|Year : 2017 | Volume
| Issue : 1 | Page : 65-70
Longitudinal assessment of the bacterial burden of buruli ulcer wounds during treatment
Grace Kpeli1, Evelyn Owusu-Mireku2, Julia Hauser3, Gerd Pluschke3, Dorothy Yeboah-Manu2
1 Department of Bacteriology, Noguchi Memorial Institute for Medical Research, Legon, Accra, Ghana; Department of Molecular Parasitology and Immunology, Swiss Tropical and Public Health Institute, Socinstrasse 57, 4051; Department of Molecular Parasitology and Immunology, University of Basel, Petersplatz 1, 4003 Basel, Switzerland
2 Department of Bacteriology, Noguchi Memorial Institute for Medical Research, Legon, Accra, Ghana
3 Department of Molecular Parasitology and Immunology, Swiss Tropical and Public Health Institute, Socinstrasse 57, 4051; Department of Molecular Parasitology and Immunology, University of Basel, Petersplatz 1, 4003 Basel, Switzerland
|Date of Web Publication||24-Jul-2017|
Noguchi Memorial Institute for Medical Research, P. O. Box LG 581, Legon, Accra
Source of Support: None, Conflict of Interest: None
Background: Bacterial contamination is common to all wounds. The bacterial burden of wounds has been found to have an inverse relationship with chronic wound healing. In seeking to develop a better understanding of the evolution of Buruli ulcer (BU) wounds, we performed a longitudinal study to quantify the bacterial burden of BU wounds during the course of streptomycin/rifampicin (SR) treatment. Methods: Twenty-one IS2404 polymerase chain reaction confirmed patients were longitudinally followed during the course of their treatment. Swab or tissue samples obtained from the lesions were quantitatively analyzed to determine the bacterial burden pre-, during, and post-SR treatment. Furthermore, the species of bacterial isolates obtained at these time points were also identified. Results: Based on the determination of the bacterial burden, 18/22 (81.8%) pretreatment, 15/25 (57.7%) during treatment, and 36/48 (75.0%) posttreatment samples were classified as superinfected, respectively. Thirty bacterial species including two species of anaerobic Clostridia (Clostridium perfringens and Clostridium sporogenes) were identified among 114 isolates. While Enterococcus faecalis, Pseudomonas aeruginosa, and Chryseomonas luteola dominated pretreatment, P. aeruginosa dominated during and posttreatment. Conclusions: Most BU patients presented with lesions with a high bacterial load which increased significantly posttreatment. Therefore, good wound care is necessary to control the microbial burden of BU wounds, especially posttreatment to minimize complications.
Keywords: Buruli ulcer, infection, longitudinal study, Pseudomonas aeruginosa, Staphylococcus aureus
|How to cite this article:|
Kpeli G, Owusu-Mireku E, Hauser J, Pluschke G, Yeboah-Manu D. Longitudinal assessment of the bacterial burden of buruli ulcer wounds during treatment. Biomed Biotechnol Res J 2017;1:65-70
|How to cite this URL:|
Kpeli G, Owusu-Mireku E, Hauser J, Pluschke G, Yeboah-Manu D. Longitudinal assessment of the bacterial burden of buruli ulcer wounds during treatment. Biomed Biotechnol Res J [serial online] 2017 [cited 2019 Aug 22];1:65-70. Available from: http://www.bmbtrj.org/text.asp?2017/1/1/65/211414
| Introduction|| |
Buruli ulcer (BU), a chronic debilitating disease caused by Mycobacterium ulcerans, affects mainly the soft tissues of the skin. If nonulcerative stages of the disease (nodules, papules, edema, or plaques) are not treated, extensive tissue destruction by the macrolide toxin mycolactone can lead to large rugged ulcers, which are the hallmark of BU. The first-line treatment for BU, daily administration of oral rifampicin (10 mg/kg) and intramuscular streptomycin (15 mg/kg) for 8 weeks (SR8), was introduced by the World Health Organization (WHO) in 2004., Early stages, especially nodules and papules, heal in most cases shortly after completion of the SR8 treatment without the need for adjunct surgical interventions. Large ulcerative lesions, however, can take as long as a year or more to heal and bacterial superinfection may occur if wound care is not optimal.
