UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An investigation of human and viral genetic factors influencing HIV disease progression and response… Brumme, Zabrina Lesley 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2006-129778.pdf [ 10.16MB ]
Metadata
JSON: 831-1.0092862.json
JSON-LD: 831-1.0092862-ld.json
RDF/XML (Pretty): 831-1.0092862-rdf.xml
RDF/JSON: 831-1.0092862-rdf.json
Turtle: 831-1.0092862-turtle.txt
N-Triples: 831-1.0092862-rdf-ntriples.txt
Original Record: 831-1.0092862-source.json
Full Text
831-1.0092862-fulltext.txt
Citation
831-1.0092862.ris

Full Text

AN INVESTIGATION OF HUMAN AND VIRAL GENETIC FACTORS INFLUENCING HIV DISEASE PROGRESSION AND RESPONSE TO TREATMENT IN THE ERA OF HIGHLY ACTIVE ANTIRETROVIRAL THERAPY (HAART) by Z A B R I N A L E S L E Y B R U M M E B . S c , The University of British Columbia, 2000 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in The Faculty of Graduate Studies (Experimental Medicine) T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A March 2006 © Zabrina Lesley Brumme, 2006 A B S T R A C T Objective: Host and viral genetic factors contribute significantly to the natural course of untreated H I V infection; however, the impact of these genetic factors on response to modern antiretroviral therapies remains poorly understood. The primary objective of this thesis is to establish the impact of specific host and viral genetic factors on H I V disease progression and treatment response in the era of Highly Active Antiretroviral Therapy ( H A A R T ) . Methods: Host genetic factors selected for investigation include polymorphisms in chemokine receptor genes known to play a role in H I V disease progression during untreated infection ( C C R 5 and C X 3 C R I ) , or genes involved in drug metabolism (MDR-1) . Vira l genetic factors selected for investigation include polymorphisms in H I V genes known to influence untreated disease progression (HIV envelope), or mutations in the viral genome known or suspected to confer resistance to antiretroviral agents. The impact of host and viral genetic factors on short and longer-term clinical outcomes were assessed using retrospective, population-based cohort and case-control analyses of HIV-infected individuals enrolled in the H I V / A I D S Drug Treatment Program in British Columbia, Canada. Results: Naturally-occurring polymorphisms in human chemokine receptor genes C C R 5 and C X 3 C R I , as well as genes involved in drug metabolism (MDR-1) appear to be associated with differential short and longer-term responses to antiretroviral therapy; however, the relative contribution of individual human genetic polymorphisms to therapy response is likely to be small. In contrast, genetic variation in the H I V envelope appears to significantly affect both short and longer-term clinical responses to H A A R T , supporting the use of H I V envelope genotypes as determinants of therapy response. A n investigation of the impact of antiretroviral resistance mutations on long-term clinical outcomes, in particular, mortality, i i suggested that drug resistance, although associated with poor short-term therapy efficacy, is not yet a major factor contributing to mortality among HIV-infected persons in British Columbia, at least on a population basis. Conclusion: Results support the potential incorporation of H I V envelope genotyping as a monitoring tool for the clinical management of HIV-infected individuals. The impact of human genetic variation on therapy response, however, is l ikely to be mediated by the contributions of polymorphisms in many different genes; a more comprehensive approach may therefore be necessary before incorporating host genotypes into routine H I V clinical management. T A B L E O F C O N T E N T S A B S T R A C T i i T A B L E O F C O N T E N T S iv LIST O F T A B L E S ix LIST O F F I G U R E S x A C K N O W L E D G E M E N T S xi i C O - A U T H O R S H I P S T A T E M E N T xi i i C H A P T E R 1 G E N E R A L I N T R O D U C T I O N A N D THESIS O B J E C T I V E S 1 1.1 H U M A N IMMUNODEFICIENCY VIRUS INFECTION: WORLDWIDE IMPACT 1 1.2 H U M A N IMMUNODEFICIENCY VIRUS INFECTION: PATHOGENESIS 1 1.3 BRIEF OVERVIEW OF HIV LIFE C Y C L E A N D GENOMIC ORGANIZATION 2 1.4 BRIEF HISTORY OF ANTIRETROVIRAL T R E A T M E N T 5 1.5 SURROGATE MARKERS FOR MONITORING OF HIV DISEASE PROGRESSION 6 1.6 IDENTIFICATION OF HOST A N D VIRAL GENETIC FACTORS WHICH INFLUENCE HIV DISEASE PROGRESSION IN UNTREATED INFECTION 8 1.7 R E L E V A N C E OF HOST A N D VIRAL GENETIC FACTORS IN T H E ERA OF HAART 10 1.8 THESIS OBJECTIVES A N D ORGANIZATION 13 1.9 BRIEF OVERVIEW OF D A T A SOURCES, A N D EVOLUTION OF OUR POPULATION-BASED COHORTS 14 1.10 REFERENCES 17 iv CHAPTER 2 L I T E R A T U R E R E V I E W : I M P A C T O F H U M A N A N D V I R A L G E N E T I C V A R I A T I O N O N H I V D I S E A S E P R O G R E S S I O N A N D T R E A T M E N T R E S P O N S E IN T H E E R A O F H I G H L Y A C T I V E A N T I R E T R O V I R A L T H E R A P Y ( H A A R T ) 30 2.1 INTRODUCTION 30 2.2 H U M A N GENETIC FACTORS: CHEMOKINE RECEPTOR POLYMORPHISMS 31 2.3 MUTATIONS IN GENES THAT MODULATE THE IMMUNE RESPONSE 32 2.4 R E L E V A N C E OF CHEMOKINE AND IMMUNOGENETIC FACTORS IN THE ERA OF HAART... .33 2.5 PHARMACOGENETICS IN THE ERA OF HAART 34 2.6 VIRAL GENETIC FACTORS: RESISTANCE TO CURRENT AND FUTURE ANTIRETROVIRAL AGENTS 36 2.7 HIV ENVELOPE A N D CO-RECEPTOR PHENOTYPE 39 2.8 R E L E V A N C E OF HIV ENVELOPE VARIATION IN THE ERA OF H A A R T 41 2.9 R E L E V A N C E OF VARIATION IN OTHER REGIONS OF THE HIV GENOME IN THE ERA OF HAART 43 2.10 REFERENCES 44 CHAPTER 3 H U M A N G E N E T I C V A R I A T I O N : I M P L I C A T I O N S O F T H E C C R 5 A 3 2 M U T A T I O N O N R E S P O N S E T O H A A R T 61 3.1 INTRODUCTION 61 3.2 MATERIALS AND METHODS 62 3.3 RESULTS 65 3.4 DISCUSSION/CONCLUSIONS 68 3.5 REFERENCES 72 CHAPTER 4 H U M A N G E N E T I C V A R I A T I O N : I M P L I C A T I O N S O F S I N G L E N U C L E O T I D E P O L Y M O R P H I S M S I N C X 3 C R 1 A N D M D R - 1 G E N E S O N I N I T I A L R E S P O N S E T O A N T I R E T R O V I R A L T H E R A P Y 75 4.1 INTRODUCTION 75 4.2 MATERIALS A N D METHODS 78 4.3 RESULTS 82 4.4 DISCUSSION/CONCLUSIONS 87 4.5 REFERENCES 96 CHAPTER5 H I V G E N E T I C V A R I A T I O N : I M P L I C A T I O N S O F A N T I R E T R O V I R A L R E S I S T A N C E O N L O N G - T E R M H A A R T O U T C O M E S 101 5.1 INTRODUCTION 101 5.2 MATERIALS A N D METHODS 102 5.3 RESULTS 106 5.4 DISCUSSION/CONCLUSIONS I l l 5.5 REFERENCES 123 vi CHAPTER 6 H I V G E N E T I C V A R I A T I O N : I M P L I C A T I O N S O F H I V p 6 G a g I N S E R T I O N S O N R E S P O N S E T O H A A R T 126 6.1 INTRODUCTION 126 6.2 MATERIALS AND METHODS 128 6.3 RESULTS 131 6.4 DISCUSSION/CONCLUSIONS 134 6.5 REFERENCES 140 CHAPTER 7 H I V G E N E T I C V A R I A T I O N : R E L E V A N C E O F H I V V 3 E N V E L O P E G E N O T Y P E O N R E S P O N S E T O H A A R T 144 7.1 INTRODUCTION 144 7.2 MATERIALS AND METHODS 146 7.3 RESULTS 152 7.4 DISCUSSION/CONCLUSIONS 156 7.5 REFERENCES 166 CHAPTER 8 H I V G E N E T I C V A R I A T I O N : R E L E V A N C E O F H I V E N V E L O P E C O R E C E P T O R U S A G E O N R E S P O N S E T O H A A R T 171 8.1 INTRODUCTION 171 8.2 MATERIALS AND METHODS 174 vii 8.3 R E S U L T S 178 8.4 D I S C U S S I O N / C O N C L U S I O N 182 8.5 R E F E R E N C E S 194 C H A P T E R 9 G E N E R A L D I S C U S S I O N A N D C O N C L U S I O N S 200 9.1 I N T R O D U C T I O N 200 9.2 H U M A N G E N E T I C V A R I A T I O N : T H E C H E M O K I N E R E C E P T O R S 202 a) C C R 5 A32 202 b) C X 3 C R 1 203 9.3 P H A R M A C O K I N E T I C S : D R U G M E T A B O L I S M A N D T H E P - G P M E M B R A N E E F F L U X T R A N S P O R T E R 205 9.4 R E L E V A N C E O F H I V G E N E T I C V A R I A T I O N I N T H E E R A O F H A A R T 208 9.5 A N T I R E T R O V I R A L R E S I S T A N C E T E S T I N G 209 9.6 H I V P 6 g a g 210 9.7 H I V E N V E L O P E 212 a) V 3 loop genotyping 213 b) H I V coreceptor phenotyping 215 9.8 U N R E S O L V E D R E S E A R C H Q U E S T I O N S S U R R O U N D I N G H I V E N V E L O P E G E N O T Y P I N G / P H E N O Y P I N G A N D E V I D E N C E F O R F U T U R E C L I N I C A L U T I L I T Y O F C O R E C E P T O R A S S A Y S 216 a) Evolution of Coreceptor phenotypes during treatment with H A A R T 216 b) H I V envelope genotype/coreceptor phenotype correlations 217 c) Evidence for adaptation of H I V to human genotypes: the C C R 5 mutation and H I V coreceptor usage 219 d) Use of H I V coreceptor phenotyping in the context of future antiretroviral agents 220 9.09 I M P L I C A T I O N S A N D F U T U R E D IRECTIONS 221 9.10 C O N C L U D I N G R E M A R K S 223 9.11 R E F E R E N C E S 224 A P P E N D I X 1: Certificates of Ethical Approval (original and updated)...... 238 L I S T O F T A B L E S Table 3.1 Influence of C C R 5 A 3 2 genotype on time to non-accidental death in the >5-year period following initiation of H A A R T : Univariate and Multivariate Analysis 70 Table 4.1 Baseline characteristics of the study population, prior to the initiation of antiretroviral therapy (N=461) 91 Table 4.2 Alle le and haplotype frequencies for the C X 3 C R 1 , M D R - 1 and C C R 5 polymorphisms..... 92 Table 4.3 Multivariate Cox Model showing effect of baseline parameters, including M D R - 1 and C X 3 C R I polymorphisms, on time to Virological and Immunological Failure 93 Table 5.1 Number of deaths and antiretroviral treatment history in this cohort, by year 116 Table 5.2 Comparison between those who died without any antiretroviral resistance compared to those who showed any antiretroviral resistance 117 Table 6.1 Distribution of N-terminal H I V p 6 G a g insertions observed in study group of antiretroviral naive individuals 137 Table 6.2 Patient Characteristics at baseline 138 ix Table 7.1 Patient Characteristics at baseline 160 Table 7.2a Influence of baseline factors on time to death after initiation of HAART. . . .161 Table 7.2b Influence of baseline factors on time to C D 4 decline below baseline after initiation of H A A R T 162 Table 8.1a Patient Characteristics at baseline, stratified by H I V co-receptor usage 186 Table 8.1b Distribution of R 5 / X 4 H I V by baseline (pre-therapy) C D 4 count strata 187 Table 8.2 Patient Characteristics at baseline, stratified by H I V co-receptor usage and human C C R 5 A 3 2 genotype 188 Table 8.3 Univariate and Multivariate predictors of the presence of X4-using H I V variants at baseline 189 Table 8.4a Influence of baseline factors on time to non-accidental death after initiation of H A A R T 190 Table 8.4b Influence of baseline factors on time to C D 4 decline below baseline after initiation of H A A R T 190 Table 8.4c Influence of baseline factors on time to p V L failure >500 copies/ml (after initial suppression of p V L <500 copies/ml) 191 L I S T O F F I G U R E S Figure 3.1 Kaplan-Meier analysis of survival following initiation of H A A R T , by C C R 5 genotype 71 Figure 4.1 Kaplan-Meier plot comparing the influence of nucleotide variation at position 3435 of the human P-glycoprotein (MDR-1) gene on time to virological failure 94 Figure 4.2 Kaplan-Meier plot comparing the influence of amino acid variation at codons 249 and 280 of the human C X 3 C R I receptor gene on time to immunological failure 95 Figure 5.1 Flowchart summarizing number of subjects in mortality group 118 Figure 5.2a Prevalence of antiretroviral resistance in the last on-therapy plasma sample, by year of death 120 x Figure 5.2b Proportion of resistant isolates, stratified by type of therapy received 121 Figure 5.3 Prevalence of antiretroviral resistance mutations, by year, in a group of living individuals enrolled in the Province's H I V / A I D S Drug Treatment program failing their current therapy (N=1220) 122 Figure 6.1 Kaplan-Meier survival analyses, showing the influence of insertions within the N-terminal region of the H I V p 6 G a g on time to virological (la) and immunological ( lb) failure 139 Figure 7.1 Association between "11/25" genotype and time to death after initiating H A A R T 163 Figure 7.2 Longitudinal analysis of population-based H I V viral loads (panel A ) and C D 4 counts (panel B) , stratified by H I V V3 loop genotype 164 Figure 8.1 Influence of H I V co-receptor usage on survival and clinical outcomes after initiation of first triple antiretroviral therapy 192 x i ACKNOWLEDGEMENTS I would like to extend my most warm and sincere thanks to my thesis supervisor, Dr. Richard Harrigan, as well as my thesis committee members, Dr. Julio Montaner, Dr. Robert Hogg and Dr. Michael O'Shaughnessy, and finally to my former supervisor Dr. Helene Cote, for their continuing direction, support, encouragement and mentorship. I would also like to thank my fellow students and coworkers who have provided invaluable expertise, advice, support, and encouragement throughout the years: Jerome Asselin, Paula Braitstein, Chanson Brumme, Keith Chan, Celia Chui , Winnie Dong, Theresa M o , Beheroze Sattha, L i l l i an Ting, and Brian Wynhoven. I gratefully acknowledge the financial support of the British Columbia Centre for Excellence in H I V / A I D S , as well as the Natural Sciences and Engineering Research Council (NSERC) , Canadian Institutes for Health Research (CIHR), the Michael Smith Foundation for Health Research ( M S F H R ) and the Izaac Walton Ki l l am Foundation for their support through graduate scholarships. x i i C O - A U T H O R S H I P S T A T E M E N T Each of chapters 3-8 consists of a manuscript that is published in an international, peer-reviewed journal. The candidate is either the lead author (chapters 3, 4, 6, 7, 8) or second author (chapter 5) on all of these manuscripts. This statement is to certify that the candidate was the major contributor to study identification and design, design of laboratory methods, collection and/or supervision of laboratory data collection, and data analysis. The candidate also wrote the manuscripts and, in the cases of chapters 7 and 8, also co-wrote the successful grant applications. Primary co-authors of these manuscripts include the thesis supervisor (Dr. Richard Harrigan, thesis committee members (Dr. Julio Montaner, Dr. Robert Hogg, Dr. Michael O'Shaughnessy), statistical analysts (Jerome Asselin, Keith Chan, Chanson Brumme, Benita Yip) , authors who provided data access and assistance with data analysis (Brian Wynhoven, Chanson Brumme), Laboratory Technical Staff (Winnie Dong, Theresa M o , Bethany Henrick) and External Collaborators. C H A P T E R 1 G E N E R A L I N T R O D U C T I O N A N D THESIS O B J E C T I V E S 1.1 Human Immunodeficiency Virus infection: worldwide impact Human Immunodeficiency Virus Type 1 (HIV-1), the virus that causes Acquired Immunodeficiency Syndrome (AIDS), has emerged as a major worldwide pandemic since its identification as a novel pathogenic retrovirus in the early 1980s [1,2,3]. A t the end of 2004, 39.4 mil l ion people worldwide (37.2 mill ion adults and 2.2 mil l ion children under 15 years of age) were estimated to be living with H I V / A I D S , with Sub-Saharan Africa bearing the major burden of these infections [4]. During the year 2004, it was estimated that a total of 3.1 mill ion individuals died of H I V / A I D S (2.6 mill ion adults and 510, 000 children) [4]. To date the total cumulative HIV/AIDS-associated death toll worldwide is estimated to exceed 25 mill ion [4,5]. Currently, an estimated 1.6 mil l ion adults and children are living with H I V / A I D S in North America [4]. The most recent estimates indicate that approximately 56,000 Canadians were l iving with H I V / A I D S at the end of 2002, approximately 30% of whom were unaware of their diagnosis [6]. 1.2 Human Immunodeficiency Virus infection: pathogenesis HIV-1 is a pathogenic retrovirus that specifically targets and destroys the CD4+ subset of T-lymphocytes of the human immune system [7,8]. If left untreated, a progressive deterioration of immune function occurs, leaving the infected individual susceptible to a wide range of opportunistic bacterial, viral, parasitic and/or fungal infections eventually leading to A I D S and death. Currently, the case definition for A I D S , as classified by the U.S . Centres for Disease Control and Wor ld Health Organisation, is defined as the presence of an A I D S -defining illness and by a C D 4 T-lymphocyte count of less than 200 cells/mm [9]. 1 1.3 Brief Overview of HIV life cycle and genomic organization. The HIV-1 genome is comprised of two single strands of positive (+)-sense R N A , each approximately 9700 base-pairs long, which encode the major structural and regulatory proteins needed for viral replication [10]. HIV-1 shares a common genomic structure and organization with all known replication-competent retroviruses: from the 5' to 3' ends of the genome are encoded the gag (group-specific antigen), pol (polymerase) and env (envelope) regions, encoding proteins needed for retroviral replication [for reviews, see 11,12]. The Gag region encodes a polyprotein precursor (Pr55 G a g ) which is cleaved by the H I V protease to release the major structural components of the virion, namely the viral matrix ( M A , p i7) , capsid ( C A , p24), nucleocapsid (NC, p7) and another small protein p 6 G a g . The H I V pol region encodes the major viral enzymes required for the replication, integration and processing of the H I V genome and gene products, including the H I V protease, reverse transcriptase, R N A s e H , and integrase enzymes. The pol genes are initially expressed as a large gag-pol polyprotein precursor ( P r l 6 0 G a g P o 1 ) , generated by a ribosomal frameshift event during translation of the viral m R N A message; individual proteins are later cleaved and released by the H I V protease. The env region encodes the envelope glycoproteins gpl20 and gp41, which form the surface (SU) and transmembrane (TM) components of the H I V envelope, respectively. Env genes are also initially produced as a polyprotein precursor (gpl60), which is cleaved into gp41 and gpl20 by a cellular protease [13]. In addition, the H I V genome encodes two regulatory proteins, tat and rev, as well as a number of accessory proteins, vif, vpr, vpu and nef, all of which are necessary for viral replication in vivo. H I V begins its replication cycle by binding to a target cell through an interaction between H I V gpl20 (envelope) and the C D 4 receptor on the surface of the target cell 2 [7,8,14]. C D 4 alone, however, is not sufficient for successful viral infection of the host cell. In the mid-1990s, the human chemokine receptors C X C R 4 [15] and C C R 5 [16,17,18] were identified as necessary co-receptors, required in addition to C D 4 , for efficient entry of H I V into target cells. The C X C R 4 receptor is expressed primarily on activated CD4+ T-lymphocytes, while the C C R 5 receptor is expressed primarily on naive and memory CD4+ T-lymphocytes as well as macrophages; thus, the major cellular targets for H I V infection are CD4+T-lymphocytes and macrophages. In rare cases, other minor co-receptors have also been shown to mediate entry of HIV-1 into target cells, at least in vitro [19,20,21]. In addition, evidence suggests that H I V may, to a lesser extent, also be able to infect CD8+ T-lymphocytes [22,23,24], although the relevance of this to the pathogenesis of H I V remains unclear. The binding of H I V gpl20 to C D 4 on the surface of the host cell induces a conformational change in the H I V envelope which exposes the co-receptor binding site on gpl20 [for a review on co-receptors and viral entry, see 25]. The subsequent binding of gpl20 to the co-receptor then triggers the virus-cell membrane fusion event, in which the transmembrane portion of the H I V envelope (gp41) changes conformation to catalyze the insertion of the "fusion peptide" (located at the N-terminus of gp41) into the host cell membrane. Following virus-cell fusion, the virus core particle (representing the remaining viral structure once the l ipid bilayer and envelope proteins are removed) enters the host cell and a poorly-understood "uncoating" process occurs whereby the viral genetic material is released into the host cell cytoplasm [11]. The H I V reverse transcriptase then converts the single-stranded R N A genome into double-stranded D N A (dsDNA) and the H I V integrase, in complex with a number of other H I V and host proteins, transports the H I V d s D N A into the 3 nucleus, where it catalyzes the insertion of the d s D N A into the host cell genome. Once integrated, the H I V D N A is referred to as a "provirus", and serves as a template for the synthesis of viral R N A s . H I V R N A expression from the provirus proceeds from the 5'Long-Terminal Repeat (LTR) , a regulatory region located at the 5' end of the genome. The first viral genes to be transcribed are those encoding the regulatory proteins Tat and Rev. Once synthesized, Tat and Rev regulate expression of the remainder of viral R N A s , which include full-length genomic R N A s (for inclusion in new retroviral particles as well as serving as templates for Gag/Pol proteins) as well as singly and multiply-spliced viral m R N A s that serve as templates for the synthesis of the remainder of the viral structural, enzymatic and accessory proteins [26]. The assembly of new retroviral particles takes place at the host cell plasma membrane, and is largely mediated by the structural proteins encoded within H I V gag [11]. As mentioned earlier, the gag (structural) and pol (enzymatic) proteins are expressed as P r 5 5 G a g and P r l 6 0 G a g P o 1 polyproteins, respectively, which are targeted to the host cell plasma membrane by the M A (matrix) domain of gag. The gag proteins are also responsible for binding and encapsulating the H I V genomic R N A and subsequently promoting the assembly of progeny virions at the host cell membrane, although the details of this process remain incompletely understood. The final steps of the viral life cycle are mediated in part by specific retroviral sequences termed "late (L) domains". In H I V , the p 6 G a g protein, located at the 3'end of Gag, is the L domain responsible for mediating the release of new virus particles from the plasma membrane [27]. After the release of immature progeny virions, a final "maturation" step occurs, in which the H I V protease cleaves and releases the individual structural and enzymatic proteins contained in the gag and gag/pol polyproteins. The 4 structural rearrangement of viral components thus results in the formation of mature infectious virus particles. 