Eur Spine J DOI 10.1007/s00586-014-3332-9
ORIGINAL ARTICLE
Dural tear and resultant cerebrospinal fluid leaks after cervical spinal trauma Soo Eon Lee • Chun Kee Chung • Tae-Ahn Jahng Chi Heon Kim
•
Received: 22 September 2013 / Revised: 23 April 2014 / Accepted: 24 April 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Purpose Traumatic cervical spinal cord injuries (SCIs) frequently develop dural tears and resultant cerebrospinal fluid (CSF) leaks. They are not usually identified with advanced imaging, and there are no reports on managing CSF leaks after cervical trauma. Hence, the authors evaluated the incidence of CSF leaks after cervical SCIs and described how to predict and manage CSF leaks. Methods An observational retrospective study was done confirming intraoperative CSF leaks among 53 patients with anterior cervical surgery after cervical spine trauma between 2004 and 2011. Results Seven patients (13.2 %) had dural tears and resultant CSF leaks intraoperatively (M:F ratio of 6:1; mean age, 44.7 years). An initial poor American Spinal Injury Association (ASIA) scale was significantly associated with CSF leaks (p = 0.009). From magnetic resonance imaging (MRI), disruption of the ligamentum flavum was correlated with CSF leaks (p = 0.02). Intraoperative application of fibrin glue on the operated site, postoperative management through the early removal of the wound drain within the first 24 h and early rehabilitation were performed in patients with CSF leaks without perioperative insertion of a lumbar drain. During the follow-up period, none of the patients developed CSF-leak-related complications. Conclusion The incidence of CSF leaks after traumatic cervical SCI is relatively higher than that of degenerative cervical spinal surgery. An initial poor neurological status
S. E. Lee C. K. Chung (&) T.-A. Jahng C. H. Kim Department of Neurosurgery, Seoul National University Hospital, Seoul National University College of Medicine, 28 Yeongun-Dong, Jongno-Gu, Seoul 110-744, Korea e-mail: [email protected]
and disruption of the ligamentum flavum on the MRI in patients were predictable factors of dural tears and CSF leaks. Keywords Spinal cord injury Cerebrospinal leaks Dural tears Cervical spine Introduction Cervical spinal trauma is increasing according to extensive literature on the descriptive epidemiology of traumatic spinal cord injury (SCI) [1–5]. Although the optimal timing for surgery is controversial, surgical techniques and approaches for traumatic cervical SCIs have been simplified by the development of the subaxial injury classification (SLIC) scoring system [4, 6]. Three categories including injury morphology, disco-ligamentous complex, and neurologic status are important in identifying the severity of the injury and making decisions about the surgery. As a diagnostic tool, magnetic resonance imaging (MRI) is considered superior to other tools for examining discoligamentous complex injury of patients with neurological deficits and is emerging as the preferred choice to evaluate damage to the spine [7, 8]. However, traumatic dural tears are not usually shown by MRIs, but identified from the resultant CSF leaks in the operative field during surgery. Although various techniques have been described to manage CSF-leak-related complications after surgery for degenerative spinal diseases, there are no reports describing the management of CSF leaks after cervical SCIs [9, 10]. Therefore, the authors evaluated the incidence of intraoperative CSF leaks after cervical SCIs, described how to predict of dural tears and the resultant CSF leaks and suggested simple and effective management for CSF leaks during the perioperative period.
123
Eur Spine J
Materials and methods One hundred thirty-four consecutive patients who underwent surgery for acute cervical spine trauma between 2004 and 2011 were identified. Of these, patients who had an anterior or anterior and posterior approach to the cervical spine and whose operative level was between C3 and C7 were further evaluated. Among them, patients with a CSF leak during the intraoperative period reported in the operative record were included. Known iatrogenic dural tears were not included. The injury mechanism was divided into high and low energy [11]. The preoperative neurologic status was stratified into American Spinal Injury Association (ASIA) scale and described neurologic level. Injury morphology was used to classify injuries in the subaxial spine as follows: compression, burst, distraction, and rotation/translation on cervical spine computed tomography (CT) according to the SLIC system [6]. Disco-ligamentous complex injury was evaluated on the cervical spine MRI.
