|
Lopinavir/ritonavir (LPV/r)- compared to nevirapine (NVP)-based ART following receipt of single-dose nevirapine (sdNVP) for prevention of mother-to-child HIV transmission in South Africa: a cost-effectiveness analysis of the OCTANE (ACTG A5208) trial
Presented by Andrea Ciaranello (United States).
A. Ciaranello1, K. Freedberg2,3, J. Chu2, S. Lockman4,5, M. Hughes6, J. Currier7, J. McIntyre8, E. Losina9,10, R. Walensky1,3,4
1Massachusetts General Hospital, Infectious Disease, Boston, United States, 2Massachusetts General Hospital, Internal Medicine, Boston, United States, 3Harvard Medical School, Center for AIDS Research, Boston, United States, 4Brigham and Women's Hospital, Infectious Disease, Boston, United States, 5Harvard School of Public Health, Immunology and Infectious Disease, Boston, United States, 6Harvard School of Public Health, Biostatistics, Boston, United States, 7University of California Los Angeles, Infectious Disease, Los Angeles, United States, 8University of the Witwatersrand, Perinatal HIV Research Unit, Johannesburg, South Africa, 9Brigham and Women's Hospital, Orthopedic Surgery, Boston, United States, 10Boston University School of Public Health, Biostatistics, Boston, United States
Background: The OCTANE trial reports superior outcomes of LPV/r- vs. NVP-based ART. However, LPV/r is 11 times more expensive than NVP.
Methods: Using the published Cost-effectiveness of Preventing AIDS Complications model of HIV infection, with data from South Africa and the OCTANE trial, we projected life expectancy (LE) and lifetime costs (2006US$) of three alternative ART sequences after sdNVP exposure: 1) no ART (for comparison); 2) 1st-line NVP/TDF/FTC (2nd-line LPV/r/ddI/ZDV); and 3) 1st-line LPV/r/TDF/FTC (2nd-line NVP/ddI/ZDV). OCTANE data included median age (31y), CD4 count (139/µL), and time since sdNVP (17m; 10th percentile 7m; 90th percentile 45m); and 24-wk ITT viral suppression rates (1st-line efficacy; Table). Efficacies of 2nd-line LPV/r and NVP were set equal to observed 1st-line NVP efficacy, failed 2nd line ART was discontinued, and RNA monitoring was unavailable. Monthly costs were: NVP $4, LPV/r $44, TDF/FTC $16, and ZDV/ddI $27. We used WHO criteria for “very cost-effective (CE):” CE ratio < 1X South Africa GDP ($5,400)/year of life saved (YLS).
Results: 1st-line LPV/r increases projected LE and costs and is very CE ($1,250/YLS) compared to 1st-line NVP (Table). 1st-line LPV/r remains very CE unless 2nd-line NVP efficacy is ≤69% or 2nd-line LPV/r efficacy is ≥95%. Policy conclusions are insensitive to availability of 3rd-line ART or RNA monitoring, continuation of failed 2nd-line ART, and LPV/r cost.
| ART strategy | 1st-line ART efficacy (% suppression, OCTANE data) | 2nd-line ART efficacy (% suppression, assumption) | LE (years) | Mean per-person lifetime costs (US$) | Incremental CE ratio (US$/YLS) | | No ART | n/a | n/a | 1.82 | 3,530 | | | 1st line NVP | 82 | 82 | 13.15 | 13,340 | 870 | | 1st line LPV/r | 96 | 82 | 13.95 | 14,350 | 1250 |
Conclusions: Long-term projections with OCTANE data suggest that in women exposed to sdNVP a median of 17m prior, 1st-line LPV/r leads to better outcomes and is very cost-effective compared to 1st-line NVP. Policy conclusions depend critically on 2nd-line ART efficacy in settings where sdNVP is used.
|