Chapter 272. Buruli Ulcer (Mycobacterium Ulcerans Infection)

Mycobacterium ulcerans causes indolent, necrotizing cutaneous lesions known as Buruli ulcers, an appellation given by Dodge and Lunn who described the first large epidemic, located in Buruli County (now called Nakasongola), Uganda.1 Today, M ulcerans infections are recognized to present a spectrum of clinical disease: nodules, plaques, severe edemas and massive ulcers in the skin, and osteomyelitis. Buruli ulcer, after tuberculosis and leprosy, is the third most common and perhaps least understood major mycobacterial infection. In contrast to tuberculosis and leprosy, Buruli ulcer is closely related to environmental factors.2

In South America, known countries in which Buruli ulcer is endemic include French Guyana, Suriname, and Peru.3,4 While Mexico is the only North American country in which it is endemic, travelers to endemic areas occasionally present to European, American, and Canadian medical centers.5-8

Since 1998, the World Health Organization (WHO) has recognized Buruli ulcer as a reemerging infectious disease in West and Central Africa with an important public health impact.9 In endemic countries, Buruli ulcer is a major public health and psychosocial problem because of potential disabling sequelae.10 The disease tends to afflict children 15 years old and under in those countries in which it is highly endemic.

Epidemiology

Endemic foci of Buruli ulcer are most common near rural permanent wetlands in warm geographic regions, especially in areas prone to seasonal flooding. Buruli ulcers have been reported from at least 27 countries, principally in the tropics.2 A few patients live in nontropical regions such as China,11 Japan,12 and southern Australia.13 The greatest number of reported patients live in West Africa (Benin, Burkina Faso, Côte d’Ivoire, Ghana, Guinea, Liberia, and Nigeria), with an estimated total annual incidence exceeding 7000 patients.2,14,15 Other known endemic countries include Angola, Cameroon, Democratic Republic of Congo, Equatorial Guinea, French Guiana, Gabon, Indonesia, Malaysia, Papua New Guinea, Peru, Suriname, Togo, and Uganda.16-19 The rapid reemergence of Buruli ulcer disease began in the early 1980s and is thought to be attributable to environmental factors such as deforestation, artificial topographic alterations (dams and irrigation systems), enlarging populations engaged in basic manual agriculture in wetlands, and possibly global climatic changes.20 In North America, 2 cases of Buruli ulcer were reported in central Mexico, the nearest location of the disease to the United States.21

Individuals of all ages are affected, but children 15 years of age or younger constitute about 75% of all cases.22 Approximately 80% of the lesions are located on the limbs, with highest frequencies involving the lower extremities. The sexes are affected equally, and racial predilection is unknown.23 Anecdotal observations of children in families of multiple parentage have suggested a possible genetic predisposition. This possibility is supported by molecular studies.24 Seasonal changes in climate affect incidence in some foci.25,26 Focal prevalence within countries varies greatly and must be assessed at the community level of geopolitical subdivisions.27

Some authorities believe that M ulcerans is a saprophytic or commensal organism that thrives in the mud, flora, or fauna of the cool microaerophilic environment at the bottom of stagnant water, well protected from the lethal ultraviolet radiation that prevails in the tropics. Marsollier et al speculate that M ulcerans may colonize biofilms in certain ecological niches.28 Although the ultimate source of M ulcerans remains obscure, the organism has been discovered in aquatic insects such as water bugs, firefly larvae, and beetles obtained from stagnant water in endemic areas of West Africa.14,29,30 In some cases, the organisms were found only in the salivary glands of aquatic insects, supporting the notion that insect bites may play a role in transmitting M. ulcerans.31,32 In Australia, M ulcerans DNA has been demonstrated in mosquitoes captured during an outbreak of Buruli ulcer.33 Koalas and possums in Australia acquire Buruli ulcer infections in the wild.14

Buruli ulcer infection is rarely, if ever, contagious. The distribution of patients, even in highly endemic foci, is random, suggesting that each patient is exposed to environmental sources such as swamps where villagers work their gardens, obtain water for domestic use, and where children play. The following additional risk factors have been determined in African populations: use of unprotected water for domestic purposes, swimming and wading in rivers, low level of schooling, and HIV infection.34,35

