This is a bilateral condition.
Some reports suggest that this entity is more commonly seen after vitrectomy than is endophthalmitis.
Dalen–Fuchs nodules are commonly seen in the retinal periphery of eyes with this entity but they are not pathognomonic for the disease.
Sympathetic ophthalmia probably is the intraocular inflammatory condition best known to practitioners outside of ophthalmology. Albeit not yet defined at that time, it has been suggested that JS Bach’s poor vision after couching for cataracts may have been due to this disorder. It has also stimulated enormous interest within our field since its clinical description by MacKenzie early in the 19th century. The number of patients having this problem each year is relatively small, but the dread of losing not only the involved eye but the contralateral, untouched eye as well in a potentially sight-threatening process is indeed great, and justifiably so. Further, the potential mechanisms to explain such a process are many, which has provided the more intellectually (and imaginatively) inclined with ample material for thought. Newer observations have certainly helped place this disorder in perspective.
Clinical appearance and prevalence
Sympathetic ophthalmia is a bilateral granulomatous uveitis that occurs after either intentional or unintentional penetrating trauma to one eye. Thereby trauma to one eye (the exciting eye) will result in an inflammatory response not only in that eye but also in the contralateral eye (the sympathizing eye). It is a disease that was known to the ancient Greeks and was referred to in the teachings of Hippocrates. It was recognized also during the early and late Middle Ages and into the 18th and 19th centuries with MacKenzie’s full clinical description. Fuchs published the classic description of the disease in 1905, establishing it as a separate entity. The disease can begin as early as several days after the penetrating insult up to decades later, with the clinical diagnosis becoming apparent in 80% of patients some 3 months after injury to the exciting eye and in 90% within 1 year of trauma. , Sympathetic ophthalmia occurs more often after nonsurgical trauma. Liddy and Stuart reported that the disorder occurred in 0.2% of eyes with nonsurgical wounds, and Holland found the disorder in 0.5% of eyes with nonsurgical trauma. Marak estimated the incidence of this disease to be fewer than 10 cases in 100 000 surgical penetrating wounds. In a review by Gass, in which 26 eye pathology laboratories were surveyed for a 5-year period from 1975 through 1980, sympathetic ophthalmia was diagnosed in 53 eyes (two of every 1000 eyes examined); 55% of these occurrences were posttraumatic. In the same report a survey of 34 retinal surgeons who had performed 14 915 vitrectomies disclosed that sympathetic ophthalmia occurred in nine eyes (incidence of 0.06%), with one eye having vitrectomy as the only operative procedure performed (incidence of 0.01%). Recent studies suggested a change in the major initiating factor of sympathetic ophthalmia, with ocular surgery, particularly vitreoretinal surgery, now being the major risk for this disorder. , This trend is being seen worldwide. In a review of 10 cases of sympathetic ophthalmia in Singapore from 1993 to 2003, the authors report that the disease occurred in three after trauma, and after surgery in seven. Of those seven, four cases were seen after vitreoretinal surgery, but two after diode laser cyclophotoablation and one after Nd:YAG laser cyclotherapy. In their commentary, Kilmartin and colleagues calculated the risk of sympathetic ophthalmia to be 1 : 1152 retinal surgical procedures and perhaps as much as 1 : 800. If this calculation reflects a true trend, the risk would be twice that reported previously for vitrectomy (0.06%). The authors raise the issue as to whether patients should be informed of this risk in the consent form, because the calculated risk is more than twice the 0.05% risk of infectious endophthalmitis after vitrectomy ( Table 23-1 ).
