Discussion of Differential Diagnosis

The patient has acute arthritis of the ankle and probably midfoot in the perioperative setting. On a statistical basis, this is most likely a result of crystal-induced inflammation, probably due to monosodium urate (gout). Because the rates of morbidity and mortality (11%) associated with bacterial infectious arthritis are high, infection should be directly excluded as a cause of the arthritis, particularly in the postoperative patient who has several potential portals of infection. In several retrospective studies, delay in diagnosis and appropriate treatment of infectious arthritis has been shown to be associated with an increased likelihood of permanent joint dysfunction.

Among 100 consecutive patients with acute monoarticular arthritis in the hospital and emergency room setting, 80% was due to microcrystalline disease (Mandell BF, unpublished information). Other potential etiologies, such as psoriasis, enteropathic, and spondylitis, are relatively uncommon and generally occur in patients with known underlying disease (e.g., psoriasis, inflammatory bowel disease, and ankylosing spondylitis). The immediate distinction in this patient must be made between crystal-induced and infectious arthritis. This distinction cannot be reliably made on clinical grounds. The presence or absence of fever, elevated peripheral white blood cell count, or elevated acute phase reactants will not distinguish crystal-induced from septic arthritis. This has been shown in several published studies, and in the postoperative setting, there are additional reasons for the vital signs and laboratory studies to be abnormal. Radiographs are of little assistance in determining the etiology of acute peripheral joint arthritis, and obtaining these studies often delays the appropriate evaluation.

The serum urate level is generally higher than the saturation point of 6.7 mg/dL at the time of an acute attack of gout, but this is not always the case. Also, patients with infectious arthritis or acute arthritis of other etiologies may also have hyperuricemia because it is so common in the general population. Additionally, an acute decrease in the serum urate level due to the initiation of a hypouricemic drug or acute intravenous hydration is a well recognized and frequent initiator of acute gouty arthritis. Thus, measurement of the serum urate level is not a sensitive or specific test to determine the etiology of acute arthritis and should not be relied on to distinguish between crystal-induced and infectious arthritis.

The test of choice in determining the etiology for acute monoarticular arthritis is synovial fluid analysis, with culture being the gold standard to diagnose bacterial infection and polarized microscopy the test of choice for crystal-induced arthritis. Anticoagulation, in this case heparin, should not be a deterrent to synovial fluid aspiration. Gram staining of the fluid to demonstrate the presence of bacteria is notoriously insensitive and fraught with the additional problem of false-positive results due to misinterpretation. Higher synovial white blood cell counts (>50,000 cells per mm ³ ) are more frequently associated with infection than with crystals, but there is too much overlap to use the total white blood cell count or percentage of neutrophils to make this distinction with confidence. The patient with inflammatory fluid and the absence of observed crystals should generally be treated as if he or she has septic arthritis until the cultures return clearly negative. The finding of crystals in the fluid confirms the diagnosis of gout (monosodium urate) or pseudogout (calcium pyrophosphate) but does not exclude the relatively uncommon coexistence of infection and crystal disease. Thus, it is reasonable to send fluid for culture even if crystals have been observed in the fluid, especially if the suspicion for infection is particularly high (i.e., in the setting of recently documented or suspected bacterial infection elsewhere in the patient or if intra-articular corticosteroid therapy is planned).

The role of polymerase chain reaction and other molecular tests to confirm the diagnosis of septic arthritis has yet to be fully defined. Synovial fluid glucose or lactate levels are not routinely useful. Crystals, although present, may occasionally not be observed on the initial evaluation; thus, subsequent aspirated fluid samples should be re-examined for the presence of crystals. The fluid can be centrifuged in a conical test tube, and the pellet evaluated for the presence of crystals. This may increase the sensitivity of the crystal analysis. Alizarin stain may be used to facilitate the recognition of calcium-containing crystals. However, this stain is not universally available, and should be filtered before use because it tends to precipitate. If necessary, synovial fluids can be stored in the absence of anticoagulant and examined at a later point in time when a more experienced observer is available.

In the described patient, monosodium urate crystals were observed in the fluid aspirated from his ankle. The midfoot tenderness also had suggested gout as the likely diagnosis, because this is one of very few anatomic areas affected by gout but generally not by infection in the acute setting. Culture of the fluid was negative. He was treated with 1.0 mg of intravenous colchicine (a therapy no longer available in many regions due to safety concerns). He had a rapid response over 24 hours, with normalization of his temperature and relief of most of his pain, avoiding the less specific antipyretic and the gastric and renal effects of nonsteroidal anti-inflammatory drugs (NSAIDs) and the hyperglycemic and (theoretical) wound healing effects of corticosteroids. Appropriately dosed intravenous colchicine does not cause nausea or diarrhea.

It is worth noting the multiple risk factors for coronary artery disease (CAD) that this patient exhibited. This is not unusual in patients with gout.