1.4 Brief History of Antiretroviral Treatment Prior to the availability of antiretroviral therapy, the median time from infection to A I D S diagnosis ranged from approximately 5 to 11 years, depending on age at seroconversion [28], and the median survival time following A I D S diagnosis was approximately 9.5 to 12 months [29,30]. The introduction of zidovudine (AZT) , a Nucleoside Reverse Transcriptase Inhibitor (NRTI), as the first antiretroviral agent in the late 1980s led to improvements in the prognosis of HIV-infected individuals, at least in the short term [31,32,33]. Shortly thereafter, the introduction of additional NRTIs didanosine (ddl) and zalcitabine (ddC) led to the use of dual-combination therapies that resulted in improved clinical responses and survival when compared to A Z T monotherapy [34,35,36,37,38]. It became quickly evident, however, that mono and dual combination therapies had only a relatively short period of effectiveness, in part due to the rapid selection of drug-resistant H I V variants [39,40,41]. It was not until the development of additional classes of antiretroviral agents and the universal recommendation of triple combination therapy (also known as "Highly Active Antiretroviral Therapy", or " H A A R T " ) in the mid-1990s, that dramatic and sustained reductions in HIV-related morbidity and mortality were observed on a population basis [42,43,44,45], resulting in a complete change in outlook for individuals infected with H I V . Currently, there are at least 19 antiretroviral agents approved by the U .S . Food and Drug Administration ( F D A ) and Health Canada for treatment of H I V infection, which may be classified into four broad antiretroviral classes based on the protein target of the 5 antiretroviral agent and the mechanism of inhibition. The first available class of antiretroviral agents, Nucleoside/Nucleotide Reverse Transcriptase Inhibitors (NRTIs), are nucleoside analogues which inhibit H I V reverse transcription by terminating synthesis of the elongating H I V D N A . To date there are at least 8 NRTIs currently approved by the U .S . FDA/Heal th Canada for treatment of H I V infection, including zidovudine, didanosine, zalcitabine, stavudine, lamivudine, abacavir, tenofovir, emtricitabine, and combinations thereof. The second class of antiretroviral agents, the Protease Inhibitors (Pis), target the viral maturation step mediated by the H I V protease enzyme. Pis were licensed for use beginning in the mid-1990s, and include ritonavir, nelfinavir, indinavir, amprenavir, saquinavir, lopinavir, and atazanavir. The third class of antiretroviral agents, the Non-Nucleoside Reverse Transcriptase Inhibitors (NNRTIs) , also target the H I V reverse transcriptase, through allosteric inhibition of the enzyme itself. N N R T I s currently licensed for treatment of H I V infection include nevirapine, delavirdine and efavirenz. The majority of current first-line antiretroviral combination therapies include a backbone of two NRTIs in combination with either a PI or an N N R T I [46]. Finally, the fourth class of antiretroviral agents, the "fusion inhibitors", inhibit the virus-cell fusion event mediated by H I V gp41. Enfuvirtide (T20) is currently the only licensed example of this class [47,48]. Antiretroviral agents currently in development include coreceptor antagonists (aimed at inhibiting the interaction between H I V gpl20 and coreceptors C C R 5 and/or C X C R 4 ) [49], integrase inhibitors (inhibiting the integration of the H I V provirus into the host cell genome) [50], R N A s e H inhibitors (inhibiting the degradation of the H I V R N A genome in the transient R N A - D N A complex that occurs during reverse transcription) [51] as well as others [52]. 1.5 Surrogate markers for monitoring of HIV disease progression 6 In the initial clinical trials evaluating the efficacy of novel antiretroviral therapies, clinical outcomes were measured primarily in terms of the reduction in the burden of opportunistic infections (or the duration of AIDS-free survival), or using death as an endpoint. The evaluation of long-term treatment outcomes, however, was replaced in later years by the use of surrogate markers, allowing more rapid monitoring and evaluation of novel antiretroviral agents in the context of clinical trials as well in the clinical management of HIV-infected individuals [53,54]. The first available surrogate marker for H I V disease progression was the C D 4 count [55], which quantifies the number of circulating CD4+ T-lymphocytes as an indirect measure of immunologic function. It was not until the mid-1990s, however, that a virologic marker of H I V disease progression was made available in the form of the reverse-transcriptase polymerase-chain-reaction (RT-PCR)-based H I V plasma viral load (pVL) assay [56]. A t the present time, additional H I V viral load assays are also available, including the nucleic acid sequence-based amplification ( N A S B A ) [57] as well as a signal amplification technology termed "branched-chain D N A " (bDNA) [58]. The plasma viral load assay, which quantifies the amount of H I V R N A in the peripheral blood, was shown to be a significant independent predictor of progression to A I D S and death in the context of untreated infection [59,60]. Furthermore, reductions in H I V plasma viral load as a direct consequence of the administration of combination antiretroviral therapy were found to be highly predictive of improved clinical prognosis, response to treatment, as well as a decreased risk of progression to A I D S and death [61,62,63,64]. A recent analysis of antiretroviral therapy outcomes from > 12000 subjects in prospective cohort studies concluded that antiretroviral treatment response could be best predicted by the changes in 7 C D 4 cell count and H I V plasma viral load observed after 6 months of treatment, regardless of pre-treatment values [65,66]. Just as C D 4 counts and H I V plasma viral loads were useful and accurate surrogate markers of H I V disease progression in untreated infection, they remain the primary tools today in the clinical management of HIV-infected individuals treated with combination antiretroviral therapies [46,67]. Current treatment guidelines recommend the use of C D 4 and, to a lesser extent, p V L to guide decisions regarding the optimal time to initiate or change antiretroviral therapy [46,67,68]. 1.6 Identification of host and viral genetic factors which influence HIV disease progression in untreated infection The natural history of H I V disease progression varies widely in infected individuals, ranging from persons who remain H I V seronegative despite multiple high-risk exposures [69,70], to persons who progress to A I D S and death within a relatively short period (months to a few years) following infection [71]. It is now appreciated that a wide spectrum of host and viral factors contribute significantly to the natural course of H I V disease progression. a) Host genetic factors Probably the best-studied example of human genetic variation playing a direct role in H I V infection and pathogenesis is a naturally-occurring 32-base pair deletion in the gene encoding the human C C R 5 chemokine receptor, one of the major co-receptors for H I V entry into target cells [72,73]. The 32-base pair deletion (CCR5A32) causes a reading-frame shift that generates a premature stop codon in the C C R 5 m R N A , resulting in a truncated, nonfunctional protein [74]. Individuals homozygous for this mutation (CCR5A32/A32, a rare genotype that occurs at a prevalence of <1% in Caucasian populations), do not express 8 C C R 5 on the surface of their cells and therefore are relatively "resistant" to infection with H I V [72,73,75]. Individuals heterozygous for this mutation (CCR5wt/A32, a genotype that occurs at a prevalence of - 1 5 % in Caucasian populations) express vastly reduced levels of C C R 5 on the surface of their cells [76], and although they may be infected with H I V , exhibit a significantly delayed disease progression when compared to C C R 5 wild-type individuals [72,73]. Variation in genes encoding other human chemokine receptors [77,78], chemokine receptor ligands [79,80], and genes involved in the control and modulation of the human immune response [81] are also now known, or suspected, to play a significant role in the body's response to H I V infection and the subsequent rate of disease progression in the context of untreated infection. b) Viral genetic factors Even before the H I V structure, genomic organization and replication cycle was fully understood, it was known that H I V variants differed in their capacity to induce disease progression [82]. One notable example is an observational clinical study from the late 1980s describing rates of H I V disease progression in haemophiliacs transfused with HIV-infected blood prior to the availability of a test to screen blood donors [83]. It was observed that recipients infected with H I V from donors who had progressed rapidly to A I D S (within 4 years) were significantly more likely to also experience a rapid disease progression, while persons infected with H I V from "slow progressors" also displayed a slower disease progression [83], strongly suggesting that viral genetic factors were important determinants of the rate of H I V disease progression. It is now known that not only does H I V have an extensive capacity for mutation and adaptation, but also that sequence variation in the H I V genome contributes significantly to 9 the natural course of H I V disease. Due to the lack of 3'-5' proofreading ability of the H I V reverse transcriptase, nucleotides are misincorporated, inserted and/or deleted during the conversion of the H I V R N A genome to D N A with an average frequency of one "mistake" per R N A - t o - D N A genomic conversion [84]. This frequency, when combined with a total in vivo production of up to 10 9 virions per day in untreated infection [84], provides H I V with an extraordinary capability for mutational diversity and adaptation. On an individual basis, genetic variation in many regions of the H I V genome, most notably env [85,86], as well as genes encoding the various H I V accessory proteins [87,88,89] is known or suspected to play a significant role in the clinical course of untreated H I V infection. On a population basis, the extreme worldwide diversity of HIV-1 is best illustrated by the existence of multiple viral classes: groups M (major), O (outlying) and N (new) [90]. Group M , which accounts for the vast majority of HIV-1 infections worldwide, may be further subdivided into multiple subtypes, or clades ( A - D , F - H and K ) , as well as numerous circulating recombinant forms (CRFs), a fact which has important implications for the development of future treatments and/or vaccines for H I V , as well as for resistance to antiretroviral agents [90,91]. 1.7 Relevance of host and viral genetic factors in the era of HAART The majority of HIV-infected individuals, at least in the developed world, now have potential access to combination antiretroviral treatment. It is therefore of importance to re-evaluate the relative contributions of known host and viral genetic determinants of untreated H I V disease progression to clinical outcomes the current era of modern antiretroviral therapies. A s H A A R T is such a potent and effective tool in the treatment of H I V infection, it is not known what residual effects, i f any, host and/or viral genetic factors may contribute above and beyond the effects of antiretroviral treatment. Population-based studies conducted 10 on HIV-infected, treated populations are therefore necessary to evaluate the influence of specific host and viral genetic variants on short and longer-term responses to therapy. The current era of H A A R T has also resulted in the need to identify factors which may more directly affect antiretroviral therapy response. The field of pharmacogenetics, or the study of how host genetic factors influence the uptake, metabolism, bioavailability, clearance, clinical response, and adverse reactions to medications, is now a major focus of H I V research. Considering the complexity of current combination therapies, as well as the lifelong nature of antiretroviral treatment, it is of importance that a greater understanding of interindividual differences in medication responses be achieved. Examples of human genes implicated in differential responses to antiretroviral therapy include a naturally-occurring polymorphism in a gene encoding a membrane multidrug efflux transporter whose substrates include H I V protease inhibitors [92,93], as well as naturally-occurring genetic variation in the H L A region, involved in the recognition of foreign antigen by cytotoxic (CD8+) T-lymphocytes, where a specific H L A - B genetic variant (HLA-B*5701) is believed to cause a severe and potentially life-threatening hypersensitivity reaction to the N R T I abacavir [94,95]. The relevance of viral genetic factors in the era of antiretroviral therapy is most clearly demonstrated by the emergence of drug-resistant H I V as a major barrier to antiretroviral treatment efficacy [96,97,98,99,100]. In fact, current guidelines for the clinical management and treatment of HIV-infected individuals recommend the routine use of H I V antiretroviral resistance genotyping assays to help guide decisions regarding therapy [101,102]. A s discussed in later chapters, it is highly likely that variation in the genes encoding current and future drug targets, as well as in other areas of the H I V genome [103], 11 wi l l play a significant role in determining an individual's response to current and future antiretroviral therapies. Finally, although plasma viral load, C D 4 counts and H I V genotypic drug resistance tests currently are, and w i l l remain important tools in the clinical monitoring of H I V disease and therapy outcomes, it is important to continue to investigate and evaluate additional markers, including genetic markers, which may independently influence therapy response. Again, the relevance of novel genetic markers to antiretroviral therapy response, as well as the potential clinical utility of these genetic markers to the management of H I V infection should be evaluated in large, observational, population-based studies of HIV-infected, treated populations, over substantial follow-up times. Although the "correction" of human genetic variation using techniques such as gene therapy remain a very distant goal, the incorporation of relevant genetic screening assays into routine clinical practise w i l l l ikely be a realistic and feasible option for clinical management of H I V infected individuals in the future. Data obtained through longitudinal, population-based studies of individuals receiving antiretroviral treatment wi l l help elucidate the individual and combined contributions of host and/or viral genetic factors to therapy outcomes, and hopefully wi l l contribute to resolving still outstanding questions regarding the optimal time to initiate antiretroviral therapy [68] and/or which therapies may be most appropriate under different circumstances. A s outlined in detail below, the main objective of this thesis is to evaluate the clinical relevance of host and viral genetic variation in the era of H A A R T in a population setting, in order to investigate the potential utility of host and viral genetic factors as prognostic indicators of response modern antiretroviral therapies. 12 1.8 Thesis objectives and organization This thesis is divided into nine chapters. This first chapter provides a general introduction to H I V pathogenesis and treatment, as well as outlines the relevance and objectives of the thesis research. Chapter two is a detailed literature review outlining the impact of host and viral genetic variation on H I V infection and disease progression, and addresses the relevance of these factors in context of the modern era of combination antiretroviral therapy. Chapters three through eight address the main thesis objectives as outlined below. Chapter nine provides a detailed discussion and summary of the research results, and comments on the implications of the thesis work. The general aim of this thesis is to evaluate the relevance of selected host and viral genetic factors to HIV disease in the modern era of combination antiretroviral therapy in a population setting. The host and viral genetic factors selected for investigation are all factors which have been shown to influence H I V infection and /or disease progression in untreated infection, or have been implicated as playing a direct role in the metabolism and/or bioavailability of specific antiretrovirals. Chapters 3,4,6,7, and 8 are population-based studies conducted using retrospective, longitudinal analysis of cohorts of HIV-infected, antiretroviral naive individuals initiating their first therapy in order to investigate the relative contributions of host chemokine receptor polymorphisms (chapters 3, 4), polymorphisms potentially affecting drug bioavailability (chapter 4), as well as viral genetic factors implicated in H I V pathogenesis (chapters 6,7,8) to short and long-term clinical responses to antiretroviral therapy. The specific objectives of these investigations are 1. to determine the prevalence of specific human and viral genetic factors in a cross-sectional analysis of the cohort population at baseline (pre-therapy); 2. to investigate sociodemographic and clinical 13 correlates of these genetic factors in the context of untreated infection; 3. to use longitudinally-collected plasma viral load and C D 4 count data in order to establish the clinical impact of these specific genotypes on initial virologic (plasma viral load) and immunologic (CD4 count) responses to antiretroviral therapy; and 4. to evaluate the long-term impact of these genotypes on response to H A A R T using mortality data. Chapter 5 employs a retrospective case-control approach and is designed to evaluate the relative contribution of viral genetic variation in the form of H I V antiretroviral resistance to mortality in the H A A R T era. The thesis will therefore address both the short and longer-term implications of host and viral genetic factors in the modern era of antiretroviral therapy. 