Amphotericin B used as a continuous infusion is safer

Introduction

Many patients in an ICU are intestinal carriers of yeasts with an inherent risk of infections if bowel perforation occurs. Resistance to azoles is increasing, while treatment with conventional amphotericin B (AM-B) is associated with potential toxicity. AM-B given in a continuous 24-hour infusion may be less toxic compared with the conventional 4-hour to 6-hour infusion rate of AM-B [1, 2]. AM-B given in a continuous 24-hour infusion is only evaluated in a tertiary care setting, with a predominance of immunocompromised patients. We evaluated the feasibility and safety of continuous AM-B treatment in critically patients with suspected or proven yeast or fungal infections.

Methods

This is an observational retrospective analysis for the side effects in consecutive patients treated with AM-B from January 2003 to December 2008 in our ICU department. During the investigation period patients received amphotericin B: 40 mg/24 hours or a lower dose if the MDRD clearance was less than 60 ml/min. During the treatment the dose was adjusted to the desired therapeutic range of 200 to 1,000 μg/l according to measured plasma levels.

Results

The mean treatment duration was 12.3 ± 6.3 days with a dose of 32.1 ± 12.2 mg/24 hours. Of the 10 patients who died, seven died after the termination of the AM-B therapy, without signs of an active yeast or fungal infection. Within the first week three patients died, necropsy was carried out in one case, demonstrating a PCP infection, and in two patients a role of antibiotic failure cannot be ruled out. All other patients with proven yeast or fungal infection demonstrated a clinical recovery of their infection. The AM-B concentration was measured in 113 blood samples, nine samples had a level <200 μg/l and 10 a level >1000 μg/l. Renal impairment, defined as more than 1.5 times the creatinine at the start of the treatment with AM-B, occurred in 9% and was not evaluable in 6% due to unresolving renal replacement therapy dependency at the start of AM-B. None of the evaluated patients developed a creatinine more than 2.0 times the baseline value. Temporary elevation of liver enzymes was seen in 23%, without the need for dose modification. Hypokalemia (K⁺ <3.0 mmol/l) was observed in one patient.

Conclusions

Compared with the conventional infusion rate of AM-B, we conclude that continuous 24-hour infusion seems a feasible and safer treatment alternative in patients with invasive yeast or fungal infections.