Introduction

Elevated intracranial pressure (ICP) may be indirectly evaluated by measuring the optic nerve sheath diameter (ONSD) [1]. The rationale resides in the anatomical continuity between the subarachnoid space of the optic nerve sheaths and that of the brain, allowing the circulation of cerebrospinal fluid (CSF) from the perichiasmatic cistern to the ocular regions [2]. With increasing ICP, the pushing pressure drives a larger amount of CSF toward the orbital spaces. As a result, the ONSDs enlarge and optic disc elevation develops over time. However, the CSF transmission may be limited at the level of the optic canals where the composition of the subarachnoid space changes, becoming subdivided in small compartments and narrowing its lumen to a capillary size [3]. This specific architecture, made of thick pillars and trabeculae, is assumed to be responsible for the quantity and speed with which the CSF can be transferred to the optic sheaths, and its large inter-individual variability may explain the documented occurrence of asymmetric ONSD enlargement and unilateral papilledema [4, 5]. Furthermore, no data on individual interocular differences have been reported.

The ONSD may be explored with various techniques, transorbital sonography being one of the most promising due to its non-invasiveness, feasibility, and high diagnostic accuracy [6]. The application fields of this method in the clinical practice have been recently described [7]. Despite the anatomical remarks cited above, most ultrasonographic studies consider ONSD measurements as an average of binocular values, rather than the maximum value between the two eyes, leading to possible errors in estimating ICP, given the potential physiological asymmetry.

The aim of this study was to evaluate the variability, in term of symmetry, of the ONSD, by using ultrasonography in healthy subjects and patients with intracranial hypertension (IH), and to establish whether the maximum value of the ONSD could be as useful as the binocular assessment.

Materials and methods

Subjects

We retrospectively reviewed ONSD data from 80 subjects, collected from a dataset available from the authors [8, 9]. All participants were ≥ 18 years old; exclusion criteria were ocular trauma, glaucoma, and significant orbital mass. Forty normal subjects who underwent general medical, ophthalmological, and neurological evaluation to rule out signs and symptoms of IH and ocular pathologies were included in the healthy group. Forty patients, matched for gender but not for age, composed the IH group, and included 20 consecutive subjects with idiopathic intracranial hypertension (IIH) and 20 consecutive patients with primary acute supratentorial intracerebral hemorrhage (ICH).

Patients with IIH were recruited at the “Franz Tappeiner” Hospital of Merano Hospital between March 2014 and November 2015; raised ICP was established (after ONSD evaluation) by lumbar puncture with a CSF opening pressure ≥ 25 cm H20 according to the Friedmann diagnostic criteria [10]. ICH patients were admitted to the “Maggiore della Carità” Hospital of Novara from January to December 2017; elevated ICP was determined by clinical conditions and cerebral CT scan. Radiologically, we considered only medium- to large-size cerebral hemorrhages (≥ 30 ml) with at least two of the following: midline shift ≥ 3 mm, collapsed 3rd ventricle, hydrocephalus, compression of cisterns, and effacement of sulci with evidence of substantial edema [11, 12]. Hematoma size was calculated using the ABC/2 score (if rounded or ellipsoid shape) or ABC/3 score (irregular or multinodular) [13, 14]. Contextually, at least one clinical sign of elevated ICP had to be present, including bradycardia, altered consciousness, vomit, headache, pupillary abnormalities, or uncontrolled hypertension.

Transorbital sonography

Transorbital ONSD sonography was performed following a previously described protocol using a Toshiba Medical System Aplio 300 (Nasu, Japan; linear probe 7.5–11 MHz with lateral resolution < 0.4 mm) for healthy subjects and ICH patients, and a Toshiba Aplio ultrasound system XG (Toshiba Medical Systems, Nasu, Japan; linear probe 7.2–14 MHz, lateral resolution < 0.4 mm) for IIH patients [15]. Ultrasonographic evaluations were performed by two skilled neurosonologists (A.N. and P.L.); all measurements were taken off-line by a single expert operator (P.L.) to reduce interobserver variability. ONSD was calculated as the distance between the external hyperechogenic borders surrounding the optic nerve, corresponding to the periorbital subarachnoid space.

ONSD was evaluated on the axial plane in triplicate for each eye. Binocular ONSD was obtained by averaging these values between the eyes, while maximum ONSD was obtained by first averaging the three measurements for each eye and then taking the maximum of the two values.

Statistical analysis

Statistical analysis was performed using the R software, version 3.6.0 [16]. For each subject, the ONSD asymmetry was evaluated as the absolute value of the difference between the mean ONSD of the two eyes; statistical significance was assessed using Welch’s t test, with multiple testing correction performed with the Benjamini-Hochberg method at a false discovery rate of 5%. For group comparison, categorical variables were expressed as numbers and percentages and compared among groups with Fisher’s exact test. Continuous variables, expressed as median and quartiles (Q1, Q3), were compared among more than two groups with the Kruskal-Wallis test, and between two groups with the Mann-Whitney U test.

Results

Demographic data, mean ONSD values, and asymmetry are summarized in Table 1. The median age of subjects was 72.5 (range 44.75–79) and was significantly lower in IIH patients (37 (29–47.25), p < 0.001), as expected for the pathology; gender did not differ between the three subgroups (p = 0.114).

