The Basic Anatomical Study Using Pedicle-lengthening Osteotomy For The Treatment Of The Lumbar Spinal Stenosis

BackgroundLumbar spinal stenosis is a common degenerative spinal disease in elderly individuals, may lead to neurogenic claudication, lower extremity pain, and lower back pain. With the rapid growth of aged population, the incidence and operability of lumbar spinal stenosis have been climbing steadily. This problem directly affects the quality of elderly people’s lives. Lumbar spinal stenosis in old people is mainly caused by degenerative changes of anatomic structures, including bulging of the intervertebral disc, and hypertrophy of the ligamentum flavum and the facet joints. Those changes lead to an anatomic reduction of the lumbar spinal canal volume or intervertebral foramen dimensions. Lumbar spinal stenosis is categorized into three types according to the anatomic sites:central canal stenosis, lateral recess stenosis, and foraminal stenosis. Central canal stenosis may result from posterior bulging of the intervertebral disc and hypertrophy of the osteophytes, or anterior hypertrophy of the ligamentum flavum. Lateral recess stenosis is caused by the dorsal superior facet joint hypertrophy, the posterior hypertrophy of the ligamentum flavum, and the ventral bulging of the intervertebral disc. Foraminal stenosis can be anteroposterior resulting from a combination of disc bulging and hypertrophy of structures anterior to the facet joint capsule, and vertical anatomical reduction resulting from a laterally degenerated intervertebral disc. As spinal stenosis progresses, degenerative changes occurred in central canal, lateral recess and intervertebral foramina, which compromises the volume of the bony and fibrous vertebral canal in one plane or multipile planes. This can induce neurogenic claudication, lower extremity pain, and lower back pain. The clinical symptoms may worsen during extension positioning because of the significantly smaller lumbar canal volume.The therapy of lumbar spinal stenosis includes conservative treatments and surgical treatments. Conservative treatments currently recommended for patients with mild stenosis include physiotherapy, activity modification, home exercise therapy, epidural injections, and medications. Surgical treatments are more effective for lumbar spinal stenosis patients with severe stenosis symptoms. The traditional surgical procedure involves laminectomy with complete decompression of the structures thought to be causing large blood loss, significant postoperative lumbar instability, and degeneration acceleration near the segment, and nerve adhesion, which may lead to long-term morbidity. Minimally invasive spine surgery is conducted by using micro-incision and percutaneous puncture technique.As such, minimally invasive spine surgery techniques are being increasingly used in the treatment of patients, with the clinical advantages of reduced tissue trauma, recovery time and complications.Pedicle-lengthening osteotomy is a new minimally invasive spine surgery for lumbar spinal stenosis. Fluoroscopic guidance and an adjustable bone saw are used to open a passage through the lumbar pedicles to be lengthened, and into the vertebral body. Pedicle osteotomies are performed bilaterally from inside the pedicle passages using a manual bone saw. Length-adjustable pedicle screws are threaded into the pedicle passages and expanded by an internal screw mechanism. This leads to an increase in the area of the spinal canal and neural foramen, and nerve root compression is relieved. The advantage of this surgery is to maintain the integrity of the posterior spinal structures without nerve adhesion. In 2010, Yang first reported the anatomy study on the intervertebral foramen through pedicle-lengthening osteotomy. This study demonstrated a new procedure to relieve foramina compression. In 2012, Kiapour et al. used a normal finite element spine model to simulate pedicle-lengthening osteotomies. The finite element spine demonstrated substantial enlargement of the cross-sectional dimensions of the spinal canal and intervetebral foramina. In 2013, Mlyavykh first reported the clinical application of pedicle-lengthening osteotomies. Treatment of a cohort of 19 patients with lumbar spinal stenosis and favorable clinical results were reported for most patients at the 12-month follow-up time point. Pedicle-lengthening osteotomy is a promising minimally invasive spine surgery for lumbar spinal stenosis.However, the complexity of the anatomic relations of lumbar pedicle-lengthening osteotomy sites and adjacent neural structures, they are not without complications, such as dural matter tear and leakage of cerebrospinal, injury of cauda equina and nerve roots. Therefore, it is necessary to avoid neural injury in pedicle-lengthening osteotomy.Previous studies have shown the effect of pedicle-lengthening osteotomy on the spinal canal and intervertebral foramen. Several