Idiopathic orbital inflammation [4–6] (IOI) is a term used to describe a syndrome of nonspecific inflammation of orbital tissue with no identifiable local or systemic cause. This condition was first described by Birch-Hirschfeld in 1905, who coined the term “pseudotumor ” [7]. Numerous other names have been proposed for this group of disorders, including idiopathic inflammatory syndrome, inflammatory orbital pseudotumor , and nonspecific orbital inflammatory syndrome [8–10]. IOI is currently a diagnosis in transition, as improving diagnostic capabilities have chipped away at what was once “idiopathic” or “nonspecific”. Sarcoidosis, granulomatosis with polyangiitis (GPA, Wegener granulomatosis), thyroid eye disease (TED), IgG4-related inflammation (IgG4-RD), Sjögren syndrome, systemic lupus erythematosus, and histiocytic syndromes (Erdheim-Chester and Rosai-Dorfman diseases) are no longer included under the umbrella term “nonspecific”; they are now classified as specific orbital inflammations with their own immunologic or histopathologic markers. Table 34.1 summarizes a common classification scheme of IOI, the more common types of specific orbital inflammations, and other disease entities which may mimic IOI and should therefore be considered in the differential diagnosis.

Although IOI is grouped among the more common orbital disorders in adults, it is postulated to be rare in the pediatric population. Early studies demonstrated that patients under the age of 18–20 years comprise between 6.6% and 16% of all IOI patients [11–13].

Our understanding of the etiology of IOI has benefited from recent advances in immunologic and histopathologic abilities to identify and measure cytokines and antibodies. The classic signs and symptoms of IOI presumably represent the clinical manifestation of one or more autoimmune and cell-mediated processes, the triggers for which are a source of current scrutiny. A “pathologic construct” proposed by Dr. Gerald Harris has correlated specific cells present in surgical specimens with particular clinical symptoms [5]. For example, macrophages activated by pro-inflammatory cytokines cause lysis of cells, escalating a cascade of inflammation and causing tissue injury and fibrosis which can result in the clinical symptoms of orbital pain, swelling, and impaired function [5].

More recently, Wladis and colleagues have demonstrated that gamma interferon and interleukin-12 levels are markedly elevated in IOI when compared to normal, noninflamed tissue profiles and that IOI may be a T-cell-mediated phenomenon [14]. The results of further studies should not only contribute to our understanding of the pathology of IOI, but also may identify targets for treatment with specific biologic agents.

The classic presentation of IOI includes the abrupt onset, often over the course of hours, of periorbital pain accompanied by eyelid edema, erythema, and chemosis (Fig. 34.1). Other common features include proptosis, diplopia, and visual changes. When supported by positive findings on appropriate imaging studies and in the absence of any other identifiable cause, this clinical presentation is considered by some to be diagnostic.

In contrast to adults, pediatric patients are more likely to present with bilateral disease [13, 15–17] (Fig. 34.2) and ptosis [18]. Constitutional signs and symptoms, such as headache, fever, and malaise, may also be more common in children [15, 18–20], and children are more likely to have recurrent disease [1, 15]. Although pain is reported in over 75% [1] of adult patients with IOI, two recent cohort studies suggest that periorbital pain may not be a meaningful indicator of IOI in children, reported in only 17% and 57% of cases, similar to other inflammatory processes [18, 19].

Many of the signs and symptoms of pediatric IOI mimic those of infection. Since orbital infection is the most common cause of acute proptosis in children and has the potential for rapid progression, it is of paramount importance to rule out an infectious etiology (especially orbital spread of adjacent paranasal sinusitis) in all patients with suspected IOI (Fig. 34.1d).

The initial symptoms and signs of IOI in children are summarized in Table 34.2.

Dedicated orbital imaging is an essential part of the diagnostic protocol for any orbital process. Imaging may be by either CT or MRI, although both modalities are associated with drawbacks particular to the pediatric population.

The preferred orbital imaging modality for the evaluation of suspected pediatric IOI remains orbital computed tomography (CT). CT is easily available, images surrounding bone in detail, and provides adequate soft tissue information. When possible, intravenous contrast should be used. Pediatric CT protocols are essential to minimize radiation exposure. Orbital inflammatory lesions have a common radiologic characteristic of an irregular margin adjacent to the focus of the inflammation, despite a variable location of involvement in the orbit, aptly described by Jakobiec as “crabgrass in the orbit”, especially when the process involves the intraconal fat [2, 8]. The involved tissues usually enhance with contrast media. Bone destruction or hypertrophy is not seen in IOI and would suggest malignancy or infection. Adjacent paranasal sinus opacification is not often seen, even though IOI may follow an upper respiratory infection (Figs. 34.1d and 34.2c); the presence of sinus opacification raises the possibility of an infectious etiology. Multi-scan CT techniques permit superb views in all planes from a rapid (<60 s) scan in the supine position.

