Mastoid effusion (ME) occurs frequently along with middle ear effusion in children as a consequence of the accumulation of transudate caused by negative pressure or inflammation in the middle ear (1). Due to the differences in the orientation, length, and function of the Eustachian tube, ME is not a common finding in healthy adults, with an incidence of approximately 1% (2, 3). However, several studies have demonstrated that the incidence of ME is higher in intensive care unit (ICU) patients than in general population. Risk factors for the development of ME in ICU patients include old age, increased mucus secretion caused by prolonged endotracheal intubation for mechanical ventilation and a nasogastric tube, prolonged ICU stay, and altered mental status, many of which can be attributed to the anatomical connection of the middle ear space and the nasopharynx via the Eustachian tube (2, 4, 5).

Anatomical continuity also exists between the subarachnoid space and the cochlear aqueduct of the middle ear, and it provides the background for the development of a non-invasive tool for monitoring intracranial pressure (ICP) by measuring the tympanic membrane pressure (6, 7). Moreover, a recent study reported that elevated levels of ICP was associated with the occurrence of ME in neuro-ICU patients (8). Therefore, it can be postulated that the occurrence of ME in patients with neurologic insults who are at a risk of ICP variation may offer additional information regarding the clinical course and outcomes.

Subarachnoid hemorrhage (SAH) due to intracranial aneurysm rupture is a disastrous event with high morbidity and mortality rates (9, 10). A rupture of an aneurysm causes an abrupt and delayed increase in ICP, resulting in impaired cerebral perfusion and temporary intracranial circulatory arrest (11). Increased ICP plays an important role in the clinical course of aneurysmal subarachnoid hemorrhage (aSAH) patients, as it is related to common complications of aSAH such as acute hydrocephalus or delayed cerebral ischemia, which mostly occur in the first 14 days after the ictus. We hypothesized that patients receiving aSAH management who developed ME in the first 14 days would demonstrate unfavorable clinical outcomes. The purpose of this study was to investigate the incidence and the clinical significance of ME in patients with aSAH.

This retrospective study was approved by our Institutional Review Board and was performed under the guidelines outlined in the Declaration of Helsinki. The diagnosis of ME was based on the criteria proposed by J. Gossner, and only cases classified as “marked” (i.e., fluid signal involving more than half of the mastoid air cells) were considered ME-positive. By including only clearly defined cases, we minimized subjectivity in the interpretation. Accordingly, there was no interobserver disagreement between the neurovascular specialist and the neuro-intensivist who independently reviewed the images. This study follows the STROBE guidelines for retrospective studies.

In this retrospective study, we found that ME occurred in 34.5% of aSAH patients within the first 14 days (mean 5.0 ± 3.5 days) and was associated with radiologic vasospasm and poor clinical outcomes. Moreover, along with older age and a poor Hunt–Hess grade, ME was an independent risk factor for poor clinical outcomes in aSAH patients.

Several studies have reported a higher incidence of ME in ICU patients than in the healthy adult population, ranging between 10.3 and 53% (2, 8, 13). Risk factors for the development of ME include thickened oropharyngeal secretions due to mucosal irritation by endotracheal intubation and nasogastric tube, and decreased mentality impairing the patient’s ability to clear excessive mucosal secretion and to open the Eustachian tube, which are hardly modifiable (2, 4, 13). Although ME can potentially cause acute mastoiditis and develop into intracranial complications such as meningitis, empyema, and brain abscess (14), incidentally detected MEs are rarely related to temporal bone disease (15). ME observed in ICU patients is considered benign in most cases, obscuring its clinical significance (2, 13). This study, however, showed that ME was independently associated with poor clinical outcomes in aSAH patients. Patients who suffer aSAH are usually hospitalized for 14 to 21 days due to the possible occurrence of delayed complications (16), allowing prognostication to be made at least 14 days after the ictus in most cases. In conjunction with the established risk factors for poor outcomes in aSAH patients, such as Hunt–Hess score, increasing age, and ruptured posterior circulation aneurysm, the occurrence of ME may offer an additional prognostic value.

