Chronic skin infections affect about 8–14% of adolescent and adult refugees seeking medical care [13,14,15,16]. Diagnosis and treatment can be challenging as many factors, such as low vaccination coverage, external environmental influences during escape or overcrowded asylum centres lead to a higher prevalence of infections that are rare in Europe. A striking increase in C. diphtheriae infections among refugees in Europe was observed starting in summer 2022 [7,8,9]. During the study period, 205 diphtheria cases were reported to the German National Health Authority. Thirty-six out of these were included in our study on adolescents. We carefully analysed the management in these CD cases with the aim to develop a treatment algorithm for refugees presenting with chronic wound infections at risk of C. diphtheriae infections.

In line with previous reports from us and others [7,8,9], escape via the Balkan route was overrepresented in refugees with CD in Germany. Chronic skin wounds were polymicrobial (S. pyogenes, S. aureus) in most cases, and among S. aureus isolates over 50% were methicillin-resistant (MRSA). This is consistent with previous reports [8, 17]. Our findings underscore that in refugees presenting with chronic skin wounds C. diphtheriae should be considered as a differential diagnosis and included in the diagnosis and management plan. In this cohort of adolescents, fever was a simple and quite specific parameter for distinguishing between severe cases of C. diphtheriae with potential for DT-related complications and cases with exclusive CD.

WHO recommends macrolides and penicillin G for C. diphtheriae [18, 19], which is commonly sensitive to a range of antibiotics [20]. Surprisingly, in our cohort, amoxicillin/clavulanic acid (n 7) was most commonly used, but failed to eradicate C. diphtheriae in two of the seven patients. In contrast, clindamycin (n 5) showed a favourable outcome in all patients. Based on this observation and since most cases of cutaneous diphtheria were co-infections with other pathogens, choosing antibiotics that cover next to C. diphtheriae other common skin pathogens seems advisable. MRSA was a common co-pathogen that should be considered when selecting a systemic antibiotic. This limits the clinically rational treatment options to clindamycin, cotrimoxazole, doxycycline and macrolides. Yet, macrolide resistance is common in MRSA [21]. Additionally, macrolide resistance was found in two of the 31 (6%) C. diphtheriae isolates in our cohort. C. diphtheriae was resistant to cotrimoxazole in 32% of the isolates in our cohort, leaving clindamycin and doxycycline as preferable treatment options. As there are some concerns with the use of doxycycline (phototoxicity, pause after calcium-containing products), which could be problematic under the circumstances, clindamycin seems preferable. However, in the ongoing diphtheria European outbreak among refugees, one (ST-377) of the major outbreak clusters showed macrolide and clindamycin resistance [7, 22]. With respect to the findings in additional European countries [23], antibiotic resistance testing should be performed for every C. diphtheriae isolate. When broadening the perspective for cutaneous diphtheria-like illnesses, infections with toxigenic strains of the zoonotic pathogen C. ulcerans must be considered. C. ulcerans infections are usually acquired via animal contact (e.g., cats and dogs, but also via a wide variety of animals) and are often resistant to clindamycin. [24] While an increasing number of C. ulcerans infections has been reported in local pet owners [25], no cases of C. ulcerans have been reported in refugees in the literature yet [24,25,26], making C. ulcerans an unusual differential diagnosis.

Focusing on the European diphtheria outbreak among refugees, the following approach can be extrapolated from our data.

In refugees with chronic skin wounds without fever and in good general condition, we suggest performing wound and throat swabs for pathogen identification and antimicrobial resistance testing. Since Corynebacteria spp. are often summarized as “skin flora”, the microbiological laboratory should be informed about the clinical circumstances and the need to test for C. diphtheriae. Topical antiseptic therapy and oral antibiotic therapy with clindamycin should be instituted. Regarding length of therapy, most patients in our cohort received 10 to 14 days of systemic antibiotics, however, shorter courses of 5 to 8 days were also associated with good outcomes. Therefore, we suggest that an initial treatment duration of 7 days might be justified. This should be followed by clinical monitoring and microbiological clearance. A longer treatment course or a change of the antibiotic agent may be necessary depending on the clinical course or if microbial clearance has not been achieved. Diphtheria antitoxin should not be administered in these cases, both from a clinical point of view for the patient (high risk of anaphylaxis) and in view of the general shortage of diphtheria antitoxin stocks [2].

