Heterotopic bone formation within the spinal canal is a known complication of bone morphogenetic protein–2 (BMP-2) and presents a clinical and surgical challenge. Imaging modalities are routinely used for operative planning in this setting. Here, the authors present the case of a 59-year-old woman with cauda equina syndrome following intraoperative BMP-2 administration. Plain film myelographic studies showed a region of severe stenosis that was underappreciated on CT myelography due to a heterotopic bony lesion mimicking the dorsal aspect of a circumferentially patent thecal sac. When evaluating spinal stenosis under these circumstances, it is important to carefully consider plain myelographic images in addition to postmyelography CT images as the latter may underestimate the true degree of stenosis due to the potentially similar radiographic appearances of evolving BMP-2–induced heterotopic bone and intrathecal contrast. Alternatively, comparison of sequentially acquired noncontrast CT scans with CT myelographic images may also assist in distinguishing BMP-2–induced heterotopic bony lesions from the thecal sac. Further studies are needed to elucidate the roles of the available imaging techniques in this setting and to characterize the connection between the radiographic and histological appearances of BMP-2–induced heterotopic bone.
Timothy Chryssikos, Kenneth M. Crandall and Charles A. Sansur
W. Jeff Elias, Charles A. Sansur and Robert C. Frysinger
The authors analyzed deep brain stimulation electrode trajectories on MR images to identify risks of cerebrovascular complications associated with the number of electrode insertions, traversal of a sulcus, and penetration of the ventricle.
Pre- and postoperative MR volumes were fused to determine the proximity of electrodes to a sulcus or ventricle and whether there were cortical, subcortical, or intraventricular complications. Complications were further classified as hemorrhagic or nonhemorrhagic and symptomatic or asymptomatic. The authors examined 258 electrode implantation for deep brain stimulation. There were 4 symptomatic events (1.6% incidence): 3 hemorrhagic and 1 nonhemorrhagic, all within the cortex. Asymptomatic events included cortical hemorrhage in 1 patient, nonhemorrhagic cortical changes in 6, pallidal hemorrhage in 1, thalamic infarction in 1, and intraventricular hemorrhage (IVH) in 5 patients.
Proximity to a sulcus was a significant risk factor for hemorrhagic and nonhemorrhagic cortical complications (p = 0.001). There was a complication rate of 10.1% within the trajectories penetrating or adjacent to a sulcus, and a 0.7% rate with trajectories clearly positioned within the gyrus. Asymptomatic IVH was observed in 5% of ventricular penetrations. A history of hypertension was a risk factor for cortical hemorrhage (p = 0.019), but not for cortical ischemic/edematous events (p = 0.605). The number of electrode penetrations did not differ between patients with and without complications (p = 0.868), and the sequence of electrode insertions was not a risk factor in bilateral surgeries.
Symptomatic cortical complications occur when electrodes traverse close to a sulcus. Asymptomatic IVH occurs infrequently with ventricular penetration. Despite intraoperative efforts to avoid cortical sulci, a higher than expected incidence of electrode proximity to the sulci was identified on careful postoperative trajectory analysis. This finding emphasizes the importance of assiduously planning trajectories and reviewing cases with thorough MR analysis.
Charles A. Sansur, John D. Heiss, Hetty L. DeVroom, Eric Eskioglu, Robert Ennis and Edward H. Oldfield
Object. The aim of this study was to evaluate the pathophysiology underlying headache associated with cough in patients with Chiari I tonsillar abnormality. The authors hypothesized that peak intrathecal pressure during coughing is higher in patients with headache aggravated by cough than in patients without or in healthy volunteers. In addition, the authors evaluated the use of intrathecal pressure during cough as a means of assessing obstruction to the free flow of cerebrospinal fluid (CSF) at the craniocervical junction.
