Proper animal experimental designs for preclinical research of biomaterials for intervertebral disc regeneration

ABSTRACT

Low back pain is a vital musculoskeletal disease that impairs life quality, leads to disability and imposes heavy economic burden on the society, while it is greatly attributed to intervertebral disc degeneration (IDD). However, the existing treatments, such as medicines, chiropractic adjustments and surgery, cannot achieve ideal disc regeneration. Therefore, advanced bioactive therapies are implemented, including stem cells delivery, bioreagents administration, and implantation of biomaterials etc. Among these researches, few reported unsatisfying regenerative outcomes. However, these advanced therapies have barely achieved successful clinical translation. The main reason for the inconsistency between satisfying preclinical results and poor clinical translation may largely rely on the animal models that cannot actually simulate the human disc degeneration. The inappropriate animal model also leads to difficulties in comparing the efficacies among biomaterials in different reaches. Therefore, animal models that better simulate the clinical charateristics of human IDD should be acknowledged. In addition, in vivo regenerative outcomes should be carefully evaluated to obtain robust results. Nevertheless, many researches neglect certain critical characteristics, such as adhesive properties for biomaterials blocking annulus fibrosus defects and hyperalgesia that is closely related to the clinical manifestations, e.g., low back pain. Herein, in this review, we summarized the animal models established for IDD, and highlighted the proper models and parameters that may result in acknowledged IDD models. Then, we discussed the existing biomaterials for disc regeneration and the characteristics that should be considered for regenerating different parts of discs. Finally, well-established assays and parameters for in vivo disc regeneration are explored.

Introduction

The Global Burden of Diseases, Injuries, and Risk Factors Study conducted in 2019 reported that low back pain (LBP) was the leading cause of loss of years to disability from 1990 through 2019, affecting 568 million individuals and with an estimated 64 million years to disability globally.¹ Among 204 countries, LBP is reportedly the leading health condition contributing to the need for rehabilitation services in 160 countries.¹ Statistically, intervertebral disc (IVD) degeneration (IDD) contributes to 40% of LBP.² With aging, the economic and social burdens imposed by IDD, which is an age-related disease, are expected to progressively increase in the coming decades.³ Traditional treatments for IDD include physiotherapy, nonsteroidal anti-inflammatory drugs, lumbar epidural steroid injections, chiropractic adjustments, decompression, spinal fusion, and discectomy.^4-6 Although these therapies, especially surgical interventions, have presented favorable outcomes in terms of pain relief and disability improvement, gastrointestinal and cardiovascular adverse effects following prolonged nonsteroidal anti-inflammatory drug administration, the incidence of reherniation and recurrent back pain after discectomy, and the adjacent disc degeneration observed in spinal fusion render traditional therapies less effective.^7-9

Therefore, several advanced strategies that emphasize the regeneration of disc integrity and modification of the unfavorable microenvironment of degenerated discs have gained momentum. First, intradiscal administration of autologous or allogeneic stem cells/mature disc cells was performed.^{10, 11} Numerous clinical trials have shown that intradiscal injection of stem cells favors pain relief, with a 1-6-year follow-up period.^11-18 However, the sample sizes in these clinical trials were extremely small (less than 30), and long-term outcomes remained debatable. Bioactive reagents present an additional option for intradiscal administration. As an avascular organ, IVDs appear to prolong the retention of injected reagents when compared with that in articular joints.¹⁹ Ideally, extended exposure to bioactive molecules prevents repetitive injections that may predispose the discs to degeneration.^20-22 The application of biomaterials, such as nanoparticles, can help alter the original drug release pattern and even regulate release based on specific stimulation in the microenvironment.^{23, 24}

Numerous biomaterials for disc regeneration have been developed to modify the intradiscal microenvironment to favor cell survival, promote cell reparative effects, and control the release of therapeutic molecules, while others with satisfactory mechanical properties aid in the mechanical repair of impaired discs.^25-28 Preclinical evaluation of these biomaterials is critical for their further application in clinical trials. Animal models that resemble the characteristics of human disc degeneration play a pivotal role in preclinical experiments. Currently, numerous animal models with either spontaneous, mechanical alteration, or disc injury have been established for the preclinical evaluation of therapeutic strategies.^29-33 However, not all models satisfactorily simulate human disc degeneration. Factors, including animal age, disc geometry, size, and mechanical properties for selected animal models, could contribute to the bias of preclinical studies and clinical applications.^{34, 35} Apart from spinal fracture-related disc injuries, age-related human disc degeneration is an overall degeneration that influences all discs, specifically the lumbar discs, which hinders the efficiency of local administration.^{36, 37} However, most existing biomaterials for disc repair cannot be systemically administered. Therefore, for the in vivo evaluation of these biomaterials, animal models with regional degeneration (e.g., disc injury) are preferred, which is not often the clinical case primarily identified in age-related disc degeneration. Therefore, selecting an appropriate animal model that not only resembles human disc degeneration but also facilitates the in vivo evaluation of novel biomaterials remains a challenge. In addition, evaluation protocols and parameters for the outcomes of in vivo disc regeneration are to yet be unified, resulting in incomparable results among different studies and limiting clinical translation.

The articles about the establishment of animal models for intervertebral disc degeneration were retrieved by the search terms: Intervertebral disc (MeSH Terms) AND (Animal (MeSH Terms) OR Models, Animal (MeSH Terms) OR Animal Experimentation (MeSH Terms)). Then, the articles about biomaterials for disc regeneration were retrieved by the search terms: Intervertebral disc (MeSH Terms) AND Biomaterials (MeSH Terms). Then, the articles related to the characteristics that determines the outcome of intervertebral disc regeneration were retrieved by the search terms: Intervertebral disc (MeSH Terms) AND (Pain (MeSH Terms) OR Hyperalgesia (MeSH Terms) OR Allodynia (MeSH Terms) OR Biocompatibility (All Fields) OR X-ray (MeSH Terms) OR computed tomography (MeSH Terms) OR CT (MeSH Terms) OR Magnetic Resonance Imaging (MeSH Terms) OR Histology (MeSH Terms) OR Anatomy (MeSH Terms) OR Mechanical Tests(MeSH Terms) OR Torsion, Mechanical (MeSH Terms) OR Stress, Mechanical (MeSH Terms) OR Adhesives (MeSH Terms)). All these searches were perfromed on PubMed, Embase, Web of Science and CNKI databases prior to Feburary, 2021. The results were further screened by title and abstract. Irrelevant articles were excluded. In the end, 810 articles were included in this review (Figure 1). This review aims to provide cues for appropriate animal experimental designs for preclinical evaluation of biomaterials for IVD regeneration. We summarized the basic pathological characteristics of human degenerated discs, animal models that resemble human disc regeneration, and discussed suitable animal models for the preclinical evaluation of specific biomaterials. We then highlighted the existing biomaterials for disc regeneration and the characteristics that should be considered for regenerating different parts of discs. Finally, we explored well-established assays and parameters for in vivo disc regeneration (Figure 2).

Figure 1. The flow diagram of enrolling articles.

Figure 2. The structural diagram of this review. AF: annulus fibrosus; CT: computed tomography; IVD: intervertebral disc; MRI: magnetic resonance imaging; NP: nucleus pulposus.

Pathological Alteration of Intervertebral Disc

Nucleus pulposus

Young and healthy human nucleus pulposus (NP) is a gel-like tissue with an 80% water content and contains two types of cells: notochordal cells and mature NP cells.³⁸ The former are large vacuolated cells that originate from the embryonic notochord and gradually disappear in an age-related manner. The latter type of cells are the major residents in the adult disc.³⁹ Both cell types play a vital role in maintaining the integrity of the NP matrix. The extracellular matrix (ECM) in healthy NP and the inner annulus fibrosus (AF) is mainly composed of loosely arranged type II collagen fibers and proteoglycans.⁴⁰ Proteoglycans and glycosaminoglycans (GAGs) maintain high osmotic pressure and hydration in NP tissues.⁴¹ Their interaction with cells and cytokines regulates cell biology through various signaling pathways.⁴²

Cellular changes during nucleus pulposus degeneration

In NP tissues, the change in cell types begins in childhood, and notochord cells gradually disappear with age. Notochord cells play a crucial role in protecting NP cells and promoting their anabolism.^{43, 44} The elimination of notochord cells is predominant in the initiation and development of IDD.⁴⁵ Reduced cell numbers and impaired cell viability in degenerative discs are closely related to the excessive activation of multiple programmed cell death pathways, including apoptosis and necroptosis. Our research group is committed to investigating NP cell death and underlying mechanisms, such as the dysfunction of autophagy and abnormal activation of heat shock protein 90.^46-48 Identifying the underlying mechanism of notochord cell reduction and cell death in an unfavorable intradiscal microenvironment provides further information and alternative targets for biomaterial design and fabrication.

In IDD, disc cells undergo cell death and demonstrate dysfunction in an age-related manner. Among dysfunctional cells, senescent cells play a crucial role in the pathology of IDD.⁴⁹ Cell senescence is characterized by a cell state of proliferating arrest and secretion of senescence-associated secretory phenotype.⁵⁰ Senescent cells are detrimental to tissue renewal and repair and are known to secrete several proinflammatory factors, including chemokines, cytokines, protein enzymes, and other bioactive factors, which further destroy the living environment of surrounding cells and place degenerated discs at risk of a vicious circle.⁵¹ Contrary to the effective self-clearance of senescent cells in other tissues, the avascular nature of discs partly limits immune-mediated clearance and causes abnormal accumulation.^{52, 53} Although senescent cells are characterized by permanent cell cycle arrest, temporary replication stagnation can be reversed when intervened during early stages with reduced risk factors.⁵⁴ Instead of reversing senescent cells and restoring their function, another strategy focusing on removing these cells, called senolysis, has been proven effective in prolonging life expectancy^55-57 and ameliorating various age-associated disorders, including cognitive impairment, vascular disease, and cardiac dysfunction.^58-60 Strategies that target cellular senescence in the field of IVD regeneration warrant further research.

During IDD, the NP cell phenotype changes as microenvironments become increasingly unfavorable, such as accumulating ECM degradation products and lactic acid.^{61, 62} Phenotypic changes in NP cells, such as the production of proinflammatory cytokines and chemokines, lead to an increase in ECM degradation, with a decrease in the level of anabolic factors and synthesis of healthy ECM.^63-67 In addition to synthesizing matrix metalloproteinases (MMPs) and a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTSs) that mediate ECM degradation, NP cells secrete chemokines, C-C chemokine ligand (CCL)2, CCL3, and (C-X-C motif) ligand 10, to stimulate the recruitment of immune cells that produce interleukin (IL)-1β and tumor necrosis factor (TNF)-α, which further stimulates the production of MMPs and ADAMTSs, thus amplifying the pro-catabolic alteration of ECM.^68-70 Additionally, inflammatory factors can promote cell apoptosis, further accelerating IDD.⁷¹ Therefore, although the IVD is considered an avascular tissue, inflammatory factors secreted by invading inflammatory cells and NP cells form a continuous inflammatory microenvironment, which plays a vital role in the development of IDD.^72-74 The alleviation of inflammation in the IVD is a crucial issue in delaying IDD.

Disruption of extracellular matrix homeostasis

The NP tissue is composed of abundant ECM that maintains a dynamic balance between synthesis and proteases-induced degradation.⁷⁵ Following IDD, ECM anabolism gradually decreases, while catabolism is accelerated, which eventually leads to an imbalance in ECM metabolism. The catabolism of NP ECM is mainly mediated by two enzymes: MMP and ADAMTS. As mentioned above, inflammatory factors can promote MMP and ADAMTS secretion, fostering ECM degradation. An increase in ECM degradation products can also induce the secretion of IL-1β, TNF-α, IL-6, and other inflammatory factors, which further promotes ECM catabolism and accelerates IDD.⁶³ The metabolic imbalance of ECM leads to a gradual decrease in the proteoglycan content, with alterations in collagen types and organization.^76-78 Eventually, the gel-like tissue is gradually replaced by a consolidated fibrous structure that fails to distribute the axial pressure evenly and limits segmental motion.^{79, 80}

Furthermore, ECM homeostasis depends on cytokines. Transforming growth factor-β (TGF-β) signaling plays a protective role in ECM homeostasis by stimulating matrix synthesis, inhibiting matrix catabolism, inflammatory response, and cell loss.⁸¹ TGF-β is also known as one of the most potent inhibitors of TNF-α-induced MMP upregulation and matrix degradation.⁸² However, excessive TGF-β activation can accelerate disc degeneration.^{83, 84} Aberrant mechanical loading resulted in excessive activation of TGF-β signaling and IDD, while suppressing TGF-β signaling attenuated IDD, which may be attributed to variations in Smad activation.^85-87 Moreover, the bone morphogenetic protein (BMP) family, including BMP-2 and growth and differentiation factor-5 (GDF-5), can reportedly promote the synthesis of ECM components and reduce MMP expression.^{88, 89}

Annulus fibrosus

AF consists of a unique and complex structure of 15-25 concentric layers that are generated from packed collagen fibers (mainly type I collagen).⁹⁰ Type I collagen increases from the inner to the outer AF, and the opposite trend can be observed with regards to the content of type II collagen and aggrecan.^{40, 91} The interlamellar matrix between adjacent layers is composed of elastic fibers, cells, water, lipids and proteoglycans, etc.⁹² Although the primary resident cell type, often referred to as AF cells, is annulocytes,⁹³ various other cell types are present within the AF, such as the AF stem/progenitor cells and cells in the interlamellar matrix with different cell morphologies, which are influenced by the mechanical loads, elastic fiber orientation, and density.^{92, 94, 95} AF cells are characterized as elongated and spindle fibroblasts with extended cytoplasmic processes. During degeneration, AF cells become more rounded and chondrocytic, developing multiple cytoplasmic processes that extend extracellularly.^{96, 97} The structural integrity of the AF tissue is critical for confining the NP, as well as for maintaining the physiological loading pressure of the spine.²⁷ With aging, the AF structure gradually becomes disordered, and the cartilage-like matrix accumulates, resulting in weakened tensile strength.⁹⁸

Mechanical disturbance of degenerated annulus fibrosus

Biomechanical “wear and tear” plays a vital role in AF tissue degeneration. When exposed to higher mechanical stress, type I collagen and aggrecan production decreased, while tissue inhibitor of metalloproteinase-1 increased, which could induce ECM remodeling to a degenerative state.⁹⁹ With progressing IDD, the mechanical stress is shifted from hydrostatic pressure to shear stress, which reportedly decreases proteoglycan production and increases apoptosis by enhancing nitric acid production.^{100, 101} A disturbed AF fiber structure and prolonged biomechanical changes lead to acute tears and fissures in AF tissues, forming a stress point where the structure around AF fissures, including fibers and NP tissue, undertake “push-out” forces that aggravate AF tissue damage and NP herniation, which is closely associated with discogenic pain.^{102, 103} AF damage caused by degeneration and biomechanical changes (e.g., overloading) are crucial factors for ECM remodeling and cellular pathology.^{98, 99} Therapeutically, it is ideal to repair annulus fissures early, thus reducing the need for future surgery.¹⁰⁴

Proinflammatory microenvironment of the injured annulus fibrosus

Mechanical stress also promotes the upregulation of proinflammatory genes in AF cells, including cyclooxygenase-2, IL-6, and IL-8.^{105, 106} The secreted chemokines and damaged tissue fragments result in the recruitment of immune cells to the wound area to eliminate tissue debris while forming an inflammatory microenvironment. Inflammatory cells, including macrophages, T lymphocytes, and mast cells, have been recognized in the region of injured AF tissue.^{93, 107-109} AF injury with a weak immune cell response results in poor healing outcomes.¹¹⁰ This phenomenon suggests that recruitment of inflammatory cells with an appropriate inflammatory phenotype and timely reversal to an anti-inflammatory healing state may be necessary to repair AF injury. However, NP herniation has been shown to induce a prolonged immune response associated with the exposure of concealed antigens from NP tissue to the immune system.^{111, 112} Herniation-related chronic inflammation is one of the key factors that induce phenotype changes and predisposes the tissue to degeneration and painful conditions.^{113, 114} Strategies that modify the inflammatory microenvironment have shown great potential for promoting disc cell survival and favoring tissue regeneration.^115-117

Endplate and bone tissue

Endplates consist of hyaline cartilage and osseous components (subchondral bone) at the cranial and caudal ends of discs. Endplates can prevent disc extrusion into the porous vertebral body and evenly distribute mechanical loads to the adjacent vertebral body.^{118, 119} With aging, the endplate gradually undergoes calcification, which affects the nutrient supply of the IVD. Conversely, as the endplate becomes thinner and bone mineral density is lost, the risk of endplate fracture increases.^{118, 120-122} The damaged endplate then extrudes into the adjacent vertebrae, increasing the NP volume and resulting in a 30-50% drop in NP pressure and uneven load distribution to the vertebral body.¹²³ These pathological and biomechanical alterations lead to endplate-driven degeneration.

