Spinal fusion with decompressive laminectomy is

Transcription

1 J Neurosurg Spine 18:24 31, 2013 AANS, 2013 Posterior dynamic stabilization in the treatment of degenerative lumbar stenosis: validity of its rationale Clinical article Kee-Yong Ha, M.D., 1 Jun-Yeong Seo, M.D., 2 Soon-Eok Kwon, M.D., 1 Il-Nam Son, M.D., 3 Ki-Won Kim, M.D., 1 and Young-Hoon Kim, M.D. 1 1 Department of Orthopaedic Surgery, College of Medicine, The Catholic University of Korea, Seoul; 2 Department of Orthopaedic Surgery, School of Medicine, Jeju National University, Jeju Self-Governing Province; and 3 Department of Orthopaedic Surgery, College of Medicine, Kwandong University, Gangwon-do, Korea Object. The authors undertook this study to investigate the validity of the rationale for posterior dynamic stabilization using the Device for Intervertebral Assisted Motion (DIAM) in the treatment of degenerative lumbar stenosis. Methods. A cohort of 31 patients who underwent single-level decompression and DIAM placement for degenerative lumbar stenosis were followed up for at least 2 years and data pertaining to their cases were analyzed prospectively. Of these patients, 7 had retrolisthesis. Preoperative and postoperative plain lumbar radiographs obtained in all patients and CT images obtained in 14 patients were analyzed. Posterior disc heights; range of motion (ROM) of proximal, distal, and implant segments; lordotic angles of implant segments; percentage of retrolisthesis; and crosssectional area and heights of intervertebral foramina on CT sagittal images were analyzed. Clinical outcomes were evaluated using visual analog scale scores and Oswestry Disability Index scores. Results. The mean values for posterior disc height before surgery, at 1 week after surgery, and at the final follow-up visits were 6.4 ± 2.0 mm, 9.7 ± 2.8 mm, and 6.8 ± 2.5 mm, respectively. The mean lordotic angles at the implant levels before surgery, at 1 week after surgery, and at the final follow-up visits were 7.1 ± 3.3, 4.1 ± 2.7, and 7.0 ± 3.7, respectively. No statistically significant difference was found between the preoperative values and values from final follow-up visits for posterior disc height and lordotic angles at implant levels (p = 0.17 and p = 0.10, respectively). There was no statistically significant difference between the preoperative and final follow-up visit values for intervertebral foramen cross-sectional area and heights on CT images. The ROMs of proximal and distal segments also showed no significant decrease (p = 0.98 and p = 0.92, respectively). However, the ROMs of implant segments decreased significantly (p = 0.02). The average 31.4-month improvement for all clinical outcome measures was significant (p < 0.001). Conclusions. Based on radiological findings, the DIAM failed to show validity in terms of the rationale of indirect decompression, but it did restrict motion at the instrumented level without significant change in adjacent-segment ROM. The clinical condition of the patients, however, was improved, and improvement was maintained despite progressive loss of posterior disc height after surgery. ( Key Words dynamic stabilization validity DIAM lumbar stenosis indirect decompression disc height Abbreviations used in this paper: AP = anteroposterior; DIAM = Device for Intervertebral Assisted Motion; ISDD = interspinous distraction device; ODI = Oswestry Disability Index; ROM = range of motion; VAS = visual analog scale. Spinal fusion with decompressive laminectomy is the treatment of choice for low-back pain combined with radiating pain and claudication related to degenerative change and instability of the lumbar spine with thecal sac compression. However, fusion has been associated with adjacent-segment degeneration, inconsistent clinical results, and complications. Recently, alternatives to conventional fusion procedures, involving the use of semirigid dynamic implants, have been introduced as minimally invasive surgical procedures. Three categories of devices have been developed based on 3 distinct principles of device implantation. One category is interspinous distraction devices (ISDDs) that are fixed between the interspinous processes, such as the X-STOP Spacer (Medtronic Spine and Biologics), the Coflex device (Paradigm Spine), the DIAM (Device for Intervertebral Assisted Motion) spinal stabilization system (Medtronic Spine and Biologics), and the Wallis implant (Zimmer Spine). Another category is pedicle screw-based systems such as the Dynesys Dynamic Stabilization System (Zimmer Spine), the TOPS (Total Posterior System, Premia Spine), and the SSCS (Segmental Spinal Correction System, Osteotech). 1,5,10 The third category is a Universal Clamp system (Zimmer Spine), which uses a polyester band that is passed under the lamina and anchored 24 J Neurosurg: Spine / Volume 18 / January 2013

