Introduction

The term ?ubdural fluid (SDF) collection?describes excess fluid in the subdural space.1) SDF collection correlates with head trauma and frequently occurs after craniotomy for deep-seated brain tumor, aneurysm repair, and microvascular decompression procedures.4)7) Indeed, SDF collection following ruptured aneurysm repair is reported more often than are other craniotomy procedures.

SDF collection has been referred to as external hydrocephalus,2) benign subdural collection,9) and subdural hygroma.10) These terms may originate from an understanding of the natural radiologic course and reflect the variable causes of SDF collection. However, SDF collection following repair of a ruptured aneurysm is distinct from subdural hygroma.14) SDF collection following ruptured aneurysm repair may be caused by CSF malabsorption4) and/or incision in the arachnoid membrane. SDF collection following an aneurysm repair has a variable natural course. Some patients experience no SDF collection while others experience SDF collection or hydrocephalus. We therefore analyzed cases of SDF collection after aneurysm repairs and reviewed the mechanisms, in order to determine proper management plans for these patients.

Materials and Methods

From January 2007 to June 2009, 288 patients underwent microsurgical aneurysm repair for ruptured aneurysms in our institution. Our study participants were 97 patients (33.7%) who experienced impaired cerebrospinal fluid (CSF) circulation, such as SDF collection or hydrocephalus. We defined SDF collection in terms of thickness (distance between the inner table of the skull and the cerebral cortex) and density (identical to CSF density on brain computed tomography [CT]). We categorized the patients into 3 groups. Group A (N = 24) comprised patients who experienced spontaneously resolved postoperative SDF collection. Group B (N = 40) comprised patients who experienced postoperative SDF collection that evolved into internal hydrocephalus. Group C (N = 33) comprised patients who experienced hydrocephalus without subdural fluid collection during the postoperative period.

To determine the differences between the three groups, we retrospectively reviewed radiographic images and clinical data, using the Hunt and Hess grade (HHG) to classify the clinical severity of each patient? aneurysmal subarachnoid hemorrhage (SAH) on admission. We used Fisher grade (FG) to classify radiological severity on each initial CT scan.

Post-operative CT scans were performed immediately and 7 and 21 days post-surgery. When patients complained of symptoms of normal pressure hydrocephalus during outpatient follow-up, they received CT scans. We defined a ventricle-size index greater than 35% on CT as hydrocephalus.5) We compared among the three groups based on age, initial clinical grade, and Fisher grade.

For statistical comparisons, we used chi-square tests, considering differences with probability values less than 0.05 to be statistically significant. Data are presented as the mean ± the standard error of the mean.

Results

Table 1 presents patient age, sex, initial HHG, and FG. There were 52 female and 45 male patients. The mean ages of the study populations were 60.0 ± 2.8, 62.3 ± 2.31, and 61.4 ± 3.1 years in groups A, B, and C, respectively. Table 2 shows the differences between the three groups on age, initial clinical grade, and CT grade. The proportions of patients over 60 years of age were 50% in group A, 52.5% in group B, and 51.5% in group C. Regarding initial clinical grade, 27 patients had HHG scores I or II, whereas 70 patients had initial HHG scores of III, IV and V. Compared to group A, group B had a greater proportion of patients whose initial HHG scores were III, IV, and V (p = 0.040). Group C also had a greater proportion of patients with III, IV, and V initial HHG scores than group A did (p = 0.004). There were 13 patients who had FG scores of I and II, whereas 84 patients had FG scores of III and IV. Compared to group A, group B had a larger proportion of patients whose initial FG score was III and IV (p = 0.020). Group C also had a larger proportion of patients with initial FG scores of III and IV than group A did (p = 0.002).

Type of treatment for CSF circulation disturbance

Among group B patients, 25 (62.5%) underwent shunt procedures, and the rest received conservative management for their internal hydrocephalus. Of the rest, 16 (40%) patients underwent simple ventriculo-peritoneal (VP) shunt procedures, 6 (15%) underwent double-gradient shunt procedures (VP and subdural-peritoneal shunt together), and 2 (5%) underwent subdural-peritoneal (SP) shunt procedures, because they had SDF collection after VP shunt insertion. One (2.5%) patient received an SP shunt first, followed by a VP shunt in the follow-up period. All patients in group C underwent VP shunt or ventriculo-atrial (VA) shunt operations.

Case illustration

Case 1

A 67-year-old man was suffering from a sudden, bursting headache. His neurological status on admission was HHG II, and CT revealed SAH in the right perisylvian area (FG II) (Fig. 1a). CT angiography showed an aneurysm in the first branch of the right posterior cerebral artery (P1) (figure not shown). He underwent aneurysm repair surgery, but 21 days after his craniotomy, he developed gait disturbance. A brain CT scan revealed subdural fluid collection in the right frontotemporal region (Fig. 1b). He subsequently underwent an SP shunt operation. Two months after the SP shunt operation, he began to develop general weakness. A repeat brain CT showed hydrocephalus (Fig. 1c), for which he underwent a VP shunt operation. Nine months later, the ventricle was reduced in size (Fig. 1d).

