Tenosynovitis caused by Mycobacterium marseillense, initially identified as Mycobacterium avium complex using AccuProbe and COBAS TaqMan

Background Mycobacterium marseillense is a new species of the Mycobacterium avium complex. There has been only a few human infections caused by M. marseillense worldwide. Case presentation We report a case of tenosynovitis caused by M. marseillense in an immunocompetent adult in Japan. The isolate was initially identified as M. intracellulare using commercial real time polymerase chain reaction assays and later identified as M. marseillense with sequencing of the the rpoB and hsp65 regions, and matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS). Conclusions This is the first case reporting on M. marseillense generating a positive result with commercial real time PCR assays targeting MAC. Human infections associated by M. marseillense might be underreported due to similarities with Mycobacterium intracellulare. To accurately identify M. marseillese, MALDI-TOF MS might provide a rapid and reliable method.


Background
The Mycobacterium avium complex (MAC) is the most common non-tuberculosis mycobacteria (NTM) causing human infections; these include chronic pulmonary infections in adults and lymphadenitis in children. M. avium and M. intracellulare are the best-known representative species of the MAC. However, with advance in molecular analyses, multiple species have been recognized as members of the MAC, resulting in frequent emendation in the taxonomy of this species complex [1]. Although there is no widely accepted definition with regard to what constitutes the MAC, a recent review proposed a definition of the MAC based on phylogenetic analyses that includes 12 valid published species [2]. Mycobacterium marseillense is among these [1]; however, little is known about its microbiological and clinical features. We here report a case of tenosynovitis with M. marseillense that was initially identified as M. intracellulare using commercial real time polymerase chain reaction (PCR) assays.

Case presentation
In March 2018, an 85-year-old man with no medical history was admitted after a 6-month history of worsening pain and swelling in the right wrist. Magnetic resonance imaging showed a large region of tenosynovitis from the Open Access *Correspondence: kokamoto-tky@umin.ac.jp 2 Department of Infectious Diseases, The University of Tokyo Hospital, 7-3-1 Hongo, Bunkyo-ku, 113-8655 Tokyo, Japan Full list of author information is available at the end of the article distal forearm to the palm, and a biopsy of the synovial membrane was performed. Histological testing showed granulomatous inflammation. After nine weeks of culture on Ogawa medium (Kyokuto Pharmaceutical Industrial CO., Japan), smooth non-pigmented colonies were observed. The colony was positive for Ziehl Neelsen stain, suggesting acid fast bacilli. Initial testing with AccuProbe (Hologic, Marlborough, MA) was positive for MAC, with a value of 197,546 relative light units (RLU). COBAS TaqMan MTB/MAI (Roche Diagnostics, Switzerland), another real-time PCR test, was positive for M. intracellulare as well. However, further identification of the isolate using matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS; MALDI Biotyper Version 2.0, Bruker Daltonics, US) suggested that the isolate was M. marseillense, with a score of 2.023 indicating a probable species identification. Because of this discordance, the rpoB and hsp65 regions were sequenced [3,4]. In the Basic Logical Alignment Search Tool (BLAST) analysis, both sequences showed a 99.7% (E value, 0.0) and 100.0% (E value, 0.0) alignment respectively with those of Mycobacterium marseillense (GenBank accession number CP023147.1).
Drug susceptibility testing was performed by a broth microdilution test, BrothMIC NTM (Kyokuto Pharmaceutical Industrial Co., Japan). Based on the breakpoints proposed by the Clinical and Laboratory Standards Institute [5], the isolate was sensitive to clarithromycin and amikacin. Subsequently, clarithromycin, ethambutol and rifampicin were initiated, and symptoms subsided gradually over the following month. Rifampicin was discontinued due to neutropenia after 2 months. Unfortunately, the patient died from an unrelated cause six months after the initiation of treatment.

Discussion
Here we describe a case of tenosynovitis of the wrist by M. marseillense in an immunocompetent adult. Thus far, five human infections of M. marseillense have been reported from Italy, China and the US (Tables 1, 2) [6][7][8][9][10].
Accurate diagnosis of the MAC to the species level is challenging as they are very similar in biochemical features [2]. In addition, they are closely related genetically, with 98-99% similarity in the 16S rRNA gene [2]. Thus, a combination of ribosomal and/or housekeeping genes has been used for identification of novel strains and species. In the reported cases, the identification was mostly made by sequencing multiple regions, such as 16S rRNA, internal transcribed spacer-1 region (ITS), hsp65, and rpoB. In our case, sequencing of the rpoB and hsp65 region and MALDI-TOF MS successfully identified the strain as M. marseillense. However, sequencing is not readily available in clinical settings; this poses a challenge to clinicians.
Commercial probes for the clinical important mycobacteria are widely used in clinical settings. Among these, AccuProbe is an FDA-approved polymerase chain reaction (PCR) test used to identify several clinically important mycobacteria, including M. tuberculosis complex, M. avium, M. intracellulare, and M. kansasii [11]. The identification is based on the hybridization of specific DNA probes to the target 16 S rRNA. By using chemiluminescence, the DNA-rRNA hybrid molecule is detected and measured in relative light units (RLU). A reliable species identification for NTM requires a cutoff value over 30,000 RLU and may require even higher values, above 80,000 for MAC [11]. Nonetheless, cross reactions with other mycobacteria including M. arosiense, M. chimaera, M. nebraskense, M. saskatchewanense, and M. colombiense have been reported [11,12]. To our knowledge, this is the first report of the positive reaction for MAC with AccuProbe and COBAS TaqMan by M. marseillense in the literature. Of note, one study reported the misidentification of M. marseillense as M. intracellulare, using a different commercial identification kit (Genotype Mycobacterium CM/AS assay, Hain Lifescience GmbH, Germany), leading to a delay in accurate diagnosis [7].
MALDI-TOF MS may be of great help for mycobacterial identification in clinical microbiology laboratories. In a study of 125 isolates from 27 different NTM species, identification using MALDI-TOF MS with a cut-off score of 1.7 had a 94.4 % (118/125) agreement between 16 S rRNA and hsp65 sequencing to the species level [13]. The seven isolates that could not be identified using MALDI-TOF MS were M. massiliense, (n = 4; not included in the database) and M. gordonae (n = 3; with no peak or reliable result) [13]. M. massilense was not included at the time of the study but has later been added in the most recent database (MALDI Biotyper Version 6). Other studies have also reported similar results reporting a 92.0 to 97.6 % agreement between MALDI-TOF MS and either 16S rRNA, rpoB or hsp65 sequencing [14,15].
In conclusion, we report on a case of tenosynovitis of the wrist by M. marseillense in an immunocompetent man. Given the limited availability of sequencing in clinical settings, cases of M. marseillense infection might have been misidentified and, therefore, underreported. To understand the epidemiology of M. marseillense and its role in human infections, accurate identification is crucial, and MALDI-TOF MS might provide a rapid and reliable identification when sequencing specific regions is not readily available.