Eur Spine J (2015) 24:348–357 DOI 10.1007/s00586-014-3297-8
ORIGINAL ARTICLE
Symptomatic epidural hematoma after lumbar decompression surgery Fu-Cheng Kao • Tsung-Ting Tsai • Lih-Huei Chen • Po-Liang Lai • Tsai-Sheng Fu • Chi-Chien Niu • Natalie Yi-Ju Ho • Wen-Jer Chen • Chee-Jen Chang
Received: 3 October 2013 / Revised: 23 February 2014 / Accepted: 28 March 2014 / Published online: 24 April 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Background context Postoperative symptomatic epidural hematoma (SEH) is a serious complication of lumbar spine surgery. Despite its rarity, this uncommon complication may result in devastating neurological sequelae, including lower limb weakness. Purpose A retrospective study was made to identify possible risk factors of postoperative spinal epidural hematoma by reviewing the clinical cases of this rare complication and analyzing the postoperative evaluations of patients. Methods From 2002 to 2010, out of 15,562 who underwent lumbar decompression procedure with/without instrumentation, 25 patients required reoperation for epidural hematoma after the initial spinal surgery. For the control group, another 75 patients were randomly selected from the pool of patients who received lumbar decompression surgery during the same period of time. The medical records of preoperative, intraoperative and postoperative factors were collected to determine possible risk
F.-C. Kao and T.-T. Tsai have contributed equally to this work. F.-C. Kao T.-T. Tsai (&) L.-H. Chen P.-L. Lai T.-S. Fu C.-C. Niu N. Y.-J. Ho W.-J. Chen Spine Section, Department of Orthopaedic Surgery, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, No. 5, Fusing St., Gueishan, Taoyuan 333, Taiwan e-mail:
[email protected] T.-T. Tsai C.-J. Chang Graduate Institute of Clinical Medical Sciences, Chang Gung University, Taoyuan, Taiwan C.-J. Chang Biostatistical Center for Clinical Research, Chang Gung Memorial Hospital, Taoyuan, Taiwan
123
factors by comparing between the cases and controls, and the postoperative evaluations of muscle power, intractable pain, saddle anesthesia, time to detection and time to evacuation were analyzed to find if there is any significant relation within the case group. Mann–Whitney U test, twosample t test, v2 test and Fisher’s exact test were used for statistical analysis. Results The incidence of postoperative symptomatic epidural hematoma is 0.16 %. After the initial procedure, 20 (80 %) patients developed progressive decrease in muscle power (MP B 3), 14 (56 %) patients had intractable pain (VAS C 7), and 19 (76 %) patients had saddle anesthesia. Preoperative diastolic blood pressure, intraoperative use of gelfoam for dura coverage and postoperative drain output were statistically significant risk factors (p \ 0.01). Within the SEH case group, postoperative symptom of decreased muscle power had significant relation with blood loss, laminectomy level and fusion level (p = 0.016, 0.021, 0.010). If the symptom of decreased muscle power or perianal anesthesia was not improved after hematoma evacuation, there was a tendency for permanent leg weakness after 1-year follow-up (p = 0.001, 0.003). Conclusions The findings suggest that preoperative diastolic blood pressure, intraoperative use of gelfoam for dura coverage and postoperative drain output are risk factors for symptomatic epidural hematoma after lumbar decompression surgery. Major blood loss and multilevel surgical procedure could result in poor recovery of muscle power. After spine decompression surgery, early detection and evacuation of hematoma are the key to avoid neurologic deterioration and have better clinical outcomes. Keywords Symptomatic epidural hematoma Spinal decompression Postoperative complication Cauda equina syndrome
Eur Spine J (2015) 24:348–357
349
Introduction The occurrence of spinal epidural hematoma is very rare, and it was first described as a clinical diagnosis by Jackson in 1869 [1]. Since then, many studies have focused on this serious complication with numerous etiologies. In a review of literature, some epidural hematomas occur spontaneously with unknown causes, but others have been associated with vascular anomalies such as arteriovenous malformation (AVM), vertebral hemangiomas, trauma (including vertebral fracture), obstetrical birth trauma, lumbar puncture and epidural procedures [2, 3]. In addition, surgical bleeding, missile injuries, hypertension, physical exertion and abnormal bleeding tendencies, whether from hematological disorders or from anticoagulant therapy have also been reported to result in spinal epidural hematoma [4, 5]. Sometimes symptomatic epidural hematoma can happen after spinal operations. Although asymptomatic epidural hematoma has been identified in 33–100 % of patients undergoing lumbar decompression surgery by computed tomography scans and magnetic resonance images [6, 7], the incidence rate of symptomatic postoperative epidural hematoma is only 0.1–0.24 % [8–10]. Although rare, it can cause devastating neurological consequences, including bowel and bladder dysfunction, saddle anesthesia, sciatica and motor weakness of lower extremities and sexual dysfunction, which are also seen in cauda equina syndrome [11]. Knowing risk factors that cause spinal epidural hematoma can help surgeons make a timely diagnosis of this serious complication with important medical and legal implication. Thus far, only a few studies have investigated risk factors of symptomatic epidural hematoma [10, 12, 13]. The aim of this study was to identify possible risk factors by reviewing the clinical cases of postoperative symptomatic epidural hematomas (SEH) and analyzing the postoperative outcomes of these patients.
