CT and MR imaging of orbital inflammation
Orbital inflammation can be idiopathic or in the context of a specific disease and it can involve different anatomical orbital structures. On imaging, inflammatory disease is frequently mistaken for infection and malignant tumors, and its underlying cause is often not determined. Through this article we aim to improve orbital inflammation diagnosis and underlying inflammatory diseases recognition.
The imaging protocols and characteristics of orbital inflammation were reviewed.
A decision tree for the evaluation of these patients is provided. First, a combination of clinical and radiological clues is used to recognize inflammation, in particular to differentiate it both from orbital infection and tumor. Subsequently, different radiological patterns are recognized, often allowing the differentiation of the several orbital inflammatory diseases.
The use of adequate imaging protocols and subsequent evaluation allow the recognition of an orbital lesion as inflammatory and the diagnosis of the underlying inflammatory disease. All in all, a proper treatment can be established, and at times, a biopsy can be avoided.
KeywordsOrbital inflammation Orbital inflammatory diseases CT MRI Diffusion-weighted imaging
Orbital inflammation may be either idiopathic or in the context of a specific inflammatory disease. It may involve different orbital structures, accounting for the different clinical presentations. Recognizing the inflammatory etiology of a lesion, identifying which structures are involved, and determining the underlying disease is mandatory in order to establish an adequate treatment .
The diagnosis of orbital inflammation is made through combining the radiological findings, laboratory data, and characteristics of other organ involvement. When the diagnosis still remains unclear, tissue characterization and/or a therapeutical test is needed.
This may be due to a number of reasons, the lack of detailed studies concerning the differential diagnosis on radiological imaging of orbital inflammatory diseases being one.
The purpose of this manuscript is to provide a comprehensive review of orbital inflammation, together with a systematic approach for the radiological evaluation of these patients, in order to improve the diagnostic accuracy of orbital inflammation.
CT and MRI protocols will be addressed first. Secondly, the specific radiological characteristics of inflammation affecting the various orbital structures will be illustrated, providing the necessary clues to differentiate orbital inflammation both from orbital infection and tumor. Thirdly, the imaging characteristics of specific inflammatory diseases will be presented, emphasizing the main features that will allow differentiation between distinct etiologies. Finally, a decision tree, combining mainly imaging features and clinical findings, will be provided, which will help in the differential diagnosis of orbital inflammatory diseases.
MRI is the modality of choice for the evaluation of orbital inflammation because of its superior soft tissue contrast and spatial resolution, as well as its possibility to generate functional images such as diffusion-weighted imaging (DWI) and perfusion-weighted imaging (PWI).
Orbital lesions should be evaluated in multiple planes, preferably at least in axial and coronal planes. However, when a lesion is located in the eyelid, in the region of the posterior wall of the globe, or in close relationship with the optic nerve, additional sagittal oblique images should be obtained for an optimal evaluation.
In general, orbital evaluation with MRI is performed by using a head coil. The MRI protocol should include T1-weighted imaging (WI) sequences and T2-WI sequences with and without a fat suppression technique, T1-WI sequences with a fat suppression technique after contrast medium administration and DWI. T1-WI and T2-WI are the standard anatomical images to be obtained. They are important to determine which orbital structures are involved and to what extent. Inflammatory lesions are hypo- to isointense on T1-WI. On T2-WI sequences, the signal intensity of inflammatory lesions depends on the balance between edema and fibrosis, edema being hyperintense and fibrosis hypointense. The use of T2-WI sequences with fat suppression will make edema more conspicuous. Fat-suppression techniques after contrast sequences allow for the differentiation between an abnormal enhancing lesion and the normal bright signal of fat on T1-WI. Enhancement pattern can be important in differentiating inflammation from tumor and infection. Inflammatory lesions will tend to show a more homogeneous enhancement pattern while tumors and infections will be more heterogeneous due to the presence of non-enhancing components such as necrosis and pus. DWI can be performed using either echo planar imaging (EPI) or non-EPI-based sequences. An EPI-based sequence is the traditional choice