SEARCH

SEARCH BY CITATION

Keywords:

  • tungiasis;
  • clinico-bacteriological study;
  • superinfection;
  • aerobic pathogens;
  • anaerobic pathogens;
  • biofilm formation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Tungiasis is caused by penetration of the female sand flea Tunga penetrans into the epidermis of its host. It is endemic in many countries in Latin America, the Caribbean and sub-Saharan Africa. Although superinfection is a common clinical observation, the frequency and the pattern of bacterial pathogens associated with tungiasis have never been investigated systematically. We conducted a prospective clinico-bacteriological study with patients living in a shantytown in Fortaleza, capital of Ceará State (Northeast Brazil), where tungiasis is hyperendemic. Swabs were taken from 78 patients with multiple lesions after surgical extraction of the parasite, and the specimens were cultured for aerobic and anaerobic microorganisms. Ninety-nine specimens were investigated for aerobic bacteria, from which 146 pathogens were identified. The most common species were Staphyloccous aureus (35.5%) and various enterobacteriaceae (29.5%). Bacillus sp., Enteroccous faecalis, Streptococcus pyogenes and Pseudomonas sp. were also isolated. Eighty-four anaerobic cultures yielded 20 pathogens: in eight cases we detected Peptostreptococcus sp., in seven cases Clostridium sp., and in five cases non-identifiable gram-negative bacilli. These results show that secondary infection is very common in tungiasis, and caused by a variety of highly pathogenic microorganisms. It is proposed that T. penetrans acts as a foreign body facilitating biofilm formation within the epidermis. To prevent spreading of pathogens to the surrounding tissue and/or the systemic circulation, sand fleas should be surgically extracted immediately after penetration.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Tungiasis is an ectoparasitic disease caused by penetration of the female sand flea Tunga penetrans into the epidermis of its host. Besides humans, various domestic and sylvatic animals are affected (Heukelbach & Feldmeier 2001). Originally, the sand flea – also called jigger flea – only occurred on the South American continent and the Caribbean Islands, but has been inadvertently introduced into sub-Saharan Africa in the late 19th century (Hesse 1899; Hoeppli 1963). Today, tungiasis is endemic in many countries in Latin America, the Caribbean and sub-Saharan Africa. In Brazil, prevalence rates reach up to 40% in the squatter camps at the outskirts of the big cities as well as in the underdeveloped rural hinterland (Heukelbach et al. 2001).

Bacterial infection of the skin around embedded fleas is a common observation in the endemic area, particularly in children (Cardoso 1990). Severe secondary infections of tungiasis lesions require immediate antibiotic treatment as the spread of pathogenic bacteria may cause abscess formation, lymphangitis, tissue necrosis or gangrene (Connor 1976; Cardoso 1990; Chadee 1994; Bush et al. 2001; Melo & Melo 2001). Chadee (1998) observed sepsis in 16 of 268 patients with tungiasis at the feet or the hands. Furthermore, there are clinical and epidemiological indications that in populations with low vaccination coverage, untreated tungiasis is a risk factor for acquiring tetanus (Soria & Capri 1953; Obengui 1989; Tonge 1989; Litvoc et al. 1991; Takimoto et al. 1998).

In view of the clinical evidence that in severely affected populations tungiasis is commonly accompanied by superinfection, it is surprising that this disease complication and the pattern of bacterial pathogens involved have never been systematically investigated. Therefore, we decided to prospectively examine patients with tungiasis who were recruited at a Primary Health Care Centre (PHCC) next to a highly endemic area in a slum in Northeast Brazil, and to swab lesions for the identification of aerobic and anaerobic bacteria.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Study area

The study was performed in the favela Serviluz, a typical shantytown at the outskirts of Fortaleza, the capital of Ceará State, Northeast Brazil. The favela is close to the beach, where poor immigrants from the arid interior have been invading the dunes and constructing houses since the 1950s. The area has a total population of about 15 000. Ninety-seven per cent of the households are connected to electric power supply, and 70% have access to piped water. Sixty per cent of the population have a monthly family income of less than two minimum wages (1 minimum wage=US $80). Adult illiteracy is 30%, and crimes of all sorts are common.

Study population

During a period of 12 weeks all patients with tungiasis, who were recruited by a community health worker of the PHCC, were examined for the presence of tungiasis at the feet or at other parts of the body. Lesions were counted, staged from I to V according to a newly elaborated staging system (M. Eisele, manuscript in preparation) and photographed. Swabs were taken from lesions which clinically showed signs of superinfection (erythema, oedema, tenderness – with or without a bulla or pustule).