Methods. Twenty-six adult patients with Chiari I malformation and syringomyelia, four adult patients with Chiari I malformation without syringomyelia, and 15 healthy volunteers were prospectively studied. Testing before surgery included the following: 1) clinical evaluation for the presence of headache associated with cough; and 2) evaluation of lumbar subarachnoid pressure at rest, during three to five coughs, while performing the Valsalva maneuver, during jugular compression, and after removal of CSF. Patients underwent suboccipital craniectomy, C-1 laminectomy, and duraplasty. Testing was repeated 6 months after surgery.
Conclusions. Peak intrathecal pressures during cough and at baseline were elevated in patients with headache associated with cough compared with either patients without headache or healthy volunteers. After surgery, intrathecal pressures during cough were significantly lower than preoperative values and headache aggravated by cough was resolved partially or completely. Headache linked to coughing in patients with Chiari I malformation is associated with sudden increased intrathecal pressure caused by obstruction to the free flow of CSF in the subarachnoid space.
Andrei F. Joaquim, Catherine C. Shaffrey, Charles A. Sansur and Christopher I. Shaffrey
The authors report a case of man-in-the-barrel (MIB) syndrome occurring after an extensive revision involving thoracoilium instrumentation and fusion for iatrogenic and degenerative scoliosis, progressive kyphosis, and sagittal imbalance. Isolated brachial diplegia is a rare neurological finding often attributed to cerebral ischemia. It has not been previously reported in patients undergoing complex spine surgery. This 70-year-old woman, who had previously undergone T11–S1 fusion for lumbar stenosis and scoliosis, presented with increased difficulty walking and with back pain. She had junctional kyphosis and L5–S1 pseudarthrosis and required revision fusion extending from T-3 to the ilium. In the early postoperative period, she experienced a 30-minute episode of substantial hypotension. She developed delirium and isolated brachial diplegia, consistent with MIB syndrome. Multiple studies were performed to assess the origin of this brachial diplegia. There was no definitive radiological evidence of any causative lesion. After a few days, her cognitive function returned to normal and she regained the ability to move her arms. After several weeks of rehabilitation, she recovered completely. Man-in-the-barrel syndrome is a rare neurological entity. It can result from various mechanisms but most commonly seems to be related to ischemia and is potentially reversible.
Akil P. Patel, Michael T. Koltz, Charles A. Sansur, Mangla Gulati and D. Kojo Hamilton
Patients requiring neurosurgical intervention are known to be at increased risk for deep vein thrombosis (DVT) and attendant morbidity and mortality. Pulmonary embolism (PE) is the most catastrophic sequela of DVT and is the direct cause of death in 16% of all in-hospital mortalities. Protocols for DVT screening and early detection, as well as treatment paradigms to prevent PE in the acute postoperative period, are needed in neurosurgery. The authors analyzed the effectiveness of weekly lower-extremity venous duplex ultrasonography (LEVDU) in patients requiring surgical intervention for cranial or spinal pathology for detection of DVT and prevention of PE.
Data obtained in 1277 consecutive patients admitted to a major tertiary care center requiring neurosurgical intervention were retrospectively reviewed. All patients underwent admission (within 1 week of neurosurgical intervention) LEVDU as well as weekly LEVDU surveillance if the initial study was normal. Additional LEVDU was ordered in any patient in whom DVT was suspected on daily clinical physical examination or in patients in whom chest CT angiography confirmed a pulmonary embolus. An electronic database was created and statistical analyses performed.
The overall incidence of acute DVT was 2.8% (36 patients). Of these cases of DVT, a statistically significant greater number (86%) were discovered on admission (within 1–7 days after admission) screening LEVDU (p < 0.05), whereas fewer were documented 8–14 days after admission (2.8%) or after 14 days (11.2%) postadmission. Additionally, for acute DVT detection in the present population, there were no underlying statistically significant risk factors regarding baseline physical examination, age, ambulatory status, or type of surgery.
The overall incidence of acute symptomatic PE was 0.3% and the mortality rate was 0%.