limitations of the former studies must be acknowledged. First, those measurements are based on bony spinal canal and neglect the effect of the bulging of the intervertebral disc and hypertrophy of the ligamentum flavum. Second, three types of lumbar spinal stenosis mainly change the nerve root canal, but there are only reports on central canal.Third, pre-operative lumbar canal size may be bigger in normal subjects than those in LSS patients and few measurements on lumbar spinal stenosis. The shape of spinal canal is irregular and each cross-sectional area is in different size. Different lumbar positions have much influence on the cross-sectional area of the spinal canal.Furthermore, the compressions on the spinal canal are also variable due to different types of spinal canal stenosis. This method did include the effect of disc budging and hypertrophy of the ligamentum flavum on spinal canal and thus we establish the spinal canal volume model by three-dimensional reconstruction techniques. Therefore, determination of the canal volume may overcome the limitations of measurements in two-dimensional axial plane and evaluate the morphological compression of the spinal canal more meaningfully.This study combines clinical anatomy and three-dimensional reconstruction techniques to simulate pedicle-lengthening osteotomies. We determine the anatomic relations between the lumbar pedicle osteotomy sites and adjacent neural structures. Our findings may decrease the risk of neurologic complications during pedicle-lengthening osteotomies. We analyze the changes of spinal canal volume and neural foramen dimension of three types of lumbar spinal stenosis after simulated pedicle-lengthening osteotomies, in order to provide parameters for the forwards finite element analysis and technical support for individual treatment.Our research includes the following two parts:Chapter.1 Anatomic study of relations between the lumbar pedicle-lengthening osteotomy sites and the adjacent neural structuresObjectiveThis study was to determine quantitatively the anatomic relationship between the lumbar pedicle-lengthening osteotomy sites and the adjacent nerural structures to provide a safe operative zone for pedicle-lengthening osteotomy, to provide anatomic parameters for the former biomechanical study and clinical applications.Materials and Methods1. MaterialsThirty embalmed cadavers aged 35 to 70 years (female, n=15; male, n=15) were obtained from the Department of Anatomy, Southern Medical University. The height of the cadavers was 155cm to 178cm. The specimens were screened by X-ray to exclude spinal deformity, joints and ligaments injury. The T12-S1 spinal segment was selected. Vernier calipers (Suce Measurement Instruments Co.LTD, accurate to 0.01 mm), Compass, Angulometer (accurate to 1°), Rongeur.2. MethodThe cadavers were placed in the supine position, all abdominal organs were removed, and the lumbar plexus was carefully dissected. The cadavers were placed in the prone position, removing all fascia and muscle tissue from the posterior aspect of the vertebrae at the middle-line of the T12-S1. The erector spinae and transversospinales were longitudinal dissected and all the muscle tissues were removed to show the whole laminae. A rongeur was used to remove spinous processes, laminae, facet joints, and ligament of the T12-S1; therefore, the lumbar pedicles, dura sac and nerve roots were exposed. The pedicle bones were ground to the junction of the vertebral body and pedicle to expose the pedicle-lengthening osteotomy sites. The lumbar nerve roots were untouched during the dissection. All dissections and measurements were performed by one researcher. The measurements included:① the distance from the superior border of the pedicle osteotomy site to the inferior limit of the adjacent nerve root (PSRD); ② the distance from the inferior border of the pedicle osteotomy site to the superior limit of the adjacent nerve root (PIRD);③the distance between the medial wall of the pedicle osteotomy site and the lateral limit of the dural sac or adjacent nerve root (PMRD);④ the distance from the lateral wall of the pedicle osteotomy site to the medial limit of the adjacent nerve root;⑤ the distance between the central axes of bilateral pedicle osteotomy sites(IPD);⑥ the distance from the insertion of the nerve root to the midline of the dural sac;⑦ the nerve root height(NRH);⑧ the angle between the midline of the dural sac and the axis of the nerve root in the frontal plane(REA). All symmetrical structures were measured bilaterally. Data were analyzed using SPSS statistical software (version 13.0, SPSS Inc.). Data were defined as x±s. Bilateral parameters were compared using the paired t-test; differences between males and females were examined using the two-sample t-test. Statistical significance was established at P<0.05.ResultsMean PSRD, PIRD, PMRD, and PLRD at all levels ranged from 4.83 to 8.75 mm, 1.09 to 2.57 mm,0.02 to 1.76 