Find synonyms Find exact match
Disclosures: David T Yu, MD Nothing to disclose. Joachim Sieper, MD Grant/Research/Clinical Trial Support: Abbvie; Pfizer; Merck; Janssen [Spondyloarthritis (adalimumab, certolizumab, etanercept, golimumab, infliximab, ustekinumab)]. Speaker’ s Bureau: Abbvie; Pfizer; Merck; UCB; Novartis [Spondyloarthritis (adalimumab, certolizumab, etanercept, golimumab, infliximab)]. Consultant/Advisory Boards: Abbvie; Pfizer; Merck; UCB; Novartis [spondyloarthritis (adalimumab, certolizumab, etanercept, golimumab, infliximab)]. Paul L Romain, MD Employee of UpToDate, Inc.
Contributor disclosures are reviewed for conflicts of interest by the editorial group. When found, these are addressed by vetting through a multi-level review process, and through requirements for references to be provided to support the content. Appropriately referenced content is required of all authors and must conform to UpToDate standards of evidence.
INTRODUCTION — Reactive arthritis is conventionally defined as an arthritis that arises following an infection, although the pathogens cannot be cultured from the affected joints. It is generally regarded as a form of spondyloarthritis (SpA).
The definition, clinical features, approach to diagnosis and differential diagnosis, and management of reactive arthritis will be reviewed here. Mechanisms that may play a role in reactive arthritis and in other spondyloarthritides are discussed separately. (See "Mechanisms for the induction of rheumatic symptoms by gastrointestinal disease" and "Pathogenesis of spondyloarthritis".)
DEFINITION — The term “reactive arthritis” was introduced in 1969 as “an arthritis which developed soon after or during an infection elsewhere in the body, but in which the microorganisms cannot be recovered from the joint” . The original definition did not specify the pathogens that were accepted as causes of reactive arthritis, and, in 1999, a panel of experts determined a specific list of gastrointestinal and urogenital pathogens that could be considered causative . These included Chlamydia trachomatis, Yersinia, Salmonella, Shigella, and Campylobacter . Escherichia coli, Clostridium difficile, and Chlamydia pneumoniae have since been added to the list [3-7].
Additional causative pathogens, alternative terms, and diagnostic and therapeutic strategies for reactive arthritis have subsequently been proposed . However, none of the newer diagnostic or therapeutic approaches or alternate names has been adequately validated. Another problem is that many of the studies generating these approaches involved patients seen in rheumatology clinics or followed outbreaks of disease after exposure to a common pathogen; such patients are not likely to be representative of the affected patients in the general community. Thus, the definition of reactive arthritis is still evolving.
Two major clinical features that characterize reactive arthritis were identified :
By convention, reactive arthritis of more than six months duration was regarded as being chronic instead of acute.
The term “reactive arthritis” has sometimes been used historically to refer to the clinical triad of postinfectious arthritis, urethritis, and conjunctivitis, which was formerly called Reiter syndrome. However, these patients represent only a subset of patients with reactive arthritis [9,10].
Patients suspected of having reactive arthritis whose features initially or subsequently satisfy the Assessment of SpondyloArthritis International Society (ASAS) criteria for axial or peripheral spondyloarthritis are also considered as having a form of spondyloarthritis (SpA). A preceding episode of genital or gastrointestinal infection is included among the ASAS criteria that may support a diagnosis of peripheral SpA and the inclusion of a patient with reactive arthritis in this group. (See "Non-radiographic axial spondyloarthritis, undifferentiated spondyloarthritis, and peripheral spondyloarthritis", section on 'ASAS classification criteria'.)
EPIDEMIOLOGY — Reactive arthritis is a very uncommon disease that typically occurs in young adults, affecting both men and women. Much of the data on the prevalence and annual incidence of reactive arthritis in populations come from the Scandinavian countries. The prevalence is estimated to be 30 to 40 per 100,000 adults, and the annual incidence is from 0.6 to 27 per 100,000 [3,4,6,11]. In a US population-based study in Oregon and Minnesota, the incidence of reactive arthritis following documented enteric bacterial infections ranged from 0.6 to 3.1 cases per 100,000, depending upon the organism .
Among patients with any of the spondyloarthritis variants seen by rheumatologists, those with reactive arthritis are a small minority. This was illustrated by two registry-based studies from Spain in which 1.2 to 1.4 percent of all patients with spondyloarthritis had been diagnosed with reactive arthritis [13,14].
