Clinical Challenge of Nucleus Replacement

Because nucleus replacement devices often are not fixed onto the endplates, there is a higher risk of implant extrusion than with TDR devices. Implant size must ensure proper load sharing between the implant and the anulus. If a nucleus device or some of the contact area of the device carries too high a load or has too high focal contact stress, it might lead to implant subsidence.

Design and Material Options of Nucleus Replacement

In contrast to the predominant ball-and-socket design with either metal-on-metal or metal-on-polyethylene for the current generation of TDR, the design and material options for nucleus replacement are quite variable. This variability is largely due to the fact that nucleus replacement preserves more of the anatomy of the motion segment, such as anulus and endplates, and more function. The first nucleus replacement with long-term clinical experience was the Fernstrom ball made of stainless steel (Fig. 52–1).^3,4 Although this design offered bipolar articulation with the endplates to provide segment motion, it caused high contact stress because of point loading and implant subsidence.

FIGURE 52–1 Fernstrom ball nucleus replacement.

Because the natural, healthy nucleus is a hydrogel with viscoelastic properties, hydrogel has become a popular material choice for nucleus replacement to mimic mechanical and physiologic properties of the natural nucleus. Preformed hydrogel and in situ curable (injectable) hydrogel have been used for nucleus replacement. Preformed hydrogel nucleus replacement has the benefit of a more consistent implant manufacturing process and more consistent and predictable implant properties, whereas the injectable hydrogel nucleus replacement has the advantage of easy implantation via a smaller annular portal.

Aquarelle developed by Howmedica (later Stryker) was the first preformed polyvinyl alcohol hydrogel used for nucleus replacement.⁵ Subsequently, several other preformed hydrogel materials have been used for nucleus replacement, including prosthetic disc nucleus (PDN) (later redesigned and named HydraFlex) (Fig. 52–2) by RayMedica⁶ and NeuDisc by Replication Medical (Fig. 52–3).⁷ PDN and NeuDisc used the same type of base hydrogel, an acrylic copolymer hydrogel (HYPAN), with PDN having the hydrogel core encased in a polyethylene jacket and NeuDisc having the hydrogel sandwiched in between the Dacron knitted meshes. The main advantage of using preformed hydrogel materials for nucleus replacement is that they not only mimic the viscoelastic and physiologic properties (imbibing and releasing water during the cyclic loading) of the nucleus, but also they have the ability to bear the mechanical load.

FIGURE 52–2 PDN nucleus replacement.

FIGURE 52–3 NeuDisc nucleus replacement.

Although there are many hydrogel materials, the mechanical loading requirement coupled with the biocompatibility and biodurability requirements make the choices of suitable hydrogel materials relatively limited. All three of the preformed hydrogel nucleus replacement devices have passed necessary static and fatigue mechanical tests for the intended application. To mimic the body fluid diffusion function, most preformed hydrogel nucleus replacement devices have tried to have the equilibrium water content in the range of 60% to 80%. The hydrogel nucleus replacement can be implanted either dehydrated or fully hydrated. The advantage of implanting at the dehydrated or semidehydrated stage is the reduced implant volume and dimension, which allows for easier implantation and reduced annular incision. If the implant is implanted at the dehydrated stage, the time required for rehydration must be taken into consideration. The main concerns and challenges for preformed hydrogel nucleus replacement devices have been potential implant extrusion because of the easy deformability and slipperiness of the hydrogel materials.

In addition to preformed hydrogel, some in situ curable hydrogel materials have also been used for nucleus replacement. Examples in this category are NuCore (Spine Wave)⁸ and BioDisc (Cryolife).⁹ NuCore is an injectable synthetic recombinant protein hydrogel, which is a sequential block copolymer of silk and elastin, with two silk blocks and eight elastin blocks per polymer sequence repeat (Fig. 52–4). The material has a water content and modulus similar to that of the natural nucleus. BioDisc is a biopolymer consisting of a protein-based hydrogel. During implantation, the dispensing device mixes the two components of predetermined ratio, and the glutaraldehyde cross-links the bovine serum albumin (BSA) molecules to each other and to the surrounding tissues.

