Introduction

Arachnoid cysts (ACs) are cerebrospinal fluid collections surrounded by an arachnoid sheet believed to be of congenital origin. They are uncommon cystic lesions that represent 1% of all intracranial masses [1]. Nearly half of pediatric intracranial arachnoid cysts are located in the sylvian fissure/middle cranial fossa [2], and 60–80% of arachnoid cysts are discovered before the age of 15 years, mainly in the male population [3,4,5].

Although ACs usually remain stable in volume, they may enlarge with different reported mechanisms [6] causing symptoms related to direct compression of specific structures or intracranial hypertension.

They may also present with symptoms deriving from cyst rupture, usually causing intracystic hemorrhage and/or subdural hematoma or hygroma [7, 8]. Rupture is usually caused by minor head trauma, but spontaneous rupture may also happen [9].

Spontaneous rupture of middle fossa arachnoid cyst (MFAC) is a rare event; only 57 cases have been reported in literature (Table 1).

Table 1 Literature review. Patients are grouped according to treatment received
Full size table

Whenever to treat them, timing of treatment and procedure of choice are still a matter of debate. Different surgical treatments have been proposed for cysts that are enlarging in volume or which become symptomatic. These procedures mainly consist in cerebrospinal fluid (CSF) diversion to peritoneum or to the physiological CSF pathways. Direct excision of the cyst, cyst fenestration, and shunting operation has been reported [27, 49, 50]. Given the rarity of the pathology, only case reports and very small case series have been reported; thus, clinical presentations of these patients have been poorly studied, and there is still uncertainty regarding the indications, feasibility, efficacy, and safety of the different surgical procedures reported. Our main purposes are to describe clinical presentation of these patients and to help physicians to critically decide which treatment will be best for these patients with more awareness of feasibility and efficacy of the different surgical procedures. To do so, we here present our series of 17 patients surgically treated for spontaneously ruptured MFAC that, to the best of our knowledge, is the largest case series in literature. We have then discussed and compared our results with an extensive literature review in order to clarify the current state of the art of treatment for this rare pathology. We have reported and discussed clinical presentation, indications, contraindications, risks, and benefits of the different treatments that have been proposed.

Materials and methods

We have retrospectively analyzed all patients operated for MFAC from 2008 to 2018 at the Neurosurgery Unit of the Istituto Giannina Gaslini Children’s Hospital, Genoa, Italy. Inclusion criteria were presence on admission CT scan of MFAC with radiological signs of rupture defined as the presence of subdural hygroma or hematoma, intracystic hemorrhage, or intraparenchymal hematoma. Exclusion criteria were history of trauma in 3 months before presentation, aggregation, or coagulation dysfunctions either iatrogenic or pathological. We have reviewed demographic parameters, clinical data, radiological exams, pathology records, and follow-up documentation. Clinical presentation, surgical treatments, complication, and outcome at follow-up are reported. Volume of AC was described using Galassi’s classification of MFAC [51].

A literature review was performed in order to find all reported cases of spontaneous rupture of arachnoid cysts of the middle cranial fossa. The latter was performed using an online database search (Medline/Pubmed) using combinations of the terms “arachnoid cyst”, “subdural”, “haemorrhage”, “hemorrhage”, “haematoma”, “hematoma”, “hygroma”, “bleeding” and “rupture”. We have excluded all cases with a history of trauma in 3 months before presentation. We have also excluded from our literature review all patients over 70 years of age, the ones with reported dementia, with blood coagulation or platelet aggregation anomalies as in these cases, other possible causes for subdural bleeding were coexisting.

Results

From 2008 to 2018, 17 patients were surgically treated in our institution for arachnoid cysts of middle cranial fossa with radiological signs of rupture without history of trauma in 3 months before presentation (Table 2). No blood coagulation or platelet aggregation disorders were found. No genetic syndromes were reported in our series.

Table 2 Case series. Clinical and radiological presentation, treatment, and outcomes at follow-up
Full size table

Mean age at presentation was 8.3 years (range 3 months–14 years).

In two cases, bilateral cysts were found; thus, 17 patients presented 19 temporal arachnoid cysts. Fifty-three percent of them were classified as Galassi’s type II, 41% were type III, and 6% were type I.