Materials and methods We retrospectively reviewed the database of all surgeries performed in spine department at our institution over an 8-year period between 2002 and 2010. A total of 15,562 lumbar spine decompression surgeries were performed, including 12,377 laminectomies and 3,185 laminotomies with/without discectomy and posterior instrumentation. In this study, patients who required revision hematoma evacuation after the initial decompression procedure and were coded with the International Classification of Diseases, Ninth revision, Clinical modification (ICD-9-CM, 998.12) for a hematoma complicating procedure were included. The diagnosis of SEH was made on the basis of
Fig. 1 The patient (No. 11) was a 56-year-old female who underwent laminectomy procedure with posterior instrumentation for degenerative spondylolisthesis. The symptoms of progressive leg weakness and perianal numbness were observed 9.5 h after the initial surgery. Under the impression of symptomatic epidural hematoma, the patient received reoperation for hematoma evacuation immediately. A hematoma accumulation was revealed around initial surgical area (left). The dural sac was compressed almost completely flat with total occlusion of cerebrospinal fluid, and unnecessary bony decompression procedure was avoided to prevent further blood oozing and hematoma formation. After gently removing hematoma by suction tube, the dural sac was re-expanded and nerves were decompressed (right). The symptoms of muscle power decrease and perianal numbness were improved after evacuation and no neurologic deficit was found at 1-year follow-up
the development of neurological deficit and was confirmed by magnetic resonance imaging (MRI) assessment or surgical finding. The intraoperative view of epidural hematoma formation during evacuation procedure was shown in Fig. 1. Patients who had infective spondylitis with epidural abscess, spinal trauma or symptomatic epidural hematoma but received the initial procedure at other hospitals were excluded. 25 patients who met our inclusion criteria were assigned as the case group, and all had developed some degree of neurological compromise. Among them, 22 patients underwent laminectomies with posterolateral fusion and other three underwent laminotomies only. The general patient data are shown in Table 1, including age, gender, diagnosis, initial surgical procedure, symptoms of epidural hematoma and neurologic status at 1-year follow-up. For the control group, another 75 patients were randomly selected from the pool of patients who received lumbar decompression surgery during the same period of time but without developing any postoperative complication. Meanwhile, there were five patients who received hematoma evacuation during this period in our hospital were excluded in this study: three who had initial spinal surgery at other hospitals were transferred to our emergency department for image survey and hematoma evacuation, and the another two suffered from spinal trauma complicated with hematoma compression. A detailed review of the medical record was carried out for each patient. The considered risk factors were categorized into three groups: preoperative, intraoperative and postoperative factors. The following data were collected to determine possible risk factors of symptomatic epidural
123
123
66/F
25
HIVD
HIVD
HIVD
Spondylolytic spondylolisthesis; Compression fx
Failed back surgery syndrome
SSS; Compression fx
SSS; DDD
SSS; DDD
SSS; Spondylolisthesis
SSS; Spondylolisthesis
SSS; HIVD
SSS
Spondylolisthesis
SSS; HIVD
Degenerative spondylolisthesis; DDD
SSS; Degenerative spondylolisthesis
SSS; Degenerative scoliosis
LAMI; DISC
LAMI; DISC
LAMI; DISC
LAMI; VP
PLF; PLIF; LAMI
LAMI; VP
PLF; TLIF; LAMI
PLF; PLIF; LAMI
PLF; LAMI
PLF; LAMI
LAMI; DISC
LAMI
PLF; LAMI
LAMI; DISC
PLF; LAMI; PLIF
PLF; LAMI; PLIF
PLF; LAMI
LAMI; DISC PLF; LAMI; PLIF
PLF; LAMI
LAMI; VP
PLF; LAMI
LAMI; DISC
PLF; LAMI
PLF; LAMI
Initial surgical procedure
VAS = 9, Sa
VAS = 8, Sa
VAS = 9, Sa
MP;, VAS = 7, Sa
MP;, Sa
MP;
MP;, Sa
MP;, VAS = 7, Sa
MP;
VAS = 7, Sa
MP;, VAS = 8, Sa
MP;
MP;, VAS = 7, Sa
MP;, VAS = 7, Sa
MP;, Sa
MP;, Sa
MP;, Sa
MP;, VAS = 7, Sa MP;, VAS = 7
VAS = 9, Sa
MP;, VAS = 8
MP;, Sa
MP;, VAS = 9
MP;, Sa
MP;, Sa
Symptoms of epidural hematoma
Numbness(i), VAS = 5
Numbness(i), VAS = 4
Numbness(i), VAS = 4
MP(n), numbness(n), VAS 5, incontinence
MP(n), numbness(n), incontinence
MP(n), incontinence
MP(n), numbness(n), incontinence
MP(i), numbness(i), VAS = 6
MP(i)
numbness(n), VAS = 7
MP(i), numbness(i), VAS = 5
MP(i)
MP(i), numbness(n), VAS = 6
MP(i), numbness(n), VAS = 8
MP(i), numbness(i)
MP(i), numbness(i)
MP(i), numbness(i)
MP(i), numbness(n), VAS = 6 MP(i), VAS = 5
VAS = 5, numbness(i)
MP(i), VAS = 4
MP(i), numbness(n)
MP(i), VAS = 4
MP(i), numbness(i)
MP(i), numbness(i)
Outcome after hematoma evacuation
MP = 5, numbness(-)
MP = 5, numbness(-)
MP = 5, numbness(-)
MP = 2, numbness(?), incontinence(?)
