Neurol India Home 

Year : 2022  |  Volume : 70  |  Issue : 4  |  Page : 1629--1634

Epidural Fluid Collections After Cranioplasty

Dimble Raju, Rajesh Bhosle, Prasad Krishnan 
 Department of Neurosurgery, National Neurosciences Centre, Kolkata, West Bengal, India

Correspondence Address:
Prasad Krishnan
Department of Neurosurgery, National Neurosciences Centre, Peerless Hospital Campus. 2nd Floor, 360 Panchasayar, Garia, Kolkata - 700 094, West Bengal


Background: Common complications following cranioplasty (CP) include infections, seizures, bone flap resorption, and intra-cranial hemorrhages. Epidural fluid collections (EFCs), often seen in the immediate post-operative scan as hypo-dense accumulations below the bone flap, have been very infrequently discussed in the literature as in the majority of the cases, they are small, get resorbed spontaneously, and usually do not cause neurological deficits. Objective: To document our experience with EFCs that needed re-operation and analyze their clinical and radiological findings. Materials and Methods: We describe a series of six cases of symptomatic EFCs following CP that necessitated re-operation in a series of 89 cases over 7 years. Conclusions: EFCs following CP have a different pathogenetic mechanism compared to post-operative epidural hemorrhages. Meticulous surgical techniques can reduce their incidence. Symptomatic EFCs can be evacuated by either re-opening the flap or placing burr holes in the replaced bone. EFCs may become symptomatic even a few days after CP.

How to cite this article:
Raju D, Bhosle R, Krishnan P. Epidural Fluid Collections After Cranioplasty.Neurol India 2022;70:1629-1634

How to cite this URL:
Raju D, Bhosle R, Krishnan P. Epidural Fluid Collections After Cranioplasty. Neurol India [serial online] 2022 [cited 2022 Dec 8 ];70:1629-1634
Available from:

Full Text

Decompressive craniectomy (DC) is a commonly performed operation to relive raised intra-cranial pressure, and patients who survive undergo interval cranioplasty[1] with the aim of providing protection to the brain as well as improving cosmesis and brain functions. Complications following cranioplasty (CP) commonly reported in large series include infections, seizures, intra-cranial hemorrhages (epidural, sub-dural and intra-parenchymal) and bone flap resorptions and are reported to occur in up to 40% of cases.[2]

The entity of post-operative epidural fluid collections (EFCs) has been infrequently addressed in the literature.[3],[4],[5],[6] A systematic review and meta-analysis of complications of CP Shepetovskey[2] reported that epidural extra-axial collections as a separate complication were reported in only three series. Kim et al.[4] have defined EFCs as the radiological presence of “low density fluid in the epidural space on brain CT scan” after CP in an attempt to differentiate it from a post-operative epidural hematoma.

 Case Series

We analyzed the records and images of 89 consecutive cases of CP performed by the corresponding author at our institution between January 2015 and December 2021 – a period of 7 years – with the primary objective of finding out how many of these patients had developed EFCs and also to audit how many of these required a second procedure to treat the same. Of the 89 patients, the stored autologous bone was used in 68 cases, polymethyl methacrylate (PMMA) bone cement was used in 16 cases, and titanium mesh was used in five cases to perform the CP. In all our cases, a sub-galeal drain was kept prior to closure, and in those patients who were on anti-platelets or anti-coagulants, these medications were stopped at least 5 days before the surgery and the procedure was carried out after ensuring that hematological parameters such as prothrombin time, activated partial thromboplastin time, and platelet count were normal on the day before operation. In all patients, post-operative computed tomography (CT) scans were performed on the first post-operative day (POD) and later as required. We defined EFCs as the presence of hypo-dense fluid collections below the replaced flap in more than 2 sections 5 mm apart. By these criteria, 48 patients (53.9%) had EFC. Forty-two of them were asymptomatic and did not need any surgical intervention. In six patients (12.5% of those with EFCs and 6.7% of all CP), there was a significant collection below the flap [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6] that caused symptoms and needed re-operation. One of these patients had a ventriculoperitoneal (VP) shunt in situ, and in one case, a lumbar puncture was performed to slacken a bulging flap to aid in placement of the bone. Two patients had significantly concave scalp defects pre-operatively even without any cerebrospinal fluid (CSF) drainage. Three patients were noted to have contralateral hemiparesis on the day after CP (i.e., POD 1), and of them, two underwent re-operation the same day. In one patient, we waited for spontaneous improvement, but as the deficit remained static over the next 2 days, he was re-operated on POD3. In three other cases, progressive neurological deterioration was noted from POD 2, and although two of them underwent re-surgery on POD3, in one case, surgery was performed on POD4.{Figure 1}{Figure 2}{Figure 3}{Figure 4}{Figure 5}{Figure 6}

