Treatment and prognosis of subdural hematoma in patients with spontaneous intracranial hypotension

Abstract

Objective

The objective of this article is to elucidate the outcome, prognostic predictors and timing of surgical intervention for subdural hematoma (SDH) in patients with spontaneous intracranial hypotension (SIH).

Methods

Patients with SDH were identified retrospectively from 227 consecutive SIH patients. Data were collected on demographics, clinical courses, neuroimaging findings, and treatment of SDH, which was later divided into conservative treatment, epidural blood patches (EBP), and surgical intervention. Poor outcome was defined as severe neurological sequelae or death.

Results

Forty-five patients (20%) with SDH (mean maximal thickness 11.9 ± 6.2 mm) were recruited. All 15 patients with SDH <10 mm achieved good outcomes by either conservative treatment or EBP. Of 30 patients with SDH ≥10 mm, patients with uncal herniation (n = 3) had poor outcomes, even after emergent surgical evacuation (n = 2), compared to those without (n = 27) (100% vs. 0%, p < 0.001). Fourteen patients underwent surgical evacuation, resulting in good outcomes in all 12 who received early intervention and poor outcomes in the remaining two who received delayed intervention after Glasgow Coma Scale (GCS) score ≤8 (100% vs. 0%, p = 0.01).

Conclusions

Uncal herniation results in poor outcomes in patients with SIH complicated with SDH. In individuals with SDH ≥10 mm and decreased GCS scores, early surgical evacuation might prevent uncal herniation.

Keywords

Spontaneous intracranial hypotension subdural hematoma epidural blood patch increased intracranial pressure uncal herniation

Introduction

Spontaneous intracranial hypotension (SIH) is caused by occult spinal cerebrospinal fluid (CSF) leak and subsequent CSF hypovolemia, which leads to brain sag, downward traction on leptomeninges and neural structures, compensatory venous engorgement, and enlargement of the subdural/subarachnoid space. The persistence of negative pressure within the subdural space elicits subdural hygroma. In addition, tearing of bridging veins or bleeding from engorged veins in the subdural space produces subdural hematoma (SDH). SDH has been reported in 20% of SIH patients in two case series (1,2).

After proper treatment, the majority of patients with SIH are supposed to have a self-limited clinical course and satisfactory outcomes. However, some cases have been reported of morbidity or mortality resulting from SDH (3 –5). The optimal management of SDH associated with SIH remains to be determined. According to the guidelines for surgical management of acute traumatic SDH proposed by Bullock et al., hematomas with thickness ≥10 mm on computed-tomography (CT) scan should be surgically evacuated as soon as possible, regardless of the patient’s Glasgow Coma Scale (GCS) score (6). For growing SDHs caused by SIH, some authors recommend urgent neurosurgical intervention for patients with complications of clinical deterioration (7) and reduced levels of consciousness (8). However, others note that subdural fluid collections can be managed safely by directing treatment at the underlying CSF leak without hematoma evacuation (1), and some authors emphasize that craniotomy might increase the risk of brain herniation (9). The aim of this study was to clarify the prognostic predictors and timing of surgical intervention for SDH in patients with SIH.

Methods

Design and data source

We reviewed 246 consecutive patients who had a diagnosis of SIH in four hospitals in Taiwan, from January 1, 1998, to August 31, 2013. Patients were included in this study based on the International Classification of Headache Disorders, second edition (ICHD-2) (code 7.2.3), criteria (10): (a) orthostatic headache associated with at least one of the following: neck stiffness, tinnitus, hypacusia, photophobia, and nausea; (b) at least one of the following: 1) evidence of low pressure on magnetic resonance imaging (MRI) such as pachymeningeal enhancement; 2) evidence of CSF leak on conventional myelography, CT myelography or cisternography; and 3) CSF opening pressure <60 mm H₂O; and (c) no history of dural puncture or other causes of CSF fistula. Of note, criterion (d), headache resolves within 72 hours after epidural blood patch (EBP), was not required because not all patients needed EBP. We excluded 19 patients (8%), including patients whose SIH diagnosis could not be established (n = 9), patients who had been lost to follow-up within the three-month period (n = 7), and patients who had prior dural punctures (n = 2) and surgical intervention potentially inducing CSF fistula (n = 1). We reviewed all brain images of SIH patients and included patients with SDH revealed in at least one series of neuroimaging studies during the whole clinical course. This retrospective review of patient data was approved by the local ethics committee.