1.9 Brief overview of data sources, and evolution of our population-based cohorts A s w i l l be discussed in detail in the following chapters, the population-based cohorts of HIV-infected, antiretroviral-treated individuals studied as a part of this thesis are drawn from the B . C . Centre for Excellence in H I V / A I D S Drug Treatment Program (DTP), based at St. Paul's Hospital in Vancouver, Canada. Briefly, antiretroviral medications in British Columbia are available free of charge to HIV-infected individuals and are centrally distributed through the D T P , which remains the only free source of antiretroviral medication in the province. The B . C . Centre for Excellence in H I V / A I D S Drug Treatment Program captures -99% of HIV-infected individuals in British Columbia receiving antiretroviral therapy [104]. Individuals are automatically entered into the D T P when they are first prescribed any antiretroviral agent. A t D T P entry and with each subsequent physician visit, the participant's complete history (if any) of antiretroviral use, C D 4 count, plasma viral load, and disease stage, as well as sociodemographic information, are recorded [105]. A s a part of 14 the current H I V / A I D S treatment guidelines, HIV-infected patients are monitored at approximately three-month intervals with respect to C D 4 counts and H I V plasma viral loads [105]. In fact, the D T P represents the largest H I V treatment-based, Canadian cohort developed to date, and one of the longest followed and best-studied cohorts in the world with the capacity to integrate epidemiological, clinical and laboratory data. The B . C . Centre for Excellence in H I V / A I D S Drug Treatment Program is therefore uniquely positioned to conduct population-based studies that reflect the virologic and clinical status of large numbers of HIV-infected individuals being treated with antiretroviral agents, and to undertake studies investigating the impact of human and viral genetic variation on antiretroviral therapy outcomes, such as described here. A s the D T P represents an open cohort in which patients are continuously being enrolled, and follow-up data continuously being collected, there has been substantial evolution of cohort composition over the course of time, reflecting changes in patient sociodemographics, treatment guidelines and other aspects of clinical practise. Also by definition, there is a ongoing accumulation of follow-up data as individuals continue to be monitored through the D T P . A t the time the thesis research was initiated, research focused on a population-based cohort of 461 individuals, of which -80% were treated with H A A R T (chapters 4 and 6), while later on, a substantially larger cohort comprising >1000 individuals, all of whom initiated H A A R T , was created (chapters, 3, 7 and 8). In addition, as time progressed, it was possible to extend follow-up times and include evaluations of long-term outcomes such as mortality (chapters 3,7 and 8). Throughout this thesis, therefore, there are some variations in cohort composition, follow-up times and available genotype data, which reflect the fact that with each study we wished to include as much longitudinal outcome data 15 as possible. Please note that this in some cases, slight inter-chapter differences in reported N , p-values, odds ratios etc. may be observed, due to maturation of the cohorts and inclusion of as much follow-up data as was available at the time. In addition, for reasons of style and composition, please note that the thesis chapters are not presented in the order in which they were undertaken. This thesis is prepared according to a manuscript-based format. Chapter two is a literature review summarizing the current knowledge surrounding the relevance of host and viral genetic factors in the era of H A A R T . Chapters three through eight each represent stand-alone manuscripts, which have all been published in international, peer-reviewed journals. Journals in which the individual thesis chapters have been published include Journal of Infectious Disease, AIDS and Antiviral Therapy. The candidate is either the lead author (chapters 3, 4, 6, 7, 8) or second author (chapter 5) on all manuscripts. The final chapter (chapter 9) discusses and summarizes the overall study results, and comments upon the implications and relevance of these findings in the context of the current state of knowledge. References are found at the end of each chapter. 16 1.10 R E F E R E N C E S 1. Hahn B H , Shaw G M , Arya S K , Popovic M , Gallo R C , Wong-Staal F. Molecular cloning and characterization of the H T L V - I I I virus associated with A I D S . Nature 1984;312:166-9. 2. A l i z o n M , Sonigo P, Barre-Sinoussi F, et al. Molecular cloning of lymphadenopathy-associated virus. Nature 1984 Dec 20-1985 Jan 2; 312:757-60. 3. Broder S, Gallo R C . A pathogenic retrovirus (HTLV-III ) linked to A I D S . N Engl J Med. 1984;311:1292-7. 4. U N A I D S A V H O . A I D S Epidemic Update 2004. 5. U N A I D S A V H O . A I D S Epidemic Update 2000. 6. Geduld J, Gatali M , Remis R S , Archibald C P . Estimates of H I V prevalence and incidence in Canada, 2002. Can Commun Dis Rep 2003; 29:197-207. 7. Dalgleish A G , Beverley P C , Clapham PR, Crawford D H , Greaves M F , Weiss R A . The C D 4 (T4) antigen is an essential component of the receptor for the A I D S retrovirus. Nature, 1984 Dec 20-1985 Jan 2; 312:763-7. 8. Klatzmann D , Champagne E , Chamaret S, et al. T-lymphocyte T4 molecule behaves as the receptor for human retrovirus L A V . Nature. 1984 Dec 20-1985 Jan 2; 312:767-8. 9. Centers for Disease Control and Prevention 1993 Revised Classification System for H I V Infection and Expanded Surveillance Case Definition for A I D S Among Adolescents and Adults. M M W R 1992 / 41(RR-17) 10. Ratner L , Haseltine W , Patarca R, et al. Complete nucleotide sequence of the A I D S virus, H T L V - I I I . Nature. 1985;313:277-84. 17 11. Freed, E O . HIV-1 replication. Somat Cel l M o l Genet 2001; 26:13-33. 12. Frankel A D , Young J A . H I V - 1 : fifteen proteins and an R N A . Annu Rev Biochem. 1998;67:1-25. 13. Moulard M , Decroly E . Maturation of H I V envelope glycoprotein precursors by cellular endoproteases. Biochim Biophys Acta. 2000;1469:121-32. 14. McDougal JS, Nicholson J K , Cross G D , Cort SP, Kennedy M S , Mawle A C . Binding of the human retrovirus H T L V - I I I / L A V / A R V / H I V to the C D 4 (T4) molecule: conformation dependence, epitope mapping, antibody inhibition, and potential for idiotypic mimicry. J Immunol 1986; 137:2937-44. 15. Feng Y , Broder C C , Kennedy PE , Berger E A . HIV-1 entry cofactor: functional c D N A cloning of a seven-transmembrane, G protein-coupled receptor. Science 1996; 272:872-7. 16. Samson M , Labbe O, Mollereau C, Vassart G , Parmentier M . Molecular cloning and functional expression of a new human CC-chemokine receptor gene. Biochemistry 1996, 35:3362-3367. 17. Deng H K , L i u R, Ellmeier W , Choe S, et al. Identification of a major co-receptor for primary isolates of H I V - 1 . Nature 1996, 381:661-666. 18. Dragic T, L i twin V , Al laway G P , et al. HIV-1 entry into CD4+ cells is mediated by the chemokine receptor C C - C K R - 5 . Nature 1996, 381:667-673. 19. Wi l ley SJ, Reeves JD , Hudson R, et al. Identification of a subset of human immunodeficiency virus type 1 (HIV-1), HIV-2 , and simian immunodeficiency virus strains able to exploit an alternative coreceptor on untransformed human brain and lymphoid cells. J V i r o l 2003; 77: 6138-52. 18 20. Zhang Y J , Dragic T, Cao Y , et al. Use of coreceptors other than C C R 5 by non-syncytium-inducing adult and pediatric isolates of human immunodeficiency virus type 1 is rare in vitro. J V i r o l 1998; 72: 9337-44. 21. Lee S, Tiffany H L , K i n g L , Murphy P M , Golding H , Zaitseva M B . C C R 8 on human thymocytes functions as a human immunodeficiency virus type 1 coreceptor. J V i ro l 2000; 74: 6946-52. 22. Flamand L , Crowley R W , Lusso P, Colombini-Hatch S, Margolis D M , Gallo R C . Activation of CD8+ T lymphocytes through the T cell receptor turns on C D 4 gene expression: implications for H I V pathogenesis. Proc Natl Acad Sci U S A 1998; 95:3111-6. 23. Livingstone W J , Moore M , Innes D , Be l l JE, Simmonds P. Frequent infection of peripheral blood CD8-positive T-lymphocytes with H I V - 1 . Lancet 1996 ; 348:649-54. 24. Mercure, L . , D . Phaneuf, and M . A . Wainberg. Detection of unintegrated human immunodeficiency virus type 1 D N A in persistently infected CD8+ cells. J Gen Vi ro l 1993; 74:2077-83. 25. Berger E A , Murphy P M , Farber J M . Chemokine receptors as HIV-1 coreceptors: roles in viral entry, tropism, and disease. Annu Rev Immunol. 1999;17:657-700. 26. Strebel K . Virus-host interactions: role of H I V proteins Vi f , Tat, and Rev. A I D S . 2003;17Suppl4:S25-34. 27. Demirov D G , Orenstein J M & Freed E O . The late domain of human immunodeficiency virus type 1 p6 promotes virus release in a cell type-dependent manner. Journal of Virology 2002; 76:105-117. 19 28. Time from HIV-1 seroconversion to A I D S and death before widespread use of highly-active antiretroviral therapy: a collaborative re-analysis. Collaborative Group on A I D S Incubation and H I V Survival including the C A S C A D E E U Concerted Action. Concerted Act ion on Seroconversion to A I D S and Death in Europe. Lancet. 2000;355:1131-7. 29. Luo K , Law M , Kaldor J M et al. The role of initial AIDS-defining illness in survival following A I D S . A I D S 1995; 9:57-63. 30. Vel la S, Giuliano M , Pezzotti P et al. Survival of zidovudine-treated patients with A I D S compared with that of contemporary untreated patients. J A M A 1992; 267:1232-1236. 31. Solomon PJ , Wilson SR, Swanson C E , Cooper D A . Effect of zidovudine on survival of patients with A I D S in Australia. Med J Aust 1990;153 :254-7. 32. Lemp G F , Payne SF, Neal D , Temelso T, Rutherford G W . Survival trends for patients with A I D S . J A M A 1990; 263:402-6. 33. Cox P H , Martin M A , Styer C M , Beall G N . Outcomes of treatment with A Z T of patients with A I D S and symptomatic H I V infection. Nurse Pract 1990; 15:36, 39-44. 34. Hammer S M , Katzenstein D A , Hughes M D , et al. A trial comparing nucleoside monotherapy with combination therapy in HIV-infected adults with C D 4 cell counts from 200 to 500 per cubic millimeter. N Engl J M e d 1996; 335:1081-90. 35. Schooley R T , Ramirez-Ronda C , Lange J M , et al. Virologic and immunologic benefits of initial combination therapy with zidovudine and zalcitabine or didanosine compared with zidovudine monotherapy. J Infect Dis 1996;173:1354-66. 20 36. Gatell J M , Leal M , Mallolas J, et al. A pilot case-control study of zidovudine compared with zidovudine plus didanosine in patients with advanced HIV-1 disease and no previous experience with antiretrovirals. Antivir Ther 1996; 1:105-12. 37. Katlama C , Ingrand D , Loveday C, et al. Safety and efficacy of lamivudine-zidovudine combination therapy in antiretroviral-naive patients. A randomized controlled comparison with zidovudine monotherapy. J A M A 1996; 276:118-25. 38. Collier A C , Coombs R W , Fischl M A , et al. Combination therapy with zidovudine and didanosine compared with zidovudine alone in HIV-1 infection. A n n Intern Med. 1993 Oct 15;119(8):786-93. 39. D A q u i l a R T , Johnson V A , Welles S L , et al. Zidovudine resistance and HIV-1 disease progression during antiretroviral therapy. A n n Intern M e d 1995; 122:401-8. 40. Shirasaka T, Yarchoan R, O'Brien M C , et al. Changes in drug sensitivity of human immunodeficiency virus type 1 during therapy with azidothymidine, dideoxycytidine, and dideoxyinosine: an in vitro comparative study. Proc Natl Acad Sci U S A 1993; 90:562-6. 41. Shirasaka T, Kavl ick M F , Ueno T, et al. Emergence of human immunodeficiency virus type 1 variants with resistance to multiple dideoxynucleosides in patients receiving therapy with dideoxynucleosides. Proc Natl Acad Sci U S A 1995;92:2398-402. 42. Palella FJ Jr, Delaney K M , Moorman A C , et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 1998; 338:853-60. 21 43. Hogg R S , Y i p B , K u l l y C , Craib K J , et al. Improved survival among HIV-infected patients after initiation of triple-drug antiretroviral regimens. Can M e d Assoc J 1999; 160:659-65. 44. Mocroft A , Ve l l a S, Benfield T L , et al. Changing patterns of mortality across Europe in patients infected with H I V - 1 . EuroSIDA Study Group. Lancet. 1998; 352:1725-30. 45. Detels R, Munoz A , McFarlane G , et al. Effectiveness of potent antiretroviral therapy on time to A I D S and death in men with known H I V infection duration. Multicenter A I D S Cohort Study Investigators. J A M A . 1998; 280:1497-503. 46. Y e n i P G , Hammer S M , Hirsch M S , et al. Treatment for adult H I V infection: 2004 recommendations of the International A I D S Society-USA Panel. J A M A 2004; 292:251-65. 47. Lazzarin A , Clotet B , Cooper D , et al. Efficacy of enfuvirtide in patients infected with drug-resistant HIV-1 in Europe and Australia. N Engl J M e d 2003; 348:2186-95. 48. Lalezari JP, Henry K , O'Hearn M , et al. Enfuvirtide, an HIV-1 fusion inhibitor, for drug-resistant H I V infection in North and South America. N Engl J Med. 2003; 348:2175-85. 49. Scozzafava A , Mastrolorenzo A , Supuran C T . Non-peptidic chemokine receptors antagonists as emerging anti-HIV agents. J Enzyme Inhib M e d Chem 2002; 17:69-76. 50. Pommier Y , Johnson A A , Marchand C. Integrase inhibitors to treat H I V / A I D S . Nat Rev Drug Discov 2005; 4:236-48. 51. Budihas SR, Gorshkova I, Gaidamakov S, et al. Selective inhibition of HIV-1 reverse transcriptase-associated ribonuclease H activity by hydroxylated tropolones. Nucleic Acids Res. 2005; 33:1249-56. 22 52. Barbaro G , Scozzafava A , Mastrolorenzo A , Supuran C T . Highly active antiretroviral therapy: current state of the art, new agents and their pharmacological interactions useful for improving therapeutic outcome. Curr Pharm Des 2005; 11:1805-43. 53. Mellors J K , Alvaro M , Janis G et al. Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med 1997; 126:946-954. 54. Vlahov D , Graham N , Hoover D , et al. Prognostic indicators for A I D S and infectious disease death in HIV-infected injection drug users: plasma viral load and CD4+ cell count. J A M A . 1998; 279:35-40. 55. Fahey J L , Taylor J M , Detels R, et al. The prognostic value of cellular and serologic markers in infection with human immunodeficiency virus type 1. N Engl J Med 1990; 322:166-72. 56. Mulder J, Resnick R, Saget B , et al. A rapid and simple method for extracting human immunodeficiency virus type 1 R N A from plasma: enhanced sensitivity. J C l i n Microbiol . 1997; 35:1278-80. 57. van Gemen B , Kievits T, Schukkink R, et al. Quantification of HIV-1 R N A in plasma using N A S B A during HIV-1 primary infection. J V i r o l Methods. 1993; 43:177-87. 58. Pachl C , Todd J A , Kern D G , et al. Rapid and precise quantification of HIV-1 R N A in plasma using a branched D N A signal amplification assay. J Acquir Immune Defic Syndr Hum Retrovirol. 1995; 8:446-54. 59. Mellors J W , Kingsley L A , Rinaldo C R Jr, et al. Quantitation of HIV-1 R N A in plasma predicts outcome after seroconversion. A n n Intern Med. 1995; 122:573-9. 23 60. Mellors J W , Rinaldo C R , Gupta P, et al. Prognosis in HIV-1 infection predicted by the quantity of virus in plasma. Science 1996;272:1167-1170. 61. O'Brien W A , Hartigan P M , Martin D , et al. Changes in plasma HIV-1 R N A and CD4+ lymphocyte counts and the risk of progression to A I D S . Veterans Affairs Cooperative Study Group on A I D S . N Engl J M e d 1996; 334:426-31. 62. Marschner IC, Collier A C , Coombs R W , et al. Use of changes in plasma levels of human immunodeficiency virus type 1 R N A to assess the clinical benefit of antiretroviral therapy. J Infect Dis 1998; 177:40-7. 63. Montaner JS, DeMasi R, H i l l A M . The effects of lamivudine treatment on HIV-1 disease progression are highly correlated with plasma HIV-1 R N A and C D 4 cell count. A I D S . 1998Mar26;12(5):F23-8. 64. O'Brien W A , Hartigan P M , Daar E S , Simberkoff M S , Hamilton JD. Changes in plasma H I V R N A levels and CD4+ lymphocyte counts predict both response to antiretroviral therapy and therapeutic failure. V A Cooperative Study Group on A I D S . A n n Intern Med. 1997; 126:939-45. 65. Chene G , Sterne J A , M a y M , et al. Prognostic importance of initial response in HIV-1 infected patients starting potent antiretroviral therapy: analysis of prospective studies. Lancet. 2003; 362:679-86. 66. Egger M , M a y M , Chene G , et al. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet. 2002; 360:119-29. 24 67. Yen i P G , Hammer S M , Carpenter C C , et al. Antiretroviral treatment for adult H I V infection in 2002: updated recommendations of the International A I D S Society-USA Panel. J A M A 2002; 288:222-35. 68. Wood E , Hogg R S , Harrigan PR, Montaner JS. When to initiate antiretroviral therapy in HIV-1-infected adults: a review for clinicians and patients. Lancet Infect Dis. 2005; 5:407-14. 69. Kulkarni PS, Butera ST, Duerr A C . Resistance to HIV-1 infection: lessons learned from studies of highly exposed persistently seronegative (HEPS) individuals. A I D S Rev 2003; 5:87-103. 70. M c N i c h o l l J M , Promadej N . Insights into the role of host genetic and T-cell factors in resistance to H I V transmission from studies of highly HIV-exposed Thais. Immunol Res. 2004; 29:161-74. 71. Phair J, Jacobson L , Detels R, Rinaldo C, Saah A , Schrager L , Munoz A . Acquired immune deficiency syndrome occurring within 5 years of infection with human immunodeficiency virus type-1: the Multicenter A I D S Cohort Study. J Acquir Immune Defic Syndr 1992; 5:490-6. 72. Huang Y , Paxton W A , Wolinsky S M , et al. The role of a mutant C C R 5 allele in HIV-1 transmission and disease progression. Nat M e d 1996, 11:1240-1243. 73. Dean M , Carrington M , Winkler C , et al. Genetic restriction of HIV-1 infection and progression to A I D S by a deletion allele of the C K R 5 structural gene. Science 1996, 273:1856-1862. 25 74. L i u R, Paxton W A , Choe S, et al. Homozygous defect in HIV-1 coreceptor accounts for resistance of some multiply-exposed individuals to HIV-1 infection. Cel l 1996; 86:367-77. 75. Samson M , Libert F , Doranz, B J , et al.: Resistance to HIV-1 infection in Caucasian individuals bearing mutant alleles of the C C R 5 chemokine receptor gene. Nature 1996, 382:722-725. 76. Benkirane M , J in D Y , Chun R F , Koup R A , Jeang K T . Mechanism of transdominant inhibition of CCR5-mediated HIV-1 infection by ccr5delta32. J B i o l Chem. 1997 Dec 5;272(49):30603-6. 77. Faure S, Meyer L , Costagliola D , et al. Rapid Progression to A I D S in HIV+ individuals with a Structural Variant of the Chemokine Receptor C X 3 C R 1 . Science 2000,287: 2274-2277. 78. Kostrikis L G , Huang Y , Moore JP, et al. A chemokine receptor C C R 2 allele delays HIV-1 disease progression and is associated with a C C R 5 promoter mutation. Nat Med 1998;4:350-3. 79. Meyer L , Magierowska M , Hubert JB , et al. CC-chemokine receptor variants, SDF-1 polymorphism, and disease progression in 720 HIV-infected patients. S E R O C O Cohort. Amsterdam Cohort Studies on A I D S . A I D S . 1999;13:624-6. 80. Gonzalez E , Kulkarni H , Bolivar H , et al. The influence of C C L 3 L 1 gene-containing segmental duplications on HIV-1 /AIDS susceptibility. Science. 2005; 307:1434-40. 81. Kaslow R A , Carrington M , Apple R, et al. Influence of combinations of human major histocompatibility complex genes on the course of HIV-1 infection. Nat Med 1996;2:405-11. 26 82. Tersmette M , Lange J M , de Goede R E , et al. Association between biological properties of human immunodeficiency virus variants and risk for A I D S and A I D S mortality. Lancet. 1989 M a y 6;l(8645):983-5. 83. Ward JW, Bush TJ , Perkins H A , et al. The natural history of transfusion-associated infection with human immunodeficiency virus. Factors influencing the rate of progression to disease. N Engl J Med. 1989; 321:947-52. 84. M a l i m M H , Emerman M . HIV-1 sequence variation: drift, shift, and attenuation. Ce l l . 2001; 104:469-72. 85. Tersmette M , Gruters R A , de W o l f F, et al. Evidence for a role of virulent human immunodeficiency virus (HIV) variants in the pathogenesis of acquired immunodeficiency syndrome: studies on sequential H I V isolates. J V i ro l 1989; 63: 2118-25. 86. Schuitemaker H , Koot M , Kootstra N A , et al. Biological phenotype of human immunodeficiency virus type 1 clones at different stages of infection: progression of disease is associated with a shift from monocytotropic to T-cell tropic virus populations. J V i r o l 1992;66:1354-60. 87. Rhodes D l , Ashton L , Solomon A , et al. Characterization of three nef-defective human immunodeficiency virus type 1 strains associated with long-term nonprogression. J V i r o l 2000, 74:10581-88. 