Mycobacterium terrae complex (Mycobacterium triviale, Mycobacterium nonchromogenicum, and M. terrae) has been described as a group of nonpathogenic slow-growing saprophytes. First described as the “radish bacillus,” having been isolated from radish washings and later from soil and vegetables, it was unclear whether M. terrae complex was pathogenic.1,2 A few case reports and literature reviews have confirmed the pathogenicity of M. terrae that range from chronic tenosynovitis in the immunocompetent patient to miliary disease in the immunodeficient population.1–3 A number of other difficult-to-diagnose and difficult-to-treat atypical mycobacteria such as Mycobacterium marinum, Mycobacterium fortuitum, and Mycobacterium avium-intracellulare can cause similar presentations.4Mycobacterium terrae of the knee has been reported in 4 previous cases,5–7 but the clinical outcomes were reported only as “improved” or with clinical failure. We report a case of septic arthritis and tenosynovitis of the knee caused by M. terrae in a 61-year-old man that was successfully treated with a 6-month multidrug regimen.
A 61-year-old African American man had a 25-year history of intermittent right knee pain that became worse in 2006 and was initially self-treated with rest and ibuprofen. The pain increased and persisted for 1 year, and consequently the patient sought an orthopedic evaluation. “Old age arthritis” was initially diagnosed, and the patient was subsequently treated conservatively with pain medication. An initial magnetic resonance imaging (May 2008) showed a right meniscal tear, and the patient underwent arthroscopy. Because his knee remained swollen and painful after 5 months, a second magnetic resonance imaging (October 2008) was performed and confirmed increased meniscal damage. A repeat arthroscopy was performed in February 2009 and according to the patient had “a lot of debris cleaned out.” An autoimmune process was suspected, so the patient consulted a rheumatologist, but diagnostic tests were unrevealing. The patient was prescribed sulfasalazine but opted to try natural methods for a period of 3 months. In June 2009, after no improvement from the natural approach, the patient was administered methotrexate, which he continued until October 2009.
A synovectomy was performed in October 2009, and knee fluid showed 56,600 white blood cells (WBCs)/mL, of which 62% were polymorphonuclear cells, 24% were lymphocytes, and 14% were monocytes; crystals were sought at this time but were not found. Fungal and bacterial cultures were obtained from the synovial fluid, but there was no growth after a week. An infectious disease consultation was obtained on November 25, 2009. The initial physical examination showed a warm, erythematous, and tender right knee. Laboratory values showed WBC count of 6.1 × 109/L, hematocrit of 37.1%, and platelet count of 264,000 × 109/L, and quantitative immunoglobulins (IgG, IgA, and IgM) were within normal limits. The erythrocyte sedimentation rate was 91 mm/hr, and C-reactive protein was 5.0 mg/dL (normal, <0.8 mg/dL). Cultures of the aspirated knee fluid taken on October 22, 2009, were reported to have grown an acid-fast bacillus on November 13, 2009 (22 days later), which was identified as M. terrae on November 25, 2009 (34 days later), with susceptibility to rifabutin, clarithromycin, and sulfamethoxazole reported on December 10, 2009, 15 days later (total of 54 days after the culture was obtained). The patient was empirically administered a course of clarithromycin (500 mg twice a day) and sulfamethoxazole-trimethoprim (1 double strength twice a day) for a 6-month course. Follow-up knee aspiration on January 15, 2010, showed a WBC cell count of 6889/mL; erythrocyte sedimentation rate was 10 mm/hr, and C-reactive protein was 0.3 mg/dL. The patient continued to gradually improve and started participating in more strenuous exercise including cycling. Follow-up on December 7, 2011, confirmed that the swelling and erythema had resolved, and no symptoms were reported by the patient. After completing a 6-month course of therapy, the patient had remained asymptomatic for 3 months.