2 Posterior dynamic stabilization for lumbar stenosis to a rod by a titanium clamp. 4 The ISDDs do not require the insertion of pedicle screws, thus resulting in less morbidity and shorter operating times. The DIAM is an ISDD that acts as a dynamic stabilizer; it consists of an X- shaped silicone interspinous process bumper enveloped by a polyester fiber sack. 5 It is secured to the adjacent spinous processes by polyester implant tethers and fixed under tension. The DIAM has been promoted as a nonfusion alternative for managing a range of degenerative spinal conditions, including spinal stenosis, acute disc herniation, and degenerative disc disease. 5 By using the DIAM, the disc height can be restored, allowing more room at the foraminal level for nerve roots in the treatment of degenerative lumbar stenosis. The DIAM has also been known to unload the intervertebral disc, restore the posterior tension band, realign facet joints, and increase foraminal height, due to its insertion between the spinous processes (Fig. 1). 11,12,14 However, little is known about the DIAM s long-term effects and induced radiographic changes in the treatment of degenerative lumbar stenosis. Thus, we prospectively analyzed changes in the cross-sectional area and height of the intervertebral foramen on CT images, as well as changes in lordotic angles of implant segments and ROM of the proximal and distal adjacent segments and the implant segments on plain standing lumbar lateral radiographs. Clinical outcomes were evaluated using VAS and ODI scores. Methods The study protocol was approved by the local institutional review board and ethics committee. Informed consent was obtained from all study participants. We prospectively analyzed 34 patients who underwent DIAM placement between April 2005 and April 2008 performed by one spine surgeon. Among the 34 patients, 31 completed the study (9 male, 22 female). Their mean age (± SD) was 62.9 ± 9.0 years (range years). Degenerative lumbar stenosis had been diagnosed in all of the patients, and they all suffered from severe neurogenic claudication and leg, buttock, or groin pain with back pain that was aggravated in the upright position and relieved by sitting or lumbar flexion. In all cases, conservative treatment had failed to control the patients symptoms. The inclusion criteria for DIAM placement in this study were the presence of single-level central stenosis with 1 or more of the following conditions: 1) foraminal stenosis, 2) pseudoprotrusion of the anulus fibrosus on MRI, causing thecal sac compression, and 3) degenerative retrolisthesis (patients with degenerative spondylolisthesis were excluded). Patients with spinal instability or degenerative deformity and patients with osteoporosis (bone mineral density T score less than 2.5) were excluded. All patients included in the study had at least 2 years of follow-up. All patients underwent decompressive laminotomy, selective foraminotomy, and single-level DIAM placement. Operative levels were L2 3 in 1 patient, L3 4 in 4 patients, and L4 5 in 26 patients. The average follow-up period was 31.4 ± 11.4 months (range months). Among the 31 patients, 7 showed retrolisthesis. Patients were evaluated preoperatively, 1 week postoperatively, and 1, 3, 6, 12, and 24 months postoperatively; at each time point, plain J Neurosurg: Spine / Volume 18 / January 2013 Fig. 1. Schematic illustration. The rationale for using a DIAM to reduce retrolisthesis is that it restricts extension motions such as extension angular instability and/or retrolisthesis. Additionally, it is expected that retrolisthesis could be reduced into the anatomical position by distraction of the interspinous space for DIAM placement. standing lumbar radiographs were obtained and ODI and VAS scores were assessed. Radiographic Measurements All radiographic assessments included plain radiography with dynamic flexion and extension standing lateral radiographs. Measurements were completed using commercial software for our hospital s picture archiving and communication system. Changes in posterior disc height, lordotic angles at implant levels, percentages of retrolisthesis, and ROM of the proximal, distal, and implant segments were analyzed. Posterior disc height was measured by determining