http://informahealthcare.com/bij ISSN: 0269-9052 (print), 1362-301X (electronic) Brain Inj, Early Online: 1–10 ! 2015 Informa UK Ltd. DOI: 10.3109/02699052.2015.1004760
Subdural hygroma following decompressive craniectomy or non-decompressive craniectomy in patients with traumatic brain injury: Clinical features and risk factors Qiang Yuan, Xing Wu, Jian Yu, Yirui Sun, Zhiqi Li, Zhuoying Du, Xuehai Wu, Liangfu Zhou, & Jin Hu
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Department of Neurosurgery, Huashan Hospital, Fudan University, Shanghai, PR China
Abstract
Keywords
Objective: Subdural hygroma (SDG) is a common complication that can occur after head trauma or secondary to decompressive craniectomy (DC). SDGs can be located not only ipsilateral or contralateral to the side of the DC, but also bilateral or unilateral in patients without DC. This study investigated the incidence and risk factors for different types of SDG in a large cohort of patients with traumatic brain injury (TBI). Methods: A retrospective study was conducted involving 379 adult patients with TBI who were admitted to Huashan Hospital, Fudan University between January 2009 and December 2013. As the outcome was dichotomous (SDG vs no SDG or hydrocephalus vs no hydrocephalus), multivariate logistic regression analyses were used to identify independent risk factors for the development of SDGs in patients without DC, ipsilateral SDG after unilateral DC, contralateral SDG after unilateral DC or SDG after bilateral DC. Risk factors for the development of hydrocephalus were also evaluated in patients with and without DC. Results: Among the 207 (54.6%) patients without DC, 30 (14.5%) had unilateral SDGs and 34 (16.4%) had bilateral SDGs. Of the 172 patients (45.4%) with DC, 134 (77.9%) underwent unilateral DC and 38 (22.1%) underwent bilateral DC. Of the 134 patients who underwent unilateral DC, 49 developed SDG, including 22 (16.4%) ipsilateral SDG, 19 (14.2%) contralateral SDG and eight (6.0%) both ipsilateral and contralateral SDGs. For patients undergoing bilateral DC, 13 (34.2%) developed a SDG. No significant difference in the incidence of SDG was observed between the patients with and without DC (36.0% vs 30.9%, p ¼ 0.291), but the characteristics of SDGs were different between the two groups. Logistic regression analysis showed that factors independently associated with the development of SDG were male sex (odds ratio [OR] ¼ 3.861; 95% CI ¼ 1.642–9.091; p ¼ 0.002), older age (OR ¼ 1.046; 95% CI ¼ 1.021–1.070; p50.001), basal cistern haemorrhage (OR ¼ 4.608; 95% CI ¼ 1.510–14.064; p ¼ 0.007), diffuse injury and swelling (OR ¼ 3.158; 95% CI ¼ 1.341–7.435; p ¼ 0.008) or diffuse injury and shift (OR ¼ 3.826; 95% CI ¼ 1.141–12.830; p ¼ 0.030) in patients without DC. Temporal haematoma or contusion in the non-DC side (OR ¼ 2.623; 95% CI ¼ 1.070–6.428; p ¼ 0.035) and traumatic SAH (OR ¼ 3.751; 95% CI ¼ 1.047–13.438; p ¼ 0.042) were independently associated with the development of ipsilateral SDG in patients who underwent unilateral DC. However, factors independently associated with the development of contralateral SDG were frontal haematoma or contusion on the non-DC side (OR ¼ 3.145; 95% CI ¼ 1.272–7.774; p ¼ 0.013) and SDH on the non-DC side (OR ¼ 7.024; 95% CI ¼ 1.477–33.390; p ¼ 0.014). Only craniectomy area (OR ¼ 1.030; 95% CI ¼ 1.008–1.052; p ¼ 0.008) was independently associated with the development of SDG in patients with bilateral DC. In the multivariate analysis, SDG in patients without DC was not associated with the development of hydrocephalus. However, SDG was significantly associated with the development of hydrocephalus for patients who underwent DC (OR ¼ 2.173; 95% CI ¼ 1.362–3.467; p ¼ 0.001). Conclusions: This study suggested that the incidence of SDG in patients who have and have not undergone DC was identical; however, the patients’ characteristics and risk factors differed. Therefore, the management and prediction of SDG should be performed according to SDG type.
Decompressive craniectomy, post-traumatic hydrocephalus, risk factor, subdural hygroma, traumatic brain injury History Received 26 August 2014 Revised 17 November 2014 Accepted 4 January 2015 Published online 23 April 2015
Introduction
Correspondence: Jin Hu, MD, PhD, Department of Neurosurgery, Huashan Hospital, Fudan University, 12 Wulumuqi Zhong Road. Shanghai 200040, PR China. Tel: +86-21-52887736; +8613003116904. Email: [email protected]
Traumatic brain injury (TBI) is an important cause of injuryrelated hospitalization, disability and death worldwide [1]. It represents a significant public health problem in China and across the world. It is estimated that an average of 1.4 million TBIs occur each year in the US, including 1.365 million emergency department visits, 275 000 hospitalizations and 52
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
2
Q. Yuan et al.
000 deaths [2, 3]. A large portion of the complications are secondary to head trauma or craniotomy, such as subdural hygroma (SDG) and post-traumatic hydrocephalus, which may further aggravate the disease, lead to poor prognosis and increase the burden of TBI. The prevention and management of complications are crucial parts of managing TBI [4]. SDG is not a rare complication that can occur after head trauma or secondary to decompressive craniectomy (DC). The incidence of SDG after head trauma is 7–12% [5]. However, the incidence increases to 23–56% of patients with a head injury if a DC is performed [6–9]. The occurrence of SDG may be due to a change in the dynamics of CSF circulation or an underlying disorder of normal CSF absorption. Moreover, SDG may be a complication of removing a bone flap, which results in formation of space in which fluid can accumulate [9–11]. SDG can be seen in the first week after injury and fluid collection may increase for up to 4 weeks [12]; most cases resolve gradually without surgical management. However, some SDGs may require surgical intervention to control the further neurological deterioration. The exact pathogenesis of SDG after TBI remains unclear and the risk factors are rarely reported or reported only for patients who undergo DC. SDGs can be located not only ipsilateral or contralateral to the side of the DC, but also bilateral or unilateral in patients without DC. It is unclear whether different types of SDG have an identical clinical course and risk factors. De Bonis et al. [13] reported that inter-hemispheric hygroma and hydrocephalus is a timeline of events for patients with TBI after DC. In addition, many studies have shown that SDG is a significant risk factor for hydrocephalus after DC [14, 15]. However, these studies focused only on craniectomy patients; therefore, whether patients without DC with TBI have identical findings remains unknown. Several factors have been associated with the development of post-traumatic hydrocephalus [16], such as older age, subarachnoid haemorrhage, CSF infection, lower Glasgow Coma Scale (GCS) score and a wide craniectomic flap [15, 17, 18]. De Bonis et al. [13] reported that craniectomy with a superior limit closer to the midline than 25 mm might pre-dispose a patient to development of hydrocephalus. It remains unclear whether these risk factors associated with post-traumatic hydrocephalus are consistent with the development of post-traumatic SDG. Therefore, it is hypothesized that, regardless of whether DC and SDG are common after TBI, the risk factors for this complication are related primarily to the severity of the primary brain injury. This study investigated the incidence and risk factors for different types of SDG in a large cohort of patients with TBI.
Methods Patients A retrospective study was conducted involving 379 adult patients with TBI who were admitted to Huashan Hospital, Fudan University between January 2009 and December 2013. This research protocol was approved by the University Hospital Medical Ethics Board. All participants provide their written informed consent at the beginning of the study and the consent procedure was approved by the University Hospital
Brain Inj, Early Online: 1–10
Medical Ethics Board. The next of kin, carers or guardians consented on the behalf of participants whose capacity to consent was compromised. The inclusion criteria were as follows: TBI with radiological signs of intracranial brain injury [epidural or subdural haematoma, intracerebral haematoma or contusion or subarachnoid haemorrhage (SAH)] documented by computed tomography (CT) scan; age 418 years; and admission within 24 hours after TBI. Patients who died in the acute phase (57 days after admission) were excluded. Among the 379 patients, 172 received DC to allow for brain swelling and 207 did not. The indications for DC were: (1) following evacuation of a mass lesion when the brain was excessively swollen; and (2) intracranial pressure (ICP) that increased to 425 mm Hg and/or reduced cerebral perfusion pressure to 560 mm Hg for longer than 30 minutes. The distance of the craniectomy from the midline and the area of the craniectomy were determined in patients who underwent DC. The incidence, clinical presentations and outcomes of SDGs related to TBI were analysed in both the DC and nonDC groups. Clinical and demographic characteristics, including age, sex, mechanism of injury, injury severity score (ISS), pupillary reaction to light, ICP monitor and GCS score at admission were recorded in all patients. Moreover, the characteristics of the initial CT scan on admission were used to assess the CT findings and Marshall CT grade. All available serial CT scans from admission to discharge were evaluated to identify SDG and subsequent hydrocephalus. There is no established definition of the minimum thickness or volume of SDGs; thus, they are defined as a low-density collection 40.5 cm in maximal depth measured from the cortical surface to the inner aspect of the scalp or skull. Posttraumatic hydrocephalus was defined by radiological evidence of progressive ventricular dilatation (Evans index40.3) with trans-ependymal oedema, together with the presence of either clinical deterioration or failure to make neurological progress over time and some evidence of clinical improvement after insertion of a ventriculo-peritoneal shunt [15]. All patients received primary care in the emergency room, as recommended by Advanced Trauma Life Support specifications. All patients were treated in accordance with the 2007 guidelines of the Brain Trauma Foundation: use of CSF drainage via ventriculostomy, mannitol treatment, hyperventilation for intracranial hypertension if needed and normovolemia maintenance with saline by central venous pressure (CVP) of 5–8 mm Hg, maintenance of CPP 460 mm Hg and ICP 520 mm Hg [19]. The neurological outcomes were determined according to the Glasgow Outcome Score (GOS) as follows: 1 ¼ dead; 2 ¼ vegetative state with an inability to interact with the environment; 3 ¼ severe disability with an inability to live independently but the ability to follow commands; 4 ¼ moderate disability with the ability to live independently but an inability to return to work or school; and 5 ¼ good recovery with the ability to return to work or school. Prognostic evaluations were determined using the GOS assessment at 3 months after the trauma. GOS evaluations were performed by physicians either in person or via telephone. A GOS of 1–3 was categorized as an unfavourable outcome, whereas a score of 4–5 was deemed a favourable outcome.