Most cases of reactive arthritis appear sporadically, but outbreaks may follow single source infections. In such outbreaks, the proportion of infected subjects who developed subsequent reactive arthritis has ranged from 0 to 21 percent [6,11].
The causative pathogens, incidence, and prevalence of reactive arthritis depend upon the geographic region. In general, among the pathogens, Chlamydia is probably the most endemic. It has been estimated that 4 to 8 percent of patients infected with Chlamydia develop arthritis .
CLINICAL MANIFESTATIONS — Patients with reactive arthritis typically present with an asymmetric oligoarthritis, usually one to four weeks following the inciting infection [2,6,8,16].The extent of the interval between infection and the onset of arthritis considered consistent with a reactive arthritis by expert consensus is a minimum of several days and a maximum of several weeks . In at least half of patients, all symptoms resolve in less than six months; in most patients, symptoms resolve within one year. The several types of clinical manifestations of reactive arthritis include:
Other than those symptoms due to the infection that has triggered the arthritis, the articular and extraarticular manifestations are similar regardless of the particular enteric or genitourinary organism or species of organism causing the disorder .
Preceding infection — The characteristic symptoms of the enteric or genitourinary infections that can cause reactive arthritis are diarrhea or urethritis. Patients with arthritis induced by enteric bacteria can also develop aseptic urethritis. In some patients with infectious urethritis, especially Chlamydia, the infection may be asymptomatic and is identified only upon studies performed because of suspicion of reactive arthritis . (See "Approach to the adult with acute diarrhea in developed countries" and "Epidemiology and causes of acute diarrhea in developed countries" and "Approach to dysuria in the adult man" and "Clinical manifestations and diagnosis of Chlamydia trachomatis infections".)
The enteric bacteria commonly associated with reactive arthritis include:
The genital pathogen commonly accepted to be the cause of reactive arthritis is Chlamydia trachomatis.
Other bacteria also reported to cause reactive arthritis include Chlamydia pneumoniae, Escherichia coli, Ureaplasma urealyticum, and Mycoplasma genitalium. Intravesical Bacillus Calmette-Guerin (BCG) treatment for bladder cancer has also been identified as a rare cause of reactive arthritis. (See 'Differential diagnosis' below and "Complications of intravesical BCG immunotherapy", section on 'Arthritis'.)
Reactive arthritis has also been reported in patients with human immunodeficiency virus (HIV) infection, in which it is generally thought to be related to other infections to which patients have been exposed, rather than to HIV itself [18,19]. An association between reactive arthritis and HLA-B27 has been noted in Caucasian HIV-infected patients but not in patients from sub-Saharan Africa, where the prevalence of HLA-B27 is much lower [20,21].
Musculoskeletal signs and symptoms — The musculoskeletal features of reactive arthritis include two major manifestations: arthritis and enthesitis .
Arthritis — The typical picture of arthritis, seen in rheumatology clinics, is an acute onset asymmetric oligoarthritis, often affecting the lower extremities (picture 1). However, about 50 percent of patients have arthritis in the upper extremities, and some have polyarthritis in the small joints . Less often, axial arthritis may occur, with areas involved including the spine at any level and also the sacroiliac joints. By convention, the minority of patients with arthritis that does not resolve within six months is defined as having chronic reactive arthritis.
Enthesitis — The enthesis is the site of insertion of ligaments, tendons, joint capsule, or fascia to bone; enthesitis (or enthesopathy), the term for inflammation around the enthesis, can occur in patients with reactive arthritis and other forms of spondyloarthritis. Swelling at the heels is among the most characteristic symptom of enthesitis. Common sites of heel involvement are at the insertions of the Achilles tendon and of the plantar fascia on the calcaneus. Pain, swelling, and local tenderness are suggestive clinical features. Estimates of the frequency of enthesitis in patients with reactive arthritis have ranged from 20 to 90 percent [12,22-25].
In one study in the US of patients with documented enteric infections and a symptom of reactive arthritis, enthesitis was more common than arthritis or inflammatory back pain, occurring in 89 percent of patients [12,22].
Enthesitis is discussed in more detail elsewhere. (See "Non-radiographic axial spondyloarthritis, undifferentiated spondyloarthritis, and peripheral spondyloarthritis", section on 'Enthesitis (enthesopathy)'.)