Recently, Jung et al. suggested that increased ICP was associated with ME occurrence in patients who underwent intracranial surgery (8). The authors revealed that the prediction model for the development of ME improved when peak ICP values were included in the model. Timely recognition and proper management of increased ICP are critical for improving clinical outcomes in patients with aSAH. However, most widely used ICP monitoring tools, such as external ventricular drainage, are invasive, and it is often a dilemma whether to place invasive ICP monitoring devices due to the risks of infection or hemorrhage (17, 18). It has been recognized that ICP is transmitted to the perilymphatic space in the middle ear, and the middle ear pressure represented by tympanic membrane displacement can be utilized for the indirect measurement of ICP (6, 19). Assuming that the occurrence of ME reflects increased ICP, a clinician may conduct additional evaluations or procedures to assess the patient’s ICP status, or ophthalmologic evaluation for the possibility of Terson’s syndrome when ME is observed. Although this finding was not statistically significant, Terson’s syndrome, a consequence of elevated ICP, was diagnosed in two patients who developed ME in this study. However, whether there is a causal relationship between elevated ICP and the development of ME is not known. A few studies attempted to explain the development of ME in patients with lateral sinus thrombosis as the result of venous congestion rather than the Eustachian tube dysfunction, as the laterality of the ME coincided with the intracranial lesions responsible for the elevated ICP (8, 20, 21). Such an explanation may not fit in other types of intracranial pathologies not involving venous congestion. From the anatomical perspective, the movement of cerebrospinal fluid from the subarachnoid space to the mastoid air cells via the middle ear due to a pressure gradient is theoretically possible. The relationship between elevated ICP and the development of ME, their mechanisms, and temporal associations should be further investigated to refine the clinical significance of ME in patients whose ICP assessments are critical.

Another finding in this study was that the occurrence of ME was associated with post-SAH vasospasm. Post-SAH vasospasm has multiple risk factors, including the amount of SAH, presence of intracranial hemorrhage and intraventricular hemorrhage, female sex, and increased ICP (22). Fukuhara et al. demonstrated that elevated ICP was associated with the development and the duration of vasospasm after aSAH (23). Similarly, Gambardella et al. showed that the use of osmotic diuretics to control ICP lowered the risk of developing delayed cerebral ischemia in patients with aSAH (24). On the other hand, Heuer et al. reported only a weak relationship between ICP and the development of angiographic or symptomatic vasospasm (25). The authors provided a few explanations for the weak link. They argue that patients with severely elevated ICP died before vasospasm could occur, or did not undergo follow-up angiographic evaluations. In our case, 8 patients (42.1%) out of 19 patients did not undergo follow-up angiography studies, as poor prognosis was expected.

The mean interval between the onset of aSAH and detection of ME was 5 days. Limited studies have reported the time for ME to develop in ICU patients. Jung et al. reported a mean of 11.1 days for the development of ME in patients who underwent intracranial surgery, and Huyett et al. described that ME was a late finding and prevalent in patients with a prolonged ICU stay (2, 8). Considering that elevated ICP contributes to the development of ME, it can be postulated that ME in patients with neurological insults such as SAH would develop at an earlier period than in patients with other etiologies. Early brain injury after aSAH is known as a pathologic process that occurs in the first 72 h, and involves the elevation of ICP, reduction of cerebral blood flow, and neuronal cell death (26). SAH-induced vasospasms usually occur 4 to 14 days after the ictus, with a peak incidence at 7 days (27). The relationship between the early phase of brain injury and the development of ME should be further described in future studies.

Although our findings suggest that mastoid effusion (ME) is associated with poor outcomes and angiographic vasospasm in patients with aSAH, the exact pathophysiological mechanisms underlying ME development in neurologically injured patients remain uncertain. It has been hypothesized that elevated intracranial pressure (ICP) or impaired cerebrospinal fluid drainage may contribute to fluid accumulation within mastoid air cells. However, this mechanism has not yet been established in the existing literature.

Given the routine use of non-contrast brain CT in the acute-phase SAH, ME may serve as a readily identifiable imaging marker that reflects disease severity or elevated ICP, particularly in patients without invasive monitoring. While ME is not a direct therapeutic target, its presence may have potential implications for early risk stratification and the intensity of subsequent monitoring or intervention. Nonetheless, this study was not designed to establish a causal mechanism or to propose a clinical algorithm based on ME findings. Further prospective studies are needed to clarify the biological basis of ME in this patient population and to determine how it may be effectively incorporated into clinical decision-making processes. We acknowledge this limitation and have highlighted the need for future research in this area.

