A Challenging Patient

A 40-year-old female presents with a complaint of right knee pain and swelling for two days. Her symptoms started suddenly. All her other joints are normal. She has been having similar problems with arthritis for 20 years, when she noticed her first attack in a shoulder joint. The frequency of attacks has increased with age and most recently she has had approximately 10 to 15 attacks per year. The attacks can last from a few hours to three days, are almost always accompanied by swelling and erythema of the involved joint, are usually monoarticular, involving any of her peripheral joints except the distal interphalangeal joints, and are not accompanied by any fever or rigors. She denies any back symptoms. The attacks resolve completely as fast as they appear. She denies any morning stiffness or any preceding signs. She has never been diagnosed with any sexually transmitted disease and has no history of psoriasis, conjunctivitis, urethritis or inflammatory bowel disease.

She takes only OTC pain medications during the attacks. Previous rheumatologic work up was negative. Family history was positive for similar kind of arthritis in her mother, grandmother and two other distant relatives who were also females. There is no family history of gout. On examination, the involved knee was swollen, tender and warm to touch. Movement of the right knee causes severe pain and she is not able to bear weight on the involved joint because of the pain. There was no involvement of other joints or any deformities in the joints. Vital signs were normal. The rest of the physical examination was normal.

Laboratory values showed all the labs within normal limits except raised ESR (90 mm/hr). The WBC count, ANA, RF, cryoglobulin, RPR, complement levels were normal but an anti-CCP antibody test was moderately positive. X-rays of the involved knee as well as the hands and wrists were normal. She was placed on ibuprofen and the arthritis resolved in two days. She believed that pain resolved because of the natural course of disease rather than because of ibuprofen.

What is the most likely diagnosis? What treatment could reduce the frequency of subsequent attacks? What chronic disease is she at risk of developing?

Introduction

Recurrent acute monoarthritis in an adult that resolves without residual joint damage is usually caused by gout or pseudogout. The diagnosis is often made by observing sodium urate crystals in patients with gout and calcium pyrophosphate crystals in pseudogout in the synovial fluid of the inflamed joint. Another uncommon cause of recurrent reversible acute monoarthritis or oligoarthritis is palindromic rheumatism (PR). While PR is often a diagnosis of exclusion, the clinical and laboratory findings in many patients may allow the physician to make this diagnosis earlier and with greater confidence, thus leading to more appropriate treatment.

Background

Hench & Rosenberg first described PR in 1944 when they reported 34 patients seen at the Mayo Clinic between 1928 and 1944 who presented with recurrent articular and peri-articular inflammation that developed rapidly over a period of hours, and resolved spontaneously and completely over a few days.(1) These patients had no evidence of crystal-induced or infectious arthritis. To express this concept of recurrence, the term palindromic, derived from the Greek "palindromos," is used, which literally means, "To come and go."

In 1959, Ansell and Bywaters published a study describing their experience of 28 patients with PR.(2) Because many of their patients developed typical rheumatoid arthritis (RA) after several years, they concluded that PR was a variant of rheumatoid arthritis. Mattingly(1966) studied 20 PR patients followed over 10 years. He noted that half of them developed low-grade RA. He also reported a positive family history of PR and a high incidence of joint diseases in the mothers of his patients.(3)

The relation of PR to RA has been debated since then and resulted in number of studies that showed that half of PR patients, if followed over years, develop RA.

Clinical Features

Joint involvement in PR is commonly monoarticular and sudden in onset with maximal intensity reached in hours, usually without any triggering event. The duration of attack varies but usually lasts for one to three days. Symptoms resolve completely after the acute attack without residual disability or deformity. Attacks recur at irregular intervals and no seasonal relation is identified. The joints involved are mostly fingers (proximal interphalangeal and metacarpophalangeal joints), wrists, knees and shoulders. The spine and the hips are rarely affected.(1)(3)(4)(5)(6)(7)

Table 1 below gives the distribution of joints affected in PR based on five large series involving 227 patients.(1)(3)(4)(5)(6)(7)

Extra-articular tissue inflammation with subcutaneous nodules is also observed in about 30% of the patients.(1)(6) Frequently, there is a history of a previous attack and no obvious inciting event can be identified.

