An eight-month-old child with cervical adenitis

FINANCIAL SUPPORT AND DISCLOSURES: No financial support was obtained for this article. All authors declare no conflicts of interest. TK is supported in part by a Career Award in the Biomedical Sciences from the Burroughs Wellcome Fund and by a CIHR Training Grant from the Canadian Child Health Clinician Scientist Program, in partnership with Sick Kids Foundation, Child & Family Research Institute (British Columbia), Women & Children’s Health Research Institute (Alberta) and the Manitoba Institute of Child Health. 
 
 
 
AUTHOR CONTRIBUTIONS: MCMM drafted the manuscript. JC, PT, JH and TK critically revised the manuscript for content. All authors saw and approved the final manuscript.


Diagnosis
Microscopic examination of surgical samples detected 1+ acid-fast bacilli. Gram and fungal stains and bacterial and fungal cultures were negative. Biopsy material underwent standard molecular testing for MTB using the Amplified M tuberculosis Direct assay (Gen-Probe, USA) and Mycobacterium avium complex (MAC) testing with the appropriate AccuProbe assay (Gen-Probe, USA). Both tests were negative. For organism identification, polymerase chain reaction analysis of the heat shock protein hsp65 was performed, yielding a 401 bp amplicon that was sequenced (1). The hsp65 sequence identified the organism (10 days after sample collection) as Mycobacterium haemophilum. After 40 days incubation, cultures became positive in the liquid Mycobacteria Growth Indicator Tube (BD, USA) mycobacterial detection system. Molecular testing confirmed the presence of M haemophilum.

DisCussion
Nontuberculous mycobacterial (NTM) adenitis in children has historically been diagnosed using a combination of clinical suspicion, TST and mycobacterial culture (which generally takes weeks to provide results) (2). Treatment is therefore empirical and based on TST and clinical presentation, in the context of possible exposure to MTB. Treatment of NTM adenitis usually involves surgical excision of lesions, unless risk of injury to adjacent structures is high, in which case antibiotic treatment with possible later surgical intervention is the norm (3). Treatment of MTB adenitis involves antimycobacterials.
MAC currently accounts for up to 80% of NTM adenitis in children (3). Studies involving MAC as the predominant isolate have suggested that a TST induration of 5 mm to 9 mm is consistent with NTM adenitis, while indurations >15 mm suggest MTB infection (4). Recently, M haemophilum, which is prevalent in the environment and linked with water exposure (5), has been described in Israel and Europe as being responsible for a significant proportion of NTM adenitis in children (5,6). This may be related to recent improvements in culturing techniques allowing detection of this organism (7). ©2012 Pulsus Group Inc. All rights reserved Unlike MAC, M haemophilum infection is associated with greater induration on TST (6), making it difficult to differentiate NTM and MTB infection. QFT-GIT and other region of difference-1 based IGRA tests are more specific for MTB versus TST (8). IGRAs do not cross-react with most NTM, including MAC and M haemophilum and, thus, make useful adjuncts in the evaluation of mycobacterial adenitis (9). However, IGRA tests may have reduced sensitivity in very young children and may not be sufficient on their own to exclude MTB infection (10).
The differential diagnosis of cervical adenitis should always include NTM and MTB as possible causative agents. Sampling of affected lymph nodes for culture and molecular diagnostic testing is crucial. Where suspicion of MTB exists, molecular testing may yield a diagnosis weeks before traditional culture methods, potentially saving children toxicity from antimycobacterial drugs. Culturing of samples is likewise essential for obtaining drug sensitivities, particularly in MTB infection. Unfortunately, the only laboratories routinely using molecular typing of NTM adenitis in Canada are the British Columbia Centre for Disease Control in Vancouver, British Columbia, and the National Microbiology Labs in Winnipeg, Manitoba.
The present patient was at high risk for tuberculosis adenitis (based on history, clinical picture, surgical sample positivity for acid-fast bacilli and TST results), and was initiated on quadruple tuberculosis therapy (ie, isoniazid, rifampin, ethambutol and pyrazinamide).
Polymerase chain reaction and sequencing results became available one week later, identifying the acid-fast bacilli as M haemophilum. Medications were changed to ciprofloxacin and rifampin 30 days before mycobacterial cultures identified the organism. This combination provided coverage of M haemophilum based on the literature (11), and tuberculosis prophylaxis in case of exposure. Due to the location, size and complexity of the masses, antibiotic treatment was chosen to reduce mass size before complete surgical excision, to be performed three months later. The temple mass resolved with antibiotic treatment alone. The patient is currently doing well.