Preliminary evaluation of new intrinsically radiopaque hydrogels for replacing the nucleus pulposus

Abstract

Treatment of early degenerative disc disease can, in some cases, be accomplished through implantation of a synthetic prosthesis for the nucleus pulposus. This treatment is attractive, since the annulus fibrosus—as well as the function of the disc—is preserved. This study reports on two new synthetic hydrogels which were specifically designed as fully radiopaque prosthetic nucleus biomaterials. Moreover, the new materials were engineered in such a way that they swell in situ (i.e., after implantation) to such an extent that they will fill the entire nucleus cavity. We describe: (i) assessment of the biocompatibility of the new biomaterials in an in vivo animal model, (ii) implantation of the new prosthesis in an ex vivo animal model (porcine spine), followed by (iii) assessment of the visibility of the entire nucleus prosthesis through both CT and MRI. The results further substantiate the idea that the concept of implantation of a prosthesis for the nucleus pulposus can benefit from contemporary insights and developments of novel synthetic biomaterials with intrinsic radiopacity.

---

1. Introduction

Degenerative disc disease (DDD) is one of the most common indications in spinal surgery. Currently, there is no real ‘golden standard’ to treat DDD. Nonoperative treatment has many proponents, presumably since operative treatment of DDD is usually complicated and the outcome may be unsure. In general, operative treatment of DDD implies spinal fusion: (part of) the affected disc is removed and the adjacent vertebrae are connected with mechanical fasteners. The separation of the vertebrae is usually secured with a cage or with an autologous or allograft bone graft. Spinal fusion, however, locally impairs the range of motion of the spine. Consequently, neighboring segments are called upon to compensate for the immobilized section, and this might result in further problems.^1–5

An alternative treatment has emerged for patients with mild (early) DDD or for some trauma cases. This treatment is based on an implant that is designed to replace the diseased or deformed nucleus pulposus, and to mimic the function of the natural healthy nucleus pulposus, while preserving the annulus fibrosus.^6–9 Nucleus replacement is a minimally invasive procedure. In order to be considered for a nucleus replacement, certain inclusion criteria must be fulfilled: (i), DDD is manifested by morphologic changes of the nucleus; (ii), the annulus fibrosus must be competent; (iii), residual disc height is at least 5 mm.¹⁰ Perhaps the most successful nucleus prosthesis is the PDN-Solo,^11,12 developed by Raymedica Inc. (Minneapolis, USA). This implant has a hydrogel core, which is packed in a woven polyethylene jacket.

We became interested in developing an alternative to the commercial PDN, based on our previous work on radiopaque biomaterials and hydrogels.^13,14 We hypothesized that the PDN concept can be improved further, especially concerning radiovisibility, taking advantage of contemporary insights into new polymeric biomaterials which combine several desirable properties (e.g., intrinsic radiopacity, controllable swelling, controllable physical-mechanical properties).¹⁵ We reasoned that it should be possible to design and fabricate a synthetic nucleus prosthesis that meets the following requirements:

1. The prosthesis consists of a hydrogel that, like the healthy nucleus, but unlike the PDN, fills up the entire cavity that is enclosed by the annulus ring.

2. The prosthesis is implanted in dry form (xerogel), allowing the opening in the annulus to be small. After insertion, the implant is allowed to swell (expand) in situ.

3. The prosthesis has excellent visibility through both X-ray fluoroscopy and magnetic resonance imaging (MRI), such that no artifacts occur.

4. The swollen hydrogel has adequate physical-mechanical properties and fatigue resistance. For instance, the prosthesis must resist continuous loading, and peak loading which occurs upon bending or lifting things.^16,17

Herein, we report on two different hydrogel biomaterials that potentially fulfil the above criteria. The envisioned procedure is depicted schematically in Fig. 1. Filling the entire cavity evidently makes the new implants more biomimetic as compared to the PDN.

Fig. 1 Schematic representation of nucleus replacement using a hydrogel: step 1 involves complete removal of the affected nucleus, step 2 is the insertion of the xerogel in the cavity, through a minimal incision in the annulus and step 3 is the subsequent in situ swelling to reach situation 4, in which the synthetic nucleus mimics the natural healthy situation.

Using both materials, nucleus prostheses prototypes were designed, manufactured and studied. We describe a series of experiments that reveal (i) biocompatibility in an in vivo model, (ii) implantability and in situ swelling, and (iii) visibility of the implant through X-ray and MR imaging. Moreover, the steps to take before the prosthesis can be introduced clinically are discussed briefly.