Localized wound infection has been identified as a significant cause of impaired healing and wound chronicity. Wound infection occurs when the growth of microorganisms within the wound is uncontrolled by host defense mechanisms and can lead to deeper and more severe pathology and sepsis. Through biofilm formation, the pathogenic effects of bacteria may be increased. Endotoxin release by Gram-negative bacteria in wounds leads to elevated levels of proinflammatory cytokines (IL-1 and tumor necrosis factor). In addition, factors including the release of free radicals, degradation of growth factors, production of metabolic products, consumption of local oxygen, and interference with collagen formation may result in a nonconducive wound environment. With high bacterial loads, the effects of these mechanisms will be increased, thereby leading to impaired wound healing.
Diagnosing wound infection is very challenging and optimally requires clinical signs and symptoms as well as quantitative and qualitative microbiological investigations involving direct microscopy and cultures. Bacterial loads above 105 CFU/g are the accepted gold standard in diagnosing localized infection.,, The quantity of bacteria in wounds has been found to have an inverse relationship with the healing of chronic wounds,,, with studies showing that wound healing progresses only when bacterial counts are below 106 CFU/ml.,,
A previous cross-sectional study identified wound infection as a probable cause of healing delay. The study reported a good correlation between clinically suspected infected lesions and results of microbiological cultures. In seeking to have a better understanding of the evolution of the wounds, we conducted a longitudinal study to quantify the bacterial burden of BU wounds during the course of SR8 treatment.
| Methods|| |
Informed consent was obtained from the patients at their first hospital attendance after the objectives of the project had been explained to them in a language they understand. In the case of minors, parental consent and child assent was obtained before the child was recruited into the study. Written informed consent in the form of a signature or thumbprint was given by all consenting participants. Sampling and confirmation of BU followed the approved WHO recommendations.
Study participants and clinical presentation of lesions
The study was conducted at the Ga West Municipal Hospital (GWMH) in the Greater Accra Region of Ghana; one of the main BU treatment facilities in the country. Participants were recruited into the study after the clinical BU diagnosis had been confirmed by at least IS2404 polymerase chain reaction (PCR) before commencement of SR8 treatment. Twenty-one patients were recruited into the study out of which 17 were inpatients while four were outpatients who took the treatment at other health centers but reported at the GWMH periodically for review. The male-to-female ratio was 8:13. Their ages ranged from 5 to 69 years with a mean of 35 ± 18, and their weights ranged from 18 to 115 kg. At their first hospital attendance, 14 patients presented with single lesions, seven presented with multiple lesions, three of whom presented with an edema and an ulcer, two presented with two ulcers, and two others presented with a nodule, plaque, and ulcer. The lesions of two were classified as Category I lesions (size of <5 cm at the widest diameter), four as Category II lesions (size between 5 and 15 cm at the widest diameter), and 15 as Category III lesions (size >15 cm at the widest diameter or multiple lesions). Eighteen of the lesions were found on the lower limbs and three on the upper limbs. Of the lesions on the lower limbs, 11 were found on the leg, four on the ankle, and three on the foot, while on the upper limb, two were found on the arm and one on the elbow. Three patients had other comorbidities, specifically, HIV, hypertension, and diabetes mellitus. Out of the 21 cases, 17 took the streptomycin/rifampicin (SR) treatment for 8 weeks and three took the treatment for 12 weeks because the treating clinician suspected the lesions to still be active after 8 weeks of treatment. One patient who had previously had BU and had undergone the 8 weeks of treatment 1 year before was prescribed a 4-week treatment course after a new lesion appeared at another location. During the course of treatment, one patient, who was positive for HIV, developed disseminated lesions on other parts of the lower limb.
The wounds were clinically assessed for signs of infection using a wound assessment chart. The clinical features assessed were the appearance of the wound (necrotic, granulation, epithelialization, and slough), the wound exudate level (low, high, medium, serous, serosanguineous, purulent, malodorous), and the appearance of the surrounding skin (macerated, edematous, erythematous, indurated, dry scaling, healthy/intact). Category II and III lesions were assessed at 2-week intervals while Category I lesions were assessed at weekly intervals.