In patients with IDD, the Modic change is acommon observation on magnetic resonance imaging (MRI), referring to the signal change of the vertebral endplate and subchondral bone, which is related to lumbar disc herniation and LBP.^124-126 Modic changes can be divided into three types based on differences in MRI signals. Type 1 changes refer to hypointensity on T1-weighted imaging (T1WI) and hyperintense on T2-weighted imaging (T2WI), which reveal the fracture or tear of the cartilage endplate, as well as revascularization and fibrous tissue formation in the adjacent cancellous bone marrow cavity. In terms of its pathological significance, the hematopoietic bone marrow is characterized by edema and inflammation. Type 2 changes refer to hyperintensity on T1WI and isointense or slightly hyperintense on T2WI. In this case, the pathological significance is that normal hematopoietic bone marrow is replaced by fatty bone marrow. Type 3 changes refer to hypointensity on both T1WI and T2WI and are considered to represent subchondral bone sclerosis.¹²⁷ Disc/endplate damage, occult discitis, and autoimmunity are potential risk factors for Modic changes, among which inflammation plays a vital role during the pathological process.^{128, 129} Studies have revealed that Modic type 1 changes are more substantially associated with LBP than other types.^{127, 130} This association may be related to inflammatory irritation of the dorsal root ganglion following injury and disc degeneration.¹²⁹ A previous study revealed that in patients with IDD and Modic type 1 vertebral endplate changes, immunoreactive nerve ingrowth and increased TNF-α expression can be observed in the vertebral endplate.¹²⁹ A recent study has reported that sensory innervation in porous endplates induced by osteoclasts may play an important role in spinal pain.¹³¹

Nerve ingrowth and vascularization

In adult human discs with no apparent histological degeneration, nerves are restricted to the outer third of the AF, while a few nerves can be observed in endplates with similar densities within various anatomical regions.^{132, 133} Damage and degenerative changes induce nerves to grow inward, resulting in an expanded distribution of nociceptive nerve endings and LBP.¹³² However, the mechanisms underlying increased nerve ingrowth during disc degeneration need to be clarified.¹³⁴ Several studies have shown that nerves are confined to proteoglycan-depleted regions of disrupted tissue, especially within annulus fissures, which may be attributed to the inhibitory effects of aggrecan on nerve ingrowth.^135-137 In addition, osteoclasts play a critical role in the ingrowth of sensory nerve porous endplates, in which Netrin-1 derived from osteoclasts is found to be involved.¹³¹ An increasing number of studies have revealed that nerve growth into the IVD is mediated by nerve growth factors secreted from IVD cells, nerve cells, and inflammatory cells.^138-140 Elevated nerve growth factor expression closely correlates with the inflammatory microenvironment of degenerated IVDs.^{138, 139, 141, 142} However, the specific mechanisms of innervation and the generation of discogenic pain are not yet well understood and could provide novel therapeutic strategies for LBP. Furthermore, a previous study revealed that although there are more ingrown nerves in endplates than in annulus fibers, many innervated endplate pathologies are undetectable on MRI.¹³³ Therefore, further research on new visualization methods is critical to better evaluate nerve ingrowth on a full scale while considering the whole disc, as well as to reveal the relationship between nerve ingrowth and LBP or the degree of IVD.

Abnormal vascularization is another pathological change associated with IDD. IVD is recognized as the largest avascular structure in the human body. In healthy adult disc tissues, a few blood vessels can be found in the outermost lamellar layers of AF, while no obvious vascularization can be identified in cartilage endplates and the NP.^{143, 144} However, abnormal vascularization has been frequently identified in damaged or disrupted ECM of cartilage endplates and inner layers of AF.¹⁴⁵ Interestingly, nerve ingrowth is often accompanied by vascularization, which provides oxygen and nutrients to the nerve.^{135, 146} Therefore, consistent with nerve ingrowth, vascular ingrowth is more likely to be localized near or within damaged tissue, which is probably due to the disruption of anti-angiogenic factors (e.g., proteoglycans and aggrecans) and the increased secretion of angiogenic growth factors and cytokines (e.g., vascular endothelial growth factor and IL-1β).^{135, 146} IL-1β can upregulate the expression of vascular endothelial growth factor, thereby triggering neovascularization in the IVD.^{141, 142} Accordingly, it can be stated that inflammation plays a vital role in IVD cell dysfunction, reduction in cell number, ECM metabolic disruption, and nerve and blood vessel growth. Therefore, inhibiting inflammation can serve as an important target for IVD therapy and delay IDD progression at multiple levels.

The harsh microenvironment in the degenerated disc

In IDD tissue, a harsh microenvironment is known to exist, including inflammation, low oxygen concentrations, acidity, and hyperosmolarity, which are detrimental to cell survival.¹⁴⁷ As a typical characteristic of IDD, the inflammatory environment is induced by changes in the NP cell phenotype, inflammatory cell infiltration, and cell senescence.^148-150 Inflammation further promotes the dysfunction and apoptosis of disc cells, aggravates the disruption of ECM metabolism, and accelerates IDD development.^{76, 151-153}

The IVD itself is inherently avascular and consequently establishes a hypoxic microenvironment, especially in the NP.¹⁵⁴ In relation to anaerobic glycolysis, IVD cells modulate their metabolic strategies to adapt to the hypoxic low-glucose environment and maintain their viability, which leads to lactic acid accumulation within IVD tissues.¹⁵⁵ Therefore, the average pH is slightly acidic (7.0-7.2) under physiological conditions. However, in mild degenerative conditions, the pH may drop to 6.5, and even to 5.6, in severely degenerated IVDs, which has a detrimental effect on NP cell viability and ECM homeostasis.^{156, 157} Hypoxic conditions facilitate energy metabolism and type II collagen production while reducing the unfavorable damage induced by oxidative stress and cell apoptosis.¹⁵⁸ More interestingly, hypoxia may favor mesenchymal stem cell survival in the hostile IVD microenvironment after implantation;¹⁵⁹ in contrast, it has been reported that prolonged exposure to severe hypoxia under serum-deprivation conditions eventually results in complete cell death.¹⁶⁰

The NP tissue is rich in GAGs characterized by negatively charged side chains of aggrecan molecules, leading to hyperosmolarity within the NP tissue. NP cells can adapt to high osmolarity by regulating the expression of tonicity enhancer-binding protein.¹⁶¹ However, with progressive loss of proteoglycans with IDD, the osmolarity declines during the degenerative process.¹⁶² Under relative hypo-osmolarity, the apoptosis of NP cells is significantly increased.¹⁶³ Furthermore, high osmolarity is detrimental to the viability and proliferation of exogenous mesenchymal stem cells, such as bone marrow stromal cells and adipose-derived stem cells, while the relative hypo-osmolarity promotes NP-derived stem cell proliferation and chondrogenic differentiation.^{164, 165} Therefore, the distinct preferences of stem cells and mature disc cells for osmolarity should be considered when designing cell delivery strategies that facilitate the survival and biofunction of implanted cells.¹⁶⁶

Animal Models

With the continuous development of IVD pathophysiology and material science, it is imperative to establish appropriate animal models that can accurately simulate the pathological and biological properties of human IVDs. Unfortunately, there is currently no recognized model that meets these requirements. The selection of animals for establishing the IDD model needs to consider the following points:

1. IVD geometry: The shape of the IVD determines the state of deformation of various parts when under stress.^167-169 An unreasonable geometric model would fail to accurately reflect the pressure on each component when the IVD is stressed, which leads to inaccurate findings. One previous study has comprehensively evaluated animal disc geometry, including axial cross-sections, shape and position of the NP, and relative disc height of the species used in the disc research.¹⁷⁰ In terms of the geometric parameters of the disc height, AF width, and NP area, the mouse lumbar, rat lumbar, and mouse tail discs most correlated with the human lumbar IVD geometry.¹⁷⁰

2. IVD disc mechanics: The force imposed on human lumbar discs comprises the axial pressure caused by the upper body weight and dynamic pressure during activities.^{90, 148, 171} Accurate simulation of this pressure mode is critical for evaluating the in vivo biomechanical properties of implanted biomaterials. Although bioreactors ideally simulate the pressure environment of human IVDs ex vivo in controllable magnitudes, the simulation of human disc dynamic pressure performed with an external device in an animal model is still limited.^{172, 173} Therefore, the mechanical characteristics of animal models are particularly important. Most animal models include rodents, rabbits, dogs, sheep, pigs, and cows, which are quadrupeds. The pressure on the IVDs of these animals is mainly caused by the paraspinal muscles and ligaments. Maintaining spine stability may require greater muscle tension and passive tension to maintain a stable horizontal state,¹⁷⁴ compared with erect state. The disc pressure caused by this position may not be less than that in the human body.¹⁷⁵ A comprehensive review has summarized the mechanical properties of human and other animal discs.³⁵ Disc axial mechanics normalized by disc height and area were similar among species. Nevertheless, the normalized stiffness of calves and pigs is slightly more than that of human discs, while that of mice and rats is significantly less than that of humans.³⁵

3. IVD size: The IVD size determines the permeability of the tissue fluid and transport of implanted drugs and nano-microparticles.¹⁷⁶ As the IVD is an organ lacking blood vessels, the nutrition obtained by the cells mainly depends on nutrient infiltration.^177-179 Simulating the osmotic dynamics will reduce deviations. Furthermore, the size of the IVD also determines the surgical approach for biomaterial implantation. Compared with larger discs, hydrogels may better repair AF defects of smaller volumes. A critical factor determining annulus repair outcomes is whether the biomaterial can be well anchored on the local defect. In a small annulus defect, the surface tension of the hydrogel may play a role in the fixation of the material, while better adhesion properties are required to achieve satisfactory reparative effects in larger annulus defects.¹⁸⁰

4. IVD components: A disc model similar to human IVD components should mainly include cellular and biochemical components. The notochord cells in the human IVD gradually decrease from birth and disappear in adulthood.¹⁸¹ Human notochord cells can proliferate and differentiate into mature NP cells that secrete the ECM.^{182, 183} It is currently believed that the reduction in human notochord cells plays a critical role in IDD with age.^{45, 184} Like humans, notochord cells of sheep, goats, horses, and cattle rapidly decrease after birth.^{175, 185, 186} However, in most other mammals, notochord cells in the NP tissue persist throughout a considerable portion of their life, including mice, rats, rabbits, and pigs.^{45, 187, 188}

   Notably, there are two types of dogs. Chondrodystrophoid (CD) dogs, such as Dachshund and Beagle, which demonstrate a shortening of the long bones and decreased notochordal cells after birth. However, notochordal cells persist in non-chondrodystrophoid (NCD) dogs, like hounds, leading to a lower incidence of IVD.¹⁸⁹ Furthermore, the biochemical composition of IVD is another factor that differs between humans and other species. For example, rodents do not express MMP-1, a general and critical matrix regulator that participates in the ECM catabolism of human IVDs.¹⁷⁵

5. Animal age: IDD is an age-related disease. With the increase in pig age (newly born: 2-3 weeks; mature: 6-9 months; older: 2-3 years), the ECM protein of the AF gradually decreases, while that of the NP first increases and then decreases.¹⁹⁰ When CD dogs are 3-7 years old, the thoracic and lumbar IVDs present degenerative morphology, while NCD dogs have similar pathophysiological changes of IDD when 6-8 years old.¹⁹¹ Clarifying the animal age for degenerative morphology or controlling the influence of age in external stimulation-induced disc degeneration (e.g., needle puncture and compression) should be carefully considered for robust experimental designs that enroll animal models of IDD.

6. Animal sex: Although human disc morphology showed no significant differences between males and females, other animals showed a unique relationship between sex and disc degeneration or related LBP.^{192, 193} The disc degeneration grade was higher in female Sprague-Dawley rats than in male rats after annular puncture injuries.¹⁹⁴ In a rat model of spontaneous degeneration, females showed a greater incidence of radiologic disc space narrowing and wedging than males.¹⁹⁵ In terms of discogenic pain, female rodents demonstrated increased sensitivity to nerve root injury, and the prevalence of LBP is greater in women than in men.^196-198 However, after annular puncture, paw withdrawal thresholds of female rats were more variable, and normalized paw withdrawal thresholds did not significantly differ between sham and injury groups; however, annular puncture induced significantly decreased paw withdrawal thresholds in male cohorts. Estrogen variation may underlie the controversial results in female models.¹⁹⁹ Therefore, male animals may be more suitable for establishing a reliable discogenic pain model.

Primates may be the most appropriate after considering all these factors. 1) Non-human primates have semi-upright and upright characteristics.^200-202 2) The shape, size, and geometric structure of the IVD are extremely similar to those of human.^201-203 3) Non-human primates share disc degeneration biomechanical properties and pathological patterns with human.²⁰² 4) Age-related disc degeneration of non-human primates simulates the pattern of human disc degeneration. For example, the notochord cells naturally degenerate, simultaneously. The aging spines of rhesus monkeys are afflicted with disc degeneration, osteophytosis, and kyphosis, while these degenerative changes are most severe in the thoracolumbar and lumbosacral zones in human^{200, 204-206} However, the large size of primate discs makes the modeling operation more varible and less stable, which may reduce the comparability among studies.²⁰⁷ Also, ethical and cost restrictions hinder the application of primate models.²⁰⁸ In fact, besides aging model, local needle puncture model and chemical stimulation with pingyangmycin and bleomycin have been performed to accelerate the progression of primates disc degeneration, while pingyangmycin and bleomycin results in more mild and slowly progressive disc degeneration.^209-218

Sheep may be another suitable candidate for the following reasons: 1) The absence of notochord cells in adulthood; 2) a disc size similar to that of humans; and 3) mechanical characteristics are similar to those of humans.^{170, 185, 219} Currently, there are various approaches to establishing animal IDD models^220-226 (Additional Table 1 ^{(316.1KB, pdf)} ).

Spontaneous

Aging

As mentioned in an earlier section, naturally occurring animal aging predisposes to IDD in certain species, and the pathological performance is substantially similar to that of humans.

