In Silico Evaluation of HIV Short-cycle Therapies with Dynamical Models

Abstract : We aim at quantifying the effect of on/off strategies for various treatment regimens. These so-called short-cycle therapies, following the FOTO trial (Cohen et al., HIV Clin. Trials, 2007), are currently tested in phase II/III such as in 4D ANRS 162 - 4/3 days on/off (DeTruchis et al., IAS, 2016) and BREATHER - 5/2 days on/off (Breather trial group, Lancet, 2016). Mechanistic models based on Ordinary Differential Equations can model HIV and CD4+ T cells trajectories. Here, we aim at predicting the results of the current trials evaluating short-cycle therapies and suggest other strategies by using in silico trials based on mechanistic models.  Using estimations from previous clinical trials such as ALBI (Prague et al., Biometrics, 2012), we show that short-cycle therapy would not be successful for old therapies based on two nucleoside analogues such as AZT+3TC or ddI+d4T. We estimated that the regimens have to be twice as potent as AZT+3TC to ensure viral load suppression using a 5/2 design. Single-round infectivity assays allow quantifying the instantaneous inhibitory potential (IIP), which is established as a measure of regimens activity. In Jilek et al., Nat. Med., 2012, efavirenz regimens are at least 2.8 times more efficient than AZT+3TC, which is enough to guarantee the success of BREATHER in most patients. We also demonstrate that 4/3 designs are likely to be more difficult to maintain in a long-term depending on patients’ characteristics at inclusion. This analysis is applied to the ANRS C03 Aquitaine observational cohort of HIV-infected patients (Prague et al., Biometrics, 2016). We focus on EFV, TDF, AZT, 3TC, ABC, FTC, LPV/r, DRV/r and ETR.
Liste complète des métadonnées

https://hal.inria.fr/hal-01579070
Contributor : Mélanie Prague <>
Submitted on : Wednesday, August 30, 2017 - 1:35:14 PM
Last modification on : Tuesday, September 18, 2018 - 4:24:02 PM

File

PRAGUE_Keystone_poster_Final.p...
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01579070, version 1

Citation

Mélanie Prague, Daniel Commenges, Rodolphe Thiébaut. In Silico Evaluation of HIV Short-cycle Therapies with Dynamical Models. Keystone Symposium, Aug 2017, Este Park, United States. ⟨hal-01579070⟩

Share

Metrics

Record views

284

Files downloads

35