Back or spine musculoskeletal impairments have been reported to represent more than one-half (51.7% or 15.4 million incidents) of the musculoskeletal impairments reported in the United States.' In the 18-84 year-old age group, back or spine impairment is the leading cause of activity limitation and results in more lost productivity than any other medical condition.1
It has been estimated that 4.4 million people aged 25-74 years report intervertebral disk problems in the United States.1 While it has been reported that 80%-90% of patients with lower back pain recover by 12 weeks with nonsurgical therapies such as bedrest and antiinflammatory medications,2 nonsurgical therapies are unsuccessful for certain injuries and pathologies including degenerative disk disease, stenosis, spondylolysis, and/or spondylolisthesis.
When conservative treatment fails, spinal fusion (arthrodesis) may be performed. In the United States, 279,000 surgeries were performed for low back pain in 1990, with 26 lumbar fusions performed per 100,000 persons.2 In 1995, approximately 160,000 spinal fusion surgeries were performed.1 In a review of 47 studies, Turner et al3 reported that 68% of patients had a satisfactory outcome after lumbar fusion, but the range was l6%-95%. Of most concern was a 20%-40% failure rate reported for lumbar spinal fusion.3
Since the approval of spinal fusion cages by the Food and Drug Administration in 1996, the use of these devices has become more prevalent.4-12 Clinically, on the basis of primarily radiographic evaluation, lumbar interbody fusion with titanium spinal fusion cages has been reported to be effective for single-level lumbar interbody fusion, with a fusion rate of s*90% at 1 to 2 years postoperatively.5,7,8,12 Fusion rates may be 70%-80% in patients with multilevel fusions or patients with risk factors such as obesity, tobacco use, or metabolic disorders.
A central question exists regarding the use of these radiopaque devices, "Is radiographic determination of fusion possible with titanium interbody fusion devices?" This question has been debated in recent literature.1314 In 2000, Cizek and Boyd13 published an experimental study that has shown that plain radiographs and computed tomographies (CT) of cage-instrumented cadavers show "considerable metallic artifact." In 2001, a panel of spine surgeons and researchers was unable to develop a consensus for successful arthrodesis following interbody fusion with titanium interbody fusion devices.14 The development of radiolucent spinal fusion devices that are mechanically competent and biocompatible would be an asset to the armamentarium of spine surgeons.
Key to Treatment and Study Design
One bioabsorbable polymer with a clinical history is polylactide (PLa). Polylactide has been used for absorbable sutures, controlled drug release, interference screws in anterior cruciate ligament surgery, bone screws for ankle fractures, biological membranes, guided tissue regeneration, and other biomedical applications.15"21
The polylactide polymer is hydrolyzed to lactic acid in vivo, a natural byproduct of anaerobic metabolism. Lactic acid is then metabolized into carbon dioxide and water. Polylactide is a chiral molecule that can exist in four stereoisomers.21 The racemic form, D,LPLa, is amorphous and degradation occurs by surface dissolution, not by the formation of crystalline degradation products.21 High molecular weight D5LPLa has a tensile strength of 35-70 MPa and a compressive strength of 60-100 MPa.21"22 DJL-PLa has an elastic modulus of 2.4 GPa,21 which is staffer than cancellous bone but more flexible than cortical bone.
A recent investigation has evaluated a PLa bioabsorbable implant for interbody fusion showing significantly greater bone growth and bone remodeling compared with titanium controls.23 Previously published studies have shown that autograft, as well as cages and other spinal fusion devices alone or packed with autograft, may not produce solid fusions.24"30 Using an Ovine model, previous studies have shown that the augmentation strategy (augmentation with rhBMP-2) has significantly increased the fusion rate of cages compared to the same implant with autograft or alone.10,24 The present study addresses the efficacy of autograft or rhBMP-2 loaded on a collagen sponge to achieve radiographic and histological fusion and an associated increase in segmental stiffness using a 70/30 poly(L-lactide-co-D,L-lactide) (70/30 D1L-PLa) device.
The goals of the present study are to evaluate the osteocompatibility of the radiolucent 70/30 D,L-PLa polymeric device, evaluate the efficacy of the 70/30 D1L-PLa device in a sheep lumbar interbody spinal fusion model using blinded radiographic, biomechanical, and histological measures, evaluate the degradation as a function of time, and evaluate the augmentation strategy of adding bone morphogenetic protein (BMP) on a collagen sponge to stimulate bony healing in conjunction with this biomaterial.
