Biomechanics of the lower thoracic spine after ... - The Spine Journal

1 downloads 0 Views 924KB Size Report
Thomas E. Mroz, MD a,b,c,*. aCleveland Clinic Center for Spine Health, Cleveland Clinic, 9500 Euclid Ave., S-40, Cleveland, OH 44195, USA. bDepartment of ...
The Spine Journal 14 (2014) 2216–2223

Basic Science

Biomechanics of the lower thoracic spine after decompression and fusion: a cadaveric analysis Daniel Lubelski, BAa,b,c, Andrew T. Healy, MDa,b, Prasath Mageswaran, PhDd, Edward C. Benzel, MDa,b,c, Thomas E. Mroz, MDa,b,c,* a

Cleveland Clinic Center for Spine Health, Cleveland Clinic, 9500 Euclid Ave., S-40, Cleveland, OH 44195, USA b Department of Neurological Surgery, Cleveland Clinic, 9500 Euclid Ave., S-40, Cleveland, OH 44195, USA c Cleveland Clinic Lerner College of Medicine, 9500 Euclid Ave., NA-21, Cleveland, OH 44195, USA d Spine Research Laboratory, Lutheran Hospital, Cleveland Clinic, 1730 W 25th St, Cleveland, OH 44113, USA Received 30 September 2013; revised 7 February 2014; accepted 16 March 2014

Abstract

BACKGROUND CONTEXT: Few studies have evaluated the extent of biomechanical destabilization of thoracic decompression on the upper and lower thoracic spine. The present study evaluates lower thoracic spinal stability after laminectomy, unilateral facetectomy, and unilateral costotransversectomy in thoracic spines with intact sternocostovertebral articulations. PURPOSE: To assess the biomechanical impact of decompression and fixation procedures on lower thoracic spine stability. STUDY DESIGN: Biomechanical cadaveric study. METHODS: Sequential surgical decompression (laminectomy, unilateral facetectomy, unilateral costotransversectomy) and dorsal fixation were performed on the lower thoracic spine (T8–T9) of human cadaveric spine specimens with intact rib cages (n510). An industrial robot was used to apply pure moments to simulate flexion-extension (FE), lateral bending (LB), and axial rotation (AR) in the intact specimens and after decompression and fixation. Global range of motion (ROM) between T1–T12 and intrinsic ROM between T7–T11 were measured for each specimen. RESULTS: The decompression procedures caused no statistically significant change in either global or intrinsic ROM compared with the intact state. Instrumentation, however, reduced global motion for AR (45 vs. 30 , p5.0001), FE (24 vs. 19 , p5.02), and LB (47 vs. 36 , p5.0001) and for intrinsic motion for AR (17 vs. 4 , p5.0001), FE (8 vs. 1 , p5.0001), and LB (12 vs. 1 , p5.0001). No significant differences were identified between decompression of the upper versus lower thoracic spine, with trends toward significantly greater ROM for AR and lower ROM for LB in the lower thoracic spine. CONCLUSIONS: The lower thoracic spine was not destabilized by sequential unilateral decompression procedures. Addition of dorsal fixation increased segment rigidity at intrinsic levels and also reduced overall ROM of the lower thoracic spine to a greater extent than did fusing the upper thoracic spine (level of the true ribs). Despite the lack of true ribs, the lower thoracic spine was not significantly different compared with the upper thoracic spine in FE and LB after decompression, although there were trends toward significance for greater AR after decompression. In certain patients, instrumentation may not be needed after unilateral decompression of the lower thoracic

FDA device/drug status: Not applicable. Author disclosures: DL: Nothing to disclose. ATH: Nothing to disclose. PM: Nothing to disclose. ECB: Royalties: Elsevier Pub (B), Thieme Pub (B); Stock Ownership: Axiomed (NA), Depuy (NA), Orthomems (NA), Turning Point (NA); Consulting: Axiomed; Speaking and/or Teaching Arrangements: Multiple (Varied); Trips/Travel: Multiple (Varied); Grants: OREF (F, Paid directly to institution), Rawlings (F, Paid directly to institution). TEM: Stock Ownership: PearlDiver Inc (No monies received); Consulting: Globus Medical (B); Speaking and/or Teaching Arrangements: AO Spine (B). The disclosure key can be found on the Table of Contents and at www. TheSpineJournalOnline.com. 1529-9430/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.spinee.2014.03.026

