Surgical Procedure
Twelve seven-month-old Yucatan mini-pigs were used in the present study, which was approved by our Institutional Animal Care and Use Committee. The number of mini-pigs was selected on the basis of a sample size and power analysis performed on the variance in spinal deformity creation in three pilot animals. Sedation was achieved by means of an intramuscular injection of a combination of ketamine (25 mg/kg), xylazine (2 mg/kg), and atropine (0.05 mg/kg). Once sedated, the animals were shaved and weighed and then dorsoventral and lateral radiographs of the spine were made. Spinal length measurements from the base of the skull to the base of the tail were also obtained. Anesthesia induction was started with propofol (2 mg/kg), administered through an intravenous line. Direct laryngoscopy was used to intubate the trachea with a 5.0 to 5.5-mm-inner-diameter cuffed endotracheal tube, and anesthesia was maintained with volatilized isoflurane (1.5% to 2.5%). The right chest was prepared with Betadine and was draped in the standard sterile fashion.
After it had been established that an adequate level of anesthesia had been achieved, a right-sided double thoracotomy was performed, with use of a sterile surgical technique, between the seventh and eighth and the tenth and eleventh rib spaces to expose the thoracic spine. Instrumentation sites were prepared over four vertebral levels from T8 to T11. At each vertebral body, the overlying pleura was incised and the segmental vessels were cauterized. Instrumentation of each segment was then performed with use of one specially designed vertebral staple and screw (with a maximum outer diameter of 7.5 mm and a length of 35 mm) (DePuy Spine, Raynham, Massachusetts) . A large-diameter cancellous screw with staple design was used at each tethered level to maximize the bone-implant interface and to minimize the risk of implant failure (screw pullout). During insertion of the staple and screw, care was taken not to disturb the intervertebral discs or growth plates. The four vertebral body staples were aligned, and an ultra-high molecular weight polyethylene ribbon (cross section, 1.5 × 7.5 mm) was placed. The tether was secured to the caudad screw, was pulled taut manually, and was secured with the proximal locking nut. The middle two screws were then captured. As these nuts were tightened, slight tensioning of the tether occurred as the slack was removed from the tether.
Routine closure of the thoracotomy sites was performed over a chest tube. After the lung was reexpanded, the chest tube was removed and the subcutaneous tissues and skin were closed with use of absorbable sutures. After each procedure, analgesia was provided with a fentanyl patch (100 µg/hr) and with subcutaneous injections of Banamine (2 mg/kg) for the treatment of breakthrough pain as needed. Postoperative antibiotics consisted of sulfamethoxazole-trimethoprim (SMZ-TMP) (50 mg/kg, administered orally twice a day for five days). Once a month, the animals were sedated with an intramuscular injection of ketamine (20 mg/kg), xylazine (2 mg/kg), and atropine (0.05 mg/kg) so that radiographs could be made and body weight and spinal length measurements could be performed.
The first six mini-pigs were allowed to grow for six months after instrumentation, and the latter six mini-pigs were allowed to grow for twelve months after instrumentation. In each group, the spines were then harvested en bloc from T6 to L1. For each survival period (six months and twelve months), the first four spines were frozen for biomechanical testing and the last two spines were placed in 10% buffered formalin for undecalcified histologic processing. In addition, paraspinal lymph nodes from the animals in the twelve-month survival group were dissected and were placed in 10% buffered formalin for histologic processing to look for evidence of a foreign-body reaction. A flowchart is presented in Figure 1.
Radiographic Evaluation
Dorsoventral and lateral radiographs of the thoracic spine were made preoperatively, immediately after surgery, and at one-month intervals throughout each animal's survival period. After a consistent level of sedation had been ensured, radiographs were made on a radiolucent table with a built-in Bucky device and a tube-to-film distance set at 40 in (101.6 cm). The radiographs were scanned and were evaluated digitally with use of Spineview 2.4 software (SurgiView, Paris, France; ENSAM-LBM, Paris, France; CRCHUM-LIO, Montreal, Quebec, Canada; Argos, Paris, France). A premeasured radiopaque name plate was placed on each radiograph to identify the animal and to calibrate the digital measuring tool. For each radiograph, coronal and sagittal deformity creation was measured (in degrees) over the thoracic segments with instrumentation with use of the standard Cobb technique. In addition, to monitor axial growth, midvertebral body height was measured (in millimeters) on the four vertebrae with instrumentation and on the immediately adjacent vertebrae, one level cephalad to and one level caudad to the implants. Finally, monthly coronal vertebral and intervertebral disc wedging was measured (in degrees) over the motion segments with instrumentation as well as one control motion segment cephalad and caudad to the instrumentation. All radiographic measurements were performed digitally with use of high-resolution monitors, magnification, and contrast enhancement to ensure accuracy.
