**Defining the Abnormal Kinematics of Lumbar Spine Instability as a Cause of Degenerative Low Back Pain-A Biomechanical Study of the Kinematics of Cadaver Lumbar Spine
Dilip K. Sengupta, MD

Purpose: Chronic low back pain (LBP) is a common musculoskeletal disorder that significantly impacts public health. However the mechanism of chronic LBP is still not fully understood. From a biomechanical point of view, it is believed that chronic LBP is related to spinal instability (abnormal motion), abnormal load transmission, or both in the degenerated motion segment. But the relationship between these factors and back pain is again not clear. There remains a need for more effective biomechanical measures that are able to characterize spine motion and load transmission, and clarify how these factors relate to pain generation. The long term goal of this study is to find effective biomechanical measures for characterizing spinal motion and load transmission, and use these measures to elucidate the biomechanical mechanism of chronic low back pain. The objective of this research is to validate in vitro that instantaneous axis of rotation (IAR) is an effective measure to define abnormal spinal motion (spinal instability), and intra-discal pressure profile (IDPP) an effective measure to define abnormal load-transmission; and further to understand the relationship between IAR and IDPP.

Hypotheses:1) Instantaneous axis of rotation (IAR) trace patters are different between normal and degenerated discs. 2) The load transmission across the disc space, as measured using intra-discal pressure profile (IDPP), is different between normal and degenerated discs. 3) There is a relationship between IAR and the load-transmission across the disc space.

Methods: An in vitro study of lumbar spine kinematics and load transmission is proposed to achieve our study aims. Ten human cadaver lumbar spine specimens will be procured. MRI scans of these spines will be performed to determine the degree of disc degeneration. Lateral radiographs of the specimens will be obtained to measure disc height, lordosis, listhesis, endplate sclerosis, etc. The specimens will then be potted with two motion segments in each potting for load-deformation tests on a six-degrees-of-freedom (6-DOF) spine tester. IAR will be determined from motion tracking data and hundreds of instances of IAR locations (continuous tracking of the locations of the IAR) will be calculated during one flexion-extension cycle. Disc pressures will be determined using two methods, with IDPP as the primary approach, to identify any abnormal load-transmission across the disc space. The two methods for measuring disc pressures are: 1)IDPP, as a continuous pressure profile across the disc space with the specimens held statically in flexion, extension and neutral posture under load-bearing; and 2)disc pressure graph, which is determined during dynamic flexion-extension motion from the center of the disc space.  Finally, both qualitative and statistical comparison of IAR and IDPP will be performed based on degree of disc degeneration to understand their relationships.

1. Findings

1.1. Project summary
We have conducted tests on 40 discs from 14 spine specimens. Before tests, MRI was conducted on specimens and degenerated discs were graded using a clinical disc degeneration grading system as described by Pfirrmann et al (2001). Discs were divided into five groups: normal (grade I), minor (grade II), mild (grade III), moderate (grade IV), and severe degeneration (grade V). Table 1 summarizes degeneration grade distribution of tested discs.

The majority of tested discs were normal (grade I), and only one grade V disc was available. It was found obtaining higher degeneration grade discs (Grade V) was difficult, because of their low availability and the frequent spontaneous fusion of such motion segments. However, we obtained sufficient numbers of discs for Grade I~IV. Clinically, symptomatic discs with degeneration equal or greater than grade III are surgically treated, therefore our tested Grade III and IV discs should be sufficient to answer our research question.

In the following sections, IAR, ROM, NZ, IDPP, and disc pressure graph (IDP) data are presented.   

1.2. Instantaneous Axis of Rotation (IAR)
Potted specimens were loaded onto a 6-DOF spine tester for load-deformation tests in flexion-extension direction. During tests, motion was captured with a motion tracking system and IAR was calculated from the continuous motion tracking data. After tests, lateral X-ray radiograph of the specimen was obtained to record the reference positions of motion tracking markers. IAR was then superimposed to the radiographs to reveal anatomic location of the IAR relative to the motion segment.

Figure 1 and Figure 2 are IAR for two typical normal discs (Grade I). Figure 3 and Figure 4 are IAR for two typical Grade II disc. Figure 5 and Figure 6 are IAR for two typical Grade III discs. Figure 7 and Figure 8 are IAR for two typical Grade IV disc. Figure 9 is the IAR result for one Grade V disc.

In summary, extensive tests have been conducted and IAR of 40 motion segments with various degrees of degeneration have been obtained.  Baseline of IAR patterns of normal discs has been established. The IAR results of degenerated discs will be analyzed and compared with normal IAR. Further results will be submitted in the final report this project.

