Effect of prosthesis endplate lordosis angles on L5-S1 kinematics after disc arthroplasty.

OBJECTIVE: We hypothesized that L5-S1 kinematics will not be affected by the lordosis distribution between the prosthesis endplates. MATERIALS AND METHODS: Twelve cadaveric lumbosacral spines (51.3 ± 9.8 years) were implanted with 6° or 11° prostheses (ProDisc-L) with four combinations of superior/inferior lordosis (6°/0°, 3°/3°, 11°/0°, 3°/8°). Specimens were tested intact and after prostheses implantation with different lordosis distributions. Center of rotation (COR) and range of motion (ROM) were quantified. RESULTS: Six-degree lordosis prostheses (n = 7) showed no difference in flexion-extension ROM, regardless of design (6°/0° or 3°/3°) (p > 0.05). In lateral bending (LB), both designs reduced ROM (p < 0.05). In axial rotation, only the 3°/3° design reduced ROM (p < 0.05). Eleven-degree lordosis prostheses (n = 5) showed no difference in flexion-extension ROM for either design (p > 0.05). LB ROM decreased with distributed lordosis prostheses (3°/8°) (p < 0.05). Overall, L5-S1 range of motion was not markedly influenced by lordosis distribution among the two prosthesis endplates. The ProDisc-L prosthesis design where all lordosis is concentrated in the superior endplate yielded COR locations that were anterior and caudal to intact controls. The prosthesis with lordosis distributed between the two endplates yielded a COR that tended to be closer to intact. CONCLUSIONS: Further clinical and biomechanical studies are needed to assess the long-term impact of lordosis angle distribution on the fate of the facet joints.

Kinematics of cervical total disc replacement adjacent to a two-level, straight versus lordotic fusion.

STUDY DESIGN: In vitro biomechanical study. OBJECTIVE: To characterize cervical total disc replacement (TDR) kinematics above two-level fusion, and to determine the effect of fusion alignment on TDR response. SUMMARY OF BACKGROUND DATA: Cervical TDR may be a promising alternative for a symptomatic adjacent level after prior multilevel cervical fusion. However, little is known about the TDR kinematics in this setting. METHODS: Eight human cadaveric cervical spines (C2-T1, age: 59 ± 8.6 years) were tested intact, after simulated two-level fusion (C4-C6) in lordotic alignment and then in straight alignment, and after C3-C4 TDR above the C4-C6 fusion in lordotic and straight alignments. Fusion was simulated using an external fixator apparatus, allowing easy adjustment of C4-C6 fusion alignment, and restoration to intact state upon disassembly. Specimens were tested in flexion-extension using hybrid testing protocols. RESULTS: The external fixator device significantly reduced range of motion (ROM) at C4-C6 to 2.0 ± 0.6°, a reduction of 89 ± 3.0% (P < 0.05). Removal of the fusion construct restored the motion response of the spinal segments to their intact state. The C3-C4 TDR resulted in less motion as compared to the intact segment when the disc prosthesis was implanted either as a stand-alone procedure or above a two-level fusion. The decrease in motion of C3-C4 TDR was significant for both lordotic and straight fusions across C4-C6 (P < 0.05). Flexion and extension moments needed to bring the cervical spine to similar C2 motion endpoints significantly increased for the TDR above a two-level fusion compared to TDR alone (P < 0.05). Lordotic fusion required significantly greater flexion moment, whereas straight fusion required significantly greater extension moment (P < 0.05). CONCLUSION: TDR placed adjacent to a two-level fusion is subjected to a more challenging biomechanical environment as compared to a stand-alone TDR. An artificial disc used in such a clinical scenario should be able to accommodate the increased moment loads without causing impingement of its endplates or undue wear during the expected life of the prosthesis.

Quantitative analysis of the long- and short-arm crescentic shelf bunionectomy osteotomies in fresh cadaveric matched pair specimens.

