Patterns of uveitis as a guide in making rheumatologic and immunologic diagnoses

358 ARTHRITIS & RHEUMATISM Vol. 40, No. 2, February 1997, pp 358-370 0 1997, American College of Rheurnatology

PATTERNS OF UVEITIS AS A GUIDE IN MAKING RHEUMATOLOGIC AND IMMUNOLOGIC DIAGNOSES

ANTONIO BANARES, JUAN A. JOVER, BENJAMIN FERN~DEZ-GUTIERREZ,

CESAR HERNANDEZ-GARCIA JOSE M. BENITEZ DEL CASTILLO, JAVIER GARCIA, EMILIO VARGAS, and

Objective. To describe the patterns of clinical presentation in a series of 407 patients with uveitis and to establish the relationship between these patterns and the final diagnosis.

Methods. Patients were referred to the Uveitis Clinic of a tertiary hospital from January 1992 to January 1996. All patients received a complete ophthalmologic examination, and a general clinical history was obtained. The current International Uveitis Study Group classification system was used for anatomic classification. To establish the final diagnosis of the most common entities causing uveitis, current diagnostic criteria were used. A discriminant analysis, with diagnostic grouping as the outcome variable and the clinical presentation features as discriminating variables, was performed.

Results. With our classification system, 66.5% of the cases could be correctly classified according to the clinical pattern and morphologic findings. By diagnostic groups, discriminant analysis showed that 75% of patients with Behget’s disease, 77.1% of those with spondylarthropathy (including inflammatory bowel disease), 33.3% of those with sarcoidosis, 97.9% of those with toxoplasmosis, 85.7% of those with Vogt-Koyanagi- Harada syndrome, 100% of those with herpes, and 50.4% of those with idiopathic uveitis were correctly

Presented in part at the 59th National Meeting of the American College of Rheumatology, San Francisco, California, Octo- ber 1995.

Supported by a grant from INSALUD (FIS 9710789). Antonio Bafiares, MD, PhD, Juan A. Jover, MD, PhD,

Benjamin FernindeL-GutiCrrez, MD, PhD, Jose M. Benitez del Castillo, MD, PhD, Javier Garcia, MD, PhD, Emilio Vargas, MD, PhD, Char Herniindez-Garcia, MD, PhD: Hospital Universitario San Carlos, Madrid, Spain.

Address reprint requests to Antonio Bafiares, MD, PhD, Servicio de Reumatologia, Ho3pital Universitario San Carlos, CiMartin Lagos SN, 28040 Madrid, Spain.

Submitted for publication November 21, 1995; accepted in revised form September 3, 1996.

classified. In the miscellaneous group, which included disease entities with fewer than 5 cases, 42.9% were correctly classified.

Conclusion. Rheumatologic evaluation of the patient with uveitis can be more cost-effective if the referring ophthalmologist follows the classification system described herein, allowing a tailored approach in which only specific and necessary diagnostic tests are used.

Uveitis, a general term used to define inflammation in the uveal tract, is a prominent manifestation of several diseases. An underlying systemic disease, often one of autoimmune origin, can be identified in up to 40% of patients with uveitis (1). Many rheumatic and infectious diseases have been associated with uveitis, and thus, internists and rheumatologists are frequently con- sulted in the diagnosis of the patient with uveitis.

Uveitis can be divided into different clinical patterns based on several parameters, such as anatomic location, clinical course, and laterality, and can be characterized by a few morphologic features. Different patterns of uveitis have been described for various underlying diseases, and different disease entities must be ruled out for each pattern of uveitis.

In the present study, we sought to describe the clinical patterns of presentation in a large series of patients with uveitis and to establish the relationship between these patterns and the final diagnosis. Based on the “tailored approach” currently proposed by most groups (2), we discuss the approach to diagnosis in each presentation pattern.

PATIENTS AND METHODS

Patients. All patients seen in our Uveitis Clinic between January 1992 and January 1996 who had a clinical diagnosis of uveitis were eligible for the study. The Uveitis

UVEITIS AS A DIAGNOSTIC MARKER 359

Clinic at our hospital is composed of a multidisciplinary team of ophthalmologists and rheumatologists. Most patients were referred by other ophthalmologists in the community or by other ophthalmology or internal medicine clinics at our hospital. The patients had been referred for diagnostic evaluation and treatment recommendations. Patients with human immu- nodeficiency virus-related disease (3) were not included.

All patients underwent a complete ophthalmologic examination that included visual acuity, slit lamp examination, funduscopy with the indirect ophthalmoscope and the +90- diopter lens, and evaluation of intraocular pressure. Other ophthalmologic tests, such as fluorescein angiography (FAG), were performed only in selected cases.

Classification of uveitis. The current International Uveitis Study Group classification system (4) was used for anatomic classification. Thus, uveitis was categorized as 1) anterior (iritis or iridocyclitis), 2) posterior when involving the choroid and, by extension, the retina (choroiditis o r chorioretinitis), 3 ) intermediate when the inflammation is confined to the vitreum, peripheral retina, and pars plana of the ciliary body, and 4) panuveitis when 2 or more of these segments were affected. Anterior uveitis (AU), with cells in the vitreum, or posterior uveitis, with a few cells in the anterior chamber (spillover), were not included in the panuveitis group. Poste- rior pole involvement was further divided into chorioretinitis, retinal vasculitis, isolated vitritis, or exudative retinal detachment, according to the anatomic and morphologic characterization on funduscopy and FAG. The predominant morphology of the posterior pole involvement was used to further classify the cases into the posterior uveitis or panuveitis groups. When two or more forms coexisted, the nonpredominant form was classified as concomitant posterior pole involvement. Vitritis accompanying some cases of AU was also included as concomitant posterior pole involvement.

The course of the uveitis was classified as acute (duration <3 months) or chronic (duration >3 months). Acute uveitis was considered to be recurrent when a new flare appeared following complete resolution of the previous flare. According to the laterality of each flare, the uveitis was classified as unilateral o r bilateral. Uveitis was considered unilateral when the disease eventually recurred in the opposite eye but both eyes were not affected simultaneously.

Further classification was established according to other morphologic features. The presence of keratitic precipitates (KP; small aggregates of inflammatory cells visible on slip lamp examination that accumulate on the endothelial surface of the cornea and on the trabecular meshwork) was included. KP were classified as granulomatous or nongranulomatous according to both size and the greasy appearance of the granulomatous KP (so-called “mouton fat”).

Finally, we also considered the presence of other ocular parameters complicating the spectrum of the uveitis, such as episcleritis, dry eye, conjunctivitis, hypopyon, elevated intraocular pressure (>21 mm Hg, measured by Goldmann tonometry), synechiae, cystoid macular edema (CME; always confirmed by FAG), band keratopathy, cataract, papillitis, keratouveitis, and snow banks in the pars plana.

Diagnostic criteria. To establish the final diagnosis of the diseases causing uveitis, the following diagnostic criteria were used. The European Spondylarthropathy Study Group criteria were used to classify a case as uveitis (SpA) (5).

Ankylosing spondylitis and Reiter’s syndrome were defined by the modified New York and the Calin criteria, respectively (6,7). Behqet’s disease was diagnosed according to the criteria of the International Study Group for Behqet’s Disease (8). For the diagnosis of juvenile rheumatoid arthritis (JRA), the 1977 American Rheumatism Association criteria were used (9). Psoriatic uveitis was diagnosed when the patient had very prominent and typical dermatologic lesions that affected the scalp and/or eyebrows. Evidence of a noncaseating granuloma was required for the diagnosis of sarcoidosis, except when uveitis presented with the classic Loffgren’s syndrome, with typical bilateral hilar or mediastinal adenopathy. In the latter case, no histologic samples were obtained (10). Patients with Sjogren’s syndrome had dry eyes (confirmed by Schirmer’s test) and typical lymphocytic infiltrates on biopsy of a minor salivary gland (11). Inflammatory bowel disease (IBD) wdS diagnosed on the basis of clinical, endoscopic, and histologic criteria. Vogt-Koyanagi-Harada’s syndrome was diagnosed according to the criteria recommended by the American Uveitis Society (12).

