Cost-Effectiveness of Antifungal Strategies in High-Risk Neutropenic Patients

Background: Invasive fungal infections (IFIs) incur significant morbidity and mortality among neutropenic patients with hematological malignancies. Prevention and early antifungal treatment include prophylaxis where antifungal agent is started along with neutropenia; empirical therapy where antifungal agent is initiated in persistently febrile patients at least four days after neutropenia onset; preemptive therapy where antifungal agent is initiated later for a suspected IFI based on clinical symptoms, lung imaging, or biological markers. Without data from clinical trials comparing antifungal strategies, determining an optimal antifungal strategy for these patients is challenging. Objective: To determine the cost-effectiveness of all possible antifungal strategies. Target Population: Adult patients with hematological malignancies in induction chemotherapy at high risk for IFIs. Interventions: Prophylaxis with either fluconazole or posaconazole, empirical strategy, and preemptive strategy with fist-line antifungal therapy being amphotericin B, liposomal amphotericin B or caspofungin (Table). Design: Cost-effectiveness decision model. The study cohort encountered seven successive chance nodes during hospital stay: having fever; in those having fever, IFI incidence; Aspergillus or Candida species among IFI; admission in intensive care unit; dying from IFI or underlying hematological malignancy; in patients alive, having severe nephrotoxicity as defined by a twofold increase in baseline serum creatinin; dying from severe nephrotoxicity. Antifungal strategies modified the probabilities of IFI and severe nephrotoxicity that depended on both the duration and the type of antifungal drugs administered. Data Sources: PREVERT trial1, effectiveness data published to December 2007, probabilities of ICU admission and in-hospital mortality according to the occurrence of IFI in the French DRG database, life expectancy of French patients with acute myeloid leukemia, and actual French hospitalization costs (2007 euros). Time Horizon: Lifetime. Perspective: Societal. Outcome Measures: Incremental cost (euros) per discounted life-year saved averaged from 100 samples of 1000 patients (second-order probabilistic Monte Carlo simulations). Results (Table): Fluconazole prophylaxis followed by ampho-B treatment was the cheapest antifungal strategy. Posaconazole prophylaxis followed by ampho-B was nearly cost-effective (59,610 € per discounted year of life gained). Other strategies were either dominated or beyond usual societal thresholds of what may be worth it. Similar results were found in sensitivity analyses among plausible ranges. Conclusions: As compared to previous studies showing that new antifungal drugs were cost-effective within a single strategy, empirical and preemptive antifungal strategies were dominated by prophylaxis strategies. Table: Incremental cost-effectiveness ratio of antifungal strategies in high-risk neutropenic patients Total cost (2007 euros) Years of life gained ICER IFI (%) Aspergillus (%) Nephro-toxicity (%) Antifungals’ cost (2007 euros) Fluconazole then amphoB 35606 2.3448 -- 3.81% 2.61% 4.22% 909 Fluconazole then L-amphoB 36025 2.3457 (extended dominance) 3.81% 2.61% 3.67% 1430 Empirical amphoB 36033 2.3433 (dominated) 3.38% 2.3% 5.98% 1914 Posaconazole then amphoB 36065 2.3525 59,610 € 1.2% 0.81% 3.67% 2646 Preemptif amphoB 36160 2.3449 (dominated) 3.89% 2.67% 4.42% 1247 Posaconazole then L-amphoB 36389 2.3532 462,857 € 1.2% 0.81% 3.21% 3055 Fluconazole then Caspo 36557 2.3459 (dominated) 3.81% 2.61% 3.53% 1985 Preemptif L-amphoB 36616 2.3455 (dominated) 3.89% 2.67% 3.84% 1809 Posaconazole then Caspo 36828 2.3533 4,390,000 € 1.2% 0.81% 3.13% 3509 Empirical L-amphoB 37308 2.3462 (dominated) 3.38% 2.3% 4.39% 3239 Preemptif Caspo 37346 2.3458 (dominated) 3.89% 2.67% 3.55% 2597 Empirical Caspo 39123 2.3473 (dominated) 3.38% 2.3% 3.47% 5080