Once-weekly liposomal amphotericin B as Candida prophylaxis in very low birth weight premature infants: a prospective, randomized, open-label, placebo-controlled pilot study

Objective: This study was conducted to evaluate once-weekly liposomal amphotericin B (L-AmB) for Candida prophylaxis in very low birth weight (VLBW) neonates.

Methods: This prospective, randomized, open-label, placebo-controlled study included neonates who were <32 weeks' gestational age, <7 days old, and weighing <1500 g at birth. Subjects were randomized to receive L-AmB 5 mg/kg per week or placebo (dextrose water) and were followed until 6 weeks of age. Surveillance cultures were obtained at baseline, at 72 hours, and weekly thereafter. Study drug was continued until 6 weeks after birth or the discontinuation of high-risk treatments and invasive devices, whichever occurred first. Blood cultures were obtained as clinically indicated. The primary end point was development of Candida colonization by 6 weeks' postnatal age; secondary end points included development of invasive candidiasis and occurrence of treatment-related adverse events. Safety variables included renal and hepatic function tests, incidence of grade III-IV intraventricular hemorrhage (IVH) and necrotizing enterocolitis (NEC), and mortality.

Results: Forty subjects were enrolled and randomized to receive L-AmB (12 males, 8 females; 50% white) or placebo (12 males, 8 females; 35% white). Subjects were evenly distributed by gestational age, age at enrollment, birth weight, race, and sex. Consent was withdrawn after completion of study treatment in 1 subject (L-AmB); 1 subject in each study arm died during the study; and 3 subjects were transferred back to their referring institutions (1 L-AmB, 2 placebo). Thus, 17 subjects in each arm completed all study procedures, although all 40 subjects were evaluable. Colonization before administration of study drug was noted in 4 L-AmB subjects (20%) and 1 placebo subject (5%); 1 (5%) and 3 (15%) subjects in the respective groups developed colonization while receiving study drug. No L-AmB subjects and 1 placebo subject developed candidiasis. One subject in each group died; these deaths were not considered related to study drug or fungal infection. There were no clinical differences between groups in the incidence of grade III-IV IVH, NEC, hypokalemia, nephrotoxicity, need for platelet or packed red blood cell transfusion, or mortality.

Conclusions: L-AmB 5 mg/kg once weekly was generally well tolerated in these VLBW infants. The data did not allow evaluation of efficacy. A larger, multicenter, randomized clinical trial of L-AmB for Candida prophylaxis that is appropriately powered is warranted.