Intracranial arachnoid cysts are a common finding among the pediatric population. Their reported prevalence (2.6%) has been increasing with the mounting availability and frequency of intracranial imaging.1 Most arachnoid cysts are considered congenital, but there are examples of acquired arachnoid cysts in the literature.2,3 The majority of these radiological findings are incidental, and their exact clinical significance is not known; however, most are thought to be of little medical consequence.4-6 Very little literature exists examining or quantifying future risks for affected children.
Despite their relatively innocuous natural history, in rare circumstances, pediatric arachnoid cysts have been reported to rupture and present with signs and symptoms of raised intracranial pressure.7-14 We define a rupture as an overlying subdural hygroma or an associated hemorrhage either in the overlying subdural space or within the cyst itself. Estimates of rupture rates of arachnoid cysts have been reported to be between 2.3% and 4.6%.7,15 These rupture rates are, by nature, overestimates, because the denominators used represent only known arachnoid cysts, which are presumably only a fraction of the true population prevalence. Risk factors for rupture have yet to be conclusively established, but recent head injury (even minor),3,12,13,16 as well as the size and location of the arachnoid cysts,17 have been suspected.
The objective of our study was to determine whether an association exists between the rupture of arachnoid cysts and cyst size, recent head injury, or altitude at residence. A case-control design was chosen to study this clinical question, because ruptured arachnoid cysts are rare and a comparative control group is available in terms of children with unruptured cysts.
PATIENTS AND METHODS
Study Design and Setting
This case-control study included children evaluated at our level I pediatric trauma center from September 1, 2005, through July 31, 2010. Cases and controls were identified through a comprehensive review of 2 surgical databases and a medical imaging billing database. An approval from the institutional review board was obtained with a waiver of consent.
Definition of Cases
Enrolled patients included children (defined as age <18 years) with intracranial computed tomography (CT) or magnetic resonance (MR) imaging identifying an arachnoid cyst. A child was classified as a case if the initial presentation of his or her arachnoid cyst was a cyst rupture; ruptures were defined as an intracystic hemorrhage or an overlying subdural hematoma or hygroma (Figure 1). After surgical fenestration of an arachnoid cyst, patients almost universally are left with an overlying subdural hygroma and sometimes a small hematoma; therefore, children with previous surgery for their arachnoid cyst were excluded.
Definition of Controls
A child was classified as a control if the initial presentation was an unruptured arachnoid cyst discovered via intracranial imaging with or without clinical symptoms. Each identified case was matched to 2 patients in the control pool based on age, sex, and side and anatomical location of the arachnoid cyst. The anatomical location was defined as either middle temporal fossa, over the convexity, within the posterior fossa, or “other location.” If multiple controls were appropriate for a single case, the 2 controls for the study were selected via a random drawing.
Clinical risk factors were evaluated from a retrospective chart review. Twenty percent of the abstracted chart data were collected in duplicate by an alternate abstractor to confirm and report abstraction accuracy. The 2 major clinical risk factors at study were preceding head injury and the altitude at place of residence. Preceding head injury was defined as a head injury within 30 days of the intracranial imaging incurred through any of the following mechanisms: fall from less than 1 story or other minor blunt trauma, fall from 1 story or greater, a motor vehicle accident, or contact sports. The altitude at place of residence was generated by using the zip code at home residence and referencing the online website http://www.wundergroud.com, which reflects the US topography index to determine the average elevation in feet above sea level for that zip code.
Race and ethnicity, age, sex, whether the subjects presented with a headache or signs and symptoms of raised intracranial pressure, and the duration of these symptoms in days were also collected from the patients' charts. Similarly, specific comorbidities that have been potentially linked with arachnoid cysts and their rupture were also collected for each patient. These include cerebral palsy, seizure disorder, coagulopathy, and hydrocephalus.
