, thepassages106 and108 between the lobes provide a relatively low-resistance flow path from the channels in the helix region68 (FIG. 6) to the downstream end of the first rotor. As discussed above, thesupport region88 andlobes90 and92 serve to provide support for the rotor within the bore and to provide an effective drive action. Surprisingly, it has been found that varying the configuration and solidity of the rotors along their axial extent, so that the helical region exhibits relatively low solidity and relatively small peripheral surfaces and the lobes have relatively high solidity and substantial circumferential surfaces, provides a particularly good combination of pumping action with adequate support and adequate magnetic linkage to the rotating flux of the stator.
As theupstream rotor56 spins about its axis, viscous drag exerted by the rotor and the blood entrained therewith on the blood immediately upstream of the rotor withinbore14 tends to impart a swirling or rotational motion to the blood upstream from the rotor, so that the blood approaching the rotor is already spinning in the forward direction of the rotor. In theory, this effect tends to reduce the pumping action imparted by the rotor. This effect could be mitigated by providing fixed axial vanes inside the bore just upstream from the rotor. However, it is believed that a significant advantage is obtained by omitting such vanes, so that the bore immediately upstream from the rotor is an unobstructed surface of revolution about thecentral axis16, with no obstruction to swirling flow. In one embodiment, the unobstructed bore extends upstream from the rotor for at least about 2 times the bore diameter. Leaving the bore unobstructed in this manner provides a gentler action at the upstream end of the rotor and thus tends to reduce hemolysis. Stated another way, limitations on rotor speed which may be imposed by hemolysis considerations are relaxed by providing such an unobstructed bore upstream from the rotor.
All of the surfaces of the rotor and the interior surface of the housing in the vicinity of the rotor are continually washed by flowing blood, so that there is no stasis or pooling of blood. This substantially mitigates the risk of thrombus formation. Moreover, because the rotor operates without wear on the rotor or the housing, the surfaces of the rotor and housing remain smooth, which further reduces thrombogenesis. The rotors constitute the only parts of the pump which move during operation. As the rotors are maintained out of contact with other parts of the pump, the pump has no moving parts which contact one another during operation. In particular, the pump has no seals which contact moving parts during operation. A pump without such seals can be referred to as a “seal-less” pump.
Becauserotor56 is a simple, two-pole magnet, the stator need provide only two flux reversals per revolution. Each flux reversal requires that the control unit and battery overcome the inductive impedance of the stator, and each flux reversal consumes power in hysteresis of the ferromagnetic material in the stator. Accordingly, the frequency of motor commutation required for a given rotational speed is lower for a two-pole rotor than for rotors with a greater number of poles.
The second ordownstream rotor58 operates in substantially the same manner as thefirst rotor56, and is suspended withinbore14 ofhousing12 by a similar combination of magnetic and hydrodynamic forces. The second rotor spins in the opposite direction from the first rotor. The blood passing downstream from theupstream rotor56 has a swirling motion in the forward or rotational direction of the first rotor, i.e., in the direction opposite to the direction of rotation of the second rotor. In the particular embodiment illustrated, the downstream or second rotor provides approximately a third of the pumping work performed on the blood passing through the pump, whereas the upstream rotor provides approximately two-thirds of the pumping work. As the magnetic fields associated with each stator apply torque to the rotors to turn the rotors about their axes, an equal but opposite torque is applied to the stators. Because the rotors turn in opposite directions, these reaction torques applied to the two stators tend to counteract one another, and thus reduce the torque load applied to stent200 (FIG. 19).
Pump10 in the embodiment described can pump approximately 3 liters per minute against a pressure differential of approximately 100 mm Hg, a typical physiological blood pressure. The pump thus provides substantial assistance to the pumping action applied by the left ventricle of the heart. Moreover, the pump provides this effective pumping assistance in a device that is small enough to be implanted in the aorta using a minimally invasive procedure, and which can operate for extended periods without wear or mechanical failure.
The leaflets of the patient's aortic valve220 (FIG. 19) seal against the exterior surface ofinflow tube32, and thus prevent backflow of blood into the left ventricle during diastole. When the ventricle contracts, during systole, blood pumped by the heart flows through the aortic valve around the inflow tube and the pump. The pump and the stent do not occlude the aorta, and do not prevent the heart from exerting its normal pumping action. Thus, in the unlikely event of a pump failure, the patient's heart can continue to provide some blood circulation. Depending upon the patient's condition, this circulation may be adequate to sustain life until corrective action can be taken.
Numerous variations and combinations of the features described above can be utilized without departing from the present invention. For example, the second or downstream rotor and the corresponding stator may be omitted to provide a smaller pump with somewhat lesser pumping capacity. Conversely, three or more rotors may be utilized. The dimensions and proportions discussed above can be varied. For example, the housing, rotors and stators can be made with a substantially larger diameter to provide more pumping capacity in a pump which is to be implanted surgically, as for example, by connection through the apex of the heart.
The pump can be positioned in other locations. In one such variant, the intake tube is omitted and the pump is positioned proximally from the position illustrated inFIG. 19, with the inlet end of the pump housing itself protruding through the aortic valve. In yet another variant, the pump may be positioned in the descending aorta, in the femoral artery or in another artery, so as to provide localized circulatory assistance. The pump also may be implanted into a pulmonary artery to provide assistance to the right ventricle.
