Eye (1988) 2, 641-645

Uveitis in Juvenile Chronic Arthritis: Incidence, Clinical Features and Prognosis

JACK l. KANSKI Windsor

Summary Three hundred and fifteen patients with anterior uveitis and juvenile chronic arthritis were reviewed in order to determine the incidence, visual prognosis, and the clinical characteristics of the intraocular inflammation. The overall incidence of uveitis was 20%. Approximately 25% of patients had relatively mild and/or transient involvement and an excellent visual prognosis. In 50% the uveitis was more severe but could be controlled with topical medication. In the remaining 25% the visual prognosis was poor due to the intractable nature of the uveitis and the subsequent development of vision-threatening complications. The majority of patients (74%) were under the age of 8 years when the uveitis was first diagnosed. Clinically, the intraocular inflammation was most frequently an asymptomatic, chronic, non-granulomatous, iridocyclitis which was bilateral in 71% of cases. Other ocular lesions, which were rare, included keratoconjunctivitis sicca and corneal melting.

Juvenile chronic arthritis (lCA) is defined as conflicting reports concerning the incidence arthritis affecting one or more joints for at of uveitis (between 2% to 21 % ).2.6-14 Because least 3 months in patients under the age of 16 JCA is a relatively rare condition, very few years, other causes of arthropathy which may ophthalmologists, even those with a special mimic the condition (e.g. systemic lupus interest in uveitis, have the opportunity of erythematosus, trauma, infection, rheumatic managing a large number of cases and it has fever etc.) having been excluded. JCA was been the author's impression that many feel formerly called Still's disease and in the that the visual prognosis is universally poor United States it is frequently referred to as and some have even doubted the value of any .15-16 juvenile 'rheumatoid' arthritis. The associa­ type of therapy The main purpose of this tion between anterior uveitis and lCA is now report is to describe a large number of well established and was firstreported by Ohm patients with JCA and uveitis followed over a in 1910.1 According to the mode of onset, the long period of time in order to determine three subgroups of JCA are pauciarticular, more accurately the incidence of uveitis and polyarticular, and systemic Although it is to show that in the majority of patients the agreed that young girls with pauciarticular visual prognosis is relatively good provided onset JCA who are also positive for circulat­ the intraocular inflammation is detected and ing antinuclear antibodies carry the highest treated early. The clinical characteristics of risk of uveitis and those with systemic onset the uveitis as well as other ocular lesions are JCA the least risk,2-5 the literature contains also described.

From: Juvenile Rheumatism Unit, Canadian Red Cross Memorial Hospital, Taplow and Division of Rheumatology, Clinical Research Centre, Northwick Park Hospital, Harrow.

Part of the thesis of MD, University of London.

Corrcspondence to: J. J. Kanski, FRCS, Bay Lodge, Wood hill Avenue, Gerrards Cross, Bucks, SL9 8DP. 642 JACKJ. KANSKI