Case 2

A 54-year-old woman experienced altered mentality. Her neurological status on admission was HHG III, and the CT revealed diffuse SAH (FG III) (Fig. 2a). CT angiography showed an anterior communicating artery (ACoA) aneurysm (figure not shown). This aneurysm was surgically repaired. She began to develop general weakness 3 months after her craniotomy. A brain CT revealed subdural fluid collection (Fig. 2b), for which she received conservative management, as we opted to ?ait and see.?A repeat CT scan 4 months after her craniotomy showed hydrocephalus (Fig. 2c). She subsequently underwent a VP shunt operation, using a medium pressure valve, for hydrocephalus. Nine months after her VP shunt operation, the ventricle size was smaller, but subdural fluid collection recurred on the contralateral side (Fig. 2d), for which she underwent SP shunt surgery, using a low-pressure valve. Finally, SDF collection was reduced 1 month following her SP shunt surgery (Fig. 2e).

Case 3

A 68-year-old woman was suffering from a sudden, bursting headache. Her neurological status on admission was HHG III, and CT revealed diffuse SAH (FG III) (Fig. 3a). CT angiography revealed a posterior communicating artery (PCoA) aneurysm (figure not shown), for which she underwent aneurysm repair surgery. She developed cognitive dysfunction 45 days after her craniotomy. A brain CT scan revealed subdural fluid collection and hydrocephalus (Fig. 3b), for which she underwent double-gradient shunt surgery. Twenty days after shunt surgery, CT showed a successful reversal of SFC and hydrocephalus (Fig. 3c).

Discussion

Subdural fluid collection is frequently encountered following surgery for a ruptured aneurysm. However, clinicians do not clearly understand its pathophysiology and natural course, and it has no known optimal management. Therefore, we attempted to clarify its natural course and determine its optimal management plan, by analyzing cases of SDF collection and reviewing the mechanism of SDF collection.

We considered hydrocephalus and SDF collection following surgery for ruptured aneurysm to be a disturbance of CSF circulation. We may explain the mechanism of disturbed CSF circulation by the ?ead space?theory,6)8) malabsorption of CSF,4) and the ?ne-way valve?theory.8)13) It has been thought that the subdural ?ead space?communicates with the major CSF cistern following craniotomy, and thus a SDF collection gradually develops. The aging process aggravates the dead space following craniotomy. Therefore, older patients with severe brain atrophy and decreased brain elastance are likely to have a larger dead space. Several study groups have supported the hypothesis that CSF circulation disturbance leads to SDF collection.4)8) First, blood clots from a subarachnoid hemorrhage lodge in the arachnoid granulation and induce fibrosis of the arachnoid granulation, which disturbs CSF absorption through the arachnoid granulation. Subsequently, hydrocephalus develops. Second, during aneurysm surgery, a subarachnoid membrane incision creates a communication between the cistern and the subdural space. Intra-operative CSF drainage and brain retraction also create an artificial subdural dead space. CSF in the ventricle and cistern subsequently move into the subdural dead space via pressure gradient.

In some cases, such as the previously mentioned Case 1, hydrocephalus develops after SP shunt surgery for SDF collection. Previous study groups8)13) described this phenomenon based on the ?ne-way valve?theory. If subarachnoid membranes act as one-way valves, then CSF accumulation in the subdural space causes compression of the brain parenchyma and the ventricle, rather than ventricular enlargement. As such, SDF collection increases. After placement of an SP shunt, narrowing and adhesions in the subdural and subarachnoid spaces lead to deterioration in CSF circulation, and, ultimately, to hydrocephalus (Fig. 4).

Age is correlated with disturbed CSF circulation. Tanaka et al.11) reported patient age was the most likely causative factor in SDF collection. In our series, patients over 60 years of age had a higher incidence of SDF collection than did patients under 40 years of age (result not shown). The reduced parenchymal elastance due to old age may be associated with SDF collection. A prior study3) suggested a correlation between brain stiffness and increasing age. Therefore, the brains of older patients are depressed immediately after craniotomy. They then have difficulty re-expanding, suggesting that older patients have a greater probability of having residual subdural space and producing SDF collection. However, age differences between the three groups were not statistically significant in our study.