Assessments

SDH was diagnosed if the subdural fluid collection showed high or iso-density on CT or high signal intensity on both T1- and T2-weighted MRI. Patients with only non-hemorrhagic subdural hygroma were excluded.

We tabulated data with regards to clinical presentation, course and outcome of all patients with SIH and SDH. A new SDH episode was defined as the formation of a new hematoma or bleeding into a non-hemorrhagic subdural fluid collection confirmed by imaging studies. A recurrent SDH episode was defined as an acute expansion of a preexistent hematoma. To identify common warning signs of SDH, we assessed the preceding clinical manifestation in close temporal relation to all SDH episodes demonstrated on brain images. Although two or more warning symptoms might precede or co-occur with SDH, we documented the most prominent presentation leading to the discovery of SDH.

To investigate the prognostic predictors of SDH in patients with SIH, we analyzed the following parameters: age, gender, degree of consciousness impairment, maximal thickness of SDH, and treatment strategy. Consciousness impairment was defined as a patient's GCS score of <15. The maximal thickness of SDH was defined as the greatest measurement of SDH in any cranial image during the clinical course and was categorized into two groups: 1–9 and ≥10 mm. Midline shift was defined as the maximal measurement between the ideal midline and the real septum pellucidum. A significant midline shift was defined on brain CT/MRI as ≥5 mm. We classified the treatment in our patients complicated with SDH into three categories: 1) conservative treatment only; 2) injection of EBP; and 3) surgical intervention with or without EBP.

Short-term outcome, based on chart review, was defined at three months post-treatment. A good outcome was defined as total relief of orthostatic headache without neurological deficit; some patients with good outcomes could have had a residual mild non-orthostatic headache without recurrence of CSF hypovolemia. A poor outcome was defined as death or severe neurological sequelae.

Statistical analyses

SPSS version 18.0 was used for the statistical analysis. Descriptive statistics were presented as means ± standard deviation or percentages. For categorical data, Fisher’s exact tests or chi-square were used to test the difference between groups. All p values were two tailed, and statistical significance was defined as a p value of less than 0.05.

Results

A total of 227 patients met the inclusion criteria of SIH, including 40 cases reported in 2007 (2). Of these, 45 (20%) had at least one episode of SDH with mean age at disease onset of 42.5 ± 9.8 years (range 26–66) and male:female ratio of 1.6:1 (28:17). The mean maximal thickness of SDH was 11.9 ± 6.2 mm (range 1–28). Significant midline shift was noted in 8 patients (18%) (range 5–12 mm). Outcomes were good in 42 patients (93%) and poor in the remaining three patients (7%), including two cases of mortality and one with severe neurological sequelae.

Symptoms/signs of acute SDH

The 45 patients had 55 SDH episodes, and nine patients (20%) had recurrent SDH episodes. Of 55 SDH episodes, 51 (93%) were symptomatic. The most common symptom leading to brain image and discovery of SDH was refractory orthostatic headache that did not respond to treatment (n = 24, 44%). Other red flag signs of SDH were: 1) transformed headache pattern or a new persistent headache substituting for the orthostatic headache (n = 11, 20%); 2) acute alteration of consciousness (n = 9, 16%); and 3) an unexpected severe headache far worse than former headaches or even a thunderclap headache (n = 7, 13%). Only four SDH episodes (7%) were asymptomatic and revealed by routine follow-up of imaging studies. Of these, one reported paradoxical improvement of orthostatic headache in close temporal relation to the formation of bilateral SDH. Documented focal neurological signs were manifestation of brainstem compression and often associated with disturbance of consciousness. Isolated focal neurological signs were noted in only two patients (4%): one had unsteadiness and another had upward gaze limitation.