2. Materials and methods

2.1 Hydrogels

Two hydrogels were prepared by copolymerizing either N-vinyl-2-pyrrolidinone (NVP) or 2-hydroxyethyl methacrylate (HEMA) with the radiopaque monomer 2-(4′-iodobenzoyl)-oxo-ethyl methacrylate (4IEMA)^18,19 in the molar ratio 94 ∶ 6. NVP and HEMA were purchased from Acros and distilled under reduced pressure before use. The monomer mixtures, together with 0.044 mol% of 2,2′-azobis(isobutyronitrile) (AIBN) were transferred to Teflon tubes. AIBN is the source of free radicals. The Teflon tubes were immersed in a temperature controlled oil bath. A temperature profile was run that keeps the oil bath for 8 h at 60 °C, followed by 4 h at 80 °C and another 4 h at 100 °C. The resulting copolymers are indicated by N94 (NVP/4IEMA; 94/6) and H94 (HEMA/4IEMA; 94/6). For both materials, the equilibrium water content (EWC) and volume swelling ratio (VSR) were determined in a swelling experiment in phosphate buffered saline (PBS) at room temperature. For this, the mass and dimensions of discs of the materials were monitored at various intervals, until equilibrium was reached. The VSR is the swollen volume divided by the dry volume and EWC is defined as the weight percentage of water in the swollen hydrogel at equilibrium. The Young's modulus of swollen cylindrical samples (9 mm diameter, 9 mm height, after equilibrium swelling) of both hydrogels was determined from the initial slope of the stress–strain curve, measured in compression (strain rate 3 × 10⁻³ sec⁻¹) using a Zwick 1445 compression bench with a 500 N load cell in a water bath, which was kept at a constant temperature (37 °C).

2.2 Biocompatibility in vivo

The animal experiments were reviewed and approved by the Animal Ethical Committee of the University of Maastricht.

Dry copolymer discs of both N94 and H94 were sterilized with ethylene oxide and swollen in sterile PBS before implantation. Eight FVB mice of approximately 11 weeks old were anaesthetized and operated under sterile conditions. Subcutaneous pockets were created on the back and in each pocket one swollen hydrogel disc of either N94 or H94 (4 mm diameter and 1 mm thickness) was implanted to study the tissue reaction to the materials. Ethanol-sterilized PMMA discs of the same dimensions were used for comparison; PMMA is a proven biomaterial. Hence, each mouse received three implants.

Four mice were sacrificed after one week, the others after three months. At five weeks, X-ray images were taken of the remaining four mice. After sacrificing the mice, the materials and some of the surrounding tissue were removed and fixated in buffered formalin. Then the retrieved implants were dehydrated using a graded alcohol series and embedded in glycol methacrylate (Technovit 7100; Heraeus Kulzer, Germany). Subsequently the embedded specimens were sectioned (3 µm) and stained with haematoxylin-eosin (HE). The histology was evaluated using light microscopy.

2.3 Porcine intervertebral disc

A lumbar spinal section was explanted from a recently deceased adult male pig. It was cleaned from soft tissues and conserved at −80 °C. For implantation of the hydrogel, a disc was isolated by sawing adjacent vertebrae in their middle. After sawing, the spinal segment was allowed to thaw overnight. The size and shape of the porcine lumbar nucleus cavity was adapted from several cadaver specimens. The xerogels were machined to this shape, and their size was diminished to incorporate their swelling ability, so that their size after swelling matches the cavity size. Clinically, MRI can be employed to determine nucleus dimensions of the affected intervertebral disc in a patient, through discriminating annulus and nucleus by their water content (T₂-relaxation), enabling the selection of an implant with the exact shape and size.

2.4 Hydrogel implantation

The annulus of the isolated disc was cut using a surgical blade posterolateral. Two parallel incisions were made and one perpendicular incision between the two incisions. This is to create a flap from the outer annulus layers. The flap was pulled away and a midannular incision was made through the annulus to reach the nucleus cavity (Fig. 2). The gelly nuclear material was removed as much as possible with a small scoop. A prosthesis made from N94 was inserted using forceps. Care was taken to insert the xerogel properly into the nucleus cavity. Next, some PBS was injected into the cavity to initiate hydration of the implant. The annulus flap was sutured to close the defect. Then the entire segment was submerged in PBS and the implant was allowed to swell overnight.