Radiological Factors at Study
Radiological factors were abstracted by a pediatric neurosurgeon. The primary risk factor at study was arachnoid cyst size, which was defined as the maximal diameter of the cyst in millimeters in any plane on either CT or MR imaging. Among the cases, we also recorded whether the imaging demonstrated a hygroma, hematoma, or intracystic hemorrhage or a combination thereof and whether the subdural hygromas and hematomas were unilateral or bilateral. The frontal and occipital horn ratio was also collected for each patient as a validated measure of ventricular size.18,19
This study used a case-control design, and, therefore, only univariate statistics were used. All nonprimary continuous and categorical variables were compared by using the Student t tests and χ2 tests and, when appropriate, the Fisher exact tests, respectively.
The Cochran-Mantel-Haenszel test was used to assess the significance of association of preceding head injury with the primary outcome of arachnoid cyst rupture, controlling for the matching of cases and controls. Continuous variables, including arachnoid cyst size and altitude at residence, were assessed for association with the primary outcome by using a linear mixed model with a term for case vs control status, and a random effect to model correlation between matched cases and controls. Statistical significance was defined as P < .05. Odds ratios were then calculated to quantify the association between arachnoid cyst rupture and both preceding head injury and arachnoid cyst size, after dichotomizing cyst size into <50 mm and ≥50 mm in maximal diameter. The 50-mm cutoff was chosen through an iterative process, because it represented the optimal balance between sensitivity and specificity. These odds ratios and confidence intervals were estimated by using exact conditional logistic regression models that accounted for the modest sample sizes and controlled for the case-control matching. All statistical analyses were performed with the use of SAS version 9.2 (SAS Institute Inc).
Case and Control Identification
A search using our medical imaging billing database and 2 separate pediatric neurosurgery operative databases determined that 16 597 children patients had either CT or MR imaging of the brain between September 1, 2005, and July 31, 2010 (Figure 2). Of these, 697 patients had an International Classification of Diseases, Ninth Revision diagnosis code of 348.0 (cerebral cysts excluding porencephaly [congenital]). This diagnostic code includes arachnoid cysts, porencephalic cysts, porencephaly (acquired), and pseudoporencephaly. Patients with only porencephaly or porencephalic cysts and no arachnoid cysts were then excluded by a direct review of all available imaging. We found 309 patients with treated and untreated arachnoid cysts for an institutional prevalence of 1.9%. Next, we excluded those patients who had surgical intervention (a craniotomy or shunt or drain placement) for their arachnoid cysts before September 1, 2005, because postoperative imaging in these cases frequently demonstrates hemorrhages and/or hygromas that represent surgical changes and not spontaneous arachnoid cyst rupture. After these exclusions, 232 children with previously untreated arachnoid cysts on intracranial imaging remained in the final cohort.
Double-chart data collection was conducted on 9 patients (21.4%) representing 153 data points. One major (incorrect head injury history) and one minor error (type of surgery) were detected and corrected for an error rate of 1.3%.
Arachnoid cyst rupture or hemorrhage was identified in 14 (6.0%) of these patients, including 6 with subdural hygromas only (2 bilateral, 4 ipsilateral), 7 with subdural hematomas only (1 bilateral, 6 ipsilateral), and 1 patient with both. In the patient with both a hygroma and subdural hematoma, both were ipsilateral to the cyst. The remaining 218 children with unruptured arachnoid cysts represented the patient pool for the control group selection process.
Each of the 14 cases of arachnoid cyst rupture were matched to 2 unruptured controls, for a total of 28 control subjects and 42 patients overall. Table 1 demonstrates the results of our matching. We had perfect matching for patient age in years, sex, and the anatomical arachnoid cyst location. Our matching based on the side of the cyst was very good (P = .74) but not perfect, with 35.7% of cases and 46.4% of controls having right-sided lesions.
Comparing the Case and Control Groups
The demographic, clinical, and radiological characteristics of the patients in both the case and control groups are outlined in Table 2. Cases and controls did not differ in terms of either race or duration of symptoms in days, but children with ruptured/hemorrhaged arachnoid cysts (cases) were significantly more likely to present with either a headache (57.1% vs 14.3%, P = .009) or with signs or symptoms of raised intracranial pressure (85.7% vs 25.0%, P < .001), such as vomiting or papilledema. In terms of comorbidities, no children in the study had preexisting shunted hydrocephalus or any coagulopathies. Cerebral palsy and seizure disorders were more common in the unruptured control group, but neither of these differences was significant. Ventricle size, as measured by the frontal and occipital horn ratio, was within normal limits and similar (P = .44) between cases and controls. Children with ruptured arachnoid cysts were more likely (74.1% vs 10.7%) to have a midline shift than their unruptured counterparts.