88. Kirchhoff F, Easterbrook PJ, Douglas N , et al. Sequence variations in human immunodeficiency virus type 1 Nef are associated with different stages of disease. J. V i r o l 1999; 73:5497-508. 27 89. L u m JJ, Cohen OJ , Nie Z , et al. Vpr R77Q is associated with long-term nonprogressive H I V infection and impaired induction of apoptosis. J C l i n Invest 2003;111:1547-1554. 90. Wainberg M A . HIV-1 subtype distribution and the problem of drug resistance. A I D S . 2004 Jun;18 Suppl 3:S63-8. 91. McMichae l A , M w a u M , Hanke T. H I V T cell vaccines, the importance of clades. Vaccine. 2002; 20:1918-21. 92. Jones K , Bray P G , Khoo S H , et al. P-Glycoprotein and transporter M R P - 1 reduce H I V protease inhibitor uptake in C D 4 cells: potential for accelerated viral drug resistance? A I D S 2001, 15: 1353-1358. 93. Hoffmeyer S, Burk O, von Richter O, et al. Functional polymorphisms of the human multidrug-resistance gene: Multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo. P N A S 2000, 97: 3473-3478. 94. Mal la l S, Nolan D , Witt C , et al. Association between presence of H L A - B * 5 7 0 1 , H L A - D R 7 , and H L A - D Q 3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet. 2002;359:727-32. 95. Hetherington S, Hughes A R , Mosteller M , et al. Genetic variations in H L A - B region and hypersensitivity reactions to abacavir. Lancet. 2002; 359:1121-2. 96. Johnson V A , Brun-Vezinet F, Clotet B , et al. Update of the Drug Resistance Mutations in H I V - 1 : 2005. Top H I V Med. 2005; 13:51-7. 97. Hirsch M S , Brun-Vezinet F, Clotet B , et al. Antiretroviral drug resistance testing in adults infected with human immunodeficiency virus type 1: 2003 recommendations of an International A I D S Society-USA Panel. C l i n Infect Dis 2003; 37:113-28. 28 98. Sista PR, Melby T, Davison D , et al. Characterization of determinants of genotypic and phenotypic resistance to enfuvirtide in baseline and on-treatment HIV-1 isolates. A I D S 2004; 18:1787-94. 99. Wainberg M A , Friedland G . Public health implications of antiretroviral therapy and H I V drug resistance. J A M A . 1998; 279:1977-83. 100. Pillay D , Taylor S and Richman D D . Incidence and impact of resistance against approved antiretroviral drugs. Rev M e d V i r o l . 2000; 10:231-53. 101. Hirsch M S , Brun-Vezinet F, Clotet B , et al. Antiretroviral drug resistance testing in adults infected with human immunodeficiency virus type 1: 2003 recommendations of an International A I D S Society-USA Panel. C l i n Infect Dis 2003; 37:113-28. 102. Vandamme A M , Sonnerborg A , Ait-Khaled M , et al. Updated European recommendations for the clinical use of H I V drug resistance testing. Antivir Ther. 2004; 9:829-48. 103. Peters S, Munoz M , Yerly S, et al. Resistance to nucleoside analog reverse transcriptase inhibitors mediated by human immunodeficiency virus type 1 p6 protein. J V i r o l 2001; 75:9644-9653. 104. Wood E , Montaner JS, Tyndall M W , Schechter M T , O'Shaughnessy M V , Hogg R S . Prevalence and correlates of untreated human immunodeficiency virus type 1 infection among persons who have died in the era of modern antiretroviral therapy. J Infect Dis . 2003; 188:1164-70. 105. Hogg R S , Y i p B , Chan K J , et al. Rates of disease progression by baseline C D 4 cell count and viral load after initiating triple-drug therapy. J A M A 2001; 286:2568-77. 29 CHAPTER 2 LITERATURE REVIEW: Impact of human and viral genetic variation on HIV disease progression and response to therapy in the era of Highly Active Antiretroviral Therapy (HAART) 2.1 Introduction The natural course of Human Immunodeficiency Virus Type 1 (HIV-1) disease progression varies widely among individuals [for a review, see 1]. Not all people exposed to HIV-1 become infected, and of those who do, the time course of untreated disease varies greatly. Although the median time from infection to A I D S diagnosis ranges from approximately 5 to 11 years, depending on age at seroconversion [2], there exist individuals who progress to A I D S in as little as one year from infection ("fast progressors"), and others who remain asymptomatic with normal C D 4 counts and low plasma viral loads for 20 years or more ("slow progressors"). Historically, studies of multiply-exposed yet uninfected individuals helped to identify factors which determine "natural" protection from H I V infection [3,4,5,6], while studies comparing characteristics of fast and slow progressors led to the identification of host and viral factors which influence the natural course of H I V disease progression [7,8,9,10,11,12]. A wide spectrum of host and viral genetic factors contribute to the extensive variability in the natural course of HIV-1 disease [7,8,9,10,11]. This review wi l l describe some of the most important findings in this area, and w i l l comment upon the relevance of these findings in context of the modern era of Highly Active Antiretroviral Therapy ( H A A R T ) [13]. 30 2.2 Human Genetic Factors: Chemokine receptor polymorphisms In the mid-1990s, the identification of the human chemokine receptors C X C R 4 [14] and C C R 5 [15,16,17] as co-receptors for H I V entry into CD4+ cells led to the discovery that genetic variation among chemokine receptor genes could directly influence susceptibility to infection with H I V , as well as the subsequent rate of disease progression. To date, the best-studied example of the contribution of chemokine receptor variation to H I V disease is the C C R 5 A 3 2 , a naturally-occurring 32 base-pair deletion in the gene encoding C C R 5 [6]. Individuals homozygous for C C R 5 A 3 2 do not express C C R 5 and are thus naturally "resistant" to H I V infection [6,22,23], although rare exceptions have been documented where CCR5A32/A32 individuals have been infected with CXCR4-us ing strains [18,19]. Individuals heterozygous for the C C R 5 A 3 2 have been shown to express C C R 5 at significantly reduced levels on the cell-surface [20], and although these individuals are equally susceptible to H I V infection as C C R 5 wild-type individuals, they appear to progress to A I D S and death at a significantly slower rate after infection than those without this mutation [21,22,23,24,25,26]. Polymorphisms in the promoter region of C C R 5 , presumably affecting expression of the receptor, have also been shown to influence H I V disease progression [27,28]. In addition, several naturally-occurring polymorphisms in minor H I V co-receptors have been reported, including a V62I mutation in the C C R 2 gene, which has been shown to confer a protective effect with respect to progression to A I D S [29,30], although not in all studies [31]. The fact that the V62I mutation is in complete linkage disequilibrium with the aforementioned C C R 5 promoter polymorphisms [30] may explain this association. More recently, two common single nucleotide polymorphisms in the human C X 3 C R 1 receptor, a 31 minor co-receptor for HIV-1 [32] have been identified [33]. HIV-1 infected individuals homozygous for the C X 3 C R I 2491 and 280M amino acid substitutions were shown to progress to A I D S and death more rapidly than those with other haplotypes [33], although these results were not confirmed in an independent study (which examined the polymorphism at codon 280 only) [34]. To date no polymorphisms relevant to H I V transmission or pathogenesis have been reported in C X C R 4 , the major co-receptor for T-cell-tropic, syncytium-inducing strains of H I V , although a polymorphism has been reported in the gene encoding its natural chemokine ligand, stromal-derived factor (SDF) - l [35]. SDF-1 may act as a natural competitive inhibitor for HIV-1 binding to CXCR4-expressing cells. The SDF-1 3 'a mutation, which lies in an untranslated region of the gene, may upregulate the expression of SDF-1 and thus potentially increase its ability to act as a competitive inhibitor [8]. However, the clinical relevance of the SDF-1 3 'a mutation has been controversial and reports have ranged from a protective effect [35] to no effect [29] to a detrimental effect [36] of this mutation on progression to A I D S in untreated H I V infection. 2.3 Mutations in genes that modulate the immune response Variation in the genes modulating the immune response also influence the course of H I V disease. There is now ample evidence that specific alleles of the Human Leukocyte Antigen ( H L A ) system, located on chromosome 6 and representing one of the most genetically diverse regions in the human genome, play a significant role in the immune control of H I V [10,37,38]. Their function is to allow the immune system to recognize and combat a wide range of pathogens, including H I V , through mediation of the cytotoxic (CD8+) T-cell response [10,37,38]. It is been suggested that H L A genetic variation 32 contributes to diversity of antigen recognition on both an individual as well as a population level [39]. Evidence strongly suggests that individuals homozygous at any of the H L A loci are at higher risk of more rapid H I V disease progression [40,41]. This is most likely because H L A homozygosity may limit the immune system's ability to respond to a diverse range of peptide antigens, thus allowing H I V to more easily adapt to avoid the host immune response [41]. In addition, several specific H L A alleles, most notably at the H L A - B locus, influence the natural course of H I V disease, including H L A - B 5 7 , H L A - B 2 7 and H L A - B 3 5 , among others [42,43,44,45,46,47,48]. In addition, H L A genetic variation directly influences viral genetic diversity: at the population level, the predominant or "consensus" H I V sequence reflects viral genetic adaptation to the most common H L A types in the infected population [49]. Finally, a combined analysis of both H L A type and corresponding viral sequence predicted pre-treatment plasma viral load [49], indicating that the H L A region is a major contributor to the natural course of H I V infection. 2.4 Relevance of chemokine and immunogenetic factors in the era of HAART Although host genetic variation plays a significant role in untreated H I V infection, the relevance of these factors in the context of H A A R T remains to be determined. Studies investigating whether the C C R 5 A 3 2 mutation confers any additional protective effect during treatment with H A A R T have produced conflicting results, with some studies reporting an independent protective effect [50,51,52,53], and others report no significant effect [54,55]. A recent study from our laboratory investigating the effect of C C R 5 A 3 2 on initial virologic and immunologic outcomes as well as survival in >1000 HIV-infected individuals initiating H A A R T reported a small protective effect of C C R 5 A 3 2 on survival, although this was not statistically significant [56]. Similarly a study from our laboratory has identified the 33 C X 3 C R I 2491 mutation as being associated with a poorer immunologic response to H A A R T , although the mechanism for this remains undetermined [57]. A small number of studies investigating other chemokine receptor and/or ligand polymorphisms in the context of H A A R T have reported variable outcomes [58,59]. Overall, studies to date appear to indicate that host chemokine genetic factors may indeed contribute to disease progression and treatment response in the context of H A A R T ; although in some cases the relative contribution may be small. Perhaps of greater interest is emerging evidence that human chemokine genotypes may act as a selective pressure mediating genetic evolution of the virus, and thus indirectly contribute to disease progression and/or response to current and future antiretroviral agents [60]. 2.5 Pharmacogenetics in the era of HAART The modern era of antiretroviral therapy has opened up an entirely new area of research in the field of H I V / A I D S : pharmacogenetics, or the study of how polymorphisms in the human genome affect the uptake, metabolism, bioavailability, clearance, clinical response as well as adverse reactions to medications. A n example of a well-studied human gene in the field of pharmacogenetics is the Multi-Drug-Resistance ( M D R ) - l gene, implicated in the development of resistance to a variety of medications [61,62], including antiretrovirals. The M D R - 1 gene encodes the P-glycoprotein (P-gp) membrane efflux transporter, which is expressed on a variety of cell types including CD4+ lymphocytes [63], and possesses a broad substrate specificity [64,65,66,67] which includes H I V protease inhibitors [68, 69]. A single C-to-T nucleotide polymorphism (C3435T) in exon 26 of M D R - 1 is significantly correlated with decreased intestinal expression levels and activity of P-glycoprotein [70]. M D R - 1 expression levels are 34 correlated with the resistance of cancer cells to chemotherapeutic agents [61,62]. Indeed it has been shown that overexpression of multidrug transporters significantly reduces the accumulation of Pis in human lymphocytes [72], and recently that intracellular nelfinavir concentrations in lymphoblastoid cell lines expressing the homozygous 3435T/T genotype were significantly higher than those with the 3435C/C genotype [71]. It is possible, therefore, that differential expression of P-gp could affect the intracellular concentration of Pis and thus influence their efficacy [63,72], at least in some populations [73]. Indeed, data have confirmed a correlation between initial C D 4 response, nelfinavir and efavirenz levels, and M D R - 1 polymorphisms in adherent patients [74]. A study conducted in our laboratory investigating the effects of the M D R - 1 C3435T polymorphism on initial clinical response to first combination antiretroviral therapy reported a trend to earlier virologic failure (defined as a rebound in plasma H I V R N A to >500 copies/ml) in subjects with the M D R - 1 3435C/C genotype [57], consistent with previous reports suggesting a potential effect of the M D R - 1 polymorphism on antiretroviral treatment efficacy [72]. In a small study subset, however, our study was unable to confirm that this earlier virologic failure might be attributed to the selection of resistance mutations to protease inhibitors [57], although this remains to be confirmed in other datasets. Somewhat in contrast, a recent study reported a significant trend to improved virologic response to nelfinavir-containing regimens in children with the heterozygous 3435 C/T genotype [75]. With respect to immunologic response to antiretroviral treatment, some studies have reported a significantly improved C D 4 response in individuals with the 3435T/T genotype after initiating therapy [74], although other studies have reported no effect [57,71]. Overall, the clinical importance of M D R - 1 genotyping for 35 antiretroviral therapy remains to be determined [71], but the search for polymorphisms in human genes that affect antiretroviral therapy response has only just begun [76]. The contribution of H L A genetic polymorphisms in the context of H A A R T has only begun to be investigated, as significant challenges are associated with teasing out the individual contributions of each specific H L A allele on an individual's response to antiretroviral treatment. A rather important, and clinically significant contribution of H L A genetic variation to treatment response, however, is demonstrated by the H L A - B * 5 7 allele, which has been shown to mediate up to 10-times increased risk of a very dramatic, and potentially life-threatening hypersensitivity reaction to the nucleoside reverse transcriptase inhibitor (NRTI) abacavir [77,78]. Due to the clinical implications of this reaction, small studies have begun to evaluate the clinical utility of "patch-testing" for abacavir hypersensitivity, (in which a small amount of the drug is placed on the skin to evaluate whether a contact reaction occurs) [79], and it is likely that clinical genetic screening programs for H L A - B * 5 7 wi l l soon follow [80,81]. It remains to be determined whether H L A allelic variation, or other human genetic variation for that matter, mediates hypersensitivity or intolerance to other antiretroviral agents. 2.6 Viral Genetic Factors: resistance to current and future antiretroviral agents The complete H I V genome sequence was published in 1985 [82]. Technical advances made in D N A sequencing technologies, most notably in the last decade, have greatly facilitated the large-scale sequence analysis of H I V genomes from diverse populations, and have resulted in an appreciation of the extreme plasticity and continuing diversification and evolution of this retrovirus [83]. Over the last two decades, variation in nearly every region of the H I V genome, whether coding or noncoding, has been shown to 36 influence virulence, infectivity and/or pathogenesis of H I V both in vitro and in vivo. The significance of viral genetic diversity to H I V pathogenesis remains one of the most pertinent questions in H I V research in the current era of H A A R T . Possibly the best example of viral genetic factors relevant in the H A A R T era is the emergence of antiretroviral-resistant H I V . Although combination antiretroviral therapies have significantly reduced HIV-related morbidity and mortality in the last decade [84,85], it is now recognised that drug-resistant H I V variants are a major cause of treatment failure and therefore a barrier to long-term antiretroviral efficacy in both antiretroviral-nai've and experienced patients [86,87,88,89]. In fact, it is estimated that virologic failure occurs in up to 20% of previously antiretroviral-naive individuals [90] and up to 30-50% of previously-treated individuals within two years of initiation of therapy [90,91]. In addition, a survey of drug resistance in the United States undertaken in the late 1990s estimated that 63% of patients receiving H I V care had plasma viremia of > 500 copies/ml, 76% of whom harbored H I V resistant to one or more antiretroviral agents [92]. A survey of antiretroviral resistance among treated individuals in England with plasma viral loads >2000 copies/ml reported resistance prevalences of 69%, 88% and 55%, respectively, over the years 1998-2000 [93]. Because of the implications of HIV drug resistance on the long-term efficacy of H A A R T , it is important to identify factors which may influence the emergence of drug resistance in clinical practise. The identification of these factors may yield interventions to minimize or at least delay the development of resistance. Our laboratory recently undertook a large-scale, retrospective, population-based analysis of >1000 HIV-infected, previously antiretroviral-nai've individuals initiating their first triple combination antiretroviral therapy between the dates of August 1996-September 1999 [94]. In this population, the incidence of drug-resistance 37 mutations was documented over a 2.5-year period following initiation of therapy, and risk factors for the development of drug resistance mutations were identified. Over the first 30 months after initiation of treatment, 25% of the study population developed antiretroviral resistance mutations, and the strongest determinants of development of resistance were high baseline viral load and substantial but incomplete levels of adherence [94]. These results are in agreement with a study of antiretroviral resistance in the United States, which reported that higher plasma H I V viral load and low C D 4 cell count were significant predictors of the selection of drug resistance mutations [92]. In the U S study [92], previous antiretroviral experience and advanced H I V disease were also associated with the presence of drug resistance. Although resistance to multiple antiretroviral classes has indeed emerged as a major challenge in the treatment of HIV-infected individuals [86,87,88,89], the long-term effects of multidrug resistance remain to be determined. As H I V therapy is a lifelong commitment, a realistic consequence of multidrug resistance is the exhaustion of available treatment options through repeated failure of antiretroviral regimens. Although the ten years that have elapsed since the introduction of H A A R T are not likely to provide sufficient time to conclusively address this issue, a recent study of HIV-infected individuals in British Columbia investigating the potential impact of antiretroviral resistance on mortality suggests that multidrug resistance is not yet a strong net contributor to HTV-related deaths in this province, at least on a population basis [95]. Whether this continues to be the case in the future remains to be determined. However, the development of novel antiretroviral agents wil l continually expand H I V treatment options, even for individuals who have developed resistance to the currently available antiretroviral agents. 