A Comparison of Mesenchymal Precursor Cells and Amnion Epithelial Cells for Enhancing Cervical Interbody Fusion in an Ovine Model

RESEARCH—ANIMAL TOPIC Research—Animal

Tony Goldschlager, MBBS, BACKGROUND: Rapid, reliable fusion is the goal in anterior cervical diskectomy and PhD*‡{‡‡ fusion. Iliac crest autograft has a high rate of donor-site morbidity. Alternatives such as Peter Ghosh, DSc, PhD§** bone graft substitutes lack osteoinductivity, and recombinant bone morphogenetic Andrew Zannettino, PhD# proteins risk life-threatening complications. Both allogeneic mesenchymal precursor cells (MPCs) and amnion derived epithelial cells (AECs) have osteogenic potential. Mark Williamson, DVM, PhD‡‡ Jeffrey Victor Rosenfeld, MD, OBJECTIVE: To compare for the first time the capacity of MPCs and AECs to promote FRACS‡k osteogenesis in an ovine model. Silviu Itescu, MD, PhD§ METHODS: Five groups of 2-year-old ewes were subjected to C3-4 anterior cervical diskectomy and fusion with a Fidji interbody cage packed with iliac crest autograft alone Graham Jenkin, PhD§§ (group A; n = 6), hydroxyapatite-tricalcium phosphate Mastergraft granules (HA/TCP) *Monash Immunology and Stem Cell alone (group B; n = 6), HA/TCP containing 5 million MPCs (group C; n = 6), or HA/TCP Laboratories, Melbourne, Victoria, Australia; containing 5 million AECs (group D; n = 5); group E was made up of age-matched ‡Department of Surgery, Monash Uni- versity, Melbourne, Victoria, Australia; nonoperative controls (n = 6). At 3 months, animals were euthanized and quantitative §Mesoblast Ltd, Melbourne, Victoria, multislice computed tomography, functional radiography, biomechanics, histology, and Australia; {Department of Neurosurgery, histomorphometry were performed. Monash Medical Centre, Clayton, Victoria Australia; kDepartment of Neurosurgery, RESULTS: No procedure- or cell-related adverse events were observed. There was sig- The Alfred Hospital, Prahran, Victoria, nificantly more fusion in the MPC group (C) than in group A, B, or D. Computed Australia; #Centre for Cancer Biology, tomography scan at 3 months revealed that 5 of 6 MPC-treated animals (83%) had Institute of Medical and Veterinary Sci- ence/SA Pathology and Robinson Insti- continuous bony bridging compared with 0 of 5 AEC-treated and only 1 of 6 autograft- tute, University of Adelaide, South and 2 of 6 HA/TCP-treated animals (P = .01). Australia, Australia; **Institute of Bone and Joint Research, Royal North Shore CONCLUSION: Implantation of allogeneic MPCs in combination with HA/TCP within an Hospital, St Leonards, New South Wales, interbody spacer facilitates interbody fusion after diskectomy. The earlier, more robust Australia; ‡‡Gribbles Pathology Pty Ltd, fusion observed with MPCs relative to autograft and HA/TCP bone substitute indicates Clayton, Victoria, Australia; §§Ritchie Centre, Monash Institute of Medical that this approach may offer a therapeutic benefit. Research, Monash University, Clayton, KEY WORDS: Amnion epithelial cell, Anterior cervical diskectomy and fusion, Biologics, Mesenchymal stem Victoria, Australia cell, Sheep, Spine fusion, Stem cells

Correspondence: Neurosurgery 68:1025–1035, 2011 DOI: 10.1227/NEU.0b013e31820d5375 www.neurosurgery-online.com Dr Tony Goldschlager, MBBS, PhD, Department of Neurosurgery, Monash Medical Centre, 246 Clayton Rd, Clayton, ultiple strategies for interbody grafting fusion because it fulfills these criteria in that it Victoria 3168, Australia. are used in anterior cervical diskectomy contains live cells and growth factors within the E-mail: 1 2 [email protected] Mand fusion surgery. The ideal graft extracellular matrix. However, autograft may should be osteogenic, osteoconductive, and os- result in short- or long-term problems such as Received, February 20, 2010. teoinductive, as well as mechanically stable and prolonged operation time, increased blood loss, Accepted, June 10, 2010. disease free.2,3 Autograft has traditionally been and donor-site morbidity.2,4,5 To alleviate these the gold standard source material for cervical difficulties, alternatives such as allograft cadav- Copyright ª 2011 by the Congress of Neurological Surgeons eric bone and synthetic bone substitutes are used. However, these alternatives are inferior to ABBREVIATIONS: AEC, amnion epithelial cell; autograft because they have limited osteoin- HA/TCP, hydroxyapatite/tricalcium phosphate; ductive properties and depend on the local IVA, intervertebral angle; LA, lordosis angle; 2,4 MPC, mesenchymal precursor stem cell in-growth of osteogenic precursor cells. More- over, the levels and bioactivity of osteogenic