mm, and 0.93 to 11.45 mm in males; and 4.72 to 8.35 mm, 1.02 to 2.35 mm,0.01 to 1.64 mm, and 0.73 to 11.13 mm in females. The measurements did not differ significantly between left and right sides (P>0.05). Females had shorter PSRD, PIRD, PLRD, and IPD but larger REA at all levels compared with males (P<0.05). Mean PMRD did not differ significantly between males and females (P>0.05). The anatomic relations in both males and females were PSRD> PIRD> PMRD.Conclusions1. Our study demonstrated that medial or inferior osteotomies of the lumbar pedicle might carry a greater risk of injury to the neural structures. During the pedicle-lengthening osteotomies, special attention should be paid to the medial inferior and lateral superior border of the lumbar pedicles, especially in women.2. Paying attention to L5 pedicle osteotomies. PSRD was longest at L5, but PMRD, PIRD, and PLRD at L5 were short. L4-L5 was most commonly affected by lumbar spinal stenosis and had higher rates of surgery. Therefore, pedicle-lengthening osteotomies at L5 level carried a greater risk.Chapter.2 Digital anatomic study of the impact of pedicle-lengthening length on spinal canal volume and neural foramen size in lumbar spinal stenosisObjectiveUsing three-dimensional reconstruction techniques to simulate pedicle-lengthening osteotomies and investigating the impact of different pedicle-lengthening osteotomies on spinal canal volume and neural foramen dimensions in three types of spinal stenosis (Central canal stenosis, Lateral recess stenosis, Foraminal stenosis) patients. To determine the quantitative relations between the pedicle-lengthening length and the increased spinal canal volume and neural foramen dimensions, in order to provide parameters for the forwards finite element analysis and technical support for individual treatment.Materials and Methods1. Data collection and experimental deviceData were obtained from 36 lumbar spinal stenosis patients who underwent lumbar CT examination at affiliated Putian hospital of southern medical university from June 2012 to June 2014. Patients with scoliosis, kyphosis, spondylolisthesis and previous lumbar surgery were excluded. Of the 36 lumbar spinal stenosis patients,12 (male 7, female 5) patients were diagnosed with L4-L5 central canal stenosis,12 (male 5, female 7) patients were diagnosed with L4-L5 foraminal stenosis, and 12 patients (male 9, female 3) were diagnosed with L4-L5 lateral recess stenosis. The central canal stenosis patients were aged from 51 to 78 years and with a mean age of 62 years. The lateral recess stenosis patients were aged from 57 to 81 years and with a mean age of 65 years. The foraminal stenosis patients were aged from 53 to 77 years and with a mean age of 63 years.12 normal adults (male 8, female 4, aged from 12-40 years, with a mean age of 36 years) were served as the control group. CT scans of the lumbar spine were obtained using a 64-channel Lightspeed scanning system (GE Healthcare Corporation). Scan parameters:Tube Voltage of 120kV, Tube Current of 264 mA, Slice Thickness of 0.625 mm. The scanned images were in contrast and grey-scale control to obtain the best bony-scale and soft-tissue scale CT images. The imported data were saved as dicom format.IdeaPad Z470 laptop was used in three-dimensional reconstructions. Main configuration:CPU:Intel Core i5-2520M@2.50GHz quad-core server processor; Mainboard:Lenovo KL6; RAM:8 GB (Ramaxel Technology DDR3 1333MHz); Primary Hard Drive:Hitachi HTS547550A9E384 (500 GB/5400 RPM); Primary Video Adapter:Nvidia GeForce GT 520M (1 GB/Lenovo); Monitor:SumSung SEC444C (14 inches).Software:Windows 7 ultimdined 64bit SP1 (Microsoft, Inc., USA), Mimics software v14.01 (Materialise Corp., Leuven, Belgium)> Rapidform XOSTM (INUS Technology Inc., Seoul, Korea).2.3D Modeling and methodsFour groups of CT scanning data were put into Mimics 14.01 software, according to the range of HU values by utilizing "Thresholding", "Region Growing", "Edit Masks", "Morphology Operations" and "Calculate 3D from Mask" to establish the L4-L5 lumbar spine model. Draw and edit the mask of the intervertebral disc based on its anatomic outline by using the "Lasso" function. At last, L4-L5 disc models were created by using function of "Calculate 3D from Mask".Four groups of L4-L5 vertebrae and disc models were saved as Binary STL and put into Rapidform software, according to "Ref. Plane", "Mesh Sketch", and "Spline" to create a mesh of the inner intervertebral foramen. At last, L4-L5 intervertebral foramen models were created by using function of "Extrude". Accordingly, the definition of inner intervertebral foramen at L4-L5 was limited superiorly by the medial margin of the caudal L4 pedicle, inferiorly by the interior margin of the cranial L5 pedicle, anteriorly by the L4 vertebral body, L4-L5 disc and L5 vertebral body, and posteriorly by the anterior margin of the facet joints.For clinical purposes, the spinal canal was considered to