MP = 3, numbness(?), incontinence(?)
MP = 3, incontinence(?)
MP = 2, numbness(?), incontinence(?)
MP = 5, numbness(?)
MP = 4
MP = 5, numbness(?)
MP = 5, numbness(-)
MP = 5
MP = 4, numbness(?)
MP = 4, numbness(?)
MP = 5, numbness(-)
MP = 5, numbness(-)
MP = 5, numbness(-)
MP = 5, numbness(?) MP = 5
MP = 5, numbness(-)
MP = 5
MP = 4, numbness(?)
MP = 5
MP = 4, numbness(-)
MP = 5, numbness(-)
Neurological status at 1-year follow-up
SSS Spinal stenosis syndrome, DDD degenerative disc disease, HIVD herniated intervertebral disc, fx fracture, PLF posterior lumbar fusion, LAMI laminectomy, DISC discectomy, VP vertebroplasty, PLIF posterior lumbar interbody fusion, TLIF transforaminal lumbar interbody fusion, MP; progressive decrease in muscle power (B3), VAS C 7 intractable pain, visual analog scale, Sa saddle anesthesia, (i) improved, (n) non-improved
32/M
82/F
22
43/M
69/M
21
23
79/F
20
24
71/F
77/F
18
54/F
17
19
73/F
60/M
15
16
62/F
77/F
13
47/M
12
14
56/F
11
Degenerative spondylolisthesis
48/F
77/M
9
35 M 62/F
7 8
10
HIVD SSS; Kypo-scoliosis
68 M
6
Degenerative spondylolisthesis
SSS; Compression fx
75/F
83/F
4
HIVD
SSS; Degenerative spondylolisthesis
SSS; Kypho-scoliosis
Diagnosis
5
78/F
71/F
2
70/F
1
3
Age/gender
Patient no.
Table 1 The general data of 25 case patients with postoperative epidural hematoma
350 Eur Spine J (2015) 24:348–357
Eur Spine J (2015) 24:348–357
hematoma by comparing between 25 case patients and 75 control patients: gender, age, body weight, diabetes, hypertension, blood type, use of anticoagulants, International normalized ratio (INR), platelet count, diastolic blood pressure (DBP), systolic blood pressure (SBP) and American Society of Anesthesiologist (ASA) physical status classification as preoperative factors; operative level, laminectomy level, spine fusion level, blood loss, operation duration, urine output and use of gelfoam or cross-link as intraoperative factors; drain output as postoperative factors. In addition, the postoperative outcomes were analyzed within the case group. For all 25 case patients, the status of muscle power, intractable pain and saddle anesthesia were evaluated before and after the reoperation for removing hematoma. The grading of muscle power (MP [ 3 or B 3) was recorded by the worst lower-extremity score, and patients with progressive decrease in postoperative muscle power were divided into two subgroups based on the recovery status of muscle power. The MP (improved) subgroup included patients with impaired muscle power (MP B 3) before hematoma evacuation but who recovered to normal strength (MP [ 3) after evacuation, and the MP (non-improved) subgroup included patients that still had incomplete recovery of muscle power (remain as MP B 3) after evacuation. The score of intractable pain (VAS C 7) was recorded by the degree of pain using visual analog pain scale. Similarly, patients with perianal numbness were also divided into two subgroups based on the improvement of numbness. The numbness (improved) subgroup included patients who suffered from perianal numbness before evacuation but whose numbness disappeared after reoperation, and the numbness (non-improved) subgroup included patients with no numbness improvement even after reoperation. Time to detection was defined as the time interval between the initial spinal surgery and the onset of symptoms and signs, and time to evacuation was the time elapsed between the onset of symptoms and the completion of evacuation procedure. If factors are continuous variables, the two-sample t test and Mann–Whitney U test were used to compare between the cases and controls and analyze the clinical outcomes within the case group. If factors are categorical variables, the v2 test and Fisher’s exact test were performed. The significant level alpha was 0.05. All analyses were performed by SPSS statistics software version 17.0 (Chicago, IL).