NEUROSURGERY VOLUME 68 | NUMBER 4 | APRIL 2011 | 1025

precursor cells within autographs decline with age, an issue that with monoclonal antibodies23 to STRO-3+ and manufactured by Lonza may reduce bone fusion in the elderly.3 Other factors that may Inc (Walkersville, Maryland) under good manufacturing practice independently decrease fusion rates include multilevel surgery, guidelines. The MPCs were derived from a single batch and culture 6 7,8 expanded to passage 4. The MPC surface marker characteristics have rheumatoid arthritis, smoking, and the concomitant use of 24 antiinflammatory medications.9 If fusion does not occur, a pseu- previously been reported. The optimal dose and matrix combination were determined previously by our group. doarthrosis results, commonly causing recurrent radiculopathy and 10 The cells were frozen and maintained in the vapor phase of a liquid neck pain that may require surgical revision. nitrogen tank until thawed and used within 30 minutes. Cellular via- Recombinant human bone morphogenetic proteins have been bility was determined before implantation using Trypan Blue exclusion tried as osteoinductive agents in the cervical interbody space and and was . 85%. have demonstrated increased fusion rates.11,12 Significantly, this advantage comes with an increase in complication rates of up to Amnion Epithelial Cells 1 24%, including life-threatening complications such as airway After institutional ethics approval (Monash Medical Centre, Clayton, 13 and neurological compression. Australia) and patient informed consent were obtained, discarded term Allogeneic mesenchymal precursor stem cells (MPCs) are placentas were collected from elective caesarean section deliveries. The derived from donor bone marrow and were recently shown to cells were harvested by the method previously described,19 culture ex- increase bone fusion rates in other indications.14-16 Allogeneic panded, and then frozen and maintained in the vapor phase of a liquid amnion epithelial cells (AECs) are readily available from discarded nitrogen tank until thawed and used within 30 minutes. Cellular via- afterbirth tissue; therefore, an invasive procedure such as bone bility was . 80% using Trypan Blue exclusion. The cells were char- marrow aspiration is not required to harvest them. AECs are acterized with monoclonal antibodies and flow cytometry as previously described (Figure 1). derived from epiblast cells before gastrulation that migrate along 17 the walls of the amniotic cavity to form the amnion epithelium. Interbody Cage and Carrier It has been hypothesized, therefore, that they retain pluri- 3 3 potency18 and are able to be differentiated down all 3 germ cell The interbody cage (7.7 12 15 mm) was made from polyetheretherketone and was packed with iliac crest autograft cancellous lineages.18,19 In relation to the present study, AECs are reported 17,19,20 bone in the autograft control group. The HA/TCP carrier was mixed to show significant osteogenic differentiation potential. with autologous blood and then packed into the cage in the HA/TCP Both AECs and MPCs exhibit low immunogenicity and an- alone group. In the cell-treated groups, the MPCs or AECs were added to 18,21 tiinflammatory properties, which diminish their potential to the carrier within the cage and allowed to soak into the granules, avoiding elicit an immune response when transplanted to an allogeneic spillage (Figure 2). host. Human AECs fail to induce a xenogeneic mixed lymphocyte reaction in animal models and exhibit minimal major Surgical Technique and Postoperative Care histocompatibility complex class I and II expression, supporting The surgical procedure has been previously described25 and was 18 the use of AECs xenogeneically. In the present study, we carried out with institutional ethics approval (School of Biomedical compared the capacity of allogeneic ovine MPCs with human Sciences, Monash University). Animals were fasted 12 to 24 hours before AECs to promote cervical interbody fusion in an ovine model of surgery and were allowed water ad libitum. Sheep were anaesthetized by cervical diskectomy. intravenous injection of thiopentone 20 mg/kg and anesthesia maintained by isoflurane (1%-3%) inhalation and were positioned supine on the operating table. Local anesthetic (0.5% Bupivicaine with 1:200 000 METHODS adrenaline) was injected before a right anterolateral approach through a longitudinal neck incision. The longus colli muscle was elevated bi- Study Design laterally with diathermy, and the position of the C3-4 level was confirmed with fluoroscopy. Distraction was achieved with 16-mm Caspar Twenty-nine 2-year-old Boarder-Leicester/Merino ewes were divided pins followed by a total diskectomy and removal of the cartilaginous end randomly into 5 groups. Four groups were subjected to C3-4 anterior plates with a high-speed drill to reveal bleeding bone. The posterior cervical diskectomy and prepared for fusion with a Fidji interbody cage longitudinal ligament was opened until the dura was visualized to di- (Abbott Spine, Austin, Texas). The C3-4 segment was used because of its 22 rectly simulate the clinical procedure and to investigate any effects of the similarities to the human cervical spine. The interbody cage was packed cells around the neural elements. In sheep receiving autograft, the left with Iliac crest autograft alone (group A; n = 6); hydroxyapatite/trical- iliac crest cortical bone was elevated, and cancellous bone was curetted cium phosphate containing 15% hydroxyapatite and 85% tricalcium and packed into the cage. All cages were inserted and countersunk by phosphate (HA/TCP; Mastergraft granules, Medtronic, Minneapolis, approximately 3 mm. The longus colli muscle was then approximated by Minnesota) alone (group B; n = 6); HA/TCP containing 5 million MPCs suture, followed by layered closure and subcuticular suture to skin. (group C; n = 6); or HA/TCP containing 5 million AECs (group D; n = 5). A fentanyl patch was administered postoperatively. After extubation, The fifth group was made up of age-matched nonoperated controls (n = 6). the sheep were transferred to a metabolic cage for observation. After 3 days, the sheep were transferred to open pastures for the duration of the Mesenchymal Precursor Cells study, and regular observations were made. The sheep were allowed to Allogeneic ovine MPCs (Mesoblast Ltd, Melbourne, Australia) were graze ad libitum and supplemented with Lucerne chaff. Clinical pa- isolated from bone marrow of outbred sheep using immunoselection thology was performed on days 0 and 1 and 3 months after surgery for

1026 | VOLUME 68 | NUMBER 4 | APRIL 2011 www.neurosurgery-online.com

FIGURE 1. Characterization of human amnion epithelial cells by flow cytometry.