CD and NCD breeds can be distinguished based on their physical appearance.²²⁷ Specifically, due to disrupted endochondral ossification, CD dog breeds (e.g., Beagles and Dachshunds) have short bowlegs, and CD dogs are closely linked with severe IDD.²²⁸ In CD dog breeds, IDD (mainly Hansen type I herniation) typically develops in the cervical or thoracolumbar spine at approximately 3-7 years of age.²²⁸ NCD dog breeds (e.g., hound) can also develop IDD (mainly Hansen type II herniation), but in the caudal cervical or lumbosacral spine at about 6-8 years old, primarily attributed to trauma or “wear and tear.” The macroscopic, histopathological, and biochemical changes, as well as the diagnostics and treatment of IVD disease, are similar in NCD and CD dogs.^{228, 229}

The mouse model is one of the most applied animal models for IDD owing to its availability, economy, ease of operation, similar genomic pattern to humans, and ease of obtaining ethical approval. The mouse spontaneously developed disc degeneration in an age-related manner. IVDs in mice less than 14-18 months of age reportedly show no significant degenerative signs, although disc degeneration was found to start from 3-6 months.^{30, 230-232} Moreover, a moderate to severe lumbar disc condition was observed by MRI analysis and histological grade in 22-month-old mice.²³⁰

Baboons are quadruped for locomotion but spend a considerable proportion of their lifetime in the upright position, which imposes chronic spinal mechanical loading.²³³ Their life expectancy is 30-45 years.²⁰¹ Reportedly, an aging baboon was found to routinely demonstrate radiographic findings of disc degeneration similar to those in humans, including disc space narrowing, endplate sclerosis, and osteophytosis.²³⁴ Statistically, the average age at which baboons developed radiologic grades 1, 2, and 3 were 17.41, 19.94, 20.05 years.²⁰⁰

Gene mutations

Gene engineering is a common tool to investigate the specific roles of certain genes, non-coding RNAs, and proteins in disease development and progression. Mutation of ECM genes, such as collagen and aggrecan, induces degenerative morphology, including NP shrinkage or disappearance and fissures in the AF, which can sometimes lead to herniation of disc material and slight osteophyte formation, along with progressive joint degeneration.^{200, 235-249}

Although genetically modified mice with ECM gene mutations have revealed the significance of these genes in maintaining IVD integrity, age-related degeneration is more relevant to the human disorder. DNA damage is a critical feature of senescence.²⁵⁰ A failure in DNA repair is a common approach for inducing progeroid syndrome. Mice deficient in the DNA repair endonuclease, ERCC1, were developed to study accelerated aging.^{251, 252} Ercc1(-/Δ) mice represent an accurate and rapid disc aging model, including premature loss of disc proteoglycan, reduced matrix proteoglycan synthesis, and enhanced apoptosis and cell senescence.²⁵³

Secreted protein, acidic and rich in cysteine, also known as osteonectin and BM-40 (40-kDa basement membrane protein), is a matricellular protein essential for tissue remodeling.²⁵⁴ Secreted protein acidic and rich in cysteine (SPARC)-null mice showed signs of movement-evoked discomfort as early as 3 months of age.²⁵⁵ More importantly, SPARC-null mice developed region-specific, age-dependent hypersensitivity to cold, icilin, and capsaicin (hind paw only), as well as axial discomfort, motor impairment, and reduced physical function.²⁵⁶ Therefore, both structural and functional alterations of SPARC-null mice suggest its superior representation of human IDD.

Mechanical alteration

Compared with the general population, drivers, athletes, and workers undertaking heavy labor are more inclined to develop LBP, in which biomechanical “wear and tear” plays a critical role in the development of IDD.^257-259 Excessive mechanical loading leads to dysfunction of the energetic metabolism of IVD cells, disc inflammation, apoptosis, necroptosis, and imbalanced catabolic and anabolic metabolism.^{47, 260-270}

Mechanical factors that induce IDD include the gravity generated by the upper body when walking upright and the torsion and shear force in activities, such as bending. Several animal models have been developed to alter disc biomechanics and induce disc degeneration, including spine instability, tail suspension, amputation of the upper limbs, tail bending, spinal shear stress, and microgravity.^{31, 271-275} However, most mechanical modifications fail to accurately simulate both static and dynamic biomechanics of the human disc.

Spinal instability

The spinal instability model involves damaging the muscles and ligaments around the spine, causing mechanical instability in the corresponding spine segment. Generally, spinous processes are resected along with the supraspinous and interspinous ligaments^{31, 83} (Figure 3). On removing these structures, the remaining muscles and ligaments form an uneven tension around the disc segments, resulting in persistent abnormalities in spinal mechanics during daily activities. With progressing days, the physiological curvature of the spine gradually disappears, along with the gradually decreasing NP tissue, increased AF microfissures, and deformed or broken endplates in severe cases.^{157, 276} The mouse model with IVD instability showed significant histological degeneration within one week of surgery. Additionally, the IDD grade in the 12^th month after establishing the mouse instability model was comparable with that observed in the 18-month age group.³¹ Additional ovariectomy aggravates degenerative morphology and promotes vascularization into the discs.²⁷⁷ Mechanical instability has been shown to promote nerve invasion into IVD tissues, resulting in hypersensitive pain, which is a critical clinical symptom of patients with IDD.²⁷⁸ Therefore, spinal instability is a reliable strategy for creating degenerated mechanical performance. Additionally, although laminectomy has been adopted to establish the ex vivo porcine or sheep lumbar disc instability models,^{279, 280} in vivo spine instability model for large animals is still lacking, and which kind of spine instability model better resembles the biomechanical properties of human degenerative discs is still a maze.²⁸⁰ Moreover, the mechanical alteration is limited to operated segments, which cannot compete with the systematic disc degeneration of spontaneous models. Furthermore, the tissue damage is markedly severe, and several spinous processes are destroyed, leading to neuralgia after tissue injury rather than discogenic pain in IDD.

Figure 3. Lumbar spine instability mouse model (LSI). Mouse L3-5 spinous processes were resected along with the supraspinous and interspinous ligaments to induce instability of lumbar spine. Reprinted by permission from Macmillan Publishers Ltd.: Bian et al.⁸³ Copyright 2016.

Static/dynamic compression

Different static/dynamic compression models are shown in Figure 4.

Figure 4. Summary of static/dynamic compression models with external apparatus. (A) Tail-suspended rat with its hind limbs off the floor.²⁷² (B) Shear loading is generated from forces of different magnitudes on the adjecant vertebrae.^{281, 282} (C) Static disc bending model based on pins and an alignment jig.²⁸³ (D) Ilizarov-type apparatus is used to produce tail torsion. (E) Surgically implanted transfixing pins and percutaneous posts allow the application of static or dynamic axial compression and distraction loading at a single level of the rabbit lumbar spine.²⁸⁴ (F) Ilizarov-type apparatus is used to produce tail axial compression.

Suspension and microgravity

Suspension simulates an enhanced tensile force on the spine. On hanging by the tail, the IVD experiences a low compressive force similar to weightlessness during space flight.^{272, 285} In turn, low hydrostatic pressure is produced. Furthermore, tail suspension creates an extensive tensile force on IVD, especially the annulus (e.g., bending stretches the posterior annulus, and twisting induces tension in the whole annulus).²⁷² Tail suspension is economical and well-established, with a string and pulley system to maintain the hind limbs off the ground (Figure 4A). Reportedly, proteoglycan levels decrease by 35% after rat tail suspension for 4 weeks.^{272, 274} Catabolic genes (MMP3 and Admts5) were significantly upregulated in NP and AF tissues after 6 weeks, but degenerative histological changes were not apparent.²⁸⁶ In mouse tail suspension, the lumbar IVD height index and matrix protein expression levels were significantly decreased, with delayed cell cycling, increased proportions of senescent cells, and senescence-associated secretory phenotype, suggesting an age-related pathological alteration.¹⁵²

Shear stress

Rotation, or body twisting, induces torsion or shear stress in various parts of the IVD. Generally, peripheral surfaces are subjected to the greatest stress, consequently developing maximum strains.^{28, 287} A stainless steel shear loading device has been developed to apply a static shear load of up to 4 N to intervertebral joints via attachment to the indicated vertebral bones of the rat in the dorsoventral direction²⁸¹ (Figure 4B). With an adjustable spring force, the shear force imposed on the disc can be easily controlled at approximately 4 ± 1 N. After shear loading, the posterior annulus initially curves into the corresponding portion of the NP. Impressively, over 2 weeks, the NP tissue completely disappeared, and continued loss of the typical lamellar architecture of the inner and middle annulus resulted in a more severely disorganized tissue after surgery. Another similar external loading device was performed on rabbits to exert an adjustable shear force to around 50 N, and induced significant disc height narrowing as well as degenerative morphology after 1-2 months.²⁸²

Bending

Spine bending is a general posture change that imposes excessive deformation and compression on the concave AF and NP; for example, when a human picks up or participates in specific activities, such as farming or lifting. Rat tail bending is mainly performed owing to availability and stability. Tail bending achieved by external devices results in different mechanical and cellular alterations on the concave and convex sides (Figure 4C). With excessive compression, aggrecan expression decreased in the concave annulus when compared with the convex annulus in both the rat bending model.²⁸³ More cell death was observed in the concave annulus (compression) than in the convex annulus (tension).²⁸⁸ Although tissue denaturation is more evident on the concave side during spine bending, NP herniation typically occurs on the convex side. Therefore, the tissue regeneration strategy should focus on the unfavorable tensile stress on the convex annulus.

Torsion

Torsion is often accompanied by shear stress. Ilizarov-type fixators are similar to the components, organization, and manufacturing processes of the Ilizarov-type apparatus that induces static axial compression (Figure 4D). By altering the angle between the carbon fiber rings, rotation of various angles can be generated. By employing motor, cyclic rotations were performed at different frequencies. Maintaining a ±30° orientation significantly promoted the expression of proinflammatory cytokines (IL-1β and TNF-α) and catabolic genes.²⁸⁹

Based on our understanding, when processing an Ilizarov-type apparatus to generate static compression, the dislocation of carbon fiber rings, in either the sagittal or coronal plane, is inevitable during initial manufacture or later animal activities. Therefore, an unpredictable amount of torsion is inevitable when using the Ilizarov-type apparatus.

Axial compression

Axial compression represents pressure along the spine while standing. For inducing compressive IVD stimulation, a dynamic-loading rabbit model has been established to perform controlled and dynamic axial loading on rabbit lumbar discs. Surgically implanted transfixing pins and percutaneous posts allow the application of controlled axial compression and distraction loading at a single level of the New Zealand white rabbit lumbar spine²⁸⁴ (Figure 4E). Also, the Ilizarov-type apparatus was employed, an external fixation device that enables mechanical force application across the IVD²⁹⁰ (Figure 4F). This device not only compresses but also immobilizes the IVD. Compressive loads on the IVD result in axial compression of the AF and bulging of the NP, which radially compresses the AF. Immobilization resulted in decreased disc thickness, axial compliance, and angular laxity, while compression induced these changes earlier and to a more severe extent.^{32, 291} Recently, a novel approach in which sutures were employed to induce compression was developed. The skin was cut along parallel lines, and 2 mm of skin was freed from the tail. Then, 4-0 silk thread was used to suture the skin and subcutaneous tissue via a simple end-to-end suture to create suture-induced compression (Figure 5).²⁹² This model is characterized by dynamic compression generated by tail movement and avoids immobilization-induced disc degeneration, which may cause bias in readout parameters. A compression dog model was established with the pairs of screws planted in lumbar discs symmetrically, then springs attached to these screws exert static compression force on adjacent discs.²⁹³ This model showed early sign of IDD with reduced cellular density and decreased proteoglycan.

Figure 5. Compressive suture-induced rat IDD model. Circumcising the skin around index discs with a width of 2 mm and anastomosing the skin impose axial compression on the tail. Reprinted from Liu et al.²⁹² Copyright ^© 2021 with permission from Elsevier. IDD: intervertebral disc degeneration.

Disc lesions/herniation

The acupuncture model is the most commonly used method for constructing a herniated disc model.^294-308 Following damage to AF integrity, the NP tissue prolapsed during exercise, resulting in morphological changes in IDD and symptoms of nerve root compression. In addition to destroying the integrity of the AF, acupuncture altered the mechanical state of the IVD, resulting in abnormal torsional and compressive biomechanics, leading to mechanical-related degeneration.¹⁰³ The most significant bias for needle puncture results from the relative needle size, depth, segments, and surgical approaches. A puncture that does not penetrate the whole layers of AF mimics the initiation of disc degeneration with AF fissures, with no apparent NP damage.^309-311 In other models, a straight penetration through the AF to the disc center or the contralateral skin can be observed, which leads to NP herniation and degeneration.^312-315 Various needle sizes (18-30G) have been used to induce IDD. The needle gauage and corresponding inner/outer diameter are summarized in Additional Table 2 ^{(316.1KB, pdf)} . Several studies have performed comparative investigations to determine the optimal needle size. van Heeswijk et al.³¹⁶ applied 18G (38% of ovine lumbar disc height) and 25G (15% of lumbar disc height) on the posterolateral annuli of healthy ovines to assess the impact of needle size on the herniation path. The results showed no association between 25G puncture and disc disruption and herniation, while nuclear material migrated through the 18G needle puncture. Accordingly, a larger needle size leads to more significant disc degeneration.^317-321 However, no consensus has been reached regarding the optimal parameters for certain species, leading to incomparable results in different studies. These parameters are summarized in Additional Table 3 ^{(316.1KB, pdf)} .^322-349

In 2008, a systematic review summarized animal studies that treated discs with a needle puncture or sham injection, using the ratio of needle diameter to disc height (diameter:height) as an important parameter to indicate the relative size of the needle or lesion.³⁵⁰ This review concluded that significant disc degeneration, in terms of histological disruption, radiological changes, or mechanical alteration, was not observed with a diameter:height ratio less than 40%. This conclusion is well supported by numerous studies^{103, 351-353} (Additional Table 3 ^{(316.1KB, pdf)} ), although a few studies have revealed alterations in radiography, MRI, and disc biochemicals in diameter:height ratios less than 40%, especially for lumbar disc injury of larger animals (e.g., pigs).^{209, 315, 354-357} The ongoing degeneration of lesions (diameter:height less than 40%) may be attributed to the stronger adjacent muscle strength of lumbar discs, causing excessive mechanical loading on injured discs. Additionally, 30G needle (34% of rat tail disc height) was reported to decrease T2-weight intensity in MRI images without inducing histological changes.³⁵⁸ Furthermore, Keorochana et al.³⁵⁹ reported that a 22G (76% of rat tail disc height) puncture induced significant histological impairment, increasing the histological grade from 2 to 6 weeks, while the grading and proteoglycan stain grading decreased after 8 weeks; this indicated spontaneous repair after injury that should not be underestimated. In contrast, punctures using 18-21G (135-87% of rat tail disc height), 27G (54% of rat lumbar disc height) needles induced a progressive disc degeneration process, with no spontaneous recovery observed after 8-12 weeks.^{320, 360-366} Although diameter:height ratios below 40% results were variable, and some significant effects were observed, disc changes were universal for diameter:height ratios exceeding 40%. Therefore, a diameter of > 40% may be a reliable and safe parameter for establishing a degeneration model of disc lesions.

In addition, approaches to needle puncture significantly influence outcomes. An open puncture was performed using a 2-cm longitudinal skin incision to expose the AF before the annulus puncture. A percutaneous puncture was performed using a radiograph-assisted IVD targeting puncture. Accordingly, the percutaneous injection induced a less severe rat NP degeneration, with no obvious NP herniation, compared with an open injection; this could be attributed to a larger tissue defect around the puncture site that facilitates the extrusion of the disc material.³⁶⁷ Similarly, rabbits with a percutaneous puncture experienced less tissue injury and showed delayed and fewer degenerative outcomes when compared with the disc exposure approach.^{368, 369} In addition, this approach determines potential herniation sites. Most studies have utilized anterior/anterolateral approaches for lumbar disc injury (Additional Table 3 ^{(316.1KB, pdf)} ). In these studies, the herniated disc tissue barely influenced the dorsal root ganglion, which is not typical in clinical cases. Due to the thinner AF layers in the dorsal area, the posterior approach induced more severe disc degeneration and more obvious discogenic pain than other approaches.^{370, 371}

For medium- and large-sized animals, such as dogs and rabbits, laminectomy or facet joint amputation is also needed, apart from disrupting the AF integrity.^{33, 372} As the muscle and tendon tension for large animals is supposedly extensive to support the stability and motion of the vertical spine, removing the adjacent vertebral attachments, like ligaments or facet joints, causes disc instability, increasing the range of motion such as a posterior extension or lateral flexion, which accelerates the NP degeneration and damage of the AF, leading to NP prolapse.

Another method that modulates the amount of herniated NP was developed by amputating the rat tail to collect the NP.³⁷³ Then, the NP tissue was quantified, and a specific amount of tissue was placed on the selected nerve roots. Although this is considered a promising strategy to generate painful radiculopathy, potential dislocation of implanted NP tissue may lead to unexpected radiculopathy in other segments. Additionally, ethical considerations regarding tail amputation may limit its application in several institutions.

