Evolution of human immunodeficiency virus type 1 coreceptor usage during antiretroviral therapy

A Bayesian approach

Christina M R Kitchen, Sean Philpott, Harold Burger, Barbara Weiser, Kathryn Anastos, Marc A. Suchard

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

There is substantial evidence for ongoing replication and evolution of human immunodeficiency virus type 1 (HIV-1), even in individuals receiving highly active antiretroviral therapy. Viral evolution in the presence of antiviral therapy needs to be considered when developing new therapeutic strategies. Phylogenetic analyses of HIV-1 sequences can be used for this purpose but may give rise to misleading results if rates of intrapatient evolution differ significantly. To improve analyses of HIV-1 evolution relevant to studies of pathogenesis and treatment, we developed a Bayesian hierarchical model that incorporates all available sequence data while simultaneously allowing the phylogenetic parameters of each patient to vary. We used this method to examine evolutionary changes in HIV-1 coreceptor usage in response to treatment. We examined patients whose viral populations exhibited a shift in coreceptor utilization in response to therapy. CXCR4 (X4) strains emerged in each patient but were suppressed following initiation of new antiretroviral regimens, so that CCR5-utilizing (R5) strains predominated. By phylogenetically reconstructing the evolutionary relationship of HIV-1 obtained longitudinally from each patient, it was possible to examine the origin of the reemergent R5 virus. Using our Bayesian hierarchical approach, we found that the reemergent R5 virus detectable after therapy was more closely related to the predecessor R5 virus than to the X4 strains. The Bayesian hierarchical approach, unlike more traditional methods, makes it possible to evaluate competing hypotheses across patients. This model is not limited to analyses of HIV-1 but can be used to elucidate evolutionary processes for other organisms as well.

Original languageEnglish (US)
Pages (from-to)11296-11302
Number of pages7
JournalJournal of Virology
Volume78
Issue number20
DOIs
StatePublished - Oct 2004

Fingerprint

Bayes Theorem
Human immunodeficiency virus 1
HIV-1
therapeutics
Viruses
viruses
Therapeutics
phylogeny
Highly Active Antiretroviral Therapy
Antiviral Agents
pathogenesis
organisms
methodology
Population

ASJC Scopus subject areas

  • Immunology

Cite this

Evolution of human immunodeficiency virus type 1 coreceptor usage during antiretroviral therapy : A Bayesian approach. / Kitchen, Christina M R; Philpott, Sean; Burger, Harold; Weiser, Barbara; Anastos, Kathryn; Suchard, Marc A.

In: Journal of Virology, Vol. 78, No. 20, 10.2004, p. 11296-11302.

Research output: Contribution to journalArticle

Kitchen, Christina M R ; Philpott, Sean ; Burger, Harold ; Weiser, Barbara ; Anastos, Kathryn ; Suchard, Marc A. / Evolution of human immunodeficiency virus type 1 coreceptor usage during antiretroviral therapy : A Bayesian approach. In: Journal of Virology. 2004 ; Vol. 78, No. 20. pp. 11296-11302.
@article{ac4bc15fa90848c0804dddc2f854034a,
title = "Evolution of human immunodeficiency virus type 1 coreceptor usage during antiretroviral therapy: A Bayesian approach",
abstract = "There is substantial evidence for ongoing replication and evolution of human immunodeficiency virus type 1 (HIV-1), even in individuals receiving highly active antiretroviral therapy. Viral evolution in the presence of antiviral therapy needs to be considered when developing new therapeutic strategies. Phylogenetic analyses of HIV-1 sequences can be used for this purpose but may give rise to misleading results if rates of intrapatient evolution differ significantly. To improve analyses of HIV-1 evolution relevant to studies of pathogenesis and treatment, we developed a Bayesian hierarchical model that incorporates all available sequence data while simultaneously allowing the phylogenetic parameters of each patient to vary. We used this method to examine evolutionary changes in HIV-1 coreceptor usage in response to treatment. We examined patients whose viral populations exhibited a shift in coreceptor utilization in response to therapy. CXCR4 (X4) strains emerged in each patient but were suppressed following initiation of new antiretroviral regimens, so that CCR5-utilizing (R5) strains predominated. By phylogenetically reconstructing the evolutionary relationship of HIV-1 obtained longitudinally from each patient, it was possible to examine the origin of the reemergent R5 virus. Using our Bayesian hierarchical approach, we found that the reemergent R5 virus detectable after therapy was more closely related to the predecessor R5 virus than to the X4 strains. The Bayesian hierarchical approach, unlike more traditional methods, makes it possible to evaluate competing hypotheses across patients. This model is not limited to analyses of HIV-1 but can be used to elucidate evolutionary processes for other organisms as well.",
author = "Kitchen, {Christina M R} and Sean Philpott and Harold Burger and Barbara Weiser and Kathryn Anastos and Suchard, {Marc A.}",
year = "2004",
month = "10",
doi = "10.1128/JVI.78.20.11296-11302.2004",
language = "English (US)",
volume = "78",
pages = "11296--11302",
journal = "Journal of Virology",
issn = "0022-538X",
publisher = "American Society for Microbiology",
number = "20",

