There are no blood studies to confirm the diagnosis of sympathetic ophthalmia. Clinical investigations in SO include fluorescein angiography (FA), indocyanine green angiography (ICG), B-scan ultrasonography (US), optical coherence tomography (OCT), and pathological sections of enucleated eyes.

FA in the early venous phase demonstrates multiple hyperfluorescent areas that then demonstrate late leakage at the level of the RPE [22, 29]. Leakage occurs in Dalen-Fuchs nodules and areas of retinal vasculitis [24]. Early blocking is seen in areas occupied by Dalen-Fuchs nodules. The optic nerve head often stains in SO even in cases without optic nerve head edema [24]. Late pooling can be seen in the posterior pole representing multiple, lobular, serous retinal detachments [32].

ICG shows multiple hypofluorescent spots without a hyperfluorescent collar in the intermediate phase. Some of these areas become isofluorescent in the late stage [33–35]. These hypofluorescent areas are thought to correspond to choroidal edema and choroidal inflammatory infiltration [24]. Large areas of hypofluorescence can be detected in the late phase of ICG [35].

US in SO is used to detect gross anatomic changes in the retina and choroid. US can demonstrate diffuse thickening of the choroid as well as serous retinal detachments [36–38].

OCT in SO can be used to detect micro-anatomic changes within the retina. Observable OCT findings in SO include serous retinal detachments, intra-retinal edema, disintegration of RPE, tears in the RPE, elongation of photoreceptors, and disorganization of the retinal layers [29, 38–40]. Dalen-Fuchs nodules on OCT appear as hyper-reflective lesions at the level of the RPE with disruption of the inner segment/outer segment (IS/OS) junction [38, 41].

Pathological sections of eyes with SO (both the injured and the secondarily involved eye) demonstrate marked swelling of the choroid that corresponds to lymphocytic infiltration of the choroid. The inflammatory response is characterized as a diffuse, non-necrotizing, granulomatous inflammation of the entire uvea [22, 29, 42, 43]. Major cell types include epithelioid cells and some giant cells [2]. However, the cellular response is variable from case to case which may explain the wide spectrum of clinical presentations [44]. In severe cases, eosinophils and plasma cells can be observed, especially around the inner choroid [45]. Severe cases are also associated with the presence of pigment in the epithelioid cells [46]. The inflammation, typically but not always, spares the choriocapillaris [21, 45, 47].

Dalen-Fuchs nodules appear in approximately one-third of SO cases [21, 22]. Dalen-Fuchs nodules on pathological sections appear as yellow-white lesions of the mid-periphery, located in the choroid, typically between the RPE and Bruch’s membrane [22, 25]. The RPE overlying the nodules is usually intact, but can vary from atrophic to hypertrophic [42, 44]. Histologically, the Dalen-Fuchs nodules are composed of lymphocytes, histiocytes, and epithelioid cells covered by an intact dome of RPE [22, 42, 44].

SO has been reported after trauma to the eye including non-penetrating trauma with hyphema, perforated corneal ulcers, penetrating foreign bodies, perforating foreign bodies, and malignant melanoma [21, 48–50]. SO has similarly been described as a sequelae from surgical procedures such as trans-scleral cyclo-destructive laser, cataract surgery, paracentesis incisions, iridectomy, irradiation of choroidal melanomas, evisceration, retinal surgeries such as pars plana vitrectomy, and scleral buckling. SO has also resulted from accidental surgical perforation of the eye and from post-surgical endophthalmitis [29, 51–60]. Regardless of the nature of the instigating injury, the ultimate initiating factor is the disruption of the immune privilege of the eye. The immune privilege of the eye is a result of blood–ocular barriers in the retinal vascular endothelium, epithelium of the RPE, retinal and ciliary blood vessels; the absence of lymphatics in the eye, except for the conjunctiva; and a host of tightly regulated molecular expression profiles and atypical immunologic cascades [25, 61–68]. Once these barriers are breached, intraocular proteins are exposed to the immune system and an immunologic reaction against these antigens is initiated. Animal studies have shown that an SO-like syndrome can be induced in mammals with the peripheral (non-ocular) injection of proteins such as rhodopsin, interphotoreceptor retinoid-binding protein, recoverin, and soluble retinal antigen (S antigen) [69–73].