38 2.7 HIV envelope and co-receptor phenotype The H I V envelope, encoding the gp41 (transmembrane) and gpl20 (surface) proteins, is believed to be the most genetically variable region in the viral genome and thus contributes significantly to H I V pathogenesis. The H I V envelope's capacity for mutation, for example, facilitates the virus' continual evasion of the humoral (antibody-mediated) immune response [for reviews, see 96,97], possibly contributing to H I V disease progression. In addition, genetic variation in the H I V envelope also defines both H I V biological phenotype as well as target cell tropism as determined by coreceptor usage. A s already outlined, H I V primarily uses host chemokine co-receptors C C R 5 and C X C R 4 , in combination with C D 4 , in order to attach, fuse and gain entry into host cells [14,15,16]. The ability of H I V to use one co-receptor over another lies in the genetics of the H I V envelope [98,99]. Coreceptor usage appears to play an important role in H I V pathogenesis. In general, most H I V variants isolated from newly-infected individuals utilize C C R 5 in combination with C D 4 to gain entry to host cells [100]. These CCR5-using H I V strains (termed "R5" variants [101]) predominantly infect activated C D 4 + T-lymphocytes as well as macrophages. While R5 variants are generally detectable over the entire course of H I V infection [102], variants able to utilize C X C R 4 ("X4" variants [101]) emerge in approximately 40-50% of infected persons over the course of disease [103], a phenomenon referred to as R5-to-X4 "phenotype switch" [104]. These X 4 variants predominantly target naive and resting C D 4 + T-lymphocytes and display different biological properties than their R5 counterparts, including increased replication rate, pathogenicity and syncytium-inducing capacity in M T - 2 T-cell culture [102,105]. In addition, dual-tropic variants capable of using both C X C R 4 and C C R 5 may also arise over the course of disease [106]. 39 Although the factors mediating the R5-to-X4 "phenotype switch" are incompletely understood, the consequences of this switch may be quite severe with respect to clinical prognosis. The emergence of CXCR4-us ing variants is associated with a rapid decline in C D 4 + T-lymphocyte counts, accelerated disease progression and reduced survival time in untreated individuals [107,108,109,110], as well as poorer response to treatment in the pre-H A A R T era [111,112,113,114,115,116]. The mechanisms whereby X 4 viruses contribute to accelerated disease progression and poorer prognosis, however, remain incompletely understood. In fact, the association may not be directly causative; it is not known whether X 4 variants are inherently more pathogenic and are directly responsible for rapid disease progression, or whether CXCR4-us ing variants may emerge as a consequence of progressive immune dysfunction [117]. Regardless of the direction of causation, the association of X 4 variants with poorer prognosis and inferior therapy response is an issue immediately relevant to the clinical management of H I V infection, and remains so in the era of H A A R T . Currently, commercial assays exist to experimentally determine H I V coreceptor phenotype in clinical isolates; however these assays are time-consuming and costly and therefore not generally used in clinical practise. However, H I V envelope genotyping, targeted at env regions responsible for coreceptor determination, is much less resource-intensive and may prove to be of clinical utility. Briefly, the capacity of the virus to switch to C X C R 4 usage lies at least partially in a change in two key amino acids in the third hypervariable region, known as the V 3 loop, of H I V gpl20 [98,99], although genetic variation at other gpl20 sites also play a role [118]. In general, X 4 variants display positively charged amino acids (namely lysine or arginine) at codons 11 and/or 25 of V 3 , while R5 variants do not [119,120]. V 3 amino acid analysis has been used to distinguish X 4 40 from R5 isolates in an oversimplified approach termed the "11/25" rule [121]. O f the available methods for sequence-based predictions of viral phenotype, the 11/25 rule has proven to be the most reliable [120], although it is by no means perfect. Studies of cloned isolates report specificities of >85% for the prediction of H I V phenotype from V 3 loop genotype [119,122], however studies employing direct ('bulk') P C R and sequencing approaches report much lower sensitivities and specificities for prediction of H I V coreceptor usage [60]. 2.8 Relevance of HIV envelope variation in the era of HAART Whether H I V envelope genotyping and/or H I V coreceptor phenotyping may be relevant as independent predictors of H I V clinical prognosis and therapy response in the current era of H A A R T remains to be determined. Two recent studies from our laboratory have attempted to address this issue [60,123]. B y analyzing clinical outcomes from a population-based cohort of >1000 HIV-infected, antiretroviral-nai've individuals initiating first H A A R T , our group attempted to establish the impact of baseline (pre-therapy) V 3 genotype and baseline H I V coreceptor usage on initial clinical response and survival following initiation of H A A R T [60,123]. Consistent with studies reporting an association of X 4 H I V and disease progression in untreated infection, we observed a dramatic correlation between the presence of X4-using variants and poorer clinical prognosis at the time of H A A R T initiation: subjects harboring X4-using variants had significantly lower C D 4 counts and higher plasma viral loads than subjects harboring exclusively R5 variants [60]. In addition we observed that subjects harboring X 4 variants at baseline had significantly poorer survival following initiation of H A A R T than subjects harboring exclusively R5 variants [60], an observation that is not surprising considering that pre-treatment p V L and especially C D 4 41 count remain the strongest predictors of response to H A A R T [124,125]. After adjustment for pre-treatment C D 4 and viral load, however, the presence of X 4 variants did not remain an independent predictor of survival following initiation of H A A R T [60], suggesting that a costly co-receptor determination assay may not provide any additional information above and beyond pre-treatment C D 4 and viral load, at least in terms of response to H A A R T [60]. The potential utility of coreceptor phenotyping in other contexts, for example treatment with novel antiretroviral agents such as coreceptor inhibitors [126], remains to be determined. O f interest, a different result was observed when baseline V 3 genotypes were evaluated with respect to H A A R T response [123]. In contrast to the coreceptor phenotypes, V 3 genotypes were not associated with pre-treatment clinical parameters, but were independently predictive of clinical outcomes following initiation of H A A R T [123]. The presence of positively-charged amino acids at codons 11 and/or 25 of the H I V V 3 loop (the "11/25" genotype) was independently predictive of poorer C D 4 response as well as mortality in the 5 years following therapy initiation, suggesting that an H I V envelope genotype may provide useful supplementary information, in addition to C D 4 count and viral load, in the clinical management of HIV-infected individuals [123]. A t present, the epidemiology of H I V coreceptor usage and envelope sequence variation is of particular relevance in light of the development of H I V co-receptor antagonists, a new class of antiretroviral agents designed to prevent H I V entry into target cells through blockade of C C R 5 and/or C X C R 4 coreceptors [117,127]. In anticipation of the approval of this new class of antiretrovirals, there is increasing interest in determining the prevalence, risk factors and effect on outcomes associated with coreceptor usage in population-based studies of both therapy naive and experienced patients [60,126]. H I V 42 coreceptor usage and envelope variation wi l l therefore continue to be of relevance in the H A A R T era and beyond. 2.9 Relevance of variation in other regions of the HIV genome in the era of HAART H I V genetic variation in regions other than those described here have also been implicated in playing a potential role in H I V disease progression, response to therapy and/or the development of antiretroviral resistance, either directly or indirectly. The HIV-1 nef gene, for example, located at the 3' end of the HIV-1 genome, encodes a 25 kDa N-myristolated protein which is a major determinant of H I V pathogenesis in untreated infection [128]. Correlations between infection with nef-deleted or ne/-defective H I V strains and dramatically decreased rates of disease progression have been reported [129,130,131], and sequence variations in nef have been shown to correlate with plasma viral load (pVL) , C D 4 counts and disease stage [132]. More recently, sequence variation in the accessory gene vpr has also been associated with H I V disease progression in untreated infection [133,134]. Finally, amino acid insertion mutations in H I V p 6 G a g have been associated with increased viral infectivity in vitro, and may also be associated with resistance to NRTIs [135], and thus influence response to therapy [136]. The relevance of these H I V genetic differences in the H A A R T era remains to be determined. In conclusion, there is abundant evidence to suggest that host and viral genetic factors wi l l continue to be of relevance in the current era of antiretroviral treatment. A s more information is made available on the clinical implications of both host and viral genetic variation in the current treatment era, and innovations in D N A sequencing technologies continue to be made [137,138], it is likely that the future of H I V clinical management wi l l feature the incorporation of novel host and viral genotyping assays. 43 2.10 R E F E R E N C E S 1. Hogan C M , Hammer S M . Host determinants in H I V infection and disease. Part 1: cellular and humoral immune responses. A n n Intern Med . 2001;134:761-76. 2. Time from HIV-1 seroconversion to A I D S and death before widespread use of highly-active antiretroviral therapy: a collaborative re-analysis. Collaborative Group on A I D S Incubation and H I V Survival including the C A S C A D E E U Concerted Action. Concerted Act ion on Seroconversion to A I D S and Death in Europe. Lancet. 2000; 355:1131-7. 3. Mazzo l i S, Trabattoni D , L o Caputo S, et al. HIV-specific mucosal and cellular immunity in HIV-seronegative partners of HIV-seropositive individuals. Nat Med. 1997;3:1250-7. 4. Bernard N F , Yannakis C M , Lee JS, Tsoukas C M . Human immunodeficiency virus (HlV)-specific cytotoxic T lymphocyte activity in HIV-exposed seronegative persons. J Infect Dis 1999;179(3):538-47. 5. Fowke K R , Nagelkerke N J , Kimani J, et al. Resistance to HIV-1 infection among persistently seronegative prostitutes in Nairobi, Kenya. Lancet 1996;348:1347-51. 6. L i u R, Paxton W A , Choe S, et al. Homozygous defect in HIV-1 coreceptor accounts for resistance of some multiple-exposed individuals to HIV-1 infection. Ce l l 1996, 86:367-377. 7. O'Brien SJ, Nelson G W . Human genes that limit A I D S . Nat Genet. 2004; 36:565-74. 8. Hogan C M , Hammer S M . Host determinants in H I V infection and disease. Part 2: genetic factors and implications for antiretroviral therapeutics. A n n Intern Med. 2001;134:978-96. 44 9. M a l i m M H , Emerman M . HIV-1 sequence variation: drift, shift, and attenuation. Ce l l . 2001 Feb 23;104(4):469-72. 10. Nolan D , Gaudieri S, John M , Mal la l S. Impact of host genetics on H I V disease progression and treatment: new conflicts on an ancient battleground. A I D S . 2004 Jun 18;18(9): 1231-40. 11. Kaslow R A , Dorak T, Tang JJ. Influence of host genetic variation on susceptibility to H I V type 1 infection. J Infect Dis 2005; 191 Suppl l:S68-77. 12. Rowland-Jones S, Pinheiro S, Kaul R. New insights into host factors in HIV-1 pathogenesis. Ce l l . 2001 Feb 23;104(4):473-6. 13. Tang J, Kaslow R A . The impact of host genetics on H I V infection and disease progression in the era of highly active antiretroviral therapy. A I D S . 2003; 17 Suppl 4:S51-60. 14. Feng Y , Broder C C , Kennedy P E , Berger E A . HIV-1 entry cofactor: functional c D N A cloning of a seven-transmembrane, G protein-coupled receptor. Science 1996; 272:872-7. 15. Samson M , Labbe O, Mollereau C, Vassart G , Parmentier M . Molecular cloning and functional expression of a new human CC-chemokine receptor gene. Biochemistry 1996, 35:3362-3367. 16. Deng H K , L i u R, Ellmeier W , Choe S, et al. Identification of a major co-receptor for primary isolates of H I V - 1 . Nature 1996, 381:661-666. 17. Dragic T, L i twin V , Al laway G P , et al. HIV-1 entry into CD4+ cells is mediated by the chemokine receptor C C - C K R - 5 . Nature 1996, 381:667-673. 45 18. Balotta C , Bagnarelli P, V i o l i n M , et al. Homozygous delta 32 deletion of the C C R -5 chemokine receptor gene in an HIV-1 -infected patient. A I D S 1997; 11 :F67-71. 19. B i t i R, Ffrench R, Young J, Bennetts B , Stewart G , Liang T. HIV-1 infection in an individual homozygous for the C C R 5 deletion allele. Nat M e d 1997; 3:252-3. 20. Benkirane M , J in D Y , Chun R F , Koup R A , Jeang K T . Mechanism of transdominant inhibition of CCR5-mediated HIV-1 infection by ccr5delta32. J B i o l Chem 1997, 272:30603-30606. 21. Huang Y , Paxton W A , Wolinsky S M , et al. The role of a mutant C C R 5 allele in HIV-1 transmission and disease progression. Nat M e d 1996, 11:1240-1243. 22. Dean M , Carrington M , Winkler C , et al.: Genetic restriction of HIV-1 infection and progression to A I D S by a deletion allele of the C K R 5 structural gene. Science 1996, 273:1856-1862. 23. Samson M , Libert F, Doranz, B J , et al. Resistance to HIV-1 infection in Caucasian individuals bearing mutant alleles of the C C R 5 chemokine receptor gene. Nature 1996, 382:722-725. 24. Stewart G , Ashton L , B i t i R, et al. Increased Frequency of C C R - 5 A32 heterozygotes among long-term non-progressors with HIV-1 infection. A I D S 1997, 11:1833-1838. 25. Mulherin S A , O'Brien T R , Ioannidis JP, et al. Effects of CCR5-Delta32 and C C R 2 -641 alleles on HIV-1 disease progression: the protection varies with duration of infection. A I D S 2003, 17:377-87. 26. Ioannidis JP, Rosenberg PS, Goedert JJ, et al. Effects of CCR5-Delta32, CCR2-64I, and SDF-1 3'A alleles on HIV-1 disease progression: A n international meta-analysis of individual-patient data. A n n Intern Med 2001, 135:782-95. 27. Martin M , Dean M , Smith M , et al. Genetic Acceleration of A I D S Progression by a Promoter Variant of C C R 5 . Science 1988, 282:1907-1911. 28. McDermott D H , Zimmerman P A , Guignard F, et al. C C R 5 promoter polymorphism and HIV-1 disease progression. Lancet 1998, 352:866-870. 29. Meyer L , Magierowska M , Hubert JB , et al. CC-chemokine receptor variants, SDF-1 polymorphism, and disease progression in 720 HIV-infected patients. S E R O C O Cohort. Amsterdam Cohort Studies on A I D S . A I D S . 1999;13:624-6. 30. Kostrikis L G , Huang Y , Moore JP, et al. A chemokine receptor C C R 2 allele delays HIV-1 disease progression and is associated with a C C R 5 promoter mutation. Nat Med 1998;4:350-3. 31. Michael N L , Louie L G , Rohrbaugh A L , et al. The role of C C R 5 and C C R 2 polymorphisms in HIV-1 transmission and disease progression. Nat M e d 1997; 3:1160-2. 32. Combadiere C , Salzwedel K , Smith E D , Tiffany H L , Berger E A , Murphy P M . Identification of C X 3 C R 1 . A chemotactic receptor for the human C X 3 C chemokine fractalkine and a fusion coreceptor for H I V - 1 . J Biol Chem 1998, 273: 23799-23804. 33. Faure S, Meyer L , Costagliola D , et al. Rapid Progression to A I D S in HIV+ individuals with a Structural Variant of the Chemokine Receptor C X 3 C R 1 . Science 2000,287: 2274-2277. 34. McDermott D H , Col la JS, Kleeberger C A , et al. Genetic Polymorphisms in C X 3 C R 1 and Risk of H I V Disease. Science 2000, 290: 203 l a . 35. Winkler C , M o d i W , Smith M W , et al. Genetic restriction of A I D S pathogenesis by an SDF-1 chemokine gene variant. Science. 1998; 279:389-93. 47 36. van Rij R P , Broersen S, Goudsmit J, Coutinho R A , Schuitemaker H . The role of a stromal cell-derived factor-1 chemokine gene variant in the clinical course of HIV-1 infection. A I D S 1998;12:F85-90. 37. Kle in J, Sato A . The H L A system. First of two parts. N Engl J Med. 2000; 343:702-9. 38. Kle in J, Sato A . The H L A system. Second of two parts. N Engl J Med 2000; 343:782-6. 39. Doherty P C , Zinkernagel R M . Enhanced immunological surveillance in mice heterozygous at the H-2 gene complex. Nature. 1975; 256:50-2. 40. Carrington M , Nelson G W , Martin M P , et al. H L A and H I V - 1 : heterozygote advantage and B*35-Cw*04 disadvantage. Science 1999; 283:1748-52. 41. Tang J, Costello C , Keet IP, et al. A I D S H L A class I homozygosity accelerates disease progression in human immunodeficiency virus type 1 infection. A I D S Res Hum Retroviruses 1999;15:317-24. 42. Altfeld M , Addo M M , Rosenberg ES , et al. Influence of H L A - B 5 7 on clinical presentation and viral control during acute HIV-1 infection. A I D S 2003;17:2581-91. 43. Kaslow R A , Carrington M , Apple R, et al. Influence of combinations of human major histocompatibility complex genes on the course of HIV-1 infection. Nat Med 1996;2:405-11. 44. Dorak M T , Tang J, Tang S, et al. Influence of human leukocyte antigen-B22 alleles on the course of human immunodeficiency virus type 1 infection in 3 cohorts of white men. J Infect Dis. 2003; 188:856-63. 48 45. Hendel H , Caillat-Zucman S, Lebuanec H , et al. New class I and II H L A alleles strongly associated with opposite patterns of progression to A I D S . J Immunol 1999;162:6942-6. 46. MacDonald K S , Fowke K R , Kimani J, et al. Influence of H L A supertypes on susceptibility and resistance to human immunodeficiency virus type 1 infection. J Infect Dis 2000; 181:1581-9. 47. Carrington M , O'Brien SJ. The influence of H L A genotype on A I D S . Annu Rev Med. 2003;54:535-51. 48. Keet IP, Tang J, Kle in M R , et al. Consistent associations of H L A class I and II and transporter gene products with progression of human immunodeficiency virus type 1 infection in homosexual men. J Infect Dis. 1999; 180:299-309. 49. Moore C B , John M , James IR, Christiansen FT , Witt C S , Mal la l S A . Evidence of HIV-1 adaptation to HLA-restricted immune responses at a population level. Science. 2002; 296:1439-43. 50. Bogner JR, Lutz B , Kle in H G , Pollerer C, Troendle U , Goebel F D . Association of highly active antiretroviral therapy failure with chemokine receptor 5 wi ld type. H I V Med 2004; 5:264-72. 51. Guerin S, Meyer L , Theodorou I, et al. C C R 5 A32 Deletion and Response to H A A R T in HIV-1-Infected Patients. S E R O C O / H E M O C O Study Group. A I D S 2000, 14:2788-2790. 52. Valdez H , Purvis S, Lederman M , Filingame M , Zimmerman P: Association of the C C R 5 A 3 2 Mutation With Improved Response to Antiretroviral Therapy. J A M A 1999, 282:734-734. 49 53. Yamashita T E , Phair JP, Munoz A , et al. Immunologic and virologic response to highly active antiretroviral therapy in the Multicenter A I D S Cohort Study. A I D S 2001; 15:735-46. 54. Bratt G , Karlsson A , Leandersson A C , Albert J, Wahren B , Sandstrom E . Treatment history and baseline viral load, but not viral tropism or C C R 5 genotype, influence prolonged antiviral efficacy of highly active antiretroviral treatment. A I D S 1998, 12:2193-2202. 55. Brumme Z L , Chan K J , Dong W , et al. CCR5Delta32 and promoter polymorphisms are not correlated with initial virological or immunological treatment response. A I D S 2001;15:2259-66. 56. Brumme Z L , Henrick B M , Brumme C J , et al. Association of C C R 5 A 3 2 mutation with clinical response and >5-year survival following initiation of first treiple antiretroviral regimen. Antiviral Therapy, in press. 57. Brumme Z L , Dong W W , Chan K J , et al. Influence of polymorphisms within the C X 3 C R 1 and M D R - 1 genes on initial antiretroviral therapy response. A I D S 2003;17:201-8. 58. Wi t F W , van Rij R P , Weverling GJ , Lange J M , Schuitemaker H . C C chemokine receptor 5 delta32 and C C chemokine receptor 2 641 polymorphisms do not influence the virologic and immunologic response to antiretroviral combination therapy in human immunodeficiency virus type 1-infected patients. J Infect Dis. 2002;186:1726-32. 