Ocular toxoplasmosis was diagnosed based on the presence of a typical ocular picture of focal chorioretinitis (with or without old choroid scars) and positive IgG anti- Toxoplasma on serology (13). Syphilis was diagnosed when there was a positive rapid plasma reagin test result confirmed by fluorescent treponemal antibody absorption and/or specific IgM antibody enzyme-linked immunosorbent assay (ELISA) (14). A positive culture for mycobacteria in any fluid or tissue sample was necessary to establish a diagnosis of tuberculosis. Herpetic uveitis was diagnosed based on the presence of corneal leukomas and/or sectorial atrophy of the iris and/or glaucoma with a history of recurrent vesicles in the mucosa and positive IgG serologic findings. We classified as nonspecific viral uveitis those cases of self-limited AU which presented with flu-like symptoms. All other infectious causes of uveitis were diagnosed on the basis of positive serologic studies of serum or aqueous humor in the presence of a suggestive ocular presentation.

Some cases of intermediate uveitis were classified as pars planitis when there was typical inflammation in the periphery of the retina, with the classic “snow banks,” peripheral retinal vasculitis, and CME, in patients with no evidence of systemic disease. Vitritis without typical precipitates in the pars plana was classified as idiopathic intermediate uveitis without snow banks. Masquerade syndrome (15) was diagnosed when an underlying malignancy or other condition, such as ischemia or congenital malformations, that mimicked true uveitis was identified. Those cases of uveitis that could not be categorized into a well-defined etiologic process were classified as idiopathic. This category does not include specific ophthalmologic forms of uveitis that could also have an unknown origin, such as the ophthalmologic A U syndromes, pars planitis, and well-defined chorioretinitis in which diagnosis is based exclusively on findings of exploratory and FAG studies.

Statistical analysis. The data were stored in a personal computer database, and the statistical analysis was performed with the SPSS PC version 5.0 program. We performed a discriminant analysis with the Wilks’ method. using 32 clinical and demographic features as discriminating (independent) variables and 9 diagnostic groups as outcome (dependent) variables. We used Wilks’ lambda (U-statistic) and the univar-

360 BANARES ET AL

Table 1. Patterns of presentation and final diagnosis in a series of 407 patients with uveitis

Acute nonrecurrent unilateral AU, n = 54 (13.3%)

Acute bilateral AU, n = 11 (2.7%)

Chronic AU, n = 43 (10.6%)

Posterior uveitis Unilateral chorioretinitis, n = 27 (6.6%)

Pattern, Final no. (%) of patients* diagnosis

Anterior uveitis (AU) Acute recurrent unilateral AU, n = 139 (34.2%) Spondylarthropathy

Idiopathic non-HLkB27-related AU Idiopathic HLA-B27-related AU Ophthalmologic AU syndromes Inflammatory bowel disease Herpes Nonspecific viral uveitis Syphilis Psoriasis Idiopathic non-HLA-B27-related AU Spondylarthropathy Idiopathic HLA-B27-related AU Ophthalmologic AU syndromes Herpes Nonspecific viral uveitis Inflammatory bowel disease Systemic lupus erythematous Sarcoidosis Idiopathic non-HLA-B27-related AU Psoriasis Spondylarthropathy Ophthalmologic AU syndromes Tubulointerstitial nephritis uveitis syndrome Idiopathic non-HLA-B27-related AU Ophthalmologic AU syndromes Sjogren’s syndrome Juvenile rheumatoid arthritis uveitis Sarcoidosis Exophthalmos Spondylarthropathy Herpes

Bilateral chorioretinitis, n = 13 (3.2%)

Retinal vasculitis, n = 14 (3.4%)

Intermediate uveitis, n = 27 (6.6%)

Panuveitis Chorioretinitis panuveitis, n = 31 (7.6%)

Vitritis panuveitis, n = 21 (5.2%)

Toxoplasmosis Ophthalmologically well-defined chorioretinitis Sarcoidosis Ophthalmologically well-defined chorioretinitis Toxoplasmosis Systemic lupus erythematous Masquerade syndromes Idiopathic retinal vasculitis Behset’s disease Masquerade syndromes Tuberculosis Sarcoidosis Pars planitis Idiopathic intermediate uveitis without snow

banks Spondylarthropathy Sarcoidosis Multiple sclerosis

Toxoplasmosis Idiopathic panuveitis Ophthalmologically well-defined chorioretinitis Herpes Vogt-Koyanagi-Harada syndrome Idiopathic panuveitis Spondylarthropathy Ophthalmologic AU syndromes Pars planitis Masquerade syndromes Sjogren’s syndrome

No. (%)t %$

67 (48.2) 34 (24.5) 16 (11.5)

6 (4.3) 6 (4.3)

l(0.7) l(0.7)

16 (29.6)

9 (16.6) 9 (16.6)

7 (5)

l(0.7)

12 (22.2)

3 (5.5) 2 (3.7) l(1.8)

l(1.8) 5 (45.4)

l(1.8)

3 (27.3) l(9.1) l(9.1) l(9.1)

18 (41.2) 14 (32.5) 2 (4.6) 2 (4.6) 2 (4.6) 2 (4.6) 2 (4.6) 1 (2.3)

25 (92.6) l(3.7) l(3.7)

l(7.7) l(7.7)

8 (60.5) 3 (23.1)

8 (57.1) 3 (21.4) l(7.1) l(7.1) l(7.1)

17 (63) 7 (25.9)

l(3.7) l(3.7) l(3.7)

19 (61.3) 5 (16.1) 5 (16.1) l(3.2) l(3.2) 9 (42.9) 3 (14.3) 2 (9.5) l(4.8) l(4.8) l(4.8)

77.9 46.6 64 21.2 60 54.5 33.3 25 25 21.9 13.9 36 27.3 27.3 66.6 10 50 11.1 6.8

75 1.2 3

100 24.7 42.4 66.7

22.2

2.3 9.1

53.2 7.1

11.1 57.1

6.4 50 33.3

100 25 33.3 50 11.1 89.5

100

100

100

1.2 11.1

100

40.4 27.8 35.7 9.1

14.3 50 3.5 6.1 5.3

33.3 33.3


Table 1. (Cont’d)

Pattern, no. (96) of patients*

Final diagnosis

NO. (”/.It

Vitritis panuveitis, n = 21 (5.2%) (Cont’d)

Retinal vasculitis panuveitis, n = 19 (4.7%)

BehGet‘s disease Sarcoidosis Inflammatory bowel disease Tuberculosis BehGet’s disease Idiopathic panuveitis Syphilis Inflammatory bowel disease Sarcoidosis Pars planitis Vogt-Koyanagi-Harada syndrome Behcet’s disease

Exudative retinal detachment panuveitis, n = 8 (1.9%)

l(4.8) 1 (4.8) l(4.8) l(4.8) 7 (36.8)

3 (15.8) 2 (10.5) 2 (10.5)

4 (21.1)

8.3 11.1 10 50 58.3 22.2 75 20 22.2

l(5.3) 5.3 6 (75) 85.7

8.3 l(12.5) Toxocariasis I (1 2.5) 100

* Percentage of the pattern in the entire series of patients. t Percentage of the specific disease within the presentation pattern. $ Percentage of the presentation pattern within the specific disease.

iate F-ratio with 7 and 333 degrees of freedom to obtain the significance of each variable. A stepwise analysis was then performed. The variables were selected by the minimum value of Wilks’ lambda. The minimum tolerance level was 0.00100. The classification function coefficients (Fisher’s linear discriminant functions) defined the correlation between each variable and the diagnostic group. The classification system predicts the theoretical group for each case (highest probability group and second highest probability group). These findings were compared with the actual group in which each case was included.

