US20110029084A1

US20110029084A1 - Foam prosthesis for spinal disc

Info

Publication number: US20110029084A1
Application number: US12/599,877
Authority: US
Inventors: Michael T. Milbocker; Jeffrey A. Wilson; Robert M. Arcangeli
Original assignee: Promethean Surgical Devices LLC
Current assignee: Promethean Surgical Devices LLC
Priority date: 2007-05-14
Filing date: 2008-05-14
Publication date: 2011-02-03
Also published as: EP2155123A1; WO2008141332A1; US8869802B2; US20100076486A1; US20090024107A1; US20110306982A1

Abstract

Disclosed herein are spinal disc implants comprising a foam adapted to completely or partially replace a nucleus pulposus within a spinal disc cavity, the foam being a nonabsorbable, closed cell and having a Poisson ratio of less than 0.5. Also disclosed are methods of implanting a foam, either as an in-situ curable material or as a preformed foam.

Description

RELATED APPLICATIONS

This application claims the benefit of priority under 35 U.S.C. §119(e) of U.S. Provisional Application No. 60/931,407, filed May 22, 2007, U.S. Provisional Application No. 60/930,064, filed May 14, 2007, and U.S. Provisional Application No. 60/930,104, filed May 14, 2007, the disclosures of which are incorporated herein by reference.

FIELD OF THE INVENTION

Disclosed herein are prostheses for replacing all or part of a nucleus pulposus in a spinal disc area.

BACKGROUND OF THE INVENTION

This disclosure relates to treatments and implant materials for treating the medical condition called degenerative disc disease (DDD) by augmenting the intervertebral disc space, with or without the removal of intradiscal tissue, the nucleus pulposus (NP), and localizing, supplementing, or replacing native NP with an in situ curing polymer (Nucleus Replacement Prosthetic or NRP) while the outer intervertebral membrane called the annulus fibrosus (AF) heals after rupture and surgical intervention commonly referred to as a discectomy procedure. The NRP is designed to treat back pain and be both physiologically and mechanically supportive of the pathological intervertebral disc (IVD) to ultimately preserve the natural motion in the vertebral segment.
Treating DDD by performing a discectomy can include surgical access through the AF and the removal of all or some the NP through an existing rupture in the AF or a surgical incision through the AF called an annulotomy. The annulotomy or disc rupture is a tunnel through individual fibro-cartilage rings, or laminae, that make up the AF. Typically the annulotomy/rupture is not closed after removal of the NP and other fibrous debris. The open annulotomy will scar over and eventually heal within weeks of the surgery. The medical literature cites a 5% to 11% risk of re-operation (depending on the amount removal of the NP) to remove additional intradiscal debris that has migrated from the intradiscal space through the annulotomy and has come into contact with sensitive spinal nerves around the surgical site before the annulotomy has scarred shut. It is often a clinical objective of the nucleus replacement prosthetic to localize any intradiscal debris after partial nuclectomy while the annulotomy scars shut, sealing the intradiscal space naturally.
Another complication of the discectomy procedure is the loss of disc height (space between two vertebral bodies at any given spine level) after surgery. Disc height loss is directly related to radiculopathy or leg pain due to nerve compression. The medical literature cites that 20% to 30% disc height loss is anticipated in the first 12 months after the discectomy procedure resulting in 5% to 30% re-operation risk within five or six years after surgery. A second but much longer clinical objective is to maintain disc height over time with the replacement of NP with NRP vs. not replacing with NRP. The ability to replenish existing implanted NRP with new NRP using non-surgical techniques is desired, thus allowing for revision of existing NRP and indefinitely maintain the functionality of the IVD for the life of the patient.
The concept of treating an intervertebral disc abnormality by placing an in vivo forming implant in the intradiscal space normally occupied by disc nucleus or has its origin at least as early as the 1962. An in vivo forming implant is any substance placed medically in the body that is intended to persist for a therapeutic interval, sometimes as long as the remaining life of a patient, which undergoes a change in its mechanical characteristics, generally a phase transition. The phase transition can be due to a condition within the body that is different from the storage condition of the implant, for example a temperature or pH change. The phase transition can be induced by combining ingredients prior to implantation resulting in polymerization, precipitation, or otherwise hardening of the implant within the body. The phase transition can be induced by an outside stimulus, such as photo-initiating light or other extra-corporeal sources of energy. In most cases the phase transition is from a liquid to a solid, or from a low modulus solid to a higher modulus solid involving a structural change in the implant.
The concept of achieving an in vivo forming transition in an implant for the purpose of treating an intervertebral disc abnormality by means of in situ curing also has its origins as early as the 1980's. The in situ curing pathway frequently chosen is chemical polymerization of simple functional groups that result in the formation of extended polymer networks within living tissue. The functional groups most frequently cited can be arranged in three broad categories: 1) functionalized polyols, 2) acrylates, and 3) functionalized or crosslinked proteins.
The category of functionalized polyols includes reactions that result in the formation polyurethanes, polyurea urethanes, and other chemistries involving active NCO end groups. For example, Garcia French patent 2,639,823 discloses as early as 1988 the use of an in situ polymerizing polyurethane mixture delivered into a enclosing device into the nuclear space of a spinal disc. Felt U.S. Pat. Nos. 5,888,220; 6,140,452; 6,248,131; 6,306,177; 6,443,988; 6,652,587 and 7,001,431 (Bao et al) disclose a two-part in situ polymerizing polyurethane composition that is injected into a molding device placed in the nuclear space of a spinal disc. Later, Bao and Felt U.S. Pat. No. 7,077,865 and U.S. published applications 2007/0038300 and 2006/0253200 disclose the use of in situ curable compositions without the use of a molding device such as a balloon. Milbocker et al, U.S. published application 2005/0070913 discloses a one-part polymerizing polyurethane composition that bonds to the tissue of the nuclear disk space as it polymerizes in situ to fill a disc nucleus, repair a disc Annulus, or localize a nuclear prosthetic.
There have since been a number of disclosures directed to filling the nuclear space of a spinal disc with a variety of in situ curing materials. Haldimann U.S. Pat. No. 6,428,576 discloses the use of in situ curing materials without specification of the implant composition. Higham U.S. published application 2006/0255503 discloses an in situ curing disc nucleus implant containing a radio-opaque agent. Kim U.S. published application 2007/0005140 discloses delivery techniques employing multiple injection of in situ curing implant. Trieu U.S. published applications 20006/0200245; 2006/0089719; 2006/0064172; 2006/0064171; 2005/0203206 disclose various combinations of in situ polymerizing implants in conjunction with nuclear prosthetics to localize a prosthetic in a spinal disc.
With respect to particular compositions, Collins U.S. published application 2006/0009851 and 2006/0009778 disclose in situ polymerizing protein compositions for treatment of a spinal disc nucleus. Umit U.S. published application 2007/0093902 discloses other in situ polymerizing protein compositions for treatment of a spinal disc nucleus. In the category of acrylate compositions of in situ curing nuclear implants, Mallupragada U.S. Pat. No. 7,183,369 and Milner U.S. Pat. No. 6,187,048 are examples.
Without specificity to spinal applications, there are a variety of injectable biomaterials disclosed in issued patents including: cross-linkable silk elastin copolymer disclosed in Stedronsky U.S. Pat. No. 6,423,333, Capello U.S. Pat. No. 6,380,154, Ferrari U.S. Pat. No. 6,355,776, Stedronsky U.S. Pat. No. 6,258,872, Ferrari U.S. Pat. No. 6,184,348, Ferrari U.S. Pat. No. 6,140,072; Stedronsky U.S. Pat. No. 6,033,654; Ferrari U.S. Pat. No. 6,018,030; Stedronsky U.S. Pat. No. 6,015,474; Ferrari U.S. Pat. No. 5,830,713; Stedronsky U.S. Pat. No. 5,817,303; Donofrio U.S. Pat. No. 5,808,012; Capello U.S. Pat. No. 5,773,577; Capello U.S. Pat. No. 5,773,249; Ferrari U.S. Pat. No. 5,770,697; Stedronsky U.S. Pat. No. 5,760,004; Donofrio U.S. Pat. No. 5,723,588; Ferrari U.S. Pat. No. 5,641,648; Capello U.S. Pat. No. 5,235,041; protein hydrogel described in Morse U.S. Pat. No. 5,318,524; Morse U.S. Pat. No. 5,259,971; Morse U.S. Pat. No. 5,219,328; polyurethane-filled balloons disclosed in Bao U.S. Pat. No. 7,077,865; Bao U.S. Pat. No. 7,001,431; Felt U.S. Pat. No. 6,306,177; Felt U.S. Pat. No. 6,248,131; Bao U.S. Pat. No. 6,224,630; collagen-PEG disclosed in Olsen U.S. Pat. No. 6,428,978; Olsen U.S. Pat. No. 6,413,742; Rhee U.S. Pat. No. 6,323,278; Wallace U.S. Pat. No. 6,312,725; Sierra U.S. Pat. No. 6,277,394; Rhee U.S. Pat. No. 6,166,130; Berg U.S. Pat. No. 6,165,489; Simonyi U.S. Pat. No. 6,123,687; Berg U.S. Pat. No. 6,111,165; Sierra U.S. Pat. No. 6,110,484; Prior U.S. Pat. No. 6,096,309; Rhee U.S. Pat. No. 6,051,648; Esposito U.S. Pat. No. 5,997,811; Berg U.S. Pat. No. 5,962,648; Rhee U.S. Pat. No. 5,936,035; Rhee U.S. Pat. No. 5,874,500; chitosan disclosed in Chemte U.S. Pat. No. 6,344,488; other polymers discussed in Boyd U.S. Pat. No. 7,004,945; Collins U.S. publication 2006/0004326; Collins U.S. publication 2006/0009851; Milner U.S. Pat. No. 6,187,048; Daniell U.S. Pat. No. 6,004,782; Urry U.S. Pat. No. 5,064,430; Urry U.S. Pat. No. 4,898,962; Urry U.S. Pat. No. 4,870,055; Urry U.S. Pat. No. 4,783,523; Urry U.S. Pat. No. 4,589,882; Urry U.S. Pat. No. 4,500,700; Urry U.S. Pat. No. 4,474,851; Urry U.S. Pat. No. 4,187,852; Urry U.S. Pat. No. 4,132,746.
None of the treatments or compositions and associated surgical methods of their use described above are entirely satisfactory from either a biocompatibility or efficacy perspective for formation of a nucleus implant in situ. Accordingly, there remains a need for continued development of implant materials and treatment methods.

SUMMARY OF THE INVENTION

One embodiment provides a spinal disc implant comprising a foam adapted to completely or partially replace a nucleus pulposus within a spinal disc cavity, the foam being a nonabsorbable, closed cell and having a Poisson ratio of less than 0.5.
Another embodiment provides a method of repairing a defect in a spinal disc space, comprising:
inserting a nonabsorbable, closed cell foam having a Poisson ratio of less than 0.5 into the defect.
A method of repairing a defect in a spinal disc space, comprising:
inserting a composition in the area of the defect, the composition comprising:

- (a) a prepolymer, and
- (b) a foaming component; and

curing the composition to form a nonabsorbable, closed cell foam having a Poisson ratio of less than 0.5.

BRIEF DESCRIPTION OF THE DRAWINGS

Various embodiments will be understood from the following description, the appended claims and the accompanying drawings, in which:

FIG. 1 is a schematic view of an incompressible substance subjected to a loaded along its z-axis;

FIG. 2A is a schematic view of a spinal disk disc in transverse cross section;

FIG. 2B is a schematic view of a disc in lateral cross section;

FIG. 3 is a schematic view of a disc in transverse cross section;

FIG. 4 is a plot of disc deflection under load as a function of hoop stress ratios;

FIG. 5 is a plot of disc deflection as a function of a spring constants;

FIG. 6 is a schematic view of in lateral cross-section a spinal disc is comprising bony endplates, annulus fibrosus, and nucleus pulposus;

FIG. 7A is a schematic view of a disc in transverse cross-section;

FIG. 7B is a schematic view of a disc in lateral cross-section;

FIG. 8 is a plot of disc height as a function of increasing nucleus radius;

FIG. 9 is a plot of disc height lost as a function of nucleus to disc radius;

FIG. 10 is a plot of load as a function of nucleus to disc radius;

FIG. 11 is a plot of disc height lost as a function of nucleus to disc radius;

FIG. 12 is a schematic view of a disc in lateral cross-section with inset 253 showing the layered structure of endplate 107 and inset 252 showing the layered structure of annulus 108;

FIG. 13 is a plot of the ratio of nucleus load/annulus load as a function of disc deflection;

FIG. 14 is a plot of load as a function of displacement;

FIGS. 15A and 15B are plots of implant modulus as a function of nucleus radius for a fixed r_n/r_d=0.7 (FIG. 15A), and unfixed r_n/r_d(FIG. 15B);

FIG. 16 is a plot of implant modulus as a function of load;

FIG. 17 is a plot of implant modulus as a function of deflection;

FIG. 18 is a plot of implant modulus as a function of load;

FIG. 19 is a plot of annulus pressure as a function of prosthetic size;

FIG. 20 is a schematic view of forces involved in prosthetic extrusion through the annulus;

FIGS. 21A-C schematically depict stresses acting on an elemental slice of the prosthetic;

FIG. 22 is a plot of failure pressure as a function of modulus;

FIG. 23 is a plot of nuclear pressure as a function of modulus;

FIG. 24 is a plot of load failure as a function of modulus;

FIG. 25 is a plot of load failure as a function of impact velocity;

FIG. 26 is a plot for the conversion of modulus (y-axis) to durometer (x-axis) for polyurethanes;

FIG. 27 is a plot of load failure as a function of modulus;

FIG. 28 is a plot of load failure as a function of modulus;

FIG. 29A is a plot of nucleus pressure as a function of defect diameter;

FIG. 29B is a plot of load as a function of defect diameter;

FIG. 30 is a plot of nucleus pressure as a function of defect diameter;

FIG. 31 is a schematic view of flexion-extension properties of a disc under load;

FIG. 32 is a schematic view of the annulus and nucleus applying forces to the endplate;

FIG. 33 is a schematic view of the tilt component and effect on the disc height;

FIG. 34 is a plot of failure pressure/modulus ratio as a function of defect diameter;

FIG. 35 is a plot of restorative force as a function of displacement;

FIG. 36 is a plot of total load as a function of displacement; and

FIG. 37 is an illustration of common pathologies of spinal discs in the lumbar region.

DETAILED DESCRIPTION

Disclosed herein are treatments and implant materials disclosed relating to optimizing the function of replacement nucleus pulposus prosthetics placed. The characteristics of a nucleus replacement prosthetic can be derived from physical properties defined by the biomechanical and biological requirements for localizing and maintaining the prosthetic in the disc space.
One embodiment provides a spinal disc implant comprising a foam adapted to completely or partially replace a nucleus pulposus within a disc nucleus space, the foam being a nonabsorbable, closed cell and having a Poisson ratio of less than 0.5.
In one embodiment, the implant that partially or wholly replaces the nucleus pulposus is delivered to and localized in a disc nucleus space. The disc nucleus space includes the nucleus pulposus and optionally the adjacent tissues, including the vertebral endplates and inner layers of the annulus fibrosus (e.g., the layers contacting or immediately surrounding the nucleus pulposus).
Implants with a Poisson ratio of 0.5 are practically incompressible under normal physiologic loads. With such implants, a volumetric decrease in implant height results in an equal volumetric expansion of implant radius when approximating the implant as a cylinder. For an implant of a given modulus, there is a fixed relationship between the axial forces applied to the implant, the radial forced applied to the annulus and the changes in implant height and radius. For example, implants with Poisson ratio 0.5 cannot change height without also changing radius, and this incompressibility of the implant directly couples the annulus to the endplates for all load frequencies. Previous implant materials include gels and liquids that have a Poisson ratio of 0.5 and have the disadvantage of incompressibility to loads.
In contrast, implants with a Poisson ratio less than 0.5 has some fraction of its total volume in the compressible form of a gas, e.g., bubbles or closed cells in a foam implant. The greater the fraction of bubbles in the implant the lower the Poisson ratio. Accordingly, in one embodiment, the implant comprises a foam, such as a closed-cell foam. The compressibility affords a looser coupling of the height and radial dimensions at high load frequencies, as opposed to the decrease in height an equal volumetric radial expansion of incompressible implants. For example, for a compressible foam that experiences a sudden impact does not subject the surrounding tissue to an immediate expansion. This compressibility feature may help preserve the annulus and/or endplates by storing transitionally some of the impact energy of the load as potential energy in the implant.
The foam can be a preformed foam having the recited properties, or a curable foam implanted as a liquid implant material and cured while in the disc space. Accordingly, disclosed herein are preformed foams and in situ curing nucleus implant foams intended to therapeutically replace or augment the natural disc nucleus space. One embodiment provides a curable implant composition, comprising:

- (a) a prepolymer, and
- (b) a foaming component.

The prepolymer and foaming component is described in greater detail below. When combined, the composition can be cured to form a nonabsorbable, closed cell foam having a Poisson ratio of less than 0.5.
Curing with respect to a liquid or deformable nucleus implant implies a phase change of the implant from liquid to solid or deformable solid to less deformable solid. One embodiment provides a liquid prepolymer mixed with water, saline or a therapeutic aqueous solution prepared outside the body and injected into a vertebral disc. The implantation can be facilitated by the mixing of liquid prepolymer and aqueous solution to provide for a low viscosity liquid implant that can be injected through a needle. In procedures directed to correcting a diseased or bulging vertebral disc, often a hole is made in the annulus in order to remove some or the entire disc nucleus. In one embodiment, the removed nucleus, whether completely or partially removed, is replaced with an implant. In such cases a liquid implant would flow out of the implantation site, and a curing step localizes the implant in the disc.
The natural nucleus is largely water, and diseased nucleus in some cases is characterized as being dehydrated. In one embodiment, a nucleus implant contains a large fraction of water; yet possess the shape retentive features of a solid. The amount of water in the cured implant determines its freestanding Young's modulus. The Young's modulus can range from as little as 0.5 MPa to as great as 10 MPa. Generally, the lower the modulus the greater the fraction of axial forces translated radially to the annulus. The minimum modulus of the cured implant can be determined by the size of a defect in the annulus. In the case where the implant is injected through a small gauge needle, 20 G or more, and no other defect exists or is created in the annulus, the modulus can range from 0.5 MPa to 1.0 MPa. The relationship between defect area and minimum implant modulus can be calculated as described in the Appendix below. Applying assumptions about the disease state of the disc and the likely surgical intervention the following table summarizes the findings of minimum implant modulus for different surgical interventions.

TABLE 1

Treatment Paradigms for Nucleus Replacement

		Disc
		Height	Increase
Indication	Bonded	Restored	Nucleus	Modulus	Procedure	Annulotomy

Thin Disc	Yes	Yes	No	0.5 MPa	Trans-axial	No
Lumbar
Thin Disc	Yes	No	No	1 MPa	Trans-axial	No
Lumbar
Thin Disc	No	Yes	No	3 MPa	Trans-	No
Lumbar					annulus
Thin Disc	No	No	No		6 MPa	Trans-	No
Lumbar					annulus
Black Disc	Yes	Either	Yes	1 MPa	annulotomy	<2.5 mm
Bulging Disc	Either	No	No		2 MPa	Annulotomy	<4.9 mm
Bulging Disc	Yes	No	No		1 MPa	Annulotomy	<4.9 mm
Bulging Disc	Yes	Yes	No	0.25 MPa	Annulotomy	<4.9 mm
Permanently	No	No	No	8 MPa	Annulotomy	Any size
compressed
Permanently	No	No	Yes	6 MPa	Annulotomy	<3.5 mm
compressed
Permanently	Yes	No	No	1 MPa	Annulotomy	<3.5 mm
compressed