NEUROSURGERY VOLUME 68 | NUMBER 4 | APRIL 2011 | 1027

then, quantitative CT was performed following the methods of Kand- ziora et al28,29 and Burkus et al.26 Specifically, the coronal slice 3.6 mm (6 slices) anterior to the posterior radiolucent marker of the cage was identified to ensure that an identical location was analyzed for each sample. A standardized elliptical region of interest, with an area of 183 mm2, was selected at this point (Figure 5). The region of interest was used to measure the bone mineral density of the callus formation in Hounsfield units. The density contribution from the HA/TCP or autograft implanted within the cage (background density values) was as- certained from radiographic control animals. These control animals were implanted with either autograft alone or HA/TCP alone and euthanized FIGURE 2. Fidji cervical polyetheretherketone interbody at 1 week to allow time for artifacts such as intraoperative air to be cage packed with Mastergraft granules to which mesenchy- reabsorbed and before the time when fusion would occur. The back- mal precursor cells or amnion epithelial cells were added. ground density was subsequently subtracted from the mean bone mineral density to give the bone mineral density of the callus formation alone, referred to as the bone callus density. standard hematological, comprehensive biochemical, and coagulation assays. Biomechanical Analysis Biomechanical testing was performed in a similar manner to the Radiographic Analysis nonconstrained method described by Gal.30 As previously described,31 Fusion was assessed by plain and functional radiography, multislice pure bending moments were applied to the cervical spine with a custom- thin-cut (0.6-mm) computed tomography (CT), and quantitative CT. made rig. Four degrees of movement, namely flexion, extension, and left Fusion was defined by continuous bridging of trabecular bone and the and right lateral bending, were tested at incremental forces of 0.75 Nm absence of radiolucent lines at 3 months on CT.26,27 Continuous to a maximum of 12 Nm after a 15-Nm preload. bridging of trabecular bone was further subdivided into , 30% or Colored markers, placed into the corpora of C1 to C7 inclusively, . 30% of the interbody cage area as assessed by a blinded consultant were detected with biplanar digital photography. A computerized motion radiologist using a semiquantitative score (see the Table and Figure 3). analysis system (Track Eye Motion Analysis 3.0; Qualisys AB, All animals had plain lateral and selected anteroposterior digital radio- Packhusgatan 6, S-411 13 Gothenburg, Sweden) was used to track and graphs (Radlink, Atomscope HF 200A, Redondo Beach, California) of measure marker positions across the total range of motion. Load-dis- the cervical spine preoperatively, within 24 hours after surgery, and at 1, placement curves were generated to determine stiffness of the C3-4 2, and 3 months postoperatively. This was performed under sedation segments for each degree of movement. Total displacement of the C3-4 using medetomidine (0.025 mg/kg IV) and reversal with atipamezole joint was calculated from these curves. (0.125 mg/kg IV). Functional radiography was conducted by the method of Kandziora Postmortem Analysis 22 et al. After the sheep were euthanized, the superficial musculature of the Clinical veterinarians performed comprehensive autopsies in a blinded explanted fresh spine was removed, carefully sparing all ligaments. T1 was fashion. Samples from all organs and tissues from the perisurgical site rigidly fixated, and a 60-N load was applied through C1. Lateral flexion were reviewed by a blinded, board-certified veterinary pathologist. and extension radiographs were taken, and the intervertebral angle (IVA) 22 and lordosis angle (LA) were measured (Figure 4) and calculated as the Histomorphological, Histomorphometric, and difference between flexion and extension by 3 blinded assessments. Fluorochrome Analyses Multiplanar images were acquired with 0.6-mm collimation on a 64-slice scanner (Siemens Sensation-64, Malvern, Pennsylvania) and The fluorochromes calcein green 10 mg/kg, oxytetracycline 50 mg/kg, reconstructed in the sagittal, axial, and coronal planes. Fusion was and alizarin complexone 30 mg/kg were administered intravenously at 3, assessed for evidence of bridging of trabecular bone as described above; 6, and 9 weeks, respectively. After the sheep were euthanized, the C3-4 segment was excised and fixed in 10% normal buffered formalin. Undecalcified bone histology was performed as previously described.32 Blocks were dehydrated in ascending concentrations of eth- TABLE. Scoring System For Computed Tomography Evidence of anol under agitation and then cleared in butanol before embedding in Fusion glycolmethacrylate (Technovit 7100; Kulzer, Wehrheim, Germany) following a slow embedding and hardening protocol.32 Grade Description For fluorochrome analysis, the midsagittal section was ground to 0 No new bone formation approximately 40-mm thickness with a diamond blade Macrotome (MR 1 New bone formation but not continuous between C3 Ltd, Cambridge, United Kingdom). For light microscopy, the 10-mm and C4 (cleft of discontinuity) sections were cut with a sledge microtome (Leitz, Wetzlar, Germany). 2 Continuous bridging new bone but make up , 30% Histological sections were stained with hematoxylin and eosin, Safranin- of fusion area O/light green, von Kossa, Alcian Blue, and Masson-Goldner, the last 3 Continuous bridging new bone formation of . 30% stain being used for histomorphometric analysis. of fusion area Slides were scanned with the Olympus dot slide System (with a BX51 microscope) at 3 2 magnification for each fluorescent label using

1028 | VOLUME 68 | NUMBER 4 | APRIL 2011 www.neurosurgery-online.com

FIGURE 3. Computed tomography (top, coronal plane; bottom, sagittal plane) showing an example of each scoring grade (see the Table).