RESULTS

From January 1992 to January 1996, 407 new patients with uveitis were evaluated at our Uveitis Clinic. The mean (“SD) age at onset was 42 ? 16 years (range 6-85), with a female preponderance (188 males, 219 females). All patients under age 18 were classified as children (n = 28). All patients over age 65 were classified as elderly (n = 40).

Uveitis presentation patterns. Table 1 shows the 12 patterns of presentation under which all uveitis cases were classified, their incidence, the incidence of each specific disease within each presentation pattern, as well as the percentage of each presentation pattern within every specific disease. Overall, 60.8% of all cases of uveitis were anterior uveitis, 13.2% were posterior uveitis, 6.6% were intermediate uveitis, and 19.4% were panuveitis.

Spondylarthropathy was the most frequent diagnosis found in the acute recurrent unilateral AU pattern (48 ankylosing spondylitis, 7 Reiter’s syndrome, 6 IBD, and 12 unclassified SPA). Sixteen cases were classified as

idiopathic HLA-B27-related AU. Within acute nonrecurrent unilateral AU, acute bilateral AU, and chronic AU, the most frequent diagnosis was idiopathic AU. Ophthalmologic AU syndromes, including Fuchs’ het- erochromic cyclitis (1 0 cases), Possner’s glaucomato- cyclitic crisis (9 cases), phacogenic uveitis (8 cases), and lens intraocular-related uveitis (6 cases), were repre- sented in all 4 presentation patterns of AU.

Unilateral chorioretinitis was associated almost exclusively with toxoplasmosis. Bilateral chorioretinitis was the predominant pattern of ophthalmologically well- defined chorioretinitis. These entities appeared not only in this pattern but also in unilateral chorioretinitis and chorioretinitis panuveitis, and included serpiginous choroiditis (3 cases), multifocal choroiditis (3 cases), acute posterior multifocal plaquoid pigment epitheliopathy (4 cases), birdshot choroiditis (2 cases), presumed ocular histoplasmosis syndrome (1 case), and inner punctata choroiditis (1 case). Isolated retinal vasculitis mostly included idiopathic uveitis, although Behqet’s disease was also frequent.

Pars planitis was the most frequent diagnosis found in the intermediate uveitis pattern, although 7 cases without snow banks were classified as idiopathic.

Finally, according to the predominant morphology of posterior pole involvement, 4 patterns of panuveitis were identified. The pattern of chorioretinitis panuveitis, again, included toxoplasmosis as the most frequent diagnosis. Idiopathic panuveitis was the most frequent diagnosis in vitritis panuveitis, i.e., vitreum inflammation with a prominent AU. The underlying disease most frequently observed in retinal vasculitis panuveitis was Behqet’s disease. Finally, the pattern of

362 BARARES ET AL

Table 2. Morphologic characteristics and final diagnosis in a series of 407 patient3 with uveitis

Finding, Final no. (%) of patients” diagnosis No. (%)t %$

Concomitant posterior pole involvement Chorioretinitis, n = 7 (1.7%)

Retinal vasculitis, n = 12 (2.9%)

Exudative retinal detachment, n = 5 (1.2%)

Vitritis, n = 76 (18.7%)

Keratitic precipitates Nongranulomatous, n = 67 (16.5%)

Granulomatous, n = 48 (11.8%)

Other morphologic findings Episcleritis, n = 11 (2.7%)

Dry eyes, n = 9 (2.2%)

Cataract, n = 78 (19.2%)

Conjunctivitis, n = 13 (3.2%)

Cystoid macular edema, n = 30 (7.4%)

Hypopyon, n = 8 (2%)

Sarcoidosis Spondylarthropathy Inflammatory bowel disease Toxocariasis Vogt-Koyanagi-Harada syndrome Idiopathic panuveitis Pars planitis Toxoplasmosis Others Idiopathic non-HLA-B27-related anterior uveitis (AU) Inflammatory bowel disease BehFet’s disease Idiopathic retinal vasculitis Toxoplasmosis Idiopathic non-HLA-B27-related AU Ophthalmologic AU syndromes Spondylarthropathy Toxoplasmosis Others

Idiopathic non-HLA-B27-related AU Ophthalmologic AU syndromes Spondylarthropathy Idiopathic €ILA-B27-related AU Others Idiopathic non-HLA-B27-related AU Spondylarthropathy Herpes Others

Spondylarthropathy Exophthalmos Systemic lupus erythematous Idiopathic panuveitis Behset’s diseasc Herpes Inflammatory bowel disease Sarcoidosis Idiopathic non-HLA-B27-related AU Idiopathic non-HLA-B27-related AU Sjogren’s syndrome Toxoplasmosis BehGet’s disease Exophthalmos Ophthalmologic AU Idiopathic non-HLA-B27-related AU Spondylarthropathy Pars planitis Others Spondylarthropathy Idiopathic I-ILA-B27-related AU Nonspecific viral uveitis Others Pars planitis Behset’s disease Spondylarthropathy Sarcoidosis Others Spondylarthropathy Idiopathic non-HLA-B277related AU Idiopathic HLA-B27-related AU

2 (28.6) 1 (14.3) 1 (14.3) 1 (14.3) l(14.3) l(14.3) 7 (58.3)

2 (16.6) 3 (25)

1(20) 1(20) 1(20) 1(20) 1 (20)

16 (21) 8 (10.5) 8 (10.5)

10 (13.1) 34 (44.9)

12 (18) 6 (9) 6 (9)

18 (27)

19 (37) 13 (27.1) 5 (10.4) 5 (10.4)

25 (52.1)

2 (18.2) 2 (18.2) 1 (9.1) l(9.1) l(9.1) l(9.1)

l(9.1) l(9.1)

l(9.1)

3 (33.3) 3 (33.3) l ( l l . 1 ) 1 (11.1) L(ll . l)

7 (9) 7 (9)

5 (38.5)

18 (23.1) 15 (19.2)

31 (39.7)

2 (15.4) 2 (15.4) 4 (30.8) 8 (26.7) 5 (16.7) 3 (10) 3 (10)

1 L (36.6) 4 (50) 2 (25) l(12.5)

22.2 1.2

10 100 14.3 5.5

36.8 6.4

1.4 10 8.3

12.5 2.1

21.9 24.2 9.3

21

24.6 36.4 7

24

17.8 5.8

45.4

2.3 100 50 5.6 8.3 9

10 11.1 1.4 4.1

2.1 8.3

50 54.5 20.5 8.1

36.8

5.8 8

66.7

42.1 45.7 3.5

33.3

4.6 2.7 4

100


Table 2. (Cont’d)

Finding, no. (%) of patients*

Final diagnosis No. (%)t %10$

Papillitis, n = 22 (5.4%)

Synechiae, n = 90 (22.1%)

Band keratopathy, n = 4 (1%)

Glaucoma, n = 46 (1 1.3%)

Snow hanks, n = 20 (4.9%)

Keratouveitis, n = 27 (6.6%)

BehGet’s disease Toxoplasmosis BehGet’s disease Syphilis Others Spondylarthropathy Idiopathic non-HLA-B27-related AU Others Juvenile rheumatoid arthritis uveitis Spondylarthropathy Idiopathic panuveitis Idiopathic non-HLA-B27-related AU Ophthalmologic AU Toxoplasmosis Others Pars planitis Others Herpes Idiopathic non-HLA-B27-related AU Others

l(12.5) 7 (31.8) 3 (13.6) 3 (13.6) 9 (40.7)

30 (33.3) 21 (23.3) 39 (43.4) 2 (50) 1(25) 1(25)

14 (30.4) 12 (26.1) 5 (10.9)

15 (32.6) 17 (85) 3 (15)

11 (40.7) 4 (14.8)

12 (54.5)

8.3 14.9 25 75

34.9 28.8

100 1.2 5.6

19.2 36.4 10.6

89.5

100 5.5

* Percentage of the morphologic finding in the entire series of patients. t Percentage of the specific disease within the morphologic finding. $ Percentage of the morphologic finding within the specific disease. 5 Cases representing <9% of diagnoses within a morphologic finding.

exudative retinal detachment panuveitis predominantly corresponded to the Vogt-Koyanagi-Harada syndrome.