Dactylitis — Some patients also develop dactylitis, which typically presents as sausage digits (picture 2) . The frequency of dactylitis in reactive arthritis has not been reported, but in our experience is most probably less than in psoriatic arthritis. (See "Non-radiographic axial spondyloarthritis, undifferentiated spondyloarthritis, and peripheral spondyloarthritis", section on 'Dactylitis (sausage digits)'.)
Extraarticular signs and symptoms — Extraarticular involvement in reactive arthritis is associated with a variety of manifestations, which may be present during the acute or chronic phase of illness [4,16]. The relative frequency of each of these manifestations has not been well-analyzed. In one European cohort of 186 patients enrolled in a clinical trial, the frequencies of eye and skin involvement were approximately 20 and 15 percent, respectively .
Extraarticular manifestations include:
None of these manifestations are specific for reactive arthritis. Balanitis, for example, may result from any of numerous causes other than reactive arthritis. (See "Balanitis and balanoposthitis in adults", section on 'Etiology'.)
LABORATORY AND IMAGING FINDINGS — Several types of findings may be present, including:
Antecedent or concommitant infection — Laboratory tests, such as stool cultures to test for Salmonella, Shigella, Campylobacter, and Yersinia, can sometimes confirm a preceding or concomitant infection with one of the pathogens that classically induce reactive arthritis. However, by the time patients develop arthritis, the diarrhea has usually resolved, and the pathogens may no longer be retrievable. Urine and genital swab testing can sometimes detect Chlamydial infection using nucleic acid amplification techniques. (See "Screening for Chlamydia trachomatis", section on 'Nucleic acid amplification tests' and 'Chronic Chlamydia-related arthritis' below.)
Serological testing is used primarily in epidemiologic studies to test for preceding infections but is generally not useful in clinical practice [8,27,28]. Infections by Yersinia, Salmonella, Campylobacter, and Chlamydia cause strong antibody responses, and the triggering enteric infections can still be identified by serology in more than 50 percent of Yersinia or Salmonella infected patients, if such testing is performed . In communities in which these infections are endemic and often clinically silent, serological testing is not specific for recent episodes of infection.
Acute inflammatory changes — Acute phase reactants such as the erythrocyte sedimentation rate or C-reactive protein may be elevated, depending upon the degree of systemic involvement and upon the acuity and activity of disease. In many patients, however, these studies may be normal; elevations in acute phase reactants were found in less than half of patients diagnosed with reactive arthritis in one study . Some patients exhibit peripheral blood leukocytosis with increased numbers of granulocytes.
Genetic predisposition — The prevalence of HLA-B27, which is increased in patients with the various forms of spondylarthritis, including reactive arthritis, is generally estimated at 30 to 50 percent in patients with reactive arthritis, although values range widely [6,16]. In hospital-based studies with more severely affected patients, frequencies as high as 60 to 80 percent have been reported ; however, estimates in population-based studies and analyses of disease outbreaks are generally much lower and occasionally have shown no increase in HLA-B27 prevalence compared with the general population . (See "Mechanisms for the induction of rheumatic symptoms by gastrointestinal disease", section on 'Role of HLA-B27'.)
Inflammatory synovitis — The findings in synovial fluid are nonspecific and are characteristic of inflammatory arthritis, with elevated leukocyte counts, predominantly neutrophils. White blood cell (WBC) counts are typically between 2000 and 64,000 WBC per mm3 . (See "Synovial fluid analysis" and "Evaluation of the adult with monoarticular pain".)
Imaging abnormalities — There are no specific findings on plain radiographs that can establish a diagnosis of reactive arthritis. Changes are usually limited to those associated with joint swelling if inflammatory arthritis is present. Patients with concurrent or past heel pain may show calcaneal spurs, but such findings are nonspecific and can be seen in other forms of arthritis and in asymptomatic individuals .
In patients with chronic joint disease, imaging studies such as ultrasonography, scintigraphy (radionuclide bone scanning), or magnetic resonance imaging (MRI) can also identify changes consistent with peripheral synovitis, enthesitis, or sacroiliitis. (See "Clinical manifestations of ankylosing spondylitis in adults", section on 'Musculoskeletal imaging'.)
