Azole-based chemoprophylaxis of invasive fungal infections in paediatric patients with acute leukaemia: an internal audit.

OBJECTIVES: Children and adolescents with acute myeloid leukaemia (AML) and recurrent acute leukaemias (RALs) are at high risk of life-threatening invasive fungal infections (IFIs). We analysed implementation, safety and efficacy of a standard operating procedure for oral, azole-based, mould-active antifungal prophylaxis. METHODS: Patients with AML and RALs aged ≥13 years received 200 mg of posaconazole three times daily and patients aged 2-12 years received 200 mg of voriconazole two times daily from the completion of chemotherapy until haematopoietic recovery. Algorithms for fever or focal findings in all patients with haematological malignancies included blood cultures, high-resolution CT and other appropriate imaging, serial serum galactomannan, invasive diagnostics and pre-emptive therapy with change in class if on antifungal medication. RESULTS: From 2006 to 2010, 40 patients (0.8-17 years; 21 males) with newly diagnosed AML (n = 31) or RAL (n = 9) were admitted, of whom 36 received a total of 149 courses of chemotherapy (reasons for exclusion: contraindications and early death ≤3 days). Azole prophylaxis was given in 87.2% (n = 130/149) of episodes. Pre-emptive antifungal therapy for pulmonary infiltrates was initiated in 5/36 (13.9%) patients or 6/130 (4.6%) episodes for a duration of 3-22 days. No proven or probable IFIs occurred. Adverse events (AEs) were common but mostly low grade and reversible. Three courses (2.3%) were discontinued due to AEs. In simultaneously admitted new patients with acute lymphatic leukaemia (ALL; n = 101) and paediatric lymphomas (n = 29) not receiving standard antifungal prophylaxis, proven/probable IFIs occurred in 4 patients with ALL (4.0%) and 7/130 patients (5.4%) received pre-emptive therapy. CONCLUSIONS: Azole-based, mould-active antifungal prophylaxis in high-risk paediatric patients with AML and RALs was satisfactorily implemented, well tolerated and effective. The low rate of IFIs in patients with ALL/lymphoma supports the lack of a general indication for prophylaxis in this population in the presence of a diagnostic and therapeutic algorithm.

Monitoring of voriconazole plasma concentrations in immunocompromised paediatric patients.

OBJECTIVES: Voriconazole is approved for management of invasive fungal diseases (IFDs) in paediatric patients. We analysed plasma trough concentrations and explored their association with endpoints of antifungal therapy. PATIENTS AND METHODS: The cohort included 74 immunocompromised patients (0.2-18 years of age) who received 101 courses of voriconazole for possible (7) and probable/proven (13) IFDs, as prophylaxis (79) or empirical therapy (2). Voriconazole was given intravenously (4), intravenously and orally (15) and orally (82) at recommended dosages until intolerance or maximum efficacy. IFDs and outcomes were assessed by EORTC/MSG consensus criteria. RESULTS: Voriconazole was administered at a median maintenance dosage of 4.8 mg/kg twice daily (range 2.2-17.4) for a median of 40 days (range 6-1002). Trough plasma concentrations at steady state (251 samples; 3.4 ±â€Š4.3/patient) ranged from <0.2 to 14.9 mg/L with high intra- and inter-individual variability and no apparent relationship to dose (P = 0.074, ANOVA). Of the samples 22%, 42% and 58% had voriconazole concentrations <0.2, ≤0.5 and ≤1.0 mg/L, respectively. Adverse events (AEs) occurred in 77/101 (76.2%) courses and were mostly grade I or II. Ten courses (9.9%) were discontinued due to AEs. Treatment success was observed in 8/20 patients (40%) with IFDs, and in 67/81 courses (82.7%) of empirical therapy/prophylaxis. There were no consistent correlations between dose, trough concentrations and laboratory/clinical AEs or treatment response, and proposed threshold values were not discriminative. CONCLUSIONS: Voriconazole had acceptable safety and useful efficacy in the management of paediatric IFDs. Pharmacokinetic variability was high and no predictable dose-concentration-effect relationships were observed.