Morphologic characterization. Table 2 details the incidence in the entire patient series of several other morphologic features of uveitis, the incidence of each specific disease within each morphologic feature, and the percentage of each morphologic feature within each specific disease. One hundred patients with or without predominant posterior pole involvement had a less- prominent posterior pole involvement accompanying the principal pattern. Bilateral chorioretinitis was the predominant pattern of ophthalmologically well-defined chorioretinitis. These entities appeared not only in this pattern but also in unilateral chorioretinitis and chorioretinitis panuveitis, and included serpiginous choroiditis (3 cases), multifocal choroiditis (3 cases), acute posterior multifocal plaquoid pigment epitheliopathy (4 cases), birdshot choroiditis (2 cases), presumed ocular histoplasmosis syndrome (1 case), and inner punctata choroiditis (1 case).

In 109 patients there were keratitic precipitates (KP), particularly nongranulomatous KP. Six patients had both granulomatous and nongranulomatous KP (4 idiopathic, 1 herpes, and 1 ophthalmologic AU). We specifically investigated 12 other morphologic findings in

order to have a complete characterization of the ophthalmologic picture.

Discriminant analysis. To validate the presentation patterns and other morphologic and demographic findings as indicators of the final diagnosis, we performed a stepwise discriminant analysis with diagnostic groups as dependent or outcome variables and the clinical presentation features as independent or discriminating variables. We applied the following 9 diagnostic groups: ophthalmologic syndromes (including ophthalmologic AU syndromes, pars planitis, and ophthalmologically well-defined chorioretinitis), idiopathic uveitis, Behqet’s disease, spondylarthropathy, toxoplasmosis, sarcoidosis, Vogt-Koyanagi-Harada syndrome, herpetic uveitis, and miscellaneous, which includes disease entities with fewer than 5 cases: psoriasis (n = 4), tuberculosis (n = 2), exophthalmos (n = 2), Sjogren’s syndrome (n = 3), JRA (n = 2), syphilis (n = 4), masquerade syndromes (n = 3), systemic lupus erythematous (SLE) (n = 2), tubulointerstitial nephritis uveitis (TINU) syndrome (n = 1), toxocariasis (n = 1), multiple sclerosis (n = l), and nonspecific viral uveitis (n = 3).

The ophthalmologic syndromes were not considered for discriminant analysis since they are exclusively diagnosed by ophthalmologic examination and usually

364 BMARES ET AL

are not encompassed in the etiologic differential diagnosis. Thus, 8 diagnostic groups and 341 patients were used in the discriminant analysis.

Table 3 presents the 33 variables analyzed, along with their significance values and the 26 variables that could finally be included in the stepwise analysis according to the minimum Wilks’ lambda. Table 4 shows Fisher’s linear discriminant functions that represent the relationship between each variable and the diagnostic grouping.

Stepwise discriminant analysis was used to classify cases. We then compared our final diagnostic group with the calculated group for which the posterior probability was highest (Table 5). With this calculation, 227 cases (66.56%) were assigned to their correct group; 114 cases were not correctly classified in the group for which the posterior probability was highest. However, of these 114 cases, 86 (75.4%) were correctly classified when assigned to the group for which the posterior probability was the second highest.

The predicted classification was correct in 75% of Behqet’s disease cases and in 77.1% of SPA cases. Most cases of SPA in which discriminant analysis was incorrect were classified as idiopathic. Among them, 10 patients had an acute nonrecurrent unilateral AU (a very nonspecific pattern) and the others included uveitis with posterior pole involvement. In sarcoidosis, only one- third of the cases were correctly classified as the group for which the posterior probability was highest, confirming the mimicking presentation form of the disease. Our system also successfully classified all but 1 case of toxoplasmosis and Vogt-Koyanagi-Harada syndrome, and all cases of herpetic keratouveitis. It is interesting to note that half of the cases that were finally classified as idiopathic were assigned to this group in the discriminant analysis based on morphologic findings.

In the miscellaneous group, many cases were misclassified. However, some conditions were correctly included in the group: all cases of Sjogren’s syndrome, JRA, or masquerade syndromes and 2 of the 4 cases of psoriasis, 1 of the 2 cases of tuberculosis, and 1 of the 2 cases of exophthalmos. The 2 other cases of psoriasis were included in SpA group, and the second case of exophthalmos and tuberculosis were classified as idiopathic. Three of the 4 cases of syphilis that presented as retinal vasculitis panuveitis were attributed to the Beh- Get’s group, and the fourth case, a keratouveitis, was attributed to herpes. The toxocariasis case, which presented as exudative retinal detachment panuveitis, was classified as Vogt-Koyanagi-Harada syndrome; 1 case of multiple sclerosis, which presented as intermediate uve-

Table 3. Variables included and not included in the stepwise discriminant analysis (summary)*

Wilks’ Wilks’ Variable lambda? F t P t lambda$

Included Exudative retinal detachment

Unilateral chorioretinitis Chorioretinitis panuveitis Keratouveitis Retinal vasculitis panuveitis Bilateral chorioretinitis Acute recurrent unilateral

anterior uveitis (AU) Snow banks Concomitant chorioretinitis Retinal vasculitis Dry eye Cystoid macular edema Cataract Children Granulomatous keratitic

Concomitant vitritis Concomitant retinal vasculitis Nongranulomatous keratitic

Synechiae Elderly Conjunctivitis Band keratopathy Glaucoma Sex Episcleritis Papillitis

Not included5 Acute bilateral AU Chronic AU Intermediate uveitis Vitritis panuveitis Acute nonrecurrent

unilateral AU

panuveitis

precipitates

precipitates

“YPoPYon Concomitant exudative

retinal detachment

0.35048 88.16 0.0000 0.35048

0.52424 43.17 0.0000 0.18410 0.74487 16.29 0.0000 0.07357 0.53054 42.10 0.0000 0.03909 0.76161 14.89 0.0000 0.02965 0.94702 2.662 0.0108 0.02264 0.67788 22.61 0.0000 0.01847

0.93148 0.92306 0.92622 0.94193 0.87302 0.93271 0.93893 0.92516

0.94985 0.94543 0.96283

0.92644 0.95703 0.97123 0.97120 0.97061 0.96142 0.96486 0.90917

3.500 3.965 3.790 2.933 6.919 3.432 3.094 3.848

2.512 2.746 1.836

3.777 2.136 1.409 1.411 1.441 1.909 1.733 4.753

0.0012 0.0004 0.0006 0.0054 0.0000 0.0015 0.0036 0.0005

0.0158 0.0088 0.0796

0.0006 0.0396 0.2006 0.2000 0.1881 0.0674 0.1005 0.0000

0.01587 0.01420 0.01292 0.01190 0.01106 0.01034 0.00970 0.00924

0.00891 0.00859 0.00830

0.00802 0.00777 0.00753 0.00729 0.0 0 7 0 9 0.00691 0.00675 0.00659

0.98976 0.4921 0.8402 0.00653 0.92740 3.724 0.0007 0.00652 0.97461 1.239 0.2804 0.00653 0.97888 1.027 0.4121 0.00648 0.95137 2.432 0.0193 0.00656

0.98367 0.7898 0.5963 0.00655 0.99292 0.3392 0.9356 0.00650

*Variables are shown in the order in which they were included according to the stepwise selection rules. Cases classified as ophthalmologic syndromes were not included in discriminant analysis. The table encompasses 341 cases and 8 diagnostic groups. Wilks’ lambda (U-statistic) and univariate F-ratio with 7 and 333 degrees of freedom were used. t At initial analysis. $ Included once in the analysis. 5 Variables did not reach minimum criteria for stepwise variable selection.

itis, was classified as sarcoidosis; one case of SLE (presented as unilateral chorioretinitis) was classified as toxoplasmosis and the other (presented as a nonrecurrent unilateral AU with keratitis) was classified as herpes. The TINU syndrome was classified as idiopathic; nonspecific viral uveitis was misclassified as idiopathic or as herpetic (1 case presented as keratouveitis).