Treatment and outcomes

Figure 1 summarizes the treatment and outcome of the 45 patients with SIH and SDH. There was no predefined protocol of the priority between medical treatment and surgical intervention for SDH. A surgical intervention decision for SDH was made by consensus among the in-charge neurologist or neurosurgeon, the consultant neurosurgeon, and the patient/family. Two patients (4%) were misdiagnosed as spontaneous SDH initially (maximal thickness 14 and 23 mm, respectively) and did not receive treatment for CSF leaks until correct diagnosis of SIH was received. As a result of the delay of treatment for CSF leaks, they both had recurrent SDH that required a second surgical evacuation and did not improve until they received proper treatment of CSF leaks. Eight patients (18%) (mean SDH maximal thickness of 9.0 ± 5.9 mm, range 1–20) were treated successfully with conservative treatment only. Twenty-two patients (49%) (mean SDH maximal thickness of 9.7 ± 5.0 mm, range 2–20) had good outcomes with conservative treatment and subsequent EBP. Nine patients (20%) (mean SDH maximal thickness of 16.3 ± 5.7 mm, range 10–28) received early evacuation of SDH because they had SDH ≥10 mm and either acute consciousness impairment (n = 4) or refractory non-orthostatic headache indicating increased intracranial pressure (IICP) (n = 5), and all of these had good outcomes. Four patients (9%) (mean SDH maximal thickness of 16.8 ± 4.9 mm, range 14–24) had brain herniation and became comatose. Of these, one had subfalcian herniation during EBP and fully recovered after emergent surgical intervention. Three had uncal herniation and poor outcomes: two received emergent evacuation of SDH, but the remaining one did not. Of 30 patients with SDH ≥ 10 mm, all patients with uncal herniation (n = 3) had poor outcomes, compared to those without (n = 27) (100% vs. 0%, p < 0.001).

Figure 1.

Treatment and outcomes of the 45 SIH patients with SDH.

Prognostic predictors

Before patients suffered from SDH, none had alterations of consciousness caused by pure CSF hypovolemia. Patients with consciousness impairment had significantly greater mean maximal thickness of SDH (p < 0.001), a higher proportion of significant midline shift (p = 0.03) and urgent surgical intervention (p < 0.001), and higher morbidity/mortality (p = 0.02), compared to those without consciousness impairment (Table 1).

Table 1.

Association between level of consciousness and severity of SDH in patients with SIH.

	All (N = 45)	GCS = 15 (N = 32)	GCS < 15 (N = 13)	p value^a
Age (mean ± SD), years	42.5 ± 9.8	42.5 ± 10.7	42.6 ± 7.4	0.34
SDH maximal thickness (mean ± SD), mm	11.9 ± 6.2 (1–28)	9.9 ± 5.4 (1 –20)	16.8 ± 5.3 (10–28)	<0.001
Significant midline shift, N (%)	8 (18%)	3 (9%)	5 (38%)	0.03
Surgical intervention, N (%)	14 (31%)	4 (13%)	10 (77%)	<0.001
Morbidity/mortality, N (%)	3 (7%)	0 (0%)	3 (30%)	0.02

SDH: subdural hematoma; SIH: spontaneous intracranial hypotension; GCS: Glasgow Coma Scale score. ^aFisher’s exact test or chi-square test.

All of the patients with SDH < 10 mm maintained full level of consciousness and had good outcomes. Increased maximal thickness of SDH (≥10 mm) was significantly associated with a higher proportion of consciousness impairment (p = 0.01) and urgent surgical intervention (p = 0.001) (Table 2).

Table 2.

Association between maximal thickness of SDH and clinical severity in patients with SIH.

		Maximal thickness of SDH
	All (N = 45)	1–9 mm (N = 15)	≥10 mm (N = 30)	p value^a
Age (mean ± SD), years	42.5 ± 9.8	38.7 ± 9.3	44.4 ± 9.7	0.05
Male gender, N (%)	28 (62%)	5 (33%)	23 (77%)	0.008
Impaired consciousness, N (%)	13 (29%)	0 (0%)	13 (43%)	0.01
Significant midline shift, N (%)	8 (18%)	2 (13%)	6 (20%)	0.70
Surgical intervention, N (%)	14 (31%)	0 (0%)	14 (47%)	0.001
Morbidity/mortality, N (%)	3 (7%)	0 (0%)	3 (10%)	0.24

SDH: subdural hematoma; SIH: spontaneous intracranial hypotension. ^aFisher’s exact test or chi-square test.