Two variations of crescentic shelf osteotomies have been described for the treatment of moderate to severe hallux abductovalgus: a short arm and a long arm. This study tested the hypothesis that the short-arm osteotomy will have a greater moment to failure and angular stiffness than the long arm. Eighteen first metatarsal specimens were dissected from 9 matched pairs of fresh frozen cadaveric specimens. One metatarsal from each pair received a short-arm osteotomy, whereas the other received a long-arm osteotomy. Each osteotomy was fixed with 2 screws. The short arm was fixed with 1 oblique screw and 1 dorsal-to-plantar screw. The long arm was fixed with 2 dorsal-to-plantar screws: 1 at the proximal aspect and 1 at the distal aspect of the shelf. Each specimen was loaded in a materials testing machine to measure moment to failure and angular stiffness. The base of the first metatarsal was potted and load applied to the plantar aspect of the metatarsal head at a constant rate until failure of the osteotomy. The mean maximum moment to failure of the short arm was significantly greater than the long arm (2.04 ± 0.96 Newton meter [Nm] vs. 1.48 ± 0.67 Nm, P = .03). The mean angular stiffness was significantly greater for short arm versus long arm (23.8 ± 19.11 Nm/radian vs. 0.98 ± 9.08 Nm/radian, P = .01). We report statistically significant data supporting the short-arm crescentic shelf osteotomy to have a greater moment to failure and angular stiffness compared with the long-arm crescentic shelf osteotomy.

Disc replacement adjacent to cervical fusion: a biomechanical comparison of hybrid construct versus two-level fusion.

STUDY DESIGN: A cadaveric biomechanical study. OBJECTIVE: To investigate the biomechanical behavior of the cervical spine after cervical total disc replacement (TDR) adjacent to a fusion as compared to a two-level fusion. SUMMARY OF BACKGROUND DATA: There are concerns regarding the biomechanical effects of cervical fusion on the mobile motion segments. Although previous biomechanical studies have demonstrated that cervical disc replacement normalizes adjacent segment motion, there is a little information regarding the function of a cervical disc replacement adjacent to an anterior cervical decompression and fusion, a potentially common clinical application. METHODS: Nine cadaveric cervical spines (C3-T1, age: 60.2 ± 3.5 years) were tested under load- and displacement-control testing. After intact testing, a simulated fusion was performed at C4-C5, followed by C6-C7. The simulated fusion was then reversed, and the response of TDR at C5-C6 was measured. A hybrid construct was then tested with the TDR either below or above a single-level fusion and contrasted with a simulated two-level fusion (C4-C6 and C5-C7). RESULTS: The external fixator device used to simulate fusion significantly reduced range of motion (ROM) at C4-C5 and C6-C7 by 74.7 ± 8.1% and 78.1 ± 11.5%, respectively (P < 0.05). Removal of the fusion construct restored the motion response of the spinal segments to their intact state. Arthroplasty performed at C5-C6 using the porous-coated motion disc prosthesis maintained the total flexion-extension ROM to the level of the intact controls when used as a stand-alone procedure or when implanted adjacent to a single-level fusion (P > 0.05). The location of the single-level fusion, whether above or below the arthroplasty, did not significantly affect the motion response of the arthroplasty in the hybrid construct. Performing a two-level fusion significantly increased the motion demands on the nonoperated segments as compared to a hybrid TDR-plus fusion construct when the spine was required to reach the same motion end points. The spine with a hybrid construct required significantly less extension moment than the spine with a two-level fusion to reach the same extension end point. CONCLUSION: The porous-coated motion cervical prosthesis restored the ROM of the treated level to the intact state. When the porous-coated motion prosthesis was used in a hybrid construct, the TDR response was not adversely affected. A hybrid construct seems to offer significant biomechanical advantages over two-level fusion in terms of reducing compensatory adjacent-level hypermobility and also loads required to achieve a predetermined ROM.

Effect of the Total Facet Arthroplasty System after complete laminectomy-facetectomy on the biomechanics of implanted and adjacent segments.

