- Email: [email protected]
Safety of Dynamic Magnetic Resonance Imaging of the Cervical Spine in Children Performed without Neurosurgical Supervision
Original Article
Safety of Dynamic Magnetic Resonance Imaging of the Cervical Spine in Children Performed without Neurosurgical Supervision Derek Yecies1, Nathaniel Fogel2, Michael Edwards1, Gerald Grant1, Kristen W. Yeom3, Samuel Cheshier4
OBJECTIVE: The need for neurosurgical supervision as well as the general safety and utility of dynamic magnetic resonance imaging (MRI) of the cervical spine in children remains controversial. We present the largest descriptive cohort study of cervical flexion-extension MRI scans in pediatric patients to help elucidate the safety and utility of this technique.
-
METHODS: We retrospectively reviewed all cervical spine MRI scans performed at Lucile Packard Children’s Hospital at Stanford from 2009 to 2015. We identified 66 dynamic cervical MRI scans performed in 45 children and 2 young adults for further study.
without direct neurosurgical supervision. We describe for the first time the use of flexion-extension MRI to clear neonates with severe congenital cervical spine abnormalities for complex operative positioning and further care in the intensive care unit.
-
RESULTS: General anesthesia was used in 43 scans. The neuroradiology team performed all scans with no direct supervision by the neurosurgery team. There were no adverse events. Dynamic MRI detected significant instability that was not clearly seen on dynamic radiographs (5 patients) and cord compression not seen on static MRI (9 patients). One patient with asymptomatic instability found on flexion-extension radiographs had no cord compression with movement on MRI and was managed conservatively. Two neonates with significant congenital malformations of the cervical spine were cleared for operative positioning for cardiac procedures based on flexion-extension MRI.
-
CONCLUSIONS: Dynamic MRI is a safe tool for evaluating the cervical spine and cervicomedullary junction in various pediatric populations and can be performed safely
-
Key words Achondroplasia - Dynamic MRI - Flexion-extension MRI - Pediatric spine - Skeletal dysplasia - Trisomy 21 -
INTRODUCTION
A
ccurate assessment of stability and dynamic neural element compression in the cervical spine and craniocervical junction of children is often a clinical challenge. Many imaging modalities can be used to assess the cervical spine and craniocervical junction; however, all have drawbacks that limit their utility. Flexion-extension radiographs are difficult to interpret in skeletally immature children. Dynamic computed tomography scanning does not allow for sensitive imaging of the spinal cord and requires significant radiation exposure. Static magnetic resonance imaging (MRI) provides excellent imaging of the neural elements but does not allow for assessment of dynamic compression of the neural elements or instability. In contrast, dynamic MRI allows for imaging of the neural elements in flexion and extension and for assessment of stability in all age groups. However, the use of dynamic MRI of the cervical spine in children is not widespread owing to concerns about the safety and utility of this technique. Several studies have suggested that dynamic MRI of the cervical spine can be performed safely in children. In 1996, Weng and Heynes1 described performing dynamic cervical MRI safely
California; 2Department of Orthopedic Surgery, Stanford University School of Medicine, Redwood City, California; 3Division of Pediatric Radiology, Department of Radiology, Stanford University School of Medicine, Stanford, California; and 4Division of Pediatric Neurosurgery, Department of Neurosurgery, Primary Children’s Medical Center, University of Utah, Salt Lake City, Utah, USA To whom correspondence should be addressed: Derek Yecies, M.D. [E-mail: [email protected]] Kristen W. Yeom and Samuel Cheshier are coesenior authors.
Abbreviations and Acronyms FSE: Fast spin echo MRI: Magnetic resonance imaging TE: Echo time TR: Repetition time
Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.05.210 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
From the 1Division of Pediatric Neurosurgery, Department of Neurosurgery, Stanford University School of Medicine and Lucile Packard Children’s Hospital at Stanford, Stanford,
WORLD NEUROSURGERY -: e1-e6, - 2018
www.WORLDNEUROSURGERY.org
e1
ORIGINAL ARTICLE DEREK YECIES ET AL.
SAFETY OF DYNAMIC MRI OF THE CERVICAL SPINE IN CHILDREN
in 17 children (16 anesthetized). More recently, Mackenzie et al.2 performed 38 dynamic MRI scans in 31 anesthetized children with skeletal dysplasia. Mukherjee et al.3 imaged 29 children with achondroplasia, Kolman et al.4 imaged 23 patients (primarily children) with Morquio syndrome, and Solanki et al.5 imaged 7 children with Morquio syndrome without any complications. Finally, Tubbs et al.6 performed flexionextension MRI in 25 children with Chiari I malformation and reported movement of the cerebellar tonsils in many of the patients and no complications related to imaging. The authors of all 5 studies suggested that the additional information provided by dynamic cervical MRI was useful in their clinical decision making. The utility of dynamic MRI of the cervical spine remains undefined across the broad range of pathologies encountered by pediatric neurosurgeons. Additionally, given the limited safety data published to date, safety concerns continue to limit the use of this technique in many institutions. Anecdotally, safety concerns have led many institutions to require direct neurosurgical supervision of this procedure. We present the largest descriptive cohort study of dynamic cervical spine MRI scans in pediatric patients and provide results on the safety and utility of this imaging technique, which in our institution is performed independently by pediatric neuroradiologists.
respiratory status was closely monitored by the attending pediatric anesthesiologist. Neuromonitoring was not used in any cases in this study.