In some West African countries, the number of Buruli ulcer patients exceeds those of leprosy and tuberculosis.16 HIV infection may render the disease highly aggressive, with rapidly spreading M ulcerans osteomyelitis developing in several patients with AIDS.36,37

MacCallum and associates in 1948 in Australia were the first to isolate the etiologic agent in culture.38M ulcerans is strongly acid-fast, with an optimal growth temperature of 30°C to 32°C on routine mycobacteriologic media such as Löwenstein-Jensen medium. The organism is a slow grower, often requiring several months’ incubation to achieve isolation in primary culture. Microaerophilic conditions promote the growth of M ulcerans, and the organism is strikingly sensitive to temperatures of 37°C or higher.39,40 After a multitude of attempts to cultivate the organism from the environment by many investigators over half a century, the first isolation of M ulcerans from nature was reported in 2008.46 The development of polymerase chain reaction (PCR) techniques for the identification of M. ulcerans, facilitated detection in the environment in Australia and West Africa.41-45

Modes of transmission to humans have not been delineated completely; however, the most plausible route is by trauma at sites of skin recently contaminated by M ulcerans.41,47 Many patients give a history of specific antecedent penetrating trauma at the site of the initial lesion, which may include wounds from a gunshot or land mine, thrown stones, human bite, and hypodermic injection (eFig. 272.1).47,48 The organism may be spread by aerosol from the surface of ponds or be carried by fomites or insects to skin surfaces. Insects may introduce M ulcerans into the skin, but this means of transmission has not been confirmed in humans.49

The spectrum of clinical and histopathologic changes after infection suggests that some patients have strong innate or acquired immunity, other patients acquire immunity at various rates, and some never develop an effective resistance.

M ulcerans elaborates a necrotizing and immunosuppressive toxin,48a-c mycolactone, a polyketide.50,51 Mycolactone profoundly suppresses both B and T cells and inhibits the production of tumor necrosis factor gamma (TNF-) and interleukin-2 (IL-2) by T lymphocytes.50,52,53

Based on the current understanding of the natural history of M ulcerans infection, pathogenesis may proceed as follows: After inoculation of the etiologic agent deep into the skin or subcutaneous tissue, there is a latent phase during which the mycobacterium proliferates slowly, possibly intracellularly at first, and begins to elaborate small amounts of toxin causing necrosis, especially of fatty tissue. This necrosis provides a microaerophilic environment and perhaps nutrients favorable for the accelerated growth of M ulcerans and elaboration of increased amounts of toxin. During this necrotic phase, immunosuppression inhibits a significant cellular response. In some patients, at this stage a subcutaneous nodule begins to develop with clusters of Mulcerans in the center surrounded by a zone of necrosis (eFig. 272.2). In a study of trauma-related lesions, the incubation period ranged from 2 weeks to 3 years following the trauma, with a mean of 3 months.47 In highly resistant patients, this lesion may self-heal, perhaps without ulceration, or it may form a small, sharply defined ulcer. In others, the skin is undermined by the necrosis and eventually breaks down into larger ulcers with widely undermined skin. In the least-resistant individuals, a nodule never develops and the necrosis spreads rapidly and widely to cover large body surface areas, but ulceration, if it takes place at all, is a late event. Eventually, the necrotic stage ceases in most patients from reduced immunosuppression and reduction of viable M ulcerans. At this time, a granulomatous stage begins to develop, followed by healing and scarring. Some of the variability in clinical presentation and progression of the disease may be related to heterogeneity in the M ulcerans genome, as well as in the plasmid that encodes the production of mycolactone.54-57

Clinical Features and Pathology

In eFigure 272.5 a proposed classification of most of the clinical forms of M ulcerans is provided. World Health Organization (WHO) designations of the various forms of lesions of M ulcerans infection include: (1) papules that are painless, elevated, measure up to 1 cm in diameter, and ulcerate early (seen only in Australia); (2) nodules that are primarily subcutaneous and firm, measure approximately 2 cm in diameter, and are painless, though often pruritic [the initial stage in most African patients (eFig. 272.6)], (3) plaques that are firm, elevated, painless, well-defined lesions more than 2 cm in the largest dimension with skin over the lesion that is reddened or discolored, these lesions may ulcerate late, producing stellate ulcers without extensive undermining (eFig. 272.7); (4) edematous lesions that do not begin in the nodular stage but spread rapidly from the initial nidus of infection and often cover wide areas like entire limbs or major portions of the face or trunk (eFigs. 272.8 and 272.9); and (5) ulcerative lesions that are more or less symmetric, have undermined edges surrounded by a zone of induration, and often have desquamation of the surrounding epidermis. These are classified as minor (eFig. 272.10) and major (eFig. 272.1) lesions. In the base of the ulcerated area, a whitish necrotic slough and, sometimes, eschar develop.58