Samples were taken from the lesions of patients for microbiological analysis to determine the bacterial burden and were also cultured to isolate and identify infecting bacteria, as previously described. During the course of SR treatment, the bacterial burden of lesions was investigated biweekly. Post-SR treatment, wound cultures were done upon advice from the responsible clinician when the lesion was suspected to be infected. Microbiological assessments were also made when patients underwent excision and debridement. A total of 96 samples were collected from the patients. These included 85 swabs and 11 tissue samples. Swab samples were collected by the Levine method pre-, during, and posttreatment from the undermined edges of lesions while tissue samples were collected posttreatment after patients had undergone surgical procedures. Two swabs were taken from each lesion; one was placed in 5 ml phosphate-buffered saline for enumeration of the bacterial burden and isolation of aerobic infecting bacteria and the second was inserted into Robertson's cooked meat media for the isolation of anaerobic bacteria. Two tissue samples were taken from patients; one was aseptically transferred into Robertson's cooked meat media while the second was transferred into sterile transport containers. The samples were transported cold from the hospital to the laboratory for processing.
Quantitative and qualitative cultures
Quantitative cultures were carried out by the pour plate method, as previously described. Values obtained from the enumeration of bacterial colonies were computed as colony-forming units per ml (CFU/ml) for swabs and colony-forming units per gram (CFU/g) for tissues. Wounds were classified as infected if bacterial counts of >106 CFU/g or ml were obtained, contaminated if bacterial counts were <106 CFU/g or ml and clean if no bacterial colonies grew on the agar plates. Comparison between bacterial loads was done by t- tests and one-way ANOVA calculations in SPSS v 24.0 (IBM Corp, Armonk, New York, United States).
For the isolation of infecting bacterial species, 10−1 dilutions of the sample suspensions were pelleted and the sediments cultured on Blood, MacConkey, and Mannitol Salt agars (Oxoid Ltd, Basingstoke, UK). Biochemical oxidase, catalase, and coagulase tests were performed to presumptively identify bacteria species. Further characterization of Gram-negative rods was done using analytical profile index (API20E) strips (bio-Merieux SA, Marcy-l'E'toile, France). The Staphylase Kit, Prolex latex agglutination system (Pro-Lab Diagnostics, Birkenhead, Bromborough, United Kingdom), was used to differentiate the catalase-positive Gram-positive bacteria, Staphylococcus aureus from other Staphylococcus species.
The Robertson's cooked meat medium containing the sample was incubated anaerobically in a glass jar with an anaerobic gas pack at 37°C for 48–73 h. Tubes with signs of growth were subcultured on blood agar and incubated anaerobically with a metronidazole antibiotic disc to check for resistance. The gram reaction of positive cultures was determined, and identification of isolated microorganisms by MALDI-TOF mass spectrometry was outsourced to Mabritec AG, Riehen, Switzerland.
| Results|| |
Clinical assessment of wounds
Pretreatment, the majority of the lesions had clinical indications of infection, seemed to be in the proliferative stage of wound healing by having granulation tissue and epithelialization, had undermined edges, and high wound exudates [Table 1]. Changes in clinical presentation during treatment are shown in [Table 1]. Posttreatment, all 15 assessed lesions presented with some granulation tissue and slough, 4/15 (26.6%) had necrotic tissues, 3/15 (20%) presented with epithelialization, while the edges of 5/15 (33.3%) were still undermined [Table 1].
Bacterial burden and assessment of infection
The bacterial burden of 22 lesions from twenty patients was determined pre-SR. The bacterial load ranged from 0 to 3.00 × 109 CFU/ml, with a mean of 2.79 × 109 CFU/ml and a median of 8.05 × 106 CFU/ml. Two (9.1%), 2 (9.1%), and 18 (81.8%) samples were classified microbiologically as clean, contaminated, and infected, respectively [Figure 1].
|Figure 1: Bacterial loads at different time points. Bacterial loads obtained at different stages of treatment classified as clean, contaminated, and infected. Blue-colored: Loads obtained pretreatment, red-colored: Loads obtained during treatment, green-colored: Loads obtained posttreatment|
Click here to view
During treatment, 26 samples were collected from the lesions of twenty patients (four patients were sampled twice and two patients had two lesions each). The bacterial load of the samples ranged from 0 to 2.06 × 109 CFU/ml, with a mean of 1.5 × 108 CFU/ml and a median of 1.13 × 106 CFU/ml. Two (7.7%) samples were clean, 8 (30.8%) were contaminated, and 15 (57.7%) were infected [Figure 1]. The load of one of the samples could not be determined as the culture got contaminated.
