A N Phillips1, M Youle, M Johnson, C Loveday. 1. Royal Free Centre for HIV Medicine, Royal Free & University College Medical School, University College London, London, UK.
Abstract
OBJECTIVE: To use a stochastic model to gain insights into the consequence for resistance development of different drug use patterns. METHODS: We consider use of three drugs (A, B and C) where for each drug one and only one viral mutation is associated with ability to replicate (effective reproductive ratio, R > 1) in the presence of that drug as monotherapy. For drug A mutation is a, etc. We define eight populations of short-lived infected cells that live 1 day: Vo with no mutations a, b, c; Va with mutation a only, Vab with mutations a and b, etc. A random number generator was used to determine whether mutations occur in any one round of replication and to sample from a Poisson distribution to determine for each cell the number of cells of the same population created in the next generation, using the R operative at that time. Values of R depended on drug exposure, cost of resistance and availability of target cells. RESULTS: Treatment strategies and the resulting percentage (over 100 runs) developing full "resistance" in 1500 days (Vabc not equal 0) were: (i) ABC 1500 days 0%; (ii) A 300 days, AB 300 days, ABC 900 days 100%; (iii) AB 300 days, ABC 1200 days 33%; (iv) ABC 2/3 1500 days 15%; (v) ABC 1/2 1500 days 100%; (vi) ABC 50 days, no drugs 50 days, for 1500 days 1%, where ABC 2/3 means on-drug for 2 days in every 3, ABC 1/2 represents on-drug for 1 day in every 2, and represents suboptimal adherence. CONCLUSIONS: This model helps to develop understanding of key principles concerning development of resistance under different patterns of treatment use.
OBJECTIVE: To use a stochastic model to gain insights into the consequence for resistance development of different drug use patterns. METHODS: We consider use of three drugs (A, B and C) where for each drug one and only one viral mutation is associated with ability to replicate (effective reproductive ratio, R > 1) in the presence of that drug as monotherapy. For drug A mutation is a, etc. We define eight populations of short-lived infected cells that live 1 day: Vo with no mutations a, b, c; Va with mutation a only, Vab with mutations a and b, etc. A random number generator was used to determine whether mutations occur in any one round of replication and to sample from a Poisson distribution to determine for each cell the number of cells of the same population created in the next generation, using the R operative at that time. Values of R depended on drug exposure, cost of resistance and availability of target cells. RESULTS: Treatment strategies and the resulting percentage (over 100 runs) developing full "resistance" in 1500 days (Vabc not equal 0) were: (i) ABC 1500 days 0%; (ii) A 300 days, AB 300 days, ABC 900 days 100%; (iii) AB 300 days, ABC 1200 days 33%; (iv) ABC 2/3 1500 days 15%; (v) ABC 1/2 1500 days 100%; (vi) ABC 50 days, no drugs 50 days, for 1500 days 1%, where ABC 2/3 means on-drug for 2 days in every 3, ABC 1/2 represents on-drug for 1 day in every 2, and represents suboptimal adherence. CONCLUSIONS: This model helps to develop understanding of key principles concerning development of resistance under different patterns of treatment use.
Authors: N M Ferguson; C A Donnelly; J Hooper; A C Ghani; C Fraser; L M Bartley; R A Rode; P Vernazza; D Lapins; S L Mayer; R M Anderson Journal: J R Soc Interface Date: 2005-09-22 Impact factor: 4.118
Authors: Jennifer A Collins; M Gregory Thompson; Elijah Paintsil; Melisa Ricketts; Joanna Gedzior; Louis Alexander Journal: J Virol Date: 2004-01 Impact factor: 5.103