The minimum implant moduli associated with the various surgical interventions of Table 1 are general guidelines for selecting an implant modulus, and can be further refined by reference back to the details of the calculations in the Appendix. The cured modulus can be varied by varying the ratio of prepolymer to aqueous solution at the time of preparation outside the body.
Another embodiment provides a cured implant comprised of solid, liquid and gaseous parts. This can be achieved with a solid polymer matrix containing both liquid and gaseous fractions. One example is a hydrogel, as described in U.S. application Ser. No. 10/020,331, published U.S. Pub. No. 2003/0135238, the disclosure of which is incorporated herein by reference. The gaseous and liquid components of the hydrogel foam implant are in equilibrium, with the volume of the implant varying with the applied load, wherein the ratio of the volumes of the liquid and gaseous components can vary according to the magnitude of the load. This feature has a direct analogue in the natural disc, where the disc height is typically less during periods of activity (high load) and expands during periods of sleep (low load). Gas in the implant can be forced into solution during protracted periods of load, reducing disc height and nuclear volume. During periods of rest, the dissolved gas comes out of solution and can fill the bubbles in the hydrogel foam. The timescale on which gas is forced into solution and then subsequently released back into the implant as a gas can be similar to the timescale for the increase in normal disc height during periods of rest. Thus, this accommodative effect is not responsive to high frequency changes in load, such as experienced while running.
In one embodiment, the water fraction of the implant is bound in the implant such that pressures of the magnitude commonly encountered in the disc do not result in water being driven out of the implant by pressure. In one embodiment, the implant contains loosely bound water. In this embodiment, the water contained in the implant can exchange with water in surrounding tissue through a statistical process, which is largely unaffected by pressure since the pressure in the implant and the pressure in the tissue are approximately equal. Even in the case where the pressure in the implant exceeds that of the surrounding tissue, water will not be driven out of the implant if the water bound in the hydrogel is localized by hydrogen bonding. This kind of localization of water within a polymer is often associated with hydrogels, which possess an affinity for water. This affinity for water is also a characteristic of normal nucleus pulposus. In one embodiment, the cured implant retains its water fraction through hydrogen bonding
One embodiment provides a minimally invasive means of delivery through a small access hole or needle injection and a second localization feature that prevents the implant from migrating through the access hole and away from the intended implantation site. The localization feature is the result of the cured implant being larger than the access hole. A curable nucleus implant can undergo two phase-transitions and also retain its original liquid phase forming a foamed hydrogel. A compressible preformed foam implant can retain its original shape once delivered through the small access hole and inserted in the disc nucleus space.
In another embodiment, the treatment involves delivery of an in situ polymerizing tissue adhesive into the treatment area of the nucleus or nuclear space. The implant, if an in-situ curing implant is delivered in fluid form typically down an access port leading to a space in the disc where formerly nucleus material resided and now air resides. For a preformed foam, the implant is delivered in compressed form and allowed to expand once inside the nuclear disc space. These procedures aim to fill a void created surgically in the disc. To fill a surgically created void, the implant can either displace the gas residing at the implant location or incorporate it in the implant volume. In either case, a large gaseous void trapped by the implant in the nucleus of the disc is typically avoided for one or more of the following reasons: 1) it provides a compressible volume into which the cured implant may shift or extrude, 2) the region in which the bubble resides is not load bearing resulting in localized forces on the vertebral endplates, 3) provides a space where infiltrating cells are likely to accumulate and cause inflammation, 4) the gas entrapped in the space is likely to eventually migrate out of the disc space and cause a loss of disc height, and 5) the space creates a discontinuity between implant and annulus predisposed to the creation of implant particulate. In one embodiment, the implant is space filling. In another embodiment, the implant is sufficiently hydrophilic and/or foaming, wherein the final cured volume is greater than the initial implant volume.
In one embodiment, the access (or opening) to the nuclear space of a disc is as small as possible. Often the therapy involves removal of some or the entire disc nucleus. The nucleus is removed and replaced by air during the nuclectomy. The constraints of the access hole dimensions can make it difficult to refill the nucleus with a liquid implant without trapping some gas in the nucleus. The region of trapped gas becomes a void in the cured implant. This void can decouple a large fraction of the implant surface from the annulus making the implant less effective in translating axial forces to radial forces applied to the annulus. The benefit of stiffening the annulus through the application of radial forces is incompletely achieved.
A large dimensional void in the implant, unlike uniformly distributed small bubbles, presents a high stress surface that can result in fracture of the implant into the void and particulate formation. The larger the radius of the void, the higher the stress accumulation on the surface of the implant and higher the likelihood the implant will degrade. It is desirable to minimize voids in the implantation volume.
A non-expanding implant will push the air pocket to the side. An expanding implant will expand into the void. In the case where a volume of gas is entrapped by the injection of a liquid implant, the initial pressure of the entrapped gas is at ambient pressure, and increases as additional implant is injected into the space. During the injection, the pressure of the liquid implant equals the pressure of the gas trapped in the void. For an expanding implant, the creation of a gas phase in the implant makes the pressure inside the implant slightly higher than the pressure in the void. The implant foams into the void. While the mass of the trapped gas does not change, its distribution within the implant becomes more uniform by becoming distributed in the bubble structure of the implant. This effect combined with the hydrophilic nature of the liquid implant tends to ensure the solid phase of the implant is in contact with the entire inner surface of the annulus and endplates.
In one embodiment, the implant fills the implantation space by a combination of increasing pressure after the implant volume has been delivered and a space-filling foaming action.
Another embodiment provides a method for preventing migration of the fully cured implant. In most surgical applications, a hole will be created surgically in the annulus of the disc through which degraded nucleus is removed and implant is delivered. It may be beneficial to include at least a temporary tissue bonding means in the chemistry of the implant to prevent extrusion of the material through the hole in the annulus. In one embodiment, the implant does not fill the hole in the annulus. In one embodiment, this bonding can achieved by devices and methods of Provisional Application No. 60/930,104, entitled “Foam Disc Prosthesis” the disclosure of which is incorporated herein by reference. A tissue bonding implant of this embodiment can be useful in one or more of the following ways: 1) bonds can be formed between the implant and inner wall of the annulus to stabilize and strengthen the annulus, 2) bonds can be formed between the implant and remaining nucleus localize the nucleus and reduce the likelihood nuclear material will escape from the disc, 3) bonds formed between the vertebral endplates and layers of the annulus can allow for a lower modulus implant without extrusion, 4) these same bonds can maintain the patency of the surgically created hole in the annulus allowing for natural annulus regeneration, and 5) the integrated resulting structure of coupled layers of annulus, nucleus, and endplates can mitigate against the need for complete removal of nuclear material and could potentially result in improved surgical outcome.
An in situ curing liquid implant that bonds to tissue renders it less likely to move around in the implantation space, especially if a void exists in the implantation space. Motion of a nucleus implant within the annulus of a disc is generally undesirable because the differential motion between tissue and implant results in tissue disruption, inflammatory response and implant abrasion. A relatively low modulus, tissue-bound implant will follow the changes in tissue geometry and avoid particulate formation caused by tissue passing across the implant surface. The ability of the implant to follow tissue motion requires an implant having a modulus similar to that of tissue; otherwise shear forces may develop between implant and tissue. In one embodiment where the modulus of the tissue matches the modulus of the implant, shear force is minimized and the implant follows tissue motion.
Since the modulus of tissue is low, an implant of tissue-like compliance will in many cases deform through a defect made in the annulus. For example, when forces are applied to the implant, and the modulus of the implant is low enough to follow tissue motion, the implant will tend to extrude through a hole made in the annulus created during implantation. The existence of a defect in the annulus can re-establish shear stress between the implant and surrounding tissue, and the implant may move relative to the tissue interface in a direction that favors extrusion of the implant from the implantation site. If the entire implant surface is free to move in this way, there is very little restorative forces present to keep the implant in its intended site. However, if some or all of the surface is bonded to surrounding tissue, bulk slippage of the implant relative to tissue is reduced or minimized. Although surface stress develops at the bond interface, the force required to extrude the implant through a defect in the annulus is greatly increased. Therefore, bonding allows for a softer, extrusion-free implant as illustrated in Table 1. In one embodiment, the implant bonds to tissue at least during the first weeks after implantation, when the defect in the annulus is open.
In one embodiment, the bond strengths range from 4 lbs/in²to about 25 lbs/in².
In another embodiment, delivery of the implant is performed through a lumen. In one embodiment the lumen is of minimal cross section, but sufficiently large to deliver the composition (e.g., liquid nucleus implant) or compressible preformed foam by conventional methods, e.g., a catheter or a syringe or similar liquid dispensing device that can be either mechanically pressurized or manually pressurized sufficiently to deliver the liquid nucleus implant to the treatment site before the liquid implant has cured.
In one embodiment, the composition is a low viscosity liquid implant. In one embodiment, the prepolymer is premixed with saline before implantation to initiate the polymerization process and reduce the viscosity of the prepolymer. Reducing the viscosity of the prepolymer can have one or more of the following features: 1) a smaller delivery lumen and a less disruptive surgical procedure is possible, e.g., with respect to preserving the structural integrity of the annulus, 2) the force required to deliver the liquid implant to the implantation site decreases with decreasing viscosity, 3) the time required to deliver the liquid implant to the implantation site decreases with decreasing viscosity, and 4) a low viscosity implant ensures the crevices and fragments of the interior disc space are uniformly and completely coated with implant. Premixing the prepolymer with saline can have one or more of the following features: 1) mixing with saline can ensure uniform activation and curing of the implant, 2) mixing with saline can ensure a more predictable cure time, 3) mixing with saline can reduce the cure time, 5) mixing with saline can establish the liquid fraction of the cured implant, 6) mixing with saline can initiate the release of the gas phase of the implant, and 7) mixing with a quantifiable volumetric ratio of prepolymer and saline can determine the final solid/liquid/gas ratio of the implant.
In one embodiment, a relatively large implant volume of a liquid in situ curing implant can be inserted or delivered through a small hole. Delivery of the implant can be performed via injection, e.g., generally through a needle. The needle can be as long as several inches. In one embodiment, the liquid implant is delivered under pressure; typically pressures attainable by manual compression of a standard 3 ml syringe are acceptable, generally less than 100 psi. Low implant viscosity has other benefits, for example, more of the injection pressure is available for pressurizing the implant to augment disc height. In some cases, it may be beneficial to be able to sense resistance when injecting into the disc, and the pressure increase due to resistance relative to the drop through the needle is related to the level of manual detection of pressure resistance.
For the range of needle lengths and diameters commonly used to inject dyes into the disc for diagnostic purposes, an implant viscosity less than 1000 cp is acceptable, such as an implant viscosity of less than 200 cp. In another embodiment, the viscosity ranges from 100 cp to 1000 cp. In one embodiment, the viscosity limit of 1000 cp is satisfied when prepolymer is mixed with water in the ratio of 70:30 or less, and 60:40 or less for the 200 cp limit.
Another embodiment provides a cure time sufficiently short to ensure the liquid implant stays localized to the implantation site and sufficiently long to allow the surgeon to deliver the implant in a beneficial way. In one embodiment, the liquid implant cures at a faster rate when heated inside the body and in the presence of tissue. In one embodiment, such an implant can result in one or more of the following: 1) shear thinning of the liquid implant as it permeates defects in the annulus can cause the implant to warm via internal body temperatures and initiate protein binding at a faster rate than the bulk polymerization of the implant, 2) this polymerization mechanism tends to encircle the implant volume with cured implant to prevent loss of the implant from the nuclear space, and 3) it can ensure that a large number of the active sites in the liquid implant are available for covalent tissue bonding and not consumed in bulk polymerization. Implants without this feature that form solid implants in situ, rely on mechanical attachment to tissue. In one embodiment, a chemically enhanced bonded implant can tolerate greater changes in implant volume and shape before implant dislocation.
In one embodiment, cure times can span as long as 1 hour and as short as 30 seconds, although longer cure times are possible. In general, access will have been made to the implantation site, and preparation of the implantation site completed before the liquid implant is prepared. In one embodiment, the implant is prepared by mixing between two syringes bridged by a female-to-female luer lok connection prepolymer in one syringe and saline or other suitable aqueous solution in the other syringe. In one embodiment, the hydrophilic nature of the prepolymer achieves homogenous mixing in approximately 10 mix cycles for mix ratios of 10-90% prepolymer. In one embodiment, all implant ratios are homogenous after 20 mix cycles. In one embodiment, the fastest cure time are achieved where the mix ratio is approximately 1:1. However, the cure time does not differ by more than 100% for all mix ratios.
The surgeon typically requires a cure time long enough to mix and inject the liquid implant and short enough to provide for in situ curing within a few minutes after implantation. In one embodiment, the cure time ranges from 1 to 10 minutes, such as from 3 to 5 minutes. In one embodiment, the cure time halves for every 10 degree centigrade increase in mixture temperature. The typical difference between body temperature and room temperature is about 10° C. Often, there is a decrease in cure time once the liquid implant is injected in the body.
Another embodiment provides an in situ cured implant that does not change volume beyond a therapeutic range through the loss or acquisition of aqueous fluid in the body. Many hydrogel compositions do not have a cross-linked structure, and when placed in an aqueous environment swell appreciably, sometimes to the point of dissolution. Polymeric swell where water enters the polymer matrix is distinguished from the phase transition aspect of this embodiment where the implant increases volume through the liberation of a gas forming bubbles in the implant, which may later fill with water. The later aspect does not compromise the tensile strength and other mechanical characteristics of the solid phase of the implant. Polymeric swell can compromise permanence of the polymer, making them susceptible to fracture, migration, particulate formation, and loss of therapeutic efficacy. In one embodiment, the mass of aqueous solution mixed with the prepolymer prior to implantation is approximately the mass of water contained in the polymer matrix after polymerization such that the ratio of polymer to water in the solid phase of the hydrogel remains approximately constant.
In one embodiment, the foaming of an in-situ curable implant pressurizes the implant volume and reduce voids in the implantation volume. The foaming may also aid in tissue infiltration and bonding. Moreover, there are many preformed nucleus implants that expand after implantation. This feature can minimize the access hole to the nuclear space by introducing a swellable preformed implant. The need for post-implantation swell is reduced for in situ cured implants since the implant volume can be pressurized prior to curing and the liquid nature of the implant provides for small access holes. The primary disadvantage to post-implantation swelling of an implant is uncontrolled forces applied to annulus and endplates that may result in annulus tears, endplate deformation, implant extrusion or unnatural disc distraction.
In general, the chemistries can be adjusted to swell when placed in the body. The tensile strength of a non-degrading hydrogel implant decreases proportionally to the change in implant volume due to a decrease in the number of cross-links per unit volume. Swelled polymers can be noticeably easier to disaggregate, leading to an increase in particle formation. This aspect of a swelled implant may be disadvantageous when all other endpoints such as disc height adjustment, implant site filling, implant modulus, as well as others are achieved in the cured state of the implant.
In one embodiment, nucleus implant volume change affecting tensile strength of the cured implant is less than 50% after 1 week for implants placed in the nuclear space of a disc. In one embodiment, the volume change is less than 20% after 1 week, or less than 10% after 1 week. These limitations can relate to but are not limited to implant swell in the body and increases of implant volume that results in decrease of a mechanical property of the implant. This limitation may apply less directly to implants designed to accommodate to the implant site within a predetermined therapeutic range, thereby providing tissues time to equilibrate with forces generated by the implant on tissue surfaces. Where implant volume decreases or increases in response to an accommodative endpoint designed into the implant, that does not affect the mechanical properties of the solid phase of the implant are contemplated later in this application. These changes refer to alteration of the gas phase of the implant, and in certain changes in the Poisson ratio of the implant after 1 week.
Another embodiment provides a nucleus pressurizing capability of the liquid nucleus implant. A liquid implant composition can liberate a gas phase during polymerization within the disc nucleus to facilitate its space filling aspects. In this embodiment, the choice of the ratio of water to prepolymer can determine the molar quantity of gas liberated per unit volume of prepolymer and water. A therapeutically beneficial volume of gas is liberated into the liquid prepolymer to cure into a larger volume foam implant to pressurize the nuclear space. Pressurizing the nuclear space for in situ polymerizing implants can be achieved mechanically by applying pressure to the implant before it has cured in situ, and can be beneficial for one or more of the following reasons: 1) the pressure can be generated within the volume of the implant, not externally, which enhances its therapeutic value by not disrupting an interface formed between tissue and implant which may be beneficial to localizing the implant, 2) the pressure can be generated isotropically without spatial gradients, 3) the pressure can directs the implant into intimate contact with the layers of the annulus, 4) the pressure can be effective in achieving a higher implant modulus that is self-tailored to the patient's condition and anatomy, and 5) the pressure can predispose the vertebral bodies to distraction thereby achieving a greater disc height.
In one embodiment, the prepolymer compositions provide for gas to cured polymer volume ratios of 3:1 greater under ambient conditions. In one embodiment, the prepolymers possess an isocyanate functionality that reacts with water to liberate carbon dioxide, resulting in a foam that forms a pressurized cured implant state.
The pressure developed in the cured state can depend on how much the implantation volume expands in response to the pressurized state of the implant. In one embodiment, a beneficial outcome of the pressurized state is that the disc endplates become distracted and the height of the disc increased. The pressurized state of the implant can increase the modulus of the implant, such that for a given volume it is possible to calculate how much gas must be liberated so that the final gaseous fraction of the implant is at a certain target pressure. These calculations involving molar quantities of gas liberating groups in the prepolymer are well known in the art.
Some of the advantages of the pressurized state of the implant may be transitory in nature, for example an initial pressurized state in the curing implant that drives the isobaric expansion of the curing implant into the small-scale features of the surrounding tissue, while the final state of the implant is insignificantly different from ambient conditions.
Another embodiment provides a composition further comprising a radio-opaque agent or illuminating marker. There are a large number of radio-opaque and radio-emitting markers that are added to medical devices to aid in their external visualization by noninvasive means. They fall into two categories: soluble additive, and insoluble additives. It is a requirement of either that they do not adversely affect the curing mechanism of the implant. It is further desirable that the additive not significantly adversely affects biocompatibility.
In one embodiment, the prepolymer is soluble in aqueous solutions, and the radio-opaque agents is soluble in water. In one embodiment, the addition of these agents to the aqueous phase of the mixture does not interfere with implant polymerization. Optionally, a solid phase radio-opaque agent may be added in the form of powder or aqueous solution, which then forms a suspension upon mixing and does not interfere with implant polymerization.
An example of a water-soluble radio-opaque agent is barium sulfate, which can be added to the liquid implant in volumetric fraction of 30% or less, or a volumetric fraction of 20% or less, or a volumetric fraction of 30% or less or 20% or less but not less than 10%. The barium sulfate can be added to the aqueous fraction of the components to be mixed to form the liquid implant. The lower fractional proportion is approximately the minimum amount of additive required to differentiate the implant from surrounding bone and tissue, greater fractional amounts of additive increase the implant contrast.
An example of an insoluble radio-opaque agent is tantalum powder. Tantalum powder can be added to the aqueous or prepolymer fractions in preparing the liquid implant. In one embodiment, the tantalum is present in a volumetric fraction of 30% or less, or a volumetric fraction of 20% or less, or a volumetric fraction of 30% or less or 20% or less but not less than 10%. The tantalum powder can be of any size that is easily transported through the delivery apparatus, e.g., sub micron in size. Although powder additions may present a potential for particulate migration, a curing aspect or non-migration aspect of the implant can mitigate against particulate loss.
Another embodiment provides an implant with a bimodular compliance characteristic. The bubbles entrapped in the cured nucleus implant have a compliance described by the state equation for a gas. The compliance of the solid phase of the implant has a distinct and independently derived compliance that is governed primarily by the degree of cross-linking in the formed implant and the amount of water comprising the hydrogel. When placed under load, the compliance aspects of each of these components sum in proportion to their volumetric ratio to yield a bimodular compliance for the implant as a whole. One feature of this bimodular aspect is that while the compliance for the solid phase of the implant is expressed as a constant the compliance of the gas phase of the implant depends on the total volume of the gas phase. As the disc height decreases under load, the implant becomes stiffer—a feature not attained with implants with a Poisson ratio of approximately 0.50. The implant can freely deform and translate axial forces to radial forces applied to the annulus under small disc compression but can become increasingly stiffer translating fractionally less of the axial forces to radial forces under greater compression of the disc. This feature is protective of the annulus, and guards against re-herniation of the annulus.
This bimodular compliance acts under small disc compression like the natural nucleus of the disc, and acts increasingly more like a solid load-bearing member under increased compression of the disc. This implant is engineered to be a compromise between distortion of the endplates typically experienced with harder nucleus implants, and herniation of the annulus or extrusion of the implant typically experienced with the natural nucleus of the disc. Further, the bimodular aspect of the implant can be tailored by the medical professional to the specific condition of a diseased disc by adjusting the ratio of aqueous and prepolymer fractions before mixing; or, the bimodular aspect can be engineered by the manufacturing concern to provide greater or lesser molar quantities of liberated gas during the polymerization of the implant in situ.
In one embodiment, the implant (e.g., a preformed or cured foam implant) contains a solid hydrogel phase and a gaseous phase enclosed in the hydrogel in the form of bubbles. In one embodiment, the solid phase comprises polyurethane and the gaseous phase is carbon dioxide. Depending on the method and amount of liquid implant delivered to the implantation site, the final state pressure of the carbon dioxide bubbles enclosed in the closed cell structure of the hydrogel varies from near ambient to several atmospheres. The long-term pressure of the carbon dioxide fraction of the implant depends on the solubility of the gas phase into the water residing in the hydrogel implant.
In one embodiment, the closed cell bubbles initially contain less than 50% of a liquid, less than 25%, or less than 10% of a liquid (or mostly gas and minimal liquid) within, e.g., 1 month or less after implantation, within one week or less after implantation, within 5 days after implantation, within 3-5 days after implantation, within 3 days after implantation, or even within 1 day of implantation. In one embodiment, the closed cell bubbles contain In one embodiment, at these time periods the Poisson ratio is less than 0.5. In another embodiment, after any of the above-mentioned time periods, the implant has a Poisson ratio approaching 0.5.
At any instance in time, the compliance of the implant is a function of the combined effects of the compliance of the gas bubbles and the Young's modulus of the hydrogel. Depending on the pressure in the gas phase of the implant, the dominant compliance can be approximately as the ideal gas law for low gas phase pressure to the Young's modulus of the hydrogel for high gas phase pressure. In general, the compliance of the gas phase is softer than the compliance of the solid phase. During compression of the implant, a point is arrive at where the pressure in the bubbles exceeds the modulus of the solid phase, and the implant is essentially incompressible, and a Poisson ratio of 0.5 is approximated.
In one embodiment, the total compliance of the implant is bimodular, where the compliance is dominates by the gas phase for small compressions and is dominated by the solid phase for large compression. The early phase compression of the implant is characteristically dependent upon the volume of the air bubbles, and thus increases with displacement. The late phase compression of the implant exhibits a characteristic constant modulus, approximately by the Young's modulus of the hydrogel fraction.
The early phase compression can provide for high frequency load absorption without significant distortion of the annulus, or at least less distortion than what would have resulted for a Poisson ratio=0.5 implant. Therefore, one can image high frequency oscillations about a mean disc height, and the mean disc height only changing when the mean load changes. This is protective of the annulus, and accommodative with regards the distribution of load between the endplates and the annulus.
Depending on the magnitude and duration of mean loads applied to the implant, a volumetric fraction of the carbon dioxide in the bubbles can absorb into the water fraction of the hydrogel matrix, thereby reducing the pressure in the bubbles of the implant. This tends to restore some of the dynamic compressive range of the implant even when relatively large loads are endured for long periods of time. However, when the implant is unloaded, the locally dissolved carbon dioxide can reappear in gaseous form and depressurizing the bubbles of the hydrogel. However, this response is dissipative with time, and eventually the carbon dioxide is lost by diffusion and the time averaged volume of the bubbles is filled with free aqueous solution derived from the surrounding tissue.
The long-term compliance of the implant is governed by the Young's modulus of the hydrogel fraction, and eventually the implant exhibits a Poisson ration of 0.5. At this point height changes of the disc can be modulated by height and radius changes of the annulus, and the implant adopts the compliance properties of the natural nucleus.
The short-term bimodular characteristic of the implant can result in isolating high frequency displacements of the vertebral endplates from displacements of the annulus, the effect of which is to provide for annulus healing, especially with respect to healing of the access hole created in the annulus for removal of nucleus and injection of liquid implant. The short-term bimodular characteristic can establish the height of the disc with respect to the partition of axial and radial forces, which is essentially locked into place when the bubbles are filled with water and the compliance of the implant, approximates a constant value. The disc height is not necessarily coupled to the compliance of the implant before this accommodation can be achieved. In one embodiment, this provides a method for axial forces to be somewhat decoupled from radial forces until a mean disc height is established, such that the range of accommodation of the annulus can be centered on the range of displacements created by loading whereby the radially and axial forces are partitioned by the Young's modulus of the implant when the bubbles become filled with a liquid fraction derived from surrounding tissue
Another embodiment provides a liquid nucleus implant, the compliance character of which is predictably controlled by the medical professional. A medical professional may decide to surgically modify a vertebral disc to correct a pathologic condition. Typical interventions include: 1) reduction or removal of a portion of the disc annulus, 2) removal of a portion or all of a disc nucleus, 3) removal of a portion of the inner layers of the disc annulus, 4) modification of the distance between endplates of a disc, and 5) various procedures intended to assess the compliance or spatial aspects of the diseased disc. These interventions can yield characteristics of the disc that can be used to assign the appropriate implant therapy. For example, the size of the annulus hole, the thickness of the annulus, the radius of the space created in the nucleus of the disc, the height of the disc, the degree of compliance of the diseased annulus, and the dimensions of the endplates are all useful inputs into a therapy paradigm. The equations involving these inputs can be used to select a set of ideal implant characteristics as set forth in the Appendix below. These equations demonstrate that the type and range of implant characteristics selectable by the medical professional is adequate to address a rather broad range of pathological disc conditions.
In one embodiment, the prepolymers of the composition are hydrophilic and readily dissolve in water. The lower limit for solid formation is approximately 5% by volume prepolymer in a mixture substantially comprise of water. The limit on Young's modulus for the in situ curing implant is the solid/liquid phase transition, which occurs in the polymerization at the 5% concentration. The upper limit can depend on the molecular weight of the prepolymer and/or its functionality, but typical prepolymers reach maximum modulus at nearly 100% prepolymer mixtures with only trace amounts of water. The water typically present in the body can be more than adequate to complete polymerization of the prepolymer. In one embodiment, the upper limit on the modulus is in the range of 10-20 MPa, e.g., from 0.5 to 20 MPa or from 0.5 to 10 MPa. In another embodiment, the upper limit on prepolymer-water mixture able to be delivered through a 20G needle with hand pressure is approximately 5 MPa, with mixtures ranging from 0.5 to 5 MPa, 1 to 5 MPa, or 1-3 MPa resulting from mixtures in the 10-80% water range. These ranges and their associated procured mixtures can depend on the cross-link density achieved with the prepolymer structure.
Prepolymers, as those disclosed herein, are adequate to meet the range of therapeutic implant moduli detailed in Table 1. Additionally, fillers such as flock or particulate added to the procured mixture will tend to shift the range of moduli to higher values.
Since the modulus varies as the cross-link density it also varies approximately uniformly with the mixture ratio. Accordingly, for any particular prepolymer a table can be generate for cured implant modulus vs. mixture ratio. This feature is useful in practicing the type of treatment guidance illustrated in Table 1.
Another embodiment provides a liquid nucleus implant that is self-sealing. In this embodiment, the implant polymerizes preferentially along the implant margins to achieve one or more of the following: 1) prevent extrusion of the implant during pressurization through the access hole made in the annulus, 2) prevent externalization of the implant through undetected defects in the annulus, and 3) provide for pressure to develop within the volume of the in situ curing implant. One or more of the following clinical benefits can be achieved: 1) the limited infiltration of the liquid implant into and along the concentric layers of the annulus to enhance the hoop stress capacity of the annulus and enhance the localization of the implant, 2) to provide an encapsulating aspect to the outer layers of the implant that act to restrain its deformation into the access hole in the annulus and thereby leave it patent to promote annulus healing, and 3) to provide a layered aspect to the nucleus implant that acts as a structural extension of the annulus and provides additional hoop stress bearing properties.
The rate of polymerization or curing of the prepolymers described herein can vary according to two environmental conditions. In certain embodiments involving in situ curable implants, the cure rate increases approximately 2 fold for every 10° C. rise in liquid implant temperature. This can be a useful feature since liquid implant injected at room temperature will cure preferentially at the margins where the implant is in contact with tissue and locally heated by the tissue. This effect can be further magnified when the implant extends beyond the bulk of the implanted mass, and begins infiltrating cracks and fissures in the tissue interface. The effect can be large enough to cause nearly instantaneous polymerization of the liquid implant as it infiltrates this tissue. In cadaver studies it was observed that in discs with partially delaminated annulus tissue, that the liquid implant followed these fault lines and acted to recouple the layers mechanically. This development has two effects: 1) the maximum hoop stress of the annulus is dramatically increased since the layers of the annulus are no longer free to slide past one another and concentrate stresses, and 2) the inward projection of the inner layers of the annulus, creating a convexo-convex cross section is defeated and the normal convexo-concave cross section is restored due to the radially directed pressure developed in the curing implant.
The second environmental condition that contributes to the self-sealing feature derives from the difference in reaction rates of polymerization. For example, where the prepolymer comprises isocyanate groups, the self-sealing feature derives from the difference in reaction rates between the reaction of the isocyanate groups on the prepolymer with water to form amine groups and the reaction rate between amine groups with isocyanate groups to form urea linkages. The latter occurs much faster than the former. Since the cure rate of the liquid implant is limited by the rate of amine formation in the implant, amines existing on the surface of tissue will react first with the implant. Thus, the cure rate at the periphery of the implant can be increased by the action of the water component of the liquid implant that generates amine groups and the preexisting proteinaceous groups residing either freely or attached to the tissue. This acceleration of the cure rate can be further increased when the implant infiltrates tissue porosities, since the ratio of implant surface area to volume is increased. In the limit the rapidly polymerizing layers of the implant come together in the narrow confines of a tissue defect causing the implant to bond tissue layers together and seal against the egress of liquid implant.
This self-sealing feature can be useful in the clinical setting where defects in the annulus may go undetected, and externalization of the implant outside the disc would be deleterious to outcome. In one embodiment, the sealing aspect is achieved prior to substantial liberation of gas within the implant, allowing the seal to be achieved before full pressurization of the implant.
Another embodiment provides a nucleus implant that forms a foamed hydrogel in the body capable of exchanging water occupying the solid phase of the implant with surrounding tissue. The water is weakly held within the polymer matrix, and can be exchanged with water molecules of the body. In one embodiment, the exchange can be one-to-one, such that the total volume of the solid phase of the implant does not change significantly. The density of the polymer matrix responsible for localizing the water is determined largely at the stage of mixing, and is determined more rigorously at the completion of polymerization. Although most tissues present a wet aspect in the body, the free water on these tissue surfaces is generally insufficient to alter significantly the ratio of water to prepolymer during the polymerization phase of the implantation. The water exchanging aspect can provide one or more of the following clinical benefits: 1) ionic or pH differences between the implant and the surrounding tissue quickly equilibrate, 2) osmotic differences between the implant and surrounding tissue quickly equilibrate, 3) accordingly, nutrient conduction to distant tissues from sources in contact with the implant, such as the vertebral endplates, is not blocked by the implant, and 4) proteins and cells responsible for marking and proliferating an inflammatory or fibrotic response are hindered by the surface washing aspect of the implant resulting from the constant thermodynamic exchange of water molecules on the implant surface.
In one embodiment, the cross-link density of an in-situ curable implant is in part determined by the mixture ratio of water to prepolymer. In another embodiment, the prepolymer is hydrophilic and all the water is incorporated into the cured implant for prepolymer ratios greater than approximately 5% water. The mode of incorporation is via loose hydrogen bonds that are reversible under statistical fluctuations and heating. However, the vacancy created by thermal excitation of a water molecule resulting in its diffusion out of the implant is preserved in the cured implant structure, making it thermodynamically more probable that the vacancy will be filled by a water molecule originating from the surrounding tissue. This feature is one reason for the negligible swell that occurs when the cured implant is placed in a bath of water and its volume stability within the body.
This constant exchange of water molecules with the environment provides for nutrients, ions and water to pass through the cured implant, principally by diffusion. This feature has the benefits of equilibrating the chemistry of the implant with the chemistry of the surrounding tissue, allowing nutrients to pass from tissue layer to tissue layer across the implant, and discourages attachment of proteins that mark implants for foreign body response. This feature can render the implant uniquely biocompatible, and can reduce a significant foreign body response without interfering with transport mechanism naturally occurring in the disc.
Accordingly, another embodiment provides a nucleus implant that forms a foamed hydrogel in the body capable of allowing complex molecule diffusion through the implant volume. This aspect can be beneficial in two ways: 1) it permits the natural flow of ionic and molecular constituents between tissues, and 2) provides for diffusion out of the implant therapeutically useful chemical structures. These structures can be distributed uniformly in the implant during mixing. These structures can be water soluble or suspended. The diffusion of these structures out of the implant is controlled by diffusion, and the rate of diffusion can be controlled by adjustment of certain characteristics of the implant. In one embodiment, the implant can act as a drug delivery device, source of cell directing proteins such as growth hormones, or an initiator of cellular responses. In another embodiment, the implant can re-hydrate the often dehydrated condition of a diseased disc.
Another embodiment provides a permanent (i.e., nonabsorbable) implant. The permanence can be superior compared with protein-based and many polyol-based in situ polymerizing implants. Implant permanence can be achieved, in one embodiment, by decreasing one or more of the magnitude of implant swell, hydrolysis, oxidation, and chain breakage.
In one embodiment, the implant is an in-situ curable composition comprising prepolymers that form polyurethane within the body and possesses permanence characteristics similar to those of polyurethanes. There are two classes of polyurethanes used commercially: polyester urethanes and polyether urethanes. In one embodiment, the cured implants comprise polyether urethanes. Polyester urethanes are known to degrade more rapidly in the body. Even among polyether urethanes, there are several chemical variations that contribute to implant permanence. Many medical grade polyurethanes are not cross-linked so they can be dissolved in solvents and applied to medical devices as coatings. Polyurethanes that are not cross-linked swell in the body and degrade faster. In one example, the prepolymer has a multifunctional structure that results in a crosslinked implant. In this example, the structure of the prepolymer comprises a small multifunctional center joined to hydrophilic arms through hydrophobic linkages. One or more of the ratio of hydrophobic to hydrophilic groups and their distribution within the prepolymer, the choice of functional end groups, the molecular weight of the prepolymer, the absence of available OH groups in the prepolymer and the fully functionalized condition of the pendant structures of the prepolymer, the structure of the bonds formed during bulk polymerization of the implant, and the presence of a biologically insignificant quantity of free functional groups, can be tailored to achieve a permanent implant.
During in situ polymerization, the chain extends to prepolymer and also forms the cross links resulting in a molecular structure that is millions of Daltons in weight. Theoretically, in situ polymerized polyurethane can form a single polymeric molecule. Highly cross-linked single macromolecules do not dissolve in solvents and do not melt with heat. This resistance to phase change results in their improved permanence in the body.