Fig. 2 The annulus incision created to insert the nucleus prosthesis. After implantation the flap in the annulus is closed with suture.

2.5 Visualization

To investigate the position of the swollen hydrogel inside the nucleus cavity, CT and MRI scans were made. CT scans of the hydrogel had to prove if the X-ray contrast is sufficient after swelling, since the iodine concentration is lowered after swelling. Due to the high water content, the implant can also be clearly visualized using a T₂-weighted MRI sequence. After the scans, the disc was carefully sectioned through the annulus fibrosus to reveal the swollen implant and photographs were taken.

3. Results and discussion

3.1 Hydrogels

Thermal radical polymerization of the monomer mixtures resulted in clear glassy polymers. The resulting copolymers (N94 and H94) were washed prior to further analysis to get rid of any possible unreacted monomer.

The swelling ability of N94 is greater than H94, because NVP is more hydrophilic than HEMA. This can be expressed in terms of VSR or EWC (Table 1). The EWC was determined at room temperature, but the difference in water content between room and body temperature is minimal (<1%). The Young's modulus as determined in a compression experiment is also shown in Table 1. Mark that, though both hydrogels have quite different water contents, their moduli are almost equal. The Young's modulus is a measure for the material's stiffness, and for this application it is necessary to be in the range of 0.2–4 MPa in order to mimic the mechanical behavior of the natural nucleus pulposus.^9,20–22

Table 1 Material properties

Material	VSR^a (v/v)	EWC^b (m/m)	Young's modulus
a VSR: volume swelling rate. b EWC: equilibrium water content.
N94	3.9	74%	0.7 MPa
H94	1.5	23%	0.9 MPa

---

3.2 In vivo biocompatibility

Macroscopic observations. The wound-healing process of the mice appeared to be normal and the implants could be observed as small bumps on the back. On the X-ray images both N94 and H94 can be distinguished (Fig. 3), N94 at a slightly tilted angle. After retrieval of the implants, no changes in visual appearance or dimensions were observed after 3 months. Obviously, for these materials to serve as permanent nucleus prostheses, degradation is unacceptable.

Fig. 3 X-Ray of the subcutaneously implanted hydrogel discs in a mouse.

Histology. The one week follow-up of the subcutaneous implantation in mice showed mild acute inflammatory reactions and the start of a fibrous capsule for all materials. After 3 months, a thin fibrous capsule was present around all implants, containing linearly organized fibroblasts, aligned parallel to the implants. Fig. 4 shows representative light micrographs of the materials after 3 months of subcutaneous implantation. The capsule thickness was comparable for all implants (±40 µm). A minimal inflammatory reaction was observed containing some macrophages and granulocytes and some newly formed blood vessels. Because PMMA is very stiff as compared to the surrounding embedded tissue, the fibrous capsule gets somewhat compressed during the cutting procedure. Basically, all three materials triggered only minimal and comparable tissue reactions. So both our swollen hydrogel materials N94 and H94 perform with equal biocompatability as the proven hydrophobic biomaterial PMMA. The histological data of N94 and H94 are comparable to the results that were obtained previously by Aldenhoff et al., who studied long-term biocompatibility of analogous, but non-swelling radiopaque biomaterials.²³

Fig. 4 HE-staining of sections of the retrieved implants after 3 months. The materials are in the top of the image and a piece of the dermis (d) is seen underneath. Around all implants a thin fibrous capsule (fc) can be observed and some inflammatory cells (arrows) are present. The scale bars represent 50 µm.

3.3 Porcine intervertebral disc

The porcine nucleus is kidney-shaped with the dimensions: 25 × 13 × 4.5 mm (l × w × t). To incorporate the swelling ability of N94 the dimensions of the implant were 15.5 × 8 × 2.8 mm (Fig. 5). The time needed for this implant to reach equilibrium swelling is about one day. Once swollen, this implant will completely fill the cavity, left after removal of the gelly nucleus material. This is expected to result in optimal transmission of compressive forces onto the annulus fibers, thus mimicking the physiological state.²⁰

Fig. 5 The swollen and dry hydrogel nucleus prosthesis. The shape is conforming to the porcine nucleus cavity.

3.4 Hydrogel implantation

Flap generation in the annulus fibrosus and removal of the natural nucleus pulposus proceeded smoothly. The nucleus prosthesis was inserted readily using a pair of tweezers, through the incision in the annulus fibrosus. Evidently, it must be inserted to align the curvature of the cavity. Through swelling pressure, the implant will probably position itself correctly within the empty nucleus cavity, since this is an unloaded situation. After implantation, PBS was injected into the cavity and the flap was sutured to close the annular defect. The implant will swell in situ to almost four times its original volume. This makes the implant much larger than the incision that was created in the annulus to implant the dry prosthesis and therefore extrusion is expected to be impeded.