MATERIALS AND METHODS
The sheep lumbar spine model was chosen specifically because of the biomechanical similarities between the sheep and human lumbar spine.31" 32 Wilke et al31 characterized the biomechanical parameters (range of motion, neutral zone, and level stiffness) of sheep spines and made comparisons with data from human specimens previously published by White and Panjabi.33
Wilke et al found that the ranges of motion of sheep spines for the different load directions are qualitatively similar in craniocaudal trends to those of human specimens reported in the literature. They concluded that "based on the biomechanical similarities of the sheep and human spines demonstrated in this study, it appears that the sheep spine.. .can serve as an alternative for the evaluation of spinal implants."31
Sandhu et al34 presented a summary of animal models for spinal instability and fixation. These animal models were used without knowledge of how the biomechanics of the animal model compares to the human spine. The sheep lumbar spinal fusion model has been shown to be biomechanically similar to the human spine through a free-body analysis of transmitted forces and histologic evaluation of trabecular structure.32 An advantage of the sheep model is that the lamellar bone growth rate is nearly equivalent to that of humans.34
Materials and Study Design
The bioabsorbable interbody fusion device was evaluated in 25 skeletally mature sheep for various survival periods (Table 1 ). This study was approved by the Institutional Animal Care and Use Committee (CSU IACUC #97- 180A01). Colorado State University complies with the US Department of Agriculture regulations promulgated under the authority of Animal Welfare Act and those of the US Public Health Service (PHS) Policy on Laboratory Animal Care as provided in the Health Research Extension Act. In addition, the program is in compliance with recommendations of the American College of Laboratory Animal Medicine and the PHS "Guide for the Care and Use of Laboratory Animals." Animals were fasted for 24 hours prior to surgery. Water was not restricted during this time. Anesthesia was induced with ketalar (4 mg/kg) and diazepam (7.5 mg total). After induction, sheep were maintained with isoflurane ( 1 .5% to 3%) in 100% oxygen (2 L/min) during the surgical procedure. Muscle relaxants were not used.
The surgical technique involved positioning the sheep in right lateral recumbency for single-level lumbar diskectomy and interbody fusion at L4-L5 via a left retroperitoneal approach. Following diskectomy, a 70/30 DX-PLa interbody fusion device (14 mm X 20 mm) was packed with morselized iliac crest cancellous autograft or InFuse bone graft substitute (Figure 1) (Medtronic Sofamor Danek, Memphis, Tenn). The 70/30 D,L-PLa polymer implants (MacroPore Biosurgery, San Diego, Calif) had the following mechanical properties: an elastic modulus of 3.15 GPA, an ultimate compressive strength of 100 MPa, an ultimate tensile strength of 58 MPa, and a ductility of 5% elongation to failure. Fabricated into the shape of the interbody fusion device (Figure 1 ), the device was able to withstand a compressive load to failure of 8230 (± 60) newtons after sterilization. InFuse bone graft substitute consisted of 0.80 mL of rhBMP-2 (Wyeth Research, Cambridge. Mass) at a concentration of 0.43 mg/mL applied to a type I collagen sponge (Helistat collagen sponge, bovine achules tendon, Integra Life Sciences, Andover, Mass).
The animals received either InFuse or autograft at the time of surgery. The key to treatment and study design is found in Table 1. Five postoperative time points of 3, 6, 12, 18, and 24 months were evaluated in the present study. Figure 1 shows the 70/30 D,L-PLa device in the interbody space after implantation. Figure 2A is a lateral radiograph showing the appearance of the defect immediately after implantation of the radiolucent 70/30 D5L-PLa device filled with rhBMP-2 on a collagen sponge. Both the device and InFuse bone graft substitute are radiolucent. All of the animals recovered well from the surgery and were examined for neurological deficits.
Pain medication after the surgical procedures included fentanyl patches at a dose of 150 pg/hour administered with a continuous percutaneous patch for 3 days. Additional pain medication included phenylbutazone at a dose of 1 g administered orally once per day for 3 days. At the conclusion of the study, all of the sheep were humanely euthanized with barbiturate overdose at 3, 6, 12, 18, or 24 months postoperatively (Table 1). The efficacy of the radiolucent device and InFuse bone graft substitute to affect lumbar interbody fusion was assessed by performing radiographic, biomechanic, and histologic analyses in a blinded fashion.
Neurologic exams were conducted daily for 7 postoperative days, at 2 months, and before euthanasia at 4months. Neurological exams were conducted using the following scale: 0= walking without any detectable ataxia, I = walking, slightly ataxic, 2= walking, but with noticeable weakness on one side or both sides, 3= able to stand on forelimbs but dragging rear limbs, 4=recumbent and unable to rise.
Radiographs were taken immediately after surgery and at regular postoperative follow-up times. High-resolution radiographs were made after biomechanical testing (posteroanterior and lateral views) using a high-resolution radiography unit (Faxitron radiograph unit, Hewlett-Packard, McMinnville, Ore) and high-resolution film (Ektascan M EM-I, Eastman-Kodak, Rochester, NY). The resulting Faxitron radiographs from the treated animals and the biomechanical sham groups were read by three blinded evaluators for fusion, bone in the cage, and implant placement. The radiographs were graded in the following manner. Grade 3 was a solid fusion with no radiolucent lines surrounding the cage. Grade 2 was a probable fusion with some radiolucent lines surrounding the cage. Grade 1 was a nonfusion with significant radiolucent lines surrounding the cage. Radiographs were also evaluated for bone present in the cage as seen from the lateral view, as well as the presence of anterior or posterior bony bridging.
Figure 1 : The appearance of the 70/30 D,LPLa device prior to implantation (A), lilac crest autograft or rhBMP-2 on a collagen sponge was packed into the thrugrowth slot of the device (B). The resorbable fusion device is implanted in the interbody space (arrow) (C).