All patients presented with concomitant chronic subdural hematoma and/or subdural hygroma. We did not find patients with radiological signs of intracystic bleeding without subdural collection nor with a different CT density between cystic and subdural fluid.

Two patients (11.7%) presented bilateral cysts, while bilateral subdural blood collections were present in 47% of patients. There were globally 25 subdural blood collections; 76% of these were hygromas, 12% were chronic hematomas, and 12% were subacute hematomas. No acute hematomas are reported in our series.

The most frequently reported symptom at presentation was headache, reported by 82% of patients. Headache was graded with the visual analogue scale (VAS), defined as mild with a pain rated from 1 to 5 and severe if the pain was rated from 6 to 10. In 93% of cases, it was reported being severe and only in 7% was reported as mild. These headaches were continuous in 36% of cases and intermittent in 64%. A fundus oculi examination was available in 15 out of 17 patients. Papilloedema was reported in only 33% of these patients. Nausea and vomiting were reported in 35.3% of patients. Consciousness alterations were reported in 2 patients (12%); both of these patients had papilloedema. No seizures were reported. There was only one completely asymptomatic patient in which the cyst and the associated subdural collection were discovered as an incidental finding.

Microsurgical fenestration of the cyst into the skull base cisterns was the procedure of choice in 76% of patients; in 2 cases, basal fenestration was judged too hazardous due to the finding of thick basal arachnoid, and only subdural fenestration was performed. We usually enter the cyst and check if the skull base arachnoid is clear enough to permit safe fenestration. If it is safely possible, we always aim to perform a basal fenestration; otherwise if it is judged too hazardous, we fenestrate the cyst only into the subdural space. We think that this procedure, when safely feasible, is the most effective in relieving symptoms and preventing re-rupture of the cyst.

We have performed shunting procedures in 2 patients in order to immediately decompress the cyst and fenestrate it in a second surgical procedure. In one case, bilateral Galassi type III cysts were first shunted into the peritoneum; then, in a second surgical time, the right cyst was endoscopically fenestrated into basal cisterns and in a third operation, the same procedure was attempted on the left side but not performed because of excessively high risks due to thick basal arachnoid. On the right side, the shunt was ligated and left in place; on the other side, it was kept functioning. The second patient was treated first with external cyst derivation and later with subdural fenestration because basal fenestration was not possible.

Only one patient of 3 months of age was treated with hematoma drainage via burr holes without approaching the cyst. We thought that in a small child, this was the safest procedure, avoiding risks of a more invasive surgery. At follow-up, the cysts disappeared and no further intervention was needed.

Interestingly, one patient presented with a spontaneous subacute hematoma, isodense to brain parenchyma, and was treated with burr holes in an urgent surgery setting; a postoperative CT scan revealed the presence of a Galassi type II temporal cyst that was then treated in a second surgical time with microsurgical fenestration into basal cisterns. We reported two surgical complications: one epidural hematoma that needed surgical evacuation and one CSF fistula that needed a revision surgery.

The average postoperative follow-up was 30.5 months. All patients experienced complete regression of symptoms at follow-up except for one patient that still complained of headaches and vomiting approximately once a month. Even if cyst volume reduction was not the purpose of treatment, reduction resulting in a lower postoperative Galassi classification type was reported in 37% of the cysts. Thirty-two percent had a slight visible reduction without lowering of the Galassi score, 16% were found to be stable at follow-up, and 2 bilateral cysts (10%) disappeared. Only 1 (5%) Galassi type 3 cyst showed slight increase in volume. The latter was anyway clinically asymptomatic at follow-up. No cases of hydrocephalus were reported at follow-up.

Discussion

Presentation

Arachnoid cysts of the middle cranial fossa account for 34% of all intracranial arachnoid cysts in adults [52] and 46% of all intracranial arachnoid cysts in children [2]. Although the majority of AC remain stable during life, natural history can quickly change by rupture of the cyst, usually defined by development of intracystic hemorrhage or subdural hematomas or hygromas. Minor head trauma is a known risk factor for AC rupture. Even if not usual, spontaneous rupture can also happen.

