Limits...
A model of HIV drug resistance driven by heterogeneities in host immunity and adherence patterns.

Bershteyn A, Eckhoff PA - BMC Syst Biol (2013)

Bottom Line: Population transmission models of antiretroviral therapy (ART) and pre-exposure prophylaxis (PrEP) use simplistic assumptions--typically constant, homogeneous rates--to represent the short-term risk and long-term effects of drug resistance.In the case of accidental PrEP use during infection, rapid transitions between adherence states and/or weak immunity fortifies the "memory" of previous PrEP exposure, increasing the risk of future drug resistance.This model framework provides a means for analyzing individual-level risks of drug resistance and implementing heterogeneities among hosts, thereby achieving a crucial prerequisite for improving population-level models of drug resistance.

View Article: PubMed Central - HTML - PubMed

Affiliation: Epidemiological Modeling Group, Intellectual Ventures Laboratory, Washington, USA. abershteyn@intven.com

ABSTRACT

Background: Population transmission models of antiretroviral therapy (ART) and pre-exposure prophylaxis (PrEP) use simplistic assumptions--typically constant, homogeneous rates--to represent the short-term risk and long-term effects of drug resistance. In contrast, within-host models of drug resistance allow for more detailed dynamics of host immunity, latent reservoirs of virus, and drug PK/PD. Bridging these two levels of modeling detail requires an understanding of the "levers"--model parameters or combinations thereof--that change only one independent observable at a time. Using the example of accidental tenofovir-based pre-exposure prophyaxis (PrEP) use during HIV infection, we will explore methods of implementing host heterogeneities and their long-term effects on drug resistance.

Results: We combined and extended existing models of virus dynamics by incorporating pharmacokinetics, pharmacodynamics, and adherence behavior. We identified two "levers" associated with the host immune pressure against the virus, which can be used to independently modify the setpoint viral load and the shape of the acute phase viral load peak. We propose parameter relationships that can explain differences in acute and setpoint viral load among hosts, and demonstrate their influence on the rates of emergence and reversion of drug resistance. The importance of these dynamics is illustrated by modeling long-lived latent reservoirs of virus, through which past intervals of drug resistance can lead to failure of suppressive drug regimens. Finally, we analyze assumptions about temporal patterns of drug adherence and their impact on resistance dynamics, finding that with the same overall level of adherence, the dwell times in drug-adherent versus not-adherent states can alter the levels of drug-resistant virus incorporated into latent reservoirs.

Conclusions: We have shown how a diverse range of observable viral load trajectories can be produced from a basic model of virus dynamics using immunity-related "levers". Immune pressure, in turn, influences the dynamics of drug resistance, with increased immune activity delaying drug resistance and driving more rapid return to dominance of drug-susceptible virus after drug cessation. Both immune pressure and patterns of drug adherence influence the long-term risk of drug resistance. In the case of accidental PrEP use during infection, rapid transitions between adherence states and/or weak immunity fortifies the "memory" of previous PrEP exposure, increasing the risk of future drug resistance. This model framework provides a means for analyzing individual-level risks of drug resistance and implementing heterogeneities among hosts, thereby achieving a crucial prerequisite for improving population-level models of drug resistance.

Show MeSH

Related in: MedlinePlus

Adherence patterns influence the dynamics of drug resistance. (a) Dynamics of drug resistance assuming 50% adherence with periodic adherent and non-adherent spans lasting 10 days (left panel), 30 days (middle panel), or 50 days (right panel). TDF-DP concentration is overlaid in gray for reference. (b) Example of how a Markov model for drug adherence could represent underlying drivers of missed doses. (c) Dynamics of drug resistance using a two-state Markov model with exponentially distributed duration spent in the adherent and non-adherent states, each with an average duration of 30 days. This is the stochastic version of the periodic model in the middle panel of (a), providing an example of how such a model could be driven stochastically to reflect this aspect of real-life adherence patterns.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3643872&req=5

Figure 5: Adherence patterns influence the dynamics of drug resistance. (a) Dynamics of drug resistance assuming 50% adherence with periodic adherent and non-adherent spans lasting 10 days (left panel), 30 days (middle panel), or 50 days (right panel). TDF-DP concentration is overlaid in gray for reference. (b) Example of how a Markov model for drug adherence could represent underlying drivers of missed doses. (c) Dynamics of drug resistance using a two-state Markov model with exponentially distributed duration spent in the adherent and non-adherent states, each with an average duration of 30 days. This is the stochastic version of the periodic model in the middle panel of (a), providing an example of how such a model could be driven stochastically to reflect this aspect of real-life adherence patterns.