Results Of 15,562 patients receiving lumbar spine surgery from 2002 to 2010, only 25 patients developed postoperative symptomatic epidural hematoma and required reoperation
351
for hematoma evacuation. The incidence rate was approximately 0.16 %, counting 0.18 % for patients that underwent laminectomy and 0.09 % for laminotomy procedure. Among the case patients, 20 (80 %) patients presented progressive loss of postoperative muscle power (MP B 3), 14 (56 %) patients had intractable pain (VAS C 7), and 19 (76 %) patients had saddle anesthesia after the initial spine surgery. After emergent hematoma evacuation, there were 16 patients as the MP (improved) subgroup with recovery of muscle power and four patients as the MP (non-improved) subgroup without recovery even after 1-year follow-up; there were 11 patients as the numbness (improved) subgroup with immediate numbness improvement and 8 patients as the numbness (non-improved) subgroup without any improvement. In this study, a statistical comparison was made between the case and control groups to identify possible risk factors of SEH (Table 2). Preoperative factor of diastolic blood pressure, intraoperative factor of using gelfoam for dura coverage and postoperative factor of drain volume proved to be statistically significant for predisposing postoperative SEH (p \ 0.01), whereas all other risk factors such as multilevel procedure, advanced age, bleeding tendency and intraoperative blood loss showed no statistical significance. In addition, statistical analyses were made within the case group for clinical evaluation of motor and sensory deficits, and the data were summarized in tables: analysis of pre-evacuation symptom in Table 3, and analysis of post-evacuation recovery status in Table 4. Comparing the symptoms before hematoma evacuation among the 25 case patients (Table 3), we found postoperative SEH presenting by decreased muscle power was significantly related to blood loss, laminectomy level and vertebral fusion level (p = 0.016, 0.021, 0.010). The results indicated that the mean blood loss (1567.50 ml) during the initial procedure in patients with postoperative muscle power\3 was greater than those with normal muscle power (156.00 ml). Also, there was a significant difference between the status of perianal numbness and time to detection (p = 0.004). The mean time to detection (33.89 h) was shorter for patients with perianal numbness than for those without numbness (124.67 h). The analysis of clinical outcomes after hematoma evacuation (Table 4) showed there was a correlation between time to evacuation and the recovery status of muscle power. The mean of time to evacuation was 7.44 h for the recovered patients and 17.88 h for the non-recovered patients. Post-evacuation recovery outcomes of muscle power and perianal numbness were significantly related to muscle power status at 1-year follow-up (p = 0.001, 0.003). At one-year follow-up, the mean muscle power was 2.50 for the MP (non-improved) subgroup and 4.69 for the MP (improved) subgroup, and the mean muscle power was
123
352
Eur Spine J (2015) 24:348–357
Table 2 A comparison between the case and control patients for risk factors of symptomatic epidural hematoma Pre-op factors
Control
Case
Mean
SD
Mean
64.24
13.48
64.6
14.47
60.66
8.39
64.66
11.33
0.06
pre-op SBP (mmHg)
134.72
24.79
138.96
18.12
0.43
pre-op DBP (mmHg)
70.96
11.32
81.16
11.50
\0.01
1.02
0.08
1
0.09
0.24
231.48
75.16
218.56
61.86
0.44
Count
%
Count
%
Count
%
Gelfoam dural coverage
SD
Body weight (kg)
Platelets (mm3)
Count
p
Age
INR
Table 2 continued
0.91
%
–
60
80
12
48
?
15
20
13
52
\0.01
Post-op factors
Mean
SD
Mean
SD
p
Drain volume (ml)
900.32
719.28
465.44
530.03
\0.01
OP Operative, DM diabetes, HTN hypertension, SBP systolic blood pressure, DBP diastolic blood pressure, ASA American society of anesthesiologist Two-sample t test for continuous factors
Gender
Pearson v2 test and
Female
54
72 %
16
64 %
Male
21
28 %
9
36 %
A
21
28
5
20
AB
3
4
3
12
B
9
12
3
12
O DM
42
56
14
56
60
80
21
84
?
15
20
4
16
0.48
Discussion
0.77
HTN –
55
73
15
60
?
20
27
10
40
0.21
Spinal OP history –
60
80
21
84
?
15
20
4
16
0.77
Use of anticoagulants –
69
92
21
84
?
6
8
4
16
0.26
ASA classification 1
I
3
4
1
4
II III
27 45
36 60
9 15
36 60
Intra-op factors
Mean
SD
Mean
SD
p
Blood loss (cc)
658.4
537.05
1285.2
1634.43
0.07
OP duration (min)
190.68
63.76
196.8
82.19
0.74
Urine output (ml)
215.2
184.4
222
183.17
0.87
Laminectomy level
2.48
1
2.72
1.43
0.44
Fusion level Number of cage
2.6 0.6
1.45 0.75
3.2 0.44
1.68 0.87
0.09 0.38
Count
%
Count
%
Use of cross-link –
66
88
23
92
?
9
12
2
8
123
Fisher’s exact test for categorical variables
3.63 for the numbness (non-improved) subgroup and 4.91 for numbness (improved) subgroup.