DIAGNOSIS — The diagnosis of reactive arthritis is a clinical diagnosis. There is no single definitive diagnostic test, nor are there validated diagnostic criteria. The diagnosis can generally be made in patients with all three of the following:
Algorithms derived from analysis of available cross-sectional data in the literature provide support for these criteria; although the probabilities described below are estimates, given the lack of a definitive diagnostic test for reactive arthritis :
The diagnostic values of inflammatory spinal pain and extraarticular features, including conjunctivitis, balanitis, and keratoderma blennorrhagica, have not been systematically evaluated in longitudinal studies.
Diagnostic evaluation — The diagnostic evaluation in a patient suspected of reactive arthritis depends upon the presenting symptoms and likely infectious etiology. In most patients, we do the following:
We perform the following studies, depending upon the history and findings:
Additional testing which may further increase or decrease support for the diagnosis, but which alone is not diagnostic or exclusionary, includes the following:
DIFFERENTIAL DIAGNOSIS — Acute inflammatory monoarthritis or oligoarthritis may occur in a variety of disorders. Particularly important and/or common causes of arthritis to differentiate from reactive arthritis include bacterial infection (eg, septic arthritis and Lyme disease, which are identified by joint fluid examination and culture and by serologic testing of patients potentially exposed while in an endemic area, respectively) and crystal-induced inflammation, which is diagnosed by the history and identification of crystals (typically monosodium urate or calcium pyrophosphate) upon examination of the joint fluid. The general approaches to the evaluation of acute monoarthritis and polyarthritis are discussed separately. (See "Evaluation of the adult with monoarticular pain" and "Synovial fluid analysis" and "Clinical manifestations and diagnosis of gout" and "Clinical manifestations and diagnosis of calcium pyrophosphate crystal deposition disease" and "Evaluation of the adult with polyarticular pain".)
The differential diagnosis of reactive arthritis can be guided, in part, by the pattern of symptoms and findings that are associated with the arthritis and that may suggest a related infectious or other systemic disorder. The following conditions should also be considered in the differential diagnosis:
TREATMENT — There are several major aspects to management. In some patients, particularly those with genitourinary infection, treatment of the infection that triggered the arthritis is indicated; arthritis and associated periarticular conditions should be treated in patients with symptomatic joint disease. Additional interventions may also be required for the treatment of extraarticular manifestations. (See 'Treatment of the infection' below and 'Treatment of arthritis' below and 'Treatment of other clinical features' below.)
Treatment of the infection — Antibiotics are not used to treat the arthritis specifically but may be indicated for treatment of the underlying infection if there is evidence of ongoing genitourinary infection or carriage of potentially pathogenic organisms. A role for antibiotic therapy in the treatment of chronic arthritis has not been established. (See 'Chronic Chlamydia-related arthritis' below.)
A systematic review and meta-analysis of randomized trials comparing antibiotic therapy to placebo or no antibiotics for treatment of reactive arthritis found that antibiotic treatment did not significantly reduce the likelihood of failing to achieve remission of the reactive arthritis; although results of the studies were very heterogeneous, trial design varied substantially, and many trials were at risk of bias . Rates of remission were able to be analyzed in 7 of the 12 trials included in the review, involving a total of 375 patients. Additionally, there were no significant effects of antibiotic therapy on pain, joint counts, or patient global assessment, but gastrointestinal side effects were nearly twice as likely in the patients receiving antibiotics.
Enteric infection — In general, antibiotics are not indicated for uncomplicated enteric infections, but some patients with active enteric infections may require treatment, depending upon their comorbidities and upon the specific organism. For example, therapy may be indicated in patients with severe gastrointestinal disease, in older adults, or in immunocompromised hosts. Treatment of enteric infections is discussed separately. (See "Approach to the adult with acute diarrhea in developed countries" and "Travelers' diarrhea".)
In patients with chronic reactive arthritis induced by enteric bacteria, the available evidence does not support the use of long-term antibiotics [22,23,34-37].
Genitourinary tract infection — In contrast with most patients with enteritis, patients with acute Chlamydia trachomatis infection of the genitourinary tract and their sexual partners should receive a standard antimicrobial treatment for Chlamydial infection of the genital tract. Treatment regimens are presented elsewhere. (See "Treatment of Chlamydia trachomatis infection".)
