Chronic lower extremity wound infection due to Kerstersia gyiorum in a patient with Buerger’s disease: a case report

Kerstersia gyiorum was first identified in 2003 by Coenye et al. as a distinct species by examination of the isolates obtained from nine clinical specimens such as leg ulcer, sputum, and faeces by cellular fatty acid analysis and 16S rRNA gene sequencing [1]. It belongs to Alcaligenaceae family and is closely related to Alcaligenes, Bordetella, Achromobacter spp. [2, 3]. After its first description, there have been publications reporting its isolation from chronic otitis media [2, 4‐6], urinary tract infection [3], chronic leg ulcer [2], post-ulcer bacteraemia and sepsis [7] and bronchoalveolar lavage fluid [8].

Here we present a case of chronic foot and ankle infection due to K. gyiorum in a 47-year-old patient with Buerger’s Disease and we have made a brief review of K. gyiorum cases in literature so far.

A 47-year-old male patient who was previously followed up at the Chronic Wound Clinic of our hospital presented with a 10 × 15 cm wound on the dorsolateral surface of the right foot and a 2 × 3 cm wound on the outside of the right ankle (Figs. 1 and 2). The patient had a history of Buerger’s Disease (Thromboangiitis obliterans) for 28 years. It has been learned that the patient continues smoking and lives in bad hygienic conditions. The patient received debridement of the wound at the Chronic Wound Clinic, and the sample was sent to our hospital’s Medical Microbiology laboratory for microscopy and culture. Before getting culture results, oral ampicillin sulbactam, ciprofloxacin, and topical mupirocin treatment were started to the patient empirically.

This time the previous graft was found to be lysed, and the wound was re-infected. The blood tests were performed and white blood cells were found 8200/μL (70% neutrophils), hemoglobin 14 g/dl, hematocrit 43.7%, and platelets 306.000 /μL. The patient’s CRP (7 mg / L) and erythrocyte sedimentation rate (65 mm / h) were found to be high.

Plenty of polymorphonuclear leukocytes and gram-negative bacilli were present in the Gram stained sample taken from the patient’s wound which was sent to our hospital’s Medical Microbiology laboratory. The material was inoculated on to blood agar and Eosin methylene blue (EMB) agar and incubated. Four different types of colonies were identified after 24 h of incubation at 37 °C on bloody agar and EMB agar. Out of these four colony types, the colonies predominantly grown on blood agar were light gray, had a dry appearance, and were large colonies with indented edges that tended to merge with each other, resembling Alcaligenes. However, unlike Alcaligenes there was no fruity odour and oxidase test was negative. These colonies were found to be indole-negative, urease-negative and strongly catalase-positive. The colonies grown on blood agar with second density were gray coloured and had swarming around the colony, were indole positive, urease positive, oxidase negative, hydrogen sulphide producing colonies. The third colony type was dirty-white-coloured, mucoid with smooth edges. They were indole negative, Voges-Proskauer test positive. The fourth colony type was smooth-edged, whitish mucoid colonies. Urease test was positive, indole positive and Voges-Proskauer test was negative. On EMB medium a mixed growth of one lactose-positive and three lactose-negative colonies corresponding to these four types of colonies was observed. Single colony passages were made from each of these four colony types in order to identify with Maldi Biotyper (Bruker Daltonics, Germany) system, a matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) system in our Medical Microbiology Laboratory. Antibiotic susceptibility tests of the bacteria were carried out on a Phoenix 100 (Beckton Dickinson, USA) device.