Risk Factors at Study: Arachnoid Cyst Size, Preceding Head Injury, and Altitude at Residence
The associations between arachnoid cyst rupture/hemorrhage and the 3 primary study risk factors of arachnoid cyst size, preceding head injury, and altitude at residence are delineated in Table 3. We found that a larger cyst size (maximal diameter in any plane) was associated with arachnoid cyst rupture/hemorrhage (P < .001). For increased clinical utility, arachnoid cyst size was dichotomized with the use of a cutoff of 5 cm. Thirteen children in the study had a cyst 5 cm or greater, and 29 children had smaller cysts. We found that 69.2% of patients with a cyst of 5 cm or greater had experienced cyst rupture and/or hemorrhage, compared with an incidence of 17.2% in patients with cysts smaller than 5 cm (P < .001). An arachnoid cyst of 5 cm or greater was 16.5 times more likely to rupture or hemorrhage (odds ratio = 16.5; confidence interval [CI] [2.5, ∞]).
Sixteen of the 42 patients in our study (38.1%) had a preceding head injury within 30 days preceding the intracranial imaging. Fifteen of these were considered minor head trauma (fall from less than 1 story or minor blunt trauma), one resulted from a contact sport injury (football), and none were a result of either a motor vehicle accident or a fall from 1 story or greater. When grouped together, a history of a preceding head injury was documented in 85.7% of cases and 14.3% of controls (P < .001). A documented preceding head injury within 30 days had an odds ratio of arachnoid cyst rupture/hemorrhage of 25.1 (CI [4.0, ∞]).
The mean altitude at home residence for the unruptured controls was 4916 feet above sea level (ranging from 3210 to 6270 feet), whereas the mean altitude for the ruptured cases was 5168 feet above sea level (ranging from 4189 to 8339 feet); however, altitude at residence was not significantly associated with arachnoid cyst rupture/hemorrhage (P = .17).
Surgical Interventions and Outcomes
Surgical intervention was performed in 10 cases (71.4%) vs 4 controls (14.3%) (P = .002). Overall, 8/14 patients underwent a craniotomy and fenestration alone, 1 had a burr hole for the insertion of a subdural/cyst drain only, and 5 began with placement of a subdural drain but went on to secondarily require a craniotomy and fenestration for the treatment of their arachnoid cyst. Therefore, the failure rate of treating with a drain upfront was 83.3% (5/6 patients). Patients with ruptured/hemorrhaged arachnoid cysts were significantly more likely to require surgical intervention, a craniotomy, a subdural/cystic drain, or both (Table 3).
There were no major neurological sequelae or mortalities in either group (Table 3). At the time of study analysis, all 42 children were alive, well, and being monitored in the pediatric neurosurgical clinic. Permanent cerebrospinal fluid shunting for hydrocephalus was necessary in 2 (14%) of the cases and none of the controls. An association was found (P < .001) when we compared age at presentation and the need for a shunt in the ruptured group, with a mean age of 1.5 years in the group with shunts and 6.6 years in the patients without shunts. One of these patients was a 19-month-old boy whose ruptured cyst was originally treated with a drain from which he could not be weaned; he had a subdural-peritoneal shunt placed 7 days later. The other patient was a 15-month-old boy who was conservatively monitored for his symptomatic cyst rupture for 10 days but then underwent a craniotomy because of continuing symptoms of raised intracranial pressure. Postoperatively, he had an improving subgaleal cerebrospinal fluid collection that dissipated within a week. He did well clinically, but after 5 months of follow-up, he was found to have accelerated head growth, a large hygroma, and increased ventricle size on CT. He then underwent placement of a subdural-peritoneal shunt.