Among the highly cross-linked polyurethanes, there are certain types that are more stable in the body. In one embodiment, the use of an aromatic isocyanate rather than an aliphatic isocyanate can improve implant permanence. In one embodiment, the prepolymers possess one or more of the following structures that contribute to implant permanence: 1) the prepolymer is functionalized with aromatic isocyanate, 2) the backbone of the prepolymer contains both polyethylene and polypropylene units, 3) the prepolymer is multi-functional, and 4) the multifunctionality is achieved by isocyanate chain extension with a low molecular weight triol.
Another embodiment provides a single component, in situ curable liquid implant, i.e., contains only one functional or reactive component. Currently, there are a number of in situ polymerizing implants used in the disc and other areas of the body that involve combinations of two or more reactive components that are necessary in the polymerization of the implant. In situ forming implants that rely on the union of two or more functional groups in order to form the implant, risk certain biocompatibility features, including loss of some fraction of each of these components to surrounding tissue. When these components become separated in the body, their characteristic chemical activity can present a biocompatibility risk. Frequently the separated components act on proteins within the body that result in immunotoxicity, sensitization, cellular toxicity, carcinogenicity, teratogenicity, and mutagenicity. These biological responses can contribute to adverse events in the clinical population.
In an embodiment that provides a single component in situ curable implant, if the prepolymer migrates from the implantation site, it can rapidly polymerize with itself into a biocompatible configuration. Its biocompatibility is not dependent upon the reaction of two or more distinct components that may become separated and migrate systemically within the patient. In one embodiment, the single component composition polymerizes in situ without the addition of water or any other reactive component. Any water addition is made to achieve aspects not central to the in situ curing feature; e.g., achieving a suitably low viscosity implant, achieving a desired modulus for the cured implant, and other aspects.
Single component in situ curing materials rely on some aspect of the implant preparation or implantation site chemistry to achieve curing, since their one-part formulation must be stable in its packaging. The advantages of single component implants is that the active structures in the pre-cured implant do not rely on the coupling together of two or more active structures to form a biocompatible implant. They achieve a cured implant by a transformation of a fraction of the implant volume into an active form. As each molecule is transformed it is rapidly consumed by the remaining un-transformed species. In one embodiment, the prepolymer cures first at the tissue interface, resulting in the uncured implant to be quickly isolated from the body before the ratio of activated structures reaches parity with the un-activated structures. Parity between activated and un-activated structures presents the opportunity for these parts to become isolated from each other within the body.
In one embodiment, the activated structure of a single component composition reacts more rapidly with inactivated structures than the rate at which the implant environment or implant preparation creates activated structure, resulting in more inactivated than activated species, at least in the beginning stage of polymerization. In another embodiment, the formation of biocompatible implant occurs first at the periphery of the implant volume so that unreacted species are contained.
Another embodiment provides an in situ polymerizing composition containing a high molecular weight polyethylene glycol. In most in situ polymerizing structures, including those disclosed herein, active groups are employed in the polymerization of the implant that are also potentially reactive with chemical, and cellular structures in the body. Attaching these active groups, which are typically low molecular weight, to the high molecular weight PEG can realize one or more of the following biocompatibility benefits: 1) larger molecules diffuse more slowly through tissue and liquids and thus are less likely to circulate systemically in the body, 2) the use of larger molecules reduces the molar fraction of active groups in a composition, and 3) the biologically active aspects of a functional group are often reduced by several order of magnitude when coupled with a large, biocompatible molecule. In one embodiment, the high molecular weight polyol has a molecular weight of several hundred or more Daltons (e.g., 200 D or more, 300 D or more, or 500 D or more). Reductions in toxicity as measured by LD50 assessments can be as large as an increase by 10,000 the mass of the compound required to achieve an LD50 level of toxicity.
In another embodiment, the prepolymer molecule, or collection of molecular structures possess a high molecular weight so that they do not rapidly diffuse into tissue and that their toxicity is reduced. It is well known that attaching polyethers to low molecular weight diisocyanates reduces their tissue toxicity, often by several orders of magnitude. The prepolymers, such as those of Example 3, can be formed by functionalizing diols of approximately 1000 Dalton with aromatic diisocyanate and chain extending with a triol to form prepolymer species with a mean molecular weight of 4,000 to 6,000 Dalton.
Another embodiment provides a prepolymer of sufficient hydrophilicity. For example, prepolymers comprising polyols functionalized with isocyanates are typically hydrophobic, and capable of retaining less than 10% by volume water in the polymerized structure. In embodiments where the prepolymer comprises isocyanate groups, depending on the choice of isocyanate group, the prepolymer is capable of forming a hydrogel comprising at least 70% water by volume. This feature can provide one or more of the following benefits, including: the strength of tissue bonding, the affinity for filling spatially small features of the implantation space and filling the space without voids, as well as enabling the ratio of water to prepolymer in the liquid implant to be adjusted by a medical professional to achieve a desired treatment outcome. The prepolymer component can be attracted to water aggressively and can be drawn into tissue features by the presence of water, e.g., when the mixed fraction of water is small or the prepolymer is used directly. When prepolymer is used alone as a pretreatment to a later implantation of mixed prepolymer, the prepolymer can rapidly coats and seals the tissue surfaces of the implant location. The presence of proteins on the tissue surface and the mixing that occurs as the prepolymer is drawn across the tissue surface can serve to accelerate polymerization and the formation of a sealing surface. When used in this way sparingly with water, the resulting modulus of the formed layers can be more rigid than an implant formed by mixing with a more generous quantity of water. In this two-step application of the implant, a compound structure can be formed in situ comprising a rigid outer shell and a soft inner core.
In one embodiment, the prepolymer is a polyurethane prepolymer. Such prepolymers do not generally form hydrogels with water content exceeding a few percent by volume. Many do not polymerize in the presence of water at all, but those that do generally form rigid solids. For example, moisture curing polyurethane varnishes form durable coatings with high durometer. In one embodiment, the prepolymers form a foaming hydrogel without employing hydrophobic components. In another embodiment, the prepolymers form foams that do not possess tensile strength suitable for industrial foams, coatings and the like. In another embodiment, the low molecular weight diisocyanate is consumed in end capping the polyether structures. In yet another embodiment, polyurethane prepolymer solutions contain between 5 and 25% or more of free, low molecular weight diisocyanate. Such a composition can form soft, pliant yet chemically resistant fully cross-linked polymers. The prepolymers can minimize biological toxicity by minimizing the free isocyanate content and binding active isocyanate groups to high molecular weight chains. In another embodiment, polyurethane prepolymer solutions contain between 1 and 5% of free, low molecular weight diisocyanate (e.g., less than 500 D, or less than 200 D).
Isocyanates are hydrophobic as a class, and the grafting on of high molecular weight polyethers not only increases their biocompatibility but also increases their hydrophilicity. This feature can achieve a homogeneous mixture of prepolymer and water in ratios significantly above a few percent. For example, combining isophorone diisocyanate with a 4500 Dalton triol composed of units of polypropylene oxide and polyethylene oxide in the ratio of 1:1 results in a prepolymer that precipitates out of solution before polymerizing at mix ratios above 50% water. However, combining toluene diisocyanate with a 4500 Dalton triol of pure ethylene glycol dissolves readily in any ratio with water, but the polymerization product swells greater than 100% by volume in water, exceeding this value for high water mix ratios. Thus, there is a competing condition between the potential for forming a high water content hydrogel and forming a volume stable implant.
Prepolymers formed using polyethers with polyethylene and polypropylene in the ratio 1:1 do not generally wet tissue well, and may not make contact with tissue sufficiently to provide for the formation of good tissue bonds. Furthermore, hydrophobic prepolymers require longer cure times; many compositions have cure times outside the 2-10 minute range required in medical applications. In balancing the hydrophilicity of the prepolymer, the characteristics of the polymeric backbone and the isocyanate can be considered together. For example, in building large prepolymer molecules to achieve an acceptable level of biocompatibility, in some embodiments the compositions intersperse the molecule with hard segments of the type obtained by chain extending two polyols with a diisocyanate. The existence of these hard segments can contribute to implant permanence and volume stability. In one embodiment, the composition comprises diols containing polyethylene and polypropylene in the ratio 75:25, respectively and chain extend them using toluene diisocyanate and trimethylol propane to achieve a trifunctional prepolymer with nitrogen contained centrally in the prepolymer molecule and a mean molecular weight of approximately 5000 Dalton (e.g., between 4000 and 6000 D, e.g., between 4500 and 5500 D).
Another embodiment provides an implant that can bind the cured foams previously formed by a first application of the prepolymer composition. Accordingly, this embodiment is directed to multiple applications (two or more) of the liquid implant to an implant site. The multiple implant materials resulting from the multiple applications can be bonded together. This can establish a proper shape and volume of an implant to achieve a desired clinical outcome; for example, an increase in disc height. A clinician may revise an initial implant by adding to an existing implant, either peri-operatively or at a future date. Such revisions may include treatment paradigms intended to maintain an implant's clinical efficacy, or to gradually direct a remodeling of an implant locus. For example, in the latter, staged additions of implant volume could be administered in order to increase disc height without causing failure in the surrounding tissue. The implant may be used without causing annular herniation, annular tearing, cracking or deformation of the endplates. The implant may be used protectively to segregate one tissue type from another. For example, the implant may be used to prevent extrusion of natural nucleus from the nuclear space of the disc. The implant may be used to reinforce an annulus of a disc, wherein a first application is directed between the layers of a disc annulus and a second application is directed into the nuclear space, with or without removing nuclear material.
The surgical applications can comprise one or more of 3 primary uses: 1) replacement of some or the entire disc nucleus, 2) repair and strengthen the disc annulus, and 3) bond or fit a preformed nucleus implant to a surgically prepared site. In one embodiment, tissue bonds take two forms: 1) mechanical bonds where the implant is anchored to tissue by infiltrating small-scale tissue structures and solidifying within these structures, and 2) chemical bonds which include covalent bonds as well as hydrophobic association and charge mitigated association. The integrity of these bonds depends on the stability of the implant.
In many applications, where the hole made in the annulus to remove nucleus is relatively small, e.g., <4 mm, the bond strength of the implant need not exceed 2-4 lb/in²in shear. However, the bond strength requirement has no upper limit and compositions that form stronger bonds are useful in a broader range of implant scenarios. In one embodiment, a tissue bonding implant has a bond strength that equals or exceeds the tensile strength of the implant. Suitable bond strength in shear is in the range of 10 lb/in²or higher, e.g., 25 lb/in²or higher.
Bonds strength and shelf life of the prepolymer can be enhanced by ensuring all the hydroxyl groups are terminated with isocyanate groups. In one embodiment, an excess of low molecular weight isocyanate is provided to ensure that all hydroxyl units are capped. The amount of these low molecular weight isocyanates in the prepolymer is a small molar fraction, ranging from 1% to 5% of the composition, and typically less than 1%. Larger fractions of low molecular weight may decrease the short-term biocompatibility of the implant, although the amines of these low molecular weight species are quickly eliminated from the body if they should escape the implantation site. In one embodiment, bond strength is generally increased up to a limit of a few percent of low molecular weight isocyanate.
Determining the precise amount of liquid implant to deliver to the repair site can be difficult to determine, and often relies on sensing back pressure in the syringe, observing spatial changes in the disc, or using noninvasive imaging means. Even with perfect determination of a therapeutic volume, the implant and/or surrounding tissue is likely to change in response to loading. In terms of long-term efficacy, tissue or implant may degrade requiring additional implant volume to maintain efficacy. The minimally invasive nature of implantation of liquid in situ curing implants affords a possibly of revision. A change in the mechanical properties of the spine may require implanting additional material over pre-existing material. To that end, multiple applications may be needed where the subsequently applied liquid implant bonds to its cured form. In one embodiment, the bond between an earlier implantation and a later implantation is seamless, such that when the combined implants are tested to failure that the failure locus is other than the joint between the two implants. In one embodiment, a textured surface presents an opportunity for mechanical bonding to a pre-existing implant. In one embodiment, the prepolymer bonds both mechanically and chemically to its cured form.
Another embodiment provides one-piece removal of an implant formed in situ. The cohesive strength of the implant after polymerization can be sufficient to allow clinicians to retrieve substantial portions, if not all, of an in situ formed implant in one piece.
In one embodiment, the bulk strength of the implant allows for clean removal of the implant if revision is required. Generally, implants sufficient to withstand the forces occurring in the nuclear space will satisfy this objective. Implants that do not satisfy this requirement are those that degrade in the body so that it is difficult to determine the boundary between implant and tissue. In general, the more permanent an implant is the easier it will be to remove. Other considerations involve the biocompatibility of the implant. An implant that induces a strong or chronic inflammatory response may be deeply embedded in fibrotic tissue such that removal of the implant risks damage to surrounding tissue, for example nerves. Other aspects of an implant that make it difficult to remove are its propensity for migrating in the nuclear space. Implants that tunnel into the annulus or endplates make removal of the implant difficult, and risk leaving behind a far more incompetent anatomy. In one embodiment, the implant features high tensile strength; biocompatibility, permanence, and low modulus to achieve an implant that can be remove years after implantation.
Another embodiment combines the features of tissue bonding, self-bonding, and affinity for water to achieve a liquid nucleus implant that can be employed in the reconstruction of the annulus. For example, a thin annulus may be reinforced by coating the inner layers of the annulus, coupling the inner layers of annulus to a nucleus implant, mechanically displacing an annular surface and bonding to an nucleus implant so as to correct a herniation, bonding together layers of an annulus to increase the annulus hoop strength or prevent extrusion of nucleus between layers of an annulus. Lastly, the liquid implant of the can be used to fill a void or defect in an annulus, correct a fissure or crack, redistribute loading forces in an annulus.
In diseased discs it is sometimes the case that the transition from the less structured nucleus to the laminated structure of the annulus becomes less distinct. The increase in disorder of the disc as a whole leads to failure of the annulus. When the disc height is in the normal range the orientation of the fibers in the layers of the annulus alternate in criss cross fashion from layer to layer. The angle of the fibers between two adjacent layers are closer to 90 degrees than 0 degrees. In the diseased case, the height of the disc can decrease and the angle between the fibers in adjacent layers of the annulus approaches zero. When this occurs, the organized structure of the annulus can be compromised by the radial forces generated by axial loading of the disc, which in the collapsed state can result in fibers of one layer passing through fibers of another layer. When the orientation of fibers in adjacent layers becomes more parallel, those fibers can migrate across layers resulting in herniation.
Whatever fraction of the radial force not contained by the first layer is passed on to successive layers. There is an advantage in strengthening the inner layers of the annulus because the inner layers are frequently the most disorganized. The containing capacity of the inner layers can be enhanced by erecting a barrier that resists fluid flow between the fibers of the annulus. An in situ bonding polymer applied to the inner surface of the annulus can both diminish the likelihood of fluid extrusion through the annulus, but also increase the maximum hoop stress of the layers. If the layers have already become disassociated, that is the layers can easily be dissection apart, the local expansion of the liquid implant limit infiltration along these layers of least resistance. Cadaver studies have shown that small amounts of liquid implant can flow between disassociated layers of an annulus and bond them together. If the disc height is restored during this bonding process, a preferred orientation of the fibers between layers is achieved.
When a disc becomes compressed and the volume of nucleus in the disc decreases the outer layers of the annulus bow away from the disc center, and the inner layers can bow toward the disc center. This condition results in a force within the annulus that tends to separate layers of the annulus. Nucleus can then enter a defect in an inner layer; travel circumferentially until it reaches another defect in the next layer and progress in this way until nuclear material escapes the disc. Nuclear material outside the disc results in a strong fibrotic response in an attempt to heal an abnormal condition of the spine. The fibrotic response is rather ineffective in containing the nuclear material and contributes to irritation of nerves surrounding the disc. In the extreme case, this growth of fibrotic tissue can affect the spinal cord. The result is a fissure in the annulus, which may be associated with multiple branching, circumferentially directed paths leading out of the disc. Removing nuclear material from the defect and filling the defect with a tissue bonding implant can repair this condition. Not only is this pathway blocked to further loss of nuclear material, but replacing the shrunken nuclear material with synthetic nuclear material prevents the inward bowing of the annulus and restores the cross sectional geometry of the annulus to a more normal shape.
Another embodiment eliminates the need for a molding element or balloon to isolate from tissue a toxic element of an in situ forming implant. Enclosing an in situ curing material within a distensible container such a balloon isolates tissue from the curing reaction. While this approach is attractive from a safety point-of-view it may defeat many of the therapeutic properties of such devices. Enclosing the liquid implant in a forming element may reduce, if not eliminate, the formation of mechanical and chemical bonds between cured implant and surrounding tissue. This observation may apply also to any advantage derived from strengthening the annulus. Such enclosures are by their nature impermeable, and thus are likely to block fluid flows within the disc. Lastly, motion of the implant within the disc can result in shear degradation of the implant, particulate formation, and fibrosis and eventually externalization of the implant.
Another embodiment provides the absence of neurotoxicity associated with both the prepolymer liquid implant and the polymerized result. Implant injected into the nucleus in a liquid state presents the chance that implant escapes the nucleus through a defect in the annulus, and cures outside the disc, possibly on the surface of surrounding nerves. Because the implants are chemically active, they could chemically interact with the nerves compromising their function. In studies of rats, prepolymer was placed on the exposed surface of rat brains and allowed to polymerize. The animals were monitored by EEG and behaviorally for 7 days, and histopathology was obtained. Accordingly, in one embodiment, the prepolymers do not compromise nervous function.
Another embodiment minimal to no migration of the implant. Migration of implants and their derived particulates requires an active interaction between implant and tissue that results in implant displacement through tissue. Factors contributing to implant migration include one or more of: 1) the difference in modulus of the implant and surrounding tissue, 2) the degree of inflammatory response initiated by the implant that can result in tissue death and faults along layers of tissue, 3) differential motion between the implant and surrounding tissue, 4) the propensity of the implant to chemically decompose or form particulates, 5) lack of homogeneity of the implant, and 6) the propensity of the implant to localize stresses on the implant surface.
The formation of bubbles in the cured implant can mitigate against implant migration. In one embodiment, the foam structure is closed cell, and occupies between 30 and 50% of the cured implant volume. In one embodiment, prepolymers of approximately 5000 Daltons mean molecular weight possessing three functional NCO groups per molecule and possessing a backbone comprising polyethylene and polypropylene in the ratio 75:25 generate adequate anti-migration characteristics when mixed with an equal volume of aqueous solution prior to implantation.
Another embodiment relates to the insensitivity of the therapeutic benefit to implantation volume, implantation technique and implant preparation. The formed nucleus implant can have a level-seeking feature that tends toward a physiologically beneficial disc height. In practice, surgical improvements in back pain can be achieved by reversing the normal reduction in disc height that occurs with age. Therefore, the therapeutic bias is towards increasing disc height. Increasing the height of a vertebral disc can be achieved by, first reducing the load applied to the disc annulus and second mechanically distracting the space between adjacent vertebral bodies. Both of these steps risk deforming the endplates of the vertebral bodies or herniation of the disc annulus. Removing load from the annulus involves shifting load toward the center of the vertebral endplates. The endplates are concave, and forces applied closer to their center reduce the load bearing capacity of the endplates. The endplates can fail suddenly or the natural concavity of the endplates can be accentuated, resulting in the peripheral surfaces of the endplates moving together. There are other structures coupled to the endplates that also can cause back pain, such as a change in the spatial relation of the facet joints.
Mechanically distracting the space between endplates is generally achieved by applying pressure to the endplates, and thus this step also risks damaging the endplates. The damage is more frequently sudden in nature, and may involve fracturing the endplates. Recently, strategies have been contemplated where an implant is placed in the nuclear space under pressure, or the implant swells to develop a pressure. In both cases the goal is distraction of the endplates. In some cases the forces are axially directed, but in other cases, especially where a balloon is used or a homogenous implant is formed in situ the forces tend to be more isotropic, applying forces equally to the endplates and the annulus. This can contribute to herniation of the annulus resulting in the painful condition of pressure being applied to nerves surrounding the disc. These risks can be categorized as acute or chronic. Their associated adverse events are linked to non-accommodative features of the therapy that generally place non-physiological forces on the spine or create non-physiologic spatial relations between spinal elements. The acute risks and associated adverse events can be avoided by applying the reconstructive therapy over a longer interval of time. Therefore, there is a need for a means to apply corrective forces to the spine over extended periods of time. However, coupled with this need is the need to guard against applying forces that are excessive, and which may create spatial relationships between the various elements of the spine that cause pain. These needs can be satisfied in an implant that is both accommodative and applies its beneficial aspects over time. An implant is accommodative if its mechanical characteristics are in part determined by the mechanical characteristics of the spine. In this respect, the bimodular aspect can be useful in applying axial loading forces only when the disc height has deteriorate to a certain degree.
Another accommodative feature is the presence of bubbles in the implant. In addition to having a nonlinear compliance, they also have the capacity to decrease in size depending on the loads applied to the implant. For example, sustained loads tend to force the gas in the bubbles to dissolve into the liquid of the solid hydrogel component of the implant. Accordingly, the bubbles collapse with no additional load needed. A factor on whether the bubbles collapse is the duration of the applied load. Conversely, when the load is reduced the gas dissolved in the implant can reform as gas, re-establishing the additional compliance constraint associated with re-inflating the bubbles in the implant. The bubbles provide additional accommodative features. For example, over several days the implant is likely to relax to a certain disc height. The degree of relaxation will depend on the ratio of the components in the prepolymer composition, the magnitude of pressure developed in the implants as it polymerizes, as well as other aspects. If the implant has relaxed to an undesirable degree, the clinician has the option to inject additional implant over the existing implant, thereby achieving a less compliant implant. But in general, the implant will establish a therapeutically beneficial disc height that balances deformation of the endplates (axial force) with deformation of the annulus (radial force). The initial ratio of radial to axial forces and the time progression of this ratio are selected by the medical professional according to the paradigm outlined in the Appendix.
Once this equilibrium is achieved in the patient, which can additionally depend on the confluence of acute and chronic forces and the way in which the implant responds to these forces, there is a need for the implant to become less accommodative. This can be achieved naturally by the implant as the bubbles in the implant fill in with water, thereby increasing the Poisson ratio of the implant. This infiltration of bubbles with water fixes both the equilibrated disc height and ratio of axial and radial forces. The principal change precipitated by the filling of the implant bubbles is the implant's accommodative nature.
In one embodiment, liquid implants curing to a solid hydrogel with a gaseous fraction possess some degree of level accommodation as the gas in the bubbles is dissolved in the aqueous phase and water migrates to fill the bubbles of the implant and a mean disc height is established. These implants will possess a Poisson ratio less than 0.5. The Poisson ratio can be controlled independently of the fraction of gas liberated by controlling the pressure at which the implant cures. In one embodiment, prepolymers with a % NCO of between 2 and 5 provide adequate gas production to be Poisson ratio adjustable to suit the treatment range by applying pressure. In another embodiment, low-pressure applications, the % NCO ranges from 1.0 to 2.5%.
Another embodiment relates to the ability of the implant to distribute load forces between axial and radial components. This distribution of forces occurs acutely as the implant polymerizes and chronically as it responds to the unique conditions of the implant environment. Accordingly, the implants tend to have a lower modulus than the nuclear implants of the past. This embodiment makes maximal use of any structural strength existing in the affected disc both during implantation and subsequently. For example, to minimize the force per unit area applied to the endplates the implant fills the nuclear space and distributes localized forces. Similarly the annulus can be made to bear more of the load when the implant applies radial forces to the annulus, making the annulus stiffer in the axial direction. In most disc disease states, the normal equal allocation of load to the central portions of the endplates and the annulus is compromised, with most of the load directed to the center of the endplates. Placing a localized implant in the nucleus that is deformable and space filling reallocates to some degree the distribution of load between the central portion of the endplates and the annulus. This reallocation can be determined by a number of factors, some of which depend on the hoop strength of the annulus, equilibrated disc height, and the degree to which the bubbles in the implant are pressurized. While the dynamics of this allocative process are complex, the mechanism for achieving reallocation of load forces can depends upon the ability of the implant to apply radial forces to the disc annulus.
For suitably soft cured implants, as those detailed in Table 1, axial forces applied to a disc can cause the implant to expand radial, apply force to the annulus and cause it to stiffen. The increased stiffness of the annulus can result in the annulus supporting a greater proportion of the total load upon further disc compression. A component of the level-seeking feature is that the distribution of forces between the nucleus and annulus is modulated by the application of radial forces by the nuclear material. An implant of very high modulus will apply no forces to the annulus, and substantially all the load will be supported by the implant. In one embodiment for a liquid implant, the axial force and radial force is equal within the implant. For Poisson ratio 0.5 implants of moderate modulus, the modulus determines the partition of axial and radial forces. For an implant with a Poisson ratio of less than 0.5, the volume displacement of the implant in the axial direction does not have to equal the volume displacement in the radial direction, and allows for the height at which the implant modulus is applied to vary. This can allow the implant to respond to a broad spectrum of loading frequencies. The bubbles in the implant absorb higher frequencies, and accordingly the implant need not dilate as far radial potentially protecting the implant from fracture. Lower frequency and mean loads set a mean compression of the implant bubbles, the gas in these bubbles eventually dissolve into the implant as liquid fills the bubbles. When the bubbles are filled, the height of the implant is now fixed at a mean height and the modulus of the hydrogel component of the implant dominates because the compressibility of the implant goes to zero. At this established height, the load and the modulus of the implant determine the axial and radial dimensions of the implant.
One embodiment provides a liquid nucleus implant sufficient to provide the therapeutic effect of strengthening and/or filling the intervertebral space and preventing extrusion of the polymerized prosthetic. Although a variety of in-situ polymerizing liquids may be used, both adhesive and non-adhesive, an exemplary in-situ polymerizing liquid is a single-component polyisocyanate based adhesive as described in U.S. Pat. Nos. 6,254,327, 6,296,607, and U.S. Pub. Nos. 2005/0129733 and 2005/0215748, the disclosures of which are incorporated herein by reference.
In one embodiment, the prepolymer comprises a polyisocyanate-capped polymeric polyol and a small amount of free poly isocyanate (e.g., 1% to 5% by weight). Such materials and their synthesis are described in detail in U.S. Pat. No. 6,524,327, the disclosure of which is incorporated herein by reference. The small amount of excess polyisocyanate, typically of molecular weight less than about 1000 Daltons, maximizes the reactivity of the polyols, and by directly and rapidly reacting with tissue, promotes bonding of the adhesive to tissue. Typically the small isocyanate contains up to about 3% of the number of active isocyanate groups on the polymer. The small isocyanate may be all or part low molecular weight capped diol. The capped polyol is multifunctional, and typically is trifunctional or tetrafunctional, or a mixture of trifunctional and/or tetrafunctional with difunctional. The polyol is can be at least in part a polyether polyol.
The polyisocyanate can be difunctional, tri- or tetrafunctional, or star forms of isocyanate. Branching (tri- or tetra-functionality) may be provided by a trifunctional polymer, or by providing a tri- or tetrafunctional low molecular weight polyol, such as glycerol, erthyritol or isomer, or trimethylolpropane (TMP). Fast reacting formulations use an aromatic diisocyanate such as toluene diisocyanate. Slow reacting formulations use an aliphatic diisocyanate such as isophorone diisocyanate. Many additional diisocyanates are potentially useful. Some are listed in U.S. Pat. No. 6,524,327, and these and others are found in chemical catalogs, for example from Aldrich Chemical. Alternatively, the polymerization time can be adjusted by selecting appropriate molecular weight polyols. The higher molecular weight polyols produce lower viscosity capped reaction products and slower reacting solutions. However, at any molecular weight of the polyol(s), the reaction rate is most significantly determined by the reactivity of the functional end group attached to the polyol.
In one embodiment, the prepolymer is an adhesive prepolymer. In other embodiments, nonadhesive prepolymers can be used. An adhesive for use in the invention can be hydrophilic in character, and can also be water-soluble before being crosslinked. This hydrophilicity enable the adhesive to be injected into tissue to polymerize in contact with, and bond to, the tissue, as adhesive and/or as local bulking agent to fill gaps or fissures, or to stabilize implants. The adhesive acts as a self-sealing fluid when injected into cavities or gaps. Once cured in situ, the hydrophilic adhesive will absorb fluid from the tissue, forming a structure that will be at least somewhat gel-like in character. The cured adhesive will swell to a controlled extent, exerting a controlled amount of local pressure.
The tensile properties of the cured adhesive can be adjusted so that the adhesive, like the native tissues of the annulus or of the nucleus, deforms under pressure while exerting a restorative force on the surrounding structures. Hence, the adhesive-tissue composite tends to return to its original shape and location after movement of the spine and is characteristically elastic. These properties can be controlled by the composition of the adhesive, or by providing a controlled degree of dilution with saline at the time of administration. This is in contrast with rigid materials, which tend to fracture rather than yield, and to flowable media, which have no tendency to return to their original shape after relaxation of stress. For example, hydrophobic adhesives tend to become rigid, favoring fracture of the cured adhesive at the surface of the tissue or implant. They also tend not to bond to tissue, which is highly hydrophilic.
Polymeric compositions other than isocyanate-capped polyols can be suitable. The cured implant may be stable in the body, or may degrade in the body to smaller, excretable molecules (“degradable”). A wide variety of linkages are known to be unstable in the body. These include, without limitation, esters of hydroxy acids, e.g., alpha and beta hydroxy carboxylic acids; esters of alpha and beta amino acids; carboxylic acid anhydrides; phosphorous esters; and certain types of urethane linkages. Generally, the cured implant is stable in the body for prolonged periods, as the fibrous materials of the annulus have very limited self-repair capabilities, and the nucleus has virtually none. However, if methods are found to enhance natural biological repair of the nucleus or annulus, then degradable adhesives or fillers could be used.
The prepolymers can have reactive groups covalently attached to them, or part of the backbone. The reactive groups are suitable for reaction with tissue, and for crosslinking in the presence of water or components of bodily fluids, for example water and protein. Suitable groups include isocyanate, isothiocyanate, anhydrides and cyclic imines (e.g., N-hydroxy succinimide, maleimide, maleic anhydride), sulfhydryl, phenolic, polyphenolic, and polyhydroxyl aromatic, and acrylic or lower alkyl acrylic acids or esters. Such reactive groups are most commonly bonded to a preformed polymer through suitable linking groups in the polymer. Commonly found linking groups include, without limitation, amines, hydroxyls, sulfhydryls, double bonds, carboxyls, aldehydes, and ketone groups. Of these groups, aliphatic hydroxyls are among the most widely used.
Thus, suitable base polymers include poly(alkyl)acrylic acids and polyhydroxyalkyl acrylates, polysaccharides, proteins, polyols, including polyetherpolyols, polyvinyl alcohol, and polyvinylpyrrolidone, and these same structures with amine or sulfur equivalents, such as polyethyleneimine, aminosugar polymers, polyalkylamine substituted polyethers, and others. Any of these polymers can be substituted with two or three reactive groups, as is required to form a crosslinkable polymer. When there are many substitutable linking groups, as with polysaccharides, only a few of the substitutable groups (here, mostly hydroxyls) should be substituted, and the derivatized polymer will have a somewhat random substitution. In one embodiment, the hydrophilic polymer will have only a few substitutable linking groups. Polyether polyols grown on glycol or amine starters will typically have reactive groups only at the end of the polyether chains, allowing for detailed control of stoichiometry. In one embodiment, the base polymer is a polymer of ethylene glycol, or a copolymer of ethylene glycol with one or more of propylene glycol, butylene glycol, trimethylene glycol, tetramethylene glycol, and isomers thereof, wherein the ratio of ethylene glycol to the higher alkanediol in the polymer is sufficient to provide substantial water solubility at room or body temperature. Such polymer substrates can be synthesized by known methods. Preformed polyetherpolyols can be purchased, optionally in a prequalified medical grade, from any of numerous catalogs or manufacturers.
The prepolymer can be liquid at room temperature (ca. 20° C.) and body temperature (ca. 37° C.), for ease of administration and of mixture with additives, etc. The prepolymer can be stable in storage at room temperature, when protected from moisture and light.
The prepolymer may be supplemented by the addition, during manufacture or at the time of administration, of ancillary materials. These may include reinforcing materials, drugs, volume or osmotic pressure controlling materials, and visualization aids for optical, fluoroscopic ultrasound or other visualization of fill locations. Reinforcing materials may include particulate materials, fibers, flocks, meshes, and other conventionally used reinforcers. These be commercial materials can be approved for in vivo medical use. Visualization materials include a wide variety of materials known in the art, such as, among others, small particles of metals or their oxides, salts or compounds for fluoroscopy, gas-filled particles for ultrasound, and dyes or reflecting particles for optical techniques.
Osmotic properties can be adjusted for immediate or long-term effects. For example, polyether polyol isocyanates have little ionic charge either before or after polymerization. However, in some situations, as described below, it may be desirable to have a controlled degree of swelling in water after curing. This can be controlled in part by the ratio of ethylene glycol to other polyols in the formulation. It can also be adjusted by adding charged groups to the formulation. A simple method is to add charged polymers or charged small molecules to the adhesive at the time of application, for example dissolved in an aqueous solution. Charged polymers, such as polyacrylic acids, will react poorly with the isocyanates, but will tend to be trapped in the polymerized matrix. They will tend to increase the swelling of the cured material. In turn, this would allow the use of higher proportions of non-ethylene glycol monomers in the polyols. Alternatively, charge could be introduced by addition of hydroxy carboxylic acids, such as lactic acid, or tartaric acid, during synthesis or during administration. Added polymers could instead be polyamines, but, to avoid rapid polymerization, should be tertiary or quaternary amines or other amine types that will not react with isocyanate. A method of increasing swelling is to incorporate higher concentration of diffusible ions, such as soluble salts—e.g., sodium chloride—into the adhesive at the time of application. The salt will attract water into the adhesive polymers; after polymerization, the salt will diffuse away and the gel will remain expanded.
The prepolymer can be adjusted in several ways to optimize its post-cure properties for the particular situation. In one embodiment, adjustment of properties is achieved by dilution of the polymer with water, saline, or other aqueous solution. A typical dilution would be in the range of 5% or less (volume of saline in liquid polymer), for formation of dense, high-tensile, low-swelling deposits, up to about 95% (19 vol. saline/vol. polymer) for readily swelling, highly compliant deposits or bonds. In formulation, allowance must be made for the amount of water that will flow into the adhesive from the tissue during reaction. This will usually be relatively small for bulk deposits, but is of more concern for thin adhesive layers. In thin layers, fast-curing compositions can be used, such as compositions with a higher proportion of aromatic diisocyanates. In general, dilution will reduce the tensile strength and the modulus. The amount of dilution will tend to be different depending on whether the modulus or tensile strength is to match that of the annulus (higher) or the nucleus (lower).
Various non-reactive ingredients can be added to the polymer solution either in the prepolymer or in the aqueous solution to alter the hydrogel mechanical properties, e.g., tensile strength, elasticity and bubble size. Inert particulate such as tantalum powder will result in bubble nucleation and a finer bubble size, increase the modulus of the hydrogel, and make the hydrogel radio opaque. Emulsifiers can be added to increase mix homogeneity, reduce bubble size, and provide a higher elongation at break. It is possible to use the same diol used to construct the prepolymer as an emulsifier. Alternatively, a higher or lower molecular weight diol may be used. The ratio of EO/PO can be altered to increased mixability, or pure forms of EO or PO can be used. When pure EO is used, the mixture of prepolymer and aqueous solution becomes non-Newtonian, and tends to take on a stringy consistency, which can further improve elasticity.
Other adjustable factors include the molecular weight of the polymer, and its degree of branching; and its hydrophilicity, which is a function of the particular polyol or polyols used in the formulation. In addition, additives, as described above, can also influence these properties.