1.3. Intra-Discal Pressure Profile (IDPP)
To measure IDPP, the specimens were held statically in flexion, extension and neutral posture under load-bearing. A needle sensor was inserted through the disc until the tip of the needle sensor is beyond the posterior annulus. Disc pressure was then recorded when the needle sensor was drawn back through the disc. Figure 10 through Figure 13 are representative IDPP results of discs with degeneration grades I though IV. For each grade, IPDD of two discs are presented.

For Grade I normal discs, IDPP profile showed relatively uniform disc pressures in the center of the discs in all loading conditions.  With increasing degeneration, IDPP profiles become more and more irregular, showing multiple pressure spikes in the annulus region in Figure 11, Figure 12 and Figure 13. Two of the 4 discs in Figure 11 and Figure 13 showed a nearly “depressurized” state in some of the loading postures, and lost completely the uniform pressure plateau of normal discs. Such pressure profiles were consisted of regions of near-zero pressures and regions of pressure spikes, indicating significant disc structural changes such as nucleus dehydration and structure disruption. The upper profiles in Figure 13 showed higher pressures in annulus region than nucleus region, and also “broadened” annulus pressure region. Such profile was also observed in McNally et al’s IDPP studies.

These pressure profiles indicate that with increasing disc degeneration grade, IDPP showed a trend to shift from the behavior of “water-filled fluid bag” as represented by uniform pressures in nucleus region, to the behavior of dehydrated and disrupted disc structure as represented by pressure spikes and depressurized states.

Figure 14 shows overall magnitude changes of disc nucleus pressures, measured from IDPP. It is found the nucleus pressure continues to drop with increasing grade of disc degeneration. Despite of inherent large variation of disc pressure measurements, statistical significance were found in all modes of load in Grade I vs. Grade IV discs, confirming the trend of decrement.

Grade I (specimen 61785)

 Grade II, (specimen 61960, 61317)

Grade III, (specimen 63160, 61960)

Grade IV, (specimen 63160)

1.4. Range of Motion (ROM) and Neutral Zone (NZ)
Rang of motion (ROM) was measured during load-deformation tests for each specimens. Figure 15 through Figure 18 are representative load-deformation curves for normal discs and discs with degeneration. The profiles of load-deformation curves, e.g. non-linearity, did not show significant changes in different degrees of degeneration.

Figure 19 shows the ROM comparison between different degeneration grades. There are some differences in ROM with increasing disc degeneration, but such differences did not demonstrate statistical significance. 

Figure 20 shows the NZ comparison between different disc degeneration grades. NZ was calculated from Hysteresis width as defined by Wilke et al in 1998. In general, NZ increases with disc degeneration from grade I through grade III, in flex-ext, lateral bending and axial rotation, which is likely caused by increasing laxity of spinal ligaments and disc annulus, due to degeneration. Grade IV discs showed smaller NZ comparing to grade III disc in lateral bending and axial rotation, indicating stabilizing effect of advanced degeneration. Statistical significance was found in flexion of grade I vs. grade III, and grade I vs. grade IV discs.

It is interesting to note that there was tendency to reduce the ROM in Grade IV, but the NZ motion increased. It may be important that NZ in % of total ROM be studied to reveal the dramatic difference that degeneration may introduce.

1.5. Disc Pressure Graphs
Disc pressure graphs (IDP) were measured using a miniature pressure sensor in the center of each disc during load-deformation tests. Figure 22 through Figure 24 are representative disc pressure graphs for normal discs and discs with various degeneration grades. With increasing degeneration, the IDP showed a trend of change from “U” shape profiles of normal disc (pressure peaks at both full flexion and extension posture) to profiles that frequently lost their pressure peaks (Figure 23, Figure 24).

Figure 25 compares disc pressure graph rise magnitude among disc groups. Although statistical significance was only found in extension of grade I-IV due to large pressure variations, with increasing degeneration, a trend of decreasing disc pressure appears to be consistent, both in flexion and extension.

Grade III, (specimen 63160, 61960)

1.6. Summary of Findings to Date:
This study defines the normal parameters of the IAR, ROM, NZ, IDP rise, and IDPP profile across the disc space. Amongst the four parameters, the IDP and IDPP profile showed consistent changes with increasing disc degeneration. Results of IAR and ROM and NZ will be further analyzed to identify any additional linkages between these parameters and disc degeneration.

Young Investigator:  Novel Scaffold Using Human Adipose-derived Stromal Cells
Francis H. Shen, MD

Abstract
Over 450,000 spinal fusions are performed in the United States a year.  Despite improved techniques and instrumentation, nonunion rates continue to occur in up to 40% of cases.  Iliac crest bone graft is considered the gold standard, however is not without significant morbidity.   Recent studies suggest that adipose-derived stromal (ADS) cells are more than space fillers and, under appropriate conditions, can dedifferentiate and return to the status of an uncommitted stromal stem cell.    The purpose for this research initiative is to determine the potential for hADS cells treated with recombinant human growth and differentiation factor-5 protein (rGDF-5) to undergo osteogenic differentiation on a novel bioengineered scaffold in vitro. 

