Corynebacterium diphtheriae and the Returned Tropical Traveler

Authors


Abstract

Background

In Western countries, nontoxigenic Corynebacterium diphtheriae is known to cause skin and soft tissue infections (SSIs), upper respiratory tract infections, and occasionally invasive disease. Its role as a skin pathogen in returned travelers from tropical destinations where the organism is endemic is often forgotten. A retrospective analysis of a large Australian private pathology laboratory's experience with C. diphtheriae was performed to identify how frequently overseas travel was associated with C. diptheriae infection/colonization.

Methods

All C. diphtheriae isolates cultured from 2002 to 2012 were reviewed. Recorded clinical information regarding recent travel, country, and cause of infection was assessed. Antibiotic susceptibility was verified on all isolates.

Results

In all there were 72 patients who had C. diphtheriae isolated on clinical specimens, and information about prior travel was available for 63. Seventy percent of these were healthy individuals with an SSI and history of recent travel to a tropical nation. Ninety-seven percent had associated copathogens. Two isolates were penicillin resistant. There was uniform susceptibility to cephalothin, clindamycin, erythromycin, and vancomycin, with 14% resistance to trimethoprim/sulfamethoxazole and 4% resistance to tetracycline. Only one isolate was a toxigenic strain.

Conclusion

The majority of C. diphtheriae isolated were from SSIs in otherwise healthy travelers returning from tropical destinations, rather than classical risk groups. Clinicians and laboratories need to be aware of this potential source of C. diphtheriae infection due to rare toxigenic strains.

The clinical syndromes of respiratory and cutaneous diphtheria, caused by toxin-producing C. dipththeriae, are now rare in Western countries, due mainly to high immunization rates in the general population. Nontoxigenic C. diphtheriae remain uncommon clinical isolates, partly as a result of decreased expertise, as most laboratories no longer screen routinely for C. diphtheriae.[1] Nontoxigenic C. diphtheriae typically cause chronic skin ulceration, upper respiratory tract infections, or rarely invasive diseases (eg, endocarditis and septic arthritis).[2] Those groups previously identified in the literature with nontoxigenic C. diphtheriae infections have mostly been individuals with underlying medical conditions or marginalized social groups, including the Australian aboriginal population, and the urban poor of Vancouver.[2, 3] This retrospective, observational case series highlights the continued presence of C. diphtheriae in skin lesions in a less-recognized demographic group—travelers from tropical destinations.[4-8]

Methods

Case Ascertainment

The study was performed using the catchment of a large private pathology laboratory based in Brisbane, Queensland, Australia, performing microbiological services for predominantly primary care facilities across the states of Queensland and northern New South Wales and also some centers in Darwin, Northern Territory. All C. diphtheriae isolates identified by the laboratory between 2002 and 2012 were included. Clinical notes were sourced via telephone contact with the referring doctor at the time of issuing the laboratory report or in a few cases retrospectively during the study period. Information obtained included basic demographic data, residence in an area historically known to be endemic for C. diphtheriae [Northern Territory and tropical Queensland (north of latitude 23.5)],[2] history of overseas travel within the previous 21 days, country traveled to, mechanism of injury, duration of symptoms, and diphtheria vaccination status. It was also ascertained if patients identified as Aboriginal or Torres Strait Islander (ATSI), an ethnic group known to have higher colonization rates of C. diphtheriae. Frequency tables were analyzed using Fisher's exact test for categorical variables.

Ethical approval was given at an institutional level in accordance with the National Health and Medical Research Council guidelines for low risk human research.[9]

Organism Identification

Nonlipophilic Gram-positive bacilli grown on blood agar from skin, soft tissue, or throat swabs were screened prior to full identification with urea, glucose, and sucrose reduction. Urea-negative, glucose-positive, sucrose-negative isolates were formally identified using the Vitek ANC card or API RAPID Coryne system (bioMérieux, Marcy l'Etoile, France), and morphological appearance on Oxoid tellurite agar. Antibiotic susceptibility testing was performed by disc diffusion using Clinical Laboratory Standards Institute (CLSI) methodology with interpretive criteria from the CLSI guidelines for Staphylococcus species,[10] as recommended by the World Health Organisation.[11] Minimum inhibitory concentration (MIC) values when required were obtained using E-test (bioMérieux) and breakpoints as defined by the CLSI guidelines for infrequently isolated or fastidious bacteria.[12] All C. diphtheriae isolates were forwarded to the State Public Health Laboratory (Queensland Health Forensic and Scientific Services, Coopers Plains, Brisbane) for confirmation. Isolates were initially confirmed phenotypically using standard laboratory testing including fermentation of sugars, catalase, and oxidase. Toxin genes were detected molecularly using a modified triplex polymerase chain reaction (PCR) with gene targets being dtxR and the toxin A and toxin B subunit genes.[13-15] Further testing to confirm toxin expression is not routinely available in Australia.[16]

Results

Seventy-two isolates were identified between 2002 and 2012. Only one isolate (from 2006) was tox gene positive, with subunits A and B detected. This organism had been isolated in a coral wound acquired 2 weeks previously in Indonesia by a young Israeli traveler whose vaccination status was unknown. The patient was lost to follow-up despite Public Health involvement, so it is unknown whether there was any clinical evidence of toxin production.

