1. Top of page
  2. Introduction
  3. Case Report
  4. Discussion

Nontuberculous mycobacterial infections such as Mycobacterium marinum are relatively rare in immunocompetent individuals. However, there have been several reports of nontuberculous mycobacterial infections in patients receiving tumor necrosis factor α (TNFα) inhibitor therapy for rheumatic disease (1–9). With the increasing use of anti-TNF agents, it is important for clinicians to recognize the clinical manifestations of nontuberculous mycobacterial infections and to be familiar with effective treatment options.

Here we describe a patient who developed an M marinum infection 1 year after receiving etanercept, a soluble p75 TNFα receptor fusion protein, which was approved in 2003 for the treatment of ankylosing spondylitis (AS). Definitive treatment was delayed because the culture results showed Candida parapsilosis. Etanercept treatment was resumed in addition to the ongoing antibiotic treatment and to date, there has been no evidence of recurrence of the infection. To our knowledge, this is the first reported case of a nontuberculous mycobacterial infection in a patient with AS who was able to resume anti-TNF therapy. Institutional Review Board approval was received to use the patient medical records in an exempt status.

Case Report

  1. Top of page
  2. Introduction
  3. Case Report
  4. Discussion

The patient, a 64-year-old man with a 20-year history of AS, was doing well while receiving etanercept (50 mg/week subcutaneously). Prior to the initiation of etanercept therapy, the patient had a negative tuberculin skin test, and a normal chest radiograph result. One year after starting etanercept therapy, he experienced a minor abrasion to his left third finger while doing woodwork. A pustular lesion developed under the fingernail and within 2 weeks, multiple, fluctuant, subcutaneous nodules extending from his finger to his distal left arm appeared. There was no fever noted. The patient was treated with an unknown antibiotic at another clinic for possible lymphadenitis.

When no improvement was noted after 2 weeks, the fingernail was removed, the pustule was drained, and wound cultures were obtained, which eventually grew C parapsilosis. Treatment with fluconazole (400 mg/day, orally) was started; however, more pustules and nodules developed over the next 4 weeks. Fluconazole was stopped after 5 weeks, and ketoconazole (200 mg twice/day, orally) and azithromycin (500 mg/day, orally) were administered. Results of a biopsy sample of 1 nodule revealed poorly formed granulomata with a moderate lymphocytic infiltrate (Figure 1). Focal necrosis was also present. Histochemical stains were negative for acid-fast bacilli and fungi. The fungal culture from the biopsy sample grew M marinum after 3 weeks. The etanercept, ketoconazole, and azithromycin were stopped and a 2-week course of trimethoprim/sulfamethoxazole (TMP/SMX; 160mg/800mg per day, orally) and doxycycline (200 mg/day, orally) was initiated.

thumbnail image

Figure 1. Subcutaneous nodule biopsy result showing poorly formed granulomata with a moderate lymphocytic infiltrate. Focal necrosis is also present (hematoxylin and eosin stained; original magnification × 600).

Download figure to PowerPoint

The patient was seen in our rheumatology clinic 5 weeks later. At that time, he reported occasionally cleaning an aquarium at home without using gloves. The size or the number of nodules had not increased, however the patient reported increasingly severe back and hip pain and morning stiffness of ≥30 minutes' duration. A physical examination revealed multiple 1–2-cm nodules extending from the dorsal left hand to an area proximal to the medial elbow. He had limited chest expansion (2.5 cm), limited occiput-to-wall distance (15.2 cm), and a normal Schober test result (6 cm). Due to the patient's relapse of AS symptoms, etanercept was restarted at a dosage of 50 mg/2 weeks, subcutaneously. An infectious disease specialist recommended continuing the TMP/SMX (160mg/800mg, orally twice per day) for an additional 3 months, and continuation of etanercept therapy at the current dose.

Three months later, there were 6 nodules located on the patient's left hand and wrist and 1 nodule was located proximal to the medial elbow. The TMP/SMX and etanercept therapy was continued. Seven months later, there was only 1 nodule, located proximal to the medial elbow. The TMP/SMX dosage was decreased to 160mg/800mg 3 times per week, and the etanercept dosage was increased to 50 mg/week, subcutaneously. Over the next 3 months, the nodule further decreased in size. Twelve months after starting TMP/SMX, there were no nodules noted and the TMP/SMX therapy was stopped. There has been no recurrence of nodules 6 months later.