Table 4. Classification function coefficients (Fisher’s linear discriminant function)

Diagnostic group

Behqet’s Vogt-Koyanagi- Variable disease SPA* Sarcoidosis Toxoplasmosis Harada Herpetic Idiopathic Miscellaneoust

Pattern Acute Recurrent Unilateral AU 3.44 5.59 Unilateral chorioretinitis 6.87 2.88 Bilateral chorioretinitis 6.97 2.34 Retinal vasculitis 14.16 2.56 Chorioretinitis panuveitis 5.07 1.72 Retinal vasculitis panuveitis 21.55 2.88 Exudative retinal detachment 14.28 0.38

panuveitis Concomitant posterior pole involvement

Chorioretinitis 1.64 2.62 Retinal vasculitis -0.588 1.16 Vitritis 1.07 1.38

Nongranulomatous 0.44 0.15 Granulomatous 2.95 1.92

Episcleritis 5.14 2.40 Dry eye 6.96 0.74 Synechiae 2.08 2.24 Conjunctivitis 2.58 2.67 Cystoid macular edema 8.11 2.74 Papillitis 1.48 0.46 Band keratopathy 2.71 2.38 Cataract 1.37 -0.28 Glaucoma 0.22 -0.49 Snow banks 11.58 -0.94 Keratouveitis 1.05 -0.4.5

Keratitic precipitates

Other morphologic finding

Demographics Sex 3.27 3.07 Child 1.71 0.76 Elderly 1.32 0.63

2.35 10.70 7.50 7.58 5.68

10.75 -7.83

17.76 -0.31 -0.33

-0.14 4.81

4.86 1.34 2.72 2.47 7.77 1.53 1.16 2.25 1.23

20.50 0.90

2.19

0.40 -0.35

2.75 52.24 35.27 4.19

39.08 3.47

11.39

1.84 -0.74

1.96

1.43 1.18

2.69 2.75 0.65 1.07 2.21 4.42 2.37

- 1.49 2.58

3.58

2.11 3.23

-0.23

-2.17

2.83 16.67 12.81 9.11

20.07 10.50

141.48

- 15.91 - 1.24 -3.33

-1.11 5.04

4.37 1.62

- 2.29 3.54 1.21

-2.04 8.85 5.21

- 1.26 -35.16

4.09

0.23

5.38 -1.99

3.68 10.16 8.89 6.74 7.66 5.46 4.82

2.44 -12.81

0.21

3.05 5.55

- 1.59 -5.61

0.85 -2.18

0.40 -0.62

5.87 -0.72 -1.83

1.57 32.59

3.05 3.00 2.1 1

3.31 4.22 3.32 3.86 2.56 2.94 1.07

1.60 -0.20

1.74

1.47 2.25

1.88 1.91 1.41 1.33 2.00 0.95 0.95 0.55 0.70

-0.74 0.94

2.49 2.21 1.00

2.56 6.78

10.98 7.9s 4.94 8.15 7.77

1.77 -2.75

0.21

1.15 1.70

4.65 6.57 0.30 4.43 1.26 2.35 8.61 1.70 0.95 5.40 6.59

2.26 6.26 3.55

* Spondylarthropathy (SpA) includes ankylosing spondylitis (n = 60), Reiter’s syndrome (n = 7), ulcerative colitis (n = 6), Crohn’s disease (n = 4), and undetermined SpA (n = 19). ‘F Miscellaneous includes disease entities with fewer than 5 cases: psoriasis (n = 4), tuberculosis (n = 2), exophthalmos (n = 2), Sjogren’s syndrome (n = 3), juvenile rheumatoid arthritis (n = 2), syphilis (n = 4), masquerade syndromes (n = 3), systemic lupus erythematosus (n = 2), tubulointerstitial nephritis uveitis syndrome (n = I), toxocariasis (n = I), multiple sclerosis (n = 1), and nonspecific viral uveitis (n = 3).

DISCUSSION

We have described here a series of 407 patients with uveitis who were subdivided into 12 presentation patterns based on the anatomic location, laterality, course, and predominant morphology of the posterior pole involvement, and we have specifically investigated 18 other morphologic findings. To validate our classification scheme as an indicator of the final diagnosis, we performed a stepwise discriminant analysis with which 66.5% of the patients without a specific ophthalmologic syndrome were correctly classified in their diagnostic group. However, we note that this number is considered optimistic because we used the same data to develop and validate our classification scheme.

With our system, 75% of the cases of Behqet’s

disease were correctly classified. Furthermore, 2 of the 3 cases that were initially misclassified were included in Behget’s disease as the group for which the posterior probability was second highest. Behqet’s disease presented mainly as “retinal vasculitis” with or without concomitant AU and vitritis, and was frequently complicated by CME and/or papillitis. The diagnosis of Beh- Get’s disease must be clinical, the presence of oral and genital ulcers, erythema nodosum, and nonerosive arthritis being the most frequent clinical symptoms. In Mediterranean countries, when atypical signs and symptoms are present, typing for HLA-Bw51 (relative risk 9.4) cannot be considered diagnostic, although its presence supports the diagnosis (16-18).

Spondylarthropathies were included in the group

366 BANARES ET AL

Table 5. Stepwise discriminant analysis: classification results

Predicted group (group for which the posterior probability is highest), no. (96)

Behqet’s Actual group (n) disease SPA* Sarcoidosis

Vogt-Koyanagi- Toxoplasmosis Harada Herpetic Idiopathic Miscellaneous? Total

BehGet’s disease (12) 9 (75) “(0)

Sarcoidosis (9) 2(22.2) O(0) Toxoplasmosis (47) 0 (0) 0 (0)

Herpetic (1 1) 0 (0) 0 (0)

Spondylarthropathy (96)* 2 (2.1) 74 (77.1)

Vogt-Koyanagi-Harada (7) 0 (0) 0 (0)

Idiopathic (131) 6 (4.6) 38 (29) Miscellaneous (28)t 3 (10.7) 2 (7.1)

Total (341)

_______

0 (0) 0 (0) l( l1.1)

0 (0)

46 (97.9) 1 (14.3)

5 (3.8) l(3.6)

O(0) 2(16.7) 0 (0) 18 (18.7) 0 (0) 3 (33.3) l(2.1) O(0) 0 (0) 0 (0)

11 (100) O(0) 5 (3.8) 66 (50.4) 4 (14.3) 4 (14.3)

227 (66.56)

* Spondylarthropathy (SpA) includes ankylosing spondylitis (n = 60), Reiter’s syndrome (n = 7), ulcerative colitis (n = 6), Crohn’s disease (n = 4), and undetermined SpA (n = 19). t Miscellaneous includes disease entities with fewer than 5 cases: psoriasis (n = 4), tuberculosis (n = 2), exophthalmos (n = 2), Sjogren’s syndrome (n = 3), juvenile rheumatoid arthritis (n = 2), syphilis (n = 4), masquerade syndromes (n = 3), systemic lupus erythematosus (n = 2), tubulointerstitial nephritis uveitis syndrome (n = l) , toxocariasis (n = I), multiple sclerosis (n = 1), and nonspecific viral uveitis (n = 3).

for which the posterior probability was highest in 77.1% of the cases and in the group for which the posterior probability was the second highest in all the misclassified cases except for 2 cases of IBD (with “retinal vasculitis” misclassified as BehCet’s disease) and 2 cases of ankylosing spondylitis (in patients over age 65 experiencing their first episode of AU). The cases that were initially misclassified included 4 cases of IBD, 8 cases of undetermined SPA, and 4 cases of ankylosing spondylitis with longstanding AU. Ankylosing spondylitis, Reiter’s syndrome, and undifferentiated SPA presented as acute recurrent unilateral AU in 77.9% of the cases (19,20). However, IBD might also present with posterior pole involvement, such as “retinal vasculitis panuveitis” (21,22). There are a few significant associated complica- tions of SpA uveitis: synechiae in up to 34.9%, CME in up to 20% of the IBD cases, and conjunctivitis in 57% (4 of 7 cases) of the Reiter’s syndrome cases. Moreover, SpA was the most frequent cause of hypopyon in our series, although it occurred in only 4 cases. Thus, sacroiliac roentgenography and HLA-B27 typing (23) are indicated in patients with acute recurrent unilateral AU.