Demographic information and site of injury is provided in Table 1. Only two isolates were from throat swabs. All others were from skin or soft tissue infections (SSIs), mostly on the lower limb. In nine cases, primary care givers were unable to confirm or exclude recent travel. Of the 63 patients for whom a travel history was available, 70% [95% confidence interval (CI) 58–80] had a history of a nonhealing wound obtained during travel to a tropical destination. Vaccination status could not be reliably ascertained for most cases. Eighteen percent of all patients were ATSI. There were significantly more ATSI patients in the group without a history of overseas travel (53% vs 4.5%; p < 0.0001), and also significantly more patients in this group from Northern Territory or tropical Queensland (53% vs 3%; p = 0.0001).

Table 1. Demographic information and site of injury for the whole cohort, plus cohort grouped by history of recent overseas travel or no travel
 Total cohort (n = 72)Travel (n = 44)No travel (n = 19)Travel history unknown (n = 9)
  1. There were nine patients whose travel history was unknown.
  2. ATSI = Aboriginal or Torres Strait Islander; F = female; M = male.
  3. aSignificantly more ATSI patients in the group with no travel history (p < 0.0001).
  4. bSignificantly more patients from an area in Australia endemic for Corynebacterium diphtheriae in the group with no travel history (p = 0.0001).
  5. cOne cervical lymph node infection and one facial lesion.
Age range (median)5–88 (43)10–74 (46)5–88 (21)31–52 (42)
Sex M/F52/2032/1211/89/0
ATSI (%)13 (18.6)2 (4.5)10 (53)a1(11)
Lives in area of endemicity for Corynebacterium diphtheriae13 (18.6)3 (6.8)10 (53)b0
Site of infection
Lower limb5536127
Upper limb5131
Throat2 2 
Otherc211 
Unspecified8611

Countries visited are shown in Figure 1. The majority of cases were from South Pacific destinations (n = 24) and Papua New Guinea (n = 10). Actual mechanism of injury was available only for 25 (35%) cases. Where the mechanism was known, coral cuts (50%) or being scratched by wood or vegetation (25%) were the most common injuries. Only one invasive infection was reported. This was a suppurative lymphadenopathy requiring surgical drainage in a 25-year-old American woman who had acquired her infection in Vanuatu, South Pacific Ocean. She had no underlying medical history.

Figure 1.

Map showing travel destination prior to documented acquisition of Corynebacterium diphtheriae for 41 patients with travel history. Numbers represent number of patients who traveled to each destination indicated. One additional patient acquired infection in Cambodia, two additional patients had previous travel to unknown destinations.

Copathogens were present in all but three patients (Table 2) with Staphylococcus aureus (n = 32) and Streptococcus pyogenes (n = 61), the two most common organisms. There was no difference in copathogens in those with or without a history of travel.

Table 2. Coinfecting organisms were found with 69 of 72 (96%) Corynebacterium diphtheriae infections
Coinfecting organismsNumber (%)
  1. aOne Prevotella, two Peptostreptococcus spp.
Streptococcus pyogenes61 (85.7)
Staphylococcus aureus32 (58.5)
Group C streptococcus4 (5.6)
Arcanobacterium haemolyticum3 (4.2)
Anaerobic spp.a3 (4.2)
Methicillin-resistant S aureus3 (4.2)
Pseudomonas aeuruginosa1 (1.4)
Group G streptococcus1 (1.4)

Two isolates were resistant to penicillin (Table 3). The penicillin MIC of one isolate was 8 mg/L. The other isolate was resistant by disc diffusion but could not be retrieved from storage for MIC testing. Both were from patients with no history of overseas travel.

Table 3. Antibiotic susceptibilities of Corynebacterium diphtheriae isolates, using CLSI breakpoints for Staphylococcus species
AntibioticNumber of isolates sensitive (%)
  1. CLSI = Clinical Laboratory Standards Institute.
Cephalothin (n = 71)71 (100)
Erythromycin/clindamycin (n = 70)70 (100)
Trimethoprim/sulfamethoxazole (n = 72)62 (86)
Penicillin (n = 72)70 (97)
Tetracycline (n = 70)67 (96)
Vancomycin (n = 54)54 (100)

All C. diphtheriae isolates were uniformly sensitive to cephalothin, clindamycin, erythromycin, and vancomycin. Trimethoprim/sulfamethoxazole had the highest rate of resistance.

