Introduction

The orbit is a bony cavity in the skull that houses the eye and its structures, among them the optic nerve (ON) and extraocular muscles. The orbital anatomy is a highly complex and intricate field of study that has important implications for various medical specialties, such as ophthalmology, neurology, otolaryngology, and plastic surgery. There are several fat compartments within the orbit that play a crucial role in maintaining the position and function of the eye, including the extraconal and intraconal fat tissues [1]. The intraconal fat is a dense and fibrous layer of adipose tissue that surrounds the ON and other vital structures in the orbit [2]. Despite its importance, the anatomy and function of different areas within the intraconal fat remain poorly understood.

The use of imaging techniques, such as magnetic resonance imaging (MRI) and computerized tomography (CT), along with pathological investigations, are powerful tools for studying orbital anatomy [3]. A comprehensive understanding of the orbits and surrounding structures mandates the combination of data provided by these various modalities. Analysis of orbital fat density, which can be readily extrapolated from CT scans, is an important part of this evaluation [4, 5]. Hounsfield units (HU) represent the density of tissue on CT. Normal fat is less dense than water, and measures in the range of −50 to −150 HU in non-contrast-enhanced (NCE)-CT scans, while they are higher in CE-CT scans [6]. Orbital inflammation or edema will cause a relative increase in intraconal fat HU values (i.e., a less negative value). Various ocular and orbital pathologies alter HU measurements: for example, they are commonly increased in Graves' orbitopathy and in orbital cellulitis [4, 7–10]. The normal range of orbital fat density of different areas within the intraconal fat have not been established, and it has been suggested that those data can assist in diagnosis of orbital pathology [10, 11]. Data on normal orbital fat density are sparse, specifically those on differences between nasal and temporal fats which have not yet been investigated [12].

As part of a recent study of orbital fat density in pre-septal and orbital cellulitis, HU values emerged as being significantly higher in nasal areas than in temporal areas, both in involved and non-involved orbits [13]. This interesting finding raised questions about the presence of such differences under normal conditions, i.e., in the absence of any associated orbital or periorbital disease. The goal of the current study, therefore, was to evaluate whether nasal and temporal intraconal fat differ in density and if so, to identify the nature of this difference in normal orbits. The information on normal baseline values obtained from this study can be beneficial for clinicians and radiologists in precisely interpreting CT scans, leading to more informed diagnoses and treatment decisions for patients with orbital pathologies.

Patients and methods

Hounsfield unit measurements

The density of intraconal orbital fat was measured by manual region of interest segmentation of the fat compartment. A single ophthalmologist (E.B.S.) performed all measurements. Six areas per orbit were measured for each patient, yielding a total of 12 measurements per patient. Three slides were used for the measurements: one at the axial equator plane of the globe, one at a section above it, and one at a section below it. The nasal and temporal HU values were measured for each orbit on each side (Fig. 1). We used axial scans exclusively since they represent the original image and had not been reformatted. As a result of this selection, we were able to make optimal comparisons between temporal and nasal sites. The HU measurements were analyzed for correlations between the nasal and temporal parts of the intraconal fat, as well as for any association between laterality of pathology, the use of contrast for imaging, and the patient's age, sex, and eye laterality.

View Full Size

Demonstration of the measurement of intraconal orbital fat density. The procedure involved manually segmenting a region of interest (ROI) in the fat compartment on axial scans at the nasal and temporal sites of the optic nerve at the axial equator plane of the globe. This process was repeated one section above and one section below the equator for each patient

Citation: Imaging 2024; 10.1556/1647.2024.00246

• Download Figure
• Download figure as PowerPoint slide

Results

The study cohort was comprised of 54 patients (29 females [54%]), with a mean ± SD age of 45.3 ± 25.5 years. NCE-CT scans were available for 36 patients (67%), and CE-CT scans for 31 patients (57%), with 13 patients having undergone both NCE-CT and CE-CT scans. The mean HU values for the left orbit was −75.9 ± 7.5 on NCE-CT and −72.8 ± 7.1 on CE-CT (Fig. 2). Comparison of those values for the 13 patients for whom there were both NCE-CT and CTE scans revealed that these differences were significant (P = 0.015, paired samples t-test). The mean HU values of each location on each side are provided in Table 1.