This study was supported in part by funds received through the Cleveland Clinic Stanley Zielony Spinal Research & Education Fund, the Neurosurgery Research and Education Foundation (NREF) and the Cleveland Clinic Research Programs Committee. * Corresponding author. Departments of Neurological Surgery, Center for Spine Health, The Cleveland Clinic, 9500 Euclid Ave., S40, Cleveland, OH 44195, USA. Tel.: (216) 445-9232; fax: (216) 363-2040. E-mail address: [email protected] (T.E. Mroz)

D. Lubelski et al. / The Spine Journal 14 (2014) 2216–2223

2217

spine; further validation and additional clinical studies are warranted. Ó 2014 Elsevier Inc. All rights reserved. Keywords:

Upper thoracic spine; Lower thoracic spine; Decompression; Costotransversectomy; Biomechanics; Cadaver

Introduction Biomechanical evaluation of spinal stability is necessary to better understand the optimal surgical treatments for a given pathology. Although the biomechanical effects of destabilization have been previously quantified in the cervical [1–3] and lumbar spine [4–10], there is limited biomechanical evidence measuring the destabilizing effects of various decompressive procedures in the thoracic region with an intact rib cage. Previous studies have shown that the costovertebral joints, rib heads, and dorsal elements are important contributors to the stability of the thoracic spine [11–15]. It is unclear if, and/or to what extent, decompression surgery destabilizes the thoracic spine. Moreover, it is unclear whether the contribution of the true ribs to the upper thoracic spine leads to more biomechanical stability as compared with the lower thoracic spine supported by only false or floating ribs. Defining stability after sequential decompressions is critical to provide an understanding of when instrumentation may be needed. Presently, guidelines that outline when instability is rendered and when fixation is needed do not exist. Furthermore, it is unclear if there are differences in stability patterns between the upper and lower thoracic spine in the intact state or after various stages of surgical decompression. In the present study, we sought to use a unique cadaveric model in which thoracic spine specimens were manipulated in a clinically relevant manner to define the destabilizing and stabilizing effects of operative decompression and instrumentation, respectively. To maximize clinical applicability, we used intact thoracic spine (ie, intact thoracic cage) and the same surgical tools that would be used clinically. Our hypothesis was that decompressing the lower thoracic spine, at the level of the floating ribs, significantly destabilizes the spine, and the increased range of motion (ROM) relative to the native state would require supplemental spinal instrumentation and fixation.

of the data by pathology or poor bony quality, computed tomography and dual-energy X-ray absorptiometry scans of each specimen were carried out to determine preexisting spinal pathology/fusion and the bone mineral density (BMD) of each specimen before biomechanical testing. Average BMD was 0.762 g/cm2, indicating a bone density within normal physiologic parameters. Specimens with previous spinal surgery, spinal implants, anatomical deformities, or structural defects/fractures were excluded from this study. One patient was documented as being diagnosed with osteoarthritis; however, no spinal deformity was found that met criteria for exclusion. Each specimen was eviscerated and dissected including the removal of the skin and subcutaneous tissue, lattismus dorsi, serratus posterior inferior, and erector spinae, while preserving the integrity of the spine, ribs, and sternum, along with the intercostal musculature and associated costosternovertebral ligaments. The intercostal muscles were left intact, as they did not obstruct the procedure and their presence was not felt to falsely immobilize or stiffen the construct. Unlike the bulky paraspinal musculature, the intercostal muscles are thin and provide connectivity to maintain a physiologic orientation. Furthermore, our primary goal in specimen preparation was to retain the integrity of ligamentous structures. Removal of the intercostal muscles could potentially disrupt ligamentous integrity as the ribs converge with the spine and sternum, both dorsally and ventrally. All specimens were then stored in a 20 C freezer at the Spine Research Laboratory at Cleveland Clinic. The day before testing, each specimen was thawed to room temperature overnight. The cranial (C7–T1) and caudal (T12–L1) levels were potted in custom test fixtures and further secured using wood screws inserted into the cranial and caudal vertebral bodies and embedded in Cereband, a liquid metal alloy (HiTech Alloys, Squamish, WA, USA). Throughout the surgical procedures, the specimens were lightly sprayed with saline solution to keep them moist. Surgical procedures

Methods Specimens This study used ten fresh-frozen human cadaveric thoracic specimens that included the sternum, ribs, and spine from C7 to L1, with all articulations intact. These specimens consisted of four men and six women, with a median age of 60.5 years (range: 43 to 70 years), height of 169 cm (range: 152 to 185 cm), and weight of 77 Kg (range: 50 to 113 Kg) (Table 1). To ensure that there was no confounding

