Intervertebral disc degeneration: biological and biomechanical factors

When the clinical application of growth factors, including BMPs, is contemplated, several factors, such as indication, mode of delivery, dose, duration, and side effects, should be taken into consideration to achieve the greatest beneficial therapeutic effect. Each growth factor may have a different effect on the NP and AF, and the action at different stages of degeneration may be different. An in vitro study using a validated method is essential to explore the therapeutic use of growth factors.

Thompson et al. first reported the positive effects of various growth factors, including TGF-β, epidermal growth factor (EGF), and basic fibroblast growth factor (bFGF),3 on PG synthesis; these effects were more pronounced in the NP. Osada et al. have shown that IGF-1 stimulated PG synthesis by bovine NP cells in serum-free conditions in a dose-dependent manner.4 Recently, Tim Yoon et al. showed that recombinant human BMP-2 (10–1000 ng/ml) increased cell proliferation, PG synthesis, mRNA expression of type II collagen, aggrecan, SOX9, and osteocalcin in the monolayer culture of rat AF cells.43 Furthermore, these authors showed that the stimulatory effect of BMP-2 (at 1–100 ng/ml) was inhibited by the addition of nicotine (10–100¼g/ml) in the culture media in rat NP cell culture.44 These results suggested that nicotine may contribute to the process of disc degeneration by a direct effect on NP cells, possibly by antagonizing the effect of BMP-2.

We have recently reported the in vitro stimulative effect of osteogenic protein 1 (OP-1, otherwise known as BMP-7) on PG and collagen metabolism in rabbit NP and AF cells cultured in alginate beads.45 In both cell types, recombinant human OP-1 (rhOP-1) stimulated the synthesis of PG and collagens in a dose-dependent manner (50–200 ng/ml) in the presence of 10% fetal bovine serum (FBS) (Fig. 1). Although the responsiveness of IVD cells to IGF-1 and TGF-β has been shown to decrease with increasing age in rabbit IVD cells,46 OP-1 effectively stimulated PG synthesis of fetal, adult, and old bovine NP and AF cells.47 In fact, the IVD cells of the older animals in this study were very responsive to OP-1. The age-related and degeneration stage-related change in the response to growth factors is of special interest for future studies because of the emerging trend to treat disc degeneration in adults or older individuals by the local injection of growth factors.48

Several in vitro feasibility studies were performed to test if growth factors can overcome a negative balance of matrix homeostasis. To model the repair process after matrix depletion in IVD cells, degeneration of the matrix of rabbit IVD cells was first induced by exposure to the proinflammatory cytokine interleukin 1 (IL-1), using the alginate culture system; subsequently, rhOP-1 was applied in culture.49 OP-1 (100 ng/ml) was effective in promoting, in vitro, the repair of the damaged matrix surrounding the NP and AF cells after exposure to IL-1. A similar study was performed using a brief exposure to the glycosidase chondroitinase-ABC (C-ABC) to deplete PGs and hyaluronan in the matrix around NP and AF cells.50 After the matrices around the cells were depleted by C-ABC digestion, the cultures were then maintained in media containing 10% FBS and OP-1 (100 ng/ml). As seen in the case of IL-1 exposure, OP-1 stimulated the repair of the C-ABC-treated matrix around NP and AF cells, actually resulting in an increase above the normal level of matrix molecules. These two studies became the fundamental basis for our strategy that growth factor injection therapy can be applied to repair the matrix of degenerated human IVDs or to promote matrix repair in conjunction with chemonucleolysis.

