• The primary clinical endpoint will be a global response consisting of a combined endpoint of toxicity and response to therapy (clinical, microbiologic and radiologic responses) 28 days after starting treatment with voriconazole. [ Time Frame: 28 days ] [ Designated as safety issue: Yes ]
  

• Overall mortality [ Time Frame: 7 and 28 days; 12 weeks ] [ Designated as safety issue: Yes ]
  
• % of serum concentrations within 2-5mg/L [ Time Frame: 7 and 28 days; 12 weeks ] [ Designated as safety issue: Yes ]
  
• % switched to salvage therapy or measured concentration level in control arm [ Time Frame: 7 and 28 days; 12 weeks ] [ Designated as safety issue: Yes ]
  
• Side effects [ Time Frame: 7 and 28 days; 12 weeks ] [ Designated as safety issue: Yes ]
  
• Time to global response [ Time Frame: 7 and 28 days; 12 weeks ]
  
• Cost-effectiveness of TDM [ Time Frame: 7 and 28 days; 12 weeks ] [ Designated as safety issue: Yes ]
  

Patients with haematological malignancies and chemotherapy-induced prolonged neutropenia are at risk for severe bacterial and fungal infections. These opportunistic infections can result in prolonged hospital stay, increases costs and greater mortality. Voriconazole has now been recommended as the first line agent for invasive pulmonary aspergillosis. Retrospective observational studies of voriconazole serum concentration suggest that serum concentration correlate with toxicity and clinical response. These observations were however made in small series of patients and data were collected retrospectively.

These inherent methodological flaws make it impossible to draw definite conclusions about the effect of voriconazole serum level monitoring on the outcome of IA, and therefore considered insufficient proof to recommend voriconazole concentration determination in blood as standard of care. The impact that so called serum concentration guided dosing of voriconazole will have on treatment success can only be evaluated through a prospective randomized clinical trial.

For this purpose, we designed a prospective stratified cluster randomized cross-over trial of therapeutic drug monitoring in patients with haematological disease who have developed IA. The order of periods (TDM or standard of care, each 12 months) will be randomized per centre. During the TDM episode, the voriconazole dosage will be adjusted to achieve trough blood concentrations in a predefined window of 2-5 mg/L. A sample size of n=192 is needed to detect a 20% absolute reduction in the number of treatment failures (40% to 20 %) compared to control.