Review Article

Stem cells in preclinical spine studiesBrian C. Werner, MD, Xudong Li, MD, PhD, Francis H. Shen, MD*Department of Orthopaedic Surgery, University of Virginia, PO Box 800159, Charlottesville, VA 22908-0159 USA

Received 21 January 2013; revised 5 July 2013; accepted 23 August 2013

Abstract BACKGROUND CONTEXT: The recent identification and characterization of mesenchymalstem cells have introduced a shift in the research focus for future technologies in spinal surgeryto achieve spinal fusion and treat degenerative disc disease. Current and past techniques use allo-graft to replace diseased tissue or rely on host responses to recruit necessary cellular progenitors.Adult stem cells display long-term proliferation, efficient self-renewal, and multipotentdifferentiation.PURPOSE: This review will focus on two important applications of stem cells in spinal surgery:spine fusion and the management of degenerative disc disease.STUDY DESIGN: Review of the literature.METHODS: Relevant preclinical literature regarding stem cell sources, growth factors, scaffolds,and animal models for both osteogenesis and chondrogenesis will be reviewed, with an emphasis onthose studies that focus on spine applications of these technologies.RESULTS: In both osteogenesis and chondrogenesis, adult stem cells derived from bone marrowor adipose show promise in preclinical studies. Various growth factors and scaffolds have also beenshown to enhance the properties and eventual clinical potential of these cells. Although its utility inclinical applications has yet to be proven, gene therapy has also been shown to hold promise in pre-clinical studies.CONCLUSIONS: The future of spine surgery is constantly evolving, and the recent advancementsin stem cell–based technologies for both spine fusion and the treatment of degenerative disc diseaseis promising and indicative that stem cells will undoubtedly play a major role clinically. It is likelythat these stem cells, growth factors, and scaffolds will play a critical role in the future for replacingdiseased tissue in disease processes such as degenerative disc disease and in enhancing host tissueto achieve more reliable spine fusion. ! 2014 Elsevier Inc. All rights reserved.

Keywords: Stem cells; Spine fusion; Degenerative disc disease; Scaffold; Growth factors; Intervertebral disc; Bone morpho-genic protein; Adipose derived stem cells

Introduction

The recent identification and characterization of mesen-chymal stem cells has introduced a shift in the research

focus for future technologies in spinal surgery to achievespinal fusion and treat degenerative disc disease. Currentand past techniques use allograft to replace diseased tissueor rely on host responses to recruit necessary cellular pro-genitors. Adult stem cells display long-term proliferation,efficient self-renewal, and multipotent differentiation. Theyalso can be genetically modified to secrete growth factorsimportant to tissue healing, thereby functioning as implant-able, long-lasting reservoirs for these molecules.

In spinal surgery, stem cells have great potential applica-bility in numerous areas. This review will focus on two im-portant applications: spine fusion and the management ofdegenerative disc disease. The treatment of spinal cord in-jury using stem cells is still experimental and outside thescope of this review. Relevant preclinical literature regard-ing stem cell sources, growth factors, scaffolds, and animal

FDA device/drug status: Investigational (Adult Mesenchymal StemCells; Growth Factors [GDF-5, others]).

Author disclosures: BCW: Nothing to disclose. XL: Nothing to dis-close. FHS: Royalties: Elsevier Publishing (B); Consulting: Synthes Spine(B), DePuy Spine (B); Speaking and/or Teaching Arrangements: DePuySpine (B); Board of Directors: MTF (B, Paid directly to institution); Fel-lowship Support: AO (D, Paid directly to institution).

The disclosure key can be found on the Table of Contents and at www.TheSpineJournalOnline.com.

* Corresponding author. Department of Orthopaedic Surgery, Univer-sity of Virginia, PO Box 800159, Charlottesville, VA 22908-0159, USA.Tel.: (434) 243-0291; fax: (434) 243-0242.

E-mail address: fhs2g@virginia.edu (F.H. Shen)

1529-9430/$ - see front matter ! 2014 Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/j.spinee.2013.08.031

The Spine Journal 14 (2014) 542–551

models for both osteogenesis and chondrogenesis will bereviewed, with an emphasis on those studies that focus onspine applications of these technologies.

Osteogenesis/spine fusion

Overview

Spinal fusion has become a universal procedure since itsintroduction a century ago, with more than 350,000 spinalfusions performed each year in the United States represent-ing a greater than 200% increase over the past decade[1–5]. Despite significant advances in surgical techniquesand instrumentation, up to 40% of fusion surgeries resultin pseudarthrosis, causing significant morbidity and oftena need for reoperation [6–9]. Although failure of spinal fu-sions is multifactorial, osteoporosis, poor bone biology,smoking, diabetes, and decreased autograft cellularity inolder patient populations all likely contribute [10–13]. So-phisticated techniques to improve fusion rates includingmodern internal fixation have greatly reduced the incidenceof symptomatic pseudarthrosis, but a 10% to 15% incidenceis still reported in recent literature [14–16]. Additionally,although autograft remains the gold standard for spinal fu-sions, it has limited availability, often low cellularity, andis associated with up to a 30% donor site morbidity rate[17–22]. These issues have led spine surgeons and re-searchers to investigate alternatives for achieving fusionand further reducing the pseudarthrosis rate, including theuse of mesenchymal stem cells.

For a successful spinal fusion to occur, several vital el-ements are necessary. They include the presence of thebone-forming cell or its precursor, the appropriate biologi-cal signals directing bone synthesis, and a biocompatiblescaffold on which the process can occur. The most criticalof these components is the bone forming cell (osteoblast) orits precursor, the mesenchymal stem cell (MSC), both ofwhich possess the ability to form bone [23].

In vitro: cells, growth factors, and scaffolds

CellsAlthough historical research investigated the use of em-

bryonic stem cells, restrictions and various moral and ethi-cal obligations on their use have led to mesenchymal stemcells as the focus of most stem cell research for spinal fu-sion. Mesenchymal stem cells have generated considerableinterest because of their ability to self-renew and multipo-tential characteristics. Furthermore, MSCs have been iden-tified in a variety of tissues including bone marrow, muscle,periosteum, and adipose tissue [16,24–28]. Most recent andpromising preclinical studies have focused on MSCs frombone marrow and adipose tissue.

Bone marrow is well established as a source of MSCs inlaboratory animals and humans, with confirmed differentia-tion into osteoblasts both in vitro and in vivo. Disadvantages

to the use of bone marrow clinically include the painful har-vest as already described and its relatively limited supply re-quiring lengthy culture expansion.

Mesenchymal stem cells from bone marrow were demon-strated to achieve successful spinal fusionmore than 10 yearsago. In 2001, Cui et al. [29] found that cloned osteoprogeni-tor cells from bone marrow produced a quicker and more ro-bust posterolateral fusion in an athymic rat model comparedwith mixed marrow cells. Over the ensuing decade, numer-ous other studies reinforced these findings. Peterson et al.[30] demonstrated that bone marrow MSCs transfected withbone morphogenic protein-2 (BMP-2) ex vivo successful in-duced posterolateral spinal fusion in an athymic rat model.In a recent study directly comparing a bone marrowMSC al-lograft with autograft, Gupta et al. [31] demonstrated similarfusion rates between the two groups in an ovine posterolat-eral lumbar spine fusion model. Similarly, Miyazaki, et al.[32] compared the effectiveness of human MSCs from bonemarrow with adipose-derived MSCs in a posterolateral fu-sion rat model. The authors found similar fusion rates withboth types of MSCs when transfected with BMP-2.

