• microbiology
• neonatology

Preterm neonates in neonatal intensive care units (NICUs) are at high risk of undergoing various invasive procedures together with immature immunity. They are more likely to develop nosocomial infections, including fungal septicaemia.1 2 Candida species have become the third most common pathogens for late-onset septicaemia in newborn infants and are associated with a mortality rate of 20–34%.3 However, Candida utilis is seldom reported as the cause of candidaemia in humans.4 As a result, there is a lack of reliable information about its clinical features, treatment and prognosis. In the present study we performed a retrospective review of premature infants with C. utilis candidaemia (CUC) occurring in our NICU.

The retrospective study was conducted in the NICU of a public hospital in Beijing. Patients with CUC (see online Appendix) during hospitalisation were enrolled. Data including demographic characteristics, medical history, invasive procedures, medications, laboratory data and outcomes were collected.

From January 2009 to December 2017, CUC occurred in 13 premature infants (0.16%) out of a total of 79 863 neonates admitted to the NICU. The mean±SD birth gestational age of the 13 preterm infants with CUC was 29.85±1.14 weeks and the mean±SD birth weight was 1329.23±409.05 g (table 1).

The clinical manifestations are shown in table 2, the most frequent signs being poor response (9/13) and SaO₂ instability (11/13). In the 13 CUC cases (online supplemental table S1), thrombocytopaenia was detected in 11 cases, granulocytopaenia in eight cases and elevated C-reactive protein levels in seven cases on the day of onset. Cultures of cerebrospinal fluid were all negative. There was no sign of bone and joint infection, fungal endophthalmitis, renal infection, carditis or involvement of other end organs. All of the isolated strains were sensitive to amphotericin B, fluconazole, F-5 fluorine cytosine, voriconazole and itraconazole.

A peripherally inserted central catheter was removed after onset of the disease. Fluconazole was administered intravenously at a therapeutic dose of 8–10 mg/kg/day until the clinical symptoms of infection disappeared and two consecutive negative blood cultures were obtained at an interval of 1 week. The mean duration of anti-infective therapy was 24 days (range 18–31), including four cases receiving normal saline or plasma volume expansion therapy and three cases receiving vasoactive drugs. All the patients recovered from CUC after fluconazole treatment.

There have been rare reports about human CUC pathologies. A few available reports of CUC infection were in old people (68–88 years) and children (0–5 years), and treatments with flucytosine, fluconazole, liposomal amphotericin B or caspofungin improved the patients’ condition.5

Candida spp, especially C. albicans, are more likely to be complicated by involvement of the end organs such as the brain, heart, kidney, eye and skeleton (online supplemental table S2)6 while, in the CUC cases in our study, no fungal endophthalmitis, renal infection, carditis, bone and joint infection or involvement of other end organs was observed. A G test was performed in seven patients and the level was elevated in only three cases, suggesting that a negative G test result does not mean the absence of CUC.

Our study reported clinical characteristics, treatment and prognosis of CUC in 13 infants. The infection was controlled successfully in all cases after intravenous administration of fluconazole. It is noteworthy that, in the three infants infected with CUC reported by Lukić-Grlić et al,7 one responded well to amphotericin B and the other two failed to respond to fluconazole and were later cured after replacement with amphotericin B and caspofungin, respectively. It was suggested that fluconazole could be used as the initial choice and, if it does not work effectively, it should be replaced by amphotericin B or echinocandins.