Misindentification of Mycobacterium kumamotonense as M. tuberculosis

To the Editor: Because of slow growth of mycobacteria, use of rapid tests to identify them is strongly recommended; rapid tests are widely used as an advanced diagnostic tool in clinical laboratories (1,2). These tests are particularly useful for diagnosing extrapulmonary mycobacterioses and identifying unusual mycobacteria as etiologic agents (3). Commercial probes are frequently used for rapid and specific identification of mycobacteria, especially Mycobacterium tuberculosis complex. However, cross-reactivity of DNA probes between mycobacterial species could result in incorrect diagnosis and treatment of patients (4,5). Misidentification could be a problem if a newly described species, such as M. kumamotonense (6), were an etiologic agent of a disease. 
 
In July 2006, we obtained a fine-needle, puncture aspiration biopsy specimen from a cervical lymph node of a 30-year-old man at Doce de Octubre Hospital (Madrid, Spain). The patient was a recent immigrant from Paraguay and was HIV positive (C2 stage of infection). A biopsy specimen from a cervical lymph node showed necrotizing granulomatous lymphadenopathy. A computed tomographic scan showed cervico-thoraco-abdominal, multiple cervical, supraclavicular, axillar, paratracheal, and mediastinal lymphadenopathies. The patient had a CD4 cell count of 219 cells/mm3 and an HIV viral load of 197,181 copies/mL. 
 
The aspiration sample was positive for acid-fast bacilli by fluorescent staining. The clinical isolate (designated 1369) obtained from the aspirate sample was grown in liquid media (MGIT Diagnostic Kit; Becton Dickinson Diagnostics, Sparks, MD, USA) and identified as M. tuberculosis complex by using the AccuProbe System (bioMerieux, Marcy l’Etoile, France). 
 
A diagnosis of lymphoid tuberculosis was made, and the patient was treated with isoniazid, rifampin, ethambutol, and pyrazinamide. After 1 month, rifampin was withdrawn because of a cutaneous exanthem. Three months later, the clinical status of the patient had improved, fever had disappeared, and sizes of cervical and axillary lymph nodes had decreased. Treatment with tenofovir, emtricitabine, and lopinavir/ritonavir was started. Two weeks later, an immune reconstitution syndrome and adenopathies developed, but these resolved in 1 month. 
 
Five months after treatment was started, susceptibility testing in a reference laboratory showed that isolate 1369 was M. kumamotonense. The isolate showed 100% identity with the 16S rRNA gene sequence of M. kumamotonense (GenBank accession no. {"type":"entrez-nucleotide","attrs":{"text":"AB239925","term_id":"110796721","term_text":"AB239925"}}AB239925). Results of PCR restriction analysis of heat shock protein 65 gene (7) (http://app.chuv.ch/prasite/index.html) were consistent with those for M. kumamotonense. The isolate was susceptible to ethambutol, rifampin, cycloserine, and ethionamide and resistant to isoniazid, streptomycin, pyrazinamide, and kanamycin. 
 
Because of the improvement in the clinical status of the patient, treatment continued without modification for 18 months. At this time, his CD4 cell count was 488 cells/mm3 and his HIV viral load was <50 copies/mL. In July 2009, the patient was asymptomatic and had a CD4 cell count of 631 cells/mm3 and an HIV viral load <50 copies/mL. 
 
To confirm misidentification of M. kumamotonense as a member of the M. tuberculosis complex, other commercial probes were tested. Isolate 1369 was also misidentified as M. tuberculosis complex by Inno-LIPA v2 (Innogenetics, Ghent, Belgium). The isolate was identified as Mycobacterium sp. by Geno-Type (Hain Lifescience, Nehren, Germany). The 3 commercial probes we used had different genome region specificities, all in the mycobacterial ribosomal operon. The AccuProbe System was specific for 16S rDNA, Inno-LIPA v2 was specific for internal transcribed spacer 1, and Geno-Type was specific for 23S rDNA. Only Geno-Type did not show cross-reactivity between M. tuberculosis complex and M. kumamotonense. The clinical isolate was identified as M. kumamotonense, a new, slow-growing mycobacterium that was first isolated from an immunocompetent patient in Japan (6). We showed that this species caused extrapulmonary disease in an HIV-positive patient. 
 