In developing the sample for this case-control study, we found an institutional incidence of 1.9% for arachnoid cyst in children undergoing intracranial imaging, and 6% of patients with a previously untreated arachnoid cyst were found to have either an associated hygroma or hemorrhage, which we defined as a rupture. Of the 3 primary risk factors examined, arachnoid cyst size (P < .001) and a history of a head injury within the previous 30 days (P < .001) were found to be associated with arachnoid cyst rupture. Residing at a higher altitude was not found to increase a patient's odds of rupture/hemorrhage.
Arachnoid cysts are not an infrequent entity in the pediatric population. Autopsy studies indicate an incidence of 0.1%,20 but radiological studies have shown a slightly higher incidence of between 0.2% and 1.7%.6,21,22 A more recent publication by Al-Holou et al1 demonstrated a pediatric prevalence, not incidence, of 2.6%. Our institutional incidence of 1.9% is consistent with these published studies. We found that 6% of children with newly diagnosed arachnoid cysts presented with a rupture. In our institutional incidence, these appear to be slightly more common than in other published works, which report rupture rates between 2.3% and 4.6%.7,15,22
Table 1 demonstrates perfect matching between cases and controls based on age, sex, and anatomical location. The matching for arachnoid cyst side (left vs right) was imperfect but still very acceptable. This level of matching of controls to cases added significant validity to the study's results. After controlling for these variables, we found a novel association between arachnoid cyst size and rupture/hemorrhage. Specifically, cysts ≥5 cm had an odds ratio of 16.5 (CI [2.5, ∞]) for rupture/hemorrhage compared with smaller cysts. Previous literature has noted that the presence of an arachnoid cyst alone may increase the risk of subdural hemorrhages or hygromas,9,11,15,16 but we could not find any previous work directly linking cyst size and rupture. Interestingly, Al-Holou et al1 found that arachnoid cysts were more likely to enlarge and require surgery in younger (age <4 years) children. Our study controlled for age and was not designed to confirm these findings; however, if enlargement is more likely to occur in younger children, then this age group may be at higher risk for rupture secondary to cyst size given our results.
A head injury within 30 days before cerebral imaging was also found to have an association with arachnoid cyst rupture/hemorrhage. In fact, this association was quite strong, with an odds ratio for rupture/hemorrhage of 25.1 (CI [4.0, ∞]); however, this potentially alarming statistic warrants temperance in its interpretation. First, we, as physicians, are potentially more apt to inquire and record recent head trauma if the patient presents with a ruptured or hemorrhaged arachnoid cyst, which may have introduced bias in this retrospective study. Second, for context, ruptures of arachnoid cysts are rare: 6% (14/232) at our institution and similarly between 2.3% and 4.6% in other published series.7,15,22 The majority (93.8%) of the head injuries in our study were seemingly trivial, including from low-level falls and direct blows such as a young sibling's slap. These types of minor head injuries are, of course, very common in children's daily lives. Therefore, the actual risk of rupture for any 1 single minor head incident is presumably quite small and not feasibly measureable. We believe our study clearly demonstrates that children with known arachnoid cyst are at higher risk of rupture with minor head injuries, but given the rare incidence of rupture and how well patients with ruptures have done neurologically and in terms of mortality, we do not, at this stage, have recommendations for the cessation of any sports or other high-impact activities. We do recommend that this risk should be discussed with patients and families, and any decisions regarding activities should be made on a case-by-case basis with the treating neurosurgeon.
On the basis of on our institution's experience, when counseling a patient with a ruptured arachnoid cyst and the family, it is important to highlight that there is a significant chance of needing surgical intervention (∼70%), a small chance of developing hydrocephalus that requires a shunt (∼14%), but, fortunately, a very low mortality rate.
Altitude was analyzed in this study because of the perceived higher incidence of arachnoid cyst rupture at our institution, which is located at higher altitude; however, no association was found. In fact, our arachnoid cyst rupture incidence was only slightly higher than that reported from other institutions.7,15,22
The majority of our patients presenting with rupture had middle fossa arachnoid cysts (86%), and none had posterior fossa cysts. This may suggest that the risks for arachnoid cysts might vary by location, but our study did not have the data to support this conclusion directly. It has been hypothesized the juxtaposition of these middle fossa arachnoid cysts to the thin temporal squamous bone and the cerebral vessels within the sylvian fissure might account for their higher rupture rate.
