Listeria-Associated Arthritis in a Patient Undergoing Etanercept Therapy: Case Report and Review of the Literature

CASE REPORT

In January 2002, a 54-year-old Caucasian woman presenting with a painful swelling of the right shoulder joint was admitted to the rheumatology ward. Apart from a largely unremarkable medical history (appendectomy and severe postoperative bleeding following a cesarean section in 1982), the patient had suffered from seropositive rheumatoid arthritis since 1998. Diagnosis of rheumatoid arthritis was made within the first 6 months of disease and was based upon fulfilling all except one (rheumatoid nodules) of the American College of Rheumatology criteria for the diagnosis of rheumatoid arthritis. The first disease-modifying antirheumatic drug (DMARD) therapy was sulfasalazine, which was later terminated due to elevated liver function tests. Next, the patient was treated with cyclosporine, which was terminated because of high blood pressure and followed by methotrexate. Because of nausea, methotrexate dosage was never increased beyond 12.5 mg/week. The patient then elected to participate in an interleukin-10 study protocol, which was stopped after 6 months because of insufficient improvement. In February 2001, she began receiving 25 mg of etanercept (Enbrel), a tumor necrosis factor alpha (TNF-α) blocker, subcutaneously twice weekly, which was well tolerated and offered sufficient control of disease activity. Five months later, however, acute swelling of the right shoulder joint started. Because septic arthritis was suspected, etanercept was replaced by oral prednisolone, and the joint was tapped several times. Cultures, however, remained sterile. Because of chronic swelling of the right shoulder joint, the patient was referred to our department 5 months later.

At presentation, the right shoulder joint was swollen and tender, with decreased range of motion. No redness or local elevation of the skin temperature was found. No fever was detectable and no night sweats were reported. The heart rate was 88/min, and blood pressure was 160/70 mm Hg. Laboratory evaluation found elevated acute-phase reactants (erythrocyte sedimentation rate, 83 mm/h; C-reactive protein, 3.56 mg/dl). Plain X-ray films of both shoulder joints were unremarkable (Fig. 1A). Sonography of the right shoulder revealed synovitis and chronic bursitis. Magnetic resonance imaging showed massive joint effusions in the subacromial and subdeltoid bursa (Fig. 1B and C). Because septic arthritis was suspected, arthrocentesis of the swollen right shoulder was performed and 70 ml of moderately cloudy inflammatory synovial fluid was aspirated and positively cultured for Listeria monocytogenes sensitive to aminopenicillins. Culture was performed on Columbia CNA agar with 5% sheep blood (Becton Dickinson Microbiology Systems, Cockeysville, MD), and identification of L. monocytogenes was done by API-Listeria kit (Biomerieux, Marcy l'Etoile, France). Consequently, antibiotic therapy with intravenous ampicillin (2 g three times daily) was started. Four additional arthrocenteses were performed, the first of which was again positive for L. monocytogenes, whereas the following three remained sterile. Blood cultures remained sterile throughout the observation period. As soon as the patient recovered, synovectomy and bursectomy were performed and a large cystic structure filled with granulomatous and necrotic masses within the rotator cuff was removed, leaving the joint largely unaffected. The patient recovered quickly and rapidly regained mobility of her right shoulder. DMARD therapy with methotrexate (15 mg/week) was initiated. The patient remained clinically stable and was discharged 1 week later.

• Open in new tab
• Download powerpoint

FIG. 1.

Imaging of the right shoulder. (A) Plain radiograph of the right shoulder shows marked joint effusion but otherwise normal appearance of the bone and joint space. (B and C) Magnetic resonance examination of the right shoulder. Axial GE T1 weighted sequence (B) shows marked effusion of the subdeltoid bursa. After intravenous administration of gadolinium, contrast media enhancement of the diffusely enlarged and thickened synovia (arrows) (C).

This case illustrates infectious arthritis caused by L. monocytogenes in a patient treated with the TNF-α blocker etanercept, which is a fusion protein of the human TNF-α receptor 2 and the Fc portion of human immunoglobulin G. L. monocytogenes is a gram-positive, facultatively intracellular bacillus that can be isolated from a large number of environmental sources. Listeria infection is relatively rare, with an estimated 2,500 cases per year in the United States. It mostly affects debilitated patients, such as those with cancer, the immunosuppressed, such as transplant recipients, and the elderly. Whereas immunocompetent individuals usually present with nonspecific flu-like symptoms, lymphadenopathy, and gastrointestinal symptoms, immunocompromised individuals can develop meningoencephalitis and sepsis, with an overall mortality as high as 15 to 30% (39). Furthermore, in pregnant women, infection with L. monocytogenes can be transmitted to the fetus, leading to septicemia and fetal loss. The majority of sporadic cases are associated with Listeria contamination of unpasteurized dairy products, although other modes of transmission may also be involved. Rare epidemic cases of listeriosis have also been traced to exposure to contaminated food (12).

Infectious arthritis caused by L. monocytogenes is a rare event. A review of the literature revealed 29 reported cases (1, 2, 4, 5, 7-9, 11, 13, 14, 19, 25-27, 30-33, 35, 36, 40, 41, 43, 45, 47) (Table 1). The median age was 55 years (range, 22 to 80 years). Typically, disease emerged as monoarthritis of a large joint, such as the hip or knee, whereas only 9 cases involved more than one joint. The majority of these patients suffered from underlying diseases, such as rheumatic diseases, diabetes mellitus, and cancer, or had undergone organ transplantation. These conditions as well as their therapeutic management led to profound alterations of the immune system, which facilitate infection with L. monocytogenes. In addition, the presence of prosthetic joints appears to increase the likelihood for articular Listeria infection, since approximately two-thirds of cases occurred in prosthetic joints. Listeria arthritis was very unusual (1 out of 29 patients) in subjects without any underlying disease or with a negative history of joint replacement, supporting the concept that immune suppression and/or joint prosthesis are critical factors for susceptibility to articular Listeria infection. Outcomes were generally good. In about half of the cases, treatment with intravenous antibiotics followed by surgical intervention resulted in complete resolution of symptoms.

