Challenges associated with the treatment of Buruli ulcer

Buruli ulcer (BU), caused by Mycobacterium ulcerans (MU), is the third most important mycobacterial diseases after tuberculosis and leprosy in immunocompetent individuals. Although the mode of transmission remains an enigma, disease incidence has been strongly linked to disturbed environment and wetlands. The blunt of the diseases is recorded in West African countries along the Gulf of Guinea, and children 15 years and below account for about 48% of all cases globally. Prior to 2004, wide surgical excisions and debridement of infected necrotic tissues followed by skin grafting was the accepted definitive treatment of BU. However, introduction of antibiotic therapy, daily oral rifampicin (10 mg/kg) plus intramuscular injection of streptomycin (15 mg/kg), for 8 weeks by the WHO in 2004 has reduced surgery as an adjunct for correction of deformities and improved wound healing. An all‐oral regimen is currently on clinical trial to replace the injectable. It is thought that a protective cloud of the cytotoxic toxin mycolactone kills infiltrating leucocytes leading to local immunosuppression and down‐regulation of the systemic immune system. Our studies of lesions from BU patients treated with SR have demonstrated treatment‐associated initiation of vigorous immune responses and the development of ectopic lymphoid tissue in the BU lesions. Despite these interventions, there are still challenges that bedevil the management of BU including paradoxical reactions, evolution of lesions after therapy, prolong viability of MU in BU lesions, and development of secondary bacterial infection. In this paper, we will mainly focus on the critical and pertinent challenges that undermine BU treatment toward effective control of BU.


INTRODUCTION
Buruli ulcer (BU), a necrotizing skin condition, is the third most important mycobacterial disease globally after tuberculosis (TB) and leprosy in immunocompetent individuals. 1,2 In highly endemic countries, such as Ghana, BU is second after TB as the most prevalent mycobacterial disease. 2

BU was first described in 1879 by a British physician, Sir
Albert Cook, and the etiologic agent of BU was later isolated from a farmer and named as Mycobacterium ulcerans (MU). 3 The name "Buruli ulcer" was designated in the 1960s after the Buruli County in Uganda where the largest number of cases was recorded then. 4 The disease is currently being reported in 33 countries globally, but highest disease burden is found in West African countries along the Gulf of Guinea and include Ivory-Coast, Ghana, Togo, Benin, and Cameroon. 5 Abbreviations: AFB, acid-fast bacilli; ART, antiretroviral therapy; BU, Buruli ulcer; IRIS, immune reconstitution inflammatory syndrome; MU, Mycobacterium ulcerans; SR, Streptomycin and rifampicin; TB, tuberculosis BU affects equally both sexes and all age groups and no ethnic preference has been reported. 6 The clinical and epidemiological aspects of cases vary considerably within and across different geographical settings, especially in Africa children 15 years or less constitute about 48% of all cases, whereas in Australia, 10% are children under 15 years and in Japan 19% are children under 15 years. 7 The mode of transmission of MU is not understood, but cases are mostly associated with disturbed environments and wetlands. 8 The initial stage presents as a nodule, papule, and plaque or in the more diffuse case as an edema.
If the early forms are not treated, extensive skin destruction leads to the formation of an ulcer. Laboratory methods for confirmation of clinical diagnosis include culturing for MU from lesion samples, direct acid-fast bacilli (AFB) detection by microscopy, and PCR to detect bacterial DNA. [9][10][11][12][13][14][15][16] The major hallmarks of MU infection, which are also used for histopathological confirmation, are the presence of coagulative necrosis, fat cell ghosts, epidermal hyperplasia, and extracellular clusters of AFB in the absence of major inflammatory infiltrates in central parts of the lesions (Fig. 1). 17,18 F I G U R E 1 Histological features of Buruli ulcer lesion. Histological sections of punch biopsies of BU lesions stained with hematoxylineosin (A-E, original magnification 50×) and Zeihl-Neelsen (F, original magnification 100×) Prior to 2004, wide surgical excisions and debridement of infected necrotic tissues followed by skin grafting to correct deformities was the mainstream management protocol for BU. 19,20 A study initiated by the WHO and conducted in Ghana indicated that BU lesions can be sterilized by treatment with streptomycin and rifampicin (SR). 19,21 Based on this finding, a treatment guideline was released in 2004; daily oral rifampicin (10 mg/kg) plus intramuscular injection of streptomycin (15 mg/kg) daily for 8 weeks, reducing surgery as an adjunct for correction of deformities and facilitating wound healing. 21,22 An all-oral regimen is currently on clinical trial. Although the introduction of antibiotics therapy has proved successful in the management of BU cases including reduction in length of hospital stay, cost of treatment, and reduction of relapse to less than 2%, 21,22 there are still challenges in BU treatment.