We have found 57 cases of spontaneous MFAC rupture reported in the literature (Table 1); 72.7% were males and 27.3% were females. Mean age was 16.7 years (range 1–47 years), higher than the mean age of our patients of 8.3 years (range 3 months–14 years) (Table 3).

Table 3 Result comparison between our series and literature
Full size table

Arachnoid cyst volume is considered a risk factor for rupture, especially those cysts larger than 5 cm in maximal diameter which have a higher risk of rupture [53]. This latter finding seems not to be confirmed by our series. We have divided patients according to Galassi’s classification (Table 2) of MFAC and compared our results with those of a large series of unruptured symptomatic MFAC [54]. In our series, 6% were type 1, 53% were type 2, and 41% were type 3, while in the reference article, they were 15% type 1, 54% type 2, and 31% type 3. We may hypothesize that type 1 MFAC is less prone to spontaneous rupture, while type 2 cysts have a slightly higher risk. Unfortunately, these data were not comparable with the data from our literature review because we found that AC volumes were rarely reported.

The most frequent MFAC radiological sign of rupture is a subdural collection; although rare, a purely intracystic bleeding is possible. While no patients from our series presented with purely intracystic hemorrhage, we found two such patients (3.5%) reported in the literature [14, 39].

In our series, the subdural collection was most frequently a hygroma while in literature (where patient’s mean age was higher than in our series), the most frequent finding was a chronic hematoma. A subacute subdural hematoma as MFAC rupture presentation is possible; conversely acute spontaneous subdural hematoma seems to be exceptional as we have found only one reported case [13].

Ruptured MFAC is more frequently found on the left side with a 2:1 ratio.

Interestingly, all patients with bilateral subdural collection had the same collection content on both sides except one patient [13] in whom there was a hygroma homolateral to the cyst and a contralateral subacute hematoma. This might be explained by a rebleeding inside the contralateral collection.

Clinical findings at presentation were unfortunately available in only 47 out of 57 patients from our literature review.

In literature, 95.9% of patients reported headache as presenting symptom; in our series, only 82% had headache at presentation. It is of note that 2 of the 3 patients that did not report headache in our series were under 1 year of age, so that the still opened skull sutures allowed enlargement of the skull compensating the raised intracranial pressure. This is confirmed by the finding of macrocrania reported in both of these patients. In agreement with this explanation, 1 of the 2 patients from the literature that did not report headache was 2 years old [17] and had macrocrania at presentation that was likely caused by the raised intracranial pressure. At presentation, he complained of nausea and vomiting. Given the fact that he had a chronic subdural hematoma, we suppose that AC rupture happened at a time when sutures were still opened leading to an increase of skull circumference; conversely by the time of presentation, cranial sutures were closed so that we may hypothesize that rebleeding or expansion of the subdural hematoma led to clinical symptom development. He was probably too young to be able to report headache, and nausea and vomiting were the only visible symptoms of intracranial hypertension.

Although we do not know from our literature review how many patients underwent a fundus oculi examination, papilloedema was reported in 11 patients (19.6%) [12, 13, 16, 18, 19, 26, 32, 35, 43, 44]. In our series, 33% of the patients who underwent a fundus oculi examination had papilloedema. Even if we do not have direct data regarding intracranial pressure neither from our experience nor from literature, we may suppose that headache in these patients is caused by meningeal irritation by blood instead of being a symptom of intracranial hypertension, because of the presence of papilloedema in only a minority of cases.

We did not find any correlation between papilloedema and the presence of nausea and vomiting neither in literature nor in our results. Thus, we hypothesize that in a substantial number of patients, these are signs of meningeal irritation as well.

Neurological deficits at presentation are usually found in a minority of patients with ruptured MFAC; in the literature, we found that the most frequently reported ones were hemiparesis, present in 8.2% of cases and ocular palsy in 10.2%. In our series, neurological deficits were even less frequently reported, possibly because of the younger age of our patients.

Treatment

Consensus regarding the procedure of choice in ruptured MFAC is lacking. Different surgical treatments were proposed that may address the cyst itself or only the subdural collection. Subdural fluid may be drained via single or multiple burr holes or craniotomy. Arachnoid cyst itself is treated with direct resection of the cyst, with shunting procedures into the peritoneum or into subcutaneously implanted reservoirs or with fenestration of cyst walls either into the subdural space or into the skull base cisterns.