View Full Size

Intraconal fat density measurements over contrast-enhanced (CE, A) and non-contrast-enhanced (NCE, B) axial CT scans of the same patient at the same anatomical location. The differences between the CE and NCE scans were significant (P = 0.015, paired samples t-test) among the 13 patients who underwent both types of scans

Citation: Imaging 2024; 10.1556/1647.2024.00246

• Download Figure
• Download figure as PowerPoint slide

Comparison of the mean HU values in the left vs. the right orbit revealed that there were no differences in the mean values in almost all locations (Table 2).

A comparison of the delta value obtained by subtracting the temporal HU from the nasal HU in each orbit revealed similar differences between the right and left orbits (3.5 ± 13.3 OD vs. 2.3 ± 3.4 OD, P = 0.6, pair-samples t-test). In addition, these delta values were not influenced by the presence of contrast (2.6 ± 4.7 CE vs. 2.2 ± 2.5 NCE, P = 0.7, paired-sample t-test).

The HU values were higher in the nasal orbit compared to the temporal orbit, both on NCE-CT (−75.8 ± 7.5 nasal vs. −78.1 ± 8.4 temporal, P < 0.001, paired-sample t-test) and CE-CT (−72.7 ± 6.7 nasal vs. −74.6 ± 8.6 temporal, P = 0.02, paired-sample t-test, Fig. 3). There was no significant sex-related difference in the mean HU values for the left and right orbits (independent samples test, P > 0.05). There was also no correlation between the patient's age and the nasal, temporal, or mean HU values in both the NCE-CT and CE-CT scans (Pearson test, P > 0.05).

View Full Size

A comparison of the mean Hounsfield unit (HU) values on non-contrast-enhanced (NCE) and contrast-enhanced (CE) CT scans of the left eye in the nasal and temporal orbits in 54 patients with normal orbits. The mean HU measurements are represented on the X-axis, and the mean HU values are shown on the Y-axis. HU values were higher in the nasal orbit compared to the temporal orbit, both on NCE-CT (P < 0.001) and CE-CT (P = 0.02)

Citation: Imaging 2024; 10.1556/1647.2024.00246

• Download Figure
• Download figure as PowerPoint slide

Discussion

We found that the density of nasal intraconal fat was higher than that of temporal intraconal fat, both on NCE-CT and CE-CT scans of normal orbits. This has important clinical implications when evaluating imaging studies in various orbital diseases.

The intraconal orbit, which comprises the ON, extraocular muscles, blood vessels, and other soft tissues, can be divided into two distinct regions based upon mid-orbit scans, specifically, the nasal and temporal orbital fat [1, 14]. Of note, these are not distinct anatomic regions, but rather parts of a continuous intraconal fat compartment around the ON.

The orbital fat and the connective tissue are not a single continuum of the same structure. For example, the preaponeurotic fat has higher carotenoid content which makes it more yellow, and it is possible that preaponeurotic fat selectively absorbs lutein and beta-carotene from plasma [15]. Bremond-Gignac et al. [16] identified two distinct types of fat in the orbit. The posterior orbit around the ON had relatively loose fibrous septa and the adipocytes were larger, with a mean diameter of 75 µm. In the anterior orbit and close to the extra ocular muscles (EOM), the septa were denser and the adipocytes were smaller with a mean diameter of 55 µm. Also, the fibrous septa were fewer around the ON, probably related to the greater movement of the nerve with ocular rotation. Such variability in microanatomy and histology in different regions of the orbit can very likely produce slight variations in HU values on CT scans. Our findings appear to indicate that these variations do not necessarily relate to pathology, and may rather be normal variations from one small area to another.