DOI: 10.3109/02699052.2015.1004760
Statistical analyses All data were analysed using the SPSS statistical software v.16 (SPSS Inc., Chicago, IL). Descriptive statistics were used to describe the incidence and other characteristics of the patients. The unpaired Student’s t-test or a non-parametric test was used to compare continuous variables and the chi-square or Fisher’s exact test to evaluate categorical variables. As the outcome was dichotomous (SDG vs no SDG or hydrocephalus vs no hydrocephalus), multivariate logistic regression analyses were used to identify independent risk factors for the development of SDGs in patients without DC, ipsilateral SDG after unilateral DC, contralateral SDG after unilateral DC or SDG after bilateral DC. Risk factors for the development of hydrocephalus were also evaluated in patients with and without DC. A p-value 50.05 was considered to indicate significance.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Results Incidence and characteristics of SDGs The study population consisted of 379 patients (287 males, 92 females; mean age ¼ 50.0 ± 15.7 years) (Figure 1). Of the initial cohort of 379 patients, 172 underwent DC to allow for brain swelling and 207 did not. Among the 207 (54.6%) patients without DC, 30 (14.5%) had unilateral SDGs (Figures 2(a) and (b)) and 34 (16.4%) had bilateral SDGs (Figures 2(c) and (d)). Of the 172 patients (45.4%) who underwent DC, 134 (77.9%) underwent unilateral DC and 38 (22.1%) underwent bilateral DC. Of the 134 patients who underwent unilateral DC, 49 developed SDG, including 22 (16.4%) ipsilateral SDG (Figures 2(g) and (h)), 19 (14.2%) contralateral SDG (Figures 2(e) and (f)) and eight (6.0%) both ipsilateral and contralateral SDGs (Figure 2(i)). For patients undergoing bilateral DC, 13 (34.2%) developed a SDG. No significant
Subdural hygroma in traumatic brain injury
3
difference in the incidence of SDG was observed between the patients with and without DC (36.0% vs 30.9%, p ¼ 0.291). The width of the SDG in the patients with DC was 1.20 ± 0.65 cm, which was significantly larger than those in patients without DC (0.76 ± 0.33 cm) (p50.001). SDGs in the no-DC group developed at 4.58 ± 4.06 days after injury, which was significantly earlier than the DC group (7.42 ± 6.95 days) (p50.001). In addition, compared with the patients who developed SDG without DC, the patients who developed SDG after DC were more likely to be younger, have a lower GCS score on admission, have one pupil react to light, have an SAH, an evacuated mass lesion, an ICP monitor, a SDH, an epidural haemorrhage (EDH), a frontal haematoma or contusion and worse neurological outcomes at 3 months. SDGs resolved gradually with conservative management in most of the patients with TBI who did not undergo DC (59/64). However, SDGs in the remaining five patients contributed to progressive clinical deterioration and burr hole drainage was performed. In the 57 patients who underwent DC, the SDG resolved gradually with conservative management and only five patients had progressive clinical deterioration and burr hole drainage performed. SDG for patients without DC Patients without DC developing a unilateral or bilateral SDG were more likely to be male, older, have a diffuse injury and swelling, a basal cistern haemorrhage, a temporal haematoma or contusion on brain CT scan and worse neurological outcomes at 3 months (Table I). In addition, patients who developed SDG were less likely to have an EDH. No differences in the mechanism of injury, intubation in the emergency room, GCS at admission, ISS, pupillary reaction
Figure 1. Flowchart showing different types of subdural hygroma in patients with and without decompressive craniectomy.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
4
Q. Yuan et al.
Brain Inj, Early Online: 1–10
Figure 2. CT scan of different types of subdural hygroma. (a, b) Unilateral subdural hygroma in patients without DC; (c, d) bilateral subdural hygroma in patients without DC; (e, f) contralateral subdural hygroma after DC; (g, h) ipsilateral subdural hygroma after DC; (i) bilateral subdural hygroma after DC. DC, decompressive craniectomy.
to light, ICP monitor, other CT findings or mortality were observed between patients with and without SDG. Logistic regression analysis showed that factors independently associated with the development of SDG were male sex (odds ratio [OR] ¼ 3.861; 95% CI ¼ 1.642–9.091; p ¼ 0.002), older age (OR ¼ 1.046; 95% CI ¼ 1.021–1.070; p50.001), basal cistern haemorrhage (OR ¼ 4.608; 95% CI ¼ 1.510–14.064; p ¼ 0.007), diffuse injury and swelling (OR ¼ 3.158; 95% CI ¼ 1.341–7.435; p ¼ 0.008) and diffuse injury and shift (OR ¼ 3.826; 95% CI ¼ 1.141–12.830; p ¼ 0.030) (Table II). SDG in patients after DC For patients with unilateral DC, the occurrence of ipsilateral SDG was associated with a temporal haematoma or contusion on the non-DC side, a traumatic SAH on the brain CT scan and worse neurological outcomes at 3 months (Table III). However, patients who developed contralateral SDG following unilateral DC were more likely to have an ICP monitor, a SDH on the non-DC side, a frontal haematoma or contusion on the non-DC side and a temporal haematoma or contusion on the non-DC side (Table IV). The occurrence of SDG was significantly associated with a lower GCS score on admission,
non-reactive pupils, a larger craniectomy area and worse neurological outcomes at 3 months in patients with bilateral DC (Table V). No differences were observed among other variables. In the multivariate logistic regression analysis, temporal haematoma or contusion in the non-DC side (OR ¼ 2.623; 95% CI ¼ 1.070–6.428; p ¼ 0.035) and traumatic SAH (OR ¼ 3.751; 95% CI ¼ 1.047–13.438; p ¼ 0.042) were independently associated with an ipsilateral SDG in patients who underwent unilateral DC (Table II). However, factors independently associated with the development of contralateral SDG were frontal haematoma or contusion on the non-DC side (OR ¼ 3.145; 95% CI ¼ 1.272–7.774; p ¼ 0.013) and SDH on the non-DC side (OR ¼ 7.024; 95% CI ¼ 1.477–33.390; p ¼ 0.014) (Table II). In addition, the logistic regression analysis showed that only craniectomy area (OR ¼ 1.030; 95% CI ¼ 1.008–1.052; p ¼ 0.008) was independently associated with an SDG in patients following bilateral DC (Table II). Relationship between SDG and hydrocephalus For patients without DC, subsequent hydrocephalus was present in six of 64 patients (9.4%) who developed a unilateral
Subdural hygroma in traumatic brain injury
DOI: 10.3109/02699052.2015.1004760
5
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Table I. Difference in characteristics between patients who developed subdural hygroma and those who did not for patients without DC. Variable
W SDG (%)
W/o SDG (%)
No. of patients Male sex Mean age (years) (mean ± SD) Mechanism of injury Motor vehicle accident Fall from a height Stumble and fall Blow to head Injured by a crashing object Others Intubation in the emergency room GCS at admission (mean ± SD) GCS 3–8 GCS 9–15 ISS (mean ± SD) Pupils react to light Both One None Injury severity (Marshall CT grade) II (Diffuse injury without swelling) III (Diffuse injury and swelling) IV (Diffuse injury and shift) V (Evacuated mass lesion) VI (Non-evacuated mass lesion 425 cc) ICP monitor CT findings EDH SDH Frontal haematoma or contusion Temporal haematoma or contusion tSAH Basal cistern haemorrhage Outcome Unfavourable Favourable Survived Died
64 54 (84.4) 57.5 ± 14.5
143 97 (67.8) 48.7 ± 16.6
39 (60.9) 7 (10.9) 15 (23.4) 1 (1.6) 0 (0.0) 2 (3.1) 9 (14.1) 10.1 ± 3.1 21 (32.8) 43 (67.2) 20.6 ± 6.6
78 (54.5) 16 (11.2) 27 (18.9) 5 (3.5) 8 (5.6) 9 (6.3) 14 (9.8) 10.9 ± 3.3 37 (25.9) 106 (74.1) 20.7 ± 7.2
60 (93.8) 4 (6.2) 0 (0.0)
139 (97.2) 4 (2.8) 0 (0.0)
p Value 0.013 50.001 0.454
0.366 0.120 0.304 0.873 0.423
0.046 12 30 7 7 8 44
(18.8) (46.9) (10.9) (10.9) (12.5) (68.8)
56 45 12 18 12 83
(39.2) (31.5) (8.4) (12.6) (8.4) (58.0)
0.144
8 23 38 45 38 13
(12.5) (35.9) (59.4) (70.3) (59.4) (20.3)
36 39 80 74 65 8
(25.2) (27.3) (55.9) (51.7) (45.5) (5.6)
0.039 0.208 0.645 0.013 0.064 0.001
36 28 62 2
(56.2) (43.8) (96.9) (3.1)
48 95 141 2
(33.6) (66.4) (98.6) (1.4)
0.002 0.774
DC, decompressive craniectomy; SDG, subdural hygroma; GCS, Glasgow coma scale; ISS, injury severity score; ICP, intracranial pressure; EDH, epidural haematoma; SDH, subdural haematoma; tSAH, traumatic subarachnoid haemorrhage. Table II. Multivariate analysis of the risk factors for developing different types of subdural hygroma. Variable Subdural hygroma w/o decompressive craniectomy Male Age Basal cistern haemorrhage Marshall CT grade(reference: II, Diffuse injury without swelling) III (Diffuse injury and swelling) IV (Diffuse injury and shift) Contralateral subdural hygroma following decompressive craniectomy Frontal haematoma or contusion in the non-DC side SDH in the non-DC side Ipsilateral subdural hygroma following decompressive craniectomy Temporal hematoma or contusion in the non-DC side tSAH Subdural hygroma following bilateral decompressive craniectomy Cranial defect size
or bilateral SDG and in six of 143 patients (4.2%) without SDG (p ¼ 0.249). However, 16 of 62 patients (25.8%) with SDG developed subsequent hydrocephalus following DC, which was significantly higher than those without SDG
Adjusted OR (95% CI)
p Value
3.861 (1.642–9.091) 1.046 (1.021–1.070) 4.608 (1.510–14.064)
0.002 50.001 0.007
3.158 (1.341–7.435) 3.826 (1.141–12.830)
0.008 0.030
3.145 (1.272–7.774) 7.024 (1.477–33.390)
0.013 0.014
2.623 (1.070–6.428) 3.751 (1.047–13.438)
0.035 0.042
1.030 (1.008–1.052)
0.008
(10.0%) (p ¼ 0.006). After adjusting for other factors (age, sex, GCS at admission, pupillary reaction to light and CT findings) in the multivariate analysis, SDG in the patients without DC was not associated with the development of