Table 1 Demographic characteristics, binocular/maximum ONSD values, and asymmetry between eyes in healthy and pathological subjects
Full size table

Forty-two out of 80 subjects (52.5%) showed a significant ONSD asymmetry between eyes at a false discovery rate of 5%, and the difference in prevalence among groups was not significant (16 for healthy subjects, 14 for ICH patients, and 12 for IIH; p = 0.28). The median asymmetry was higher in patients with IH than in normal subjects (0.45 mm vs 0.23 mm, p = 0.007, Fig. 1a); when stratifying by pathology, the significance was confirmed for ICH patients (p = 0.012), while the result for IIH patients was close to significance (p = 0.07) when compared with the healthy group (Fig. 1b). There was no significant difference between ICH and IIH patients (p = 0.58). For ICH patients, there was no significant correspondence between the hemorrhage site and the side of the larger ONSD (p = 0.41).

Fig. 1
figure 1

Degree of asymmetry (expressed in millimeters) in healthy subjects and patients with intracranial hypertension (a) and subgroup analysis dividing the cases by pathology (b). ICH, intracerebral hemorrhage; IIH, idiopathic intracranial hypertension

Full size image

Both the binocular and maximum ONSD values were significantly higher in the IH group (independently of the pathology) compared with the healthy group (both p < 0.001, Fig. 2a, b), with no significant differences among the two pathologies (p = 0.12 and p = 0.15, respectively).

Fig. 2
figure 2

Binocular (a) and maximum (b) ONSD values in patients with intracranial hypertension and healthy subjects. ONSD, optic nerve sheath diameter; ICH, intracerebral hemorrhage; IIH, idiopathic intracranial hypertension

Full size image

Discussion

This study documented for the first time in a clinical context the occurrence and degree of interocular ONSD asymmetry by using transorbital sonography in healthy and pathological subjects with IH. Ultrasonography of the ONSD has been proposed as a non-invasive technique to evaluate ICP and ONSD behavior in various conditions. The sheath diameter may increase or shrink in response to ICP changes, but above a dilation threshold the enlargement seems to be irreversible [17, 18]. However, the exact reaction of the ONSD and the optic disc upon ICP variation remains elusive and different mechanisms may be involved in the orbital response, including optic atrophy, CSF and ocular pressure, and changes in the lamina cribrosa or optic nerve structures [19]. Regarding the anatomical aspects, some authors emphasized the role of the optic canals in the transmission of the CSF toward the optic sheaths [4]. Specifically, the thick arachnoid trabeculae and pillars that compose the complex meshwork of the optic canals may limit or even prevent the flow of CSF in the course of IH, determining a disproportionate expansion of ONSDs and unilateral papilledema [20,21,22]. In addition, an elevated inter-individual variability of optic canal conformation among humans and animals has been described [5, 21]. However, to our knowledge, no data are available about the interocular asymmetry of the ONSD between eyes both in normal conditions and following elevated ICP, particularly in a clinical setting where only isolated descriptions of asymmetry are mentioned [22]. Nevertheless, most ultrasonographic and radiological studies related to this topic evaluated the binocular ONSD value, without considering this anatomical variability, which may limit the enlargement of the optic sheaths, thus potentially leading to an inaccurate ICP estimation.

Our results documented an asymmetry of the ONSD of the two sides, both in healthy individuals and in patients with IH. The asymmetry increased following raised ICP, confirming the anatomical finding of a different architecture of the optic canals: If the CSF may not be adequately conveyed toward one of them because of a different composition in the trabecular network thickness, the two ONSDs may enlarge asymmetrically. No differences where observed between IIH and ICH subjects, suggesting that the asymmetry was independent on the cause and mechanism of intracranial hypertension. In fact, while in IIH increased ICP may be related to a pathological process involving the whole brain, in ICH it is a consequence of a focal event. Interestingly, there was no association between the side of cerebral hemorrhage and the maximal ONSD dilation, further suggesting that anatomical aspects may influence the ONSD behavior.

The binocular ONSD assessment is a well-known technique for assessing increased ICP; conversely, only few studies evaluated the role of maximum ONSD for this purpose [23,24,25]. In our series, both binocular and maximum ONSD were significantly higher in patients than in healthy subjects; thus, both measures could be useful for the detection of elevated ICP. However, considering the asymmetry and the possibility of a missing or reduced enlargement of the ONSD, ideally maximum ONSD could be a more precise and quicker measurement, especially when asymmetry is present.

There are some limitations to this study, mainly related to its retrospective design and to the small sample size. In ICH patients, diagnosis of IH was based on clinical and radiological findings, without directly measuring ICP. Finally, ONSD values were recorded by using two different sonographic machines. Further studies in a larger population are needed to confirm these results.

Conclusions

An interocular ONSD asymmetry exists in normal subjects and patients with elevated ICP, with a larger extent in IH. If present, the asymmetry should be considered for correct ICP estimation. Both binocular ONSD measurements and maximum ONSD may be useful for the evaluation of raised ICP.