The passages made from the first colony type produced dry, light gray, opaque colonies, with irregularly indented, protruding edges, and swarming around the colonies on blood agar (Fig. 3), and lactose-negative colonies on EMB. Gram-negative, short bacilli were seen in Gram staining of colonies from blood agar (Fig. 4). These colonies were identified as Kerstersia gyiorum by Maldi Biotyper (Bruker Daltonics, Germany) system with 2.348 Biotyper score (excellent identification). The identification of K. gyiorum with Maldi Biotyper was confirmed by 16S ribosomal RNA gene sequencing according to the predefined methodology [9, 10]. Briefly, the following steps were taken for this procedure. DNA was isolated from pure culture using DNeasy Blood & tissue kit (Qiagen, Germany). 16S rRNA gene amplification was performed using universal 27F (AGAGTTTGATCMTGGCTCAG) and 1492R (GGTTACCTTGTTACGACTT) primers. In the PCR reaction, 1XTaq buffer, 2 mM MgCl 2, 0.2 mM dNTP, 0.4 pmol of primers and 1.25 U Taq polymerase (Thermo Fisher Scientific, USA) were used in a volume of 50 μL. DNA sequencing of the resulting product was performed using the Bigdye Cycle Sequencing Kit v.3.1 (Applied Biosystems, USA). Sequence analyzes were compared to the GenBank NCBI gene library data using the BLAST program (https://​blast.​ncbi.​nlm.​nih.​gov/​Blast.​cgi). The sequence was found 100% identical to K. gyiorum. Antibiotic susceptibility tests for K. gyiorum were performed using the Phoenix 100 (Beckton Dickinson, USA) device and the E-test (Liofilchem, Italy) for six antimicrobial agents (ceftriaxone, ciprofloxacin, colistin, gentamicin, imipenem, trimethoprim-sulfamethoxazole) on Mueller-Hinton agar in accordance with the manufacturer’s specifications. The results were evaluated according to the MIC breakpoints established by Clinical and Laboratory Standards Institute for other non-Enterobacteriaceae [11]. The results of antibiotic susceptibility tests for K. gyiorum are shown in Table 1.

Other microorganisms were identified as Proteus vulgaris, Enterobacter cloacae, Morganella morganii by Maldi Biotyper system, respectively. Antibiotic susceptibility results of P. vulgaris, E. cloacae, M. morganii are shown in Table 2.

Oral ampicillin-sulbactam and ciprofloxacin treatment was continued because the microorganisms were susceptible to these antibiotics. Hyperbaric oxygen therapy was used. Wound debridement and skin grafting were applied. The patient was discharged after one month of hospitalization. The patient was doing well in follow-up examination three-months later (Fig. 5). Figure 6 shows a timeline of events.

K. gyiorum’s name was derived from the Greek word “gyion”, which means “limb” since it was often isolated from leg and ankle wounds when it was first described [1‐3, 6, 7]. K. gyiorum belongs to Alcaligenaceae family, and it is related to Alcaligenes, Bordatella, Achromobacter, and Pigmentiphaga generas [1, 3‐7]. K. gyiorum colonies are known to show an appearance similar to Alcaligenes faecalis. However, K. gyiorum isolates are oxidase negative and lack the characteristic fruity odor [3, 7]. K. gyiorum is highly catalase positive but urease and β-galactosidase negative [1].

When we conducted a search for “Kerstersia gyiorum case report” on PubMed and Medline we found that there were 9 cases since the first time it was defined in 2003 [2‐8] Table 3 shows clinical features of these case reports. In 2012, Vandamme et al. identified Kerstersia similis, a close species, from the neck abscess of a 54-year-old patient [12].

In our case K. gyiorum grew with other types of microorganisms from wound specimen. This was also observed by previous researchers who have reported cases of K. gyiorum [2‐8]. In seven of the nine cases that have been reported so far there were polymicrobial infections. K. gyiorum was isolated from urinary tract infection together with P. vulgaris [3], from chronic otitis media with Corynebacterium amycolatum, from chronic lower extremity wound with Morganella spp. [2], from bronchoalveolar lavage fluid with Pseudomonas aeruginosa and Stenotrophomonas maltophilia [8], from chronic suppurative otitis media with P. aeruginosa in a case report [5] from our country, and in two different cases from chronic suppurative otitis media with Proteus mirabilis and from chronic suppurative otitis media with Staphylococcus aureus and Escherichia coli [6]. In our case, K. gyiorum was seen to be grown more dominantly than other microorganisms in the culture plate. However, since there are a very limited number of reports in the literature on this microorganism, it is difficult to predict the effect of K. gyiorum in the disease process in a polymicrobial infection [2]. The virulence factors of this microorganism should be investigated [6]. In our case, the patient’s condition got better after antimicrobial therapy. Ogawa et al. reported that K. gyiorum is prone to cause infection in situations predisposing to polymicrobial infection because Achromobacter and Alcaligenes spp. which are closely related have the same tendency [3].