|Cause||Post trauma||Post surgery (especially vitreoretinal)|
|Patients||Males and children (reflecting trauma peaks)||No sex preference (reflects positive impact of injury prevention programs) and increasingly elderly patients (reflects impact of ocular surgery)|
|Incidence||Considered disappearing 30 years ago||Probably increasing (underdiagnosed?)|
|Onset||For 65% within 2–8 weeks (for 90%, onset <1 year from trauma or most recent surgery)||Many delayed presentations (for significant percentage onset >1 year from trauma or most recent surgery)|
|Presentation||Granulomatous panuveitis||Any clinical uveitis|
|Outcome||Enucleation within 2 weeks for prevention of SO||Enucleation solely of SO prevention questionable|
|Visual prognosis||Poor||Reasonable due to moden immunosuppression|
It is interesting to note that no cases of sympathetic ophthalmia were reported in the 467 perforated globes observed by the US military during World War II, the Korean War, and the Vietnam War. However, there was one recent case of a soldier injured in the Iraq conflict. These findings are in contrast to the older literature of the last century, which suggested an incidence of about 2%, and it was believed that 16% of ocular injuries during the American Civil War resulted in sympathetic ophthalmia. In a recent prospective surveillance study in the UK and the Republic of Ireland drawn from a surveillance of 59 million citizens, 17 patients with newly diagnosed sympathetic ophthalmia were reported in a 12-month period, and a minimum estimated incidence of 0.03 in 100 000 was calculated.
Sympathetic ophthalmia is thought to be more common in males – almost certainly because of their higher incidence of ocular trauma – but this demographic is changing as the causes of sympathetic ophthalmia change (see below). Also, it is thought to be more common in lighter-skinned ethnic groups, but this may be because of better reporting and recognition of the disease.
We reviewed 32 patients with sympathetic ophthalmia seen at the National Eye Institute during a 10-year period. They represented 1.4% of the 2287 patients with uveitis seen by us during that period. There was no sex preponderance (16 males and 16 females) and 23 of the 32 patients had their disease develop after accidental injury. The patients’ ages spanned from 3 years to 80 years. In our group of patients, sympathetic ophthalmia developed within 1 year of injury in 56%, with approximately 31% of them developing disease between 2 weeks and 3 months after the ocular insult, whereas one patient’s disease developed some 46 years later. Hakin and associates reported their experience with 18 patients: 15 of their cases followed trauma, 11 of which occurred after a known penetrating wound, and the other four occurred after a posterior scleral rupture. Three cases occurred after retinal detachment, with patients having on average three surgical interventions. Surgical trauma, particularly vitrectomy, is well recognized as an antecedent event to sympathetic ophthalmia and, as mentioned above, appears to be more common than accidental trauma as the initiating event ( Fig. 23-1 ).
In the classic presentation the inflammatory response seen in the anterior chamber is granulomatous, with mutton-fat keratic precipitates on the corneal endothelium and findings of an acute anterior uveitis. However, the anterior chamber inflammatory response can be relatively mild, with the inflammation taking on a nongranulomatous appearance as well. There is generally a moderate to severe vitritis accompanied by changes in the posterior segment. Of particular note are multiple white-yellow lesions in the periphery, which sometimes become confluent ( Fig. 23-2 ). These represent the clinical appearance of a Dalen–Fuchs nodule, a histologic finding. Swelling of the disc (papillitis) can be prominent, and circumpapillary choroidal lesions can also occur. Retinal involvement is not an uncommon finding in sympathetic ophthalmia, although this involvement, classically thought of as a choroidal disease, has been amply corroborated by histologic findings (see following discussion). Although rare, subretinal neovascularization has been reported to occur in sympathetic ophthalmia, as it has been in the Vogt–Koyanagi–Harada syndrome (VKH) (see Chapter 24 ) (see Fig. 24-2 ).
A small number of patients with sympathetic ophthalmia report or demonstrate extraocular findings not dissimilar from those seen in VKH (see Chapter 24 ), such as cells in the cerebrospinal fluid, hearing disturbances such as high-frequency deafness, and hair and skin changes. These extraocular findings are rare in sympathetic ophthalmia compared to VKH, in which they are fairly common symptoms that are important indicators in making the diagnosis. They are also important in the conceptualization of autoimmune disease, because sympathetic ophthalmia is the only disease for which we know when a specific event leading to autosensitization has occurred. The development of extraocular signs must be secondary to ocular immunization.