IBD, however, must be ruled out when there is a positive history of abdominal symptoms (22). We have previously demonstrated that subclinical chronic intesti- nal inflammation might be a part of the spectrum of AU, especially in HLA-B27 negative patients, as well as in patients with more than 2 episodes of AU per year (24), which suggests that IBD must be specifically ruled out in patients with highly recurrent acute unilateral AU. Search for gram-negative bacteria in the aqueous humor

could help to confirm Reiter’s syndrome in patients with AU, but currently, this approach should be reserved for research protocols (25).

We cannot predict the diagnosis of sarcoidosis with our discriminant system: only 3 of 9 cases were correctly assigned. Sarcoidosis presented in 7 of the 12 patterns we defined. In this series, retinal vasculitis and the presence of a chronic granulomatous AU were the outstanding characteristics of the disease. Our findings confirm the mimicking character of the ocular involvement of this disease. Sarcoidosis should be kept in mind in all undiagnosed cases of uveitis, and all patients with uveitis should undergo a chest roentgenogram as a mandatory screening test (26-28). Normal findings on chest roentgenograms can obviate other costly tests that are of limited specificity and/or uncertain sensitivity. More aggressive procedures, such as conjunctival (29), lacrimal, transconjunctival, or transbronchial (30) biop- sies, for confirming suspected ocular sarcoidosis must be reserved for patients with positive findings on noninva- sive tests in which the diagnosis might modify the therapeutic approach.

Exudative retinal detachment is the hallmark of the Vogt-Koyanagi-Harada syndrome (12,31) and had the most significant Wilks’ lambda value in our discriminant analysis. A granulomatous AU can also be present. Lumbar puncture is the confirmatory test in all patients presenting with panuveitis and exudative retinal detachment, since Vogt-Koyanagi-Harada syndrome is always associated with a lymphocytic meningitis in the acute phase, although a full clinical meningeal syndrome may


or may not be present and the other classic extraocular manifestations are rarely present as the first event.

In our classification system, 97.9% of the patients with toxoplasmosis were correctly assigned to the group for which the posterior probability is highest. The misclassified case, a chorioretinitis panuveitis with concomitant keratitis, which was classified as herpes, was correctly assigned to the second diagnosis group. Toxoplasmosis presented as a very characteristic isolated chorioretinitis that was predominantly unilateral (53.2% of the cases) or accompanied by concomitant AU and vitritis (40.4%). It was frequently complicated by elevated intraocular pressure or papillitis. In the diagnosis of toxoplasmosis uveitis, a positive serologic test result (IgG ELISA) has a very low specificity, but a negative result essentially excludes the diagnosis (13). Serology of aqueous humor can be helpful in doubtful cases or in primary infection (32-34). Herpes simplex and herpes zoster were also correctly classified in all cases in the discriminant analysis. The presence of keratouveitis with granulomatous and/or nongranulomatous KP was very characteristic. The presentation pattern is generally of acute recurrent unilateral AU and less frequently associated with chorioretinitis or vitritis. In the event of keratouveitis or in doubtful cases, to initiate specific antiviral therapy (32), serology of herpetic serum or aqueous humor can be useful.

For discriminant analysis, we assigned all diseases with fewer than 5 cases to a miscellaneous group, which could give rise to nonsignificant associations that might be avoided in larger series that permitted some specific diagnoses to be classified separately. However, despite a limited number of patients, 42.9% of the cases in this group were correctly classified. Furthermore, in the miscellaneous group, in the case of some disease entities, all patients were correctly classified in the discriminant analysis. Three cases of Sjogren’s syndrome- associated uveitis (35) presenting in elderly subjects as chronic AU with occasional vitritis in a dry eye were correctly classified. Two cases of uveitis associated with JRA, both with chronic AU and band keratopathy, were also correctly classified. JRA uveitis presents character- istically in children previously diagnosed as having a joint condition. However, the incidence of uveitis pre- ceding the development of arthritis ranges from 6% to 24%, with a delay of 4-9 years (36-39). Finally, all 3 cases of masquerade syndromes were also correctly classified: 1 with antiphospholipid syndrome presenting as retinal vasculitis, 1 with chorioretinal metastasis from a lung cancer presenting as bilateral chorioretinitis, and 1 with Coats’ syndrome with a vascular malformation,

prominent vitritis, and cells in anterior chamber. Anti- cardiolipin antibodies and/or lupus anticoagulant should be investigated in vasoocclusive disease mimicking retinal vasculitis (40,41). A neoplasm must be ruled out, especially in children and in the elderly, and if there is a poor response to adequate antiinflammatory therapy (15). In these patients, it may be necessary to perform cytology of aqueous or vitreous samples (42) or a chorioretinal biopsy (43).

Conversely, all 4 cases of syphilis were misclassified. The most significant finding for the disease was papillitis, which presented in 3 of the 4 cases. As we have already mentioned when referring to sarcoidosis, the mimicking character of the ocular manifestation of syphilis makes it mandatory that we recommend routine serologic screening in all patients with uveitis (14,22,23,28,44).

Several other diseases were misclassified in the discriminant analysis because of the small number of patients and/or the lack of specificity of their presenting patterns. Three of the 4 patients with psoriasis had an acute bilateral AU with nongranulomatous PK (a very nonspecific pattern), and the other had acute recurrent unilateral AU. Half of them were correctly classified. Two patients with SLE were misclassified in the discriminant analysis as having toxoplasmosis and herpes, respectively. Both patients were diagnosed as having SLE based on clinical findings. A thorough medical history and examination would point to a systemic autoimmune disease as a cause of the uveitis in these patients (45). We routinely search for the presence of specific autoantibodies or cryoglobulins if there are any symptoms of SLE or systemic vasculitis or if there are any abnormalities in the results of the complete blood cell count, blood chemistry profile, urinalysis, or eryth- rocyte sedimentation rate. Because of the high rate of false-positive findings, the presence of autoantibodies should not be investigated if there are no other clinical symptoms (46-48). A few patients with SLE could develop uveitis, but uveitis is rarely the presenting symptom of SLE. We found only one case of TINU syndrome (49), which presented as a bilateral acute AU. This syndrome must be ruled out in patients with azotemia who develop uveitis.

One patient with multiple sclerosis and 1 patient with toxocariasis were incorrectly classified in the discriminant analysis. In multiple sclerosis, uveitis mimicking pars planitis occurs in as many as 20% of cases (50). Thus, in all patients with intermediate uveitis, an exten- sive neurologic history and examination are needed to rule out previously undiagnosed neurologic disease (51).