Chemical stimulation

In addition to physical damage, chemical agents that impair the IVD matrix or cellular integrity have been employed to develop IDD. Injection of degradative enzymes, including chymopapain and chondroitinase ABC (ChABC), is a traditional approach for digesting proteoglycan or breaking GAG chains. Chymopapain causes the destruction of proteoglycan and GAG, which help maintain the hydration of NP tissues, leading to decreased intradiscal hydrostatic pressure and altered biomechanical stability.³⁷⁴ On administering an intradiscal injection of chymopapain to Macaca fasciculata, pain-related brain areas, the secondary somatosensory cortex and insular cortex, were provoked 1 and 3 days after chymopapain treatment, suggesting a reliable acute discogenic LBP model.³⁷⁵ As a proteolytic enzyme, chymopapain has been approved to treat herniated discs by degrading aggrecan and decreasing the herniated tissue volume, thus reducing pressure on the spinal cord or nerve root.^{376, 377} However, it was withdrawn owing to serious complications, including anaphylaxis, paraplegia, and subarachnoid hemorrhage.³⁷⁸

ChABC is another digestive enzyme that cleaves the protein core of proteoglycans with GAG side chains. Proteoglycans determine the hydration level of the IVD tissue as well as mediate interactions between collagen and other connective tissues.³⁷⁹ Injecting ChABC into the lumbar IVD of goats decreased disc height, reduced T2 and T1p, aggravated collagen disruption, downregulated matrix content, and altered disc dynamic viscoelastic mechanics.^380-382 Interestingly, although ChABC digestion decreased GAG content, the released GAG fragments, such as hyaluronic acid fragments, have been reported to promote an anabolic effect on surrounding cells.³⁸³ Meanwhile, using ChABC during cartilage development increased the tensile properties of this tissue.^{384, 385} Thus, ChABC is beneficial for matrix-rich tissue regeneration. Moreover, ChABC increased the elastic toughness and total shear energy of the high-density collagen (HDC) gel-AF interface by 88% and 46%, respectively, and enhanced the adhesion of the HDC gel to the AF without significantly decreasing native AF cell viability.³⁸⁶ However, spontaneous repair may occur after digestive enzymes are degraded, indicating that the degenerative process may be short.

Diabetic mellitus is reportedly associated with IDD.^{387, 388} In a mouse model, diabetes was found to induce pathological changes in the structure and composition of IVDs and vertebrae.³⁸⁹ Advanced glycation end-products (AGEs), a heterogeneous group of molecules, are major biochemical benchmarks in diabetes mellitus.³⁹⁰ Chronically elevated blood glucose levels lead to AGE formation in patients with diabetes, with AGE localization in IVDs.³⁹¹ Accumulation of AGEs in IVD results in disruption of IVD cell metabolism, senescence, death, and matrix destruction.^392-395 Intradiscal injection of AGEs (200 μg/mL, 2 μL) into rat tails resulted in significantly low disc height and decreased T2-weight intensity, increased apoptosis, and a disturbed IVD structure after 4 and 8 weeks.^{396, 397} Furthermore, chronic exposure to dietary AGEs in mice resulted in age-accelerated IVD degeneration and vertebral alterations involving ectopic calcification.³⁹³

Complete Freund’s adjuvant (CFA) is a water in oil mixture of Mycobacterium tuberculosis that promotes inflammatory reaction at the local injection site. A rat disc degeneration model was developed to serve as a discogenic pain model by injecting CFA into discs.³⁹⁸ By provoking an inflammatory reaction, CFA injection resulted in significant dehydration of NP and blurred boundaries between the NP and AF.³⁹⁹ Moreover, CFA injection leads to a sustained increase in pain-related neurotransmitters and receptor expression, resulting in prolonged central sensitization for 8-10 weeks.^398-400

Fibronectin is elevated and frequently presents as fragments, corresponding to the degenerative process of IDD. In IDD, most fragments were large (around 240 kDa), with relatively few fragments presenting molecular weights less than 40 kDa.⁴⁰¹ Fibronectin fragments inhibit matrix synthesis and upregulate the synthesis of some metalloproteases, leading to decreased tissue integrity.^{402, 403} These fibronectin fragments serve as damage-associated molecular patterns to activate toll-like receptors and promote the proinflammatory phenotype to induce disc degeneration.⁴⁰⁴ Injection of 30 kDa N-terminal fibronectin fragments into the central region can induce a progressive, degenerative process resembling degenerative disc disease over 16 weeks, supported by histological changes where the NP region appeared to shrink rapidly, replaced by fibrous tissue.^{405, 406}

The inflammatory microenvironment is a primary characteristic of disc degeneration. Proinflammatory cytokines secreted by disc cells and infiltrated immune cells, such as IL-1β, TNF-α, IL-6, and IL-12, impair matrix synthesis, cell survival, and disc integrity.^{407, 408} Local administration of IL-1β, the predominant cytokine involved in the pathological process of IDD, accelerated the disc degeneration process, demonstrating rapid NP herniation, a robust immune response, and neuropathic pain in less than 3 weeks of the intradiscal injection.⁴⁰⁹

Autoimmunity against proteoglycan and aggrecan can induce spondylitis and erosive polyarthritis in BALB/c mice.^{153, 410-412} Systemic immunity to proteoglycan and aggrecan resulted in early mononuclear cell infiltration in the ligamentous tissue and entheses adjacent to the annulus. Later, structural disruption and bone erosion were observed in association with mononuclear cell infiltration.^{413, 414} However, this systemic immunity could induce other joint diseases and enthesitis, which may cause unexpected bias in the readout parameters.¹⁵³

Others

Fusion

Spinal fusion immobilizes the adjacent IVD at the index level, which is often applied to immobilize spines and decompress the impaired discs after removing herniated disc tissues. However, spinal fusion usually alters the biomechanics of adjacent discs. For example, the adjacent segment (L3-4) after spinal fusion (L4-S1) in patients revealed significantly increased intradiscal pressure by 0.29 (0.13, 0.47) MPa.⁴¹⁵ During surgery, manual contouring of the spinal rods is often required for proper rod alignment within the pedicle screw heads, while dedicated reduction devices correct residual misalignments.⁴¹⁶ However, pulling forces up to 1.0 kN were required to correct the induced misalignments. Accordingly, the adjacent facet, discs, and vertebrae experienced abnormally asymmetrical high forces, leading to progressive disc degeneration and vertebral disease.⁴¹⁷ Patients who underwent anterior cervical instrumented fusion experienced severe adjacent-level ossification and showed significantly increased osteophyte growth and decreased disc height with a minimum 2-year follow-up.⁴¹⁸ Therefore, an animal model of lumbar fusion was established to simulate the adjacent disc dysfunction.^{419, 420} For example, the rabbit model underwent spinal fusion at the proximal (L4-5) and caudal (L7-S1) levels. After 3 months, IVDs adjacent to the fusion showed a loss of parallel collagen bundle arrangement within the annular lamellae. Furthermore, the disc structure was wholly replaced by disorganized fibrous tissue, with annular tears observed after 9 months.^{419, 421, 422}

Loss of nutrient supply

In the NP and AF, nutrient supply is primarily achieved by solute transport from the endplate structure, which largely relies on the blood supply of adjacent vertebrae.^{145, 423} IVD allograft transplantation requires revascularization of the endplate and AF to reestablish the nutrient supply, while insufficient nutrient diffusion could lead to transplantation failure.⁴²⁴ Endplate perforation on pig lumbar discs was performed, which led to a reactive response in the early phase, including infiltration of inflammatory cells, fibroblast-like cell dominance, and reactive bone formation around the drill canal.^425-427 However, endplate perforation also leads to the disruption of NP and/or AF tissues by drill puncture, and numerous blood vessels were found to grow into NP tissues through the drill hole.⁴²⁶ Therefore, endplate perforation may not be an ideal model to mimic nutrient restriction. Ossification of the endplate is a crucial factor that leads to restricted nutrient diffusion. Destruction of the blood supply to adjacent vertebrae is another approach to induce nutrient restriction. After injection of ethanol and cement into the vertebral body, bone sclerosis was found to develop in the endplate, and NP cells gradually changed from predominantly vacuolar cells to chondrogenic cells and eventually fibrocartilaginous cells, presenting NP fibrosis and AF rupture.^{428, 429} Although a nutrient restriction model has been established, the interruption by either perforation, ethanol, or cement injection does not represent age-related nutrient insufficiency in disc tissues and appears irreversible.²²² Therefore, appropriate regeneration strategies may not be suitable for investigation in these models.

Nerve and vessel ingrowth

The outer AF structure is disrupted during IDD, probably allowing the inappropriate entry of nerves and blood vessels, ultimately inducing pain.^{430, 431} In addition to structural destruction, increased levels of inflammatory mediators, neurotrophins, and angiogenic factors induce nerve ingrowth.^{139, 432} Furthermore, matrix disorganization, for example, aggrecan, inhibits nerve fiber growth.^{137, 433, 434} Melrose et al.⁴³⁵ developed an AF defect sheep model, demonstrating increased blood vessel and nerve ingrowth and infiltration of cells through the original defect 3-12 months post-surgery, primarily associated with proteoglycan depletion. Bioactive molecules also contribute to nerve ingrowth. Puncture-induced mouse AF defects resulted in increased nerve growth factor expression, elevated from 2 weeks and maintained at 12 weeks post-surgery.⁴³⁶ Interestingly, to control the leakage of nucleus content, Xin et al.³³⁶ established a nerve ingrowth model that induced annular injury using a 16G needle puncture, followed by sealing with poly (lactic-co-glycolic acid)/fibrin gel, promoting detrimental nerve and blood vessel growth into deeper regions of the injured disc.

Intervertebral Disc Biomaterials

Annulus fibrosus regeneration

During IDD, reduced intradiscal hydrostatic pressure poses an excessive loading burden on AF tissues, including elevated compression, torsion, and shear force.^{90, 437} These aberrant biomechanical characteristics lead to irreversible structural damage such as fissures or small tears that may progress over time to larger defects, allowing nuclear content leakage to develop disc herniation.^{438, 439} Surgical interventions mainly focus on the removal of herniated tissues and decompression of intradiscal pressure.^{440, 441} However, untreated AF defects may cause unwanted reherniation and imbalanced mechanical properties.¹⁰⁴ Moreover, patients with annular defects of at least 6 mm in width experience reherniation and necessitate reoperation at more than twice the rate of those with smaller annular defects.⁴⁴² Therefore, the current biomaterial strategies have been developed to prevent AF destruction and regenerate tissues to maintain lamellar integrity and alignment at certain levels.^{25, 443} The AF defect model induced by needle puncture or incision is mainly employed to evaluate the regenerative effects of these biomaterials.

Biomaterials designed for AF regeneration should consider the following criteria: (1) repair of axial, torsional, and viscoelastic motion segment responses (e.g., regenerated AF can withstand multidirectional and multitype motion with minimal risk of re-disruption); (2) repair of mechanical strength and the corresponding herniation risk (e.g., reconstruction of AF can repair abnormal mechanics and maintain tissue integrity under anisotropic force); (3) durability of the repair strategy (e.g., AF repair materials can seal local defects for an extended period, allowing native tissue regeneration).

Mesh

Typically, a physical block of AF fissures is a traditional approach to maintain the outer lamellar integrity and prevent progressive nuclear content herniation. Numerous closure and repair systems have been used in clinical trials.

The Barricaid^® annular closure device consists of a woven polyethylene terephthalate, flexible fabric component attached to a titanium alloy (Ti-6Al-4 V ELI) intravertebral bone anchor, designed to treat large AF defects (4-6 mm tall and 6-10 mm wide).⁴⁴⁴ This device received U.S. Food and Drug Administration (FDA) premarket approval in 2019, and its repair effects have been evaluated in several clinical trials and comparative clinical studies.^445-447 A systematic review of four controlled studies (801 patients with large AF defects) has reported that the risks of symptomatic reherniation and reoperation were approximately 50% lower in patients who received additional treatment with the Barricaid^® device after a 2-year follow-up period.⁴⁴⁷ However, the Barricaid^® device is unsuitable for treating AF defects less than 6 mm in width, as the reherniation risk in these patients is relatively low, and this device does not encourage tissue regeneration to prevent progressive AF tearing.⁴⁴² Although the 2-year follow-up data showed favorable outcomes, long-term comparative data are warranted to comprehensively evaluate the clinical applications. A 5-year follow-up study was recently completed,⁴⁴⁸ but data remain unpublished.

The Anulex-Xclose^TM device comprises tension band(s), each with two tissue anchors placed on either side of the annular defect on the AF surface, allowing repair of the defect opening in a single band or multiple band pattern.⁴⁴⁹ This device received FDA clearance in 2009.⁴⁴⁷ A 2-year follow-up clinical trial has reported no significant benefit in reducing reherniation, although no increased adverse effects were reported with the use of the Xclose^TM device.⁴⁵⁰

These suture repair systems require additional damage to AF tissues or adjacent vertebrae to fix the systems on the outer layer of the AF rings. Moreover, they do not integrate with AF lamella and fail to promote disc ECM regeneration and cellular biofunction, which could explain the limited success.

Hydrogels

Although some closure devices have focused on reducing the reherniation rate post-surgery, most patients still complained of residual leg pain (around 70%).⁴⁵¹ The could be attributed to the fractured AF lamellae structure, which fails to provide a physical barrier against nerve ingrowth.^{430, 431} Therefore, it remains crucial to develop materials that can integrate with the AF tissue and restore the native ECM architecture to an intact state.

Adhesive hydrogels have gained popularity in IVD tissue regeneration, depending on their processability, injectability, water retention, and cell-laden capability.⁴⁰ Hydrogels injected or placed in lamella surface defects are exposed to asymmetric axial and sagittal forces that expel the implanted hydrogel from the lesion.^{452, 453} Therefore, the adhesive property is a critical characteristic that helps hydrogels maintain their location and integrate with surrounding tissues, which should be seriously considered for establishing reparative AF biomaterials. However, minimal attention has been paid to properties that influence the adhesion and interface between the index tissue and adhesive hydrogel system when fabricating AF hydrogels.¹⁸⁰ When considering the ideal integration of implanted hydrogels, the optimized adhesion to the tissue of interest plays a critical role. Three mechanisms of interactive adhesion should be emphasized for proper selection when developing AF hydrogels. (1) Physical interaction: This occurs when the interface of tissue and adhesive hydrogels forms a key-lock structure, and a topological match is achieved, which usually requires rough topography on both surfaces.⁴⁵⁴ (2) Electrostatic interaction: atoms distributed on the surface of both interfaces show different eletronegativities, and therefore generate an electrostatic state where one surface performs positive charge and the other performs negative charge. This electrical attraction contributes to the interaction between charged surfaces.⁴⁵⁵ (3) Chemical interaction: an intrinsic concentration difference of chemicals or polymers between the adhesive hydrogels and the corresponding tissue surface forces the initiation of chemical diffusion, thus leading to hydrogel integration;^{456, 457} non-covalent binding, including hydrogen bonding or van der Waals forces, is another common chemical interaction that anchors hydrogels on tissue surfaces or lesions, also referred to as physisorption, which generates covalent bonds that form strong linkages among chemical functional groups and results in various types of covalent bonds (including imine bonds, amide bonds, urea bonds, and N−N bonds resulting in hydrazine derivatives, bonds arising from Michael addition, and disulfide bridges arising from thiol oxidation).^{454, 458-460} Strong adhesive properties, mostly dependent on selected components, are essential for AF hydrogels to remain adherent to defects. Natural, synthetic, or complex polymers containing both of these two component types have been fabricated into adhesive hydrogels. However, the lack of a standardized protocol for adhesive measurement renders comparison among AF hydrogels from different studies nearly impossible.