  1. Top of page
  2. Introduction
  3. Case Report
  4. Discussion

TNF is vital in the host defense against mycobacteria, especially Mycobacterium tuberculosis (10). TNF inhibition in animal models reverses the formation and maintenance of granulomas, and it alters macrophage function by decreasing microbicidal activity and expression of TNF (10, 11). M tuberculosis infection is a known adverse effect of anti-TNF therapy, with susceptibility rates varying among agents. Less is known about infection with nontuberculous mycobacteria.

In humans, M marinum is usually contracted following contamination of minor skin wounds while cleaning fish aquariums or while swimming. The growth of M marinum is inhibited at 37°C, so infections are usually confined to the extremities. After an incubation period of 2 to 6 weeks, papulonodular skin lesions develop, which may progress to pustules or ulcers. The lesions usually spread to the contiguous areas of skin or along local lymphatics. Osteomyelitis, tenosynovitis, and disseminated infection can occur in more serious cases (12). The differential diagnosis of M marinum infection includes other nontuberculous mycobacteria, sporotrichosis, Bartonella henselae, tularemia, Nocardia species, and Leishmania species. A history of aquatic exposure is useful in diagnosing M marinum (12).

In general, nontuberculous mycobacterial infections are not known to cause latent infection (13), unlike M tuberculosis, which can reactivate especially in the setting of anti-TNF therapy. Skin test studies show that many adults may have had prior infection with nontuberculous mycobacteria; however, species-specific skin testing for nontuberculous mycobacteria is not available (13). Up to 95% of tissue biopsy cultures have negative acid-fast bacillus smears (13). A diagnosis is based on tissue biopsy culture results, which should be species specific; susceptibility testing is only done in select cases (13). Synovial biopsy findings of granulomas have been confused with findings found in rheumatoid arthritis, which can cause further delays in proper treatment (1, 2).

We found 9 reported cases (1–9) of nontuberculous mycobacterial infections in patients with rheumatic disease who were taking anti-TNF agents (Table 1). Two of these 9 cases had M marinum infections (1, 2). Most of these cases involved patients with rheumatoid arthritis. The mean duration of the anti-TNF therapy before the infection presented was 9 months. There may be a delay in diagnosis and treatment because of the lack of clinical suspicion and because of diagnostic difficulties. In one case, treatment was started 2.5 years after the initial presentation (3). An infection can be confused with a rheumatic disease, as illustrated in 1 case where draining granulomas were mistaken for rheumatoid nodules (2). In the 9 cases we reviewed, 6 had good outcomes, and 3 cases were complicated by pulmonary infections resulting in death (4, 5, 9).

Table 1. Nontuberculous mycobacterial infections in patients with rheumatic disease receiving anti-TNF agents*
Author, yearAge, years/sexDiagnosisAnti-TNF agentDuration of anti-TNF therapy, monthsOrganismInfection siteAntibiotic, treatment durationSurgeryOutcome
  • *

    Anti-TNF = anti–tumor necrosis factor; RA = rheumatoid arthritis; Etan = etanercept; M = Mycobacterium; CLA = clarithromycin; INH = isoniazid; RIF = rifampin; PZA = pyrazinamide; ETB = ethambutol; AZI = azithromycin; Inflix = infliximab; sc = subcutaneous; PM = polymyositis; CIP = ciprofloxacin; AMK = amikacin; CFO = cefoxitin; AS = ankylosing spondylitis; Doxy = doxycycline; TMP/SMX = trimethoprim/sulfamethoxazole.

  • Prior to infection.

  • Patient was initially treated with a macrolide for 2–3 weeks. He relapsed 18 months later, but was unable to tolerate rifampin, ethambutol, and clarithromycin combination therapy so he received intermittent single agent therapy for 9 months.