6
Q. Yuan et al.
Brain Inj, Early Online: 1–10
Table III. Difference in characteristics between patients who developed ipsilateral subdural hygroma following unilateral DC and those who did not.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Variable No. of patients Male sex Mean age (years) (mean ± SD) Mechanism of injury Motor vehicle accident Fall from a height Stumble and fall Blow to head Injured by a crashing object Others Intubation in the emergency room GCS at admission (mean ± SD) GCS 3–8 GCS 9–15 ISS (mean ± SD) Pupils react to light Both One None Injury severity (Marshall CT grade) II (Diffuse injury without swelling) III (Diffuse injury and swelling) IV (Diffuse injury and shift) V (Evacuated mass lesion) ICP monitor CT findings EDH in the non-DC side SDH in the non-DC side Frontal haematoma or contusion in the non-DC side Temporal haematoma or contusion in the non-DC side EDH in the DC side SDH in the DC side Frontal haematoma or contusion in the DC side Temporal haematoma or contusion in the DC side tSAH Basal cistern haemorrhage The area of the craniectomy (cm2) (mean ± SD) The distance of the craniectomy from the midline (cm) (mean ± SD) Outcome Unfavourable Favourable Survived Died
W ipsilateral SDG (%)
W/o ipsilateral SDG (%)
30 25 (83.3) 46.9 ± 16.5
104 79 (76.0) 48.8 ± 14.2
16 (53.3) 6 (20.0) 6 (20.0) 0 (0.0) 0 (0.0) 2 (6.7) 4 (13.3) 6.8 ± 2.3 24 (80.0) 6 (20.0) 25.1 ± 7.0
63 (60.6) 19 (18.3) 11 (10.6) 3 (2.9) 4 (3.8) 4 (3.8) 12 (11.5) 7.4 ± 2.6 74 (71.2) 30 (28.8) 23.2 ± 6.2
14 (46.7) 14 (46.7) 2 (6.7)
54 (51.9) 40 (38.5) 10 (9.6)
0 4 5 21 26
2 9 25 68 90
p Value 0.393 0.538 0.583
1.000 0.328 0.336 0.171 0.741
0.747 (0.0) (13.3) (16.7) (70.0) (86.7)
2 (6.7) 2 (6.7) 11 (36.7) 12 (40.0) 6 (20.0) 18 (60.0) 21 (70.0) 23 (76.7) 27 (90.0) 0 (0.0) 103.2 ± 28.1 2.0 ± 1.4 25 5 27 3
(83.3) (16.7) (90.0) (10.0)
(1.9) (8.7) (24.0) (65.4) (86.5)
1.000
8 (7.7) 6 (5.8) 29 (27.9) 20 (19.2) 21 (20.2) 54 (51.9) 65 (62.5) 67 (64.4) 72 (69.2) 2 (1.9) 97.7 ± 38.0 2.3 ± 1.2
1.000 1.000 0.354 0.019 0.982 0.434 0.450 0.208 0.023 1.000 0.462 0.194
50 54 97 7
(48.1) (51.9) (93.3) (6.7)
0.001 0.837
DC, decompressive craniectomy; SDG, subdural hygroma; GCS, Glasgow coma scale; ISS, injury severity score; ICP, intracranial pressure; EDH, epidural haematoma; SDH, subdural haematoma; tSAH, traumatic subarachnoid haemorrhage.
hydrocephalus. However, SDG was significantly associated with the development of hydrocephalus in patients with DC (OR ¼ 2.173; 95% CI ¼ 1.362–3.467; p ¼ 0.001).
Discussion This was the first study to investigate the incidence and risk factors for different types of SDG in a large cohort of patients with TBI. No significant difference was found in the incidence of SDG between patients who did and did not undergo DC, but the patients between the two groups had different characteristics. The patients who underwent DC and subsequently developed SDG were more likely to have greater thickness, later onset time, younger age, lower GCS score on admission, one pupil reacting to light, a SAH, an evacuated mass lesion, a ICP monitor, a SDH, a EDH, a frontal haematoma or contusion and worse neurological outcomes at 3 months than those who did not undergo DC but developed
SDG. Therefore, the patients who underwent DC and subsequently developed SDG may have had more severe primary brain injury and more severe representation of SDG than those who did not undergo DC but developed SDG. Honeybul et al. [20] have reported the association between injury severity and the development of SDG in the patients with DC. However, the comparison between the patients with and without DC is still lacking. SDG is a well-known complication of TBI. The incidence of SDG after TBI in patients who have not undergone DC is 5–21% in several clinical series. Ohno et al. [21] reported 43 SDGs (6%) in 715 patients and the incidence of SDG in the Kaufman et al. [22] study was 38 (4%) of 881 patients with a closed-head injury. Born et al. [23] described 16 patients (15%) with SDGs among 109 with blunt TBI. However, this study found that SDGs developed in 64 (30.9%) of 207 patients who had not undergone DC, which was higher than previous reports. The reported incidence appears to increase
Subdural hygroma in traumatic brain injury
DOI: 10.3109/02699052.2015.1004760
7
Table IV. Difference in characteristics between patients who developed contralateral subdural hygroma following unilateral DC and those who did not.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Variable No. of patients Male sex Mean age (years) (mean ± SD) Mechanism of injury Motor vehicle accident Fall from a height Stumble and fall Blow to head Injured by a crashing object Others Intubation in the emergency room GCS at admission (mean ± SD) GCS 3–8 GCS 9–15 ISS (mean ± SD) Pupils react to light Both One None Injury severity (Marshall CT grade) II (Diffuse injury without swelling) III (Diffuse injury and swelling) IV (Diffuse injury and shift) V (Evacuated mass lesion) ICP monitor CT findings EDH in the non-DC side SDH in the non-DC side Frontal haematoma or contusion in the non-DC side Temporal haematoma or contusion in the non-DC side EDH in the DC side SDH in the DC side Frontal haematoma or contusion in the DC side Temporal haematoma or contusion in the DC side tSAH Basal cistern haemorrhage The area of the craniectomy (cm2) (mean ± SD) The distance of the craniectomy from the midline (cm) (mean ± SD) Outcome Unfavourable Favourable Survived Died
W contralateral SDG (%) W/o contralateral SDG (%) p Value 27 22 (81.5) 52.1 ± 15.0
107 82 (76.6) 47.4 ± 14.5
14 (51.9) 4 (14.8) 3 (11.1) 2 (7.4) 1 (3.7) 3 (11.1) 3 (11.1) 6.8 ± 2.1 22 (81.5) 5 (18.5) 23.5 ± 4.5
65 (60.7) 21 (19.6) 14 (13.1) 1 (0.9) 3 (2.8) 3 (2.8) 13 (12.1) 7.4 ± 2.7 76 (71.0) 31 (29.0) 23.7 ± 6.8
15 (55.6) 10 (37.0) 2 (7.4)
53 (49.5) 44 (41.1) 10 (9.3)
1 5 4 17 27
1 8 26 72 89
0.589 0.135 0.139
1.000 0.339 0.274 0.847 0.900
0.149 (3.7) (18.5) (14.8) (63.0) (100.0)
0 (0.0) 5 (18.5) 14 (51.9) 11 (40.7) 7 (25.9) 14 (51.9) 18 (66.7) 18 (66.7) 21 (77.8) 0 (0.0) 99.4 ± 25.7 2.4 ± 1.4 17 10 25 2
(63.0) (37.0) (92.6) (7.4)
(0.9) (7.5) (24.3) (67.3) (83.2)
0.048
10 (9.3) 3 (2.8) 26 (24.3) 21 (19.6) 20 (18.7) 58 (54.2) 68 (63.6) 72 (67.3) 78 (72.9) 2 (1.9) 98.8 ± 38.2 2.2 ± 1.3
0.214 0.009 0.005 0.021 0.402 0.827 0.763 0.951 0.606 1.000 0.937 0.581
58 49 99 8
(54.2) (45.8) (92.5) (7.5)
0.413 1.000
DC, decompressive craniectomy; SDG, subdural hygroma; GCS, Glasgow coma scale; ISS, injury severity score; ICP, intracranial pressure; EDH, epidural haematoma; SDH, subdural haematoma; tSAH, traumatic subarachnoid haemorrhage.
following DC, with reports ranging from 26–60%. Yang et al. [9] documented 23 cases (21.3%) of SDGs in 108 patients who had undergone DC for severe head injury. Aarabi et al. [12] reported that SDGs developed in 39 (57.4%) of 68 patients who survived41 month after TBI and DC. This study found that SDGs developed in 62 (36.0%) of 172 patients who underwent DC. In a 2009 study, Yang et al. [24] reported 11 cases (6.5%) of contralateral SDG among 169 patients who had undergone DC for severe head injury. Wang et al. [25] reported that the overall incidence of contralateral subdural effusion was 7.3%. In this study, of the 134 patients who underwent unilateral DC, 30 (22.4%) had ipsilateral SDG and 27 (20.1%) had contralateral SDG, which is also higher than previous reports. It was also found that the incidence of SDG after TBI in patients who had undergone DC was not significantly higher than that in patients who had not undergone DC (36.0% vs 30.9%). The main reason is likely use of different inclusion and diagnostic criteria in the
previous study; however, uniform inclusion and diagnostic criteria were used in this work. Jeon et al. [26] reported that radiological factors indicate that a midline shift 45 mm, SAH, delayed hydrocephalus, compression of basal cisterns and tearing of the arachnoid membrane may be risk factors associated with SDG after DC. However, in this study, the risk factors for SDG in patients without DC, ipsilateral SDG after unilateral DC, contralateral SDG after unilateral DC and SDG after bilateral DC varied. In this study, temporal haematoma or contusion on the nonDC side and traumatic SAH were independently associated with the development of ipsilateral SDG in patients who underwent unilateral DC. Frontal haematoma or contusion on the non-DC side and SDH on the non-DC side were associated with contralateral SDG. A basal cistern haemorrhage was associated with SDG in patients without DC. All of these results indicate that tearing of the arachnoid membrane may be a risk factor associated with SDG. Previous studies
8
Q. Yuan et al.
Brain Inj, Early Online: 1–10
Table V. Difference in characteristics between patients who developed subdural hygroma following bilateral DC and those who did not.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Variable No. of patients Male sex Mean age (years) (mean ± SD) Mechanism of injury Motor vehicle accident Fall from a height (3 m) Stumble and fall Blow to head Injured by a crashing object Others Intubation in the emergency room GCS at admission (mean ± SD) GCS 3–8 GCS 9–15 ISS (mean ± SD) Pupils react to light Both One None Injury severity (Marshall CT grade) II (Diffuse injury without swelling) III (Diffuse injury and swelling) IV (Diffuse injury and shift) V (Evacuated mass lesion) ICP monitor CT findings EDH SDH Frontal haematoma or contusion Temporal haematoma or contusion tSAH Basal cistern haemorrhage The area of the craniectomy (cm) (mean ± SD) Outcome Unfavourable Favourable Survived Died
W/SDG (%)
W/o SDG (%)
13 12 (92.3) 51.8 ± 17.3
25 20 (80.0) 45.8 ± 13.3
10 (76.9) 2 (15.4) 0 (0.0) 0 (0.0) 0 (0.0) 1 (7.7) 2 (15.4) 6.5 ± 2.1 10 (76.9) 3 (23.1) 24.9 ± 9.9
15 (60.0) 3 (12.0) 3 (12.0) 0 (0.0) 1 (4.0) 3 (12.0) 1 (4.0) 8.8 ± 3.1 10 (40.0) 15 (60.0) 20.8 ± 7.3
5 (38.5) 8 (61.5) 0 (0.0)
18 (72.0) 4 (16.0) 3 (12.0)
p Value 0.604 0.240 0.849
0.548 0.023 0.031 0.156 0.015
0.261 0 2 0 11 13
(0.0) (15.4) (0.0) (84.6) (100.0)
5 (38.5) 7 (53.8) 11 (84.6) 11 (84.6) 13 (100.0) 0 (0.0) 151.5 ± 63.4 12 1 12 1
(92.3) (7.7) (92.3) (7.7)
1 10 1 13 23
(4.0) (40.0) (4.0) (52.0) (92.0)
0.538
5 (20.0) 7 (28.0) 24 (96.0) 14 (56.0) 18 (72.0) 2 (8.0) 94.0 ± 27.9
0.402 0.225 0.548 0.160 0.095 0.538 0.007
12 13 23 2
(48.0) (52.0) (92.0) (8.0)
0.020 1.000
DC, decompressive craniectomy; SDG, subdural hygroma; GCS, Glasgow coma scale; ISS, injury severity score; ICP, intracranial pressure; EDH, epidural haematoma; SDH, subdural haematoma; tSAH, traumatic subarachnoid haemorrhage.