Three of these patients were treated by re-craniotomy and evacuation with bone flap replacement, two underwent burr hole and evacuation by placement of a ventricular drain in the epidural space, and in one case, the bone flap was removed and re-cranioplasty was performed at a later date. The details of these patients are given in [Table 1].{Table 1}


The true incidence of EFCs following CP is unclear as definitions of EFCs vary and many series report only symptomatic collections [Table 2]. Lee et al.[3] have described an overall rate of 37.3% (22 out of 59 cases), and although they have included all cases where any hypo-density is found in post-operative CT scans, regardless of the amount, it is still less than our series which considers a case as having an EFC only if it seen in more than 2 sections 5 mm apart. Kim et al.[4] found EFCs in 49 out of 117 patients (41.8%) undergoing CP, whereas Jeong et al.[5] found it in 79.2% of their cases (65 out of 82 cases undergoing CP).{Table 2}

Several reasons have been speculated for the formation of EFCs following CP. These include leak of CSF because of violation of the neo-dura covering the brain during either elevation of the skin or placement of hitch sutures,[3],[5] stiffness of the neo-dura because of calcification[3] (probably because it may prevent its expansion with the resultant formation of a dead space between the dura and replaced bone), foreign body reaction to the hydroxyapatite material used to create an artificial bone,[7] and (counter-intuitively) using an autologous bone rather than bone cement[4] for cranioplasty, large bone flaps[5] (presumably because excessive dissection might lead to pooling of the tissue fluid from the dissected site), the absence of sub-galeal drain,[6] and intra-cranial hypotension because of the presence of VP shunt[4] where the sunken brain might not expand to abut the replaced bone.

Post-operative CT scans in three of our six patients who needed re-operation showed the presence of an “air bubble” below the replaced flap. Lee et al.[3] and Kim et al.[4] have shown the presence of “air bubble” in the post-operative scans in their series to correlate significantly with the presence of EFCs.

In our series, none of the patients who were re-operated showed any evidence of CSF leak, nor did any of them have significant dural calcifications. The autologous bone had been used for CP in four cases, and PMMA bone cement had been used in two cases. There were three patients with concave flaps prior to CP, of which one was following a VP shunt. One patient with a full flap underwent lumbar puncture (LP) after induction of anesthesia to slacken it and aid better seating of the bone. By definition, these collections are supposed to be non-infectious and non-hemorrhagic,[2] and in all our patients too, during re-operation for removal of EFCs, a watery fluid tinged with blood was expressed. No obvious bleeding point in the scalp or neo-dura was demonstrable in cases that underwent re-craniotomy.

A male preponderance has been reported in the formation of EFCs after cranioplasty by Lee et al.[3] All our patients in this series were males too. It is possible that this finding may be because of the fact that the most common cause of de-compressive craniotomies in our institution is resultant to trauma which has a definite male preponderance.

In our series, the time interval between DC and CP in patients that needed re-operations because of EFCs was 5.7 months on average. Kim et al.[4] speculate that early cranioplasty (when the brain is relatively more swollen) may have a lesser incidence of this complication. One of our patients had a VP shunt, and another underwent an LP at the time of cranioplasty. It has been reported[5] that EFC incidence is higher in patients with shunts than in those without shunts (81.8% against 42.1%).

Chang et al.[6] report a lower incidence (2.4%) of EFCs in patients with drains compared to those without drains (8.6%). All our patients as well as those in the two other series[3],[4] had sub-galeal drains placed at the time of CP closure, and hence, although this may decrease the incidence, it is not completely protective. Nguyen et al.[8] have argued for fenestrating the bone flap to prevent extra-axial collections. Although we have not made multiple holes in any of the flaps we replaced, we have never seen this complication in the few cases where we have used titanium mesh for CP, suggesting that this technique may have some protective value.