Because the cellular phenotypes of young animals, especially rat and rabbit cells, are different from those of adult human cells, studies using human disc cells are essential for the clinical development of growth factor therapy. There is considerable evidence in the recent literature that human disc cells, healthy or degenerated, will respond to growth factors and BMPs. Gruber et al. reported that human AF cells in three-dimensional cultures (agarose gel and alginate beads) produced and accumulated more abundant extracellular matrix molecules, such as PGs, than those cultured in monolayer; under those conditions; they also showed that cell proliferation was stimulated by TGF-β.51

Imai et al. reported that OP-1 (100–200 ng/ml) enhanced the production and accumulation of PGs by human NP and AF cells cultured in alginate beads in the presence of 10% FBS.52 Interestingly, AF cells, which are more fibrochondrocytic, strongly responded to OP-1, suggesting that OP-1 might be beneficial not only for nucleus repair but for anulus repair as well.

Recombinant human BMP-2 (rhBMP-2) has been shown to stimulate PG synthesis at relatively high doses (67% at 300 ng/ml and 200% at 1500 ng/ml rhBMP-2).53 Ahn et al. recently reported that both BMP-2 and BMP-12 stimulated PG and collagen synthesis by human NP cells from degenerated discs cultured in monolayer in the absence of serum.54

A decrease in cell number is one characteristic of aging IVD tissue.55 Growth factor therapy may be beneficial to maintain functional cells in IVD tissues. Gruber et al. reported a high incidence of apoptosis in AF tissues.56 Furthermore, they found that the surviving AF cells were not synthetically inactive but were, rather, producing inappropriate matrix molecules during aging and degeneration. Based on these initial findings, they studied the effects of IGF-1 and platelet-derived growth factor (PDGF) on apoptosis in human AF cells grown in monolayer culture.57 In monolayer cultures of human degenerative IVD cells, Wehling recently reported that the combination of autologous IL- 1 receptor antagonist (IL-1ra)/IGF-1/PDGF proteins, which was produced in vitro by stimulating monocytes and thrombocytes, reduced the percentage of apoptosis and the production of biochemical markers of disc degeneration, such as IL-1 and IL-6.58

Walsh et al. reported the in vivo effects of a single injection of other growth factors, such as bFGF, growth differentiation factor 5 (GDF-5), IGF-1, or TGF-β, in the mouse caudal disc with degeneration induced by static compression.59 An injection of GDF-5 was effective in promoting disc regeneration. Multiple injections (four injections, once per week) of TGF-β showed a stimulatory effect, although the other growth factors did not show a significant enhancement of their original effect.59

As the first feasibility in vivo study on the effect of OP-1, we reported that the in vivo intradiscal administration of the growth factor OP-1 resulted in an increase in disc height and in the PG content of the NP in normal rabbits.60 At 2 weeks after the injections of OP-1 (2¼g in saline), the mean disc height index of the OP-1-injected discs was 15% greater than that of the saline group, and the increase was sustained for up to 8 weeks. Biochemically, a significant increase in the PG content of the NP was observed in OP-1-injected discs when compared to the saline-injected group at the 2-week time point. Using a defined-gauge needle puncture of the AF,61 we conducted a study of the radiographic and magnetic resonance imaging (MRI) changes in the rabbit IVD after injection of OP-1 into the NP in the anular-puncture disc degeneration model. Six weeks after the OP-1 injection, a restoration of disc height and MRI changes were observed and sustained for the entire experimental period, up to 24 weeks (Fig. 2).62 In another degeneration model using C-ABC, 4 weeks following the injection of C-ABC, OP-1 (100¼g/disc) injection restored the disc height up to 12 weeks after the OP-1 injection.63 These data suggest that OP-1 may be utilized in patients with disc degeneration who had received chemonucleolysis in the past that resulted in some loss of disc height.

