W. Bradley Jacobs, Richard J. Bransford, Carlo Bellabarba and Jens R. Chapman
Charcot spinal arthropathy (CSA) is an uncommon disorder that occurs in the setting of conditions with decreased protective sensation of the vertebral column, resulting in vertebral joint degeneration, pain, and deformity. Historically, CSA treatment has been fraught with high failure rates. Over time, the authors' institution has trended toward a CSA treatment paradigm of intralesional debridement, circumferential fusion, and four-rod lumbopelvic fixation. As such, the overall objectives of this study were to define the specific clinical characteristics of this rare condition and to determine whether the authors' treatment paradigm has decreased the incidence of revision due to hardware failure/presumed pseudarthrosis or the development of a new CSA over the course of the study and in comparison with historical controls.
The authors performed a retrospective review of the clinical and radiographic records for all patients with CSA treated by the Spine Service at the University of Washington between 1997 and 2009.
Twenty-three patients with CSA were identified. The mean age at presentation was 43.1 years, and the mean latency between spinal cord injury and CSA diagnosis was 19.6 years. The mean follow-up was 33.1 months. Pain and progressive deformity were the major presenting symptoms. Concomitant infection was identified in 17% of patients. Patients with CSA were noted to have long initial fusion constructs spanning an average of 8.4 vertebral levels. Charcot spinal arthropathy did not occur above the level of neurological injury. The vast majority of CSA cases occurred caudally along the spinal axis, with 65% occurring within 1 level of the caudal end of the index fusion construct and 35% occurring even farther distally. Revision due to hardware failure or the development of a new CSA level occurred in 35% of patients. Rates of treatment failure requiring revision significantly decreased over the course of the study, with revision occurring in 6 (66%) of 9 patients who underwent surgery before 2002, in comparison with only 2 (14%) of 14 treated between 2002 and 2009. During a mean follow-up period of 34 months, no treatment revision occurred in the subgroup of 9 patients who underwent four-rod lumbopelvic fixation.
This study represents the largest reported modern surgical series of CSA patients. While revision rates were initially high and comparable to previous reports, the authors' multimodal treatment paradigm, which includes the use of bone morphogenetic protein and four-rod lumbopelvic fixation, dramatically reduced the incidence of treatment failure requiring revision over the course of the study period and represents a significant improvement in the treatment of CSA.
Doniel Drazin, Jens R. Chapman, Andrew Dailey and John Street
Gregory C. Wiggins, Sohail Mirza, Carlo Bellabarba, G. Alex West, Jens R. Chapman and Christopher I. Shaffrey
Anterior decompression and stabilization for thoracic spinal tumors often involves a thoracotomy and can be associated with surgical approach–related complications. An alternative to thoracotomy is surgery via a costotransversectomy exposure.
To delineate the risks of surgery, the authors reviewed their prospective database for patients who had undergone surgery via either of these approaches for thoracic or thoracolumbar tumors. The complications were recorded and graded based on severity and risk of impact on patient outcome.
Between September 1995 and April 2001, the authors performed 29 costotransversectomies (Group 1) and 18 thoracolumbar or combined (Group 2) approaches as initial operations for thoracic neoplasms. The age, sex, pre-operative motor score, and preoperative Frankel grade did not significantly differ between the groups. In the costotransversectomy group there were greater numbers of metastases, upper thoracic procedures, and affected vertebral levels; additionally, the comorbidity rate based on Charlson score, was higher. The mean Frankel grades at discharge were not significantly different whereas the discharge motor and last follow-up motor scores were better in Group 2. There were 11 Group 1 and seven Group 2 patients who suffered at least one complication. The number or patients with complications, the mean number of complications, and severity of complications did not differ between the groups.
Compared with anterior or combined approaches, the incidence and severity of perioperative complications in the surgical treatment of thoracic and thoracolumbar spinal tumors is similar in patients who undergo costotransversectomy. Costotransversectomy may be the preferred operation in patients with significant medical comorbidity or tumors involving more than one thoracic vertebra.