have suggested that SDG and hydrocephalus following DC may be due to a reduction in pulsatile CSF flow and venous outflow when the medial margin of a unilateral craniotomy is close to the midline [13]. However, this study does not confirm these hypotheses. A number of mechanisms could be responsible for the development of SDG. First, the formation of a one-way flap valve due to the tearing or disruption of the arachnoid may result in CSF leakage and accumulation in the subdural space [24, 27]. Second, absorption disturbance of the CSF circulation would also increase the risk of accumulation of the effusion through a torn arachnoid [28]. The close correlation between SAH on the initial CT scan and subsequent hydrocephalus may explain the underlying disturbance of CSF circulation after TBI that promoted the formation of the SDG in some of the patients. Third, when there initially was a possible rupture in the arachnoid layer after TBI, the rapid decrease in ICP resulting from external ventricular drainage would form a pressure gradient and result in enlargement of the subdural space and accumulation of SDG [29, 30]. Fourth, a pressure gradient also could be formed when intraoperative tissue retracts and is not able to reshape, thus, the subsequent
brain shift plays an important role in development of the SDG [31]. However, SDGs in patients who have undergone DC have different characteristics from those who have not undergone DC. Therefore, other mechanisms may be responsible for the development of SDGs. When the effusion is not large or when there is no mass effect or deterioration in clinical presentation, conservative treatment is recommended for the treatment of an SDG. However, surgery should be performed when there is clinical evidence of midline shift, neurological deterioration or the enlargement of effusion. The type of operation includes burr hole drainage or placement of a subduropleural or subduroperitoneal shunt [32]. Drainage using a burr hole is considered a simple and effective procedure for subdural effusion. In this series, all patients who needed surgery received a burr hole. Honeybul et al. [15] found that SDG is a significant risk factor for hydrocephalus after DC, which is consistent with these findings. However, this study also found that SDG in patients without DC was not associated with the development of hydrocephalus. Therefore, DC combined with SDG may be an independent risk factor for hydrocephalus.
DOI: 10.3109/02699052.2015.1004760
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
It has been shown that different types of SDG in patients with TBI were associated with a worse long-term neurological outcome, as has been reported previously [15]. Although this association could be a result from residual confounding because of their association with the severity of the primary brain injury, it is possible that SDGs exert considerable pressure on the brain and, thus, reduce neurological recovery. There were several limitations to this study. First, although it included a reasonable number of patients with TBI, relatively few had SDG and hydrocephalus, which limited the power to identify multiple risk factors for SDG and hydrocephalus. Second, this was a retrospective study that relied on the accuracy of chart records or recall of individuals. Therefore, the findings should be subjected to more rigorous testing in a further well-designed, prospective study.
Conclusion The incidence of SDG in patients who have and have not undergone DC was identical; however, the patients’ characteristics and risk factors differed. SDG was an independent risk factor for hydrocephalus only in patients who had undergone DC. Therefore, the management and prediction of SDG should be performed according to SDG type.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. This work was supported by the National Natural Science Foundation of China (NSFC Grants 30371454, 81271375 and 81171133), The Science and Technology Commission of Shanghai Municipality Project (10JC1402300) and Shanghai Nature Science Foundation (08411952000).
References 1. Summers CR, Ivins B, Schwab KA. Traumatic brain injury in the United States: An epidemiologic overview. Mount Sinai Journal of Medicine 2009;76:105–110. 2. CDC grand rounds: Reducing severe traumatic brain injury in the United States. MMWR Morbidity & Mortality Weekly Report 2013;62:549–552. 3. Pearson WS, Sugerman DE, McGuire LC, Coronado VG. Emergency department visits for traumatic brain injury in older adults in the United States: 2006-08. Western Journal of Emergency Medicine 2012;13:289–293. 4. Honeybul S, Ho KM. Long-term complications of decompressive craniectomy for head injury. Journal of Neurotrauma 2011;28: 929–935. 5. Stone JL, Lang RG, Sugar O, Moody RA. Traumatic subdural hygroma. Neurosurgery 1981;8:542–550. 6. Yang XJ, Hong GL, Su SB, Yang SY. Complications induced by decompressive craniectomies after traumatic brain injury. Chinese Journal of Traumatology 2003;6:99–103. 7. Huang Q, Dai WM, Wu TH, Jie YQ, Yu GF, Fan XF. Comparison of standard large trauma craniotomy with routine craniotormy in treatment of acute subdural hematoma. Chinese Journal of Traumatology 2003;6:305–308. 8. Jiang JY, Xu W, Li WP, Xu WH, Zhang J, Bao YH, Ying YH, Luo QZ. Efficacy of standard trauma craniectomy for refractory intracranial hypertension with severe traumatic brain injury: A multicenter, prospective, randomized controlled study. Journal of Neurotrauma 2005;22:623–628.
Subdural hygroma in traumatic brain injury
9
9. Yang XF, Wen L, Shen F, Li G, Lou R, Liu WG, Zhan RY. Surgical complications secondary to decompressive craniectomy in patients with a head injury: A series of 108 consecutive cases. Acta Neurochirurgica (Wien) 2008;150:1241–1247; discussion 1248. 10. Aarabi B, Hesdorffer DC, Ahn ES, Aresco C, Scalea TM, Eisenberg HM. Outcome following decompressive craniectomy for malignant swelling due to severe head injury. Journal of Neurosurgery 2006; 104:469–479. 11. Stiver SI. Complications of decompressive craniectomy for traumatic brain injury. Neurosurgical Focus 2009;26:E7. 12. Aarabi B, Chesler D, Maulucci C, Blacklock T, Alexander M. Dynamics of subdural hygroma following decompressive craniectomy: A comparative study. Neurosurgical Focus 2009; 26:E8. 13. De Bonis P, Sturiale CL, Anile C, Gaudino S, Mangiola A, Martucci M, Colosimo C, Rigante L, Pompucci A. Decompressive craniectomy, interhemispheric hygroma and hydrocephalus: A timeline of events? Clinical Neurology & Neurosurgery 2013; 115:1308–1312. 14. Kaen A, Jimenez-Roldan L, Alday R, Gomez PA, Lagares A, Alen JF, Lobato RD. Interhemispheric hygroma after decompressive craniectomy: Does it predict posttraumatic hydrocephalus? Journal of Neurosurgery 2010;113:1287–1293. 15. Honeybul S, Ho KM. Incidence and risk factors for post-traumatic hydrocephalus following decompressive craniectomy for intractable intracranial hypertension and evacuation of mass lesions. Journal of Neurotrauma 2012;29:1872–1878. 16. Poca MA, Sahuquillo J, Mataro M, Benejam B, Arikan F, Baguena M. Ventricular enlargement after moderate or severe head injury: A frequent and neglected problem. Journal of Neurotrauma 2005;22:1303–1310. 17. Jiao QF, Liu Z, Li S, Zhou LX, Li SZ, Tian W, You C. Influencing factors for posttraumatic hydrocephalus in patients suffering from severe traumatic brain injuries. Chinese Journal of Traumatology 2007;10:159–162. 18. Choi I, Park HK, Chang JC, Cho SJ, Choi SK, Byun BJ. Clinical factors for the development of posttraumatic hydrocephalus after decompressive craniectomy. Journal of Korean Neurosurgical Society 2008;43:227–231. 19. Bullock MR, Povlishock JT. Guidelines for the management of severe traumatic brain injury. Editor’s Commentary. Journal of Neurotrauma 2007;24(Supp 1):2 (pages preceding S1). 20. Honeybul S, Ho KM. Decompressive craniectomy for severe traumatic brain injury: The relationship between surgical complications and the prediction of an unfavourable outcome. Injury 2014; 45:1332–1339. 21. Ohno K, Suzuki R, Masaoka H, Matsushima Y, Inaba Y, Monma S. Chronic subdural haematoma preceded by persistent traumatic subdural fluid collection. Journal of Neurology, Neurosurgery & Psychiatry 1987;50:1694–1697. 22. Kaufman HH, Childs TL, Wagner KA, Bernstein DP, Karon M, Khalid M, Gildenberg PL. Post-traumatic subdural hygromas: Observations concerning a surgical enigma. Acta Neurochirurgica (Wien) 1984;72:197–209. 23. Born JD, Hans P, Flandroy P, Milbouw G, Bonnal J. [Incidence and treatment of subdural hygroma in severe head injuries]. Acta Neurologica Belgica 1984;84:174–182. 24. Yang XF, Wen L, Li G, Zhan RY, Ma L, Liu WG. Contralateral subdural effusion secondary to decompressive craniectomy performed in patients with severe traumatic brain injury: Incidence, clinical presentations, treatment and outcome. Medical Principles & Practice 2009;18:16–20. 25. Wang HK, Lu K, Liang CL, Tsai YD, Wang KW, Liliang PC. Contralateral subdural effusion related to decompressive craniectomy performed in patients with severe traumatic brain injury. Injury 2012;43:594–597. 26. Jeon SW, Choi JH, Jang TW, Moon SM, Hwang HS, Jeong JH. Risk factors associated with subdural hygroma after decompressive craniectomy in patients with traumatic brain injury: A comparative study. Journal of the Korean Neurosurgical Society 2011;49: 355–358. 27. Su TM, Lee TH, Chen WF, Lee TC, Cheng CH. Contralateral acute epidural hematoma after decompressive surgery of acute subdural hematoma: Clinical features and outcome. Journal of Trauma 2008; 65:1298–1302.
10
Q. Yuan et al.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
28. Su FW, Ho JT, Wang HC. Acute contralateral subdural hygroma following craniectomy. Journal of Clinical Neurosciences 2008;15: 305–307. 29. Lee KS. The pathogenesis and clinical significance of traumatic subdural hygroma. Brain Injury 1998;12:595–603. 30. St John JN, Dila C. Traumatic subdural hygroma in adults. Neurosurgery 1981;9:621–626.
Brain Inj, Early Online: 1–10
31. Wetterling T, Demierre B, Rama B, Spoerri O. The clinical course of surgically treated posttraumatic subdural hygromas. Acta Neurochirurgica (Wien) 1986;83:99–104. 32. Kilincer C, Simsek O, Hamamcioglu MK, Hicdonmez T, Cobanoglu S. Contralateral subdural effusion after aneurysm surgery and decompressive craniectomy: Case report and review of the literature. Clinical Neurology & Neurosurgery 2005;107:412–416.