The initial presentation of the intraocular antigens to immune cells is via major histocompatibility molecules (MHC) and the process is regulated via several cytokines. It is hypothesized that certain MHC molecules, as a result of differential inter-molecular interactions, present intraocular proteins more successfully to immune cells. Similarly, certain cytokine variants are more or less able to induce a successful immune reaction. Thus, individuals with certain MHC and/or cytokine types are more likely to develop SO. It follows that the severity of the manifestations of SO might also be affected by the specific types of cytokines or MHC molecules that the individual possesses. Several variants of cytokines and their associated proteins, for example tumor necrosis factor (TNF)-α, TNF-β, TNF receptor 2, and cytotoxic T-lymphocyte-associated protein (CTLA) 4 and interleukin 10 (IL-10), have been shown to be either promote or protect against SO. Patients with these cytokine variants differ both in the severity of their clinical presentation and the amount of steroids they required to control their disease [74, 75]. Similarly, patients with certain MHC variants, such as HLA-DR4, HLA-DRw54, HLA-Bw54, HLA-DRB1*04, and HLA-DQA1*03, are more susceptible to SO and develop more severe variants of SO [76, 77].

There are essentially two broad approaches to the management of SO: preventative by removal of the inciting ocular tissues and/or therapeutic with immunosuppressive/anti-inflammatory treatment.

Enucleation of the injured eye as a treatment for SO was first advocated for by Pritchard in 1851 who suggested that enucleation be performed once the uninjured eye showed signs of serious inflammation [78]. This recommendation was controversial because in some cases enucleation did not impact the course of the disease [2]. A study by Reynard et al. [79] demonstrated that early enucleation, which was defined as enucleation within 2 weeks of the inciting injury, resulted in better visual acuity in the uninjured eye, irrespective of treatment with corticosteroids. A subsequent multivariate logistic regression analysis of these data reaffirmed the conclusion that enucleation prior to 2 weeks after the injury resulted in reduced rates of SO. However, it was noted that eyes with good visual potential should not be enucleated [80]. Generally, enucleation of injured eyes with poor visual potential within 2 weeks of injury reduces but does not eliminate the risk of SO. Evisceration of the eye has also been used as a method to prevent SO [81]. There is a healthy ongoing controversy about the relative benefits of evisceration (improved cosmesis, surgical ease, and surgical risk) vs. enucleation (reduced risk of SO) [82–88].

Immunosuppression is used to treat sympathetic ophthalmia after it becomes manifest. Prior to the use of corticosteroids, approximately 50–60 % of eyes affected by SO became permanently blind [7, 23, 47]. By the last 1970 (well after the introduction of corticosteroids), as many as 64 % of patients who had been treated with corticosteroids had vision 20/60 or better. The cost of visual preservation in these patients was the steroid-associated side effects, with most patients developing Cushing’s syndrome [7]. More modern corticosteroid treatment involves high-dose oral corticosteroids (e.g., 1.0–2.0 mg/kg/day prednisone) continued for a period of 3 months. This is administered with adjunctive topical steroids and cycloplegics as dictated by anterior chamber inflammation. Steroids are subsequently tapered off and the response to treatment is evaluated. Pulsed intravenous steroids (methylprednisolone 1 g/day for 3 days), followed by oral steroids, may be beneficial in severe cases [22, 89].

In an effort to reduce the systemic side effects of corticosteroids, the use of intravitreal steroids may enable reduction of the quantity of systemic steroids that may be required [22].

Some cases of SO are refractory to steroids or require high doses of steroids for prolonged duration that can cause systemic side effects. These cases have been managed with steroid sparing immunomodulatory therapy including cyclosporine, tacrolimus, chlorambucil, cyclophosphamide, methotrexate, mycophenolate mofetil, and azathioprine [22, 25, 90–95]. The preceding drugs can be toxic and can have severe long-term sequelae including infertility and secondary malignancies [28]. Moreover, these agents require the physician to have competence in prescribing these agents and managing complications that may ensue. Current therapy has shifted toward less toxic and more directed immunomodulatory molecules. There have been several reports of SO patients that have responded to treatment with anti-TNF-α monoclonal antibodies including adalimumab and infliximab [96–98].

In the future, with the advent of “personalized medicine,” patients may be genotyped to risk stratify those who are more likely to develop severe variants of SO and treated more aggressively to prevent ocular damage.

Sympathetic ophthalmia is a relatively rare, bilateral granulomatous panuveitis that occurs more commonly after penetrating trauma but also described after intraocular surgical procedures. The initiating event compromises the immune privilege of the eye and induces intraocular inflammation. Systemic corticosteroids are the initial treatment implemented, but these patients often require steroid sparing immunomodulatory therapy to control the uveitis and avoid steroid-associated toxicity.