Discussion of Differential Diagnosis

This patient was referred for management of seropositive rheumatoid arthritis (RA) and osteoarthritis; she had experienced a less than ideal response to methotrexate, and now was experiencing greater difficulties in performing daily activities due to joint stiffness. The presence of diabetes and CKD complicated the choice of medications.

The symmetrical polyarthritis is consistent with rheumatoid disease, as is the positive rheumatoid factor in the past. A positive ANA can be found in patients with RA as well as in those with autoimmune thyroid disease. The radiographs were consistent with osteoarthritis but did not exclude the coexistence of RA or an alternative inflammatory arthritis affecting the wrists. The lack of a complete response to methotrexate does not exclude the diagnosis of RA. Chronic hepatitis C, which can cause a “pseudorheumatoid” polyarthritis without erosions, was excluded by the absence of viral antibodies. There was nothing in the history or on examination to suggest psoriatic, enteropathic, or spondylitis-associated peripheral arthritis (these entities often, although not always, cause a fairly more asymmetric arthritis). The time course and examination did not suggest bacterial infection. The clinical wrist involvement, in the absence of a history of wrist trauma, made the diagnosis of generalized osteoarthritis affecting the wrists unlikely. The arthropathy preceded the dialysis; hence, dialysis-associated amyloidosis was not tenable.

Although no calcinosis was seen on radiographs, the diagnoses of chronic pseudogout and hemochromatosis were considered. However, the olecranon nodules (thought to be tophi) and the thumb lesion, a characteristic intradermal deposit of urate, strongly suggested that the arthritis was due to chronic progressive gout with likely coexistent osteoarthritis. On careful questioning, most patients with chronic gout, or their family members, will recall a history of intermittent flares at the outset of the arthritis. As in this case, occasional patients with chronic established disease have symptoms and findings of chronic swelling and discomfort which can mimic rheumatoid or psoriatic (more asymmetric) arthritis. The diagnosis of gout was confirmed by aspiration of a few drops of fluid from the olecranon bursa demonstrating monosodium urate crystals.

Retrospective studies suggest that the frequency of gout flares may decrease with the development of end stage renal disease, perhaps due to decreased inflammatory cell reactivity to the crystals. But as in this patient, symptoms from chronic proliferative gouty synovitis may not abate. The serum urate level was between 9 and 10 mg/dL on two occasions. Because renal transplantation was being considered and it was believed that the use of calcineurin antagonist therapy would likely accelerate the gouty arthritis, it was believed that it was imperative to initiate hypouricemic therapy.

She was started on celecoxib prophylactically to reduce the likelihood of a gout flare with the introduction of hypouricemic therapy. A cyclooxygenase-2 (COX-2) selective NSAID was chosen because of its slightly decreased gastrointestinal (GI) bleeding risk and lack of antiplatelet effect compared with a nonselective NSAID in a patient taking (cardioprotective) aspirin and likely having kidney failure–associated platelet dysfunction. Because the patient was already on dialysis, there was no concern over worsening the renal function with celecoxib. Low-dose allopurinol (50 mg) was subsequently added and increased to a dose that decreased the serum urate to less than 6.0 mg/dL. The patient was monitored for systemic hypersensitivity reaction or rash as the dose of allopurinol was increased. It has been suggested by some (but not all) authors that the risk of hypersensitivity reactions is higher in patients with renal dysfunction. It should be noted that the promulgated guidelines for allopurinol dosing in patients with renal insufficiency have not been rigorously validated and, if adhered to, will result in suboptimal control of serum urate levels in a significant majority of patients.

TREATMENT OF THE ACUTE ATTACK

The acute gout attack is a dramatic inflammatory response to urate crystals. In addition to being membranolytic, recent studies have elucidated a mechanism by which crystals trigger inflammatory cells to release interleukin 1 (IL-1) via a toll-like receptor ( Fig. 22-2 ), which then activates a polymeric protein complex in the cytoplasm (inflammasome). The inflammasome generates active IL-1 beta, which then triggers and amplifies the local and systemic inflammatory response to crystals. The primary role of IL-1 can be demonstrated by the ability of IL-1 antagonists to block animal responses to urate crystals and it has been used to successfully treat human gout. Downstream from the effect of IL-1, other inflammatory mediators are released including prostaglandins, leukotrienes, and interleukin 8.

Uric acid crystals activate mononuclear cells to release interleukin 1 via toll-like receptor and the NALP3 inflammasome.

(From So A. Uric acid crystals activate mononuclear cells to release interleukin 1 via toll like receptor and the NALP3 inflammasome. Arth Res Therapy 2008;10:221–7.)

