Introduction
Non-tuberculous mycobacterial cervicofacial lymphadenitis infection is a rare disease but a common cause of chronic cervical lymphadenopathy in the immunocompetent child. Presentation is classical with persistent cervical lymphadenopathy, usually unilateral, that does not respond to antibiotics. As the disease progresses, the skin changes to characteristic cherry red and the lymph nodes liquefy, resulting in a fluctuating swelling that will often form a sinus that will drain for a prolonged period, with eventual resolution and scarring.
Non-tuberculous mycobacteria are ubiquitous in the environment, with over 125 species having been identified.Reference Spinelli, Mannelli, Arcuri, Venturini, Chiappini and Galli1 The incidence worldwide is quoted as between 0.8 and 3.1 cases per 100 000, with geographical variability,Reference Moe, Rajan, Caltharp and Abramowicz2, Reference Loizos, Soteriades, Pieridou and Koliou3 and is widely acknowledged to be on the increase. This has been demonstrated in the UKReference Meoli, Deolmi, Iannarella and Esposito4 with a rise from 0.9 to 2.9 cases per 100 000 between 1995 and 2006. The most frequent causative subspecies identified is Mycobacterium avium.Reference Spinelli, Mannelli, Arcuri, Venturini, Chiappini and Galli1–Reference Ford, Silcock and Holland9
Management of this condition is controversial, with a limited evidence base and no universally agreed treatment. Proposed treatment options broadly include surgical, medical (often combined) or a watch-and-wait approach. We looked to survey UK paediatric tertiary centres to gain an understanding of the current practice in the management of these patients.
Materials and methods
An 18-question online survey was sent to all 21 UK paediatric tertiary centres. The questions asked aimed to evaluate each unit’s case load, involvement of cases, other specialty contributions, investigation and treatment choices as well as follow-up decisions. The survey was sent to a nominated consultant in each paediatric ENT department. The questions were accompanied by multiple-choice answers to be selected, with 11 of the questions incorporating additional free text to provide further information where deemed appropriate. The results were then received and analysed.
Results and analysis
We received 19 responses from the 21 centres surveyed. The majority of centres (58 per cent) see at least 4 cases a year, with 89 per cent of centres receiving referrals from neighbouring ENT departments for the management of this condition. Otolaryngology is almost always involved in the management of these patients (95 per cent), often as the lead specialty (58 per cent), alongside multiple other specialties, including infectious diseases (79 per cent), general paediatrics (47 per cent) and microbiology (32 per cent).
The survey revealed that the most common method (89 per cent) of diagnosis was based on ‘clinical suspicion alone, from history and examination’. Other investigations are variably used, including, but not exhaustive of, ‘microbiology/ histopathology from excisional biopsy or fistulous drainage’ (37 per cent), ‘AFB\ stains’ (37 per cent), ‘lymphadenopathy serology screen’ (37 per cent), ‘tuberculin skin testing, in a child without risk factors for M. tuberculosis and absence of previous BCG vaccination’ (37 per cent). In addition, 67 per cent of centres selected their investigation choice to be ‘variable depending on presentation’ (see Figure 1).
Figure 1. Investigation and diagnosis of cases of suspected non-tuberculous mycobacterial cervicofacial lymphadenitis by UK tertiary paediatric ENT centres. BCG = bacilli Calmette-Guérin (vaccine); AFB = Acid-Fast bacillus; FBC = Full Blood Count; CRP = C-reactive protein; EBV = Epstein–Barr virus.
It was found that 63 per cent of centres routinely requested an ultrasound and 11 per cent requested a chest X-ray, while 32 per cent of responders would not request any routine imaging.
Management of non-tuberculous mycobacterial cervicofacial lymphadenitis was assessed by presenting four common clinical scenarios, ranging from a milder presentation to a more extensive or problematic one, with the responders asked how they would manage each scenario. The scenarios and results are summarised in Figure 2.
Figure 2. Treatment selection dependent on different clinical presentations by UK tertiary paediatric ENT centres for non-tuberculous mycobacterial cervicofacial lymphadenitis.
A surgical procedure was the most common treatment choice on the whole, with a less extensive procedure subsequently opted for when more adverse features were present, for example close proximity to the facial nerve and/or skin involvement. Non-surgical treatment was also common, chosen as the preferred management option (16–47 per cent) depending on the clinical presentation. Observation was shown to be a common treatment choice, with rates between the clinical scenarios ranging from 16 to 32 per cent. Antimycobacterial monotherapy was selected by between 0 and 21 per cent of centres. Where ‘other’ was selected, commonly responders would comment that their treatment of choice may change related to variations and unique circumstances between individual cases.