Computed tomography scans, including three-dimensional coronal and sagittal reconstructions, were performed on each spine after harvest. Vertebral body and intervertebral disc heights were measured (in millimeters) on midcoronal and midsagittal images, for each vertebra with instrumentation, with use of AMICAS Image Viewer software (version 5.1; AMICAS, Boston, Massachusetts). All measurements were made from three consecutive computed tomography images, and the values were averaged. The posterior-anterior and left-right differences in vertebral body and intervertebral disc heights were calculated.
T1 and T2-weighted magnetic resonance imaging was also performed on each harvested spine. These images were analyzed with regard to the presence of osteophytes, disc extension beyond the interspace, the shape of the nucleus pulposus, the presence of anular tears, end plate cartilage irregularities, the presence of Schmorl nodes, T2-weighted signal intensity changes, and signal intensity changes in the vertebral body marrow next to the end plate. In addition, midcoronal T2-weighted magnetic resonance images were used to measure migration of the nucleus pulposus for each tethered motion segment as the distance from the center of the intervertebral disc to the center of the nucleus pulposus (OsiriX Imaging Software, version 3.1; OsiriX, Geneva, Switzerland).
Histologic Evaluation
Two spines (eight vertebral bodies) from each survival period were prepared for undecalcified histologic analysis by Vet Path Services (Mason, Ohio) to evaluate the bone-screw interface. The vertebrae were first fixed in 10% buffered formalin for at least forty-eight hours and then were dehydrated with use of acetone and 70%, 95%, and 99% alcohol solutions. They were then infiltrated and embedded in methylmethacrylate. The polymerized sections were cut with use of a custom, water-cooled, high-speed, cutoff saw into a single 5-mm-thick midcoronal slab. These sections were then ground down to a thickness of 100 µm, were glued on plastic slides, and then were polished to an optical finish with use of a variable-speed grinding wheel (Buehler, Lake Bluff, Illinois). Each section was stained with Sanderson Rapid Bone Stain (Surgipath Medical Industries, Richmond, Illinois) and counterstained with Acid Fuchsin.
Paraspinal lymph node specimens from the animals in the twelve-month survival group were prepared for routine histologic analysis by Vet Path Services. The specimens were fixed in 10% buffered formalin for at least forty-eight hours; were dehydrated with use of acetone and 70%, 95%, and 99% alcohol solutions; and were embedded in paraffin. Next, 2.5-µm sections of the lymph node tissue were placed on glass slides and were stained with hematoxylin and eosin. These sections were evaluated for the presence of foreign bodies and for signs of infection or inflammation.
Biomechanical Evaluation
Nondestructive biomechanical testing was performed with use of a biaxial servohydraulic MTS 858 machine (MTS Systems, Eden Prairie, Minnesota). Before mechanical testing, the apical motion segment of each specimen was prepared for kinematic analysis. Kirschner wires (1.5 mm) were inserted in the midline in the T9 and T10 vertebral bodies, parallel to the plane of motion, and four 4-mm reflective markers were attached to each wire to define a rigid coordinate system.
Data on construct stiffness (measured in Newton-millimeters/degree) and range of motion (measured in degrees) were obtained, with the tether intact and then with the tether cut, in right and left torsion, right and left lateral bending, flexion, and extension. For flexion-extension and lateral bending, a cantilever beam applied unconstrained moments of ±4.5 Nm with use of a displacement control rate of 2 mm/sec. For axial torsion, each specimen was mounted in-line with the machine actuator and a 100-N axial load was applied with cyclic moments between ±2 Nm at an angle control rate of 2 deg/sec. Mechanical data for displacement (measured in millimeters), force (Newtons), angle (degrees), and torque (Newton-meters) were sampled at 10 Hz for the duration of each test.