We studied the osteogenesis of hADS cells in vitro with the induction of rGDF-5. The osteogenesis differentiation was compared with von Kossa staining after treatment with osteogenic medium.  The gene expressions were analyzed by real-time PCR.  The cells cultured with different concentrations of GDF-5 were measured for an optimized application of the growth factor. Gene expression confirmed the osteogenic differentiation of hADS cells.  Furthermore, real-time PCR revealed that the gene expression of VEGF groups demonstrated different reactions to the treatment of ODM with GDF-5 and confirmed the relationship between VEGF and BMP signaling pathways. This suggests that future studies investigating cell signal pathways during the stem cell differentiation may hold significant benefit for the investigation of spinal fusions and osteogenesis.

These experiments also investigated the ability of hADS cells to adhere and proliferate on novel bioengineered scaffolds in vitro.  The sintered microsphere matrix (SMM) were fabricated using the solvent evaporation technique.  Using micron sieves, selected diameter size ranges of 500–600 mm were packed in a stainless steel mold and heated at 80 oC.  The hADS cells demonstrated a good compatibility and similar level of osteogenesis on PLGA scaffold compared with monolayer culture. Next, nanofiber PLGA scaffold were developed utilizing a flexible nanofiber PLGA scaffold, which could be bent while maintaining its strength.   The hADS cells were then co-cultured on the spun nanofibers (fiber sizes, 700-1000nm), while the cell density was 2000cells/scaffold.  Using SEM, the results demonstrate both cell viability and growth between nanofibers. 

Results of SMM and spun nanofibers suggest that fabrication of these novel bioengineered scaffolds are feasible using PLGA and currently available tissue engineering techniques in our laboratory.  Furthermore, human ADS cells appear to be compatible on the bioengineered, biodegradable scaffold and that growth appears viable in an in vitro environment.  Moreover, this nanofiber scaffold has the capability to be combined with the microsphere scaffold for use during in vivo bone formation in the spine.  The ability for differentiation and migration of hADS cells on these novel scaffolds are still unclear and warrant further investigation.  Furthermore, extensive in vivo experiments will be required in order to further confirm biocompatibility between scaffold and osteogenic cells and ability to undergo a spinal fusion. 

  
**Young Investigator:  Intervertebral Disc Regeneration from Co-cultured Disc and Stem Cells in Biomimetic Engineered Extracellular Matrix Stimulated by Mechanically Active Bioreactor
Wan-Ju Li, Phd; Paul A. Anderson, MD

Status of Research Progress
The project started in August 2008, and due to uncontrollable incidents (laboratory relocation and unavailability of new culture facility), the project actually did not start until April 2009. We were kindly granted with a year extension by NASS research committee.

During the extended year, we have made significant progress for this project; we have accomplished the specific aim 1, and are working toward to finishing up the specific aim 2. Specially, we have optimized fabrication of the IVD biomaterial scaffold, and successfully produced working scaffolds for IVD regeneration as we proposed in the specific aim 2. We also investigated human mesenchymal stem cell (hMSC), annulus fibrosus cell (AFC) and nucleus pulposus cell (NPC) co-culture and determined the optimized ratios of hMSCs/hAFCs and hMSCs/hNPCs for regenerating AF and NP tissues.

In the proposed study plan for the specific aim 2, using the multi-specimen BioDynamic bioreactor to enhance engineered IVD properties is the most critical component of the study. The powerful bioreactor is a custom-made instrument by Bose Corporation. Since the instrument is a large item purchase (> $120,000) and requires a great number of hours to build, it took 5 months to approve the purchase and then fabricate the instrument. By June 2010, we finally have the running bioreactor in our laboratory for this project.

Because of the delay of the bioreactor availability for this project, we would like to ask NASS research committee to consider granting us with another year extension. With the additional time, we will be able to accomplish the specific aim 2, finish up the project, and collect convincible results for NIH funding opportunities and publications.

Findings to Date
We proposed to accomplish two specific aims; Specific Aim 1 is to construct a structurally mimicking IVD biomaterial scaffold, and Specific Aim 2 is to regenerate biologically and biomechanically functioning IVD in a bioreactor. We have successfully accomplished the Specific Aim 1, and are currently studying the Specific Aim 2.