In all patients a thorough medical history should be taken, including duration of the lesion, route of escape and vaccination status, as well as a detailed clinical examination including a throat examination.

Less than half of the afebrile patients in our cohort were isolated to prevent transmission. Although preventive isolation measures are recommended for C. diphtheriae infected individuals [10, 27], difficulties in implementing isolation, such as overcrowding and lack of resources, may explain this. Isolated skin colonization, as in all our afebrile patients, carry a low risk of transmission if the wounds are appropriately covered. In the current outbreak, whole genome sequence analysis of 363 European cases of diphtheria infection identified three major clusters, yet transmission likely occurred outside Europe [22, 23, 28]. The phylogenetic analysis of 42 cases in Germany in 2022 yielded similar results, since no secondary cases could be identified [7]. It seems noteworthy that diphtheria vaccination coverage is high in Germany (96.9%, according to Robert Koch Institute) and across Europe (97%, according to WHO) [29, 30]. Balancing risks and resources, it seems justified not to isolate patients suspected of having isolated cutaneous diphtheria if adequate hygienic measures, such as covering wounds, the avoidance of close direct or indirect contact to the wound and fomites are in place. A recommendation to wear a mask until microbiology results are available is a resource-saving alternative. This approach could be implemented and adapted in conjunction with local health authorities. On the other hand, patients with suspected pharyngeal diphtheria, fever and poor general condition should definitely be isolated. For determination of the duration of isolation, we recommend microbiological eradication control after 5 days.

Patients with cutaneous lesions and fever on initial presentation should be thoroughly examined and hospital admission must be considered. In the reported patients, a variety of antibiotic classes was employed, including penicillin G, amoxicillin/clavulanic acid, azithromycin, clarithromycin, doxycycline and combinations of these. Two patients with systemic diphtheria toxin-related symptoms also received antitoxin. In addition to a wound and throat swab, blood cultures must be taken, and if there are clinical signs of pharyngeal diphtheria or toxaemia, diphtheria antitoxin treatment (DAT) and antibiotic treatment according to WHO recommendations, i.e., macrolides or high-dose penicillin G, should be administered before results are available [11]. For febrile patients without signs of pharyngeal diphtheria, close clinical monitoring, thorough sampling and intravenous antibiotic therapy, preferably with clindamycin, may be sufficient in certain settings with suspected polymicrobial wounds and known clindamycin-sensitivity for C. diphtheriae in an outbreak setting. Our recommendation derived from this study and analysis of literature is summarized in Fig. 4.

Proposed algorithm for treatment and diagnostics for refugees with chronic, purulent (erosive) skin wounds. DTP: Diphtheria, Tetanus, Pertussis vaccine; iv.: intravenous

This is to our knowledge the largest, detailed report on antimicrobial treatment for CD in central Europe. Although we were unable to obtain complete information on all points for all patients, it is noteworthy that only two patients in the afebrile group were missing data on treatment outcome. Due to the retrospective design only those individuals in whom C. diphtheriae was isolated from swabs could be included. There is a risk that cases may have been missed as Corynebacteria spp. may have been hidden under ‘skin flora’ or masked by faster growing bacteria, especially in mixed cultures. Furthermore, our report lacks information on the vaccination status of the patients, a common problem in refugee management, as in most cases vaccination records are not available. As vaccination coverage in countries of origin is sometimes low or incomplete [29], vaccinations should be given generously to all patients and to close contacts in conjunction with ECDC and national guidelines [10, 31].