Each specimen underwent a series of three sequential decompressive procedures from T8–T9, followed by instrumentation at T7–T11. These procedures included laminectomy, unilateral factectomy, unilateral costotransversectomy, and pedicle screw fixation. Laminectomies were performed at T8 and T9 and the inferior 20% to 30% of the T7 spinous process was removed. This was performed without disrupting ligamentous integrity of cranial to T7 and caudal to T10–11. Facetectomy was performed on the right side and involved complete

2218

D. Lubelski et al. / The Spine Journal 14 (2014) 2216–2223

Table 1 Specimen demographics and average bone density from DEXA scans Specimen

Age (y)

Height (cm)

Weight (Kg)

Sex

Smoker

Cause of death

Conditions

BMD

1 2 3 4 5 6 7 8 9 10

62 43 69 69 57 70 69 56 55 59

157 178 175 173 173 152 163 157 180 185

86 113 87 63 58 50 73 76 — 86

Female Female Male Male Female Female Female Female Male Male

Yes No Yes Yes Yes Yes Yes Yes Yes Yes

Cirrhosis Atherosclerosis Cardiac arrest Brain damage CHF COPD Kidney disease Stroke Natural causes NA

Chemotherapy, asthma, jaundice HTN HTN, PVD, CVD CVA, seizure COPD, PNA, Chronic Bronchitis Hypokalemia Anemia, HTN Hemorrhagic CVA Osteoarthritis HTN

0.707 0.723 0.145 1.083 0.852 0.642 1.15 0.787 0.603 0.930

DEXA, dual-energy X-ray absorptiometry; BMD, bone mineral density; CHF, coronary heart failure; COPD, chronic obstructive pulmonary disease; HTN, hypertension; PVD, peripheral vascular disease; CVD, cardiovascular disease; CVA, cardiovascular accident; PNA, pneumonia; NA, not available. Note: BMD represents average bone mineral density in g/cm2.

resection of the T8 inferior articular process and T9 superior articular process. Costotransversectomy included removal of the T9 transverse process, pedicle, and 3 to 4 cm of rib including its articulations with the vertebral column. The instrumentation procedure required decortications of an entry point, following a ‘‘straight forward’’ technique [18], with a high speed drill. Then the appropriate trajectory through cancellous bone was determined with a Medtronic ‘‘pedicle finder’’ blunt probe. This trajectory was then tapped with a 4.0 mm cortical tap and adequate pedicle placement was confirmed with a sound probe. Pedicle screws were joined by a 5.5 cm titanium rod. Spinal implants and associated surgical tools were supplied by Medtronic (Medtronic Sofamor Danek, Memphis, TN, USA). Medtronic screws (5.035 mm) were placed in the bilateral pedicles of the two supraadjacent and subadjacent levels, using the free hand technique [18]. The aforementioned procedures were performed by a fellowship trained spine surgeon (TEM) with the assistance of a neurosurgical resident (ATH) at the Cleveland Clinic.

The system has a measurement accuracy of 60.1 mm in translation and 60.13 in rotation. The associated Optotrak software provided readily extractable smart marker position and orientation data that was then processed to obtain the relative ROM between each vertebral body. The robotic system and motion capture system were synchronized to begin testing and data recording at the same time to obtain matching loading and motion data. Posttest analyses were used to determine global ROM between T1 and T12 and intrinsic intersegmental ROM between T7 and T11. The test configurations are shown in Figs. 1 and 2.

Multidirectional biomechanical testing An industrial robot (KUKA Robotics GmbH, Augsburg, Germany) at the Spine Research Laboratory at Cleveland Clinic, was used to carry out multidirectional biomechanical flexibility tests on the 10 specimens. Applied load was monitored continuously using a six-axis force-moment sensor system (Gamma; ATI Industrial Automation, Apex, NC, USA). The sensor also provided feedback to ensure that a pure moment was applied along the primary axis of motion. Relative three-dimensional (3D) vertebral motion was measured using an optoelectronic motion capture system (Optotrak; Northern Digital, Inc., Waterloo, Ontario, Canada.) at a rate of 20 Hz. The motion capture system determines the 3D kinematics of the spine by tracking the position and orientation of a smart marker placed on the spinous process of a vertebra. These smart markers are basically a set of three infrared emitting diodes, mounted onto a ‘‘marker carrier,’’ which in turn is rigidly attached to the spinous process using a 3.5 mm stainless steel threaded rod.

Fig. 1. Photograph of the cadaveric thoracic torso with instrumented fusion.