for DWI. It has high temporal resolution but is sensitive to susceptibility artifacts and image distortion, especially present at air-tissue and bone-tissue interfaces, making it a challenging technique in the orbit. A non-EPI technique takes longer but does not show image distortions and susceptibility artifacts. DWI helps distinguish benign from malignant lesions. In the study from Sepahdari et al. with 189 cases, orbital masses were likely to be malignant (> 90% probability) when ADC < 0.93 × 10–3 mm2/s and likely to be benign (> 90% probability) when ADC > 1.35 × 10–3 mm2/s. Inflammatory lesions due to its higher free water content will have less diffusion restriction and therefore will show high ADC values. Meanwhile, malignant tumors having higher cellular content will restrict water diffusion and show low ADC values. In cases of bacterial infection, the presence of pus will be responsible for restricting diffusion and consequently high signal on DWI, matching the non-enhancing portion of the mass [8, 9, 10]. PWI can be performed in the orbit but few studies have been published. Most used a dynamic contrast-enhanced technique (DCE) in which serial T1-weighted images are acquired before, during, and after contrast administration. It provides data in the wash-in and wash-out contrast kinetics within a lesion. In DCE-MRI, the qualitative evaluation of the time intensity curve (TIC) pattern seems to be a complementary investigation in distinguishing benign from malignant lesions. In the study from Yuan et al., a persistent TIC pattern (type I curve) suggests a benign lesion, a wash-out TIC pattern (type III curve) mostly suggests malignancy, and a plateau TIC pattern (type II curve) occurs both in benign and malignant lesions .
In an emergency setting, computed tomography (CT) is often the first-line imaging modality because of its availability, high temporal resolution, and allowing oftentimes the diagnosis of a mass lesion. It may also identify a metallic foreign body that could become harmful during MRI examination. However, CT diagnostic performance compares negatively with MRI, namely in the differentiation between inflammation and tumor, as it lacks the information obtained through DWI and due to its worse soft tissue contrast and spatial resolution.
Ultrasonography is another alternative imaging method to diagnose inflammation or tumors of the globe in selected cases, but the technique is operator dependent and shows limited capacity in the evaluation of the retrobulbar structures [12, 13].
Characteristics of orbital inflammation involving different orbital structures
Although infection may cause inflammation, in this paper, when referring to inflammatory disease, we are considering non-infectious inflammation.
Imaging characteristics of inflammation involving different orbital structures
Orbital inflammation/orbital structure
Scleritis (Fig. 1)
Scleral enhancement, scleral thickening, no DWI restriction, focal periscleral cellulitis
Uveitis (Fig. 2)
Uveal tract increased enhancement, uveal tract thickening, no DWI restriction, subretinal effusions, vitreous humor signal abnormalities
Dacryoadenitis (Fig. 3)
Lacrimal gland enlargement both involving the orbital and palpebral lobes, no DWI restriction, surrounding cellulitis
Optic perineuritis (Fig. 4)
Optic nerve sheath enhancement, no DWI restriction, surrounding cellulitis
Optic nerve enhancement and hyperintensity on T2 and FLAIR, DWI restriction possible
Myositis (Fig. 5)
Muscle enlargement, with increased enhancement, no DWI restriction, tubular/fusiform configuration, surrounding cellulitis
Cellulitis (Fig. 6)
Preseptal fat thickening, pre and postseptal fat infiltration and enhancement, no DWI restriction
Characteristics of orbital inflammation due to different inflammatory diseases
Main imaging characteristics of the most common orbital inflammatory diseases
Orbital inflammatory disease
Main imaging characteristics
IOI (Fig. 8a–d)
Pain, can involve all orbital structures, myositis with tubular configuration, cellulitis
Sarcoidosis (Fig. 8e)
Similar to IOI but: pain unusual, uveitis most common manifestation, predilection for antero-inferior quadrant
Graves’ D (Fig. 9a, b)
Bilateral, myositis with predilection for inferior and medial quadrant and with fusiform configuration, increased orbital fat, no cellulitis
IgG4 RD (Fig. 9c–e)
Bilateral, chronic course, predilection for lateral and superior quadrant with myositis and dacryoadenitis, myositis with fusiform configuration, cellulitis, infraorbital nerve involvement
Granulomatosis with polyangiitis (Fig. 10)
Predilection for extraconal and conal compartments, chronic sinonasal involvement with bone destruction
ISOI (Fig. 11)
Chronic course, predilection for lateral and superior quadrant with myositis and dacryoadenitis, enophthalmus possible