U-MNIBA3, U-WIG3, and U-MWU2 filters and a Peltier-cooled high- with Masson-Goldner in a blinded fashion. Consistent inclusive sensitivity camera at consistent exposure. The images were then uploaded threshold mapping was then used to measure the threshold percentage into the Metamorph quantitative analysis software program (version 7.6; within the total cage area, within the section, containing red staining Molecular Devices, MDS Inc, Sunnyvale, California), and the area regions representing osteoid volume and green staining regions repre- within the cage was identified and demarcated with the Metamorph senting mineralized bone volume, respectively34 (Figure 6). proprietary software package. The intensity of each fluorescent label within the cage was measured and expressed as intensity per unit area (square microns). This gave a quantitative assessment of the amount of Statistical Analysis bone deposition at each time point. Comparison of nonparametric data was evaluated by the Kruskal- A certified veterinary pathologist performed a semiquantitative anal- Wallis test on the median values followed by the Dunn multiple- ysis of the histological sections in a blinded manner. A score, using the comparison test when significant differences were observed. Parametric criteria of Zdeblick et al,33 to assess fusion was assigned separately to the data were analyzed with 1-way analysis of variance followed by the cage-vertebra interface and the tissue inside the cage using the following Dunnett multiple-comparison test when significant differences were system: empty (score 0), fibrous tissue (score 1), and bone (score 2). Four observed. The 2-tailed Student t test was used for comparison of points indicate a successful fusion, and 3 points represent a developing parametric data, and the Fisher exact test was used for contingency data. fusion. Prism 5.0 (Graph Pad Software) was used for the statistical analysis. Histomorphometric analysis was performed with quantitative image Values are expressed as means and ranges unless otherwise stated. All data analysis for percentage of osteoid formation relative to mineralized bone figures show mean 6 SD unless otherwise specified. A value of P , .05 within the cage area. This was conducted on the midsagittal slice stained was considered statistically significant.

NEUROSURGERY VOLUME 68 | NUMBER 4 | APRIL 2011 | 1029

FIGURE 4. Functional radiography demonstrating the measurement of the intervertebral and lordosis angles in flexion and extension.

RESULTS 5 mm. Veterinary assessment, clinical pathology, and gross pathological and histopathological analyses showed no differences Adverse Events between groups. No cell-related adverse events were observed during the study or at postmortem. One ewe in the MPC group developed Radiographic Results pneumonia and hypoproteinemia that resolved under close vet- By CT scan at 3 months, 5 of 6 MPC-treated animals (83%) erinary treatment. Two ewes, both from the autograft group, lost had continuous bony bridging compared with no bony bridging significant weight during the study period. One ewe in the AEC in any of the AEC-treated animals (P = .01). In the control group had anterior protrusion of the cage by approximately animals, there was continuous bony bridging in 1 of 6 autograft- treated animals and 2 of 6 animals treated with HA/TCP alone (Figure 7). This finding was confirmed by objective quantitative

FIGURE 6. A 10-mm sagittal section of the C3-4 interbody cage at 3 2 magnification stained with Masson Goldner trichrome. Red and green component FIGURE 5. Quantitative computed tomography in the coronal plane showing colors are separated for quantitative histomorphometric assessment. In this ex- the region of interest measuring bone mineral density. ample, bridge of bone (left) is shown in green and osteoid (right) is shown in red.

1030 | VOLUME 68 | NUMBER 4 | APRIL 2011 www.neurosurgery-online.com

interbody space extending immediately above and below the anterior aspects of the vertebral bodies. This was found at a mean maximal distance of 6.58 mm anterior to the cage in the AEC group compared with 3.84 mm in the MPC-treated group (P = .03). Similar reactive bone formation was seen in the autograft and HA/TCP controls with a mean of 6.10 and 6.68 mm, respectively. There was no evidence of calcification within the muscles or ligaments anterior to the spine.

Biomechanical Results Significantly more stiffness at the C3-4 level was observed in the operated groups compared with nonoperated controls in FIGURE 7. Graph of computed tomography results at 3 flexion (P , .015); however, this did not achieve significance in months. APC, amnion epithelial cell; MPC, mesenchymal extension or lateral bending. No significant difference between precursor cell. operated groups was observed in any of the 4 degrees of motion.

Histological Results CT, which revealed that MPC-treated animals had significantly There was no evidence of inflammatory or neoplastic changes higher bone callus density compared with AEC-treated animals in any specimen within the fusion area as reviewed by the vet- (P = .01; Figure 9). erinary pathologist. The semiquantitative histological scoring Functional radiography scores (Figure 8) showed a reduction system revealed that 5 of 6 of the MPC-treated animals (83%) in the IVA in all operated groups compared with nonoperated had either a developing or complete fusion compared with 2 of 5 controls (P = .0003). MPC-treated animals had a significant AEC-treated animals (40%). In the controls, 3 of 6 (50%) treated reduction in IVA compared with HA/TCP (P = .01) and au- with HA/TCP alone and 1 of 6 (17%) in the autograft group had tograft (P = .03). The IVA was also significantly reduced (P = a developing or complete fusion. This finding was supported by .005) in MPC-treated animals compared with AEC-treated an- the histomorphometric analysis (Figure 10), which demonstrated imals, and there was no reduction in the IVA of AEC-treated significantly more mineralized bone in the MPC-treated group animals compared with either operated control. compared with the AEC-treated group (P = .04). There was The LA was significantly reduced in cell-treated animals a significant difference between the MPC-treated group and compared with nonoperated controls (P = .02). MPC-treated autograft-treated group (P , .0001) and between the MPC animals had a significantly reduced LA compared with HA/TCP group and the HA/TCP group (P = .06). There was, however, no (P = .02) but not compared with autograft- or AEC-treated difference between the AEC-treated group and either control. animals; in addition, LA was not reduced in AEC-treated animals At 3 weeks postoperatively, fluorochrome analysis showed compared with operated controls. more prominent deposition of mineralized bone in the MP- There was no radiological evidence of ectopic bone formation treated group compared with the AEC-treated group (P = .03). posteriorly around the spinal canal or at the neural exit foramina Although there was no significant difference between the cell- in any animals. Bone callus formation was noted anterior to the treated groups at 6 weeks, by 9 weeks, significant differences were observed for the MPC-treated group compared with the AEC- treated group (P = .05)