However, the natural history of an EFC is usually of regression and disappearance,[3],[5] and the same happened in 42 of our 48 patients (87.5%) too. Lee et al.[3] have remarked that although small EFCs are common, it is rare for an EFC to have enough volume to produce a detectable change of neurological status. In the lack of any existing guidelines, we had arbitrarily chosen the presence of extra-axial hypo-density in CT scans in more than 2 cuts 5 mm apart as a criterion for labeling the case as having an EFC, thereby hoping to draw a line between a mere radiological finding and something that might be considered a “complication”.

A noteworthy finding in our series is that not all patients had neurological deterioration on the day following CP itself. This was found in three patients only, and although two of them underwent re-surgery on the same day (POD1), in one case, re-operation was performed as the deficit remained static and did not improve over the next 48 hours. With this limited series, we cannot say if a further trial of “watchful expectancy” might be beneficial to patients who have static protean deficits which may be expected to resolve with time. In three other cases, although collection was noted on CT scan performed on POD 1, neurological deterioration started from POD 2 and was progressive. This would imply that fluid accumulation below the replaced flap may be an ongoing process in some patients and that discharge must be deferred for a few days if significant EFCs are found even in the absence of immediate post-operative deficits.


There are only a few series reporting EFCs following CP. Most EFCs are small, get re-absorbed spontaneously, and cause no deficits, and in such cases, it can be debated whether to consider them as complications or mere radiological findings. However, apart from the fact that they are hypo-dense on imaging and found to be serohemorrhagic on evacuation, they appear to have a distinctly different pathogenetic mechanism from post-operative epidural hemorrhages as their origin is not attributable to bleeding from the scalp or neo-dura, and in several instances, mere catheter drainage via a burr hole on the flap may suffice. As they have been described as a potential cause of post-operative surgical site infection,[9] every attempt must be made to prevent EFCs. Placement of a sub-galeal drain with holes near the temporobasal gap of the replaced flap, avoidance of unnecessary dissection of the temporalis muscle, care to prevent CSF leak by breeching neo-dura, and placement of hitches through the replaced bone during CP (particularly when the flap is concave or in shunted patients) may decrease the incidence of EFCs. Finally, it must be borne in mind that EFCs may become symptomatic and cause neurological deficits at some interval after CP and cases with significant collections on imaging must be observed carefully even if they are asymptomatic in the immediate aftermath of a CP.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

1Krishnan P, Bhattacharyya AK, Sil K, De R. Bone flap preservation after decompressive craniectomy- Experience with 55 cases. Neurol India 2006;54:291-2.
2Shepetovsky D, Mezzini G, Magrassi L. Complications of cranioplasty in relationship to traumatic brain injury: A systematic review and meta-analysis. Neurosurg Rev 2021;44:3125-42.
3Lee JW, Kim JH, Kang HI, Moon BG, Lee SJ, Kim JS. Epidural fluid collection after cranioplasty: fate and predictive factors. J Korean Neurosurg Soc 2011;50:231-4.
4Kim SP, Kang DS, Cheong JH, Kim JH, Song KY, Kong MH. Clinical analysis of epidural fluid collection as a complication after cranioplasty. J Korean Neurosurg Soc 2014;56:410-8.
5Jeong SH, Wang US, Kim SW, Ha SW, Kim JK. Symptomatic epidural fluid collection following cranioplasty after decompressive craniectomy for traumatic brain injury. Korean J Neurotrauma 2016;12:6-10.
6Chang V, Hartzfeld P, Langlois M, Mahmood A, Seyfried D. Outcomes of cranial repair after craniectomy. J Neurosurg 2010;112:1120-4.
7Imaizumi S, Owada K, Onuma T, Kamii H, Nakajima T. Epidural fluid collection after cranioplasty using hydroxyapatite ceremics following bone cement Jpn J Neurosurg (Tokyo) 2000;9:44-7.
8Nguyen HS, Doan N, Wolfla C, Pollock G. Fenestration of bone flap during interval autologous cranioplasty. Surg Neurol Int 2015;6:190. doi: 10.4103/2152-7806.172535.
9Kim MJ, Lee HB, Ha SK, Lim DJ, Kim SD. Predictive factors of surgical site infection following cranioplasty: A study including 3d printed implants. Front Neurol 2021;12:745575. doi: 10.3389/fneur. 2021.745575.