Relative motion between the reflective markers was measured with use of a four-camera noncontact motion-measurement system (Qualisys, Gothenburg, Sweden) to obtain kinematic data. A custom software program (The MathWorks, Natick, Massachusetts) utilizing Euler angle calculation for the plane of motion being tested was used to measure apical range of motion in flexion-extension, right lateral bending, left lateral bending, and axial torsion. Kinematic data (measured in degrees) were sampled at 10 Hz for the duration of each test.
Statistical Methods
Preoperative, six-month, and twelve-month data on the coronal and sagittal Cobb angles and coronal vertebral body and intervertebral disc wedging were compared with use of repeated-measures analysis of variance. A Bonferroni correction for multiple comparisons was used to set the alpha value at p < 0.01. For the progression of coronal deformity, vertebral body wedging, and intervertebral disc wedging, a regression analysis was conducted to test the slope over time against zero (p < 0.05). Computed tomography measurements of vertebral body and intervertebral disc height differences were compared between six and twelve months with use of a univariate analysis of variance (p < 0.05). Biomechanical data for construct stiffness, construct range of motion, and apical range of motion were compared with use of multivariate analysis of variance (p < 0.05) with the duration of tethering (six or twelve months) and the state of the tether (intact or cut) as the two independent variables.
Perioperative Data
The average surgical time (and standard deviation) was 85 ± 17 minutes (range, sixty to 110 minutes), and the estimated blood loss was <10 mL in each case. There were no intraoperative or immediate postoperative complications, and all animals recovered from anesthesia without difficulty. Of note, one animal in the twelve-month survival group was found to have a deep abscess over the entire length of the instrumentation during harvest. Postmortem cultures demonstrated growth of Streptococcus dysgalactiae; however, the animal had been clinically healthy throughout its survival period, with no signs or symptoms of infection or pain. This animal was excluded from all radiographic and biomechanical analyses.
The average monthly body mass and spinal length measurements are shown in a table in the Appendix. At the end of six months the average increase in spinal length was 15%, whereas at twelve months the average increase in spinal length was 30%. The average monthly change in spinal length was calculated to be 2.0 ± 0.3 cm, and the average monthly increase in body mass was calculated to be 2.2 ± 0.4 kg.
Radiographic Data
For all radiographic data, six animals were included in the six-month survival group and five animals were included in the twelve-month survival group. Coronal and sagittal Cobb angles for T8 through T11 were measured on preoperative, immediate postoperative, and monthly radiographs (Figs. 2-A through 2-D). The average preoperative, six-month postoperative, and twelve-month postoperative values are recorded in Table I. A positive coronal Cobb value refers to an apex left curve, and a positive sagittal Cobb value refers to an apex posterior or kyphotic curve. Coronal spinal deformity was consistently created in each animal, with progression over the survival period (Fig. 3). However, there was no significant difference in the sagittal Cobb measurements at the three time-points (p = 0.81).
Monthly midvertebral body height measurements were performed on the four vertebrae with instrumentation and on control vertebrae cephalad and caudad to the instrumentation (see Appendix). In both groups, preoperative midvertebral body height measurements demonstrated a gradual increase in size from T7 (about 22 mm) to T12 (about 26 mm). The average changes in midvertebral body height for all eleven animals after six months of growth are shown in Figure 4. The apical vertebrae that had undergone instrumentation (T9 and T10) were found to have a significantly decreased change in height and were below the expected untethered range (p < 0.04).