Nishida et al. injected an adenovirus construct encoding the TGF-β gene (Ad/CMV-hTGF-β₁) into lumbar discs of skeletally mature rabbits and showed the feasibility of using the gene therapy approach to induce biological changes in IVD tissues in vivo. The NP tissues from the discs injected with Ad/CMV-hTGF-β₁ showed a significant increase in production of active and latent TGF-β₁. They found that these tissues exhibited an increase of 100% in PG synthesis when compared to intact control tissues.64 However, the effect of TGF-β on PG synthesis remains controversial. Wallach et al., from the same group, recently reported the inhibition of PG synthesis by the transduction of the TGF-β ₁ gene.65 The authors suggested that the inhibitory response to TGF-β₁ may have resulted from increased cell-to-cell interactions within the three-dimensional pellet culture, which in turn may have produced higher, and potentially toxic, TGF-β₁ concentrations in the microcellular environment. More biologically relevant in vivo studies, such as those using a large animal model, might be required to reach a conclusion on the effect of TGF-β₁. Ahn et al. recently reported that the introduction of the BMP-12 gene into human IVD cells in pellet culture by an adenoviral vector significantly increased cell proliferation and matrix production (PG and collagen).66 The increase in mRNA for collagen types I, II, and VI in the AF cells was more pronounced than that in NP cells.

Matsumoto et al. have also shown that the nonviral gene transfer of the human OP-1 gene into bovine NP and AF cells, using the gene gun (particle-mediated gene transfer method), can stimulate PG synthesis.67 On day 3 after transfection of the human OP-1 gene, transfected cells synthesized PGs at a faster rate in both the AF (+24%) and NP (+44%). In an in vivo experiment, the injection of OP-1-transfected cells (allogenic cells from a donor rabbit) into the NP induced a significant preservation of disc height and retention of PG content when compared to the control vector-transfected group.68

Wang et al. showed that transfection of the GDF-5 gene into rabbit and human IVD cells with adenovirus encoding the GDF-5 gene or the TGF-β gene induced a significant increase of GDF-5 and TGF-β protein production,69 respectively.

Paul et al. reported the successful adenovirus-mediated gene transfer of the SOX9 gene and the BMP-2 gene, both with green fluorescent protein (GFP), into IVD cells obtained from human degenerated IVD tissues70 with a steady expression over the 80% level for 7 days in culture and also with an increased level of type II collagen synthesis. The in vivo transfection of the SOX9 gene by adenoviral vector using the rabbit stab model resulted in maintenance of a chondrocytic appearance over 5 weeks.70

Wallach et al. also reported that gene transfer of the tissue inhibitor of metalloproteinase 1 (TIMP-1), an inhibitor of catabolic enzymes, can increase PG accumulation in pellet cultures of human IVD cells.71 This study introduced a new approach to modulate a course of disc degeneration by inhibiting the catabolic pathway rather than stimulating the anabolic pathway. The combined use of anabolic factors with an inhibitor of a catabolic enzyme is one way to induce regeneration of IVD disc undergoing degenerative changes.

Yoon et al. reported that overexpression of LIM mineralization protein 1 (LMP-1),72 an intercellular protein, using an adenoviral vector (AdLMP-1) in rat IVD cells resulted in an increase of BMP-2 and BMP-7 gene expression and PG synthesis.73 BMP-2 and BMP-7 protein levels increased by two- to threefold. The mRNA levels of BMP-2 and BMP-7 were also upregulated. In addition, the in vivo injection of AdLMP-1 resulted in a significant elevation of mRNA levels of LMP-1, BMP-2, and BMP-7. Based on these findings, the authors suggested that LMP-1 upregulates the gene expression of BMP-2 and BMP-7 and that the increased level of BMP production resulted in an increased PG synthesis.

Based on these studies, gene therapy has the potential to upregulate matrix synthesis and downregulate catabolic processes by introducing cDNAs that encode specific proteins. With any gene therapy, safety issues should be resolved, as well as transfection efficiency and duration of action. In addition, further research using degenerated human IVD cells may be necessary before taking this approach to a clinical setting.