Apump310 in accordance with a further embodiment (FIG. 20) includes agripper300 having acollar302 attached to the outflow fitting or shroud of the pump adjacent the downstream or outlet end of the pump, and a set offingers304 attached tocollar302 at locations spaced apart from one another circumferentially around the axis of the pump. Each finger has atip306 remote from the collar and a crown section having a relatively large circumferential extent disposed near the tip. Each finger also includes abeam section311 between thecrown section308 andcollar302. In the expanded condition illustrated inFIG. 20, thebeam sections311 project outwardly from thecollar302, and the crown sections have a curvature so that thetips306 point inwardly toward the axis of the pump. Thecrown sections308 bear on the interior wall of an artery (not shown), whereas thetips306 are maintained out of contact with the artery wall. Thebeam sections311 have relatively low resistance to flow of blood in the axial direction.Fingers304 may be formed from a shape memory alloy as discussed above, and have a collapsed condition in which they project axially and thus lie flat against the exterior surfaces ofpump310. In a variant, thetips306 may project inwardly toward one another at the upstream or inlet end of the pump to facilitate movement of the pump along the artery during placement.
Referring toFIG. 21, apump410 according to yet another embodiment of the invention includes a gripper in the form of twostrips402 and406. In the collapsed condition,strip406 is wound in a helix, closely overlying the exterior surface of the pump. The downstream end of thestrip406, closest to thedownstream end412 of the pump, is affixed to the pump body. Theupstream end414 is free, and is secured to the pump only by the other turns of the helix. Thesecond strip402 has itsupstream end409 affixed to the pump body, and itsdownstream end408 free to move relative to the pump body, constrained only by the remaining turns of the helix. When the pump is implanted,strip402 assumes the expanded position shown at402′. In this condition,strip402 is generally in the form of a spiral extending clockwise about the axis ofpump410, as seen from thedownstream end412 of the pump.Strip406 assumes the shape shown at406′ and approximates a spiral with the opposite direction fromspiral402′, i.e., extending counter-clockwise from its juncture with the pump body to itsfree end414, again as seen from thedownstream end412 of the pump.
Grippers described herein can be configured at intervals along a driveline that extends from the pump to a battery or a controller. Such driveline is downstream of the pump and a configuration of grippers along its length can be used to maintain driveline position in a main path of blood flow and away from arterial walls. The combination of gripper support of the pump and gripper support of the driveline eases the removability of the pump if, for example, repair is needed or the pump is no longer needed by the patient. Driveline gripper supports may be attached to the driveline in intervals of roughly 0.23-0.46 inches along the length of the driveline.
As discussed above, the proportions of the rotors can be varied. More than two lobes and helix sections may be employed. In one embodiment, the number of lobes is equal to the number of helix sections. However, other configurations can be employed. Also, the rotors discussed above have the same major diameter over the length of the pump body so that, considering the major diameter only, the rotor is generally cylindrical. This also is not essential. For example, a rotor may have a helix section with a greater major diameter than the support section. Such a rotor may be used with a tapered housing. The stator may surround only the region of the housing which receives the support region, so that the pump as a whole has a small diameter.
Merely by way of example, a rotor according to a further embodiment of the invention (FIGS. 22 and 23) includes ahelix section568 as discussed above, but includes asupport section588 in the form of a hollow tubular shell having a singleinterior bore502, best seen inFIG. 23, which depicts a cross section view looking from the downstream end of the rotor. The shell also defines a support surface in the form of a cylinder with a full 360° extent around theaxis564 of the rotor. Thechannels574 and576 of the helix section communicate withpassage502 of the support section throughports506 at the juncture between the support section and the helix section. The particular configuration of theports506 inFIGS. 22 and 23 is schematic. Desirably, the ports would have surfaces merging smoothly with the channels defining the helix section. The cylindrical support surface can act as a hydrodynamic bearing surface. To assure continuous washing of thesupport surface504 and the interior bore of the housing, holes508 may be provided through the wall of the tubular support section.
Referring toFIG. 24, arotor600 according to a further embodiment of the invention includes ahelix region602 withperipheral surface areas606 andchannels608 similar to those discussed above. The rotor also includes asupport region604. In this embodiment, the support region includes anupstream portion610, adownstream portion620 spaced-apart from the upstream portion in the axial direction and ashaft630 extending between these portions.Upstream portion610 haslobes614 andchannels616 between the lobes. Here again,lobes614 have radially outwardly facing support surfaces which incorporate hydrodynamic bearing surfaces618. Thedownstream portion620 haslobes624 andpassages626 between the lobes.Lobes624 define further hydrodynamic bearing surfaces628.Shaft630 has a relatively small cross-sectional area.Passages616 and626 communicate with the space surrounding the shaft. The small cross-sectional area of the shaft provides a relatively low resistance to axial fluid flow. The axial spacing of the upstream and downstream portions from one another provides a greater axial distance between the hydrodynamic bearing surfaces, and thus provides greater resistance to yaw or tilting of the rotor in directions transverse to the central axis of the housing.
In the embodiments discussed above, each rotor is formed entirely as a unitary body of a single ferromagnetic material. However, this is not essential. For example, the rotor could be formed from a ferromagnetic material such as iron or an iron-nickel alloy, which has desirable ferromagnetic properties, but which is far less compatible with the blood. The rotor may be plated with a metal having desirable blood compatibility of properties such as platinum, with or without one or more intermediate plating layers. In yet another variant, the helix section of each rotor may be formed from a non-ferromagnetic material which is bonded to a support section incorporating a ferromagnetic material.
As these and other variations and combinations of the features discussed above can be utilized, the foregoing description of the preferred embodiments should be taken by way of illustration rather than by way of limitation of the invention as defined by the claims.