Improved magnetic resonance imaging (MRI) technology has led to more rapid scanning times, albeit considerably longer than typical multi-scan CT, making MRI feasible for pediatric patients and avoiding any radiation exposure; younger children may still require sedation to minimize motion artifacts during longer MRI sequences. While MRI does often reveal more detailed information about involved soft tissue, it does not usually provide much more information about IOI lesions when compared to CT , provides less information about bone changes, and is usually more expensive. B-scan ultrasonography can be used for evaluating anterior orbital lesions, but provides no information about surrounding bones, does not image the posterior one-third of the orbit, and is usually more difficult to interpret than CT and MRI.

The evaluation of patients suspected of having IOI is summarized in Table 34.3. The work-up should include a review of systems and a physical exam, as there may be systemic symptoms caused by the orbital inflammation or by a preceding upper respiratory infection; the symptoms may also herald the diagnosis of an underlying systemic autoimmune disease. Headache, vomiting, sore throat, anorexia, abdominal pain, lethargy, and weight loss have been noted in some cases. IOI alone should not cause a significant fever, although a low-grade fever may be present in some cases. Baseline laboratory studies for patients with IOI may include a complete blood cell count (CBC) with differential (cytopenias are seen in inflammation and malignancy), erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), comprehensive metabolic panel (CMP), urinalysis (possible urine protein/creatinine and urine calcium/creatinine ratios), antinuclear antibody (ANA) (if positive, should be followed by evaluation for specific antibodies), rheumatoid factor (RF), angiotensin converting enzyme, anti-neutrophil cytoplasmic antibody (ANCA) [21], thyroid function studies (TSH, free T3, free T4, thyroid-stimulating immunoglobulin), EBV serologies, placement of PPD, fecal calprotectin, serum total immunoglobulin G, and/or Lyme serology in patients living in or having traveled to endemic areas. The eosinophil count may be elevated in some cases of pediatric IOI. While these studies cannot “rule in” IOI, they may be helpful in ruling out other entities. Of note, not all patients require an extended work-up, especially if the clinical and radiographic presentation is classic for IOI.

The role of orbital biopsy in the diagnosis of IOI has been debated at length in the literature [5, 6, 16, 22]. Briefly, some experts posit that an orbital biopsy should be attempted in all patients before beginning steroid treatment of presumed IOI, provided the tissue in question is easily accessible, arguing that “inflammation” is not a diagnosis but may be a sign of a potentially dangerous underlying tissue process [6]. Others counter that a corticosteroid trial is part of a diagnostic and therapeutic protocol in patients who present in a manner typical for IOI and who have corroborating imaging studies, and that orbital exploration may expose the typical IOI patient to unnecessary surgical risk [1, 5, 15]. These experts reserve orbital biopsy for patients with an atypical presentation, for those who do not experience an immediate and sustained response to corticosteroids, and for those whose symptoms recur.

Philosophic differences aside, clinical practice typically marries axioms of both camps. Most agree that in the case of orbital myositis or an inflammatory lesion of the orbital apex, the risks of iatrogenic injury should be weighed against the benefits of histologic confirmation [23]. Similarly, advocates of biopsy argue that there is an unacceptably high incidence of malignancy in lacrimal gland masses [8, 24] and that any easily accessible involved tissue should be biopsied. In addition, any atypical presentations (subacute onset of symptoms, recurrence of symptoms on or after steroid taper, bone erosion on imaging) or a history of local or distant malignancy should prompt a biopsy.

In the pediatric population, where the specter of malignancy includes metastatic neuroblastoma, rhabdomyosarcoma, and leukemic lesions, among others (see Chap. 36), which may present in a manner similar to IOI, few would argue with the importance of tissue biopsy of accessible tissue, especially the lacrimal gland.

When performed, histologic evaluation should include consideration for malignancy, granulomas (e.g. sarcoidosis), inflammatory bowel disease and tuberculosis, vasculitis (e.g., ANCA-associated vasculitis, polyarteritis nodosa)), atypical infection (fungus, mycobacteria), and IgG4 staining. If tissue biopsy (especially of the lacrimal gland) for suspected IOI is performed, consideration should be given to debulking with or without intralesional corticosteroid injection. A recent study concluded that this may decrease recurrence of IOI [25]. If possible, systemic corticosteroid treatment should be withheld until biopsy has been performed to minimize any masking of serious pathology by corticosteroid therapy.

Treatment of IOI is usually initiated based on characteristic clinical and orbital imaging findings. A bacterial infection must be excluded if there is one or more of the following findings: pyrexia, increased white blood cell count (WBC) with neutrophilia and left shift, or CT- or MRI-documented sinusitis. One should also consider the possibility of infection or inflammation from an occult, organic foreign body in children, who frequently do not admit to a history of previous trauma. If infectious orbital cellulitis or dacryoadenitis is a possibility, it is best to obtain blood cultures and treat with broad-spectrum systemic antibiotics for 48 h (see Chap. 33, Orbital Infections) before initiating corticosteroids; of note, there is no clear evidence that corticosteroids, when used in conjunction with systemic antibiotics, worsen the course of pediatric orbital cellulitis [26].

IOI can be classified into six groups based on anatomic location and clinical presentation: anterior, diffuse, diffuse sclerosing, apical, myositic, and lacrimal [8]. Anterior, myositic, and lacrimal are the most common presentations and also usually have the best response to CS therapy.