the length of a perpendicular line from the posterior superior border of the lower vertebra to the posterior inferior border of the upper vertebra. The angle between the lower endplate of an upper vertebra and the upper endplate of a lower vertebra at the insertion level was defined as the lordotic angle at the implant level (Fig. 2 left). Seven patients had spinal stenosis with retrolisthesis. The percentage slip was calculated in patients with retrolisthesis by determining the ratio between the AP diameter of the top of the lower vertebra and the distance that the upper vertebra had slipped posteriorly (Fig. 2 right). We measured the ROM of the implant segment as well as that of the proximal and distal adjacent segments on flexion and extension standing lateral radiographs. Angles between a line parallel to the upper endplate and a line parallel to the lower endplate on plain lumbar flexion and extension radiographs were defined as the angles of flexion and extension, respectively. The difference between the corresponding flexion and extension angles was defined as the ROM (Fig. 3). The crosssectional area and height of the intervertebral foramina were calculated on CT images. Of the 31 patients, 14 underwent CT after surgery at a mean of 14.4 ± 4.8 months (range months) postoperatively. To obtain accurate data, we used a foramen view with CT sagittal reconstruction images that bisected the width of the pedicle and were parallel to the long axis of the pedicle, because conventional CT sagittal reconstruction images could not provide images with the same location of the intervertebral foramina preoperatively and postoperatively. To calculate intervertebral foramen sizes and heights, we used a picture archiving and communication system (M-View, version 5.1, Marotech; Fig. 4). 25

3 K. Y. Ha et al. Fig. 2. Radiographic images showing measurement of the lordotic angle of the implant level, the posterior disc height, and the percentage of retrolisthesis. Left: The angle between the lower endplate of an upper vertebra and the upper endplate of a lower vertebra was defined as the lordotic angle of the implant level; X indicates the posterior disc height. Right: The percentage of retrolisthesis was calculated based on the AP diameter of the top of the lower vertebra (a) and the distance that the upper vertebra has slipped posteriorly (b) as follows: b/a 100. Clinical Assessments Clinical outcomes were quantified using low-back and leg pain VAS and ODI scores before surgery; at 3, 6, 12, and 24 months postoperatively; and at final follow-up. Pain was measured on a VAS scale from 0 to 10, with 0 indicating no pain and 10 indicating maximum pain, and ODI was measured on a scale from 0 to 100, with 0 indicating no disability and 100 indicating maximum disability. The VAS and ODI scores were recorded by the patients. Statistical Analyses Normal distribution of data was confirmed using the Kolmogorov-Smirnov method. Pairwise comparison between groups for back VAS, leg VAS, ODI, ROM, posterior disc height, angle of implantation segments, and retrolisthesis were performed using the paired t-test with the appropriate Bonferroni corrections for multiple comparisons. Between-groups comparisons of foraminal height and cross-sectional area of CT images were performed using the Mann-Whitney U-test. Relationships of posterior disc height and back VAS, leg VAS, and ODI score were analyzed by linear regression. A p value of less than 0.05 was deemed to indicate statistical significance. The SPSS software was used for the database and statistics. after surgery, and 6.8 ± 2.5 mm at the final follow-up. The mean posterior disc height was greater at 1 week after DIAM placement than preoperatively but then gradually decreased; no statistically significant difference was found between preoperative and final follow-up posterior disc heights (p = 0.17) (Fig. 5A), but there was a statistically significant difference between the preoperative disc height and disc height 1 year after surgery (p = 0.04). Lordotic Angles of Implanted Segments. The mean lordotic angles at implant levels preoperatively; at 1 week after surgery; at 1, 3, 6, 12, and 24 months after surgery; and at final follow-up visits were 7.1 ± 3.3, 4.1 ± 2.7, 4.6 ± 3.0, 4.7 ± 3.0, 5.3 ± 3.3, 5.6 ± 2.8, 6.9 ± 3.3, and 7.0 ± 3.7, respectively (Fig. 5B). The mean lordotic angle of implant segments had decreased substantially at 1 week after surgery; it then gradually increased, although it remained below the preoperative level at the final follow-up. The