Subdural hygroma following decompressive craniectomy or non-decompressive craniectomy in patients with traumatic brain injury: Clinical features and risk factors Qiang Yuan, Xing Wu, Jian Yu, Yirui Sun, Zhiqi Li, Zhuoying Du, Xuehai Wu, Liangfu Zhou, & Jin Hu
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Department of Neurosurgery, Huashan Hospital, Fudan University, Shanghai, PR China
Abstract
Keywords
Objective: Subdural hygroma (SDG) is a common complication that can occur after head trauma or secondary to decompressive craniectomy (DC). SDGs can be located not only ipsilateral or contralateral to the side of the DC, but also bilateral or unilateral in patients without DC. This study investigated the incidence and risk factors for different types of SDG in a large cohort of patients with traumatic brain injury (TBI). Methods: A retrospective study was conducted involving 379 adult patients with TBI who were admitted to Huashan Hospital, Fudan University between January 2009 and December 2013. As the outcome was dichotomous (SDG vs no SDG or hydrocephalus vs no hydrocephalus), multivariate logistic regression analyses were used to identify independent risk factors for the development of SDGs in patients without DC, ipsilateral SDG after unilateral DC, contralateral SDG after unilateral DC or SDG after bilateral DC. Risk factors for the development of hydrocephalus were also evaluated in patients with and without DC. Results: Among the 207 (54.6%) patients without DC, 30 (14.5%) had unilateral SDGs and 34 (16.4%) had bilateral SDGs. Of the 172 patients (45.4%) with DC, 134 (77.9%) underwent unilateral DC and 38 (22.1%) underwent bilateral DC. Of the 134 patients who underwent unilateral DC, 49 developed SDG, including 22 (16.4%) ipsilateral SDG, 19 (14.2%) contralateral SDG and eight (6.0%) both ipsilateral and contralateral SDGs. For patients undergoing bilateral DC, 13 (34.2%) developed a SDG. No significant difference in the incidence of SDG was observed between the patients with and without DC (36.0% vs 30.9%, p ¼ 0.291), but the characteristics of SDGs were different between the two groups. Logistic regression analysis showed that factors independently associated with the development of SDG were male sex (odds ratio [OR] ¼ 3.861; 95% CI ¼ 1.642–9.091; p ¼ 0.002), older age (OR ¼ 1.046; 95% CI ¼ 1.021–1.070; p50.001), basal cistern haemorrhage (OR ¼ 4.608; 95% CI ¼ 1.510–14.064; p ¼ 0.007), diffuse injury and swelling (OR ¼ 3.158; 95% CI ¼ 1.341–7.435; p ¼ 0.008) or diffuse injury and shift (OR ¼ 3.826; 95% CI ¼ 1.141–12.830; p ¼ 0.030) in patients without DC. Temporal haematoma or contusion in the non-DC side (OR ¼ 2.623; 95% CI ¼ 1.070–6.428; p ¼ 0.035) and traumatic SAH (OR ¼ 3.751; 95% CI ¼ 1.047–13.438; p ¼ 0.042) were independently associated with the development of ipsilateral SDG in patients who underwent unilateral DC. However, factors independently associated with the development of contralateral SDG were frontal haematoma or contusion on the non-DC side (OR ¼ 3.145; 95% CI ¼ 1.272–7.774; p ¼ 0.013) and SDH on the non-DC side (OR ¼ 7.024; 95% CI ¼ 1.477–33.390; p ¼ 0.014). Only craniectomy area (OR ¼ 1.030; 95% CI ¼ 1.008–1.052; p ¼ 0.008) was independently associated with the development of SDG in patients with bilateral DC. In the multivariate analysis, SDG in patients without DC was not associated with the development of hydrocephalus. However, SDG was significantly associated with the development of hydrocephalus for patients who underwent DC (OR ¼ 2.173; 95% CI ¼ 1.362–3.467; p ¼ 0.001). Conclusions: This study suggested that the incidence of SDG in patients who have and have not undergone DC was identical; however, the patients’ characteristics and risk factors differed. Therefore, the management and prediction of SDG should be performed according to SDG type.
Decompressive craniectomy, post-traumatic hydrocephalus, risk factor, subdural hygroma, traumatic brain injury History Received 26 August 2014 Revised 17 November 2014 Accepted 4 January 2015 Published online 23 April 2015
Introduction
Correspondence: Jin Hu, MD, PhD, Department of Neurosurgery, Huashan Hospital, Fudan University, 12 Wulumuqi Zhong Road. Shanghai 200040, PR China. Tel: +86-21-52887736; +8613003116904. Email: [email protected]
Traumatic brain injury (TBI) is an important cause of injuryrelated hospitalization, disability and death worldwide [1]. It represents a significant public health problem in China and across the world. It is estimated that an average of 1.4 million TBIs occur each year in the US, including 1.365 million emergency department visits, 275 000 hospitalizations and 52
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
2
Q. Yuan et al.
000 deaths [2, 3]. A large portion of the complications are secondary to head trauma or craniotomy, such as subdural hygroma (SDG) and post-traumatic hydrocephalus, which may further aggravate the disease, lead to poor prognosis and increase the burden of TBI. The prevention and management of complications are crucial parts of managing TBI [4]. SDG is not a rare complication that can occur after head trauma or secondary to decompressive craniectomy (DC). The incidence of SDG after head trauma is 7–12% [5]. However, the incidence increases to 23–56% of patients with a head injury if a DC is performed [6–9]. The occurrence of SDG may be due to a change in the dynamics of CSF circulation or an underlying disorder of normal CSF absorption. Moreover, SDG may be a complication of removing a bone flap, which results in formation of space in which fluid can accumulate [9–11]. SDG can be seen in the first week after injury and fluid collection may increase for up to 4 weeks [12]; most cases resolve gradually without surgical management. However, some SDGs may require surgical intervention to control the further neurological deterioration. The exact pathogenesis of SDG after TBI remains unclear and the risk factors are rarely reported or reported only for patients who undergo DC. SDGs can be located not only ipsilateral or contralateral to the side of the DC, but also bilateral or unilateral in patients without DC. It is unclear whether different types of SDG have an identical clinical course and risk factors. De Bonis et al. [13] reported that inter-hemispheric hygroma and hydrocephalus is a timeline of events for patients with TBI after DC. In addition, many studies have shown that SDG is a significant risk factor for hydrocephalus after DC [14, 15]. However, these studies focused only on craniectomy patients; therefore, whether patients without DC with TBI have identical findings remains unknown. Several factors have been associated with the development of post-traumatic hydrocephalus [16], such as older age, subarachnoid haemorrhage, CSF infection, lower Glasgow Coma Scale (GCS) score and a wide craniectomic flap [15, 17, 18]. De Bonis et al. [13] reported that craniectomy with a superior limit closer to the midline than 25 mm might pre-dispose a patient to development of hydrocephalus. It remains unclear whether these risk factors associated with post-traumatic hydrocephalus are consistent with the development of post-traumatic SDG. Therefore, it is hypothesized that, regardless of whether DC and SDG are common after TBI, the risk factors for this complication are related primarily to the severity of the primary brain injury. This study investigated the incidence and risk factors for different types of SDG in a large cohort of patients with TBI.
Methods Patients A retrospective study was conducted involving 379 adult patients with TBI who were admitted to Huashan Hospital, Fudan University between January 2009 and December 2013. This research protocol was approved by the University Hospital Medical Ethics Board. All participants provide their written informed consent at the beginning of the study and the consent procedure was approved by the University Hospital
Brain Inj, Early Online: 1–10
Medical Ethics Board. The next of kin, carers or guardians consented on the behalf of participants whose capacity to consent was compromised. The inclusion criteria were as follows: TBI with radiological signs of intracranial brain injury [epidural or subdural haematoma, intracerebral haematoma or contusion or subarachnoid haemorrhage (SAH)] documented by computed tomography (CT) scan; age 418 years; and admission within 24 hours after TBI. Patients who died in the acute phase (57 days after admission) were excluded. Among the 379 patients, 172 received DC to allow for brain swelling and 207 did not. The indications for DC were: (1) following evacuation of a mass lesion when the brain was excessively swollen; and (2) intracranial pressure (ICP) that increased to 425 mm Hg and/or reduced cerebral perfusion pressure to 560 mm Hg for longer than 30 minutes. The distance of the craniectomy from the midline and the area of the craniectomy were determined in patients who underwent DC. The incidence, clinical presentations and outcomes of SDGs related to TBI were analysed in both the DC and nonDC groups. Clinical and demographic characteristics, including age, sex, mechanism of injury, injury severity score (ISS), pupillary reaction to light, ICP monitor and GCS score at admission were recorded in all patients. Moreover, the characteristics of the initial CT scan on admission were used to assess the CT findings and Marshall CT grade. All available serial CT scans from admission to discharge were evaluated to identify SDG and subsequent hydrocephalus. There is no established definition of the minimum thickness or volume of SDGs; thus, they are defined as a low-density collection 40.5 cm in maximal depth measured from the cortical surface to the inner aspect of the scalp or skull. Posttraumatic hydrocephalus was defined by radiological evidence of progressive ventricular dilatation (Evans index40.3) with trans-ependymal oedema, together with the presence of either clinical deterioration or failure to make neurological progress over time and some evidence of clinical improvement after insertion of a ventriculo-peritoneal shunt [15]. All patients received primary care in the emergency room, as recommended by Advanced Trauma Life Support specifications. All patients were treated in accordance with the 2007 guidelines of the Brain Trauma Foundation: use of CSF drainage via ventriculostomy, mannitol treatment, hyperventilation for intracranial hypertension if needed and normovolemia maintenance with saline by central venous pressure (CVP) of 5–8 mm Hg, maintenance of CPP 460 mm Hg and ICP 520 mm Hg [19]. The neurological outcomes were determined according to the Glasgow Outcome Score (GOS) as follows: 1 ¼ dead; 2 ¼ vegetative state with an inability to interact with the environment; 3 ¼ severe disability with an inability to live independently but the ability to follow commands; 4 ¼ moderate disability with the ability to live independently but an inability to return to work or school; and 5 ¼ good recovery with the ability to return to work or school. Prognostic evaluations were determined using the GOS assessment at 3 months after the trauma. GOS evaluations were performed by physicians either in person or via telephone. A GOS of 1–3 was categorized as an unfavourable outcome, whereas a score of 4–5 was deemed a favourable outcome.
