Shigella dysentery is recognized to trigger arthritis and ocular inflammation in predisposed HLAB-27–positive patients. In this context, ocular findings are usually limited to conjuctivitis and uveitis. We describe a 15-year-old HLA-B27–positive boy with infectious gastroenteritis complicated by polyarthritis and a severe painful keratitis.
A 15-year-old boy developed bilateral conjunctivitis with copious purulent discharge accompanied by bloody diarrhea, urethritis, and painful swelling of the right wrist 2 weeks after returning from Kenya. His travel companion, who also developed bloody diarrhea, had laboratory-confirmed Shigella dysentery, but our patient's stool cultures were negative for Salmonella, Shigella, Yersinia, Campylobacter, and Escherichia coli O157:H7. His urine samples were negative for Chlamydia and Neisseria gonorrhea. The conjunctivitis was treated by his pediatrician with topical fucidic acid and he was given oral ciprofloxacin for his presumed bacterial systemic infection.
His medical history was significant for enthesitis-related arthritis, a type of juvenile idiopathic arthritis. He was diagnosed at 7 years of age and treated with methotrexate for 2 years. He had been medication-free for the past 6 years. He was HLA-B27 positive and negative for rheumatoid factor and anti-nuclear antibody. There was no history of contact lens wear and no history of ocular trauma or contact with stagnant or dirty water in Africa.
A week after resolution of the conjunctivitis the bloody diarrhea recurred, accompanied by severe right ocular pain. On referral to the ophthalmology clinic, his visual acuity was 20/20 in each eye and slit-lamp examination of the right eye showed anterior stromal infiltration with an overlying 2 × 2.3 mm epithelial defect. A low-grade (+) anterior chamber reaction was noted. Corneal scrapings were taken for bacteria, fungi, Acanthamoeba, and herpes simplex virus polymerase chain reaction.
Intensive topical moxifloxacin and cycloplegic drops were prescribed for the right eye. The left eye was uninvolved.
He was admitted into hospital for gastroenterology investigation and pain control. Opiates were necessary to control his severe ocular pain. Over the next 24 hours, the area of corneal ulceration enlarged and the anterior stroma adopted a ground-glass appearance. As the infiltrates became increasingly opaque, satellite polymorphic subepithelial infiltrates appeared in the adjacent cornea. There was a mild anterior chamber reaction, no keratitic precipitates, and no hypopyon (Figure 1). The visual acuity in the right eye deteriorated to 20/40.
Paracentral anterior stromal keratitis in the right eye with overlying epithelial defect with characteristic ground-glass appearance.
After a total of 36 hours of topical antibiotic treatment, topical prednisolone 1% was added six times per day. Within 12 hours there was a significant improvement in the cornea. Progressive re-epithelialization was observed over the next 36 hours and the stromal infiltrate gradually cleared over subsequent weeks.
The prednisolone 1% eye drop was substituted by 0.5% unpreserved prednisolone on day 2 and gradually tapered over 3 months, with no recurrence of corneal inflammation. The topical moxifloxacin was tapered over the 6 days of hospital stay. At 4 months, the visual acuity was 20/20 in each eye. The cornea had cleared apart from mild residual subepithelial haze in the area of the previous epithelial defect. Culture of the joint aspirate grew no organism and pediatric rheumatology made a diagnosis of reactive arthritis. The persistent right wrist monoarthritis required continued oral naproxen 500 mg two times per day, supplemented by intra-articular triamcinolone (40 mg) injection 10 weeks after diagnosis. A second stool culture grew Clostridium difficile toxin, thought to be secondary to systemic antibiotics rather than the inciting pathogen.
Unlike conjunctivitis and anterior uveitis, keratitis is a rare finding in reactive arthritis,1 generally occurring at the time of recurrent ocular inflammation rather than at the time of initial conjunctivitis.1,2 Reported descriptions of keratitis in reactive arthritis include pleomorphic anterior stromal keratitis, subepithelial infiltrates, peripheral corneal location, and epithelial defect.1 Our patient's keratitis was more centrally located than other described cases and was associated with intense pain controlled only by morphine infusion. This case is a reminder that reactive arthritis is a rare differential diagnosis for an atypical paracentral corneal ulcer with anterior stromal infiltrates mimicking microbial ulcer or an infected corneal erosion.