Polymeric Compositions

Disclosed herein are liquid preparations for use in medicine. The liquid preparation contains a reactive polymer, which comprises a “base polymer” or “backbone polymer”, reactive groups on the backbone polymer, and a slight excess of “free” (low molecular weight) polyreactive molecules. The liquid composition can be prepared by a method requiring no catalysts and essentially no solvent. The reactive liquid polymer is self-curing when applied to tissue, by absorption of water and other reactive molecules from the tissue. The cured polymer can seal tissue to tissue, or to devices; to apply a protective coating to tissue; to form an implant within or upon tissue; to deliver drugs. The cured polymer can be provided with biodegradable groups, and has a controllable degree of swelling in bodily fluids.

Backbone Polymers

In one embodiment, the backbone polymer comprises a polymeric segment, of molecular weight about 500 D or more, e.g., about 1000 to about 10,000 D, optionally up to about 15 kD or 20 kD. The backbone polymer can contain groups that can be easily derivatized (“capped”) to form the final reactive group. Such groups can be alcohols or amines, or optionally sulfhydryls or phenolic groups. Examples include polymers such as a polymeric polyol, or optionally a polymeric polyamine or polyamine/polyol. In one embodiment, the polyols are polyether polyols, such as polyalkylene oxides (PAOs), which may be formed of one or more species of alkylene oxide. The PAO, when comprising more than one species of alkylene oxide, may be a random, block or graft polymer, or a polymer combining these modes, or a mixture of PAO polymers with different properties. Exemplary alkylene oxides are ethylene oxide and propylene oxide. Other oxiranes may also be used, including butylene oxide. PAOs are typically made by polymerization onto a starter molecule, such as a low molecular weight alcohol or amine, e.g., a polyol. Starting molecules with two, three, four or more derivatizable alcohols or other derivatizable groups can be used. The multi-armed PAOs obtained from such starters will typically have one arm for each group on the starter. PAOs with two, three or four terminal groups can be used. Mixtures of PAOs or other backbone polymers, having variable numbers of arms and/or variation in other properties can be used.
Common polyols useful as starters are aliphatic or substituted aliphatic molecules containing a minimum of 2 hydroxyl or other groups per molecule. Since a liquid end product is desired, the starters can be of low molecular weight containing less than 8 hydroxyl or other groups. Suitable alcohols include, for illustration and without limitation, adonitol, arabitol, butanediol, 1,2,3-butanetriol, dipentaerythritol, dulcitol, erythritol, ethylene glycol, propylene glycol, diethylene glycol, glycerol, hexanediol, iditol, mannitol, pentaerythritol, sorbitol, sucrose, triethanolamine, trimethylolethane, trimethylolpropane. Small molecules of similar structures containing amines, sulfhydryls and phenols, or other groups readily reactive with isocyanates, are also useable.
The PAO, or other backbone polymer, may optionally incorporate non-PAO groups in a random, block or graft manner. Non-PAO groups are optionally used to provide biodegradability and/or absorbability to the final polymer. Groups providing biodegradability are well known. They include hydroxy carboxylic acids, aliphatic carbonates, 1,4-dioxane-2-one (p-dioxanone), and anhydrides. The hydroxy carboxylic acids may be present as the acid or as a lactone or cyclic dimmer, and include, among others, lactide and lactic acid, glycolide and glycolic acid, epsilon-caprolactone, gamma-butyrolactone, and delta-valerolactone. Amino acids, nucleic acids, carbohydrates and oligomers thereof can be used to provide biodegradability. Methods for making polymers containing these groups are well known, and include, among others reaction of lactone forms directly with hydroxyl groups (or amine groups), condensation reactions such as esterification driven by water removal, and reaction of activated forms, such as acyl halides. The esterification process involves heating the acid under reflux with the polyol until the acid and hydroxyl groups form the desired ester links. The higher molecular weight acids are lower in reactivity and may require a catalyst making them less desirable.
The backbone polymers may also or in addition carry amino groups, which can likewise be functionalized by polyisocyanates. Thus, the diamine derivative of a polyethylene glycol could be used. Low molecular weight segments of amine containing monomers could be used, such as oligolysine, oligoethylene amine, or oligochitosan. Low molecular weight linking agents, as described below, could have hydroxyl functionality, amine functionality, or both. Use of amines will impart charge to the polymerized matrix, because the reaction product of an amine with an isocyanate is generally a secondary or tertiary amine, which may be positively charged in physiological solutions. Likewise, carboxyl, sulfate, and phosphate groups, which are generally not reactive with isocyanates, could introduce negative charge if desired. A consideration in selecting base polymers, other than PAOs or others that react only at the ends, is that the process of adding the reactive groups necessarily requires adding reactive groups to every alcohol, amine, sulfhydryl, phenol, etc. found on the base polymer. This can substantially change the properties, e.g., the solubility properties, of the polymer after activation.

Reactive Groups

The base or backbone polymer is then activated by capping with low molecular weight (LMW) reactive groups. In one embodiment, the polymer is capped with one or more LMW polyisocyanates (LMW-PIC), which are small molecules, typically with molecular weight below about 1000 D, more typically below about 500 D, containing two or more reactive isocyanate groups attached to each hydroxyl, amine, etc of the base molecule. After reaction of the LMW-PIC with the backbone, each capable group of the backbone polymer has been reacted with one of the isocyanate groups of the LMW-PIC, leaving one or more reactive isocyanates bonded to the backbone polymer via the PIC. The LMW-PIC are themselves formed by conjugation of their alcohols, amines, etc. with suitable precursors to form the isocyanate groups. Starting molecules may include any of those mentioned above as starting molecules for forming PAOs, and may also include derivatives of aromatic groups, such as toluene, benzene, naphthalene, etc. The LMW-PIC for activating the polymer can be di-isocyanates, e.g., toluene diisocyanate (TDI) and isophorone diisocyanate, both commercially available. When a diisocyanate is reacted with a capable group on the base polymer, one of the added isocyanates is used to bind the diisocyanate molecule to the polymer, leaving the other isocyanate group bound to the polymer and ready to react. As long as the backbone polymers have on average more than two capable groups (hydroxyl, amine, etc.), the resulting composition will be crosslinkable.
A wide variety of isocyanates are potentially usable as LMW-PICs. Suitable isocyanates include 9,10-anthracene diisocyanate, 1,4-anthracenediisocyanate, benzidine diisocyanate, 4,4′-biphenylene diisocyanate, 4-bromo-1,3-phenylene diisocyanate, 4-chloro-1,3-phenylene diisocyanate, cumene-2,4-diisocyanate, cyclohexylene-1,2-diisocyanate, cyclohexylene-1,4-diisocyanate, 1,4-cyclohexylene diisocyanate, 1,10-decamethylene diisocyanate, 3,3′dichloro-4,4′biphenylene diisocyanate, 4,4′diisocyanatodibenzyl, 2,4-diisocyanatostilbene, 2,6-diisocyanatobenzfuran, 2,4-dimethyl-1,3-phenylene diisocyanate, 5,6-dimethyl-1,3-phenylene diisocyanate, 4,6-dimethyl-1,3-phenylene diisocyanate, 3,3′-dimethyl-4,4′diisocyanatodiphenylmethane, 2,6-dimethyl-4,4′-diisocyanatodiphenyl, 3,3′-dimethoxy-4,4′-diisocyanatodiphenyl, 2,4-diisocyantodiphenylether, 4,4′-diisocyantodiphenylether, 3,3′-diphenyl-4,4′-biphenylene diisocyanate, 4,4′-diphenylmethane diisocyanate, 4-ethoxy-1,3-phenylene diisocyanate, ethylene diisocyanate, ethylidene diisocyanate, 2,5-fluorenediisocyanate, 1,6-hexamethylene diisocyanate, isophorone diisocyanate, lysine diisocyanate, 4-methoxy-1,3-phenylene diisocyanate, methylene dicyclohexyl diisocyanate, m-phenylene diisocyanate, 1,5-naphthalene diisocyanate, 1,8-naphthalene diisocyanate, polymeric 4,4′-diphenylmethane diisocyanate, p-phenylene diisocyanate, 4,4′,4″-triphenylmethane triisocyanate, propylene-1,2-diisocyanate; p-tetramethyl xylene diisocyanate, 1,4-tetramethylene diisocyanate, toluene diisocyanate, 2,4,6-toluene triisocyanate, trifunctional trimer (isocyanurate) of isophorone diisocyanate, trifunctional biuret of hexamethylene diisocyanate, and trifunctional trimer (isocyanurate) of hexamethylene diisocyanate.
In general, aliphatic isocyanates will have longer cure times than aromatic isocyanates, and selection among the various available materials will be guided in part by the desired curing time in vivo. In addition, commercial availability in grades suitable for medical use will also be considered, as will cost. Toluene diisocyanate (TDI) and isophorone diisocyanate (IPDI) can be used. The reactive chemical functionality of the liquids can be isocyanate, but may alternatively or in addition be isothiocyanate, to which all of the above considerations will apply.