DOI: 10.3109/02699052.2015.1004760
Statistical analyses All data were analysed using the SPSS statistical software v.16 (SPSS Inc., Chicago, IL). Descriptive statistics were used to describe the incidence and other characteristics of the patients. The unpaired Student’s t-test or a non-parametric test was used to compare continuous variables and the chi-square or Fisher’s exact test to evaluate categorical variables. As the outcome was dichotomous (SDG vs no SDG or hydrocephalus vs no hydrocephalus), multivariate logistic regression analyses were used to identify independent risk factors for the development of SDGs in patients without DC, ipsilateral SDG after unilateral DC, contralateral SDG after unilateral DC or SDG after bilateral DC. Risk factors for the development of hydrocephalus were also evaluated in patients with and without DC. A p-value 50.05 was considered to indicate significance.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Results Incidence and characteristics of SDGs The study population consisted of 379 patients (287 males, 92 females; mean age ¼ 50.0 ± 15.7 years) (Figure 1). Of the initial cohort of 379 patients, 172 underwent DC to allow for brain swelling and 207 did not. Among the 207 (54.6%) patients without DC, 30 (14.5%) had unilateral SDGs (Figures 2(a) and (b)) and 34 (16.4%) had bilateral SDGs (Figures 2(c) and (d)). Of the 172 patients (45.4%) who underwent DC, 134 (77.9%) underwent unilateral DC and 38 (22.1%) underwent bilateral DC. Of the 134 patients who underwent unilateral DC, 49 developed SDG, including 22 (16.4%) ipsilateral SDG (Figures 2(g) and (h)), 19 (14.2%) contralateral SDG (Figures 2(e) and (f)) and eight (6.0%) both ipsilateral and contralateral SDGs (Figure 2(i)). For patients undergoing bilateral DC, 13 (34.2%) developed a SDG. No significant
Subdural hygroma in traumatic brain injury
3
difference in the incidence of SDG was observed between the patients with and without DC (36.0% vs 30.9%, p ¼ 0.291). The width of the SDG in the patients with DC was 1.20 ± 0.65 cm, which was significantly larger than those in patients without DC (0.76 ± 0.33 cm) (p50.001). SDGs in the no-DC group developed at 4.58 ± 4.06 days after injury, which was significantly earlier than the DC group (7.42 ± 6.95 days) (p50.001). In addition, compared with the patients who developed SDG without DC, the patients who developed SDG after DC were more likely to be younger, have a lower GCS score on admission, have one pupil react to light, have an SAH, an evacuated mass lesion, an ICP monitor, a SDH, an epidural haemorrhage (EDH), a frontal haematoma or contusion and worse neurological outcomes at 3 months. SDGs resolved gradually with conservative management in most of the patients with TBI who did not undergo DC (59/64). However, SDGs in the remaining five patients contributed to progressive clinical deterioration and burr hole drainage was performed. In the 57 patients who underwent DC, the SDG resolved gradually with conservative management and only five patients had progressive clinical deterioration and burr hole drainage performed. SDG for patients without DC Patients without DC developing a unilateral or bilateral SDG were more likely to be male, older, have a diffuse injury and swelling, a basal cistern haemorrhage, a temporal haematoma or contusion on brain CT scan and worse neurological outcomes at 3 months (Table I). In addition, patients who developed SDG were less likely to have an EDH. No differences in the mechanism of injury, intubation in the emergency room, GCS at admission, ISS, pupillary reaction
Figure 1. Flowchart showing different types of subdural hygroma in patients with and without decompressive craniectomy.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
4
Q. Yuan et al.
Brain Inj, Early Online: 1–10
Figure 2. CT scan of different types of subdural hygroma. (a, b) Unilateral subdural hygroma in patients without DC; (c, d) bilateral subdural hygroma in patients without DC; (e, f) contralateral subdural hygroma after DC; (g, h) ipsilateral subdural hygroma after DC; (i) bilateral subdural hygroma after DC. DC, decompressive craniectomy.
to light, ICP monitor, other CT findings or mortality were observed between patients with and without SDG. Logistic regression analysis showed that factors independently associated with the development of SDG were male sex (odds ratio [OR] ¼ 3.861; 95% CI ¼ 1.642–9.091; p ¼ 0.002), older age (OR ¼ 1.046; 95% CI ¼ 1.021–1.070; p50.001), basal cistern haemorrhage (OR ¼ 4.608; 95% CI ¼ 1.510–14.064; p ¼ 0.007), diffuse injury and swelling (OR ¼ 3.158; 95% CI ¼ 1.341–7.435; p ¼ 0.008) and diffuse injury and shift (OR ¼ 3.826; 95% CI ¼ 1.141–12.830; p ¼ 0.030) (Table II). SDG in patients after DC For patients with unilateral DC, the occurrence of ipsilateral SDG was associated with a temporal haematoma or contusion on the non-DC side, a traumatic SAH on the brain CT scan and worse neurological outcomes at 3 months (Table III). However, patients who developed contralateral SDG following unilateral DC were more likely to have an ICP monitor, a SDH on the non-DC side, a frontal haematoma or contusion on the non-DC side and a temporal haematoma or contusion on the non-DC side (Table IV). The occurrence of SDG was significantly associated with a lower GCS score on admission,
non-reactive pupils, a larger craniectomy area and worse neurological outcomes at 3 months in patients with bilateral DC (Table V). No differences were observed among other variables. In the multivariate logistic regression analysis, temporal haematoma or contusion in the non-DC side (OR ¼ 2.623; 95% CI ¼ 1.070–6.428; p ¼ 0.035) and traumatic SAH (OR ¼ 3.751; 95% CI ¼ 1.047–13.438; p ¼ 0.042) were independently associated with an ipsilateral SDG in patients who underwent unilateral DC (Table II). However, factors independently associated with the development of contralateral SDG were frontal haematoma or contusion on the non-DC side (OR ¼ 3.145; 95% CI ¼ 1.272–7.774; p ¼ 0.013) and SDH on the non-DC side (OR ¼ 7.024; 95% CI ¼ 1.477–33.390; p ¼ 0.014) (Table II). In addition, the logistic regression analysis showed that only craniectomy area (OR ¼ 1.030; 95% CI ¼ 1.008–1.052; p ¼ 0.008) was independently associated with an SDG in patients following bilateral DC (Table II). Relationship between SDG and hydrocephalus For patients without DC, subsequent hydrocephalus was present in six of 64 patients (9.4%) who developed a unilateral
Subdural hygroma in traumatic brain injury
DOI: 10.3109/02699052.2015.1004760
5
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Table I. Difference in characteristics between patients who developed subdural hygroma and those who did not for patients without DC. Variable
W SDG (%)
W/o SDG (%)
No. of patients Male sex Mean age (years) (mean ± SD) Mechanism of injury Motor vehicle accident Fall from a height Stumble and fall Blow to head Injured by a crashing object Others Intubation in the emergency room GCS at admission (mean ± SD) GCS 3–8 GCS 9–15 ISS (mean ± SD) Pupils react to light Both One None Injury severity (Marshall CT grade) II (Diffuse injury without swelling) III (Diffuse injury and swelling) IV (Diffuse injury and shift) V (Evacuated mass lesion) VI (Non-evacuated mass lesion 425 cc) ICP monitor CT findings EDH SDH Frontal haematoma or contusion Temporal haematoma or contusion tSAH Basal cistern haemorrhage Outcome Unfavourable Favourable Survived Died
64 54 (84.4) 57.5 ± 14.5
143 97 (67.8) 48.7 ± 16.6
39 (60.9) 7 (10.9) 15 (23.4) 1 (1.6) 0 (0.0) 2 (3.1) 9 (14.1) 10.1 ± 3.1 21 (32.8) 43 (67.2) 20.6 ± 6.6
78 (54.5) 16 (11.2) 27 (18.9) 5 (3.5) 8 (5.6) 9 (6.3) 14 (9.8) 10.9 ± 3.3 37 (25.9) 106 (74.1) 20.7 ± 7.2
60 (93.8) 4 (6.2) 0 (0.0)
139 (97.2) 4 (2.8) 0 (0.0)
p Value 0.013 50.001 0.454
0.366 0.120 0.304 0.873 0.423
0.046 12 30 7 7 8 44
(18.8) (46.9) (10.9) (10.9) (12.5) (68.8)
56 45 12 18 12 83
(39.2) (31.5) (8.4) (12.6) (8.4) (58.0)
0.144
8 23 38 45 38 13
(12.5) (35.9) (59.4) (70.3) (59.4) (20.3)
36 39 80 74 65 8
(25.2) (27.3) (55.9) (51.7) (45.5) (5.6)
0.039 0.208 0.645 0.013 0.064 0.001
36 28 62 2
(56.2) (43.8) (96.9) (3.1)
48 95 141 2
(33.6) (66.4) (98.6) (1.4)
0.002 0.774
DC, decompressive craniectomy; SDG, subdural hygroma; GCS, Glasgow coma scale; ISS, injury severity score; ICP, intracranial pressure; EDH, epidural haematoma; SDH, subdural haematoma; tSAH, traumatic subarachnoid haemorrhage. Table II. Multivariate analysis of the risk factors for developing different types of subdural hygroma. Variable Subdural hygroma w/o decompressive craniectomy Male Age Basal cistern haemorrhage Marshall CT grade(reference: II, Diffuse injury without swelling) III (Diffuse injury and swelling) IV (Diffuse injury and shift) Contralateral subdural hygroma following decompressive craniectomy Frontal haematoma or contusion in the non-DC side SDH in the non-DC side Ipsilateral subdural hygroma following decompressive craniectomy Temporal hematoma or contusion in the non-DC side tSAH Subdural hygroma following bilateral decompressive craniectomy Cranial defect size
or bilateral SDG and in six of 143 patients (4.2%) without SDG (p ¼ 0.249). However, 16 of 62 patients (25.8%) with SDG developed subsequent hydrocephalus following DC, which was significantly higher than those without SDG
Adjusted OR (95% CI)
p Value
3.861 (1.642–9.091) 1.046 (1.021–1.070) 4.608 (1.510–14.064)
0.002 50.001 0.007
3.158 (1.341–7.435) 3.826 (1.141–12.830)
0.008 0.030
3.145 (1.272–7.774) 7.024 (1.477–33.390)
0.013 0.014
2.623 (1.070–6.428) 3.751 (1.047–13.438)
0.035 0.042
1.030 (1.008–1.052)
0.008
(10.0%) (p ¼ 0.006). After adjusting for other factors (age, sex, GCS at admission, pupillary reaction to light and CT findings) in the multivariate analysis, SDG in the patients without DC was not associated with the development of