MATERIALS AND METHODS Institutional review board approval under protocol 28674 was obtained for retrospective chart review and collection of data on patient treatment, demographics, and outcomes. Data were initially accessed using the electronic medical record and then anonymized. All cervical spine MRI scans performed at Lucile Packard Children’s Hospital at Stanford from 2009 to 2015 were retrospectively reviewed. We identified 66 dynamic cervical MRI scans performed in 45 children and 2 young adults for further study. General anesthesia was used in 43 scans. In all dynamic cervical MRI scans, T2 sagittal fast spin echo (FSE) images were obtained in neutral, flexion, and extension positions at either 1.5T or 3T magnets (GE Medical Systems, Milwaukee, Wisconsin, USA). At 3T, the parameters for sagittal T2 FSE were 3-mm slice thickness, 1-mm skip, and repetition time (TR)/echo time (TE) 3670 ms/95 ms. The parameters for axial T2 FSE images were 5-mm slice thickness, 0.5-mm skip, and TR/TE 5500 ms/100 ms. At 1.5T, the parameters for sagittal T2 FSE images were 3-mm slice thickness, 1-mm skip, and TR/ TE 3600 ms/160 ms. The parameters for axial T2 FSE images were 4-mm slice thickness, 1-mm skip, and TR/TE 3000 ms/100 ms. With the exception of 2 intubated neonates, flexion and extension of all patients was performed by the attending neurosurgeon in the clinic or on the inpatient ward and the patients were observed for any concerning signs before ordering dynamic MRI. During imaging, all flexion and extension positions were performed under direct supervision of an attending pediatric neuroradiologist (K.W.Y.). Positioning was achieved by placing several layers of towels under the occiput for flexion and under the cervical spine for extension. For studies that required anesthesia, movement of the head was directly performed by the attending pediatric neuroradiologist at the scanner, and
e2
www.SCIENCEDIRECT.com
RESULTS In 45 children and 2 young adults, 66 dynamic cervical MRI scans (43 under general anesthesia) were performed. One child with known obstructive sleep apnea had an episode of apnea with flexion during MRI. This child had no long-term sequelae, and there were no other adverse events related to dynamic MRI. Specifically, documented neurologic examinations from the clinic visit and/or progress notes preceding and following dynamic cervical MRI were assessed and demonstrated no alterations in neurologic examination. Mean age of subjects was 101 months (95% confidence interval, 80.5e121.8 months; age range, 0e249 months) (Table 1). All scans were performed owing to clinical concern for cervical or craniovertebral junction instability. No scans were performed in the setting of acute trauma. Patients had a wide variety of underlying conditions, including trisomy 21 (n ¼ 9), achondroplasia (n ¼ 9), Chiari I (n ¼ 8), connective tissue disorder (n ¼ 4), skeletal dysplasia (n ¼ 3), and mucopolysaccharidoses (n ¼ 2). In all cases, dynamic MRI of the cervical spine provided information thought to be clinically useful by the treatment team. Instability was definitively ruled out in 34 patients. Five patients had instability that had not been seen on dynamic radiographs. Nine patients had dynamic cord compression not seen on static MRI. Two neonates with significant congenital cervical spine malformations were cleared for operative positioning and further clinical care based on dynamic MRI results. One asymptomatic patient with trisomy 21 was found to have instability without cord compression and subsequently was managed conservatively with serial imaging. Based on imaging results, 11 patients (23.4%) were offered surgery with the surgical approach guided by the results of dynamic MRI. Surgeries offered were posterior C1-2 fusion (n ¼ 2), bony posterior fossa decompressions with C1 laminectomy for patients with achondroplasia (n ¼ 2), multilevel laminectomies without fusion in patients with no instability on dynamic imaging (n ¼ 2), posterior fossa decompressions with duraplasty for Chiari I (n ¼ 2), posterior C2-6 fusion in a patient who had previously had C1-2 fusion (n ¼ 1), anterior/ posterior occiput-C2 fusion (n ¼ 1), and posterior occiput-C3 fusion (n ¼ 1).
Table 1. Age of Patients Age (years)
Number of Patients
<1
7
1e6
15
7e12
11
13e18
12
18e20
2
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.05.210
ORIGINAL ARTICLE DEREK YECIES ET AL.