FIGURE 52–4 NuCore nucleus replacement.

Although the injectable hydrogel nucleus replacement has the advantage of being able to implant the device through a small annular incision, the main challenge for injectable hydrogel nucleus replacement devices is whether the cavity can be adequately filled during surgery without risking the implant extrusion, especially when the disc is fully loaded during normal activities. Although both these devices are designed to bond to the surrounding tissues, their adhesion strength and durability in a clinical setting, especially when some of the bonding is to nucleus pulposus, have not yet been proven.

In addition to hydrogel elastomers, nonhydrogel elastomers have also been used for nucleus replacements. Nonhydrogel elastomer nucleus replacements can be preformed and cured in situ. Nonhydrogel elastomer nucleus replacements have the same benefits as hydrogel nucleus replacements with low modulus, but they do not have the ability to imbibe and release body fluid during the cyclic loading. Several nonhydrogel elastomers with implantable history for other applications, such as polyurethane and silicone, have been attempted for nucleus replacement. The only preformed nonhydrogel elastomer nucleus replacement with limited clinical experience is NewDisc by Sulzer (later Zimmer). NewDisc is made from polyurethane and shaped in the form of a spiral coil. An insertion instrument was designed to implant the device by stretching it into a strip and inserting it into the disc cavity, where it is coiled back to a circle shape inside the disc space.

DASCOR (Disc Dynamics) is the most clinically advanced in situ cured nonhydrogel elastomer nucleus replacement and is the only injectable nucleus device with a balloon for containment (Fig. 52–5).¹⁰ The injectable polymer is a two-part in situ curable polyurethane. The balloon is also made from polyurethane to facilitate adhesion between the curable polymer and the balloon. There are several clear advantages of using the balloon. First, it allows the polymer to be injected under certain pressure to fill the disc cavity completely without the risk of polymer leaking through the defected anulus. Second, the balloon prevents the direct contact between the uncured or semicured polymer and the surrounding wet tissue, so it avoids the leach of uncured monomers. The isolation of body fluid from the uncured polymer also can ensure better polymerization and a more consistent and stronger mechanical property for the final implant.

FIGURE 52–5 DASCOR nucleus replacement.

Another in situ curable nucleus replacement device is Percutaneous Nucleus Replacement (PNR; Trans1).¹¹ The PNR consists of two threaded titanium vertebral body anchors connected by a cylindric silicone rubber membrane. The two titanium anchors are fixed to the adjacent vertebrae via a trans-sacral approach, and the in situ curable silicone rubber is injected through the sacral anchor into the silicone membrane until the disc cavity is filled.

Some nonelastomeric materials have also been used for nucleus replacement based on the favorable clinical data of the Fernstrom ball. The main advantages of using these nonelastomers for nucleus replacement are their better mechanical strength and durability. The design of the Regain (Biomet) is similar to a modified Fernstrom ball and made of pyrolytic carbon (Fig. 52–6). Although the endplate contacting surfaces of Regain are still convex, they have a much larger radius so that the device has a larger initial contact area and smaller initial contact stress than the Fernstrom ball. Because of the mismatch of the surface contour between the implant and endplates, some subsidence is still inevitable.

FIGURE 52–6 Regain nucleus replacement.

To maintain the articulating feature of the Fernstrom ball while intending to increase the contact area, the NUBAC (Pioneer Surgical Technology) adopted the design similar to many TDR devices (Fig. 52–7).¹² It consists of two plates with outer surfaces contoured to the general geometry of the endplates and has an inner ball-and-socket articulation. This inner ball-and-socket articulation allows free motion within all physiologic ranges of motion in all major axes. This articulation feature is also believed to be a major factor in minimizing the shear force between the endplate and implant surface during bending motion and reducing the risk of implant extrusion. The implant is made from polyetheretherketone (PEEK) OPTIMA, which has well established biocompatibility, superb biodurability, and a history of being used as permanent spinal implants.

FIGURE 52–7 NUBAC nucleus replacement.