Blood type
–
0.45
0.73
Postoperative epidural hematoma is one of the most important factors causing spinal canal compromise and nerve root compression after lumbar decompression surgery. It can form a space occupying lesion and cause lumbosacral nerve root compression below the conus within the spinal canal. Nerve root dysfunction is usually incomplete and due to ischemia of direct compression. Thus, patients after lumbar decompression surgery with symptomatic epidural hematoma mostly would complain about some neurological deficits including unilateral or bilateral sciatica, motor weakness of lower extremities, sensory disturbance in saddle area, detrusor dysfunction and loss of anal tone [14]. Based on the data presented in this study, we interestingly found the following as significant risk factors for causing postoperative symptomatic epidural hematoma: preoperative diastolic blood pressure, intraoperative use of gelfoam for dura coverage and postoperative drain output. Other risk factors, such as advanced age ([60 years), use of anticoagulants, multilevel procedure and blood loss, have been identified as predisposing causes for postoperative SEH in previous studies [8, 10–12]; however, these factors did not show any statistical significance in our work. A side-by-side comparison is shown in Table 5 to better understand and explain the differences in the results between our study and Awad’s study [8]. In both studies, the use of anti-coagulants was not associated with an increased risk of hematoma formation. The volume of postoperative drain showed significant difference in our study, but not in Awad’s study. Many studies have reported the relationship between hypertension and blood viscosity [15, 16]. In a study by de
Eur Spine J (2015) 24:348–357
353
Table 3 Pre-evacuation symptom analysis within the case group Factors
Muscle power before hematoma evacuation
Perianal numbness before hematoma evacuation
MP [ 3 Mean ± SD
MP B 3 Mean ± SD
p
No numbness Mean ± SD
Numbness Mean ± SD
p
Blood loss (cc)
156.00 ± 197.94
1567.50 ± 1716.58
0.016*
483.33 ± 461.16
1538.42 ± 1794.63
0.726
Urine output (ml)
160.00 ± 74.16
237.50 ± 199.89
0.657
206.67 ± 173.63
226.84 ± 190.41
0.898
Drain volume (ml) Laminectomy level
498.00 ± 717.37 1.20 ± 1.64
457.30 ± 496.16 3.10 ± 1.12
0.434 0.021*
393.50 ± 303.45 3.00 ± 1.10
488.16 ± 588.84 2.63 ± 1.54
0.610 0.948
Fusion level
1.20 ± 1.64
3.70 ± 1.30
0.010*
3.33 ± 1.51
3.16 ± 1.77
0.871
Discectomy level
0.60 ± 0.55
0.75 ± 0.91
0.941
0.50 ± 0.84
0.79 ± 0.85
0.405
Initial OP duration (h)
2.20 ± 1.15
3.55 ± 1.31
0.052
3.00 ± 1.26
3.37 ± 1.42
0.559
Time to detection (h)
48.20 ± 60.85
57.55 ± 63.84
0.946
124.67 ± 67.24
33.89 ± 42.44
0.004*
MP Muscle power, SD standard deviation, OP operative Two-sample t test and Mann–Whitney test for continuous variables * Represents a statistical significance (p \ 0.05)
Table 4 Post-evacuation recovery status analysis within the case group
MP Muscle power, SD standard deviation, OP operative, F/U follow-up Two-sample t test and Mann– Whitney test for continuous variables * Represents a statistical significance (p \ 0.05)
Post-evacuation status
Muscle power B3 before hematoma evacuation
Perianal numbness before hematoma evacuation
Improved Mean ± SD
Improved Mean ± SD
Non-improved Mean ± SD
p
Non-improved Mean ± SD
p
Laminectomy level
3.19 ± 0.83
2.75 ± 2.06
0.729
2.55 ± 1.75
2.75 ± 1.28
0.931
Fusion level
3.50 ± 1.15
4.50 ± 1.73
0.157
2.91 ± 2.12
3.50 ± 1.20
0.673
Discectomy level
0.81 ± 0.91
0.50 ± 1.00
0.441
0.82 ± 0.87
0.75 ± 0.89
0.858
59.50 ± 68.39
49.75 ± 48.23
0.850
25.09 ± 44.21
46.00 ± 39.37
0.083
Time to evacuation (h)
7.44 ± 3.43
17.88 ± 13.02
0.129
17.18 ± 21.86
9.06 ± 4.92
0.590
MP at 1-year F/U
4.69 ± 0.48
2.50 ± 0.58
0.001*
4.91 ± 0.30
3.63 ± 1.19
0.003*
Time to detection (h)
Simone and Devereux, [17] they examined the relation of whole blood viscosity with cardiovascular risk factors in 128 normal adults and found diastolic blood pressure was an independent predictor of whole blood viscosity. According to our study, diastolic blood pressure was associated with the formation of epidural hematoma, so we hypothesized that high diastolic blood pressure may lead to an increase in whole blood viscosity which can easily cause blood clots and drain dysfunction. Therefore, venous blood oozed out and accumulated in epidural space and caused lumbar root compression by mass effect. There is a possible correlation between diastolic blood pressure and hematoma formation, so hypertensive patients should be carefully monitored to avoid blood accumulation. Gelfoam was initially introduced in the 1940s as a hemostatic agent showing results in preventing scar adhesion by LaRocca and Macnab [18], and it has been widely used to control bleeding during spinal surgery [19]. However, it can possibly cause mass effect with compression by absorbing blood and becoming surrounded with clot for as
long as 3 days, and the maximum tissue response to its presence usually occurs by postoperative day 12 [20]. Large pieces of gelfoam placed over dura may form a barrier which limits blood oozing out of epidural vein and causes blood to accumulate around dura and nerve roots. This not only could cause severe mass effect for producing compression to dura and nerve roots, but also drain dysfunction. Until now, no previous study has identified the use of gelfoam during spinal surgery as a predisposing risk factor for postoperative epidural hematoma, but this intraoperative variable demonstrated an increased risk of hematoma formation in our study. Besides causing epidural hematoma, using gelfoam during brain and spine surgeries could also cause other postoperative complications such as cauda equina syndrome, myelopathy, acute quadriparesis and infection with neurological deterioration [20–24]. Care should be taken when covering the exposed dura with gelfoam because overuse of gelfoam may delay bioresorption and cause compression to spinal cord due to mass effect.