Mentions: Unlike suppressive therapy, for which pauses in treatment lead to subclinical drug concentrations that drive resistance, accidental PrEP use during infection can drive resistance even with perfect adherence. Even so, we found that the dynamics of adherence determine the proportion of time spent with drug resistance. For a fixed fraction of doses taken, this depends on the duration of adherent and non-adherent spans another variable is the degree of periodicity or stochasticity in these durations. In the examples shown, we used a simple assumption of 50% adherence with immune pressure p fixed at 10-5, a value at which resistance occurs in 37 days of constant monotherapy, and reversion after prolonged monotherapy occurs after 24 days. We compared a perfectly periodic schedule to the opposite extreme of a Markov process. For periodic 10-day transitions between adherence states, resistance cannot develop in a single span of monotherapy, but leads to eventual and sustained resistance with no opportunity for reversion (Figure 5a, left panel). Longer durations of adherence states permit reversion of resistance, leading to a smaller proportion of time spent in the resistant state (Figure 4a). In a stochastically driven model, elaborate Markov chains can be constructed with this model to represent different causes of adherence that may lead to different durations of dose-taking and dose-missing, as shown in Figure 5b. A simple, two-state Markov model with transition rates of (30 days)-1 allows for wide stochastic variation in the time spent in adherence states, including occasional longer intervals spent in states that favor drug resistance (Figure 5c). These longer dwell times may be the main drivers of resistance in individuals whose time to resistance is long compared to the average transition rate between states (e.g., due to a higher immune pressure p).


A model of HIV drug resistance driven by heterogeneities in host immunity and adherence patterns.

Bershteyn A, Eckhoff PA - BMC Syst Biol (2013)

Adherence patterns influence the dynamics of drug resistance. (a) Dynamics of drug resistance assuming 50% adherence with periodic adherent and non-adherent spans lasting 10 days (left panel), 30 days (middle panel), or 50 days (right panel). TDF-DP concentration is overlaid in gray for reference. (b) Example of how a Markov model for drug adherence could represent underlying drivers of missed doses. (c) Dynamics of drug resistance using a two-state Markov model with exponentially distributed duration spent in the adherent and non-adherent states, each with an average duration of 30 days. This is the stochastic version of the periodic model in the middle panel of (a), providing an example of how such a model could be driven stochastically to reflect this aspect of real-life adherence patterns.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3643872&req=5

Figure 5: Adherence patterns influence the dynamics of drug resistance. (a) Dynamics of drug resistance assuming 50% adherence with periodic adherent and non-adherent spans lasting 10 days (left panel), 30 days (middle panel), or 50 days (right panel). TDF-DP concentration is overlaid in gray for reference. (b) Example of how a Markov model for drug adherence could represent underlying drivers of missed doses. (c) Dynamics of drug resistance using a two-state Markov model with exponentially distributed duration spent in the adherent and non-adherent states, each with an average duration of 30 days. This is the stochastic version of the periodic model in the middle panel of (a), providing an example of how such a model could be driven stochastically to reflect this aspect of real-life adherence patterns.
Mentions: Unlike suppressive therapy, for which pauses in treatment lead to subclinical drug concentrations that drive resistance, accidental PrEP use during infection can drive resistance even with perfect adherence. Even so, we found that the dynamics of adherence determine the proportion of time spent with drug resistance. For a fixed fraction of doses taken, this depends on the duration of adherent and non-adherent spans another variable is the degree of periodicity or stochasticity in these durations. In the examples shown, we used a simple assumption of 50% adherence with immune pressure p fixed at 10-5, a value at which resistance occurs in 37 days of constant monotherapy, and reversion after prolonged monotherapy occurs after 24 days. We compared a perfectly periodic schedule to the opposite extreme of a Markov process. For periodic 10-day transitions between adherence states, resistance cannot develop in a single span of monotherapy, but leads to eventual and sustained resistance with no opportunity for reversion (Figure 5a, left panel). Longer durations of adherence states permit reversion of resistance, leading to a smaller proportion of time spent in the resistant state (Figure 4a). In a stochastically driven model, elaborate Markov chains can be constructed with this model to represent different causes of adherence that may lead to different durations of dose-taking and dose-missing, as shown in Figure 5b. A simple, two-state Markov model with transition rates of (30 days)-1 allows for wide stochastic variation in the time spent in adherence states, including occasional longer intervals spent in states that favor drug resistance (Figure 5c). These longer dwell times may be the main drivers of resistance in individuals whose time to resistance is long compared to the average transition rate between states (e.g., due to a higher immune pressure p).

