Efficacy and safety of voriconazole in real-life setting in Belgium: results from the Voribel-1 trial

Abstract number: 1733_781

Jacobs F., Selleslag D., Aoun M., Sonnet A., Gadisseur A., Damas P., Laterre P., Van Der Meeren O.

Objective: Voriconazole (VRC) is standard of care for invasive aspergillosis (IA) and a key treatment option for candidaemia and invasive candidiasis (CC). VRC has been granted reimbursement by the Belgian Health System in 2003 under the condition to confirm its safety, efficacy and cost-efficacy through an observational study in a local real-life setting. The clinical results are presented here.

Methods: A national multicentre prospective non-interventional study was initiated after approval by ethics committees. All patients (p) receiving VRC and presenting with proven/probable IA or CC were allowed to enter a 12-week follow-up. A sample size of at least 120p was needed to test the hypothesis that based on average cost, the real-life setting is not inferior to the pharmacoeconomic model assumed in the reimbursement dossier.

Results: 141p started treatment with VRC at 14 sites between 3/2004 and 9/2005. Most p were adults with proven/probable IA (116); 9 adults had CC, 5p were children (all with CC) and 11p had a primary diagnosis outside the scope of the study. Most p (90%) were immunocompromised and suffered from haematological conditions (leukaemia 43%, stem-cell transplantation 21%) or had undergone solid organ transplantation (9%). In adult p with IA, a complete or partial clinical response was observed in 50% of p after median treatment duration of 50 days; the mortality rates by day 14, 42 and 84 were 14%, 27% and 42% respectively. Most patients (64%) were treated with VRC only and didn't switch, combine or interrupt therapy; 41% of p received the oral formulation only and 29% received both the IV and oral forms sequentially. Serious adverse events regarded as possibly drug-related were reported in 3p (2%); 23p (16%) discontinued VRC for safety issues. No new or unexpected adverse event was reported.

Discussion: In p with proven/probable IA, the response rate of 50% in this real-life setting confirms the results of the landmark randomised trial by Herbrecht et al (53%); this result is achieved despite a shorter median treatment duration (50 vs 77 days). Not surprisingly, the death rate at week 12 in our non-selected p population is slightly higher than in the randomised series (42% vs 29%).

Conclusions: In this large real-life cohort, VRC demonstrates very similar safety and efficacy for patients with IA than in published randomised trial. The shorter treatment duration and the broad use of oral therapy are remarkable from an economic perspective.