DISCUSSION

These results demonstrate the safety of allogeneic implantation of both human AECs and ovine MPCs when implanted ante- riorly into the cervical interbody space. The low immunogenicity and antiinflammatory properties of these cells have previously been described18,21; however, to the best of our knowledge, the present study is the first to test both cell types within the cervical interbody space adjacent to the spinal cord, trachea, esophagus, and other neck structures. Despite the safety profile common to both cell types, efficacy FIGURE 8. Graph of quantitative computed tomography in facilitating new bone formation was distinct. The multiple (qCT) results at 3 months. APC, amnion epithelial cell; MPC, mesenchymal precursor cell. modalities of assessment used in this study revealed significantly improved fusion with MPCs compared with both autograph and

NEUROSURGERY VOLUME 68 | NUMBER 4 | APRIL 2011 | 1031

FIGURE 9. Functional radiology results. Intervertebral angle (left) and lordosis angle (right). Difference between flexion and extension of intervertebral angle (degrees) is shown. APC, amnion epithelial cell; MPC, mesenchymal precursor cell.

HA/TCP controls, which represent the current standard of care, xenogeneic reactions.18,36-39 Because of the difficulties inherent and animals implanted with AECs. At the 3-month time point, with undecalcified tissue, it was not possible to definitively no animals implanted with AECs had bridging of trabecular bone identify human AECs in the histological sections in the present between the operated vertebral bodies, and the AECs appeared to study; however, the identification of human AECs in tissues in have a negative effect on fusion. The fusion results with MPCs other animal models and the lack of inflammation at the peri- were similar to those reported for growth factors, including bone operative site in the present study suggest that the inability of morphogenetic protein, when applied to an ovine cervical model AECs to contribute to interbody fusion is unlikely to have at the 3-month end point.28 Other studies in lumbar postero- a xenogeneic basis. lateral or interbody spinal fusion have shown superior effects with The different techniques used for isolation of human AECs bone morphogenetic proteins.35 and MPCs could account for their difference in efficacy. The Human amnion-derived cells have been evaluated in a variety monoclonal antibody selection technique used for MPC isolation of animal models of human disease without any evidence of yields a homogeneous population of cells with batch-to-batch consistency.40 In contrast, enzymatic digestion, followed by density gradient and plastic adherence selection of the AECs, yields a heterogeneous population of cells, which introduces variability between batches.20,41 Although the cell marker characterization of the batch of cells used in the present study (Figure 1) closely resembles the characteristics of other batches of AECs and is similar to the cell surface marker profile defining mesenchymal stem cell populations,42 it is possible that different results may have been obtained with another batch of AECs. Studies by Bilic et al20 have demonstrated different outcomes in osteogenic differentiation of AECs between batches. Further work in identifying the characteristics of those AECs with highest osteogenic potential and then improving selection of this sub- population should be undertaken before further testing this cell type for the present indication. It is also possible that a higher does of AECs in vivo may have a superior effect on fusion. FIGURE 10. Graph of histomorphometric results of min- Alternatively, it is possible that AECs may be inhibitory to fusion eralized bone volume (Md. V) within the cage at the altogether. Should this be the case, a fusion inhibition effect midsagittal section stained with Mason Goldner trichrome. would be beneficial in other applications such as motion pres- APC, amnion epithelial cell; MPC, mesenchymal precursor 43 cell. ervation. Further studies with AECs in these applications are underway.