Although several adhesive hydrogels have been developed to permeate AF lesions, few have undergone rigorous mechanical testing to assess the biomechanical compatibility between the hydrogel and native AF tissues.²⁸ Discs sealed using genipin-crosslinked fibrin hydrogel matched the torque range and stiffness of intact discs, with restored the stress-relaxation parameters, including the effective hydraulic permeability.^{461, 462} Another study revealed that genipin-crosslinked fibrin hydrogel fully restored compressive stiffness to intact levels, validating organ culture findings, with only partial restoration of tensile and torsional stiffness obtained.⁴⁵² The addition of polymer materials is a feasible approach to enhance mechanical performance, while biocompatibility should be considered. Moreover, a genipin-crosslinked decellularized AF hydrogel showed a compressive modulus similar to that of native AF tissues and alleviated the continuous loss of the NP tissue during IDD progression.⁴⁶³ The additive poly (D,L-lactide-co-glycolide) improved mechanical outcomes of genipin-crosslinked fibrin hydrogel, including adhesive strength (∼5 to 35 kPa), shear moduli (∼10 to 110 kPa), and compressive moduli (∼25 to 150 kPa), overlapping with native AF properties; however, tensile moduli (∼300 kPa) were still five times lower than the native AF circumferential tensile moduli (∼30 MPa).⁴⁶⁴ The combination of hydrogels and other scaffolds is another feasible strategy. Biocomposite laminates composed of long collagen fibers in unidirectional and angle-plied ±30° orientations, embedded in alginate hydrogel, can duplicate the entire stress-strain mechanical behavior of the AF lamellae in the longitudinal and circumferential directions, allowing comparison between in vitro results and human AF literature data.⁴⁶⁵ To our knowledge, no existing material completely possesses all biomechanical parameters of native AF. Determining which mechanical property should be considered most critical for material behavior outcomes of in vivo regeneration remains controversial. Moreover, a comprehensive review has summarized parameters of human IVD motion segment stiffness, apparent modulus values, and strain measurements, which should be matched for better material selection and performance outcomes.²⁸

Scaffolds

Hydrogel-based matrices usually lead to unorganized tissue regeneration.⁴⁵² An organized structure that mimics the aligned distribution of AF lamellae plays an important role in delivering guidance cues for biomaterial integration with surrounding tissues.^{90, 466-468} It is critical to regenerate the organized AF structure to appropriately distribute the push-out force transduced from the inner NP to prevent reherniation. Scaffolds that can be conveniently modified to establish ideal microstructures have been developed to achieve structural reconstruction and mechanical repair.

Various techniques have been developed to fabricate AF scaffolds, including silk fiber winding, freezing, and collagen gel contraction.^469-472 In an attempt to simulate the collagen fiber alignment and related multilamellar AF structural hierarchy, a silk winding machine was used to wrap silk fibers around the central hydrogel, such that the fibers lay parallel to each other at an angle of approximately 30° to the vertical axis in a single layer, oriented at alternating angles in successive layers. This three-dimensional (3D) silk fibroin scaffold allowed cells to align along the fibers and produced an oriented cartilaginous matrix.⁴⁷⁰ A directional freezing technique, with a pre-designed polydimethylsiloxane mold, was adopted to prepare lamellar scaffolds encircled in alternate directions to mimic a disc-like angle-ply construction. This scaffold not only promoted stem cell proliferation, oriented matrix deposition, and differentiation to an AF-specific phenotype but also matched the compressive modulus of native AF tissues.⁴⁷¹

Electrospinning is a controllable technique for generating ideal structural characteristics according to the reparative needs.^{466, 473, 474} The polymer solution present in a syringe is attracted by the voltage between the syringe and collector, shrinking along its trajectory to form a mass of fibers on the metallic collector.^{475, 476} By controlling the molecular weight, concentration, and viscosity of the solution, as well as the flow rate, voltage, distance between the needle and collector, and environmental conditions, the electrospun fibers can be modified into microfibers or nanofibers.⁴⁷³ The sequential rotation of the collector should be aligned with fiber distribution, and its morphology should be carefully designed.^{477, 478} Electrospun poly L-lactic acid fibrous scaffolds mimic key structural features (fiber size and alignment) of native AF tissue, promoting differentiation of AF-derived stem cells to a specific AF cell phenotype, which provides a solid basis for designing novel strategies for improved AF repair and regeneration using the physical cues of scaffolds.⁴⁷⁹ Aligned nanoyarn scaffolds generated from gelatin/poly(L-lactide-co-caprolactone) solution showed tensile properties similar to the native AF tissue and substantially promoted the expression of AF-associated ECM (type I collagen) when compared with hybrid scaffolds with randomly organized nanofibrous.⁴⁸⁰ Compared with the disorganized and scarcer fibrous tissue in a randomly organized control fibrous scaffold, electrospun-aligned polycaprolactone formed a biomimetic multilayer fibrous scaffold integrated with the surrounding tissue and homogeneously aligned collagen fiber organization within each lamella after implantation into ovine box AF defects.⁴⁶⁷

Decellularization tissue matrix (DTM) removes the material’s immunogenicity and maintains the matrix composition, microscopic nanostructure, and biological properties of the native tissue to the greatest extent.^{481, 482} DTM materials can promote cell proliferation, differentiation, and migration through various mechanisms, including micro-nanostructure, cytokines, matrix-bound nanovesicles, and peptides produced during preparation.⁴⁸³ For example, decellularized tissue sheets were gently dried with tissue paper to identify a clearly defined collagen fiber-preferred/aligned direction and then cut out into squares. The fiber-aligned direction of each square was then oriented ±30° (verified via a protractor) relative to a stationary grid containing a common horizontal axis. The established angle-ply multi-laminate AF repair patch demonstrated structural and mechanical properties comparable with those of native human AF tissue.⁴⁸⁴ The combination of a decellularized matrix with synthetic polymers promoted both mechanical and biological repair. Decellularized AF was conjugated with poly(ether carbonate urethane) urea using coaxial electrospinning technology and revealed appropriate mechanical properties and significant promotion of ECM secretion.⁴⁸⁵ Additional in vivo data regarding the degradation, adhesive properties, and the time and extent of its integration with surrounding tissues are required to further evaluate whether the implanted AF would allow for timely annulus regeneration before degradation.

The microarchitecture pore size influences implanted cell morphology, cellular adhesion, and distribution of cellular skeleton, which predominantly affects cellular interaction, migration, growth, stem cell differentiation, and inflammatory phenotype.^486-488 Larger pore sizes, providing a large survival room, good nutrient supply, and metabolite discharge, encourage cell proliferation;^489-492 smaller pore sizes better mimic intimate cellular interactions and 3D cell aggregation, which favor chondrogenesis and matrix secretion.^493-496 For example, an AF biomimetic structure with a pore size of 343.0 ± 88.25 μm could provide an ideal scaffold for adipose-derived stem cell proliferation.⁴⁹¹ Polylactic-co-glycolic acid (PLGA) scaffolds were fabricated by solvent casting/salt-leaching with pore sizes of 90-180, 180-250, 250-355, and 355-425 μm; among these, pore sizes of 90-250 μm showed better effects on ECM production.⁴⁹⁶ Inflammatory cells, including macrophages, monocytes, and adaptive cells, reportedly infiltrate into the IVD through AF defects, among which macrophages play a critical role in initiating local inflammation.^497-499 Macrophages should be appropriately modified to avoid excessive proinflammatory phenotypes that interfere with the regenerative outcomes of implanted biomaterials.¹¹⁶ The properties of porous scaffolds significantly influence macrophage activation. Specifically, compared with 34 μm pore size, the 160 μm ± 12% pore size can better promote macrophages toward an anti-inflammatory phenotype (M2-type).^{500, 501} If the pore size was increased from 0.3 to 1.5 mm, the release of proinflammatory cytokines from macrophages was significantly decreased when cells were implanted on either alginate, glass, or polystyrene.⁵⁰² Smaller pores allow for greater interaction between cells and the pore wall, encouraging macrophages to recognize antigenic epitopes, limiting M2-type polarization, while larger pores would reduce the interaction and favor the M2 phenotype. However, excessive pores may cause scaffold collapse and premature degradation.⁵⁰⁰ Pore shape determines cellular morphology, which significantly alters M1 and M2 gene expression profiles.⁵⁰³ Specifically, box-shaped scaffolds elongate cell phenotypes to trigger murine polarization into M2-like macrophages, while restricted cellular elongation leads to significant M1-like polarization.^503-505 Therefore, both pore size and shape should be cautiously designed to modify cell fate and favor tissue integration and inflammatory response after a scaffold has been implanted at the repair site.

Scaffolds mainly fabricated for void filling on AF defects rarely achieve in vivo regeneration; this could be attributed to their instability after implantation into the defect area, considering the relatively lower adhesive strength when compared with hydrogels and delayed integration with surrounding tissues. Moreover, an additional process may be needed during surgery, as scaffolds need to be polished to match the size and shape of the AF defect. Moreover, scaffolds usually require an additional patch to anchor them at the site of implantation.⁴⁶⁷ A combination of scaffold and hydrogel may be a feasible approach to enhance the adhesive properties of materials, facilitate nutrition/cell infiltration, and achieve the ideal mechanical repair.

Nucleus pulposus regeneration

IDD is closely related to NP cell degeneration and a decrease in the ECM, which leads to structural changes and functional defects of the spine.⁵⁰⁶ The unfavorable microenvironment in the degenerated IVD can cause autophagy, mitochondrial dysfunction, and even programmed or non-programmed death in NP cells.^{260, 507} Dysfunction of NP cells results in an inability to effectively maintain the content and structure of the ECM. Decreased NP anabolism directly leads to the loss of ECM integrity, including reduced proteoglycan and collagen and the substitution of GAGs.⁵⁰⁸ Typically, the ECM of NP has a high charge and proteoglycan concentration. A well-functioning proteoglycan should possess an abundance of sulfated GAG chains and hyaluronic acid, which can form highly hydrophilic aggregates with aggrecan.⁴² However, during IDD, proteoglycans in the ECM of the NP are gradually replaced by truncated proteoglycan molecules (formed by the aggregation of fewer and short chondroitin sulfate chains).^{508, 509} Accumulation of the truncated proteoglycan form weakens its ability to bind to hyaluronic acid and impact its distribution, leading to ECM fragmentation.⁵¹⁰ Given the two major pathophysiological processes (cell degeneration and imbalanced ECM metabolism), many tissue engineering materials have successfully restored NP tissue hydration and alleviated the ongoing IDD by repairing cell functions and introducing natural proteoglycan and/or artificial replacement components. Similar to the evaluation of AF regeneration, the currently applied biomaterials for NP regeneration are primarily assessed at single or specific IVD levels. Therefore, animal models based on disc lesions are the most frequently applied models.

Nucleus pulposus device for clinical trials

Many NP engineering materials have been employed in clinical trials. For these materials, the design concept involves in situ hydration. The objective is to simulate the hydration properties of the NP tissue to restore its water content, which facilitates disc distraction and allows the disc to cushion the load during weight-bearing activities, as well as restores the disc height for spinal motion. The prosthetic disc nucleus, PDN^® (Raymedica, Inc., Minneapolis, MN, USA), a material comprised of a woven polyethylene jacket encasing a copolymer hydrogel composed of hydrophobic polyacrylonitrile and hydrophilic polyacrylamide, reportedly allows the hydrogel to absorb surrounding fluid, with the jacket constraining its swelling.⁵¹¹ DiscMaxx HydroGel^TM (Replication Medical, Inc., Monmouth Junction, NJ, USA), also known as Gelstix^TM, is composed of hydrolyzed polyacrylonitrile that absorbs the surrounding fluid to restore NP hydration and biomechanical properties.^{512, 513} BioDisc^TM, composed of albumin and glutaraldehyde, is injectable through a dual syringe delivery system, with polymerization occurring during the delivery process.⁵¹⁴ NuCore^® injectable nucleus (Spine Wave, Shelton, CT, USA), comprising a sequential block copolymer of silk and elastin components, is first mixed with a crosslinker and immediately injected into the NP through the AF defect, allowing polymerization at the surrounding tissue. In particular, this technique seals the AF defect while filling the NP void and prevents gel extrusion.⁵¹⁵ Novocart^® Disc (TETEC, Reutlingen, Germany) is composed of two main components: (1) a suspension with 3.6 to 4.4 million autologous IVD cells contained in a solution of modified human albumin, human serum, culture media components, chondroitin sulfate, and hyaluronic acid; (2) a bis thio-polyethylene glycol solution. The mentioned NP devices for clinical trials have achieved their design purpose of maintaining NP hydration and restoring disc height and mobility.^{511, 515, 516} Moreover, Novocart^® Disc improves the bioactivity of the NP tissue and increases ECM deposits by introducing autologous cells, exerting anti-inflammatory, anti-angiogenic, and anti-osteogenic properties.^{517, 518}

However, for swelling materials such as PDN^® and Gelstix^TM, it is difficult to ensure that the swollen device ideally matches the NP lesion. An oversized device may impose extensive pressure on the endplates or adjacent facets, contributing to the ongoing degeneration.^{25, 513} However, without a restraining device, fragments of Gelstix^TM hydrogels were found to dislocate and compress the spinal root.⁵¹³ Therefore, the safety and efficacy of existing commercial devices should be further assessed with more extensive trials and a larger sample size.

Hydrogels

Several bioactive components have been fabricated to produce hydrogels for NP regeneration, including natural materials (collagen, hyaluronic acid, fibrin, gelatin, alginate, and chitosan),^{116, 519-523} synthetic materials such as poly(ethylene glycol) diglycidyl ethers, polyvinyl pyrrolidone, poly(ethylene glycol) dimethacrylate, poly(ethylene argininylaspartate diglyceride),^524-526 and biosynthetic materials that combine the bioactive properties of natural components with the characteristics of synthetic polymers, including cross-linking, strength, and easy modification.⁵²⁷ Owing to the native gel-like structure of NP tissues, hydrogels are the most commonly fabricated bioactive materials for NP regeneration. Hydrogels have become popular materials in the field of IVD tissue regeneration, dependent on their processability, injectability, water retention, and cell-laden capability.^{40, 528} A fine needle puncture can be used to inject the material into the NP tissue. Compared with cutting a window on AF and placing the graft into the intradiscal area, the injectable material causes little damage to the AF tissue. Therefore, it can be ideally employed for treating early NP degeneration.

The injectable material can be transformed from liquid to gel or solid-like material after injection into the NP tissue via chemical cross-linking, temperature, and pH control. For example, a fully synthetic, thermoresponsive poly(glycerol monomethacrylate)-poly(2-hydroxypropyl methacrylate) diblock copolymer worm gel, mimicking the structure of hydrophilic GAGs, can form highly anisotropic worms at 21°C to create an ECM network that can differentiate stem cells into the NP phenotype, despite the addition of growth factors (TGF-β3 or GDF-6).⁵²⁹ Cross-linking agents can inhibit protein degradation and improve the mechanical properties of the material, to meet the requirements of IVD repair.⁵³⁰ Genipin, derived from geniposide following hydrolysis by β-glucosidase, is a commonly used natural biological cross-linking agent.^{530, 531} The addition of genipin to the decellularized matrix hydrogel or a collagen/hyaluronic acid hydrogel can quickly promote the transformation of the material from a liquid to a gel state, significantly increasing the mechanical properties of the material.^{463, 532-534} Furthermore, genipin cross-linking can reduce the enzymatic degradation rate of the material, thereby ensuring sufficient time for regeneration of surrounding tissues and significantly increasing the water content and disc height.⁵³⁵ Interestingly, the genipin concentration controls the release of cytokines such as TGF-β3.⁵³¹ Hydrogels directly loaded with cytokines usually experience burst release kinetics driven by the largest gradients during the initial stage.⁵³⁶ Genipin cross-linking efficiently reduced the porosity and mesh size to limit the release of encapsulated cytokines. Gelation occurred in approximately 20 minutes without an initiator; laminin-111 functionalized poly(ethylene glycol) hydrogel achieved ideal gelation within 20 minutes with the temperature adjusted to 37°C and pH adjusted to 7.4. This hydrogel maintained the cellular viability of NP cells in the IVD explants.⁵³⁷ Chitosan/hyaluronic acid hydrogels conjugated with kartogenin, a chondrogenic and chondroprotective reagent, can achieve continuous kartogenin release for days, promoting cell proliferation and ECM deposition (type II collagen, aggrecan).⁵³⁸ Hydrogels that mimic the natural NP microenvironment for exogenous and resident cells allow convenient intradiscal delivery and serve as an ideal carrier for bioactive reagents and cell delivery.