3.5 Visualization

After one day, the swelling of an implant of the material N94 with the dimensions of the nucleus cavity is complete. CT and MRI scans were made of the porcine vertebrae containing the nucleus prosthesis. The CT and MRI image are shown in Figs. 6A and 6B respectively. Both techniques can clearly distinguish the implant in the nucleus cavity. Apparently, the X-ray contrast of the swollen hydrogel is sufficient to visualize it in between two vertebrae. During surgery, X-ray fluoroscopy will probably be used instead of CT, but in terms of X-ray contrast there is no difference. Earlier, it was seen that these types of material can be visualized in situ using X-ray fluoroscopy.¹⁵ CT provides extra information, since it is three-dimensional.

Fig. 6 Visualization of the hydrogel implant, using (a) CT and (b) T₂-weighted MRI. Both techniques clearly show the correctly positioned hydrogel implant.

Also, due to the high water content, it can be clearly seen on the MRI image. An implant made out of H94 has a much lower water content, but will probably also appear on a MRI image: the difference in water content and subsequent T₂-relaxation, between annulus and prosthesis, provides the contrast. On the contrary, due to its lower water content, H94 will exhibit more contrast using X-rays.

Then, the intervertebral disc was sectioned to reveal the swollen hydrogel within the nucleus cavity (Fig. 7). It confirmed the imaging techniques; the swollen implant completely filled the nucleus cavity.

Fig. 7 Picture of the swollen implant in the nucleus cavity; it perfectly fills the entire cavity.

4. Concluding remarks

It is a challenge to design a radiopaque hydrogel that can withstand the harsh environment of the spinal column. Both radiopaque hydrogels, N94 and H94, fulfill a set of requirements to serve as prosthetic biomaterials for replacement of the nucleus pulposus. In particular, they offer the clinically significant advantage of integral visibility through CT and MRI. The absence of any artifacts in the MRI images is believed to be important, especially in view of the fact that there is a trend towards higher magnetic field strengths; these lead to improved quality of the MR images, but also to amplification of artifacts, such as those caused by metallic X-ray markers.

Moreover, it was verified that implants of N94 and H94 can be introduced into the nucleus cavity through a so-called annulus flap; this is an established technique in spinal surgery. N94 has a larger swelling capacity as compared to H94, which is theoretically beneficial. Still, extensive biomechanical tests will be performed in the near future, especially since implant extrusion and fatigue behavior are important issues in this field. Preliminary fatigue tests seem promising. With our novel method we may decrease the risk of implant extrusion due to the improved and customized implant fit. Further experiments are planned to explore which material performs best in vivo. Biocompatibility of N94 and H94 was tested in an animal model (subcutaneous implantation for 3 months). Both materials were well tolerated: there were no signs of severe adverse tissue reactions. Histology of the tissues surrounding N94 and H94 were comparable to the histology of the tissues around the control material PMMA.

The results of this study provide further support for our earlier hypothesis, that these radiopaque hydrogels hold promise as improved biomaterials to replace a diseased or damaged nucleus pulposus in cases of a competent annulus fibrosus.¹⁵ It is clear that more research is mandatory before clinical application of the new radiopaque hydrogels is possible. The nucleus prosthesis is a surgically invasive implant for long term use (>30 days), which falls into Class IIb of the EC Medical Device Directive.²⁴ First, physical-mechanical testing for durability is necessary (fatigue resistance preferably in a confined model²⁵ and creep behavior). Secondly, tests of the implant in a representative animal model need to be executed.^26–29 Thirdly, provided that the outcome of all experiments is positive, a clinical pilot study needs to be done. Then, the combined experimental data need to be examined by one of the Notified Bodies, which can then award the CE mark. It is important to consider that the two new radiopaque hydrogels are non-classical biomaterials, especially since 4IEMA was used as one of the building blocks. Another implant, the so-called ScrewFinder,³⁰ which is also based on 4IEMA, already received a CE mark in 2001 (2012476CE01).