Chopra et al, 200261/MRAEtan8M marinumRight handCLA, 3 monthsYesImprovement
Yim et al, 200449/MRAEtan24M xenopiLeft hip, thoracic spineINH, RIF, PZA, ETB, 4 months; AZI, ETB, 7 monthsYesImprovement
Mufti et al, 200567/FRAInflix13M abscessusSkin, sc fat, left thighCLA, 12 monthsNoImprovement
Marie et al, 200568/MPMInflix2.5M pereginumLungRIF, INH, ETB, PZA, 9 daysNoDeath
Lam et al, 200566/MRAInflix6M marinumSkin, right wristRIF, CLA, ETB, 6 monthsYesImprovement
Maimon et al, 200771/MRAEtan12M xenopiLungMacrolide, 2–3 weeks; RIF, ETB, CLANoDeath
Van Ingen et al, 200754/MRAEtan24M szulgaiLungRIF, ETB, CLA, 18 monthsYesImprovement
Boulman et al, 200673/FRAInflix3M fortuitumLiverCIP, 3 monthsNoImprovement
Thomas et al, 200656/MRAEtan15M abscessusLung, pleural spaceAMK, CFO, CLA, 43+ daysNoDeath
Current case64/MASEtan12M marinumSkinDoxy, 2 weeks; TMP/SMX, 12 monthsNoImprovement

Five of the 9 patients we reviewed were ≥65 years old, and our patient was 64 years old. Although the safety and efficacy of infliximab and adalimumab in elderly patients (age ≥65 years) have not been specifically addressed, there is no significant difference in the incidence of serious infections in elderly patients versus patients <65 years of age who are treated with etanercept (14).

Although there are no specific guidelines for treating M marinum in immunocompromised patients, a suggested approach is to treat the patient with 2 active agents for up to 2 months after the resolution of symptoms (13). In a review of nonimmunosuppressed patients with M marinum infections (15), the most commonly prescribed antibiotics were clarithromycin, the cyclines, rifampin, and ethambutol, and there were good results in a majority of these cases. The duration of treatment ranged from 1–25 months (median 3.5 months). To our knowledge, no studies have compared the effectiveness of different antibiotic regimens, nor have there been studies regarding the optimal duration of treatment (15). Our patient's antibiotic regimen of 2 weeks of doxycycline and TMP/SMX for more than a year was not used in any of the cases that we reviewed. We chose to use TMP/SMX based on the patient's lack of response after the initial 2-week course of doxycycline plus TMP/SMX along with the recommendations of our infectious disease consultants. In cases of deeper soft tissue infection, surgical debridement plays an important role in hastening the recovery (1).

There are also no guidelines regarding the safety of resuming anti-TNF therapy during or after treatment of either M tuberculosis or nontuberculous mycobacterial infections. In most cases, the anti-TNF agent was stopped when infection was suspected. The etanercept treatment was resumed in our patient after 5 weeks of antibiotics due to a flare in AS symptoms. Etanercept treatment was also resumed in a case involving Mycobacterium xenopi affecting the hip joint and thoracic spine without evidence of recurrence (8). In a patient with peritoneal M tuberculosis, there was no recurrence seen after 12 months despite the resumption of infliximab after 4 months of antibiotics (16). It is likewise unknown if switching to another anti-TNF agent is safe. No complications were encountered after switching a patient with rheumatoid arthritis who contracted a Mycobacterium szulgai lung infection while on etanercept to adalimumab after 18 months of antibiotic therapy (3). In all cases, the antibiotics were continued in a parallel fashion. We suggest that further studies are needed to address these concerns.

Interestingly, there has been a study regarding the use of etanercept as an adjuvant immunotherapy for M tuberculosis (17). This phase I study introduced etanercept on day 4 of the anti-tuberculosis treatment in 16 subjects with M tuberculosis associated with human immunodeficiency virus type 1. The sputum cultures became negative a median of 7 days earlier in the patients treated with etanercept than in the controls. The CD4 counts increased, and there were trends toward superior resolution of lung infiltrates and cavities in the etanercept-treated patients (17). To date, other anti-TNF agents have not been studied as adjunctive therapy for M tuberculosis infections in etanercept-treated patients.


  1. Top of page
  2. Introduction
  3. Case Report
  4. Discussion

Dr. Dare had full access to all of the data in the study and takes responsibility for the integrity of the data.

Acquisition of data. Dare, Jahan, Torralba.

Analysis and interpretation of data. Dare, Jahan, Torralba.

Manuscript preparation. Dare, Jahan, Hiatt, Torralba.