Fat density can vary due to several factors, including differences in blood supply, the presence of fibrous tissue, and the arrangement of the fat cells themselves. Alternatively, the proximity of the medial rectus muscle to the nasal side of the orbit could result in compression of the neighboring fat cells, leading to increased density. The figures quoted in Koornneef's anatomical and histological investigation published in 1977 [12] suggested that the nasal intraconal orbit exhibits a smaller size and a higher density than the temporal intraconal side. It is plausible that this structural variation may correspond to the density measurements obtained from contemporary imaging scans, but further research is warranted to explore this theory. In the mid-orbit, there are more anatomic structures supported by fascial connective septa, such as the superior ophthalmic vein and the nasociliary nerve in the nasal orbit than in the temporal orbit. Moreover, because the medial fat region is narrower on a CT scan, the medial rectus muscle pulley system is more likely to be incorporated in a HU measurement than the lateral rectus muscle pulley in the lateral orbit, calling for an alternative explanation.

A noteworthy discovery is that metastatic cancer in the orbit often affects the nasal quadrant [17, 18]. In one study, the medial orbit was the most commonly affected quadrant (23%), followed by the lateral orbit (18%), superior orbit (14%), and orbital apex (14%) [17], despite the fact that the nasal quadrant takes up less space. This observation can be partially attributed to the notion that the medial orbit may have a more abundant vascular supply. However, it must be noted that there is no conclusive evidence supporting differences in vascularity among orbital quadrants. The blood supply to the orbit is intricate and comprises diverse vessels that deliver blood to various regions of the orbit, and there may be variations in blood supply among individuals [19]. In the present investigation, the delta values representing the difference in HU values between the nasal and temporal intraconal fat compartments were not influenced by the administration of contrast material. This finding suggests that even if there were variations in vascular supply between the two compartments, they did not significantly influence the density measurements obtained on the CT scans.

We also demonstrated that there were no significant discrepancies in HU values between the left and right orbits. This outcome is consistent with a Korean study which showed no significant variations in measurements of orbital structures between the two orbits, including the extraocular muscle diameter, optic nerve width, and globe position [20], and with the study by Forbes et al. who reported comparable mean values of fat volume in the right and left orbit [21], both further supporting our findings.

We found no notable variations in HU values between sexes or across age groups, implying that differences in HU values may derive from individual variations in intraconal fat density rather than demographic factors. Volumetric measurements in an earlier investigation of Graves' orbitopathy revealed that orbital fat density was not dependent upon age in orbits that are not afflicted by the disease.¹⁶ Patients with Graves' orbitopathy were shown to exhibit a higher average density of orbital fat compared to healthy individuals, and this density did not appear to alter with age [22]. Contrarily, a study conducted among healthy East Asians observed that the volume of orbital fat increased with age in both males and females [23] and that males generally have larger volumes of orbital tissue, except for orbital fat volumes [23]. Another study by Forbes et al. quantified the volume of normal orbital soft tissue and reported slight variations between males and females, but negligible disparities between the right and left orbits [24]. Based upon our own current findings, HU values appear to serve as a dependable indicator of intraconal fat density, independent of age, sex, and laterality. Of note, while we did not find any influence of age on HU values, it is important to acknowledge that measuring intraconal fat density in a large pediatric population may yield different results. Children's anatomical characteristics and developmental stages could potentially impact the density measurements, thus warranting further investigations specifically focused upon pediatric populations to better ascertain any age-related variabilities in orbital CT imaging.

Intraconal orbital fat plays an important role in maintaining the position of the globe. Several studies have reported cases of proptosis caused by increased orbital fat volume, among them patients with Graves' orbitopathy, Cushing disease/syndrome, and obesity without endocrinopathy [25]. Clinical indications for fat decompression have traditionally focused upon reducing exophthalmos, orbital pain, and congestion [26–28]. However, it has been shown that removing orbital fat can also reduce enough orbital pressure to reverse optic neuropathy in some cases [29], most likely due to the decreased intra-compartment pressure resulting from severing the orbital septa during decompression surgery.