Male gender was significantly associated with increased severity of SDH, involving greater mean maximal thickness (13.9 ± 6.0 mm vs. 8.5 ± 5.1 mm, p = 0.004) and more frequent emergent surgical intervention (43% vs. 6%, p = 0.015).

Timing of surgical intervention

None of the patients with SDH <10 mm required surgical intervention. We analyzed the differences between surgical and non-surgical patients with SDH ≥10 mm after excluding the single patient who did not receive surgery despite the recommendation of a neurosurgeon. Compared to non-surgery patients, surgery patients had higher frequencies of GCS scores <15 (71% vs. 13%, p = 0.003) and recurrence of SDH (50% vs. 13%, p = 0.05) (Table 3).

Table 3.

Comparison of surgery vs. non-surgery patients with SDH ≥10 mm.

	Surgery (N = 14)	Non-surgery (N = 15)	p value^a
Age (mean ± SD), years	43.1 ± 8.1	45.7 ± 11.4	0.48
Male predominance, N (%)	13 (93%)	9 (60%)	0.08
Maximal thickness of SDH (mean ± SD), mm	16.9 ± 5.4	13.5 ± 3.9	0.062
Recurrence of SDH, N (%)	7 (50%)	2 (13%)	0.05
Significant midline shift, N (%)	5 (29%)	1 (7%)	0.08
Impaired consciousness, N (%)	10 (71%)	2 (13%)	0.003

SDH: subdural hematoma. Definition of consciousness impairment is Glasgow Coma Scale score <15. ^aFisher’s exact test or chi-square test.

Discussion

Our study showed that SIH patients with SDH maximal thickness <10 mm had good outcomes without the need for surgical intervention, and the risk of neurological deterioration increased dramatically in patients with acute SDH maximal thickness ≥10 mm. Refractory orthostatic headache, change of headache characteristics, acute change in consciousness, and an unexpected severe headache were the most noticeable warning symptoms of IICP arising from a growing SDH. Surgical intervention of critically symptomatic SDH was not detrimental to patients with SIH, but was necessary to relieve life-threatening IICP and avoid uncal herniation. Patients who received early surgical intervention had good outcomes, including those with preoperative misdiagnosis of SIH and timely postoperative EBP injections. However, a negative outcome was irreversible once uncal herniation had initiated, even after an emergent evacuation of SDH. These data support adherence to the guidelines for surgical management of acute traumatic SDH (6) when it is caused by SIH.

Some authors oppose surgical intervention for SDH in SIH patients. Schievink et al. reported eight SIH patients with SDH (with thickness range 10–30 mm) who were successfully treated by non-surgical management and concluded that SDH could be safely managed by directing treatment at the underlying CSF leak without the need for hematoma evacuation (1). Some authors have reported ineffectiveness of surgery or even postoperative acute neurological worsening (4,11 –15) or brain herniation (9,16). These cases had the same histories: preoperative misdiagnosis of the primary cause of SDH, i.e. SIH. Another report showed an SIH patient who received three EBP treatments and two surgical evacuation of SDH, but did not improve until a thoracic CSF leak was surgically repaired (17). According to a review of the literature and our study, poor outcomes or ineffectiveness of the surgical intervention for SDH in SIH patients is often caused by: 1) preoperative misdiagnosis of underlying SIH; 2) delayed timing of surgical intervention; and/or 3) postoperative persistence of CSF leaks, rather than the surgery itself.

One study reported four patients with SIH complicated by large SDHs (8). Before successful surgical intervention of SDH, three of these patients received partial or ineffective treatment, and the remaining one had misdiagnosis of an underlying CSF leak. This case report revealed that incomplete treatment or misdiagnosis of underlying CSF leaks potentially contributed to the formation of SDHs, and surgical intervention did no harm as a rescue procedure. Several case reports of successful surgical treatment of SDH in patients with SIH who had acute consciousness changes also support that timely surgical intervention can stop further enlargement of SDH and avoid uncontrollable IICP and herniation (18 –20).