Ex vivo biomechanical testing was performed to quantify the flexibility of the treated motion segment by measuring load displacement behavior. The treated lumbar motion segments were dissected from the harvested lumbar spine and cleaned of extraneous soft tissues leaving the ligamentous and osseous tissues intact. Unconstrained biomechanical testing was performed in a nondestructive manner on all treated spines. Specially designed loading and base frames were secured on the caudal and cranial vertebrae, respectively. Three retroreflective markers were attached to each vertebra. Pure moments (0, 0.5, 2.5, 4.5, 6.5, and 8.5 Nm) were applied in the following loading directions: flexion, extension, right and left lateral bending, and left and right axial rotation. The location of the markers was recorded at each load using three infrared video cameras (VICON cameras, Oxford Metrics, Oxford, England). The three-dimensional coordinate data were then analyzed to obtain the rotation angles and the flexibility of each motion segment. In addition to the treated animals, seven "biomechanical sham" (polymeric device implanted in normal cadaver sheep spine using same surgical technique) motion segments were tested in the same manner. The rationale for the biomechanical sham is that it allows for comparison of the biomechanics of the treated survival groups to the instrumented sham levels. A fused level would then have an increased stiffness and decreased flexibility compared to the instrumented sham levels. Biomechanical testing data of the biomechanical shams provide a better comparison to the nonfused survival implant than untreated normal motion segments. Biomechanical shams provide an estimate of the immediate postoperative stiffness in all six loading directions of the stabilized spinal construct due to the implantation of the interbody device.10,26
Summary of Mean Radiographic Fusion Scores for the Biomechanical Sham Croup and the Treated Survival Groups
Figure 2: Postoperative lateral radiographs showing the appearance of the defect immediately after implantation of the radiolucent 70/30 D,L-PLa device filled with rhBMP-2 on a collagen sponge. Both the device and bone graft substitute are radiolucent (A), Faxitron lateral radiographs were taken of levels treated with autograft (B) and rhBMP-2 at 3 months postoperative sacrifice time (C). Faxitron lateral radiographs were also taken of two levels treated with autograft at 24 months postoperative sacrifice time (D and E) and of two levels treated with rhBMP-2 at 24 months postoperative sacrifice time (F and G).
Immediately after biomechanical testing, the specimens were fixed in 10% neutral buffered formalin and bisected midsagitally to produce right and left halves. These halves were sequentially dehydrated in alcohols, cleared in xylene, and embedded in graded catalyzed methylmethacrylate for undecalcified histological studies. After polymerization was complete, sections were cut continuously through the explant on a diamond saw (IsomeL Buehler, Lake Bluff, 111) to an approximate thickness of 1.50-400 urn. Approximately 10-15 sections were made in the sagittal plane through each half of the bisected level. The thickness of each section was measured with a metric micrometer. Differential staining with a trichrome stain was used to permit both histological and cytological differentiation. With this staining method, the following tissues can be differentiated on the basis of color: bone is stained blue/green, articular cartilage and fibrocartilage are stained dark purple, and fibrovascular tissue is stained pink. Staining of cellular and nuclear detail by the trichrome stain is similar to staining with hematoxylin and eosin, and allows cytological differentiation.
In addition to stained undecalcified sections, four to eight undecalcified sections from each treated level were radiographed using a Faxitron radiography unit and spectroscopic film (EM-I film). The thickness of the sections was measured with a metric micrometer (Fowler, Japan) to determine the exposure time. Sections were labeled with Ultra-fine permanent markers and then exposed to the x-ray source at 20 kV and 3 m-A for approximately 45 seconds for each 1 OO urn of section thickness. The samples were placed on the sheet of spectroscopic film and the film was placed on a rectangular film holder. The loaded cassette assembly was inserted into the Faxitron radiograph unit and exposed to the radiation as described. The films were then developed, fixed, and analyzed for ossification using standard optical microscopy.
The histological slides and microradiograpbs were used to evaluate histologic fusion and determine the presence of pseudarthroses. The criterion used to assess histological fusion was a continuous bony bridge from the superior to the inferior vertebra. A solid fusion existed if >50% of the sections and corresponding microradiographs showed continuous bony bridging through the thrugrowth region of the 70/30 D,L-PLa device. A partial fusion existed if <50% of the sections and corresponding microradiographs showed continuous bony bridging through the thrugrowth region of the 70/30 D,L-PLa device. A nonfusion existed if none of the sections and corresponding microradiographs showed continuous bony bridging through the thrugrowth region of the 70/30 D,L-PLa device.
Figure 3: Level stiffness (Nm/deg) for biomechanical shams and treated functional spinal units (FSU) as a function of survival time in response to moments applied in flexion and extension (A). Level stiffness (Nm/deg) for biomechanical shams and treated FSUs as a function of survival time in response to moments applied in right and left lateral bending (B). Level stiffness (Nm/deg) for biomechanical shams and treated FSUs as a function of survival time in response to moments applied in right and left axial rotation (C).
Analysis of the stained undecalcified sections was also used to determine the histological and cytological response to the treatments and osteocompatibility of the polymeric device. The quality and quantity of bone were evaluated both within and in contact with the bioabsorbable implant.