Posttreatment, 48 samples were collected between weeks 8 and 75 from 25 lesions of 20 patients. Ten lesions were sampled once, nine lesions twice, four lesions thrice, and two lesions four times. The bacterial load of the samples ranged from 0 to 3.00 × 109 CFU/ml, with a mean of 3.49 × 108 CFU/ml and a median of 1.69 × 108 CFU/ml. One (2.1%), 11 (22.9%), and 36 (75.0%) samples were classified as clean, contaminated, and infected, respectively [Figure 1].
Bacterial loads were obtained for 17 patients (18 lesions) at all three stages. During treatment, the bacterial load of 12 lesions decreased and that of six lesions increased compared to the values obtained pretreatment. Posttreatment, the loads of 16 lesions increased from the values obtained during treatment while that of two lesions decreased. Overall, most patients presented with bacterial loads which reduced during treatment but increased dramatically posttreatment [Figure 2]. The difference between the bacterial loads at all three time points was significant (P < 0.05). Pairwise comparison of the bacterial loads, however, showed that a statistically significant difference existed only between the loads during and posttreatment.
|Figure 2: Bacterial loads of individual patients. Bacterial loads of 17 patients at all three stages showing the evolution of the bacterial burden|
Click here to view
Isolated bacterial flora
One hundred and fourteen bacterial isolates were recovered from qualitative cultures made up of 28 different species of aerobic bacteria and two species of Clostridia (Clostridium sporogenes and Clostridium perfringens). At pretreatment, 14 distinct species were identified among 28 isolates, dominated by Enterococcus faecalis (5/28, 17.8%), Pseudomonas aeruginosa (4/28, 14.3%), and Chryseomonas luteola (4/28, 14.3%) [Table 2]. Twelve species were identified among 29 isolates recovered during treatment and the main species was P. aeruginosa (12/29, 41.4%). Fifty-seven bacterial isolates were recovered posttreatment, out of which 20 distinct bacterial species, dominated by P. aeruginosa (16/57, 28.1%), were identified [Table 2].
Post-SR8 treatment follow-up
After antibiotic treatment, 12 of the patients underwent surgical interventions. Three of them had excisions only while nine patients underwent excision and skin grafting. The grafts of three of the patients failed and two had a second skin grafting. The lesions of 13 patients healed with an average healing time of 27–28 weeks' posttreatment. The rest of the patients were still dressing their wounds at the time of preparation of this manuscript, and the time of dressing ranged at that time between 4 weeks and 75 weeks' posttreatment.
| Discussion|| |
We investigated the bacterial burden of PCR confirmed BU wounds to gain an understanding into the evolution of microbial flora in BU wounds during treatment, which could have implications for the healing process. Our findings show that bacterial loads and bacterial diversity are high before SR treatment, decrease during treatment, and increase dramatically posttreatment. About thirty distinct bacterial species were found colonizing BU wounds with Gram-negative rods dominating.
We previously established that BU wounds could be infected by bacterial pathogens contrary to formerly held beliefs that they were sterile as a result of the presence of mycolactone. However, it has recently been demonstrated that mycolactone is inactive against bacterial species such as Streptococcus pneumoniae, Neisseria meningitidis, Escherichia coli as well as the yeast Saccharomyces cerevisiae and ameba Dictyostelium discoideum. The results obtained from this study confirm our previous finding and findings from others, on secondary infection of BU wounds.
A decrease in bacterial load during treatment compared to pre- and post-treatment was observed. This appears to be due to administration of the broad-spectrum antibiotics within that period. A study by Gardner et al. also reported decreased bacterial loads in study participants who were on systemic antibiotics during the time of study, compared to those not on antibiotics. This result was also in good agreement with observed clinical signs, such as fewer lesions presenting with slough, necrosis, and high wound exudate production.