50 59. O'Brien T R , McDermott D H , Ioannidis JP, et al. Effect of chemokine receptor gene polymorphisms on the response to potent antiretroviral therapy. A I D S 2000; 14:821-6. 60. Brumme Z L , Goodrich J, Mayer H B , et al. Molecular and Clinical Epidemiology of CXCR4-us ing HIV-1 in a large population of therapy-naive individuals. J Infect Dis 2005; 192:466-74. 61. Gottesman M M , Pastan I, Ambudkar V . P-glycoprotein and multidrug resistance. Curr Opin Genet Dev 1996, 6: 610-617. 62. Ambudkar S V , Dey S, Hrycyna C A , Ramachandra M , Pastan I, Gottesman M M . Biochemical, cellular, and pharmacological aspects of the multidrug transporter. Annu Rev Pharmacol Toxicol 1999, 39: 361-398. 63. Huisman M T , Smit JW, Schinkel A H . Significance of P-glycoprotein for the pharmacology and clinical use of H I V protease inhibitors. AIDS 2000, 14: 237-242. 64. Gottesman M M , Pastan I. Biochemistry of multidrug resistance mediated by the multidrug transporter. Annu Rev Biochem 1993, 62: 385-427. 65. Leveque D , Jehl F: P- glycoprotein and pharmacokinetics. Anticancer Res. 1995, 15: 331-336. 66. Sarkadi B , Mul ler M , Homolya L , et al. Interaction of bioactive hydrophobic peptides with the human multidrug transporter. F A S E B J 1994, 8: 766-770. 67. Sharom FJ , Y u X , DiDiodato G , Chu JW. Synthetic hydrophobic peptides are substrates for P-glycoprotein and stimulate drug transport. Biochem J 1996, 320: 421-428. 51 68. K i m A E , Dintaman J M , Waddell D S , Silverman J A . Saquinavir, an H I V protease inhibitor, is transported by P-glycoprotein. J Pharmacol Exp Ther 1998, 286: 1439-1455. 69. Lee C G , Gottesman M M , Cardarelli C O , et al. HIV-1 protease inhibitors are substrates for the M D R - 1 multidrug transporter. Biochemistry 1998, 37: 3594-3601. 70. Hoffmeyer S, Burk O, von Richter O, et al. Functional polymorphisms of the human multidrug-resistance gene: Multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo. P N A S 2000, 97: 3473-3478. 71. Zhu D , Taguchi-Nakamura H , Goto M , et al. Influence of single-nucleotide polymorphisms in the multidrug resistance-1 gene on the cellular export of nelfinavir and its clinical implication for highly active antiretroviral therapy. Antivir Ther 2004; 9:929-35 72. Jones K , Bray P G , Khoo S H , et al. P-Glycoprotein and transporter M R P - 1 reduce H I V protease inhibitor uptake in C D 4 cells: potential for accelerated viral drug resistance? A I D S 2001, 15: 1353-1358. 73. Schaeffeler E , Eichelbaum M , Brinkmann U , et al. Frequency of C3435T polymorphism of M D R 1 gene in African people. Lancet 2001, 358: 383-384. 74. Fellay J, Marzol ini C , Meaden E R , et al. Response to antiretroviral treatment in H I V -1 infected individuals with allelic variants of the multidrug resistance transporter 1: a pharmacogenetics study. Lancet 2002, 359:30-36. 75. Saitoh A , Singh K K , Powell C A , et al. A n MDR1-3435 variant is associated with higher plasma nelfinavir levels and more rapid virologic response in HIV-1 infected children. A I D S . 2005; 19:371-80. 52 76. Quirk E , M c L e o d H , Powderly W . The pharmacogenetics of antiretroviral therapy: a review of studies to date. C l i n Infect Dis. 2004; 39:98-106. 77. Mal la l S, Nolan D , Witt C , et al. Association between presence of H L A - B * 5 7 0 1 , H L A - D R 7 , and H L A - D Q 3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet. 2002;359:727-32. 78. Hetherington S, Hughes A R , Mosteller M , et al. Genetic variations in H L A - B region and hypersensitivity reactions to abacavir. Lancet. 2002; 359:1121-2. 79. Phillips E J , Sullivan JR, Knowles SR, Shear N H . Uti l i ty of patch testing in patients with hypersensitivity syndromes associated with abacavir. A I D S . 2002; 16:2223-5. 80. Hughes D A , Vi la r FJ , Ward C C , Alfirevic A , Park B K , Pirmohamed M . Cost-effectiveness analysis of H L A B*5701 genotyping in preventing abacavir hypersensitivity. Pharmacogenetics. 2004; 14:335-42. 81. Nolan D , Gaudieri S, Mal la l S. Pharmacogenetics: a practical role in predicting antiretroviral drug toxicity? J H I V Ther 2003; 8:36-41. 82. Ratner L , Haseltine W , Patarca R, et al. Complete nucleotide sequence of the A I D S virus, H T L V - I I I . Nature. 1985; 313:277-84. 83. M a l i m M H , Emerman M . HIV-1 sequence variation: drift, shift, and attenuation. Ce l l . 2001; 104:469-72. 84. Palella FJ Jr., Delaney K M , Moorman A C , et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J M e d 1998; 338:853-60. 85. Hogg R S , Y i p B , K u l l y C, et al. Improved survival among HIV-infected patients after initiation of triple-drug antiretroviral regimens. C M A J 1999; 160:659-65. 53 86. Wainberg M A , Friedland G . Public health implications of antiretroviral therapy and H I V drug resistance. J A M A . 1998; 279:1977-83. 87. Pillay D , Taylor S and Richman D D . Incidence and impact of resistance against approved antiretroviral drugs. Rev M e d Vi ro l . 2000; 10:231-53. 88. Hirsch M S , Brun-Vezinet F, D A q u i l a R T , et al. Antiretroviral drug resistance testing in adult HIV-1 infection: recommendations of an International A I D S Society-USA Panel. J A M A 2000; 283:2417-26. 89. Yeni P G , Hammer S M , Hirsch M S , et al. Treatment for adult H I V infection: 2004 recommendations of the International A I D S Society-USA Panel. J A M A . 2004;292:251-65. 90. Ledergerber B , Egger M , Opravil M , et al. Cl inical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Swiss H I V Cohort Study. Lancet 1999; 353:863-8. 91. Deeks S G , Hecht F M , Swanson M , et al. H I V R N A and C D 4 cell count response to protease inhibitor therapy in an urban A I D S clinic: response to both initial and salvage therapy. A I D S 1999; 13:F35-43. 92. Richman D D , Morton S C , W r i n T, et al. The prevalence of antiretroviral drug resistance in the United States. A I D S 2004; 18:1393-401. 93. Scott P, Arnold E , Evans B , et al. Surveillance of H I V antiretroviral drug resistance in treated individuals in England: 1998-2000. J Antimicrob Chemother. 2004; 53:469-73. 54 94. Harrigan PR, Hogg R S , Dong W W , et al. Predictors of H I V drug-resistance mutations in a large antiretroviral-naive cohort initiating triple antiretroviral therapy. J Infect Dis 2005; 191:339-47. 95. Recsky M A , Brumme Z L , Chan K J , et al. Antiretroviral resistance among H I V -infected persons who have died in British Columbia, in the era of modern antiretroviral therapy. J Infect Dis. 2004; 190:285-92. 96. Johnson W E , Desrosiers R C . Vira l persistance: HIV's strategies of immune system evasion. Annu Rev Med. 2002;53:499-518. 97. da Silva J. The evolutionary adaptation of HIV-1 to specific immunity. Curr H I V Res. 2003; 1:363-71. 98. De Jong JJ, De Ronde A , Keulen W , Tersmette M , Goudsmit J. Min imal requirements for the human immunodeficiency virus type 1 V 3 domain to support the syncytium-inducing phenotype: analysis by single amino acid substitution. J Virol 1992; 66:6777-80. 99. Fouchier R A , Groenink M , Kootstra N A , et al. Phenotype-associated sequence variation in the third variable domain of the human immunodeficiency virus type 1 gpl20 molecule. J Virol 1992; 66:3183-7. 100. van't Wout A B , Kootstra N A , Mulder-Kampinga G A , et al. Macrophage-tropic variants initiate human immunodeficiency virus type 1 infection after sexual, parenteral, and vertical transmission. J C l i n Invest. 1994; 94(5):2060-7. 101. Berger E A , Doms R W , Fenyo E M , et al. A new classification for H I V - 1 . Nature. 1998 Jan ; 391:240. 55 102. Schuitemaker H , Kootstra N A , de Goede R E , de W o l f F , Miedema F, Tersmette M . Monocytotropic human immunodeficiency virus type 1 (HIV-1) variants detectable in all stages of HIV-1 infection lack T-cell line tropism and syncytium-inducing ability in primary T-cell culture. J V i ro l 1991; 65: 356-63. 103. Berger E A , Murphy P M , Farber J M . Chemokine receptors as HIV-1 coreceptors: roles in viral entry, tropism, and disease. Annu Rev Immunol 1999; 17: 657-700. 104. Regoes R R , Bonhoeffer S. The H I V coreceptor switch: a population dynamical perspective. Trends Microbiol . 2005; 13:269-77. 105. Connor RI, Ho D D . Human immunodeficiency virus type 1 variants with increased replicative capacity develop during the asymptomatic stage before disease progression. J V i r o l 1994; 68: 4400-08. 106. Glushakova S, Y i Y , Grivel JC , et al. Preferential coreceptor utilization and cytopathicity by dual-tropic HIV-1 in human lymphoid tissue ex vivo. J C l in Invest 1999; 104: R7-11. 107. Tersmette M , Gruters R A , de W o l f F, et al. Evidence for a role of virulent human immunodeficiency virus (HIV) variants in the pathogenesis of acquired immunodeficiency syndrome: studies on sequential H I V isolates. J V i r o l 1989; 63: 2118-25. 108. Schuitemaker H , Koot M , Kootstra N A , et al. Biological phenotype of human immunodeficiency virus type 1 clones at different stages of infection: progression of disease is associated with a shift from monocytotropic to T-cell tropic virus populations. J V i r o l 1992; 66:1354-60. 56 109. Koot M , Keet IP, Vos A H . Prognostic value of HIV-1 syncytium-inducing phenotype for rate of CD4+ cell depletion and progression to A I D S . A n n Int M e d 1993; 118: 681-88. 110. Richman D D , Bozzette S A . The impact of the syncytium-inducing phenotype of human immunodeficiency virus on disease progression. J Infect Dis 1994; 169: 968-74. 111. Rusconi S, De Pasquale M P , Mainini F, et al. Vi ra l load, viral phenotype modification, zidovudine susceptibility and reverse transcriptase mutations during the first 6 months of zidovudine monotherapy in HIV-1-infected people. Antivir Ther 1996; 1:211-19. 112. Katzenstein T L , Nielsen C , Bruun L , Mathiesen L R , Pedersen C , Gerstoft J. Quantification of HIV-1 R N A during antiretroviral therapy: association with viral phenotype and development of resistance. Antivir Ther 1996; 1: 246-54. 113. Vida l C , Garcia F , Gatell J M , et al. Predictive factors influencing peak viral load drop in response to nucleoside reverse transcriptase inhibitors in antiretroviral-naive H I V -1-infected patients. J Acquir Immune Defic Syndr Hum Retrovirol 1998; 19: 55-60. 114. Davey R T Jr, Chaitt D G , Reed G F , et al. Randomized, controlled phase I/II, trial of combination therapy with delavirdine (U-90152S) and conventional nucleosides in human immunodeficiency virus type 1-infected patients. Antimicrob Agents Chemother 1996; 40:1657-64. 115. Hughes M D , Johnson V A , Hirsch M S , et al. Monitoring plasma HIV-1 R N A levels in addition to CD4+ lymphocyte count improves assessment of antiretroviral therapeutic 57 response. A C T G 241 Protocol Virology Substudy Team. A n n Intern M e d 1997; 126: 929-38. 116. Karlsson A , Parsmyr K , Aperia K , Sandstrom E , Fenyo E M , Albert J. M T - 2 cell tropism of human immunodeficiency virus type 1 isolates as a marker for response to treatment and development of drug resistance. J Infect Dis . 1994; 170:1367-75. 117. Shaheen F, Collman R G . Co-receptor antagonists as HIV-1 entry inhibitors. Curr Opin Infect Dis 2004; 17: 7-16. 118. Jansson M , Backstrom E , Scarlatti G , et al. Length variation of glycoprotein 120 V 2 region in relation to biological phenotypes and coreceptor usage of primary H I V type 1 isolates. A I D S Res Hum Retroviruses. 2001; 17:1405-14. 119. Fouchier R A , Brouwer M , Broersen S M , Schuitemaker H . Simple determination of Human Immunodeficiency Virus Type 1 Syncytium-Inducing V 3 Genotype by P C R . J Clin Micb 1995; 33:906-11. 120. Hoffman N G , Seillier-Moiseiwitsch F, A h n J, Walker J M , Swanstrom R. Variability in the human immunodeficiency virus type 1 gpl20 Env protein linked to phenotype-associated changes in the V 3 loop. J V i ro l 2002; 76:3852-64. 121. Resch W , Hoffman N , Swanstrom R. Improved success of phenotype prediction of the human immunodeficiency virus type 1 from envelope variable loop 3 sequence using neural networks. Virology 2001; 288:51-62. 122. McDonald R A , Chang G , Michael N L . Relationship between V 3 genotype, biologic phenotype, tropism, and coreceptor use for primary isolates of human immunodeficiency virus type 1. J Hum V i r o l . 2001; 4:179-87. 58 123. Brumme Z L , Dong W W , Y i p B , et al. Clinical and immunological impact of H I V envelope V 3 sequence variation after starting initial triple antiretroviral therapy. A I D S 2004; 18:Fl-9. 124. Hogg R S , Y i p B , Chan K J , et al. Rates of disease progression by baseline C D 4 cell count and viral load after initiating triple-drug therapy. J A M A 2001; 286:2568-77. 125. Egger M , M a y M , Chene G , et al. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet 2002; 360:119-29. 126. Moyle G J , Wildfire A , Mandalia S, et al. Epidemiology and predictive factors for chemokine receptor use in HIV-1 infection. J Infect Dis . 2005;191:866-72. 127. Moore JP, Kitchen S G , Pugach P, Zack J A . The C C R 5 and C X C R 4 coreceptors-central to understanding the transmission and pathogenesis of human immunodeficiency virus type 1 infection. A I D S Res Hum Retroviruses 2004; 20:111-26. 128. Cullen B R . HIV-1 auxiliary proteins: making connections in a dying cell. (1998) Cell, 93:685-92. 129. Kirchhoff F, Greenough T C , Brettler D B , Sullivan J L , Desrosiers R C . Brief report: absence of intact nef sequences in a long-term survivor with nonprogressive HIV-1 infection. N Engl J M e d 1995; 332:228-32. 130. Mariani R, Kirchhoff F, Greenough T C , Sullivan J L , Desrosiers R C , Skowronski J. High frequency of defective nef alleles in a long-term survivor with nonprogressive human immunodeficiency virus type 1 infection. J V i r o l 1996; 70:7752-64. 59 131. Rhodes D l , Ashton L , Solomon A , et al. Characterization of three nef-defective human immunodeficiency virus type 1 strains associated with long-term nonprogression. J V i r o l 2000, 74:10581-88. 132. Kirchhoff F, Easterbrook PJ , Douglas N , et al. Sequence variations in human immunodeficiency virus type 1 Nef are associated with different stages of disease. J V i ro l 1999; 73:5497-508. 133. Wang B , Ge Y C , Palasanthiran P, et al. Gene defects clustered at the C-terminus of the vpr gene of HIV-1 in long-term nonprogressing mother and child pair: in vivo evolution of vpr quasispecies in blood and plasma. Virology 1996; 223:224-232. 134. L u m JJ, Cohen OJ , Nie Z , et al. Vpr R77Q is associated with long-term nonprogressive H I V infection and impaired induction of apoptosis. J C l i n Invest 2003;111:1547-1554. 135. Peters S, Munoz M , Yer ly S, et al. Resistance to nucleoside analog reverse transcriptase inhibitors mediated by human immunodeficiency virus type 1 p6 protein. J V i r o l 2001; 75:9644-9653. 136. Kaufmann G R , Suzuki K , Cunningham P, et al. Impact of H I V type 1 protease, reverse transcriptase, cleavage site, and p6 mutations on the virological response to quadruple therapy with saquinavir, ritonavir, and two nucleoside analogs. A I D S Res HumRetrov 2001; 17:487-497. 137. Kl ing , J. Ultrafast D N A sequencing. Nature Biotechnology 2003; 21:1425 - 1427 138. Simons JF, Egholm M , Lanza J, et al. Ultra-deep sequencing of H I V from drug resistant patients. Abstract 142, X I V International H I V Drug Resistance Workshop, June 7-11 2005, Quebec City, Canada. Antiviral Therapy 2005; 10:S157 60 CHAPTER 3 HUMAN GENETIC VARIATION: IMPLICATIONS OF THE CCR5A32 MUTATION ON RESPONSE TO HAART 1 3.1 I N T R O D U C T I O N The C C R 5 chemokine receptor is a key cell-surface co-receptor for HIV-1 [1,2]. A naturally-occurring 32-base pair deletion in C C R 5 confers protective effects with respect to H I V infection and disease progression in untreated individuals [3,4], and in some studies, improved response to antiretroviral therapy [5,6,7] although this remains controversial [8,9,10]. Our laboratory previously investigated the effect of the C C R 5 A 3 2 genotype on initial virologic and immunologic response to first antiretroviral treatment in a cohort of 436 subjects initiating dual (20%) and triple (80%) therapy [9]. We observed no significant influence of C C R 5 A 3 2 genotype on initial virologic or immunologic response over a two-year follow-up [9]. Here we investigate the influence of C C R 5 A 3 2 on first virologic and immunologic response to triple therapy in a cohort of >1000 antiretroviral naive adults initiating their first triple therapy in British Columbia, Canada, over a median >5-year follow-up period. In addition we investigate the association between C C R 5 A 3 2 and survival following initiation of treatment, which was not possible in the previous study [9] due to shorter duration. 1 This chapter has been published as: Brumme Z L , Henrick B M , Brumme CJ, Hogg RSH, Montaner JSG, Harrigan PR. Short communication. "Association of CCR5A32 mutation with clinical response and >5-year survival following initiation of first triple antiretroviral regimen " Antivir Ther 2005; 10:849-53. 61 3.2 M A T E R I A L S A N D M E T H O D S The B. C. Centre for Excellence in HIV/AIDS Drug Treatment Program In the province of British Columbia, Canada, antiretrovirals are distributed free of charge to HIV-infected individuals through a centralized drug treatment program based at the B . C . Centre for Excellence in H I V / A I D S (the Centre). Antiretrovirals are prescribed according to specific guidelines set by the B . C . Therapeutic Guidelines Committee, which are revised regularly and are in accordance with international guidelines [11]. Patients who enroll in the program may provide informed consent and participate in a survey that collects sociodemographic data. Routine clinical monitoring of patients takes place at approximately 3-month intervals at which time plasma H I V viral load testing (Roche Amplicor Monitor Assay) and C D 4 counts are performed. These data are stored in the Centre's Drug Treatment Program database. The HOMER cohort The H A A R T Observational Medical Evaluation and Research ( H O M E R ) cohort includes all HIV-positive, antiretroviral naive adults who started triple antiretroviral therapy (consisting of two nucleoside reverse transcriptase inhibitors [nRTIs] and either a protease inhibitor [PI] or a non-nucleoside reverse transcriptase inhibitor [NNRTI]) through the B . C . Drug Treatment Program between August 1996 and September 1999 (N=1188). This cohort has been the focus of a number of population-based studies and has been described in detail previously [11,12]. In the present study, H O M E R cohort subjects were followed over a median of 5.8 years following therapy initiation. 62 CCR5 genotyping PCR primers flanking CCR5A32 were used to amplify this region from nucleic acid material extracted from a single blood or plasma sample from each subject, as described in [9]. Amplified products were visualised by agarose gel electrophoresis and CCR5A32 genotypes confirmed by automated DNA sequencing on an ABI 3700 DNA sequencer . Note that there is some overlap between the current and previous [9] cohorts (N=322, 27.0%). Statistical Analysis Associations between CCR5A32 and baseline sociodemographic and clinical parameters (age, gender, prior AIDS diagnosis, history of injection drug use, pVL and CD4 count) were determined using the Chi-Squared test (binary variables) or the Wilcoxon Rank-sum test (continuous variables). Associations between CCR5A32 and treatment outcomes were assessed using Kaplan-Meier survival analysis and Cox proportional hazards regression. The primary endpoint investigated was time to non-accidental death (occurring on or before September 30, 2004, the date of last linkage with the British Columbia Vital Statistics Agency). Deaths were classified according to the International Classification of Diseases, Tenth revision (ICD-10). Additional surrogate clinical endpoints included time to first virologic suppression (time to the first of 2 consecutive pVL <400 copies/ml from date of therapy initiation), time to subsequent virologic rebound (time elapsed between pVL suppression and to the first of two pVL > 400 copies/ml), and time to CD4 decline (time to the first of two CD4 counts below baseline), consistent with definitions in [9]. Event-free subjects were censored on the draw date of the last pVL or CD4 sample up to and including September 30, 2004. Subjects were followed for a median of 5.8 years following therapy 63 initiation. Ethical approval was granted by the institutional Research Ethics Board (REB) . Individual consent from subjects was not obtained as samples were analyzed in an anonymized manner consistent with the current R E B guidelines. 