All sheep recovered from anesthesia uneventfully and were standing and walking without signs of neurological deficits. All sheep received a score of 0 (ie, walking without any detectable ataxia) for limb use at 7 days and 2 months postoperatively, and before euthanasia.
Mean radiographic fusion scores for the biomechanical sham group and all postoperative survival time groups were made by the three blinded evaluators (Table 2). A lateral radiograph showing the appearance of the defect immediately after implantation of the radiolucent 70/30 D,L~PLa device filled with rhBMP-2 on a collagen sponge is seen in Figure 2A. Both the device and bone graft substitute are radiolucent (Figure 2A). Biomechanical shams levels had a mean radiographic fusion score of 1 .06, indicating that the three blinded radiograph evaluators were able to detemiine nonfusions in biomechanical sham radiographs mixed with treated survival levels (Figure 3). Faxitron lateral radiographs showed two levels treated with autograft and rhBMP-2 at 3 months, postoperative survival time (Figures 2B and 2C). Faxitron lateral radiographs of four levels treated with autograft and rhBMP-2 at 24 months postoperative survival time are found in Figures 2D, 2E, 2F, and 2G. Marked radioJucencies were not observed surrounding the cages in any of the treatment groups.
Summary of Histological Fusion Results for the Treated Survival Croups: Fusion Incidence and Percent
Figure 4: Stained undecalcafied section (A) and mkroradiograph (B) from a level treated with 70/30 D,L-PLa and rhBMP-2 at 3 months postoperative sacrifice time. This level was read as fused as continuous bony bridging had occurred. Stained undecalcified section (C) and microradiograph (D) treated with 70/30 D,L-PLa and rhBMP-2 at 3-months postoperative sacrifice time. Although bone formation from superior and inferior is seen within the thrugrowth region of the device, a continuous bony bridge has not formed at this time period and this level was read as not fused. The stained fibrocartilage (C) corresponds to the radiolucency on the microradiograph (D).
Polytomous logistic regression models were used to compare radiographic fusion scores for the five treated survival groups and the biomechanical sham group. Four of the treated survival groups (6, 12, 18, and 24 months) showed significantly higher odds ratios for having better fusion scores than the sham group by all three evaluators (P<.01 13). A significant difference in odds ratios between 3 months and 24 months was found in the scores from two evaluators (P=.01). The difference between 6 and 24 months was also significant by the scores from the same two evaluators (P =.02 and P-. 03) and was marginally significant by the third evaluator (?=.09).
Ex vivo Biomechanical Testing Results
Biomechanical flexibility data were presented as stiffness in Nm/degrees (mean ± standard deviation) for flexion and extension, right and left lateral bending, and right and left axial rotation for the biomechanical sham group. Stiffness data for levels treated with autograft and rhBMP-2 were combined for each postoperative survival time. Biomechanics flexibility data were nonparametric. Therefore, differences in the stiffness among all groups were statistically analyzed using the nonparametric Kruskal-Wallis Test for all six loading directions. If statistical significance was found, nonparametric posthoc pair-wise comparisons were performed. A statistical significance level of P<.05 was used for all tests.
A trend of increased fusion stiffness with time was demonstrated from 3-24 months, with the exception of a slight drop at the 1 8-month time point in five of the six loading directions. The stiffness of the fusion segments from the 18- and 24-month groups was significantly higher in all six loading directions compared to the biomechanical sham group. The 12-month treatment group showed significantly higher stiffness than the biomechanical sham group in five of the six directions. This difference was not significant in right axial rotation. The stiffness of the 6-month group was higher than the biomechanical sham group in both left and right lateral bending and in extension. Stiffness differences between the 3-month survival group and the biomechanical sham group were not significant. These data correlated with radiographic and histological nonfusions in the 3-month survival group.
The 24-month group was significantly stiffer than the 3-month group in all six loading directions. The 24-month group was also stiffer than the 6-month group in all six directions except left lateral bending. Finally, the 24-month group was stiffer than the 12-month group in right axial rotation. In flexion, both the 18-month and 12- month groups were significantly stiffer than the 3-month group. The 12month group was also stiffer than the 6month group. Finally, the 12-month group demonstrated higher stiffness in right lateral bending than the 3-month group.
In the three earliest postoperative survival groups (3, 6, and 12 months), levels treated with rhBMP-2 showed a higher stiffness than those treated with autograft in all six loading directions of the flexibility tests. For these parameters, the low number of sheep treated with rhBMP-2 and autograft in each group was not amenable to statistical analysis. At 18 and 24 months, the autograft and rhBMP-2 treatments resulted in similar stiffness in all six loading directions.
Histological fusion data for the three treatment groups were gathered (Table 3). Radiographic, biomechanic, and histologic measures for fusion were highly correlated in this study. Only one out of four sheep achieved histological fusion at three months (Figure 4). Three out of four sheep had histological nonfusions at the 3-month survival time because 3 months may be too early to expect histological fusion in this model.