368 BmARES ET AL

Table 6. Diagnostic approach to the patient with uveitis, by pattern of presentation*

Specific disease entity Specific disease entity to Uveitis pattern to rule out rule out occasionally

Acute recurrent unilateral AU

Acute nonrecurrent unilateral AU

Acute bilateral AU Chronic AU

Unilateral chorioretinitis Bilateral Chorioretinitis

Retinal vasculitis

Intermediate uveitis

Chorioretinitis panuveitis

Vitritis panuveitis

Retinal vasculitis panuveitis

Exudative retinal detachment panuveitis

Spondylarthropathy

Toxoplasmosis

Toxoplasmosis

IBD, if abdominal symptoms present Herpes, if keratouveitis present Spondylarthropathy, if vertebral symptoms present Herpes, if keratouveitis present TINU syndrome, if systemic symptoms present JRA uveitis, in children and/or if band keratopathy present Sjogren’s syndrome, if dry eyes present Sarcoidosis, if granulomatous uveitis or systemic symptoms present

Toxoplasmosis, if typical morphology present Masquerade (neoplasia), in children or elderly, if no response to

Connective tissue diseases, if systemic symptoms are present Masquerade (antiphospholipid syndrome), in retinal venooclusive

Behset’s disease, if mucocutaneous symptoms are present Sarcoidosis, if clinical symptoms are present Multiple sclerosis, if neurologic signs or symptoms are present Masquerade (neoplasia), in children or elderly, if no response to

Lyme disease, if joint, skin, or neurologic signs or symptoms or in

Herpes, if keratouveitis is present Toxocariasis, if typical morphology is present Tuberculosis, in immunocompromised patient, if granulomatous

AU is present or if no response to systemic corticosteroids Masquerade (neoplasia), in children or elderly, if no response to

adequate antiinflammatory therapy Sarcoidosis, if granulomatous AU present and/or if clinical

suspicion Masquerade (neoplasia), in children or elderly, if no response to

adequate antiinflammatory therapy Sarcoidosis, if granulomatous AU present and/or if clinical

suspicion BehGet’s disease, if mucocutaneous symptoms present IBD, if abdominal symptoms present Sarcoidosis, if granulomatous AU present and/or if clinical

Behset’s disease, if mucocutaneous symptoms present IBD, if abdominal symptoms present Sarcoidosis, if granulomatous AU present and/or if clinical

adequate antiinflammatory therapy

disease

adequate antiinflammatory therapy

endemic areas

suspicion Vogt-Koyanagi-Harada

syndrome

suspicion ~~ ~ ~

* Data are based on the results of using the classification system described herein, as well as on the approach recommended in the current literature (60, 61). AU = anterior uveitis; IBD = inflammatory bowel disease; TINU = tubulointerstitial nephritis uveitis; JRA = juvenile rheumatoid arthritis. Always perform a complete ophthalmologic examination and include each case in one specific pattern, together with its characteristic morphologic findings. Rule out specific ophthalmologic syndromes and obtain a complete clinical history, a chest roentgenogram, and a syphilis serology.

When an infectious origin is suspected, such as in the case of toxocariasis (52), it may be necessary to perform serologic and microbiologic studies of the aqueous humor. Tuberculosis presents as retinal vasculitis (the most frequent presentation pattern of uveitis in tuberculosis) and as panuveitis with granulomatous vitritis. However, tuberculosis is now a rare cause of uveitis; the incidence of uveitis in patients with demonstrated tuberculosis is very low and tuberculous uveitis without extraocular involvement is exceptional (28). Chest roentgenography

rules out the presence of pulmonary tuberculosis in most cases, and skin testing with purified protein derivative tuberculin if of very limited value without other findings (48). The likelihood of tuberculosis increases in the debilitated patient, the patient with granulomatous AU, and/or the patient whose uveitis does not respond to systemic corticosteroids.

We classified 3 cases of unilateral AU (1 of them recurrent) as nonspecific viral uveitis because they presented with flu-like symptoms, but they could also have

UVEITIS AS A DIAGNOSTIC MARKER

been classified as idiopathic. We have not observed other rare causes of uveitis, such as those related to sulfonamide treatment (53), Lyme disease (4434-57), or Whipple’s disease (58). In these, extraocular symptoms habitually precede the development of uveitis. We currently reserve serology in patients with Lyme disease for those who have additional symptoms or live in endemic areas, and we recommend routine confirmation of IgG-IgM ELISA serology with Western blotting.

Finally, 131 cases (32.2%) were classified as idiopathic. Idiopathic cases presented in all but 3 of our uveitis patterns (unilateral and bilateral chorioretinitis and exudative retinal detachment panuveitis) and in all but 1 of the other morphologic findings (snow banks). It is interesting to note that half of the cases that were finally classified as idiopathic were assigned to this group in the discriminant analysis, indicating the existence of certain forms of presentation that are more characteristic of idiopathic conditions.

In summary, we have proposed a scheme for the etiologic diagnosis of uveitis (Table 6). We have described 12 discriminative patterns of clinical presentation, together with characteristic morphologic findings, and have established the relationship between these findings and the final diagnosis in these patients. With our results and with those of other published series (59-61), we can conclude that a rheumatologic evaluation can be more cost-effective if the referring ophthalmologist adheres to the classification system as recommended here, thus allowing a tailored approach in which only specific and necessary diagnostic tests are used. Further studies applying this classification system to another series of similar patients will be needed to validate the proposed diagnosis system.

ACKNOWLEDGMENT

The authors thank Dr. Isabelle Runkle d e la Vega for technical support.

REFERENCES 1.

2.

3.

4.

5.

Rosenbaum JT, Nozik RA: Uveitis: many diseases, one diagnosis. Am J Med 79:545-547, 1985 Heiden D, Nozik RA: A systematic approach to uveitis diagnosis. Ophthalmol Clin North Am 6:13-22, 1993 Freeman WR, Lerner CW, Mines JA, Lash RS, Nadel AJ, Stan MB, Tapper M L A prospective study of the ophtbalmologic findings in the acquired immune deficiency syndrome. Am J Ophthalmol 97:133-142, 1984 Bloch-Michel E, Nussenblatt RB: International Uveitis Study recommendations for the evaluation of intraocular inflammatory disease. Am J Ophthalmol 103:234-235, 1987 Dougados M, van der Linden S, Juhlin R, Huitfeldt B, Amor B, Calin A, Cats A, Dijkmans B, Olivieri 1, Pasero G, Veys E, Zeidler

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

369

H, The European Spondylarthropathy Study Group: The Euro- pean Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis Rheum 34:1218- 1227, 1991 Goei The HS, Steven MM, van der Linden SM, Cats A: Evaluation of diagnostic criteria for ankylosing spondylitis: a comparison of the Rome, New York and modified New York criteria in patients with a positive clinical history and screening tests for ankylosing spondylitis. Br J Rheumatol 24:242-249, 1985 Calin A, Fox R, Gerber RC, Gibson DJ: Prognosis and natural history of Reiter’s syndrome. Ann Rheum Dis 38:28-31, 1979 International Study Group for Behcet’s Disease: Criteria for diagnosis of Behcet’s disease. Lancet 335:1078-1080, 1990 Cassidy JT, Levinson JE, Brewer EJ Jr: The development of classification criteria for children with juvenile rheumatoid arthritis. Bull Rheum Dis 38:l-7, 1989 Winterbauer RH, Belic N, Moores KD: A clinical interpretation of bilateral hilar adenopathy. Ann Intern Med 78:65-71, 1973 Daniels TE: Salivary histopathology in diagnosis of Sjogren’s syndrome. Scand J Rheumatol 61 (Suppl):36-43, 1986 Beniz J, Forster DJ, Lean JS: Variations in clinical features of the Vogt-Koyanagi-Harada syndrome. Retina 11:275-280, 1991 Phaik CS, Seah S, Guan OS, Chandra MT, Hui SE: Anti- toxoplasma serotitres in ocular toxoplasmosis. Eye 5:636-639, 1991 Tamesis RR, Foster CS: Ocular syphilis. Ophthalmology 97:1281- 1287, 1990 Weissman SS: Masquerade syndromes. Ophthalmol Clin North