SAFETY OF DYNAMIC MRI OF THE CERVICAL SPINE IN CHILDREN
Figure 1. Preoperative (A) extension and (B) flexion and postoperative (C) extension and (D) flexion sagittal T2 magnetic resonance imaging of patient P.V.
Case Illustrations Case 1. P.V., a 5-year-old boy with a history of trisomy 21, perimembranous ventricular septal defect, reactive airway, seizure disorder, visual abnormality, sleep apnea, developmental delay,
WORLD NEUROSURGERY -: e1-e6, - 2018
hypoplastic dens, severe developmental disability, delayed gross motor development with wide-based gait, poor fine motor coordination, and poor balance was found to have compression of the spinal cord with flexion on flexion-extension
www.WORLDNEUROSURGERY.org
e3
ORIGINAL ARTICLE DEREK YECIES ET AL.
SAFETY OF DYNAMIC MRI OF THE CERVICAL SPINE IN CHILDREN
Figure 2. (A) Extension and (B) flexion sagittal T2 magnetic resonance imaging of patient C.B.
MRI (Figure 1A and B). The patient underwent a posterior instrumented fusion (Harms technique) with good radiographic results and subsequently had improvement in balance, gait, and fine motor coordination (Figure 1C and D). Case 2. C.B., a female neonate with multiple cranial and limb abnormalities (ultimate diagnosis was Goldenhar syndrome); multiple congenital heart defects including a large muscular ventricular septal defect, patent foramen ovale, hypoplastic transverse arch, dextrocardia, and interrupted aortic arch; and an abnormal cervicothoracic junction with levoscoliosis was transferred to Lucile Packard Children’s Hospital at Stanford from an outside hospital. Given her cardiac diagnoses, she needed early cardiac surgical intervention. Neurosurgery was consulted for clearance of the cervical spine given multiple abnormalities. Flexion-extension MRI revealed no instability or dynamic cord compression (Figure 2) and the patient was cleared for further care without spinal precautions. Case 3. Y.G., a female infant with achondroplasia and central and obstructive sleep apnea, was initially referred to the neurosurgery service with concerns by the referring provider for hydrocephalus and sagittal synostosis. She was thought to have neither condition, but MRI scan of the brain revealed significant foramen magnum stenosis without cord signal change. She underwent
e4
www.SCIENCEDIRECT.com
flexion-extension MRI of the cervical spine at 13 months of age, which demonstrated stenosis, but normal cord signal, no dynamic change in canal diameter, and no direct compression of the cord (Figure 3). The patient did not have neurologic symptoms at any time and has since been followed conservatively without clinical worsening. DISCUSSION This is the largest study of pediatric flexion-extension MRI scans published to date and demonstrates that the technique is both useful and safe. All scans presented in this study were performed without direct neurosurgical supervision, suggesting that dynamic MRI can be performed safely by experienced pediatric neuroradiologists. Flexion-extension MRI was performed in patients with varied conditions, including trisomy 21, achondroplasia, Chiari I, connective tissue disorder, skeletal dysplasia, and mucopolysaccharidoses. We found flexion-extension MRI to be useful for assessment of children for dynamic spinal cord and brainstem compression, definitive assessment of instability in skeletally immature children, and guidance of surgical planning in complex craniocervical junction cases. In this study, 30% (14 of 47) of patients had dynamic cord compression or instability not identified on prior imaging. The remainder had instability definitively ruled out. This study also represents the first reported use of this
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.05.210
ORIGINAL ARTICLE DEREK YECIES ET AL.
SAFETY OF DYNAMIC MRI OF THE CERVICAL SPINE IN CHILDREN
Figure 3. (A) Extension and (B) flexion sagittal T2 magnetic resonance imaging of patient Y.G.