Table 1 Demographics of the patients CSF-leak group (n = 7)
Non-CSF-leak group (n = 46)
p value
Sex (male:female)
6:1
39:7
1.00
Age (years)
44.7 ± 19.2
49.8 ± 18.0
Injury mechanism
0.49 0.66
High energy injury
6
33
Low energy injury
1
13
Initial ASIA scale
0.009
A B
6 1
13 5
C
0
3
D
0
25
Neurologic level
0.29
C2
0
4
C3
0
8
C4
2
17
C5
3
15
C6
2
3
ASIA American Spinal Injury Association
Results The anterior or anterior and posterior operative approach was used in 53 patients after cervical spinal trauma. Seven patients (13.2 %) were identified as having a CSF leak during the intraoperative period from the operative record (CSF-leak group, Table 1). This group consisted of 6 men and 1 woman, and the mean age at operation was 44.7 ± 19.2 years (range 17–67 years). The non-CSF-leak group had 46 patients (39 men and 7 women) with a mean age of 49.8 ± 18.0 years (range, 19–84 years). The injury mechanism, initial American Spinal Injury Association (ASIA) scale and neurological level are presented in Table 1. Age, sex, injury mechanism and neurologic level were not different, but the initial ASIA scale was statistically significantly different between the two groups (p = 0.009). In addition, the mean follow-up period was 9.7 ± 5.7 months (range 1–18) in the CSF-leak group and 22.81 ± 25.37 months (range 1–104) in the non-CSF-leak group. Preoperative imaging Injury classification on the cervical spine CT and MRI findings are presented in Table 2. All seven patients in the CSF-leak group showed a disrupted disco-ligamentous complex and had an abnormal finding on the MRI as follows: disruption of the ALL, PLL, ligamentum flavum and interspinous ligament, and high signal intensity on the intervertebral disc and spinal cord. In contrast, intermediate changes consisting of isolated interspinous
123
widening or a signal change in the disco-ligamentous complex was only seen in 5 patients, and a disrupted disco-ligamentous complex was observed in 41 patients in the non-CSF-leak group. Injury morphology and discoligamentous complex injury were not different between the two groups (p = 0.40 and p = 0.58, respectively), and the disruption of the LF had statistical significance (p \ 0.05) (Fig. 1). Intraoperative findings and perioperative management With the standard anterior cervical approach, the intraoperative findings of the disco-ligamentous structures and vertebral body are presented in Table 2. In the CSFleak group, seven patients had PLL injury, and a dural tear was simultaneously observed in these patients after a discectomy or corpectomy. Because the margin of the dural tear was rough and complex, primary dural repair could not be attempted. After removal of the ruptured disc material, further resection of the PLL was not performed, and fibrin glue (Green plastÒ, Green Cross, Seoul, Korea) was applied at the dural defect site in the CSF-leak group. During closure, a submuscular drain, a Jackson-Pratt (JP), was inserted, and the wound was closed layer by layer. A postoperative lumbar CSF drain was not inserted in all the patients. After the operation, the patients were kept in a head elevated position and allowed to do activities they could perform, and rehabilitation started as soon as possible in the intensive care unit.
Eur Spine J Table 2 Injury morphology at preoperative images and intraoperative findings CSF-leak group (n = 7)
Non-CSF-leak group (n = 46)
Injury morphology on CT
p value
0.78
Compression
1
8
Burst
1
8
Distraction
1
8
Translation/rotation
5
21
Disco-ligamentous complex injury
1.00
Intermediate
0
5
Disrupted
7
41
6
36
Abnormality on MRI Anterior longitudinal ligament (ALL)
0.65
Intervertebral disc
6
41
0.79
Posterior longitudinal ligament (PLL)
7
33
0.10
Vertebral body
2
16
0.74
Spinal cord
6
31
0.33
Ligamentum flavum
7
26
0.02
Interspinous ligament
6
35
0.57
Intraoperative finding ALL injury
5
20
0.24
Intervertebral disc injury
7
37
0.46
PLL injury
7
9
0.00
Vertebral body injury
3
15
0.67
In the CSF-leak group, the JP drain was removed within the first postoperative day, and the drainage amount was 70.6 cc ± 51.7 (range 31–160). In the non-CSF-leak group, the JP drain was removed within 1.24 days ± 0.52 (range 1–3), and the drainage amount was 30.0 cc ± 16.61 (range 10–69, p \ 0.05).
Fig. 1 Preoperative MRI of patients with traumatic cervical spinal cord injury at C6-7. Sagittal T2-weighted image shows disruption of the ALL and intervertebral disc (A).The disruption of the PLL is shown and (B) the disruption of the ligamentum flavum (C) is also shown. This patient was intraoperatively identified as having a dural tear and resultant CSF leak
Discussion Incidence of CSF leaks after traumatic cervical SCIs
Hospital course The following perioperative complications developed: pneumonia in 9 patients, urinary tract infection (UTI) in 11 patients, bacteremia in 1 patient, and wound oozing in 1 patient. In the CSF-leak group, the following complications developed: pneumonia in 4 patients, UTI in 1 patient and bacteremia in 1 patient. Moreover, complications frequently developed in this group (p \ 0.05). However, none of the patients developed complications related to CSF leaks such as fluid collection, wound dehiscence, meningitis, and pseudomeningocele. One patient in the CSF-leak group who developed bacteremia and sepsis died 1 month after the operation.
