Cost-Effectiveness of Dolutegravir in HIV-1 Treatment-Experienced (TE) Patients in France

OBJECTIVES To evaluate the cost-effectiveness of a new generation integrase inhibitor (INI), dolutegravir (DTG), in France, in treatment-experienced (TE) and INI-naïve HIV-infected adults with at least two classes resistance compared to raltegravir (RAL), by adapting previously published Anti-Retroviral Analysis by Monte Carlo Individual Simulation (ARAMIS) model. METHODS ARAMIS is a microsimulation Markov model with a lifetime time horizon and a monthly cycle length. Health states are defined as with or without opportunistic infection and death. In the initial cohort, efficacy and safety data were derived from a phase III study comparing DTG to RAL. Antiretroviral treatment algorithms, accounting for patient history, were based on French guidelines and experts opinion. Costs are mainly including treatment costs, routine HIV and opportunistic infection care, and death. Utilities depend on CD4+ cell count and the occurrence of opportunistic infections. RESULTS The ARAMIS model indicates in the TE population that DTG compared to RAL over a life time is associated with 0.35 additional quality-adjusted life years (QALY; 10.75 versus 10.41) and additional costs of €7,266 (€390,001 versus €382,735). DTG increased costs are mainly related to a 9.1-month increase in life expectancy for DTG compared with RAL, and consequently a longer time spent on ART. The incremental cost-effectiveness ratio (ICER) for DTG compared with RAL is €21,048 per QALY gained. About 83% and 14% of total lifetime costs are associated with antiretroviral therapy and routine HIV care respectively. Univariate deterministic sensitivity analyses demonstrate the robustness of the model. CONCLUSION DTG is cost-effective in the management of TE INI naive patients in France, from a collective perspective. These results could be explained by the superior efficacy of DTG in this population and its higher genetic barrier to resistance compared to RAL. These data need to be confirmed with longer-term real life data.