D. Lubelski et al. / The Spine Journal 14 (2014) 2216–2223

2219

as the midpoint of the two vertebral origins. We used a digitizing probe to realign the camera coordinate system to the local spine-specific coordinate system as recommended by International Society of Biomechanics standard. Analysis Posttest analyses were used to determine global (T1–T12) and intrinsic intersegmental (T7–T11) 3D rotations across supra- and subadjacent levels at the peak applied moment. Data for each surgical state was normalized against the intact spine condition. Each specimen thus acted as its own control and thereby, interspecimen variability was addressed. Statistical analysis was performed using Minitab 16 (Minitab, Inc., State College, Pennsylvania, USA). A repeated measures analysis of variance was used to compare the segmental ROM after each surgical treatment with the intact segmental ROM. Post hoc Tukey-Kramer analysis was used for pairwise comparisons. Fisher exact tests were used to compare between the upper and lower thoracic spine intrinsic ROMs (T3–T7 vs. T7–T11) and to compare the global ROM after surgery of T4–T5 (upper thoracic) versus surgery at T8–T9 (lower thoracic). Statistical significance was set at p#.05.

Fig. 2. Photograph of the robotic spine testing system (Kuka Robot) along with a cadaveric thoracic torso fitted to the custom fixture.

Biomechanical tests were carried out in three orthogonal directions for each of the five aforementioned states: intact specimens; laminectomy; laminectomy with unilateral factectomy; laminectomy with unilateral costotransversectomy; laminectomy with unilateral costotransversectomy plus pedicle screw fixation. These three test directions correspond to flexion-extension (FE), bilateral lateral bending (LB), and bilateral axial rotation (AR) of the thoracic spine. Each specimen was subjected to a pure moment of 65 Nm in each primary axis of motion, while continuously minimizing the off-axis loads. As previously described [17], our test system minimized off-axis forces and moments through an active feedback system that communicated through the control software. Pure moment loads were continuously applied, while there was simultaneous dynamic optimization of the motion path that minimized off-axis loads. To eliminate viscoelastic effects, the specimens were initially preconditioned before actual testing. The specimens were unconstrained so as to allow for natural coupled rotation motion of the spine. Additional information regarding the programming of the robot and coordinate system analysis has been described previously [17]. Briefly, the coordinate system was based on the International Society of Biomechanics 2002 standard [19], in which the origin is defined as the intersection of the proximal and distal y-axes in the reference, neutral position. In our system, the midpoint between adjacent end plates was estimated

Results A total of 10 specimens were tested in the intact state, followed by the aforementioned surgical procedures. Global ROM was measured between T1 and T12 in all 10 specimens, and intrinsic ROM from T7 to T11 was measured in seven of the specimens (Optoelectric sensors were initially placed at T1 and T12, however after testing the initial three specimens, we found minimal change between intact and costotransversectomy conditions. It was felt that measurement across a more focal segment, using sensors (T7 and T11), would better capture any postsurgical instability. These measurements were included in the subsequent seven specimens.). We found that in all three planes of motion, the sequential decompressive procedures caused no statistically significant change in motion between T1 and T12 (Table 2) or T7 and T11 (Table 3) when compared with intact. In comparing between intact and instrumented specimens, our study found that instrumentation reduced global motion (T1–T12) for AR (45 vs. 30 , p5.0001), FE (24 vs. 19 , p5.02), and LB (47 vs. 36 , p5.0001), and intrinsic motion (T7–T11) for AR (17 vs. 4 , p5.0001), FE (8 vs. 1 , p5.0001), and LB (12 vs. 1 , p5.0001). The biomechanical global and intrinsic ROM data obtained for the lower thoracic decompression and fusion were compared with the data obtained in a previous, parallel study [16] by our group, investigating the upper thoracic spine biomechanics (Tables 4 and 5). The lower thoracic spine was observed to be associated with trends (p#.1) toward greater intrinsic axial ROM than the upper thoracic

2220

D. Lubelski et al. / The Spine Journal 14 (2014) 2216–2223

Table 2 Mean global ROM in axial rotation, flexion-extension, and lateral bending Global ROM (n510)

Axial rotation (  )

Flexion-extension (  )

Lateral bending (  )

Intact Laminectomy Facetectomy Costotransversectomy Fusion

44.91618.79 44.592618.88 (p5.99) 44.64619.5 (p5.99) 44.52619.28 (p5.99) 30.35616.22 (p5.0001)

23.72611.36 24.3368.28 (p5.99) 24.7667.84 (p5.72) 25.4768.44 (p5.94) 19.3767.81 (p5.02)

47.44614.77 47.14614.44 47.46614.55 47.64614.69 35.59612.61

(p5.99) (p5.99) (p5.99) (p5.0001)

ROM, range of motion. Note: All ROM values are presented as mean6standard deviation. p Values were calculated with respect to the intact state; p#.05 was considered statistically significant.

spine, in the intact and decompressed state. In contrast, FE and LB ROMs were greater, although not significantly, in the upper as compared with the lower thoracic spine. In addition, the percent decrease in global axial ROM from the intact to the instrumented state was significantly greater for the lower thoracic spine than the upper thoracic spine at the level of true ribs.