We acknowledge some limitations of this study. The study was retrospective and depended on proper annotation of the risk factors of interest. In particular, it may be that patients in the control group were less likely to be asked about a recent history of trauma, which introduces a form of recall bias. Selection bias may have occurred, because we were only able to identify those cases (ruptured) that were hospitalized and those controls (unruptured) that were imaged for any reason. There may be a referral bias creating a higher prevalence of cases vs controls, because our institution receives the bulk of pediatric neurosurgical referrals in the region.
This case-control study demonstrated that children with larger intracranial arachnoid cysts were at increased risk of cyst rupture (intracystic hemorrhage or an overlying subdural hematoma or hygroma), especially if they had a cyst larger than 5 cm in maximal diameter. A head injury within the previous 30 days, even if considered relatively minor, was also associated with cyst rupture; however, minor head trauma was potentially underreported in the unruptured or control sample. We found no association between altitude and cyst rupture. We believe our center is representative of most North American tertiary pediatric neurosurgical centers, suggesting that our findings may be used as an adjunct in counseling pediatric patients with known arachnoid cysts.
The authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article.
The authors thank Kristin Kraus, MSc, for her invaluable assistance in editing this manuscript.
1. Al-Holou WN, Yew AY, Boomsaad ZE, Garton HJ, Muraszko KM, Maher CO. Prevalence and natural history of arachnoid cysts in children. J Neurosurg Pediatr. 2010;5(6):578–585.
2. Guzel A, Tatli M, Kilincer C, Yilmaz F. Posttraumatic intraventricular arachnoid cyst accompanied by pseudomeningoencephalocele in a child. J Clin Neurosci. 2007;14(12):1210–1213.
3. Peyser E, Weissberg D. Post-traumatic arachnoidal cyst. Report of an unusual case. J Neurosurg. 1961;18:551–553.
4. Kim BS, Illes J, Kaplan RT, Reiss A, Atlas SW. Incidental findings on pediatric MR images of the brain. AJNR Am J Neuroradiol. 2002;23(10):1674–1677.
5. Morris Z, Whiteley WN, Longstreth WT Jr, et al.. Incidental findings on brain magnetic resonance imaging: systematic review and meta-analysis. BMJ. 2009;339:b3016.
6. Weber F, Knopf H. Incidental findings in magnetic resonance imaging of the brains of healthy young men. J Neurol Sci. 2006;240(1-2):81–84.
7. Bilginer B, Onal MB, Oguz KK, Akalan N. Arachnoid cyst associated with subdural hematoma: report of three cases and review of the literature. Childs Nerv Syst. 2009;25(1):119–124.
8. Domenicucci M, Russo N, Giugni E, Pierallini A. Relationship between supratentorial arachnoid cyst and chronic subdural hematoma: neuroradiological evidence and surgical treatment. J Neurosurg. 2009;110(6):1250–1255.
9. Gelabert-González M, Fernández-Villa J, Cutrín-Prieto J, Garcìa Allut A, Martínez-Rumbo R. Arachnoid cyst rupture with subdural hygroma: report of three cases and literature review. Childs Nerv Syst. 2002;18(11):609–613.
10. Gündüz B, Yassa MI, Ofluoğlu E, et al.. Two cases of arachnoid cyst complicated by spontaneous intracystic hemorrhage. Neurol India. 2010;58(2):312–315.
11. Offiah C, St Clair Forbes W, Thorne J. Non-haemorrhagic subdural collection complicating rupture of a middle cranial fossa arachnoid cyst. Br J Radiol. 2006;79(937):79–82.
12. Pillai P, Menon SK, Manjooran RP, Kariyattil R, Pillai AB, Panikar D. Temporal fossa arachnoid cyst presenting with bilateral subdural hematoma following trauma: two case reports. J Med Case Rep. 2009;3:53.
13. Sener RN. Arachnoid cysts associated with post-traumatic and spontaneous rupture into the subdural space. Comput Med Imaging Graph. 1997;21(6):341–344.
14. Ziaka DS, Kouyialis AT, Boviatsis EJ, Sakas DE. Asymptomatic massive subdural hematoma in a patient with bitemporal agenesis and bilateral temporal arachnoid cysts. South Med J. 2008;101(3):324–326.
