Is Idelalisib Cost-Effective for Refractory/Relapsed Chronic Lymphocytic Leukemia? a Decision Analysis in the Second-Line Setting

Background and Aims:

A number of therapeutic options proved to be effective for relapsed or refractory chronic lymphocytic leukemia (R/R CLL), but the optimal sequence of therapies as well as their economic impact is still to be ascertained. Idelalisib, a first-in-class PI3k inhibitor, has been recently approved in combination with Rituximab (IR) for R/R CLL. Therefore, we aimed at assessing the cost-effectiveness of IR as compared with fludarabine-cyclophosphamide-rituximab (FCR), bendamustine-rituximab (BR), bendamustine-ofatumomab (BO), rituximab (R) and ofatumomab (O) for CLL patients requiring a second-line treatment, in the perspective of the Italian national health-care system.

Methods:

A life-long Markov model simulated patients moving through 5 health states: progression-free on intravenous therapy, progression-free on oral extended therapy, progression-free off-therapy, progressed disease, death. Pairwise comparisons between 2^nd -3^rd line treatment sequences were conducted: each pair included a sequence adopting IR for the 2^nd line therapy, followed either FCR, BR, BO, R or O for the 3^rd line, versus a sequence modelling IR for the 3^rd line therapy. Progression-free survival, toxicities, treatment schedules and duration of treatment were obtained from phase II-III studies (Furman et al. 2014, Byrd et al 2014, Fisher et al. 2011, Cortelezzi et al. 2014, Awan et al. 2014). The probability of progression reported by published studies was adjusted for the median number of treatment lines according to the reported hazard ratio of 1.4 (Fisher et al. 2011, Holzer-Goor et al. 2014). Progressed disease and progression-free health states were assigned a utility of 0.65 and 0.85, respectively (Beusterien et al. 2010). Quality of life decrements and costs (Capri et al. 2007) were applied to the portion of patients incurring grade 3-4 adverse events (anemia, neutropenia, pneumonia, thrombocytopenia, diarrhea). The following unit costs (average wholesale price) were considered: idelalisib 150 mg (€66.67), rituximab 100 mg (€277), ofatumumab 100 mg (€241). Bendamustine, fludarabine and cyclophosphamide were not valued because they are not refunded. Median charge for intravenous drugs administration (€288/day) was obtained from a published retrospective analysis (Pasdera et al. 2013). Future life gains and costs were discounted at a 3% annual rate. First and second-order sensitivity analysis was performed.

Results:

Second-line treatment with IR was expected to improve quality-adjusted life expectancy (QALE) to 4.49 quality-adjusted life years (QALYs), that is prolongation of 0.94-1.68 QALYs as compared with combination therapies and 2.09-2.62 QALYs versus single agent B or O. Cumulative discounted health-care cost for patients receiving second line treatment with IR (eventually followed by 3rd treatment with BR or BO) were expected to be €169,291-171,217 but incremental cost-utility ratio (ICUR) of IR versus BR and BO resulted to be lower than €40,000/QALY with a probability higher than 90%. Second line IR (eventually followed by 3^rd line FCR) as compared with second line FCR (eventually followed by 3^rd line IR) was calculated to prolong life expectancy by 0.94 QALYs at an incremental cost lower than €40,000/QALY in 81% of the simulations (Monte-Carlo analysis). Comparison between IR and single-agent immunotherapies (R and O), showed a prolonged life expectancy at an incremental cost lower than €20,000/QALY. Results were sensitive to the time horizon of the analysis and idelalisib unit cost.

Conclusions:

The lack of head-to-head studies in the different CLL treatment lines makes indirect comparisons among treatments possible only using simulation analyses. Hence, a decision analysis allowed us to compare IR with five treatments in the second-line setting: IR was expected to provide a reasonable value for money as compared with FCR, BR, BO, R and O and the results were quite robust at sensitivity analysis.