Microscopically, the active ulcer shows extensive coagulation necrosis of the subcutaneous tissue down to and often including the fascia (eFig. 272.3).18,59 Marked edema is present, and fat cells are enlarged and dead, leaving only their cellular ghost outlines (eFig. 272.4).

Bone involvement occurs due to both contiguous or metastatic osteomyelitis in about 10% of all patients.60 Reactive osteitis occasionally develops beneath destroyed overlying skin and soft tissue. Bone subjacent to the Buruli ulcer lesion may become devitalized and necrotic, with the development of sequestra (eFigs. 272.11 and 272.12).60 Metastatic osteomyelitis most likely results from lymphohematogenous spread of M ulcerans from a cutaneous lesion (eFig. 272.13). The overlying skin ordinarily is intact, but swelling and inflammation develop over the site of bone involvement. Bone scans, if available, are helpful in diagnosis.61 If the condition goes untreated, a draining fistula may develop. Both contiguous and metastatic osteomyelitis often result in deformity and amputation (eFigs. 272.14 and 272.15).

Diagnosis

An accurate clinical diagnosis often can be made by an experienced observer.58,62 In the ulcerative forms, a Ziehl-Neelsen stain of exudate from the undermined edge obtained with a cotton swab will reveal clusters of extracellular acid-fast bacilli (eFig. 272.21). The same material, obtained by swab after decontamination, may be used for culture on Löwenstein-Jensen or other suitable mycobacterial media. The incubation temperature must be 30°C to 32°C. If culture cannot be performed locally, transport media may be inoculated with material from the cotton swab and maintained at 4°C while in transport to a specialized laboratory. The transport medium is composed of Middlebrook 7H9 broth supplemented with polymyxin B, amphotericin B, nalidixic acid, trimethoprim, azlocillin, and 0.5% agar.63 Molecular biologic analysis techniques consisting of PCR for the identification of M ulcerans are available, convenient, and growing in popularity.45,54,64 Tissue for histopathologic analysis should be obtained from the edge of the ulcer or presumed center of edematous or plaque lesions and must include all levels of the integument, including the fascia. Fixation in 10% buffered formalin is adequate.

Other diagnosis that can be mistaken for Buruli ulcers include papules from insect bites, verruca vulgaris, pityriasis, granuloma annulare; nodules from lipomas, sebaceous cysts, onchocerciasis, furuncle; plaques from leprosy, mycosis, necrobiosis, psoriasis, Mkar disease (granuloma multiforme); edema from bacterial cellulitis, necrotizing fasciitis, actinomycosis, elephantiasis, pyomyositis; and ulcers from tropical phagedenic ulcer, noma, stasis ulcer, leishmaniasis, or injection abscesses (eFig. 272.22)

Treatment

Treatment options for Buruli ulcer are antibiotics and surgical intervention. The choice is usually based on the morphology and extent of the lesions, and the availability of antibiotics and surgical facilities. Favorable anecdotal results of antibiotic therapy have been known for many years, especially for early lesions, nodular and ulcerative. Current WHO guidelines are available, both for the surgical and antibiotic management of Buruli ulcer.60,65 Physiotherapeutic evaluation and management is imperative for all Buruli ulcer patients.66 Minor Buruli ulcers are believed to self-heal, but this form needs more clinical and pathogenetic study because these ulcers may demonstrate fundamental elements of high host resistance to M ulcerans disease.