Adipose tissue provides an additional well-researchedand promising source of MSCs for spine applications(Fig. 1). Adipose tissue is easily procured from patientsand large numbers of MSCs can be obtained from relativelysmall amounts of adipose tissue, in contrast to bone marrow[33]. Shen et al. confirmed the osteoblastic differentiationof rat adipose–derived MSCs when cultured with growthand differentiation factor-5 (GDF-5). [33,34] Subsequentstudies have yielded similar promising results. Hsu et al.[35] found that MSCs from adipose demonstrate potentialas cellular delivery vehicles for recombinant proteins suchas BMP-2 using an athymic rat posterolateral fusion model.Similarly, Miyazaki et al. [32] demonstrated that BMP-2producing human adipose–derived MSCs, created using ad-enoviral gene transfer, induced spinal fusion in athymicrats, demonstrating that adipose-derived MSCs are osteo-genic and enhance spinal fusion as effectively as bone mar-row MSCs. Gimble et al. [36] reported that syngeneic andallogeneic rat adipose–derived MSCs on a tricalcium phos-phate and collagen I scaffold accelerated spinal fusion ina rat model.

Growth factorsThe complex process of bone formation can be affected

by a variety of biologic signals, including mechanical loadsand electromagnetic and chemical factors. The effective-ness of stem cells in promoting spinal fusion is heavilydependent on osteoinductive factors, specifically growthfactors that enhance the osteogenic capability of osteopro-genitors such as MSCs [23]. Although many growth factorshave been investigated for their role in the osteogenicdifferentiation of MSCs, only a select few have been re-searched for their role in inducing spinal fusion with MSCsor through gene therapy with MSCs. Table 1 providesa summary of these growth factors.

543B.C. Werner et al. / The Spine Journal 14 (2014) 542–551

Despite recent controversy regarding negative clinicalside effects, BMP-2 has been the subject of considerable re-search for use as an osteoinductive factor with MSCs forspinal fusion. As early as 2003, researchers realized the po-tential harm in the large doses of BMP required to inducea spinal fusion in humans, which suggested that deliveryof such osteoinductive proteins could be improved. Wanget al. [37] used ex vivo adenoviral gene transfer to createBMP-2–producing bone marrow MSCs, which successfullyproduced an intertransverse fusion in a rat spine model. Pe-terson et al. [30] reported similar results in their experi-ments in which human bone marrow MSCs were infectedwith a BMP-2 containing adenovirus, yielding sufficientbone to fuse the lumbar spine in a rat model. Similar resultswith ex vivo gene therapy to overproduce BMP-2 were re-ported by Miyazaki et al. [32], Hsu et al. [35], and Sheynet al. [38] Fu et al. [39] also recognized the need to reducethe dose of BMP-2 necessary to achieve fusion. These au-thors found that combining low doses of BMP-2 with bonemarrow MSCs achieved this goal in a rabbit fusion model.

BMP-7, also known as osteogenic protein-1, is anotherclinically available growth factor that has undergone consid-erable preclinical evaluation with stem cells. Hidaka et al.[40] assessed the enhancement of spine fusion usingbone marrow stem cells modified by an adenovirus vectorencoding BMP-7 seeded onto an allograft scaffold in a ratmodel. The authors found that the addition of adenovirusBMP-7–modified marrow stem cells significantly enhancedallograft spinal fusion. Zhu et al. [41] found that combinedgene transfer of BMP-2 and BMP-7 using bone marrow stem

cells in a rat model was significantly more effective in induc-ing osteoblastic differentiation and spine fusion than indi-vidual BMP gene transfer. Numerous studies have alsobeen conducted successfully using BMP-7 as an inductiveagent without the use of stem cells, which is not the subjectof this review [42,43].

As an alternative to BMP-2, GDF-5 has also been stud-ied recently for its use as an osteoinductive agent for stemcells in spine fusion. Shen et al. [34] confirmed thatadipose-derived stem cells, when treated with GDF-5, arecapable of adhering to a bioengineered scaffold while re-maining viable and demonstrating the ability to migrate,proliferate, and subsequently undergo osteogenic differenti-ation. Similarly, Zeng and coauthors [33,44] proved in twostudies that GDF-5 promotes the differentiation of rat adi-pose–derived stromal cells into osteogenic lineages.

ScaffoldsThe importance of an effective osteoconductive scaffold

should not be underestimated. Despite the fact that scaf-folds do not contain osteogenic cells or intrinsic biologicsignals, an appropriately designed scaffold can providethe framework for the migration and attachment of osteo-genic cells and a suitable environment for the synthesisof extracellular matrix [23]. Allograft is currently the mostcommonly used osteoconductive scaffold and is typicallycombined with autograft to provide the necessary osteo-genic osteoinductive and osteogenic factors and cells. Un-fortunately, the extensive preparation and treatment thatallograft is subject to leads to diminished biologic and me-chanical properties as well as increased expense, which hasled researchers to investigate other scaffolds for preclinicalstudies of stem cells for spinal fusion.

A scaffold for stem cells should maximize the osteoin-ductive and osteogenic effects of cells delivered to the siteof interest, retain growth factors at that site for optimal timeof release, function as an osteoconductive scaffold for bonein-growth with appropriate sized pores for cellular and vas-cular passage, not compete with or limit bone formation,and limit inflammatory response by biocompatibility [45].Important scaffold properties to consider include porosity,pore size, geometry, and material. Numerous materials

Table 1Growth factors used with mesenchymal stem cells for spine fusion

Growth factor MSC type Model References

BMP-2 Adipose, BM Rat, rabbit [30,32,35,37–39]BMP-6 Adipose, BM Mouse, rabbit [123]BMP-7 BM Rat [40–43]BMP-9 BM Rat [124]GDF-5 Adipose In vitro [34,44]FGF BM Rabbit [125]

BM, bone marrow; BMP, bone morphogenic protein; FGF, fibroblastgrowth factor; GDF, growth and differentiation factor; MSC, mesenchymalstem cell.

Fig. 1. Human adipose derived stem cells (Left) undergo effective osteogenic differentiation (Right) as demonstrated by alizarin red staining in this in vitroexperiment.

544 B.C. Werner et al. / The Spine Journal 14 (2014) 542–551

are currently available for scaffolds, including natural orsynthetic polymers, bioactive materials, and osteophilicmaterials such as ceramics or composites [46].

Type I collagen is a natural polymer and represents thesimplest of scaffolds meeting all described criteria. It hasa long history of use in surgical applications because of itsbiocompatibility, ease of degradation, and known interac-tion with other molecules including proteins. Gabbay et al.[47] found a progressively stimulatory effect on adipose-derived stem cells with regard to osteogenesis when culturedin a three-dimensional collagen I gel compared with a two-dimensional monolayer. Hsu et al. [35] successfully useda type I collagen matrix for their study investigating spinalfusion in an athymic rat model with BMP-2–producingadipose-derived stem cells. Similarly, Miyazaki and coau-thors [32] found that both human adipose–derived and bonemarrow–derived mesenchymal stem cells on a type I colla-gen sponge induced spinal fusion. Other natural polymersincluding matrigel [29] and hydrogels [48] have been inves-tigated. Although these polymers are suitable for carriers ofcells and growth factors, they do not possess the mechanicalstrength of stiffer polymers and thus have been used morefor intervertebral disc applications.