Coronal vertebral body and intervertebral disc wedging were also measured (in degrees) on preoperative, postoperative, and monthly dorsoventral radiographs. Figures 5-A and 5-B demonstrate the average coronal wedging observed in each vertebral body and intervertebral disc with instrumentation as well as in cephalad and caudad control motion segments for all eleven animals. Vertebral body wedging, with decreased height on the side of the tether, was observed starting one month postoperatively and progressed over each animal's survival period. Preoperative vertebral body wedging was significantly less than that observed at six or twelve months (p = 0.001), and wedging observed at twelve months was significantly greater than that observed at six months (p = 0.001). Modest intervertebral disc wedging with decreased height on the side of the tether was present on the immediate postoperative radiographs, likely because of slight intraoperative tensioning of the tether during the tightening procedure. As time progressed, however, reverse disc wedging was observed, with increased disc height on the side of the tether (as indicated by the negative values in Figure 5-B) (p = 0.001). These findings were corroborated by computed tomography measurements of vertebral body and intervertebral disc heights.
Serial midcoronal and midsagittal postharvest computed tomography images were used to measure vertebral body and intervertebral disc heights. Differences in height in the coronal (left side-right side) and sagittal (posterior-anterior) planes were calculated (see Appendix). In the coronal plane, all vertebrae with instrumentation (T8, T9, T10, and T11) were noted to be taller on the left side as compared with the right. In the sagittal plane, the posterior aspect of the vertebral body of all four vertebrae with instrumentation was noted to be taller than the anterior aspect of the vertebral body. On the average, after six months of tethering, the left side and posterior aspect of the vertebrae were 4.4 ± 1.3 mm and 2.8 ± 1.7 mm taller than the right side and the anterior aspect, respectively. Thus, the anterolateral placement of the tether had caused the vertebrae to become wedged. Unlike the vertebral bodies, the intervertebral discs were noted to be wedged toward the tether at both six and twelve months. After six months of tethering, the right side and the anterior aspect of the disc were noted to be taller than the left side and the posterior aspect by 1.2 ± 1.3 mm and 3.7 ± 0.8 mm, respectively. After twelve months of growth, the discs were still wedged toward the tether; however, the difference in height was less obvious: 0.6 ± 1.4 mm in the coronal plane and 3.0 ± 0.7 mm in the sagittal plane.
T1 and T2-weighted magnetic resonance imaging was performed on each harvested spine, and the findings were evaluated by a radiologist. There was no evidence of intervertebral disc degeneration in any of the eleven specimens. It was not possible to distinguish T2 signal intensity between discs with and without instrumentation or between specimens that had undergone six months and twelve months of tethering. No osteophytes or obvious tears were observed in the anular structure of the discs. The end plates appeared to be undamaged, and there was normal signal intensity in the vertebral body next to the end plate. Of note, the nucleus pulposus was found to have shifted toward the tether in all specimens after six and twelve months of spinal growth modulation (Figs. 6-A and 6-B).
Migration of the nucleus pulposus was measured (in millimeters) for each tethered motion segment on midcoronal T2-weighted magnetic resonance images and was expressed as a percentage of total intervertebral disc width (see Appendix). Between six and twelve months, there was a significant difference in migration of the nucleus pulposus (1.1 ± 0.5 mm at six months, compared with 4.0 ± 0.5 mm at twelve months; p = 0.001) and in the migration as a percentage of total disc width (4.8% ± 0.4% at six months, compared with 19.7% ± 1.1% at twelve months; p = 0.001).
Histologic Data
All vertebral staples and screws were noted to have been placed without penetrating the intervertebral disc or damaging the physis. Evaluation of the vertebral body-screw interface revealed that there was good contact with cancellous bone, without evidence of implant failure (screw plowing or pullout). The paraspinal lymph node analysis revealed no evidence of foreign-body reaction, infection, or inflammation.
Biomechanical Data
Data on construct stiffness, construct range of motion, and apical range of motion were collected for each specimen with the tether intact and then with the tether cut (see Appendix). In flexion, extension, left lateral bending, and right lateral bending, construct stiffness increased significantly (p < 0.04) and construct range of motion decreased significantly (p < 0.006) between the six-month and twelve-month time-points; however, stiffness and range of motion in right and left torsion were not found to be significantly different (p > 0.1). After the tether was cut, construct stiffness in left lateral bending (away from the tether) decreased significantly (p = 0.04) at both six and twelve months and range of motion in left lateral bending increased significantly (p = 0.004) at both of these time-points. For the kinematic analysis, apical range of motion in right lateral bending (toward the tether) was found to decrease significantly (p = 0.001) between the six and twelve-month time-points. The interaction terms considering the duration of tethering and the state of the tether were not found to be significantly different for any of the variables collected, with the numbers studied.