Previous studies have suggested a potential correlation between mastoid effusion (ME) and elevated intracranial pressure (ICP), implying that ME may serve as an indirect radiological marker of increased ICP in patients with acute brain injury (8). One proposed mechanism is that a pressure gradient between the subarachnoid space and the middle ear cavity could facilitate the movement of cerebrospinal fluid (CSF) into the mastoid air cells, resulting in effusion. Furthermore, elevated ICP has been linked to both the occurrence and persistence of cerebral vasospasm following aSAH, potentially explaining the association between ME and poor outcomes observed in this study.

To further explore this hypothesis, we analyzed available ICP data in relation to ME status (Table 3). The mean ICP was 20.9 ± 11.4 mmHg in the ME-negative group and 19.2 ± 9.6 mmHg in the ME-positive group, with no statistically significant difference. This finding should be interpreted cautiously for several reasons. First, the timing of ICP measurement and the detection of ME on imaging did not always coincide, limiting a temporal correlation. Second, most patients underwent intensive ICP-lowering interventions, including osmotherapy, targeted temperature management, and coma therapy, which may have modified or masked actual ICP values. These factors likely contributed to the lack of observed statistical significance.

Nevertheless, we believe that ME may still reflect transient or unrecorded ICP elevations, particularly in patients without invasive monitoring. Given that non-contrast brain CT is routinely performed in the acute phase of aSAH, the detection of ME may provide a non-invasive and easily accessible clue suggesting raised ICP, especially in resource-limited settings.

Importantly, the pathophysiology of ICP elevation in aSAH differs from that of other neurological conditions such as traumatic brain injury or ischemic stroke. In aSAH, abrupt ICP elevation can occur due to aneurysmal rupture and subsequent hydrocephalus and can fluctuate rapidly with events such as rebleeding, vasospasm, or therapeutic interventions. These unique dynamics complicate the interpretation of single-timepoint ICP data but underscore the potential value of adjunctive imaging markers such as ME.

This study has several limitations. Due to its retrospective design and small sample size, selection and information bias were unavoidable, as the details of management and follow-up strategies for aSAH patients varied. For example, the follow-up non-contrast brain CT imaging intervals varied between patients, obscuring the precise timing of ME development, and only 83.3% (95/114) of the patients underwent follow-up angiography studies at irregular intervals to be evaluated for the occurrence of vasospasm. In asymptomatic patients with longstanding mastoid effusion resulting from conditions such as chronic otitis media, the absence of prior imaging may lead to misclassification of ME as a newly developed finding. While ME could be ruled out through comparison with previous imaging in patients who underwent serial follow-up studies, this was not feasible in newly admitted patients without prior imaging. The likelihood of pre-existing mastoiditis was presumed to be comparable between the two groups. Furthermore, as mastoid effusion grade 2 or higher was defined as ME-positive in this study, cases of mild mastoiditis were likely excluded. Patients with a documented history of otitis media were also excluded based on a thorough review of medical records. In our dataset, only 5 out of the 40 patients with ME showed evidence of ME on their initial brain CT. Although it was not possible to determine whether these cases represented pre-existing ME, all five patients demonstrated progression of ME on follow-up imaging.

Another limitation is the potential subjectivity in the radiologic diagnosis of ME. To address this limitation, only clearly defined cases meeting the “marked” criteria proposed by Gossner—defined as a fluid signal involving more than half of the mastoid air cell cavity—were included (28). All imaging was independently assessed by a neurovascular specialist and a neuro-intensivist, and there was complete agreement between the two evaluators. Therefore, a formal consistency test was not conducted. Additionally, although a quantitative threshold such as the Hounsfield unit (HU) could be useful, variability in CT protocols and scanner calibration limited its application in this retrospective analysis. Finally, some factors that may influence the development of ME—such as the presence of nasogastric tubes or ICP values—were not included in this study. In particular, the absence of ICP data may limit the ability to fully explore the relationship between ME and elevated intracranial pressure. Despite these limitations, ME was found to be independently associated with poor outcomes and cerebral vasospasm in patients with aSAH. Further studies are warranted to clarify the underlying mechanisms linking ME with elevated levels of ICP and to validate its role as a prognostic marker.

This study demonstrated that the occurrence of ME in aSAH patients was independently associated with tracheostomy, vasospasm, and poor clinical outcomes. Given that ME can be easily identified on non-contrast brain CT early in the clinical course, it may serve as an additional imaging marker for prognosis in aSAH patients.