In Cases 2 and 3, we may explain SDF collection following VP shunt for hydrocephalus and SDF collection combined with hydrocephalus in terms of cortical atrophy and compartmentalization of CSF flow. In elderly patients or patients with atrophied brains, the dura-arachnoid interface combined with adequate potential subdural space may separate after VP shunt surgery, due to reduced parenchymal elastance. Subdural space, intraventricular space, and subarachnoid space, including several cisterns, may have compartmentalized due to arachnoid adhesion in a patient who has undergone surgery for a ruptured aneurysm. For example, as presented in Case 2, one ventricle collapsed, and unilateral SDF collection occurred after the VP shunt. CSF flow obstruction between both sides of the CSF spaces caused this phenomenon. The pressure gradient between the subdural space and intraventricular space also contributed to this phenomenon. Therefore, to manage SDF collection in Case 2, we chose an SP shunt using a low-pressure valve. In Case 3, hydrocephalus was combined with SDF collection. We took the patient? age and potential pressure gradient between the subdural space and intraventricular space into consideration when building the management plan for Case 3. Finally, the SDF collection combined with hydrocephalus was resolved by a double gradient shunt.

The initial clinical and radiologic grades correlate with disturbed CSF circulation. A patient with poor neurologic status is at high risk for CSF circulation disturbance.12) Previous series12) have shown that patients with higher HHG have a greater need for post-operative shunting. In our series, after adjusting for age, patients with initial HHG scores of III, IV, and V had a higher incidence of hydrocephalus than did patients with initial HHG scores I and II (results not shown). Patients with initial FG scores III or IV had a higher tendency to develop hydrocephalus than did patients with initial FG scores I and II (results not shown).

If SDF collection develops, there are three possible clinical courses. First, the SDF collection resolves spontaneously. In this case, the patient does not need any further intervention. Second, the SDF collection persists without a change in volume. In such a case, the patient requires an SP shunt, which is effective and in wide use for treating intractable chronic subdural hematomas or hygromas. Third, SDF collection evolves into internal hydrocephalus. In such a case, the patient will require a VP shunt or a double-gradient shunt (VP shunt and SP shunt together). In this study, patients with higher HHG or FG scores were more likely to experience SDF collection evolving into internal hydrocephalus. We considered such patients for shunt operations (Fig. 5).

Conclusion

Subdural fluid collection following surgery for ruptured aneurysm is not merely caused by the tearing of arachnoid membrane. Instead, it may be understood in the context of disturbed CSF circulation. Based on our understanding of SDF collection? clinical course and mechanism, we carefully suggest that clinicians consider the presence of cortical atrophy, the initial CT grade, and the clinical grades when establishing a treatment plan for SDF collection. Unless symptoms of hydrocephalus develop, VP shunt should be deferred until and unless SDF collection evolves into internal hydrocephalus. In a high-grade SAH patient with an atrophied brain, when SDF collection is combined with hydrocephalus, we recommend the double gradient shunt.

References

1)   Amason O, Jakobsson KE, Lindgren S. Extracerebral nonhemorrhagic complications of intracranial surgery. Acta Neurochir 67:231-238, 1983

2)   Chapman PH. External hydrocephalus. Concept Pediatr Neurosurg 4:102-118, 1983

3)   Fukuhara T, Gotoh M, Asari S, Ohmoto T, Akioka T. The relationship between brain surface elastance and brain re-expansion after evacuation of chronic subdural hematoma..Surg Neurol 45:570-574, 1996

4)   Greitz D. Cerebrospinal fluid circulation and associated intracranial dynamics. A radiologic investigation using MR imaging and radionuclide cisternography. Acta Radiol Suppl 386:1-23, 1993

5)   Heinz ER, Ward A, Drayer BP, Dubois PJ. Distinction between obstructive and atrophic dilatation of ventricles in children. J Comput Assist Tomog 4(3) :320-325, 1980

6)   Koizumi H, Fukamachi A, Nukui H. Postoperative subdural fluid collection in  neurosurgery. Surg Neurol 27:147-153, 1987

7)   Naffziger HC. Subdural fluid accumulations following head injury. JAMA 82:1751-1752, 1924

8)   Park HS, Kim SH, Lee KC. Subdural fluid collection following ruptured aneurysm surgery as a manifestation of disturbed cerebrospinal fluid circulation. J Korean Neurosurg Soc 25 :769-777, 1996

9)   Robertson WC Jr, Chun RWM, Orrison WW, Sackett JF. Benign subdural collections of infancy. J Pediatr 94:382-385, 1980

10)   Stone JL, Lang RGR, Sugar O, Moody RA. Traumatic subdural hygroma. Neurosurgery 8:542-550, 1981

11)   Tanaka Y, Mizuno M, Kobayashi S, Sugita K. Subdural fluid collection following craniotomy. Surgical Neurol 27:353-356, 1987

12)   Vale FL, Bradley EL, Fisher WS. The relationship of subarachnoid hemorrhage and the need for postoperative shunting. J Neurosurg 86:462-466, 1997

13)   Yamada H, Hanada Y, Nagao T. Hydrocephalus after treatment of post-craniotomy subdural fluid collection in patient with subarachnoid hemorrhage. Current treatment for hydrocephalus 3:17-23, 1993

14)   Yuhei Y, Susumu W, Masaaki H. External hydrocephalus after aneurysm surgery: paradoxical response to ventricular shunting. J Neurosurg 88:485-489, 1998