difference between the mean preoperative lordotic angle of the implanted segments and the mean lordotic angle at last follow-up was not significant (p = 0.10), but there was a statistically significant differ- Radiological Outcomes Results Posterior Disc Height. Posterior disc height was assessed with standing neutral lateral radiographs preoperatively; at 1 week postsurgery; at 1, 3, 6, and 12 months postsurgery; at 2 years postsurgery; and at the final follow-up visit. The mean posterior disc height was 6.4 ± 2.0 mm preoperatively, 9.7 ± 2.8 mm 1 week after surgery, 8.1 ± 2.6 mm 1 month after surgery, 7.5 ± 2.4 mm 3 months after surgery, 7.3 ± 2.4 mm 6 months after surgery, 7.1 ± 2.8 mm 12 months after surgery, 6.9 ± 2.3 mm 24 months Fig. 3. Angles between a line parallel to the upper endplate and a line parallel to the lower endplate on plain lumbar flexion and extension radiographs were defined as the angles of flexion and extension, respectively. Differences between flexion and extension angles were defined as ROMs. OP = operative level (level of implant). 26 J Neurosurg: Spine / Volume 18 / January 2013

4 Posterior dynamic stabilization for lumbar stenosis heights of the right intervertebral foramina at the corresponding times of measurement were 18.7 ± 2.9 mm, 20.4 ± 3.2 mm, and 19.2 ± 2.6 mm, respectively, and those of the left intervertebral foramina were 18.0 ± 2.9 mm, 20.0 ± 3.2 mm, and 18.5 ± 2.8 mm, respectively; however, there was no statistically significant difference between the postoperative values and those obtained at last followup for either side (right, p = 0.94; left, p = 0.87) (Fig. 6). Fig. 4. Measurement of the intervertebral foramen. Left: Axial CT image. A thick dotted line was drawn through the center of an intervertebral foramen for purposes of measurement. Right: The crosssectional area of an intervertebral foramen was measured from the foramen view at the plane of the thick dotted line in the axial CT image. ence between the mean preoperative angle and the mean angle at 1 year after surgery (p = 0.02; Fig. 5B). Percentage of Retrolisthesis. Seven patients had combined retrolisthesis. Percentages of slip were measured preoperatively; at 1 week after surgery; at 1, 3, 6, 12, and 24 months after surgery; and at final follow-up visits, and the mean values were 7.5% ± 4.4%, 4.6% ± 3.7%, 7.3% ± 3%, 7.6% ± 2.8%, 8.1% ± 2.6%, 8.4% ± 1.5%, 9.1% ± 1.6%, and 9.6% ± 1.9%, respectively. At 1 week after surgery, the percentage of slip was lower than the preoperative value, but somewhat higher than that at final followup. However, the difference between the preoperative and final follow-up values was not statistically significant (p = 0.13) (Fig. 5C). Range of Motion. The ROMs of the implanted segment and proximal and distal adjacent segments were measured preoperatively, 6 months after surgery, 1 and 2 years postsurgery, and at the final follow-up visit. The respective mean values for the implanted segments were 4.3 ± 2.7, 1.8 ± 1.2, 2.1 ± 1.5, 2.4 ± 1.8, and at the final follow-up was 2.3 ± 2.1 ; the difference between the preoperative and final follow-up values was statistically significant (p = 0.018; Fig. 5D). The respective mean ROM values for the proximal adjacent segments were 3.9 ± 2.5, 3.7 ± 1.8, 4.1 ± 2.1, 3.9 ± 2.4, and 3.9 ± 2.6, and the respective mean ROM values for the distal adjacent segments were 4.6 ± 2.4, 4.1 ± 2.3, 4.2 ± 2.6, 4.6 ± 2.3, and 4.6 ± 2.0. At final follow-up, the ROM seemed to have decreased somewhat, but the mean preoperative and final follow-up values did not differ significantly for either the proximal or distal adjacent segments (p = 0.98 and p = 0.92, respectively; Fig. 5D). Cross-Sectional Area and Height of Intervertebral Foramina. The mean cross-sectional areas calculated from the preoperative CT, 2-week postoperative CT, and follow-up CT (performed at a mean of 14.4 months after surgery), respectively, were ± 23.1 mm 2, ± 21.8 mm 2, and ± 29.7 mm 2 for the right intervertebral foramen and ± 34.4 mm 2, ± 19.0 mm 2, and ± 29.3 mm 2 for the left. There was no statistically significant difference between the mean preoperative values for cross-sectional area and those from the last follow-up visit (right, p = 0.85; left, p = 0.93). The mean J Neurosurg: Spine / Volume 18 / January 2013 Clinical Outcomes The mean VAS scores for back pain obtained preoperatively; at 3, 6, 12, and 24 months postoperatively; and at final follow-up were 6.1 ± 2.1, 3.6 ± 1.9, 3.0 ± 2.1, 3.7 ± 3.1, 4.0 ± 3.5, and 3.7 ± 3.2, respectively. The corresponding mean