Performed within 1 week of admission in patients who will undergo neurosurgical intervention, LEVDU is effective in screening for acute DVT and initiating treatment to prevent PE, thereby decreasing the overall mortality rate. Routine LEVDU beyond this time point may not be needed to detect DVT and prevent PE unless a change in the patient's physical examination status is detected.
Charles A. Sansur, Nicholas M. Caffes, David M. Ibrahimi, Nathan L. Pratt, Evan M. Lewis, Ashley A. Murgatroyd and Bryan W. Cunningham
Optimal strategies for fixation in the osteoporotic lumbar spine remain a clinical issue. Classic transpedicular fixation in the osteoporotic spine is frequently plagued with construct instability, often due to inadequate cortical screw–bone purchase. A cortical bone trajectory maximizes bony purchase and has been reported to provide increased screw pullout strength. The aim of the current investigation was to evaluate the biomechanical efficacy of cortical spinal fixation as a surgical alternative to transpedicular fixation in the osteoporotic lumbar spine under physiological loading.
Eight fresh-frozen human spinopelvic specimens with low mean bone mineral densities (T score less than or equal to –2.5) underwent initial destabilization, consisting of laminectomy and bilateral facetectomies (L2–3 and L4–5), followed by pedicle or cortical reconstructions randomized between levels. The surgical constructs then underwent fatigue testing followed by tensile load to failure pullout testing to quantify screw pullout force.
When stratifying the pullout data with fixation technique and operative vertebral level, cortical screw fixation exhibited a marked increase in mean load at failure in the lower vertebral segments (p = 0.188, 625.6 ± 233.4 N vs 450.7 ± 204.3 N at L-4 and p = 0.219, 640.9 ± 207.4 N vs 519.3 ± 132.1 N at L-5) while transpedicular screw fixation demonstrated higher failure loads in the superior vertebral elements (p = 0.024, 783.0 ± 516.1 N vs 338.4 ± 168.2 N at L-2 and p = 0.220, 723.0 ± 492.9 N vs 469.8 ± 252.0 N at L-3). Although smaller in diameter and length, cortical fixation resulted in failures that were not significantly different from larger pedicle screws (p > 0.05, 449.4 ± 265.3 N and 541.2 ± 135.1 N vs 616.0 ± 384.5 N and 484.0 ± 137.1 N, respectively).
Cortical screw fixation exhibits a marked increase in mean load at failure in the lower vertebral segments and may offer a viable alternative to traditional pedicle screw fixation, particularly for stabilization of lower lumbar vertebral elements with definitive osteoporosis.
Christopher M. Maulucci, Charles A. Sansur, Vaneet Singh, Alexandra Cholewczynski, Snehal S. Shetye, Kirk McGilvray and Christian M. Puttlitz
Nerve root decompression to relieve pain and radiculopathy remains one of the main goals of fusion-promoting procedures in the subaxial cervical spine. The use of allograft facet spacers has been suggested as a potential alternative for performing foraminotomies to increase the space available for the cervical nerve roots while providing segmental stiffening. Therefore, the goal of this cadaveric biomechanical study was to determine the acute changes in kinetics and foraminal area after the insertion of cortical bone facet spacers into the subaxial cervical spine.
Allograft spacers (2 mm in height) were placed bilaterally into cadaveric cervical spine specimens (C2-T1, age of donors 57.5 ± 9.5 years, n = 7) at 1 (C4–5) and 3 (C3–6) levels with and without laminectomies and posterior lateral mass screw fixation. Standard stereophotogrammetry under pure moment loading was used to assess spinal kinetics. In addition, the authors performed 3D principal component analysis of CT scans to determine changes in foraminal cross-sectional area (FCSA) available for the spinal nerve roots.
Generally, the introduction of 2-mm-height facet spacers to the cervical spine produced mild, statistically insignificant reductions in motion with particular exceptions at the levels of implantation. No significant adjacent-level motion effects in any bending plane were observed. The addition of the posterior instrumentation (PI) to the intact spines resulted in statistically significant reductions in motion at all cervical levels and bending planes. The same kinetic results were obtained when PI was added to spines that also had facet spacers at 3 levels and spines that had been destabilized by en bloc laminectomy. The addition of 2-mm facet spacers at C3–4, C4–5, and C5–6 did produce statistically significant increases in FCSA at those levels.