REVIEW OF THE LITERATURE
Description of M. terrae Complex and Epidemiology
Mycobacterium terrae complex is a Runyon group III, nonpigmented, slow-growing group of organisms consisting of M. terrae, M. nonchromogenicum, and M. triviale.1 From 1966 to 2000, a total of 54 cases of M. terrae complex causing clinical disease were reported, with synovial and joint infections of the upper extremities representing the majority (32/54 [59%]) of cases.5,6,8,9 Although the most severe cases caused by M. terrae complex are in immunocompromised hosts, the organism is fully capable of causing disease in the immunocompetent patient, with a spectrum that includes tenosynovitis, septic arthritis, and osteomyelitis of the upper and lower extremities. It has also been reported to cause gastrointestinal tract, urinary tract, pulmonary, and disseminated infections.2,3,6,10,11
The majority of cases of M. terrae complex infection occur in the upper extremities, usually involving the fingers, hands, or wrists, but lower-extremity infections in the hips and knees have been reported as well.5–7 We report the fifth case of septic knee arthritis due to M. terrae and review the literature (Table 1). Whereas penetrating injury in a marine or agricultural environment or inoculation by a contaminated needle has been associated with the introduction of M. terrae into the synovial space or tendon sheath, in many cases, an inciting event is not found.1,4,10,12,13 It appears that all the knee cases have had either intra-articular steroids (4 cases) or systemic corticosteroids.
The onset of disease is usually insidious, with symptoms ranging from mild monoarticular swelling and stiffness of the joint without pain or erythema to painful and significant polyarticular (but unilateral) swelling, erythema, numbness, or tingling in the distal extremity and decreased range of motion.14–17 Duration of symptoms until diagnosis for knee cases ranged from 3 to 39 months until the diagnosis was established and symptoms continued from 3 weeks to 6 years of little to no improvement with standard osteoarthritic or rheumatologic treatments10 for other types of presentations. Our patient had symptoms for approximately 36 months and had tried many other forms of therapy until the diagnosis was established. Synovial fluid aspiration has yielded between 5300 WBCs/μL with 90% polymorphonuclear cells to 9540 WBCs/μL and 78% polymorphonuclear cells, findings that may be indistinguishable from other etiologies but clearly show an inflammatory process.
Atypical mycobacterial infections of the joints, including M. terrae infection, are a cause of chronic monoarticular/polyarticular swelling and may be confused with or a complication of degenerative, rheumatologic, and infectious arthritides.5,6 Relative or absolute immunodeficiency through disease process or iatrogenic means can mask or even aid mycobacterial pathogenicity.2,6,11,18,19 Treatment for other osteoarticular or autoimmune conditions without an accurate or with an equivocal diagnosis may worsen the patient’s condition, due to the anti-inflammatory properties of intra-articular or systemic corticosteroid use1,6,10,12,14,15,17; introduction may occur through an improperly sterilized needle into joint spaces or through contaminated fluids used for arthroscopy or unsterilized arthroscopic devices.10,11 The organism has been identified as a contaminant in water supplies and laboratory equipment, causing pseudo-outbreaks and confusing the clinical picture.5,20Mycobacterium terrae should remain a diagnostic consideration in cases of chronic and insidious monoarticular swelling of unknown etiology with no obvious cause of autoimmunity or joint disease on routine clinical, radiographic, or laboratory findings.
Histological and Microbiological Findings
Mycobacterium terrae complex has been isolated from a variety of sources5,6,10,11 but is most successfully recovered from synovial and tendinous material after tenosynovectomy.10,12,14 However, M. terrae complex, as well as other mycobacterial pathogens, is frequently unidentified even with Ziehl-Neelsen preparation, and only indirect findings of mycobacterial infection (granulomatous inflammation with or without caseation, giant cells, lymphocytic infiltration, and focal fibrinoid necrosis) are found, which may serve as the initial suspicion of closed-space atypical mycobacterial infection.1,3–6,10,11,18,19 Our patient in the case report did not have histological analysis of synovectomy tissue, with clinicians unfortunately opting for only direct fungal culture of aspirated knee fluid.