Physical Properties of the Cured Implant

The polymerizable materials are typically liquids at or near body temperature (i.e., below about 45° C.), and can be liquid at room temperature, ca. 20-25° C., or below. The liquids are optionally carriers of solids. The solids may be biodegradable or absorbable. The liquid polymerizable materials are characterized by polymerizing upon contact with tissue, without requiring addition of other materials, and without requiring pretreatment of the tissue, other than removing any liquid present on the surface(s) to be treated. A related property of the polymerizable materials is that they are stable for at least 1 year when stored at room temperature (ca. 20-25 degrees C.) in the absence of water vapor. This is because the material has been designed so that both the reaction that polymerizes the polymers, and the reactions that optionally allow the polymer to degrade, both require water to proceed.
In contrast to previous formulations, the polymeric polyisocyanates contain a low residual level of low molecular weight (LMW) polyisocyanates (PIC). For example, the final concentration of LMW-PIC isocyanate groups in the formulation, expressed as the equivalent molarity of isocyanate groups attached to LMW compounds, is normally less than about 1 mM (i.e., 1 mEq), e.g., less than about 0.5 mEq, or even less than about 0.4 mEq. In one embodiment, the level of LMW isocyanate groups is finite and detectable, for example greater than about 0.05 mEq, or greater than about 0.1 mEq. It is believed that having a low but finite level of LMW-PIC molecules tends to promote adherence between the applied polymer formulation and the tissue being treated. However, decreased levels of LMW-PIC may tend to decrease tissue irritation during application and cure of the liquid polymer preparation. It is believed that the range of about 1 mEq to about 0.05 mEq is approximately optimal. In situations requiring tissue adherence in the presence of significant biological fluid, or in adherence to difficult tissues, greater levels of LMW-PIC isocyanate groups may be used.

Exemplary Polymer Structures

There are several ways in which the above-recited steps can be used to obtain a suitable liquid reactive polymer system. In a simple system, a polymeric polyol with a number of end groups on average greater than two is treated with a slight excess of a LMW-PIC, such as toluene diisocyanate. The reaction product is formed under nitrogen with mild heating, e.g., by the addition of the LMW-PIC to the polymer. The product is then packaged under nitrogen, typically with no intermediate purification.
An exemplary biodegradable polyol composition includes a trifunctional hydroxy acid ester (e.g., several lactide groups successively esterified onto a trifunctional starting material, such as trimethylolpropane, or glycerol). This is then mixed with a linear activated polyoxyethylene glycol system, in which the PEG is first capped with a slight excess of a LMW-PIC, such as toluene diisocyanate. Then mixing together the activated polyoxyethylene glycol and the lactate-triol forms the activated polymer. Each lactate triol binds three of the activated PEG molecules, yielding a prepolymer with three active isocyanates at the end of the PEG segments, and with the PEG segments bonded together through degradable lactate groups. In the formed implant, the lactate ester bonds gradually degrade in the presence of water, leaving essentially linear PEG chains that are free to dissolve or degrade. Interestingly, in this system, increasing the percentage of degradable crosslinker increases rigidity, swell and solvation resistance in the formed polymer.
Other polyol systems include hydroxy acid esterified linear polyether and polyester polyols optionally blended with a low molecular weight diol. Similarly, polyester and polyether triols esterified with hydroxy acid are useful. Other polyol systems include the use of triol forming components such as trimethylolpropane to form polyols having three arms of linear polyether chains.
Exemplary materials are described in U.S. Pat. No. 6,254,327, and U.S. Pub. Nos. 2003/0135238 and 2004/0068078, the disclosures of which are incorporated herein by reference.
These embodiments provides methods and compositions for treating intervertebral disc disorders by providing a disc nucleus implant having one or more of the following: 1) the cured implant comprises solid, liquid and gas phases, 2) the cured implant fills the entire volume of an implantation formed in a vertebral disc, 3) the liquid implant bonds to tissue as it cures, 4) the liquid implant is of sufficiently low viscosity to be delivered by conventional means via a tube directed to the implantation site, 5) the liquid implant provides a cure time short enough to ensure the liquid implant is localized to an intended implantation site and long enough to allow the implant to be delivered to the site clinically, 6) the cured implant does not change volume beyond a therapeutic range, 7) the liquid implant releases a gas phase while curing that acts to pressurize the implantation site as the implant cures, 8) the liquid implant contains a radio-opaque or illuminating marker, 9) the cured implant possesses a bimodular compliance, 10) the compliance of the cured implant can be adjusted in a predictable manner by the medical professional, 11) the liquid implant seals against tissue surfaces within the disc and at the implantation access to prevent loss of the native nucleus pulposus or implant while it cures, 12) the cured implant can exchange its water phase with surrounding tissue, 13) the cured implant allows nutrient diffusion through the implant, 14) the cured implant is clinically permanent or at least resides functionally useful in the body longer than other described nucleus implants, 15) the liquid implant is comprised of one functional part, 16) the functional part of the liquid implant comprises chemical species that are stable when store together and the minimum molecular weight is large enough to improve biocompatibility, 17) the liquid implant is hydrophilic, 18) the liquid implant bonds to existing implant, 19) the cured implant is clinically removable in one piece, 20) the liquid implant provides clinically significant reinforcement of the disc annulus when cured, 21) the liquid implant does not require a molding element or balloon in order to be safe and effective, 22) the liquid and cured implant is not neurotoxic, and more generally is biocompatible, 23) the cured implant possesses a biocompatibility and modulus that reduces the chance of implant degradation and migration through the annulotomy, 24) the liquid implant is self-adjusting within the body so as to result in a therapeutic volume that makes the clinical outcome insensitive to method of implantation, and 25) the cured implant translates axial forces originating at the vertebral endplates into radial forces applied to the disc annulus.
Another embodiment provides the use of preformed foams, which is distinct from the use of an in situ curable implant material. The preformed foams can have one or more of the same properties as the cured foam materials as described herein. These foams are sufficiently compressible to allow implantation through an opening having dimensions smaller than that of the disc space. Exemplary preformed foams include the nonabsorbable foams comprising polyurethanes, polytetrafluoroethylene, silicone foams, epoxies, and polyvinyl chloride foam.
One embodiment provides a method of repairing a defect in a spinal disc space, comprising:
inserting a nonabsorbable, closed cell foam having a Poisson ratio of less than 0.5 into the defect.
The inserting can be performed with a delivery catheter having a lumen of sufficient large diameter for inserting a preformed foam. In one embodiment, prior to the inserting, the method further comprises removing some or all of the nucleus pulposus within the spinal disc space, and the inserting results in replacement of the removed nucleus pulposus with the foam.
In one embodiment, the removing further comprises removing portions of the annulus fibrosus in the vicinity of the nucleus pulposus.

EXAMPLES

Example 1

Preparation of Prepolymer

In this example an isocyanate terminated diol is trifunctionalized to yield a slow curing tissue adhesive. The type and amount of isocyanate to be used is 326.27 g of isophorone diisocyanate (IPDI). A suitable IPDI is Desmodur I. The type and amount of diol to be used is 749.94 g of 75:25 diol comprised of 75% polyethylene glycol and 25% polypropylene glycol. A suitable diol is Ucon 75-H-450, with a molecular weight of 978 Daltons and hydroxyl number of 119.4. The type and amount of triol to be used is 23.79 g of trimethylol propane. The theoretical target for completion of the diol termination steps is % NCO=5.23%. The theoretical target for completion of the trifunctionalization step is % NCO=3.09%. Final temperature pre-TMP was 80° C. The NCO levels at various times are: at 28 hrs 6.197%, at 56 hrs 5.468%, at 78 hrs 5.421, and at 126 hrs 5.23%. The TMP was added at hour 127. The final NCO of % NCO=3.09% was reached at hour 271. The viscosity at 34° C. was 103 Kcps.
The TMP and glycols should be deionized and dried. All of the diol and isocyanate are to be added at once.
The temperature in the reacting chamber should follow the schedules described above, and the % NCO at the described time points should follow the values recorded above. The reaction should be conducted under vacuum with a trickle flow of argon.
The reactor is a standard cylindrical glass 1 Liter reactor with a stir rod comprising 2 reactor blades of 55 mm diameter with 5 blades oriented 45° from the axis. The rate of mixing is 220 rpm.
Under these conditions the prepolymer is comprised of a broad distribution of chain lengths in the diol termination phase with a minimum of side reactions. This distribution cannot be achieved solely by adding diols of molecular weights in the ratio obtained in the final synthesis product, since the actual synthesis process can affect the final chain length distribution. Adding the diols in this ratio at the beginning of the synthesis process results in a prepolymer that is unusable as a tissue adhesive. Calling the single chain length of 978 Dalton the monomer, the following distribution is obtained after the diol termination process.


	Actual Value	Useful Range

Monomer	28.2%	+/−10%
Dimer	20.0%	+/−10%
Trimer	14.8%	+/−5%
Tetramer	10.9%	+/−2%
Pentamer	9.9%	+/−1%
Hexamer	6.8%	+/−1%
Heptamer	5.2%	+/−1%
Octamer	3.3%	+/−1%
Nonamer	1.8%	+/−0.5%

Example 2

Prepolymer Synthesis

The synthesis is the same as Example 1, except that the diol is added in 1% increments rather than all at once to the isocyanate. Each 1% increment of diol added to the reacting isocyanate is made after the exotherm of the previous addition is complete. This step-wise addition yields the following distribution of terminated diols:


	Actual Value	Useful Range

Monomer	55.3%	or more
Dimer	27.1%	or less
Trimer	8.5%	or less
Tetramer	4.7%	or less
Pentamer	2.5%	or less
Hexamer	1.3%	or less
Heptamer	0.6%	or less

Example 3

Prepolymer

In this example an isocyanate terminated diol is trifunctionalized to yield a fast curing tissue adhesive. Fast adhesives cure within 5 minutes when used neat and applied to tissue. Slow adhesives cure after this time, generally 5 to 10 times longer. The type and amount of isocyanate to be used is 270.26 g of toluene diisocyanate (TDI). A suitable TDI is Rubinate, a mixture of 80% 2-4 and 20% 2-6 isomers. The type and amount of diol to be used is 870.53 g of Ucon 75-H-450. The type and amount of triol to be used is 9.21 g of trimethylol propane. The theoretical target for completion of the diol termination steps is % NCO=4.55%. The theoretical target for completion of the trifunctionalization step is % NCO=3.76%. Final temperature pre-TMP was 50° C. The NCO levels at 25 hrs 4.78% and at 75 hrs 4.55%. Then the TMP was added at hour 76. The final NCO of % NCO=3.67% was reach at hour 100. The viscosity at 31° C. was 24.5 Kcps.

Example 4

Prepolymer

In this example an isocyanate terminated diol is trifunctionalized to yield a fast curing tissue adhesive with a ratio of soft-to-hard centers greater than that achieved in Example 3. The type and amount of isocyanate to be used is 231.65 g of toluene diisocyanate (TDI). The type and amount of diol to be used is 870.53 g of Ucon 75-H-450. The type and amount of triol to be used is 9.21 g of trimethylol propane. The theoretical target for completion of the diol termination steps is % NCO=3.90%. The theoretical target for completion of the trifunctionalization step is % NCO=2.69%. Final temperature pre-TMP was 50° C. The NCO levels at 23 hrs 3.80% and at 75 hrs 4.55%. Then the TMP was added at hour 23. The final NCO of % NCO=2.69% was reach at hour 72. The viscosity at 30° C. was 48 Kcps.

Example 5

Prepolymer

In this example two isocyanate terminated diols are randomly trifunctionalized to yield a fast curing, absorbable tissue adhesive. The type and amount of isocyanate to be used is 270.26 g of toluene diisocyanate (TDI). The types and amounts of diol to be used is 870.53 g of Ucon 75-H-450 and 25 g poly(DL-lactide-co-glycolide) (50:50). The average molecular weight of the copolymer is 50,000 Dalton. The type and amount of triol to be used is 9.21 g of trimethylol propane. The theoretical target for completion of the diol termination steps is % NCO=4.55%. The theoretical target for completion of the trifunctionalization step is % NCO=3.00%. Final temperature pre-TMP was 50° C. The NCO levels at 96 hrs 4.99% and at 312 hrs 4.41%. Then the TMP was added at hour 312. The final NCO of % NCO=2.93% was reach at hour 528. The viscosity at 32° C. was 240 Kcps.

Example 6

Prepolymer

In this example a high molecular weight diol is terminated and randomly trifunctionalized to yield a slow curing, low viscosity implant. The type and amount of isocyanate to be used is 171.29 g of isophorone diisocyanate (IPDI). The type and amount of diol to be used is 824.93 g of Ucon 75-H-1400. The molecular weight of 75-H-1400 is 2500 Dalton. The type and amount of triol to be used is 12.49 g of trimethylol propane. The theoretical target for completion of the diol termination steps is % NCO=3.3%. The theoretical target for completion of the trifunctionalization step is % NCO=2.2%. Final temperature pre-TMP was 80° C. The NCO levels at 168 hrs 4.54% and at 624 hrs 3.32%. Then the TMP was added at hour 625. The final NCO of % NCO=2.2% was reach at hour 824. The viscosity at 32° C. was 150 Kcps.

Example 7

Prepolymer

In this example a high molecular weight diol is terminated and randomly trifunctionalized to yield a fast curing, low viscosity tissue adhesive. The formula for Example 6 is used substituting molar equivalents of TDI.

Example 8

In this example, any of Examples 1-7 where the triol, TMP, is substituted with a molar equivalent of TONE polyol 0301 manufactured by Union Carbide. The molecular weight of this triol is 300 Dalton with a hydroxyl number of 560.

Example 9

Liquid Prosthetic

In some medical applications a tissue bonding adhesive that does not appreciably swell during polymerization is useful. Applications include disc nucleus replacement, disc annulus augmentation, and any application where large static forces predominate. For these applications an adhesive of low % NCO can be used. It is also advantageous to initiate polymerization outside the body by pre-mixing the tissue adhesive with water. The amount of water added determines cure time and cured modulus. A useful adhesive for these applications can be prepared by mixing Example 7 in the following ratios with water.


% prepolymer	% water	cure time	Modulus

33	67	1 minute	disc nucleus like	1-2 N/cm ²
50	50	2 minutes	disc annulus like	2-3 N/cm ²
70	30	3 minutes	hardest	3-4 N/cm²

Example 10

Liquid Prosthetic

The cured modulus of an adhesive can be increased by adding a particulate. When 0.3 micron tantalum powder is added, the material can be made radio-opaque. A higher modulus implant can be made by adding 10% by volume tantalum powder to the mixtures of Example 9.

Example 11

Two human lumbar spine specimens were utilized in a study of load failure. The study compared nucleus replaced by a gel vs nucleus replaced by a foam of Example 3. The specimens were screened for gross anatomical defects. The age of the donors did not exceed 85 years.

TABLE 2

Specimen Levels Used

	L4/5
	L2/3
	T12/L1
	L3/L4

	L = Lumbar,
	T = Thoracic

The specimens were thawed to room temperature and all residual musculature removed via careful dissection. Throughout preparation and testing, the specimens were kept moist with a wrapping of saline-soaked gauze. A total of four two-level spinal lumbar spine segments were harvested. To create Anterior Column Units, the pedicles were transected and the posterior elements removed. Care was taken to preserve all remaining ligamentous attachments and maintain segmental integrity. For each segment, the cephalad and caudad vertebrae were rigidly embedded in a urethane potting compound. The segments were potted so that the mid-plane of the intervertebral disc was horizontal. Sufficient space was left for injection of the nucleus replacement device (Gel or Foam).
The spinal segments were tested using a standard compression protocol developed by Rhode Island Hospital with custom fixtures in an MTS 810 servohydraulic load frame. The upper compression platen was not allowed to rotate. The segments were loaded to failure in compression at a rate of 0.167 mm/sec. The mode of failure and maximum load were recorded.


Implant Type	Mean Failure (N)	Failure Mode

Gel	4100 +/− 2400	endplate fracture
Foam	7500 +/− 1000	endplate fracture

From the higher mean failure value, it can be seen that the foam implants having a Poisson ratio <0.5 allows the disc to withstand higher loads without failure compared to preformed liquid/gel implants.

APPENDIX

Ideal physical characteristics of a disc nucleus replacement prosthetics are presented. Emphasis is placed on in situ polymerizing prosthetics, e.g., hydrogel forming prepolymers. One feature is identification of the optimal range of prosthetic moduli for various disc dimensions and loading conditions that satisfy the derived endplate limit and load deflection requirements. Formulae for matching prosthetic moduli to various pathological conditions of the annulus fibrosus are disclosed.
Disclosed generally herein is the formation of implants for the repair of lesions in the human spinal disc, especially in its nucleus, and materials and material properties for such implants. Disc repair implants may be either preformed or formed in the body. Also disclosed is the selection of optimal mechanical characteristics for nucleus replacement materials. Also disclosed are clinically effective ranges for mechanical properties, e.g., the modulus, for nucleus replacement devices placed in the nuclear space of a spinal disc with a defective annulus fibrosis.

BACKGROUND

A. Treatment of Spinal Disc Abnormalities

Intervertebral disc abnormalities are common in the population and cause considerable pain, particularly if they affect adjacent nerves. Disc abnormalities result from trauma, wear, metabolic disorders and the aging process and include degenerative discs, localized tears or fissures in the annulus fibrosis, localized disc herniation with contained or escaped extrusions, and chronic, circumferential bulging discs. Disc fissures occur as a degeneration of fibrous components of the annulus fibrosis. Rather minor activities such as sneezing, bending or simple attrition can tear degenerated annulus fibers and create a fissure. The fissures may be further complicated by extrusion of nucleus pulposus material into or beyond the annulus fibrosis. Difficulties can still present even when there is no visible extrusion, due to biochemicals within the disc irritating surrounding structures and nerves. Initial treatment includes bed rest, painkillers and muscle relaxants, but these measures rarely correct the underlying cause. Surgical treatments include reduction of pressure on the annulus by removing some of the interior nucleus pulposus material by percutaneous nuclectomy. Surgical treatments meant to cure the underlying cause include spinal fusion with screws, rods and fusion cages. Devices and procedures involving screws, rods and plates are disclosed in the following U.S. patents, as well as others: Errico U.S. Pat. Nos. 37,665; 5,733,286; 5,549,608; 5,554,157; 5,876,402; 5,817,094; 5,690,630; 5,669,911; 5,647,873; 5,643,265; 5,607,426; 5,531,746 and 5,520,690; Metz-Stavenhagen U.S. Pat. No. 6,261,287; Puno U.S. Pat. No. 5,474,555; Byrd U.S. Pat. No. 5,446,237; Biedermann U.S. Pat. Nos. 5,672,176 and 5,443,467; Cotrel U.S. Pat. Nos. 4,815,453 and 5,005,562; Jackson U.S. Pat. No. 5,591,165; Harms U.S. Pat. Nos. 4,946,458; 5,092,867; 5,207,678 and 5,196,013; Mellinger U.S. Pat. No. 5,624,442; Sherman U.S. Pat. Nos. 5,885,286; 5,797,911 and 5,879,350; Morrison U.S. Pat. No. 5,891,145; Tatar U.S. Pat. No. 5,910,142; Nicholas U.S. Pat. No. 6,090,111; and Yuan U.S. Pat. No. 6,565,565. Fusion cages and related procedures are disclosed in Bagby U.S. Pat. No. 4,501,269; Michelson U.S. Pat. Nos. 5,015,247 and 5,797,909; Ray U.S. Pat. No. 6,042,582 and Kuslich U.S. Pat. Nos. 5,489,308; 6,287,343 and 5,700,291. Proposed disc replacement devices are disclosed in the following U.S. patents: Middleton U.S. Pat. No. 6,315,797; Marnay U.S. Pat. No. 5,314,477; Stubstad U.S. Pat. No. 3,867,728; Keller U.S. Pat. No. 4,997,432; and Buettner-Janz U.S. Pat. No. 4,759,766.
A contained disc herniation is not associated with free nucleus fragments migrating to the spinal canal. However, a contained disc herniation can still protrude and irritate surrounding structures, for example by applying pressure to spinal nerves. Escaped nucleus pulposus can chemically irritate neural structures. Current treatment methods include reduction of pressure on the annulus by removing some of the interior nucleus pulposus material by percutaneous nuclectomy. See, for example, Kambin U.S. Pat. No. 4,573,448. Complications include disc space infection, nerve root injury, hematoma formation, and instability of the adjacent vertebrae and collapse of the disc from decrease in height. It has been proposed to treat weakening due to nucleus pulposus deficiency by inserting preformed hydrogel implants. See, Ray U.S. Pat. Nos. 4,772,287; 4,904,260 and, 5,562,736 and Bao U.S. Pat. No. 5,192,326.
Circumferential bulging of the disc also can result in chronic disc weakening. The joint can become mechanically less stable. As the bulging disc extends beyond its normal circumference, the disc height is compromised and nerve roots are compressed. In some cases osteophytes form on the outer surface of the disc and further encroach on the spinal canal and channels through which nerves pass. The condition is known as lumbar spondylosis. Continued disc degeneration can resulting in one vertebral body segment approaching and possibly contacting an adjacent vertebral body segment.
Treatment for segmental instability include bed rest, pain medication, physical therapy and steroid injection. Spinal fusion is the final therapy performed with or without discectomy. Other treatment includes discectomy alone or disc decompression with or without fusion. Nuclectomy can be performed by removing some of the nucleus matter to reduce pressure on the annulus. Complications include disc space infection, nerve root injury, hematoma formation, and instability of adjacent vertebrae. New fixation devices include pedicle screws and interbody fusion cages. Studies on fixation show success rates between 50% and 67% for pain improvement, and a significant number of patients have more pain postoperatively.
Delivery of tissue adhesives to the spine in a minimally invasive manner has been disclosed, including procedures for restoring structural integrity to vertebral bodies. See Scribner U.S. Pat. Nos. 6,241,734 and 6,280,456; Reiley U.S. Pat. Nos. 6,248,110 and 6,235,043; Boucher U.S. Pat. No. 6,607,554 and Bhatnagar U.S. Pat. No. 6,395,007. Methods of repairing the spinal disc or portions thereof are disclosed in Cauthern U.S. Pat. No. 6,592,625, Haldimann U.S. Pat. No. 6,428,576, Trieu U.S. Pat. No. 6,620,196 and Milner U.S. Pat. No. 6,187,048.

B. Surgical Approaches to the Spine

The spine may be approached in open surgery using posterior, anterior or lateral approaches. The following is a brief description of several proposed surgical approaches, which may be used to gain access to the spine in a less invasive manner to treat spinal insufficiency.
Posterior Lateral Approach
Methods for disc access include laminectomy, a procedure wherein a channel is made from the dorsal side of the patient's back to the lumbar lamina of the disc. Blood vessels, ligaments, major back support muscles and spinal nerves located around the dural sac are retracted. Once the channel has been cleared, the standard procedure is to cut a hole in the disc capsule and pass instruments into the disc interior. This approach creates a defect that is oriented toward spinal nerves, thus typically the nucleus is completely removed to prevent extrusion of nuclear material and subsequent pressure on these nerves. Alternatively, under visual magnification with an operating microscope or operating loupe, small diameter microsurgical instruments can access the disc without cutting bone. It is possible to bypass the nerves and blood vessels entirely by inserting a cannula through the patient's side above the pelvic crest to reach a predetermined position along the lumbar portion of the spine. This procedure can be guided with use of fluoroscopy.
Kambin U.S. Pat. No. 4,573,448 describes a posterior lateral approach performed under local anesthesia by the insertion of a cannulated trocar over a guide wire extending through the patient's back toward a target disc at an angle of approximately 35 degrees with respect to the patient's perpendicular line. A hollow needle with a stylet can be inserted at a location spaced from the midline so as to form a 35-degree angle in an oblique direction. When the needle reaches the annulus fibrosis it is withdrawn after a guide wire is introduced through the needle to the disc. A cannulated, blunt-tipped trocar is passed over the guide wire until the tip reaches the annulus. The guide wire is withdrawn. A closely fitting, thin-walled cannula is passed over the trocar until it reaches the annulus. The trocar can be withdrawn. Cutting instruments or a punch can be used to expose the nucleus.
Paramedian Transabdominal Procedure
In this procedure the patient is in the supine or lithotomy position. This transabdominal procedure involves splitting the paramedian rectus, retracting the bowel, incising the peritoneum on the posterior wall of the abdominal cavity and accessing the anterior spine. Alternatively, the anterior rectus sheath is exposed of the left rectus muscle. The anterior rectus sheath is incised to expose the body of the rectus muscle. The rectus muscle is then mobilized over an adequate length, preferably symmetrical with the incision, and the rectus is retracted medially. The posterior rectus sheath is cut to expose the peritoneum. The peritoneum is pushed aside and dissected to expose the psoas muscle. The ureter and the left iliac vessels are mobilized. The rectus muscle, ureter, iliac vessels, and peritoneum are retracted laterally to expose the lumbar region. For repair to lumbar vertebrae L3-4 and L4-5, access should be made to the left of the aorta and inferior vena cava, between the aorta and the psoas muscle, and through the posterior peritoneum and fatty tissue. In some cases it may be necessary to transverse the psoas muscle. For access to sites between L5 and S-1, the dissection is closer to the midline between the iliac branches of the great vessels.
Lateral Retroperitoneal Procedure
The retroperitoneal procedure involves placing the patient in the right lateral recumbent position and making an incision in the abdomen at the border of the rectus muscle and subsequent dissection down to identify the peritoneum. Dissection can be performed bluntly or may be facilitated using a balloon cannula or expanding cannula as described by Bonutti (U.S. Pat. No. 5,514,153). The resulting retroperitoneal cavity can be held open with a retractor positioned to elevate the wall of the cavity adjacent to the patient's left side. The retractor may be a balloon retractor, see for example Moll U.S. Pat. No. 5,309,896 and Bonutti U.S. Pat. Nos. 5,331,975; 5,163,949; 6,277,136; 6,171,236; and 5,888,196. The peritoneum is dissected away from the abdominal wall in first a lateral and then a posterior direction until the spine is exposed. Under endoscopic visualization the iliopsoas muscle is dissected or retracted to facilitate disc repair.
Alternatively, dissection of the peritoneum can be accomplished using gas pressure into the preperitoneal and retroperitoneal space, thereby expanding the space and dissecting the peritoneal lining from the abdominal wall while relocating the peritoneal lining toward the midline of the abdomen. Access devices that may be used to gain minimally invasive access to the spine in several of the foregoing surgical approaches to the spine include expanding cannula structures such as Dubrul U.S. Pat. Nos. 5,183,464 and 5,431,676, Bonutti U.S. Pat. Nos. 5,674,240 and 5,320,611, and Davison U.S. Pat. Nos. 6,652,553 and 6,187,000.
Laparoscopic Approach
It is also known to approach the lumbar spine anteriorly using a laparoscopic approach. See, for example, Green U.S. Pat. Nos. 5,755,732 and 5,620.458. Techniques for laparoscopic placement of spinal fusion cages are shown and described in Kuslich U.S. Pat. No. 5,700,291 and Castro U.S. Pat. No. 6,004,326. Implementing the laparoscopic approach requires that one or more laparoscopic access devices, commonly referred to as trocars (see for example Moll U.S. Pat. Nos. 4,601,710 and 4,654,030) are introduced into the abdominal cavity and that the cavity is insufflated to create working space. A laparoscope is inserted through one of the trocar ports to provide visualization of the abdominal cavity and surgical instruments may be introduced either through another trocar port or through a working channel of the laparoscope to dissect, manipulate and retract tissue to gain access to the posterior wall of the abdomen adjacent to the spine. Retractors, including balloon retractors, may be used to retract organs and tissue to maintain a clear working path. Care is taken to avoid damage to the major blood vessels, the aorta and femoral arteries, and the posterior wall of the peritoneum is opened to access the desired spinal vertebral body or disc segment.