Clinical Experience of Different Nucleus Replacement Devices

The early short-term and long-term clinical data of the Fernstrom ball have provided fundamental scientific evidence that nucleus replacement can be effective in relieving discogenic back pain. Fernstrom³ reported his relative short-term clinical data (6 months to 2.5 years) on 125 patients who received the device after discectomy and 100 patients with discectomy alone. He divided his patient population into two different groups: back pain only and back pain with leg pain. The success rates for patients with back pain only were 60% for the Fernstrom ball group versus 12% for the discectomy group. The success rates for patients with back pain with leg pain were 88% for the Fernstrom ball group versus 40% for the discectomy group. Motion, although not being quantitatively measured, was observed in the index level for patients with a Fernstrom ball. A different degree of subsidence of 1 to 3 mm was noticed for most of the Fernstrom ball group.

In 1995, McKenzie⁴ reported long-term (average 17 years) follow-up on 103 patients. The success rates were 75% for patients with back pain only and 83% for patients with back pain and leg pain. After the procedure, 95% of patients returned to work. Between these two studies with a total of 344 implants, only one implant extrusion was reported (0.3%). Although late criticism of the Fernstrom ball nucleus replacement device has been mainly on the subsidence, this does not seem to have a significant impact on the clinical success. The Fernstrom ball has shown similar results to other surgical treatments, including TDR and fusion, for either back pain only or back pain with leg pain.

PDN has the most clinical experience among all nucleus replacement devices. More than 4000 patients have been implanted with various versions of PDN since 1996. Although the base polymer and the jacket material remained the same, the PDN device has gone through multiple design changes, from its original HYPAN 68 base polymer with a pair of square-shaped pellets to PDN-SOLO (a single device) and ultimately the softer HydraFlex. All these changes were mainly intended to reduce the rate of implant extrusion and the risk of subsidence.

Klara and Ray¹³ reported the clinical data of the first four phases from 1996 to early 2000. The surgical success, which has not been well defined by the authors but is different from the commonly used term clinical success, varied from 62% to 91% during these four phases with implant extrusion up to 26% in phase II when the design changed to HYPAN 80 (with a higher water content than the original HYPAN 68) and a pair of angle-shaped devices. In addition to the design change, the surgical approach was changed from the initial posterior approach to the anterolateral transPsoatic approach.

Bertagnoli and Schoenmayr¹⁴ reported a large cohort study with 243 patients having a good clinical outcome. The average Oswestry Disability Index (ODI) score declined from 52.7 preoperatively to 9 at 2-year follow-up, and the average visual analog scale (VAS) score declined from 7.1 preoperatively to 1.8 at 2-year follow-up. The average disc height increased from 8.1 mm to 10.2 mm at 2-year follow-up. Shim and colleagues¹⁵ reported similar clinical outcomes with their series of 48 patients at 1-year follow-up. They found 19.6% subsidence, however. They also observed 60.9% endplate sclerosis and 82.8% Modic change. In a separate study, Shim and colleagues¹⁶ reported good results on their 75 PDN-SOLO patients with ODI and VAS scores decreased from 53 and 7.3 preoperatively to 14 and 2.2 at 2-year follow-up and only 2.7% implant extrusion. Regardless of the high incidence of endplate sclerosis and Modic change, these changes did not seem to be correlated with the clinical outcome and could have been caused by restoring the mechanical load on the endplate in the nucleus region.

Other than the Fernstrom ball, PDN is the only other nucleus replacement device that has some long-term follow-up results. The long-term clinical results of the PDN have been mixed, however. Schoenmayr and Kliniken,¹⁷ who implanted the first series of 11 PDN devices, reported 8-year follow up on the original 11 patients using the first-generation PDN device with fairly good long-term results. The average ODI scores were 45 preoperatively and 14 8 years postoperatively. Of the 11 patients, only 1 had extruded (9%), and 8 were satisfied with the surgery. Pimenta and colleagues¹⁸ reported high adverse event rates with 9 years’ follow-up, however, on 80 patients with the more recent version of PDN-SOLO, even though there were still significant reductions in ODI and VAS scores for patients who were not revised. Of these 80 patients, 38 (47.5%) were revised during the 9-year follow-up, of which 23 (28.8%) were due to implant expulsion, and 15 (18.8%) were due to subsidence.