Cell delivery microspheres

The application of microspheres in in vitro cell expansion, with or without external stimulation, and as an efficient delivery system periodically releasing drugs or bioactive reagents have been well-established.⁵³⁹ Microspheres with a large specific surface area facilitate nutrient infiltration and release of internal reagents. Many commercial microspheres are available to facilitate biomaterial fabrication, while alginate, chitosan, silk, and gelatin are selective alternatives for developing microspheres.^540-544

Numerous biomaterials based on microspheres have been developed as injectable materials for NP regeneration.^545-548 For example, NP cells encapsulated in decellularized small intestinal submucosal matrix microspheres proliferated and produced ECM components to modify microspheres. Then, following decellularization, the microsphere containing NP ECM was injected into the degenerated rabbit IVDs. Consequently, the T2WI imaging intensity revealed that the IVD water content increased along with the increased disc height.⁵⁴⁹ Adipose-derived stem cell-seeded PLGA microspheres supplemented with dexamethasone and TGF-β3 promoted disc regeneration with evident ECM production, as well as restored NP hydration and disc mobility.⁵⁵⁰ Similarly, biodegradable PLGA microspheres can achieve sustained release of recombinant human GDF-5 for more than 40 days, effectively maintaining recombinant human GDF in the IVD. The slow-release of the active molecule significantly promoted expression levels of type II collagen and aggrecan and restored the height of the degenerated IVD.³⁶¹

Microfluidic technology allows for simultaneous loading of reagents and cells and enables convenient delivery via a minimally invasive approach.^{551, 552} An injectable “peptide-cell-hydrogel” microsphere was constructed using a microfluidic technique with the covalent coupling of active peptide APETx2 and further loading of NP cells, which inhibited local inflammatory cytokine storms, regulating the metabolic balance of the ECM.⁷⁶ The established microsphere system can potentially provide gradients for electrospinning and 3D printing to develop biomaterials with required properties and may even serve as units for fabricating organoids.⁵³⁹

Nano particles

Nanoparticle materials serve as carriers for active biomolecules and therapeutic agents, maintaining molecular activity while achieving controlled drug release, with superior potential for IDD treatment.^{553, 554} Commercial nanoparticle carriers (MaxSuppressor In Vivo RNALANCEr II Kit and Lipid Extruder, BIOO Scientific, Austin, TX, USA) equipped with micro-RNA-141 were injected into the mouse tail IVD and alleviated NP degeneration caused by the puncture.⁵⁵⁵ Albumin/heparin nanoparticles, as a carrier of stromal cell-derived factor-1α, can effectively maintain the activity of stromal cell-derived factor-1α, significantly promote the expression of aggrecan, type II collagen, and other matrix proteins, and improve the integrity of the NP structure.²⁴ Intradiscal injection of chitosan/poly-(γ-glutamic acid) nanoparticles with an anti-inflammatory drug (diclofenac) not only reduces the release of inflammatory cytokines from IVD cells but also promotes native matrix production.²³ Heparin/poly(L-lysine) nanoparticles embedded with dexamethasone and growth factor showed continuous release of dexamethasone and TGF-β3/basic fibroblast growth factor to achieve minimal implantation-associated inflammation and promoted stem cell differentiation toward the NP cell phenotype and matrix deposition.^{550, 556} Moreover, targeting strategies can often be achieved by adequately designed nanosized delivery systems through the appropriate selection of nanoparticle types and surface molecules for particular cell targeting.^557-559

Although several nanoparticle-mediated bioactive reagent delivery systems have been developed, maintaining a prolonged local drug presence may be challenging. The monocyte-macrophage system is efficient in removing exogenous particles, especially particles < 10 μm, and nanoparticles (10-200 nm) are also taken up via endocytosis by various cells.^{560, 561} The amount and approach of uptake depend on the cell type and nanoparticle characteristics (e.g., size and surface).^{562, 563} Moreover, the infiltrated vessels during disc degeneration also contribute to the removal of nanoparticles.⁵⁶¹ Chitosan nanoparticles showed shorter retention than their microsphere counterparts.⁵⁶⁴ Nanoparticles embedded in microspheres or hydrogels may withdraw their limited release profiles while maintaining the advantages of both nanoparticles and microspheres or hydrogels.^{565, 566}

Combined regeneration

Various studies have attempted to replace damaged IVDs with integrated disc materials. Instead of simply regenerating the degenerated or injured AF or NP tissues, AF and NP combined strategies allow for the replacement of damaged tissue with an intact and complete functional unit; this could theoretically reduce the progressive degeneration of adjacent facet joints and prevent stenosis and spondylosis. Moreover, AF or NP individual strategies cannot replace the whole IVD, which means that the remaining tissue may still undergo age-related degeneration. IVD replacement with bioactive components with/without cellular delivery reduces the reoccurrence potential of degeneration.

The main challenges in establishing and delivering an IVD implant are as follows: (1) the implanted IVD must be functional to maintain the mobility of segments and allow for proper spinal flexion, torsion, and extension;⁵⁶⁷ (2) the implanted IVD must integrate with adjacent vertebrae mechanically or biologically to remain at the implanted location;^568-570 and (3) the implanted IVD must withstand the complex mechanical loading of the disc space.⁵⁷¹

IVD transplantation from healthy donors has been performed to entirely restore the diseased IVD. Initial attempts to perform IVD transfer in dogs and rhesus monkeys rarely presented satisfactory outcomes, which might be attributed to the poor natural remodeling potential.^{572, 573} Later, fresh-frozen disc allografts requiring dimethyl sulfoxide, which is typically used to maintain cellular viability and metabolic activity during the cell freezing process, appeared to promote bony remodeling with adjacent vertebral bone.^{574, 575} Five patients with cervical disc herniation underwent transplantation of fresh-frozen composite disc allografts after disc excision. Within 3 months, the graft endplates showed good integration with adjacent vertebral bones, and the motion and stability of the spinal unit were preserved after 5 years of follow-up, despite signs of mild age-related degeneration.⁵⁷⁵ Moreover, four cases confirmed the safety and rationality of disk allografting after a follow-up of 10 years.⁵⁷⁶ IVD transplantation requires size matching to ensure graft retention and prevent disc migration. Considerable efforts are required to perform careful pre-surgery preparation to ensure size matching and donor sources.

Notably, several artificial discs have gained FDA approval for lumbar total disc replacement, including ProDisc-L and activL.^{577, 578} Maverick^TM, has completed the investigational device exemption and is awaiting FDA approval.^{579, 580} Favorable outcomes have been witnessed with the clinical application of commercial artificial discs and are summarized in a comprehensive review.⁵⁸¹ A lumbar disc device, named ActivL^®, was developed with an inferior and superior cobalt-chromium plate anchored in the endplate, presenting an ultra-high molecular weight polyethylene insert that may translate 2 mm in the anterior-posterior direction on the inferior endplate.^{578, 582} The 5-year results of a randomized controlled trial with 324 enrolled patients reported that the ActivL^® group presented a significantly better range of motion for flexion-extension rotation, flexion-extension translation, and disc angle than the groups treated with the previous generations (either Pro Disc-L^® or Charité^®).⁵⁸³

However, complications related to the technique, specific device, and approach may lead to transplant failure. Technique-related complications, including facet joint degeneration, endplate mispositioning, and vertebral body split fractures, can attribute to the general features of artificial discs that fail to match expected disc sizes, alter the biomechanical properties, and disturb adjacent vertebral discs for anchoring.^{584, 585} Specific devices, such as Maverick^TM, induced severe persistent pain 1 year after initial implantation, with metallosis around articulation surfaces of the disc prosthesis, which was attributed to the persistent release of metal ions such as chromium and cobalt.^586-588 Metal-on-metal devices may lead to metallic debris and ion release, causing immunoreactions against implanted devices and connective tissue diseases.^{588, 589} Approach complications result from damage to adjacent tissues during surgery, including retrograde ejaculation, ureteral injury, and vascular injury.^{590, 591} Injury to the superior hypogastric plexus and anterior vessels have been reported on employing an approach similar to anterior lumbar interbody fusion, leading to retrograde ejaculation and increased risk of postoperative hematoma.^{592, 593} XL TDR (NuVasive, Inc., San Diego, CA, USA) is a lumbar disc replacement instrument that allows a lateral approach, effectively resulting in pain relief and improved quality of life at mid- to long-term follow-up; this could be attributed to the superiority of the lateral approach over the anterior approach in reducing vessel injury.^{594, 595}

Based on existing complications, several studies have attempted to achieve total disc engineering with or without cellular delivery. Layers of electrospun poly(e-caprolactone) and interspersed poly(ethylene oxide) were combined with cell-seeded hydrogels to form disc-like angle-ply structures. This structure was used to replace the rat disc by employing a stable fixation system. Accordingly, good integration was achieved, as endogenous cells populated the full thickness of the implant and produced a collagenous network.⁵⁹⁶ Decellularization is a promising approach to remove cells while preserving the ECM structure and components. A decellularized whole IVD xenograft was fabricated with proper detergents, ultrasonication, freeze-thaw cycles, and nucleases. Although the native angle-ply collagen microarchitecture and collagen contents were preserved, the linear region moduli, peak stress, and equilibrium moduli were significantly reduced.^{597, 598} The contracted-collagen (AF)/alginate (NP) technique was applied to fabricate a tissue-engineered total disc implant that can adequately withstand mechanical loads, producing an integrated and mechanically functional ECM similar to the native IVD in rat and canine spines;^{472, 599} the implant-endplate interface demonstrated progressive integration with only small discontinuities at boundaries between the endplate and engineered tissue.⁴⁷² Although many bioactive disc replacement devices have been developed and used for in vivo regeneration, long-term reparative strategies are still required to gain a comprehensive understanding of nutrient reestablishment and potential degeneration of the implanted disc.

Acellular and cellular repair strategies

Numerous studies have attempted to modify the local microenvironment and promote native cell function using bioactive hydrogels. Acellular regeneration of the annulus impairs native cellular infiltration and subsequent remodeling of matrix tissues. Collagen gel is a well-known material used in the field of acellular repair. HDC gel has been fabricated to repair annular defects in rat or sheep models generated by annular puncture and NP discectomy.^{600, 601} The HDC gel was maintained in the defects for up to 18 weeks, preserving approximately 70% of the NP tissue.⁶⁰⁰ Moreover, collagen gel appeared to enhance intrinsic healing by attracting host fibroblast cells and forming aligned, fibrous tissue.^{600, 601} However, the defect size and shape could influence the reparative effects of the HDC gel, especially in large animal models. A rectangular defect may better facilitate gel adhesion than a round defect, which is attributed to increased tension.^601-604 Fibrinogen is another material that possesses excellent tissue adhesion, which remains in the injected area with or without membrane anchoring.⁶⁰⁵ An ex vivo test in bovine discs has reported that fibrinogen crosslinked with genipin restored IVD height and compressive properties, as well as partially restored other biomechanical behaviors of IVD motion segments under a range of physiological loads, without herniation.⁴⁵² However, in vivo long-term tests showed equivalent outcomes between fibrinogen and control groups, which may be attributed to the unexpected disruption of the endplate during surgery, limited cell infiltration, and no aligned fiber lamellae formation.⁶⁰⁵

Identifying signals toward which AF cells possess sensitivity could aid in designing biomaterials that facilitate cellular recruitment strategies. IDD-associated inflammation promotes the expression of C-C chemokine ligand receptors (CCRs), such as CCR1, CCR2, and CCR5, in native AF and NP cells. CCRs can bind CCLs to stimulate cell migration and matrix production.^606-608 However, in vivo data revealed that CCL5 might not recruit AF cells to repair the defect area after disc puncture treated with fibrin gel delivery.⁶⁰⁹ In contrast, CCLs have been used to recruit stem cells for tissue regeneration.^{610, 611} A sequential chemokine delivery system that releases CCL-5, TGF-β1, and GDF-5 sequentially uses CCL-5 to recruit stem/progenitor cells and TGF-β1 and GDF-5 to induce the synthesis of a type II collagen- and aggrecan-rich ECM, resulting in promising disc regeneration.⁶¹² As CCRs undergo degeneration-related upregulation for promoting cellular assembly, the application of CCLs may be a promising strategy for designing acellular regeneration, while their in vivo recruiting effects on native AF and NP cells needs further evaluation. Moreover, there are other concerns regarding the use of CCLs. Certain CCLs, especially CCL4, present liver toxicity and have been used to establish a liver fibrosis model.^{613, 614} CCLs possess oncologic potential, including cell proliferation, drug resistance, migration, invasion, and organ-specific metastasis of tumor cells.^615-617 As a proinflammatory cytokine, CCL also showed chemo-attraction toward macrophages, which may aggravate disc inflammation.^{618, 619}

Supplementary bioactive molecules are an efficient approach to enhance native cell recruitment. Cytokines, TGF-β3, BMP-2, BMP-3, insulin-like growth factor 1, and osteogenic protein-1, showed positive regulatory effects on the amelioration of disc anabolism.^620-624 Sustained TGF-β1 release induced an anabolic stimulus in AF cells while mimicking the 3D ECM environment of the AF tissue.⁶²⁵ Other reagents also modify the function of AF cells to enhance tissue self-repair. However, it should be noted that TGF-β1 and BMPs in regenerative therapy can lead to the generation of osteophytes at the repair site, resulting in an exacerbated spinal pathology.^{626, 627} Hyaluronan oligosaccharides upregulated AF proMMP-2 and MMP-9 and downregulated MMP-13, ADAMTS1, ADAMTS4, ADAMTS5, aggrecan, and type II collagen; simultaneously, hyaluronan oligosaccharides promoted the upregulation of MMP-1, MMP-13, and ADAMTS1 and the anabolic matrix repair genes aggrecan, type I collagen, and type II collagen in the NP.^{625, 628} These modulations are expected to promote clearance of granulation/scar tissues from AF defects, as well as matrix replenishment. The in vivo test revealed that AF defect sites contained enlarged annular lamellae in response to the hyaluronan oligosaccharides, consistent with an active repair response.

Delivery of exogenous cells is an attractive approach to effectively control the number, type, quality, and genetic or chemical modification of delivered cells. Comprehensive reviews have discussed cell-based strategies for IVD repair.^{25, 26, 629} Although several cell delivery systems have achieved ideal regenerative outcomes for IVD repair, no significant difference has been observed between the cell-loaded and no-cell-loaded systems; this could be attributed to inappropriate selection of loading substances and limited nutrition at the repair site. Following IDD development, an unfavorable microenvironment, including oxidative stress and the release of inflammatory factors, aggravates cell death and matrix decline, causing excessive cellular autophagy, apoptosis, and necrosis of stem/progenitor cells, thereby limiting the repair efficiency.^{630, 631} Accordingly, modifying the local microenvironment is required to promote the reparative potential of implanted cells in vivo. Additionally, primary stem cell sources for clinical treatment are autologous sources, such as bone marrow stromal cells or adipose stem cells extracted from the patient’s bone marrow or adipose tissue.⁶³² However, these sources may encounter limited donor tissues, with high economic cost and time for in vitro culture, potential cellular infection by pathogens, and additional invasive operations in donors.^{633, 634} The donor age and disease status, in vitro cell preservation, and cell processing during surgery may affect clinical outcomes after auto-transplantation. Moreover, xenogeneic or allogeneic stem cells may result in unexpected host-versus-graft reactions.⁶³⁵ In addition, reparative biomaterial systems with cellular delivery may delay their clinical translation.

Systemic regeneration

Systemic administration of therapeutic biomaterials that targeting IVD, including oral, intravenous and intraperitoneal administration, potentially achieve the regeneration of general IDD related to aging. A new functionalized nanofullerene conjugated with a peptide that binds specifically to a formyl peptide receptor-1 expressed on activated macrophages was developed and denoted FPR-1 targeted C60 nanoparticle (FT-C60). By preferentially binding to formyl peptide receptor-1, FT-C60 significantly attenuated the mRNA expression of proinflammatory cytokines, which are critical components for inflammation and discogenic pain. Furthermore, FT-C60 alleviated pain in a mouse model of lumbar radiculopathy established by puncture-induced nucleus protrusion toward the L5 nerve root following abdominal administration. The systemic application of FT-C60 showed targeting properties to the local injury site.⁶³⁶

Systemic administration of nanoparticles, such as alginate oligosaccharides, can be employed as a feasible strategy to reduce clinical complications (infection, prolonged pain) after lumbar fusion surgery.⁶³⁷

PLGA nanoparticles are widely applied carriers for delivering reagents. A previous study has evaluated the safety of a PLGA nanoparticle delivery system by assessing superoxide dismutase and catalase in healthy dogs. A typical complement activation-related pseudoallergy was observed, widely known to be associated with nanoparticle-based drug delivery, including a combination of bradycardia, hypotension, hypersalivation, pale gums, and involuntary urination, within 7 days, while no long-term clinical signs and pathologies were recorded.⁶³⁸

However, systemic regeneration for IDD is still debating, based on the fact that IVD is an avascular organ with less chance of local assembling of systemic administrated biomaterials.⁶³⁹ To solve the problem, robust IVD tissue-specific markers, proper vascularization and well designed targeting biomaterials should be determined and established. Then, the systemic and early intervention of IDD may be achieved with no need of invasive operations.