1032 | VOLUME 68 | NUMBER 4 | APRIL 2011 www.neurosurgery-online.com

The bone formation observed anterior to the cage was similar a sponsored research agreement from Mesoblast Ltd. Author Itescu is employee to that shown in previous preclinical studies.28,33,44 Zdeblick and has stock in Mesoblast. Author Ghosh is consultant and has stock in Mes- et al33 suggested that the countersinking of the cages, as was oblast. Author Zannettino is consultant to Mesoblast. Author Goldschlager has received travel support from Mesoblast. undertaken in the present study, results in exposed reamed bone lacking cartilaginous endplates that acts as a source for new bone formation. It is also possible, however, that the mechanism of this REFERENCES bone formation is related, in part, to the instability of the ver- 1. Ryken TC, Heary RF, Matz PG, et al. Techniques for cervical interbody grafting. tebral joints, as has been invoked for the development of os- J Neurosurg Spine. 2009;11(2):203-220. 45 2. Kalfas IH. Principles of bone healing. Neurosurg Focus. 2001;10(4):E1. teophytes in cervical spondylosis. In the present study, there 3. Pilitsis JG, Lucas DR, Rengachary SS. Bone healing and spinal fusion. Neurosurg was significantly more anterior bone formation in the AEC group Focus. 2002;13(6):e1. compared with the MPC group. The former group failed to fuse, 4. Chau AM, Mobbs RJ. Bone graft substitutes in anterior cervical discectomy and and functional assessment revealed increased mobility, which may fusion. Eur Spine J. 2009;18(4):449-464. 5. Silber JS, Anderson DG, Daffner SD, et al. Donor site morbidity after anterior iliac account in whole or in part for this difference. crest bone harvest for single-level anterior cervical discectomy and fusion. Spine The beneficial effects on fusion resulting from the MPCs (Phila Pa 1976). 2003;28(2):134-139. combined with HA/TCP observed in this study are possibly due 6. Crawford CH III, Carreon LY, Djurasovic M, Glassman SD. Lumbar fusion to 2 separate mechanisms of action: direct osteogenic differen- outcomes in patients with rheumatoid arthritis. Eur Spine J. 2008;17(6):822-825. 46,47 7. Mooney V, McDermott KL, Song J. Effects of smoking and maturation on tiation of the MPCs at the fusion site or a paracrine effect on long-term maintenance of lumbar spinal fusion success. J Spinal Disord. endogenous osteogenic precursors because the MPCs secrete 1999;12(5):380-385. many growth factors, including bone morphogenetic proteins.15 8. Andersen T, Christensen FB, Laursen M, Hoy K, Hansen ES, Bunger C. Smoking as a predictor of negative outcome in lumbar spinal fusion. Spine (Phila Pa 1976). In combination with the HA/TCP, MPCs are therefore likely 2001;26(23):2623-2628. osteogenic, osteoconductive, and osteoinductive. Although au- 9. Lumawig JM, Yamazaki A, Watanabe K. Dose-dependent inhibition of diclofenac tograft also possesses these 3 attributes, it is only weakly osteo- sodium on posterior lumbar interbody fusion rates. Spine J. 2009;9(5):343-349. genic and osteoinductive.4 The low number of osteogenic 10. Kuhns CA, Geck MJ, Wang JC, Delamarter RB. An outcomes analysis of the treatment of cervical pseudarthrosis with posterior fusion. Spine (Phila Pa 1976). precursor cells or stem cells contained within fresh autogenous 2005;30(21):2424-2429. 15 grafts is likely to account for the difference in fusion outcome 11. Baskin DS, Ryan P, Sonntag V, Westmark R, Widmayer MA. A prospective, compared with MPCs at this time point. randomized, controlled cervical fusion study using recombinant human bone In addition to these fusion-enhancing properties, the immu- morphogenetic protein-2 with the CORNERSTONE-SR allograft ring and the ATLANTIS anterior cervical plate. Spine (Phila Pa 1976). 2003;28(12):1219-1224. nological properties of the MPCs may have additional benefits. 12. Leach J, Bittar RG. BMP-7 (OP-1) safety in anterior cervical fusion surgery. J Clin MPCs lack the ability to induce an allogeneic mixed lymphocyte Neurosci. 2009;16(11):1417-1420. reaction21,48 and secrete multiple antiinflammatory and immu- 13. Food and Drug Administration. FDA public health notification: life-threatening 21 complications associated with recombinant human bone morphogenetic protein in nosuppressive cytokines such as interleukin-10. These antiin- cervical spine fusion. US Food and Drug Administration, Center for Devices and flammatory properties could conceivably reduce radiculopathy or Radiological Health. Accessed July 2008. http://www.fda.gov/cdrh/safety/ myelopathy, which are the common underlying problems ne- 070108-rhbmp.html. cessitating cervical fusion. The potential beneficial properties of 14. Liu G, Zhao L, Zhang W, Cui L, Liu W, Cao Y. Repair of goat tibial defects with bone marrow stromal cells and beta-tricalcium phosphate. J Mater Sci Mater Med. MPCs reported in this animal study support further evaluation in 2008;19(6):2367-2376. a human clinical trial. 15. Arthur A, Zannettino A, Gronthos S. The therapeutic applications of multipo- tential mesenchymal/stromal stem cells in skeletal tissue repair. J Cell Physiol. 2009;218(2):237-245. CONCLUSION 16. NIH. Study of 3 doses of NeoFuse combined with MasterGraft Granules in subjects requiring posterolateral lumbar fusion (PLF). Accessed 2007. http:// MPCs facilitated a more robust fusion relative to controls clinicaltrials.gov/ct2/show/NCT00549913. representing the current standard of care in this study. In con- 17. Ilancheran S, Moodley Y, Manuelpillai U. Human fetal membranes: a source of stem cells for tissue regeneration and repair? Placenta. 2009;30(1):2-10. tradistinction, AECs reduced fusion. The osteogenic, osteo- 18. Bailo M, Soncini M, Vertua E, et al. Engraftment potential of human amnion and conductive, and osteoinductive potentials of MPCs obviate the chorion cells derived from term placenta. Transplantation. 2004;78(10):1439- donor-site morbidity associated with autograft harvest and the 1448. potential life-threatening complications associated with the use of 19. Ilancheran S, Michalska A, Peh G, Wallace EM, Pera M, Manuelpillai U. Stem cells derived from human fetal membranes display multilineage differentiation bone morphogenetic proteins. The potential mechanisms by potential. Biol Reprod. 2007;77(3):577-588. which MPC mediated spinal fusion could include promotion of 20. Bilic G, Zeisberger SM, Mallik AS, Zimmermann R, Zisch AH. Comparative osteoblast differentiation and their secretion of trophic factors characterization of cultured human term amnion epithelial and mesenchymal stromal cells for application in cell therapy. Cell Transplant. 2008;17(8):955-968. such as multiple bone morphogenetic proteins and other growth 21. Tyndall A, Walker U, Cope A, et al. Immunomodulatory properties of mesen- factors, which are known to upregulate osteogenesis. chymal stem cells: a review based on an interdisciplinary meeting held at the Kennedy Institute of Rheumatology Division, London, UK, 31 October 2005. Disclosure Arthritis Res Ther. 2007;9(1):301-311. 22. Kandziora F, Pflugmacher R, Scholz M, et al. Comparison between sheep and This work was partially funded through project grant 491145 from the human cervical spines: an anatomic, radiographic, bone mineral density, and National Health and Medical Research Council of Australia and partially through biomechanical study. Spine (Phila Pa 1976). 2001;26(9):1028-1037.