BACKGROUND CONTEXT: Lumbar fusion is traditionally used to restore stability after wide surgical decompression for spinal stenosis. The Total Facet Arthroplasty System (TFAS) is a motion-restoring implant suggested as an alternative to rigid fixation after complete facetectomy. PURPOSE: To investigate the effect of TFAS on the kinematics of the implanted and adjacent lumbar segments. STUDY DESIGN: Biomechanical in vitro study. METHODS: Nine human lumbar spines (L1 to sacrum) were tested in flexion-extension (+8 to -6Nm), lateral bending (+/-6Nm), and axial rotation (+/-5Nm). Flexion-extension was tested under 400 N follower preload. Specimens were tested intact, after complete L3 laminectomy with L3-L4 facetectomy, after L3-L4 pedicle screw fixation, and after L3-L4 TFAS implantation. Range of motion (ROM) was assessed in all tested directions. Neutral zone and stiffness in flexion and extension were calculated to assess quality of motion. RESULTS: Complete laminectomy-facetectomy increased L3-L4 ROM compared with intact in flexion-extension (8.7+/-2.0 degrees to 12.2+/-3.2 degrees, p<.05) lateral bending (9.0+/-2.5 degrees to 12.6+/-3.2 degrees, p=.09), and axial rotation (3.8+/-2.7 degrees to 7.8+/-4.5 degrees p<.05). Pedicle screw fixation decreased ROM compared with intact, resulting in 1.7+/-0.5 degrees flexion-extension (p<.05), 3.3+/-1.4 degrees lateral bending (p<.05), and 1.8+/-0.6 degrees axial rotation (p=.09). TFAS restored intact ROM (p>.05) resulting in 7.9+/-2.1 degrees flexion-extension, 10.1+/-3.0 degrees lateral bending, and 4.7+/-1.6 degrees axial rotation. Fusion significantly increased the normalized ROM at all remaining lumbar segments, whereas TFAS implantation resulted in near-normal distribution of normalized ROM at the implanted and remaining lumbar segments. Flexion and extension stiffness in the high-flexibility zone decreased after facetectomy (p<.05) and increased after simulated fusion (p<.05). TFAS restored quality of motion parameters (load-displacement curves) to intact (p>.05). The quality of motion parameters for the whole lumbar spine mimicked L3-L4 segmental results. CONCLUSIONS: TFAS restored range and quality of motion at the operated segment to intact values and restored near-normal motion at the adjacent segments.

Altered disc pressure profile after an osteoporotic vertebral fracture is a risk factor for adjacent vertebral body fracture.

This study investigated the effect of endplate deformity after an osteoporotic vertebral fracture in increasing the risk for adjacent vertebral fractures. Eight human lower thoracic or thoracolumbar specimens, each consisting of five vertebrae were used. To selectively fracture one of the endplates of the middle VB of each specimen a void was created under the target endplate and the specimen was flexed and compressed until failure. The fractured vertebra was subjected to spinal extension under 150 N preload that restored the anterior wall height and vertebral kyphosis, while the fractured endplate remained significantly depressed. The VB was filled with cement to stabilize the fracture, after complete evacuation of its trabecular content to ensure similar cement distribution under both the endplates. Specimens were tested in flexion-extension under 400 N preload while pressure in the discs and strain at the anterior wall of the adjacent vertebrae were recorded. Disc pressure in the intact specimens increased during flexion by 26 +/- 14%. After cementation, disc pressure increased during flexion by 15 +/- 11% in the discs with un-fractured endplates, while decreased by 19 +/- 26.7% in the discs with the fractured endplates. During flexion, the compressive strain at the anterior wall of the vertebra next to the fractured endplate increased by 94 +/- 23% compared to intact status (p < 0.05), while it did not significantly change at the vertebra next to the un-fractured endplate (18.2 +/- 7.1%, p > 0.05). Subsequent flexion with compression to failure resulted in adjacent fracture close to the fractured endplate in six specimens and in a non-adjacent fracture in one specimen, while one specimen had no adjacent fractures. Depression of the fractured endplate alters the pressure profile of the damaged disc resulting in increased compressive loading of the anterior wall of adjacent vertebra that predisposes it to wedge fracture. This data suggests that correction of endplate deformity may play a role in reducing the risk of adjacent fractures.

Vertebroplasty comparing injectable calcium phosphate cement compared with polymethylmethacrylate in a unique canine vertebral body large defect model.