In the treatment of degenerative disc diseases, no clinical option currently exists between conservative therapies and more aggressive therapies, such as fusion or disc replacement. The prevention of disc degeneration and the stimulation of the biological disc repair process by the injection of growth factors, cytokine inhibitors, enzyme inhibitors, and cells transfected with therapeutic genes will create a new category of therapy for degenerative disc diseases. In addition, two or more different approaches might be applied, resulting in the prevention of disc degeneration and/or the active stimulation of the repair process of the degenerated disc. These therapies could produce an amelioration of pain and restructuring of the degenerated disc.

The basic concept of spinal instability is that excessive motion beyond normal constraints causes either compression or stretching of the neural elements or causes abnormal deformations of ligaments, joint capsules, annular fibers, or endplates, which are known to have a significant number of nociceptors.80 Even though several studies have indicated that excessive motion on flexion-extension radiographs is associated with degenerative disc disease,83 other studies cite decreased motion in patients with degenerative changes.84,85 Lumbar segmental instability may be associated with a spectrum of clinical manifestations of degenerative changes in the IVD.77,86,87 Intervertebral disc degeneration has been studied using magnetic resonance (MR) imaging, and grades of degeneration have been reported.87‐89 There also have been numerous biomechanical studies on disc degeneration and instability.85,90‐101 The relationship between the types (or grades) of disc degeneration and kinematic characteristics of the motion segment has been studied using cadaveric spinal motion segments. 102‐107 Despite variations in results, likely because of different loading conditions and methods of grading degenerative disc changes, the overall results of these studies indicate that the biomechanical characteristics of the motion segment can become significantly altered when degenerative changes develop in the IVD. In vitro biomechanical studies by Fujiwara et al. determined the changes in the rotational mobility of the lumbar motion segment in relation to the degenerative changes in the disc and facet joints.102 The authors found that segmental motion increased with increasing severity of disc degeneration to grade 4 but decreased when the disc degeneration advanced up to grade 5. Such segmental motion changes were much greater in axial rotation when compared to those in lateral bending, flexion, and extension, demonstrating the importance of torsional instability in diagnosing spinal instability. The results of these studies are important for understanding the kinematic changes in relation to the types or grades of disc degeneration. However, these results were obtained in vitro from cadaveric specimens. Therefore, we developed a noninvasive in vivo segmental lumbar motion measurement system and clarified the relationship between the lumbar segmental motion and disc degeneration, as indicated in the following sections.

Many methods have been used in an attempt to quantify segmental instability in patients, but the most typical method has been flexion-extension lateral radiographs. The use of radiographs has been shown to be limited by poor accuracy, and radiographic studies are impractical to measure out-of-plane rotation.108,109 In addition, the range of motion measured in many radiographic studies is affected by the variability in voluntary efforts that the subject applies at the time of examination and can also be limited because of pain. Other two-dimensional (2D) methods for measuring axial rotation, as opposed to flexion/extension, have involved MR imaging scanning of subjects in various rotated positions.109,110 Although these studies were noninvasive and controlled for voluntary motions, they could only determine changes in vertebral motion around one axis. It has been suggested that coupled motions could play an important role in determining spinal instability. To measure these coupled motions, studies have been conducted to measure 3D motions in vivo. More invasive techniques involve inserting wires into the spinous process of subjects to determine 3D motion.94 Although this method has proven more accurate than radiographs, its invasive nature limits its widespread clinical use. Other researchers have used biplanar radiography, in which subjects are filmed from two directions simultaneously and 3D motions are determined from anatomical landmarks in corresponding images.111,112 There has been some concern about the accuracy of determining anatomical landmarks for biplanar radiography, as well as a lack of equipment for this method in typical clinical settings. To combat some of these limitations to 3D motion measurement, Lim et al. developed a 3D imaging technique using parallel computed tomography (CT) scans to determine rotations and translations in individual cadaveric cervical vertebrae.108 The authors illustrated that accurate measurements (±1mm and ±1°) can be made using CT in vitro. We have expanded on this technique to measure vertebral segmental rotations and translations, the motion between adjacent vertebral bodies, in human lumbar spines in vivo.