123
354
Eur Spine J (2015) 24:348–357
Table 5 A side-by-side comparison of our study and Awad study for the case group Pre-op factors
Our study (n = 25)
Awad’s study [8] (n = 32)
Incidence rate
0.16 %
0.2 %
Age (yr)
64.6 (32–82)
60.47 (17–87)
Female
16 (64 %)
18 (56.3 %)
Male
9 (36 %)
14 (43.7 %)
A AB
5 (20 %) 3 (12 %)
8 (25 %) 1 (3.1 %)
B
3 (12 %)
4 (12.5 %)
O
Gender
Blood type
14 (56 %)
19 (69.4 %)
INR
1 (0.87–1.3)
1 (0.8–1.4)
DM
4 (16 %)
5 (15.6 %)
HTN
10 (40 %)
15 (46.9 %)
Spinal OP history
4 (16 %)
16 (50 %)
Use of anticoagulants
4 (16 %)
1 (3.1 %)
Platelets (103/ll)
218.56 (80–338)
281 (95–632)
Intra-op factors Blood loss (ml)
1285.2 (50–5200)
OP duration (h)
3.28 (1–5.5)
2065.8 (50–10000) 4.7 (1.5–8.5)
Fusion level
3.2 (0–6)
4.58 (0–12)
OP Operative, DM diabetes, HTN hypertension, INR International normalized ratio Presented as mean (range or percentage)
A closed wound suction drainage system (Hemovac drain) is often used in spine surgery for removing any blood and other fluids that might build up in surgical area, which can prevent postoperative epidural blood accumulation. Mirzai et al. [25] evaluated 50 patients undergoing lumbar disc surgery with or without insertion of drains in the epidural space. The MRI results showed that the use of a drain could decrease the incidence rate and reduce the size of postoperative epidural hematoma [25], especially it is more favorable to place drains under the fascia then above the fascia. In our study, hemovac drains were inserted in all patients except for ones that underwent discectomy procedure in the initial surgery. Comparing between the cases and controls, intraoperative blood loss showed no significant difference but low postoperative drain volume was statistically associated with SEH. Sufficient drainage after spine surgery may play an important role in preventing postoperative SEH. In our study, symptomatic epidural hematoma occurred at an overall rate of 0.16 %, which seemed to fall within the range from 0.1 to 0.2 %, as previously reported [16– 19]. According to Kebaish and Awad [9], the clinical
123
Fig. 2 a The patient (No. 3) was a 71-year-old female that received laminectomy and discectomy with intraoperative use of gelfoam for covering the dura. Progressive motor weakness was found 4 days after surgery, and MRI was immediately assessed. MRI evaluation revealed a hyperacute epidural hematoma formation: T2-weighted sagittal view (left) and axial view (lower right) of L–S spine showed hyperintense lesion and T1-weighted axial view (upper right) revealed hypointense lesion and demonstrated a significant degree of mass effect with compression causing severe postoperative spinal stenosis at L5–S1 level. b The patient (No. 14) who received laminectomy procedure without using gelfoam during the initial surgery and MRI evaluation was assessed on postoperative day 6. The appearance of hypointense lesions on T2-weighted sagittal view (left) and both T1- and T2-weighted axial view (right) of L–S spine indicated acute hematoma formation, which caused severe postoperative spinal stenosis at L4–L5 with mass effect
evaluation was the most important tool for early diagnosis of spinal epidural hematoma and the clinical outcomes of patients were dependent on the rate of development of symptoms, interval to surgery, level of spinal involvement
Eur Spine J (2015) 24:348–357
355
Fig. 3 A medical algorithm is designed to follow in our clinical practice for early diagnosis and management of postoperative symptomatic epidural hematoma
and degree of neurological deficit. Therefore, the clinical profiles of the SEH patients between the initial spinal surgery and reoperation for hematoma evacuation were advisedly examined in our study. Progressive decrease in muscle power (MP B 3) with leg weakness was the most common symptom for SEH, and it was followed by saddle anesthesia. Previous studies stated that blood loss more than 1,000 ml (major blood loss) was significantly associated with increased risk of postoperative neurological deficits due to spinal epidural hematoma [8, 9]. Awad et al. [8] reviewed the Johns Hopkins series of 14,932 patients undergoing spinal surgery during a period of 18 years, and they reported that blood loss [ 1 L, the involvement of more than five operative levels and intraoperative hemoglobin \10 g/dL were associated with an increased risk of postoperative SEH. Kou et al. [10] organized a case–control retrospective analysis of 12 patients who developed postoperative spinal epidural hematoma and 404 control patients, and they suggested that patients who required multilevel surgical procedures are at a significantly higher risk for this complication. Thus, perianal numbness was more common for patients with shorter time to detection, which indicated that the symptom of saddle anesthesia occurred less often in delayed epidural hematoma in our study. The results of clinical observations after hematoma evacuation pointed out that if the symptom of impaired