NEUROSURGERY VOLUME 68 | NUMBER 4 | APRIL 2011 | 1033

23. Gronthos S, Fitter S, Diamond P, Simmons PJ, Itescu S, Zannettino AC. A novel Paper presented at: Orthopedic Research Society 55th Annual Meeting; 2009; Las monoclonal antibody (STRO-3) identifies an isoform of tissue nonspecific alkaline Vegas, NV. phosphatase expressed by multipotent bone marrow stromal stem cells. Stem Cells 47. Goldschlager T, Itescu S, Ghosh P, et al. Cervical interbody fusion is enhanced by Dev. Dec 2007;16(6):953-963. allogeneic mesenchymal precursor cells in an ovine model. Spine (Phila Pa 1976). 24. McCarty RC, Gronthos S, Zannettino AC, Foster BK, Xian CJ. Characterisation December 9, 2010. DOI: 10.1097/BRS.0b013e3181dfcec9; Accessed December and developmental potential of ovine bone marrow derived mesenchymal stem 29, 2010. ,http://meta.wkhealth.com/pt/pt-core/template-journal/lwwgateway/ cells. J Cell Physiol. 2009;219(2):324-333. media/landingpage.htm?doi=10.1097/BRS.0b013e3181dfcec9. 25. Goldschlager T, Abdelkader A, Kerr J, Boundy I, Jenkin G. Undecalcified bone 48. Ryan JM, Barry FP, Murphy JM, Mahon BP. Mesenchymal stem cells avoid preparation for histology, histomorphometry and fluorochrome analysis. J Vis Exp. allogeneic rejection. J Inflamm (Lond). 2005;2:8-19. 2010 Jan 8;(35). pii: 1707. doi: 10.3791/1707. 26. Burkus JK, Dorchak JD, Sanders DL. Radiographic assessment of interbody fusion Acknowledgments using recombinant human bone morphogenetic protein type 2. Spine (Phila Pa 1976). 2003;28(4):372-377. We would like to thank Drs Anne Gibbon and Christine Mackay for their 27. Buchowski JM, Liu G, Bunmaprasert T, Rose PS, Riew KD. Anterior cervical professional work throughout this study. We also wish to thank Toula Papadi- fusion assessment: surgical exploration versus radiographic evaluation. Spine (Phila mitrakis, Jill McFadyean, Alex Satrango, Dr George Thouas, Jeremy Weinstein, Pa 1976). 2008;33(11):1185-1191. and Raphael Weidenfield for their assistance during this study; Dr Amany Ab- 28. Kandziora F, Pflugmacher R, Scholz M, et al. Comparison of BMP-2 and delkadar and Ian Boundy for histological preparation; and Professor Frank combined IGF-I/TGF-ss1 application in a sheep cervical spine fusion model. Eur Kandziora and his team for their expert advice. We wish to acknowledge the Spine J. 2002;11(5):482-493. 29. Kandziora F, Schmidmaier G, Schollmeier G, et al. IGF-I and TGF-beta1 application advice and resources of Monash Micro Imaging. by a poly-(D,L-lactide)-coated cage promotes intervertebral bone matrix formation in the sheep cervical spine. Spine (Phila Pa 1976). 2002;27(16):1710-1723. 30. Gal JM. Mammalian spinal biomechanics, I: static and dynamic mechanical COMMENTS properties of intact intervertebral joints. J Exp Biol. 1993;174(1):247-280. 31. Goldschlager T, Thouas G, Edward G, Weinstein J, Jenkin G. A novel and simple ultiple materials have been studied to replace the use of autograft method for assessing spinal biomechanics. Paper presented at: SGI 56th A.S.M– Reproductive Bioengineering–Modelling; 2008; Glasgow, Scotland. M implanted by itself or in a vehicle in the interbody space after 32. Goldschlager T, Rosenfeld JV, Young IR, Jenkin G. Anterior cervical discectomy cervical diskectomy. As the authors mention, autograft has many ad- and fusion in the ovine model. J Vis Exp. 2009 Oct 5;(32). pii: 1548. doi: vantages, but obtaining the graft is associated with its own disadvantages. 10.3791/1548. Consequently, the authors studied mesenchymal precursor cells (MPCs) 33. Zdeblick TA, Ghanayem AJ, Rapoff AJ, et al. Cervical interbody fusion cages: an and amnion epithelial cells (AECs) placed in interbody cages. They animal model with and without bone morphogenetic protein. Spine (Phila Pa studied 5 groups of 2-year-old sheep undergoing the following treat- 1976). 1998;23(7):758-765. 34. Parfitt AM, Drezner MK, Glorieux FH, et al. Bone histomorphometry: stan- ments: (1) iliac autograft placed alone in an interbody cage (n = 6), (2) dardization of nomenclature, symbols, and units: report of the ASBMR Histo- hydroxyapatite/tricalcium phosphate (HA/TCP) (n = 6), (3) HA con- morphometry Nomenclature Committee. J Bone Miner Res. 1987;2(6):595-610. taining MPC (n = 6), (4) HA and TCP containing AECs (n = 5), and (5) 35. Carlisle E, Fischgrund JS. Bone morphogenetic proteins for spinal fusion. Spine J. HA nonoperative control (n = 6). 2005;5(6)(Suppl):240S-249S. At 3 months, the animals were euthanized, and the spines were 36. De Coppi P, Bartsch G Jr, Siddiqui MM, et al. Isolation of amniotic stem cell lines studied with computed tomography, functional radiography, spinal with potential for therapy. Nat Biotechnol. 