**7. Population Microdiversity and Role of Different Species of** *Candida*

*Candida* spp. are distributed differentially according to age: *C. albicans* and *C. parapsilosis* are prevalent in neonates [69], whereas adults are mainly affected with *C. albicans* and *C. glabrata* [70]. In the largest study to date (EUROCANDY), involving 23 pediatric centres, *C. albicans* (52.5%) and *C. parapsilosis* (28%) were the predominant species, followed by *C. tropicalis, C. glabrata, C. krusei* and other rare species (including *C. dubliniensis, C. pulcherrima, C. blankii, C. famata, C. guilliermondii, C. lusitaniae, C. magnolia, C. orthopsilosis, C. zeylanoides*). *C. albicans* was prevalent among neonates (60.2%), while highest infection rates of *C. parapsilosis* were observed among infants (42%), with significantly lower prevalence in neonates (26%) [71]. Similar data were reported by a multicenter pediatric and neonatal study (involving 23 centers in the United States and 19 in 15 other countries), with 48% *C. albicans* isolates and 28% *C. parapsilosis* isolates in newborns [72]. Focusing on patients of surgical intensive care units of the EUROCANDY cohort, 72.2% episodes were due to *C. albicans* while the remaining cases were ascribed to *C. parapsilosis*. However, the number of neonates, infants, and children who underwent major surgery was not specified [71]. High-risk neonates become colonized with *Candida* spp. not only

vertically during vaginal birth from their mothers, who may be receiving an azole for vaginal candidiasis, but also horizontally from colonized hospital-workers during their stay in NICU.

Although *C. albicans* remains the most common isolate in NICU, a shift to infections caused by *C. parapsilosis* and *C. tropicalis* has occurred during the last decades, and it has been associated with decreased mortality [12].

Among all ICIs, *C. albicans*, *C. parapsilosis, C. tropicalis, C. glabrata,* and *C. krusei* account for nearly 90% of isolates from blood or other sterile site cultures. Candidemia caused by other uncommon species, such as *C. guilliermondii*, and *C. lusitaniae*, is less well-known. It seems, though, to have a poorer response to antifungal treatment (frequently due to antifungal minimal inhibitory concentration -MIC- above the epidemiologic cut-off value) and a longer duration of candidemia [25].

Whereas specific inflammatory and tissue-destructive histopathologic features were found in most neonatal *C. albicans* cases, the mechanisms underlying cases of species other than *C. albicans* are still poorly understood. According to autopsy-based data, species other than *C. albicans* could involve both the gastrointestinal tract and pulmonary airways and their incidence is often underrated [73].

### **8. Epidemiology of Fungal Infections in NICUs**

Although there is an inter-site variability in the incidence of candidemia [22,71], prevention of ICIs should be an achievable evidence-based goal for every NICU [74]. NICU and PICU admissions were considered as significant predictors for mortality, with an odds ratio of 4.67 and 8.325, respectively, in the EUROCANDY cohort [71]. However, most data involve extremely preterm infants.

In specific subgroups of patients (e.g., abdominal surgical patients), ICIs are also frequent [30–33,75], but to date there are no large epidemiological studies on the incidence of ICIs in neonates who have undergone major surgery. *Candida* spp., within four weeks from admission in intensive care units, colonize the skin and mucous membranes of about 64% of critically ill neonates and can progress to invasive infection [76].

ICIs are a major cause of morbidity and mortality among critically ill patients [31,77,78] and impose an important economic burden mainly due to prolonged ICU stay, cost of antifungal drugs, and overall use of hospital resources [79,80].

In case of nosocomial ICI outbreaks, a cluster of infections could be defined when at least two cases of severe neonatal infection (i.e., bloodstream infection) occur within a defined time interval in one center with the same pathogen species in different patients: *Candida albicans* is one of the most frequently occurring microorganisms, according to a recent German surveillance system [81].

Therefore, a nosocomial ICI outbreak within a NICU could have important clinical and economic repercussions. A contaminated environment has been identified as a possible source of the outbreak: the colonized locations included wiping cloths, faucets, sinks, an operating table, puddles in the bathroom, a ventilator, and an ultrasonic probe in a recent outbreak of *Candida parapsilosis* fungemia in a Chinese hospital [82]. An emergency plan should be promptly scheduled with environmental surveillance and comprehensive interventions, such as hand hygiene and disinfection techniques. Improving both disinfection and isolation, as well as interrupting the pathway of transmission, resulted to be the key to controlling the spread of infection [83].

New methods (such as fingerprinting analysis of *Candida* isolates) can help to identify the identical strains, to investigate suspected outbreaks and to help therapeutic decisionmaking [84].

### **9. Prophylaxis of Fungal Infections**

The high-risk population of critically ill neonates benefits greatly from prompt, effective treatment and prophylactic measures. A prompt antifungal treatment is one of the most important determinants for mortality reduction. In addition, antifungal prophylaxis

given to critically ill patients at high risk for ICIs may have a positive impact on patients' outcomes, given ICIs' high morbidity and mortality rates [85,86].

Fluconazole prophylaxis has been proven to be safe and effective in neonates, reducing ICIs by more than 80% and *Candida*-related mortality by 90%, especially in highrisk preterm infants, without significant side effects or emergence of resistant fungal species [75]. Considering its long half-life plasma elimination, which allows an intermittent administration schedule, fluconazole should be administered at 3 mg/kg once a day, two times a week in the first two weeks of life whereas, from the third week of life, prophylaxis should be administered every other day [4]. The benefits of prophylaxis are less clear when incidence of ICIs is lower than 2%, and the administration should be discussed case by case, in relation to the presence of risk factors for ICIs.

There is currently clear evidence on the efficacy of fluconazole prophylaxis in the prevention of ICIs in preterm infants [87–91], but not in surgical newborns. In these neonates, fluconazole prophylaxis is not clearly suggested, although they are considered at risk of ICI as explained above. A major concern regarding a larger prophylactic use of antifungal agents, even in term infants with risk factors, is the emergence of resistant species. However, resistance to fluconazole or echinocandins in newborns is reported as rare: fluconazole-resistant *C. albicans* was seen among 1.6% of the isolates, while no echinocandins-resistant *C. albicans* was observed [23].
