Antifungals in systemic neonatal candidiasis

Fungal infections are common in the newborn period, especially among premature neonates, and are responsible for considerable morbidity and mortality. Currently, three classes of antifungals are commonly used in the treatment of systemic fungal infections in neonates: the polyene macrolides (e.g. amphotericin B [deoxycholate and lipid preparations]); the azoles (e.g. fluconazole); and the fluorinated pyrimidines (e.g. flucytosine). The echinocandins (e.g. caspofungin and micafungin) are a newer class of antifungals which shows promise in this population. The available kinetic data on amphotericin B deoxycholate in neonates are derived from very small studies and exhibit considerable variability. There are no kinetic data available for the use of lipid preparations in this population and, again, much has been inferred from adult studies. The information available for flucytosine is also limited but appears similar to what is observed in adults. Fluconazole has the most neonatal pharmacokinetic data, which show slightly less variability than the other antifungals. Genomic factors which affect the metabolism of amphotericin B and fluconazole may explain some of the observed variability. Most of the data for the efficacy of antifungal drugs in neonates are derived from retrospective studies and case reports. The data for amphotericin B deoxycholate and flucytosine are limited. There are more data for the liposomal and lipid complex preparations of amphotericin B and for fluconazole in this population. These support the use of these drugs in neonates, but because of their largely noncomparative nature they can not define the optimal dosage or duration of therapy. Amphotericin B deoxycholate is primarily nephrotoxic. It also induces electrolyte abnormalities and is to a lesser degree cardiotoxic. This toxicity in neonates appears similar to published data in older children and adults. While the lipid preparations of amphotericin B owe their existence to a presumed decrease in toxicity, the observed toxicity in neonates appears to be equal to that seen with the deoxycholate, although it should be noted that the lipid preparations are usually given at much higher dosages. Fluconazole toxicity appears to be milder and less frequent in this population than is seen with amphotericin B. In the final analysis, we do not have sufficient data to define the pharmacokinetic profiles, optimal dose or duration of therapy, or toxicity for any of these compounds in neonates. Further studies are necessary if the optimisation of antifungal therapy in this population is to continue.