Most patients with segmental instability have disc degeneration, but the relationship between instability and degeneration has not been clarified. Disc degeneration has been extensively studied, particularly with MR imaging.113‐116 Yu et al.116 described the MR imaging appearance of degenerating discs as a change in signal intensity associated with a radial tear of the AF. Schiebler et al.115 showed that before the obvious MR imaging signal changes, such changes as infolding of the central fibers of the outer anulus and the central dot may be early signs of disc degeneration. Recently, using MR imaging, Hedman and Fernie113 reported that certain patterns such as a dark nucleus with annular tears were highly associated with positive symptoms on discography. These studies indicate the ability of MR imaging to identify anatomical changes in the IVD and possibly identify dynamic changes of the disc with different loadings, and even, potentially, to identify symptomatic discs. Takekuchi et al.117 presented a study using MR images in which T₁ relaxation time was decreased in degenerative discs and the energy dissipated to axial loading was linearly correlated with T₁ relaxation time. The authors attempted to correlate the intrinsic biomechanical properties of the disc with MR imaging findings, but no information could be derived about the segmental motion characteristics from this study.

Results of the in vitro studies of segmental motion characteristics and disc degeneration done by Fujiwara et al. demonstrated that torsional motion was most significantly affected by the degenerative changes in disc and facet joints. In addition, some investigators advocate the importance of torsional loads and stability on the injuries and degeneration of the motion segments.118,119 However, torsional instability in relation to the degenerative changes in the disc had not been well investigated in vivo, mostly because of difficulties in measurements.

To resolve the accuracy problems in spinal motion measurements and to establish a meaningful relationship between segmental instability and degenerative changes in the disc and facet joints, we have established a method to measure 3D vertebral motion noninvasively using CT images of the vertebrae.108 We have developed an in vivo measurement system of lumbar segmental motion with an accuracy of less than 0.2mm in translation and 0.2° in rotation (Fig. 3). We also developed a fixture to apply axial rotation to the lumbar spine in a controlled manner with minimal voluntary effort of the subjects (Fig. 4). Using the new techniques, we found that a relationship exists between the severity of IVD degeneration and increases in torsional and flexion-extension movements in vivo, which was previously demonstrated only in cadaveric studies.120

Biomechanical alterations in motion resulting from IVD degeneration and their ensuing effects on facet kinematics have been largely ignored. The facet joints are one of the primary structures for stability in the spinal motion segment.90,99,100,121,123‐125,130,136,140,145‐156 Load-displacement relationships in the intervertebral motion segments are a complex interaction involving at least six degrees of freedom and structural and material nonlinearity.126,157 As such, alteration in the load-bearing role of the facet joint most likely occurs with IVD degeneration because of the alterations in the structural properties of the disc.97,135,141,158

However, it was not until the work of Fujiwara et al. that the association between segmental motion characteristics of the IVD and the facet joints was analyzed.102,159‐161 The most pronounced association was found between segmental motion and the severity of subchondral sclerosis. The authors believed that the severity of subchondral sclerosis reflected the magnitude of applied stresses on the involved facet joints. However, a mechanistic relationship between disc degeneration and facet joint degeneration still remains unclear. Further studies are needed to understand facet joint degeneration associated with disc degeneration.

The newly developed noninvasive 3D analytical method to determine segmental motion of the lumbar spine permitted the demonstration of a relationship that exists between the severity of IVD degeneration and increases in segmental lumbar motion in vivo. Abnormal motions could accelerate the facet degeneration processes and may lead to facet osteoarthritis. A better understanding of the effects of IVD degeneration on alternations in facet kinematics and load-bearing would provide a clear understanding of the pathogenesis of the facet degeneration derivative of IVD degeneration. A clarification of the relationship between structure and the mechanical properties of the degeneration of the disc and facet joints would also provide a strategy for treatment modality selection at different IVD degeneration levels.