VAS scores for leg pain were 6.9 ± 2.0, 3.9 ± 2.4, 3.8 ± 2.8, 3.6 ± 3.2, 3.7 ± 3.4, and 3.5 ± 3.3. The corresponding ODI scores were 51.0 ± 17.2, 39.3 ± 20.6, 29.6 ± 15.6, 26.8 ± 15.6, 26.4 ± 17.4, and 26.0 ± The VAS and ODI scores decreased after surgery, and there was a statistically significant difference between preoperative and final follow-up scores for back VAS, leg VAS, and ODI scores (p < 0.001). The relationships between posterior disc height and ODI score as well as between posterior disc height and back and leg VAS score were analyzed by linear regression. There was no correlation between them (p = 0.46 ODI, 0.05 back VAS, and 0.15 leg VAS, respectively). Thus, despite progressive loss of posterior disc height after surgery, clinical outcomes were improved and maintained (Fig. 7). Discussion According to a letter to the editor by Thomas E. White sides Jr., 21 Dr. Fred L. Knowles developed a steel plug to be placed between spinous processes to hold the flexed posture of the lumbar spine before The device often dislodged and needed to be removed. Recently, the use of ISDDs has shed new light on managing lumbar disease such as spinal stenosis, lumbar disc disease, and lumbar segmental instability, and a lot of ISDDs have been designed and introduced to the spine surgeon. 2 The devices are categorized, based on their design, as static or dynamic. 1 Static ISDDs include the X-STOP Spacer (Medtronic Spine and Biologics), the Wallis implant (Zimmer Spine), and ExtenSure (NuVasive), and they are noncompressible spacers. The implant fit can be loosened in flexion and can be tightened in extension. The Coflex device (Paradigm Spine), formerly Intraspinous U, and DIAM belong to the category of dynamic ISDDs, which are compressible between adjacent interspinous processes. The key rationale of the ISDD in the treatment of degenerative lumbar stenosis is that ISDDs have been proposed as a dynamic stabilization alternative to rigid instrumented fusion. These devices can provide indirect decompression by distracting the posterior spinal column. In vivo and in vitro studies have shown that ISDDs can prevent narrowing of the spinal canal and foramina, and they have been developed with the intention of preventing reimpingement of the neural structures at the decompressed level. 1,13,16 Initial biomechanical stud- 27

5 K. Y. Ha et al. Fig. 5. Graphs depicting mean posterior disc height (A), angles of the implanted segment (B), percentage of retrolisthesis (C), and ROM (D) for the implanted and proximal and distal adjacent segments at each time point from before surgery (pre-op) to final follow-up (F/U). PO = postoperatively. ies indicate that the interspinous process spacer reduces intradiscal pressure and posterior annular pressure at the implanted level. 17 Vaga et al. 19 reported the quantification of glycosaminoglycan concentration within instrumented segments with Dynesys (Zimmer Spine) and adjacent levels by means of the delayed Gd-enhanced MRI of cartilage protocol. Dynamic stabilization of the lumbar spine using Dynesys is claimed to be able to stop and partially reverse disc degeneration, especially in seriously degenerated discs. Dynesys differs from ISDDs by design but is considered a dynamic stabilization system, and ISDDs have been known to show similar effects. One fascinating advantage of insertion of an ISDD is the prevention of adjacent-segment degeneration by preservation of spinal motion of the implant-treated segments. 7 Lindsey et al. 8 also reported that ISDDs did not significantly alter the kinematics of mobile segments adjacent to instrumented levels. Another rationale of ISDDs is to provide a primary stabilizing function directly after decompression. 4,15 Ilharreborde et al. 4 reported on the comparison between the Universal Clamp and Wallis systems. The results confirmed that both implants restricted motion at the instrumented level without significant change in the ROM at the adjacent segments. Schulte et al. 15 investigated the influence of spinal decompression alone, as well as decompression in conjunction with the Wallis and Dynesys semirigid systems. The ROM of fresh-frozen human lumbar spine segments with these 2 systems was also investigated. According to their results, implantation of the Wallis and Dynesys devices following decompression led to a restriction of ROM in all motion planes investigated. Flexion-extension was most affected by both implant systems. In our study, changes in lordotic angles and posterior disc heights at the implant segments showed a