muscle power or perianal anesthesia showed no improvement after evacuation, then there was a tendency to have permanent lower extremity weakness after 1-year followup. At 1-year follow-up, there were four patients who had permanent impaired muscle power and they also had neurologic sequelae of urine retention and stool incontinence. Although motor weakness, urinary and stool incontinence are the key signs of cauda equina syndrome caused by SEH, Foley catheter insertion, postoperative wound pain and bedridden status make its detection difficult postoperatively. When SEH was suspected, further examination of voluntary anal contraction by digital anal exam is needed to make an affirmative diagnosis of cauda equina syndrome. The symptoms of spinal epidural hematoma usually occur within 24 h after the initial surgery [13]. Most authors agreed that emergent surgical evacuation should be performed as soon as possible when neurological deterioration is detected and postoperative SEH is suspected. Kebaish et al. [9] concluded that if patients underwent evacuation operation within 6 h of symptom onset of postoperative deficit, then the neurologic outcomes were usually better. In a study of 30 cases of epidural hematoma with neurologic deficit due to various causes such as spinal surgery and anticoagulation medications, Lawton et al. [26] revealed that the rapidity of taking patients to reoperation was correlated with recovery outcomes, and better
123
356
neurological recovery was achieved when surgical decompression was performed within 12 h of symptom onset. In our study, time to evacuation in patients with nonimproved muscle power was longer than in patients with improved muscle power after removing hematoma. The time interval to hematoma evacuation is critical because the earlier the hematoma evacuation was performed, the better the neurological improvements in patients. The MRI study provides excellent clarity for differential diagnosis of epidural compression, including recurrent disc herniation, epidural abscess, durotomy with CSF leakage, and hematoma as well. As the hematoma ages, the oxygenation of hemoglobin also changes into different forms. The formation of hematoma is generally classified into different stages and can be distinguished by different signal intensity on MRI [27]. The MRI features of hypointense signal on T1 and hyperintense signal on T2 weighted images revealed an epidural hemorrhage formation in the hyperacute stage (\24 h-old) due to the intracellular oxyhemoglobin in fresh oxygenated blood (Fig. 2a). In the acute stage (1–3 days old), the presence of intracellular deoxyhemoglobin was represented by isointense on T1 and hypointense on T2 weighted images (Fig. 2b). In the subacute stage, the result of the oxidation of deoxyhemoglobin to methemoglobin increased the signal intensity to hyperintense on T1 weighted image. The chronic stage ([2 week-old) began to covert the extracellular methemoglobin to paramagnetic hemosiderin and ferritin within macrophages and appeared to be hypointense on both T1 and T2 weighted images. We designed a medical algorithm (Fig. 3) for the diagnosis and management of postoperative symptomatic epidural hematoma. After lumbar spine operation, if immediate motor weakness is noted during recovery from anesthesia, intraoperative screw malposition, nerve roots injury, neurapraxia, or vascular compromise is highly suspected. When patients suffered from progressive decrease in muscle power and urinary incontinence within the first 3 days after lumbar spine surgery, SEH should be first considered and immediate surgical decompression is suggested. When the symptoms of intractable pain and/or saddle anesthesia are observed, MRI evaluation is first suggested for a definite diagnosis and then surgical evacuation is indicated if the clinical symptoms and signs of SHE is well supported by the result of image studies. After postoperative day 3, the differential diagnosis of patients with neurologic symptoms includes infection, recurrent disc, cage dislodgement and delayed epidural hematoma, and the exact diagnosis is confirmed by MRI study [28]. Whenever epidural hematoma is diagnosed, an emergent decompression surgery should be performed as early as possible.
123
Eur Spine J (2015) 24:348–357
The limitation of this study is its retrospective design due to the rarity of postoperative spinal epidural hematoma and small sample size. The uncommon occurrence of this complication precludes any other possible study designs, unless a multicenter effort is organized. Due to the retrospective nature of the study, bias and confounders are difficult to control and some essential patient data may not be available because data collection relies on the written records. In addition, our sample size was too small to establish a desired power of 0.80 for all risk factors. In the future, a multicenter study can be performed to fully investigate predisposing risk factors of postsurgical epidural hematoma to improve the postoperative outcomes of patients and even prevent this complication. Conclusions The findings suggest that preoperative diastolic blood pressure, intraoperative use of gelfoam for dura coverage and postoperative drain output are risk factors for symptomatic epidural hematoma after lumbar decompression surgery. Major blood loss and multilevel surgical procedure could result in poor recovery of muscle power. After spine decompression surgery, early detection and evacuation of hematoma is the key to avoid neurologic deterioration and have better clinical outcomes. Conflict of interest
None.