64 3.3 R E S U L T S Association of CCR5A32 with baseline parameters C C R 5 A 3 2 genotyping was successful for 1174 of 1188 subjects (98.8%). The heterozygous CCR5wt /A32 genotype was observed at a prevalence of 14.6% (n=171 of 1174). No CCR5A32/A32 individuals were observed. A t baseline, subjects with and without the C C R 5 A 3 2 mutation were comparable with respect to sociodemographic and clinical parameters including gender (overall cohort 84% male), history of injection drug use (30%), and proportion previously diagnosed with an AIDS-defining illness (13%). However, CCR5wt /A32 subjects were on average slightly older at the time of therapy initiation when compared to CCR5wt/wt subjects (median 38 vs. 37 years p=0.04), had significantly higher baseline C D 4 counts when compared to CCR5wt/wt subjects (CD4 median 320 cells/mm 3 ; Interquartile range [IQR 140-470] vs. 270 cells/mm 3 [IQR 130-420], p=0.04) and slightly lower baseline plasma viral loads when compared CCR5wt/wt subjects ( p V L median 110,000 copies H I V R N A / m l [IQR 35,000-240,000] vs. 130,000 copies H I V R N A / m l [IQR 43,000-320,000], p=0.06). Association of CCR5A32 with shorter-term clinical surrogate outcomes In multivariate analyses adjusting for baseline parameters as well as time spent on antiretrovirals in the first year of therapy (estimated by dividing the number of months of prescriptions dispensed by months of follow-up in the first year, a surrogate of adherence [13]), we observed a significant association between the CCR5wt /A32 genotype and a shorter time to initial virologic suppression <400 copies H I V R N A / m l (multivariate H R [95% CI] 1.20 [1.01-1.44], p=0.044). The percentages of CCR5wtAvt and CCR5A32/wt subjects who achieved initial virologic suppression were 76% and 84%, respectively (p=0.02, chi-squared 65 test). Among subjects who achieved virologic suppression, the median time to suppression was 10.8 months for the C C R 5 wt/wt group vs. 9.6 months for the CCR5A32/wt group. However, no significant difference was observed between CCR5wt /A32 and CCR5wt/wt genotype groups and subsequent time to first virologic rebound >400 copies H I V R N A / m l after initial suppression (multivariate p>0.1). Similarly, no significant association was observed between CCR5wt /A32 and CCR5wt/wt genotype groups and immunologic H A A R T outcomes, as measured by the time to the first decline of C D 4 count below baseline (multivariate p>0.1). Association of CCR5A32 with survival following initiation of triple therapy A total of 179 non-accidental deaths were recorded over study follow-up; 16 (9.4%) among CCR5wt /A32 and 163 (16.3%) among CCR5wt/wt subjects. O f the non-accidental deaths among CCR5wt /A32 and CCR5wt/wt subjects, 75% and 72%, respectively, were coded as directly H I V and/or AIDS-related. In univariate analyses we observed improved survival in the (median) >5-year period following therapy initiation among CCR5wt/A32 subjects (univariate Hazard Ratio [HR] 0.56, 95% Confidence Interval [CI] 0.34-0.94, p=0.03) (Figure and Table). This trend remained significant after adjustment for baseline age, C D 4 , p V L , and presence of AIDS-defining illnesses (p=0.03, data not shown). However, in a multivariate model additionally controlling for time spent on antiretrovirals in the first year of therapy [13], a surrogate of adherence, the association between CCR5wt/A32 and improved survival did not remain statistically significant (multivariate H R [95% CI] 0.64 [0.38- 1.07], p= 0.09) (Table). In addition an analysis controlling for adherence over the entire course of study follow-up yielded comparable results (data not shown). Finally, a 66 survival analysis taking into consideration deaths from all causes produced results entirely consistent with those of the original analysis (data not shown). 67 3.4 D I S C U S S I O N A N D C O N C L U S I O N Studies investigating whether the C C R 5 A 3 2 mutation confers any protective effect during treatment with H A A R T have reported somewhat conflicting results. Some studies have reported a significant association between the heterozygous CCR5wt /A32 genotype and an increased likelihood of plasma virus suppression after treatment with H A A R T [6,7,14], improved short-term C D 4 responses [6,15], as well as improved 6-and 12-month virologic responses to H A A R T in subjects with advanced disease [5]. In contrast, other studies have reported no significant correlation between C C R 5 A 3 2 genotype and response to treatment [8,9,10]. In most cases, however, these studies were limited to the evaluation of the effect of C C R 5 A 3 2 on shorter-term (<2-year) clinical outcomes. The present study evaluated the effect of the C C R 5 A 3 2 mutation on survival and treatment outcomes in a cohort of >1000 antiretroviral naive individuals initiating their first H A A R T , over a median >5-year follow-up. After adjustment for baseline sociodemographic and clinical factors known to affect therapy response, individuals heterozygous for the C C R 5 A 3 2 deletion experienced significantly more rapid initial suppression of plasma H I V R N A below 400 copies/ml. However consistent with our previous findings [9], no significant association between C C R 5 A 3 2 and the subsequent duration of initial virologic response was observed. Similarly we observed no association between C C R 5 A 3 2 and initial immunologic response, as measured by the time to a decline of C D 4 count to below baseline. Over the longer-term (median >5-year) follow-up, a trend to improved survival in CCR5wt/A32 subjects was observed; however, the magnitude of difference was relatively small and most importantly, the association did not remain significant after controlling for adherence. Among the parameters investigated, the strongest predictors of survival over the 68 study follow-up period were age at therapy initiation, baseline C D 4 count and adherence (see Table). As it has been firmly established that the CCR5wt/A32 allele does indeed confer a protective effect with respect to H I V disease progression in the context of untreated infection [3,4,16,17], the survival advantage observed among CCR5wt/A32 subjects in this study may possibly be explained by a "natural history effect" in this cohort, which is minimized after controlling for adherence. These results emphasize the need to include data on adherence in studies evaluating the effects of potential prognostic markers on treatment response, and may partially explain the contrasting reports on the relevance of C C R 5 A 3 2 to treatment outcomes thus far. These results indicate that C C R 5 A 3 2 is likely not a clinically significant predictor of longer-term clinical responses or survival in the context of H A A R T . Results therefore suggest that C C R 5 A 3 2 is likely not relevant from a clinical management perspective and results do not support the utility of the C C R 5 A 3 2 genotype as an independent clinical prognostic marker for response to H A A R T . However, the potential relevance of CCR5A32 genotypes evaluated in combination with other host genetic factors [18,19] as well as the relevance of C C R 5 A 3 2 genotypes in context of future antiretroviral agents, (including co-receptor antagonists) to therapeutic outcomes, remains to be determined 69 Table 3.1: Influence of CCR5A32 genotype on time to non-accidental death after initiation of HAART (median >5-year follow-up) Risk Factor Univariate Multivariate Hazard Ratio 95% CI P Hazard Ratio 95% CI P CCR5 wt/A32 genotype 0.56 0.34-0.94 0.03 0.64 0.38-1.07 0.09 male gender 0.93 0.62-1.38 0.7 - - -age (per 10 year increment) 1.46 1.26-1.69 <0.0001 1.50 1.29-1.76 <0.0001 baseline HIV RNA (per log, 0 increment) 1.67 1.29-2.16 <0.0001 1.37 1.06-1.76 0.02 baseline CD4 cell count (per 100 cell decrement) 1.23 1.14-1.34 <0.0001 1.19 1.09-1.30 0.0002 baseline AIDS diagnosis 1.68 1.17-2.43 0.005 1.29 0.86-1.92 0.2 History of injection drug use 1.08 0.79-1.47 0.6 - - -PI vs. NNRTI -containing initial regimen 1.38 0.96-2.00 0.09 - - -95% adherent (yes vs. no)* 0.38 0.28-0.52 <0.0001 0.31 0.22-0.42 <0.0001 * Adherence was estimated by determining the proportion of time spent on antiretrovirals in the first year of therapy. This parameter was calculated by dividing the number of months of antiretroviral prescriptions dispensed by the number of months of follow-up in the first year [13] 70 Figure 3.1: Kaplan-Meier analysis of survival following initiation of HAART, by CCR5 genotype CCR5wt/A32 subjects are represented by circle (o) symbols while CCR5wt/wt individuals are represented by cross (+) symbols. Symbols indicate time points at which events occur (in this case, only nonaccidental deaths are considered events. Note that accidental deaths are not considered events and subjects censored on date of death). In this Kaplan-Meier analysis, a significant association between CCR5wt /A32 genotype and improved survival after initiating H A A R T was observed (p<0.05). However it is important to note that this association did not remain statistically significant after controlling for the effects of adherence (see Table 3.1). 71 3.5 R E F E R E N C E S 1. Samson M , Labbe O, Mollereau C, Vassart G , Parmentier M . Molecular cloning and functional expression of a new human CC-chemokine receptor gene. Biochemistry 1996; 35: 3362-3367. 2. Deng H , L i u R, Ellmeier W , et al. Identification of a major co-receptor for primary isolates of H I V - 1 . Nature 1996; 381: 661-666. 3. Huang Y , Paxton W A , Wolinsky S M , et al. The role of a mutant C C R 5 allele in HIV-1 transmission and disease progression. Nat M e d 1996; 11: 1240-1243. 4. Dean M , Carrington M , Winkler C , et al. Genetic restriction of HIV-1 infection and progression to A I D S by a deletion alleleof the C K R 5 structural gene. Science 1996;273:1856-1862. 5. Guerin S, Meyer L , Theodorou I, et al. C C R 5 A32 Deletion and Response to H A A R T in HIV-1-Infected Patients. A I D S 2000; 14:2788-2790. 6. Kasten S, Goldwich A , Schmitt M , et al. Positive influence of the Delta32CCR5 allele on response to highly active antiretroviral therapy ( H A A R T ) in HIV-1 infected patients. Eur J M e d Res 2000; 5:323-8. 7. Bogner JR, Lutz B , Kle in H G , et al. Association of highly active antiretroviral therapy failure with chemokine receptor 5 wi ld type. H I V Med . 2004; 5:264-72. 8. Bratt G , Karlsson A , Leandersson A C , et al. Treatment history and baseline viral load, but not viral tropism or C C R 5 genotype, influence prolonged antiviral efficacy of highly active antiretroviral treatment. A I D S 1998; 12: 2193-2202. 72 9. Brumme Z L , Chan K J , Dong W , et al. CCR5Delta32 and promoter polymorphisms are not correlated with initial virological or immunological treatment response. A I D S 2001; 15:2259-2266. 10. Wi t F W , van Rij R P , Weverling G J , Lange J M , Schuitemaker H . C C chemokine receptor 5 delta32 and C C chemokine receptor 2 641 polymorphisms do not influence the virologic and immunologic response to antiretroviral combination therapy in human immunodeficiency virus type 1-infected patients. J> Infect Dis 2002; 186: 1726-1732. 11. Hogg R S , Y i p B , Chan K J , et al. Rates of disease progression by baseline C D 4 cell count and viral load after initiating triple-drug therapy. J A M A 2001; 286: 2568-2577. 12. Brumme Z L , Dong W W , Y i p B , et al. Clinical and immunological impact of H I V envelope V 3 sequence variation after starting initial triple antiretroviral therapy. A I D S 2004; 18: F l - 9 . 13. Hogg R S , Heath K , Bangsberg D , et al. Intermittent use of triple-combination therapy is predictive of mortality at baseline and after 1 year of follow-up. A I D S 2002; 16: 1051-58. 14. Valdez H , Purvis S, Lederman M , Filingame M , Zimmerman P. Association of the C C R 5 A 3 2 Mutation Wi th Improved Response to Antiretroviral Therapy. J A M A 1999, 282:734-734. 15. Yamashita T E , Phair JP, Munoz A , et al. Immunologic and virologic response to highly active antiretroviral therapy in the Multicenter A I D S Cohort Study. A I D S 2001; 15:735-46. 73 16. Samson M , Libert F, Doranz B J , et al. Resistance to HIV-1 infection in Caucasian individuals bearing mutant alleles of the C C R 5 chemokine receptor gene. Nature 1996, 382:722-725. 17. Stewart G J , Ashton L J , B i t i R A , Ffrench R A , et al. Increased Frequency of C C R - 5 A32 heterozygotes among long-term non-progressors with HIV-1 infection. A I D S 1997, 11:1833-1838. 18. Gonzalez E , Kulkarni H , Bolivar H , et al. The influence of C C L 3 L 1 gene-containing segmental duplications on H I V - 1 / A I D S susceptibility. Science 2005; 307:1434-40. 19. Tang J, Wilson C M , Meleth S, et al. Host genetic profiles predict virological and immunological control of HIV-1 infection in adolescents. A I D S 2002; 16: 2275-84. 74 CHAPTER 4 HUMAN GENETIC VARIATION: IMPLICATIONS OF SINGLE NUCLEOTIDE POLYMORPHISMS IN CX3CR1 AND MDR-1 GENES ON INITIAL RESPONSE TO ANTIRETROVIRAL THERAPY1 4.1 I N T R O D U C T I O N Many recent studies have focused on the influence of naturally occurring polymorphisms within the human genome on HIV-1 infection and clinical disease progression. Probably the most notable polymorphism lies within the gene encoding the human chemokine receptor C C R 5 , a key cell-surface co-receptor for R5 strains of HIV-1 [1-3]. A homozygous 32-base pair deletion in the coding region of C C R 5 (CCR5A32/A32) has been associated with a protective effect with respect to HIV-1 infection [4,5], while the heterozygous genotype (CCR5wt/A32) has been associated with slower HIV-1 disease progression [4, 6-8] in untreated individuals. C X 3 C R I , a leukocyte chemotactic and adhesion receptor for the human chemokine fractalkine, has also been defined as a minor co-receptor for HIV-1 [9, 10]. Two common single nucleotide polymorphisms (SNPs) within the human C X 3 C R I gene have recently been identified: HIV-1 infected patients homozygous for the C X 3 C R I 2491 and 280M amino acid substitutions (a G-to-A nucleotide substitution encoding a V-to-I amino acid substitution at position 249, and a C-to-T nucleotide substitution encoding a T- to -M amino acid substitution at position 280) were shown to progress to A I D S and death more rapidly than those with other haplotypes [11], although these results were not confirmed in an independent study ' This chapter has been published as: Brumme Z L , Dong W W , Chan KJ, Hogg RS, Montaner JS, O'Shaughnessy M V , Harrigan PR. "Influence of polymorphisms within the CX3CR1 and MDR-1 genes on initial antiretroviral therapy response." AIDS. 2003;17:201-8. 75 (which examined the polymorphism at codon 280 only) [12]. Although the influence of C C R 5 polymorphisms on response to antiretroviral therapy has been examined [13-15], to date, no published studies appear to have examined whether polymorphisms in the C X 3 C R I gene are associated with altered response to antiretroviral therapy. SNPs in other genes may also affect response to antiretroviral therapy more directly. The human multidrug-resistance ( M D R ) - l gene encodes the ATP-dependent P-glycoprotein (P-gp) efflux membrane transporter which possesses a broad substrate specificity [17, 18] and therefore the ability to export a wide range of hydrophobic molecules from cells [19, 20]. The HIV-1 protease inhibitors (Pis) are substrates for the P-gp drug efflux pump in vitro [21-23], a finding which could potentially affect the bioavailability of HIV-1 protease inhibitors and limit the therapeutic efficacy of Pis in P-gp expressing cells [24], including CD4+ cells [25]. A single C-to-T nucleotide polymorphism in exon 26 (C3435T) of the M D R - 1 gene is significantly correlated with decreased intestinal expression levels and activity of P-glycoprotein [26]. In other disease models, mainly cancer, the level of M D R - 1 expression is correlated with the resistance of tumour cells against chemotherapeutic agents [27, 28]. It has also been found that overexpression of multidrug transporters significantly reduces the accumulation of Pis in human lymphocytes [24], and recently shown that individuals homozygous for the 3435C allele exhibited increased efflux of the P-gp substrate rhodamine 123 from CD56+ natural killer cells [29]. Based on these findings and evidence of racial differences in 3435C allele distribution, it was hypothesized that differential P-gp expression could affect the utility of Pis in African populations [30]. Finally, data have confirmed a correlation between initial C D 4 response, nelfinavir and efavirenz levels, and M D R - 1 polymorphisms in adherent patients [31]. 76 To address whether previously identified SNPs in C X 3 C R I or M D R - 1 affect response to antiretroviral therapy in an unselected patient population, we retrospectively determined the prevalence of the C X 3 C R I 2491 and 280M amino acid substitutions, and the M D R - 1 C3435T polymorphisms in a group of 461 antiretroviral naive individuals initiating their first H I V therapy in British Columbia, Canada, between June 1996 and August 1998, and assessed the potential effects of these natural human polymorphisms on the initial virological and immunological response to antiretroviral therapy. 77 4.2 M A T E R I A L S A N D M E T H O D S Study Design and Patients In the province of British Columbia, Canada, antiretroviral drugs are distributed free of charge to HIV-infected individuals through a centralised drug treatment program based at the British Columbia Centre for Excellence in H I V / A I D S (the B . C . Centre). Briefly, B . C . treatment guidelines recommend that plasma viral load testing be performed at baseline, again at 4 weeks after initiating or changing therapy, and approximately every 3 months thereafter [see reference 32 for more details of policies and treatment guidelines]. The median time between plasma viral load measurements for subjects included in this study was three months (Interquartile Range 2-4 months). Less than 5% of sequential viral load measurements had an elapsed time of > 8 months. A l l antiretroviral naive individuals first seeking treatment at the B . C . Centre between June 1, 1996 and August 31, 1998, were eligible for analysis in this retrospective study (N=479). O f these, 461 (96.2%) had a pre-therapy plasma sample available for testing. Greater than 80% of the antiretroviral naive patients were prescribed triple therapy including two nucleoside analogues and one protease inhibitor, consistent with treatment guidelines existing at the time. Adherence to medication was evaluated by analyzing patients' prescription records and determining the frequency of prescriptions filled on time during the first year of therapy [33, 34], with "adherence" being defined here as fill ing 100% of prescriptions on schedule. Ethical Considerations Ethical approval for this study was obtained from the local ethics board at St. Paul's Hospital, Vancouver. A l l patient samples were assigned new anonymous code identifiers 78 prior to genotypic analyses. Laboratory analyses were performed using the coded identifiers only. Plasma Viral Load and CD4 Cell Count C D 4 cell counts were measured by standard techniques. Plasma HIV-1 R N A levels (viral loads) were determined using the Roche Amplicor Monitor assay (Roche Diagnostics, Laval, Quebec, Canada) using either the standard method or the ultrasensitive adaptation. The ultrasensitive adaptation to this assay was not routinely available over the time period of this study; for this reason a constant cut-off of 500 H I V R N A copies/ml was used throughout the analysis. Determination of CX3CRI, MDR-1, CCR5 A32 polymorphisms, and HIV-RNA Genotyping for drug resistance. Plasma nucleic acids were extracted and precipitated with guanidine thiocyanate and isopropanol followed by two ethanol washes. The region of C X 3 C R I containing the polymorphisms of interest was amplified in a single round of P C R , using the oligonucleotide primers C X 3 C R 1 F l (5' C T G A A T C A G T G A C A G A A A A C T T T 3') and C X 3 C R 1 R l (5' G T A G A C A C A A G G C T T T G G G A T T C 3'), to generate a 1.1 K b product. In the case of the M D R - 1 polymorphism in exon 26, one round of P C R using oligonucleotide primers M D R - 1 F l (5' C A A T T A T G A C C T T G T T G G G T T A A 3') and M D R - 1 R l (5' A G A T G C T T G T A T A C A G G T A A G G 3') was performed, to generate a 0.5 K b product. P C R products were then sequenced directly in both the 5' and 3' directions using these primers and the BigDye dye terminator cycle sequencing kit (Applied Biosystems, Foster City, C A ) and resolved on an A B I 3700 automated sequencer. C C R 5 A 3 2 status [13] and HIV-1 drug-resistance genotyping for the protease gene [35] were performed as previously described. 