https://orcid.org/0000-0001-5418-3330
Abstract
Antibiotics are highly effective in curing Mycobacterium ulcerans lesions, but are associated with significant toxicity. In those not undergoing surgery, we compared 6 weeks with the currently recommended 8 weeks of combination antibiotic therapy for small M. ulcerans lesions.
Mycobacterium ulcerans cases from an observational cohort at Barwon Health, Victoria, treated with antibiotics alone from 1 October 2010 to 31 March 2018 were included. The 6-week antibiotic group received ≥28 days and ≤42 days and the 8-week antibiotic group received ≥56 days of antibiotic therapy, respectively. Only World Health Organization category 1 lesions were included.
207 patients were included; 53 (25.6%) in the 6-week group and 154 (74.4%) in the 8-week group. The median age of patients was 53 years (interquartile range [IQR], 33–69 years) and 100 (48.3%) were female. Lesions were ≤900 mm2 in size in 79.7% of patients and 93.2% were ulcerative. Fifty-three patients (100%) achieved treatment cure in the 6-week group compared with 153 (99.4%) in the 8-week group (P = .56). No patients died or were lost to follow-up during the study. Median time to heal was 70 days (IQR, 60–96 days) in the 6-week group and 128 days (IQR, 95–173 days) in the 8-week group (P < .001). Two (3.8%) patients in the 6-week group experienced a paradoxical reaction compared with 39 (25.3%) patients in the 8-week group (P = .001).
For selected small M. ulcerans lesions, 6 weeks may be as effective as 8 weeks of combined antibiotic therapy in curing lesions without surgery.
Mycobacterium ulcerans causes destructive lesions of skin and soft tissue known as Buruli ulcer and is recognized in 33 countries worldwide, mainly involving remote and rural communities in tropical Central and Western Africa [1]. It is associated with high rates of morbidity and long-term disability [2]. Currently, the recommended first-line treatment is a course of rifampicin-based combination antibiotic therapy for 8 weeks, sometimes combined with surgery to aid wound healing and prevent deformities [3, 4]. However, in Australian populations, where the median age of patients with M. ulcerans is older than in Africa and ranges between 55 and 60 years [5], there is a high rate of significant toxicity; ~40% of those aged 65 years or older develop an antibiotic complication severe enough to require cessation of at least 1 antibiotic and complications lead to hospitalization in up to 4% of cases [6]. The risk of complications persists throughout the antibiotic course, and thus shorter courses of antibiotics have the potential to significantly reduce the complication rate. In addition, there is a significant cost and inconvenience to more prolonged antibiotics [7] and increasing evidence that the extent of change to the human microbiome from antibiotic treatment depends on the duration of therapy [8].
Previous research from our group has reported that short-course antibiotic treatment can be effective for M. ulcerans, especially when combined with surgery [9]. In that study, all 6 patients who received between 28 and 42 days of antibiotics without surgery were cured, suggesting that a 6-week course of antibiotics may be effective in curing some M. ulcerans lesions. Furthermore, all 5 patients in that study who had tissue cultured postexcision after 28–38 days of antibiotics were culture negative. Other studies in which treatment involved 8 weeks of rifampicin-based antibiotic regimens have reported that swabs [10] or excised tissue [11, 12] from M. ulcerans lesions were culture negative after 4–6 weeks of antibiotics. Further supporting evidence for the effectiveness of 6-week antibiotic regimens comes from mouse studies that show that the combination of rifampicin and clarithromycin sterilizes mouse footpad lesions after only 4 weeks of daily antibiotics [13, 14].
The recommendation of 8 weeks of antibiotic treatment for M. ulcerans is made regardless of lesion size [3, 4]. However, in smaller lesions there may be a lower bacterial burden, and therefore smaller lesions may need shorter antibiotic durations to achieve sterilization and cure. Furthermore, it has also been shown that some lesions heal even if live organisms are present at the end of treatment [15, 16], and small lesions can heal spontaneously [17], suggesting that the immune system may be able to heal lesions once the majority of organisms are killed by antibiotics.
We aimed to assess, in those not undergoing surgery, whether 6 weeks compared with the currently recommended 8 weeks of combination antibiotic therapy was an effective treatment for small M. ulcerans lesions.