The addition of allograft cervical facet spacers should be considered a potential option to accomplish indirect foraminal decompression as measured in this cadaveric biomechanical study. However, 2-mm spacers without supplemental instrumentation do not provide significantly increased spinal segmental stability.
D. Kojo Hamilton, Justin S. Smith, Charles A. Sansur, Aaron S. Dumont and Christopher I. Shaffrey
The originally described technique of atlantoaxial stabilization using C-1 lateral mass and C-2 pars screws includes a C-2 neurectomy to provide adequate hemostasis and visualization for screw placement, enable adequate joint decortication and arthrodesis, and prevent new-onset postoperative C-2 neuralgia. However, inclusion of a C-2 neurectomy for this procedure remains controversial, likely due in part to a lack of studies that have specifically addressed whether it affects patient outcome. The authors' objective was to assess the surgical and clinical impact of routine C-2 neurectomy performed with C1–2 segmental instrumented arthrodesis in a consecutive series of elderly patients with C1–2 instability.
Forty-four consecutive patients (mean age 71 years) underwent C1–2 instrumented fusion, including C-1 lateral mass screw insertion. Bilateral C-2 neurectomies were performed. Standardized clinical assessments were performed both pre- and postoperatively. Numbness or discomfort in a C-2 distribution was documented at follow-up. Fusion was assessed using the Lenke fusion grade.
Among all 44 patients, mean blood loss was 200 ml (range 100–350 ml) and mean operative time was 129 minutes (range 87–240 minutes). There were no intraoperative complications, and no patients reported new postoperative onset or worsening of C-2 neuralgia postoperatively. Outcomes for the 30 patients with a minimum 13-month follow-up (range 13–72 months) were assessed. At a mean follow-up of 36 months, Nurick grade and pain numeric rating scale scores improved from 3.7 to 1.0 (p < 0.001) and 9.4 to 0.6 (p < 0.001), respectively. The mean postoperative Neck Disability Index score was 7.3%. The fusion rate was 97%, and the patient satisfaction rate was 93%. All 24 patients with preoperative occipital neuralgia reported relief. Seventeen patients noticed C-2 distribution numbness only during examination in the clinic, and 2 patients reported C-2 numbness, but it did not affect their daily function.
In this series of C1–2 instrumented arthrodesis in elderly patients, excellent fusion rates were achieved, and patient satisfaction was not negatively affected by C-2 neurectomy. In the authors' experience, C-2 neurectomy enhanced surgical exposure of the C1–2 joint, thereby facilitating hemostasis, placement of instrumentation, and decortication of the joint space for arthrodesis. Importantly, with C-2 neurectomy in the present series, no cases of new onset postoperative C-2 neuralgia occurred, in contrast to a growing number of reports in the literature documenting new-onset C-2 neuralgia without C-2 neurectomy. On the contrary, 80% of patients in the present series had preoperative occipital neuralgia and in all of these patients this neuralgia was relieved following C1–2 instrumented arthrodesis with C-2 neurectomy.
Randall Schultz Jr., Andrew Steven, Aaron Wessell, Nancy Fischbein, Charles A. Sansur, Dheeraj Gandhi, David Ibrahimi and Prashant Raghavan
Dorsal arachnoid webs (DAWs) and spinal cord herniation (SCH) are uncommon abnormalities affecting the thoracic spinal cord that can result in syringomyelia and significant neurological morbidity if left untreated. Differentiating these 2 entities on the basis of clinical presentation and radiological findings remains challenging but is of vital importance in planning a surgical approach. The authors examined the differences between DAWs and idiopathic SCH on MRI and CT myelography to improve diagnostic confidence prior to surgery.
