Consensus recommendations for the diagnosis, treatment and control of Mycobacterium ulcerans infection (Bairnsdale or Buruli ulcer) in Victoria, Australia

Bairnsdale or Buruli ulcer (BU) had been a rare disease in Victoria, but its incidence has risen markedly since 1990. Clinicians are increasingly called upon to diagnose and treat BU, but there is little information on the best approach. These guidelines reflect contemporary clinical practice in Victoria. They may not be applicable in countries with less developed health infrastructure. Existing guidelines, approved by the World Health Organization, are recommended for such areas.1

In late 2004, the Victorian Department of Human Services called for targeted research on BU. The development of consensus guidelines was one component of the successful grant application.

Plastic surgeons, general practitioners, laboratory scientists, pathologists, infectious diseases physicians and public health experts known to have experience with BU were invited to participate. Of 85 invitees, 46 attended the conference on 10 February 2006.

Before the conference, participants received a draft document written by the conference conveners. In the morning, 10 speakers reviewed the published literature and presented and discussed recent cases. In the afternoon, the draft document was systematically reviewed, extended and improved through interactive round table discussion, with the help of a professional facilitator. The next draft was circulated to all participants after the conference, and final comments and suggestions were incorporated. The penultimate document underwent peer review and was edited for publication. The level of evidence throughout this document is level 4/5 (observational case series/expert opinion), except where specific references are cited.

Mycobacterium ulcerans was discovered in 1948 by Australian scientists who were investigating a cluster of patients with unusual skin ulcers in the Bairnsdale region of eastern Victoria.2 M. ulcerans is related to the causative agents of tuberculosis and leprosy, but is transmitted from the environment rather than from person to person.3 The major virulence factor is a lipid toxin, mycolactone, which causes necrosis of fat and subcutaneous tissue.4

M. ulcerans infection is not fatal, but can result in significant morbidity and is expensive to treat.5,6 It has been reported in more than 30 countries, and is currently a significant public health problem in sub-Saharan Africa (Buruli ulcer).3 In Australia, there are active foci in coastal Victoria (Bairnsdale ulcer),7 Far North Queensland (Daintree ulcer),8 and near Rockhampton.9 Single cases also occur elsewhere in the Wet Tropics. The reason for this patchy distribution is unknown, but molecular typing has shown differences between strains isolated from patients in different regions.10

BU continues to occur in low numbers in the Bairnsdale/Gippsland Lakes area. In 1980, a new focus was noted near Tooradin and Warneet (Box 1) on Westernport,7 and this was followed by significant outbreaks at East Cowes on Phillip Island (1992–1995),7,11,12 Frankston and Langwarrin (1990–1998),7 St Leonards (1998–2001), Point Lonsdale (since 2002)13 and Barwon Heads and Ocean Grove (since 2005). There are also frequent single cases from other parts of the Mornington and Bellarine Peninsulas.

The incidence of BU in Victoria has increased sharply in recent years (Box 2). It is unlikely that this increase is due to improved diagnosis, as there has been considerable publicity since 1994, and all diagnostic polymerase chain reaction (PCR) assays have been available at a single institution since 1995.14 BU was made formally notifiable in Victoria from January 2004. Although the incidence in Victoria is currently only 1.4 per 100 000 overall, it is estimated that up to 6% of the permanent population of East Cowes (1992–1995)15 and 1% of the permanent population of Point Lonsdale have required treatment. Visitors to endemic areas are also at risk, and brief contact may be sufficient to become infected.

Patients who present with BU are usually otherwise well, and may be male or female, child or adult. Most patients will be residents or visitors to a known endemic area, but isolated cases occur.

BU typically presents as a slowly progressive skin papule that undergoes necrosis and evolves to a deeply undermined ulcer. Lesions are usually painless or minimally painful. Less commonly, the initial lesion may resemble a plaque or necrotic patch of skin. Systemic symptoms (fever, malaise) are rare.

Fully developed ulcers are characteristically deeply undermined, meaning a probe can be passed easily under the edge into the space left by necrotic liquefied fat tissue. Lesions can occur anywhere and may be multiple, but a single lesion on the leg or arm is most common. Infections on the buttock, abdominal wall, ear and face have all occurred in Victoria in the past 10 years. There is also an unusual oedematous presentation that can cause extensive swelling (eg, involving most of a limb).