METHODS
Mycobacterium ulcerans cases from a prospective observational cohort at Barwon Health, Victoria, treated with antibiotics and without surgery from 1 October 2010 to 31 March 2018 were included in the study. Patients were retrospectively included in the 6-week antibiotic group if they received antibiotics for at least 28 days up to 42 days and were included in the 8-week antibiotic group if they received antibiotics for 56 days or more. Only World Health Organization (WHO) category 1 lesions were included. As it was an observational cohort, patients were not randomized into treatment groups; patients were included in the 6-week group either due to having their antibiotic treatment ceased earlier than planned due to an antibiotic complication or by electively having a 6-week course by joint agreement between the patient and treating clinician.
An M. ulcerans case was defined as the presence of a lesion clinically suggestive of M. ulcerans plus any of (1) a culture of M. ulcerans from the lesion, (2) a positive polymerase chain reaction (PCR) result from a swab or biopsy of the lesion, or (3) histopathology of an excised lesion showing a necrotic ulcer with the presence of acid-fast bacilli consistent with acute M. ulcerans infection. The initial size of the lesion was determined by measuring the induration diameter of lesions and calculating the surface area in millimeters squared. The WHO category was assigned according to published definitions, with category 1 being a single lesion less than 5 cm in diameter [3]. Prior to 2015, lesions were assigned a WHO category but exact induration size was not recorded. Wound healing was defined as complete re-epithelialization of the wound. Healing time was determined as the time from commencement of antibiotic treatment until full healing of the lesion.
Standard dosages for antibiotics used included rifampicin 10 mg/kg per day (up to a maximum of 600 mg daily), clarithromycin 7.5 mg/kg twice daily (up to 500 mg twice daily), and ciprofloxacin 500 mg twice daily. A complication of medical therapy was defined as an adverse event attributable to an antibiotic that required its cessation. Immune suppression was defined as current treatment with immunosuppressive medication (eg, prednisolone) or active malignancy.
Treatment cure was defined as complete healing of the lesion following treatment without the occurrence of a culture-positive M. ulcerans recurrent lesion appearing within 12 months of commencing antibiotic treatment. Paradoxical reactions were determined by the treating clinician and defined by the presence of 1 or both of the following features: (1) clinical, an initial improvement on antibiotic treatment in the clinical appearance of an M. ulcerans lesion followed by a clinically significant deterioration of the lesion or its surrounding tissues or the appearance of a new lesion(s); and (2) histopathology, examination of excised tissue from the clinical lesion showing evidence of an intense inflammatory reaction consistent with a paradoxical reaction [18].
Data Analysis
Data were collected prospectively using Epi-info 6 (Centers for Disease Control and Prevention) and analyzed using STATA 14 (StataCorp). Median values for nonparametric variables were compared using the Wilcoxon rank sum test.
Ethical Approval
This study was approved by the Barwon Health Human Research and Ethics Committee. All previously gathered human medical data were analyzed in a de-identified fashion.
RESULTS
A total of 207 patients fulfilled the inclusion criteria for the study and were included in the analysis: 53 (25.6%) in the 6-week group (Figure 1) and 154 (74.4%) in the 8-week group. The median age of patients was 53 years (interquartile range [IQR], 33–69 years) and 100 (48.3%) were female (Table 1). Of patients, 79.7% had lesions that were 900 mm2 or less in size and 93.2% of lesions were ulcerative. Only 6.8% of patients had diabetes and only 7.7% were immune suppressed. Initial antibiotic regimens used included rifampicin and clarithromycin in 66.7% and rifampicin and ciprofloxacin in 31.4% of cases, respectively. Patients in the short-course group were significantly more likely to be female (62.3% compared with 43.5%, P = .02) and have significantly smaller lesions at diagnosis (≤900 mm2 in size, 95.3% compared with 69.2%; P < .001).
