Reader Comments
Post a new comment on this article
Post Your Discussion Comment
Please follow our guidelines for comments and review our competing interests policy. Comments that do not conform to our guidelines will be promptly removed and the user account disabled. The following must be avoided:
- Remarks that could be interpreted as allegations of misconduct
- Unsupported assertions or statements
- Inflammatory or insulting language
Thank You!
Thank you for taking the time to flag this posting; we review flagged postings on a regular basis.
closeIncorrect predictions of the potential impact of antiretroviral chemoprophylaxis (PrEP) on HIV-1 epidemics
Posted by sblower on 11 Oct 2007 at 01:35 GMT
The authors have modeled an extremely important topic. They have modeled the potential impact of PrEP on HIV-1 epidemics in resource-limited settings.
Unfortunately, the authors have made incorrect assumptions when constructing their model; these incorrect assumptions are apparent from examination of the equations that specify their model.
There is a major concern that drug-resistant strains of HIV could be selected by drug pressure from individuals who are infected with drug-sensitive strains and taking PrEP. However when constructing their model the authors assumed incorrectly that:
(1) drug-resistant strains of HIV could evolve in individuals who are NOT infected with HIV and taking PrEP.
(2) drug-resistant strains of HIV could NOT evolve in individuals who are infected with drug-sensitive strains of HIV and taking PrEP.
Hence, unfortunately, the authors’ predictions of the potential impact of PrEP on HIV epidemics are incorrect.
Professor Sally Blower
Head of Disease Modeling Group
David Geffen School of Medicine at UCLA
THE AUTHORS REPLY
uabbas replied to sblower on 18 Oct 2007 at 23:28 GMT
The comments posted by Professor Blower are incorrect. First, our model does NOT assume that drug-resistant strains of HIV could evolve in individuals who are not infected with HIV and taking PrEP. Second, our model DOES assume that drug-resistant strains of HIV could evolve in individuals who are infected with drug-sensitive strains of HIV and taking PrEP. Our model equations and assumptions are detailed in the Appendix S1 and are elaborated further below.
Our model represents both transmitted and acquired drug resistance in individuals with HIV infection, as well as different stages of HIV infection including recent infection. The nature and timing of selection of drug resistance in individuals who become infected while on PrEP with source partner having wild-type virus are unknown. Nevertheless, there is concern that resistant virus could be selected if PrEP is continued during recent HIV infection in individuals who acquire wild-type, drug sensitive virus from their source partner. In addition, there is some concern that PrEP may select for the transmission of drug-resistant virus even though the dominant species from the source partner is drug susceptible. To represent these two issues, our model assumes that a fraction (50% - 100%) of individuals who become infected while on PrEP from a source partner having wild-type virus will develop resistant virus. Our model also represents the duration of persistence of drug resistant virus in these individuals both on and off PrEP.
Ume L. Abbas, MD
John W. Mellors, MD
University of Pittsburgh, Pittsburgh, PA, USA
Roy M. Anderson, Ph.D.
Imperial College London, London, UK
RE: THE AUTHORS REPLY
sblower replied to uabbas on 26 Oct 2007 at 17:39 GMT
I thank the authors for replying to my previous posting so that we can debate these important issues.
I again suggest that if the authors examine carefully their mathematical model (i.e., their specific equations and flow diagram) they will see that my comments posted on 11 October 2007 remain correct.
As I have said previously, from examination of the equations proposed by Abbas et al. it can be seen that in their model:
(1) drug-resistant strains of HIV can evolve in individuals who are NOT infected with HIV and taking PrEP (see the flow-diagram of the Abbas et al. model included in their paper).
(2) drug-resistant strains of HIV can NOT evolve in individuals who are infected with drug-sensitive strains of HIV and taking PrEP (see the flow diagram of the Abbas et al. model included in their paper).
If the authors continue to disagree with my statements then I would like to suggest - so that the interested reader can evaluate who is correct - that the authors specify which arrows in their flow diagram represent acquired and transmitted resistance to PrEP. This should clarify matters. I hope that the authors will respond to this suggestion.
Professor Sally Blower
Head of Disease Modeling Group
David Geffen School of Medicine at UCLA
THE AUTHORS’ FINAL REPLY
uabbas replied to sblower on 26 Nov 2007 at 18:14 GMT
The statements in the second comment posted by Professor Blower are again incorrect First, our model does not allow the evolution of drug-resistant strains of HIV in uninfected individuals. Second, our model does allow that drug-resistant strains of HIV could evolve in individuals who are infected with drug-sensitive strains of HIV and taking PrEP.
We have detailed our model equations and assumptions in our manuscript and Appendix S1 with further clarification in our first reply. We leave interested readers to draw their own conclusions from the published work.
We do not believe this dialogue is productive for the readership, hence we will not respond to subsequent postings by Professor Blower.
Ume L. Abbas, MD
John W. Mellors, MD
University of Pittsburgh, Pittsburgh, PA, USA
Roy M. Anderson, Ph.D.
Imperial College London, London, UK
RE: THE AUTHORS’ FINAL REPLY
sblower replied to uabbas on 27 Nov 2007 at 01:38 GMT
Dear Authors,
You state the correct biological assumptions about the emergence of acquired resistance to PrEP in the text of your paper and when you describe the assumptions that you intended to use when you constructed your model of PrEP implementation BUT the problem is that unfortunately you did not write down the correct equations that incorporated these correct assumptions. Hence - inadvertently - when you wrote down your equations you constructed a PrEP implemention model that actually reflects the incorrect assumptions that I have previously outlined.
If you examine your flow-diagram (which reflects and matches the equations that specify your model) you will quickly see that you have modeled the emergence of acquired resistance to PrEP incorrectly.
The majority of interested readers of PLoS ONE can not understand equations. However a flow-diagram is rather easy for anyone to follow - they do not need to understand any mathematics or examine any equations. As I have said previously it is quite simple for you to explain how you have modeled acquired resistance to PrEP simply by specifying which arrows in your flow diagram represent the development of acquired resistance to PrEP. This is an extremely reasonable suggestion as it: (a) would take less than a minute for the authors to do, and (b) would allow interested readers to evaluate the Abbas et al. model. I do hope that the authors will respond to my suggestion as it would enable interested readers to evaluate who is correct.