15. Wester K, Helland CA. How often do chronic extra-cerebral haematomas occur in patients with intracranial arachnoid cysts? J Neurol Neurosurg Psychiatry. 2008;79(1):72–75.
16. Mori K, Yamamoto T, Horinaka N, Maeda M. Arachnoid cyst is a risk factor for chronic subdural hematoma in juveniles: twelve cases of chronic subdural hematoma associated with arachnoid cyst. J Neurotrauma. 2002;19(9):1017–1027.
17. Bristol RE, Albuquerque FC, McDougall C, Spetzler RF. Arachnoid cysts: spontaneous resolution distinct from traumatic rupture. Case report. Neurosurg Focus. 2007;22(2):E2.
18. Kulkarni AV, Drake JM, Armstrong DC, Dirks PB. Measurement of ventricular size: reliability of the frontal and occipital horn ratio compared to subjective assessment. Pediatr Neurosurg. 1999;31(2):65–70.
19. O'Hayon BB, Drake JM, Ossip MG, Tuli S, Clarke M. Frontal and occipital horn ratio: a linear estimate of ventricular size for multiple imaging modalities in pediatric hydrocephalus. Pediatr Neurosurg. 1998;29(5):245–249.
20. Shaw C, Alford E Jr. Congenital arachnoid cysts and their differential diagnosis. In: Vinkin PJ, Bruyn GW, eds. Handbook of Clinical Neurology. Vol 31. Amsterdam: North Holland Publishing Company; 1977:75–136.
21. Eskandary H, Sabba M, Khajehpour F, Eskandari M. Incidental findings in brain computed tomography scans of 3000 head trauma patients. Surg Neurol. 2005;63(6):550–553; discussion 553.
22. Parsch CS, Krauss J, Hofmann E, Meixensberger J, Roosen K. Arachnoid cysts associated with subdural hematomas and hygromas: analysis of 16 cases, long-term follow-up, and review of the literature. Neurosurgery. 1997;40(3):483–490.
The authors have assembled a matched case-control study to answer a very specific question: in children presenting with intracranial arachnoid cysts, are recent head injury, cyst size, and altitude at residence risk factors for presentation with ruptured vs nonruptured cyst. They defined rupture as an intracystic hemorrhage or an overlying subdural hematoma/hygroma. Given the relative rarity of ruptured arachnoid cysts (6% of all newly diagnosed arachnoid cysts in their population), the relatively small sample size is not surprising (14 cases matched to 28 controls). Nevertheless, the authors have applied solid methodology and analysis to conclude that larger cyst size and history of recent head injury are risk factors for presentation with a ruptured arachnoid cyst. As mentioned by the authors, one of the main limitations in interpreting their results is that there is an inherent bias for treating physicians to more likely search for, and document, a history of trauma when faced with a child presenting with a known ruptured cyst. Therefore, the association is likely being overestimated to some degree. As well, as the authors point out, many of the recorded head injuries were really quite trivial, so the causal link is somewhat tenuous. On this basis, the authors correctly warn against overinterpretation. Specifically, this study cannot truly quantify the risk of rupture of a known arachnoid cyst following a head injury event, limiting our ability to fully counsel families with respect to risk-related activities, such as contact sports. Their data, nevertheless, add further to our knowledge about this condition and represent one of the methodologically stronger studies in the literature.
Abhaya V. Kulkarni
Toronto, Ontario, Canada
1. An 8 year old boy presents to the ER after falling off his bike onto his head. A head CT only reveals an incidental unruptured temporal lobe arachnoid cyst. What percentage of arachnoid cysts present with a rupture?Supination
2. What single factor most significantly increases the risk of arachnoid cyst rupture?
a. Cyst size greater than 5 cm in diameter
c. High altitude home residence
d. Minor head trauma within past 30 days
e. Age >18 years
3. A child is found to have an incidental 5.5 cm arachnoid cyst in the right middle temporal fossa. What age range would significantly increase the risk of arachnoid cyst rupture?
a. <4 years
b. 4-8 yrs
c. 8-12 yrs
d. 12-16 yrs
e. >16 yrs