It should be noted that the clinical appearance of sympathetic ophthalmia comprises a spectrum that ranges from the very mild to the severe. Lewis and coworkers reported a patient with sympathetic ophthalmia after trauma and vitrectomy that was confined to the posterior segment in such a way that it resembled focal lesions of acute posterior multifocal placoid pigment epitheliopathy (APMPPE) (see Chapter 29 ). A mild anterior uveitis has also been reported, but this has not been our experience. Carrying the concept to the extreme, some have proposed the notion of a unilateral sympathetic ophthalmia. It is quite possible that the sympathizing response may be so mild as to disappear with minimal or no sequelae, and the spectrum of the histologic response seen would support this concept. However, if no inflammatory disease is noted, or if none is seen on a histologic specimen, it seems inappropriate to use the term sympathetic ophthalmia to describe the disease under observation. Marak describes a phenomenon he has called sympathetic irritation, which occurs in the fellow eye after injury to or disease in one eye. About one-third of these patients may have cells and flare in the anterior chamber, with 67% of patients complaining of photophobia and a smaller number having perilimbal or conjunctival injection or both. The distinction between this entity and ‘mild’ sympathetic ophthalmia seems to us quite unclear. There has been a suggested variant of sympathetic ophthalmia with diffuse subretinal fibrotic changes. In one report of a patient who developed the disease after multiple operations for the repair of a retinal detachment, there was fast detererioration of vision with the fibrotic changes becoming more extensive ( Fig. 23-3 ). This eye had many CD20+ cells (B cells) along with CD3+ (T cells) and CD68+ cells (macrophages). In another report of a multifocal granuloma and fibrosis, PCR was negative for herpes viruses, and types III, IV, V, and VI collagens were noted in the fibrosis; antibodies to photoreceptors were also seen.
Several ocular entities have been associated with sympathetic ophthalmia. Phacoanaphylaxis has been observed, and Blodi noted its occurrence in 23% of eyes with sympathetic ophthalmia. He believed that the presence of the two entities in the same eye reflected the fact that these patients have a predisposition for autoimmune phenomena. In their excellent review of 105 cases, Lubin and coworkers found an incidence of 46%. It should be noted that both entities can be found in the sympathizing (noninjured) eye as well, and the phacoanaphylaxis can be a confounding element in attempting to make the diagnosis of sympathetic ophthalmia (see later discussion). It had been suggested in the past that infection would ‘protect’ from sympathetic ophthalmia. However, this disorder has been reported after a central corneal perforation secondary to an Aspergillosis fumigatus keratitis.
Malignant melanoma has also been associated with sympathetic ophthalmia. In 400 cases of sympathetic ophthalmia evaluated by Easom and Zimmerman at the Armed Forces Institute of Pathology, seven were found to involve a malignant melanoma. Fries and colleagues reported sympathetic ophthalmia after helium ion irradiation of a choroidal melanoma, whereas Margo and Pautler reported the development of a granulomatous bilateral uveitis after therapy with proton-beam irradiation. In the latter case the authors stressed the fact that the sutures securing the four tantalum rings to the sclera did not penetrate the uvea.
Sympathetic ophthalmia has been reported to occur after cyclocryotherapy. Harrison discussed the development of sympathetic ophthalmia in a patient who underwent cyclocryotherapy for neovascular glaucoma. No penetration of the sclera was noted in that patient. However, in one case of sympathetic ophthalmia after cyclocryotherapy it was felt that the procedure resulted in microperforations with inflammatory cells as well as uveal pigment found in the subconjunctival space. Edward and colleagues reported a case after noncontact neodymium yttrium–aluminum–garnet (Nd:YAG) laser cyclotherapy. Lam and coworkers reported three others, and the problem was discussed by Tessler and colleagues. There may still be the rare case of no obvious clinical penetration into the globe, such as the case of penetrating trauma resulting in hyphema. Of course it is possible that a micropentatrating wound did occur.