Misidentification of M. kumamotonense as M. tuberculosis complex by commercial DNA probes has serious clinical implications. Once a patient is given a diagnosis of tuberculosis, he or she will be treated with specific drugs for a long period and be prone to adverse side effects. Furthermore, M. kumamotonense is resistant to many drugs used during typical treatment. After a diagnosis of tuberculosis, patient contacts need to be investigated to identify new cases. Emerging mycobacterial pathogens, such as M. kumamotonense, may also cause pulmonary and extrapulmonary infections that are also caused by other members of this genus and could be misidentified as M. tuberculosis.

To the Editor: Because of slow growth of mycobacteria, use of rapid tests to identify them is strongly recommended; rapid tests are widely used as an advanced diagnostic tool in clinical laboratories (1,2). These tests are particularly useful for diagnosing extrapulmonary mycobacterioses and identifying unusual mycobacteria as etiologic agents (3). Commercial probes are frequently used for rapid and specifi c identifi cation of mycobacteria, especially Mycobacterium tuberculosis complex. However, crossreactivity of DNA probes between mycobacterial species could result in incorrect diagnosis and treatment of patients (4,5). Misidentifi cation could be a problem if a newly described species, such as M. kumamotonense (6), were an etiologic agent of a disease.
In July 2006, we obtained a fi neneedle, puncture aspiration biopsy specimen from a cervical lymph node of a 30-year-old man at Doce de Octubre Hospital (Madrid, Spain). The patient was a recent immigrant from Paraguay and was HIV positive (C2 stage of infection). A biopsy specimen from a cervical lymph node showed necrotizing granulomatous lymphadenopathy. A computed tomographic scan showed cervico-thoraco-abdominal, multiple cervical, supraclavicular, axillar, paratracheal, and mediastinal lymphadenopathies. The patient had a CD4 cell count of 219 cells/mm 3 and an HIV viral load of 197,181 copies/mL.
The aspiration sample was positive for acid-fast bacilli by fl uorescent staining. The clinical isolate (designated 1369) obtained from the aspirate sample was grown in liquid media (MGIT Diagnostic Kit; Becton Dickinson Diagnostics, Sparks, MD, USA) and identifi ed as M. tuberculosis complex by using the AccuProbe System (bioMérieux, Marcy l'Etoile, France).
A diagnosis of lymphoid tuberculosis was made, and the patient was treated with isoniazid, rifampin, ethambutol, and pyrazinamide. After 1 month, rifampin was withdrawn because of a cutaneous exanthem. Three months later, the clinical status of the patient had improved, fever had disappeared, and sizes of cervical and axillary lymph nodes had decreased. Treatment with tenofovir, emtricitabine, and lopinavir/ritonavir was started. Two weeks later, an immune reconstitution syndrome and adenopathies developed, but these resolved in 1 month.
Five months after treatment was started, susceptibility testing in a reference laboratory showed that isolate 1369 was M. kumamotonense. The isolate showed 100% identity with the 16S rRNA gene sequence of M. kumamotonense (GenBank accession no. AB239925). Results of PCR restriction analysis of heat shock protein 65 gene (7) (http://app.chuv.ch/prasite/index.html) were consistent with those for M. kumamotonense. The isolate was susceptible to ethambutol, rifampin, cycloserine, and ethionamide and resistant to isoniazid, streptomycin, pyrazinamide, and kanamycin.
Because of the improvement in the clinical status of the patient, treatment continued without modifi cation for 18 months. At this time, his CD4 cell count was 488 cells/mm 3 and his HIV viral load was <50 copies/mL. In July 2009, the patient was asymptomatic and had a CD4 cell count of 631 cells/mm 3 and an HIV viral load <50 copies/mL.