We included 78 patients in the study. In 13 patients severely affected by tungiasis, swabs were taken from several lesions to determine whether different pathogens were present in lesions at different topographic sites.

After disinfection of the lesion and the surrounding skin with 70% alcohol for 1 min, the flea was extracted with a sterile needle. For aerobic bacteria, a cotton swab wetted with sterile physiological saline was moved in and out of the remaining cavity. The swab was transferred to a sterile transportation tube which then was temporarily stored in a box with ice.

For the identification of anaerobic bacteria a dry sterile cotton swab was gently pressed into the cavity. The swab was immediately transferred into a special transportation tube filled with Cary and Blair PRAS medium for anaerobic pathogens (Summanem et al. 1993) and kept in a box at ambient temperature. Total transfer time to the bacteriology laboratory was not longer than 2 h. A total of 99 swabs for aerobic culture and 84 swabs for anaerobic culture were collected.

Pathogenic organisms were isolated according to standard microbiological techniques. For aerobic bacteria, the specimen was transferred to brain heart infusion agar supplemented with 5% sheep blood, and MacConkey agar. The specimens were also incubated in brain heart infusion broth. Identification of colonies was done by Gram staining and biochemical tests as described in Konemann et al. (1997). For anaerobic bacteria, the specimen was plated on Bacteroides–bile–esculin agar, phenylethyl alcohol sheep blood agar, brain heart infusion agar supplemented with 5% sheep blood, and brain heart infusion broth. All media were supplemented with hemin (5 μg/ml) and menadione (1 μg/ml) (Sigma Chemicals, São Paulo, Brazil). The cultures were incubated in anaerobic jars at 37 °C for 2–5 days. An anaerobic environment was obtained using a commercially available kit (DIFCO, São Paulo, Brazil). Colonies were identified by Gram staining and aerotolerance testing according to Summanem et al. (1993). Bacillus species were identified using the Vitek System (bioMerieux, São Paulo, Brazil). Because of technical and financial constraints, we could not look for coryneform gram-positive rods (Corynebacteria, Brevibacterium species and Actinomyces species), microorganisms which may be found in infected skin.

Six fleas encountered on the skin, either running or in statu penetrandi, were caught with a sterile forceps, squashed and transferred to a transportation tube for aerobic culture.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Demographic and clinical characteristics of the patients are summarized in Table 1. The median age of 11 years of our patients reflects the age distribution of tungiasis in the area. Most swabs (78%) were taken from lesions situated at the periungual area of the toes, the predilection site of T. penetrans.

Table 1.  Demographic and clinical characteristic of patients (n=78) Thumbnail image of

From the 99 aerobic specimens 146 pathogens were isolated (Table 2). Staphylococcus aureus was the predominant species (35.5%), followed by Enterobacteriaceae of seven species (29.5%). Bacillus species were identified in 6.2% of the isolates. Enterococcus faecalis and Pseudomonas species were isolated in two and five cases, respectively.

Table 2.  Results of aerobic cultures of 99 specimens Thumbnail image of

Seventy-six per cent of the 84 anaerobic specimens gave a negative result. In eight cases (40%), Peptostreptococcus species were identified, Clostridium in seven (35%), and non-identifiable gram-negative bacilli in five cases (25%) (Table 3). There was no relationship between the age of the patient and the pattern of pathogens isolated, nor between the stage of the lesion and microorganisms identified.

Table 3.  Results of anaerobic cultures of 84 specimens Thumbnail image of

To verify whether swabs taken from several lesions from the same patients yielded different results, patients were grouped according to the topographic sites swabbed (Table 4). Only when swabs were taken from lesions situated at the same toe was the probability to isolate an identical species high. The further apart the lesions were, the higher the percentage of diverging results. The considerable diversity of bacteriological isolates in a single patient is illustrated in Table 5. Four aerobic and one anaerobic pathogen species were cultured from five lesions.