contrary pattern of changes in postoperative and final follow-up values, which were similar to preoperative values (not statistically significantly different). Additionally, CT showed that there was an increase in intervertebral foramen cross-sectional areas and heights immediately after DIAM implantation but that it subsequently decreased in extent. Thus, no significant difference was evident between preoperative and final follow-up values. It could be concluded from this study that the DIAM increased spinal canal and intervertebral foramen sizes and local kyphosis with increasing posterior disc height at implanted segments immediately after surgery, but that these returned to their preoperative levels during follow-up, starting 12 months after surgery. Additionally, in 14 of 31 patients, we used CT sagittal reconstruction images to calculate changes in preand postoperative cross-sectional areas and heights of intervertebral foramina. The postoperative increases in intervertebral foramen cross-sectional areas and heights were not maintained, however. It was evident that DIAM might have no effect on the maintenance of intervertebral disc heights during the follow-up period. However, Siddiqui et al. 16 reported that the X-STOP interspinous device improved the degree of central and foraminal stenosis in vivo as determined by MRI. Richards et al. 13 also reported that X-STOP prevented narrowing of the spinal 28 J Neurosurg: Spine / Volume 18 / January 2013

6 Posterior dynamic stabilization for lumbar stenosis Fig. 6. Graphs depicting the mean cross-sectional area (left) and height (right) of intervertebral foramina (IVF) on CT images. canal and foramina in extension. The X-STOP device was developed with the same rationale as the DIAM. However, X-STOP is categorized as a static spacer, made of titanium, but DIAM is dynamic spacer as mentioned above. The DIAMs failed to maintain the regained height of the neural foramina at the last follow-up in our study. The main difference between our results and those obtained with X-STOP was that the authors of the X-STOP study observed the neural foramina and spinal canal areas just 6 months after surgery. In our study, we evaluated the neural foramina for a longer period (average 14.4 months). Decompressive laminectomy can sometimes lead to further collapse of the intervertebral disc height, resulting in postlaminectomy instability. The rate of revision surgery for postlaminectomy instability is reported to be 2.7% 17%. 3,22 Our study, however, showed that the loss of local kyphosis and regained disc height after DIAM placement were not aggravated more than the preoperative status during the follow-up period. Thus, DIAM can prevent the further collapse of the intervertebral disc height. We had no patients with recurrent spinal stenosis until the end of the study, despite the loss of disc height and local kyphosis after DIAM placement. Kim et al. 6 also reported that placement of the DIAM, as an interspinous spacer, did not alter disc height or sagittal alignment 12 months postoperatively. Their findings are consistent with ours. In the present study, 7 patients showed preoperative degenerative retrolisthesis. Two types of sagittal instabilities exist extension instability and flexion instability. 9 Furthermore, an increase in retrolisthesis with extension may be demonstrated radiographically. Verhoof et al. 20 reported that an ISDD using X-STOP showed an extremely high failure rate for the treatment of degenerative spondylolisthesis, and suggested that the use of an ISDD may be contraindicated in degenerative spondylolisthesis, because distraction might aggravate sagittal instability in flexion. Thus, our rationale for using a DIAM in the treatment of degenerative retrolisthesis with stenosis was that it restricted extension motions, such as extension angular instability and/or retrolisthesis (Fig. 8). However, in the present study, no statistically significant difference was found between preoperative measurements and those from the last follow-up visit with respect to posterior slippage of the cephalad vertebrae, which shows that DIAMs may not maintain preoperatively realized retrolisthesis reductions (Fig. 9). In the present study, the ROM at the implanted segments decreased after DIAM insertion (p = 0.02). However, the ROM of the proximal and distal adjacent segments was unchanged after DIAM placement, and no significant difference was found between preoperative and final follow-up ROM values. Even though the ROM decreased at the implanted segments, motion was preserved at the proximal and distal adjacent levels. The study also showed that DIAM permitted restricted ROM in flexion and extension. However, a longer follow-up study is required to investigate this issue, because some Fig. 7. Graphs showing clinical outcomes from the preoperative time point to the final follow-up evaluation. There was a statistically significant difference between preoperative and final follow-up values with respect to back pain VAS scores (Back VAS), leg pain VAS scores (Leg VAS) (left), and ODI (right) (p < 0.001). J Neurosurg: Spine / Volume 18 / January