Patients with a history of Chlamydia-induced arthritis should be evaluated for recurrent genitourinary infection if arthritis or genitourinary symptoms recur and should be retreated with antibiotics if testing for Chlamydia infection is positive. Antibiotic treatment of the infection might prevent relapses of arthritis in patients with recurrent genitourinary tract symptoms alone . Prompt treatment of acute Chlamydia infections prior to the development of reactive arthritis, in both patients and partners, may lower the probability of developing reactive arthritis, although this has not been proven.
The approach to the evaluation and treatment of persistent or recurrent genitourinary tract symptoms despite initial antibiotic therapy is discussed in detail separately. (See "Treatment of Chlamydia trachomatis infection", section on 'Persistent or recurrent symptoms'.)
Chronic Chlamydia-related arthritis — We do not advise the routine use of long-term antibiotics to treat established reactive arthritis. Results are mixed in randomized trials of long-term therapy with single antibiotics, and most do not show benefit [23,34-37,39].
In contrast, one study (included in the meta-analysis) provides limited evidence to suggest that combination antibiotics might be useful in the subset of patients with polymerase chain reaction (PCR)-proven Chlamydia-induced arthritis . This trial involving 42 patients with reactive arthritis of at least six months’ duration, who tested positive for C. trachomatis or C. pneumoniae by a special PCR technique in blood or affected synovial tissue . Patients were randomly assigned to receive six months of treatment with rifampin plus either doxycycline or azithromycin or to receive placebo. The composite endpoint for improvement in musculoskeletal signs and symptoms was reached more often by the patients receiving combination antibiotics (17 of 27 [63 percent] versus 3 of 15 [20 percent], p = 0.01). Adverse events were mild and comparable between groups. However, there were limitations to the study: the numbers in each group were too small to compare the different antibiotic regimens, and the study was limited to patients with evidence of persistent Chlamydial infection by a special PCR using peripheral blood or synovial samples instead of routine urine samples. Serologic tests did not correlate with these PCR results, and the PCR assays that were used in the study are not yet available for routine clinical use and are not of proven clinical utility. Further studies with larger numbers of patients are required to confirm these findings. There is no evidence of benefit for such treatment in patients with any other form of reactive arthritis.
Treatment of arthritis — The treatment of arthritis with antiinflammatory and immunosuppressive agents can be divided into two stages: the treatment of acute reactive arthritis and the treatment of refractory (chronic) reactive arthritis, usually defined as disease of greater than six months’ duration. (See 'Acute reactive arthritis' below and 'Chronic reactive arthritis' below.)
The initial treatment goal is symptomatic relief of the arthritis, because the disease is self-limited in a substantial majority of patients, and joint injury with significant continued symptoms is infrequent. (See 'Prognosis' below.)
The approach to therapy is based upon both clinical experience in reactive arthritis and upon evidence of benefit for these therapies in other inflammatory arthritides, particularly other forms of spondyloarthritis, given the paucity of randomized trials and longitudinal observational studies in patients with reactive arthritis. (See "Non-radiographic axial spondyloarthritis, undifferentiated spondyloarthritis, and peripheral spondyloarthritis" and "Assessment and treatment of ankylosing spondylitis in adults" and "Treatment of psoriatic arthritis".)
In clinical practice, disease activity and the response to therapy are assessed by the swollen or tender joint counts, by the intensity of joint pain, by the presence and severity of enthesitis, and by global assessments by the clinician and patient. There are no validated indices for assessment of disease activity in reactive arthritis.
Acute reactive arthritis — In patients with acute reactive arthritis, we initiate treatment with nonsteroidal antiinflammatory drugs (NSAIDs), which are usually the principal form of therapy, and may administer intraarticular and/or systemic glucocorticoids in patients with disease resistant to NSAIDs. (See 'Initial therapy' below and 'Intraarticular glucocorticoids' below and 'Systemic glucocorticoids' below.)
Initial therapy — We suggest treatment with antiinflammatory doses of NSAIDs (eg, naproxen 500 mg two to three times daily, diclofenac 50 mg three times daily, or indomethacin 50 mg three to four times daily) for symptomatic therapy for most patients, unless contraindicated (eg, due to allergy, cardiovascular, or renal status). Maximum antiinflammatory doses and continuous use may be necessary to control pain and inflammation. An adequate trial of a given NSAID is usually at least two weeks in duration. Individual responses vary, and more than one NSAID may need to be tried before identifying an effective agent. There is no evidence that NSAIDs shorten or otherwise affect the course of the disease.