References 1. Jackson R (1869) Case of spinal apoplexy. Lancet 2:5–6 2. Boukobza M, Guichard JP, Boissonet M, George B, Reizine D, Gelbert F, Merland JJ (1994) Spinal epidural haematoma: report of 11 cases and review of the literature. Neuroradiology 36(6):456–459 3. Dickman CA, Shedd SA, Spetzler RF, Shetter AG, Sonntag VK (1990) Spinal epidural hematoma associated with epidural anesthesia: complications of systemic heparinization in patients receiving peripheral vascular thrombolytic therapy. Anesthesiology 72(5):947–950 4. Foo D, Rossier AB (1981) Preoperative neurological status in predicting surgical outcome of spinal epidural hematomas. Surg Neurol 15(5):389–401 5. Gundry CR, Heithoff KB (1993) Epidural hematoma of the lumbar spine: 18 surgically confirmed cases. Radiology 187(2):427–431 6. Ikuta K, Tono O, Tanaka T, Arima J, Nakano S, Sasaki K, Oga M (2006) Evaluation of postoperative spinal epidural hematoma after microendoscopic posterior decompression for lumbar spinal stenosis: a clinical and magnetic resonance imaging study. J Neurosurg Spine 5(5):404–409. doi:10.3171/spi.2006.5.5.404 7. Kotilainen E, Alanen A, Erkintalo M, Helenius H, Valtonen S (1994) Postoperative hematomas after successful lumbar microdiscectomy or percutaneous nucleotomy: a magnetic resonance imaging study. Surg Neurol 41(2):98–105
Eur Spine J (2015) 24:348–357 8. Awad JN, Kebaish KM, Donigan J, Cohen DB, Kostuik JP (2005) Analysis of the risk factors for the development of post-operative spinal epidural haematoma. J Bone Jt Surg Br 87(9):1248–1252. doi:10.1302/0301-620X.87B9.16518 9. Kebaish KM, Awad JN (2004) Spinal epidural hematoma causing acute cauda equina syndrome. Neurosurg Focus 16(6):e1 pii 160601 10. Kou J, Fischgrund J, Biddinger A, Herkowitz H (2002) Risk factors for spinal epidural hematoma after spinal surgery. Spine (Phila Pa 1976) 27(15):1670–1673. doi:10.1097/0007632200208010-00016 11. Fraser S, Roberts L, Murphy E (2009) Cauda equina syndrome: a literature review of its definition and clinical presentation. Arch Phys Med Rehabil 90(11):1964–1968. doi:10.1016/j.apmr.2009. 03.021 12. Sokolowski MJ, Garvey TA, Perl J 2nd, Sokolowski MS, Cho W, Mehbod AA, Dykes DC, Transfeldt EE (2008) Prospective study of postoperative lumbar epidural hematoma: incidence and risk factors. Spine (Phila Pa 1976) 33(1):108–113. doi:10.1097/BRS. 0b013e31815e39af 13. Aono H, Ohwada T, Hosono N, Tobimatsu H, Ariga K, Fuji T, Iwasaki M (2011) Incidence of postoperative symptomatic epidural hematoma in spinal decompression surgery. J Neurosurg Spine 15(2):202–205. doi:10.3171/2011.3.SPINE10716 14. Canale ST, Campbell WC (2003) Campbell’s operative orthopaedics, 10th edn. Mosby, St. Louis 15. Sandhagen B, Frithz G, Waern U, Ronquist G (1990) Increased whole blood viscosity combined with decreased erythrocyte fluidity in untreated patients with essential hypertension. J Intern Med 228(6):623–626 16. Devereux RB, Case DB, Alderman MH, Pickering TG, Chien S, Laragh JH (2000) Possible role of increased blood viscosity in the hemodynamics of systemic hypertension. Am J Cardiol 85(10):1265–1268 (pii S0002-9149(00)00744-X) 17. de Simone G, Devereux RB, Chien S, Alderman MH, Atlas SA, Laragh JH (1990) Relation of blood viscosity to demographic and physiologic variables and to cardiovascular risk factors in apparently normal adults. Circulation 81(1):107–117
357 18. LaRocca H, Macnab I (1974) The laminectomy membrane. Studies in its evolution, characteristics, effects and prophylaxis in dogs. J Bone Jt Surg Br 56B(3):545–550 19. Light RU, Prentice HR (1945) Gelatin sponge; surgical investigation of a new matrix used in conjunction with thrombin in hemostasis. Arch Surg 51:69–77 20. Friedman J, Whitecloud TS 3rd (2001) Lumbar cauda equina syndrome associated with the use of gelfoam: case report. Spine (Phila Pa 1976) 26(20):E485–E487 21. Epstein NE, Silvergleid RS, Hollingsworth R (2009) Increased postoperative cervical myelopathy and cord compression resulting from the use of Gelfoam. Spine J 9(2):e19–e21. doi:10.1016/j. spinee.2008.03.009 22. Herndon JH, Grillo HC, Riseborough EJ, Rich JC Jr (1972) Compression of the brain and spinal cord following use of gelfoam. Arch Surg 104(1):107 23. Alander DH, Stauffer ES (1995) Gelfoam-induced acute quadriparesis after cervical decompression and fusion. Spine 20(8):970–971 24. Knowlson GT (1974) Gel-foam granuloma in the brain. J Neurol Neurosurg Psychiatry 37(8):971–973 25. Mirzai H, Eminoglu M, Orguc S (2006) Are drains useful for lumbar disc surgery? A prospective, randomized clinical study. J Spinal Disord Tech 19(3):171–177. doi:10.1097/01.bsd. 0000190560.20872.a7 26. Lawton MT, Porter RW, Heiserman JE, Jacobowitz R, Sonntag VK, Dickman CA (1995) Surgical management of spinal epidural hematoma: relationship between surgical timing and neurological outcome. J Neurosurg 83(1):1–7. doi:10.3171/jns.1995.83.1.0001 27. Braun P, Kazmi K, Nogues-Melendez P, Mas-Estelles F, ApariciRobles F (2007) MRI findings in spinal subdural and epidural hematomas. Eur J Radiol 64(1):119–125. doi:10.1016/j.ejrad. 2007.02.014 28. Uribe J, Moza K, Jimenez O, Green B, Levi AD (2003) Delayed postoperative spinal epidural hematomas. Spine J 3(2):125–129 (pii S1529943002005351)
123
