Lymph node disease is the most common manifestation of NTM infection in children and has been extensively described (Schaad et al., 1979; Colville, 1993; Hazra et al., 1999; Saggese et al., 2003; Panesar et al., 2003; Mandell et al., 2003; Haverkamp et al., 2004). There are many retrospective case series making comparisons of different surgical treatments in children with significant lymphadenopathy. However, evidence on appropriate drug treatment is confined to small case series. Most NTM isolated from lymph nodes are M. avium, with M. malmoense increasingly identified in the UK (Colville, 1993).
Typically NTM infection occurs in young children, between 1 and 8 years, as unilateral, chronic, cervical or submandibular lymphadenopathy. Affected individuals are generally well, apyrexial, with no systemic upset, no local node tenderness and do not respond to antimicrobial agents that target streptococci and staphylococci (Schaad et al., 1979; Panesar et al., 2003). The lymph node(s) is initially firm and mobile, may appear suddenly or gradually over a few weeks, and with time there is often progression to overlying skin involvement with a reddish purple color.
Occasionally NTM-associated lymphadenopathy may become very large and disfiguring. If left untreated there will eventually be spontaneous resolution of NTM infection over months or years, though most, before resolving, will progress to abscess formation, discharge and sinus formation before healing with scarring (Mandell et al., 2003; Schaad et al., 1979).
Assessment should include investigation for other possible differential diagnoses. Although an underlying immuncomprimising illness should be considered, routine immune studies are not generally indicated. An acute bacterial infection is usually clinically apparent but a normal full blood count and C-reactive protein may be reassuring. Potential viral etiologies such as Epstein-Barr virus, adenovirus, cytomegalovirus and mumps should be evaluated. Bartonella henselae and Toxo-plasma gondii should be considered particularly in those children exposed to cats. In patients from endemic areas or with a contact history, TB should be excluded. A chest radiograph is usually normal in both TB or NTM lymphadenitis. Almost all healthy children with TB adenitis will have a positive tuberculin reaction, however, up to 30% with NTM lymphadenitis will have 10 mm+ of induration with PPD. Diagnosis maybe aided by comparing tuberculin with NTM skin tests (Daley and Isaacs, 1999; Saggese et al., 2003) but these are not standardized nor widely commercially available. Novel in vitro T cell stimulation tests (ELISPOT or whole blood assays) may offer better discrimination in the future (Rolinck-Werninghaus et al.,
CT and MRI scanning of patients with NTM lymphadenitis may show ring-enhancing lesions with minimal inflammatory stranding of the subcutaneous fat (Nadel et al., 1996; Hazra et al., 1999). Although imaging does not reliably distinguish NTM from TB infection, these may be valuable in defining the extent of disease and planning surgical intervention (see below).
Diagnosis should initially be based on clinical suspicion but there is frequently a delay. In many cases a diagnosis may only be arrived at after suggestive histology (caseating granulomas with or without acid-alcohol fast bacilli) or definitive culture of NTM following surgical intervention. Where tissue for histology and culture is acquired by biopsy or incision and drainage, there is often poor wound healing, continuing discharge and complicated by sinus formation (Schaad et al., 1979; Mandell et al., 2003; Panesar et al., 2003). Biopsy or incision and drainage should therefore be avoided whenever possible and are usually only undertaken where there are difficulties in initial diagnosis. There is published experience with fine needle aspiration not producing sinus formation but this remains controversial (Alessi and Dudley, 1988; Mandell et al., 2003). Many authorities therefore advocate complete surgical excision which is associated with cure rates of 81 to 92% (Saggese et al., 2003; Panesar et al., 2003). Surgery, however, is not without risks and may not be feasible in extensive disease or where there is facial nerve or parotid gland involvement. It has been argued that the evidence in favor of surgical treatments is biased by an over representation of the most severe cases (Haverkamp et al.,
2004). Medical approaches to treatment using macrolides (azithromycin or clarithromycin) either alone or in combination with rifabutin ± ethambutol have been evaluated in some centers. However, antimicrobial therapy alone, particularly monotherapy, has not been entirely successful (Panesar et al., 2003; Hogan et al., 2005;
Hazra et al., 1999; Haverkamp et al., 2004; Saggese et al., 2003) and may be associated with significant drug side effects. Because the natural history of NTM mycobacterial lymphadenopathy is that of resolution, comparative trials require careful design. A randomized, open-label cohort trial, comparing surgery versus 3 months medical treatment has recently been completed in the Netherlands and will be reported in the near future (Jaap T. van Dissel, personal communication). It is important to recognize that whatever the initial therapeutic approach, scarring, sinus tract formation and mycobacterial reactivation of residual infection may occur.
Was this article helpful?