BACKGROUND CONTEXT: Vertebroplasty was developed to mechanically reinforce weakened vertebral bodies. Polymethylmethacrylate (PMMA) bone cement has been most commonly used but carries risks of thermal injury and respiratory and cardiovascular complications. Calcium phosphate (CaP) offers the potential for biological resorption and replacement with new bone, restoring vertebral body mass and height. PURPOSE: To compare compressive strength, elastic modulus of the adjacent motion segments, and histologic response of vertebral bodies injected with either CaP or PMMA in a canine vertebroplasty model. STUDY DESIGN: By using a canine vertebroplasty model, two level vertebroplasties were performed at L1 and L3 and studied for 1 month (n=10) and 6 months (n=10). In each canine, one vertebral defect was randomly injected with either CaP cement (BoneSource; Stryker, Freiberg, Germany) or PMMA. METHODS: Twenty dogs had an iatrogenically created cavitary lesion at two nonadjacent levels injected with either CaP or PMMA. Canines from each group were tested mechanically (n=5) and histologically (n=5). Histology consisted of axial sections of the L1 and L3 vertebral bodies and high-resolution contact radiographs. Sections from each specimen were embedded in plastic without decalcification to study the bone-cement interface. Bone-cement interfaces were compared for evidence of necrosis, fibrosis, foreign body response, cement resorption, and new bone formation between the PMMA and CaP treatments groups. Mechanical compression testing was performed on specimens from the 1-month (n=5) and 6-month (n=5) time periods. The T13 vertebral body was used as an intact control for the destructive compression testing of L1 and L3. Each vertebral body was compressed to 50% of its original height under displacement control at 15 mm/min to simulate a nontraumatic loading situation. Force and displacement data were recorded in real time. RESULTS: Vertebral sites containing PMMA were characterized by a thin fibrous membrane. PMMA was detected within the trabeculae, vascular channels, and the spinal canal. Unlike PMMA, CaP underwent resorption and remodeling with vascular invasion and bone ingrowth. Woven and lamellar bone was found on the CaP cement surface, within the remodeled material, and on the surrounding trabeculae. Vertebral body compression strength testing revealed no significant difference in vertebral body height and compressive strength between PMMA and CaP. There was a trend for CaP-treated vertebrae to increase in compressive strength from 1 month to 6 months, whereas PMMA decreased compressive strength when compared with adjacent nontreated vertebrae. CONCLUSION: For both short and intermediate time periods, the injection of CaP cement can be an effective method to treat large vertebral defects. Early results indicate that CaP remodeling might result in the resorption of the majority of the cement with replacement by lamellar bone.

Enhancing the stability of anterior lumbar interbody fusion: a biomechanical comparison of anterior plate versus posterior transpedicular instrumentation.

STUDY DESIGN: Biomechanical study using human cadaver spines. OBJECTIVE: To assess the stabilizing effect of a supplemental anterior tension band (ATB, Synthes) plate on L5-S1 anterior lumbar interbody fusion (ALIF) using a femoral ring allograft (FRA) under physiologic compressive preloads, and to compare the results with the stability achieved using FRA with supplemental transpedicular instrumentation. SUMMARY OF BACKGROUND DATA: Posterior instrumentation can improve the stability of ALIF cages. Anterior plates have been proposed as an alternative to avoid the additional posterior approach. METHODS: Eight human specimens (L3 to sacrum) were tested in the following sequence: (i) intact, (ii) after anterior insertion of an FRA at L5-S1, (iii) after instrumentation with the ATB plate, and (iv) after removal of the plate and adding transpedicular instrumentation at the same level. Specimens were tested in flexion-extension, lateral bending, and axial rotation. Flexion-extension was tested under 0 N, 400 N, and 800 N compressive follower preload to simulate physiologic compressive preloads on the lumbar spine. RESULTS: Stand-alone FRAs significantly decreased the range of motion (ROM) in all tested directions (P < 0.05); however, the resultant ROM was large in flexion-extension ranging between 6.1 +/- 3.1 degrees and 5.1 +/- 2.2 degrees under 0 N to 800 N preloads. The ATB plate resulted in a significant additional decrease in flexion-extension ROM under 400 N and 800 N preloads (P < 0.05). The flexion-extension ROM with the ATB plate was 4.1 +/- 2.3 under 0 N preload and ranged from 3.1 +/- 1.8 to 2.4 +/- 1.3 under 400 N to 800 N preloads. The plate did not significantly decrease lateral bending or axial rotation ROM compared with stand-alone FRA (P > 0.05), but the resultant ROM was 2.7 +/-1.9 degrees and 0.9 +/- 0.6 degrees , respectively. Compared with the ATB plate, the transpedicular instrumentation resulted in significantly less ROM in flexion-extension and lateral bending (P < 0.05), but not in axial rotation (P > 0.05). CONCLUSION: The ATB plate can significantly increase the stability of the anterior FRA at L5-S1 level. Although supplemental transpedicular instrumentation results in a more stable biomechanical environment, the resultant ROM with the addition of a plate is small, especially under physiologic preload, suggesting that the plate can sufficiently resist motion. Therefore, clinical assessment of the ATB plate as an alternative to transpedicular instrumentation to enhance ALIF cage stability is considered reasonable.