Comparison of Baseline Characteristics for 6-Week and 8-Week Antibiotic Treatment Groups in the Barwon Health Cohort, 1 October 2010 to 31 March 2018
| Overall | 6-Week Group | 8-Week Group | P Value | |
|---|---|---|---|---|
| Gender | ||||
| Male | 107 (51.7) | 20 (37.7) | 87 (56.5) | .02 |
| Female | 100 (48.3) | 33 (62.3) | 67 (43.5) | |
| Age | ||||
| ≤15 y | 20 (9.7) | 4 (7.6) | 16 (10.4) | .72 |
| 16–64 y | 116 (56.0) | 32 (60.4) | 84 (54.6) | |
| ≥65 y | 71 (34.3) | 17 (32.8) | 54 (35.1) | |
| Median duration of symptoms prior to diagnosis, d | 42 (28–84) | 60 (28–90) | 42 (28–75) | .28 |
| Lesion size | ||||
| ≤ 400 mm2 | 45 (41.7) | 28 (65.1) | 17 (26.2) | <.001 |
| 401–900 mm2 | 41 (38.0) | 13 (30.2) | 28 (43.1) | |
| 901–1600 mm2 | 22 (20.4) | 2 (4.7) | 20 (30.8) | |
| Type of lesion | ||||
| Ulcer | 193 (93.2) | 51 (96.2) | 142 (92.2) | .57 |
| Nodule | 13 (6.3) | 2 (3.8) | 11 (7.1) | |
| Oedema | 1 (0.5) | 0 (0.0) | 1 0.7) | |
| Diabetes | ||||
| Yes | 14 (6.8) | 2 (3.8) | 12 (7.8) | .32 |
| No | 193 (93.2) | 51 (96.2) | 142 (92.2) | |
| Immune suppression | ||||
| Yes | 16 (7.7) | 2 (3.8) | 14 (9.1) | .21 |
| No | 191 (92.3) | 51 (96.2) | 140 (90.9) | |
| Initial antibiotic regimen | ||||
| Rifampicin/clarithromycin | 138 (66.7) | 39 (73.6) | 99 (64.3) | .18 |
| Rifampicin/ciprofloxacin | 65 (31.4) | 12 (22.6) | 53 (34.4) | |
| Other | 4 (1.9) | 2 (3.8) | 2 (1.3) |
| Overall | 6-Week Group | 8-Week Group | P Value | |
|---|---|---|---|---|
| Gender | ||||
| Male | 107 (51.7) | 20 (37.7) | 87 (56.5) | .02 |
| Female | 100 (48.3) | 33 (62.3) | 67 (43.5) | |
| Age | ||||
| ≤15 y | 20 (9.7) | 4 (7.6) | 16 (10.4) | .72 |
| 16–64 y | 116 (56.0) | 32 (60.4) | 84 (54.6) | |
| ≥65 y | 71 (34.3) | 17 (32.8) | 54 (35.1) | |
| Median duration of symptoms prior to diagnosis, d | 42 (28–84) | 60 (28–90) | 42 (28–75) | .28 |
| Lesion size | ||||
| ≤ 400 mm2 | 45 (41.7) | 28 (65.1) | 17 (26.2) | <.001 |
| 401–900 mm2 | 41 (38.0) | 13 (30.2) | 28 (43.1) | |
| 901–1600 mm2 | 22 (20.4) | 2 (4.7) | 20 (30.8) | |
| Type of lesion | ||||
| Ulcer | 193 (93.2) | 51 (96.2) | 142 (92.2) | .57 |
| Nodule | 13 (6.3) | 2 (3.8) | 11 (7.1) | |
| Oedema | 1 (0.5) | 0 (0.0) | 1 0.7) | |
| Diabetes | ||||
| Yes | 14 (6.8) | 2 (3.8) | 12 (7.8) | .32 |
| No | 193 (93.2) | 51 (96.2) | 142 (92.2) | |
| Immune suppression | ||||
| Yes | 16 (7.7) | 2 (3.8) | 14 (9.1) | .21 |
| No | 191 (92.3) | 51 (96.2) | 140 (90.9) | |
| Initial antibiotic regimen | ||||
| Rifampicin/clarithromycin | 138 (66.7) | 39 (73.6) | 99 (64.3) | .18 |
| Rifampicin/ciprofloxacin | 65 (31.4) | 12 (22.6) | 53 (34.4) | |
| Other | 4 (1.9) | 2 (3.8) | 2 (1.3) |
Data are presented as no. (%) unless otherwise indicated.