Cataract surgery has been mentioned rarely as a cause of this entity. In one report the disease occurred 3 months after an intracapsular cataract extraction and anterior chamber lens placement in which the immediate postoperative problems of flat chamber, pupillary block, and wound dehiscence occurred, with one of the haptics protruding into the subconjunctival space. Vitrectomy and removal of the lens resulted in control of the disease and improvement of vision. A similar case was reported by El-Asrar and Al-Obeidan. We saw a patient who had a history of retinal detachment repair but whose sympathetic ophthalmia was diagnosed weeks after having received laser therapy for a retinal tear in the other eye, as did El-Asrar et al. reporting a case after retinal detachment surgery and vitrectomy.
There is an ongoing debate as to whether an evisceation or enucleation should be performed. Many have claimed that evisceration carries a risk of sympathetic ophthalmia. Du Toit et al. reviewed 1392 patients who had penetrating eye trauma at the Groote Schuur Hospital in South Africa: 491 (35.5%) had a primary evisceration and no cases of sympathetic ophthalmia were seen in this group. However, we do see sporadic cases develop after this procedure, as with the patient from the Iraq conflict who had an evisceration.
The sequelae of the inflammation noted in sympathetic ophthalmia are quite variable, depending on the severity of the ocular inflammation and whether therapy has been instituted (see later discussion of therapy). Secondary glaucoma and cataract can result from the inflammatory process, as well as retinal and optic atrophy in association with retinal detachment and/or subretinal alterations such as fibrosis and underlying choroidal atrophy. Disc neovascularization has been reported, which initially regressed with steroid therapy but was finally stabilized only with the addition of methotrexate.
Tests and immunologic characteristics
Clinical testing of patients will help to put into perspective the amount and severity of sympathetic ophthalmia but will not help make the diagnosis. Fluorescein angiography is useful in evaluating the degree of posterior segment disease. Spitznas noted that the fluorescein angiogram obtained during the acute stage of sympathetic ophthalmia reflects an exudative process, with multiple subretinal enlarging hyperfluorescent spots and pooling of dye ( Fig. 23-4 ). This observation is further emphasized by OCT. Abu Azar et al. reported the presence of multiple exudative retinal detachments and areas of hyperreflectivity at the level of the RPE ( Fig. 23-5 ). Although Spitznas thought that the retinal vasculature was not involved, we have seen evidence of late staining of the retinal vessels. During the cicatricial phase there is a coarsening of the pigment pattern in the fundus. The areas that are believed to correspond to Dalen–Fuchs nodules will, in the later stages of the disease, become atrophic and appear on the angiogram as window defects.
Sympathetic ophthalmia has been studied using indocyanine green angiography. Two patterns of fluorescence were observed. The first was a pattern of hypofluorescence in the intermediate phase of the angiogram, followed by a fading, and the second was a hypofluorescent pattern that persisted throughout the course of the study. The first was interpreted as showing active lesions, whereas the persistent pattern was that of a cicatricial or atrophic lesion. Saatci and colleagues used ICG to follow the therapeutic response of a patient who developed sympathetic ophthalmia after a perforated anterior staphyloma. During active disease there were multiple hyperflourescent dots which coalesced and then with adequate therapy there was a minimal residual noted ( Fig. 23-6 ).
As might be expected, findings on both electroretinography and electrooculography can be affected by the disease. Dreyer and associates observed changes in the photopic cone b-wave amplitudes and scotopic rod b-wave amplitudes during acute disease, with improvement in these functions after successful therapy.