Where medical treatment was used, the majority of respondents (56 per cent) were unsure of the antimycobacterial regimen used in their respective hospitals because this decision was made by other specialities. Where known, the most common choice (17 per cent) was dual therapy with rifampicin and clarithromycin. The duration of antituberculous treatment also varied. Where known, most centres typically gave 3 months of treatment (39 per cent). Others provided 6 months of treatment (17 per cent) or until clinical resolution (17 per cent).
Follow-up protocols were also identified. If observation alone was chosen, most commonly monthly follow up was undertaken (58 per cent), followed by 3 monthly (21 per cent) and 6 monthly (5 per cent) follow up. Some centres advocated patient-initiated follow up. If surgery was undertaken, regular follow up until the wound had healed was the most common choice (58 per cent), with some respondents only scheduling one follow up to check the wound. Where antituberculous therapy had been completed, most respondents (42 per cent) would still arrange regular interval follow up until the wound had healed. If this management protocol was followed, 26 per cent of respondents stated that follow up would be undertaken by a different specialty.
It was found that 68 per cent of the respondents said they incorporated the views of the patient’s parents when deciding on treatment. Furthermore, 26 per cent of centres responded that their management of this condition had been altered by adverse outcomes or complications in their own practice or in their department. Where comments were shared, universally a change from a surgical approach to considering and using more conservative surgery or non-surgical choices appeared to be the norm.
Only 21 per cent of the centres that responded had local management guidelines and 79 per cent of them had the opinion that management of this condition would be improved by written national guidelines.
Discussion
The objective of this survey was to gain an up-to-date understanding of the way in which non-tuberculous mycobacterial cervicofacial lymphadenitis is currently managed in the UK. The management of non-tuberculous mycobacterial cervicofacial lymphadenitis is controversial and its optimum management is often debated both locally and internationally. There is a lack of high-level evidence, which likely contributes to this difference in opinion. The lack of evidence can be partly explained by the challenge in obtaining such evidence.
There is a wide variation of the heterogeneity of patients, with variable anatomical locations and extent of disease. Thus, there are potential subsequent differences in the risk profile associated with the investigation and management of the condition for each individual based on these factors.
There are only three randomised control trialsReference Lindeboom, Kuijper, van Coppenrat ESB and Lindeboom10–Reference Lindeboom12 on the treatment of this condition, each with their own limitations. In addition, there is one systematic review and analysis from 2015 by Zimmerman et al.Reference Zimmerman, Tebruegge, Curtis and Ritz13 This paper provided data from nearly 2000 patients and forms the largest and most robust body of evidence in the management of non-tuberculous mycobacterial cervicofacial lymphadenitis.
A demonstration of the lack of evidence to inform a definitive management plan was highlighted when the International Pediatric Otolaryngology GroupReference Roy, Balakrishnan, Boudewyns, Cheng, Chun and Daniel14 published their recommendations in 2023. Although guidance was provided, complete consensus was not achieved. Importantly, this guidance does recognise the role of other medical specialities, namely infectious diseases and paediatrics, encouraging a multidisciplinary approach to management. This approach is demonstrated in our survey results. Most recently, further guidance has been provided by the British Association of Paediatric Otolaryngology15 for the management of this condition.
For a definitive diagnosis of non-tuberculous mycobacterial cervicofacial lymphadenitis, a positive mycobacterial culture or polymerase chain reaction test is required. This is desirable, but there is a classical presentation where a probable diagnosis can be made. This is of an immunocompetent, systemically well young child with a chronic unilateral, single, above-the-hyoid neck lesionReference Spinelli, Mannelli, Arcuri, Venturini, Chiappini and Galli1, Reference Neven, Van der Linden, Hainaut and Schmitz16 with a red or purple discolouration. Our survey shows that clinicians are happy to make the diagnosis based on this classical presentation. However, other microbiology and virology tests are considered and used in the UK. These investigations have often been performed by primary or secondary care providers, prior to presentation to tertiary centres, in an effort to confirm diagnosis as well as to rule out other differential diagnosis.
There is a common concern that diagnostic fine needle aspiration may be a risk for fistula formation, with limited evidenceReference Zimmerman, Tebruegge, Curtis and Ritz13 to support this complication. However, this, along with other invasive investigatory and therapeutic choices (e.g. incision and drainage, lymph node excision), is difficult to prove owing to the natural progression of disease, which can demonstrate development of a fistula in the absence of any of the above.
