Molecular surveillance of healthcare-associated Candida infections in a rehabilitation centre for patients with severe acquired brain injuries

Objectives 
Healthcare-associated infections (HAIs) affect an estimated 4.1 million patients in the EU annually and cause considerable increases in illness, mortality, and costs. In European intensive care units (ICUs), about 8.4% of the bloodstream infections (BSIs) are caused by pathogenic Candida species acquired by patients during their stay in the hospital. The most part of these HAIs is linked to the use of invasive devices (e.g. vascular and urinary catheters) by underlining the key role of hospital environments in the cross-transmission of microbial pathogens. 
The aim of this study was to describe the genetic pattern, and population structure, of different pathogenic Candida species recovered from blood samples, central venous catheters (CVCs), hands of healthcare workers (HCWs) and hospital environments (air, medical equipment and various surfaces) in one of the most important hospitals of Southern Italy in the treatment of patients with severe acquired brain injuries.

Methods 
Twenty-six blood and 28 CVCs Candida isolates were examined in this study. Two hundred-eight additional isolates, 161 collected from ward environments and 47 from HCWs, were also included.
All isolates were initially identified using CHROMagar medium and Vitek2 yeast identification system and, subsequently, subjected to molecular identification using species-specific PCRs and/or ITS-sequencing.
The multilocus sequence typing (MLST) scheme used for C. albicans, C. glabrata and C. tropicalis genotyping was based on sequence analysis of 6-7 housekeeping genes as previously described whereas for C. parapsilosis (66 total isolates examined), genotyping was performed using four highly-polymorphic microsatellite markers.

Results 
In this study, 23 C. tropicalis, 20 C. parapsilosis, 7 C. albicans and 4 C. glabrata isolates were isolated from blood cultures and CVCs of hospitalized patients.
C. parapsilosis was the most common species recovered from environmental and HCWs samples (179 isolates) followed by C. albicans (14) C. glabrata (10) and C. tropicalis (5).
Genotyping data revealed 18 different MLST-genotypes (or diploid sequence types; DSTs) for C. albicans, 3 for C. glabrata and 8 for C. tropicalis. Except for C. glabrata, 14 C. albicans and 6 C. tropicalis DSTs were completely new and were submitted in the respective public MLST databases. For C. tropicalis, we found that 4 of the new DSTs (747, 748, 749 and 759) composed a single well-defined Italian clonal cluster (CC33) with the DST747 as the putative founding genotype. For C. parapsilosis, a total of 28 microsatellites genotypes were detected among 66 isolates genotyped. 
The availability of genetic and epidemiological data supported the evidence of recurrent outbreaks of Candida infection in our hospital.

Conclusion 
This study shows that, in our hospital, candidaemia is predominantly caused by non-albicans Candida species colonizing also various hospital surfaces including air, medical devices and HCWs, and therefore the implementation of surveillance programmes is imperative to prevent the spread of HAIs. The study also shows the increase of C. tropicalis BSIs and highlights that the population structure of this yeast is far from being fully elucidated as its complexity increases as different categories of patients and geographic areas are examined.
This work was funded by the Italian Ministry of Health (project code: GR-2011-02347606).

Molecular surveillance of healthcare-associated Candida infections in a rehabilitation center for patients with severe acquired brain injuries