IMMUNE SUPPRESSION ASSOCIATED WITH BU DISEASE
The pathogenesis and host immune response mechanism are not clearly understood; however, most of the observed pathology is linked to the secretion of a polyketide macrolide toxin called mycolactone.
Upon entry into the host, MU is confined under the skin and the long incubation period, which has been suggested to be between 2.0-4.5 months, favors its proliferation within the dermis. 23 The temperature requirement of MU offers optimal conditions for the development of lesions in cooler tissues, particularly, the skin and subcutaneous tissues. 23,24 BU may manifest initially as a painless nodule, papule, nodule, plaque, or edema. Subsequent obliteration of the subcutaneous adipose tissue results in the breakdown of the epidermis and formation of characteristic ulcers with undermined edges. 25,26 It is thought that a protective cloud of the cytotoxic toxin mycolactone kills infiltrating leucocytes leading to local immunosuppression and down-regulation of systemic immune response. 24,27 Mycolactone caused cytopathic effects on cultured L929 murine fibroblasts and inoculation of puri-fied mycolactone into guinea pigs intradermally also produced lesions that were histologically like BU with necrosis of subcutaneous fat. 27 In contrast to other pathogenic mycobacteria, which are intracellular pathogens of macrophages, histology of MU lesion finds extracellular clusters of MU bacilli lying within areas of coagulative necrosis that extend some distance from the site of bacterial colonization. 28,29 Nevertheless, inoculation of an isogenic toxin-negative mutant of MU caused a granulomatous lesion typical of the inflammatory response to other mycobacteria with phagocytozed MU visible within macrophages and none of the characteristic fat necrosis. 27 Despite some antiphagocytic activity of mycolactone, other studies have shown that phagocytes can internalize MU in vitro. Coutanceau et al. using mouse models found that MU was initially captured by phagocytes and transported to draining lymph nodes within host cells; however upon ulceration, tissue necrosis and extracellular bacteria as seen in human BU were seen. 30 Torrado et al. also demonstrated that mycolactone-producing MU isolates are efficiently phagocytozed by murine macrophages. 31 The authors further note that MU multiplies inside cultured mouse macrophages when low multiplicities of infection are used to prevent early mycolactone-associated cytotoxicity and subsequently induced lysis of the infected host cells to become extracellular. 31 The innate response involves the release of proinflammatory cytokines, such as IL-6 and lipids, to recruit and activate other immune cells and apoptosis. [32][33][34][35][36][37] If the infection persists, the phagocytes stimulate the adaptive immune system by presenting Ags to activated T and B cells. 38 It appears that adaptive immune responses associated with IFN-secretion may be crucial. The cytopathic effect of the macrolide toxin causes apoptosis of mammalian cells 39  that infection with MU is associated with T cell anergy as PBMCs from individuals with BU exhibited reduced lymphoproliferation and production of IFN-following stimulation with living or heat-killed mycobacteria. 43 These same authors investigated cytokine profiles of PBMCs from patients and household contacts and showed that BU patients mounted a Th2-type response, which was manifested by the production of mRNA for IL-4, IL-5, IL-6, and IL-10, whereas unaffected contacts responded mainly with the Th1 cytokines IFN-and IL- 12. 44 This suggests that a Th1-type immune response to MU may prevent the development of BU in people exposed to MU. 44 In Guyana, Prevot et al. demonstrated that in active BU patients, in vitro production of IL-10 in PBMCs after stimulation with MU was significantly increased compared to tuberculin positive controls and the reverse was true for IFN-. 45 The production of distinct cytokines was also found to be dependent of lesion stage. In resected tissues, the level of IFN-mRNA was higher, and IL-10 mRNA was lower in nodular lesions than ulcerative lesions after stimulation with heat-killed MU. 45 Westernbrink et al. using in a whole blood assay demonstrated a systemic reduction in IFN-production in response to purified protein derivative Ags in patients with early lesions compared to those with later Mycobacterial material ("acid-fast debris") was primarily located inside phagocytes. The role of antibody-mediated immunity is not evident, but BU patients were able to produce antibodies against MU culture filtrate. The mechanism of suppression is not clearly understood. proteins that transit through the endoplasmic reticulum. 50 George et al. further showed that addition of mycolactone to macrophages and fibroblast affected the organization of the cytoskeleton that leads to growth arrest and apoptosis. 27 Furthermore, IL-2 production from activated T lymphocyte was blocked by the toxin. 27

SECONDARY LESIONS OCCURRING AFTER BU TREATMENT
The occurrence of new lesions at the site of infection during the era of surgical excision was high and it can be up to 47% due to insufficient excision of infected tissues. 51 However, this has been reduced greatly following antibiotic therapy and most secondary lesions are not due to relapse but rather increased immune reactivity. In mycobacterial infections, such as TB and leprosy, studies have shown that effective antimicrobial killing may be accompanied by clinical deterioration 52-54 ; a phenomenon normally described as paradoxical reaction. Although commonly it occurs in severely immunosuppressed patients, it can also occur in immunocompetent hosts. 52,55 This reaction may be due to increased exposure to mycobacterial Ags, a decrease in suppressor mechanisms, or improved host cell-mediated immunity following antimycobacterial therapy. 54,56 Recently, paradoxical reactions have been recognized to complicate up to 20% of patients receiving BU therapy, and sometimes leads to evolution of multifocal BU lesions. 13