Although surgical intervention of SDH should be considered when maximal thickness of SDH is ≥10 mm, it is important to highlight that the severity of SDH in an individual patient does not depend solely on SDH maximal thickness. According to our study, the most useful sign is change of consciousness, which is suggestive of neurological deterioration. Despite equal thickness of SDH, some patients had no change of consciousness, and others suffered from dramatic neurological changes and even became comatose. The discrepancy between imaging findings and clinical presentations may result from different rates of SDH formation. Patients who have chronic SDH may manifest only mild neurological deficits (21). As a result of the slow enlargement of the hematoma, the period of spatial compensation is probably sufficient to neutralize the IICP (22). However, acute SDH may elevate the intracranial pressure dramatically because of insufficient time for spatial compensation of the brain. Mihara et al. reported extremely IICP and revealed an absence of correlation between the hematoma thickness and CSF pressure in patients with acute traumatic SDH with maximal thickness ≥10 mm (23). Ahn et al. demonstrated unchanged CSF pressure with a slow intrathecal infusion of artificial CSF and a CSF pressure boost after sequential bolus infusions in a patient with SIH (24).

In our study, most acute SDH episodes were symptomatic and could be discovered early by close monitoring for changes in headache profile and consciousness. We suggest that instant brain imaging follow-up is necessary if any warning symptoms occur.

This study is limited by its retrospective nature. Our study did not reveal the risk factor of recurrence of SDH in patients with SIH. According to the experiences of two misdiagnosed cases, delay of treatment for CSF leaks might be the major cause of recurrence of SDH. However, the case number might be too small to reach a conclusion. Further study is needed to explore this issue.

Conclusions

This study demonstrated the prognostic predictors of good and poor outcomes for SIH patients with SDH. Our proposed surgical indication, based on a clear-cut imaging finding (maximal thickness of SDH ≥10 mm) and easily detectable clinical manifestations (change of consciousness), could help to determine when surgical intervention is necessary to avoid irreversible neurological damage or even death. Further prospective studies are needed to validate the surgical indicators identified in this study.

Clinical implications

In patients with spontaneous intracranial hypotension (SIH), refractory orthostatic headache, transformed headache pattern, acute change in consciousness, and an unexpected severe headache are common warning symptoms of increased intracranial pressure (IICP) arising from a growing subdural hematoma (SDH).

Our study supports early surgical intervention for acute SDH caused by SIH, in the presence of SDH maximal thickness ≥10 mm and any change of consciousness, because a negative outcome is irreversible once uncal herniation induced by IICP has initiated, even after an emergent evacuation of SDH.

Poor outcomes or ineffectiveness of the surgical intervention for SDH in SIH patients are caused by preoperative misdiagnosis of underlying SIH, delayed timing of surgical intervention and/or postoperative persistence of cerebrospinal fluid leaks, rather than the surgery itself.

Footnotes

Funding

This work was supported by the Ministry of Science and Technology of Taiwan (NSC 101-2314-B-075-037-MY2, 101-2314-B-075 A-013, 102-2321-B-010-030), Taipei and Taichung Veterans General Hospital (V103E9-006, V103E3-005, V103C-083, TCVGH-1033403C), Ministry of Science and Technology support for the Centre for Dynamical Biomarkers and Translational Medicine, National Central University, Taiwan (NSC 102-2911-I-008-001), Brain Research Center, National Yang-Ming University and a grant from the Ministry of Education, Aim for the Top University Plan.

Conflict of interest

S.-J. Wang has served on the advisory boards of Allergan and Eli Lilly Taiwan. He has received speaking honoraria from local companies (Taiwan branches) of Pfizer, Eli Lilly and GSK. He has received research grants from the Taiwan National Science Council, Taipei-Veterans General Hospital and Taiwan Headache Society. J.-L. Fuh is a member of a scientific advisory board of Novartis, and has also received research support from the Taiwan National Science Council and Taipei Veterans General Hospital. The other authors have nothing to declared.

References

Schievink

Maya

Moser

, et al. Spectrum of subdural fluid collections in spontaneous intracranial hypotension. J Neurosurg 2005; 103: 608–613.