Sarcoidosis is a systemic inflammatory disease of unknown etiology, characterized by the presence of granulomas in the affected organs . Lungs and skin are most commonly affected. Sinonasal involvement is rare . Orbital involvement is seen in 25–60% of patients with systemic sarcoidosis . Sarcoidosis can involve any orbital compartment , similar to IOI. However, in sarcoidosis, uveitis is the most common manifestation, the antero-inferior orbital quadrant is involved to a greater degree, the cavernous sinus can be affected as well and isolated myositis is rare (Fig. 8e) . The involved orbital structures are hypointense on T2, enhance and no DWI restriction is expected. Clinical presentation is subacute evolving from months to years. Pain is not a typical feature. Orbital involvement is unilateral in 75% of cases. Isolated orbital granulomatous involvement, in the absence of systemic disease, should not be called orbital sarcoidosis, as it may represent an idiopathic granulomatous orbital inflammation, probably a different entity, affecting especially men in the fourth decade and in 50% of cases affecting the lacrimal gland [26, 29]. Serum angiotensin-converting enzyme (ACE) is increased in 60 to 90% of patients with active disease and reflects its severity [1, 30]. Oral steroids are the mainstay of treatment. Cytotoxic agents (v.g. methotrexate) are used as second line. Surgical excision may be considered for localized orbital disease, namely the eyelid .
Immunoglobulin G4-related disease (IgG4 RD) is a recently described systemic inflammatory process of unknown etiology . Any organ can be involved but there is a predilection for the orbits, salivary glands, lymph nodes, pancreas, and hepatobiliary system. Mickulicz disease, previously thought to be a subtype of Sjögren’s syndrome, is now considered part of the IgG4 RD [34, 35]. IgG4 RD of the orbit has an indolent chronic course with symptoms evolving on average for 45 months at time of diagnosis . Pain is not a characteristic finding [24, 36]. On IgG4 RD of the orbit, the extraocular muscles are the most common orbital structure involved (89%). Myositis is mostly bilateral (88%). The lateral rectus is the most affected muscle (76%) and typically enlarged to the greatest degree. The tendon is spared in 96% of cases, giving the muscle a fusiform configuration . The lacrimal gland is the second most common orbital structure involved (70%) and its involvement is more common bilateral (58%) (Fig. 9 c–e) [24, 34]. Cellulitis is present in 44%, either pre- or postseptal or uni- or bilateral . Perineural involvement has been reported, mostly affecting branches of the trigeminal nerve, the infraorbital nerve being involved in 30% and mostly unilateral [24, 35, 37, 38, 39]. There is expansion of the foramina . In 89% of patients with IgG4 RD, there is sinusal disease as well . At imaging, orbital IgG4 RD lesions are diffuse or tumefactive, homogeneous, hypointense on T2-WI, enhancing, with no DWI restriction [37, 40]. Bone remodeling is possible . Increased IgG4 levels in serum will help in making the diagnosis , but serum IgG4 can be normal in up to 40% of patients with biopsy-proven disease [32, 38]. The definitive diagnosis is histopathologic typically with abundant IgG4-positive plasma cells and fibrosis [38, 41]. Lymphoma can be a complication of IgG4-related disease [35, 36, 38]. Although the most effective therapy of IgG4-RD has yet to be defined, rituximab is a promising alternative to glucocorticoids [3, 42, 43].
Decision tree in orbital inflammation
Orbital inflammation presents on radiological imaging as a solid-enhancing lesion, mostly as an ill-defined or infiltrative lesion. The differential diagnosis of an orbital solid-enhancing lesion is however vast including not only inflammation but also infection, benign and malignant tumors, and vascular malformations (e.g., cavernous hemangioma). In the presence of an orbital enhancing solid mass one should first recognize its inflammatory nature and second try to determine the underlying inflammatory disease.
Population is first dichotomized according to the DWI and subsequently whether pain and cellulitis are present. Diffusion restriction with no pain and no cellulitis point to malignant tumor, and biopsy should be envisaged. Facilitated diffusion together with no pain and no cellulitis can still correspond to inflammation, but other diagnosis such as a benign tumor or a vascular malformation should be kept in mind. The absence of diffusion restriction, in the presence of pain and/or cellulitis, favors an inflammatory or infectious process. The differentiation between inflammation and infection is mainly based on clinical features, with long-standing symptoms pointing to inflammation, while the presence of fever, high infectious parameters (v.g. leukocytosis, elevated CRP), and pus, in which case DWI restriction should be expected, suggesting infection. On imaging, the presence of sinusitis and/or an abscess points to infection, while scleritis suggests inflammation.
