Background

Buruli ulcer (BU) caused by Mycobacterium ulcerans is endemic in parts of West Africa [1, 2]. The causative organism produces the toxin called mycolactone [3], which has been shown to be cytotoxic to a range of immune and nonimmune cells [4]. Although the toxin mediates a predominantly necrotizing cutaneous and subcutaneous tissue infection, it may occasionally involve deeper structures such as the bone, causing osteomyelitis. BU begins as a small, painless, raised skin papule, nodule, plaque, or edema that becomes indurated, crusted, and eventually ulcerates. The formation of the characteristic ulcer results from destruction of the subcutaneous adipose tissue leading to collapse of the epidermis. Over a period of weeks, the ulcer enlarges, with widespread thickening of subcutaneous tissue. Subsequently, there is necrosis that produces the characteristic undermining of the ulcer margins. In larger lesions, there may be extensive induration adjacent to the ulcer. Beyond this area of thickening, however, the skin appears normal, with no cellulitis or lymphadenopathy. Necrosis may extend down to muscle fascia and even to bone, but actual osteomyelitis is rare. Advanced lesions display massive tissue destruction and minimal inflammation, with extracellular microcolonies of M. ulcerans in the superficial necrotic areas [5, 6]. Characteristically, the lesions are not associated with pain or tenderness. The disease is rarely fatal but can lead to significant debilitation as well as enormous social and economic consequences in afflicted persons [2]. BU is most prevalent among the age group 5–15 years but is uncommon among neonates [1]. The exact mode of transmission and the incubation period of the causative organism of the disease are not known. We report cases of confirmed BU disease in two human immunodeficiency virus (HIV)-unexposed, vaginally delivered term neonates from the Ashanti region of Ghana and discuss the implications for understanding the incubation period of M. ulcerans.

Case presentation

The clinical characteristics of the two patients are presented in Table 1.

Table 1 Clinical characteristics of two Ghanaian neonates with Buruli ulcer disease
Full size table

Patient 1

This patient was from the Ashanti ethnic group in Ghana. Two weeks after hospital delivery, patient 1’s mother noticed a papule with associated edema on the right anterior chest wall and neck that later ulcerated and was associated with fever and poor feeding. Clinical examination revealed an 11-cm × 7-cm ulcer with undermined edges on the right anterior chest wall that extended to involve the right neck and pre- and postauricular regions and submental regions (Fig. 1). There was no restriction of neck movements. An initial clinical diagnosis of necrotizing fasciitis was made, and empiric treatment with gentamicin was started, based on the local culture and sensitivity patterns. The ulcer was secondarily infected by Staphylococcus aureus, which was sensitive to gentamicin. The diagnosis of BU was confirmed on the basis of a swab sample that was polymerase chain reaction (PCR)-positive for the IS2404 repeat sequence of M. ulcerans.

Fig. 1
figure 1

Pre- and post-treatment Buruli ulcer (BU) lesion in patient 1. Buruli ulcer located on the anterior chest of patient 1 before BU-specific antibiotic treatment (a). The ulcer had completely healed after 8 weeks of treatment with rifampicin and clarithromycin in addition to appropriate daily wound dressings (b)

Full size image

Patient 2

This patient was from the Ashanti ethnic group in Ghana. The mother of patient 2 reported noticing a swelling in the baby’s left gluteal region 4 days after birth. The lesion progressively increased in size to involve almost the entire left gluteal region. Around the same time, the mother noticed a second, smaller lesion on the forehead and left side of the neck. The patient had a 6 × 5-cm left gluteal ulcer with undermined edges, firm base, and floor covered with slough and purulent discharge (Fig. 2). An initial clinical diagnosis of necrotizing fasciitis was made, and empiric treatment with gentamicin was initiated. A wound swab culture result yielded Escherichia coli, which was sensitive to gentamicin. The diagnosis of BU was confirmed by PCR when the child was aged 4 weeks.