SCIs are frequently accompanied by dural tears and resultant CSF leaks [12, 13]. Traumatic dural tears are commonly associated with thoracic-lumbar spinal trauma in the range of 18–36 % [12, 13]. However, traumatic dural tears are not usually predicted with advanced imaging tools such as MRI, but are identified in the operative field during surgery either by decompression or CSF leakage around the injured segments [13]. Because incomplete cervical spinal cord injury is the most common clinical feature and accounts for up to 70 % of all traumatic SCIs, it is important to predict dural tears and manage CSF leaks after cervical spinal trauma [14, 15]. Although various techniques have been described to manage CSF-leak-related complications after surgery for degenerative cervical spinal
123
Eur Spine J
disease [9, 10, 16], there are no reports on managing CSF leaks in cervical spine trauma. The incidence of CSF leaks after anterior decompression surgery ranges from 0.5 to 3 % in degenerative cervical spinal disease [10]. For ossification of the posterior longitudinal ligament (OPLL), the incidence is much higher, ranging from 4.3 to 32 %, because the most common cause of CSF leaks is injury to the dura, while resecting the PLL [9, 10]. In addition, risk factors for intraoperative CSF leaks were revision anterior surgery and the male gender [9, 10]. However, in trauma cases, dural tears mostly developed at the moment of injury. In the present study, seven patients (13.2 %) had a dural tear and the resultant CSF leak, and the incidence after cervical spine trauma was higher than that of degenerative cervical spinal disease and as frequent as OPLL. Although age, sex, injury mechanism, and injury morphology were not associated with dural tears, an initial poor ASIA scale was correlated with CSF leaks. In addition, disruption of the ligamentum flavum on the initial MRI was also a predictive factor of CSF leaks. Therefore, the authors suggest the possibility of intraoperative CSF leaks in patients who present with a poor neurologic status and simultaneously show a disruption of the ligamentum flavum on a preoperative MRI after cervical spinal trauma. Perioperative management of CSF leaks after cervical trauma Our strategy for the management of CSF leaks in spinal surgery for degenerative spinal disease and spinal tumors is as follows: primary dural repair and no wound drain insertion or early removal of the drain and early mobilization. In our trauma cases, PLL injury and dural tears simultaneously developed at the moment of injury, and the resultant CSF leaks already existed before the operation. The PLL was believed to provide stability to the cervical spine, and the removal of PLL could lead to a decrease in cervical stability after surgery [17]. However, the removal of the PLL was helpful to get more decompression in the anterior approach for cervical spondylotic myelopathy (CSM), although it was more technically demanding [18]. In the operative procedure, further resection of the PLL was not meaningful because the purpose of the operation was to provide stability to the unstable spine in trauma cases. Therefore, if CSF leak was observed during the discectomy or the PLL was already injured, we did not try any further resection of the PLL to prevent further dural defects and maintain spinal stability. In addition, primary dural repair was not accomplished, and duraplasty was also not achieved due to a wide and rough defect of the dura. In this situation, the authors managed the CSF leaks by applying fibrin glue over the operated site. Although a
123
wound drain was inserted during closure, the drain was removed if CSF appeared in the drain tube or bag, and the tube was clamped and removed regardless of the drainage amount within the first 24 h [19]. In addition, a lumbar drain was not inserted after the operation. Instead, the patients were not allowed bed rest and encouraged early ambulation using a wheel chair. If CSF accumulated under the surgical wound, it was observed unless the symptoms were related to an airway obstruction. As time went by, CSF accumulation gradually disappeared [19]. Wound drainage and lumbar drainage were prescribed to manage CSF leaks during the postoperative period [10, 20]. However, significant controversy exists over the use of wound drains due to the fear of developing duracutaneous fistulas [20]. The use of lumbar drains is also associated with infections such as meningitis, over-drainage of the CSF and pneumocephalus with brainstem compression [21, 22]. In addition, because the standard position of patients with lumbar drainage is bed rest to prevent over-drainage of the CSF, bed rest may increase the risk of postoperative morbidity due to the development of deep venous thrombosis and pulmonary embolism after prolonged immobilization [23]. The present study involved a relatively small number of patients and a short follow-up period. The authors had to include patients who underwent an anterior cervical surgical approach because the intraoperative findings were used as a reference tool for comparison to the preoperative MRI findings. Moreover, patients with a poor neurologic status had difficulty in continuing their follow-up after transfer to another hospital. Although this study is valuable in identifying the predictive factors of CSF leaks after cervical spinal trauma and suggests a simple management for CSF leaks, the study has limitations in concluding an ultimate management strategy for CSF leaks in cervical spinal surgery.
Conclusion The incidence of CSF leaks after cervical spinal trauma is relatively higher than that of degenerative cervical spinal disease. ASIA A or B as an initial neurological status and disruption of the ligamentum flavum on the initial MRI in patients were predictive factors of CSF leaks. The authors suggest a simple and effective management of CSF leaks by early removal of wound drains and not using lumbar drains. Acknowledgments The authors would like to express their appreciation for the statistical analysis done by the Medical Research Collaborating Center at Seoul National University Hospital and the Seoul National University College of Medicine. This work was supported by the Global Frontier R and D Program on \Human-centered
Eur Spine J Interaction for Coexistence[ funded by the National Research Foundation of Korea grant funded by the Korean Government (MEST) (NRF-2012M3A6A3055889). Conflict of interest
None.