References

1. G. Ghiselli, J. C. Wang, N. N. Bhatia, W. K. Hsu and E. G. Dawson, J. Bone Jt. Surg., Am. Vol., 2004, 86-A, 1497–1503.
2. F. M. Phillips, J. Reuben and F. T. Wetzel, J. Bone Jt. Surg., Br. Vol., 2002, 84, 289–294.
3. A. S. Hilibrand and M. Robbins, Spine J., 2004, 4, 190S–194S.
4. Q. B. Bao, G. M. McCullen, P. A. Higham, J. H. Dumbleton and H. A. Yuan, Biomaterials, 1996, 17, 1157–1167.
5. P. Gillet, J. Spinal Disord. Tech., 2003, 16, 338–345.
6. Q. B. Bao and H. A. Yuan, Spine, 2002, 27, 1245–1247.
7. C. D. Ray, Eur. Spine J., 2002, 11, S137–142.
8. J. L. Husson, A. Korge, J. L. Polard, T. Nydegger, S. Kneubuhler and H. M. Mayer, J. Spinal Disord. Tech., 2003, 16, 405–411.
9. J. Thomas, A. Lowman and M. Marcolongo, J. Biomed. Mater. Res., 2003, 67A, 1329–1337.
10. R. D. Guyer and D. D. Ohnmeiss, Spine, 2003, 28, S15–23.
11. D. Jin, D. Qu, L. Zhao, J. Chen and J. Jiang, J. Spinal Disord. Tech., 2003, 16, 331–337.
12. A. Carl, E. Ledet, H. Yuan and A. Sharan, Spine J., 2004, 4, 325S–329S.
13. C. S. J. van Hooy-Corstjens, L. E. Govaert, A. B. Spoelstra, S. K. Bulstra, G. M. Wetzels and L. H. Koole, Biomaterials, 2004, 25, 2657–2667.
14. C. C. Peerlings, H. H. Hanssen, R. T. Bevers, E. J. H. Boelen, B. J. Stelt, E. J. Korthagen and L. H. Koole, J. Biomed. Mater. Res., 2002, 63, 692–698.
15. E. J. Boelen, C. S. van Hooy-Corstjens, S. K. Bulstra, A. van Ooij, L. W. van Rhijn and L. H. Koole, Biomaterials, 2005, 26, 6674–6683.
16. P. Ghosh, The Biology of the Intervertebral Disc, CRC Press, Boca Raton, FL, 1988.
17. A. A. White and M. M. Panjabi, Clinical Biomechanics of the Spine, Lippincott, Philadelphia, 1978.
18. M. A. Kruft, A. Benzina, R. Blezer and L. H. Koole, Biomaterials, 1996, 17, 1803–1812.
19. C. S. J. van Hooy-Corstjens, Y. B. Aldenhoff, M. L. Knetsch, L. E. Govaert, E. Arin, H. Erli and L. H. Koole, J. Mater. Chem., 2004, 14, 3008–3013.
20. J. R. Meakin, J. E. Reid and D. W. Hukins, Clin. Biomech., 2001, 16, 560–565.
21. M. J. Fagan, S. Julian, D. J. Siddall and A. M. Mohsen, Proc. Inst. Mech. Eng., Part H, 2002, 216, 299–314.
22. J. R. Meakin, J. Mater. Sci.: Mater. Med., 2001, 12, 207–213.
23. Y. B. Aldenhoff, M. A. B. Kruft, A. P. Pijpers, F. H. van der Veen, S. K. Bulstra, R. Kuijer and L. H. Koole, Biomaterials, 2002, 23, 881–886.
24. http://europa.eu.int/comm/enterprise/medical_devices/meddev/ .
25. A. Joshi, G. Fussell, J. Thomas, A. Hsuan, A. Lowman, A. Karduna, E. Vresilovic and M. Marcolongo, Biomaterials, 2006, 27, 176–184.
26. H. J. Wilke, A. Kettler and L. E. Claes, Spine, 1997, 22, 2365–2374.
27. T. H. Smit, Eur. Spine J., 2002, 11, 137–144.
28. J. E. Reid, J. R. Meakin, S. P. Robins, J. M. Skakle and D. W. Hukins, Clin. Biomech., 2002, 17, 312–314.
29. J. H. Schimandle and S. D. Boden, Spine, 1994, 19, 2474–2477.
30. S. K. Bulstra, A. Benzina, R. Kuijer and I. L. Koole, Med. Device Technol., 1999, 10, 38–39.

---

Footnote

† The Center for Biomaterials Research of the University of Maastricht is part of the Dutch Research School “Integrated Biomedical Engineering” (IBME)

---