technique to clear neonates with congenital spinal abnormalities for intraoperative positioning and clinical care without spinal precautions. Similar to Mukherjee et al.,3 we found flexion-extension MRI to be particularly helpful in deciding whether surgery was warranted in infants and young children with achondroplasia,7 although we did not use cine MRI to assess cerebrospinal fluid flow. These patients often present for neurosurgical evaluation with a picture of mixed central and obstructive sleep apnea, which are typical to a degree in this patient population. Cervical stenosis and delayed motor development and hypotonia are also typical. MRI evaluation of the cervical spine is often warranted; however, in some cases, cord compression may not be seen in the neutral position and is seen only with dynamic imaging. The presence of T2 signal change in the high cervical cord is of unclear significance, as a study by Brouwer et al.8 reported that 40% of an asymptomatic (primarily adult) cohort of Dutch patients with achondroplasia had T2 signal change in the high cervical cord. We have generally avoided surgery except in cases of clear neurologic symptoms and neurologic compression on dynamic imaging. Accordingly, 7 of 9 patients with achondroplasia evaluated in our study were found to have stenosis without cord
WORLD NEUROSURGERY -: e1-e6, - 2018
compression and were managed conservatively with no development of neurologic deficits. Flexion-extension imaging in anesthetized patients with potential cervical spine instability raises reasonable safety concerns given the potential for spinal cord compression and injury. The available literature in addition to our study has demonstrated no serious adverse events in 143 pediatric flexion-extension MRI scans.1-3,5 None of these scans were done in the setting of acute trauma, and we do not advise performing flexion-extension MRI in that setting given the high false-positive rate of MRI in children in the acute trauma setting.9 Neuromonitoring has not been reported as an adjunct in these studies and was not used in our study, although it could potentially have a role in increasing safety in some circumstances. We believed the safety of the technique was increased by close coordination and fastidious communication with the pediatric anesthesiology and neuroradiology teams regarding the specific clinical concerns for each patient. Additionally, with the exception of 2 intubated neonates, the patients in this study were screened before imaging by performing cervical flexion and extension under neurosurgical supervision while awake to ensure that no neurologic deficits or concerning signs were elicited. There are
www.WORLDNEUROSURGERY.org
e5
ORIGINAL ARTICLE DEREK YECIES ET AL.
SAFETY OF DYNAMIC MRI OF THE CERVICAL SPINE IN CHILDREN
children in whom flexion-extension imaging of any type would neither be warranted nor safe owing to severe instability. In the vast majority of such cases, it is likely that imaging findings from static MRI, computed tomography, or plain radiographs would present clear findings of instability that would obviate the need for dynamic cervical MRI. Despite being the largest study of pediatric flexion-extension cervical spine MRI scans published to date, our study is limited by its small sample size and small numbers of patients with
specific pathologies. It is also limited by being a retrospective and single-institution study.
REFERENCES 1. Weng MS, Haynes RJ. Flexion and extension cervical MRI in a pediatric population. J Pediatr Orthop. 1996;16:359-363. 2. Mackenzie WG, Dhawale AA, Demczko MM, Ditro C, Rogers KJ, Bober MB, et al. Flexionextension cervical spine MRI in children with skeletal dysplasia: is it safe and effective? J Pediatr Orthop. 2013;33:91-98. 3. Mukherjee D, Pressman BD, Rakow DK, Rimoin DL, Danielpour M. Dynamic cervicomedullary cord compression and alterations in cerebrospinal fluid dynamics in children with achondroplasia: review of an 11-year surgical case series. Clinical article. J Neurosurg Pediatr. 2014;14: 238-244. 4. Kolman SE, Ohara SY, Bhatia A, Feygin T, Colo D, Baldwin KD, et al. The clinical utility of flexionextension cervical spine MRI in 22q11.2 deletion syndrome. J Pediatr Orthop. 2017. https://doi.org/ 10.1097/BPO.0000000000000994. [Epub ahead of print].
e6
www.SCIENCEDIRECT.com
CONCLUSIONS Further reports on the safety of this technique will be beneficial to assess for the possibility of rare events of neurologic compromise. However, based on our experience and the current literature, dynamic MRI represents a safe and useful technique.
5. Solanki GA, Martin KW, Theroux MC, Lampe C, White KK, Shediac R, et al. Spinal involvement in mucopolysaccharidosis IVA (Morquio-Brailsford or Morquio A syndrome): presentation, diagnosis and management. J Inherit Metab Dis. 2013;36: 339-355. 6. Tubbs RS, Kirkpatrick CM, Rizk E, Chern JJ, Oskouian RJ, Oakes WJ. Do the cerebellar tonsils move during flexion and extension of the neck in patients with Chiari I malformation ? A radiological study with clinical implications. Childs Nerv Syst. 2016;32:527-530.
9. Brockmeyer DL, Ragel BT, Kestle JRW. The pediatric cervical spine instability study: a pilot study assessing the prognostic value of four imaging modalities in clearing the cervical spine for children with severe traumatic injuries. Childs Nerv Syst. 2012;28:699-705.
Conflict of interest statement: The authors declare that the article content was composed in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
7. Danielpour M, Wilcox WR, Alanay Y, Pressman BD, Rimoin DL. Dynamic cervicomedullary cord compression and alterations in cerebrospinal fluid dynamics in children with achondroplasia. J Neurosurg Pediatrics. 2007;107: 504-507.
This work was presented at the AANS/CNS Section on Pediatric Neurological Surgery Annual Meeting, December 5e8, 2016, in Orlando, Florida.
8. Brouwer PA, Lubout CM, van Dijk JM, Vleggeertlankamp CL. Cervical high-intensity intramedullary lesions in achondroplasia: Aetiology, prevalence and clinical relevance. Eur Radiol. 2012;22: 2264-2272.