Conservative treatment with follow up alone or with medical therapy with acetazolamide or steroids has also been proposed. Spontaneous resolution of MFAC has been reported for unruptured cysts since 1985 [55]. We have found 9 papers reporting conservative treatment for ruptured MFAC [7, 10,11,12,13,14,15,16,17]. Good clinical outcomes are reported but long-term follow-up is lacking in most of these reports. One of these cases needed surgical fenestration due to enlargement of subdural hygroma with midline shift after 2 months of observation along with acetazolamide therapy [16].

Surgery may address only the subdural collection which is usually thought to be the cause of brain compression leading to symptoms. Burr holes or craniotomy may be used to evacuate the hematoma. This treatment is a less invasive option with shorter surgical time, but it does not resolve the cyst itself that may remain excluded from CSF circulation and therefore carries the risk rupture recurrence. We have generally found good outcomes reported except for few cases that reported persistent symptoms at follow-up even if improved (Table 1).

Cysts may be decompressed by means of shunting procedures that may aim to create a shunt directly between cyst and peritoneum or between subdural space and peritoneum. These procedures allow immediate decompression of the subdural collection preventing at the same time future complications deriving from re-rupture or cyst expansion. Disadvantages are risks of shunt failure and lifelong shunt dependence.

Treatment may directly address the cyst by means of fenestration of the capsule decompressing the cyst in order to reduce mass effect and risk of rupture. Fenestration may be performed either into the subdural space or into the skull base cisterns. In these procedures, the subdural collection is encountered on the surgical route and therefore drained.

All the patients from our series experienced complete symptom relief at follow-up, except for the single one that still complained of rare headaches (Table 2).

Microsurgical fenestration of the cyst has been reported in literature (Table 1) even if technical limitations due to patient anatomical differences have been scantly described. Regarding surgical fenestration of ruptured MFAC, we have found good outcomes reported in the literature (Table 1). These data however lack of a consistent follow-up period (mean 0.5 months for subdural fenestration procedures and 7.2 months for basal fenestration procedures). All cases for whom follow-up data were available showed full recovery without procedure-related complications or rupture recurrence. Our series, considering only microsurgical basal cyst fenestration procedures, with a mean follow-up of 29.9 months, confirms the effectiveness and safety of these techniques adding an adequate postoperative follow-up time to the already reported cases. Even if the current best level of evidence regarding the treatment of choice for these patients is still low, based on our results and our literature review, we think that microsurgical fenestration of the cyst in the skull base cisterns, when safely feasible, is an efficient and safe procedure that addresses both the acute compression and the risk of rupture recurrence. This procedure also avoids the risks related to implantation of permanent devices that is needed in shunting procedures.

Due to the retrospective nature of our study and the rarity of the pathology, our recommendation is still based on weak evidences and future prospective, and more targeted studies are needed. Current evidences due to the limited reported follow-up do not seem to clearly show that conservative treatment is a good option for these patients; thus, at today, we tend to discourage this approach. Further studies with better evidences and longer follow-up times are mandatory to evaluate this option.

Conclusions

Spontaneous rupture of MFAC is a rare but possible event. Development of a subdural collection is the most frequent radiological sign of rupture. Headache and nausea are the most frequent symptoms at presentation; focal neurological deficits are relatively rare but possible.

From our experience and from the presented literature review, we think that spontaneously ruptured MFAC will benefit from operative rather than conservative treatment. Of all the surgical procedures reported, we suggest that microsurgical fenestration of the cyst into the skull base cisterns, when safely feasible, is the most effective and safe procedure addressing both the acute compression and the risk of rupture recurrence.

Limitations

Patients from our series have been treated according to surgeon’s preference based on the single cases; the absence of standardization of treatment may limit the comparability of our results.

Even if to our knowledge this is the largest reported series of patients treated for spontaneous rupture of MFAC, due to the rarity of the pathology, numbers are still relatively small.

Given the retrospective nature of our study, we think that this topic needs future prospective randomized trials investigating the best treatment for these patients.