C. Imaging Techniques

A variety of tools exist to assist the surgeon in assuring the desired access and treatment are achieved without compromising or adversely affecting adjacent healthy tissue. Treatment of the spine is usually planned based on CT or MR scans and fluoroscopy is commonly used during surgery to assure proper positioning and placement of surgical tools and devices. Image guided spinal surgery has been proposed and is contemplated for use with the surgical treatments proposed herein. See, for example, Cosman U.S. Pat. Nos. 5,662,111; 5,848,967; 6,275,725; 6,351,661; 6,006,126; 6,405,072; Bucholz U.S. Pat. Nos. 5,871,445; 5,891,034; 5,851,183; and Heilbrun U.S. Pat. Nos. 5,836,954 and 5,603,318. The position of instruments typically is detected using a camera and markers on the surgical tool, and an image of the working portion of the instrument is super-imposed upon a pre-operative image, such as a CT, MRI or ultrasound image to show the surgeon where the working instrument is located relative to anatomical landmarks and the tissue to be treated. As imaging techniques and equipment improve, it is contemplated that image guided surgery will evolve to using real time intraoperative images and that the position of the surgical instrument will be shown relative to these real-time intra-operative images in addition to or in place of pre-operative images.

D. Adhesives and Other Repair Materials.

Numerous patents describe previous approaches to disc repair. These include U.S. Pat. No. 6,332,894, Stalcup et al., which describes an orthopedic implant for implanting between adjacent vertebrae comprising an annular bag and a curable polymer and hard particulate with the bag. The polymer is cured after implantation to make it harder and to fuse the hard particulate into a single mass. U.S. Pat. No. 6,264,659, Ross et al., describes a process of injecting a thermoplastic material within an annulus fibrosis of a selected intervertebral disk. U.S. Pat. No. 6,127,597, Beyar et al., describes a solid phase formation device for orthopedic application. The expandable device includes a material that polymerizes after implantation. U.S. Pat. No. 6,419,706, Graf, describes a disc prosthesis comprising a preformed polymer core surrounded by a rigid material coating. U.S. Pat. No. 6,569,442, Gan et al., describes polymer foam prepared outside the body for intervertebral disc reformation.
U.S. Pat. No. 6,022,376, Assell et al., describes a capsule-shaped prosthetic spinal disc nucleus for implantation into a human intradiscal space, made of a substantially inelastic constraining jacket surrounding a pre-formed amorphous polymer core. U.S. Pat. No. 6,132,465, Ray et al., describes a device similar to the device described in U.S. Pat. No. 6,022,376 with certain shape modifications. U.S. Pat. No. 6,306,177, Felt et al., describes an in situ polymerizing fluid used in tissue repair in the absence of a constraining structure, such as a balloon. The polymerizing materials comprise a quasi-prepolymer component and a curative component containing chain extenders, catalysts and the like. U.S. Pat. No. 4,743,632, Marinovic, discloses the use of a two-part adhesive for use in surgery, where a diisocyanate material is mixed with a polyamine or similar material to produce an in situ cure. Exemplary materials are described in U.S. Pat. No. 6,254,327, and our pending applications US 2003-0135238 and US-2004-0068078, the disclosures of which are incorporated herein by reference.

E. Other References Providing Background Information

These include U.S. Pat. Nos. Re. 33,258 (Onik et al.), 4,573,448 (Kambin), 5,192,326 (Bao et al.), 5,195,541 (Obenchain), 5,197,971 (Bonutti), 5,285,795 (Ryan et al.), 5,313,962 (Obenchain), 5,514,153 (Bonutti), 5,697,889 (Slotman et al.), 5,755,732 (Green et al.), 5,772,661 (Michelson), 5,824,093 (Ray et al.), 5,928,242 (Kuslich et al.), 6,004,326 (Castro et al.), 6,187,048 (Milner et al.), 6,226,548 (Foley et al.), 6,416,465 (Brau), WO 01/32100, and FR 2 639 823.
Disclosed herein is the creation of disc nucleus replacement prosthetics that effectively transfer load on the nucleus, in the form of pressure on the nucleus, to load on the annulus fibrosis in the form of hoop stress. This results in a reduction in nucleus pressure, increase in disc height, and places the elements of the disc in a more normal structural and load-bearing relationship. Certain ranges of materials properties are described that can lead to successful nucleus replacements, depending on the size of the replacement, the condition of the annulus, and the disc height. The insights disclosed herein are equally applicable to prosthetics formed outside the body and prosthetics formed inside the body.
The mechanical properties of the nucleus have been modeled in order to determine what sorts of materials we should use as nucleus replacements. The modeling involves making assumptions, but has produced useable predictions for improved materials.

General Considerations for Nucleus Replacement Prosthetics

We start by considering incompressible nucleus replacement prosthetics. By incompressible we mean the implant does not change volume when a force is applied. Referring to FIG. 1, an incompressible substance 101 is loaded along its z-axis 106 as depicted by force vector 102. The result is that the solid increases dimensionally along lines 103 in the directions of the x-axis 104 and y-axis 105. The condition of incompressibility is given by
V=K or ∀V=0,
where V is the volume of the substance, K is a constant, ∀ is the divergence operator, yielding the divergence of V.
Accordingly, the effect of stress induced by load 102 on a small unit volume within the nucleus replacement prosthetic 101 can be seen by taking the divergence of the product of unit vectors x, y, and z. Since load is in one direction, and the substance is isotropic, x is equivalent to y and we need only take the divergence in one direction x.
d/dz(zy ² =V)→y ²+2zydy/dz=0→y=−2dyz/dz→dy/y/dz/z=−½
The last expression is the Poisson ratio, v, where
ε_z =dz/z and ε_y =dy/y and v=−ε _y/ε_z
and ε_yand ε_zare the dimensionless strains in the transverse 103 and longitudinal 102 directions, respectively. What these relations demonstrate is that a prosthetic that is incompressible under normal physiologic loads must have a Poisson ratio near ½; and for every percent decrease in prosthetic height along z 106 there's a ½ percent increase in prosthetic thickness along axes x 104 and y 105. The ratio of strains in the lateral and transverse directions holds for homogeneous compressible prosthetics as well, but v<0.5. This means for compressible compared to incompressible prosthetics there is more strain, or decrease in height, in the direction of load for the same strain, or increase in width, perpendicular to the direction of load. It will be shown later that compressibility (v<0.5) increases the fraction of the total load at the center of the endplates of the disc.

Disc Kinematics

The natural, healthy spinal disc is comprised of bony endplates defining a disc height, the edges of which are sealed with an annulus bridging the endplates and defining a nuclear space. The nucleus is structured, but comparing its modulus to that of the annulus, its modulus is essentially zero. The capacity of the nucleus as an energy storage mechanism is essentially zero and stress applied in the axial direction is immediately translated to stress in the lateral directions, as if it were a liquid, i.e., σ_x=σ_y=σ_z. Solids, on the other hand, resist shape change and for incompressible solids 2σ_x=2σ_y=σ_zfor forces applied in the z-direction. This difference means solids store energy in compression, liquids do not.
When a load is applied to a disc, work is done in the form of Lε_zD_disc(Force×Distance) and this energy is stored in the disc as potential energy. Since the natural nucleus has no energy storing capability, forces must all be transferred and stored in the annulus. Define the potential (stored) energy in the annulus as E_p=E_hoop+E_comp., where E_hoopis the energy stored as hoop stress and E_compis the energy stored as compression stress.
The first mechanism for storing energy in the annulus is hoop stress. Any unconstrained, homogeneous object with internal pressure assumes the shape of a sphere, or if constrained as in the case of the annulus, circular in cross section. Because the annulus is not circular in cross section, but flattened where it contacts the endplates, the degree of flattening or bowing out of the walls represents stored energy in the form of E_hoop.
The second mechanism for storing energy in the annulus originates in the cross hatched orientation of the fibers in the annulus. These fibers are normally at 30 degree angles with respect to the plane of the disc and are interconnected by elastic Type I collagen. Compressive forces reduce the inter-fiber angle and stretch the inter-connecting collagen fibers. The consequence is store energy in the form of Hooke's Law, E=½kε_z ², where k is the spring constant of the fibers.
Given the above conditions of a healthy natural disc, it appears from the analysis that any such disc whose nucleus has been replaced by a material with E>0, will be stiffer (N/mm) than a natural disc. Since such materials are typical of materials generally used in disc replacement, it is likely that studies that report reduced or equal stiffness in a prosthetic-implanted disc have either damaged one of the energy storage mechanisms of the disc or have left a compressible void in the nuclear space. Our analysis indicates that, in contrast to reports in the literature, a nucleus replacement device having with E>0 (solid) is not ideal for a healthy, normal annulus. Accordingly, literature studies reporting ideal nucleus prosthetic moduli E in the range of approximately 100 kPa up to the annulus modulus, based on replacing the nucleus of normal, healthy discs, may be reporting experimental artifact. Similarly, finite element models found in the literature that yield such results for a normal, healthy disc are believed to reflect computational errors, perhaps introduced by a different choice of boundary conditions than hat used in the present analysis.
A justification for using a prosthetic with E>0 must presuppose that the energy storage capacity of E_hoopor E_comp, is reduced from normal, i.e., damaged. To see this, consider a model containing a prosthetic with E>0, resulting in direct energy storage in the prosthetic by an amount Eε_z ²A, where A is the area of the prosthetic in contact with the endplates.
The total energy, Load times Compression, was formerly stored as E_p=E_hoop+E_comp.=constant, in the case where E=0 for the nucleus. Hence, with an E>0 prosthetic, some of E_pis stored in the prosthetic as Eε_z ²A, reducing E_hoopand/or E_comp. Since E_hoopis rather robust, a likely explanation is that nearly all the loss in the energy storage capacity of the pathological disc is due to a breakdown in the inter-layer collagen connections in the annulus responsible for E_comp.
In the theoretical treatment that follows, it is assumed E_comp=0. This provides the justification for nonzero prosthetic modulus. The physical properties for prosthetics designed to replace E_compover the entire spectrum of natural disc geometries are derived. These solutions are uniquely obtained by matching E's required to replace lost energy storage capacity in the natural yet pathologic disc with E's required to prevent prosthetic extrusion through the annulotomy made in the disc during prosthetic insertion. In addition, composite structures are considered as well as the importance of fixating the prosthetic in the nuclear space.
One consideration of disc annulus kinematics involves balancing the hoop stresses in the annulus. It will be shown that a consequence of balancing the hoop stresses in the annulus is that a unique ideal disc deflection ε_zis obtained. Referring now to FIGS. 2A and 2B, the disc is shown in lateral cross section 300 and transverse cross section 301. The outer layers of the annulus experience two hoop stresses; σ _θ 351 and σ _φ 352 associated with r _d 350, and r _a 353, respectively. The angles θ and φ are the usual orthogonal angles in spherical coordinates. Accordingly the hoop stresses are given by
σ_θ =P _a r _d/τ and σ_φ =P _a r _d/τ
To balance these stresses, consider that annulus tissue can support more stress in the direction θ, than in φ. It will be shown that this difference, while useful in preventing the annulus from rupturing, is relatively unimportant in determining the ideal disc deflection. Nevertheless, measurements performed on swine annulus indicate that the ratio of tensile strength Tθ/Tφ=2.
Therefore, the hoop stresses are balanced when σ_θ=2σ_φ, which gives r_a=½r_dor more generally r_a=(Tφ/Tθ)r_d
Referring now to FIG. 3, the annulus in transverse cross section has disc height D′_disc 361 under mean load and radius of curvature r _a 362. A bisector x 363 passing through the origin 364 intersects 361 at a right angle and forms angle φ 365.
It follows from trigonometry that
Sin φ=D′ _disc/2r _a
Now, D_disc(loaded) is related to the disc height D_disc(unload) by
D _disc(unload)=κD _disc(loaded)
where κ is the stretch factor or spring constant under mean load. Measurements on swine annulus give a value of κ=0.6.
Then it further follows from trigonometry that
(2φ/π)×2πr _a =D _disc, φ in radians.
Combining these equations yields
2 Sin⁻¹(D′ _disc/2r _a)×2r _a =D _disc(loaded).
Now D′_disc=D_disc(1−ε_z) and r_a=(Tφ/Tθ)r_d, D_disc(unload)=κD_disc(loaded), substituting yields
2 Sin⁻¹(D _disc(1−ε_z)/2(Tφ/Tθ)r _d)×2(Tφ/Tθ)r _d =D _disc/κ
or
Sin⁻¹(D _disc(1−ε_z)/2(Tφ/Tθ)r _d)=D _disc/(4κ(Tφ/Tθ))r _d)
which gives
Sin(D _disc(4κ(Tφ/Tθ)r _d)=D _disc(1−ε_z)/2(Tφ/Tθ)r _d
Now substituting D_disc=9 mm, Tφ/Tθ=½, r_d=14 mm gives
Sin(9/28κ)=9(1−ε_z)/14
For κ=0.60, it follows that ε_z=0.21. A disc deflection of 21% under mean load will exactly satisfy σ_θ=2σ_φ.
To show the ideal disc deflection is insensitive to Tθ/Tφ, let Tθ/Tφ range from 1 to 5, which is equivalent to letting σ_θ=Sσ_φ, where S runs from 1 to 5.
FIG. 4 illustrates insensitivity of the ideal disc deflection under load for a large range of ratios of hoop stress, (Tθ/Tφ) for κ=0.6.
FIG. 5 illustrates the sensitivity of the ideal loaded disc deflection as a function of a narrow range of spring constants of the annulus, κ, for (Tθ/Tφ)=2. It is interesting to note that for an inelastic annulus, (Tθ/Tφ)=2, the ideal loaded deflection is 50%.
In conclusion, an analysis of the spinal disc purely from the point of view of energy storage and stress balance yields the following conclusions:

- 1. The justification for the use of a prosthetic nucleus with E>0 is when the energy storage capacity of the annulus has been impaired. Where the potential energy stored as compression energy is a minimum and substantially all the stored energy in the annulus is due to hoop stress.
- 2. Under this condition it can be useful to balance hoop stress in the disc plane and hoop stress in the transverse plane of the annulus. This balance yields an ideal loaded deflection of the disc of approximately 20%.