The bacterial species isolated from the BU wounds are representative of the spectrum of bacterial species usually isolated from chronic wounds. Colonization of most wounds by Gram-negative bacteria indicates that they were at an advanced stage of bacterial colonization since early colonizing bacteria in chronic wounds are mainly Gram-positive organisms, notably staphylococci, and beta-hemolytic Streptococci, which are replaced by Gram-negative organisms as the wound microbiota becomes more complex. Many BU patients report late to the formal health centers for treatment. The period between the time of ulceration and the time of first presentation thus provides contaminating bacteria with adequate time to multiply and establish a complex microbiota accounting for the high number of Gram-negatives isolated from the lesions.
P. aeruginosa was dominant at all time points. This organism is detected in wounds during their later natural history when a complex microbial flora is being formed and is also frequently cited as a source of wound healing delay and wound infection.,, Two anaerobes were isolated from the lesions of two patients' pre- and during treatment. The presence of anaerobic bacteria indicates that the wounds are at a stage of irreversible adhesive colonization as anaerobes are only known to colonize lesions after a hypoxic state has been created due to increased multiplication of the wound microflora. Bacteria at this stage also develop inherent resistance to antibiotics which could impact on wound healing. Previous work at the study health center has shown high levels of antibiotic resistance among bacteria isolated from lesions of hospitalized BU patients, including methicillin-resistant S. aureus and multidrug-resistant P. aeruginosa, Most of the isolated bacteria are known nosocomial pathogens, and since BU treatment is associated with long hospital stays, BU patients are at high risk for the acquisition of these pathogens in the health centers.
A potential limitation of this study is that it did not formally assess the impact of the bacterial burden on wound healing outcome. However, various studies have shown that an increased bacterial load negatively impacts wound healing, and the presence of bacteria in wounds even in the absence of obvious clinical signs can inhibit the normal wound healing process. Therefore, BU wound management guidelines should consider the bacterial load and entail strategies for decreasing bacterial load at all time points during the course of the infection.
We are grateful to the following collaborators and health centers for their involvement in the study: Dr Albert Paintsil, Korle-Bu Teaching Hospital; Dr Joseph Tuffour, Ga West Municipal Hospital, Amasaman, the Nurses on the Buruli Ward of the Ga West Municipal Hospital, and the patients who trusted us and agreed to work with us on this study.
Financial support and sponsorship
This work was supported by the Stop Buruli Initiative of the UBS Optimus foundation and the Volkswagen Foundation. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Conflicts of interest
There are no conflicts of interest.
| References|| |
Debacker M, Aguiar J, Steunou C, Zinsou C, Meyers WM, Guédénon A, et al. Mycobacterium ulcerans
disease (Buruli ulcer) in rural hospital, Southern Benin, 1997-2001. Emerg Infect Dis 2004;10:1391-8.
WHO. Provisional Guidance on the Role of Specific Antibiotics in the Management of Mycobacterium Ulcerans
Disease (Buruli ulcer); 2004.
Converse PJ, Nuermberger EL, Almeida DV, Grosset JH. Treating Mycobacterium ulcerans
disease (Buruli ulcer): From surgery to antibiotics, is the pill mightier than the knife? Future Microbiol 2011;6:1185-98.
Tarnuzzer RW, Schultz GS. Biochemical analysis of acute and chronic wound environments. Wound Repair Regen 1996;4:321-5.
Robson MC, Heggers JP. Bacterial quantification of open wounds. Mil Med 1969;134:19-24.
Gardner SE, Frantz RA, Doebbeling BN. The validity of the clinical signs and symptoms used to identify localized chronic wound infection. Wound Repair Regen 2001;9:178-86.
Schraibman IG. The significance of beta-haemolytic streptococci in chronic leg ulcers. Ann R Coll Surg Engl 1990;72:123-4.
Lookingbill DP, Miller SH, Knowles RC. Bacteriology of chronic leg ulcers. Arch Dermatol 1978;114:1765-8.
Daltrey DC, Rhodes B, Chattwood JG. Investigation into the microbial flora of healing and non-healing decubitus ulcers. J Clin Pathol 1981;34:701-5.
Edwards R, Harding KG. Bacteria and wound healing. Curr Opin Infect Dis 2004;17:91-6.