Table 4.  Results of aerobic swabs according to topographic site swabbed (26 swabs from 13 patients) Thumbnail image of
Table 5.  Pathogens isolated from five different sites of patient N.S.S. (male, 50 years, 34 lesions) Thumbnail image of

Cultures from squashed fleas yielded unexpected results. Only coagulase-negative staphylococci (nine cultures) and Bacillus species (seven cultures) were isolated. Of the latter cultures, four yielded B. thuringiensis, two B. sphaericus and one Bacillus cereus.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Superinfection of lesions caused by T. penetrans is a common observation of clinicians working in areas where the ectoparasitosis is prevalent. However, the empirical evidence has never been substantiated by systematic bacteriological investigations. In fact, in the only textbook on tropical microbiology, tungiasis is not even mentioned (Montefiore et al. 1984). So far, only Chadee (1994, 1998) reported the isolation of K. pneumoniae in one patient, and of a variety of pathogens in 16 patients with signs of sepsis whose lesions were swabbed.

This is the first study in which patients with severe tungiasis were systematically examined clinically and bacteriologically. The results confirm that the skin around embedded fleas is infected with bacterial pathogens in most cases. Although we expected S. aureus to be the predominant species, the high frequency of other pathogens, particularly of the many enterobacteriaceae, came as a surprise.

Because of lack of hygiene and sewage disposal, the environment of the favela is expected to be highly contaminated with faecal material. It is therefore tempting to speculate that the high frequency of enterobacteriacae and other intestinal organisms in aerobic isolates reflects such contamination. Furthermore, as many inhabitants, particularly the children, walk barefoot, this should favour the colonization of the skin, especially of the feet, with such microorganisms. However, we are at a loss to explain the complete absence of Escherichia coli, the most common intestinal microorganism, in tungiasis lesions.

The isolation of Bacillus thuringiensis, B. sphaericcus and B. cereus in 6% of the isolates confirms a previous observation by Chadee (1998). These soil microorganisms, when found in traumatic wounds, are considered to be pathogens (Penn & Klotz 1998). As the pattern of bacillus species isolated from lesions of our patients was very similar to the pattern of these organisms cultured from free-living fleas, T. penetrans seems to be a natural host for these bacilli. This is of entomological interest as B. thuringiensis is highly toxic for a number of important disease-transmitting insects, including mosquitoes and blackflies (Curtis 1991; Berenbaum 1995). On the other hand, it cannot be excluded that despite disinfection with alcohol spores of Bacillus species remained viable on the skin around the lesions.

In comparison with aerobic cultures, anaerobic cultures rarely yielded positive results. This is not surprising in view of the erythema frequently surrounding a lesion, a clinical finding pointing to increased blood flow with a high oxygen concentration in the neighbouring tissue which, in turn, should inhibit the growth of anaerobic pathogens. In fact, in histological sections of lesions the presence of erythema correlated to vasodilatation and new formation of blood vessels in the dermis underneath the embedded fleas (M. Eisele, 2001, unpublished observation). An alternative explanation is the lower number of anaerobic compared with aerobic microorganisms on the skin or in the environment, which explains why wound infections are caused by aerobic pathogens in the vast majority of cases.

Although we did not isolate Clostridium tetani from the specimens, the presence of other clostridium species in the lesions of our patients clearly illustrates that T. penetrans may serve as an entrance for pathogenic anaerobic microorganisms. In fact, highly pathogenic species of clostridia (C. perfringens, C. botulinum and C. difficile) have been isolated from abscesses formed around foreign bodies in the skin (Brook 1995). Therefore, although only indirectly, these findings corroborate the clinical observation of an association between tungiasis and tetanus in individuals without tetanus vaccination (Soria & Capri 1953; Obengui 1989; Tonge 1989; Litvoc et al. 1991). Because of limited resources we were not able to look for atypical mycobacteriae and fungi such as Nocardia brasiliensis, microorganisms that also may cause abscesses in the skin after trauma with or without the implantation of a foreign body (Holland et al. 1997; Bhalodia et al. 1998).

The normal microflora of the skin develops a biofilm which is removed by the shedding of corneocytes and skin cleaning, for example, during washing or bathing (Larson 2001). Subsequently, new cells are colonized by biofilm-forming bacteria without giving rise to any symptoms. However, when the skin surface is connected by means of an implanted foreign body with underlying tissue compartments, translocation of the normal microflora easily occurs, eventually leading to acute infection and the formation of new biofilms on the foreign body (Hoiby et al. 1994). Although in this study we did not look for the existence of extracellular polymers produced by pathogens adhering to a submerged surface, circumstantial evidence makes us suggest that embedded tunga fleas should be considered as a foreign body with a high probability of biofilm formation.