If we look at the results of antibiotic susceptibility, our K. gyiorum isolate was found to be susceptible to aminoglycosides, ciprofloxacin, imipenem and meropenem and broad spectrum cephalosporins. These results were similar to Coenye et al. [1] and Almuzara et al.’s results [4]. However, Pence et al. [2], Uysal et al. [5], Mwalutende et al. [6], Bostwick et al. [7], Deutscher et al. [8], found resistance to ciprofloxacin in K. gyiorum isolates.

Our patient was a smoker. He had Buerger’s Disease (Thromboangiitis obliterans) for 28 years due to smoking. Thromboangiitis obliterans is a non-atherosclerotic inflammatory disease affecting the small and medium vessels in the limbs and causes circulatory problems [13]. In this disease, ulcers that do not heal especially at the distal end of the extremities are seen [13, 14]. The chronic suppurative otitis media case due to K. gyiorum that Mwalutende et al. reported was a chronic smoker [6]. Pence et al. also reported chronic cigarette smoking in their case [2]. We believe that there may be a relationship between chronic smoking and infection due to K. gyiorum. However, further work is needed in order to test this hypothesis.

Previous researchers who reported case reports of K. gyiorum reported that the associated infection usually develops on a long-standing inflammatory condition [2‐4, 6, 8]. Our patient had long-term lower extremity ulcers due to Buerger’s Disease.

In our case, we observed that on blood agar K. gyiorum forms colonies with a irregular spreading edge morphology. Pence et al., Deutscher et al., and Bostwick et al. reported that they observed colonies with similar appearance to ours [2, 7, 8]. We think that these phenotypical features may be typical for K. gyiorum. This morphology can be used to distinguish K. gyiorum from Acinetobacter spp., which is also a non-fermentative and oxidase-negative bacterium as suggested by previous researchers [2, 8]. Pence et al. referred to the formation of lavender pigment on MacConkey agar [2]. We did not see such pigment formation in our own case. However we used EMB instead of MacConkey agar. In our opinion, the appearance of such a colony type after inoculation of a specimen from a site of chronic inflammation should suggest K. gyiorum infection.

Introduction of molecular and genetic identification methods to clinical microbiology has increased the detection of new and rare bacteria in clinical specimens. MALDI-TOF MS and 16S rRNA gene sequencing are two of these new and advanced methods [2, 6, 8]. In our case we detected the presence of K. gyiorum by using MALDI-TOF MS. The reliance on solely biochemical identification methods or automated identification systems may lead to misidentification of K. gyiorum, because it has common biochemical features with more commonly detected pathogens such as Acinetobacter spp. [5‐7]. Also K. gyiorum is phenotypically similar to Alcaligenes faecalis and it may go unnoticed if proper identification methods are not used [5]. The increase in the identification of new and previously unrecognized bacteria has led to two important situations in clinical microbiology laboratories. The first is to determine whether these microorganisms are pathogens or contaminants, and the second is to correctly determine the antibiotic susceptibilities of these microorganisms [2]. The co-operation of the clinician and the laboratory is required to achieve these. As the data in the literature will increase with the studies to be done on this subject, a more accurate perspective will be obtained for approaching these cases. Besides, it is difficult to identify such rare microorganisms by conventional methods. Therefore, when the pathogen cannot be identified by routine biochemical methods in cases of chronic infections that do not heal, MALDI-TOF MS and 16S rRNA gene sequencing are emerging as fast and reliable alternatives [5, 8].