A stained undecalcified section and corresponding microradiograph from a level treated with 70/30 D,LPLa and autograft at 3 months demonstrating histological nonfusion is seen in Figures 4C and 4D, respectively. At 6 months, two of the four sheep showed histological fusions, although one level treated with 70/30 D1L-PLa plus autograft had a partial fusion with continuous bone in the anterior margin, but not in the thrugrowth region of the device. One device showed fracture at six months. Figure 5A shows a stained undecalcified section from a level treated with 70/30 D,L-PLa and autograft at the 6-month postoperative sacrifice time. The section was taken lateral to the thrugrowth region and showed a fracture through the device. Fibrovascular ingrowth in fissures confirms that fracture did not occur during mechanical testing (Figure 5B). The level in Figure 5D was not fused as demonstrated by the dark purple fibrocartilagenous pseudarthrosis inferior to the device. The device in Figure 5 D appears to have been placed adjacent to the inferior cartilaginous endplate.
At 12 months, three of the four 70/30 D,L-PLa devices packed with autograft and one of the two devices packed with rhBMP-2 were efficacious in achieving histological fusion. The microradiograph from a level treated with 70/30 D,L-PLa and autograft at the 12-month postoperative sacrifice time shows a fusion in the thrugrowth region of the device, as well as through the anterior and posterior margins (Figure 6).
Figure 5: Stained undecalcified section from a level treated with 70/30 D,L-PLa and autograft at 6-months postoperative sacrifice time (A). A section taken lateral to the thrugrowth region shows a fracture through the device with fibrovascular ingrowth in fissures and degradation products (B). A macrophage response was found in the vicinity of degradation products (C). A stained undecalcified section from a level treated with 70/30 D, L-PLa and rhBMP-2 at 6-months postoperative sacrifice time shows no fusion by the dark purple fibrocartiagenous pseudarthrosis inferior to the device (D).
A stained section, also from a level treated with 70/30 D,L-PLa and autograft at 12 months, shows that the polymer is undergoing degradation and that a nonfusion is present. The microradiograph from a level treated with 70/30 D,L-PLa and rhBMP-2 at the 12-month postoperative sacrifice time period shows a fusion in the thrugrowth region of the device as well as through the anterior and posterior margins. AU five levels treated with the 70/30 D.L-PLa device and autograft or rhBMP-2 from the 18and 24-month time periods showed histological fusions. Histological fusions occurred in the thrugrowth regions of the device, as well as the anterior and in some cases posterior margins (Figures 7 and 8).
The Fisher's Exact Test was used to compare histologic fusion rates for the five treated survival groups. The histological fusion rate of the 3-month survival group was significantly lower than the histological fusion rate for the other four treated survival groups (P=. 003 for the 6-month group; P<.001 for the 12-, 18-, and 24-month groups). The fusion rate of the 6-month group was in turn significantly lower than that of the 12-, 18-, and 24-month groups (P<.001). The difference in histological fusion rate between the 18- and 12-month groups was not significant (P=. 77). The histological fusion rate of the 24-month group was significantly higher than both the 12-month (P =.003) and 18-month (P= .02) groups.
Device degradation. The use of rhBMP-2 in this model did not appear to retard or enhance the 70/30 D,L-PLa implant degradation rate. However, correlations were observed between implant degradation and fusion status. Specifically, nonfusion and device degradation were highly correlated at the earlier time points (6 and 12 months). For example, at the 12 months postoperative time point, sheep 104 (70/30 D1L-PLa + auto) and 109 (70/30 D.L-PLa + rhBMP-2) did not achieve histologic fusion. Both of these devices showed significant degradation, fissures, and device fragments. However, levels with histologic fusions showed minimal device fragmentation. Also at the 12-month postoperative time point, sheep 107 (70/30 D,L-PLa + rhBMP-2) and sheep 108 (70/30 D,L-PLa + auto) achieved histologic fusion but showed minimal fissures and degradation. For the most part, the 70/30 D1L-PLa devices were intact in all levels demonstrating histological fusion at 6 and 12 months. In addition, these levels had a fibrovascular capsule surrounding the implant and adjacent to host bone. The degree of implant degradation and replacement by host bone was similar for the 1 8-month time points and all earlier time points. The 24-month postsurvival levels showed a higher degree of implant degradation. In addition, surface staining of the polymer was observed at only the 24-month time point. This may indicate that the polymer was more porous and absorbed the dye.
Figure 6: Microradiographs from a fevel treated with 70/30 D,L-PLa and autograft at 12-months postoperative sacrifice time. Level shows a fusion in the thrugrowth region of the device as well as through the anterior and posterior margins (A). The stained section, also from a level treated with 70/30 D,L-PLa and autograft at 12 months, shows that the polymer is undergoing degradation and that a nonfusion is present (B). An intervening fibrous capsule (pink) is present adjacent to the polymer (white) (C). Also shown are microradiographs from a level treated with 70/30 D,L-PLa and rhBMP-2 at 12-month postoperative sacrifice time. Level shows a fusion in the thrugrowth region of the device as well as through the anterior and posterior margins (D). Also from a level treated with 70/30 D,L-PLa and rhBMP-2 at 12 months, an intervening fibrous capsule (pink) is present adjacent to the polymer (white) (E and F).