Establishing the diagnosis of M. terrae infection by culture may take between 4 and 11 weeks after mycobacterial culture has been obtained (Table 1). It grows best on Lowenstein-Jensen agar and yields cream-colored nonphotochromogenic colonies.2,6 Growth characteristics include growth at 25°C, 30°C, and 37°C on Lowenstein-Jensen agar and 24°C and 37°C growth on glycerol agar. Growth on Lowenstein-pyruvate and Lowenstein-thiopen agar, positive catalase activity at 68°C, and no growth on 5% NaCl or MacConkey agar are characteristic of the organism.1,2,7,15 Our case report established the diagnosis of atypical mycobacterial infection through fungal culture of aspirated knee fluid, but detected only after 22 days of growth.
In Vitro Antibiotic Susceptibility and Treatment of M. terrae Complex
The optimal treatment of M. terrae infections has not been defined. Although antimycobacterial agents rifampin and isoniazid have been used to treat M. terrae complex infections, they have not been established as a standard of care or as empiric therapy because of the paucity of data and variable success. Some cases of M. terrae complex infection treated successfully with antituberculous chemotherapy have shown resistance in vitro to the same agents.2,6 In addition, some cases of M. terrae joint infection have resolved with surgery alone1,4,14,16 or with surgery and antituberculosis chemotherapy.5–7,9,12,17,19 Previous studies have shown M. terrae complex, as well as other mycobacterial pathogens responsible for monoarticular septic arthritis, to be most susceptible to clarithromycin, ethambutol, and rifampicin. Mycobacterium terrae complex has also shown susceptibility to ethionamide, azithromycin, amikacin, streptomycin, linezolid, ciprofloxacin, cycloserine, trimethoprim-sulfamethoxazole, and some aminoglycosides.5,6,10,12,16–19 Chemotherapy agents with ethambutol and rifamycin, the macrolides, aminoglycosides, fluoroquinolones, and oxazolidinones may all play a significant role in resolving an M. terrae complex joint infection. Most authors advocate avoiding monotherapy, and the patient in our case report was treated using a combination of clarithromycin and sulfamethoxazole-trimethoprim for 6 months.
Several case reports in nonknee cases have shown successful outcomes with tenosynovectomy or debridement alone without recurrence of disease, whereas others show that surgery alone led to persistent, progressive, and debilitating disease, which even led to amputation.5–7,12,15,17,19 Milne and colleagues5 described their case of a 21-year-old man with left knee infectious arthritis caused by M. terrae treated with multiple arthroscopies, synovectomies, and an antimycobacterial regimen of clarithromycin, ciprofloxacin, linezolid, and ethambutol over the course of 8 years, with the patient ultimately ceasing all antibiotic medication and no clinical improvement achieved. Smith and collegues6 described the case of a 42-year-old woman with M. terrae infection in the right index finger in which numerous synovectomies and chemotherapeutic agents were used and alternated without clinical improvement, ultimately resulting in a ray amputation of the right index finger with no recurrence of M. terrae infection since amputation. Smith and colleagues6 also described statistics regarding 29 upper-extremity cases of the 52 total cases in their literature review. Of the 29 patients infected with M. terrae in the wrist, finger, or hand, 21% had persistent or recurrent infection, and 10% ultimately required amputation, with no correlation to a comorbid condition in these cases. Petrini and colleagues15 described their case series of 6 patients who received repeated excisions and extirpations of tenosynovium and drainage of pus in addition to chemotherapeutic intervention, resulting in permanently impaired range of motion, osteitis, arthritis, and a persistent fistula in 1 patient. Even with reports of surgical and chemotherapeutic failures in the treatment of M. terrae, surgical intervention helps in decreasing the diseased tissue burden and area of necrosis to a level where both chemotherapeutic agents and the immune system can reach the area of infection.6,20 The patient in our case report had been treated with synovectomy and multiple arthroscopies; however, improvement did not occur through surgical means alone.
Previous treatment for suspected rheumatologic or orthopedic conditions such as intra-articular corticosteroid injections delayed diagnosis as well as promoted proliferation of M. terrae and exacerbated the patient’s condition. Clinicians should consider atypical mycobacteria, specifically M. terrae complex, in the differential diagnoses of monoarticular pain and swelling when common rheumatologic and osteoarticular conditions have been ruled out and before initiating empiric corticosteroid therapy. There is no clear answer regarding the efficacy of surgical treatment alone or combined with antibacterial therapy for M. terrae knee infections. The patient in our case report underwent both surgical and chemotherapeutic modalities with resolution of the disease process after initiation of a multidrug antimycobacterial regimen.