Synthetic polymers including polylactic-co-glycolic acid(PLGA), its derivatives, chitosan, and others have been in-vestigated in preclinical studies as well (Fig. 2). PLGA isosteoconductive, is able to deliver osteoinductive growthfactors, and has established biocompatibility. Lee et al.[49] and Shen et al. [34] both successfully used PLGA intheir studies. Similarly, numerous studies in the past 5 yearshave established the use of chitosan as an effective and

viable scaffold for osteogenic differentiation and deliveryof stem cells in preclinical studies [50–52].

Various biomaterials also have served as suitable scaf-folds in preclinical studies, the most popular of which arecalcium phosphate derivatives including hydroxyapatite(HA) and beta tricalcium phosphate (b-TCP). Both naturaland synthetic forms of calcium phosphate have been devel-oped as materials for bone repair and augmentation. Theyclosely resemble the mineral composition, properties, andmicroarchitecture of human cancellous bone and thus havea high affinity for binding proteins. Commercially availableHA is brittle, carries minimal mechanical strength, and isslowly resorbed in vivo, but has still been the subject ofseveral preclinical studies involving stem cells. Chistoliniet al. [53] studied porous HA scaffolds as carriers of bonemarrow MSCs. The authors found that cell-loaded implantswere stronger than cell-free implants with notable bone for-mation. More recently, Minamide et al. [54] and Seo et al.[55] both reported that bone marrow MSCs on HA scaf-folds with and without growth factors induced posterolat-eral fusion in rat and rabbit models.

Beta tricalcium phosphate is another calcium phosphatebiomaterial that serves as a purely osteoconductive scaffold.It has greater solubility than HA and is rapidly resorbed. Theresorption rate of b-TCP closely matches the course of nor-mal cancellous bone remodeling, making it an ideal candi-date for stem cell–based scaffolds. It has recently beenstudied in rat and ovine models in preclinical studies forthe delivery of both bone marrow– and adipose-derivedMSCs for posterolateral spine fusion with uniformly excel-lent results comparable to that of autograft [31,36,56].

Table 2Animal models of mesenchymal stem cells for spinal fusion

Animal model Mesenchymal stem cell type Type of fusion References

Rat Bone marrow Posterolateral lumbar spine [29]Rabbit Bone marrow Posterolateral and interbody lumbar spine [39,54,123,125]Sheep Bone marrow Posterolateral lumbar spine [31]Rhesus monkey Bone marrow Interbody lumbar spine [126]

Fig. 2. Scaffolds are an integral element in the development of cell-based strategies for spine fusion and disc regeneration. Polylactic-co-glycolic acid(PLGA) microspheres (Left) and a PLGA nanofiber scaffold (Right) are two examples used in spine applications.

545B.C. Werner et al. / The Spine Journal 14 (2014) 542–551

In vivo: animal models

Although in vitro studies lay the important foundation forthe use of stem cells to achieve spine fusion, the transition toin vivo studies is a key step in the eventual translation ofstem cell allografts to clinical use. Various animal modelshave been used to investigate the application of mesenchy-mal stem cells in achieving spine fusion. Tables 2 and 3 sum-marize various animal models for spine fusion using MSCsand gene therapy using MSCs, respectively.

Chondrogenesis/intervertebral disc degeneration

Overview

Intervertebral disc (IVD) degeneration is a complex pro-cess that is a result of morphologic and molecular changeswithin the disc itself compounded by progressive structuralfailure and instability of the vertebral motion segments thatsurround the affected disc [57,58]. The resultant clinicalmanifestations of degenerative disc disease have a huge im-pact on society and the economy, with estimated direct andindirect costs of $50 to $90 billion per year in the UnitedStates alone [57,59].

Traditional management of low back pain, which is oftenattributed to IVD degeneration, has focused on conservativeoptions including lifestyle changes, physical therapy, painmedication, and rehabilitation as well as surgical optionsthat focus on eliminating motion through disc arthroplastyor arthrodesis [60,61]. The clinical results of these proce-dures are often suboptimal. Both nonsurgical and surgicaltherapies do not deal with the inherent loss of functionalnative disc tissue and therefore fail to regenerate or curethe degenerated, painful disc tissue that is the essence ofthe disease process [62].

The intervertebral disc is a two-part structure composedof the tough annulus fibrosis (AF) on the periphery andthe amorphous nucleus pulposus (NP) as the central core[60,63]. The cellular content of the AF and NP differ signif-icantly. Cells in the AF are fibroblast-like with robust col-lagen fibrils. The extracellular matrix (ECM) in the AFis predominantly composed of type I collagen and con-tains relatively low amounts of proteoglycan and water[60,64]. Within the NP, cells become more rounded witha chondrocyte-like shape. ECM in the NP is an amorphousarrangement of type II collagen. It has a high proteoglycanconcentration with inverse proportions of collagen and pro-teoglycan compared with the AF [61,64].

The relationship between the fluid-like NP and the struc-tural lattice of the AF provides the biomechanical propertiesnecessary for spinal stability [65]. Any disturbance of theNP or AF that compromises this balance leads to disc degen-eration. Degeneration is marked by a loss of disc height andblurring of the gross morphologic characteristics of the AFand NP, resulting from the inability of the disc to regenerate,avascularity, and numerous biochemical changes within thedisc due to aging [66,67].

Alterations in the cell population of the IVD have beenparticularly implicated in the pathogenesis of IVD degener-ation [68]. Early disc degeneration is characterized by theloss of notochordal cells, which play a key role in maintain-ing the integrity of the disc and matrix stabilization [69].Loss of chondrogenic NP cells and their replacement byfibroblast-like cells has also been implicated in the laterstages of IVD degeneration [70].

Although considerable research has been dedicated tostrategies that boost the ability of native disc cells to syn-thesize and maintain the ECM, the introduction of exoge-nous cells such as stem cells to supplement or replenishthe declining disc cell population is gaining popularity asa potential treatment for IVD degeneration [68].

In vitro and in vivo: cells, growth factors, scaffolds,and animal models

CellsSimilar to bone regeneration, preclinical studies of stem

cells for the IVD focus on MSCs from a variety of sources,including bone marrow [68,71–74], adipose tissue [75–80],muscle tissue [81], synovium [82–84], and even the NP it-self (Table 4).