The patients are usually afebrile and their serum uric acid levels are normal. Joint fluid aspirations show an inflammatory synovial fluid with negative cultures and no crystals on polarizing microscopic examination.(8)

The following criteria have been proposed for the diagnosis of PR:

(Adapted from Pasero and Barbieri and Hannonen et al.)(6)(9)

A 6-month history of brief sudden-onset and recurrent episodes of monoarthritis or rarely polyarthritis or of soft tissue inflammation. Direct observation of one attack by a physician. Three or more joints involved in different attacks. Absence of erosions on radiographs. Exclusion of other types of arthritis.

In order to make a diagnosis of PR, all five criteria should be met.

Investigations

While the diagnosis of PR is essentially clinical, laboratory tests and X-rays may help in excluding other rheumatologic conditions. Elevations of acute phase reactants like ESR and CRP are observed during the attacks but can be normal between the attacks.(1)(3)(4)(5) Hemoglobin and white cell counts are usually normal. Serum complement levels are normal and immune complexes are not raised.(5)(10)

In one study, synovial fluid analysis during palindromic episodes had varying leukocyte counts (150-12,700/mm³) and composed of 2% to 66% polymorphonuclear cells. The number of WBCs in the synovial fluid did not correlate with the severity of symptoms or with the intensity of the attacks.(8) The synovial fluid did not show any birefringent crystals. These synovial fluid studies were not useful in predicting which patients with PR would evolve into RA.

X-rays of joints usually show no erosions or joint space narrowing.(1)(3)(5)(9) A relatively recent study showed that MRI would be a useful tool to differentiate early RA from PR, since early RA frequently reveals pannus formation and bone erosions on MRI (when x-rays could not detect such changes), while these findings were not seen in PR.(11)

A case report of two PR patients studied the utility of ultrasonography to predict the conversion of PR to RA. In the patient who progressed to RA, the ultrasound of the involved joint showed synovial proliferation along with effusion, while in the patient who did not progress, the ultrasound revealed only synovial effusion.(12)

In PR, the ANA test is usually negative. Recent studies have reported that the anti-CCP antibody test (anti-cyclic citrullinated peptide antibodies), which is thought to be highly specific for RA, is positive in 55% of PR patients -- a prevalence similar to chronic RA’s. Anti-CCP was also recorded in similar proportion in PR patients (42%) without serum RF, similar to that found in seronegative RA (38%). The prevalence of antikeratin antibodies (AKA) in PR is also high (36%), though lower than in patients with RA (61%).(13)

PR patients may be positive for rheumatoid factor (~42%) but the percentage positive is, compared to early RA (~75%), markedly lower. RF was more frequently positive in anti-CCP and AKA positive PR patients.(13)

Differential Diagnosis(4)

At this point, it is worthwhile to know the diseases that would mimic PR (Table 2).

Epidemiology

PR is a rare condition. It affects males and females equally and can affect people from the second to eighth decade. Family studies of PR have suggested a genetic component to its development.(3)(5)(14)(15) An association between HLA DRB1 and PR has been suggested and is a useful genetic marker for susceptibility to PR.(14) HLA B 27 is usually negative.(1)

Etiology

The etiology of PR is not yet clear. Initially, the sudden appearance and quick disappearance of symptoms and signs of synovitis, accompanied by a frequent history of food sensitivity reactions, were suggestive of an allergic etiology. But none of the allergic reactions occurred regularly in association with the articular reactions, and the eosinophil count during attacks was normal.(1) No evidence of a reduction in levels of complement factors found in serum sickness could be detected.(5)(10) As anti-CCP antibodies specific for RA are frequently present in PR, the syndrome is now considered to be an abortive form of RA.(13) Two mutations in tumor necrosis factor receptor 1, a cytokine system involved in the pathogenesis of the inflammation in RA and some autoinflammatory disorders (TRAPS), have been associated with PR.(16)

