We would like to thank Dr Mattei for his thoughtful letter based on our study “Risk Factors for Pediatric Arachnoid Cyst Rupture/Hemorrhage: A Case-Control Study” and the editor for inviting this response.
First, Dr Mattei seems to be freely equating the expansion of the subarachnoid space (like we see in benign intracranial collections of the newborn) with the subdural space. This is particularly evident in his included Figures 1 and 2. The infants in these figures clearly have benign intracranial collections of the newborn and an arachnoid cyst, but neither has a subdural hygroma because the excess “extracerebral collections” are subarachnoid. The anatomy would be more definitively demonstrated on magnetic resonance imaging. Therefore, neither of these patients would be considered as having a “ruptured cyst” as defined in our study and, thus, both would have been considered controls not cases. Similarly, we would not treat either the “extracerebral collections” or the arachnoid cyst in these 2 patients.
Dr Mattei also states that our outcome was “devoid of concomitant clinical evaluation.” This is unfortunately misleading, because all of the rupture cases had either a headache (P = .009) or symptoms of raised intracranial pressure (P < .001), which was significantly different than control patients, who generally had imaging for other reasons (Table 2 in our article). In fact, in all of our patients with subdural hygromas, these lesions were adjacent to the arachnoid cyst, creating mass effect with gyral effacement and, in many cases, even midline shift.
Dr Mattei clearly states, when detailing his theory of arachnoid cyst development, that his concerns center primarily on the young infants under the age of 2 years presenting with, by our standards, a subdural hygroma. Therefore, we did additional analysis to see whether these effects were biasing our results. When patients were categorized as less than or greater than 2 years, young age had no univariate association (P = .67) with a subdural hygroma presentation vs that of an intracystic or subdural hemorrhage. Furthermore, we had 3 patients with ruptured cysts who met his criteria of <2 years and had hygromas. When these 3 children and their associated 6 controls were eliminated from the analysis, we found nearly identical results. A recent head injury (82% of cases, 9% of controls, P < .001) and a cyst larger than 50 mm (64% of cases, 14% of controls, P = .006) remained associated with arachnoid cyst rupture. Therefore, even if his theory is correct, it has no influence on the study's results or conclusions.
To further evaluate these data, we divided the entire sample into children less than 2 years of age (all of which were between 1 and 2 years of age) and those older than 2 years and ran separate primary analyses. For the children <2 years, a recent head injury (P = .003) and large cyst size (60% of cases, 22% of controls, P = .26) trended with cyst rupture. For the children older than 2 years, a recent head injury (P = .003) and large cyst size (67% of cases, 11% of controls, P = .005) were significantly associated with cyst rupture. Therefore, these sensitivity analyses demonstrate the robustness of our results, with neither young age (less than 2 years) nor rupture type (hygroma vs hemorrhage) influencing our conclusions.
We are intrigued with his theory of arachnoid cyst development; however, given the explanations above and the expanded reanalyses, our results stand as published.
The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.