Bendy RH Jr., Nuccio PA, Wolfe E, Collins B, Tamburro C, Glass W, et al
. Relationship of quantitative wound bacterial counts to healing of decubiti: Effect of topical Gentamicin. Antimicrob Agents Chemother (Bethesda) 1964;10:147-55.
Pruitt BA Jr., McManus AT, Kim SH, Goodwin CW. Burn wound infections: Current status. World J Surg 1998;22:135-45.
Krizek TJ, Robson MC, Kho E. Bacterial growth and skin graft survival. Surg Forum 1967;18:518-9.
Yeboah-Manu D, Kpeli GS, Ruf MT, Asan-Ampah K, Quenin-Fosu K, Owusu-Mireku E, et al
. Secondary bacterial infections of Buruli ulcer lesions before and after chemotherapy with streptomycin and rifampicin. PLoS Negl Trop Dis 2013;7:e2191.
WHO. Buruli Ulcer-Diagnosis of Mycobacterium Ulcerans
Disease. Global Buruli ulcer Initiative, Communicable Diseases, World Health Organization, 1211 Geneva 27-Switzerland; 2016.
Levine NS, Lindberg RB, Mason AD Jr., Pruitt BA Jr. The quantitative swab culture and smear: A quick, simple method for determining the number of viable aerobic bacteria on open wounds. J Trauma 1976;16:89-94.
IBM Corp. IBM SPSS Statistics for Windows. Version 24.0. Armonk, NY: IBM Corp; 2016.
Scherr N, Gersbach P, Dangy JP, Bomio C, Li J, Altmann KH, et al
. Structure-activity relationship studies on the macrolide exotoxin mycolactone of Mycobacterium ulcerans
. PLoS Negl Trop Dis 2013;7:e2143.
Komenan K, Elidjé EJ, Ildevert GP, Yao KI, Kanga K, Kouamé KA, et al
. Multifocal Buruli ulcer associated with secondary infection in HIV positive patient. Case Rep Med 2013;2013:348628.
Barogui YT, Klis S, Bankolé HS, Sopoh GE, Mamo S, Baba-Moussa L, et al
. Towards rational use of antibiotics for suspected secondary infections in Buruli ulcer patients. PLoS Negl Trop Dis 2013;7:e2010.
Gardner SE, Hillis SL, Frantz RA. Clinical signs of infection in diabetic foot ulcers with high microbial load. Biol Res Nurs 2009;11:119-28.
Bowler PG, Duerden BI, Armstrong DG. Wound microbiology and associated approaches to wound management. Clin Microbiol Rev 2001;14:244-69.
Percival SL, Dowd SE. The Microrbiology of Wounds. Percival SL, Cutting K, editors. Boca Raton FL: CFC Press: 2010.
Ahorlu CK, Koka E, Yeboah-Manu D, Lamptey I, Ampadu E. Enhancing Buruli ulcer control in Ghana through social interventions: A case study from the Obom sub-district. BMC Public Health 2013;13:59.
Heggers JP. Defining infection in chronic wounds: Methodology. J Wound Care 1998;7:452-6.
Danielsen L, Balslev E, Döring G, Høiby N, Madsen SM, Agren M, et al
. Ulcer bed infection. Report of a case of enlarging venous leg ulcer colonized by Pseudomonas aeruginosa
. APMIS 1998;106:721-6.
Twum-Danso K, Grant C, al-Suleiman SA, Abdel-Khader S, al-Awami MS, al-Breiki H, et al
. Microbiology of postoperative wound infection: A prospective study of 1770 wounds. J Hosp Infect 1992;21:29-37.
Kpeli G, Darko Otchere I, Lamelas A, Buultjens AL, Bulach D, Baines SL, et al
. Possible healthcare-associated transmission as a cause of secondary infection and population structure of Staphylococcus aureus
isolates from two wound treatment centres in Ghana. New Microbes New Infect 2016;13:92-101.
Grietens KP, Boock AU, Peeters H, Hausmann-Muela S, Toomer E, Ribera JM. “It is me who endures but my family that suffers”: Social isolation as a consequence of the household cost burden of Buruli ulcer free of charge hospital treatment. PLoS Negl Trop Dis 2008;2:e321.
[Figure 1], [Figure 2]
[Table 1], [Table 2]