NSAIDs are used initially because treatment of the infection does not alleviate the signs or symptoms of arthritis, because the disease is most frequently self-limited, and because most patients will not require disease-modifying antirheumatic drugs (DMARDs) to control inflammation or to prevent erosive joint changes. The use of NSAIDs is based upon their ability to provide significant symptomatic relief in other forms of spondyloarthritis and upon clinical experience with their use in reactive arthritis [6,16,41-43]; there have been few formal trials of NSAIDs in reactive arthritis [44,45]. (See "Assessment and treatment of ankylosing spondylitis in adults", section on 'Nonsteroidal antiinflammatory drugs'.)
Patients receiving NSAIDs should be cautioned regarding the risk of gastrointestinal (GI), renal, hepatic, and cardiovascular adverse effects, including the risk of NSAID gastropathy and GI bleeding. Complete blood counts, renal function, and hepatic aminotransferases should be measured after two months on a given agent and then every three to six months thereafter in patients continuing to require antiinflammatory dosing regimens of NSAIDs, depending upon the comorbidities present. (See "NSAIDs: Therapeutic use and variability of response in adults" and "Nonselective NSAIDs: Overview of adverse effects" and "Overview of selective COX-2 inhibitors".)
Inadequate response to NSAIDs — Intraarticular glucocorticoids can be administered in patients with persistent symptoms despite treatment with NSAIDs, and oral glucocorticoids may be of benefit in patients who do not respond to these therapies.
Intraarticular glucocorticoids — In patients who do not respond adequately to NSAIDs, we suggest injection of major affected joints with intraarticular glucocorticoids. In our experience, intraarticular injection with triamcinolone acetonide (40 mg for a large joint, such as the knee, and lower doses for smaller joints), or an alternative agent in equivalent doses, is generally effective in reducing joint inflammation and in providing symptomatic relief similar to that seen in other forms of spondyloarthritis or in rheumatoid arthritis. These may result in symptomatic relief sufficient to avoid the need for oral glucocorticoids or DMARD therapy and the side effects associated with use of these agents. (See "Use of glucocorticoids in the treatment of rheumatoid arthritis", section on 'Intraarticular therapy'.)
There is no evidence in our experience, nor are there any reports, of a greater frequency of adverse events or of disease worsening associated with the use of intraarticular or systemic glucocorticoids, despite the infectious etiology of reactive arthritis and the detection of the causative microbes in the joints of some patients with reactive arthritis. This issue has not been systematically evaluated. Pain and swelling of the plantar fascia can also be relieved by local glucocorticoid injections. (See "Intraarticular and soft tissue injections: What agent(s) to inject and how frequently?" and "Use of glucocorticoids in the treatment of rheumatoid arthritis" and "Joint aspiration or injection in adults: Technique and indications" and "Joint aspiration or injection in adults: Complications".)
Systemic glucocorticoids — In patients who do not respond adequately to NSAIDs and intraarticular glucocorticoid injections or in those with a large number of involved joints, we suggest low to moderate doses of oral glucocorticoids (eg, a starting dose of prednisone 20 mg daily), which should be reduced gradually to the lowest dose required to control symptoms. In our experience, use of these agents for a limited period and in a tolerable dose range may preclude need for use of DMARD therapy. There are no randomized trials of systemic glucocorticoids in reactive arthritis. The use of glucocorticoids for inflammatory joint disease, the adverse effects of these agents, and strategies for the prevention of glucocorticoid-induced osteoporosis are described in detail elsewhere. (See "Use of glucocorticoids in the treatment of rheumatoid arthritis" and "Major side effects of systemic glucocorticoids" and "Prevention and treatment of glucocorticoid-induced osteoporosis".)
Resistant to NSAIDs and glucocorticoids — Treatment with a DMARD is indicated in patients who do not respond adequately to initial therapies. The duration prior to beginning DMARDs depends upon the degree of disease activity and upon the relative risks and benefits of NSAIDs and glucocorticoids in a given patient, given the comorbidities that are present and the dose of glucocorticoids required to control symptoms. We generally use DMARDs in patients who have not responded adequately to at least two different NSAIDs over a total of four weeks and who require ongoing therapy with more than 7.5 mg of prednisone or equivalent for more than three to six months. (See 'Chronic reactive arthritis' below.)