Journal homepage: www.WORLDNEUROSURGERY.org
Received 15 March 2018; accepted 28 May 2018 Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.05.210
Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.05.210
Safety of Dynamic Magnetic Resonance Imaging of the Cervical Spine in Children Performed without Neurosurgical Supervision Derek Yecies1, Nathaniel Fogel2, Michael Edwards1, Gerald Grant1, Kristen W. Yeom3, Samuel Cheshier4
OBJECTIVE: The need for neurosurgical supervision as well as the general safety and utility of dynamic magnetic resonance imaging (MRI) of the cervical spine in children remains controversial. We present the largest descriptive cohort study of cervical flexion-extension MRI scans in pediatric patients to help elucidate the safety and utility of this technique.
-
METHODS: We retrospectively reviewed all cervical spine MRI scans performed at Lucile Packard Children’s Hospital at Stanford from 2009 to 2015. We identified 66 dynamic cervical MRI scans performed in 45 children and 2 young adults for further study.
without direct neurosurgical supervision. We describe for the first time the use of flexion-extension MRI to clear neonates with severe congenital cervical spine abnormalities for complex operative positioning and further care in the intensive care unit.
-
RESULTS: General anesthesia was used in 43 scans. The neuroradiology team performed all scans with no direct supervision by the neurosurgery team. There were no adverse events. Dynamic MRI detected significant instability that was not clearly seen on dynamic radiographs (5 patients) and cord compression not seen on static MRI (9 patients). One patient with asymptomatic instability found on flexion-extension radiographs had no cord compression with movement on MRI and was managed conservatively. Two neonates with significant congenital malformations of the cervical spine were cleared for operative positioning for cardiac procedures based on flexion-extension MRI.
-
CONCLUSIONS: Dynamic MRI is a safe tool for evaluating the cervical spine and cervicomedullary junction in various pediatric populations and can be performed safely
-
Key words Achondroplasia - Dynamic MRI - Flexion-extension MRI - Pediatric spine - Skeletal dysplasia - Trisomy 21 -
INTRODUCTION
A
ccurate assessment of stability and dynamic neural element compression in the cervical spine and craniocervical junction of children is often a clinical challenge. Many imaging modalities can be used to assess the cervical spine and craniocervical junction; however, all have drawbacks that limit their utility. Flexion-extension radiographs are difficult to interpret in skeletally immature children. Dynamic computed tomography scanning does not allow for sensitive imaging of the spinal cord and requires significant radiation exposure. Static magnetic resonance imaging (MRI) provides excellent imaging of the neural elements but does not allow for assessment of dynamic compression of the neural elements or instability. In contrast, dynamic MRI allows for imaging of the neural elements in flexion and extension and for assessment of stability in all age groups. However, the use of dynamic MRI of the cervical spine in children is not widespread owing to concerns about the safety and utility of this technique. Several studies have suggested that dynamic MRI of the cervical spine can be performed safely in children. In 1996, Weng and Heynes1 described performing dynamic cervical MRI safely
California; 2Department of Orthopedic Surgery, Stanford University School of Medicine, Redwood City, California; 3Division of Pediatric Radiology, Department of Radiology, Stanford University School of Medicine, Stanford, California; and 4Division of Pediatric Neurosurgery, Department of Neurosurgery, Primary Children’s Medical Center, University of Utah, Salt Lake City, Utah, USA To whom correspondence should be addressed: Derek Yecies, M.D. [E-mail: [email protected]] Kristen W. Yeom and Samuel Cheshier are coesenior authors.
Abbreviations and Acronyms FSE: Fast spin echo MRI: Magnetic resonance imaging TE: Echo time TR: Repetition time
Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.05.210 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
From the 1Division of Pediatric Neurosurgery, Department of Neurosurgery, Stanford University School of Medicine and Lucile Packard Children’s Hospital at Stanford, Stanford,
WORLD NEUROSURGERY -: e1-e6, - 2018
www.WORLDNEUROSURGERY.org
e1
ORIGINAL ARTICLE DEREK YECIES ET AL.
SAFETY OF DYNAMIC MRI OF THE CERVICAL SPINE IN CHILDREN
in 17 children (16 anesthetized). More recently, Mackenzie et al.2 performed 38 dynamic MRI scans in 31 anesthetized children with skeletal dysplasia. Mukherjee et al.3 imaged 29 children with achondroplasia, Kolman et al.4 imaged 23 patients (primarily children) with Morquio syndrome, and Solanki et al.5 imaged 7 children with Morquio syndrome without any complications. Finally, Tubbs et al.6 performed flexionextension MRI in 25 children with Chiari I malformation and reported movement of the cerebellar tonsils in many of the patients and no complications related to imaging. The authors of all 5 studies suggested that the additional information provided by dynamic cervical MRI was useful in their clinical decision making. The utility of dynamic MRI of the cervical spine remains undefined across the broad range of pathologies encountered by pediatric neurosurgeons. Additionally, given the limited safety data published to date, safety concerns continue to limit the use of this technique in many institutions. Anecdotally, safety concerns have led many institutions to require direct neurosurgical supervision of this procedure. We present the largest descriptive cohort study of dynamic cervical spine MRI scans in pediatric patients and provide results on the safety and utility of this imaging technique, which in our institution is performed independently by pediatric neuroradiologists.
respiratory status was closely monitored by the attending pediatric anesthesiologist. Neuromonitoring was not used in any cases in this study.