6
Q. Yuan et al.
Brain Inj, Early Online: 1–10
Table III. Difference in characteristics between patients who developed ipsilateral subdural hygroma following unilateral DC and those who did not.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Variable No. of patients Male sex Mean age (years) (mean ± SD) Mechanism of injury Motor vehicle accident Fall from a height Stumble and fall Blow to head Injured by a crashing object Others Intubation in the emergency room GCS at admission (mean ± SD) GCS 3–8 GCS 9–15 ISS (mean ± SD) Pupils react to light Both One None Injury severity (Marshall CT grade) II (Diffuse injury without swelling) III (Diffuse injury and swelling) IV (Diffuse injury and shift) V (Evacuated mass lesion) ICP monitor CT findings EDH in the non-DC side SDH in the non-DC side Frontal haematoma or contusion in the non-DC side Temporal haematoma or contusion in the non-DC side EDH in the DC side SDH in the DC side Frontal haematoma or contusion in the DC side Temporal haematoma or contusion in the DC side tSAH Basal cistern haemorrhage The area of the craniectomy (cm2) (mean ± SD) The distance of the craniectomy from the midline (cm) (mean ± SD) Outcome Unfavourable Favourable Survived Died
W ipsilateral SDG (%)
W/o ipsilateral SDG (%)
30 25 (83.3) 46.9 ± 16.5
104 79 (76.0) 48.8 ± 14.2
16 (53.3) 6 (20.0) 6 (20.0) 0 (0.0) 0 (0.0) 2 (6.7) 4 (13.3) 6.8 ± 2.3 24 (80.0) 6 (20.0) 25.1 ± 7.0
63 (60.6) 19 (18.3) 11 (10.6) 3 (2.9) 4 (3.8) 4 (3.8) 12 (11.5) 7.4 ± 2.6 74 (71.2) 30 (28.8) 23.2 ± 6.2
14 (46.7) 14 (46.7) 2 (6.7)
54 (51.9) 40 (38.5) 10 (9.6)
0 4 5 21 26
2 9 25 68 90
p Value 0.393 0.538 0.583
1.000 0.328 0.336 0.171 0.741
0.747 (0.0) (13.3) (16.7) (70.0) (86.7)
2 (6.7) 2 (6.7) 11 (36.7) 12 (40.0) 6 (20.0) 18 (60.0) 21 (70.0) 23 (76.7) 27 (90.0) 0 (0.0) 103.2 ± 28.1 2.0 ± 1.4 25 5 27 3
(83.3) (16.7) (90.0) (10.0)
(1.9) (8.7) (24.0) (65.4) (86.5)
1.000
8 (7.7) 6 (5.8) 29 (27.9) 20 (19.2) 21 (20.2) 54 (51.9) 65 (62.5) 67 (64.4) 72 (69.2) 2 (1.9) 97.7 ± 38.0 2.3 ± 1.2
1.000 1.000 0.354 0.019 0.982 0.434 0.450 0.208 0.023 1.000 0.462 0.194
50 54 97 7
(48.1) (51.9) (93.3) (6.7)
0.001 0.837
DC, decompressive craniectomy; SDG, subdural hygroma; GCS, Glasgow coma scale; ISS, injury severity score; ICP, intracranial pressure; EDH, epidural haematoma; SDH, subdural haematoma; tSAH, traumatic subarachnoid haemorrhage.
hydrocephalus. However, SDG was significantly associated with the development of hydrocephalus in patients with DC (OR ¼ 2.173; 95% CI ¼ 1.362–3.467; p ¼ 0.001).
Discussion This was the first study to investigate the incidence and risk factors for different types of SDG in a large cohort of patients with TBI. No significant difference was found in the incidence of SDG between patients who did and did not undergo DC, but the patients between the two groups had different characteristics. The patients who underwent DC and subsequently developed SDG were more likely to have greater thickness, later onset time, younger age, lower GCS score on admission, one pupil reacting to light, a SAH, an evacuated mass lesion, a ICP monitor, a SDH, a EDH, a frontal haematoma or contusion and worse neurological outcomes at 3 months than those who did not undergo DC but developed
SDG. Therefore, the patients who underwent DC and subsequently developed SDG may have had more severe primary brain injury and more severe representation of SDG than those who did not undergo DC but developed SDG. Honeybul et al. [20] have reported the association between injury severity and the development of SDG in the patients with DC. However, the comparison between the patients with and without DC is still lacking. SDG is a well-known complication of TBI. The incidence of SDG after TBI in patients who have not undergone DC is 5–21% in several clinical series. Ohno et al. [21] reported 43 SDGs (6%) in 715 patients and the incidence of SDG in the Kaufman et al. [22] study was 38 (4%) of 881 patients with a closed-head injury. Born et al. [23] described 16 patients (15%) with SDGs among 109 with blunt TBI. However, this study found that SDGs developed in 64 (30.9%) of 207 patients who had not undergone DC, which was higher than previous reports. The reported incidence appears to increase
Subdural hygroma in traumatic brain injury
DOI: 10.3109/02699052.2015.1004760
7
Table IV. Difference in characteristics between patients who developed contralateral subdural hygroma following unilateral DC and those who did not.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Variable No. of patients Male sex Mean age (years) (mean ± SD) Mechanism of injury Motor vehicle accident Fall from a height Stumble and fall Blow to head Injured by a crashing object Others Intubation in the emergency room GCS at admission (mean ± SD) GCS 3–8 GCS 9–15 ISS (mean ± SD) Pupils react to light Both One None Injury severity (Marshall CT grade) II (Diffuse injury without swelling) III (Diffuse injury and swelling) IV (Diffuse injury and shift) V (Evacuated mass lesion) ICP monitor CT findings EDH in the non-DC side SDH in the non-DC side Frontal haematoma or contusion in the non-DC side Temporal haematoma or contusion in the non-DC side EDH in the DC side SDH in the DC side Frontal haematoma or contusion in the DC side Temporal haematoma or contusion in the DC side tSAH Basal cistern haemorrhage The area of the craniectomy (cm2) (mean ± SD) The distance of the craniectomy from the midline (cm) (mean ± SD) Outcome Unfavourable Favourable Survived Died
W contralateral SDG (%) W/o contralateral SDG (%) p Value 27 22 (81.5) 52.1 ± 15.0
107 82 (76.6) 47.4 ± 14.5
14 (51.9) 4 (14.8) 3 (11.1) 2 (7.4) 1 (3.7) 3 (11.1) 3 (11.1) 6.8 ± 2.1 22 (81.5) 5 (18.5) 23.5 ± 4.5
65 (60.7) 21 (19.6) 14 (13.1) 1 (0.9) 3 (2.8) 3 (2.8) 13 (12.1) 7.4 ± 2.7 76 (71.0) 31 (29.0) 23.7 ± 6.8
15 (55.6) 10 (37.0) 2 (7.4)
53 (49.5) 44 (41.1) 10 (9.3)
1 5 4 17 27
1 8 26 72 89
0.589 0.135 0.139
1.000 0.339 0.274 0.847 0.900
0.149 (3.7) (18.5) (14.8) (63.0) (100.0)
0 (0.0) 5 (18.5) 14 (51.9) 11 (40.7) 7 (25.9) 14 (51.9) 18 (66.7) 18 (66.7) 21 (77.8) 0 (0.0) 99.4 ± 25.7 2.4 ± 1.4 17 10 25 2
(63.0) (37.0) (92.6) (7.4)
(0.9) (7.5) (24.3) (67.3) (83.2)
0.048
10 (9.3) 3 (2.8) 26 (24.3) 21 (19.6) 20 (18.7) 58 (54.2) 68 (63.6) 72 (67.3) 78 (72.9) 2 (1.9) 98.8 ± 38.2 2.2 ± 1.3
0.214 0.009 0.005 0.021 0.402 0.827 0.763 0.951 0.606 1.000 0.937 0.581
58 49 99 8
(54.2) (45.8) (92.5) (7.5)
0.413 1.000
DC, decompressive craniectomy; SDG, subdural hygroma; GCS, Glasgow coma scale; ISS, injury severity score; ICP, intracranial pressure; EDH, epidural haematoma; SDH, subdural haematoma; tSAH, traumatic subarachnoid haemorrhage.
following DC, with reports ranging from 26–60%. Yang et al. [9] documented 23 cases (21.3%) of SDGs in 108 patients who had undergone DC for severe head injury. Aarabi et al. [12] reported that SDGs developed in 39 (57.4%) of 68 patients who survived41 month after TBI and DC. This study found that SDGs developed in 62 (36.0%) of 172 patients who underwent DC. In a 2009 study, Yang et al. [24] reported 11 cases (6.5%) of contralateral SDG among 169 patients who had undergone DC for severe head injury. Wang et al. [25] reported that the overall incidence of contralateral subdural effusion was 7.3%. In this study, of the 134 patients who underwent unilateral DC, 30 (22.4%) had ipsilateral SDG and 27 (20.1%) had contralateral SDG, which is also higher than previous reports. It was also found that the incidence of SDG after TBI in patients who had undergone DC was not significantly higher than that in patients who had not undergone DC (36.0% vs 30.9%). The main reason is likely use of different inclusion and diagnostic criteria in the
previous study; however, uniform inclusion and diagnostic criteria were used in this work. Jeon et al. [26] reported that radiological factors indicate that a midline shift 45 mm, SAH, delayed hydrocephalus, compression of basal cisterns and tearing of the arachnoid membrane may be risk factors associated with SDG after DC. However, in this study, the risk factors for SDG in patients without DC, ipsilateral SDG after unilateral DC, contralateral SDG after unilateral DC and SDG after bilateral DC varied. In this study, temporal haematoma or contusion on the nonDC side and traumatic SAH were independently associated with the development of ipsilateral SDG in patients who underwent unilateral DC. Frontal haematoma or contusion on the non-DC side and SDH on the non-DC side were associated with contralateral SDG. A basal cistern haemorrhage was associated with SDG in patients without DC. All of these results indicate that tearing of the arachnoid membrane may be a risk factor associated with SDG. Previous studies
8
Q. Yuan et al.
Brain Inj, Early Online: 1–10
Table V. Difference in characteristics between patients who developed subdural hygroma following bilateral DC and those who did not.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
Variable No. of patients Male sex Mean age (years) (mean ± SD) Mechanism of injury Motor vehicle accident Fall from a height (3 m) Stumble and fall Blow to head Injured by a crashing object Others Intubation in the emergency room GCS at admission (mean ± SD) GCS 3–8 GCS 9–15 ISS (mean ± SD) Pupils react to light Both One None Injury severity (Marshall CT grade) II (Diffuse injury without swelling) III (Diffuse injury and swelling) IV (Diffuse injury and shift) V (Evacuated mass lesion) ICP monitor CT findings EDH SDH Frontal haematoma or contusion Temporal haematoma or contusion tSAH Basal cistern haemorrhage The area of the craniectomy (cm) (mean ± SD) Outcome Unfavourable Favourable Survived Died
W/SDG (%)
W/o SDG (%)
13 12 (92.3) 51.8 ± 17.3
25 20 (80.0) 45.8 ± 13.3
10 (76.9) 2 (15.4) 0 (0.0) 0 (0.0) 0 (0.0) 1 (7.7) 2 (15.4) 6.5 ± 2.1 10 (76.9) 3 (23.1) 24.9 ± 9.9
15 (60.0) 3 (12.0) 3 (12.0) 0 (0.0) 1 (4.0) 3 (12.0) 1 (4.0) 8.8 ± 3.1 10 (40.0) 15 (60.0) 20.8 ± 7.3
5 (38.5) 8 (61.5) 0 (0.0)
18 (72.0) 4 (16.0) 3 (12.0)
p Value 0.604 0.240 0.849
0.548 0.023 0.031 0.156 0.015
0.261 0 2 0 11 13
(0.0) (15.4) (0.0) (84.6) (100.0)
5 (38.5) 7 (53.8) 11 (84.6) 11 (84.6) 13 (100.0) 0 (0.0) 151.5 ± 63.4 12 1 12 1
(92.3) (7.7) (92.3) (7.7)
1 10 1 13 23
(4.0) (40.0) (4.0) (52.0) (92.0)
0.538
5 (20.0) 7 (28.0) 24 (96.0) 14 (56.0) 18 (72.0) 2 (8.0) 94.0 ± 27.9
0.402 0.225 0.548 0.160 0.095 0.538 0.007
12 13 23 2
(48.0) (52.0) (92.0) (8.0)
0.020 1.000
DC, decompressive craniectomy; SDG, subdural hygroma; GCS, Glasgow coma scale; ISS, injury severity score; ICP, intracranial pressure; EDH, epidural haematoma; SDH, subdural haematoma; tSAH, traumatic subarachnoid haemorrhage.