Comparison of Baseline Characteristics for 6-Week and 8-Week Antibiotic Treatment Groups in the Barwon Health Cohort, 1 October 2010 to 31 March 2018
| Overall | 6-Week Group | 8-Week Group | P Value | |
|---|---|---|---|---|
| Gender | ||||
| Male | 107 (51.7) | 20 (37.7) | 87 (56.5) | .02 |
| Female | 100 (48.3) | 33 (62.3) | 67 (43.5) | |
| Age | ||||
| ≤15 y | 20 (9.7) | 4 (7.6) | 16 (10.4) | .72 |
| 16–64 y | 116 (56.0) | 32 (60.4) | 84 (54.6) | |
| ≥65 y | 71 (34.3) | 17 (32.8) | 54 (35.1) | |
| Median duration of symptoms prior to diagnosis, d | 42 (28–84) | 60 (28–90) | 42 (28–75) | .28 |
| Lesion size | ||||
| ≤ 400 mm2 | 45 (41.7) | 28 (65.1) | 17 (26.2) | <.001 |
| 401–900 mm2 | 41 (38.0) | 13 (30.2) | 28 (43.1) | |
| 901–1600 mm2 | 22 (20.4) | 2 (4.7) | 20 (30.8) | |
| Type of lesion | ||||
| Ulcer | 193 (93.2) | 51 (96.2) | 142 (92.2) | .57 |
| Nodule | 13 (6.3) | 2 (3.8) | 11 (7.1) | |
| Oedema | 1 (0.5) | 0 (0.0) | 1 0.7) | |
| Diabetes | ||||
| Yes | 14 (6.8) | 2 (3.8) | 12 (7.8) | .32 |
| No | 193 (93.2) | 51 (96.2) | 142 (92.2) | |
| Immune suppression | ||||
| Yes | 16 (7.7) | 2 (3.8) | 14 (9.1) | .21 |
| No | 191 (92.3) | 51 (96.2) | 140 (90.9) | |
| Initial antibiotic regimen | ||||
| Rifampicin/clarithromycin | 138 (66.7) | 39 (73.6) | 99 (64.3) | .18 |
| Rifampicin/ciprofloxacin | 65 (31.4) | 12 (22.6) | 53 (34.4) | |
| Other | 4 (1.9) | 2 (3.8) | 2 (1.3) |
| Overall | 6-Week Group | 8-Week Group | P Value | |
|---|---|---|---|---|
| Gender | ||||
| Male | 107 (51.7) | 20 (37.7) | 87 (56.5) | .02 |
| Female | 100 (48.3) | 33 (62.3) | 67 (43.5) | |
| Age | ||||
| ≤15 y | 20 (9.7) | 4 (7.6) | 16 (10.4) | .72 |
| 16–64 y | 116 (56.0) | 32 (60.4) | 84 (54.6) | |
| ≥65 y | 71 (34.3) | 17 (32.8) | 54 (35.1) | |
| Median duration of symptoms prior to diagnosis, d | 42 (28–84) | 60 (28–90) | 42 (28–75) | .28 |
| Lesion size | ||||
| ≤ 400 mm2 | 45 (41.7) | 28 (65.1) | 17 (26.2) | <.001 |
| 401–900 mm2 | 41 (38.0) | 13 (30.2) | 28 (43.1) | |
| 901–1600 mm2 | 22 (20.4) | 2 (4.7) | 20 (30.8) | |
| Type of lesion | ||||
| Ulcer | 193 (93.2) | 51 (96.2) | 142 (92.2) | .57 |
| Nodule | 13 (6.3) | 2 (3.8) | 11 (7.1) | |
| Oedema | 1 (0.5) | 0 (0.0) | 1 0.7) | |
| Diabetes | ||||
| Yes | 14 (6.8) | 2 (3.8) | 12 (7.8) | .32 |
| No | 193 (93.2) | 51 (96.2) | 142 (92.2) | |
| Immune suppression | ||||
| Yes | 16 (7.7) | 2 (3.8) | 14 (9.1) | .21 |
| No | 191 (92.3) | 51 (96.2) | 140 (90.9) | |
| Initial antibiotic regimen | ||||
| Rifampicin/clarithromycin | 138 (66.7) | 39 (73.6) | 99 (64.3) | .18 |
| Rifampicin/ciprofloxacin | 65 (31.4) | 12 (22.6) | 53 (34.4) | |
| Other | 4 (1.9) | 2 (3.8) | 2 (1.3) |
Data are presented as no. (%) unless otherwise indicated.
Duration of antibiotics in those included in the 6-week treatment group in the Barwon Health cohort, 1 October 2010 to 31 March 2018.
Mycobacterium ulcerans infection was confirmed in 202 (97.6%) cases on the basis of a positive IS2404 PCR, 3 (1.4%) on positive histology, and 2 (1.0%) case on the basis of a positive M. ulcerans culture.
Median durations of antibiotics in the 6-week and 8-week groups were 42 days (IQR, 37–42 days) and 56 days (IQR, 56–56 days), respectively. In the short-course group, 41 (77.4%) patients had electively ceased antibiotics and 12 (22.6%) had treatment ceased due to antibiotic complications.
Fifty-three patients (100%) achieved treatment cure in the 6-week group compared with 153 (99.4%) in the 8-week group (P = .56). The patient who failed treatment in the 8-week group was a 17-month-old boy with a WHO category 1 nodular lesion who received 56 days of rifampicin and clarithromycin as previously described [19]. No patients died or were lost to follow-up. Median time to heal was 70 days (IQR, 60–96 days) in the 6-week group compared with 128 days (IQR, 95–173 days) in the 8-week group (P < .001). Two (3.8%) patients in the 6-week group experienced a paradoxical reaction compared with 39 (25.3%) patients in the 8-week group (P = .001). Paradoxical reactions occurred a median of 56 days (IQR, 36–80 days) after antibiotic treatment commenced. An example of an M. ulcerans lesion cured with 42 days of antibiotics can be seen in Figure 2.