1. Halla JT, Gould JS, Hardin JG. Chronic tenosynovial hand infection from Mycobacterium terrae
. Arthritis Rheum
. 1979; 22: 1386–1390.
2. Krisher KK, Kallay MC, Nolte FS. Primary pulmonary infection caused by Mycobacterium terrae
complex. Diagn Microbial Infect Dis
. 1988; 11: 171–175.
3. Peters EJ, Morice R. Miliary pulmonary infection caused by Mycobacterium terrae
in an autologous bone marrow transplant patient. Chest
. 1991; 100: 1449–1450.
4. Marchevsky AM, Damsker B, Green S, et al.. The clinicopathological spectrum of non-tuberculous mycobacterial osteoarticular infections. J Bone Joint Surg Am
. 1985; 67: 925–929.
5. Milne BW, Arnold MH, Hudson B, et al.. Infectious arthritis of the knee caused by Mycobacterium terrae
: a case report. J Orthop Surg
. 2009; 17: 103–108.
6. Smith DS, Lindholm-Levy P, Huitt GA, et al.. Mycobacterium terrae
: case reports, literature review, and in vitro antibiotic susceptibility testing. Clin Infect Dis
. 2000; 30: 444–453.
7. Chen HW, Lai CC, Tan CK. Arthritis caused by Mycobacterium terrae
in a patient with rheumatoid arthritis. Int J Infect Dis
. 2009; 13: 145–147.
8. Richmond L, Cummings MM. An evaluation of methods of testing the virulence of acid-fast bacilli. Am Rev Tuberc
. 1950; 62: 632–637.
9. Kozin SH, Bishop A. Atypical Mycobacterium
infections of the upper extremity. J Hand Surg
. 1994; 19A: 480–487.
10. Dijkmans BAC, Mouton RP, Macfarlane JD, et al.. Bacterial arthritis caused by Mycobacterium terrae
. 1981; 9: 204–207.
11. Chan TH, Ng KC, Ho A, et al.. Urinary tract infection caused by Mycobacterium terrae
complex. Tubercle Lung Dis
. 1996; 77: 555–557.
12. Fodero J, Chung KC, Ognenovski VM. Flexor tenosynovitis in the hand caused by Mycobacterium terrae
. Ann Plast Surg
. 1999; 42: 330–332.
13. Elhassan BT, Wynn SW, Gonzalez MH. Atypical infections of the hand. J Am Soc Surg Hand
. 2004; 4: 42–49.
14. Deenstra W. Synovial hand infection from Mycobacterium terrae
. J Hand Surg
. 1988; 13: 335–336.
15. Petrini B, Svartengren G, Hoffner SE, et al.. Tenosynovitis of the hand caused by Mycobacterium terrae
. Eur J Clin Microbial Infect Dis
. 1989; 8: 722–724.
16. Karthigasu KT, Spagnolo DV, Gow BL. Mycobacterium terrae
. 1990; 22: 106–107.
17. Zenone T, Boibieux A, Tigaud S, et al.. Non-tuberculous mycobacterial tenosynovitis: a review. Scand J Infect Dis
. 1999; 31: 221–228.
18. Yang DH, Chang WC, Cheng MF, et al.. Peripheral arthritis caused by Mycobacterium avium-intracellulare
in a patient with ankylosing spondylitis. J Clin Rheumatol
. 2009; 15: 323–324.
19. Bracewell C, Wright D. A 10-year history of Mycobacterium malmoense
septic arthritis of the wrist. J Clin Rheumatol
. 2009; 15: 433–434.
20. Bettiker RL, Axelrod PI, Fekete T, et al.. Delayed recognition of a pseudo-outbreak of Mycobacterium terrae
. Am J Infect Control
. 2006; 34: 343–347.
© 2012 Lippincott Williams & Wilkins, Inc.