Discussion Dorsal and dorsolateral decompression procedures of the thoracic spine are used to treat degeneration, lateral disc herniations, trauma and fractures, dorsal, dorsolateral, and ventrolateral tumors, and infectious processes such as retropleural abscesses [20,21]. Yet, there are limited guidelines regarding the destabilizing effects of these decompressive procedures and when subsequent instrumentation is needed. This study demonstrated that the sequential decompression of laminectomy, facetectomy, and costotransversectomy did not significantly destabilize the lower thoracic spine as compared with the intact state. Our hypothesis that decompression at the level of the floating ribs would destabilize and increase ROM of the lower thoracic spine was proved wrong. This was true for AR, FE, and LB, when measuring both global ROM (from T1 to T12) and intrinsic ROM (from T7 to T11). In contrast, after pedicle screw fixation, there was a statistically significant decrease in ROM in all measured axes. These findings are concordant with previous studies investigating sequential destabilization of the thoracic spine. Feiertag et al. [11] conducted a similar cadaveric analysis using specimens with intact spinal musculature,

rib cage, and sternum. They measured the biomechanical effects of discectomy, rib head resection, and facetectomy on T8–T9, using mechanical testing (weights, pulleys, digital goniometers). They found that even with the combined rib head resection and discectomy, there was no significant increase in ROM. The previous studies that have investigated the various stabilizing contributions to the thoracic spine have found that the boney articulations account for a minority of the stability, whereas the intervertebral discs, supra/interspinous ligaments (SIL), and rib cage contribute majority of the stability [13,22–24]. The SIL has been shown to substantially contribute to the flexion ROM in the thoracic spine [23]. The previously mentioned study by Feiertag et al. [11], and our own present study, demonstrate that after removal of the rib heads, ROM is not significantly affected. We additionally measured the effects of a laminectomy, and removal of the SIL, on the stability of the lower thoracic spine. Here too, our results are concordant with a previous study by Allen et al. [25] that showed no change in ROM or stability after removal of the lamina and SIL. The studies that have investigated the individual biomechanical contributions of components of the thoracic spine have typically emphasized the stabilizing effect of the rib cage, or ‘‘the fourth column’’ as coined by Berg [26], which has been shown to contribute up to 78% of the thoracic spinal stability [27], contributing approximately 40% to FE, 35% to LB, and 31% to AR [28]. Oda et al. [12] used canine thoracic spines with rib cages and demonstrated the largest increase in neutral zone (NZ), LB, and AR after sectioning of the costovertebral joints at T7 and destruction of the rib cage. Recently, Brasiliense et al. [27] used cadaveric models and removed 25%, 50%, 75%, and 100% of the

Table 3 Mean intrinsic ROM in axial rotation, flexion-extension, and lateral bending Intrinsic ROM (n57)

Axial rotation (  )

Flexion-extension (  )

Lateral bending (  )

Intact Laminectomy Facetectomy Costotransversectomy Fusion

17.2867.41 18.3768.06 18.1868.01 18.4268.09 4.0862.00

8.4963.79 8.2663.29 8.5563.52 8.8264.13 0.6460.31

11.7365.13 11.8165.43 11.8265.55 12.0465.66 0.9560.47

(p5.96) (p5.95) (p5.98) (p5.0001)

(p5.99) (p5.99) (p5.99) (p5.0001)

ROM, range of motion. Note: All ROM values are presented as mean6standard deviation. p Values were calculated with respect to the intact state; p#.05 was considered statistically significant.

