Potential for Spread of Drug Resistance Due to PrEP

A major concern in the use of ARV-drug in HIV-uninfected individuals to prevent HIV infection (Pre-Exposure Prophylaxis, PrEP) is the risk of generation of drug-resistant virus that compromises the efficacy of current first line treatment regimens.


Mathematical models have been developed to estimate the potential for the generation and spread of drug-resistant virus due to PrEP but since these models require many detailed assumptions about the virology, immunology and epidemiology, about which there exists substantial uncertainties, there has so far been a lack of consensus in the overall projections formed. There is also a need for this research to feed insights into other models examining the overall cost-effectiveness of progams. Furthermore, new data about patterns of pill-taking behaviour, the interaction between adherence, degree of protection and the performance of diagnostics (for patient screening) and propensity for pill sharing behaviour now pose important new questions for existing models.


The HIV Modelling Consortium convened a  two-day meeting in Seattle to gather key expert biologists and mathematical modellers in this area to compare and discuss existing data, model structures and key results. Following this meeting, the Secretariat coordinated a survey of virologists to produce a report about the current understanding of the dynamics of resistance and the implications for model assumptions as a resource for modellers, and undertook a modelling exercise to compare models' predictions about how a widespread PrEP provision would affect the overall burden of resistance, finding general agreement amongst models that while averting new HIV infections, PrEP would likely have a minimal impact on the volume of resistance compared to that expected as a result of therapeutic ARV provision. 

Aims of Meeting

  • Mapping of questions that policy-makers have regarding resistance, identifying which have received attention and a strategy for tackling others
  • Establish if there is overall agreement in the potential “threat” posed by the generation and spread of drug-resistant HIV due to PrEP, across biologists, clinicians and modellers. If not, identify reasons for this.
  • Agree on major sources of uncertainty in model results and consider ways to gather suitable data.
  • To explore implications of new data for model development and new analyses (from iPrex trial and pre-marketing research, and from usage of rapid tests)
  • To agree on best use of these emerging insights and results for cost-effectiveness modelling done by other groups.
  • Discussion of plans for modeling of next generation products (e.g., TMC278 and DPV ring)


This collaborative work was submitted to the US Food and Drug Administration (FDA) for consideration as a part of the regulatory process resulting in the approval of Truvada for PrEP.


  • David van de Vijver, Brooke Nichols, Ume Abbas, Charles Boucher, Valentina Cambiano, Jeffrey W Eaton, Robert Glaubius, Katrina Lythgoe, John Mellors, Andrew Phillips, Kim Sigaloff, Timothy B Hallett. Preexposure prophylaxis will have a limited impact on HIV-1 drug resistance in sub-Saharan Africa: a comparison of mathematical models. AIDS. 2013 Aug 12.
  • Dimitrov D, et al. Analytic review of modeling studies of ARV based PrEP interventions reveals strong influence of drug. PLOSMedicine, 2013

Funding issued

MC 2.1: PI Dobromir Dimitrov - Fred Hutchinson Cancer Research Center

MC 2.1: PI David van de Vijver - Erasmus Medical Centre