The finding that HU values are a reliable indicator of intraconal fat density can aid in the diagnosis and management of orbital pathologies. For instance, they can be used to quantify changes in fat density due to various pathologies or treatments, such as orbital cellulitis [30]. Additionally, our understanding of the role of intraconal fat in maintaining globe position and its potential impact on other contained structures, such as the ON, underscores the importance of accurately measuring and monitoring intraconal fat in patients with orbital pathologies. This can facilitate better informed decision-making regarding the need for interventions, such as fat decompression surgery. Overall, our study provides valuable new insights into the characteristics of intraconal fat and its potential clinical implications, and highlights the use of objective measures to identify subtle differences in orbital CT scans that are imperceptible to the human eye.

Despite its many advantages, including excellent examination of bone anatomy, high availability and short acquisition time, the widespread use of CT over the past three decades has resulted in a significant increase in the cumulative radiation exposure of the general population [3, 31–33]. This rise may potentially lead to a higher incidence of radiation-related conditions, such as cataracts and malignancies, in the foreseeable future [32]. An even greater concern is the growing use of CT scans among pediatric patients [34]. The estimated lifetime cancer mortality risks from radiation in pediatric CT examinations are considerably higher than those in adults [35]. MRI is the preferred imaging modality for children as it avoids ionizing radiation exposure [36]. It is frequently used in the follow-up of various ocular and orbital pathologies, providing superior soft tissue resolution [31].

Various imaging modalities can be employed to characterize orbital structure and volume, as well as to evaluate orbital growth and the impact of pathological processes and treatments on both developing and mature orbits. When orbital volume analysis is necessary, CT remains the preferred modality due to its superior depiction of bony anatomy compared to other options, such as magnetic resonance imaging (MRI) and ultrasonography [3, 31]. Measurement of intraconal fat can also be of interest and a compelling topic for future studies. Currently, various studies have demonstrated the mean orbital volume in adults to be around 30 cm³ [33], with the upper limit of normal orbital fat on CT being 14.4 cm³ [24]. One study measured the mean and standard deviations for orbital fat volume in both control and proptosis groups, as well as the etiologic subdivisions of the proptosis group, finding the control group to have a mean fat volume of 8.16 cm³ and the proptosis group 11.04 cm³ [25]. In patients with Graves' disease, the mean fat volume in the right and left orbits ranges from 3.77 to 22.63 cm³ on CT scans [21]. Additionally, a study found that orbital fat density in patients with Graves' Orbitopathy was significantly higher than in controls and was negatively correlated with fat volume [22].

The current study has several limitations that should be considered when interpreting the results. One limitation is the relatively small sample size, which may limit the generalizability of the findings. Another limitation is the use of patients who underwent CT scans for various clinical indications, which may introduce some selection bias. Furthermore, we did not assess the potential impact of selective manual measurement errors on the HU measurements in our study, which could affect the accuracy of our results. Finally, our study was limited to HU measurements, and future studies may benefit from incorporating additional imaging modalities or biomarkers to further explore the relationships between intraconal fat density and various pathologies.

In summary, the findings of this study suggest that there are distinct differences in the density of nasal and temporal intraconal fat on CT scans of normal eyes, and that these differences may aid in the identification and differentiation of orbital pathologies. They also indicate that there are no significant differences in HU values between the sexes, age groups, or laterality. Further studies are needed to validate these findings and explore their clinical applications in greater depth.

Authors' contribution

BS Einav: Data curation, Formal analysis, Writing – original draft, Writing – review & editing.

G Ghal: Conceptualization, Data curation, Writing – original draft, Writing – review & editing.

S Gangadhara: Formal analysis, Writing – original draft, Writing – review & editing.

JBS Guy: Formal analysis, Writing – original draft, Writing – review & editing.

LP Daphna: Conceptualization, Data curation, Formal analysis, Writing – original draft, Writing – review & editing.

All authors reviewed the final version of the manuscript and agreed to submit it to IMAGING for publication.

Funding sources

No financial support was received for this study.

Conflict of interests

The authors have no conflict of interest to disclose.

Ethical statement

The described research adhered to the tenets of the Declaration of Helsinki, and Institutional Review Board (IRB) approval was obtained (decision/protocol number: SMC-9972-22). Patient consent was waived for this anonymized retrospective study.