Ultrasound is routinely used to aid diagnosis. This is not surprising given its favourable attributes of being easily accessible, non-invasive and able to demonstrate characteristic findings to support a diagnosis of non-tuberculous mycobacterial cervicofacial lymphadenitis. The use of other imaging modalities is not favoured routinely, which matches the available literature, with an additional magnetic resonance imaging scan often reserved for surgical planning in more extensive cases or where there is concern of deeper involvement.
The definitive management of this condition generally falls into three categories: surgical excision, medical management with antimycobacterial therapy and observation. In the UK, the most common practice appears to be surgical excision, historically the preferred treatment choice in the literature. However, a more limited procedure is often preferred in the setting involving more risk, such as skin breakdown or proximity to a branch of the facial nerve.
Our survey showed that alternatives to surgery are certainly commonplace in current UK practice. In fact, observation alone is commonly used in a variety of clinical presentations seen by non-tuberculous mycobacterial cervicofacial lymphadenitis and is shown to be the treatment of choice when there is concern about the risk of damage to the facial nerve with surgery.
The literature showsReference Zimmerman, Tebruegge, Curtis and Ritz13 there is a temporary facial nerve palsy rate of 10 per cent and a permanent palsy rate of 2 per cent. This perhaps demonstrates a change in practice, certainly from the literature, which has historically demonstrated surgery to be the mainstay of treatment. This suggestion was further supported by our survey results when evaluating changes of practice over one’s career. According to our survey, where practice has changed it has transitioned from a pro-surgical to a more conservative approach in the management of this condition. This appears to be based on previous adverse outcomes as a consequence of a surgical approach.
Where antimycobacterial therapy is used, the regimen of choice is varied. This once again may be a reflection of the lack of published evidence on any recommended regimen and the duration of antimycobacterial agents, aside from the consensus that at least dual therapy for an extended period is advocated owing to the known resistance profile of the organisms. This is a significant undertaking for the patient. In addition, side effects of the recommended antimycobacterial agents are commonly experienced,Reference Lindeboom, Kuijper, van Coppenrat ESB and Lindeboom10, Reference Lindeboom11 such as teeth discolouration, fatigue, fever and abdominal pain. The economic cost for long-term antibiotics should also be considered.
Some argue that the benefits thought attributable to long-term antimycobacterial agents are more a reflection of the natural course of the pathology, therefore preferring observation alone. This theory is supported in some evidenceReference Zimmerman, Tebruegge, Curtis and Ritz13 where there was demonstrated to be no improved cure rate between antimycobacterial therapy when compared with observation. With the great variability of choice and duration of antimycobacterial therapies used in the limited available literature, this is area of potential research.
Another area of potential investigation is qualitative research to evaluate the patient’s and parents’ experience of their clinical journey. As far as the authors are aware there are no publications on this. Our survey demonstrated that the views of parents influenced management choice in 68 per cent of patients.
• There is no universally agreed management of paediatric non-tuberculous mycobacterial cervicofacial lymphadenitis, with limited high-level evidence that likely contributes to the different approaches in the management
• Paediatric non-tuberculous mycobacterial cervicofacial lymphadenitis is broadly managed surgically and medically, often with prolonged multi-agent antimicrobials or through observation
• The available literature suggests that surgery was the mainstay of treatment historically
• Non-surgical management of paediatric non-tuberculous mycobacterial cervicofacial lymphadenitis is now commonly used in the UK
• The management of paediatric non-tuberculous mycobacterial cervicofacial lymphadenitis in the UK is in keeping with current international consensus
Conclusion
Our survey showed that local guidelines are infrequent, even though there is a demonstrated appetite for national guidance. Respondents felt that national guidelines would improve the care of this patient group.
The results of our survey, with data collected prior to the publishing of the International Pediatric Otolaryngology Group paper,Reference Roy, Balakrishnan, Boudewyns, Cheng, Chun and Daniel14 appears to demonstrate that the management of non-tuberculous mycobacterial cervicofacial lymphadenitis in the UK is comparable to the current international consensus. The paper makes recommendations of establishing a ‘probable non-tuberculous mycobacteria lymphadenitis’ diagnosis based solely classical clinical presentation, an approach heavily favoured in UK, as established above. Perhaps where current UK management mirrors the international consensus the most is around treatment choice, although the paper states ‘clinical resolution was best achieved with complete surgical excision… some institutions strongly favour non-operative management.’ With regards to treatment, the paper acknowledges that when medical management is pursued, ‘there is high variability amongst institutions both in antibiotics regimen and duration’.
Competing interests
None declared.