References 1. Devivo MJ (2012) Epidemiology of traumatic spinal cord injury: trends and future implications. Spinal cord 50:365–372 2. Aarabi B, Alexander M, Mirvis SE, Shanmuganathan K, Chesler D, Maulucci C, Iguchi M, Aresco C, Blacklock T (2011) Predictors of outcome in acute traumatic central cord syndrome due to spinal stenosis. J Neurosurg Spine 14:122–130 3. Hohl JB, Lee JY, Horton JA, Rihn JA (2010) A novel classification system for traumatic central cord syndrome: the central cord injury scale (CCIS). Spine 35:E238–E243 4. Dvorak MF, Fisher CG, Fehlings MG, Rampersaud YR, Oner FC, Aarabi B, Vaccaro AR (2007) The surgical approach to subaxial cervical spine injuries: an evidence-based algorithm based on the SLIC classification system. Spine 32:2620–2629 5. Lamothe G, Muller F, Vital JM, Goossens D, Barat M (2011) Evolution of spinal cord injuries due to cervical canal stenosis without radiographic evidence of trauma (SCIWORET): a prospective study. Annal phys rehabil med 54:213–224 6. Vaccaro AR, Hulbert RJ, Patel AA, Fisher C, Dvorak M, Lehman RA Jr, Anderson P, Harrop J, Oner FC, Arnold P, Fehlings M, Hedlund R, Madrazo I, Rechtine G, Aarabi B, Shainline M (2007) The subaxial cervical spine injury classification system: a novel approach to recognize the importance of morphology, neurology, and integrity of the disco-ligamentous complex. Spine 32:2365–2374 7. Flanders AE, Schaefer DM, Doan HT, Mishkin MM, Gonzalez CF, Northrup BE (1990) Acute cervical spine trauma: correlation of MR imaging findings with degree of neurologic deficit. Radiology 177:25–33 8. Zhuge W, Ben-Galim P, Hipp JA, Reitman CA (2013) Efficacy of MRI for assessment of spinal trauma: correlation with intraoperative findings. J spinal disord techni. doi:10.1097/BSD.0b013 e31827734bc 9. Joseph V, Kumar GS, Rajshekhar V (2009) Cerebrospinal fluid leak during cervical corpectomy for ossified posterior longitudinal ligament: incidence, management, and outcome. Spine 34:491–494 10. Mazur M, Jost GF, Schmidt MH, Bisson EF (2011) Management of cerebrospinal fluid leaks after anterior decompression for ossification of the posterior longitudinal ligament: a review of the literature. Neurosurg Focus 30:E13
11. Co T (1999) Resource for the optimal care of the injured patient. Comm Trauma, Chicago 12. Aydinli U, Karaeminogullari O, Tiskaya K, Ozturk C (2001) Dural tears in lumbar burst fractures with greenstick lamina fractures. Spine 26:E410–E415 13. Silvestro C, Francaviglia N, Bragazzi R, Piatelli G, Viale GL (1991) On the predictive value of radiological signs for the presence of dural lacerations related to fractures of the lower thoracic or lumbar spine. J Spinal Disord 4:49–53 14. Merriam WF, Taylor TK, Ruff SJ, McPhail MJ (1986) A reappraisal of acute traumatic central cord syndrome. J bone joint surg Br 68:708–713 15. Bosch A, Stauffer ES, Nickel VL (1971) Incomplete traumatic quadriplegia. A 10-year review. JAMA J Am Med Assoc 216:473–478 16. Eismont FJ, Wiesel SW, Rothman RH (1981) Treatment of dural tears associated with spinal surgery. J bone joint surg Am 63:1132–1136 17. Chen TY, Crawford NR, Sonntag VK, Dickman CA (2001) Biomechanical effects of progressive anterior cervical decompression. Spine 26:6–13 discussion 14 18. Wang X, Chen Y, Chen D, Yuan W, Zhao J, Jia L, Zhao D (2009) Removal of posterior longitudinal ligament in anterior decompression for cervical spondylotic myelopathy. J spinal disord tech 22:404–407 19. Sohn S, Chung CK, Kim CH (2012) Is closed-suction drainage necessary after intradural primary spinal cord tumor surgery? Eur Spine J 22:577–583 20. Cain JE Jr, Rosenthal HG, Broom MJ, Jauch EC, Borek DA, Jacobs RR (1990) Quantification of leakage pressures after durotomy repairs in the canine. Spine 15:969–970 21. Coplin WM, Avellino AM, Kim DK, Winn HR, Grady MS (1999) Bacterial meningitis associated with lumbar drains: a retrospective cohort study. J Neurol Neurosurg Psychiatry 67:468–473 22. Shapiro SA, Scully T (1992) Closed continuous drainage of cerebrospinal fluid via a lumbar subarachnoid catheter for treatment or prevention of cranial/spinal cerebrospinal fluid fistula. Neurosurgery 30:241–245 23. Weill-Engerer S, Meaume S, Lahlou A, Piette F, Saint-Jean O, Sachet A, Beinis JY, Gallinari C, Grancher AS, Vincent JP, Naga H, Belmin J, Salvatore R, Kazes M, Pautas E, Boiffin A, Piera JB, Duviquet M, Knafo D, Piau A, Miric D, Jean A, Bellamy V, Tissandier O, Le Blanche AF (2004) Risk factors for deep vein thrombosis in inpatients aged 65 and older: a case-control multicenter study. J Am Geriatr Soc 52:1299–1304
123
ORIGINAL ARTICLE