Bone marrow is an excellent source of mesenchymalstem cells for most orthopedic applications including chon-drogenic differentiation for the diseased intervertebral disc.Yamamoto et al. [85] investigated the use of bone marrow–derived stromal cells for upregulation of the viability ofnative nucleus pulposus cells in a direct contact coculturesystem. The authors found significant upregulation of cellproliferation, DNA synthesis and proteoglycan synthesisin the NP cells which had direct cell-to-cell contact withbone marrow MSCs. Stoyanov et al. [86] compared stan-dard chondrogenic differentiation protocols using trans-forming growth factor (TGF)-b with a novel techniqueutilizing hypoxia, GDF-5 and coculture with NP cells.The authors found that both hypoxia and GDF-5 were suit-able for directing bone marrow MSCs toward an

Table 3Animal models of gene therapy with mesenchymal stem cell for spinal fusion

Animal model Mesenchymal stem cell type Type of fusion References

Mouse Bone marrow, adipose Posterolateral lumbar spine [38]Rat Bone marrow, adipose Posterolateral lumbar spine [30,32,35,37,40,41,124]Rabbit Bone marrow Posterolateral and interbody, lumbar spine [127]Pig Bone marrow Interbody lumbar spine [128]

546 B.C. Werner et al. / The Spine Journal 14 (2014) 542–551

intervertebral disc-like phenotype. Allon et al. [72] ex-tended these in vitro findings of successful co-culture toa rat in vivo model. Utilizing a novel bilaminar co-culturepellet of bone marrow MSCs and NP cells, the authorsdemonstrated prevention of disc degeneration in vivo.

Similar to bone applications of MSCs, adipose derivedstem cells have gained significant attention in recent years,given the ease of procurement and larger number of mesen-chymal stem cells that can be obtained compared tobone marrow or other sources (Fig. 3). Liang et al. [87] re-cently conducted an in vitro study of both adolescent andadult adipose-derived stem cells and found that despitethe harsh microenvironment of the IVD, adipose derivedstem cells are a viable option for IVD regeneration. Choiand co-authors [80] investigated the use of human adiposederived stem cells co-cultured with human nucleus pulpo-sus cells using a porous membrane. The authors foundthe most ideal culture conditions for chondrogenic differen-tiation of the ASCs to be ASC pellets cultured with normalNP cells. ASCs cultured in monolayer and the use of hu-man NP cells from degenerative discs had inferior results.

Yang et al. [88] utilized genetic engineering to induce ec-topic expression of Sox-9 in humanASCs. The authors foundimproved chondrogenic differentiation in the engineeredASCs, indicating potential application in the treatment of de-generative disc disease. Lu et al. [89] reported that co-cultureof human ASCs and NP cells in a micro- mass cultureresulted in differentiation of ASCs into an NP cell-like

phenotype. Tapp et al. [78] revealed that either treatment ofTGF-b or co-culture with human disc cells could signifi-cantly stimulate expression of proteoglycan and type I colla-gen in 3D-cultured sand rat ADSCs. Gaetani et al. [75]presented data indicating that co-culture of human NP andASCs improved the quality of the invitro reconstructed tissuein term of matrix production and 3D cell organization [62].

Other stem cell sources have been investigated for IVDrepair. Vadala et al. [81] investigated the use of adult mus-cle tissue as a stem cell source. Others have recently inves-tigated the use of synovial cells as a stem cell source forIVD tissue engineering. [82–84] In these in vitro andin vivo studies, the authors found promise with the use ofsynovial stem cells for their inductive capabilities and theirability to differentiate to a chondrogenic phenotype. Recentliterature has also revealed the potential for human annulusfibrosis cells [90] and nucleus pulposus cells [91] to serveas mesenchymal stem cells.

Growth factorsDisc cell metabolism is modulated by a variety of growth

factors acting in both paracrine and autocrine roles [92,93].These factors function to increase the synthesis of extracel-lular matrix components, block their breakdown, or a com-bination of these roles. Growth factors can be applied inIVD tissue engineering via delivery of the ‘‘naked’’ or‘‘embedded’’ proteins as well as prolonged supplement bycell-based gene therapy [62,92]. Numerous growth factorshave been used for IVD applications (Table 5), includingtransforming growth factor-b [94,95], insulin-like growthfactor-1 [96,97], GDF-5 [98–101], platelet-derived growthfactor [96], osteogenic protein-1/BMP-7 [102–104],BMP-2 and 12 [105–107], and fibroblast growth factor-2[108].

In a 2005 study, Steck, et al. [109] demonstrated theability of adult bone marrow MSCs to differentiate toward

Fig. 3. Chondrogenic differentiation of human adipose–derived stem cells for disc regeneration. Cells cultured in basal medium demonstrated no positivestaining (A). Safranin-O staining indicated deposition of proteoglycan in cells cultured in chondrogenic medium (B). Supplementing the chondrogenic me-dium with transforming growth factor-b (C), and increasing concentrations of growth and differentiation factor-5 (D–F) augmented chondrogenicdifferentiation.

Table 4Mesenchymal stem cell sources for regeneration of the intervertebral disc

Cell source References

Bone marrow [68,71–74]Adipose tissue [75–80]Muscle tissue [81]Synovium [82–84]Nucleus pulposus [62]

547B.C. Werner et al. / The Spine Journal 14 (2014) 542–551

the molecular phenotype of human IVD cells. Using TGF-b mediated generation, the authors found that MSC spher-oids preferentially differentiated toward an IVD phenotyperather than articular cartilage, making them an attractivesource to obtain IVD-like cells.

Yang et al. [110] strengthened the potential role of TGF-b and MSCs as therapy for IVD degeneration. In their study,bone marrow MSCs with TGF-b were transplanted into de-generative discs of rabbits. The authors found that MSCs canslow the rate at which the degenerative process occurs, pos-sibly due to the inhibition of apoptosis by the MSCs.

Stoyanov and co-authors [86] utilized bone marrowMSCs and compared standard chondrogenic differentiationprotocols using TGF-b with hypoxia, GDF-5 and co-culturewith bovine NP cells. Their results suggest that hypoxia andGDF-5 may be suitable for directing MSCs toward an IVD-like phenotype.

McCanless et al. [71] investigated the effects of BMP-2and synthetic peptide B2A on cell proliferation and ECMsynthesis by NP-like differentiated bone marrow MSCs.The authors found that B2A induces proliferation, aggrecansynthesis and stabilized collagen accumulation consistentwith cells of the young, healthy NP, indicating potentialuse of B2A in MSC-based NP regeneration therapy.

Feng et al. [100] examined the effects of GDF-5 onchondrogenesis of adipose-derived stem cells (ADSCs) us-ing genetically engineered rat ADSCs in an in vitro pelletculture model. The authors found that GDF-5 is a potent in-ducer of chondrogenesis in ADSCs, and that ADSCs can begenetically engineered to express pro-chondrogenic growthfactors as a promising therapeutic cell source for IVD tissueengineering.

ScaffoldsScaffolds in tissue engineering can help to retain cells in

the desired location and provide appropriate mechanicalproperties and/or biochemical signals. [62] To date, a vari-ety of biomaterials have been used for fabricating scaffoldsin both annulus fibrosis (AF) and nucleus pulposus (NP) tis-sue engineering.

The choice of scaffold is critical as it directly affects thetype of tissue that forms. Scaffold fiber diameter and

stiffness can influence cell function, proliferation and orien-tation, all of which can have profound effects on the cell[111]. Numerous biomaterials have been used for scaffolds,including PLGA, poly-d-lactide, chitosan, alginate, fibrin,collagens, gels, calcium polyphosphate, demineralizedbone matrix, and many others [61,112–122].