Effect of uncovertebral joint excision on the motion response of the cervical spine after total disc replacement.

STUDY DESIGN: In vitro biomechanical study. OBJECTIVE: To quantify the effects of uncinatectomy on cervical motion after total disc replacement (TDR). SUMMARY OF BACKGROUND DATA: The effect of uncinatectomy on TDR motion is unknown. Partial uncinatectomy may be required to decompress the foramen; however, the residual uncinates can potentially limit TDR motion and serve as a source of progressive spondylosis. Complete resection of the uncinates may decrease this risk yet endanger destabilizing the segment. METHODS: Seven human cervical spines (C3-C7) (age, 63.4 +/- 6.9 years) were tested first intact and then after implantation of a metal-on-polyethylene ball-and-socket semiconstrained prosthesis at C5-C6. Following this, gradually increased uncinatectomy was performed in the following order: 1) right partial-posteromedial (two thirds), 2) right complete, and 3) bilateral complete resection. Specimens were tested in flexion-extension, lateral bending, and axial rotation (+/-1.5 Nm). Flexion-extension was tested under 150 N follower preload. RESULTS: TDR without uncinatectomy increased C5-C6 flexion-extension range of motion from 8.4 degrees +/- 3.5 degrees to 11.6 degrees +/- 3.4 degrees, but statistical significance was not reached (P > 0.05). Lateral bending decreased from 6.2 degrees +/- 2.2 degrees to 3.1 degrees +/- 1.4 degrees, with a trend for statistical significance (P = 0.07). Axial rotation decreased from 5.5 degrees +/- 2.4 degrees to 4.3 degrees +/- 1.4 degrees after the implantation (P > 0.05). Both right partial and right complete uncinatectomy resulted in nearly symmetrical restoration of lateral bending to intact values and significantly increased flexion-extension compared with intact (P < or = 0.05); however, axial rotation still did not differ from intact (P > 0.05). Complete bilateral resection also restored lateral bending to intact values (7.3 degrees +/- 2.7 degrees, P > 0.05); however, it resulted in significant increase in range of motion in flexion-extension (14.1 degrees +/- 3.0 degrees, P < or = 0.05) and axial rotation (8.7 degrees +/- 2.4 degrees, P < or = 0.05). CONCLUSION: Unilateral complete or even partial uncinatectomy can normalize lateral bending after TDR. Bilateral complete uncinatectomy is not necessary to restore lateral bending and may result in significantly increased range of motion in flexion-extension and axial rotation compared with intact values.

Novel model to analyze the effect of a large compressive follower pre-load on range of motions in a lumbar spine.

A 3-D finite element model (FEM) of the lumbar spine (L1-S1) was used to determine the effect of a large compressive follower pre-load on range of motions (ROM) in all three planes. The follower load modeled in the FEM produced minimal vertebral rotations in all the three planes. The model was validated by comparing the disc compression at all levels in the lumbar spine with the corresponding results obtained by compressing 10 cadevaric lumbar spines (L1-S1) using the follower load technique described by Patwardhan et al. [1999. A follower load increases the load-carrying capacity of the lumbar spine in compression. Spine 24(10), 1003-1009]. Further validation of the model was performed by comparing the lateral bending and torsion response without pre-load and the flexion-extension response without pre-load and with an 800 N follower pre-load with those obtained using cadaver lumbar spines. Following validation, the FEM was subjected to bending moments in all three planes with and without compressive follower pre-loads of up to 1200 N. Disc compression values and the flexion-extension range of motion under 800 N follower pre-load predicted by the FEM compared well with in vitro results. The current model showed that compressive follower pre-load decreased total as well as segmental ROM in flexion-extension by up to 18%, lateral bending by up to 42%, and torsion by up to 26%.