First, during penetration the flea breaks up the stratum corneum allowing the spread of microcolonies present on the skin surface, which will result in contamination of the surrounding squamous cells. Alternatively, microorganisms on the outer surface of the flea may be carried into the epidermis. This assumption is corroborated by the observation that B. thuringiensis, B. sphaericus and B. cereus, rather uncommon microorganisms, were isolated in a similar pattern from tungiasis lesions and from fleas captured before penetration.

Secondly, during growth and maturation, the ectoparasite remains connected to the air through its rear end, an opening of 250–500 μm, through which it breathes, releases eggs and ejects faeces. Obviously, this permanent sore in the skin will facilitate the contamination of deeper layers of the epidermis with pathogenic microorganisms. As, immediately after penetration, tungiasis becomes very itchy, patients usually scratch their lesions, which, in turn, should favour the entry of bacteria through the existing opening of the epidermis.

Thirdly, shortly after penetration the piercing–sucking mouthparts of the ectoparasite bore through the lower layer of the epidermis and are placed into subepidermal blood vessels, thereby connecting the outer surface of the skin with the systemic circulation (Geigy & Herbig 1949). Fourthly, histopathological sections of biopsies taken from tungiasis lesions consistently showed the formation of microabscesses in stage II, i.e. 1–2 days after penetration (S. Sahebali, manuscript in preparation). Finally, as the continuously expanding body of the flea – the surface of which increases by the factor 150 in about 1 week – is covered by a rather smooth chitinous exoskeleton, the embedded flea becomes an inert structural matrix to which microorganisms could easily adhere (Donlan 2001).

Biofilms on foreign bodies situated in the skin frequently are formed by S. aureus, E. faecalis, Pseudomonas aeruginosa and enterobacteriaceae such as K. pneumoniae and Proteus mirabilis (Donlan 2001), pathogens isolated in 70% of our specimens. Furthermore, as foreign bodies aggravate the virulence of coagulase-negative staphylococci (CNS), the finding of CNS in 35% of the aerobic cultures supports the assumption that under particular circumstances these microorganisms may indeed be considered as true pathogens (Lapins et al. 1999).

If biofilm formation around a foreign body is indeed the pathophysiological basis for the development of severe bacterial infection in tungiasis lesions, this would explain why the infection persists until the death of the parasite and the complete rejection of its carcass from the skin (Costerton et al. 1999).