Figure 7: Microradiographs from a level treated with 70/30 D,L-PLa and rhBMP-2 at 18months postoperative sacrifice time. Level shows a fusion in the thrugrowth region of the device as well as through the anterior margin (A). The stained section, also from a level treated with 70/30 D,L-PLa and rhBMP-2 at 18 months, shows a fusion in the anterior margin (B). The level treated with 70/30 D,L-PLa and autograft at 18 months shows a fusion in the thrugrowth region of the device as well as through the anterior margin (C).
Host response. Table 1 of the ASTM F981-93 standard was used to quantify the cytological response to the 70/30 D,L-PLa device implanted material at all postoperative time periods in the current study. The result was a rating of 0 to 2.0 for all inflammatory cells at all time periods for all implants on a scale of 0 to 3.0. At all time periods, a vigorous host response was not observed, even in the face of numerous implant fragments in periimplant tissues. There was no acute inflammatory phase present (no neutrophils) at 3 and 6 months. Some acute inflammation was observed in devices from sheep sacrificed at 12 months. There was neither a humoral nor cell-mediated immune response at any time period. Mostly macrophages with a few foreign body giant cells were observed in periimplant tissues. A mild to moderate chronic inflammatory response was present at all sacrifice times. Significant degradation was observed at the 24-month period. This degradation was associated with an increased inflammatory response, but no adverse host response was observed. Bone was found to replace the space occupied by the degrading polymer at 24 months. Bone mineralization adjacent to the polymeric device was found to be normal by staining (Trichrome stain is similar to Masson's stain) and microradiography. No osteolysis was found as a result of polymeric degradation products in peri-implant tissues.
Only one of four sheep (25%) achieved histological fusions at 3 months. Similar to postoperative fusion status clinically, 3 months is too early to expect histological fusion in the ovine lumbar interbody fusion (LIF) model. For comparison, at 4 months in the ovine LIF model, tricortical iliac crest autograft in a titanium cage (BAK device, SulzerSpineTech, Minneapolis, Minn) produced a 29% fusion rate.26 In the current study, for both autograft and InFuse, the histological fusion rate was 50% at 6 months. For comparison, at 6 months in the ovine LIF model, tricortical iliac crest autograft in a titanium cage (InterFix, Medtronic Sofamor Danek) produced a 37% fusion rate.24
Also in that study, the InFuse bone graft substitute produced a 100% histological fusion rate at 6 months in the Interfix cage; however, the amount of rhBMP-2 delivered was 20% greater in that study compared to the current study (the dose concentration itself was identical at 0.43 mg/mL but 1.0 mL was delivered).24 The biomechanical stability of a single cage implanted via a lateral retroperitoneal approach in the ovine LIF model provides a challenging healing environment. However, a better animal model does not exist at this time. Therefore, fusion rates must be viewed within the context of this challenging animal model.
For the radiolucent bioabsorbable polymeric device under investigation, the three blinded radiograph evaluators were able to determine fusion and nonfusion that correlated with histological findings. At 3 months, the evaluators saw intra-device radiolucencies associated with pseudarthrosis and continuous trabecular bridging associated with fusion. In several experimental studies reported in the literature,10,23,24,29,42 nonfusions consisted of pseudarthroses within titanium spinal fusion cages, as opposed to pseudarthroses along the inferior or superior device interface which would generate radiolucencies surrounding the cages. These pseudarthroses associated with titanium fusion cages are not detectable by current radiographic techniques. In the same model, the mean radiographic fusion score generated by the three blinded radiograph evaluators for the biomechanical sham group consisting of a titanium BAK cage packed with autograft was 2.00 (indicating probable fusion).26 In the current study, the mean radiographic fusion score for the biomechanical sham group was 1 .06 (recall that grade 1 is a nonfusion). Therefore, blinded evaluators could determine the presence or absence of radiographic fusion associated with the radiolucent polymeric device.
In a study by van Dijk et al,23 similar experiences evaluating the fusion construct were observed. At 3 months, bone ingrowth was observed in the resorbable polymer implants (PLLA), but with a radiolucency in the fusion mass. At 6 months, this group observed solid arthrodesis in four of six (67%) of the PLLA implants, whereas none of the three titanium implants of similar design had full fusion.23 The titanium implants demonstrated ingrowth of bone but with a radiolucent discontinuity in the fusion mass.23
In the current study, regardless of graft or graft substitute, failure to achieve histologic fusion was associated with a decrease in segment stiffness and thus an increase in segment mobility. Segment mobility of nonfused devices was associated with mechanical degradation of the devices. Correlations were observed between implant degradation and fusion status. Thus, although other degradation mechanisms may be operable,21 we hypothesize that degradation of these devices is also related to mechanical forces and the resulting increased stresses generated in unstable motion segments. Also, it appears that placement of these devices may be a variable in predicting the fusion sucess. Due to the sclerotic nature of the ovine endplate, vascular supply may be restricted, thus inhibiting transport of cytokines and growth factors in the healing process. As a result, fusion may be delayed or restricted. None of the lumbar fusions in this model had any supplemental posterior fixation for increased stability.