Three-dimensional in vivo measurement of lumbar spine segmental motion.

STUDY DESIGN: Fifteen asymptomatic volunteers were externally rotated and CT scanned to determine lumbar segmental motion. OBJECTIVES: To measure three-dimensional segmental motion in vivo using a noninvasive measurement technique. SUMMARY OF BACKGROUND DATA: Spinal instability has been implicated as a potential cause of low back pain, especially, axial rotational instability. Typically, flexion-extension lateral radiographs were used to quantify instability, but inaccurately measured translations and inability to capture out-of-plane rotations are limitations. METHODS: Using a custom-calibrated rotation jig, L1-S1 CT reconstructions were created of volunteers in each of 3 positions: supine and left and right rotations of the torso with respect to the hips. Segmental motions were calculated using Euler angles and volume merge methods in three major planes. RESULTS: Segmental motions were small (< 4 degrees or 6 mm) with the greatest motions seen in axial rotation (range, 0.6 degrees to 2.2 degrees ), lateral bending (range, -3.6 degrees to 3.0 degrees ), and frontal translation (-1.2 mm to 5.4 mm). Largest motions were in the levels: L1-L2 to L3-L4. CONCLUSIONS: Complex coupled motions were measured due to external torsion and could be indicative of instability chronic patients with low back pain. The presented data provide baseline segmental motions for future comparisons to symptomatic subjects.

Augmentation of pedicle screw fixation strength using an injectable calcium phosphate cement as a function of injection timing and method.

STUDY DESIGN: Axial pullout tests using fresh cadaveric thoracolumbar vertebral bodies. OBJECTIVES: To evaluate the effect of a new injectable calcium phosphate cement on the axial pullout strength of both revised and augmented pedicle screws in comparison with polymethyl methacrylate and in terms of injection method. SUMMARY OF BACKGROUND DATA: Failure of pedicle screws by loosening and back out remains a significant clinical problem and is of particular concern for patients with low bone quality. Polymethyl methacrylate was shown to significantly improve the screw pullout strength. However, polymethyl methacrylate is known to have a high polymerization temperature, which may damage surrounding tissues, and a short handling time, and it lacks long-term biocompatibility. Bone mineral cements such as calcium phosphate have a longer working time, very low thermal effect, and are biodegradable as well as having good mechanical strength. Recently, new calcium phosphate cement with improved infiltration properties for better injectability has been introduced, but its performance in augmenting the pedicle screw fixation has not been tested yet. METHODS: The bone mineral densities of 52 vertebral bodies (T11-L5) were measured using dual-energy x-ray absorptiometry. In each vertebral body, a 6.5-mm-diameter and 45 +/- 5-mm-long pedicle screw was inserted into either the right or left pedicle, representing an initial intact implantation. These intact screws were pulled axially until failure at 10 mm/min. Following failure of the intact pedicle, 3.0 cc of cement was injected into the failed screw hole, representing a revision case, and the prepared screw hole in the contralateral intact pedicle representing an augmentation case. The cement was injected either to the distal tip of the screw hole (calcium phosphate-1 group, n = 19) or along the entire length of the screw hole (calcium phosphate-2 group, n = 20), and the screws were inserted. The cement was then allowed to cure for 24 hours at room temperature before both screws were pulled to failure. In 13 specimens, polymethyl methacrylate was injected along the entire length of the screw hole (polymethyl methacrylate group). Kruskal-Wallis and Mann-Whitney tests were used to compare the screw pullout strengths for study groups, whereas linear relationships between variables were assessed with scatter plots and Spearman correlation coefficients with a significance level of 0.05. RESULTS: Mean bone mineral densities of all groups were similar. A significant positive correlation was seen between bone mineral density and intact pullout strength. In revision, the pullout strength of calcium phosphate-1 was similar to that of intact, whereas the pullout strength of calcium phosphate-2 and polymethyl methacrylate was significantly greater than that of intact. In augmentation, all 3 injection methods significantly improved the pullout strength over intact. Injection of the calcium phosphate cement along the entire screw length was found to produce significantly higher pullout strengths than injection only at the distal tip of the screw in revision case. Injection of polymethyl methacrylate produced significantly higher pullout strengths than the injection of calcium phosphate by either method in both revision and augmentation. CONCLUSION: Results of this study demonstrate that the new calcium phosphate cement can improve the axial pullout strength of revised and augmented pedicle screws when injected along the entire length of the screw. This suggests that the injection method may be crucial for revision of failed pedicle screws. Considering inherent properties more favorable for in vivo application, such as nonexothermal polymerization and longer working time, and significant improvement in pullout strength, the new calcium phosphate cement may be a good alternative to polymethyl methacrylate for the augmentation of pedicle screw fixation.