(p5.99) (p5.99) (p5.99) (p5.0001)

D. Lubelski et al. / The Spine Journal 14 (2014) 2216–2223 Table 4 Comparison of global (T1–12) ROM after surgery in T4–T5 versus T8–T9 Global ROM Axial rotation Intact Laminectomy Facetectomy Costotransversectomy Fusion % decrease intact to fusion Flexion-extension Intact Laminectomy Facetectomy Costotransversectomy Fusion % decrease intact to fusion Lateral bending Intact Laminectomy Facetectomy Costotransversectomy Fusion % decrease intact to fusion

T4–T5 (n59)

T8–T9 (n510)

p

43.69616.87 43.90619.3 44.30619.71 44.13619.54 36.59615.36 16.47

44.91618.79 44.59618.88 44.64619.5 44.52619.28 30.35616.22 32.5

.9 .9 .9 .9 .4 .03*

26.90610.00 28.7869.43 29.1269.30 29.2169.41 23.6867.12 8.5

23.72611.36 24.3368.28 24.7667.84 25.4768.44 19.3767.81 8.4

.5 .3 .3 .4 .2 .99

42.06619.04 40.84618.08 43.21617.30 43.47617.10 34.31612.56 8.4

47.44614.77 47.14614.44 47.46614.55 47.64614.69 35.59612.61 22.4

.5 .4 .6 .6 .8 .4

ROM, range of motion. Note: % decrease intact to fusion represents the average percent decrease in ROM from intact to fusion states, based on the % decrease of the individual cadavers (not based on % decrease of the average values). All ROM values are presented as mean6standard deviation. ROM values for T4–T5 were derived from the study by Healy et al. [16]. p Values were calculated with respect to the intact state; p#.05 was considered statistically significant. * p!.05.

rib cage to determine that the rib cage added 181% increase to flexion and 702% to extension stiffness. In the present study, we sought to destabilize the thoracic spine using clinically relevant decompression procedures, including laminectomy, facetectomy, and costotransversectomy. We found no difference in ROM after decompression of the lower thoracic spine (despite the lack of true ribs stabilizing this region). Similarly in a previous parallel study of the upper thoracic spine (at the level of the true ribs), we also did not find a destabilization after decompressive procedures [18]. The intrinsic degrees of motion in the intact state did vary slightly between upper and lower thoracic specimens. In AR, T7–T11 moved 17.28 67.41 versus T3–T7, which moved only 11.15 64.34 (p5.1). This slight difference remained constant and there was no difference in the percent change in ROM after decompression between the two groups. Although no significant differences were found between the upper and lower thoracic spine after decompressive procedures, the lower thoracic spine may have a trend toward greater flexibility in terms of AR. Notably, after pedicle screw fixation, the lower thoracic spine had a significantly greater percent decrease in axial ROM than did the upper thoracic spine. In a recent 3D imaging study by Fujimori et al. [29], the authors similarly demonstrated greater AR in the lower thoracic (T6–T11) relative to the upper thoracic spine (T1–T6). Further

2221

investigation, including larger sample sizes, is needed to validate these findings. The present study sought to shed light on the clinical situations that require laminectomy with unilateral facetectomy or costotransversectomy, in which there is disruption of the SIL, unilateral facet, and the costovertebral joint, but with sparing of the anterior column. The segmental disruption that occurs during a costotransversectomy has not been previously investigated using thoracic spines with intact rib cages. The present study demonstrated that there is no significant loss of stability or increase in ROM as compared with the intact state after unilateral laminectomy, unilateral facetectomy, or unilateral costotransversectomy. As predicted, a statistically significant decrease in range of AR, FE and LB for both the whole spine and the segmental levels was shown after instrumentation. The data presented herein suggest that instrumentation may not be needed after unilateral facetectomy or costotransversectomy (in either the upper or lower thoracic spine) and may expose the patient to unnecessarily increased operative time, blood loss, morbidity, and a decreased ROM, especially at the lower thoracic levels. Importantly, although this study suggests that instrumentation may not be necessary for stability after these decompressive procedures, it is not designed to assess the clinical ramifications of such surgery. Specifically, it is unclear if patients would have pain related to unilateral facetectomy or constotransversectomy, and it is unclear if decompression without instrumentation would lead to progressive kyphosis; a clinical study is required to define these aspects. Table 5 Comparison of intrinsic (T3–T7/T7–T11) ROM after surgery in T4–T5 versus T8–T9, respectively Global ROM Axial rotation Intact Laminectomy Facetectomy Costotransversectomy Fusion Flexion-extension Intact Laminectomy Facetectomy Costotransversectomy Fusion Lateral bending Intact Laminectomy Facetectomy Costotransversectomy Fusion

T4–T5 (n55)

T8–T9 (n57)

p

11.1564.34 10.6064.94 10.7964.78 10.9464.50 1.9762.74

17.2867.41 18.3768.06 18.1868.01 18.4268.09 4.0862.00

.1 .09 .1 .09 .2

10.6265.27 11.4166.02 11.7866.00 11.8465.94 0.72160.54

8.4963.79 8.2663.29 8.5563.52 8.8264.13 0.6460.31

.4 .3 .3 .3 .5

17.9667.09 16.4368.04 16.7367.33 16.7667.14 1.6560.37

11.7365.13 11.8165.43 11.8265.55 12.0465.66 0.9560.47

.1 .3 .2 .2 .02*

ROM, range of motion. Note: All ROM values are presented as mean6standard deviation. ROM values for T4–T5 were derived from the study by Healy et al. [16]. p Values were calculated with respect to the intact state; p#.05 was considered statistically significant.