Lai

Fuh

Lirng

, et al. Subdural haematoma in patients with spontaneous intracranial hypotension. Cephalalgia 2007; 27: 133–138.

Chi

Wang

Lirng

, et al. Transtentorial herniation with cerebral infarction and duret haemorrhage in a patient with spontaneous intracranial hypotension. Cephalalgia 2007; 27: 279–282.

Schievink

. Stroke and death due to spontaneous intracranial hypotension. Neurocrit Care 2013; 18: 248–251.

Loya

Mindea

, et al. Intracranial hypotension producing reversible coma: A systematic review, including three new cases. J Neurosurg 2012; 117: 615–628.

Bullock

Chesnut

Ghajar

, et al. Surgical management of acute subdural hematomas. Neurosurgery 2006; 58(3 Suppl): S16–S24.

Mokri

. Spontaneous low pressure, low CSF volume headaches: Spontaneous CSF leaks. Headache 2013; 53: 1034–1053.

de Noronha

Sharrack

Hadjivassiliou

, et al. Subdural haematoma: A potentially serious consequence of spontaneous intracranial hypotension. J Neurol Neurosurg Psychiatry 2003; 74: 752–755.

Kelley

Johnson

. Sinking brain syndrome: Craniotomy can precipitate brainstem herniation in CSF hypovolemia. Neurology 2004; 62: 157–157.

10.

The International Classification of Headache Disorders: 2nd edition. Cephalalgia 2004; 24(Suppl 1): 9–160.

11.

Evan

Mokri

. Spontaneous intracranial hypotension resulting in coma. Headache 2002; 42: 159–160.

12.

Whiteley

Al-Shahi

Myles

, et al. Spontaneous intracranial hypotension causing confusion and coma: A headache for the neurologist and the neurosurgeon. Brit J Neurosurg 2003; 17: 456–458.

13.

Sayer

Bodelsson

Larsson

, et al. Spontaneous intracranial hypotension resulting in coma: Case report. Neurosurgery 2006; 59: E204–E204. discussion E204.

14.

Rahman

Bidari

Quisling

, et al. Spontaneous intracranial hypotension: Dilemmas in diagnosis. Neurosurgery 2011; 69: 4–14.

15.

Dhillon

Rabinstein

Wijdicks

. Coma from worsening spontaneous intracranial hypotension after subdural hematoma evacuation. Neurocrit Care 2010; 12: 390–394.

16.

Chotai

Kim

, et al. Brain herniation induced by drainage of subdural hematoma in spontaneous intracranial hypotension. Asian J Neurosurg 2013; 8: 112–115.

17.

Ghavanini

Scott

Chan

, et al. Management of patients with spontaneous intracranial hypotension causing altered level of consciousness: Report of two cases and review of literature. Cephalalgia 2013; 33: 43–51.

18.

Takeuchi

Takasato

Masaoka

, et al. Progressive subdural hematomas after epidural blood patch for spontaneous intracranial hypotension. J Anesth 2010; 24: 315–316.

19.

Mao

Dong

. Delayed subdural hematoma and cerebral venous thrombosis in a patient with spontaneous intracranial hypotension. Neurol Sci 2011; 32: 981–983.

20.

Liu

Lin

Cho

, et al. Burr hole drainage for complicated spontaneous intracranial hypotension syndrome. Kaohsiung J Med Sci 2008; 24: 498–502.

21.

Gelabert-González

Iglesias-Pais

García-Allut

, et al. Chronic subdural haematoma: Surgical treatment and outcome in 1000 cases. Clin Neurol Neurosurg 2005; 107: 223–239.

22.

Tanaka

Nakayama

Yoshinaga

. Cerebral blood flow and intracranial pressure in chronic subdural hematomas. Surg Neurol 1997; 47: 346–351.

23.

Mihara

Dohi

Nakamura

, et al. Novel method for emergency craniostomy for rapid control and monitoring of the intracranial pressure in severe acute subdural hematoma. Neurol Med Chir (Tokyo) 2010; 50: 1039–1044.

24.

Ahn

Berman

Dillon

. Spontaneous intracranial hypotension-hypovolemia associated with tacrolimus. Headache 2010; 50: 1386–1389.