Taken together, our findings indicate that in severe tungiasis bacterial superinfection is extremely common. Therefore, we suggest the surgical extraction of fleas as soon as they have penetrated in order to prevent local and eventually systemic bacterial infections. Because of the considerable diversity of pathogenic bacteria occurring in the lesions, a broad spectrum antibiotic is required in patients in whom superinfection has already developed.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This study was supported in part by the Ärztekomittee für die Dritte Welt, Frankfurt (Germany), the German Brazilian Academic Cooperation Programme (PROBRAL project D/01/12357), Medosan, Schwarzenbach (Switzerland), Bayer Animal Health Division, Leverkusen-Monheim (Germany), VARIG Brazilian Airlines and Merck do Brazil, Rio de Janeiro (Brazil). We appreciate very much the fruitful discussions with Dr Rômulo César Sabóia Moura from the Primary Health Care Center Aída Santos e Silva, Fortaleza (Brazil). We thank Dr J. Ungeheuer for his many suggestions and his constructive criticism. The data are part of a medical thesis by Margit Eisele.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Berenbaum MR (1995) Bugs in the System. Insects and Their Impact on Human Affairs. Helix Books, Reading, p. 294.
  • 2
    Bhalodia AM, Lertzmann BH, Kantor GR & Granick MS (1998) Localized cutaneous Nocardia brasiliensis mimicking foreign body granuloma. Cutis 61, 161163.
  • 3
    Brook I (1995) Clostridial infection in children. Journal of Medical Microbiology 42, 7882.
  • 4
    Bush AO, Fernández JC, Esch GW & Seed JR (2001) Parasitism. The Diversity and Ecology of Animal Parasites. Cambridge University Press, Cambridge, pp. 265–268.
  • 5
    Cardoso AEC (1990) Tungiase. Anais Brasileira de Dermatologia 65, 29S33S.
  • 6
    Chadee DD (1994) Distribution patterns of Tunga penetrans within a community in Trinidad, West Indies. Journal of Tropical Medicine and Hygiene 97, 167170.
  • 7
    Chadee DD (1998) Tungiasis among five communities in south-western Trinidad, West Indies. Annals of Tropical Medicine and Parasitology 92, 107113.
  • 8
    Connor DH (1976) Tungiasis. In: Pathology of Tropical Extraordinary Diseases (eds CH Binford & DH Connor) Armed Forces Institute of Pathology, Washington, pp. 610–614.
  • 9
    Costerton JW, Stewart PS & Greenberg EP (1999) Bacterial biofilms: a common cause of persistent infection. Science 284, 13181322.DOI: 10.1126/science.284.5418.1318
  • 10
    Curtis CF (1991) Control of Disease Vectors in the Community. Wolfe, London, pp. 140–141.
  • 11
    Donlan RM (2001) Biofilms and device-associated infections. Emerging Infectious Diseases 7, 277281.
  • 12
    Geigy R & Herbig A (1949) Die Hypertrophie der Organe beim Weibchen von Tunga penetrans. Acta Tropica 6, 246262.
  • 13
    Hesse P (1899) Die ausbreitung des sandflohs in Afrika. Geographische Zeitschrift (Hettner) 522530.
  • 14
    Heukelbach J & Feldmeier H (2002) Tungiasis in poor communities of Northeast Brazil: an important public health problem. Public Health, in press.
  • 15
    Heukelbach J, Oliveira FAS, Hesse G & Feldmeier H (2001) Tungiasis: a neglected health problem of poor communities. Tropical Medicine and International Health 6, 267272.
  • 16
    Hoeppli R (1963) Early references to the occurrence of Tunga penetrans in Tropical Africa. Acta Tropica 20, 143152.
  • 17
    Hoiby N, Espersen F, Fomsgaard A et al. (1994) Biofilm, foreign bodies and chronic infections. Ugeskr Laeger 156, 59986005.
  • 18
    Holland J, Smith C, Childs PA & Holland AJ (1997) Surgical management of cutaneous infection caused by atypical mycobacteria after penetrating injury: the hidden dangers of horticulture. Journal of Trauma 24, 337340.
  • 19
    Konemann WK, Allen SD, Janda WM, Schreckenberger PC & Winn WC (1997) Color Atlas and Textbook of Diagnostic Microbiology. Lippincott, New York.
  • 20
    Lapins J, Jarstrand C & Emtestam L (1999) Coagulase-negative staphylococci are the most common bacteria found in cultures from the deep portions of hidradenitis suppurativa lesions, as obtained by carbon dioxide laser surgery. British Journal of Dermatology 140, 9095.DOI: 10.1046/j.1365-2133.1999.02613.x
  • 21
    Larson E (2001) Hygiene of the skin: when is clean too clean? Emerging Infectious Diseases 7, 225230.
  • 22
    Litvoc J, Leite RM & Katz G (1991) Aspectos epidemiológicos do tétano no estado do São Paulo (Brasil). Revista Do Instituto de Medicina Tropical de São Paulo 33, 477484.
  • 23
    Melo CR & Melo IS (2001) Linfedema elefantoíde verrucoso associado a infestacao macica por Tunga penetrans. Anais Brasileira de Dermatologia 64, 3537.
  • 24
    Montefiore DG, Alausa KO & Tomori O (1984) Medicine in the Tropics: Tropical Microbiology. Churchill Livingstone, Edinburgh.
  • 25
    Obengui (1989) La tungose et le tétanos au C.H.U. de Brazzaville. Dakar Médical 34, 4448.
  • 26
    Penn CC & Klotz SA (1998) Bacillus anthracis and other aerobic spore formers. In: Infectious Diseases, 2nd edn (eds LG Gorbach, JG Bartlett & NR Blachlow) W.B. Saunders, Philadelphia, pp. 1747–1750.
  • 27
    Soria MF & Capri JJ (1953) Tetanos y `pique'. La Prensa Medica Argentina 40, 411.
  • 28
    Summanem P, Baron EJ, Citron DM et al. (1993) Wadsworth Anaerobic Bacteriology Manual, Star Publishing Co., Belmont, California.
  • 29
    Takimoto S, Waldman EA, Moreira RC et al. (1998) Enterovirus 71 infection and acute neurological disease among children in Brazil (1988–1990). Transactions of the Royal Society of Tropical Medicine and Hygiene 92, 2528.
  • 30
    Tonge BL (1989) Tetanus from chigger flea sores. Journal of Tropical Pediatrics 35, 9494.