Dural tear and resultant cerebrospinal fluid leaks after cervical spinal trauma Soo Eon Lee • Chun Kee Chung • Tae-Ahn Jahng Chi Heon Kim
•
Received: 22 September 2013 / Revised: 23 April 2014 / Accepted: 24 April 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Purpose Traumatic cervical spinal cord injuries (SCIs) frequently develop dural tears and resultant cerebrospinal fluid (CSF) leaks. They are not usually identified with advanced imaging, and there are no reports on managing CSF leaks after cervical trauma. Hence, the authors evaluated the incidence of CSF leaks after cervical SCIs and described how to predict and manage CSF leaks. Methods An observational retrospective study was done confirming intraoperative CSF leaks among 53 patients with anterior cervical surgery after cervical spine trauma between 2004 and 2011. Results Seven patients (13.2 %) had dural tears and resultant CSF leaks intraoperatively (M:F ratio of 6:1; mean age, 44.7 years). An initial poor American Spinal Injury Association (ASIA) scale was significantly associated with CSF leaks (p = 0.009). From magnetic resonance imaging (MRI), disruption of the ligamentum flavum was correlated with CSF leaks (p = 0.02). Intraoperative application of fibrin glue on the operated site, postoperative management through the early removal of the wound drain within the first 24 h and early rehabilitation were performed in patients with CSF leaks without perioperative insertion of a lumbar drain. During the follow-up period, none of the patients developed CSF-leak-related complications. Conclusion The incidence of CSF leaks after traumatic cervical SCI is relatively higher than that of degenerative cervical spinal surgery. An initial poor neurological status
S. E. Lee C. K. Chung (&) T.-A. Jahng C. H. Kim Department of Neurosurgery, Seoul National University Hospital, Seoul National University College of Medicine, 28 Yeongun-Dong, Jongno-Gu, Seoul 110-744, Korea e-mail: [email protected]
and disruption of the ligamentum flavum on the MRI in patients were predictable factors of dural tears and CSF leaks. Keywords Spinal cord injury Cerebrospinal leaks Dural tears Cervical spine Introduction Cervical spinal trauma is increasing according to extensive literature on the descriptive epidemiology of traumatic spinal cord injury (SCI) [1–5]. Although the optimal timing for surgery is controversial, surgical techniques and approaches for traumatic cervical SCIs have been simplified by the development of the subaxial injury classification (SLIC) scoring system [4, 6]. Three categories including injury morphology, disco-ligamentous complex, and neurologic status are important in identifying the severity of the injury and making decisions about the surgery. As a diagnostic tool, magnetic resonance imaging (MRI) is considered superior to other tools for examining discoligamentous complex injury of patients with neurological deficits and is emerging as the preferred choice to evaluate damage to the spine [7, 8]. However, traumatic dural tears are not usually shown by MRIs, but identified from the resultant CSF leaks in the operative field during surgery. Although various techniques have been described to manage CSF-leak-related complications after surgery for degenerative spinal diseases, there are no reports describing the management of CSF leaks after cervical SCIs [9, 10]. Therefore, the authors evaluated the incidence of intraoperative CSF leaks after cervical SCIs, described how to predict of dural tears and the resultant CSF leaks and suggested simple and effective management for CSF leaks during the perioperative period.
123
Eur Spine J
Materials and methods One hundred thirty-four consecutive patients who underwent surgery for acute cervical spine trauma between 2004 and 2011 were identified. Of these, patients who had an anterior or anterior and posterior approach to the cervical spine and whose operative level was between C3 and C7 were further evaluated. Among them, patients with a CSF leak during the intraoperative period reported in the operative record were included. Known iatrogenic dural tears were not included. The injury mechanism was divided into high and low energy [11]. The preoperative neurologic status was stratified into American Spinal Injury Association (ASIA) scale and described neurologic level. Injury morphology was used to classify injuries in the subaxial spine as follows: compression, burst, distraction, and rotation/translation on cervical spine computed tomography (CT) according to the SLIC system [6]. Disco-ligamentous complex injury was evaluated on the cervical spine MRI.