Conclusions

The future of spine surgery is constantly evolving, andthe recent advancements in stem cell–based technologiesfor both spine fusion and the treatment of degenerative discdisease are both promising and indicative that stem cellswill undoubtedly play a major role clinically. In both osteo-genesis and chondrogenesis, adult stem cells derived frombone marrow or adipose show promise in preclinical stud-ies. Various growth factors and scaffolds have also beenshown to enhance the properties and eventual clinical po-tential of these cells. Although its utility in clinical applica-tions has yet to be proven, gene therapy has also beenshown to hold promise in preclinical studies. It is likely thatthese stem cells, growth factors, and scaffolds will playa critical role in the future for replacing diseased tissue indisease processes such as degenerative disc disease and inenhancing host tissue to achieve more reliable spine fusion.

References

[1] Deyo RA, Gray DT, Kreuter W, et al. United states trends in lumbarfusion surgery for degenerative conditions. Spine 2005;30:1441–5;discussion 1446–7.

[2] Gray DT, Deyo RA, Kreuter W, et al. Population-based trends involumes and rates of ambulatory lumbar spine surgery. Spine2006;31:1957–63; discussion 1964.

[3] Katz JN. Lumbar spinal fusion. surgical rates, costs, and complica-tions. Spine 1995;20(24 Suppl):78S–83S.

[4] Lee CK, Langrana NA. A review of spinal fusion for degenerativedisc disease: need for alternative treatment approach of disc arthro-plasty? Spine J 2004;4(6 Suppl):173S–6S.

[5] Taylor VM, Deyo RA, Cherkin DC, Kreuter W. Low back pain hos-pitalization. Recent United States trends and regional variations.Spine 1994;19:1207–12; discussion 1213.

[6] Mok JM, Cloyd JM, Bradford DS, et al. Reoperation after primaryfusion for adult spinal deformity: rate, reason, and timing. Spine2009;34:832–9.

[7] Pichelmann MA, Lenke LG, Bridwell KH, et al. Revision ratesfollowing primary adult spinal deformity surgery: six hundredforty-three consecutive patients followed-up to twenty-two yearspostoperative. Spine 2010;35:219–26.

[8] Kim YJ, Bridwell KH, Lenke LG, et al. Pseudarthrosis in primaryfusions for adult idiopathic scoliosis: incidence, risk factors, andoutcome analysis. Spine 2005;30:468–74.

[9] Kim YJ, Bridwell KH, Lenke LG, et al. Pseudarthrosis in long adultspinal deformity instrumentation and fusion to the sacrum: preva-lence and risk factor analysis of 144 cases. Spine 2006;31:2329–36.

[10] Dipaola CP, Bible JE, Biswas D, et al. Survey of spine surgeons onattitudes regarding osteoporosis and osteomalacia screening andtreatment for fractures, fusion surgery, and pseudoarthrosis. SpineJ 2009;9:537–44.

[11] Cavagna R, Tournier C, Aunoble S, et al. Lumbar decompressionand fusion in elderly osteoporotic patients: a prospective study using

Table 5Growth factors used with mesenchymal stem cells for regeneration of theintervertebral disc

Cell source References

Transforming growth factor-b [94,95]IGF-1 [96,97]Growth and differentiation factor-5 [98–101]Platelet-derived growth factor [96]OP-1/BMP-7 [102–104]BMP-2, BMP-12 [105–107]FGF-2 [108]

BMP, bone morphogenic protein; FGF, fibroblast growth factor; IGF,insulin-like growth factor-1; OP, osteogenic protein.

548 B.C. Werner et al. / The Spine Journal 14 (2014) 542–551

less rigid titanium rod fixation. J Spinal Disord Tech 2008;21:86–91.

[12] Keaveny TM, Yeh OC. Architecture and trabecular bone - towardan improved understanding of the biomechanical effects of age,sex and osteoporosis. J Musculoskelet Neuronal Interact 2002;2:205–8.

[13] Hart RA, Prendergast MA. Spine surgery for lumbar degenerativedisease in elderly and osteoporotic patients. Instr Course Lect2007;56:257–72.

[14] Zdeblick TA. A prospective, randomized study of lumbar fusion.preliminary results. Spine 1993;18:983–91.

[15] McGuire RA, Amundson GM. The use of primary internal fixationin spondylolisthesis. Spine 1993;18:1662–72.

[16] Miyazaki M, Tsumura H, Wang JC, Alanay A. An update on bonesubstitutes for spinal fusion. Eur Spine J 2009;18:783–99.

[17] Kim DH, Rhim R, Li L, et al. Prospective study of iliac crest bonegraft harvest site pain and morbidity. Spine J 2009;9:886–92.

[18] Schaaf H, Lendeckel S, Howaldt HP, Streckbein P. Donor site mor-bidity after bone harvesting from the anterior iliac crest. Oral SurgOral Med Oral Pathol Oral Radiol Endod 2010;109:52–8.

[19] Singh JR, Nwosu U, Egol KA. Long-term functional outcome anddonor-site morbidity associated with autogenous iliac crest bonegrafts utilizing a modified anterior approach. Bull NYU Hosp JtDis 2009;67:347–51.

[20] Pitzen T, Kranzlein K, Steudel WI, Strowitzki M. Complaints andfindings at the iliac crest donor site following anterior cervical fu-sion. Zentralbl Neurochir 2004;65:7–12.

[21] Sasso RC, LeHuec JC, Shaffrey C, Spine Interbody ResearchGroup. Iliac crest bone graft donor site pain after anterior lumbarinterbody fusion: a prospective patient satisfaction outcome assess-ment. J Spinal Disord Tech 2005;18(Suppl):S77–81.

[22] Silber JS, Anderson DG, Daffner SD, et al. Donor site morbidity af-ter anterior iliac crest bone harvest for single-level anterior cervicaldiscectomy and fusion. Spine 2003;28:134–9.

[23] Shen FH, Samartzis D, An HS. Cell technologies for spinal fusion.Spine J 2005;5(6 Suppl):231S–9S.

[24] Caplan AI, Bruder SP. Mesenchymal stem cells: building blocks formolecular medicine in the 21st century. Trends Mol Med 2001;7:259–64.

[25] Jackson KA, Mi T, Goodell MA. Hematopoietic potential of stemcells isolated from murine skeletal muscle. Proc Natl Acad Sci US A 1999;96:14482–6.

[26] Williams JT, Southerland SS, Souza J, et al. Cells isolated fromadult human skeletal muscle capable of differentiating into multiplemesodermal phenotypes. Am Surg 1999;65:22–6.

[27] Nakahara H, Dennis JE, Bruder SP, et al. In vitro differentiation ofbone and hypertrophic cartilage from periosteal-derived cells. ExpCell Res 1991;195:492–503.

[28] Zuk PA, Zhu M, Mizuno H, et al. Multilineage cells from humanadipose tissue: Implications for cell-based therapies. Tissue Eng2001;7:211–28.

[29] Cui Q, Ming Xiao Z, Balian G, Wang GJ. Comparison of lumbarspine fusion using mixed and cloned marrow cells. Spine2001;26:2305–10.

[30] Peterson B, Iglesias R, Zhang J, et al. Genetically modified humanderived bone marrow cells for posterolateral lumbar spine fusion inathymic rats: beyond conventional autologous bone grafting. Spine2005;30:283–9; discussion 289–90.

