Ralstonia mannitolilytica infections in an oncologic day ward: description of a cluster among high-risk patients

The genus Ralstonia comprises a group of non-fermentative, Gram-negative bacteria (NFGN) found in moist environments, such as water, soil and plants [1]. Three Ralstonia species, Ralstonia pickettii, Ralstonia insidiosa and Ralstonia mannitolilytica, formerly designated Burkholderia pickettii, Burkholderia solanacearum and Pseudomonas thomasii, respectively, have been recognized as opportunistic human pathogens [1]. Their relevance has been currently re-evaluated because of their ability to survive in different types of disinfectants and to pass through 0.2-μm filters that are used to sterilize solutions [1, 2].

Ralstonia spp. is reported as a causative agent of bacteremia [3, 4], meningitis [5, 6], and sepsis [7, 8] in immunocompromised patients and of central venous catheter (CVC)-associated bacteremia in oncology patients [1–4, 9]. Several hospital outbreaks have been described that were associated with contaminated solutions, including water for injection, saline solutions, disinfectants and antiseptics [1]. Multidrug resistance in NFGN is widely reported in the literature [10, 11] and is causing increasing concern because such bacteria may have a role not only as human pathogens but also as potential reservoirs of resistance genes, particularly when they are found in hospital settings. Several studies have described resistance to fluoroquinolones, 3^rd generation cephalosporin and carbapenems [1] in isolates belonging to all the three Ralstonia species.

R. pickettii is the Ralstonia species most frequently reported in the literature while only a limited number of infections are attributed to R. insidiosa and R. mannitolilytica [1]. The clinical importance of these two species is probably underestimated because their biochemical patterns are similar to that of R. pickettii, making it impossible their distinction based on conventional microbiological tests only [1, 4, 12, 13].

Here we describe an outbreak caused by R. mannitolilytica in patients attending a day ward unit in an oncology hospital in Rome occurred from September 2013 – June 2014.

The microbiological and epidemiological investigation of this outbreak did not detect the source of the contamination, however the molecular typing of the pathogen strongly supports the hypothesis of a common source of contamination. Failure to identify the culprit is probably due to the fact that when the investigation was started samples from disinfectants, antiseptics and saline solutions used at the beginning of the outbreak were not available for microbiological investigation. In fact, disinfectants, antiseptics and multi-dose bottles of saline solution have been described in the literature as one of the main sources for Ralstonia spp. contamination [1], therefore their prompt removal from DW-A immediately after the beginning of the outbreak was the most understandable precaution to adopt from the patients’ safety viewpoint. However, in spite of the multi-dose bottles of saline solution removal, the outbreak lasted 9 months. Indeed, the revision of the medical records showed that CVC flushing with saline solution was a common procedure adopted with all the patients and that, at the beginning of the outbreak, multi-dose bottles of saline solution were in use in DW-A. Thus it is very likely that patients were exposed to one or more contaminated bottle(s) of saline. The majority of the patient developed clinical symptoms immediately thereafter, while others likely had a CVC colonization by R. mannitolilytica. Any subsequent procedure through CVC, chemotherapy or new saline flushing, may have caused detachment and dissemination of R. mannitolilytica from CVC, causing fever and chills. This might have happened even after several weeks from CVC colonization, possibly accounting for the protracted duration of the outbreak. A similar finding have been reported by Raveh et al. [19] in which, a patient with a CVC developed fever and chills and his blood culture was positive for R. pickettii. The patient was treated with antibiotic therapy and the CVC was left in place. However 15 months later the patient developed another bloodstream infection with the same organism likely colonizing CVC.

The ability to grow as biofilm on abiotic surfaces plays an important role in colonization of hospital equipment and indwelling medical devices such as CVCs [20]. Ability to form biofilm has been described in different Ralstonia species including also R. mannitolylitica [21–24]. It is conceivable that biofilm formation might have played a role for these strains to allow adherence to CVCs and subsequent dissemination in the host, following the flushing procedures of the device.

The prolonged duration of this outbreak is not unprecedented for this pathogen. Daxboeck et al. [4] reported isolation of R. mannitolilytica in 30 patients attending 15 different wards between February 2002 and March 2004; in this study also the source was not identified. Other outbreaks have been reported in pediatric patients, associated with oxygen delivery devices [25, 26] and in oncology patients [1, 2]. In the outbreak described by Gröbner et al. [2], which lasted 11 weeks, the source was not identified but it was hypothesized that a contaminated solution administered through CVC could have been the culprit.

It is important to underline that the misuse of multidose vials, as in our outbreak, has been frequently reported as a cause of numerous outbreaks [27–30] in different European countries as well as in USA where the CDC is aware of at least 49 outbreaks occurred in 13 years [31]. Therefore the correct use of single dose vials, for a single patient, is an effective and important way to prevent outbreaks as reported by specific guidelines [32].

R. mannitolilytica is frequently recognized as a multiresistant microorganism [4, 33–35] but limited data are available about carbapenem-resistance. To the best of our knowledge, only a limited number of studies so far, have tested susceptibility to this drug: in four studies, all the strains tested were resistant (five in one study and one strain each in the other three) [2, 17, 33] while Daxboeck et al. [4] reported carbapenem-resistance in 12 strains out of 30. In our isolates no carbapenemase genes were detected, but they were positive for the chromosomal genes bla _OXA-443 and bla _OXA-444, that have been previously described in one R. mannitolilytica carbapenem-resistant strain [17] and that bear close similarity to the bla _OXA-22 and bla _OXA-60 genes found in R. pickettii [36]. bla _OXA-22 and bla _OXA-60 code for two oxacillinase, a narrow spectrum oxacillinase and an inducible carbapenemase, respectively, although none of them was found to be responsible for carbapenem resistance in R. pickettii [36, 37]. In addition, the cloxacillin test suggested that the R. mannitolilytica isolates were positive for the presence of AmpC β-lactamases. These enzymes are clinically important because they may confer resistance to a wide variety of β-lactam drugs, narrow-, expanded-, and broad-spectrum cephalosporins, β-lactam-β-lactamase-inhibitor drugs combinations, as well as to aztreonam and carbapenems in case of association with altered porins and/ or efflux mechanisms [38]. Taken together, our data suggest that meropenem resistance was likely due to overproduction of AmpC β-lactamase, possibly in synergy with a second mechanism (e.g. decreased production of the porin channel and/or activation of efflux systems). Recently, a serin-hydrolase class C family β-lactamase has been identified in R. mannitolilytica (GenBank accession number: WP_045219476) [15], therefore production of this enzyme may contribute to carbapenem resistance.

Although R. mannitolilytica did not cause life-threatening infections in this outbreak as well as in previously reported outbreaks [2, 26], the increased detection of Ralstonia spp. in healthcare settings coupled with the emergence of multi-resistant strains of R. mannitolilytica, represent a reason of concern, particularly in case of vulnerable patients which may require an antimicrobial therapy.

We reported the first outbreak due to R. mannitolilytica in oncology patients bearing CVC in Italy. Although the source of the outbreak could not be readily identified, the investigation suggested that contaminated saline solution used for CVC flushing may have been the source of the outbreak. In order to prevent possible infections we recommend the removal of any CVC that is not longer necessary and the use of single dose solutions for any parenteral treatment of cancer patients.