2007;25(1):100-106. mechanics, histology, and histomorphometry. Implantation of the al- 37. Wei JP, Zhang TS, Kawa S, et al. Human amnion-isolated cells normalize blood glucose in streptozotocin-induced diabetic mice. Cell Transplant. 2003;12(5):545-552. logeneic MPCs in combination with the HA/TCP in the interbody 38. Kakishita K, Nakao N, Sakuragawa N, Itakura T. Implantation of human amniotic spacer facilitated interbody fusion. Interestingly, no fusion occurred in epithelial cells prevents the degeneration of nigral dopamine neurons in rats with the AEC group. Consequently, the authors suggest that MPCs in an 6-hydroxydopamine lesions. Brain Res. 2003;980(1):48-56. interbody cage might be used as an alternative for autograft. The authors 39. Kakishita K, Elwan MA, Nakao N, Itakura T, Sakuragawa N. Human amniotic found no undue reactions to either the MPCs or AECs. epithelial cells produce dopamine and survive after implantation into the striatum These observations are quite interesting. It is to be hoped that further of a rat model of Parkinson’s disease: a potential source of donor for transplantation studies by these authors or others will clarify the benefit of MPCs in therapy. Exp Neurol. 2000;165(1):27-34. 40. Gronthos S, Zannettino AC, Hay SJ, et al. Molecular and cellular characterisation promoting fusion. of highly purified stromal stem cells derived from human bone marrow. J Cell Sci. Volker K. H. Sonntag 2003;116(pt 9):1827-1835. 41. Dominici M, Paolucci P, Conte P, Horwitz EM. Heterogeneity of multipotent Phoenix, Arizona mesenchymal stromal cells: from stromal cells to stem cells and vice versa. Transplantation. 2009;87(9)(Suppl):S36-S42. oldschlager et al used an ovine model of anterior cervical dis- 42. Dominici M, Le Blanc K, Mueller I, et al. Minimal criteria for defining multi- kectomy and fusion at C3-4 to evaluate the capacity of allogeneic potent mesenchymal stromal cells: the International Society for Cellular Therapy G position statement. Cytotherapy. 2006;8(4):315-317. mesenchymal precursor stem cells (MPCs) and human allogeneic am- 43. Goldschlager T, Ghosh P, Zannettino A, et al. Cervical motion preservation using nion epithelial cells (AECs) in promoting cervical fusion compared with mesenchymal progenitor cells and a novel chondrogenic agent, pentosan poly- autograft or Mastergraft alone. Polyetheretherketone cages of the same sulfate: a preliminary study in an ovine model. Neurosurg Focus. 2010;28(6):E4. size were used in all experimental groups. The authors monitored all 44. Slivka MA, Spenciner DB, Seim HB, Welch WC, Serhan HA, Turner AS. High rate animals for adverse effects such heterotopic bone formation, neoplastic of fusion in sheep cervical spines following anterior interbody surgery with absorbable transformation of transplanted cells, immunogenic reaction of host and nonabsorbable implant devices. Spine (Phila Pa 1976). 2006;31(24):2772-2777. 45. Roh JS, Teng AL, Yoo JU, Davis J, Furey C, Bohlman HH. Degenerative disorders animals, and surgical complications and used radiographic, bio- of the lumbar and cervical spine. Orthop Clin North Am. 2005;36(3):255-262. mechanical, and histological assessment of bone fusions. 46. Goldschlager T, Itescu S, Ghosh P, et al. Allogeneic mesenchymal precursor cells The results demonstrate an excellent fusion-promoting capacity of safely and effectively increase the rate and robustness of cervical interbody fusion. the MPC-Mastergraft combination in contrast to lack of fusion in the

1034 | VOLUME 68 | NUMBER 4 | APRIL 2011 www.neurosurgery-online.com

AEC-treated animals. There were corresponding biomechanical changes genetic proteins in promoting anterior cervical fusion without the in all fusion animals, especially in the MPC-treated animals, which had complications encountered with the use human bone morphogenetic a significant reduction in intervertebral and lordotic angles, although proteins. Further studies are required to understand the molecular change in stiffness was not significant. Histological assessment of fusion mechanisms of the fusion-promoting properties of the MPCs compared also confirmed the superiority of MPC treatment with regard to evidence with AECs. of bone fusion. This well-done study looks at the important question of whether Wale Sulaiman allogeneic stem cells are viable alternatives to human bone morpho- New Orleans, Louisiana

NEUROSURGERY VOLUME 68 | NUMBER 4 | APRIL 2011 | 1035