[31] Gupta MC, Theerajunyaporn T, Maitra S, et al. Efficacy of mesen-chymal stem cell enriched grafts in an ovine posterolateral lumbarspine model. Spine 2007;32:720–6; discussion 727.

[32] Miyazaki M, Zuk PA, Zou J, et al. Comparison of human mesenchy-mal stem cells derived from adipose tissue and bone marrow forex vivo gene therapy in rat spinal fusion model. Spine 2008;33:863–9.

[33] Zeng Q, Li X, Beck G, et al. Growth and differentiation factor-5(GDF-5) stimulates osteogenic differentiation and increases

vascular endothelial growth factor (VEGF) levels in fat-derived stro-mal cells in vitro. Bone 2007;40:374–81.

[34] Shen FH, Zeng Q, Lv Q, et al. Osteogenic differentiation of adipose-derived stromal cells treated with GDF-5 cultured on a novel three-dimensional sintered microsphere matrix. Spine J 2006;6:615–23.

[35] Hsu WK, Wang JC, Liu NQ, et al. Stem cells from human fat as cel-lular delivery vehicles in an athymic rat posterolateral spine fusionmodel. J Bone Joint Surg Am 2008;90:1043–52.

[36] Lopez MJ, McIntosh KR, Spencer ND, et al. Acceleration of spinalfusion using syngeneic and allogeneic adult adipose derived stemcells in a rat model. J Orthop Res 2009;27:366–73.

[37] Wang JC, Kanim LE, Yoo S, et al. Effect of regional gene therapywith bone morphogenetic protein-2-producing bone marrow cells onspinal fusion in rats. J Bone Joint Surg Am 2003;85-A:905–11.

[38] Sheyn D, Ruthemann M, Mizrahi O, et al. Genetically modifiedmesenchymal stem cells induce mechanically stable posterior spinefusion. Tissue Eng Part A 2010;16:3679–86.

[39] Fu TS, Chen WJ, Chen LH, et al. Enhancement of posterolaterallumbar spine fusion using low-dose rhBMP-2 and cultured marrowstromal cells. J Orthop Res 2009;27:380–4.

[40] Hidaka C, Goshi K, Rawlins B, et al. Enhancement of spine fusionusing combined gene therapy and tissue engineering BMP-7-expressing bone marrow cells and allograft bone. Spine 2003;28:2049–57.

[41] Zhu W, Rawlins BA, Boachie-Adjei O, et al. Combined bone mor-phogenetic protein-2 and -7 gene transfer enhances osteoblastic dif-ferentiation and spine fusion in a rodent model. J Bone Miner Res2004;19:2021–32.

[42] Jenis LG, Wheeler D, Parazin SJ, Connolly RJ. The effect of oste-ogenic protein-1 in instrumented and noninstrumented posterolat-eral fusion in rabbits. Spine J 2002;2:173–8.

[43] Bomback DA, Grauer JN, Lugo R, et al. Comparison of posterolat-eral lumbar fusion rates of Grafton putty and OP-1 putty in an athy-mic rat model. Spine 2004;29:1612–7.

[44] Zeng Q, Li X, Choi L, et al. Recombinant growth/differentiationfactor-5 stimulates osteogenic differentiation of fat-derived stromalcells in vitro. Connect Tissue Res 2006;47:264–70.

[45] Kwon B, Jenis LG. Carrier materials for spinal fusion. Spine J2005;5(6 Suppl):224S–30S.

[46] Bueno EM, Glowacki J. Cell-free and cell-based approaches forbone regeneration. Nat Rev Rheumatol 2009;5:685–97.

[47] Gabbay JS, Heller JB, Mitchell SA, et al. Osteogenic potentiation ofhuman adipose-derived stem cells in a 3-dimensional matrix. AnnPlast Surg 2006;57:89–93.

[48] Kimelman-Bleich N, Pelled G, Sheyn D, et al. The use of a syntheticoxygen carrier-enriched hydrogel to enhance mesenchymal stemcell-based bone formation in vivo. Biomaterials 2009;30:4639–48.

[49] Lee JH, Rhie JW, Oh DY, Ahn ST. Osteogenic differentiation of hu-man adipose tissue-derived stromal cells (hASCs) in a porous three-dimensional scaffold. Biochem Biophys Res Commun 2008;370:456–60.

[50] Machado CB, Ventura JM, Lemos AF, et al. 3D chitosan-gelatin-chondroitin porous scaffold improves osteogenic differentiation ofmesenchymal stem cells. Biomed Mater 2007;2:124–31.

[51] Natesan S, Baer DG, Walters TJ, et al. Adipose-derived stem celldelivery into collagen gels using chitosan microspheres. TissueEng Part A 2010;16:1369–84.

[52] Costa-Pinto AR, Reis RL, Neves NM. Scaffolds based bone tissueengineering: the role of chitosan. Tissue Eng Part B Rev 2011;17:331–47.

[53] Chistolini P, Ruspantini I, Bianco P, et al. Biomechanical evaluationof cell-loaded and cell-free hydroxyapatite implants for the recon-struction of segmental bone defects. J Mater Sci Mater Med1999;10:739–42.

[54] Minamide A, Yoshida M, Kawakami M, et al. The use of culturedbone marrow cells in type I collagen gel and porous hydroxyapatitefor posterolateral lumbar spine fusion. Spine 2005;30:1134–8.

549B.C. Werner et al. / The Spine Journal 14 (2014) 542–551

[55] Seo HS, Jung JK, Lim MH, et al. Evaluation of spinal fusion usingbone marrow derived mesenchymal stem cells with or without fibro-blast growth factor-4. J Korean Neurosurg Soc 2009;46:397–402.

[56] Gan Y, Dai K, Zhang P, et al. The clinical use of enriched bone mar-row stem cells combined with porous beta-tricalcium phosphate inposterior spinal fusion. Biomaterials 2008;29:3973–82.

[57] Frymoyer JW, Cats-Baril WL. An overview of the incidences andcosts of low back pain. Orthop Clin North Am 1991;22:263–71.

[58] Sakai D. Stem cell regeneration of the intervertebral disk. OrthopClin North Am 2011;42:555–62. [viii-ix].

[59] Phillips FM, An H, Kang JD, et al. Biologic treatment for interver-tebral disc degeneration: summary statement. Spine 2003;28(15Suppl):S99.

[60] Kepler CK, Anderson DG, Tannoury C, Ponnappan RK. Interverte-bral disk degeneration and emerging biologic treatments. J Am AcadOrthop Surg 2011;19:543–53.

[61] O’Halloran DM, Pandit AS. Tissue-engineering approach to regen-erating the intervertebral disc. Tissue Eng 2007;13:1927–54.

[62] Yang X, Li X. Nucleus pulposus tissue engineering: a brief review.Eur Spine J 2009;18:1564–72.

[63] Huang S, Tam V, Cheung KM, et al. Stem cell-based approaches forintervertebral disc regeneration. Curr Stem Cell Res Ther 2011;6:317–26.

[64] Choi YS. Pathophysiology of degenerative disc disease. Asian SpineJ 2009;3:39–44.

[65] Buckwalter JA. Aging and degeneration of the human intervertebraldisc. Spine 1995;20:1307–14.