MATERIALS AND METHODS Institutional review board approval under protocol 28674 was obtained for retrospective chart review and collection of data on patient treatment, demographics, and outcomes. Data were initially accessed using the electronic medical record and then anonymized. All cervical spine MRI scans performed at Lucile Packard Children’s Hospital at Stanford from 2009 to 2015 were retrospectively reviewed. We identified 66 dynamic cervical MRI scans performed in 45 children and 2 young adults for further study. General anesthesia was used in 43 scans. In all dynamic cervical MRI scans, T2 sagittal fast spin echo (FSE) images were obtained in neutral, flexion, and extension positions at either 1.5T or 3T magnets (GE Medical Systems, Milwaukee, Wisconsin, USA). At 3T, the parameters for sagittal T2 FSE were 3-mm slice thickness, 1-mm skip, and repetition time (TR)/echo time (TE) 3670 ms/95 ms. The parameters for axial T2 FSE images were 5-mm slice thickness, 0.5-mm skip, and TR/TE 5500 ms/100 ms. At 1.5T, the parameters for sagittal T2 FSE images were 3-mm slice thickness, 1-mm skip, and TR/ TE 3600 ms/160 ms. The parameters for axial T2 FSE images were 4-mm slice thickness, 1-mm skip, and TR/TE 3000 ms/100 ms. With the exception of 2 intubated neonates, flexion and extension of all patients was performed by the attending neurosurgeon in the clinic or on the inpatient ward and the patients were observed for any concerning signs before ordering dynamic MRI. During imaging, all flexion and extension positions were performed under direct supervision of an attending pediatric neuroradiologist (K.W.Y.). Positioning was achieved by placing several layers of towels under the occiput for flexion and under the cervical spine for extension. For studies that required anesthesia, movement of the head was directly performed by the attending pediatric neuroradiologist at the scanner, and
e2
www.SCIENCEDIRECT.com
RESULTS In 45 children and 2 young adults, 66 dynamic cervical MRI scans (43 under general anesthesia) were performed. One child with known obstructive sleep apnea had an episode of apnea with flexion during MRI. This child had no long-term sequelae, and there were no other adverse events related to dynamic MRI. Specifically, documented neurologic examinations from the clinic visit and/or progress notes preceding and following dynamic cervical MRI were assessed and demonstrated no alterations in neurologic examination. Mean age of subjects was 101 months (95% confidence interval, 80.5e121.8 months; age range, 0e249 months) (Table 1). All scans were performed owing to clinical concern for cervical or craniovertebral junction instability. No scans were performed in the setting of acute trauma. Patients had a wide variety of underlying conditions, including trisomy 21 (n ¼ 9), achondroplasia (n ¼ 9), Chiari I (n ¼ 8), connective tissue disorder (n ¼ 4), skeletal dysplasia (n ¼ 3), and mucopolysaccharidoses (n ¼ 2). In all cases, dynamic MRI of the cervical spine provided information thought to be clinically useful by the treatment team. Instability was definitively ruled out in 34 patients. Five patients had instability that had not been seen on dynamic radiographs. Nine patients had dynamic cord compression not seen on static MRI. Two neonates with significant congenital cervical spine malformations were cleared for operative positioning and further clinical care based on dynamic MRI results. One asymptomatic patient with trisomy 21 was found to have instability without cord compression and subsequently was managed conservatively with serial imaging. Based on imaging results, 11 patients (23.4%) were offered surgery with the surgical approach guided by the results of dynamic MRI. Surgeries offered were posterior C1-2 fusion (n ¼ 2), bony posterior fossa decompressions with C1 laminectomy for patients with achondroplasia (n ¼ 2), multilevel laminectomies without fusion in patients with no instability on dynamic imaging (n ¼ 2), posterior fossa decompressions with duraplasty for Chiari I (n ¼ 2), posterior C2-6 fusion in a patient who had previously had C1-2 fusion (n ¼ 1), anterior/ posterior occiput-C2 fusion (n ¼ 1), and posterior occiput-C3 fusion (n ¼ 1).
Table 1. Age of Patients Age (years)
Number of Patients
<1
7
1e6
15
7e12
11
13e18
12
18e20
2
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.05.210
ORIGINAL ARTICLE DEREK YECIES ET AL.