Materials and methods (b) Visual prognosis The incidence of uveitis in lCA was deter­ Although there was some overlap, it was pos­ mined by studying 122 patients with JCA all sible to divide the 103 patients into four of whom fulfilledthe following criteria: groups according to the severity and duration (a) they were followed up for at least 5 years of intraocular inflammation as well as the (b) they were initially evaluated within three presence or absence of vision-threatening months of the onset of arthritis, complications. (c) they had regular slitiamp biomicroscopy by the author or one of his assistants Group 1. The seven patients in this group when they attended the combined 'joint had very mild uveitis without keratic precipi­ eye' clinic. tates and never more than + 1 cells in the In order to evaluate the visual prognosis, a aqueous humour. In only one patient did the total of 103 patients with JCA and uveitis intraocular inflammation persist for longer were studied. All patients fulfilledthe follow­ than 12 months. None of the patients were ing criteria: treated with topical corticosteroids as it was (a) they had a minimal follow up of two felt that the risk of steroid-induced complica­ years from the time of diagnosis of tions (glaucoma and cataract) was greater uveitis; the mean follow-up period was 7 than the risk of vision-threatening complica­ years 3 months (range 2 to 14 years), tions due to uveitis. (b) the uveitis was initially diagnosed prior to the development of complications (i.e. Group 2. The seventeen patients in this posterior synechiae, band keratopathy, group had one attack of uveitis which lasted secondary inflammatory glaucoma, and less than 4 months. The severity of inflamma­ complicated cataract), tion varied from +2 to +4 cells. In all (c) they were all personally examined and patients the uveitis was controlled with topi­ treated by the author although some cal dexamethasone drops. All patients were patients seen prior to January 1975 were also treated with twice daily instillation of initially managed by the author's pre­ tropicamide 1 % drops. When last examined decessor. none of the patients had developed posterior In order to establish the clinical charac­ synechiae, band keratopathy, secondary teristics of uveitis the medical records of 315 inflammatory glaucoma, or complicated patients with JCA and uveitis were reviewed. cataract. The vast majority had been examined by the author personally during a ten year period Group 3. The fifty-two patients in this group (January 1975 to December 1984), either at had moderate to severe (+2 to +4 cells) the Juvenile Rheumatism Unit, Canadian uveitis which persisted for more than 4 Red Cross Memorial Hospital, Taplow, or at months. The majority had acute exacerba­ the Division of Rheumatology, Clinical tions of intraocular inflammation which usu­ Research Centre, Northwick Park Hospital, ally could be controlled by a three day course Harrow. Apart from routine slitlamp biomic­ of very frequent (hourly or two-hourly) instil­ roscopy, applanation tonometry, and exami­ lation of dexamethasone drops. Three nation of the fundi with direct and indirect patients who were unresponsive to topical ophthalmoscopy, special investigations to medication were treated by anterior sub­ detect the possibility of asymptomatic com­ Tenon's injections of methylprednisolone plications unrelated to uveitis were not per­ (Depomedrone). When last examined most formed (e.g. staining of the conjunctiva with patients were still on topical medication, four rose bengal). had developed visually inconsequential lens opacities and one patient had lens opacities Results as well as secondary inflammatory glaucoma. (a) Incidence of uveitis. Twenty five patients out of 122 (20'/'0) of the Group 4. All 27 patients in this group had patients developed uveitis. severe (+ 3 to +4 cells) and prolonged uveitis UVEITIS INJUVENILE CHRONIC ARTHRITIS: INCIDENCE, CLINICAL FEATURES AND PROGNOSIS 643 which responded poorly to topical corticos­ two patients during acute and severe exacer­ teroid medication, Because of the intractable bations. Posterior synechiae were common in nature of the intraocular inflammation and eyes with longstanding neglected uveitis and the fear of blindness, three patients were in some cases the iris surface contained treated with systemic chlorambucil. When dilated blood vessels which extended onto last examined, all patients in this group had the lens. Keoppe iris nodules were seen in a had cataract surgery in at least one eye, and few cases. In eyes with moderate to severe thirteen patients also had secondary inflam­ uveitis, cells in the anterior vitreous were fre­ matory glaucoma, quent. In this entire group of 103 patients, the (vi) Posterior segment signs. Cystoid macular incidence of complicated cataract was 31% oedema was rare in eyes that had not under­ (32 patients) and 14% (14 patients) had sec­ gone cataract surgery. ondary inflammatory glaucoma. (d) Other ocular lesions (c) Clinical fe atures of uveitis Apart from uveitis, other