have suggested that SDG and hydrocephalus following DC may be due to a reduction in pulsatile CSF flow and venous outflow when the medial margin of a unilateral craniotomy is close to the midline [13]. However, this study does not confirm these hypotheses. A number of mechanisms could be responsible for the development of SDG. First, the formation of a one-way flap valve due to the tearing or disruption of the arachnoid may result in CSF leakage and accumulation in the subdural space [24, 27]. Second, absorption disturbance of the CSF circulation would also increase the risk of accumulation of the effusion through a torn arachnoid [28]. The close correlation between SAH on the initial CT scan and subsequent hydrocephalus may explain the underlying disturbance of CSF circulation after TBI that promoted the formation of the SDG in some of the patients. Third, when there initially was a possible rupture in the arachnoid layer after TBI, the rapid decrease in ICP resulting from external ventricular drainage would form a pressure gradient and result in enlargement of the subdural space and accumulation of SDG [29, 30]. Fourth, a pressure gradient also could be formed when intraoperative tissue retracts and is not able to reshape, thus, the subsequent
brain shift plays an important role in development of the SDG [31]. However, SDGs in patients who have undergone DC have different characteristics from those who have not undergone DC. Therefore, other mechanisms may be responsible for the development of SDGs. When the effusion is not large or when there is no mass effect or deterioration in clinical presentation, conservative treatment is recommended for the treatment of an SDG. However, surgery should be performed when there is clinical evidence of midline shift, neurological deterioration or the enlargement of effusion. The type of operation includes burr hole drainage or placement of a subduropleural or subduroperitoneal shunt [32]. Drainage using a burr hole is considered a simple and effective procedure for subdural effusion. In this series, all patients who needed surgery received a burr hole. Honeybul et al. [15] found that SDG is a significant risk factor for hydrocephalus after DC, which is consistent with these findings. However, this study also found that SDG in patients without DC was not associated with the development of hydrocephalus. Therefore, DC combined with SDG may be an independent risk factor for hydrocephalus.
DOI: 10.3109/02699052.2015.1004760
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
It has been shown that different types of SDG in patients with TBI were associated with a worse long-term neurological outcome, as has been reported previously [15]. Although this association could be a result from residual confounding because of their association with the severity of the primary brain injury, it is possible that SDGs exert considerable pressure on the brain and, thus, reduce neurological recovery. There were several limitations to this study. First, although it included a reasonable number of patients with TBI, relatively few had SDG and hydrocephalus, which limited the power to identify multiple risk factors for SDG and hydrocephalus. Second, this was a retrospective study that relied on the accuracy of chart records or recall of individuals. Therefore, the findings should be subjected to more rigorous testing in a further well-designed, prospective study.
Conclusion The incidence of SDG in patients who have and have not undergone DC was identical; however, the patients’ characteristics and risk factors differed. SDG was an independent risk factor for hydrocephalus only in patients who had undergone DC. Therefore, the management and prediction of SDG should be performed according to SDG type.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. This work was supported by the National Natural Science Foundation of China (NSFC Grants 30371454, 81271375 and 81171133), The Science and Technology Commission of Shanghai Municipality Project (10JC1402300) and Shanghai Nature Science Foundation (08411952000).
References 1. Summers CR, Ivins B, Schwab KA. Traumatic brain injury in the United States: An epidemiologic overview. Mount Sinai Journal of Medicine 2009;76:105–110. 2. CDC grand rounds: Reducing severe traumatic brain injury in the United States. MMWR Morbidity & Mortality Weekly Report 2013;62:549–552. 3. Pearson WS, Sugerman DE, McGuire LC, Coronado VG. Emergency department visits for traumatic brain injury in older adults in the United States: 2006-08. Western Journal of Emergency Medicine 2012;13:289–293. 4. Honeybul S, Ho KM. Long-term complications of decompressive craniectomy for head injury. Journal of Neurotrauma 2011;28: 929–935. 5. Stone JL, Lang RG, Sugar O, Moody RA. Traumatic subdural hygroma. Neurosurgery 1981;8:542–550. 6. Yang XJ, Hong GL, Su SB, Yang SY. Complications induced by decompressive craniectomies after traumatic brain injury. Chinese Journal of Traumatology 2003;6:99–103. 7. Huang Q, Dai WM, Wu TH, Jie YQ, Yu GF, Fan XF. Comparison of standard large trauma craniotomy with routine craniotormy in treatment of acute subdural hematoma. Chinese Journal of Traumatology 2003;6:305–308. 8. Jiang JY, Xu W, Li WP, Xu WH, Zhang J, Bao YH, Ying YH, Luo QZ. Efficacy of standard trauma craniectomy for refractory intracranial hypertension with severe traumatic brain injury: A multicenter, prospective, randomized controlled study. Journal of Neurotrauma 2005;22:623–628.
Subdural hygroma in traumatic brain injury
9
9. Yang XF, Wen L, Shen F, Li G, Lou R, Liu WG, Zhan RY. Surgical complications secondary to decompressive craniectomy in patients with a head injury: A series of 108 consecutive cases. Acta Neurochirurgica (Wien) 2008;150:1241–1247; discussion 1248. 10. Aarabi B, Hesdorffer DC, Ahn ES, Aresco C, Scalea TM, Eisenberg HM. Outcome following decompressive craniectomy for malignant swelling due to severe head injury. Journal of Neurosurgery 2006; 104:469–479. 11. Stiver SI. Complications of decompressive craniectomy for traumatic brain injury. Neurosurgical Focus 2009;26:E7. 12. Aarabi B, Chesler D, Maulucci C, Blacklock T, Alexander M. Dynamics of subdural hygroma following decompressive craniectomy: A comparative study. Neurosurgical Focus 2009; 26:E8. 13. De Bonis P, Sturiale CL, Anile C, Gaudino S, Mangiola A, Martucci M, Colosimo C, Rigante L, Pompucci A. Decompressive craniectomy, interhemispheric hygroma and hydrocephalus: A timeline of events? Clinical Neurology & Neurosurgery 2013; 115:1308–1312. 14. Kaen A, Jimenez-Roldan L, Alday R, Gomez PA, Lagares A, Alen JF, Lobato RD. Interhemispheric hygroma after decompressive craniectomy: Does it predict posttraumatic hydrocephalus? Journal of Neurosurgery 2010;113:1287–1293. 15. Honeybul S, Ho KM. Incidence and risk factors for post-traumatic hydrocephalus following decompressive craniectomy for intractable intracranial hypertension and evacuation of mass lesions. Journal of Neurotrauma 2012;29:1872–1878. 16. Poca MA, Sahuquillo J, Mataro M, Benejam B, Arikan F, Baguena M. Ventricular enlargement after moderate or severe head injury: A frequent and neglected problem. Journal of Neurotrauma 2005;22:1303–1310. 17. Jiao QF, Liu Z, Li S, Zhou LX, Li SZ, Tian W, You C. Influencing factors for posttraumatic hydrocephalus in patients suffering from severe traumatic brain injuries. Chinese Journal of Traumatology 2007;10:159–162. 18. Choi I, Park HK, Chang JC, Cho SJ, Choi SK, Byun BJ. Clinical factors for the development of posttraumatic hydrocephalus after decompressive craniectomy. Journal of Korean Neurosurgical Society 2008;43:227–231. 19. Bullock MR, Povlishock JT. Guidelines for the management of severe traumatic brain injury. Editor’s Commentary. Journal of Neurotrauma 2007;24(Supp 1):2 (pages preceding S1). 20. Honeybul S, Ho KM. Decompressive craniectomy for severe traumatic brain injury: The relationship between surgical complications and the prediction of an unfavourable outcome. Injury 2014; 45:1332–1339. 21. Ohno K, Suzuki R, Masaoka H, Matsushima Y, Inaba Y, Monma S. Chronic subdural haematoma preceded by persistent traumatic subdural fluid collection. Journal of Neurology, Neurosurgery & Psychiatry 1987;50:1694–1697. 22. Kaufman HH, Childs TL, Wagner KA, Bernstein DP, Karon M, Khalid M, Gildenberg PL. Post-traumatic subdural hygromas: Observations concerning a surgical enigma. Acta Neurochirurgica (Wien) 1984;72:197–209. 23. Born JD, Hans P, Flandroy P, Milbouw G, Bonnal J. [Incidence and treatment of subdural hygroma in severe head injuries]. Acta Neurologica Belgica 1984;84:174–182. 24. Yang XF, Wen L, Li G, Zhan RY, Ma L, Liu WG. Contralateral subdural effusion secondary to decompressive craniectomy performed in patients with severe traumatic brain injury: Incidence, clinical presentations, treatment and outcome. Medical Principles & Practice 2009;18:16–20. 25. Wang HK, Lu K, Liang CL, Tsai YD, Wang KW, Liliang PC. Contralateral subdural effusion related to decompressive craniectomy performed in patients with severe traumatic brain injury. Injury 2012;43:594–597. 26. Jeon SW, Choi JH, Jang TW, Moon SM, Hwang HS, Jeong JH. Risk factors associated with subdural hygroma after decompressive craniectomy in patients with traumatic brain injury: A comparative study. Journal of the Korean Neurosurgical Society 2011;49: 355–358. 27. Su TM, Lee TH, Chen WF, Lee TC, Cheng CH. Contralateral acute epidural hematoma after decompressive surgery of acute subdural hematoma: Clinical features and outcome. Journal of Trauma 2008; 65:1298–1302.
10
Q. Yuan et al.
Brain Inj Downloaded from informahealthcare.com by Nyu Medical Center on 06/16/15 For personal use only.
28. Su FW, Ho JT, Wang HC. Acute contralateral subdural hygroma following craniectomy. Journal of Clinical Neurosciences 2008;15: 305–307. 29. Lee KS. The pathogenesis and clinical significance of traumatic subdural hygroma. Brain Injury 1998;12:595–603. 30. St John JN, Dila C. Traumatic subdural hygroma in adults. Neurosurgery 1981;9:621–626.
Brain Inj, Early Online: 1–10
31. Wetterling T, Demierre B, Rama B, Spoerri O. The clinical course of surgically treated posttraumatic subdural hygromas. Acta Neurochirurgica (Wien) 1986;83:99–104. 32. Kilincer C, Simsek O, Hamamcioglu MK, Hicdonmez T, Cobanoglu S. Contralateral subdural effusion after aneurysm surgery and decompressive craniectomy: Case report and review of the literature. Clinical Neurology & Neurosurgery 2005;107:412–416.