Chronic reactive arthritis — In patients who develop refractory reactive arthritis, usually defined as disease lasting greater than six months, or who are resistant to initial therapy for acute arthritis with NSAIDs and glucocorticoids, we initiate therapy with a nonbiologic (traditional) DMARD, usually sulfasalazine or, alternatively, methotrexate. In patients with disease that is refractory to a nonbiologic DMARD, we administer therapy with a tumor necrosis factor (TNF) inhibitor. The use of these medications for reactive arthritis is based upon limited data and upon clinical experience. (See 'Resistant to NSAIDs and glucocorticoids' above and 'Nonbiologic DMARD use' below and 'Resistant to nonbiologic DMARDs' below.)
Nonbiologic DMARD use
Sulfasalazine — We suggest the use of sulfasalazine (SSZ) in patients with refractory reactive arthritis or with reactive arthritis resistant to initial therapy for acute arthritis. We initiate treatment with 500 mg once daily and increase the daily dose stepwise (in increments of 500 mg each week) to 1000 mg twice daily. The dose can be increased to a maximum of 3000 mg/day (taken in two or three divided doses) if required. (See 'Resistant to NSAIDs and glucocorticoids' above and 'Chronic reactive arthritis' above.)
The use of SSZ in patients with reactive arthritis is based primarily upon one randomized trial in patients with reactive arthritis  and is further supported by a meta-analysis and randomized trials comparing it with placebo in patients with ankylosing spondylitis and with psoriatic arthritis; and by randomized trials comparing it with placebo and other nonbiologic DMARDs in rheumatoid arthritis. We prefer SSZ over methotrexate because of its better documented efficacy in patients with peripheral spondyloarthritis, including psoriatic arthritis. (See "Assessment and treatment of ankylosing spondylitis in adults", section on 'Sulfasalazine' and "Treatment of psoriatic arthritis", section on 'Sulfasalazine' and "Sulfasalazine in the treatment of rheumatoid arthritis".)
Benefit from SSZ was suggested by a 36-week randomized trial involving 134 patients with chronic reactive arthritis who had not responded adequately to NSAIDs . Response in the trial was defined by a composite measure that included patient self-assessment, clinician assessment, and improvement in joint pain or tenderness and in joint swelling. Results for the primary outcome measure showed a trend toward benefit with SSZ that was not statistically significant but that might be clinically meaningful if real. SSZ was well-tolerated, with minor gastrointestinal side effects in some patients.
In general, the most common adverse effect causing discontinuation of therapy with SSZ is gastrointestinal upset, but central nervous system symptoms, rash, and other infrequent side effects may also occur. The use and monitoring of SSZ, as well as its adverse effects, are discussed in detail separately. (See "Sulfasalazine in the treatment of rheumatoid arthritis", section on 'Adverse effects'.)
Methotrexate — Methotrexate (MTX) may be used as an alternative nonbiologic DMARD for patients who may be allergic to or intolerant of SSZ or who do not respond to SSZ treatment. However, MTX has not been formally studied in patients with reactive arthritis, and there is a lack of evidence to support the use of MTX for axial disease in ankylosing spondylitis, even though clinical experience suggests it may be of benefit for treating peripheral arthritis in such patients. (See "Assessment and treatment of ankylosing spondylitis in adults", section on 'Other DMARDs'.)
The use of MTX is based upon clinical experience with its use for peripheral spondyloarthritis, including psoriatic arthritis, and upon evidence in meta-analyses and in randomized trials for its efficacy in rheumatoid arthritis. In reactive arthritis, we use MTX in the same fashion and doses (15 to 25 mg given on one day weekly) as are employed in psoriatic arthritis and in rheumatoid arthritis. The use and adverse effects of MTX are described in detail elsewhere. (See "Treatment of psoriatic arthritis", section on 'Methotrexate' and "Use of methotrexate in the treatment of rheumatoid arthritis" and "Major side effects of low-dose methotrexate".)
Duration of therapy — The nonbiologic DMARDs are continued for at least four months (SSZ) or three months (MTX) at the maximally tolerated therapeutic dose (up to 3 g/day SSZ or up to 25 mg/week MTX) to determine if there is a response to therapy and are then discontinued three to six months after the patients have entered into remission, with resolution of clinical signs and symptoms of disease activity. Some patients who do not respond to the initial DMARD therapy may respond to the alternative agent, in our experience; however, whether to use a second nonbiologic DMARD or to proceed directly to the use a biologic agent in such patients has not been formally evaluated. If disease recurs, we resume therapy with the previously effective agent.