7 K. Y. Ha et al. Fig. 8. Serial plain lateral lumbar spine radiographs obtained in a single patient before surgery (A), and 1 week (B), 1 year (C), and 38 months (D) after DIAM placement. The images show posterior disc heights of 7.52 mm, 9.87 mm, 9.05 mm, and 6.02 mm, respectively. The posterior disc height had increased at 1 week after surgery, but this increase was not maintained at 38 months follow-up. restricted motion at the DIAM placement segment could have aggravated adjacent-segment motion as compensation for the loss of motion. Back and leg pain VAS scores and ODI scores were lower at the final follow-up evaluation. Taylor et al. 18 reported that pain levels, as recorded by physicians, improved in 88.5% of patients; however, 20 of 104 patients who underwent DIAM implantation experienced adverse events, and 13 of those 20 patients required a second lumbar surgery. In our study, back and leg pain VAS scores improved from 6.1 to 3.7 and from 6.9 to 3.5, respectively. Thus, the improvement for all clinical outcomes measures at a mean follow-up of 31.4 months was significant (p < 0.001). We believe that improvement of the clinical outcomes could have resulted from some restriction of painful motion and decompressive laminotomy, although the distracting effect was not maintained over the followup period. Why would DIAM lose its distracting effect? It is a compressible dynamic interspinous process device as opposed to a static device. The cause of progressive subsidence was thought to be the result of disproportionate load shifting to the posterior column caused by the loss of viscoelasticity of the DIAM device. Biomechanical studies using cadavers have limitations in that they address only the immediate stability of the motion segment after the insertion of the DIAM device. 12 Loosening caused by a laxity of the tether on the spinous processes over time may affect stability. Additionally, erosion of the spinous processes may also have caused a loss of distraction. Potential limitations of this study include a relatively small sample size and the lack of a control group. Although the issue of small sample size limits study accuracy in terms of validity of the rationale for DIAM placement, we felt that more than 2 years of radiographic and CT-supported follow-up would be sufficient to evaluate DIAM efficacy in terms of rationale, based on radiological findings. In conclusion, no change in cross-sectional area or in heights of intervertebral foramina on CT sagittal reconstruction images was observed at final follow-up. Furthermore, no change in the heights of posterior intervertebral discs, lordotic angles of implanted segments, or reductions in retrolisthesis was observed at final followup on plain lumbar lateral radiographs. The ROM values for the implanted segments were reduced below the normal motion range after DIAM placement, but no change in proximal or distal adjacent-segment ROM values was observed. Conclusions Based on radiological findings, the DIAM failed to show validity in terms of the rationale of continued indirect decompression, although it was found to restrict motion at the instrumented level without significant change in adjacent-segment ROM. Assessment of clinical outcome, however, showed improvement and maintenance of improvement despite loss of posterior disc height after surgery. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: Ha, Seo. Acquisition of data: Ha, Seo, Kwon. Analysis and interpretation of data: Ha, Seo, Kwon, Son, YH Kim. Drafting the article: Ha, Seo, Kwon, Son, YH Kim. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version Fig. 9. Preoperative and postoperative radiographs from a representative case showing change in retrolisthesis. A: Preoperatively, 12.5% retrolisthesis was shown. B: One week after surgery, retrolisthesis was reduced. C: One year after surgery, 12.3% retrolisthesis was shown. D: At final follow-up, 13.4% retrolisthesis was shown. 30 J Neurosurg: Spine / Volume 18 / January 2013