Richard Bransford, Fangyi Zhang, Carlo Bellabarba, Mark Konodi and Jens R. Chapman
Symptomatic thoracic disc herniations (TDHs) are relatively uncommon and are typically treated with an anterior approach. Various posterior surgical approaches have been developed to treat TDH, but the gold standard remains transthoracic decompression. Certain patients have comorbidities and herniation aspects that are not optimally treated with an anterior approach. A transfacet pedicle-sparing approach was first described in 1995, but outcomes and complications have not been well described. The objective of this study was to assess outcomes and complications in a consecutive series of patients with TDH undergoing posterior transfacet decompression and discectomy with posterior instrumentation and fusion.
Eighteen consecutive patients undergoing operative management of TDH were identified from a tertiary care referral database. All patients underwent a transfacet pedicle-sparing decompression and segmental instrumentation with interbody fusion. Outcomes and complications were retrospectively assessed in this patient series. Clinical records were scrutinized to assess levels and types of disc herniation; blood loss; pre- and postoperative motor scores, Nurick grades, and visual analog pain scale scores; and complications such as wrong-level surgery, infection, seroma, and neurological changes. Pre- and postoperative imaging studies were reviewed to assess levels and types of herniation, alignment, and accuracy of instrumentation.
Of the 18 patients, 9 had TDHs at multiple levels. The patients presented with symptoms including myelopathy, axial back pain, urinary symptoms, and radiculopathy and radiological evidence of 29 compressive TDHs ranging from T1–2 to T12–L1. Discs were classified as central (10) or paracentral (19). All discs were successfully removed with no incidence of wrong-level surgery or CSF leak. The mean estimated blood loss was 870 ml with no dural tears. Nurick grades improved on average from 2.5 to 1.9. All patients reported improvement in symptoms compared with preoperative status. The mean visual analog scale score improved from 59 to 21. Sixteen of the 18 patients spent an average of 4.2 days in the hospital; the 2 other patients spent 58 and 69 days. The average duration of follow-up was 12.2 months in 14 patients; 4 patients were lost to follow-up. Twelve patients had no complications. Five patients developed postoperative wound infections or seromas requiring additional operative debridement. One patient had a misplaced screw and suboptimally positioned interbody graft requiring revision. One transient neurological deterioration (American Spinal Injury Association [ASIA] D to ASIA B) occurred postoperatively associated with an inferior segment fracture 20 days after surgery. This necessitated extending the fusion caudally; the patient subsequently experienced a full return to better-than-baseline neurological status.
A modified transfacetal pedicle-sparing approach combined with short segmental fusion offers a safe means of achieving concurrent decompression and segmental stabilization and is an option for certain subtypes of TDH. Although 6 patients required additional surgery for postoperative complications, all patients experienced improvement relative to their preoperative status.
Jens R. Chapman, Paul A. Anderson, Christopher Pepin, Sean Toomey, David W. Newell and M. Sean Grady
✓ Fractures, tumors, and other causes of instability at the cervicothoracic junction pose diagnostic and treatment challenges. The authors report on 23 patients with instability of the cervicothoracic region, which was treated with posterior plate fixation and fusion between the lower cervical and upper thoracic spine. During operation AO reconstruction plates with 8- or 12-mm hole spacing were affixed to the spine using screws in the cervical lateral masses and the thoracic pedicles. Postoperative immobilization consisted of the patient's wearing a simple external brace for 2 months. The following parameters were analyzed during the pre- and postoperative treatment period: neurological status, spine anatomy and reconstruction, and complications. Follow up consisted of clinical and radiographic examinations (mean duration of follow up, 15.4 months; range, 6–41 months).
No neurovascular or pulmonary complications arose from surgery. All patients achieved a solid arthrodesis based on flexion-extension radiographs. There was no significant change in angulation during the postoperative period, but one patient had an increase in translation that was not clinically significant. There were no hardware complications that required reoperation. One patient requested hardware removal in hopes of reducing postoperative pain in the cervicothoracic region. One postoperative wound infection required debridement but not hardware removal. The authors conclude that posterior plate fixation is a satisfactory method of treatment of cervicothoracic instability.