The five postimplant evaluation periods permit a glimpse of the dynamic host response to the bioabsorbable polymer. Inflammatory responses to bioabsorbable polymers have been reported in the literature.35"40 Some inflammatory responses were associated with adverse host responses such as osteolysis. Although many of these responses occurred shortly after implantation to polymers with short degradation times, polymers with longer degradation times may incite an inflammatory response years after implantation.35"40 Therefore, biological characterization of implanted polymers should be made relative to the degradation rate. In the current study, significant degradation was observed at the 24-month period. This was associated with an increased inflammatory response, but no adverse host responses were observed. Bone was found to replace the space occupied by the degrading polymer. Microradiography demonstrated that bone mineralization was not altered adjacent to the polymer. No osteolysis was found in the vicinity of the degradation products.
Figure 8; Stained undecalcified sections and corresponding microradiographs from two levels treated with 70/30 D, L- PLa and autograft at 24-months postoperative sacrifice time. These levels show a fusion in the thrugrowth region of the device as well as through the anterior margin (A through D).
The host response observed clinically may be different from what was observed in the ovine model. No animal model can fully predict clinical results.
Titanium lumbar interbody spinal fusion devices have been reported to be 90% effective for single-level lumbar interbody fusion, although radiographic determination of fusion has been debated in the literature. Using blinded radiographic, biomechanic, histologic, and statistical measures, we evaluated a radiolucent 70/30 D1L-PLa interbody fusion device packed with autograft or rhBMP-2 on a collagen sponge in 25 sheep at 3, 6, 12, 1 8, and 24 months. A trend of increased fusion stiffness, radiographic fusion, and histologic fusion was demonstrated from 3 to 24 months. Device degradation was associated with a mild to moderate chronic inflammatory response at all postoperative sacrifice times. Based on these results, D,L-PLa may be a viable alternative to metals in interbody fusion cages due to the polymers of the increased radiolucency and decreased stiffness.
1. Praemer A, Furner S, Rice DP. Musculoskeletal Conditions in the United States. Park Ridge, ITI: American Academy of Orthopaedic Surgeons. 1999.
2. Andersson GB. Epidemiological features of chronic low-back pain. Lancet. 1999; 354:581-585.
3. Turner JA. Ersek M, Herron L, et al. Patient outcomes after lumbar spinal fusions. JAMA. 1992;268:907-911.
4. Bagby GW. Arthrodesis by the distractivecompression method using a stainless steel implant. Orthopedics 1988; 1 1:931-934.
5. Kuslich SD, Ulstrom CL, Griffith SL, Ahem JW, Dowdle JD. The Bagby and Kuslich method of lumbar interbody fusion. History. techniques, and 2-year follow-up results of a United States prospective, mult icenter trial. Spine 1998;23:1267-1278.
6. Marte EO, Goel VK, Pope MH, Park JB, Materials and design of spinal implants - a review. J Appi Biomater. 1997; 38:267-288.
7. McAfee PC Regan JJ Geis WP, Fedder IL. Minimally invasive anterior retroperitoneal approach to the lumbar spine. Emphasis on the lateral BAK. Spine. 1998; 23:1476-1484.
8. Ray CD. Threaded titanium cages for lumbar interbody fusions. Spine. 1997; 22:667680.
9. Sandhu HS, Turner AS, Kabo JM, el al. Distractive properties of a threaded interbody fusion device. An in vivo model. Spine. 1996; 21:1201-1210.
10. Toth JM, An HS. Biomaterials, bone grafts and substitutes, biomaterials, and osteoinductive substances in spinal instrumentation. In: An HS, Coder J, eds. Spinal Instrumentation. 2nd ed. Baltimore, Md: Lippincott, Williams & Wilkins; 1999:73-84.
11. Weiner BK, Fraser RD. Spine Update: Lumbar interbody cages. Spine. 1998; 23:634640.
12. Whitecloud TS, Castro FP, Brinker MR, Hartzog CW, Ricciardi JE, Hill C. Degenerative conditions of the lumbar spine treated with intervertebral titanium cages and posterior instrumentation for circumferential fusion. J Spinal Disord. 1998; 1 1 :479-486.
13. Cizek GR, Boyd LM. Imaging Pitfalls of Interbody Spinal Implants. Spine. 2000; 25:26332636.
14. McAfee PC. Boden SD, Brantigan JW, et al. Symposium: a critical discrepancy-a criteria of successful arthrodesis following interbody spinal fusions. Spine. 2001; 26(3) :32Q-334.
15. Bostman O, Pihlajamaki H. Clinical biocompatibility of biodegradable orthopaedic implants for internal fixation: a review. Biomaterials. 2000; 21:261 5-262 1 .
16. Hoilinger JO, Battistone GC: Biodegradable bone repair materials. Synthetic polymers and ceramics. Clinical Orthopaedics Related Research. 1986; 207:290-305.
17. Vernino AR, Ringeisen TA. Wang HL, et al. Use of biodegradable polylactic acid barrier materials in the treatment of grade II periodontal furcation defects in humans, I: A multicenter investigative clinical study. International Journal of Periodontics & Restorative Dentistry. 1988: 18:572-585.