2222

D. Lubelski et al. / The Spine Journal 14 (2014) 2216–2223

There are several limitations of the present study that must be considered when interpreting the data. The most important point is the age range of the specimens studied. Given the nature of using cadaveric specimens, our specimens ranged from 43 to 70 years. Despite seeking out younger specimens, we were only able to obtain specimens with a median age of 60.5 years. There is likely some correlation to segmental motion, age, and degeneration [30– 32]; however, this has not been directly answered in the thoracic spine. Before any confident statement can be made regarding the biomechanical stability of this region, we feel a similar study would need to be carried out with specimen age ranging from 20 to 40 years. Furthermore, the investigation used cadaveric specimens that differ from human patients because of the tissue handling process. Moreover, many of these patients had comorbidities that may have affected their spinal integrity. To account for this, BMD of all specimens was measured before use and computed tomography images were taken to ensure that there were no anomalous aspects to these spine/rib specimens. Another limitation is that only 10 specimens were used for the aforementioned procedures and biomechanical analysis. Although this is in line with many of the previous biomechanical studies, it is still a relatively low sample size and coupled with the large standard deviation and variability between specimens, statistical significance of smaller, more subtle changes would not be detected. Similarly, optical markers were placed at only four points, and a construct that allowed for additional markers/sensors would have enabled more specific measurements. However, the presence of the rib cage made it difficult to place optical markers ventrally, and we were unable to place the markers at the levels directly above and below and had to place it closest to the level where dorsal elements were undisturbed. The loading force vectors must also be considered when interpreting the data. Although a uniform load distribution was applied along the spine and rib cage, the lack of anterior musculature (as is the standard in these biomechanical models) will have limitations in its ability to reflect the physiologic state. Lastly, preoperative ankylosis was not accounted for, and with increasing age, ankylosis, and stiffness, there is a concomitant decrease in ROM. We are focusing future biomechanical studies to evaluate differences between spines of different ages and determine whether severities of ankylosis affect ROM or degree of destabilization after decompression procedures.

Conclusion Thoracic spine stability was not significantly affected by sequential decompression procedures in lower thoracic segments in all three planes of motion in intact specimens. Pedicle screw fixation increased specimen rigidity and reduced AR to a significantly greater extent than was seen after pedicle screw fixation of upper thoracic levels. The

potential avoidance of thoracic instrumentation after unilateral decompression has significant clinical implications and warrants further investigation.