Derhaag PJ, van der Horst AR, de Waal LP, Feltkamp TE: HLA- B27+ acute anterior uveitis and other antigens of the major histocompatibility complex. Invest Ophthalmol Vis Sci 30:2160- 2164, 1989 Yazici H, Akokan G, Yal~in B, Miiftuoglu A: The high prevalence of HLA-BS in Behcet’s disease. Clin Exp Immunol 30:259-261, 1977 Feltkamp TEW. Ophthalmologic significance of HLA associated uveitis. Eye 4:839-844, 1990 Rosenbaum JT: Characterization of uveitis associated with spon- dyloarthritis. J Rheumatol 16:792-796, 1989 Edmunds L, Elswood J, Calin A New light on uveitis in ankylosing spondylitis. J Rheumatol 1850-52, 1991 Rodriguez A, Akova AA, Pedroza-Seres M, Foster CS: Posterior segment ocular manifestations in patients with HLA-B27- associated uveitis. Ophthalmology 101:1267-1274, 1994 Salmon JF, Wright JP, Murray A D N Ocular inflammation in Crohn’s disease. Ophthalmology 98:480-484, 1991 Linssen A, Rothova A, Valkenburg HA, Dekker-Saeys AJ, Luy- endijk L, Kijlstra A, Feltkamp TEW: The lifetime cumulative incidence of acute anterior uveitis in a normal population and its relation to ankylosing spondylitis and histocompatibility antigen HLA-B27. Invest Ophthalmol Vis Sci 322568-2578, 1991 Baiiares A, Jover JA, Fernandez-Gutitrrez B, Benitez del Castillo JM, Garcia J, Gonzalez F, L6pez JA, Hernandez-Garcia C Bowel inflammation in anterior uveitis and spondyloarthropathy. J Rheu- matol 22:1112-1117, 1995 Wakefield D, Stahlberg TH, Toivanen A, Granfors K, Tennant C: Serologic evidence of Yersinia infection in patients with anterior uveitis. Arch Ophthalmol 108:219-221, 1990 Kijlstra A The value of laboratory testing in uveitis. Eye 4:732- 736, 1990 Rosenbaum JT, Wernick R: Selection and interpretation of laboratory tests for patients with uveitis. Int Ophthalmol Clin 30:238- 243, 1990 Copeland RA: The classics: tuberculosis, syphilis, and sarcoidosis. Ophthalmol Clin North Am 6:69-80, 1993 Nichols CW, Eagle RC, Yanoff M, Menocal NG: Conjunctival

Am 6~127-137, 1993

370 BA~~ARES ET AL

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

biopsy as an aid in the evaluation of the patient with suspected sarcoidosis. Ophthalmology 87:287-291, 1980 Ohara K, Okubo A, Kamata K, Sasaki H, Kobayashi J, IQtamura S: Transbronchial lung biopsy in the diagnosis of suypected ocular sarcoidosis. Arch Ophthalmol 111:642-644, 1993 Michelson JB: Diffuse uveitis. Ophthalmol Clin North Am 655- 68, 1993 Kijlstra A, van den Horn GJ, Luyendijk L, Baarsma GS, Schweitzer CM, Zaal MJ: Laboratory tests in uveitis: new devel- opments in the analysis of local antibody production. Doc Oph- thalmol 75:225-231, 1990 Rothova A, van Knapen F, Baarsma GS, Kruit PJ, Loewr-Sieger DH, Kijlstra A Serology in ocular toxoplasmosis. Br J Ophthalmol

Aouizerate F, Cazenave J, Poirier L, Verin P, Cheyrou A, Be- gueret J, Lagoutte F: Detection of Toxoplasma gondii in aqueous humour by the polymerase chain reaction. Br J Ophthalmol

Rosenbaum JT, Bennett RM: Chronic anterior and posterior uveitis in primary Sjogren’s syndrome. Am J Ophthalmol 104:346- 352, 1987 Rosenberg AM: Uveitis associated with juvenile rheumatoid arthritis. Semin Arthritis Rheum 16:158-173, 1987 Leak AM: Ophthalmological screening in seronegative juvenile chronic arthritis: a personal view. Br J Rheumatol 31:631-632, 1992 Sothwood TR, Ryder CAJ: Ophthalmological screening in juvenile arthritis: should the frequency of screening be based on the risk of developing chronic iridocylitis? Br J Rheumatol31:633-634, 1992 Rosenberg AM, Oen KG: The relationship between ocular and articular disease activity in children with juvenile rheumatoid arthritis and associated uveitis. Arthritis Rheum 29:797-800, 1986 Rosenbaum JT, Robertson JE, Watzke RC: Retinal vasculitis-a primer. West Med J 154:182-185, 1991 Jabs DA, Fine SL, Hochberg MC: Severe retinal vaso-occlusive disease in systemic lupus erythematosus. Arch Ophthalmol 104:

Michelson JB, Grossman KR, Lozier JR: Iridocyclitis masquerade syndrome. Sum Ophthalmol 31:125-130, 1986 Martin DF, Chan CC, de Smet MD, Palestine AG, Davis JL, Whitcup SM, Burnier MN, Nussenblatt RB: The role of chorioretinal biopsy in the management of posterior uveitis. Ophthal- mology 100:705-714, 1993 Zierhut M, Kreissig I, Pickert A: Panuveitis with positive serolog- ical tests for syphilis and Lyme disease. J Clin Neuroophthalmol

Stafford-Brady FJ, Urowitz MB, Gladman DD, Easterbrook M:

701615-622, 1986

77:107-109, 1993

558-560, 1986

9171-75, 1989

Lupus retinopathy: patterns, associations, and prognosis. Arthritis Rheum 31:llOS-1110, 1988

46. Hundert I, Bakimer R, Amital-Teplizki H, Slor H, Yassur Y, Palestine A, Nussenblatt KB, Shoenfeld Y: Antinuclear autoantibodies in uveitis. Clin Exp Rheumatol 7:237-241, 1989

47. Young DW: The antineutrophil antibody in uveitis. Br J Ophthal- mol 75:208-211, 1991

48. Rosenbaum JT, Wernick T: The utility of routine screening of patients with uveitis for systemic lupus erythematosus or tuberculosis: a Bayesian analysis. Arch Ophthalmol 108:1291-1294, 1990

49. Rosenbaum JT: Bilateral anterior uveitis and interstitial nephritis. Am J Ophthalmol 105534-537, 1988

50. Breger BC, Leopold IH: The incidence of uveitis in multiple sclerosis. Am J Ophthalmol 62540-545, 1966

51. Zicrhut M, Foster CS: Multiple sclerosis, sarcoidosis and other diseases in patients with pars planitis. In, Intermediate Uveitis Bev Ophthalmol. Edited by WRF Boke, KF Manthey, RB Nussenblatt. Basel, Karger, 1992

52. Benitez del Castillo JM, Guilltn JL, Fenoy S, Bafiares A, Garcia J: Bilateral ocular toxocariasis demonstrated by aqueous humor ELISA assay. Am J Ophthalmol 119514-516, 1995

53. Tilden ME, Rosenbaum JT, Fraunfelder FT: Systemic sulfon- amides as a cause of bilateral, anterior uveitis. Arch Ophthalmol

54. Breedveld J, Rothova A, Kuiper H: Intermediate uveitis and Lyme borreliosis. Br J Ophthalmol 76:181-182, 1992

55. Isogai E, Isogai H, Kotake S, Yoshikawa K, Ichiishi A, Kosaka S, Sato N, Hayashi S, Oguma K, Ohno S: Detection of antibodies against Borrelia burgdorferi in patients with uveitis. Am J Oph- thalmol 112:23-30, 1991

56. Aaberg TM: The expanding ophthalmologic spectrum of Lyme disease. Am J Ophthalmol 107:77-80, 1989

57. Rosenbaum JT, Rahn DW: Prevalence of Lyme disease among patients with uveitis. Am J Ophthalmol 112462-463, 1991

58. Rickman LS, Freman WR, Green WR, Feldman ST, Sullivan J, Russack V, Relman DA: Uveitis caused by Tropheryma whippelii (Whipple’s bacillus). N Engl J Med 332:363-366, 1995

59. Rosenbaum JT: Uveitis: an internist view. Arch Intern Med

60. Rothova A, Buitenhuis HJ, Meenken C, Brinkman CJJ, Linssen A, Alberts C, Luyendijk L, K~jlstra A Uveitis and systemic diseases. Br J Ophthalmol 76:137-141, 1992

61. Rosenbaum JT: An algorithm for the systemic evaluation of patients with uveitis: guidelines for the consultant. Semin Arthritis Rheum 19:248-257, 1990

109:67-69, 1991

14911 173-1 176, 1989

Patterns of uveitis as a guide in making rheumatologic and immunologic diagnoses

Documents

Transcript of Patterns of uveitis as a guide in making rheumatologic and immunologic diagnoses