  1. Top of page
  2. Introduction
  3. Case Report
  4. Discussion
  • 1
    Chopra N, Kirschenbaum A, Widman D. Mycobacterium marinum tenosynovitis in a patient on etanercept therapy for rheumatoid arthritis. J Clin Rheumatol 2002; 8: 2658.
  • 2
    Lam A, Toma W, Schlesinger N. Mycobacterium marinum arthritis mimicking rheumatoid arthritis. J Rheumatol 2006; 33: 8179.
  • 3
    Van Ingen J, Boeree M, Janssen M, Ullmann E, de Lange W, de Haas P, et al. Pulmonary Mycobacterium szulgai infection and treatment in a patient receiving anti-tumor necrosis factor therapy. Nat Clin Pract Rheumatol 2007; 3: 4149.
  • 4
    Maimon N, Brunton J, Chan A, Marras TK. Fatal pulmonary Mycobacterium xenopi in a patient with rheumatoid arthritis receiving etanercept. Thorax 2007; 62: 73940.
  • 5
    Marie I, Heliot P, Roussel F, Herve F, Muir JF, Levesque H. Fatal Mycobacterium peregrinum in refractory polymyositis treated with infliximab [letter]. Rheumatology (Oxford) 2005; 44: 12012.
  • 6
    Boulman N, Rozenbaum M, Slobodin G, Rosner I. Mycobacterium fortuitum infection complicating infliximab therapy in rheumatoid arthritis [letter]. Clin Exp Rheumatol 2006; 24: 723.
  • 7
    Mufti A, Toye B, Mckendry R, Angel JB. Mycobacterium abscessus infection after use of tumor necrosis factor α inhibitor therapy: case report and review of infectious complications associated with tumor necrosis factor α inhibitor use. Diagn Microbiol Infect Dis 2005; 53: 2338.
  • 8
    Yim K, Nazeer S, Kiska D, Rose FB, Brown D, Cynamon MH. Recurrent Mycobacterium xenopi infection in a patient with rheumatoid arthritis receiving etanercept. Scand J Infect Dis 2004; 36: 1503.
  • 9
    Thomas JE, Taoka CR, Gibbs BT, Fraser SL. Fatal pulmonary Mycobacterium abscessus infection in a patient using etanercept. Hawaii Med J 2006; 65: 125.
  • 10
    Flynn J, Goldstein M, Chan J, Triebold KJ, Pfeffer K, Lowenstein CJ, et al. Tumor necrosis factor α is required in the protective immune response against Mycobacterium tuberculosis in mice. Immunity 1995; 2: 56172.
  • 11
    Roach DR, Bean AG, Demangel C. TNF regulates the chemokine induction essential for cell recruitment, granuloma formation, and clearance of mycobacterial infection. J Immunol 2002; 168: 462070.
  • 12
    Adhikesavan LG, Harrington TM. Local and disseminated infections caused by Mycobacterium marinum: an unusual cause of subcutaneous nodules. J Clin Rheumatol 2008; 14: 15660.
  • 13
    Griffith DE, Aksamit T, Brown-Elliott BA, Catanzaro A, Daley C, Gordin F, et al. An official ATS/IDSA statement: diagnosis, treatment and prevention of nontuberculous mycobacterial diseases. Am J Respir Crit Care Med 2007; 175: 367416.
  • 14
    Fleischmann R, Iqbal I. Risk: benefit profile of etanercept in elderly patients with rheumatoid arthritis, ankylosing spondylitis, or psoriatic arthritis. Drugs Aging 2007; 24: 23954.
  • 15
    Aubry A, Chosidow O, Caumes E, Robert J, Cambau E. Sixty-three cases of Mycobacterium marinum infection: clinical features, treatment, and antibiotic susceptibility of causative isolates. Arch Intern Med 2002; 162: 174652.
  • 16
    Matsumoto T, Tanaka T, Kawase I. Infliximab for rheumatoid arthritis in a patient with tuberculosis [letter]. N Engl J Med 2006; 355: 7401.
  • 17
    Wallis RS, Kyambadde P, Johnson JL, Horter L, Kittle R, Pohle M, et al. A study of the safety, immunology, virology and microbiology of adjunctive etanercept in HIV-1-associated tuberculosis. AIDS 2004; 18: 25764.