Radiculopathy Caused by Osteoporotic Vertebral Fractures in the Lumbar Spine

Neurol Med Chir (Tokyo) 51, 484 489, 2011 Radiculopathy Caused by Osteoporotic Vertebral Fractures in the Lumbar Spine Manabu SASAKI, 1 Masanori AOKI, 2 Kazuya NISHIOKA, 3 and Toshiki YOSHIMINE 4 1 Department of Neurosurgery, Iseikai Hospital, Osaka; 2 Department of Neurosurgery and Spine Surgery, Yukioka Hospital, Osaka; 3 Department of Neurological Surgery, Wakayama Medical University, Wakayama; 4 Department of Neurosurgery, Osaka University Graduate School of Medicine, Suita, Osaka Abstract The clinical features of radiculopathy caused by osteoporotic vertebral fractures (OVFs) in the lumbar spine were investigated in 66 patients treated for pain caused by OVFs from January 2006 to December 2009. Ten of the patients complained of persistent radiculopathy. The cause of radiculopathy was initially diagnosed as lumbar canal stenosis (LCS) in seven patients, lumbar foraminal stenosis (LFS) in two, and both in one. One patient with LFS had reduced pain with conservative treatment, and the other nine needed surgical treatments. LCS was treated with posterior decompression, and LFS complicated with LCS at the same spinal level was treated with posterior lumbar interbody fusion (PLIF). Vertebroplasty was performed for one patient with LFS to attain indirect nerve root decompression achieved as a result of vertebral reconstruction and stabilization. Four of the patients treated with posterior decompression for LCS suffered from residual radiculopathy postoperatively, which was caused by LFS at the same level or the level below the treated level. Two patients underwent second operative procedure (PLIF) for recurrent radiculopathy. The Japanese Orthopedic Association and Visual Analogue Scale scores of the pain improved after operations, but the scores of the patients treated without spinal fusion gradually worsened during the follow-up period, whereas the scores of the patients treated with PLIF remained stable at various levels. Seven of the ten patients developed LFS following OVF, suggesting that radiculopathy following OVF involves LFS with high frequency. Key words: osteoporotic lumbar compression fracture, radiculopathy, surgery, lumbar canal stenosis, lumbar foraminal stenosis Introduction The frequency of osteoporotic vertebral fractures (OVFs) will increase with the proportion of aged people in the population. The major symptom of OVF is back pain, but neurological compromises sometimes develop. The most frequently reported neurological involvements are myelopathy and cauda equina syndrome. These symptoms are attributed to neural compression due to retropulsation of the posterior vertebral wall and progression of segmental kyphosis. Kyphotic changes are prompted by wedge vertebral deformities which frequently develop at the thoraco-lumbar junction. 3,5,11) These neurological signs include such severe symptoms as Received December 6, 2010; Accepted March 23, 2011 paraparesis and bladder dysfunction, so challenging surgical treatments are often required in aged patients with fragile spines. 3,5,6,12) OVFs in the lower lumbar spine occasionally induce radiculopathy. 2) In our experience, radiculopathy develops in patients with OVF much more frequently than in patients with myelopathy or cauda equina syndrome. This radicular pain often limits the activities of daily living (ADL). However, radiculopathy in such patients is not well understood. The present study investigated the clinical features of radiculopathy caused by OVFs in the lumbar spine. Materials and Methods A total of 66 patients, 22 males and 44 females aged from 56 to 95 years (median 75.2 years), were admit- 484