SAFETY OF DYNAMIC MRI OF THE CERVICAL SPINE IN CHILDREN
Figure 1. Preoperative (A) extension and (B) flexion and postoperative (C) extension and (D) flexion sagittal T2 magnetic resonance imaging of patient P.V.
Case Illustrations Case 1. P.V., a 5-year-old boy with a history of trisomy 21, perimembranous ventricular septal defect, reactive airway, seizure disorder, visual abnormality, sleep apnea, developmental delay,
WORLD NEUROSURGERY -: e1-e6, - 2018
hypoplastic dens, severe developmental disability, delayed gross motor development with wide-based gait, poor fine motor coordination, and poor balance was found to have compression of the spinal cord with flexion on flexion-extension
www.WORLDNEUROSURGERY.org
e3
ORIGINAL ARTICLE DEREK YECIES ET AL.
SAFETY OF DYNAMIC MRI OF THE CERVICAL SPINE IN CHILDREN
Figure 2. (A) Extension and (B) flexion sagittal T2 magnetic resonance imaging of patient C.B.
MRI (Figure 1A and B). The patient underwent a posterior instrumented fusion (Harms technique) with good radiographic results and subsequently had improvement in balance, gait, and fine motor coordination (Figure 1C and D). Case 2. C.B., a female neonate with multiple cranial and limb abnormalities (ultimate diagnosis was Goldenhar syndrome); multiple congenital heart defects including a large muscular ventricular septal defect, patent foramen ovale, hypoplastic transverse arch, dextrocardia, and interrupted aortic arch; and an abnormal cervicothoracic junction with levoscoliosis was transferred to Lucile Packard Children’s Hospital at Stanford from an outside hospital. Given her cardiac diagnoses, she needed early cardiac surgical intervention. Neurosurgery was consulted for clearance of the cervical spine given multiple abnormalities. Flexion-extension MRI revealed no instability or dynamic cord compression (Figure 2) and the patient was cleared for further care without spinal precautions. Case 3. Y.G., a female infant with achondroplasia and central and obstructive sleep apnea, was initially referred to the neurosurgery service with concerns by the referring provider for hydrocephalus and sagittal synostosis. She was thought to have neither condition, but MRI scan of the brain revealed significant foramen magnum stenosis without cord signal change. She underwent
e4
www.SCIENCEDIRECT.com
flexion-extension MRI of the cervical spine at 13 months of age, which demonstrated stenosis, but normal cord signal, no dynamic change in canal diameter, and no direct compression of the cord (Figure 3). The patient did not have neurologic symptoms at any time and has since been followed conservatively without clinical worsening. DISCUSSION This is the largest study of pediatric flexion-extension MRI scans published to date and demonstrates that the technique is both useful and safe. All scans presented in this study were performed without direct neurosurgical supervision, suggesting that dynamic MRI can be performed safely by experienced pediatric neuroradiologists. Flexion-extension MRI was performed in patients with varied conditions, including trisomy 21, achondroplasia, Chiari I, connective tissue disorder, skeletal dysplasia, and mucopolysaccharidoses. We found flexion-extension MRI to be useful for assessment of children for dynamic spinal cord and brainstem compression, definitive assessment of instability in skeletally immature children, and guidance of surgical planning in complex craniocervical junction cases. In this study, 30% (14 of 47) of patients had dynamic cord compression or instability not identified on prior imaging. The remainder had instability definitively ruled out. This study also represents the first reported use of this
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.05.210
ORIGINAL ARTICLE DEREK YECIES ET AL.
SAFETY OF DYNAMIC MRI OF THE CERVICAL SPINE IN CHILDREN
Figure 3. (A) Extension and (B) flexion sagittal T2 magnetic resonance imaging of patient Y.G.