Resistant to nonbiologic DMARDs — For patients who have reactive arthritis that is refractory to NSAIDs and glucocorticoids and who do not respond to or have a contraindication to the use of sulfasalazine and methotrexate, we use an anti-tumor necrosis factor (TNF) agent (eg, etanercept 50 mg/week administered by subcutaneous injection or infliximab 3 to 5 mg/kg administered intravenously on weeks zero, two, and six and then every eight weeks). In patients who do not respond to an initial trial of one agent after three months of therapy, another TNF inhibitor can be tried instead. An effort can be made to discontinue therapy in patients who have been in remission induced by a TNF inhibitor for at least three months, but treatment should be resumed with the medication if disease then recurs.
The use of TNF inhibitors in patients with reactive arthritis is directly supported only by case reports and by small case series. Their use, rather than the use of other nonbiologic or biologic DMARDs, is also supported by clinical experience and by their effectiveness in the treatment of other spondyloarthropathies, including axial and peripheral spondyloarthritis and the various forms of psoriatic arthritis. The dosing and agents, the evidence supporting their use in patients with other types of spondyloarthritis, and the adverse effects of TNF inhibitors are described in detail elsewhere. (See "Assessment and treatment of ankylosing spondylitis in adults", section on 'Tumor necrosis factor alpha antagonists' and "Treatment of psoriatic arthritis", section on 'TNF inhibitors' and "Tumor necrosis factor-alpha inhibitors: An overview of adverse effects".)
Evidence directly supporting the use of TNF inhibitors in reactive arthritis is illustrated by the following examples [47,48]:
There are no studies of patients with reactive arthritis refractory to the above therapies, although other approaches to the treatment of chronic inflammatory arthritis have been evaluated in spondyloarthritis, including psoriatic arthritis, and in rheumatoid arthritis. In the very uncommon patients who are resistant to treatment with the agents noted above, including at least two TNF inhibitors, the diagnosis should be reevaluated. (See "Assessment and treatment of ankylosing spondylitis in adults", section on 'Other DMARDs' and "Treatment of psoriatic arthritis", section on 'Experimental therapies' and "Treatment of rheumatoid arthritis resistant to initial DMARD therapy in adults", section on 'Resistant to two TNF inhibitors' and "Treatment of rheumatoid arthritis resistant to initial DMARD therapy in adults", section on 'Resistant to standard therapies'.)
Treatment of other clinical features — Some extraarticular manifestations, including ocular involvement and mucous membrane and skin manifestations, require additional interventions. The management approaches are based upon the treatments used for these or for similar manifestations in patients with other disorders and upon clinical experience [49-51].
None of these interventions has been systematically evaluated in randomized trials or has been the subject of observational studies in patients with reactive arthritis, and most treatment studies of antiarthritic or antibiotic therapy for reactive arthritis and reviews of this disease do not discuss the effects of such therapies on these manifestations or the recommended approaches for their treatment. Extraarticular manifestations that may require treatment include the following:
PROGNOSIS — The course of reactive arthritis varies considerably, probably depending upon the triggering pathogen and the genetic background of the host [6,8]. The typical disease duration is three to five months. Most patients either remit completely or have little active disease within 6 to 12 months after presentation, but 15 to 20 percent may experience more chronic persistent arthritis. After entering remission of peripheral joint arthritis, pain is occasionally still noted in the joints, at entheses, or in the spine. Representative results are illustrated by the following:
Some patients with chronic reactive arthritis later develop features characteristic of another of the spondyloarthritides, eg, psoriatic arthritis, ankylosing spondylitis, or the arthritis associated with inflammatory bowel disease. HLA-B27 testing has been associated with a worse prognosis in some, but not all studies, with findings suggesting that patients who are HLA-B27 positive are more likely to develop a chronic spondyloarthropathy with radiographic changes [8,23,56]. Patients with the triad of postinfectious arthritis, urethritis, and conjunctivitis may also have a poorer prognosis [57,58].
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, “The Basics” and “Beyond the Basics.” The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on “patient info” and the keyword(s) of interest.)
SUMMARY AND RECOMMENDATIONS
All topics are updated as new information becomes available. Our peer review process typically takes one to six weeks depending on the issue.