We tested C. difficile type 078 isolates from humans and piglets for their minimal inhibitory concentration against eight antimicrobial agents and the results were evaluated using the wild type distribution data of EUCAST and the breakpoint values recommended by CLSI. In general, the results obtained indicate that type 078 isolates have similar antimicrobial susceptibility profiles as other PCR-ribotypes, though moxifloxacin resistance was found somewhat less in the examined strains, since only 16% of them were resistant to this antibiotic. Thirty seven percent of the isolates were resistant to four or more antimicrobials, a percentage comparable to those already described for other C. difficile types .
A study concerning type specific risk factors for human CDI in The Netherlands found that prior use of fluoroquinolones (mainly ciprofloxacin) was associated with CDI due type 078 . Of type 078 isolates included is this study, 94% was resistant to ciprofloxacin (second generation fluoroquinolone), but moxifloxacin (third generation) resistance was low, which is in concordance with the low resistance (27%) found by Salomon et al., (2011). During the rapid emergence of type 027 a decade ago, resistance of type 027 against moxifloxacin was frequent , which was also in line with the association of type 027 and fluoroquinolones in epidemiological studies. Moxifloxacin is not among the most frequently used human fluoroquinolones and is not available for veterinary use. It was however possible to establish an epidemiological link between resistance to moxifloxacin and the administration of fluoroquinolones in general (moxifloxacin and others) in human patients (p<0.01) and piglets (p=0.02). Since resistance can be linked with treatment with fluoroquinolones, but resistance is infrequent compared to other PCR-ribotypes isolated from human disease (Figure 1), we doubt that increased fluoroquinolone use alone explains the recent emergence of C. difficile type 078 in humans. In piglets, however, type 078 is virtually the only PCR ribotype that causes disease, indicating that a species border exists. The resistance against fluoroquinolones could have contributed to emergence of Type 078 in pigs after the introduction of type 078 in pig farms and the subsequent spread of type 078, as fluoroquinolone use was present in many pig farms and no other C. difficile PCR ribotypes were found.
In the Netherlands, there is a remarkable difference in antimicrobial use in humans versus animals. The use of antimicrobials in human medicine in the Netherlands is among the lowest in the European Union, while veterinary use of antimicrobials is among the highest [22, 23]. The most frequent used antibiotics for all types of disease in sows and suckling piglets are tetracyclines, followed by trimethoprim/sulfonamides (co-trimoxazole) and penicillins . This difference in antimicrobial use was however not reflected in different resistance profiles in our study.
The only discrepancy in human and porcine antimicrobial resistance profile was found for imipenem, with resistance in 29% of the human isolates and in 50% of the porcine isolates. Since imipenem is not used in porcine medicine, a higher level of resistance to this antimicrobial in porcine isolates compared to human isolates was unexpected. In the United states where imipenem is neither used in porcine medicine 100% of piglet isolates had MIC values of ≥16 mg/L . No explanation for this higher level of resistance can be found. Cross-resistance with other antibiotics has not been described for imipenem in gram-positive bacteria. Besides considerable overlap in susceptibility profiles, human and porcine 078 strains also had great genetic similarity when evaluated by multilocus variable-number tandem repeat analysis (MLVA), whole-genome analysis and pulsed-field gel electrophoresis (PFGE) [7, 26–28]. Using these methods, numerous human and porcine type 078 isolates were indistinguishable and were therefore suggested to have a high-level of genetic relatedness. Together with the increasing incidence of type 078, the association with community-associated CDI and the presence of type 078 in more than 90% of the piglets with colonization with C. difficile, the hypothesis arose that human and piglet type 078 had a common origin. The results of our study contribute to this hypothesis since resistance patterns highly overlapped despite different antimicrobial pressure . However, as many resistance genes are situated on transposons, antimicrobial susceptibility studies cannot give definitive insight in the (possible) common source of human and porcine CDI.
In this study, resistance to moxifloxacin was associated with the aminoacid substitution Thr82 to Ile in GyrA, as already observed in the majority of European C. difficile clinical isolates resistant to fluoroquinolones [19, 30, 31]. Surprisingly, no tetracycline resistant isolates were found in our study. However, MIC values between 8 and 12 mg/L were observed in less than 20% of the isolates analysed. It has been shown that C. difficile strains with reduced susceptibility to tetracycline often carry a tet(M) gene and are inducibly resistant when exposed to sub-inhibitory concentrations of this antibiotic . A recent Irish study found that MLSB resistance in type 078 is less frequently associated with erm(B) than other types . Our results support this observation since only a minority of the erythromycin resistance (up to 10%) can be explained by the presence of an erm(B) gene. Anyway, we did not observe any clindamycin resistant, erythromycin susceptible 078 isolate, as reported in the study mentioned above. What causes the resistance to erythromycin in 078 isolates is currently unknown. Although other erm(B) negative anaerobes possess one of the other erm genes (frequently erm classes A, F or Q) or overexpress an efflux pump, these have not been determined in C. difficile type 078 .
The strengths of our study in comparison with other reports on antimicrobial susceptibility data of C. difficile type 078 are that a large number of isolates of different origin were included and eight different antimicrobials were tested [6, 7, 25, 33]. Additionally, we compared the antimicrobial susceptibility data with previous antimicrobial treatments of patients and piglets. Limitations of the study are the lack of more detailed data on previous antimicrobial use (preferably expressed as daily defined dosages) and the fact that iinformation in piglets was obtained by interviewing the farmer, which enables reporting bias. Currently, attempts are underway to use defined daily dosage (DDD) as a measure for antimicrobial use, applicable for both humans and animals.