Historically, surgical intervention has been the standard treatment for all forms of Buruli ulcer disease (eFigs. 272.18 and 272.19).58,60 In recent times, to minimize the possibility of M ulcerans spread, many surgeons give clarithromycin and rifampin for 1 to 2 days before surgery and continue antibiotic therapy for several weeks after surgery. However, this antibiotic adjunct to surgery is of variable efficacy.67

Papules and preulcerative nodules are seldom diagnosed, even in endemic areas; however, when they present, wide excision and primary closure are usually curative.68 Plaques and edematous forms are excised widely down to fascia or through the fascia if it is necrotic. Muscle usually is not damaged, but if it is, the excision is extended into muscle. The lateral extent of excision often is difficult to determine. By careful palpation, the physician can establish an approximate limit of the disease. Exploratory incisions and blunt dissection may help determine the limit of induration and necrosis. Use of real-time PCR for determining extent of disease is under evaluation.69 Very small ulcers can also be excised and closed primarily. Large ulcers are excised widely. Again, the required extent of surgery may be determined by exploratory lateral excision and blunt dissection. Split-skin autografts are normally applied after a bed of granulation tissue has formed. Recurrences after surgical treatment are common, but rates vary widely and the reasons are not well established.70

Heat therapy without surgical excision has been successful for appropriate lesions but must be applied assiduously with all necessary controls to prevent iatrogenic burns.39

Bone lesions are difficult to manage and should be referred to specialists to minimize disabilities.60,66 Antibiotic treatment as recommend by the WHO65 and though still somewhat empiric, currently includes at least an 8-week course of daily oral rifampin (10 mg/kg) and streptomycin (15 mg/kg) given daily as an intramuscular injection. eFigures 272.16 and 272.17 show a patient treated successfully with rifampin. Published reports on treatment results in Africa and Australia are encouraging. Short-term follow-up of 208 patients treated with antibiotics revealed only a 1.44% recurrence.71

Rifampin is supplied in tablet and syrup forms, the latter helpful for small children. Streptomycin and, to a lesser extent, rifampin, are associated with rare but important side effects, some of which require stopping treatment. For streptomycin, treatment should be stopped if hearing impairment or vertigo with nystagmus develop. For rifampin, treatment should be stopped if hepatitis, jaundice, or acute renal failure develops; these side effects are generally associated with intermittent therapy of doses over 10 mg/kg. Streptomycin is also contraindicated during pregnancy. In the authors’ opinion, antibiotic therapy for advanced lesions remains a subject for inquiry, and the regimens described above, although quite promising, do not constitute established recommendations at this time.

Infection may traverse the deep fascia and damage tendons, nerves, joints, and genitalia. When eyelids and periorbital tissues are destroyed, enucleation of the eye may be required if expert reconstructive surgery is not available. Healing leads to fibrosis and scarring and can severely limit movement, causing lifestyle alterations.72 The scar may form keloids and often causes major contraction deformities, especially in lesions that cross joints (eFig. 272.20). Squamous cell carcinoma may develop in healed lesions, especially those that are nonpigmented. Skin grafting and physiotherapy will prevent most of these complications. Most disease-related deaths result from septicemia, gas gangrene, or tetanus.

Without treatment, Buruli ulcer often leads to deforming depressed scars, contracture deformities, or amputations. A few deaths attributed to severe edematous Buruli ulcer have been observed. The stigma of the deformities and the socioeconomic burden of the disease often are marked.87 With early appropriate treatment, including excision and grafting, the prognosis usually is excellent. However, metastatic lesions and local recurrences occur frequently enough to warrant vigilant follow-up.88

In an endemic tropical, rural setting, where children usually are scantily attired, preventing contamination of the skin from environmental sources is virtually impossible. Wearing trousers seems to prevent development of infection.22 Protected water supplies in villages would reduce exposure somewhat; however, such protective measures usually are futile in rural areas of developing countries. Some other effective preventive measures include frequent use of soap for washing, treating injuries with soap and/or application of antibiotic powder, contact with rapidly flowing water, use of bed nets and insect repellent.35,49,75

Vaccination with bacille Calmette-Guérin (BCG) has a moderate protective effect against M ulcerans infections for 6 to 12 months.76 BCG vaccination has a prophylactic effect against osteomyelitis in Buruli ulcer in children.77 Studies are proposed to determine whether repeated BCG vaccination may render the population more immune to M ulcerans infection, as was found for leprosy.78Other vaccines based on DNA engineering and virulence factors of M ulcerans (eg, the toxin) are being studied.79