These findings will be applied in the calculations that follow.
The Disc without a Nucleus
Referring now to FIG. 6, the spinal disc is comprised of bony endplates 107, annulus fibrosis 108 and nucleus 109. The annulus is comprised of rings of fibrous tissue concentric with the center of the disc. The tensile strength of these rings drops off as their radii decrease. The optimal configuration would be a drop off in annular strength of the layers of the annulus fibrosis consistent with the hoop stress given by
σ_hoop =Pr/τ
where σ_hoopis the hoop stress, P is the nuclear pressure, r is the radius from the center of the disc to the relevant fibrous layer of the annulus, and τ is the thickness of that layer. This configuration minimizes the shear stress between layers of the annulus fibrosis, which is a condition for optimal mechanical stability of the disc.
When the inner layers of the annulus fracture or lose their collagen content, they no longer develop hoop stress and become essentially an extension of the incompressible nucleus 109. Therefore, inner layer 110 develops essentially no hoop stress and must be supported by pressure from the nucleus. The outer layer 111 develops essentially all the hoop stress that is responsible for the pressure in the annulus 108. It will be shown below that there are two sources for pressure in the annulus, and they add together to support the overall disc height.
Referring now to FIGS. 7A and 7B, the annulus 108 approximates a torus (as shown in FIG. 7A) with Poisson ratio 0.5 placed between two endplates 107 (as shown in FIG. 7B) that creates a flat portion 116 and a bulging portion with a radius of curvature 112. Referring to FIG. 7A, the radius of the disc is 113, the radius of the nucleus is 114 and a load placed along the z-axis 106 creates stress in the annulus perpendicular to 112 (hoop stress) and 115 (longitudinal stress). Further, it is recognized that the disc height is approximately twice the annulus radius of curvature 112. When the nucleus is removed we have the following static equations of state:
P _a=σ_hoop τ/r _a
P _a =L/[π(r _d ² −r _n ²)]
where P_ais the pressure in the annulus, σ_hoopis the hoop stress in the annulus, τ is the thickness of the outer annulus layer 111, r_ais the radius of the curvature of the annulus 112, L is the load, r_dis the radius of the disc 113, r_nis the radius of the nucleus 114.
The variables r_nand r_dwill be used to calculate areas over which loads will be distributed. However, the shape of the disc in lateral cross section, as shown for example in FIG. 7A, is not a circle, but rather a bent ellipse. So, for purposes of application, one needs to relate the surgically measurable quantities, the minor axes of the inner and outer ellipses, r_n,surgand r_d,surg, respectively to the effective radii.
r_n=1.5r_n,surgr_d=1.5r_d,surg
Accordingly, the measured quantities should be increased by 50% for determining optimal prosthetic parameters, based on the findings disclosed here. For the case of the normal, healthy disc the nuclear volume equals the annulus volume. This determines the ratio of r_nto r_d
r_d=1.4r_nor r_n=0.7r_d
This ratio determines normal, healthy load sharing between annulus and nucleus. Without nucleus, the disc height, h_d, can be expressed as a function of the other parameters.
σ_hoop τ/r _a =L/[π(r _d ² −r _n ²)]→r _a=σ_hoopτ[π(r _d ² −r _n ²)]/L→h _d=σ_hoopτπ(r _d ² −r _n ²)/L
It can be seen that as load, L, increases the disc height decreases. The effect of decreasing the annulus width (increasing the nucleus width), which can occur during nuclectomy, can be dramatic in a disc without nucleus, as shown below.
FIG. 8 is a plot of the disk height (y-axis) at constant load as the nucleus radius (x-axis, % total disc) is increased. In FIG. 8, the disc height is normalized to 1 when the annulus spans the entire disk diameter yielding the maximum load bearing capacity. The nucleus thickness or radius is r_n=1 when the annulus width is 0 (r_d−r_n).
The Disc with an Incompressible, Zero Modulus Nucleus
Now let's derive the same relation when the disc is filled with a natural nucleus, with Poisson ratio 0.5, and Modulus 0 (fluid). The modulus 0 condition is valid under static loads, since the nucleus will always flow to make stress in the z-direction 106 zero. Or stated another way, the pressure or stress in the nucleus is everywhere the same when constrained, and zero when not constrained. Then we have the following state equations:
L_n=πr_n ²P_n
L _a=[π(r _d ² −r _n ²)]P _a
L=L _a +L _n
P _a=σ_hoop τ/r _a
where L_nis the load supported by the nucleus, and L_ais the load supported by the annulus. Then summing gives the total load, L.
L=τr _n ² P _n+[π(r _d ² −r _n ²)]P _a
Now for nuclear modulus 0, P_a=P_n, then
L=π(r _d ²)σ_hoop τ/r _a
which gives
h _d=σ_hoopτπ(r _d ²)/L disc with nucleus
h _d=σ_hoopτπ(r _d ² −r _n ²)/L disc without nucleus
FIG. 9 is a plot of % disc height lost (y-axis) when the nucleus is removed, as indicated by nucleus to disc radius, r_n/r_d(x-axis). FIG. 9 illustrates why replacing the nucleus after nuclectomy is performed.
Thus, we see the disc height for disc with-nucleus is greater than the disc height for disc without-nucleus for all loads L. The relative contributions, R_h, of the annulus and nucleus to supporting disc height are given by
$\begin{matrix} R_{h} = height contribution from nucleus / \\ height contribution from annulus \\ = \\ = (h_{n} - h_{a}) / h_{a} = r_{n}^{} / (r_{d}^{2} - r_{n}^{2}) \end{matrix}$
where h_nis the disc height for a disc with nucleus, h_ais the height contribution from the annulus alone. For normal discs, r_d≅1.4 r_n, so
R_h=1
This expression illustrates the importance of filling the nucleus post-nuclectomy, and that filling an empty nucleus with an incompressible fluid (E=0) improves the disc height by 100% under all physiologic loading conditions.
Returning to our previous equations:
L_n=πr_n ²P_n
L _a=[π(r _d ² −r _n ²)]P _a
The load ratio, R_I, is given by
R _I =L _n /L _a =r _n ²/(r _d ² −r _n ²)=1
when r_n/r_d=0.7
FIG. 10 illustrates that load is distributed between the annulus and nucleus according to the ratio of the radii of the nucleus and disc. FIG. 11 shows the normal load distribution in an isolated lumbar (L5) vertebral segment from a 200 lb pig. This data validates the relation r_d=1.4r_n.
As expected, for a healthy disc r_n/r_d=0.7, the loads are distributed as a linear superposition, with about 50% of the load supported by the annulus alone. It will be demonstrated below that the annulus contribution to the load support drops as the modulus of the nucleus increases.
The Disc with Incompressible Nucleus Having Modulus Greater than Zero
The pressure ratio R_pfor an E=0 nucleus, as stated before as P_a=P_n, is 1. The pressure ratio R_pwill be greater than 1 when a nucleus with a non-zero modulus, E, is used to replace the natural nucleus. In other words, as the modulus of the nucleus increases, more of the load is transmitted to the soft center of the endplates where the nucleus is positioned with less being transmitted to the harder periphery of the endplates where the annulus is located. To see this, the concept of a modulus must be introduced. Simply stated, the modulus determines the distribution of stresses in a solid based on strain, for example
E=σ _xx/ε_xx
where E is Young's modulus, σ_xxis the stress along the x-axis and ε_xxis the strain. For convenience, stress and strain will only be given along the principal axes x, y, z, so the double subscript can be replaced by a single letter. From the Poisson relation, v=−ε_x/ε_z, we get
v=0.5=−ε_x/ε_z=−σ_x/σ_z
So there's twice the amount of stress and strain in the direction of load than at the bulging edges. In terms of a disc under load L in direction z, ε_zis the fractional disc height shortening, Δ(disc height)/(disc height), called here disc deflection. Therefore, we have the following state equations:
L _n =πr _n ²(Eε _z +P _a)
L _a=π(r _d ² −r _n ²)P _a
L=L _a +L _n
Note that when E=0 the state equations for the natural nucleus are retrieved. If Eε_zand P_aare left uncoupled, these equations have solutions that place unsafe stress on the endplates of the vertebral bodies. To prevent this from happening, a boundary condition is applied, called here the Endplate Limit. It is a direct consequence of filling the nucleus to capacity such that the inner layers of the annulus and the prosthetic are in equilibrium. When this condition is fulfilled, P_a=Eε_x. This is a rigid requirement, but there is a physiological reason as well.
In FIG. 12, a disc 250 is shown in transverse cross section with endplates 107 and annulus 108. Prosthetic 251 resides in contact with annulus 108. To define the Endplate Limit, consider first the geometry of the annulus. Depicted in out-take 252 is the layer orientation of the annulus. The endplates are made up of a layer of fibrous tissue that is essentially a continuation of the annulus, which forms a type of enveloping sack around the nucleus. Underneath this layer is bone. The orientation of the fibrous layers on the endplates is depicted in out take 253. Together, the tissues depicted in 252 and 253 form a continuous structure, the layers of which are approximately oriented perpendicular to a vector with origin at the center of the nucleus. It is this direction 254 for the annulus 108 and 255 for the endplates that provides the most resistance to nuclear pressure. These layers are weakest in the in-plane directions 255 and 256. In the case of the annulus, the force aligned in the in-plane direction is the hoop stress σ_hoop=P_ar_a/τ, where P_ais the annulus pressure, r_ais the radius of curvature, and τ is the wall thickness. In the case of the endplates, the in-plane stress due to nuclear pressure is counter-balanced by adhesion to the adjacent bony layer of the vertebral body. However, with the addition of prosthetic 251, there is an additional stress Eε_x, and it is this stress in the weak in-plane direction, and not Eε_zin the strong direction 255 that requires limitation. The Endplate Limit essentially establishes a balance between forces that promote endplate failure, Eε_x, and forces that promote annulus failure, σ_hoop. Mathematically, this is expressed as
σ_hoop=Eε_x.
The annulus is essentially a solid inner tube, where the wall thickness equals the radius of curvature, r_a=τ. Substituting this relation into the equation σ_hoop=P_ar_a/τ=Eε_xgives the Endplate Limit
P_a=Eε_xEndplate Limit
Now, Eε_z=2Eε_x., for v=0.5, and Eε_x=Eε_y, substituting these expressions we arrive at an alternative form of the Endplate Limit
Eε _z =Eε _x +P _a
The Endplate Limit has a rather elegant consequence that for the normal nucleus case r_d=1.4r_n, the load on the annulus is reduced by 50%, from 50% of total load to 25%. This ratio, 25% for the annulus and 75% for the nucleus, is constant for all disc deflections, as will be shown below using the following equation derived from the state equations.
L _n /L _a=(r _n ² /r _d ² −r _n ²){ε_z(E,L _n)/[ε_z(E,L _n)−ε_x(E,L _n)]+1}.
FIG. 13 is a plot showing load on the nucleus/load on annulus (L_n/L_a, y-axis) versus disc deflection (x-axis). Comparing FIG. 10 to FIG. 13, replacing the natural nucleus (E=0) with a nucleus prosthetic (E>0) shifts load to the nucleus. This is a practical necessity in cases where the annulus has already demonstrated insufficiency under load. For the case r_n/r_d=0.7, the load ratio L_n/L_ashifts from 1 (E=0) to 3 (E>0). Note that by obeying the Endplate Limit constraint the load ratio is a function of r_n/r_dand not the prosthetic modulus or the load.
The disc deflection ε_zis dependent on the modulus E and the load L on the total disc. It is useful to separate the nucleus load components in terms of nuclear pressure L_p(P_n=P_a) and the load supported by the modulus of the prosthetic L_m.
$L_{m} = π r_{n}^{} E ɛ_{z}$ $L_{p} = π r_{n}^{} P_{a}$ $Then$ $\begin{matrix} L_{m} / L = π r_{n}^{} ɛ_{z} / {π r_{d}^{} [ɛ_{z} (E, L_{n}) - ɛ_{x} (E, L_{n})] + π r_{n}^{} ɛ_{z}} \\ = r_{n}^{} ɛ_{z} / [(r_{n}^{} + r_{d}^{2}) ɛ_{z} - r_{d}^{} ɛ_{x}] \\ = R_{m} \end{matrix}$
The fraction of the entire load L supported by the prosthetic, and therefore also by the soft centers of the endplates is explicitly determined by the ratios of r_nand r_d. This is a clinically useful result since the surgeon can match the modulus of the implant based on the amount of viable annulus remaining and the total deflection ε_zdesired under a given load L. The Young's modulus is then given by
E=LR _m /πr _n ²ε_z
For 1.4r_n=r_d, then R_m=0.505 and E=0.161 L/r_n ²ε_z. E is in megapascals if r_nis in mm and L is in Newtons.
The correct modulus for a disc prosthetic can be determined by considering the normal Load vs. Displacement relation of a disc, as illustrated in FIG. 14, which shows the displacement (x-axis) as a function of load (y-axis). When compressing a disc by amount D_discε_zfor a given nucleus modulus, L is proportional to ε_z/(1−ε_Z), as depicted in FIG. 14, where typically L=1000 N when D_discε_z=1 mm and ε_z=20%.
However, the equation above gives L proportional to ε_zfor fixed E. This is a direct consequence of incorporating the Endplate Limit into the equation for selecting appropriate modulus E. Accordingly, for any target pair (ε_z,L) as shown in FIG. 14 by the dashed lines, a prosthetic modulus E is specified by the equation that satisfies the Endplate Limit.
Considering a collection of points (ε_z,L) on the line graphed in FIG. 14, for each point there is a unique and different prosthetic modulus E proportional to (1−ε_z)⁻¹, that balances the loads such that the Endplate Limit is satisfied. For a trace such as given in FIG. 14 there is only one value E corresponding to each paired value (ε_z,L) for which the Endplate Limit is satisfied. When the clinician selects E for values of L=1000 N and ε_z=0.20 (1 mm) he is selecting a modulus for the prosthetic which will result in the implanted disc reproducing this normal disc relation between L and ε_z. These are the load and displacement values that are typical for the loaded spine.
For discs ranging from normal to pathologic, r_n>=0.7 r_d, there is no non-zero value for E which gives R_L=1 for any values of L and ε_z. The condition R_L=1 is the loading condition for the healthy disc at r_n=0.7r_d. The only way to achieve R<3 for r_n>=0.7 r_dis to use a liquid/solid composite prosthetic. This case will be covered in the section entitled, the Disc with Bubble Entrapped Nucleus.
FIGS. 15-17 illustrate additional consequences of the Endplate Limitation restriction.
FIG. 15A illustrates that for fixed r_n/r_d=0.7, the absolute size of the nucleus determines implant modulus when the desired deflection and load are specified.
FIG. 15B illustrates the requirements on implant modulus when the ratio r_n/r_dis not fixed, but r_d=20 mm is fixed. The modulus requirement drops with increasing implant radius, r_n.
FIG. 16 illustrates the modulus required for the normal disc condition r_d=1.4r_nas a function of anticipated maximum load. This figure generally illustrates that for fixed deflection, and r_n/r_d=0.7, the modulus requirement increases linearly with increasing load.
FIG. 17 illustrates the sensitivity of the modulus requirement on the target deflection under a fixed load of 1000 N. A disc height of 5 mm has been used in these calculations to represent the state of a pathological disc. Normally, in the lumbar region disc height is between 9 and 12 mm. Restoring a disc to this height by some means of distraction will afford a larger range of deflection under load. For example, a surgeon may elect to choose a prosthetic modulus corresponding to deflections of 40-50%. The above FIG. 17 illustrates that doing so will halve the required modulus.
The target deflection under load will be determined to some extent by annulus health. Annulus health is reflected in the ratio r_n/r_d, which has been taken to be a healthy 0.7 in the example above. Although increasing r_nby removing annulus decreases the requirements on the modulus, the ratio R_Lincreases as r_n/r_dincreases. This is shown in FIG. 13. These illustrations serve to emphasize the importance of surgically preserving as much of the viable annulus as possible to obtain the lowest possible value of r_n/r_dand R_L. The surgeon will likely find FIG. 13, where r_n/r_dvaries, to be most useful enabling him to select the ratio consistent with the outcome of the nuclectomy.
In the limit, E=0, the natural disc places no uncompensated in-plane stress on the endplates, and the entire load on the nucleus is carried in the z-direction of the endplates. Disclosed herein are the requirements of a failing annulus. The choice was made to transfer some of the stress normally in the annulus to in-plane stress in the endplates. Utilizing this aspect of the endplates provides support for disc height that is analogous to hoop stress, without placing that stress in the annulus. This is achieved by filling the nuclear space to capacity such that the prosthetic is in pressure equilibrium with the annulus. Strategies that do not establish equilibrium with the annulus place 100% of the restorative force generated by the prosthetic in the z-direction on the endplates, and concentrate those forces in an area smaller than the cross sectional area of the nucleus.
It should be recognized that the Endplate Limits established in the preceding are for disc heights of 5 mm. This is a typical height for a diseased disc. Therapies that do not increase disc height must provide for these high moduli. It is one object of these analyses to demonstrate and emphasize the importance of combining a nucleus replacement therapy with either disc height augmentation or prosthetic bonding to the endplates. When disc height augmentation is performed, prosthetic moduli are typically reduced by at least 50%. When prosthetic bonding is achieved even greater reductions in prosthetic modulus are possible.
The Disc with a Nucleus Prosthetic Bonded to its Endplates
The foregoing assumed the nucleus, natural or replaced, is homogenous in modulus. Accordingly, the stress in the nucleus is isotropic in a plane transverse to the direction of the load σ_x=σ_yfor all planes in z. Due to cyclic boundary conditions imposed by the toroidal geometry of the annulus, it follows ε_x=ε_yfor all planes in z. In the case where the prosthetic is bonded to the endplates, the transverse strain varies as a second order polynomial of z, and the effective modulus is no longer homogenous in z.
Under the condition σ_x<bond strength in shear, Hooke's law may be applied, where the elastic modulus tensor C_i,j,k,lis summed over k and l. However, the bond at the endplates yields this simplified form of the stress
σ_ij=C_i,j,k,lε_kl =C ₁₁₁₁ε₁₁ +C ₁₁₂₂ε₂₂ +C ₁₁₃₃ε₃₃ =C ₁₁₁₁ε₁₁
so the effective stiffness for constrained compression is C₁₁₁₁.
Solving for C₁₁₁₁=σ_z/ε_z, which is the modified modulus E, in terms of E starts with the elementary form of the isotropic Hooke's law:
ε_xx=(1/E){σ_xx −vσ _yy −vσ _zz}
ε_yy=(1/E){σ_yy −vσ _xx −vσ _zz}
ε_zz=(1/E){σ_zz −vσ _xx −vσ _yy}
Then impose the constraint ε_xx=ε_yy=0, then
σ_yy =vσ _xx −vσ _zz
σ_xx =vσ _yy −vσ _zz
Substituting
σ_yy=σ_xx=σ_zz {v(1+v)/(1+v)(1−2v)}
Then
C ₁₁₁₁=σ_z/ε_z =E{(1−v)/(1+v)(1−2v)}
Here, v=0 at the bonded ends and approximates v=r_n/D_discwhere D_discis the disc height or prosthetic height in the direction z. And the equations of state become
L _n =πr _n ² [Eε _z{(1−v)/(1+v)(1−2v)}+P _a]
L _a=π(r _d ² −r _n ²)P _a
Eε _z =Eε _x +P _a
And the load ratio R_Iis
R _I =L _n /L _a =[r _n ² /r _d ² −r _n ²]{[(1−v)ε_z/(ε_z−ε_x)(1+v)(1−2v)]+1} bound
R _I=(r _n ² /r _d ² −r _n ²){ε_z(E,L _n)/[ε_z(E,L _n)−ε_x(E,L _n)]+1}. unbound
This is illustrated in FIG. 18. Here, the Poisson ratio v is the effective ratio due to prosthetic bonding to the endplates, and is greater than ½. The result of endplate bonding of the prosthetic is that more load is shifted to the endplates when compared to the load distribution of the same prosthetic not bound to the endplates. Recall that decreasing the modulus of the prosthetic shifts load to the annulus. Therefore, when the endplates are bonded a lower modulus prosthetic may be used to achieve the same distribution of loads achieved for a free prosthetic of higher modulus.
The Disc with Bubble Entrapped Nucleus
Gaseous inclusions in a water permeable prosthetic will exchange gases with the surrounding fluids in the tissue. During periods of minimal loading the spring constant of the bubbles will result in their gaseous volume being replaced with a liquid volume. Let the total volume of the liquid component of the implant be expressed as a fraction of the total implant volume, v_L. Let v_Pbe the non-liquid volumetric fraction of the prosthetic, where v_L+v_P=1. The equations of state are
L_L=πr_n ²P_av_L
L_M=πr_n ²Eε_zv_P
L_P=πr_n ²P_a
L _a=π( r _d2−r _n ²)P _a
L _n =L _L +L _p
where L_Lis the load supported by the liquid in the prosthetic, L_Mis the load supported by the modulus of the prosthetic, L_Pis the load supported by the nuclear pressure. Then
L _n /L _a =r _n ²(1+v _L)/r _d ² −r _n ² +r _n ²ε_z(E,L _n)v _P/(r _d ² −r _n ²)[ε_z(E,L _n)−ε_x(E,L _n)]
where
R _m =v _P/[(r _d ² /r _n ²+1)−(r _d ² /r _n ² +v _L)]v
E=LR _m /πr _n ²ε_z v _P
The effect of the bubbles is to reduce the load on the center of the endplates.
Disc with Partially Filled Nucleus
Disclosed herein are optimal mechanical properties for a nucleus prosthetic intended to entirely replace the natural nucleus. The following details the disadvantage of partially, rather than completely filling the nuclear space.
In earlier attempts to provide a replacement for the nucleus some portion of the natural nucleus was left in the disc. These nucleus replacements had modulus greater than 0, and consequently, natural nucleus would extrude around the implant and out a defect in the annulus. In the case where the annulus is sealed, the results for this partial replacement of the nucleus would be similar to those findings reported in the section The Disc with Bubble Entrapped Nucleus.
In the case where the entire natural nucleus is removed and only a portion of the nuclear space is filled by a replacement prosthetic and a space remains, then we have the equations of state:
L_n=πr_p ²Eε_z
L _a=π(r _d ² −r _n ²)P _a
where the radius of the prosthetic r_p<r_n. Now, the pressure in the annulus is
P _a=(L−πr _p ² Eε _z)/π(r _d ² −r _n ²) prosthetic radius<nuclear radius
Compared with
P _a =L−πr _n ² Eε _z/πr_d ²prosthetic radius=nuclear radius
FIG. 19 illustrates r_d=14 mm, r_n=10 mm, E=3 MPa, L=1000 N, ε_z=0.2. FIG. 19 shows that the burden placed on the annulus, P_a, is much greater when the prosthetic does not fill the nuclear space. This is partially due to the fact that the portion of the load supported by the prosthetic decreases with smaller prosthetic size (the sloping portion), and partially due to decoupling of the nucleus from the annulus (the offset). Even a prosthetic that fills 90% of the nuclear space would achieve a further reduction in annulus pressure of about 50% if coupled to the annulus with water.

Nucleus Extrusion Pressure and Prosthetic Failure

It should not be ignored that often discs that undergo nuclectomy require an annulotomy or already possess a defect that must be removed. Therefore, any nucleus replacement must be designed to prevent extrusion of the nucleus prosthetic through the annulotomy.
The forces involved in prosthetic extrusion through the annulus can best be understood by referring to FIG. 20. Note a labeling change where the main direction of force is the z-axis, which for extrusion in the disc annulus is in the former xy-plane. An equilibrium is maintained between the forces acting on the elemental slice 200 of the prosthetic as it extrudes through static material zone 203. The stresses acting on element 200 are shown in FIGS. 21A-21C. The equilibrium equation is given by
−(p _z +dp _z)π(D+dD)²/4+p _z πD ²/4+p _r πDds sin α+τ_f πDds cos α=0
where τ_fis the frictional stress in the dead-zone 213, p_ris the radial pressure 210, α is the dead-zone angle 214, p_zis the pressure in the z-direction 211, p_z+dp_zis 212, ds is the length 217, D is the length 215, D+dD is length 216, 202 is the disc height D_disc, 205 is the element width dz, the length z is 204, 201 is the diameter of the annulus defect D_D.
The equilibrium equation can be further simplified by using the following relationships among dz 205, dD, and ds 217:
ds sin α=dz tan α=dD/2
ds cos α=dz=dD/2 tan α
The von Mises yield criterion is
p _r =p _z+σ and τ_f=σ/3^1/2
where σ is the flow stress in the prosthetic. Substituting and neglecting higher order differentials, the equilibrium equation is obtained in the integral form
dp _z/[σ(1+cot α/3^1/2)]=2dD/D
Assuming flow stress is constant during extrusion, the integration of the equation yields
log_e D ² C=p _z/[σ(1+cot α/3^1/2)]
where C is the integration constant.
The boundary conditions D=D_D(diameter of annulus defect) and p_z=0, give an expression for C
C=D_D ⁻²
Substituting back into the equilibrium equation for the constant C, the average extrusion pressure is
P _ave,z=0=2σ(1+cot α/3^1/2)log_e(D _disc /D _D)
where D_discis the disc height.
The nucleus/defect interface friction must be included to determine the total pressure required for extrusion. The equation for static equilibrium in the z-direction is
[(p _z +dp _z)−p _z ]πD _disc ²/4=τ_f πD _disc dz
where τ_fis the frictional force at the nucleus/defect interface. In the integral form
dp _z/τ_f=4dz/D _disc
Integrating, and substituting the boundary condition at z=0, p_z=P_ave,z=0, the extrusion pressure is
p _z=4τ_f z/D _disc +P _ave,z=0
Now substituting the expression for P_ave,z=0and τ_fat yield, the average extrusion pressure is
P _ave=2σ(1+cot α/3^1/2)log_e(D _disc /D _D)+4σz/3^1/2 D _disc
Now σ=Eε where
ε={(12VD _disctan α)(D _disc ³ −D _D ³)}2 log_e(D _disc /D _D)
and ε is the mean velocity strain, V is the impact speed in mm/s of the vertebral load.
Also, from geometric considerations
$α = \tan^{- 1} [\frac{1}{2} (D_{disc} - D_{D}) / (r_{d} - r_{n})]$ $\begin{matrix} z = r_{n} - r_{a} \\ = 0 iff r_{n} <= r_{a} \end{matrix}$
where r_dis the disc radius, r_ais the annulus radius of curvature and r_nis the nucleus radius.
The explicit expression for the average extrusion pressure is
P _ave=2E{{(12VD _disctan α)/(D _disc ³ −D _D ³)}2 log_e(D _disc /D _D)}××{(1+cot α/3^1/2)log_e(D _disc /D _D)+2(r _n −r _a)/3^1/2 D _disc}
It is worthwhile to note that P_aveis the nuclear pressure P_nin earlier calculations, and not the load. The load required to cause prosthetic extrusion through the annulus is
L=L _n +L _a =πr _n ²(Eε _z +P _ave)+π(r _d ² −r _n ²)P _ave
Let D_D=2.5 mm, D_disc=5 mm, r_n=10 mm, r_a=2.5 mm and it follows that r_d=14 mm, z=7.5 mm, tan α=0.3125, cot α=3.2. Then P_aveand the corresponding Load at failure is
P_ave=2.21EV unbound
P_ave=19.9EV bound
L _fail=314Eε _z+1360EV unbound
L _fail=314Eε _z+12,200EV bound
For the “at rest” case, let V=1 mm/s, then the failure pressure P_avefor all deflections as a function of the modulus E is illustrated below.
FIG. 22 illustrates that failure pressure rises linearly with prosthetic modulus when everything else is fixed and the implant is not bonded to the endplates. A high failure pressure is desirable.
FIG. 23 illustrates that the faster the load is applied the higher the failure pressure. For static load, the disc has a characteristic relaxation time to reach equilibrium. The relaxation time T_rdivided by the change in disc height ε_zD_discis the impact velocity V in the static load case.
V_static≅ε_zD_disc/T_r≅1 mm/s
Older discs will typically equilibrate faster, making V_staticlarger.
FIG. 24 is load failure as opposed to nuclear pressure failure, illustrated previously. Fairly large load failures are generated by relatively modest implant modulus.
Now consider a large annulus defect. Let D_D=4.9 mm, D_disc=5 mm, r_n=10 mm, r_a=2.5 mm and it follows that r_d=14 mm, z=7.5 mm, tan α=0.0125, cot α=80.
P_ave=0.0221EV
L=314Eε _z+13.6EV
FIG. 25 illustrates a large decrease in load failure threshold for larger defect size. Despite the fact that a disc with a large defect (4.9 mm) will have a high V, this consequence does not appreciably improve the load failure threshold. On the other hand, for a healthier annulus (defect=2.5 mm) the impact velocity has a large effect on the load failure threshold for near static loading. The effect of prosthetic bonding is also illustrated
FIG. 26 provides a convenient conversion from modulus (MPa) to durometer (Shore A) for polyurethane prosthetics.
FIGS. 27 and 28 illustrate load thresholds for free prosthetic for various deflections where D_D=2.5 mm. These plots describe a cone, within which the spectrum of allowed spinal motion is bound. The cone narrows for increasing V.
Now, consider the following conditions D_disc=5 mm, r_n=10 mm, r_a=2.5 mm, r_d=14 mm, deflection=0.20, z=7.5 mm where D_D=0.5, 1, 2.5, 3.5, 4.9.
Failure Thresholds for Various Defect Sizes

Failure Pressure (MPa) Failure Load (N)

P_ave,0.5=15.9EV L=314Eε _z+9807EV
P_ave,1.0=8.10EV L=314Eε _z+4987EV
P_ave,2.5=2.21EV L=314Eε _z+1360EV
P_ave,3.5=0.624EV L=314Eε _z+384EV
P_ave,4.9=0.0221EV L=314Eε _z+13.6EV
P_ave≅□□0.84[(Ddisc/D_D)^1.3−1]EV L=314Eε _z+264[(Ddisc/D _D)^1.3−1]EV
For r_d=1.4r_n
For bonded
P_ave ^bond=9.0P_aveand L_bond=9.0L.
This is illustrated in FIGS. 29A and 29B.

The Disc in Flexion/Extension

To this point the load vector was always perpendicular to the plane of the disc. In what follows, we treat the case of flexion-extension under load. Referring to FIG. 31, the midline 220 makes an angle Ω 221 to the plane of the endplates 222. The load L, 223, is a vector with perpendicular components 224, 225. Component 224 is in the plane of the endplate and does not contribute to compression of the disc. Somewhere, at a distance X 231 from the edge of the endplate is a pivot 230 coplanar with the endplate. Then the torques 227, 228 and 229 balance. Referring to FIG. 32, the annulus 231 and nucleus 232 apply forces to the endplate 222 proportional to the area, where φ is the sweeping angle 233 defining infinitesimal element 234 and the forces in integral form are:
Annulus: 2(r_d−r_n)P_a(φ)sin φdx
Nucleus: 2r_n[P_n+Eε_z(φ)] sin φdx
where x=2r_d+X−r_d(1+cos φ) and dx=r_dsin φdφ
The equilibrium equation then is
Int(φ,0→180){[2r _d +X−r _d(1+cos φ)]2r _d ²sin² φP _a dφ}−−Int(φ,0→180){[X+r _n +r _d −r _n(1+cos φ)]2r _n ²sin² φP _a dφ}++Int(φ,0→180){[X+r _n +r _d −r _n(1+cos φ)]2r _n ²sin² φ[P _n +Eε _z(φ)]dφ}==L cos φ[X+r _d/2]
This equation separates into a perpendicular component (α=0) and a tilt component (α>0). Referring to FIG. 33, without tilt the disk height would be 240, D_disk,meanand with tilt it would increase on the posterior side by 241, r_dsin Ω and decrease by the same amount on the anterior side 242. This is the case because there is no force exerted at the pivot in the equation above, in fact X can be taken to be any length. Now coupling the mean annulus pressure P_a,meanand annulus radius of curvature r_a,meanto the mean deflection ε_zwe have:
P _a,mean=(1/πr _n ²)L−Eε _z
r _a,mean=(D _disc −D _discε_Z)/2
Then we get for posterior and anterior annulus pressures:
P _a,post=(r _a,mean /[r _a,mean +r _dsin Ω])P _a,mean
P _a,ant=(r _a,mean /[r _a,mean −r _dsin Ω])P _a,mean
The nuclear pressure has three components in the tilt case, an equilibrium component equal to P_a,mean, and two translational components Eε_zsin Ω and (P_a,ant−P_a,post).
P _n={(1/πr _n ²)L−Eε _z }+Eε _zsin Ω+(r _a,mean r _dsin Ω/[r ² _a,mean −r _d ²sin²Ω])[(1/πr _n ²)L−Eε _z]
To avoid extrusion
P_n<P_ave
Note, the translational component changes the previous physical interpretation of P_n. Now calculate P_nfor the following conditions D_disc=5 mm, r_n=10 mm, r_d=14 mm, D_D=2.5, deflection=20% and Ω=7 degrees. Than r_a,mean=2.0 mm and using R_m=r_n ²ε_z/[(r_n ²+r_d ²)ε_z−r_d ²ε_x]=0.5
E=LR _m /πr _n ²ε_z=0.00795L
P _n=0.0123L−0.77E=0.776 E tilt
P_n=0.177E without tilt
P_ave,2.5=2.12E failure threshold, D_D=2.5, V=1 mm/s
For the nucleus not to extrude through a 2.5 mm diameter hole in the annulus under 7% flexion, Pn, the internal nuclear pressure must be less than the failure threshold for that defect size, P_ave,2.5. The failure pressure is not exceeded, and the implant is not extruded. Therefore, flexion/extension does not exceed the Endplate Limit. For lesser constraints, the implant is likely to fail under flexion because the nuclear pressure increases 438%.
It appears counter intuitive that the nuclear pressure should increase with E, but one should keep in mind that this equation is valid for only one value of deflection=20%, therefore to maintain the same deflection for higher E the load must be increased, hence rising nuclear pressure.
The graph of FIG. 30 illustrates that any defect less than approximately 3.5 mm in diameter would not fail if a free prosthetic has a modulus of 1 MPa or greater and the tilt angle is less than 7 degrees. With prosthetic bonding, the failure threshold increases to a defect size of 4.8 mm in diameter.