Spinal growth modulation with an anterolateral flexible tether has been demonstrated previously in animal models. Braun et al. concluded that a bone anchor ligament tether could correct only coronal deformity in an experimentally created scoliosis goat model; however, this effect was lost over time21. Newton et al. used a flexible metallic cable with a vertebral staple-double screw construct in an immature bovine model and demonstrated the creation of coronal and sagittal spinal deformity over a six-month growth period18. The rapid vertebral growth rate in the bovine model, however, was thought to overexaggerate deformity creation, and additional testing with an anterolateral tether in an animal model with a growth rate more similar to that of adolescents was suggested. The purpose of the present study was to assess the ability of a novel vertebral staple-tapered screw construct and ultra-high molecular weight polyethylene ribbon to modulate spinal growth in the immature porcine model for six and twelve months. Deformity creation, vertebral and intervertebral disc wedging, spinal flexibility, intervertebral disc health, and the quality of the bone-screw interface were assessed.
An immature porcine model was chosen in the present study because a pilot analysis of three Yucatan mini-pigs demonstrated that the size and anatomy of the vertebrae, intervertebral discs, and thoracic cavity nearly approximated those in juvenile and adolescent humans. In addition, the pigs were thought to have a growth rate more similar to adolescents than the previously studied growing calf model18,20. In the present study, the pigs experienced an average 104% increase in body weight and an average 30% increase in spinal length over the twelve-month survival period (average vertebral body growth, 4 mm/year). In comparison, adolescents, on the average, experience a 70% increase in body weight and a 30% increase in sitting height over their two to three-year growth spurt. Average vertebral body growth in the adolescent population has been reported to be about 1 mm/year22. The bovine model, on the other hand, demonstrated an average 270% increase in body weight, an average 60% increase in spinal length, and average vertebral body growth of about 18 mm/year20.
The novel vertebral staple-screw-tether construct used in the present study was able to consistently create coronal spinal deformity and vertebral body-disc wedging in the coronal and sagittal planes in the porcine model. The construct maintained adequate osseous fixation, with no evidence of staple-screw pullout or screw loosening within the vertebral body identified on midcoronal undecalcified histologic sections. A previous study in the bovine model18 demonstrated that a single-screw construct resulted in screw back-out or plowing and levering within the vertebral body. This likely was due to the implant-bone fixation mechanism or to the rapid bovine vertebral growth rate. In the porcine model, monthly radiographs confirmed the creation of a coronal plane spinal deformity that was proportional to the duration of tethering.
Spinal deformity creation, however, was not observed in the sagittal plane. This finding was not expected and is different from the kyphosis creation reported following anterolateral spinal tethering in the bovine model18,19. In the porcine model, slight anterior wedging of the vertebral bodies was likely balanced by posterior wedging of the intervertebral discs. The ability to control spinal growth modulation in both the coronal and sagittal planes will be important during the clinical application of this tether; however, additional evaluation of the effect of compression created by the tether on vertebral growth plates is required. On the other hand, the normal changes in the porcine sagittal spinal profile during growth have not been described previously, and, as the present study did not include a sham control group, it is possible that the tether modulated spinal growth in the sagittal plane by limiting the development of normal lordosis.
Vertebral and intervertebral disc wedging resulting from fusionless spinal growth modulation has been reported previously in other animal studies19,23. In the current study, coronal vertebral body and intervertebral disc wedging measurements were obtained from preoperative, postoperative, and monthly radiographs. Vertebral body wedging, with decreased vertebral height on the side of the tether, was observed starting one month after surgery and progressed throughout the survival period (Fig. 5-A). Intervertebral disc wedging, however, did not follow the same trends (Fig. 5-B). Immediately postoperatively, the discs appeared to be wedged, with decreased height on the side of the tether (a positive measurement). The narrowing of discs on the tethered side of the spine was likely due to slight tensioning of the tether during the tightening procedure. However, after one month, there was no observable wedging, and from then on the disc became wedged in the reverse direction, with increased height on the side of the tether (a negative measurement). This disc height difference was more pronounced after six months of growth modulation (average, 1.2 ± 1.3 mm and 3.7 ± 0.8 mm in the coronal and sagittal planes, respectively) than after twelve months (average, 0.6 ± 1.4 mm and 3.0 ± 0.7 mm in the coronal and sagittal planes, respectively), despite progressively increasing disc angulation.