The first published account of keratitis in reactive arthritis described a ground-glass appearance of the cornea,3 as was seen in our case and as seen in other presumed antigen-antibody complex–mediated stromal processes such as herpes simplex immune stromal keratitis.4 These changes are slow to clear, disappearing after months. The term reactive arthritis has replaced the eponymous “Reiter's syndrome” since the latter's involvement in Nazi wartime experiments has come to light.5
Reactive arthritis is an autoimmune condition that develops soon after an infection elsewhere in the body, usually acquired transmucosally such as through gastroenteritis or sexually transmitted urogenital infection. Non-validated classification criteria exist.6 Clinically, a monoarthritis or oligoarthritis follows enteric or genitourinary infection by 2 to 4 weeks (range: 1 day to 6 weeks).6,7 It is not necessary to have the full triad of arthritis, non-gonococcal urethritis, and conjunctivitis for a diagnosis of reactive arthritis, and although HLA-B27 is present in the majority of cases, it is not essential for the diagnosis either; individuals with the latter are liable to have a protracted clinical course, such as recurrent arthritis.6 Enteropathic organisms known to induce reactive arthritis include those tested in our patient's stool specimen as described above. Chlamydial genitourinary infection is another cause. The arthritis in reactive arthritis by definition should be “sterile” or “aseptic” (otherwise septic arthritis would be the diagnosis) but joint aspirates from patients with reactive arthritis have recently been found to harbor viable Chlamydia, leading to trials of systemic antibiotics for Chlamydia-induced reactive arthritis with promising results.8 Interestingly, only the ocular serotypes of Chlamydia (viz. C. trachomatis serovars A–C; etiologic agent for trachoma) had been identified in reactive arthritis aspirates, but not the genital serovars D–K.4,9
A small proportion of urogenital chlamydial infections are due to the ocular serovars.9 Chromosomal DNA of enteropathic organisms known to trigger reactive arthritis has been identified by polymerase chain reaction of synovial tissues of patients with reactive arthritis but viable organisms are not present in synovial fluid, with the possible exception of Yersinia.7 In the future, corneal specimens of patients with keratitis related to reactive arthritis could be studied by polymerase chain reaction to look for the presence of enteropathic organism markers, potentially improving our understanding of keratitis related to reactive arthritis. Our patient had a background of enthesitis-related arthritis, an inflammation of tendon insertion. Enthesitis-related arthritis is a subtype of juvenile idiopathic arthritis.10 Previous authors have suggested that enthesitis-related arthritis could be a forme fruste of reactive arthritis, possibly triggered by a subclinical enteritis; in one study, a lympho-proliferative response to reactive arthritis-triggering enteropathic antigens has been identified in the synovial fluid of affected joints of children with enthesitis-related arthritis.11
- Kozeis N, Trachana M, Tyradellis S. Keratitis in reactive arthritis (Reiter syndrome) in childhood. Cornea. 2011;30:924–925. doi:10.1097/ICO.0b013e3182000916 [CrossRef]
- Kiss S, Letko E, Qamruddin S, Baltatzis S, Foster CS. Long-term progression, prognosis, and treatment of patients with recurrent ocular manifestations of Reiter's syndrome. Ophthalmology. 2003;110:1764–1769. doi:10.1016/S0161-6420(03)00620-1 [CrossRef]
- Cass E. Interstitial keratitis occurring in a case of Reiter's disease. Br J Ophthalmol. 1949;33:454–455. doi:10.1136/bjo.33.7.454 [CrossRef]
- Holland E, Brilakis HS, Schwartz GS. Herpes simplex keratitis. In: Krachmer JH, Mannis MJ, Holland EJ, eds. Cornea, 2nd ed. St. Louis: Elsevier; 2005:1043–1074.
- Wallace DJ, Weisman MH. The physician Hans Reiter as a prisoner of war in Nuremberg: a contextual review of his interrogation (1945–1947). Semin Arthritis Rheum. 2003;32:208–230. doi:10.1053/sarh.2003.49995 [CrossRef]
- Hannu T. Reactive arthritis. Best Pract Res Clin Rheumatol. 2011;25:347–357. doi:10.1016/j.berh.2011.01.018 [CrossRef]
- Carter JD, Inman RD. Chlamydia-induced reactive arthritis: hidden in plain sight?Best Pract Res Clin Rheumatol. 2011;25:359–374. doi:10.1016/j.berh.2011.05.001 [CrossRef]
- Carter JD, Espinoza LR, Inman RD, et al. Combination antibiotics as a treatment for chronic Chlamydia-induced reactive arthritis: a double-blind, placebo-controlled, prospective trial. Arthritis Rheum. 2010;62:1298–1307. doi:10.1002/art.27394 [CrossRef]
- Gerard HC, Stanich JA, Whittum-Hudson JA, Schumacher HR, Carter JD, Hudson AP. Patients with Chlamydia-associated arthritis have ocular (trachoma), not genital, serovars of C. trachomatis in synovial tissue. Microb Pathog. 2010;48:62–68. doi:10.1016/j.micpath.2009.11.004 [CrossRef]
- Colbert RA. Classification of juvenile spondyloarthritis: enthesitis-related arthritis and beyond. Nat Rev Rheumatol. 2010;6:477–485. doi:10.1038/nrrheum.2010.103 [CrossRef]
- Saxena N, Misra R, Aggarwal A. Is the enthesitis-related arthritis subtype of juvenile idiopathic arthritis a form of chronic reactive arthritis?Rheumatology (Oxford). 2006;45:1129–1132. doi:10.1093/rheumatology/kel056 [CrossRef]