technique to clear neonates with congenital spinal abnormalities for intraoperative positioning and clinical care without spinal precautions. Similar to Mukherjee et al.,3 we found flexion-extension MRI to be particularly helpful in deciding whether surgery was warranted in infants and young children with achondroplasia,7 although we did not use cine MRI to assess cerebrospinal fluid flow. These patients often present for neurosurgical evaluation with a picture of mixed central and obstructive sleep apnea, which are typical to a degree in this patient population. Cervical stenosis and delayed motor development and hypotonia are also typical. MRI evaluation of the cervical spine is often warranted; however, in some cases, cord compression may not be seen in the neutral position and is seen only with dynamic imaging. The presence of T2 signal change in the high cervical cord is of unclear significance, as a study by Brouwer et al.8 reported that 40% of an asymptomatic (primarily adult) cohort of Dutch patients with achondroplasia had T2 signal change in the high cervical cord. We have generally avoided surgery except in cases of clear neurologic symptoms and neurologic compression on dynamic imaging. Accordingly, 7 of 9 patients with achondroplasia evaluated in our study were found to have stenosis without cord
WORLD NEUROSURGERY -: e1-e6, - 2018
compression and were managed conservatively with no development of neurologic deficits. Flexion-extension imaging in anesthetized patients with potential cervical spine instability raises reasonable safety concerns given the potential for spinal cord compression and injury. The available literature in addition to our study has demonstrated no serious adverse events in 143 pediatric flexion-extension MRI scans.1-3,5 None of these scans were done in the setting of acute trauma, and we do not advise performing flexion-extension MRI in that setting given the high false-positive rate of MRI in children in the acute trauma setting.9 Neuromonitoring has not been reported as an adjunct in these studies and was not used in our study, although it could potentially have a role in increasing safety in some circumstances. We believed the safety of the technique was increased by close coordination and fastidious communication with the pediatric anesthesiology and neuroradiology teams regarding the specific clinical concerns for each patient. Additionally, with the exception of 2 intubated neonates, the patients in this study were screened before imaging by performing cervical flexion and extension under neurosurgical supervision while awake to ensure that no neurologic deficits or concerning signs were elicited. There are
www.WORLDNEUROSURGERY.org
e5
ORIGINAL ARTICLE DEREK YECIES ET AL.
SAFETY OF DYNAMIC MRI OF THE CERVICAL SPINE IN CHILDREN
children in whom flexion-extension imaging of any type would neither be warranted nor safe owing to severe instability. In the vast majority of such cases, it is likely that imaging findings from static MRI, computed tomography, or plain radiographs would present clear findings of instability that would obviate the need for dynamic cervical MRI. Despite being the largest study of pediatric flexion-extension cervical spine MRI scans published to date, our study is limited by its small sample size and small numbers of patients with
specific pathologies. It is also limited by being a retrospective and single-institution study.
REFERENCES 1. Weng MS, Haynes RJ. Flexion and extension cervical MRI in a pediatric population. J Pediatr Orthop. 1996;16:359-363. 2. Mackenzie WG, Dhawale AA, Demczko MM, Ditro C, Rogers KJ, Bober MB, et al. Flexionextension cervical spine MRI in children with skeletal dysplasia: is it safe and effective? J Pediatr Orthop. 2013;33:91-98. 3. Mukherjee D, Pressman BD, Rakow DK, Rimoin DL, Danielpour M. Dynamic cervicomedullary cord compression and alterations in cerebrospinal fluid dynamics in children with achondroplasia: review of an 11-year surgical case series. Clinical article. J Neurosurg Pediatr. 2014;14: 238-244. 4. Kolman SE, Ohara SY, Bhatia A, Feygin T, Colo D, Baldwin KD, et al. The clinical utility of flexionextension cervical spine MRI in 22q11.2 deletion syndrome. J Pediatr Orthop. 2017. https://doi.org/ 10.1097/BPO.0000000000000994. [Epub ahead of print].
e6
www.SCIENCEDIRECT.com
CONCLUSIONS Further reports on the safety of this technique will be beneficial to assess for the possibility of rare events of neurologic compromise. However, based on our experience and the current literature, dynamic MRI represents a safe and useful technique.
5. Solanki GA, Martin KW, Theroux MC, Lampe C, White KK, Shediac R, et al. Spinal involvement in mucopolysaccharidosis IVA (Morquio-Brailsford or Morquio A syndrome): presentation, diagnosis and management. J Inherit Metab Dis. 2013;36: 339-355. 6. Tubbs RS, Kirkpatrick CM, Rizk E, Chern JJ, Oskouian RJ, Oakes WJ. Do the cerebellar tonsils move during flexion and extension of the neck in patients with Chiari I malformation ? A radiological study with clinical implications. Childs Nerv Syst. 2016;32:527-530.
9. Brockmeyer DL, Ragel BT, Kestle JRW. The pediatric cervical spine instability study: a pilot study assessing the prognostic value of four imaging modalities in clearing the cervical spine for children with severe traumatic injuries. Childs Nerv Syst. 2012;28:699-705.
Conflict of interest statement: The authors declare that the article content was composed in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
7. Danielpour M, Wilcox WR, Alanay Y, Pressman BD, Rimoin DL. Dynamic cervicomedullary cord compression and alterations in cerebrospinal fluid dynamics in children with achondroplasia. J Neurosurg Pediatrics. 2007;107: 504-507.
This work was presented at the AANS/CNS Section on Pediatric Neurological Surgery Annual Meeting, December 5e8, 2016, in Orlando, Florida.
8. Brouwer PA, Lubout CM, van Dijk JM, Vleggeertlankamp CL. Cervical high-intensity intramedullary lesions in achondroplasia: Aetiology, prevalence and clinical relevance. Eur Radiol. 2012;22: 2264-2272.
Journal homepage: www.WORLDNEUROSURGERY.org
Received 15 March 2018; accepted 28 May 2018 Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.05.210
Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.05.210