[66] Risbud MV, Shapiro IM, Vaccaro AR, Albert TJ. Stem cell regener-ation of the nucleus pulposus. Spine J 2004;4(6 Suppl):348S–53S.

[67] Risbud MV, Guttapalli A, Tsai TT, et al. Evidence for skeletal pro-genitor cells in the degenerate human intervertebral disc. Spine2007;32:2537–44.

[68] Sobajima S, Vadala G, Shimer A, et al. Feasibility of a stem celltherapy for intervertebral disc degeneration. Spine J 2008;8:888–96.

[69] Aguiar DJ, Johnson SL, Oegema TR. Notochordal cells interactwith nucleus pulposus cells: regulation of proteoglycan synthesis.Exp Cell Res 1999;246:129–37.

[70] Kim KW, Lim TH, Kim JG, et al. The origin of chondrocytes in thenucleus pulposus and histologic findings associated with the transi-tion of a notochordal nucleus pulposus to a fibrocartilaginous nu-cleus pulposus in intact rabbit intervertebral discs. Spine 2003;28:982–90.

[71] McCanless JD, Cole JA, Slack SM, et al. Modeling nucleus pulpo-sus regeneration in vitro: mesenchymal stem cells, alginate beads,hypoxia, bone morphogenetic protein-2, and synthetic peptideB2A. Spine 2011;36:2275–85.

[72] Allon AA, Aurouer N, Yoo BB, et al. Structured coculture of stemcells and disc cells prevent disc degeneration in a rat model. Spine J2010;10:1089–97.

[73] Hiyama A, Mochida J, Iwashina T, et al. Transplantation of mesen-chymal stem cells in a canine disc degeneration model. J Orthop Res2008;26:589–600.

[74] Jandial R, Aryan HE, Park J, et al. Stem cell-mediated regenerationof the intervertebral disc: cellular and molecular challenge. Neuro-surg Focus 2008;24:E21.

[75] Gaetani P, Torre ML, Klinger M, et al. Adipose-derived stem celltherapy for intervertebral disc regeneration: an in vitro recon-structed tissue in alginate capsules. Tissue Eng Part A 2008;14:1415–23.

[76] Lu ZF, Zandieh Doulabi B, Wuisman PI, et al. Differentiation of ad-ipose stem cells by nucleus pulposus cells: configuration effect. Bi-ochem Biophys Res Commun 2007;359:991–6.

[77] Hoogendoorn RJ, Lu ZF, Kroeze RJ, et al. Adipose stem cells forintervertebral disc regeneration: current status and concepts forthe future. J Cell Mol Med 2008;12:2205–16.

[78] Tapp H, Deepe R, Ingram JA, et al. Adipose-derived mesenchymalstem cells from the sand rat: transforming growth factor beta and 3D

co-culture with human disc cells stimulate proteoglycan and colla-gen type I rich extracellular matrix. Arthritis Res Ther 2008;10:R89.

[79] Jeong JH, Lee JH, Jin ES, et al. Regeneration of intervertebral discsin a rat disc degeneration model by implanted adipose-tissue-derived stromal cells. Acta Neurochir (Wien) 2010;152:1771–7.

[80] Choi EH, Park H, Park KS, et al. Effect of nucleus pulposus cellshaving different phenotypes on chondrogenic differentiation ofadipose-derived stromal cells in a coculture system using porousmembranes. Tissue Eng Part A 2011;17:2445–51.

[81] Vadala G, Sobajima S, Lee JY, et al. In vitro interaction betweenmuscle-derived stem cells and nucleus pulposus cells. Spine J2008;8:804–9.

[82] Chen S, Emery SE, Pei M. Coculture of synovium-derived stemcells and nucleus pulposus cells in serum-free defined medium withsupplementation of transforming growth factor-beta1: a potentialapplication of tissue-specific stem cells in disc regeneration. Spine2009;34:1272–80.

[83] He F, Pei M. Rejuvenation of nucleus pulposus cells using extracel-lular matrix deposited by synovium-derived stem cells. Spine2012;37:459–69.

[84] Miyamoto T, Muneta T, Tabuchi T, et al. Intradiscal transplantationof synovial mesenchymal stem cells prevents intervertebral disc de-generation through suppression of matrix metalloproteinase-relatedgenes in nucleus pulposus cells in rabbits. Arthritis Res Ther2010;12:R206.

[85] Yamamoto Y, Mochida J, Sakai D, et al. Upregulation of the viabil-ity of nucleus pulposus cells by bone marrow-derived stromal cells:significance of direct cell-to-cell contact in coculture system. Spine2004;29:1508–14.

[86] Stoyanov JV, Gantenbein-Ritter B, Bertolo A, et al. Role of hypoxiaand growth and differentiation factor-5 on differentiation of humanmesenchymal stem cells towards intervertebral nucleus pulposus-like cells. Eur Cell Mater 2011;21:533–47.

[87] Liang C, Li H, Tao Y, et al. Responses of human adipose-derivedmesenchymal stem cells to chemical microenvironment of the inter-vertebral disc. J Transl Med 2012;10:49.

[88] Yang Z, Huang CY, Candiotti KA, et al. Sox-9 facilitates differen-tiation of adipose tissue-derived stem cells into a chondrocyte-likephenotype in vitro. J Orthop Res 2011;29:1291–7.

[89] Lu ZF, Doulabi BZ, Wuisman PI, et al. Influence of collagen typeII and nucleus pulposus cells on aggregation and differentiationof adipose tissue-derived stem cells. J Cell Mol Med 2008;12:2812–22.

[90] Feng G, Yang X, Shang H, et al. Multipotential differentiation of hu-man anulus fibrosus cells: an in vitro study. J Bone Joint Surg Am2010;92:675–85.

[91] Mark Erwin W, Islam D, Eftekarpour E, et al. Intervertebral disc-derived stem cells: Implications for regenerative medicine and neu-ral repair. Spine 2013;38:211–6.

[92] Paesold G, Nerlich AG, Boos N. Biological treatment strategies fordisc degeneration: potentials and shortcomings. Eur Spine J2007;16:447–68.

[93] Yang SH, Wu CC, Shih TT, et al. In vitro study on interaction be-tween human nucleus pulposus cells and mesenchymal stem cellsthrough paracrine stimulation. Spine 2008;33:1951–7.

[94] Matsunaga S, Nagano S, Onishi T, et al. Age-related changes in ex-pression of transforming growth factor-beta and receptors in cells ofintervertebral discs. J Neurosurg 2003;98(1 Suppl):63–7.

[95] Specchia N, Pagnotta A, Toesca A, Greco F. Cytokines and growthfactors in the protruded intervertebral disc of the lumbar spine. EurSpine J 2002;11:145–51.

[96] Gruber HE, Norton HJ, Hanley EN Jr. Anti-apoptotic effects of IGF-1 and PDGF on human intervertebral disc cells in vitro. Spine2000;25:2153–7.

[97] Okuda S, Myoui A, Ariga K, et al. Mechanisms of age-related de-cline in insulin-like growth factor-I dependent proteoglycan synthe-sis in rat intervertebral disc cells. Spine 2001;26:2421–6.