The immunogenetics of the disease have been studied and appear to be identical to those of the Vogt–Koyanagi–Harada syndrome. Davis and coworkers reported HLA typing results in a small series of patients with sympathetic ophthalmia that appeared to be the same as that of the larger VKH group. Haplotypes HLA-DR4, DRw53, and DQw3 were found more often in both groups than in control subjects. Further work using high-resolution DNA-based HLA typing has been done. In a study reported by Kilmartin and coworkers the highest relative risk was found for the haplotype HLA-DRB1*0404-DQA1*0301. Patients with that haplotype were more prone to develop sympathetic ophthalmia earlier and needed more immunosuppressive therapy to control the disease. It is significant that this association appears to be the same in both Japanese patients and those of European origin. Further work suggests that there was a significant association between the IL-10–1082 SNP and disease recurrence, whereas the GCC IL-10 promoter haplotype (IL-10-1082G, -819C, and -592C) was seen to be protective against disease recurrence. Of interest, yet another single nucleotide polymorphism (SNP) of IL-10 (-1082A) was found to be associated with unremitting sympathetic ophthalmia.
The histologic features of sympathetic ophthalmia, which are rather distinctive, were first described by Fuchs in 1905. In the full-blown picture the choroid is markedly thickened by a sea of lymphocytes, with a granulomatous nonnecrotizing reaction as the predominant feature ( Fig. 23-7 ). In general, the choriocapillaris tends to be spared, although this is not always the case. The infiltrate in the choroid can also have large numbers of eosinophils. Although plasma cell infiltration into the choroid was not thought to be a part of the disease, these cells were noted in 85.7% of the cases reviewed by Lubin and coworkers. Marak and colleagues reported the presence of more choroidal granulomas and eosinophils in African-American than in white patients, but no differences have been noted between Chinese patients and those from the United States. As already noted, retinal changes are not rare. Croxatto and coworkers described retinal detachment and retinal perivasculitis in 50% of the 100 eyes with sympathetic ophthalmia they examined, whereas 18% had mild inflammatory infiltrates into the retina.
A typical but by no means pathognomonic finding in sympathetic ophthalmia is Dalen–Fuchs nodules ( Fig. 23-8 ), which occur in about one-third of eyes with sympathetic ophthalmia. , On fluorescein angiography the lesion may block early (i.e., become hypofluorescent) and then stain late (i.e., hyperfluorescence). As the lesions mature, with ultimate disruption of the retinal pigment epithelium (RPE) and probably a decrease in the inflammatory process, these lesions may simply appear as window defects. These lesions were thought by Fuchs to represent RPE cells that had migrated and transformed. Although early Dalen–Fuchs nodules may represent such a phenomenon, Jakobiec and coworkers and Chan and colleagues , used monoclonal antibodies and structural analysis to demonstrate that Dalen–Fuchs nodules were made up of a mixture of Ia+, OKM1+ cells (presumably histiocytes), and depigmented RPE cells, which were Ia − and OKM1 − . These authors’ conclusions, which were similar for the epithelioid cells seen in the choroid itself, were corroborated by Rao and coworkers. The subsets of infiltrating lymphocytes into the choroid have been identified with the use of immunohistochemical techniques. Jakobiec and colleagues noted that the predominant T cell was of the CD8+ subset (suppressor/cytotoxic cell) in the choroid of an eye removed 1.5 years after the initial surgical trauma. Chan and coworkers evaluated an eye enucleated for the disease only several months after the initial nonsurgical trauma and found that the predominant T cell was the CD4 (Leu-3a) subset, the helper/inducer subset. The sympathizing eye of the same patients was eventually studied, and in this eye the predominant T-cell subset had changed to the suppressor/cytotoxic (CD8+) subset, as had been seen in the study of Jakobiec and coworkers. The interpretation of these observations is that the human eye’s inflammatory response in this disease parallels that seen in the experimentally induced uveitis model in lower mammals (see Chapter 1 ), that is, there is a dynamic, changing cellular immune response initiated by the helper T cell. As the disease progresses the body attempts to downregulate the immune response, and this is done by an influx of large numbers of the suppressor subset of T cells. Others have emphasized the presence of B cells in the infiltrate, found in more chronic conditions.