Characteristics of Intervertebral Disc Biomaterials

Clinical manifestations

LBP is often described as pain, muscle tension, or stiffness in the body region below the costal margin and above the inferior gluteal fold, with or without limb pain.^640-642 Generally, diseases affecting the anatomical structure around the lumbar spine, including vertebrae, ligaments, muscles, facet joints, and IVD, can lead to LBP, among which IDD is the leading cause contributing to 40% of LBP cases.^{2, 258} Pain originating from degenerated IVD is referred to as discogenic pain.⁶⁴³

Evaluation of pain, which is often neglected, is necessary to comprehensively demonstrate the regenerative effects of IVD biomaterials. A systematic review has summarized different pain types, including neuropathic/nociceptive pain, acute/chronic pain, evoked/spontaneous pain, and hyperalgesia/allodynia.⁶⁴⁴ Numerous behavioral assays have been performed on rodent models to evaluate pain associated with LVD degeneration (Additional Tables 4 ^{(316.1KB, pdf)} and 5 ^{(316.1KB, pdf)} ). As the plantar surface of the rodent hind paw is primarily innervated by the tibial nerve, which is composed of spinal nerve roots from L4-S2, pain sensitivity detected from the hind paw can represent a measurement of LBP.645 In addition, LBP related disc discomfort indicated by movement-evoked hypersensity and spontaneous painful bahavior is also an important sign of LBP. Therefore, the evaluation of rodent pain sensitivity is mainly composed of stimuli-evoked hyperalgesia on mainly on hind paw, movement-evoked hyperalgesia and spontaneous tests.

Stimuli-evoked hyperalgesia evaluation

Evoked tests involve external stimuli to initiate pain and evaluate responses to controlled stimulation, including mechanical and thermal stimulation (Additional Table 4 ^{(316.1KB, pdf)} ). Mechanical pain is usually based on the von Frey assay, which applies mechanical force by calibrated microfilaments.^646-648 When force is applied to the hind paws, a positive response is defined as the brisk withdrawal of the probed foot. The up and down” method provides a solution to calculate the 50% mechanical withdrawal threshold by repeatedly adjusting the mechanical degree of the probing filaments.^{649, 650} An algometer used to measure tenderness is a reliable approach for evaluating mechanical hyperalgesia. This device is valuable and can be applied to various locations, including the tail, hind paw, and spine. Indeed, it has been employed to assess pressure pain thresholds in clinical settings.^{651, 652} Thus, algometer data from animal models are comparable with clinical situations. Thermal pain, either with hot or cold stimulation, can be measured to determine pain sensitivity.^{30, 314, 653-662} On applying heat or cold, the time when rodents first display avoidance reactions such as foot withdrawal, paw lifting, or jumping is recorded as paw withdrawal latency. Interestingly, mice with progressive disc degeneration present normal sensitivity to mechanical stimuli applied to the hind paw, with hypersensitivity to cold stimuli applied to the hind paw.^{30, 255, 256, 663}

Movement-evoked hyperalgesia evaluation

In disc degeneration, the tolerance to axial stretch is assumed to be decreased, indicating a bending hyperalgesia in human.^{30, 256} Grip Force assay is based on a Grip Strength Meter (Stoelting Co., Wood Dale, IL, USA) that can record the stretching force performed on the metal bar.³⁰ Initially, it was designed to determine the neuromuscular function of animals.^664-666 This assay involves stretching a mouse tail back with the mouse forelimb grasping the wire gauzes and recording the maximal force when the mouse releases the gauze. Decreased maximal force is an indication of impaired tolerance to axial stretch and discomfort^{30, 667} (Additional Table 5 ^{(316.1KB, pdf)} ). Notably, the tolerance is not attenuated in all degenerative model. For example, a spontaneous degenerative model, SPARC-null mice, showed significantly dimished grip force, while needle puncture-induced degeneration model revealed no obvious alteration.^{30, 256, 657} The efficacy of grip force in other degeneration model shall be further evaluated.

Tail suspension is another assay to measure axial discomfort.²⁵⁵ Specifically, mice are individually suspended by the tail underneath a platform with adhesive tape. The duration of time spent in a) immobility (not moving but stretched out), b) rearing (trying to reach the underside of the platform), c) full extension (actively reaching for the floor), and d) self-supported (holding either the base of its tail or the tape), is analyzed by a digital software over the entire testing period around 180 seconds. Deceased immobility is indicative of axial discomfort.⁶⁶³ This assay is often used as a traditional measure of depression in mice.⁶⁶⁸ It has been proven to reliably measure signs of axial pain in mice.^{256, 663}

Besides, FlexMaze is designed to measure lateral flexion–induced discomfort. It force the mice to undergo lateral flexion in order to explore the maze. Then, the total distance and velocity covered by the mice indicate the movement-evoked hyperalgesia with other variables controlled.²⁵⁶

Spontaneous behaviors

The measurement of spontaneous pain does not require external stimulation. Instead, certain inconspicuous behaviors can be recorded to reveal signs of pain. Unilateral injury or related pain models have shown weight-bearing differences in paws, which can be detected by an expert investigator.⁶⁶⁹ Furthermore, if the animal experiences discomfort, the explored area is reduced. Therefore, an open-field test was used to record the exploratory behavior for assessing spontaneous pain.^{670, 671} When rodents experience pain, they demonstrate high-frequency ultrasonic vocalization, like screaming in humans.⁶⁷² Ultrasonic vocalization is often accompanied by audible vocalization, which challenges the usefulness of measuring ultrasonic vocalization.⁶⁷³ Finally, analysis based on animal expression and altered movements is another strategy to assess spontaneous pain. The Mouse Grimace Scale was initially established to evaluate spontaneous pain based on mouse facial expression utilizing five features (orbital tightening, nose bulge, cheek bulge, ear position, and whisker change).⁶⁷⁴ Similarly, the Rat Grimace Scale, based on four units: orbital tightening, nose/cheek flattening, ear changes, and whisker change, was employed to analyze pain according to rat expression⁶⁷⁵ (Figure 6). The reliability of these scales has been verified in numerous spontaneous pain models by experienced observers.^{674, 675} However, this approach may not be highly reliable, as judgments regarding spontaneous expression and movement are subjective and affected by the observers’ proficiency.⁶⁷⁶ Moreover, baseline scores significantly differed between men and women, whereas live scores were significantly lower than retrospective scores from images.⁶⁷⁷

Figure 6. Scales for spontaneous pain evaluation based on rodent facial expressions. Copyright ^©2011 Sotocinal et al.⁶⁷⁵ Reprinted from BioMed Central Ltd.

Biocompatibility and degradation

Biocompatibility is a characteristic of materials that determines cellular, tissue, and organ responses to the respective materials.⁶⁷⁸ Biocompatibility is mainly determined by the nature and application of the material. Material properties, including shape, size, surface roughness, residual toxic low-molecular substances during material polymerization or preparation, material processing pollution, and material degradation products, are associated with biocompatibility.^{679, 680} On considering a material for application in the biomedical field, biocompatibility is an important indicator that needs to be considered and evaluated.

Material components possess critical characteristics that may lead to cellular toxicity and host immunological responses. For example, in the DTM, residual cellular components after decellularization include high mobility group box 1 (HMGB1), DNA, and gal antigen epitopes. HMGB1, an intracellular protein that binds DNA, is one of the most common damage-associated molecular patterns following cell rupture.⁶⁸¹ The current decellularization protocols cannot comprehensively eliminate HMGB1, and the residual content depends on the tissue source, decellularization strategies, and application of cross-linking agents.⁶⁸² HMGB1 induces inflammation by promoting the release of inflammatory factors such as CCL2 and CCL4 and activating proinflammatory signaling pathways, such as the toll-like receptor pathway.⁶⁸³ Apart from the irritating inflammatory response, HMGB1 is reportedly chemotactic and promotes the proliferation of bone marrow mesenchymal stem cells and keratinocytes.^{684, 685} Therefore, damage-associated molecular patterns not only activate resident inflammation but also induce cell aggregation and proliferation, consequently affecting the graft-versus-host reaction and overall repair.

Reportedly, α-(1,3) epitopes of α-1, 3-galactosyl transferase are abundant on the cellular membrane in almost all species, except humans.⁶⁸⁶ However, large amounts of the oligosaccharide galactose-alpha-1,3-galactose (α-Gal) antibodies exist in the human circulatory system.⁶⁸⁷ Therefore, the immune response triggered by α-Gal epitopes is a major concern after transplantation of tissue-derived biomaterials. Porcine AF tissue was treated by freeze-thawing in liquid nitrogen, incubated in a hypotonic buffer at 37°C for 24 hours, and decellularized in 0.1% sodium dodecyl sulfate, 0.1% ethylenediamine tetraacetic acid, and 10 KIU/mL aprotinin, resulting in the removal of nearly 80% of α-Gal epitopes, with good immuno-compatibility and a decrease in mononuclear cells after implantation on AF defects at 14 days.⁶⁸⁸ Galactosidase removes 60-75% of α-Gal epitope residues following tendon decellularization, resulting in excellent histocompatibility after subcutaneous implantation of decellularized tendon.^{689, 690}

Un-α-Gal antibody refers to antibodies that target epitopes other than α-Gal. In humans, they may bind with non-homologous proteins in the ECM polypeptide sequences. If these antibodies bind to the xenograft, graft rejection is induced by activating the complete cascade.⁶⁸⁵ For example, allograft transplantation of the heart and kidney with knocked out α-Gal epitopes resulted in xenograft rejection induced by un-α-Gal antibodies at 6 months and 1 month, respectively.⁶⁹¹ The remaining major histocompatibility complex antigen after decellularization presents another group of un-α-Gal antigen epitopes.⁶⁸⁶ Major histocompatibility complex staining is essential for evaluating the decellularization efficacy when fabricating materials with low immunogenicity.⁶⁹² The presence of residual major histocompatibility complex I/II antigens in implanted materials can lead to macrophage infiltration.^{693, 694} Furthermore, an excessive humoral reaction induced by the un-α-Gal antigen epitope in DTM may block adhesion sites and inhibit adhesion and host cell infiltration.⁶⁸⁷ Hence, the presence of un-α-Gal antibodies may induce graft rejection or hinder the interaction between cells and DTM, leading to failed repair.⁶⁹⁵

Following decellularization, DNA fragments released from cells tend to adhere to the surface of decellularized matrix owing to its adhesiveness. Resident or infiltrating macrophages derived from monocytes recognize DNA fragments and phagocytose them.⁶⁸⁹ This procedure is a primary step in innate immunity for the removal of damaged cells and tissues. Excessive DNA residues result in a relative lack of DNase in macrophages, leading to the accumulation of DNA in cells, which may activate the nuclear factor-κB pathway, resulting in inflammatory cell aggregation and aggravating local inflammation.^{696, 697} Notably, if the tissue was contaminated during processing, the residual DNA may be derived from viruses or prions, which can cause severe consequences by infecting host cells after transplantation.

Products released during material degradation should promote local tissue regeneration and must nontoxic. Several polymers utilized in IVD regeneration are reportedly degradable. By hydrolytic scission of the ester bonds, polylactic acid results in monomeric lactic acid, which can worsen the low pH of the disc microenvironment.⁶⁹⁸ The persistently acidic microenvironment leads to increased cell glycolysis and overexpression of the acid-sensing ion channel family.^{64, 155, 699, 700} Sequential energetic exhaustion and calcium overload induction by acid-sensing ion channel results in cellular dysfunction and death.^{155, 699} Multiple studies have unraveled a shift from M1-type to M2-type within 1 to 2 weeks of implanting biological scaffolds composed of mammalian ECM.^{687, 701, 702} ECM degradation appears to be necessary for the transition from M1 to M2 phenotypes, suggesting that decomposition products may be essential for this transition. Hydrogels composed of ECM biological scaffolds can promote the transformation to the M2-type when compared with materials lacking ECM.⁵⁰⁴ Several studies have reported that DTM scaffolds demonstrate an adequate effect of promoting the polarization of infiltrating macrophages toward the M2-type.^{504, 505, 527}

One barrier that hinders the development of biomaterials with idea-degrading properties is the discrepancy between in vitro and in vivo degradation profiles. A typical approach to establish an in vitro degradation profile can be evaluated by immersing biomaterials in phosphate-buffered saline with or without enzymes or detergents.^{703, 704} Nevertheless, the in vitro condition is completely differs from in vivo situation. After implantation, biomaterials interact with surrounding tissue fluid, and infiltrated cells play critical roles in modifying biomaterials and mediating their degradation. Considerable attention has been paid to optimizing the regulatory effects of biomaterials on cells and tissues. Furthermore, data on how cellular responses and tissue reactions influence the degradation and other modification processes of biomaterials are lacking.^{705, 706} The host response to biomaterials involves sophisticated biochemical changes that modulate biomaterial erosion. A typical response to biomaterials is the formation of a fibrous capsule that may interfere with tissue fluid changes and influence the degradation rate.⁷⁰⁷ Therefore, the in vivo evaluation of degradation is relatively more valuable when considering biomaterials for tissue regeneration. As the subcutaneous immunological response is considered timely and intensive, biomaterials are subcutaneously implanted to evaluate their biocompatibility and degradation.^708-711 After certain days or weeks, histological images are obtained to evaluate the amount of residual biomaterials.⁷⁰⁴ However, compared with a non-vascularized structure such as IVD, skin tissues possess an abundance of vessels and lymph nodes to provide effortless and rapid tissue fluid changes, which may exaggerate the degradation rate of materials used for IVD regeneration.^{712, 713} Moreover, a subcutaneously intensive immunological response leads to the formation of a fibrous capsule after material implantation and may impede material degradation.^{714, 715} Other studies have attempted to evaluate degradation by histological analysis of tissues with implanted biomaterials. Hematoxylin-eosin staining showed retention of implanted decellularized spinal cord ECM in spinal cord lesions after 8 weeks.⁷¹⁶ Another study recorded time-dependent changes in hydrogel thickness in the rat abdominal wall using histological images.⁷¹⁷ Intradiscal implantation is necessary for accurately measuring the degradation properties of these materials.

However, the analysis based on histological images is semi-quantitative and largely depends on the obtained section, which can sometimes greatly influence tissue and implanted material morphology. A fluorescent dye may be a feasible approach to track hydrogel retention time at repair sites. RGD-biotin encapsulated with the fluorescent dye cyanine 5.5 was used to track the location of the hydrogel implanted in the kidney injury site.⁷¹⁸ By employing epitope as a labeling agent, the degradation of elastin-like polypeptide gels can be monitored through enzyme-linked immunosorbent assay, estimating the amount of leached-out antibodies.⁷¹⁹ With a real-time imaging system, the degradation of the hydrogel can be easily recorded and analyzed.⁷¹⁸ However, the fluorescence from conventional fluorophores quickly decays and requires ultraviolet or visible light as the excitation source, which fails to sufficiently penetrate deep biological tissues.^{720, 721} Lanthanide-doped upconverting nanoparticles may be a feasible alternative, as they can convert near-infrared (NIR) to ultraviolet or visible-NIR light via a sequential multiphoton absorption process referred to as upconversion. As NIR light can penetrate up to a few centimeters inside tissues, upconverting nanoparticles allow the tracking of gel degradation with photoluminescence spectroscopy and NIR imaging after implantation in discs.⁷²²

Imaging

Imaging technologies are useful for clinical IDD diagnosis and surgical planning.⁷²³ For preclinical research, imaging technologies are reliable methods for identifying animal disc degeneration and the degree of degeneration.

Radiographic technology, based on X-ray or computed tomography discography, clearly illustrates the vertebrae but does not depict soft tissues clearly, requiring contrast to detect intradiscal abnormalities. With a clear illustration of the vertebrae, an image intensity adjustment technique was developed to precisely evaluate the relative disc height.⁷²⁴ Vertebral body height and disc height were measured along different axes (denoted as A, B, C…, I) using an image analysis program. IVD height was expressed as the disc height index.^{312, 313, 725} The disc height index is a calculated relative index comparable among different studies, regardless of the differences in background and intensity (Figure 7).