8 Posterior dynamic stabilization for lumbar stenosis of the manuscript on behalf of all authors: Ha. Statistical analysis: Ha, Kwon. Administrative/technical/material support: Ha. Study supervision: Ha, Seo. References 1. Bono CM, Vaccaro AR: Interspinous process devices in the lumbar spine. J Spinal Disord Tech 20: , Christie SD, Song JK, Fessler RG: Dynamic interspinous process technology. Spine (Phila Pa 1976) 30 (16 Suppl):S73 S78, Hopp E, Tsou PM: Postdecompression lumbar instability. Clin Orthop Relat Res 227: , Ilharreborde B, Shaw MN, Berglund LJ, Zhao KD, Gay RE, An KN: Biomechanical evaluation of posterior lumbar dynamic stabilization: an in vitro comparison between Universal Clamp and Wallis systems. Eur Spine J 20: , Kim DH, Albert TJ: Interspinous process spacers. J Am Acad Orthop Surg 15: , Kim KA, McDonald M, Pik JH, Khoueir P, Wang MY: Dynamic intraspinous spacer technology for posterior stabilization: case-control study on the safety, sagittal angulation, and pain outcome at 1-year follow-up evaluation. Neurosurg Focus 22(1):E7, Korovessis P, Papazisis Z, Koureas G, Lambiris E: Rigid, semirigid versus dynamic instrumentation for degenerative lumbar spinal stenosis: a correlative radiological and clinical analysis of short-term results. Spine (Phila Pa 1976) 29: , Lindsey DP, Swanson KE, Fuchs P, Hsu KY, Zucherman JF, Yerby SA: The effects of an interspinous implant on the kinematics of the instrumented and adjacent levels in the lumbar spine. Spine (Phila Pa 1976) 28: , Moon MS: Preliminary design and experimental studies of a novel soft implant for correcting sagittal plane instability in the lumbar spine. Spine (Phila Pa 1976) 24: , 1999 (Letter) 10. Morishita Y, Ohta H, Naito M, Matsumoto Y, Huang G, Tatsumi M, et al: Kinematic evaluation of the adjacent segments after lumbar instrumented surgery: a comparison between rigid fusion and dynamic non-fusion stabilization. Eur Spine J 20: , Nockels RP: Dynamic stabilization in the surgical management of painful lumbar spinal disorders. Spine (Phila Pa 1976) 30 (16 Suppl):S68 S72, Phillips FM, Voronov LI, Gaitanis IN, Carandang G, Havey RM, Patwardhan AG: Biomechanics of posterior dynamic stabilizing device (DIAM) after facetectomy and discectomy. Spine J 6: , Richards JC, Majumdar S, Lindsey DP, Beaupre GS, Yerby SA: The treatment mechanism of an interspinous process implant for lumbar neurogenic intermittent claudication. Spine (Phila Pa 1976) 30: , Schlegel JD, Smith JA, Schleusener RL: Lumbar motion segment pathology adjacent to thoracolumbar, lumbar, and lumbosacral fusions. Spine (Phila Pa 1976) 21: , Schulte TL, Hurschler C, Haversath M, Liljenqvist U, Bullmann V, Filler TJ, et al: The effect of dynamic, semi-rigid implants on the range of motion of lumbar motion segments after decompression. Eur Spine J 17: , Siddiqui M, Karadimas E, Nicol M, Smith FW, Wardlaw D: Influence of X Stop on neural foramina and spinal canal area in spinal stenosis. Spine (Phila Pa 1976) 31: , Swanson KE, Lindsey DP, Hsu KY, Zucherman JF, Yerby SA: The effects of an interspinous implant on intervertebral disc pressures. Spine (Phila Pa 1976) 28:26 32, Taylor J, Pupin P, Delajoux S, Palmer S: Device for intervertebral assisted motion: technique and initial results. Neurosurg Focus 22(1):E6, Vaga S, Brayda-Bruno M, Perona F, Fornari M, Raimondi MT, Petruzzi M, et al: Molecular MR imaging for the evaluation of the effect of dynamic stabilization on lumbar intervertebral discs. Eur Spine J 18 (Suppl 1):40 48, Verhoof OJ, Bron JL, Wapstra FH, van Royen BJ: High failure rate of the interspinous distraction device (X-Stop) for the treatment of lumbar spinal stenosis caused by degenerative spondylolisthesis. Eur Spine J 17: , Whitesides TE Jr: The effect of an interspinous implant on intervertebral disc pressures. Spine (Phila Pa 1976) 28: , 2003 (Letter) 22. Yaşar B, Simşek S, Er U, Yiğitkanli K, Ekşioğlu E, Altuğ T, et al: Functional and clinical evaluation for the surgical treatment of degenerative stenosis of the lumbar spinal canal. Clinical article. J Neurosurg Spine 11: , 2009 Manuscript submitted April 17, Accepted September 24, Please include this information when citing this paper: published online November 9, 2012; DOI: / SPINE Address correspondence to: Kee-Yong Ha, M.D., Department of Orthopaedic Surgery, Seoul St. Mary s Hospital, College of Medicine, The Catholic University of Korea, 505 Banpo-Dong, Seocho- Gu, Seoul , Korea. kyh@catholic.ac.kr. J Neurosurg: Spine / Volume 18 / January