Acute attacks of gout are generally responsive to high doses of several different anti-inflammatory therapies. The clinical choice of therapeutic agents is generally dependent on the risk of side effects in a given patient based on his or her comorbidities and concomitant medications, as well as the personal preferences of the prescribing physician (and patient). It has been suggested that the earlier an attack is treated, the easier it is to treat. Some authors have described their experience with having patients abort an incipient attack if low doses of oral colchicine or an NSAID are taken at the first twinge of joint pain that can be recognized by the patient as gout. Once established, an attack may need to be treated for a longer period of time until it completely resolves and does not recur when the treatment is discontinued. In one retrospective review of 90 hospitalized patients treated with intravenous colchicine, the duration of symptoms before therapy did not influence the response (Khurana, PS and Mandell, BF presented at the Soc Gen Int Med annual meeting, 1999), but there is little other information to refute the common concept that the longer an attack persists, the harder it is to quickly get it to resolve. Agents that are effective in treating the acute gout attack include virtually all NSAIDs (aspirin is generally not used for this) including some COX-2 selective ones, corticosteroids (oral, parenteral and intra-articular), and colchicine (oral and previously intravenous). Narcotics blunt some of the pain but will not resolve the attack, and some clinicians (including myself) believe that they are not as effective as the anti-inflammatory medications in relieving the pain. Recently, there have been a few case reports on the efficacy of specific biologic therapies including anti-tumor necrosis factor (TNF) and anti-IL1 agents. These are briefly discussed later in this chapter.

There have been few controlled trials studying acute gout; even fewer have included a placebo (or narcotic) arm. The historic approach to the treatment of gout attacks with repeated hourly (or bihourly) doses of oral colchicine was demonstrated to be more effective than placebo within 24 hours. However, virtually all patients suffered GI side effects with this regimen, and most clinicians avoid this approach. Using just a few oral colchicine pills (one taken hourly) at the first onset of an attack can be effective, without GI side effects, according to some authors (and patients). Intravenous colchicine was safe and effective in several retrospective studies (one published in full form ) totaling approximately 250 patients. Despite some benefits of using low dose intravenous (IV) administration (limited GI side effects with appropriate dosing; no effect on bleeding; no general antipyretic effect; no leukocytosis, which can cause diagnostic confusion; and no effect on glucose levels or renal function); the multiple published case reports of morbidity and deaths attributed to IV colchicine use (many with therapeutic overdose) have led to removal of this formulation from the marketplace.

NSAIDs have been a popular choice for treating acute gout attacks. Indomethacin has had a time-honored role as the gold standard therapy. It is believed to be quickly absorbed and rapidly effective. But given its proclivity to cause headache and confusion in the elderly, as well as gastric bleeding, and its potency in adversely affecting renal function, it has slowly and partially been replaced in general use by other NSAIDs. In a controlled double-blind comparison between indomethacin (50 mg every 8 hours) and the COX-2 selective NSAID etoricoxib (120 mg), significant pain relief was achieved with both agents by 4 hours after the first dose. Naproxen (500 mg twice daily) was shown in a different controlled trial as equivalent in efficacy to prednisolone (35 mg once daily). Pain was reduced to 50% of baseline with both agents by approximately 48 hours. NSAID therapy should be continued for several days after complete resolution of the attack to avoid resumption of the inflammation and pain. The well-accepted efficacy of NSAIDs in treating acute gout is tempered by the frequent presence of comorbidities in gouty patients, which often make the use of high-dose NSAIDs less than ideal. The demography of patients with gout includes a high prevalence of hypertension, renal insufficiency, metabolic syndrome, coronary disease and alcohol use. Owing to this situation and the increased awareness of the gastric toxicity of NSAIDs, many clinicians now use corticosteroids as first-line therapy for acute attacks of gout, particularly in hospitalized patients.

For years, there was a dearth of data supporting what clinicians already knew, that steroids effectively can treat the acute gout attack. Textbooks stated that use of corticosteroids was associated with “rebound” attacks of gout on discontinuation. My own interpretation of this hackneyed caveat is that steroid therapy was frequently used for too short of a period of time, and if therapy (steroids or NSAIDs) of an attack of gout is stopped too soon, the attack is not resolved and symptoms resume. In 1990, a small experience was reviewed that documented the efficacy and lack of rebound attacks with steroid therapy. There is still no uniformly accepted dose, and many clinicians use some permutation of a scheme that involves initial treatment with approximately 40 mg daily of prednisone or the equivalent until symptoms resolve and then slowly taper the steroid over another 7 days. Adrenocorticotropic hormone is also effective, with animal models suggesting that its mechanisms of action include a direct peripheral anti-inflammatory effect as well as eliciting cortisol release from the adrenal gland. A single dose is generally not sufficient for the reasons noted earlier, and it is expensive therapy. Intra-articular steroid is effective, assuming the joint is accessible to injection. But, even though some animal studies and a study in pediatric patients with septic arthritis suggest that systemic steroids are not harmful, and may be beneficial when provided with appropriate antibiotics, there remains a strong theoretical concern with injecting a deposit formulation of steroid into an infected joint because once injected, it cannot be removed. The difficulties in rapidly and reliably distinguishing infection from crystal-induced arthritis were discussed earlier. Thus, I have some reluctance to use intra-articular steroids to treat gout attacks if there is any increased concern over the possibility of co-existent infection.