18. Sallum EA, Sallum AW, Nocili FH Jr, Marcantonio RA, de Toledo S. New attachment achieved by guided tissue regeneration using a bioresorbable polylactic acid membrane in dogs, lnt J Periodontics Restorative Dent. 1998; 18:502-510.
19. McGuire DA, Barber FA, Elrod BF, Paulos LE. Bioabsorbable interference screws for graft fixation in anterior cruciate ligament reconstruction . A rthroscopy. 1 999; 1 5 :463-473 .
20. Bucholz RW; Henry S, Henley MB. Fixation with bioabsorbable screws for the treatment of fractures of the ankle. J Bone Joint Surg Am. 1994;76:319-324.
21. Kohn J, Langer R. Bioresorbable and bioerodibie materials. In: Ramer BD, Hoffman AS, Schoen FJ, Lemons JE, eds. Biomaterials Science; An Introduction to Materials in Medicine. New York, NY: Academic Press; 1996:64-73.
22. Thomas KA. Biomechanics and biomaterials. In: Brinker MR, ed. Review of Trauma. Philadelphia, Pa: WB Saunders; 2001:43-51.
23. van Dijk M. Smit TH, Sugihara S, Burger EH, Wuisman PI. The effect of cage stiffness on the rate of lumbar interbody fusion: an in vivo model using poly(L-lactic Acid) and titanium cages. Spine. 2002; 27:682-688.
24. Sandhu HS, Toth JM, Diwan AD, et al. Histological evaluation of the efficacy of rhBMP2 compared with autograft bone in sheep spinal anterior interbody fusion. Spine. 2002; 27:567575.
25. Sandhu HS. Kanim LEA, Kabo JM, et al. Evaluation of rhBMP-2 with an OPLA carrier in a canine posterolateral (transverse process) spinal fusion model. Spine. 1995; 20:2669-2682.
26. Toth JM. Seim III HB, Schwardt JD, Humphrey WB, Wallskog JA, Turner AS. Direct current electrical stimulation increases the fusion rate of spinal fusion cages. Spine. 2000; 25:25802587.
27. Brekke JH, Toth JM. Principles of tissue engineering applied to programmable osteogenesis. J Appi Biomater. 1998;43:380-398.
28. Sandhu HS, Boden SD. Biological enhancement of spinal fusion. Orthop Clin North Am. 1998;29:621-631.
29. Boden SD, Martin GJ, Horton WC, Truss TL, Sandhu HS, Laparoscopic anterior spinal arthrodesis with rhBMP-2 in a titanium interbody threaded cage. J Spinal Disord. 1998; 11:95-101.
30. Zdeblick TA. Ghanayem AJ. Rapoff AJ, et al. Cervical interbody fusion cages: an animal model with and without bone morphogenetic protein. Spine. 1998; 23:758-766.
31. Wilke H-J, Kettler A, Ciaes LE. Are sheep spines a valid biomechanical model for human spines? Spine. 1997; 22:2365-2374.
32. Smit TH. The use of a quadruped as an in vivo model for the study of the spine - biomechanical considerations. Eur Spine J. 2002; 11:137-144.
33. White AA, Panjabi MM. eds. Clinical Biomechanics of the Spine. 2nd ed, Philadelphia, Pa: Lippincott; 1990.
34. Sandhu HS. Kanim LEA, Girardi F, Cammisa FP, Dawson EG. Animal models of spinal instability and spinal fusion. In: An YH, Friedman RJ. eds. Animal Models in Orthopaedic Research. Boca Raton, Fla: CRC Press, 1999.
35. Bostman OM. Pihlajamaki HK. Late foreign-body reaction to an intraosseous bioabsorbable polylactic acid screw, A case report. J Bone Joint Surg Am. 1 998; 80: 1 79 1 - 1 794.
36. Bergsma, JE. de Bruijn WC, Rozema FR, Bos RR, Boering G. Late degradation tissue response to poly(L-lactide) bone plates and screws. Biomaterials. 1995; 16:25-31.
37. Bostman O, Hirvensalo E, Makinen J, Rokkanen P. Foreign-body reactions to fracture fixation implants of biodegradable synthetic polymers. J Bone Joint Surg Br. 1990; 72:592596.
38. Bostman OM. Osteoarthritis of the ankle after foreign-body reaction to absorbable pins and screws: a three- to nine-year follow-up study. J Bone Joint Surg Br. 1998; 80:333-338.
39. Tegnander A, Engebretsen L, Bergh K, Eide E, Holen KJ, Iversen OJ. Activation of me complement system and adverse effects of biodegradable pins of polylactic acid (Biofix) in osteochondritis dissecans. Acta Orthop. Scandinavica. 1994; 65:472-475.
40. Weiler A. Helling HJ, Kirch U, ZirbesTK, Rehm KE. Foreign-body reaction and the course of osteolysis after polyglycolide implants for fracture fixation. Experimental study in sheep. J Bone Joint Surg Br. 1 996; 78:369-376.
Key to Treatment and Study Design
Summary of Mean Radiographic Fusion Scores for the Biomechanical Sham Croup and the Treated Survival Groups
Summary of Histological Fusion Results for the Treated Survival Croups: Fusion Incidence and Percent