Review of the picture archiving and communication system (PACS) database between 2005 and 2015 identified 6 patients with DAW and 5 with SCH. Clinical data including demographic information, presenting symptoms and neurological signs, and surgical reports were collected from the electronic medical records. Ten of the 11 patients underwent MRI. CT myelography was performed in 3 patients with DAW and in 1 patient with SCH. Imaging studies were analyzed by 2 board-certified neuroradiologists for the following features: 1) location of the deformity; 2) presence or absence of cord signal abnormality or syringomyelia; 3) visible arachnoid web; 4) presence of a dural defect; 5) nature of dorsal cord indentation (abrupt “scalpel sign” vs “C”-shaped); 6) focal ventral cord kink; 7) presence of the nuclear trail sign (endplate irregularity, sclerosis, and/or disc-space calcification that could suggest a migratory path of a herniated disc); and 8) visualization of a complete plane of CSF ventral to the deformity.
The scalpel sign was positive in all patients with DAW. The dorsal indentation was C-shaped in 5 of 6 patients with SCH. The ventral subarachnoid space was preserved in all patients with DAW and interrupted in cases of SCH. In no patient was a web or a dural defect identified.
DAW and SCH can be reliably distinguished on imaging by scrutinizing the nature of the dorsal indentation and the integrity of the ventral subarachnoid space at the level of the cord deformity.
Harry Mushlin, Daina M. Brooks, Joshua Olexa, Bryan J. Ferrick, Stephen Carbine, Gerald M. Hayward II, Brandon S. Bucklen and Charles A. Sansur
The sacroiliac joint (SIJ) is a known source of low-back pain. Randomized clinical trials support sacroiliac fusion over conservative management for SIJ dysfunction. Clinical studies suggest that SIJ degeneration occurs in the setting of lumbosacral fusions. However, there are few biomechanical studies to provide a good understanding of the effect of lumbosacral fusion on the SIJ. In the present study, researchers performed a biomechanical investigation to discern the effect of pelvic versus SIJ fixation on the SIJ in lumbosacral fusion.
Seven fresh-frozen human cadaveric specimens were used. There was one intact specimen and six operative constructs: 1) posterior pedicle screws and rods from T10 to S1 (PS); 2) PS + bilateral iliac screw fixation (BIS); 3) PS + unilateral iliac screw fixation (UIS); 4) PS + UIS + 3 contralateral unilateral SIJ screws (UIS + 3SIJ); 5) PS + 3 unilateral SIJ screws (3SIJ); and 6) PS + 6 bilateral SIJ screws (6SIJ). A custom-built 6 degrees-of-freedom apparatus was used to simulate three bending modes: flexion-extension (FE), lateral bending (LB), and axial rotation (AR). Range of motion (ROM) was recorded at L5–S1 and the SIJ.
All six operative constructs had significantly reduced ROM at L5–S1 in all three bending modes compared to that of the intact specimen (p < 0.05). In the FE mode, the BIS construct had a significant reduction in L5–S1 ROM as compared to the other five constructs (p < 0.05). SIJ ROM was greatest in the FE mode compared to LB and AR. Although the FE mode did not show any statistically significant differences in SIJ ROM across the constructs, there were appreciable differences. The PS construct had the highest SIJ ROM. The BIS construct reduced bilateral SIJ ROM by 44% in comparison to the PS construct. The BIS and 6SIJ constructs showed reductions in SIJ ROM nearly equal to those of the PS construct. UIS and 3SIJ showed an appreciable reduction in unfused SIJ ROM compared to PS.
This investigation demonstrated the effects of various fusion constructs using pelvic and sacroiliac fixation in lumbosacral fusion. This study adds biomechanical evidence of adjacent segment stress in the SIJ in fusion constructs extending to S1. Unilateral pelvic fixation, or SIJ fusion, led to an appreciable but nonsignificant reduction in the ROM of the unfused contralateral SIJ. Bilateral pelvic fixation showed the greatest significant reduction of movement at L5–S1 and was equivalent to bilateral sacroiliac fusion in reducing SIJ motion.