References [1] DeVries NA, Gandhi AA, Fredericks DC, et al. Biomechanical analysis of the intact and destabilized sheep cervical spine. Spine 2012;37:E957–63. [2] Lee M-Y, Chang G-L, Chang J-H, et al. Biomechanical evaluation of cervical spine fixation after healing in a destabilized cervical spine model in sheep: a comparison of the anterior plating and posterior wiring techniques. J Trauma 2006;60:1307–14. [3] Kretzer RM, Hsu W, Hu N, et al. Adjacent-level range of motion and intradiscal pressure after posterior cervical decompression and fixation: an in vitro human cadaveric model. Spine 2012;37:E778–85. [4] Schmoelz W, Erhart S, Unger S, Disch AC. Biomechanical evaluation of a posterior non-fusion instrumentation of the lumbar spine. Eur Spine J 2012;21:939–45. [5] Harris BM, Hilibrand AS, Savas PE, et al. Transforaminal lumbar interbody fusion: the effect of various instrumentation techniques on the flexibility of the lumbar spine. Spine 2004;29:E65–70. [6] Kiapour A, Ambati D, Hoy RW, Goel VK. Effect of graded facetectomy on biomechanics of Dynesys dynamic stabilization system. Spine 2012;37:E581–9. [7] Wilke H-J, Schmidt H, Werner K, et al. Biomechanical evaluation of a new total posterior-element replacement system. Spine 2006;31: 2790–6; discussion 2797. [8] Healy AT, Lubelski D, Mageswaran P, et al. Biomechanical analysis of the upper thoracic spine after decompressive procedures. Spine 2013 Dec 1; http://dx.doi.org/10.1016/j.spinee.2013.11.035. pii: S15299430(13)01930-X. [Epub ahead of print]. [9] Pintar FA, Cusick JF, Yoganandan N, et al. The biomechanics of lumbar facetectomy under compression-flexion. Spine 1992;17:804–10. [10] Goel VK, Goyal S, Clark C, et al. Kinematics of the whole lumbar spine. Effect of discectomy. Spine 1985;10:543–54. [11] Feiertag MA, Horton WC, Norman JT, et al. The effect of different surgical releases on thoracic spinal motion. A cadaveric study. Spine 1995;20:1604–11. [12] Oda I, Abumi K, L€u D, et al. Biomechanical role of the posterior elements, costovertebral joints, and rib cage in the stability of the thoracic spine. Spine 1996;21:1423–9. [13] Oda I, Abumi K, Cunningham BW, et al. An in vitro human cadaveric study investigating the biomechanical properties of the thoracic spine. Spine 2002;27:E64–70. [14] White AA 3rd, Hirsch C. The significance of the vertebral posterior elements in the mechanics of the thoracic spine. Clin Orthop Relat Res 1971;81:2–14. [15] Yoganandan N, Maiman DJ, Pintar FA, et al. Biomechanical effects of laminectomy on thoracic spine stability. Neurosurgery 1993;32: 604–10. [16] Healy AT, et al. Biomechanical analysis of the upper thoracic spine following decompressive procedures. Spine J 2013. [17] Mageswaran P, Techy F, Colbrunn RW, et al. Hybrid dynamic stabilization: a biomechanical assessment of adjacent and supraadjacent levels of the lumbar spine. J Neurosurg Spine 2012;17:232–42. [18] Kim YJ, Lenke LG. Thoracic pedicle screw placement: free-hand technique. Neurol India 2005;53:512–9. [19] Wu G, et al. ISB recommendation on definitions of joint coordinate system of various joints for the reporting of human joint motion: part I: ankle, hip, and spine. J Biomech 2002;35:543–8. [20] Lubelski D, et al. Lateral extracavitary, costotransversectomy, and transthoracic thoracotomy approaches to the thoracic spine: review of techniques and complications. J Spinal Disord Tech 2013;26: 222–32.http://dx.doi.org/10.1097/BSD.0b013e31823f3139.

D. Lubelski et al. / The Spine Journal 14 (2014) 2216–2223 [21] Lubelski D, et al. Lateral extracavitary vs costotransversectomy approaches to the thoracic spine: reflections on lessons learned. Neurosurgery 2012;71:1096–102. [22] Horton WC, et al. The role of the sternum, costosternal articulations, intervertebral disc, and facets in thoracic sagittal plane biomechanics: a comparison of three different sequences of surgical release. Spine 2005;30:2014–23. [23] Anderson AL, McIff TE, Asher MA, et al. The effect of posterior thoracic spine anatomical structures on motion segment flexion stiffness. Spine 2009;34:441–6. [24] Kretzer RM, et al. The effect of spinal instrumentation on kinematics at the cervicothoracic junction: emphasis on soft-tissue response in an in vitro human cadaveric model. J Neurosurg Spine 2010;13:435–42. [25] Allen BL Jr, Tencer AF, Ferguson RL. The biomechanics of decompressive laminectomy. Spine 1987;12:803–8.

2223

[26] Berg EE. The sternal-rib complex. A possible fourth column in thoracic spine fractures. Spine 1993;18:1916–9. [27] Brasiliense LBC, et al. Biomechanical contribution of the rib cage to thoracic stability. Spine 2011;36:E1686–93. [28] Watkins R 4th, et al. Stability provided by the sternum and rib cage in the thoracic spine. Spine 2005;30:1283–6. [29] Fujimori T, et al. Kinematics of the thoracic spine in trunk rotation: in vivo 3-dimensional analysis. Spine 2012;37:E1318–28. [30] Lindahl O. Determination of the sagittal mobility of the lumbar spine. A clinical method. Acta Orthop Scand 1966;37:241–54. [31] Kong MH, et al. Lumbar segmental mobility according to the grade of the disc, the facet joint, the muscle, and the ligament pathology by using kinetic magnetic resonance imaging. Spine 2009;34:2537–44. [32] Kirkaldy-Willis WH, Farfan HF. Instability of the lumbar spine. Clin Orthop Relat Res 1982;110–23.