ocular lesions were (i) Bilateral involvement. Seventy one per found in nine patients. Five had keratocon­ cent (233 of 315) of patients had bilateral junctivitis sicca characterised by corneal fila­ uveitis. In most cases both eyes were either ments and the typical staining pattern of the involved simultaneously or within a few conjunctiva with rose bengal. Surprisingly, months of each other. It was unusual for none of the patients had symptoms and for patients with initially unilateral uveitis to this reason treatment with tear substitutes develop involvement of the second eye after was felt to be unnecessary. In three eyes of more than one year. In those with bilateral two patients inactive chorioretinal scars were uveitis, the severity of intraocular inflamma­ seen at the posterior pole. One patient tion was usually symmetrical. developed severe unilateral corneal thinning which was not associated with scleritis. The (ii) Age at diagnosis of uveitis. Because of eye was subsequently enucleated as the the frequently asymptomatic nature of the cornea had perforated. Another patient intraocular inflammation it was not possible developed transient unilateral inferior to give a precise age of the patients at the superficialcorneal stromal infiltrates. time of its commencement. For this reason only the age at which the uveitis was initially Discussion and Conclusions detected is given. The mean age was 6 years The true incidence of uveitis in patients with (range 1 to 35 years), and 74% (233) of the JCA is dependent on the length of follow up patients were under the age of 8 years when and the care with which the eyes are the uveitis was initially diagnosed. examined. The incidence of 20% in this (iii) Symptoms. The onset of uveitis was report is still probably a slight underestima­ invariably asymptomatic and detected on tion, because a few patients develop uveitis routine slitlamp biomicroscopy. Even during many years after the onset of arthritis. The acute exacerbations with +4 cells in the incidence of uveitis in other series varies aqueous humour it was rare for patients to from 2% to 21 % .2,6-14 This discrepancy is complain, although a few older children probably due to variations in the length of noticed an increase in floaters. follow up and the fact that in some reports the patients were initially examined by non­ (iv) External signs. Acute exacerbations are ophthalmologists and no particular attention not associated with a ciliary flush. was paid to the eyes of asymptomatic (v) Slitlamp signs. Keratic precipitates were patients. usually small to medium in size and were It is evident that, as in the case of joint located mainly in the inferior half of the involvement, the intraocular inflammation cornea. A few patients had large keratic pre­ can be subdivided into several groups accord­ cipitates which in one case were of the 'mut­ ing to severity and duration. It is, however, ton fat' variety. Hypopyon was seen in only only after a prolonged follow up that the 644 JACK 1. KANSKI characteristics of these groups can be patients at regular intervals so that the condi­ delineated accurately. It is certainly true that tion can be treated prior to the development until the advent of cortisone the only availa­ of complications. Although a few patients ble treatment for uveitis was with mydriatics had Koeppe nodules and one had 'mutton and non-specific agents which had no proven fat' keratic precipitates, the uveitis was non­ anti-inflammatory action, and until Blegvad granulomatous in the majority of cases. The in 194117 reported the first case of JCA with frequent incidence of bilateral involvement uveitis unassociated with band keratopathy it of 71% is in the general agreement with other was generally assumed that most patients reports of between 67% and 89%.10,11.14.18.19 eventually became blind. Smiley (1974)18 felt Of more practical interest to the ophthal­ that although in the majority of patients the mologist and parents is the finding that it was visual prognosis was poor, there were a few unusual for children with initially unilateral with relatively mild involvement and a good uveitis to develop involvement of the second prognosis. He stated that the facts determin­ eye after a period of more than twelve ing this benign course in some patients were months. still unknown, treatment alone was not the Although no specific tests were routinely only explanation since many children who performed to detect keratoconjunctivitis had received identical therapy over a similar sicca, there were five patients with dry eyes. period of time had slowly deteriorated. It is, Chylack and co-workers2 claim to be the first however, important to emphasise that during to report keratoconjunctivitis sicca in associ­ the fifties and early sixties, when Smiley ation with JCA and uveitis. All three of their treated most of his patients, only relatively patients had symptoms in contrast to the five weak topical corticosteroids preparations patients in this study who were all such as cortisone and