550 B.C. Werner et al. / The Spine Journal 14 (2014) 542–551

[98] Chujo T, An HS, Akeda K, et al. Effects of growth differentiationfactor-5 on the intervertebral disc–in vitro bovine study andin vivo rabbit disc degeneration model study. Spine 2006;31:2909–17.

[99] Cui M, Wan Y, Anderson DG, et al. Mouse growth and differentia-tion factor-5 protein and DNA therapy potentiates intervertebral disccell aggregation and chondrogenic gene expression. Spine J 2008;8:287–95.

[100] Feng G, Wan Y, Balian G, et al. Adenovirus-mediated expression ofgrowth and differentiation factor-5 promotes chondrogenesis of ad-ipose stem cells. Growth Factors 2008;26:132–42.

[101] Li X, Lee JP, Balian G, Greg Anderson D. Modulation of chondro-cytic properties of fat-derived mesenchymal cells in co-cultures withnucleus pulposus. Connect Tissue Res 2005;46:75–82.

[102] Kawakami M, Matsumoto T, Hashizume H, et al. Osteogenicprotein-1 (osteogenic protein-1/bone morphogenetic protein-7) in-hibits degeneration and pain-related behavior induced by chroni-cally compressed nucleus pulposus in the rat. Spine 2005;30:1933–9.

[103] Masuda K, Imai Y, Okuma M, et al. Osteogenic protein-1 injectioninto a degenerated disc induces the restoration of disc height andstructural changes in the rabbit anular puncture model. Spine2006;31:742–54.

[104] Wei A, Brisby H, Chung SA, Diwan AD. Bone morphogeneticprotein-7 protects human intervertebral disc cells in vitro from apo-ptosis. Spine J 2008;8:466–74.

[105] Gilbertson L, Ahn SH, Teng PN, et al. The effects of recombinanthuman bone morphogenetic protein-2, recombinant human bonemorphogenetic protein-12, and adenoviral bone morphogeneticprotein-12 on matrix synthesis in human annulus fibrosis and nu-cleus pulposus cells. Spine J 2008;8:449–56.

[106] Huang KY, Yan JJ, Hsieh CC, et al. The in vivo biological effects ofintradiscal recombinant human bone morphogenetic protein-2 onthe injured intervertebral disc: an animal experiment. Spine2007;32:1174–80.

[107] Shen B, Wei A, Tao H, et al. BMP-2 enhances TGF-beta3-mediatedchondrogenic differentiation of human bone marrow multipotentmesenchymal stromal cells in alginate bead culture. Tissue Eng PartA 2009;15:1311–20.

[108] Tsai TT, Guttapalli A, Oguz E, et al. Fibroblast growth factor-2maintains the differentiation potential of nucleus pulposus cellsin vitro: implications for cell-based transplantation therapy. Spine2007;32:495–502.

[109] Steck E, Bertram H, Abel R, et al. Induction of intervertebral disc-like cells from adult mesenchymal stem cells. Stem Cells 2005;23:403–11.

[110] Yang H, Wu J, Liu J, et al. Transplanted mesenchymal stem cellswith pure fibrinous gelatin-transforming growth factor-beta1 de-crease rabbit intervertebral disc degeneration. Spine J 2010;10:802–10.

[111] Kandel R, Roberts S, Urban JP. Tissue engineering and the interver-tebral disc: the challenges. Eur Spine J 2008;17(4 Suppl):480–91.

[112] Chang G, Kim HJ, Kaplan D, et al. Porous silk scaffolds can be usedfor tissue engineering annulus fibrosus. Eur Spine J 2007;16:1848–57.

[113] Chang G, Kim HJ, Vunjak-Novakovic G, et al. Enhancing annulusfibrosus tissue formation in porous silk scaffolds. J Biomed MaterRes A 2010;92:43–51.

[114] Gruber HE, Leslie K, Ingram J, et al. Cell-based tissue engineeringfor the intervertebral disc: in vitro studies of human disc cell geneexpression and matrix production within selected cell carriers. SpineJ 2004;4:44–55.

[115] Halloran DO, Grad S, Stoddart M, et al. An injectable cross-linkedscaffold for nucleus pulposus regeneration. Biomaterials 2008;29:438–47.

[116] Calderon L, Collin E, Velasco-Bayon D, et al. Type II collagen-hyaluronan hydrogel–a step towards a scaffold for intervertebraldisc tissue engineering. Eur Cell Mater 2010;20:134–48.

[117] Mercuri JJ, Gill SS, Simionescu DT. Novel tissue-derived biomi-metic scaffold for regenerating the human nucleus pulposus. J Bi-omed Mater Res A 2011;96:422–35.

[118] Roughley P, Hoemann C, DesRosiers E, et al. The potential ofchitosan-based gels containing intervertebral disc cells for nucleuspulposus supplementation. Biomaterials 2006;27:388–96.

[119] Nesti LJ, Li WJ, Shanti RM, et al. Intervertebral disc tissue engi-neering using a novel hyaluronic acid-nanofibrous scaffold(HANFS) amalgam. Tissue Eng Part A 2008;14:1527–37.

[120] Wang J, Liu J, Tian H, et al. Comparison study on injectable tissueengineered nucleus pulposus constructed by different cells and chi-tosan hydrogel. Zhongguo Xiu Fu Chong Jian Wai Ke Za Zhi2010;24:806–10.

[121] Feng G, Jin X, Hu J, et al. Effects of hypoxias and scaffold architec-ture on rabbit mesenchymal stem cell differentiation towards a nu-cleus pulposus-like phenotype. Biomaterials 2011;32:8182–9.

[122] Bertolo A, Mehr M, Aebli N, et al. Influence of different commer-cial scaffolds on the in vitro differentiation of human mesenchymalstem cells to nucleus pulposus-like cells. Eur Spine J 2012;21(6Suppl):826–38.

[123] Valdes M, Moore DC, Palumbo M, et al. rhBMP-6 stimulated osteo-progenitor cells enhance posterolateral spinal fusion in the NewZealand white rabbit. Spine J 2007;7:318–25.

[124] Dumont RJ, Dayoub H, Li JZ, et al. Ex vivo bone morphogeneticprotein-9 gene therapy using human mesenchymal stem cells in-duces spinal fusion in rodents. Neurosurgery 2002;51:1239–44; dis-cussion 1244–5.

[125] Minamide A, Yoshida M, Kawakami M, et al. The effects of bonemorphogenetic protein and basic fibroblast growth factor on cul-tured mesenchymal stem cells for spine fusion. Spine 2007 1;32:1067–71.

[126] Wang T, Dang G, Guo Z, Yang M. Evaluation of autologous bonemarrow mesenchymal stem cell-calcium phosphate ceramic com-posite for lumbar fusion in rhesus monkey interbody fusion model.Tissue Eng 2005;11:1159–67.

[127] Riew KD, Wright NM, Cheng S, et al. Induction of bone formationusing a recombinant adenoviral vector carrying the human BMP-2gene in a rabbit spinal fusion model. Calcif Tissue Int 1998;63:357–60.

[128] Riew KD, Lou J, Wright NM, et al. Thoracoscopic intradiscal spinefusion using a minimally invasive gene-therapy technique. J BoneJoint Surg Am 2003;85-A:866–71.

551B.C. Werner et al. / The Spine Journal 14 (2014) 542–551