Radiculopathy Due to OVFs in the Lumbar Spine 485 ted to our hospital due to severe back pain and/or leg pain caused by OVFs in the lumbar spine from January 2006 to December 2009 (Table 1). All patients underwent routine radiography at first examination, and magnetic resonance (MR) imaging in the early period of their hospital stay. Fresh vertebral fractures were identified by signal changes in the broken vertebra on MR imaging and the loss of vertebral height on follow-up radiography. Radiculopathy was observed in 10 of the 66 patients, two males and eight females aged from 65 to 87 years (mean 76.7 years) (Table 2). Radiculopathy was caused by lumbar canal stenosis (LCS) and/or lumbar foraminal stenosis(lfs).thepatientswereexaminedwithmr imaging and computed tomography (CT) focusing on the spinal canal and the vertebral foramen adjacent to the fractured vertebrae (Fig. 1). The disturbed nerve roots were identified by spinal imaging showing compression of the nerve root responsible for the area of the leg pain and neurological abnormalities. Final confirmation was attained by selective nerve root infiltration. All patients initially Table 1 Summary of the 66 patients with severe pain caused by osteoporotic vertebral fractures in the lumbar spine Level No. of cases Type of deformity Wedge Concave Flat Radiculopathy L1 22 18 2 2 0 L2 17 9 5 3 0 L3 9 2 5 2 1 L4 12 2 7 3 4 L5 6 0 3 3 5 Total 66 31 22 13 10 received conservative treatment for more than a month. Nine of the ten patients needed surgical manipulations for pain uncontrollable with conservative treatments. We performed three types of surgical manipulation as follows. Posterior decompression included partial laminectomy and medial facetectomy, performed for LCS. Bilateral spinal decompression with unilateral laminotomy was performed for bilateral radiculopathy. 9) Posterior lumbar interbody fusion (PLIF) was performed for LFS associated with LCS at the same spinal level. In addition to partial laminectomy for LCS, facetectomy was undertaken for decompression of the intervertebral foramen, followed by PLIF with posterior instrumentation for spinal stabilization. PLIF was performed with carbon cages (Fig. 2A) or autologous iliac bone struts (Fig. 2C). Before pedicle screw insertion, hydroxyapatite (HA) sticks (MedtronicSofamorDanek,Memphis,Tennessee,USA) were implanted near the anterior and posterior walls of the vertebral body for stabilization of the screws (Fig. 2B). Vertebroplasty was performed as reported previously. 7) HA blocks (Medtronic Sofamor Danek) were impacted through the bilateral pedicles. Vertebroplasty was performed in Case 9 to achieve indirect nerve root decompression through reconstruction and stabilization of the fractured vertebral body. 2) Vertebroplasty was performed in Case 8 to stabilize pseudoarthrosis of the fractured vertebrae, which was expected to provide better spinal condition following PLIF procedures (Fig. 2C, D). Surgical outcomes were evaluated with Japanese Orthopaedic Association (JOA) scores and Visual Analogue Scale (VAS) scores of the pain taken at the Table 2 Summary of the 10 patients with radiculopathy following osteoporotic vertebral fracture (OVF) Case No. Age (yrs) Sex Level of OVF Type of deformity Level of LCS Level of LFS Follow up (mos) Operation 1st 2nd Postoperative use of analgesics (Level of new OVFs 1 66 F L4 flat L3-4 - 46 PD - - - 2 65 F L4 flat L3-4 - 15 PD - - - 3 74 F L5 concave L4-5 - 8 PD - - T12 4 81 M L5 flat L4-5 L5-S1 48 PD - + - 5 79 F L5 flat L4-5 L4-5 42 PD - + - 6 76 F L4 concave L4-5 L4-5 26 PD PLIF - L3 7 81 F L5 concave L3-4 L4-5 27 PD PLIF - - 8 82 M L5 flat L4-5 L4-5 30 PLIF - - L1 9 76 F L4 concave - L4-5 14 VP - + L3 10 87 F L3 flat - L3-4 2 - - + - Mean 76.7 25.8 F: female, LCS: lumbar canal stenosis, LFS: lumbar foraminal stenosis, M: male, PD: posterior decompression, PLIF: posterior lumbar interbody fusion, VP: vertebroplasty.