Mycobacterium ulcerans lesion on inner thigh (size, 625 mm2) at (A) baseline and (B) following healing 2.5 months after 42 days of antibiotics.
DISCUSSION
The results from our study suggest that in Australian patients the antibiotic duration required to achieve cure of selected small M. ulcerans lesions without surgery may be effectively reduced from the currently recommended 8 weeks to 6 weeks. To obtain the highest level of evidence that accounts for the potential of unknown confounders, ideally these findings should be confirmed in a prospective randomized controlled study. However, we feel the 100% cure rate in more than 50 patients treated with 6 weeks of combined antibiotics provides very promising data to suggest that this treatment is effective for small M. ulcerans lesions. The resulting 25% reduction in antibiotic treatment duration has the potential to achieve substantial benefits to patients, including reductions in the toxicity, cost, and inconvenience of antibiotic treatment, and a reduction in the impact on the human microbiome that occurs with more prolonged antibiotic courses [6, 8]. Although our study results only apply to WHO category 1 lesions, this is relevant for ~80% of Australian patients at diagnosis [20], and thus the potential benefits of reducing treatment duration in this population are significant. However, the potential benefits in African populations where only approximately one-fifth of patients present with WHO category 1 lesions would be less pronounced [21].
The effectiveness of 6-week antibiotic regimens is supported by mouse model studies showing sterilization of footpad lesions after 4 weeks of oral rifampicin and clarithromycin regimens [13, 14]. There are also limited human data that show that 6 weeks of antibiotic therapy can sterilize lesions [9–12]. Further human data reveal that lesions may still heal despite organisms remaining viable at the end of treatment [15, 16], and we have recently demonstrated that small lesions can spontaneously heal without antibiotic or surgical treatment [17]. Therefore, it is likely that the immune system, through a robust T-helper type 1 (TH1) response [22], can overcome the immune-suppressive effects of mycolactone in treated or small lesions to achieve cure even if a small number of live organisms persist. Nevertheless, to our knowledge, this is the first study involving a large cohort treated without surgery that examines the effectiveness of 6 weeks of antibiotics in human M. ulcerans lesions.
Importantly, two-thirds of lesions where baseline size was available were 400 mm2 or less and 95% were 900 mm2 or less, meaning the strongest evidence for effectiveness was for the smallest lesions. Additionally, due to low numbers of patients in their respective groups, caution should be exercised in interpreting these results in children, those with nodular lesions, and those with immune suppression or diabetes. It is possible that in those with immature immune systems (children) or those with suboptimally functioning immune systems (immune suppression, diabetes) the host may not be as effective at supporting the cure of M. ulcerans disease. Therefore, it would be important to assess a larger number of patients before drawing any firm conclusions on the effectiveness of 6-week antibiotic courses in these groups. Furthermore, there were very few patients in the 6-week group who experienced paradoxical reactions, likely related to clinicians and patients preferring longer courses of antibiotics in those whose lesions clinically appeared to become worse on treatment. Therefore, caution needs to be exercised in using 6 weeks of treatment in this subpopulation. However, it seems unlikely that paradoxical reactions would increase the failure rate of 6-week antibiotic regimens as they likely result from a robust immune response to effective killing of organisms by antibiotic treatment [18], theoretically increasing their chance of treatment cure.
Our findings may not be generalizable to African patients, as small genetic differences between strains, including the potential of a more virulent mycolactone in African strains that may provide stronger inhibition of the host immune response [23], may reduce the effectiveness of 6-week antibiotic courses. In addition, African cohorts have a higher proportion of children, nonulcerative lesions, and immune suppression related to coinfection with human immunodeficiency virus [21, 24]. Therefore, 6-week regimens should be trialed in African patients to assess their effectiveness prior to advocating their use in these settings.
Conclusions
For selected small M. ulcerans lesions in Australian patients, 6 weeks may be as effective as the currently recommended 8 weeks’ duration of combined antibiotic therapy in curing lesions without surgery. Although this finding needs confirmation in prospective controlled trials, it offers the potential to reduce treatment toxicity and cost without compromising effectiveness.
Note
Potential conflicts of interest. The authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