Richard J. Bransford, Jens R. Chapman, Andrea C. Skelly and Ellen M. VanAlstyne
The purpose of this paper was to systematically review and critically appraise the evidence for whether there are differences in outcomes or recovery after thoracic spinal cord injuries (SCIs) based on the spinal level, the timing of intervention, or cause of SCI.
Systematic searches were conducted using PubMed/MEDLINE through January 5, 2012. From 486 articles identified, 10 included data on the population of interest. Included studies were assigned a level of evidence (LOE) rating based on study quality, and an overall strength of evidence was assessed. To estimate the effect of injury level on patient outcomes, the relative risk and risk difference were calculated when data were available.
From 486 citations identified, 3 registry studies and 7 retrospective cohort studies met the inclusion criteria. All were rated as being of poor quality (LOE III). Limited literature exists on the epidemiology of traumatic and nontraumatic SCI. Few studies evaluated outcomes based on SCI level within the thoracic spine. Pulmonary complications and thromboembolic events were less common in persons with lower thoracic SCI (T7–12) than in those with higher thoracic SCI (T1–6) in 2 large studies, but no differences were found in functional outcomes in 4 smaller studies. Patients undergoing earlier surgery (< 72 hours) may have fewer ventilator, ICU, and hospital days than those undergoing later surgery. One small study of SCI during repair of aortic aneurysm compared with traumatic SCI reported similar outcomes for both groups. There are substantial deficiencies in the scientific literature on thoracic SCI in regard to assessment, outcomes ratings, and effectiveness of therapy.
The overall strength of evidence for all outcomes reported is low. Definitive conclusions should not be drawn regarding the prognosis for outcome and recovery after thoracic SCI. From a physiological standpoint, additional methodologically rigorous studies that take into consideration various levels of injury in more anatomically and physiologically relevant form are needed. Use of validated, comprehensive outcomes tools are important to improve our understanding of the impact of thoracic SCI and aid in examining factors in recovery from thoracic SCI.
Gerald A. Grant, Sohail K. Mirza, Jens R. Chapman, H. Richard Winn, David W. Newell, Dolors T. Jones and M. Sean Grady
Object. The authors retrospectively reviewed 121 patients with traumatic cervical spine injuries to determine the risk of neurological deterioration following early closed reduction.
Methods. After excluding minor fractures and injuries without subluxation, the medical records and imaging studies (computerized tomography and magnetic resonance [MR] images) of 82 patients with bilateral and unilateral locked facet dislocations, burst fractures, extension injuries, or miscellaneous cervical fractures with subluxation were reviewed. Disc injury was defined on MR imaging as the presence of herniation or disruption: a herniation was described as deforming the thecal sac or nerve roots, and a disruption was defined as a disc with high T2-weighted signal characteristics in a widened disc space. Fifty-eight percent of patients presented with complete or incomplete spinal cord injuries. Thirteen percent of patients presented with a cervical radiculopathy, 22% were intact, and 9% had only transient neurological deficits in the field.
Early, rapid closed reduction, using serial plain radiographs or fluoroscopy and Gardner—Wells craniocervical traction, was achieved in 97.6% of patients. In two patients (2.4%) closed reduction failed and they underwent emergency open surgical reduction. The average time to achieve closed reduction was 2.1 ± 0.24 hours (standard error of the mean).
The incidence of disc herniation and disruption in the 80 patients who underwent postreduction MR imaging was 22% and 24%, respectively. However, the presence of disc herniation or disruption did not affect the degree of neurological recovery, as measured by American Spinal Injury Association motor score and the Frankel scale following early closed reduction. Only one (1.3%) of 80 patients deteriorated, but that occurred more than 6 hours following closed reduction.
Conclusions. Although disc herniation and disruption can occur following all types of traumatic cervical fracture subluxations, the incidence of neurological deterioration following closed reduction in these patients is rare. The authors recommend early closed reduction in patients presenting with significant motor deficits without prior MR imaging.