}

TY - JOUR

T1 - Evolution of human immunodeficiency virus type 1 coreceptor usage during antiretroviral therapy

T2 - A Bayesian approach

AU - Kitchen, Christina M R

AU - Philpott, Sean

AU - Burger, Harold

AU - Weiser, Barbara

AU - Anastos, Kathryn

AU - Suchard, Marc A.

PY - 2004/10

Y1 - 2004/10

N2 - There is substantial evidence for ongoing replication and evolution of human immunodeficiency virus type 1 (HIV-1), even in individuals receiving highly active antiretroviral therapy. Viral evolution in the presence of antiviral therapy needs to be considered when developing new therapeutic strategies. Phylogenetic analyses of HIV-1 sequences can be used for this purpose but may give rise to misleading results if rates of intrapatient evolution differ significantly. To improve analyses of HIV-1 evolution relevant to studies of pathogenesis and treatment, we developed a Bayesian hierarchical model that incorporates all available sequence data while simultaneously allowing the phylogenetic parameters of each patient to vary. We used this method to examine evolutionary changes in HIV-1 coreceptor usage in response to treatment. We examined patients whose viral populations exhibited a shift in coreceptor utilization in response to therapy. CXCR4 (X4) strains emerged in each patient but were suppressed following initiation of new antiretroviral regimens, so that CCR5-utilizing (R5) strains predominated. By phylogenetically reconstructing the evolutionary relationship of HIV-1 obtained longitudinally from each patient, it was possible to examine the origin of the reemergent R5 virus. Using our Bayesian hierarchical approach, we found that the reemergent R5 virus detectable after therapy was more closely related to the predecessor R5 virus than to the X4 strains. The Bayesian hierarchical approach, unlike more traditional methods, makes it possible to evaluate competing hypotheses across patients. This model is not limited to analyses of HIV-1 but can be used to elucidate evolutionary processes for other organisms as well.

AB - There is substantial evidence for ongoing replication and evolution of human immunodeficiency virus type 1 (HIV-1), even in individuals receiving highly active antiretroviral therapy. Viral evolution in the presence of antiviral therapy needs to be considered when developing new therapeutic strategies. Phylogenetic analyses of HIV-1 sequences can be used for this purpose but may give rise to misleading results if rates of intrapatient evolution differ significantly. To improve analyses of HIV-1 evolution relevant to studies of pathogenesis and treatment, we developed a Bayesian hierarchical model that incorporates all available sequence data while simultaneously allowing the phylogenetic parameters of each patient to vary. We used this method to examine evolutionary changes in HIV-1 coreceptor usage in response to treatment. We examined patients whose viral populations exhibited a shift in coreceptor utilization in response to therapy. CXCR4 (X4) strains emerged in each patient but were suppressed following initiation of new antiretroviral regimens, so that CCR5-utilizing (R5) strains predominated. By phylogenetically reconstructing the evolutionary relationship of HIV-1 obtained longitudinally from each patient, it was possible to examine the origin of the reemergent R5 virus. Using our Bayesian hierarchical approach, we found that the reemergent R5 virus detectable after therapy was more closely related to the predecessor R5 virus than to the X4 strains. The Bayesian hierarchical approach, unlike more traditional methods, makes it possible to evaluate competing hypotheses across patients. This model is not limited to analyses of HIV-1 but can be used to elucidate evolutionary processes for other organisms as well.

UR - http://www.scopus.com/inward/record.url?scp=4644243007&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4644243007&partnerID=8YFLogxK

U2 - 10.1128/JVI.78.20.11296-11302.2004

DO - 10.1128/JVI.78.20.11296-11302.2004

M3 - Article

VL - 78

SP - 11296

EP - 11302

JO - Journal of Virology

JF - Journal of Virology

SN - 0022-538X

IS - 20

ER -