Fabrication of engineered IVD scaffolds
The fabricated IVD scaffold is a complex, dual-architectural structure that mimics the anatomic structure of a native IVD; layers of alternatively oriented fibers resembling AF enclose a gel substance as NP. To fabricate the scaffold, we fabricated aligned poly(-caprolactone) (PCL) nanofibers using the electrospinning technique, following the experimental steps published previously (Fig. 1). Aligned nanofibrous mats were collected from the surface of the shaft after 8 hours of electrospinning. 

Figure 1. Setup of electrospinning apparatus for aligned nanofibers. Aligned nanofibrous mats were collected from the setup with a rotary metal shaft.

The size of 15.7 x 1 cm of nanofibrous strips were prepared by cutting the collected, 1-mm thick aligned nanofibrous mats according to the cutting patterns described in Figure 2. Two types of strips were made: one with aligned nanofibers obliquely oriented from the right to the left, and the other with nanofibers obliquely oriented from the left to the right.

Figure 2. Cutting patterns for preparation of strips with obliquely oriented nanofibers
 
The two nanofibrous strips with different fiber orientation were stacked and rolled to form a concentric ring with a hollow space in it. The structure was 1.5 cm in diameter and 1.00 cm in height (Fig. 3A). A hyaluronan-based hydrogel was added into the open space of the ring structure to form a composite structure with nanofiber layers (AF) outside and a hydrogel (NP) inside (Fig. 3B). Finally, the composite structure was closed up by nanofibrous layers on the top and at the bottom to form an IVD scaffold (Fig. 3C) for tissue regeneration.        

  
 
Figure 3. (A) A concentric, hollow ring made of nanofibrous strips. (B) A composite structure with outer layers of nanofibers and an inner core of hydrogel. (C) A constructed IVD scaffold with nanofiber layers covering the composite structure.

Optimization of hMSC, hAFC, and hNPC co-culture
Human MSCs and hAFCs or hMSCs and hNPCs at the ratio of 1:1 mixed and cultured in scaffolds for AF and NP regeneration was originally proposed. However, based on Dr. James Kang’s recent finding published in Spine Journal, he used rabbit cells and suggested the cell ratio has effects on cell behavior (Spine J. 2008). To investigate the effect with human cells, the study of optimization of hMSC, hAFC, and hNPC co-culture was carried out. In this study, MSCs were isolated from human bone marrow, AFCs from human AF, and NPCs from human NP. The human tissues were obtained from surgery with the IRB approval from University of Wisconsin-Madison. These isolated cells were expanded in tissue culture flasks, collected, and replated into different scaffolds for AF and NP regeneration; hMSCs/hAFCs with different ratios were seeded onto nanofibrous strips whereas hMSCs/hNPCs with different ratios were seeded into hyaluronan hydrogels. Both cell-laden scaffolds were cultured in chondrogenic medium for 21 days and collected for real-time RT-PCR analysis.

The real-time RT-PCR result showed that at Day 21, the hMSC culture expressed the highest level of collagen type I mRNA, followed by the 1:1 (hNPC:hMSC) culture, the 1:2 (hNPC:hMSC) culture, and then the hNPC and 2:1 (hNPC:hMSC) culture (Fig. 4A). On the other hand, the 1:2 culture showed the highest level of collagen type II mRNA, followed by the 1:1, MSC, 2:1, and NP culture (Fig. 4B). For the expression of the transcription factor, Sox9, critical for chondrogenesis regulation, the 1:1 culture outperformed all other culture (Fig. 4C). These results suggest that hNPCs and hMSCs were mixed at the ratio of 1:1 or 1:2 culture was effective for chondrogenesis. Notably, the ratio of collagen types II to I in engineered cartilage is a key property of hyaline cartilage. The collagen type II to I ratio in the 1:2 culture is higher than that in the 1:1 culture. Therefore, hNPCs will be mixed with hMSCs at the ratio of 1:2 for NP regeneration.

Figure 4. Real time RT-PCR analysis of hNPC and hMSC co-cultured in hyaluronan hydrogels. (A) Collagen type I; (B) Collagen Type II; (C) Sox9.

The multi-specimen BioDynamic bioreactor
The bioreactor was custom-made by Bose Corporation and installed in our laboratory (Fig. 5A). The bioreactor system (Fig. 5B) was precisely controlled and operated by a computer to culture cell-laden scaffolds. Fabricated scaffolds described in Figure 3 will be loaded with co-cultured hMSCs, hAFCs, and hNPCs described in Figure 4 and then cultured in bioreactor chambers (Fig. 5C).        

Figure 5. (A) The multi-specimen BioDynamic bioreactor and control system. (B) The multi-specimen BioDynamic test instrument. (C) A bioreactor chamber with a specimen holder.    

  *Abstracts/permission forms not received at the time of publication
**Current and ongoing research