The Disc in Torsion

Torsion principally acts to reduce disc height. The height is reduced by torsion angle Ψ where
D_disc=D_disccos Ψ
For angle ψ=12 degrees, D_discis decreased by 2%, a negligible amount. Torsion does not significantly alter prosthetic modulus characteristics derived by applying the Endplate and Extrusion Limits.

Surgical Insights

The following are procedural elements that can be useful in the surgical replacement of a natural nucleus with a prosthetic nucleus:

- 1. The following conversion factors are to be applied for converting measured disc radius, r_d,surg, and nucleus radius, r_n,surg, to their effective equivalents used in the formulae developed here.

r_n=1.5r_n,surgr_d=1.5r_d,surg

- 2. The Endplate Limit, P_a=Eε_x, is applied to all tables for selecting ideal prosthetic properties. This condition ensures that if a prosthetic were placed in a healthy nucleus that load would be optimally supported by the annulus and the endplates. It is derived from balancing the in-plane stress in the annulus with the in-plane stress on the endplates.
- 3. The disc Deflection Limit, ε_z=20% at maximum (1000 N) load, is applied to all tables for selecting ideal prosthetic properties. This condition ensures that under normal loading conditions orthogonal hoop stresses in the annulus are balanced and a disc with its nucleus replaced with a prosthetic deflects the same amount as a natural healthy disc. Under this constraint, interference between structural elements of the spine, bones, muscle, nerves, is avoided.
- 4. The load bearing capacity of the annulus is directly dependent upon its ability to develop hoop stress. Accordingly, inner layers of annulus that are disorganized or fractured should be treated as part of the nucleus and removed to prevent tissue extrusion through the annulotomy. The radial dimension of the space created should be considered r_n,surgfor use with the prosthetic selection procedures herein described.
- 5. The disc height, D_disc, to be used here is the greatest height reasonably attainable by mechanical distraction, pressurized injection of an in situ polymerizing disc prosthetic, patient orientation, or any combination of these.
- 6. The following table is to be used to determine the minimum modulus necessary to preserve the annulus (Endplate Limit) with a nucleus prosthesis using surgically derived r_n, where r_d=1.4 r_n:

TABLE 1

Removed	Implant
Nucleus	Modulus (MPa)

Radius (mm)	Free	Bonded

7	16.43	1.83
8	12.58	1.40
9	9.94	1.10
10	8.05	0.89
11	6.65	0.74
12	5.59	0.62
13	4.76	0.53
14	4.11	0.46
15	3.58	0.40
16	3.14	0.35
17	2.79	0.31
18	2.48	0.28
19	2.23	0.25
20	2.01	0.22

The following table is to be used to determine the minimum modulus necessary to preserve the annulus (Endplate Limit) of nucleus prosthesis using r_n, where r_d=14:

TABLE 2

Removed	Implant
Nucleus	Modulus (MPa)

Radius (mm)	Bonded	Free

7	1.20	10.8
8	1.09	9.8
9	0.99	8.9
10	0.89	8.0
11	0.81	7.3
12	0.73	6.6
13	0.66	6.0
14	0.60	5.4

- 7. The nucleus Extrusion Limit, P_n<P_ave,defect, must be applied by measuring the annulus defect diameter and using the table below to determine whether prosthetic bonding is necessary. See 8 if the disc height is more than 5 mm.

TABLE 3

Maximum Static Loads for Free Prosthetic

Implant Modulus

Defect Diameter (mm)

(MPa)	0.5 mm	1.0 mm	2.5 mm	3.5 mm	4.9 mm

0.25	2467.45	1262.45	355.7	111.7	19.1
0.5	4934.9	2524.9	711.4	223.4	38.2
1	9869.8	5049.8	1422.8	446.8	76.4
1.5	14804.7	7574.7	2134.2	670.2	114.6
2	19739.6	10099.6	2845.6	893.6	152.8
2.5	24674.5	12624.5	3557	1117	191
3	29609.4	15149.4	4268.4	1340.4	229.2
3.5	34544.3	17674.3	4979.8	1563.8	267.4
4	39479.2	20199.2	5691.2	1787.2	305.6
4.5	44414.1	22724.1	6402.6	2010.6	343.8
5	49349	25249	7114	2234	382
5.5	54283.9	27773.9	7825.4	2457.4	420.2
6	59218.8	30298.8	8536.8	2680.8	458.4

TABLE 4

Maximum Static Loads for Bonded Prosthetic

Implant Modulus

Defect Diameter (mm)

(MPa)	0.5 mm	1.0 mm	2.5 mm	3.5 mm	4.9 mm

0.25	22207.05	11362.05	3201.3	1005.3	171.9
0.5	44414.1	22724.1	6402.6	2010.6	343.8
1	88828.2	45448.2	12805.2	4021.2	687.6
1.5	133242.3	68172.3	19207.8	6031.8	1031.4
2	177656.4	90896.4	25610.4	8042.4	1375.2
2.5	222070.5	113620.5	32013	10053	1719
3	266484.6	136344.6	38415.6	12063.6	2062.8
3.5	310898.7	159068.7	44818.2	14074.2	2406.6
4	355312.8	181792.8	51220.8	16084.8	2750.4
4.5	399726.9	204516.9	57623.4	18095.4	3094.2
5	444141	227241	64026	20106	3438
5.5	488555.1	249965.1	70428.6	22116.6	3781.8
6	532969.2	272689.2	76831.2	24127.2	4125.6

- 8. If the surgeon elects to increase the disc height from ITS diseased height by some form of distraction, then the modulus required to satisfy the Endplate Limitation decreases by the factor D_corrected/D_disease. The modulus required to satisfy the Extrusion Limit changes as shown in FIG. 34, where P_ave(disc height=5 mm)=1. FIG. 34 is represented by Table 5.

TABLE 5

Defect	Disc Height (mm)

Diameter (mm)	6 mm	7 mm	8 mm	9 mm	10 mm	11 mm	12 mm

0.5	0.875035	0.781545	0.708979	0.650957	0.603432	0.563726	0.53
1	0.924871	0.860423	0.805617	0.758783	0.718403	0.683248	0.65235
1.5	0.97868	0.947851	0.914914	0.882715	0.85231	0.824012	0.797824
2	1.041422	1.052389	1.048177	1.036207	1.020305	1.002529	0.984021
2.5	1.119619	1.185582	1.220976	1.238102	1.243916	1.24254	1.236519
3	1.225811	1.369662	1.463281	1.524658	1.564563	1.589739	1.604568
3.5	1.389185	1.656403	1.844834	1.980157	2.078436	2.150148	2.202366
4	1.697321	2.201312	2.575099	2.857635	3.074185	3.241797	3.372348
4.5	2.588091	3.782284	4.702029	5.422998	5.995734	6.455359	6.827046
4.9	9.604787	16.24868	21.49539	25.70658	29.12825	31.93488	34.25413

Exemplary Prepolymers

Suitable prepolymers form non-absorbable hydrogels. Examples of suitable nonabsorbable hydrogel compositions are described in U.S. Pat. No. 6,296,607. Other compositions that have the appropriate strength, and that bond to tissue when required to obtain appropriate mechanical properties are also suitable.
Non-Absorbable Prepolymers
Prepolymers of polyurethanes can be used as hydrogels. They are formed by encapping triols, or triolized diols with diisocyanate and then reacting these with excess quantities of water. When the polyol component contains approximately 75% polyethylene oxide and 25% polypropylene oxide the resulting hydrogel can contain up to 90% water and achieve desirable stability and strength characteristics.
Exemplary prepolymers are the product of reacting about 20% by weight to about 40% by weight TDI, 65% by weight to about 85% by weight diol and about 0.5% by weight to about 2% by weight TMP. In one embodiment, the composition is the product of reacting in weight ratios about 20% to about 25% TDI, 70% to about 80% diol and about 0.7% to about 1.2% TMP. In another embodiment, the composition is the result of reacting about 23% to about 25% TDI, about 73% to about 77% diol and about 0.7% to about 1.0% TMP. In yet another embodiment, the composition is the result of reacting about 24% TDI, 75% diol and about 0.7% to 1.0% TMP. In yet another embodiment, the diol is 75% polyethylene glycol and 25% polypropylene glycol.
Other suitable compositions are the product of reacting about 20% by weight to about 40% by weight IPDI, 65% by weight to about 85% by weight diol and about 1% by weight to about 10% by weight TMP. In one embodiment, the composition is the product of reacting in weight ratios about 25% to about 35% IPDI, 70% to about 80% diol and about 2% to about 8% TMP. In another embodiment, the composition is the result of reacting about 25% to about 30% IPDI, about 70% to about 75% diol and about 1% to about 8% TMP. In another embodiment, the composition is the result of reacting about 25% IPDI, 70% diol and about 1% to 2% TMP. In yet another embodiment, the diol is 75% polyethylene glycol and 25% polypropylene glycol.
Hydrogels can formed by mixing the above prepolymers with up to 90% water by volume, e.g., 50% water by volume.

Animal Studies

Synthesis of the Implant
Seven hundred grams of Diol UCON 75-H-1400 (Dow Chemical) is heat to 49° C. and stirred under a continuous flow of argon for 24 hours. The prepared diol is cooled to room temperature (22° C.) and 113.40 g of Toluene Diisocyanate added. The mixture is stirred under an argon blanket and the temperature of the solution increased linearly to 60+/−2° C. over a 2 hour period. The mixture is maintained at these condition until the % NCO drops to 2.95%. When this target is reached 6.26 g of Trimethylolpropane is added, and the mixture stirred under argon at 60+/−2° C. until the % NCO=2.21.
The composition above was use directly (100% polymer) and mixed with equal parts by volume of water (50:50 polymer). These compositions were injected into an isolate lumbar segment of a 200 lb pig through a 2.5 mm diameter annulotomy. FIG. 35 illustrates disc compliance for the natural disc, the natural disc with nuclectomy, and the same disc filled with 100% and 50:50 versions of the polymer. FIG. 36 illustrates the load bearing capability of one embodiment of an implant via a plot of % total load (y-axis) versus displacement (x-axis).

Treatment of Thin Discs by Trans-Axial Approach

FIG. 37 shows common pathologies of spinal discs located in the lumbar region. Lumbar discs have the largest radius, r_d(9-14 mm), and disc height, D_disc(9-12 mm), of all the spinal discs. In the diseased state [the] A disc can be highly compressed, and in this example the disc is compressed from a normal 9 mm to 3 mm. Otherwise, the disc does not show signs of annulus rupture.
In this case, the absence of bulging or nucleus leakage mitigates the need for an annulotomy. There are two options: 1) access the nuclear space via a trans vertebral body approach or 2) inject the prosthetic posterior laterally with a small gauge (18 G) needle through the annulus.
The first approach will be considered in this example. The trans-axial approach comprises the steps of 1) forming a passage through the vertebral body endplates, 2) inserting a coring device into the nuclear space that sweeps out the nucleus in a radius r_n=12 mm, 3) mechanically distracting the endplates to achieve an 8 mm disc height, and 4) filling the space formed with a nonabsorable in situ polymerizing agent.
Since the nucleus is completely removed in this procedure, it is beneficial to use a nucleus prosthetic that can be delivered as a liquid and can bond to the endplates. Any polymer disclosed herein can sufficiently bond the endplates in this application. Referring now to Table 1, the Endplate Limit specification for the implant modulus is 731 kPa. The target disc is 8 mm, so the value obtained from Table 1 must be decreased as specified in Step 9, 731 kPa×5/8=457 kPa.

Treatment of Thin Discs by Trans-Annulus Approach

If the annulus is in tact, the surgeon may elect not to create a defect in it in order to perform a nuclectomy. In this case, the liquid prosthetic is injected over the existing nucleus. The goal is to restore some disc height, so a pressurized injection will be used to help distract and restore disc height. The nuclear material is in an unknown state and likely highly fractured and disordered. It acts as a loose coating on the endplates preventing bonding of the prosthetic to a rigid structure. Thus, the surgeon will use the column in Table 1 corresponding to a free prosthetic.
This section of Table 1 requires high prosthetic modulus, and the surgeon recognizes that the patient is elderly and has reduced bone density. In this case, the Endplate Limit should be strictly respected. Since the annulus will not be damaged by the procedure, the surgeon elects to distract the space to 12 mm. This reduces the modulus from 6.57 MPa to 6.57×5/12=2.7 MPa. Using a lower modulus prosthetic will shift load to the annulus, this implant modulus represents a balance between stresses on the annulus and those on the endplates.

Treatment for Black Disc

FIG. 37 also depicts common pathologies of the mid spinal region. One disc depicts nucleus dehydration and or degeneration. Disc degeneration results in irritating and often painful degradation products, which must be removed. The current standard of care is removal of the nuclear material without replacement. This will result in an eventual loss of 50% disc height, causing the bony structures of the spine to close around nerves and cause pain.
The treatment consists of a 2.5 mm annulotomy made in the annulus, followed by removal of all the degenerated material. Once the nuclear space is open to atmospheric pressure, the disc height is more easily increased by positioning the patient. The target disc height is 9 mm. Due to the highly diseased state of the nucleus, some of the annulus is also to be removed, and imaging reveals about r_n=12 of the r_d=14 mm of the disc is to be removed. Then from Table 2 (bonded) we obtain a prosthetic modulus of 730 kPa, and this is to be reduced by 5/9 to yield 406 kPa. This is the Endplate Limit.
Looking now to Table 4, for free prosthetic, the minimum modulus is 1 MPa. From Table 5, we see there is marginal benefit in increasing the disc height from an Extrusion Limit perspective. In this case, after the height adjustment is made, the Extrusion Limit dictates the choice of prosthetic modulus.

Treatment for Bulging Disc

An aneurysm of the disc is a bulge that impinges on nerves, causing pain. Because this represents a weak portion of the annulus, merely filling the nucleus to a higher height will not draw the aneurysm in because once the load is re-established the weak portion will bulge to the same extent, since the nuclear pressure under load is not significantly improved. This is a portion of the annulus where E_hoopis small compared to the rest of the annulus, and therefore can be removed without loss of annulus energy storage capacity. It must be removed, if pain is to be alleviated.
In this case, the side of the annulus defect is governed by the extent of the aneurysm and not the size of the surgeon's tools. In this case the defect will be 4.9 mm in diameter. Given the size of the defect, restoring disc height significantly decreases the modulus requirement. Here, the target disc height will be 9 mm.
As is the case with any defect, the nucleus must be removed. One advantage of a large annulotomy is that the nucleus can be thoroughly removed, and the prosthetic can be effectively bonded to the endplates.
Imaging reveals the nucleus is relatively healthy with r_n=10 and r_d=14. In this case, the surgeon will want to avoid removing and healthy annulus material. From Table 1, bonded, the Endplate Limit is 893 kPa reduced by the disc height improvement by 5/9 to 496 kPa. From Table 4 the Extrusion Limit is between 1.5-2 MPa. From Table 5, the increase in disc height decreases the Extrusion Limit by 1/7, and yields a modulus of 240 kPa. In this case the Endplate Limit determines the choice of prosthetic.

Treatment for Permanently Compressed Disc

In this case the achievable distraction is only 5 mm. Here the extrusion minimum is likely to dominate. The surgeon may elect to try to minimize the annulotomy and use the free condition of Table 1. Table 1 gives 8.4 MPa. Given a small annulotomy, the Endplate Limit now dominates. At this hardness, the prosthetic is nearly as hard as the endplates. This is a consequence of the shortness of the disc height, where further shortening may cause pain and a normal 20% deflection is a relatively small distance.
Decreasing the prosthetic modulus below the Endplate Limit risks pain-inducing deflections under load, especially since the procedure involves creating a defect in the annulus.
One option is to enlarge the nuclear space using Table 2, while decreasing the annulotomy. Once the annulotomy has been made and measured, then the surgeon can determine the amount of nucleus to remove. In this case, the surgeon is approximating a total disc prosthetic.
Given the disc's compressed state, the surgeon ends up with a 3.5 mm defect. Looking on Table 3, this gives an extrusion limit of 6 MPa. Now turning to Table 2, r_nmust be enlarged from its present 10 mm to 13 mm to get the Endplate Limit down to 6 MPa.

Generalized Indicators for Nucleus Prosthesis Modulus

Table 6 provides general guidelines for choosing a disc prosthesis based on the state of the disc and achievable end points.

TABLE 6

Treatment Paradigms for Nucleus Replacement

Exemplary embodiments are provided below.
1. A polymerizable spine disk repair implant, wherein said implant is selected to have a modulus, after polymerization, that distributes the stress caused by a load on the spine evenly between the annulus and the end plates.
2. The implant of claim 1 where the stress is distributed in a range between about 3:1 to 1:3 between the annulus and the end plates.
3. An implant according to claim 1 in which the implant adheres to the walls and endplates of the defect into which it is implanted.
4. An implant according to claim 1 in which the implant does not adhere to the walls or endplates of the defect into which it is implanted.
5. The implant of claim 1 wherein the modulus of the implant is selected from any of Tables 1, 2, 3 or 4 of this application, according to the sizes of various features of the disk being prepared.
6. A method for selection of a prosthesis material for repair of a spinal disc, the method comprising:
determining the radius of the disc;
determining the radius of the space that will be filled by the prosthesis;
selecting whether the implant will adhere to the disc and end plates, or not;
and selecting a defect radius for a defect formed while removing nucleus from said disc; and then
using a table having the same information as Table 1 or Table 2 of this application to determine the required modulus for the material used for repair, after it is cured;
and use a table having the same information as Table 3 or Table 4 of this application to ensure that the repair material of the selected modulus has a value of maximum static load that is greater than a criterion, at the selected defect radius;
and if so, to use the selected modulus material as the repair material;
and if not, to use a table equivalent to table 3 or table 4 to select a material of a higher modulus that will be greater than the criterion.
7. The method of claim 6 where the criterion is a maximum static load that is greater than 1000 Newtons.
8. The method of claim 6 where a selected material is provided by dilution of a stock material with a physiologically-compatible solution.
9. The method of claim 6 wherein the modulus is adjusted for disc height as a function of defect radius by multiplying it by a factor from a table containing the information of table 5 of this application.
10. The method of claim 6 wherein all of the parameters of tables 1-5 are contained in a computer program which will calculate the required modulus upon entry of the values selected into said computer program.
11. The method of claim 10 where a required disk height adjustment as described in claim 14 is also entered into said program and its effects included in the calculation.
12. The method of claim 10 wherein the output of the program is an instruction of how much to dilute a stock material with a physiologically-compatible solution.
13. An implant for repair of a nucleus of a spinal disk wherein the physical properties of the annulus are selected to conform to the Endplate Limit, wherein the pressure in the annulus is selected to be equal to the hoop stress, as described herein.
14. An implant for repair of a nucleus of a spinal disk wherein the material properties of the implant are selected so that after repair, about 25% of the load is carried by the annulus and about 75% of the load is carried by the nucleus.
15. An implant according to claim 13 or 14 wherein the implantation of the implant is accompanied by an adjustment of disk height to a greater value.
16. An implant according to claim 15 wherein the implant adheres to adjacent tissue.
17. A kit for spine repair, the kit comprising a polymerizing material for replacing at least a part of a disc nucleus, a device for delivering the polymerizing material, and one or more tables for selection of the degree of dilution of the polymerizing material according to the conditions required to produce a limitation of the load on the annulus to about 25% of the total load.
18. A procedure for nucleus repair, wherein the procedure contains a step requiring the use of a material with the ability to adhere sufficiently to surrounding tissue to lower exterior pressure by a factor of 3 or more.

Claims

1. A spinal disc implant comprising a foam adapted to completely or partially replace a nucleus pulposus within a disc nucleus space, the foam being a nonabsorbable, closed cell and having a Poisson ratio of less than 0.5.

2. The implant of claim 1, wherein the foam comprises 30 to 50% by volume of closed cell bubbles.

3. The implant of claim 2, wherein the closed cell bubbles contain less than 50% of a liquid within 1 week or less after implantation.

4. The implant of any preceding claim, wherein the foam further comprises a radio-opaque marker.

5. The implant of claim 4 wherein the marker is selected from tantalum and barium sulfate.

6. The implant of any preceding claim, wherein the foam adheres to at least one of the annulus fibrosus, walls, and endplates of the spinal disc cavity.

7. The implant of claim 6, wherein the foam adherence takes the form of a covalent bond.

8. The implant of claim 7, wherein the bond strength ranges from 4 lbs/in²to 24 lb/in².

9. The implant of any preceding claim, wherein the foam has a modulus ranging from 0.5 to 10 MPa.

10. The implant of any preceding claim, wherein the foam has a modulus ranging from 0.5 to 5 MPa.

11. The implant of any preceding claim, wherein the foam comprises a polyurethane.

12. The implant of any preceding claim, wherein the foam is sufficiently hydrophilic to wet tissue surfaces.

13. The implant of any preceding claim wherein the foam has a bimodular compliance.

14. The implant of any preceding claim, wherein the foam is prepared from:

a polyurethane prepolymer comprising a polymeric polyol end-capped with diisocayanate, and

a low molecular weight polyisocyanate.

15. The implant of claim 14, wherein the polymeric polyol comprises polyethylene oxide and polypropylene oxide.

16. The implant of claim 15, wherein the polymeric polyol comprises polyethylene oxide in an amount ranging from 70 to 90% by weight and polypropylene oxide in an amount ranging from 10 to 30% by weight.

17. The implant of claim 15, wherein the polymeric polyol comprises 75% polyethylene oxide and 25% polypropylene oxide.

18. The implant of any one of claims 14-17, wherein the polyurethane prepolymer is a trifunctional polyol capped with diisocyanate, the trifunctional polyol being formed by trimerizing polymeric diols with trimethylol propane.

19. The implant of claim any one of claims 14-18, the polyurethane prepolymer has a molecular weight ranging from 4500 D to 5500 D.

20. any one of claims 14-19, the low molecular weight polyisocyanate has a molecular weight of 1000 D or less.

21. A method of repairing a defect in a spinal disc space, comprising:

inserting a nonabsorbable, closed cell foam having a Poisson ratio of less than 0.5 into the defect.

22. The method of claim 21, wherein prior to the inserting, the method further comprises removing some or all of the nucleus pulposus within the spinal disc space, and the inserting results in replacement of the removed nucleus pulposus with the foam.

23. The method of claim 22, wherein the removing further comprises removing portions of the annulus fibrosus in the vicinity of the nucleus pulposus.

24. A method of repairing a defect in a spinal disc space, comprising:

inserting a composition in the area of the defect, the composition comprising:

(a) a prepolymer, and

(b) a foaming component; and

25. The method of claim 24, wherein the cured foam expands to fill the spinal disc space.

26. The method of claim 24 or 25, wherein prior to the inserting, the method further comprises removing some or all of the nucleus pulposus within the spinal disc space, and the inserting results in replacement of the removed nucleus pulposus with the foam.

27. The method of any one of claims 24-26, wherein the cured foam comprises 30 to 50% by volume of closed cell bubbles.

28. The method of any one of claims 24-27, wherein the inserting comprises injecting the composition.

29. The method of any one of claims 24-28, wherein the composition has a viscosity ranging from 100 cp to 1000 cp.

30. The method of any one of claims 24-29, wherein the curing occurs over a period of time ranging from 2 to 5 minutes.

31. The method of any one of claims 24-30, wherein the cured foam does not swell more than 30% by volume.

32. The method of any one of claims 24-31, wherein the foaming component comprises an aqueous solution.

33. The method of claim 32, wherein prior to the inserting, the composition is adjusted by mixing the prepolymer with the aqueous solution in a ratio that achieves a desired foam compliance.

34. The method of claim 33, wherein the prepolymer and aqueous solution are mixed in a volumetric ratio of 1:1.

35. The method of any one of claims 24-34, wherein the prepolymer comprises free isocyanate groups and free amine groups.

36. The method of claim 35, wherein the composition cures through the interaction of isocyanate groups with amine groups, some of which are also present in the tissue of the implantation site.

37. The method of any one of claims 24-36, wherein the cured foam exchanges water with the surrounding tissue.

38. The method of any one of claims 24-37, wherein the cured foam is a permanent implant.

39. The method of any one of claims 24-38, wherein the inserting further comprises a first and second application of the composition, wherein the second application bonds to the first application of the cured foam.

40. The method of any one of claims 24-39, wherein the cured foam can be surgically removed from the implantation site.

41. The method of any one of claims 24-40, wherein the cured foam provides clinically significant reinforcement to the annulus fibrosus.

42. The method of any one of claims 24-41, wherein the cured foam changes height in response to spinal loads.

43. The method of any one of claims 24-42, wherein the cured foam begins with a Poisson ratio less than 0.5, changes height in response to spinal loads, and subsequently attains a Poisson ratio of approximately 0.5.

44. The method of claim 43, wherein the cured foam has a Poisson ratio of less than 0.5 at least within 5 days after implantation.

45. The method of any one of claims 24-44, wherein the cured foam translates axial forces originating at the vertebral endplates into radial forces applied to the disc annulus.

46. The method of any one of claims 24-45, wherein gas in the foam is replaced with fluids from surrounding tissues within about one month or less.

47. The method of any one of claims 24-46, wherein a bond strength between the prosthetic and tissue ranges from 4 lbs/in²to 24 lb/in².