Discussion

Few laboratories in the developed world continue to screen throat swabs for C. dipththeriae,[17, 18] and a recent survey confirmed that throat carriage rates in highly immunized countries remains low.[17] Although the world-wide incidence of diphtheria has declined as a result of childhood immunization, the condition remains endemic in some eastern European countries of the former Soviet Union and many third world countries.[1, 19] C. diphtheriae is under-recognized as a skin pathogen in travelers returning from these endemic areas. Increasing global travel will continue to allow exposure of healthy adults to C. diphtheriae, despite high immunization rates in Western countries.[19] For example, the last respiratory diphtheria case in New Zealand (1998) was thought to have been acquired by the index case from a relative with an abrasion received in Indonesia.[20]

While small case series of imported cutaneous diphtheria from toxigenic C. diphtheriae have been reported (Table 4), it is difficult to estimate true numbers of nontoxigenic C. diphtheriae infections in returned travelers from endemic areas, particularly because isolation of nontoxigenic C. diphtheriae is not notifiable in many countries including Australia. A review of C. diphtheriae isolates in the UK[21] which included nontoxigenic isolates revealed few imported isolates in contrast to this study, in which 70% of patients with available history had recently traveled to an endemic area. While cutaneous lesions can potentially be a source of toxigenic C. diphtheriae, only one of the 71 isolates reviewed in this study was toxigenic. Therefore, importation of a toxigenic strain is likely to be a rare event.

Table 4. Recent literature describing Corynebacterium diphtheriae in returned travelers
ReferencesCountryYearNumber of patientsToxigenic isolates
  1. aOne toxigenic isolate only.
Sing and Heesemann[4]Germany1997–20036Yes
de Benoist[5]UK1995–200217Yes
Hart et al.[7]UK19962Yes
Edwards et al.[18]New Zealand1987–20122Yes
Orouji et al.[6]Germany20111No
Lindhusen-Lindhé[8]Sweden20121Yes
This studyAustralia2002–201244Noa

Given the nonspecific nature of the presentation, it is likely that many C. diphtheriae infections go unrecognized, either due to failure of laboratory detection or lack of investigation. Therefore it is important to raise awareness of the need for microbiological surveillance of nonhealing lesions in returned tropical travelers.

In skin infections, C. diphtheriae tends to cause primary lesions (punched-out, gray-based lesions) in endemic countries (Figure 2), and be a super-infection of an existing lesion in nonendemic countries, with no characteristic features.[19] Infected lesions may persist for weeks to months.[22] While immunization is generally associated with reduced severity of disease and reduced colonization with toxigenic C. diphtheriae,[17, 23] it may not prevent cutaneous infection.[5]

Figure 2.

Skin ulceration in a wound infected with Corynebacterium diphtheriae.

The finding of frequent coinfection with S. pyogenes, S. aureus, and Arcanobacterium haemolyticum was consistent with previous reports in the literature.[3, 24] Treatment of C. diphtheriae is traditionally with penicillin or erythromycin, however, all isolates also appeared to be susceptible to first generation cephalosporins which may commonly be given for SSIs. There were two penicillin-resistant isolates of C. diphtheriae, both of which appear to be locally acquired. The lack of clinically defined breakpoints for C. diphtheriae complicates interpretation of sensitivity testing, but penicillin and erythromycin resistance has been reported by others.[25, 26]

The relationship between vaccination and colonization with nontoxigenic C. diphtheriae needs further clarification. While in general, it is believed that vaccination will not prevent colonization or infection with nontoxigenic strains of C. diphtheriae,[27] Bergamini and colleagues demonstrated an association between nonprotective levels of antitoxin antibodies and increased nasopharyngeal carriage of Corynebacterium species.[28] The significance of these findings requires further investigation to determine relevance to clinical practice.

Regardless, especially in view of the death of an unvaccinated Australian due to diphtheria in recent years[29] and the ability of nontoxigenic C. diphtheriae to acquire the tox gene through horizontal transfer from toxin-carrying strains,[30] recognition of this source of C. diphtheriae infection and circulation is important. Primary care givers should be encouraged to update vaccination status of all travelers. In Australia, the latest national diphtheria antibody serosurvey data from 2005 suggested good protection in children, but waning immunity levels in up to one fourth of the adult population.[31]

This study is retrospective, so inevitably some of the clinical data is incomplete. It is possible that some individuals with a history of recent travel may also have some of the more accepted risk factors for nontoxigenic C. diphtheriae infection. However, overseas recreational travel makes significantly poor health or extremely low financial status unlikely for the majority. Of note there were only two ATSI subjects in the group with a history of travel, which means this factor is unlikely to confound the results. While C. diphtheriae can also be found in tropical areas of northern Australia including the Northern Territory, few of the patients with a history of travel were from these areas. This suggests two distinct populations at risk of C. diphtheriae colonization/infection, those with traditional risk factors living in endemic areas, and those who have most likely acquired the organism through traumatic inoculation during overseas travel.

In conclusion, C. diphtheria can colonize or cause SSI in otherwise healthy travelers returning from tropical destinations. Clinicians and laboratories need to be aware of this potential source of C. diphtheriae infection due to rare toxigenic strains.

Acknowledgments

The authors would like to acknowledge the assistance of Dr Amy Jennison and Ms Helen Smith of the Queensland Health Forensic and Scientific Services laboratory for their assistance in providing details regarding the confirmatory testing of C. diphtheriae isolates in Queensland. The authors declare no external funding or writing assistance during the preparation of this data and the manuscript.

Declaration of Interests

The authors state they have no conflicts of interest to declare.

Ancillary