Imaging may also play an important role on establishing the diagnosis of the underlying inflammatory disease. This is especially important if the patient is not known to harbor any inflammatory systemic disease. In cases the inflammatory process shows predilection for the muscles at the inferior/medial quadrants of the orbits, the involved muscles have a fusiform configuration, and no cellulitis is present, Graves’ disease should be considered. Graves’ disease is mostly bilateral and symmetric and there can be increased intraorbital fat. In cases of IgG4-related disease or sclerosing orbital inflammation, the lateral/superior quadrants of the orbit are preferentially involved and both have an indolent course. IgG4-related disease is mostly bilateral, the involved muscles have a fusiform configuration, and in 30% of the cases, there is enlargement of the infraorbital nerve, which when present is very suggestive of the diagnosis. On sclerosing orbital inflammation, enophthalmus can exist suggesting the fibrotic process. If there is a predilection for the extraconal space, with or without chronic sinonasal involvement, with bone destruction, consider granulomatosis with polyangiitis. History of uveitis and a predilection for the antero/inferior quadrant suggests sarcoidosis. Sarcoidosis is mostly unilateral and with a subacute presentation. When pain is a predominant feature, the involved muscles having a tubular configuration and cellulitis is present, consider idiopathic orbital inflammation. Idiopathic orbital inflammation is mostly unilateral, and when involving the muscles, it affects especially the medial followed by the superior and lateral recti.
The definitive diagnosis of the orbital inflammatory disease is made by combining the radiological pattern with the laboratory findings and characteristics of other organ involvement. The radiological pattern can be specific for a certain type of orbital inflammation such as in Graves’ disease or in granulomatosis with polyangiitis. However, sometimes, these patterns are shared between different etiologies making the imaging pattern not specific. Still the evaluation of the radiological pattern will shorten the differential diagnosis. That is helpful as it can guide the laboratory evaluation and eventual imaging of other organs. When the diagnosis is still unclear, tissue characterization and/or a therapeutical test is needed. An orbital biopsy is easily considered for accessible orbital lesions such as dacryoadenitis. Locations where surgery is difficult or dangerous, such as the orbital apex or around the optic nerve, may confer a higher threshold for biopsy [51, 52].
Orbital inflammation is frequently mistaken either for orbital infection or malignant tumors, and its underlying cause is often overlooked. Imaging findings obtained through appropriate protocols and knowledge of the most common orbital inflammatory diseases will help shorten the differential diagnosis, with important therapeutic and prognostic consequences. We have therefore combined different imaging and clinical clues that will allow one to recognize an orbital solid-enhancing lesion as inflammatory. Subsequently we have shown how the different radiological patterns will help in differentiate the possible orbital inflammatory diseases. Overall these considerations enable the treating physician to establish an adequate treatment, and at times, a biopsy can be avoided.
The authors thank Gerrit Kracht for his technical support in the development of this article and Prof. Albert De Roos for the article revision.
Compliance with ethical standards
No funding was received for this study.
Conflict of interest
The authors declare that they have no conflict of interest.
All procedures were performed in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
For this type of study formal consent is not required.
- 41.Deshpande V, Zen Y, Chan JK, Yi EE, Sato Y, Yoshino T, Kloppel G, Heathcote JG, Khosroshahi A, Ferry JA, Aalberse RC, Bloch DB, Brugge WR, Bateman AC, Carruthers MN, Chari ST, Cheuk W, Cornell LD, Fernandez-Del Castillo C, Forcione DG, Hamilos DL, Kamisawa T, Kasashima S, Kawa s KM, Lauwers GY, Masaki Y, Nakanuma Y, Notohara K, Okazaki K, Ryu JK, Saeki T, Sahani DV, Smyrk TC, Stone JR, Takahira M, Webster GJ, Yamamoto M, Zamboni G, Umehara H, Stone JH (2012) Consensus statement on the pathology of IgG4-related disease. Mod Pathol 25:1181–1192CrossRefGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.