Figure 7. Measurements and calculations of vertebral body height and IVD height based on radiographs. The IVD height should be quantified using a relative value, %DHI, which measures changes in the DHI of punctured discs. DHI = 2 × (D + E + F) /(A + B + C + G + H + I); %DHI = post-punctured DHI /pre-punctured DHI × 100. DHI: disc height index; IVD: intervertebral disc.³¹³ .

MRI is a typical diagnostic method and a reliable surgical indication for IDD.⁷²⁶ Currently, conventional T1WI-, T2WI-, and proton density-weighted vessel wall MRI have been successfully employed in the clinical work-up of patients with LBP and suspected (advanced) IDD.⁷²⁷ Proteoglycans are critical IVD components to maintain tissue fluid, and their downregulation occurs concurrently with the loss of hydration, particularly that of the NP.⁷²⁸ The Pfirrmann grading system based on the mid-sagittal plane of T2WI is widely applied for IDD grading.⁷²⁹ Pfirrmann provides a grade from I to V, according to structural morphology (homogeneity within NP, signal intensity, disc height) illustrated on T2WI⁷²⁹ (Additional Table 6 ^{(316.1KB, pdf)} ). The Pfirrmann grading system requires subjective evaluation of T2WI and requires independent and skilled observers to perform the grading process.^{730, 731} Water content alteration and ECM degradation usually occur before the appearance of detectable morphological MRI findings.^{728, 732} Therefore, changes in T2WI and Pfirrmann grading are usually apparent in advanced IDD.^{733, 734} Additionally, T2WI- and proton density-weighted vessel wall MRI provides limited information regarding the mechanical functioning of the disc or the cause of degeneration and pain.^735-737 Moreover, the Pfirrmann grading system combined with IVD micro-nano structural changes more comprehensively reflects the extent of disc degeneration, including GAG assay, histological analysis, evaluation of bony endplates by scanning electron microscopy, atomic force microscopy imaging, and nano-mechanical testing.⁷³⁸

Advanced MRI techniques sensitive to proton-matrix interaction (proteoglycans-bound water), matrix-organization, and water diffusion, rather than water content only, could provide more meaningful findings to identify early IDD.^739-741 Quantitative T2* axial maps based on quantitative T2* MRI provide information regarding the interaction of water within the macromolecular network.⁷⁴² The combination of surface volumes and quantitative T2* axial maps provides insight into the initial degeneration stages.⁷⁴³ T1rho is an MRI relaxation time parameter and a promising MRI contrast for imaging proteoglycan-rich nucleus regions, demonstrating a superior correlation with proteoglycans than T2.^{744, 745} Interestingly, T1rho values in the NP were found to be strongly associated with GAG content and mechanical properties, including swelling pressure.⁷⁴⁶ Diffusion-tensor imaging is another MRI contrast that enables quantification of the water apparent diffusion constant (ADC), which indicates the integrity of IVD tissues.^{747, 748} ADC values from diffusion-tensor imaging of IVD can distinguish time-dependent fluid changes and regional fluid-flow directional shifts.^748-750 Additionally, ADC identified by diffusion-weighted imaging was found to be effective in diagnosing early IDD with tiny variations.⁷⁵¹ A comparative study showed that the T1rho nucleus value was superior to T2 and ADC, correlating with GAG content, histological degeneration, and disc mechanical properties.⁷⁵² Chemical exchange saturation transfer is a method to directly detect exchangeable solute protons in tissues by constant irradiation and saturation of their chemically shifted magnetization, based on the MRI protocol of a T2W sequence.⁷⁵³ GAGs chemical exchange saturation transfer, using the Spin-Lock technique (chemical exchange saturation transfer) and WASABI (WaterSaturation and B1) methods for B0 and B1 field inhomogeneity correction, generating a color-coded GAGs chemical exchange saturation transfer map with high GAG content in blue and low GAG content in red in the lumbar spine, correlating with either LBP or radiculopathy.⁷⁵⁴

Apart from the indication of disc degeneration, videography also enables non-invasive monitoring of implanted components. A study has attempted to monitor mesenchymal stem cell survival after implantation into an IVD with surgically induced degeneration. Compared with MRI, the positron emission tomography reporter probe, 9-(4-[18F]-fluoro-3-hydroxymethylbutyl)-guanine, was more sensitive and identified the longest survival data at 3 weeks after implantation.⁷⁵⁵ However, the cost-benefit ratio of positron emission tomography is relatively low. Cells labeled with ferumoxides (Endorem^®) or protamine sulfate (USPIONs) appear as hypointense regions on MR images and demonstrate significant signal intensity loss and contrast on T2*-weighted images; thus, they can be distinguished from the surroundings.^{756, 757} Radiopaque zirconia nanoparticles enable long-term non-invasive assessment of the implanted hydrogel, as well as scaffold performance and distribution, without impairing the viability and biofunction of encapsulated cells.^{758, 759}

Histology

Histological evaluation is the most convincing for disc degeneration, assessing factors such as the annulus integrity, endplate disruption, fiber alignment, nucleus arrangement, and tissue components. In addition, histological images can reliably indicate the penetration of nerves, vessels, and inflammatory cells. However, the subjective observations performed by individual researchers were incomparable between studies and species. Therefore, although several grading scales have been developed, no consensus has been reached. Herein, we list the most commonly used grading scales.

Nomura et al.⁷⁶⁰ developed a grading system based on the classification of AF established by Nishimura and Mochida.⁷⁶¹ The grading system contained grades of only NP and AF tissues (Additional Table 7 ^{(316.1KB, pdf)} ). This method is relatively precise and straightforward. However, the classification is limited to two primary IVD components and neglects the degenerative morphology of the endplate and adjacent vertebral body.^{334, 362}

Masuda et al.⁷²⁵ developed a more detailed grading scale that separates the grades of NP cellular components and the matrix and emphasizes the border between the NP and AF^{317, 725, 762, 763} (Additional Table 8 ^{(316.1KB, pdf)} ).

Han et al.³¹³ upgraded the grading scale of Masuda et al. by grading cellular components of the AF tissue (Additional Table 9 ^{(316.1KB, pdf)} ). Clear interpretation and easy grading systems have been reported by several studies for histological evaluation.^764-766

Thompson et al.⁷⁶⁷ provided a comprehensive grading system widely employed for histological grading of human discs, distributing equal weights to the nucleus, annulus, endplates, and vertebral body (Additional Table 10 ^{(316.1KB, pdf)} ). Both intra-observer and inter-observer agreement values in the initial Thompson grading scheme were 85%,⁸⁰ and these levels were maintained when the scheme was applied across species.^{79, 80}

The grading scales described by Boos et al.⁷⁶⁸ are more detailed, containing 23 items, including the evaluation of the IVD and vertebral endplate region. It allows for a comprehensive evaluation of these criteria across the entire IVD, with no provision for separate evaluation of the AF or NP regions. All grades are clearly claimed and explained (Additional Table 11 ^{(316.1KB, pdf)} ). Interrater reliability estimates for the assessment of histologic features generally showed good to excellent rater agreement. The agreement of each detailed content exceeded 80%. As a comprehensive grading system and detailed explanation are provided, several studies evaluating disc degeneration have utilized this grading system.⁷⁶⁹ However, it is not highly practical, and considerable effort is required to evaluate histological images. Boyd et al.²³² extracted 11 criteria from those described by Boos et al.⁷⁶⁸ and formed a new grading system^{232, 770} (Additional Table 12 ^{(316.1KB, pdf)} ). Each criterion was graded from 0-4. However, the definition of each grade is missing, which may restrict the reliability and repeatability.

Although many grading scales have been established, these scales have been cited recently. Nevertheless, the established grading system barely considers nerve ingrowth into the disc, which signifies the ongoing degenerative process and persistent discogenic pain.^{136, 278, 771} We recommend that the grading system should meet the following characteristics: 1) Comprehensiveness: the grading system should evaluate the degenerative degree of all substructures of IVDs and symptom-related structures, such as innervation; 2) Simplicity: easy and quick evaluation that requires less effort will contribute to the consistency of results, and help save manpower and material resources; 3) Repeatability: the agreement rate should be evaluated by independent and proficient or less proficient observers to ensure the repeatability of a certain grading system among researches and species; 4) Universality: the grading system should be suitable for different species to facilitate the comparison of biomaterials used in various species. Although the Boos et al.⁷⁶⁸ grading scale is currently the most comprehensive and detailed system for histological evaluation with good repeatability, grading systems with less comprehensiveness, such as Han et al.³¹³ and Masuda et al.⁷²⁵ are still widely employed.^{762, 764-766} Adequate simplification and inclusion of innervation for the Boos et al.⁷⁶⁸ grading scale are required to facilitate reliable histological evaluation and a robust experimental design.

Adhesive

The adhesive property is a critical characteristic for biomaterials to be maintained at the implantation site and integrate with adjacent tissues, especially for AF regeneration. However, studies that fabricate AF biomaterials often pay less attention to adhesive evaluation,^{463, 772} which may partially be attributed to a lack of guidance documents providing a systematic framework for screening the biomechanical performance of newly established AF materials. A comprehensive review has summarized approaches for adhesive properties.¹⁸⁰ Here, we updated the methods and provided detailed protocols to facilitate the experimental design and selection of relevant testing configurations^{117, 773-776} (Additional Table 13 ^{(316.1KB, pdf)} ).

Mechanical evaluation

The spine is a critical weight-bearing organ that can withstand gravity generated by the upper body. Therefore, biomaterials designed for IVD regeneration must satisfy mechanical demands to support tissue regeneration and spine motion. NP materials that cannot withstand hydrostatic pressure within the disc are likely to cause excessive stress on the surrounding AF and endplate, leading to progressive disc degeneration.⁵¹³ Furthermore, mechanical restoration is a tremendous challenge for AF reparative materials, as AF suffers from an asymmetric “push-out” force transduced from the NP tissue and an axial force from endplates. Therefore, additional studies should be performed to satisfy the mechanical requirements of AF regeneration. In addition, standards of robust experimental design and comparative mechanical outcomes should be met among studies. Many reviews have compared the mechanical properties among different species, and suggested the parameters for IVD regeneration.^{203, 777} Here, we adapted a paradigm, including series of screening tests, has been illustrated to rapidly evaluate if the materials meet required mechanical properties^{28, 778-780} (Additional Table 14 ^{(316.1KB, pdf)} ). Correspondingly, the recommended parameters for mechanical properties after biomaterials implantation have been summarized as a benchmark for disc regeneration²⁸ (Additional Table 15 ^{(316.1KB, pdf)} ), which has been acknowledged in many research studies.^{28, 453, 462, 475} Whereas, many factors, including the testing environment, machine No. and loading rate etc. can cause bias to the obtained mechanical results. The same specimen tested in three institutions (two in UK, one in USA) with different testing machines by the same protocol came out with ∼35% difference in compressive stiffness, after normalizing for disc geometry and adjusting for system compliance.⁷⁸¹ Therefore, as it is difficult to unify the testing environment and machine in different researches, a unified control group, such as health bovine/sheep lumbar discs, should be selected to enable the comparison of the mechanical reparative efficiency among different biomaterials.

A real-time monitoring system containing two fluorophore particle probes (blue or NIR emitting) can remotely, with minimal invasiveness, measure the stain exhibited by load-supporting gels. The compressive deformation ratio of gels is equal to the photoluminescence intensity from the blue-emitting probe, while the deformation ratio with a transition from compression to tension is linear to the normalized ratio of the photoluminescence intensity for the blue and NIR probes.⁷⁸² This system may be an important tool for remotely monitoring the compression and tension alterations in the implanted gel in a real-time manner.

Conclusion & Perspectives

Although several animal models have been established, there is a lack of consensus regarding the selection of specific models for evaluating therapeutic strategies. The spontaneous degenerative model from non-human primates, especially the aging model, is probably the most satisfying model that resembles both the components and mechanical characteristics of human disc degeneration.^{200, 206} However, this model is hindered by ethical restrictions, time consumption, and financial burden. Instead, sheep may be a reliable species for developing animal models that simulate the natural disappearance of notochord cells in adulthood, with similar disc size and mechanical characteristics to humans.^{170, 185, 219} Furthermore, established methods for pain detection in large animals remain unavailable and need to be developed to assess the pain-relieving effects of implanted biomaterials.

Currently, the most commonly used degenerative models are induced by acute injuries, such as needle puncture and mechanical alteration established by external apparatus.^{32, 136, 290, 435, 783} In contrast, the age-related degenerative process is a long-term condition with gradual changes in the mechanical properties of the disc and microenvironment.^{506, 784} The occurrence of replication-related senescent cells also plays a critical role in the disc degeneration process by producing senescence-associated secretory phenotypes.^{152, 785, 786} However, the acute injury-induced disc degenerative model cannot replicate this situation. Therefore, animal models developed by mechanical alteration in an elderly animal may be a situational simulation of the most common clinical cases (approximately 40%), with excessive or acute labor increasing intradiscal pressure on age-related degenerative discs, resulting in disc herniation and nerve root irritation.^{787, 788}

Most repair materials that achieved ideal reparative effects were applied at the initial stage of the degenerative model.^{463, 467, 758, 759, 789} The time required for biomaterial implantation approximates the time required to establish a degenerative model. Therefore, in these cases, disc degeneration is prevented but not reversed; this does not match the clinical situation, where disc degeneration is usually established before patients visit the clinic. Therefore, searching for better treatment options is more practical than focusing on the prevention of IDD.⁵⁴⁶ Furthermore, when evaluating the reparative effects, many studies have failed to address pain relief, which is the main reason why patients approach a physician. Future studies should evaluate the pain sensitivity of degenerative models before and after treatment to assess the pain-relieving effects of biomaterials.

The surgical approach is another challenge that needs to be highlighted. Due to the lordotic alignment of disc spaces, AF fissures often occur at the posterolateral site during the range of motion. The fissure is near the spinal cord and dorsal root ganglion, which largely influences the surgical approach to implant repair materials.³⁷⁰ However, annulus defects of lumbar discs are often developed by anterior puncture or incision, and the injury and herniation sites are not likely to irritate the dorsal root ganglion, which is not relevant to the clinical case in terms of disc degeneration and herniation.^{337, 601, 790}

Scaffolds and AF defects must be reshaped and polished to allow for sturdy implantation. Therefore, additional surgery is required, and an enlarged AF damage may occur during surgery. Scaffolds or hydrogels that lack suitable adhesive properties are unstable, and dislocation of these materials will cause further damage to surrounding tissues along with nerve irritation. Moreover, hydrogels fabricated for AF regeneration do not provide cues for organized tissue repair, and the resulting tissue usually lacks an ideal organization matching the original tissue; this may alter the mechanical features of AF and increase the rate of reherniation.^{452, 531, 791}

All human discs undergo age-related degeneration to different extents.^{792, 793} Disc herniation often occurs in lumbar discs of certain levels (L2-4) but usually not in all discs.⁷⁹³ Does herniation in adjacent discs only occur when herniated segments are treated? Studies have shown that spinal fusion aggravates the degeneration of adjacent discs.^{8, 794, 795} The number of levels that should be treated remains a debatable question among surgeons. Most biomaterials designed for disc regeneration are meant to be employed on local discs but do not focus on overall disc protection. To achieve overall repair, systemic administration with disc targeting properties seems to be more feasible than a comprehensive surgical approach or multi-level injection. Systemic administration of biomaterials, such as nanoparticles with functional domains that can target degenerative disc cell markers or ECM components, may fail to produce an adequate response owing to the limited blood supply of the IVD.^796-799 However, increased vascularization of degenerative discs may facilitate the penetration of systemically administered biomaterials, which helps avoid biomaterial assembly in relatively healthy discs.

Funding Statement

This work was supported by the Major Research Plan of National Natural Science Foundation of China (No. 91649204), the National Key Research and Development Program of China (No. 2016YFC1100100), the National Natural Science Foundation of China (No. 81974352), the Scientific Research Training Program for Young Talents from Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, National Natural Science Foundation of China (No. 82002333), and Zhejiang Provincial Natural Science Foundation of China (No. LQ21H060004). These fundings were not involved in the collection, analysis, or interpretation of data in the study.