hydrocortisone were asymptomatic. As far as is known, this is the available in contrast to the potent corticos­ first report of peripheral corneal infiltrates teroids which are now at our disposal such as and corneal thinning in patients with uveitis dexamethasone, betamethasone, and pre­ and JCA. The significanceof transient limbal dnisone. infiltrates in one patient is difficult to inter­ This study has shown that in about 25% of pret although it is tempting to speculate that cases the prognosis is very good and the the severe peripheral corneal thinning which uveitis either requires no specific treatment eventually led to perforation and enucleation or responds readily to topical corticosteroids. in the other patients was related to JCA. At the other end of the spectrum are patients However, the finding of old chorioretinal with very severe and prolonged inflammation scars in the other two patients is almost cer­ which respond poorly to treatment. This tainly coincidental. group accounts for about 25% of cases and In adults there appears to be very little the vast majority of patients eventually overlap between the ocular complications of require surgery for complicated cataract and/ sero-positive (for rheumatoid factor) or secondary glaucoma. The largest group, rheumatoid arthritis and the sero-negative accounting for about 50% of patients, is arthritides (e.g. ankylosing spondylitis, Rei­ characterised by moderate to severe uveitis ter's syndrome etc,). In rheumatoid arthritis, which responds reasonably well to topical the ocular lesions are invariably confined to corticosteroids although treatment may have the outer eye (keratoconjunctivitis sicca, to be continued for months and even years. keratitis and scleritis) and the incidence of The long term visual prognosis is good uveitis is no different than in the general although a few eventually develop complica­ population. In contrast the main ocular com­ tions. plication of the sero-negative arthritides is The intraocular inflammation in patients anterior uveitis. It is therefore surprising that with JCA was typically an insidious, five patients with JCA, which is a sero-nega­ asymptomatic, chronic iridocyclitis. Because tive arthritis, developed keratoconjunctivitis of the initially asymptomatic nature of the sicca and another keratitis. This finding uveitis it is extremely important to screen suggests that in children the overlap between UVEITIS IN JUVENILE CHRONIC ARTHRITIS: INCIDENCE, CLINICAL FEATURES AND PROGNOSIS 645 the ocular manifestations of the sero-positive 8 Edstrom G: Band-shaped keratitis in juvenile and sero-negative may be more common rheumatoid arthritis, Acta Rheumatol Scan, than in adults. 1961; 7: 169-73, 9 Smiley WK: Iridocyclitis in Still's disease. Trans References Ophthalmol Soc UK, 1965; 85: 351-6, 1 Ohm J: Bandformige Hornhauttrubung bei 10 Stewart AJ and Hill BM: Ocular manifestations einen neunjahrigen Madchen und ihre in juvenile rheumatoid arthritis. Can J Behandlung mit sUbkonjunktivalen Jod­ Ophthalmol, 1967; 2: 58-62. kaliumeinspritzungen, Klin Mbl Augenheilk 11 Kazdan JJ, McCulloch JC, Crawford JS: Uveitis 1910,48; 243-45. in children. J Ass Med Can, 1967; 96: 385-91. 2 Chylack LT Jr, Bienfang DC, Bellows AR, 12 Schaller J and Wedgwood RJ: Juvenile Stillman JS: Ocular manifestations of juvenile rheumatoid arthritis: A review, Pediatrics, rheumatoid arthritis. Am J Ophthalmol, 1975; 1972, 50: 940--53. 79: 1026-33. 13 Goel, KM and Shanks RA: Follow-up study of 3 Ohno S, Char DH, Kimura SJ, O'Connor GR: 100 cases of juvenile rheumatoid arthritis. HLA antigens and antinuclear antibody titres Ann Rheum Dis, 1974;33: 25-31. in juvenile chronic iridocyclitis. Br J Ophthal­ 14 Spaiter HF: The visual prognosis in juvenile mol, 1977, 61: 59-61. rheumatoid arthritis, Trans Am Ophthalmol 4 Kanski JJ and Shun-Shin GA: Systemic uveitis Soc, 1975;3:554-7. syndromes in childhood: an analysis of 340 15 Tressler H: Uveitis. In Peyman GA, Sanders cases, Ophthalmology, 1984; 91: 1247-52. DR, Goldberg MF eds. Principles and prac­ 5 Schaller JG, Johnson GD, Holborow EJ, tice of ophthalmology. Philadelphia. WB Ansell BM, Smiley WK: The association of Saunders 1980: 1581. 16 antinuclear antibodies with the chronic Smith RE and Nozik RM: Uveitis. A clinical iridocyclitis of juvenile rheumatoid arthritis approach to diagnosis and management. Bal­ (Still's disease). Arthritis Rheum, 1974, 17: timore. Williams and Wilkins 1983: 156. 409-16, 17 Blegvag 0: Iridocyclitis and disease of the joints 6 Vesterdal E and Sury B: Iridocyclitis and of children. Acta Ophthalmol, 1941; 19: 219- bandshaped corneal opacity in juvenile 36, rheumatoid arthritis. Acta Ophthalmol, 1950; 18 Smiley WK: The eye in juvenile rheumatoid 28:321-37. arthritis. Trans Ophthalmol Soc UK, 1974; 94: 7 Smiley WK, May E, Bywaters EGL: Ocular 817-29. presentations of Still's disease: Its complica­ 19 Key SN and Kimura SJ: Iridocyclitis associated tions and treatment. Ann Rheum Dis, 1957; with juvenile rheumatoid arthritis. Am J 16: 371-83. Ophthalmol, 1975; 80: 425-29.