Over the years, the PDN has gone through multiple design changes in an attempt to solve the implant extrusion problem. All these changes have been insufficient to correct the implant extrusion problem. In the past 10 years, RayMedica has also initiated several investigational device exemption (IDE) feasibility studies in the United States with different versions of the PDN device, including the newest version of HydraFlex. None of these studies have proceeded to the pivotal study, however, likely because of the unacceptable implant extrusion rate. From the various reported clinical results, it seemed that although the PDN device was effective in improving patient function and reducing back pain, the main challenges remain to be implant extrusion and subsidence.

Since its first clinical use at the end of 2004, NuBac has progressed quickly among all other nucleus replacement devices and become the second most implanted nucleus device in more than 300 patients so far. At the 2009 Spine Arthroplasty Society annual meeting, Pioneer Surgical Technology reported the worldwide multicenter clinical study that included patients enrolled outside of the United States and within the United States as a part of an IDE feasibility study with up to 2 years’ follow-up.¹⁹ The study has enrolled 219 patients using all three different surgical approaches—posterior, lateral, and anterolateral. Significant improvement in function and pain relief were observed with ODI scores being reduced from 53 preoperatively to 30, 25, 24, 22, and 20 at 6 weeks, 3 months, 6 months, 12 months, and 24 months postoperatively. VAS scores were reduced from 77 preoperatively to 32, 29, 23, 31, and 32 at 6 weeks, 3 months, 6 months, 12 months, and 24 months postoperatively.

In a separate study, Balsano and colleagues²⁰ reported the clinical data of 45 NuBac patients using only the posterior surgical approach. ODI score decreased from 51 preoperatively to 31, 27, 24, and 23 at 6 weeks, 3 months, 6 months, and 12 months postoperatively, and VAS score decreased from 76 preoperatively to 31, 31, 31, and 27 at 6 weeks, 3 months, 6 months, and 12 months postoperatively. In the U.S. IDE feasibility study, 20 patients were enrolled using the lateral surgical approach. Similar ODI improvement and VAS reduction were observed in these patients with no implant extrusions. Based on these clinical results, in 2008 NuBac became the first nucleus replacement device to gain the approval of the U.S. Food and Drug Administration (FDA) to start the pivotal study.

More than 200 DASCOR devices have been implanted since 2003, using anterolateral, lateral, posterolateral, and endoscopic approaches. Ahrens and colleagues reported an interim 2-year European clinical study of 85 patients using lateral and anterolateral approaches.²¹ Mean preoperative VAS and ODI scores improved significantly by the 6-week postoperative follow-up point throughout the 2 years. Mean VAS back pain scores showed a 56.6% reduction at 24 months from a preoperative mean of 7.6. Mean ODI score showed a 59.7% reduction at 24 months from a preoperative mean of 57.5. No expulsions were observed in this 2-year clinical cohort. DASCOR is presently undergoing the FDA IDE feasibility study with VAS and ODI results showing statistically significant similarity to the European cohort.

Several other nucleus replacement devices with smaller patient numbers and relatively short follow-up have been also reported. Berlemann and Schwarzenbach²² reported 2-year follow-up results on 12 NuCore patients who had back pain and leg pain. Decompression is part of the procedure to remove a herniated disc. Owing to the decompression, the patient’s leg pain was significantly reduced from 6.8 preoperatively to 1.1 at 2 years. Back pain VAS score was also reduced from 32 to 6. Pawulski and colleagues²³ reported on 11 BioDisc patients with up to 2 years of follow-up with the average ODI and VAS scores reduced from 47.5 and 5.9 preoperatively to 17.7 and 2.3 at 2 years postoperatively. Two patients had recurrent herniation, and both had reherniation removed along with the device without incident.

Since 2005, NeuDisc has been undergoing a two-arm European pilot study.⁷ The first arm is an open approach using the anterolateral transPsoatic approach, and the second arm is a posterolateral endoscopic approach. In the pilot study, 15 patients were implanted, but the details of follow-up data were not reported. Based on the initial lessons learned from this pilot study, the device has been modified to a less hydrophilic and reconfigured kidney-shaped design. A new pilot study with posterior and endoscopic transforaminal approaches has been initiated.