486 M. Sasaki et al. Fig. 1 Case 8. A, B: Preoperative T 2 -weighted magnetic resonance (MR) images showing left foraminal stenosis at the L4 5 level (arrow). C, D: Computed tomography (CT) scans showing left L4 nerve root compression (arrow). E G: T 2 -weighted MR images (E, F) and CT scan (G) showing left lateral recess stenosis at the L4 5 level. preoperative stage, at 1 3 postoperative months, at 6 9 postoperative months, and at the final follow-up oneormoreyearslater.cases6and7receivedsecondary operative procedures (PLIF) for recurrent radiculopathy at the 8th and 9th postoperative months, respectively, and were evaluated again before and after the second operations. The mean follow-up period was 25.8 months (Table 2). Results Fig. 2 Postoperative computed tomography scans of Case 6 (A, B) and Case 8 (C, D) showing posterior lumbar interbody fusion was performed with carbon cages (A) or autologous iliac bone struts (C). Before screw insertion, hydroxyapatite sticks were inserted near the anterior and posterior vertebral walls (B, arrow). Vertebroplasty with hydroxyapatite blocks was performed to stabilize the fractured vertebral body manifesting as pseudoarthrosis (D). Table 1 shows the spinal levels of OVFs, types of the vertebral deformities, and incidence of radiculopathy in the 66 patients with OVFs in the lumbar spine. Wedge vertebral deformities occurred frequently in the thoraco-lumbar junction, whereas concave and flat deformities were frequent in the lower lumbar spine. Severe kyphotic changes were not observed in the lower lumbar spine. None of the patients presented with myelopathy or cauda equina syndrome, but 10 patients suffered from radiculopathy following the OVFs. Radiculopathy developed in OVFs at levels L3 and below (Table 1); one at L3, four at L4, and five at L5. The incidence of radiculopathy was more frequent in patients with OVFsatlowerlumbarlevels. At the initial examination, the cause of the radiculopathy was diagnosed as LCS in seven patients (Cases 1 7), LFS in two (Cases 9 and 10), and both LCS and LFS in one (Case 8). LCS and LFS

Radiculopathy Due to OVFs in the Lumbar Spine 487 developed at the spinal level adjacent to the OVFs. After conservative treatments for more than one month, one patient with LFS (Case 10) showed reduced pain which was controllable with analgesics. She underwent follow up at the outpatient clinic of another hospital. The other nine patients needed surgical treatments. Posterior decompression was performed in seven patients diagnosed with LCS at initial examinations (Cases 1 7). Six of the patients (Cases 1 3 and 5 7) were relieved from pain. However, one patient (Case 4) complained of residual radicular pain. He again received detailed spinal imaging, nerve root infiltration, and radiculography. The pain was caused by LFS one level below the original level, which was overlooked in the preoperative imaging. Three of the six patients (Cases 5 7) developed radiculopathy caused by LFS at the same level as the LCS in the late postoperative period. In these patients, LFS was detected in the preoperative sagittal MR images, but was considered to be asymptomatic according to preoperative examinations; distributions of the pain and neurological abnormalities were concordant to the descending nerve root at that lumbar level, and infiltration of this nerve root obtained temporary pain relief. No remarkable change was observed in the postoperative MR images. On the other hand, CT demonstrated subtle postoperative changes in the spine of these patients. Pre-existing spondylolisthesis adjacent to the OVFs had worsened a little after operations in two patients (Cases 5 and 6) (Fig. 3A, B). Deformity of the fractured vertebra progressed postoperatively in the other patient (Case 7) (Fig. 3C, D). These changes probably enhanced pre-existing LFS conditions, so becoming symptomatic. Two patients (Cases 4 and 5) refused additional surgical treatments and controlled the pain with oral analgesics. The other two (Cases 6 and 7) needed second operations (PLIF) at the 8th and 9th postoperative months, respectively. These patients were free from pain after second operations. PLIF was performed as the initial operation for one patient (Case 8) who presented with bi-radicular involvement caused by LCS and LFS at the same spinal level. Dynamic radiography had demonstrated pseudoarthrosis in the broken vertebra, so vertebroplasty was performed before a PLIF procedure to stabilize the vertebral body (Fig. 2C, D). He was relieved from radiculopathy postoperatively, and returned to almost previous ADL levels. Spinal stabilization was still preserved at the final followup examination. Vertebroplasty was performed for one patient (Case 9) who was in poor general condition, to obtain indirect nerve root decompression achieved by Fig. 3 Computed tomography scans of Case 6 (A, B) and Case 7 (C, D) before and after the posterior decompression showing spondylolisthesis at the L4 5 level (A), which was aggravated after operation (B), and preoperative deformity of the L5 vertebra (C), which progressed postoperatively (D). Table 3 Time course of the Japanese Orthopaedic Association score before and after the operation Without spinal fixation: Case No. Preop. 1 3 Mos 6 9 Mos Final 1 10 23-19 2 9 22 18 14 3 17 25 15-4 6 9-4 5 10 21-16 6 21 25 18 18 7 13 17 12 12 9 16 27 20 17 Average 12.8 21.1 16.6 14.3 With spinal fixation: Case No. Preop. 1 3 Mos 6 9 Mos Final 6 18 25 22 25 7 12 16 22 17 8 9 19 17 18 Average 13.0 20.0 20.3 20.0 Preop.: preoperative, : postoperative.