Table 1 Demographics of the patients CSF-leak group (n = 7)
Non-CSF-leak group (n = 46)
p value
Sex (male:female)
6:1
39:7
1.00
Age (years)
44.7 ± 19.2
49.8 ± 18.0
Injury mechanism
0.49 0.66
High energy injury
6
33
Low energy injury
1
13
Initial ASIA scale
0.009
A B
6 1
13 5
C
0
3
D
0
25
Neurologic level
0.29
C2
0
4
C3
0
8
C4
2
17
C5
3
15
C6
2
3
ASIA American Spinal Injury Association
Results The anterior or anterior and posterior operative approach was used in 53 patients after cervical spinal trauma. Seven patients (13.2 %) were identified as having a CSF leak during the intraoperative period from the operative record (CSF-leak group, Table 1). This group consisted of 6 men and 1 woman, and the mean age at operation was 44.7 ± 19.2 years (range 17–67 years). The non-CSF-leak group had 46 patients (39 men and 7 women) with a mean age of 49.8 ± 18.0 years (range, 19–84 years). The injury mechanism, initial American Spinal Injury Association (ASIA) scale and neurological level are presented in Table 1. Age, sex, injury mechanism and neurologic level were not different, but the initial ASIA scale was statistically significantly different between the two groups (p = 0.009). In addition, the mean follow-up period was 9.7 ± 5.7 months (range 1–18) in the CSF-leak group and 22.81 ± 25.37 months (range 1–104) in the non-CSF-leak group. Preoperative imaging Injury classification on the cervical spine CT and MRI findings are presented in Table 2. All seven patients in the CSF-leak group showed a disrupted disco-ligamentous complex and had an abnormal finding on the MRI as follows: disruption of the ALL, PLL, ligamentum flavum and interspinous ligament, and high signal intensity on the intervertebral disc and spinal cord. In contrast, intermediate changes consisting of isolated interspinous
123
widening or a signal change in the disco-ligamentous complex was only seen in 5 patients, and a disrupted disco-ligamentous complex was observed in 41 patients in the non-CSF-leak group. Injury morphology and discoligamentous complex injury were not different between the two groups (p = 0.40 and p = 0.58, respectively), and the disruption of the LF had statistical significance (p \ 0.05) (Fig. 1). Intraoperative findings and perioperative management With the standard anterior cervical approach, the intraoperative findings of the disco-ligamentous structures and vertebral body are presented in Table 2. In the CSFleak group, seven patients had PLL injury, and a dural tear was simultaneously observed in these patients after a discectomy or corpectomy. Because the margin of the dural tear was rough and complex, primary dural repair could not be attempted. After removal of the ruptured disc material, further resection of the PLL was not performed, and fibrin glue (Green plastÒ, Green Cross, Seoul, Korea) was applied at the dural defect site in the CSF-leak group. During closure, a submuscular drain, a Jackson-Pratt (JP), was inserted, and the wound was closed layer by layer. A postoperative lumbar CSF drain was not inserted in all the patients. After the operation, the patients were kept in a head elevated position and allowed to do activities they could perform, and rehabilitation started as soon as possible in the intensive care unit.
Eur Spine J Table 2 Injury morphology at preoperative images and intraoperative findings CSF-leak group (n = 7)
Non-CSF-leak group (n = 46)
Injury morphology on CT
p value
0.78
Compression
1
8
Burst
1
8
Distraction
1
8
Translation/rotation
5
21
Disco-ligamentous complex injury
1.00
Intermediate
0
5
Disrupted
7
41
6
36
Abnormality on MRI Anterior longitudinal ligament (ALL)
0.65
Intervertebral disc
6
41
0.79
Posterior longitudinal ligament (PLL)
7
33
0.10
Vertebral body
2
16
0.74
Spinal cord
6
31
0.33
Ligamentum flavum
7
26
0.02
Interspinous ligament
6
35
0.57
Intraoperative finding ALL injury
5
20
0.24
Intervertebral disc injury
7
37
0.46
PLL injury
7
9
0.00
Vertebral body injury
3
15
0.67
In the CSF-leak group, the JP drain was removed within the first postoperative day, and the drainage amount was 70.6 cc ± 51.7 (range 31–160). In the non-CSF-leak group, the JP drain was removed within 1.24 days ± 0.52 (range 1–3), and the drainage amount was 30.0 cc ± 16.61 (range 10–69, p \ 0.05).
Fig. 1 Preoperative MRI of patients with traumatic cervical spinal cord injury at C6-7. Sagittal T2-weighted image shows disruption of the ALL and intervertebral disc (A).The disruption of the PLL is shown and (B) the disruption of the ligamentum flavum (C) is also shown. This patient was intraoperatively identified as having a dural tear and resultant CSF leak
Discussion Incidence of CSF leaks after traumatic cervical SCIs
Hospital course The following perioperative complications developed: pneumonia in 9 patients, urinary tract infection (UTI) in 11 patients, bacteremia in 1 patient, and wound oozing in 1 patient. In the CSF-leak group, the following complications developed: pneumonia in 4 patients, UTI in 1 patient and bacteremia in 1 patient. Moreover, complications frequently developed in this group (p \ 0.05). However, none of the patients developed complications related to CSF leaks such as fluid collection, wound dehiscence, meningitis, and pseudomeningocele. One patient in the CSF-leak group who developed bacteremia and sepsis died 1 month after the operation.