Biomechanical evaluation of an injectable calcium phosphate cement for vertebroplasty.

STUDY DESIGN: Destructive biomechanical tests using fresh cadaveric thoracolumbar vertebral bodies. OBJECTIVES: To evaluate the compression strength of human vertebral bodies injected with a new calcium phosphate (CaP) cement with improved infiltration properties for augmentation of the vertebral bodies before compression fracture and also for vertebroplasty in comparison with polymethylmethacrylate (PMMA) injection. SUMMARY OF BACKGROUND DATA: Vertebroplasty is the percutaneous injection of PMMA cement into the vertebral body. While PMMA has high mechanical strength, it cures fast and thus allows only a short handling time. Other potential problems of using PMMA injection may include damage to surrounding tissues by a high polymerization temperature or by the unreacted toxic monomer, and the lack of long-term biocompatibility. Bone mineral cements, such as calcium carbonate and CaP cements, have longer working time and low thermal effect. They are also biodegradable while having a good mechanical strength. However, the viscosity of injectable mineral cements is high, and the infiltration of these cements into vertebral body has been questioned. Recently, the infiltration properties of a CaP cement have been significantly improved, which is ideal for the transpedicular injection to the vertebral bodies for vertebroplasty or augmentation of osteoporotic vertebral body strength. METHODS: The bone mineral densities of 30 vertebral bodies (T2-L1) were measured using dual-energy x-ray absorptiometry. Ten control specimens were compressed at a loading rate of 15 mm/min to 50% of their original height. The other specimens had 6 mL of PMMA (n = 10) or the new CaP (n = 10) cement injected through the bilateral pedicle approach before being loaded in compression. Additionally, after the control specimens had been compressed, they were injected with either CaP (n = 5) or PMMA (n = 5) cement using the same technique, to simulate vertebroplasty. Loading experiments were repeated with the displacement control of 50% vertebral height. Load to failure was compared among groups and analyzed using analysis of variance. RESULTS: Mean bone mineral densities of all five groups were similar and ranged from 0.56 to 0.89 g/cm2. The size of the vertebral body and the amount of cement injected were similar in all groups. Load to failure values for PMMA, the new CaP, and vertebroplasty PMMA were significantly greater than that of control. Load to failure of the vertebroplasty CaP group was higher than control but not statistically significant. The mean stiffness of the vertebroplasty CaP group was significantly smaller than control, PMMA, and the new CaP groups. The mean height gains after injection of the new CaP and PMMA cements for vertebroplasty were minimal (3.56% and 2.01%, respectively). CONCLUSION: Results of this study demonstrated that the new CaP cement can be injected and infiltrates easily into the vertebral body. It was also found that injection of the new CaP cement can improve the strength of a fractured vertebral body to at least the level of its intact strength. Thus, the new CaP cement may be a good alternative to PMMA cement for vertebroplasty, although further in vivo animal and clinical studies should be done. Furthermore, the new CaP may be more effective in augmenting the strength of osteoporotic vertebral bodies for preventing compression fractures considering our biomechanical testing data and the known potential for biodegradability of the new CaP cement.