Chemokines that have been found in sympathetic ophthalmia include matrix metalloproteinase-9, stromal cell-derived factor-1, and monocytic chemotactic protein-1. None of these is surprising, all being found in the context of an inflammatory response.
Preservation of the choriocapillaris
One of the classic findings in both sympathetic ophthalmia and the Vogt–Koyanagi–Harada syndrome is the preservation of the choriocapillaris. This is seen even with the choroid filled with inflammatory cells. It has been suggested that the RPE can play both a pro- and an antiinflammatory role in the eye. Although RPE cells in culture may generate proinflammatory cytokines such as interleukin (IL)-1 and IL-6, they may suppress the inflammatory response by releasing factors such as retinal pigment epithelial protective protein, which is believed to suppress neutrophil superoxide generation. ,
The foregoing scenario, however, may not explain all cases. Several authors have stressed the fact that the classic findings just outlined represent one point in a spectrum of disease recognized clinically as sympathetic ophthalmia. An immunohistochemical study was completed in 29 patients with sympathetic ophthalmia. In 20 eyes, T cells predominated in the choroidal infiltration, whereas in four eyes – all from male patients – B cells were predominantly found. The presence of B cells was correlated to long-term disease (more than 9 months) and phthisical changes. It has classically been taught that the presence of B cells in the choroid of an eye with sympathetic ophthalmia is extremely rare. The authors of this study suggested that these findings may reflect a secondary pathologic process. They further speculated on these findings from a therapeutic point of view. It could be argued that in light of these observations, cytotoxic agents rather than more T-cell-specific drugs (e.g., ciclosporin, tacrolimus (FK506), or perhaps daclizumab) may be a more logical choice for patients whose disease does not respond to steroid therapy.
There is no established and accepted laboratory test for sympathetic ophthalmia. Sen and coworkers measured the level of serum β 2 -microglobulin levels in the blood of patients with sympathetic ophthalmia. They found levels to be high in these patients compared to those with traumatic uveitis and in control subjects. The levels appeared to be higher in the initial stages and reflected the severity of the disease. These results need to be corroborated. Immunohistochemistry of eight eyes with sympathetic ophthalmia showed an increased expression of TNF-α in the photoreceptor layer, whereas iNOS was found in the mitochondria, suggesting that photoreceptor oxidative stress occurs in the absence of a leukocyte infiltrate and that this may lead to vison loss. A recent report by Chan’s group has shown the presence of IL-17 in the macrophages and granuloma of sympathetic ophthalmia eyes. This is in agreement with the concept that the IL-17 is a major player in propagating chronic inflammatory disease.
Sympathetic ophthalmia and vogt–koyanagi–harada syndrome
The question arises whether sympathetic ophthalmia can be distinguished histologically from VKH because it may be the same from an immunogenetic perspective. It would appear that the two entities represent the spectrum of very similar inflammatory responses. It is thought that ‘classic’ sympathetic ophthalmia neither involves the choriocapillaris nor has chorioretinal scarring, both of which are common in VKH. Although these changes can be seen in sympathetic ophthalmia, they probably occur in fewer patients. In addition, the lack of chorioretinal scarring and obliteration of the choriocapillaris in eyes with sympathetic ophthalmia (a phenomenon seen late) may reflect the fact that eyes with sympathetic ophthalmia are usually removed early in the course of the inflammatory disease. In contrast, the eyes with VKH that have been studied were removed late in the course of the disease. Is the inflammatory response seen in sympathetic ophthalmia unique to this entity? Chan and coinvestigators evaluated the cellular composition of granulomas of sympathetic ophthalmia and sarcoidosis. Most of the cells within the lesions were found to be identical and were mainly cells from bone-derived monocytes ( Table 23-2 ). It has been suggested that Dalen–Fuchs nodules in sarcoidosis are larger.