Arcobacter prevalence in human stool samples
This is the first prospective study addressing the prevalence of Arcobacter in stool samples from outpatients and hospitalized patients in Germany by applying an Arcobacter-specific detection method. Overall, Arcobacter spp. were isolated from 36 out of a total of 4646 (0.77%) examined specimens. This isolation rate is in concordance with studies from New Zealand and Belgium, where Arcobacter spp. were detected in 0.9% (12/1380) and 1.31% (89/6774) of human diarrheal fecal samples, respectively (10, 20), whereas slightly different prevelances (as low as 0.2 or up to 3.6%) were found in other studies from Belgium, Turkey, Portugal, India and Chile (12, 21-24). These differences could be attributed to various factors, such as patient populations, geographical aspects, examined sample sizes, and in particular, to the different microbiological methods applied. The impact of the detection method has been demonstrated in several studies (25-28). The authors each compared different cultural isolation strategies with varying incubation and medium conditions revealing differences in Arcobacter isolation frequency ranging from 7% to 36%. Notably, our study revealed a higher Arcobacter prevalence in an analyzed subgroup by using Arcobacter-specific enrichment (0.97%) than determined by non-specific methods used in the three routine laboratories (0.49%). Future studies should address whether patients with Arcobacter spp. at low quantities that can only be detected by applying specific enrichment methods differ clinically from those patients in whom the pathogen is easily detected within the routine culture based procedures.
Furthermore, we determined a higher Arcobacter prevalence in stool samples of outpatients than of hospitalized patients (i.e., 0.85% (33/3884) and 0.40% (3/752), respectively). Thus, in most patients, Arcobacter spp. most likely do not cause serious infections requiring hospitalization. Likewise, in a previous German study, patients who were hospitalized for severe gastroenteritis (n = 104) were found to be positive mainly for norovirus or Campylobacter spp.; in contrast, no Arcobacter was isolated by using routine diagnostics (29).
Among the 36 Arcobacter isolates obtained in our study, A. butzleri was the most prevalent species (n = 24) followed by A. cryaerophilus (n = 10), which is in line with other studies (10, 21, 24). In addition, to best of our knowledge, this is the first report of A. lanthieri isolation from human specimens (n = 2) which might point towards its role as gastrointestinal pathogen. However, the applied selective enrichment media as well as the multiplex PCR are validated for the detection of the three species A. butzleri, A. cryaerophilus and A. skirrowii only, and could therefore bias the result according to species diversity (16, 18).
Overall, in the analyzed subgroup Arcobacter spp. were the second most frequently isolated pathogens (0.97%) after Campylobacter spp. (4.39%), followed by Salmonella enterica (0.75%). Our results are supported by a previous study demonstrating Arcobacter spp. as fourth most commonly isolated pathogens from diarrheal patients (1.31%), after Campylobacter spp. (5.61%), Salmonella spp. (2.04%) and C. difficile (1.61%), albeit prevalences of the enteropathogens were higher than in our study (10).
Antimicrobial susceptibility
Data regarding antimicrobial susceptibilities of Arcobacter spp. are scarce, mainly due to missing standardized protocols and defined breakpoints, which makes it difficult to interpret results and to define antimicrobial resistance. In previous studies, MIC results have been compared with breakpoints for Enterobacteriaceae or Staphylococcus spp. as defined by the Clinical Laboratory Standards Institute (CLSI), with breakpoints for Campylobacter as defined by the U.S. National Resistance Monitoring System criteria or with EUCAST breakpoints for Enterobacteriaceae, Campylobacter or non-species related breakpoints (5, 30, 31). In our study, we compared the MICs with ECOFFs defined by EUCAST for C. jejuni (32). For ciprofloxacin, gentamicin and ampicillin the C. jejuni ECOFFs appear to be applicable for Arcobacter as well, which has also been shown by Riesenberg et al. (2017) (33). However, our data suggests that Arcobacter ECOFFs for erythromycin, tetracycline and azithromycin may be higher than those of C. jejuni. All of our isolates displayed MICs for azithromycin above the ECOFF of C. jejuni (0.25 µg/ml), which, however, is comparable with data from a Belgian study (34). Although erythromycin and azithromycin are both macrolides, the bimodal distribution for azithromycin but not for erythromycin was remarkable. Van den Abeele et al. (2016) have also detected MICs > 8µg/ml for azithromycin in 50% of A. butzleri isolates, which is in line with our results (34). Likewise, other studies revealed elevated MICs for azithromycin in up to 95% of A. butzleri and in 20% of A. cryaerophilus strains isolated from poultry products (30, 35). Similar to our results, other studies on antimicrobial susceptibility revealed also low MICs for Arcobacter spp. to erythromycin whereas some studies reported resistance rates up to 62% (5, 36, 37). In contrast to our study, those studies used disc diffusion assays with 15µg/disc and applied resistance criteria for Enterobacteriaceae according to CLSI 2010. In Campylobacter, there is usually cross-resistance betweeen azithromycin and erythromycin. Single isolates, however, may display susceptibility to erythromycin and resistance to azithromycin, and whole genome sequencing analysis revealed an amino acid substitution in ribosomal protein L22 (leading to azithromycin resistance), but no mutations in the 23S rRNA gene, which explains the susceptibility to erythromycin (38). Further analyses are needed to determine the genomic background being responsible for the divergent MIC distributions observed by us for Arcobacter spp.
As mentioned before, 86% of the investigated Arcobacter isolates showed low MICs for ciprofloxacin ranging from 0.032-0.50 µg/ml, which is comparable to other studies (36). In contrast, clinical Campylobacter isolates displayed high resistance rates (MICs ≥ 4 µg/ml) ranging from 45 to 71.4% (39, 40). Notably, we found elevated MICs for ciprofloxacin predominantly in A. cryaerophilus strains similar to a Belgian study (34). Thus, ciprofloxacin might be the drug of choice, if antibiotic treatment of A. butzleri-infection is required.
In accordance with our data, only low resistance rates from 0-4% of Arcobacter spp. to gentamicin have been reported before (36). Similarly, susceptibility to tetracycline might be common, although one recent study from retail food in Portugal demonstrated high resistance (95%) in A. butzleri (5, 41). Furthermore, 42% of our A. butzleri isolates displayed high MICs for ampicillin (24-64 µg/ml), which is similar to previous studies where 50 to 100% isolates with high ampicillin MICs have been shown (20, 22, 31, 34).