Bottom Line: Population transmission models of antiretroviral therapy (ART) and pre-exposure prophylaxis (PrEP) use simplistic assumptions--typically constant, homogeneous rates--to represent the short-term risk and long-term effects of drug resistance.In the case of accidental PrEP use during infection, rapid transitions between adherence states and/or weak immunity fortifies the "memory" of previous PrEP exposure, increasing the risk of future drug resistance.This model framework provides a means for analyzing individual-level risks of drug resistance and implementing heterogeneities among hosts, thereby achieving a crucial prerequisite for improving population-level models of drug resistance.

View Article: PubMed Central - HTML - PubMed

Affiliation: Epidemiological Modeling Group, Intellectual Ventures Laboratory, Washington, USA. abershteyn@intven.com

ABSTRACT

Background: Population transmission models of antiretroviral therapy (ART) and pre-exposure prophylaxis (PrEP) use simplistic assumptions--typically constant, homogeneous rates--to represent the short-term risk and long-term effects of drug resistance. In contrast, within-host models of drug resistance allow for more detailed dynamics of host immunity, latent reservoirs of virus, and drug PK/PD. Bridging these two levels of modeling detail requires an understanding of the "levers"--model parameters or combinations thereof--that change only one independent observable at a time. Using the example of accidental tenofovir-based pre-exposure prophyaxis (PrEP) use during HIV infection, we will explore methods of implementing host heterogeneities and their long-term effects on drug resistance.

Results: We combined and extended existing models of virus dynamics by incorporating pharmacokinetics, pharmacodynamics, and adherence behavior. We identified two "levers" associated with the host immune pressure against the virus, which can be used to independently modify the setpoint viral load and the shape of the acute phase viral load peak. We propose parameter relationships that can explain differences in acute and setpoint viral load among hosts, and demonstrate their influence on the rates of emergence and reversion of drug resistance. The importance of these dynamics is illustrated by modeling long-lived latent reservoirs of virus, through which past intervals of drug resistance can lead to failure of suppressive drug regimens. Finally, we analyze assumptions about temporal patterns of drug adherence and their impact on resistance dynamics, finding that with the same overall level of adherence, the dwell times in drug-adherent versus not-adherent states can alter the levels of drug-resistant virus incorporated into latent reservoirs.

Conclusions: We have shown how a diverse range of observable viral load trajectories can be produced from a basic model of virus dynamics using immunity-related "levers". Immune pressure, in turn, influences the dynamics of drug resistance, with increased immune activity delaying drug resistance and driving more rapid return to dominance of drug-susceptible virus after drug cessation. Both immune pressure and patterns of drug adherence influence the long-term risk of drug resistance. In the case of accidental PrEP use during infection, rapid transitions between adherence states and/or weak immunity fortifies the "memory" of previous PrEP exposure, increasing the risk of future drug resistance. This model framework provides a means for analyzing individual-level risks of drug resistance and implementing heterogeneities among hosts, thereby achieving a crucial prerequisite for improving population-level models of drug resistance.

Show MeSH
Related in: MedlinePlus