Granulomatosis with polyangiitis (GPA) is a necrotizing granulomatous vasculitis, usually affecting small vessels. It classically presents with constitutional symptoms and affects the respiratory tract, the sinuses, and the kidneys, although it can affect any organ system [67, 68]. It rarely occurs before adolescence. In a recently published pediatric cohort, the large majority of patients had sino-nasal and/or ear involvement along with constitutional symptoms and respiratory involvement [69]. Subglottic stenosis is not an uncommon complicating feature of pediatric disease. If untreated, the systemic variant of GPA may have a mortality of >90%. The disease may burn out in 1–2 years but may persist for longer. Although no single finding or laboratory test is diagnostic, the majority of patients with systemic GPA will have positivity for anti-neutrophil cytoplasmic antibodies (ANCA) . A child or adolescent is said to have GPA by EULAR/PRINTO/PRES 2008 criteria if at least two of the following criteria are present: (1) granulomatous inflammation on biopsy, (2) upper airway involvement, (3) laryngo-tracheo-bronchial stenosis, (4) pulmonary involvement, (5) ANCA positivity, or (6) renal involvement [70].

Ocular involvement occurs in 52–61% of patients at some point during the disease course [71], with fewer than 10% involving adnexal structures and 7–30% involving the orbit. The frequency may be lower in the pediatric population [21, 72]. Cases of GPA involving the orbit are typically part of a more limited form of the disease (also known as sino-orbital GPA) affecting the ear, nose, throat and upper airway, but sparing the kidneys [73, 74], and it is important for physicians to appreciate that few cases with disease limited to the orbit progress to systemic disease (Fig. 34.13) [75]. This “limited” form may have a chronic remitting coarse, in contrast to the fulminant “systemic” form which progresses quickly to multisystem damage and renal failure [71, 75]. Ophthalmic involvement can result from contiguous extension of respiratory system involvement in the nose and sinuses or from discrete ocular lesions. Nasal and sinus involvement can spread to the nasolacrimal duct and lacrimal sac leading to epiphora and/or dacryocystitis. Bone destruction on imaging is the rule. GPA can also lead to orbital inflammation, mimicking IOI. Discrete focal lesions not caused by contiguous spread can affect the extraocular muscles, retina, uvea, conjunctiva, lids, cornea, and sclera [72]. Ocular or facial pain was a presenting symptom in 75% of patients in one large series [73]. A third of patients with orbital GPA had associated ocular inflammation, which is less common in IOI [75]. Also in contrast to IOI, GPA often exhibits bilateral disease [73, 75].

Radiologically, GPA can show bone destruction, a highly atypical feature for IOI (Fig. 34.13). Additionally, GPA lesions appear hyperintense on T2-weighted MRI, in contrast to many other orbital lesions, both neoplastic and inflammatory [76]. This discrepancy is thought to be due to the abundance of fibrocollagenous tissue present in the granulomatous lesions. It is important to note, however, that chronic inflammation, as seen in long-standing IOI, may also appear hyperintense on T2 images [77].

Serum ANCA is a sensitive, although not specific, indicator of ANCA-associated vasculitis (AAV) [78]. Immunofluorescence for cytoplasmic-ANCA (c-ANCA) (as opposed to perinuclear ANCA), along with antigen-specific anti-proteinase 3 (Pr3) positivity (as opposed to myeloperoxidase positivity), increases sensitivity and specificity for GPA. The antigen-specific anti-proteinase 3 Ab, often a c-ANCA pattern, is positive in 96% of patients with active generalized disease [79, 80]. It is positive in a lower percentage of patients with limited (sino-orbital) disease [81]. False-positive results of proteinase antibodies are rare in contrast to the rate seen with ANCA testing. Because atypical ANCAs can occur in inflammatory bowel disease, another granulomatous disease, it can sometimes be difficult to distinguish the two diseases. In certain patients, ANCA titers can parallel disease activity. In those in whom this is the case, ANCA titers can be useful in monitoring the response to medical therapy and may help determine relative quiescence when surgical intervention (e.g., dacryocystorhinostomy) might be safest [78].

A biopsy is recommended when feasible in all cases of GPA to confirm the diagnosis, even when c-ANCA is positive [82]. Occasionally, orbital biopsies in patients with GPA show minimal inflammation. Nasal biopsies are more likely than orbital biopsies to show the pathologic triad of vasculitis, granulomatous inflammation with giant cells, and tissue necrosis. For this reason, biopsies are often taken from multiple sites.

Children with orbital inflammation resulting from GPA should be under the care of a pediatric rheumatologist. Other than diagnostic biopsies, reconstructive orbital or lacrimal surgery should be deferred until the clinical course and/or serologic testing suggest inactivity, which can often be induced by high-dose CS; a variety of treatments are used to maintain remission (see Therapeutic discussion). While azathioprine was frequently used in the past [83], currently, rituximab is becoming the treatment of choice [84–87], especially for orbital disease [88]. Many practitioners also use trimethoprim and sulfamethoxazole (co-trimoxazole) as a disease suppressant [89, 90].