488 M. Sasaki et al. reconstruction and stabilization of the fractured vertebral body. She became almost free from the pain in the early postoperative period, but began to feel slight recurrent pain during follow up. JOA scores improved in all patients after the operations (Table 3). However, the scores of the patients treated with posterior decompression or vertebroplasty gradually decreased during the followup period. On the other hand, the scores of the patients treated with PLIF were maintained at a certain level at least until the last follow-up examination over one year later. The VAS score of the pain decreased in most patients after the operations, but increased in some patients. This deterioration was chiefly attributed to back pain. The time-course of the VAS scores shows that postoperative improvements in the score were preserved in patients treated with PLIF, whereas early improvements were gradually lost in patients who underwent posterior decompression or vertebroplasty. Continuous use of oral analgesics became unnecessary after operations for three of the patients treated with posterior decompression and all patients treated with PLIF (Table 2). No serious complications, including instrumentation failure, were observed in the present series. One patient (Case 7) suffered infection on the surface layer of the operative wound 5 months after the operation. New OVFs occurred in four patients more than 6 months after surgery. The spinal levels of these OVFswereT12inonepatient(Case3),L1inone (Case 8), and L3 in two (Cases 6 and 9). Discussion The present study indicates that radiculopathy is more inducible following OVF. The pathology of radiculopathy seems quite different to that of myelopathy and cauda equina syndrome. The latter is usually induced by neural compression due to progression of segmental kyphosis. This change is based on wedge vertebral deformity which frequently develops as a result of OVF at the thoraco-lumbar junction. On the other hand, radiculopathy tends to be induced in the lower lumbar spine (Table 2). OVF in this area frequently results in concave or flat vertebral deformities. 5) As the present study shows, the incidence of radiculopathy was more frequent in OVFs at lower lumbar levels (Table 1). LCS and LFS caused by spinal degeneration also frequently occurred in the lower lumbar spine. Radiculopathy following OVF potentially develops as a result of the progression of pre-existing LCS and/or LFS due to vertebral deformities and subsequent instability. The present study shows that radiculopathy following OVF involves LFS with extremely high frequency, compared to radiculopathy caused by lumbar degeneration. Symptomatic LFS was present at the onset in four of the ten patients (Cases 4 and 8 10) (Table 2). The three patients treated with posterior decompression for LCS (Cases 5 7) developed symptomatic LFS at the same spinal level as thelcsinthelatepostoperativeperiod.preoperative MR imaging detected the LFS, but this was considered to be asymptomatic according to preoperative physical examinations. Postoperative CT showed subtle progression of the pre-existing spondylolisthesis or vertebral deformity in these patients (Fig. 3). These postoperative changes were thought to be the reason for the LFS becoming symptomatic. As OVFs disturb the stability of the anterior column, removal of the posterior spinal elements through decompression procedures has the potential to induce postoperative changes. In the present study, seven of the ten patients developed LFS following OVF. Recent advances in medical imaging technology provide improved detection of nerve root compression in the intervertebral foramens, 1,2,4,8) but diagnosis of LFS is still difficult with only routine spinal imaging. If a certain nerve root involvement is suggested by the location of leg pain and neurological findings, that nerve root should be observed with detailed spinal imaging focusing on both the spinal canal and the vertebral foramen. Subsequently, the disturbed nerve roots should be confirmed by selective nerve root infiltration. 