Pathology

In the synovial membranes of the acutely affected joints, there is the transient presence of an acute inflammatory cell infiltrate, dominated largely by polymorphnuclear leukocytes. Between the attacks, histological examination of joints is grossly normal. In contrast to RA, no pannus formation or cartilage destruction is seen in PR.(1)(8)

Schumacher et al. studied the synovial membranes from inflamed joints of patients with PR by electron microscopy. They found a neutrophilic infiltrate with a small number of perivascular lymphocytes and an absence of plasma cells. Fibrin and cellular thrombi were noted in some blood vessels, along with some perivascular fibrosis.(8)

The subcutaneous nodules show neither low-grade, nonspecific inflammation without central necrosis or palisading cells, as found in rheumatoid nodules, nor fibrinoid degeneration, as seen in Aschoff nodules of rheumatic fever.(1)

Evolution of PR

On long-term follow up of patients with PR, approximately one-third to one-half of patients progress to RA. This observation, together with the high prevalence of anti-CCP that is considered highly specific for RA, the occurrence of nodules similar to RA and the occasional response to DMARD treatment support the contention that PR might be a variant, prodrome or abortive form of RA. However, some features of PR, such as the different pattern of distribution of the arthritis, the frequent remissions, the absence of significant constitutional symptoms, radiographic changes and different HLA associations suggest that it is a distinctive entity.

Treatment

Because of PR’s relative rarity no major randomized controlled studies have been performed to evaluate treatment efficacy. The variable prognosis of PR and the self-limiting attacks of the disease also contribute to the difficulty in evaluating drug interventions. A number of different treatments have been used with inconsistent results.

As the attacks are self-limiting, treatment is often symptomatic. NSAIDs have been prescribed to treat and abort acute episodes. Local steroid injections can also be efficacious. Temporary splinting to rest the affected joint during an attack is beneficial. Physiotherapy is usually of no benefit.

There is no consensus with respect to the use of the RA disease-modifying antirheumatic drugs (DMARDs) such as hydroxychloroquine, gold, methotrexate, leflunomide and sulfasalazine. If the attacks are frequent and are not controlled satisfactorily with long courses of NSAIDs and simple analgesics, an anti-rheumatoid DMARD agent should be considered. However, the results of DMARD therapy have been mixed. Chrysotherapy using gold salts has been tried, with varying response rates, as measured by reduction in the frequency of acute attacks, from 20% to 70%.(3)(4)(6)(17) D penicillamine was reported to have variable success. In one series, it was very effective with complete remission in four of five patients.(18) Response to antimalarials like chloroquine and hydroxychloroquine varied in different studies, from 15% to 80%.(4)(17)(19) Some rheumatologists believe that the use of DMARDs in PR patients may prevent or delay the eventual development of RA. According to our literature review, however, none of the newer biologic agents currently approved for the treatment of RA has been used to treat PR.

Prognosis

One-third to one-half of PR patients develop RA.(3)(4)(5)(6)(7)

Certain prognostic factors have been identified that increase the risk for the development of RA in patients with PR. Studies have shown that most patients with PR who are or become seropositive for rheumatoid factor develop RA. Those who remain seronegative for more than 10 years have little or no chronic synovitis or permanent disability. Involvement of PIP and wrist joints, female sex and an older age also confer a higher likelihood of evolution of PR to classic RA.(5)(20)

Whether anti-CCP or AKA antibodies predict the development of RA is unknown.(13) RA that develops from PR has a better prognosis than classical RA in some studies.(3)(7)

Conclusion

PR is an episodic recurring arthritis with unknown aetiopathogenesis. The diagnosis is usually made in retrospect and by exclusion.

The first line of treatment should be NSAIDs and, if necessary occasional local steroid injections. The second line of treatment is an anti-rheumatoid agent.

About 50% of the patients if followed over years develop RA.