Charles Kuntz IV, Sohail K. Mirza, Abel D. Jarell, Jens R. Chapman, Christopher I. Shaffrey and David W. Newell
The optimum treatment of Type II odontoid fractures in the geriatric population remains controversial. Coexisting medical conditions encountered in the elderly patient often increase operative risk and make cervical immobilization difficult to tolerate. Previous studies have shown increased morbidity and mortality and decreased fusion rates for Type II odontoid fractures treated with cervical orthoses in the geriatric population, whereas low morbidity and mortality rates with operative management have recently been documented. To investigate the role of surgical and nonsurgical treatment, a retrospective analysis was performed of patients with Type II odontoid fractures who were at least 65 years old and were consecutively admitted to a single medical center from 1994 to 1998. Twenty patients met inclusion criteria. In 12 patients nonsurgical management with a cervical orthosis was attempted. The nonsurgical management failed early in six patients, with one associated death. Eleven patients were treated surgically with either anterior odontoid screw fixation or posterior C1–2 transarticular screw fixation and modified Gallie fusion. Postoperatively one patient required revision of the C1–2 transarticular screws, and there was one death. In conclusion Type II odontoid fractures in this elderly population were associated with early 10% morbidity and 20% mortality rates. Nonsurgical management of Type II odontoid fractures failed early in six (50%) of 12 patients, whereas surgical treatment failed early in one of 11 (9%) patients. Both the nonsurgical and surgical treatments resulted in approximately 10% morbidity and 10% mortality rates.
Wyatt L. Ramey, Alexander von Glinski, Andrew Jack, Ronen Blecher, Rod J. Oskouian and Jens R. Chapman
The surgical treatment of osteomyelitis and discitis of the spine often represents a challenging clinical entity for a multitude of reasons, including progression of infection despite debridement, development of spinal deformity and instability, bony destruction, and seeding of hardware. Despite advancement in spinal hardware and implantation techniques, these aforementioned challenges not uncommonly result in treatment failure, especially in instances of heavy disease burden with enough bony endplate destruction as to not allow support of a modern titanium cage implant. While antibiotic-infused polymethylmethacrylate (aPMMA) has been used in orthopedic surgery in joints of the extremities, its use has not been extensively described in the spine literature. Herein, the authors describe for the first time a series of patients treated with a novel surgical technique for the treatment of spinal osteomyelitis and discitis using aPMMA strut grafts with posterior segmental fusion.
Over the course of 3 years, all patients with spinal osteomyelitis and discitis at a single institution were identified and included in the retrospective cohort if they were surgically treated with spinal fusion and implantation of an aPMMA strut graft at the nidus of infection. Basic demographics, surgical techniques, levels treated, complications, and return to the operating room for removal of the aPMMA strut graft and placement of a traditional cage were examined. The surgical technique consisted of performing a discectomy and/or corpectomy at the level of osteomyelitis and discitis followed by placement of aPMMA impregnated with vancomycin and/or tobramycin into the cavity. Depending on the patient’s condition during follow-up and other deciding clinical and radiographic factors, the patient may return to the operating room nonurgently for removal of the PMMA spacer and implantation of a permanent cage with allograft to ultimately promote fusion.
Fifteen patients were identified who were treated with an aPMMA strut graft for spinal osteomyelitis and discitis. Of these, 9 patients returned to the operating room for aPMMA strut graft removal and insertion of a cage with allograft at an average of 19 weeks following the index procedure. The most common infections were methicillin-sensitive Staphylococcus aureus (n = 6) and methicillin-resistant S. aureus (n = 5). There were 13 lumbosacral infections and 1 each of cervical and thoracic infection. Eleven patients were cured of their infection, while 2 had recurrence of their infection; 2 patients were lost to follow-up. Three patients required unplanned return trips to the operating room, two of which were for wound complications, with the third being for recurrent infection.
In cases of severe infection with considerable bony destruction, insertion of an aPMMA strut graft is a novel technique that should be considered in order to provide strong anterior-column support while directly delivering antibiotics to the infection bed. While the active infection is being treated medically, this structural aPMMA support bridges the time it takes for the patient to be converted from a catabolic to an anabolic state, when it is ultimately safe to perform a definitive, curative fusion surgery.