To confi rm misidentifi cation of M. kumamotonense as a member of the M. tuberculosis complex, other commercial probes were tested. Isolate 1369 was also misidentifi ed as M. tuberculosis complex by Inno-LIPA v2 (Innogenetics, Ghent, Belgium). The isolate was identifi ed as Mycobacterium sp. by Geno-Type (Hain Lifescience, Nehren, Germany). The 3 commercial probes we used had different genome region specifi cities, all in the mycobacterial ribosomal operon. The Ac-cuProbe System was specifi c for 16S rDNA, Inno-LIPA v2 was specifi c for internal transcribed spacer 1, and Geno-Type was specifi c for 23S rDNA. Only Geno-Type did not show crossreactivity between M. tuberculosis complex and M. kumamotonense. The clinical isolate was identifi ed as M. kumamotonense, a new, slow-growing mycobacterium that was fi rst isolated from an immunocompetent patient in Japan (6). We showed that this species caused extrapulmonary disease in an HIV-positive patient.
Misidentifi cation of M. kumamotonense as M. tuberculosis complex by commercial DNA probes has serious clinical implications. Once a patient is given a diagnosis of tuberculosis, he or she will be treated with specifi c drugs for a long period and be prone to adverse side effects. Furthermore, M. kumamotonense is resistant to many drugs used during typical treatment. After a diagnosis of tuberculosis, patient contacts need to be investigated to identify new cases. Emerging mycobacterial pathogens, such as M. kumamotonense, may also cause pulmonary and extrapulmonary infections that are also caused by other members of this genus and could be misidentifi ed as M. tuberculosis. To the Editor: Mycobacterium fortuitum complex members are rapidly growing mycobacteria found in water and soil (1). These opportunistic pathogens are responsible for posttraumatic skin and soft tissue infections. They also account for 60%-80% of postsurgical wound infections caused by rapidly growing mycobacteria (2), particularly after breast surgery (with or without prosthetic implants) (3). M. conceptionense, an emerging member of the M. fortuitum complex, was initially described in a case of osteomyelitis that occurred after an open fracture of the tibia (4). We report a case of M. conceptionense infection that occurred after breast surgery.
A woman 58 years of age had a left mastectomy with lymph node dissection and chemotherapy for breast carcinoma in March 2004. Three years later, she underwent breast reconstruction that used a cutaneomuscular latissimus dorsi fl ap with a prosthetic implant. Immediately after surgery, a fever (39°C) developed, but 3 blood cultures remained sterile. No treatment was administered, and she became afebrile within 3 days.
At day 15 after surgery, a serous discharge appeared in the tip of the skin fl ap. By day 21, the patient was again febrile, and the wound discharge was swabbed for analysis. On day 27, she underwent surgical revision with ablation of the breast implant, drainage, and sample collection. The leukocyte count was normal. However, the C-reactive protein level was 99 mg/L, and the erythrocyte sedimentation rate was 111 mm (fi rst hour). Treatment with intravenous amoxicillin/clavulanic acid was started. Although the biologic parameters normalized, the serous discharge continued. Micro-scopic examination of specimens from days 21 and 27 yielded no bacteria in Gram-and Ziehl-Nielsen-stained pus specimens, and standard bacteriologic cultures remained sterile. M. conceptionense, identifi ed by partial rpoB gene sequencing (100% identity with GenBank accession no. AY859695.1) (4), grew in both specimens after 8 days of incubation at 37°C under a 5% CO 2 atmosphere in Coletsos medium (bioMérieux, La Balme-les-Grottes, France). By the Etest method (4), both isolates were susceptible to several antimicrobial drugs, including clarithromycin, amikacin, ciprofl oxacin, and doxycycline. The patient was treated with ciprofl oxacin, azythromycin, and amikacin for 3 weeks, followed by ciprofl oxacin and azythromycin for 4 weeks.
At patient's relapse 3 months later, M. conceptionense exhibiting identical antimicrobial drug susceptibility pattern was again isolated from the wound fl uid. The patient was then treated with ciprofl oxacin, azythromycin, and doxycycline for 6 months; subsequently, doxycycline alone was given for a total of 18 months. Results from the 2-month follow-up examination were unremarkable.
M. conceptionense was unambiguously identifi ed by partial rpoB gene sequencing, a fi rst-line tool for accurate identifi cation of nontuberculous mycobacteria (5). A pathogenic role for M. conceptionense was supported by 1) its repetitive isolation from the wound;