In patients presenting with unresolved cellulitis8 or a suspected necrotising spider bite, BU should be considered. At present, there is no evidence connecting spider bites to M. ulcerans.16

Once an ulcer is present, diagnosis is usually straightforward, provided BU is considered. Delays are most likely when patients present outside endemic areas, or when the patient has a non-ulcerative presentation (eg, nodule, papule, plaque or acute oedema).

Acid fast bacilli (AFB) smear-positive swabs or specimens from skin ulcers are strongly suggestive of M. ulcerans infection in patients from Victoria who have not travelled overseas.

IS2404 PCR,8,14,17 which can be performed directly from ulcer swabs, approaches 100% sensitivity and specificity (P D R J, J A M F, independent personal observations). Culture is also diagnostic, but generally takes several weeks. A positive M. ulcerans PCR result is sufficient evidence to proceed to treatment.

A negative M. ulcerans PCR result, when AFB are detected on the smear, is likely to indicate an alternative diagnosis.

A negative AFB smear and negative M. ulcerans PCR result make the diagnosis of BU unlikely, but it should not be completely excluded.

If an ulcer is present, obtain samples deep to the undermined edge using two standard cotton-tipped swabs. These swabs may be dry or of the standard transport medium type (swabs stored in charcoal transport medium should be avoided). Ensure that some material is visible on each swab. The swabs (one labelled “for PCR only”) should be sent to your usual pathology service for microscopy for AFB, M. ulcerans PCR, and culture. “Possible Bairnsdale ulcer” should be written on the clinical notes. A smear result is usually available within the same working day. Separate swabs or specimens should be obtained for other pathogens or conditions, depending on the clinical situation.

Other specimen types suitable for PCR include fresh tissue biopsies (preferably not stored in large volumes of buffer or saline, nor wrapped in gauze) and paraffin-embedded fixed tissue sections (not suitable for culture).

If an ulcer is not present but there is a suspicious plaque, necrotic patch, nodule or acute oedematous presentation in an at-risk patient, an incisional or excisional biopsy should be performed. Some patients notice an apparent increase in the rate of progression of the disease following biopsy, and this should be carefully discussed.

Small lesions sometimes resolve spontaneously, but this is thought to be uncommon. The rate of progression varies among patients. Even though host immunity progressively develops during infection, ulcers may become very extensive. Relapse after treatment is not uncommon,18 so regular follow-up is recommended.

Cure with medical therapy alone is possible, and there is increasing interest in this approach in western Africa.1,19 However, in Victoria, where there is ready access to surgery, we believe that surgery or combined surgical and medical therapy is the most efficient way of effecting cure. The trend in management is towards conservative surgery with macroscopic removal of necrotic tissue and the use of adjuvant antibiotics. Patients may be best managed by a team, with surgeons working with infectious diseases specialists, GPs and allied health practitioners.

For difficult or recurrent disease or when antibiotics and surgery are failing, other modes of therapy, including continuous heat therapy or hyperbaric oxygen, may be considered.

It is not understood why M. ulcerans outbreaks occur in new areas or why the disease has spread westwards from the original endemic area near the Gippsland Lakes. There are no public health interventions that can remove M. ulcerans from the environment, although there may be a natural cycle of human infection that lasts several years, after which the incidence of new cases in a given area abates. There is circumstantial evidence that nutrient enrichment of very low-lying coastal environments may be a factor in the appearance of outbreaks.7,12,28

M. ulcerans has been detected by PCR in plant material and mud obtained from swamps in endemic areas,29 a golf course irrigation system that used recycled water,28,29 and from aquatic water insects in Africa.30 At Point Lonsdale, M. ulcerans has been detected by PCR in soil and leaf litter from a stormwater drainage system, mud from a lake, and about 0.5% of more than 10 000 mosquitoes trapped during 2004–2006 (unpublished data). It is yet to be determined whether mosquitoes play a role in human disease.

Studies from African endemic areas have reported that farming activities close to rivers31 and swimming32 may be risk factors, and that wearing trousers31 and a shirt33 when working outdoors appears protective. A recent case–control study performed on the Bellarine Peninsula has shown that direct exposure to the environment and to mosquitoes are risk factors, and wearing protective clothing and insect repellent appears protective (unpublished data).

Morbidity and cost can be reduced by early diagnosis. Therefore, local doctors are central to early detection. When an outbreak occurs in a new area it may take some time for this disease to be recognised, so public health authorities need to provide timely information to both clinicians and the general public about disease activity and the location of new endemic areas.