10) Conservative treatment is usually ineffective for radiculopathy following OVF. In the present study, only one patient with LFS (Case 10) had pain controlled by selective nerve root infiltration and oral analgesics. The other 9 patients needed surgical interventions. The surgical methods were determined after consideration of the cause of the radiculopathy and general condition. JOA scores and VAS scores of the pain improved in the early postoperative periods (Table 3), but the scores of patients treated with posterior decompression (Cases 1 7) or vertebroplasty (Case 9) gradually worsened in the late postoperative periods, whereas the scores of the patients treated with PLIF (Case 8) were maintained during the postoperative follow-up period. Preservation of these scores was also observed in the two patients treated with PLIF as second operations (Cases 6 and 7). Our observations suggest that PLIF indicates better surgical outcome over a year compared to other surgical methods without spinal fixation. Late deterioration of the scores in the patients treated with posterior decompression was attributed to spinal instability caused by removal of posterior spinal elements. In Case 9 treated with vertebroplas-

Radiculopathy Due to OVFs in the Lumbar Spine 489 ty for LFS, late deterioration was caused by loss of the vertebral height gained by the operation. Recurrent radicular pain developed during the short-term follow-up period at mean 9.1 months after vertebroplasty for symptomatic LFS following OVF in two of seven previously reported patients. 2) The present results suggest that spinal fusion with instrumentation may provide better surgical outcomes, but application is often limited in patients with OVF because of their poor general condition and spinal fragility. In addition, new OVFs developed postoperatively in two of the three patients treated with PLIF, indicating that spinal fixation can potentially influence the incidence of new OVFs. Long-term observation is necessary for patients with OVF treated with spinal fusion. In the present study, some patients achieved reduced residual pain with oral analgesics after operation without spinal fixation. Therefore, surgical methods should be selected after consideration of the general condition following accurate diagnosis of the cause of radiculopathy. We believe that the present study provides advanced information for the diagnosis and treatment of radiculopathy following OVF. References 1) Aota Y, Niwa T, Yoshikawa K, Fujiwara A, Asada T, Saito T: Magnetic resonance imaging and magnetic resonance myelography in the presurgical diagnosis of lumbar foraminal stenosis. Spine (Phila Pa 1976) 32: 896 903, 2007 2) Chung SK, Lee SH, Kim DY, Lee HY: Treatment of lower lumbar radiculopathy caused by osteoporotic compression fracture: the role of vertebroplasty. J Spinal Disord Tech 15: 461 468, 2002 3) Kaneda K, Asano S, Hashimoto T, Satoh S, Fujiya M: The treatment of osteoporotic-posttraumatic vertebral collapse using the Kaneda device and a bioactive ceramic vertebral prosthesis. Spine (Phila Pa 1976) 17(8 Suppl): S295 303, 1992 4) Kunogi J, Hasue M: Diagnosis and operative treatment of intraforaminal and extraforaminal nerve root compression. Spine (Phila Pa 1976) 16: 1312 1320, 1991 5) Mochida J, Toh E, Chiba M, Nishimura K: Treatment of osteoporotic late collapse of a vertebral body of thoracic and lumbar spine. J Spinal Disord 14: 393 398, 2001 6) Muhlbauer M, Pfisterer W, Eyb R, Knosp E: Minimally invasive retroperitoneal approach for lumbar corpectomy and anterior reconstruction. Technical note. J Neurosurg 93: 161 167, 2000 7) Nishioka K, Imae S, Kitayama M, Miki J, Okawa T, Itakura T: Percutaneous vertebroplasty using hydroxyapatite blocks for the treatment of vertebral body fracture. Neurol Med Chir (Tokyo) 49: 501 506, 2009 8) Rothman SL, Glenn WV Jr, Kerber CW: Multiplanar CT in the evaluation of degenerative spondylolisthesis. A review of 150 cases. Comput Radiol 9: 223 232, 1985 9) Sasaki M, Abekura M, Morris S, Akiyama C, Kaise K, Yuguchi T, Mori S, Iwatsuki K, Yoshimine T: Microscopic bilateral decompression through unilateral laminotomy for lumbar canal stenosis in patients undergoing hemodialysis. J Neurosurg Spine 5: 494 499, 2006 10) Sasso RC, Macadaeg K, Nordmann D, Smith M: Selective nerve root injections can predict surgical outcome for lumbar and cervical radiculopathy: comparison to magnetic resonance imaging. JSpinalDisord Tech 18: 471 478, 2005 11) Shen M, Kim Y: Osteoporotic vertebral compression fractures: a review of current surgical management techniques. Am J Orthop 36: 241 248, 2007 12) Suk SI, Kim JH, Lee SM, Chung ER, Lee JH: Anteriorposterior surgery versus posterior closing wedge osteotomy in posttraumatic kyphosis with neurologic compromised osteoporotic fracture. Spine (Phila Pa 1976) 28: 2170 2175, 2003 Figure Legends Address reprint requests to: Manabu Sasaki, MD, PhD, Department of Neurosurgery and Spine Surgery, Iseikai Hospital, 6 2 25 Sugahara, Higashiyodogaka ku, Osaka 533 0022, Japan. e-mail: mana-nsu@umin.net