SCIs are frequently accompanied by dural tears and resultant CSF leaks [12, 13]. Traumatic dural tears are commonly associated with thoracic-lumbar spinal trauma in the range of 18–36 % [12, 13]. However, traumatic dural tears are not usually predicted with advanced imaging tools such as MRI, but are identified in the operative field during surgery either by decompression or CSF leakage around the injured segments [13]. Because incomplete cervical spinal cord injury is the most common clinical feature and accounts for up to 70 % of all traumatic SCIs, it is important to predict dural tears and manage CSF leaks after cervical spinal trauma [14, 15]. Although various techniques have been described to manage CSF-leak-related complications after surgery for degenerative cervical spinal
123
Eur Spine J
disease [9, 10, 16], there are no reports on managing CSF leaks in cervical spine trauma. The incidence of CSF leaks after anterior decompression surgery ranges from 0.5 to 3 % in degenerative cervical spinal disease [10]. For ossification of the posterior longitudinal ligament (OPLL), the incidence is much higher, ranging from 4.3 to 32 %, because the most common cause of CSF leaks is injury to the dura, while resecting the PLL [9, 10]. In addition, risk factors for intraoperative CSF leaks were revision anterior surgery and the male gender [9, 10]. However, in trauma cases, dural tears mostly developed at the moment of injury. In the present study, seven patients (13.2 %) had a dural tear and the resultant CSF leak, and the incidence after cervical spine trauma was higher than that of degenerative cervical spinal disease and as frequent as OPLL. Although age, sex, injury mechanism, and injury morphology were not associated with dural tears, an initial poor ASIA scale was correlated with CSF leaks. In addition, disruption of the ligamentum flavum on the initial MRI was also a predictive factor of CSF leaks. Therefore, the authors suggest the possibility of intraoperative CSF leaks in patients who present with a poor neurologic status and simultaneously show a disruption of the ligamentum flavum on a preoperative MRI after cervical spinal trauma. Perioperative management of CSF leaks after cervical trauma Our strategy for the management of CSF leaks in spinal surgery for degenerative spinal disease and spinal tumors is as follows: primary dural repair and no wound drain insertion or early removal of the drain and early mobilization. In our trauma cases, PLL injury and dural tears simultaneously developed at the moment of injury, and the resultant CSF leaks already existed before the operation. The PLL was believed to provide stability to the cervical spine, and the removal of PLL could lead to a decrease in cervical stability after surgery [17]. However, the removal of the PLL was helpful to get more decompression in the anterior approach for cervical spondylotic myelopathy (CSM), although it was more technically demanding [18]. In the operative procedure, further resection of the PLL was not meaningful because the purpose of the operation was to provide stability to the unstable spine in trauma cases. Therefore, if CSF leak was observed during the discectomy or the PLL was already injured, we did not try any further resection of the PLL to prevent further dural defects and maintain spinal stability. In addition, primary dural repair was not accomplished, and duraplasty was also not achieved due to a wide and rough defect of the dura. In this situation, the authors managed the CSF leaks by applying fibrin glue over the operated site. Although a
123
wound drain was inserted during closure, the drain was removed if CSF appeared in the drain tube or bag, and the tube was clamped and removed regardless of the drainage amount within the first 24 h [19]. In addition, a lumbar drain was not inserted after the operation. Instead, the patients were not allowed bed rest and encouraged early ambulation using a wheel chair. If CSF accumulated under the surgical wound, it was observed unless the symptoms were related to an airway obstruction. As time went by, CSF accumulation gradually disappeared [19]. Wound drainage and lumbar drainage were prescribed to manage CSF leaks during the postoperative period [10, 20]. However, significant controversy exists over the use of wound drains due to the fear of developing duracutaneous fistulas [20]. The use of lumbar drains is also associated with infections such as meningitis, over-drainage of the CSF and pneumocephalus with brainstem compression [21, 22]. In addition, because the standard position of patients with lumbar drainage is bed rest to prevent over-drainage of the CSF, bed rest may increase the risk of postoperative morbidity due to the development of deep venous thrombosis and pulmonary embolism after prolonged immobilization [23]. The present study involved a relatively small number of patients and a short follow-up period. The authors had to include patients who underwent an anterior cervical surgical approach because the intraoperative findings were used as a reference tool for comparison to the preoperative MRI findings. Moreover, patients with a poor neurologic status had difficulty in continuing their follow-up after transfer to another hospital. Although this study is valuable in identifying the predictive factors of CSF leaks after cervical spinal trauma and suggests a simple management for CSF leaks, the study has limitations in concluding an ultimate management strategy for CSF leaks in cervical spinal surgery.
Conclusion The incidence of CSF leaks after cervical spinal trauma is relatively higher than that of degenerative cervical spinal disease. ASIA A or B as an initial neurological status and disruption of the ligamentum flavum on the initial MRI in patients were predictive factors of CSF leaks. The authors suggest a simple and effective management of CSF leaks by early removal of wound drains and not using lumbar drains. Acknowledgments The authors would like to express their appreciation for the statistical analysis done by the Medical Research Collaborating Center at Seoul National University Hospital and the Seoul National University College of Medicine. This work was supported by the Global Frontier R and D Program on \Human-centered
Eur Spine J Interaction for Coexistence[ funded by the National Research Foundation of Korea grant funded by the Korean Government (MEST) (NRF-2012M3A6A3055889). Conflict of interest
None.