Fig. 2
figure 2

Pre- and post-treatment Buruli ulcer (BU) lesion in patient 2. Buruli ulcer on the gluteal region of patient 2 before BU-specific antibiotic treatment (a). The ulcer had completely healed after 8 weeks of treatment with rifampicin and clarithromycin in addition to appropriate daily wound dressings (b)

Full size image

Treatment

Both patients were treated with oral rifampicin and clarithromycin (at doses of 10 mg/kg and 15 mg/kg, respectively) administered daily for 8 weeks in addition to appropriate daily wound dressing. Antibiotic treatment was well tolerated by both patients. Treatment adherence was monitored using the standard Buruli ulcer 01 (BU 01) form used during the routine care of patients with BU. Their ulcers healed completely by the end of antibiotic treatment, but they developed contracted scars. Surgery to correct the deformities was planned by a plastic surgeon.

Discussion

To the best of our knowledge, there has not been any previous report of confirmed BU in neonates in Ghana. Our report details two cases of PCR-confirmed BU in children whose lesions appeared at age 14 and 4 days, respectively. BU caused by M. ulcerans is endemic in parts of West Africa and usually presents as a painless subcutaneous nodule or plaque or as a more aggressive edematous lesion. Over time, all the lesions ulcerate and progressively enlarge. In Africa, the disease is most prevalent among the age group 5–15 years [1].

The mode of transmission of M. ulcerans infection is unknown, so the incubation period is difficult to estimate. Studies of people who have become infected after moving from a nonendemic to an endemic area have provided some insight into this. The Uganda Buruli Group studied refugees from Rwanda, where there are no reported cases, and found new Buruli lesions occurring 4–10 weeks after their arrival from an endemic area [7]. An incubation period of 6 weeks was estimated in a Nigerian physician who had visited an endemic area and later developed BU in New York [8]. In two recent studies among patients developing BU after a short visit to an endemic area in Australia, the median incubation periods were estimated as 135 and 143 days, respectively, with the shortest period given as 32 days [9, 10]. These studies provide a fair idea of the possible incubation period of BU in adults, but in neonates it may range from a few days to several weeks. A study of 13 cases of BU in Port Moresby identified the shortest known incubation period of 2–3 weeks in a 6-week-old baby born in an endemic region [11]. Cases of BU have been observed in 18-day-old [12] and 9-month-old [13] babies in the Democratic Republic of Congo and Uganda, respectively.

In the present study, lesions appeared 4 days after birth in patient 2. Unless infection was acquired in utero, this would be the shortest incubation period ever recorded. There was no evidence of BU in the mother, so it must be assumed that the baby was exposed to an M. ulcerans-contaminated water source shortly after birth. Because neither the mothers nor the babies in this report had HIV infection, the short incubation period probably resulted from immaturity of the neonatal immune system. Another explanation would be that the babies were exposed to a high infection dose, but there is no way of confirming this.

Bacille Calmette-Guérin (BCG) vaccine was thought to have a short-term protective effect against M. ulcerans infection in Uganda [14], and having a BCG vaccination scar was reported to provide significant protection against M. ulcerans osteomyelitis [15]. A recent study, however, did not find significant evidence of a protective effect of routine BCG vaccination on the risk of developing either BU or severe forms of BU [16]. The vaccine used now may not be exactly the same, and certainly it did not protect the babies in this report, who had both received it at birth.

Traditionally, combination antibiotic therapy with intramuscular streptomycin and oral rifampicin for 8 weeks has been used to treat BU disease, and this has been shown to give excellent cure rates [17, 18]. Due to the disadvantages of streptomycin [19] (prolonged intramuscular injections and risk of ototoxicity), daily oral treatment with rifampicin and clarithromycin is now recommended for treatment, especially in children.

There was evidence of symptomatic secondary bacterial infection in both the patients reported here, with S. aureus in patient 1 and E. coli in patient 2. This would account for the fever and poor feeding reported by both mothers and the systemic signs of infection that are not associated with BU. However, the ulcers did not start to heal until specific therapy for M. ulcerans was administered. Secondary bacterial infections with S. aureus and E. coli have been reported as common isolates, usually sensitive to gentamicin, as was the case in our patients [20].

Conclusions

This report presents the first cases of BU in neonates from Ghana, a BU-endemic country in West Africa. Oral treatment with rifampicin and clarithromycin administered daily for 8 weeks proved effective in the management of BU in these neonates, thereby avoiding the use of painful daily streptomycin injections with their associated ototoxicity. BU should be included in the differential diagnosis of neonates who present with characteristic lesions. The incubation period of BU in this age group is probably shorter than is reported for adults.