Immunity to Listeria infection is complex; antigen-specific T cells that produce macrophage-activating cytokines, such as gamma interferon, are thought to play a significant role. However, there is evidence for other important factors involved, since athymic mice can initially protect themselves against Listeria infection and since human immunodeficiency virus-infected patients appear not to be at an increased risk for listeriosis (18). In contrast, mice treated with steroids were predisposed to Listeria infection (28), and the administration of methotrexate resulted in increased Listeria multiplication (42). TNF-α, which mediates macrophage activation, is considered to be important for immunity against L. monocytogenes (16, 17, 29, 37). Aside from playing a central role in the pathogenesis of inflammatory and erosive joint disease, the physiological function of TNF-α is strongly involved in the defense against pathogens, intracellular pathogens in particular (10). Several points of evidence from experimental animal models and clinical studies support this concept. Experimental models have shown that a sublethal infectious challenge with Listeria leads to a local increase in the production of TNF-α in the infected organ (16). If TNF-α production is blocked, these animals fail to survive, whereas substitution of recombinant TNF-α yields excellent protection against Listeria infection (17). It is presently unclear how TNF-α and other cytokines, like interleukin-1 and interleukin-6, gamma interferon, and macrophage colony-stimulating factor, affect Listeria infection. However, T-cell-dependent immunity does not appear to be critical, since TNF-α is protective and TNF-α blockade also increases the severity of Listeria infection in T-cell-deficient mice (17).

In addition, much of the knowledge of the role of TNF in host defense has been gained as a result of infections emerging upon clinical use of TNF-α blocking agents. Beneficial as TNF-α blockers have turned out to be for a variety of rheumatic diseases, their efficiency in blocking TNF-α also clearly increases the risk for infections like tuberculosis, histoplasmosis, or listeriosis, as well as the risk of an accelerated and severe course of these infections (21-24, 46). Moreover, there is evidence that severe bacterial infections, noncandida fungal infections, and hepatitis B are among those that are commonly kept under control with the help of TNF-α (44). Listeriosis is one of the opportunistic infections which has a higher incidence in patients treated with TNF-α inhibitors. In 2001, the Food and Drug Administration's postlicensure safety surveillance of TNF-α antagonists listed 12 cases of listeriosis after treatment with TNF-α antagonists (38). Thus, Listeria infection is still a relatively rare event in patients receiving TNF-α blockers. However, its severity, as illustrated by the fact that 5 (41%) out of these 12 patients had a fatal outcome, makes it clinically important. So far, 31 cases of L. monocytogenes infections under TNF-α-neutralizing agents have been described in the literature (3, 6, 15, 20, 34, 40) (Fig. 2). Twenty-eight cases were associated with infliximab (Remicade), a chimeric antibody against human TNF-α, and three cases were associated with etanercept treatment. TNF-α blocker therapy was initiated for rheumatoid arthritis in 18 cases, inflammatory bowel disease in 9 cases, and juvenile rheumatoid arthritis and psoriatic arthritis in 1 patient each. For two patients, the indication for infliximab is unknown. Eight patients (26%) died from the infection. The most common clinical manifestations were septicemia (60%) and meningitis (50%).

• Open in new tab
• Download powerpoint

FIG. 2.

Listeria infection in 31 patients treated with infliximab and etanercept. The x axis indicates the percentage of total. (A) Type of TNF-α blocker. (B) Indication of TNF-α blocker. (C) Manifestation of Listeria infection. (D) Concomitant medication. Other immunosuppressive drugs include azathioprine and mercaptopurine. Three patients with methotrexate received combination therapy (one each with hydroxychloroquin, mycophenolat mofetil, and cyclosporine A). (E) Outcome of Listeria infection. Note that categories C and D contain overlapping factors.

Infectious complications with Listeria were observed more frequently in patients undergoing infliximab treatment than etanercept treatment (38, 40). This can be attributed to several different factors, and our understanding is still not complete enough at the moment to decide their relative weights. Apart from pharmacological differences concerning the way to block TNF-α by using either soluble receptors, such as etanercept, which also block lymphotoxin-beta and preferentially bind the soluble form of TNF-α, or antibodies, such as infliximab, which bind both soluble and transmembrane TNF-α but not lymphotoxin-beta, environment-based factors need to be considered. For instance, the apparently higher rate of tuberculosis infection upon infliximab treatment might be due to the fact that most patients treated with etanercept live in the United States, where tuberculosis is by far less common than in Europe, while infliximab has been more frequently used worldwide outside the United States. Thus, the increased probability of a history of exposure could therefore also be a variable contributing to higher numbers of Listeria infection with infliximab than with etanercept. Nevertheless, reports of Listeria infection upon etanercept treatment are rare, and Listeria arthritis has so far been reported only for a patient receiving infliximab therapy for rheumatoid arthritis (40).

In summary, the reported case of Listeria arthritis in a patient undergoing etanercept and corticosteroid therapy without concomitant methotrexate points to the need for a higher index of suspicion of intracellular infections in rheumatoid arthritis patients treated with anti-TNF-α drugs. Infection with L. monocytogenes can clearly occur during etanercept treatment and is a diagnostic challenge.