References 1. Devivo MJ (2012) Epidemiology of traumatic spinal cord injury: trends and future implications. Spinal cord 50:365–372 2. Aarabi B, Alexander M, Mirvis SE, Shanmuganathan K, Chesler D, Maulucci C, Iguchi M, Aresco C, Blacklock T (2011) Predictors of outcome in acute traumatic central cord syndrome due to spinal stenosis. J Neurosurg Spine 14:122–130 3. Hohl JB, Lee JY, Horton JA, Rihn JA (2010) A novel classification system for traumatic central cord syndrome: the central cord injury scale (CCIS). Spine 35:E238–E243 4. Dvorak MF, Fisher CG, Fehlings MG, Rampersaud YR, Oner FC, Aarabi B, Vaccaro AR (2007) The surgical approach to subaxial cervical spine injuries: an evidence-based algorithm based on the SLIC classification system. Spine 32:2620–2629 5. Lamothe G, Muller F, Vital JM, Goossens D, Barat M (2011) Evolution of spinal cord injuries due to cervical canal stenosis without radiographic evidence of trauma (SCIWORET): a prospective study. Annal phys rehabil med 54:213–224 6. Vaccaro AR, Hulbert RJ, Patel AA, Fisher C, Dvorak M, Lehman RA Jr, Anderson P, Harrop J, Oner FC, Arnold P, Fehlings M, Hedlund R, Madrazo I, Rechtine G, Aarabi B, Shainline M (2007) The subaxial cervical spine injury classification system: a novel approach to recognize the importance of morphology, neurology, and integrity of the disco-ligamentous complex. Spine 32:2365–2374 7. Flanders AE, Schaefer DM, Doan HT, Mishkin MM, Gonzalez CF, Northrup BE (1990) Acute cervical spine trauma: correlation of MR imaging findings with degree of neurologic deficit. Radiology 177:25–33 8. Zhuge W, Ben-Galim P, Hipp JA, Reitman CA (2013) Efficacy of MRI for assessment of spinal trauma: correlation with intraoperative findings. J spinal disord techni. doi:10.1097/BSD.0b013 e31827734bc 9. Joseph V, Kumar GS, Rajshekhar V (2009) Cerebrospinal fluid leak during cervical corpectomy for ossified posterior longitudinal ligament: incidence, management, and outcome. Spine 34:491–494 10. Mazur M, Jost GF, Schmidt MH, Bisson EF (2011) Management of cerebrospinal fluid leaks after anterior decompression for ossification of the posterior longitudinal ligament: a review of the literature. Neurosurg Focus 30:E13
11. Co T (1999) Resource for the optimal care of the injured patient. Comm Trauma, Chicago 12. Aydinli U, Karaeminogullari O, Tiskaya K, Ozturk C (2001) Dural tears in lumbar burst fractures with greenstick lamina fractures. Spine 26:E410–E415 13. Silvestro C, Francaviglia N, Bragazzi R, Piatelli G, Viale GL (1991) On the predictive value of radiological signs for the presence of dural lacerations related to fractures of the lower thoracic or lumbar spine. J Spinal Disord 4:49–53 14. Merriam WF, Taylor TK, Ruff SJ, McPhail MJ (1986) A reappraisal of acute traumatic central cord syndrome. J bone joint surg Br 68:708–713 15. Bosch A, Stauffer ES, Nickel VL (1971) Incomplete traumatic quadriplegia. A 10-year review. JAMA J Am Med Assoc 216:473–478 16. Eismont FJ, Wiesel SW, Rothman RH (1981) Treatment of dural tears associated with spinal surgery. J bone joint surg Am 63:1132–1136 17. Chen TY, Crawford NR, Sonntag VK, Dickman CA (2001) Biomechanical effects of progressive anterior cervical decompression. Spine 26:6–13 discussion 14 18. Wang X, Chen Y, Chen D, Yuan W, Zhao J, Jia L, Zhao D (2009) Removal of posterior longitudinal ligament in anterior decompression for cervical spondylotic myelopathy. J spinal disord tech 22:404–407 19. Sohn S, Chung CK, Kim CH (2012) Is closed-suction drainage necessary after intradural primary spinal cord tumor surgery? Eur Spine J 22:577–583 20. Cain JE Jr, Rosenthal HG, Broom MJ, Jauch EC, Borek DA, Jacobs RR (1990) Quantification of leakage pressures after durotomy repairs in the canine. Spine 15:969–970 21. Coplin WM, Avellino AM, Kim DK, Winn HR, Grady MS (1999) Bacterial meningitis associated with lumbar drains: a retrospective cohort study. J Neurol Neurosurg Psychiatry 67:468–473 22. Shapiro SA, Scully T (1992) Closed continuous drainage of cerebrospinal fluid via a lumbar subarachnoid catheter for treatment or prevention of cranial/spinal cerebrospinal fluid fistula. Neurosurgery 30:241–245 23. Weill-Engerer S, Meaume S, Lahlou A, Piette F, Saint-Jean O, Sachet A, Beinis JY, Gallinari C, Grancher AS, Vincent JP, Naga H, Belmin J, Salvatore R, Kazes M, Pautas E, Boiffin A, Piera JB, Duviquet M, Knafo D, Piau A, Miric D, Jean A, Bellamy V, Tissandier O, Le Blanche AF (2004) Risk factors for deep vein thrombosis in inpatients aged 65 and older: a case-control multicenter study. J Am Geriatr Soc 52:1299–1304
123