Superinfection of BU wounds
The occurrence of secondary infection in BU disease was previously believed to be uncommon 7 and therefore was not well characterized and documented. Mycolactone secretion by MU during active disease was formerly hypothesized to exert a sterilizing effect on the wounds thus preventing secondary infection due to the fact that other macrolides have broad spectrum activity against many bacterial species. 63 Recent studies however have shown that secondary infection is more common than formerly thought (Fig. 4).
Studies documenting the occurrence of secondary infection with the isolation of infecting pathogens 60,63-68 and the growth of microbial pathogens in the presence of mycolactone 64

BU and HIV co-infection
The synergy between Mycobacterium tuberculosis infection and HIV/AIDS is well established but not so in BU. HIV commonly presents with clinical anemia as part of a pan-cytopenic cell line presentation. [76][77][78] The severity of the anemia correlates with the extent of immunosuppression as expressed by declining CD4 count. 79 Some studies have revealed association of mycobacteria infections with worsening peripheral blood cytopenia even though such has not been proven with MU infection. [80][81][82][83] What is rather proven as a common cause of anemia in BU patients is nutritional anemia because the disease is mainly associated with persons of lower socioeconomic status. 84,85 Susceptibility to BU is associated with polymorphism in the gene for the iron transporter protein NRAMP1. 86 There are models we recommend further studies to ascertain the most appropriate time to commence ART in relation to SR treatment to minimize paradoxical reactions. 85 The goal of BU wound management is to achieve wound closure within the shortest possible time devoid of extensive limb restrictions and deformities. 85,88,104 Various studies have shown that there are challenges in achieving timely wound closure of BU/HIV wounds. 85,104 Wound closure duration of BU/HIV cases could be more than 2 times the duration to healing of HIV negative BU cases. 85,101 This could occur irrespective of the diligent wound care practices adhered to in the management of these ulcers. It is believed that the immunosuppressive state has a tendency to stall the progress of certain phases of the wound healing process, thereby leading to wounds failing to heal. 105,106

PROLONG VIABILITY OF MU IN BU LESIONS
The SR8 treatment regimen was based on observational study of patients with early lesions, which were excised after SR treatment for 2, 4, 8, or 12 weeks. 19 All lesions were culture positive until 2 weeks but thereafter all were culture negative. 19

Improper wound care management
The median time to healing of early limited BU lesions has been reported to be about 18 weeks. 104,107 The long healing times therefore implies that wound management is an important component of BU wound management especially postantibiotic therapy. Good wound management is believed to reduce time to healing ultimately decreasing the risk for secondary infection, pain, and morbidity. 117 According to the authors, in several health centers in endemic countries, improper wound management was practiced. 117 Reported practices, including not washing of wounds and surrounding intact skin, removal of old dressings without moistening exposing the wounds to trauma, wounds being cleaned by rubbing cotton wool soaked with dressing solutions on the wound instead of the application of moderate-pressure irrigation, the choice of using different topical antiseptics instead of normal saline for wound cleaning, frequency of dressing changes not based on wound characteristics but on hospital policy, and the use of unsterilized materials for dressing wounds, 117 have the potential of negatively impacting the healing of patients, increasing the risk of secondary infection of wounds, and ultimately delaying their reintegration into their families and society. 117 The WHO guidelines for wound management exist and reports have shown that health workers have adequate knowledge and training on these guidelines 105,117 ; however, in some health facilities, adherence to the guidelines and delivery of proper wound care is hampered by the lack of adequate infrastructure, equipment, and wound dressing supplies. 105,109 Providing appropriate facilities and tools to these health centers will empower them to be able to provide good care to patients. Periodic training of health care workers on the guidelines for wound management and monitoring to ensure compliance will go a long way to ensure compliance and increase the standard of wound care.

CONCLUSION
Antibiotic therapy has proven to be essential in the management of BU requiring surgery only as an adjunct to correct deformities. Nevertheless, some patients experience poor clinical outcomes in the course of treatment. In this review, we have explored several probable factors that challenge BU case management. There is the need for more biomedical and behavioral studies for understudying the evolution of BU wounds during treatment and adherence to proper wound care for improved case management, ultimately reducing the associated long hospital stays.

ACKNOWLEDGMENTS
The authors acknowledge the support provided by all Buruli ulcer patients. They are also grateful to Prince Asare for generating the figures used in this paper.

DISCLOSURES
The authors declare no conflicts of interest.