constitute three zones, including a central canal between the medial margins of the facet joints and two lateral canals beneath the facet joints and entering the neural foramen. Accordingly, the definition of spinal canal volume at L4-L5 was limited superiorly by the superior rim of the bilateral L4 pedicles, inferiorly by the inferior rim of the bilateral L5 pedicles, anteriorly by the L4 vertebral body, L4-L5 vertebral disc and L5 vertebral body, and posteriorly by the ligamentum flavum, laterally by the inner borders of L4, the interior margin of the intervertebral foramen (disc level) and the inner borders of L5. According to the above standards, the L4-L5 segmental spinal canal volume model was calculated in Mimics14.01 by using Edit Masks, Morphology Operations, Region Growing and Calculate 3D from Mask under soft tissue scale.The pedicle-lengthening osteotomies were simulated in Mimics 14.01 software. The L4-L5 vertebrae model was modified by cutting the bilateral pedicle bones at the junction of the vertebral body and pedicles by using "Cut Orthogonal to Screen" and "Reposition". Four groups of L4-L5 vertebrae models were simulated 1,2,3,4, or 5 mm pedicle-lengthening procedures at L4 and/or L5. Then, four groups of lumbar pedicle-lengthening models saved as Binary STL and put into Rapidform software to establish the pedicle-lengthening intervertebral foramen models. Four groups of spinal canal volume and neural foraminal dimension were measured in the intact state and after the simulated pedicle-lengthening procedures.ResultsIn the initial state, spinal canal volume and neural foraminal dimensions were significantly smaller in the stenosis models compared to the normal models (P<0.05). The spinal canal volume was in this order:Normal group (21590.6±2638.5mm3)> Foraminal stenosis group (15175.7±2785.3 mm3)> Lateral recess stenosis group (13032.0±1583.8 mm3)> Central canal stenosis group (11979.5±3353.1 mm3). The neural foraminal dimension was in this order:Normal group (134.3±18.5 mm2)> Central canal stenosis group (93.3±9.5mm2)> Lateral recess stenosis group(87.7± 9.7 mm2)> Foraminal stenosis group (74.8±8.6 mm2). In all of the lengthened states, the percentage increase in spinal canal volume and neural foraminal dimensions were larger in the lumbar spinal stenosis groups compared to the normal group (P<0.05). After lengthening at L4, the percentage increase in spinal canal volume was Lateral recess stenosis group> Central canal stenosis group> Foraminal stenosis group> Normal group, the percentage increase in neural foraminal dimensions were Foraminal stenosis group> Central canal stenosis group> Lateral recess stenosis group> Nonnal group. After lengthening at L5, the percentage increase in spinal canal volume was Lateral recess stenosis group> Foraminal stenosis group> Central canal stenosis group> Normal group, the percentage increase in neural foraminal dimensions were Central canal stenosis group> Lateral recess stenosis group> Normal group> Foraminal stenosis group. After lengthening at L4-L5, the percentage increase in spinal canal volume was same as lengthening at L5, the percentage increase in neural foraminal dimensions were same as lengthening at L4. After lengthening at L4, the spinal canal volume increased by 4.3%,5.9%,7.2%, and 5.3% per millimeter in the Normal, Central canal stenosis group, Lateral recess stenosis group, and Foraminal stenosis group, respectively, the neural foraminal dimensions increased by 8.7%,12.9%,12.4%, and 14.3% per millimeter in the Normal, Central canal stenosis group, Lateral recess stenosis group, and Foraminal stenosis group, respectively. After lengthening at L5, the spinal canal volume increased by 4.6%, 5.0%,7.3%, and 5.7% per millimeter in the Normal, Central canal stenosis group, Lateral recess stenosis group, and Foraminal stenosis group, respectively, the neural foraminal dimensions increased by 7.2%,10.4%,9.7%, and 6.7%. After lengthening at L4-L5, the spinal canal volume increased by 8.8%,10.0%,13.1%, and 10.8% per millimeter in the Normal, Central canal stenosis group, Lateral recess stenosis group, and Foraminal stenosis group, the neural foraminal dimensions increased 14.0%, 15.5%,14.6%, and 16.9% per millimeter in the Normal, Central canal stenosis group, Lateral recess stenosis group.Conclusions1. Pedicle-lengthening osteotomy produced larger L4-L5 spinal canal volume and neural foraminal dimensions in lumbar spinal stenosis groups than normal group. Lateral recess stenosis patients might benefit from larger increases in spinal canal volume than other stenosis patients. These data indicated that Lateral recess stenosis patients were the most suitable for treatment with pedicle-lengthening osteotomy.2. For enlarging the L4-L5 neural foramina, lengthening L4 pedicles was superior to lengthening L5 pedicles. Lengthening L4 pedicles might be the best option for relieving foramina compression in lumbar spinal stenois patients.