Magnetic resonance images of the harvested spines showed that the nucleus pulposus had shifted toward the tether (1.1 ± 0.5 mm at six months and 4.0 ± 0.5 mm at twelve months). This response was unexpected and may indicate a compensatory mechanism to maintain spinal alignment and to counter the vertebral wedging and spinal deformity created by the compressive forces of the tether. This phenomenon has not been reported previously; however, future studies will evaluate this finding further with biochemical assays and nuclear magnetic resonance spectroscopy to quantify the changes in disc composition that may be occurring during compression-induced spinal growth modulation.
Mechanical loading (compression) has been shown to accelerate intervertebral disc degeneration24-26. No evidence of disc degeneration was identified on magnetic resonance images of the harvested porcine spines after six and twelve months of growth. This could be because changes in disc water content, required to identify changes in T2 signal intensity, were not yet present in the porcine discs. Longer-term compressive loads or further magnetic resonance evaluation with T1-rho or T2-mapping studies will be required to further evaluate the early degenerative changes that may be occurring in these discs27-29.
Biomechanical testing showed that the spines in the present study were significantly stiffer and had significantly less range of motion in flexion, extension, and right and left lateral bending after twelve months of tethering as compared with after six months. Kinematic analysis demonstrated a similar trend in right lateral bending of the apical motion segment. Stiffness and range of motion in torsion, however, did not change significantly between the two time-points. It is not clear from these data if the increased stiffness and decreased range of motion were due to natural changes seen with increasing age or whether they were due to the tether. Biomechanical data on normal controls (animals followed for twelve months without a tether) are needed to determine the etiology of these findings. When the tether was cut, construct stiffness decreased significantly and range of motion increased significantly for left lateral bending at both survival time-points. These findings are expected and indicate that the tether primarily limited motion and increased stiffness in lateral bending away from the tether. Future studies should compare the effect of various fusionless implants (tether, staples, and growing rod) on the biomechanics of the spine over time.
Several limitations of the present study and the porcine model should be considered. Primarily, the present study lacked animals that could serve as negative controls (sham surgery) to compare the natural progression of spinal alignment and disc health for this animal model. Instead, comparisons in the present study were made between treated animals at the preoperative, six-month, and twelve-month time-points to evaluate the effect of spinal growth modulation over time. In addition, differences in spinal mechanical forces secondary to postural differences between a porcine quadruped and a bipedal human may affect the ability to modulate growth. The evaluation of disc health was performed in the present study after only six and twelve months of growth modulation, whereas these tethers would be in place in adolescents for several years before skeletal maturity is reached. The effect of a long-term compressive load and motion limitation on disc health will need to be studied. Finally, this is an animal model that creates deformity rather than corrects it. As such, the extent to which such a growth-modulating approach will be successful in arresting or correcting progressive scoliosis in the growing juvenile or adolescent patient with scoliosis is unclear. If these growth effects on the nonscoliotic porcine model translate to the human as we believe they will, the outlook for the treatment of scoliosis in growing patients may change substantially.
In conclusion, the present study confirms the ability of a flexible tether, attached to the vertebral bodies anteriorly, to limit vertebral growth adjacent to the tether much in the way physeal-bridging staples alter growth in long bones. Vertebral staples have been applied across the disc spaces in children and adolescents with scoliosis with the similar goal of limiting convex spinal growth. Fusionless techniques allowing the spine to grow straight over time remain the underlying goal that has prompted this research. We believe that the data from this animal model of growth modulation support the clinical concept of tethering the spinal growth in patients with scoliosis, yet further questions regarding the response of the disc to this mechanical perturbation remain, particularly with the identification of reverse disc wedging. It remains unclear how the discs of a patient with scoliosis will respond to such a tether.