79 Statistical Analysis Time to virological success was defined as the time to achieve a plasma viral load (pVL) < 500 copies HIV-1 R N A / m l from baseline. Time to virological failure was defined as the time elapsed between virological success and the second of two consecutive p V L measurements > 500 copies/ml. If a person never achieved virological success, they were excluded from the analysis of time to virological failure. If a patient reached therapeutic success and did not subsequently fail then they were censored at the time of their last viral load date (occurring prior to December 1, 2000). Time to immunological failure was defined as the time to the second of two successive C D 4 cell counts below baseline. The effects of the C X 3 C R I , M D R - 1 , and combined C X 3 C R I and C C R 5 polymorphisms on time to virological and immunological success and/or failure were analyzed by Kaplan-Meier methods. Group differences were assessed using the log-rank test. Cox proportional hazard regression was used to model the effect of the C X 3 C R I and M D R - 1 polymorphisms on time to virological and immunological failure while adjusting for age, gender, baseline p V L , baseline C D 4 cell count, adherence, type of antiretroviral therapy at initiation (triple therapy with 2 NRTIs and a PI vs. dual therapy), and "time to virological success" in the virological failure model only. Finally, Fisher's exact test was used to assess the potential correlation between M D R - 1 genotype and the development of HIV-1 resistance to protease inhibitors in the first two years of therapy (an analysis restricted to those patients initiating therapy including a protease inhibitor). A l l of the 461 patients in the study were included in the "time-to-virological success" analyses. In order to be included in the "time-to-virological failure" analysis, patients had to register at least one post-baseline H I V p V L < 500 copies/ml; this included 374 of 461 80 individuals (81.1%). 382 of 461 individuals (82.9%) had at least one post-baseline C D 4 count, and were therefore included in the time-to-immunological-failure analysis. 81 4.3 R E S U L T S Baseline characteristics of the 461 HIV-1 infected individuals examined for C X 3 C R I 2491 and 2 8 0 M amino acid substitutions and the M D R - 1 C3435T polymorphisms are presented in Table 1. There were no significant differences with respect to age, gender, baseline plasma viral load, C D 4 count, A I D S defining-illness or history of injection drug use between the 479 patients eligible for the study and the 461 (96.2%) patients for whom plasma samples were available for testing. However, patients in the study group (N=461) were more likely to have been prescribed triple therapy (p=0.03). There were no statistically significant differences observed between any baseline parameter and presence or absence of SNPs (p>0.1, data not shown). The median follow up time after starting therapy was 40 months. P C R products and C X 3 C R I genotypes were obtained for a total of 451 of 461 individuals (97.8%). A t C X 3 C R I codon 249, 282 individuals (62.5%) were homozygous G / G (encoding amino acid 249V), 54 individuals (12.0%) were homozygous A / A (encoding amino acid 2491), and the remaining 115 (25.5%) were heterozygous. A t C X 3 C R I codon 280, 339 (75.2%) were homozygous C / C (encoding amino acid 280T), 27 (6.0%) were homozygous T/T (encoding amino acid 280M), and the remaining 85 (18.8%) were C/T heterozygous. The C X 3 C R I 249G allele (encoding amino acid 249V) was in complete linkage disequilibrium with the 280C allele (encoding amino acid 280T) [11]. P C R products and M D R - 1 genotypes were obtained for 455 of 461 individuals (98.7%). A total of 137 (30.1%) individuals were 3435 C / C , 165 (36.3%) were 3435 T/T, and the remaining 153 (33.6%) were 3435 C/T. A s described elsewhere, patients were also genotyped for the C C R 5 A 3 2 [13]. The allele and haplotype frequencies for all polymorphisms in the C X 3 C R I , M D R - 1 and C C R 5 genes are summarized in Table 2. There was no genetic linkage detected 82 between any of the polymorphic groups (with the exception of the linkage observed within the C X 3 C R I gene). Kaplan-Meier analysis of the influence of M D R 1 genotype on time elapsed to virological failure (defined as the time elapsed to the second of two successive p V L > 500 copies/ml from initial success) is shown in figure 1. The proportion of individuals not achieving virological success was between 18-20% for all three groups. There was a trend to earlier virological failure in the homozygous 3435C7C group which did not attain statistical significance (p=0.07). No effect of the 3435 C/T nucleotide polymorphism in the M D R - 1 gene on immunological failure, defined as the time elapsed to the second of two consecutive C D 4 counts below baseline, was detected (p=0.46, data not shown). When only the homozygous groups (3435C/C and 3435T/T) were considered in the analysis, the correlation between 3435C/C and earlier virological failure reached statistical significance (p=0.046), but no effect was detected on immunological failure (p=0.97). In a secondary Kaplan-Meier analysis, there was no detectable influence of M D R - 1 genotype on the time elapsed to virological success (time to H I V p V L < 500 copies/ml) (p=0.371, data not shown). In addition, we wished to determine whether the polymorphism in M D R - 1 could be correlated with the development of resistance to H I V protease inhibitors [24]. A search of our HIV-1 genotype database identified a subset of 124 (of 461) individuals (26.9%) who had at least one HIV-1 antiretroviral resistance genotype requested by a physician performed during the course of study follow-up, at least 6 months after initiation of therapy. 13 of 124 patients (10.5%) displayed decreased susceptibility to one or more protease inhibitors in at least one post-therapy genotype using a "virtual" phenotype analysis [35] where "decreased susceptibility" was defined as a > 3-, 3.5-, 4-, 2.5-, 2-, or 2.5-fold decrease in susceptibility to 83 indinavir, ritonavir, nelfinavir, saquinavir, amprenavir and lopinavir, respectively [36]. Samples for which genotypes were available revealed no correlation between the M D R - 1 polymorphism and development of resistance to protease inhibitors using the Fisher's exact test, in an analysis restricted to those patients initiating therapy including a protease inhibitor (n=83, p=0.99). Pi-resistance was evenly distributed between the M D R - 1 3435 C / C , T/T, and C/T genotype groups. For the Kaplan-Meier analyses of the influence of the C X 3 C R I genotype on time elapsed to virological and immunological failure, patients were classified into categories based on their amino acid polymorphisms at codons 249 and 280. The first analysis included homozygous amino acid combinations only ( C X 3 C R I genotypes 2491 280T, 2491280M and 249V 280T). The proportion of individuals censored in the time to virological failure analysis was similar for all three groups. No significant influence was observed between the 3 groups with respect to time elapsed to virological failure (p=0.99, data not shown), however, a tendency to earlier immunological failure was observed in the two groups containing the 2491 polymorphism which achieved statistical significance (p=0.02, figure 2). When heterozygous allele combinations were included in the analysis, the effect on virological outcome remained not statistically significant (p=0.92, data not shown), and a trend for earlier time to immunological failure for the two groups containing the homozygous 2491 polymorphism remained (p=0.07, data not shown). Kaplan-Meier survival curves for the heterozygous allele groups fell near those of the homozygous 249V 280T group. Finally, when both homozygous and heterozygous allele combinations at codons 249 (I, I/V, V ) and 280 (T, T / M , M ) were analyzed independently (N=451), no effects of variation at codons 249 or 280 on virological outcome could be demonstrated (p=0.78 and p=0.52, respectively). 84 Similarly, no effect of variation at codon 280 alone [12] on immunological outcome could be demonstrated (p=0.167). However, a statistically significant trend for earlier immunological failure in the homozygous 2491 group was observed (p=0.036, data not shown). In a secondary analysis, no influence of CX3CRI genotype on the time elapsed to virological success was observed (p=0.89, data not shown). No significant association between the C X 3 C R I 2491 or 249V genotypes and CCR5A32 status was observed (N=446, p>0.1, data not shown). No effect of the combined C X 3 C R I and C C R 5 genotypes on time to virological success or virological or immunological failure was observed (p>0.6, data not shown). Cox proportional hazard regression was used to model the effect of the CX3CRI and M D R - 1 polymorphisms on time to virological and immunological failure while adjusting for age, gender, baseline p V L , baseline C D 4 cell count, adherence, type of antiretroviral therapy at initiation, and "time to virological success" in the virological failure model only (table 3). The results were consistent with the previous analyses. The M D R - 1 virological analysis reached statistical significance (p=0.04) (risk ratios [and 95% confidence intervals] were 0.67 [0.45-0.99] for the M D R - 1 3435 T/T and 0.71 [0.47-1.06] for the M D R - 1 3435 C/T (Y) genotypes), and results of the CX3CRI immunological analysis also remained significant (p=0.05) (risk ratios were 1.92 [1.00-3.69] for the C X 3 C R 1 2491280T and 1.46 [0.80-2.70] for the 2491 2 4 0 M genotypes). O f interest, the only baseline variable that was consistently protective in the multivariate analyses was "100% adherence" (p=0.0001, Table 3), which was evaluated using prescription refill records [33,34], with "adherence" being defined as filling 100% of prescriptions on schedule. 85 In order to control for non-adherence, a secondary subset analysis was performed on only those individuals who were most likely highly adherent to therapy based on the above definition (240 of 461 subjects, or 52.1%). The adherence subset analyses were consistent with previous results (p>0.05 for C X 3 C R I and M D R - 1 virological failure and M D R - 1 immunological failure, and p <0.05 for C X 3 C R I immunological failure). 86 4.4 D I S C U S S I O N A N D C O N C L U S I O N S This study investigated the effects of single nucleotide polymorphisms (SNPs) within the human C X 3 C R I and M D R - 1 genes on time to virological and immunological treatment failure in a large cohort of antiretroviral naive patients initiating therapy. We report a small but statistically significant correlation between the C X 3 C R I 2491 polymorphism and slightly earlier immunological failure; and a trend to earlier virological failure in the M D R - 1 3435C/C genotype group. The role of M D R - 1 gene polymorphisms and P-glycoprotein expression in H I V pathogenesis remains largely unresolved. It has been suggested that overexpression of multidrug transporters may accelerate the development of HIV-1 resistance to protease inhibitors [24], and that the M D R - 1 3435C/C genotype may be a factor restricting access of HIV-1 protease inhibitors to target cells expressing P-gp [30]. The M D R - 1 3435C/C genotype has a much higher frequency among individuals of African origin (between 60-80%) [30] than in individuals of Caucasian or Asian origin (> 30%) [26, 30, 37]. Although complete ethnic background data is not available for our cohort (individuals are not obligated to answer any questions pertaining to ethnic origin), partial data indicates that the cohort is mainly comprised of persons of Caucasian and aboriginal descent. Based on this partial ethnic data, the frequency of the homozygous M D R - 1 3435C/C genotype (30.1%) was consistent with the frequency previously reported for Caucasian and Asian populations [26, 30, 37]. In support of previous reports suggesting a potential effect of the M D R - 1 polymorphism on the effectiveness of HIV-1 protease inhibitors [24], a strong trend to earlier virological failure in the M D R - 1 3435C/C genotype group was observed. However, there was no detectable effect of the M D R - 1 polymorphism on initial immunological response to 87 antiretroviral therapy using our a priori definition, or, in a limited subset, HIV-1 drug resistance development. These observations extend the findings from a recent study of 123 Caucasian, drug-naive patients initiating antiretroviral therapy, which reported a beneficial effect of the M D R - 1 3435T/T genotype on short-term C D 4 cell count increases [31]. Differences in virological response could not be detected over the 6-month follow-up period [31], compared to the 40 months of follow-up reported here. The C X 3 C R I haplotype frequencies determined in this study are consistent with those previously reported [11]: 249V 280T was the most common, with a haplotype frequency of 0.75 (compared to 0.678-0.730 in French cohorts), followed by 2491 280M, at 0.15 (France 0.126-0.198), and finally 2491 280T, at 0.1 (France 0.123-0.142). Similarly, we observed complete linkage disequilibrium within C X 3 C R I SNPs, and no discernible genetic linkage between the C X 3 C R I and C C R 5 polymorphisms [11]. Thus, despite the fact that the C X 3 C R I and C C R 5 genes are found in close proximity on chromosome 3 [11], any effects of the C X 3 C R I polymorphisms on disease progression and therapy outcome are not due to linkage to the C C R 5 A 3 2 allele. C X 3 C R I 2491280M was previously identified as a recessive H I V genetic risk factor in untreated individuals, with 280M (but not necessarily 2491) independently associated with faster progression to clinical A I D S and more rapid C D 4 decline [11]. However, McDermott et al. examined C X 3 C R I 2 8 0 M alone and could not confirm these results [12]. Although not directly addressed here, the discrepancy between the two studies may be explained by our observation that the polymorphism at codon 249, and not codon 280, may be largely responsible for the differences observed in response to therapy. The homozygous 2491 polymorphism was associated with reduced immunological response alone and in combination with variation at codon 280, while independent variation 88 at codon 280 was not significant in our analyses. Variation in C X 3 C R I genotype at either codon 249 or 280 however, did not influence virological response in any of our analyses. A limitation of this study relates to the frequency of viral load and C D 4 tests performed on the study group. Although the majority of patients had tests performed at baseline, one month, and every three months thereafter (the median time between viral load tests in this study was 3 months), a minority of patients may have had viral load tests performed more or less often. It is possible therefore, that the therapy follow-up data may be biased towards longer times to success and/or failure, due to the possibility that some study subjects may have large gaps between viral load or C D 4 tests. Another possible limitation of this study relates to the potential confounding effect provided by incomplete adherence. We addressed this using a previously validated definition based on prescription refills [33,34]. Although we cannot rule out the possibility of incomplete adherence, it is unlikely that this would be unbalanced between S N P groups. Other potential limitations of this study include its retrospective nature, the fact that the cohort, although large, represents only a geographically limited population, and the fact that the resistance subset for the M D R - 1 analysis is non-random, and therefore potentially subject to bias. Finally, the biological mechanism of the effects of the C X 3 C R I and M D R - 1 polymorphisms is unclear. In support of previous studies identifying the C X 3 C R I 2491280M genotype as a genetic risk factor in A I D S progression in untreated individuals [11], and linking the M D R - 1 3435 C / C genotype to reduced C D 4 cell responses [31], we have observed a correlation between the C X 3 C R I 2491 polymorphism and the M D R - 1 3435C/C genotype and slightly earlier immunological and virological failure, respectively. The magnitude of these effects is relatively minor despite the fact that they remain statistically significant after adjusting for multiple factors. 89 The actual utility of determining these genotypes in routine clinical practice is therefore not clear. 90 Table 4.1: Baseline characteristics of the study population, prior to the initiation of antiretroviral therapy (N=461). Characteristics median IQR Total sample size age (years) Gender(%) plasma viral load (x103 copies/ml) CD4 count (x106 cells/I) Baseline AIDS diagnosis % Triple Therapy 461 36 31-42 391 Male (84.8%) 70 Female (15.2%) 89 30-280 310 140-460 49(10.6%) 385 (83.5%) The distribution of baseline parameters was similar for all the individual C X 3 C R I and M D R -1 genotype groups (p>0.1). (IQR = interquartile range) 91 Table 4.2: Allele and haplotype frequencies for the CX3CR1, MDR-1 and CCR5 polymorphisms. CX3CR1 Genotypes at codon 280 # of patients % of patients Independent haplotype codon 249 (N = 451) allele frequencies frequencies A/A (249I) C/C (280T) 21 4.7 249A=0.25 249I280T = 0.1 A/A (249I) "ITT (280M) 27 6.0 249G=0.75 249I 280M = 0.15 A/A (249I) C/T (280T/M) 6 1.3 G/G (249V) C/C (280T) 282 62.5 280C=0.85 249V 280T = 0.75 G/A(249V/I) C/C (280T) 36 8.0 280T=0.15 G/A(249V/I) C/T (280T/M) 79 17.5 MDR-1 Genotypes at nucleotide 3435 # of patients % of patients allele frequencies (N = 455) 3435 C/C 137 30.1 3435C=0.47 3435 T/T 165 36.3 3435T=0.53 3435 C/T 153 33.6 C C R 5 CCR5 Genotype # of patients % of patients (N = 461) CCR5 wt/wt 392 85.0 CCR5 A32/wt 68 14.8 CCR5 A32/A32 1 0.2 92 Table 4.3: Multivariate Cox Model showing effect of baseline parameters, including MDR-1 and CX3CRI polymorphisms, on time to Virological and Immunological Failure. Time to virological failure Time to immunological failure Baseline Variable Risk ratio 95% CI P Risk ratio 95% CI P MDR-1 C X 3 C R I Age 0.91 0.83-0.99 0.03 1.05 0.96-1.15 0.3 Male sex 0.77 0.48-1.22 0.3 0.83 0.52-1.34 0.4 plasma viral load 1.14 0.91-1.43 0.3 0.95 0.75-1.22 0.7 CD4 count 1.09 1.02-1.15 0.007 1.14 1.08-1.20 <0.001 100% adherence 0.84 0.79-0.90 <0.001 0.85 0.82-0.89 <0.001 Triple Therapy 0.39 0.26-0.59 <0.001 0.61 0.39-0.94 0.03 MDR-1 3435T/T 0.67 0.45-0.98 0.04 1.02 0.68-1.54 0.9 MDR-1 3435C/T (Y) 0.71 0.47-1.06 0.09 0.86 0.56-1.31 0.5 time to virological success 1.00 0.97-1.03 0.9 Age 0.90 0.80-1.01 0.07 0.96 0.85-1.08 0.5 Male sex 0.96 0.54-1.71 0.9 0.93 0.53-1.63 0.8 plasma viral load 1.36 1.03-1.81 0.03 0.93 0.69-1.24 0.6 CD4 count 1.09 0.99-1.19 0.07 1.14 1.06-1.23 <0.001 100% adherence 0.84 0.78-0.90 <0.001 0.85 0.81-0.90 <0.001 Triple Therapy 0.37 0.22-0.64 <0.001 0.57 0.31-1.06 0.08 C X 3 C R I 249I 280T 0.96 0.39-2.33 0.9 1.89 0.98-3.64 0.05 C X 3 C R I 249I 280M 1.13 0.59-2.20 0.7 1.48 0.80-2.74 0.2 time to virological success 1.00 0.96-1.03 0.8 CI, confidence interval 93 Figure 4.1: Kaplan-Meier plot comparing the influence of nucleotide variation at position 3435 of the human P-glycoprotein (MDR-1) gene on time to virological failure. The number of patients in each genotype group (3435 C / C , C/T , and T/T) still experiencing virological success at 24 months was (55, 72 and 74). (p=0.07) o 0.2 -a MDR1 3435 C/C (n=112) o MDR1 3435 C/T(n=125) x MDR1 3435 T/T (n=133) 0.0 -h ; T 1 1 " i—: r 0 10 20 30 40 50 Time (months) 94 Figure 4.2: Kaplan-Meier plot comparing the influence of amino acid variation at codons 249 and 280 of the human CX3CRI receptor gene on time to immunological failure. For clarity, only those genotype groups involving homozygous amino acid combinations are shown. The number of patients in each genotype group ( C X 3 C R 1 2491 280T, 2491 280M, and 249V 280T) still experiencing immunological success at 24 months was (9, 10, 133). (p=0.02) Time (months) 95 4.5 R E F E R E N C E S 1. Samson M , Labbe O, Mollereau C, Vassart G , Parmentier M . Molecular cloning and functional expression of a new human CC-chemokine receptor gene. Biochemistry 1996, 35:3362-3367. 2. Deng H , L i u R, Ellmeier W , Choe S, et al. Identification of a major co-receptor for primary isolates of H I V - 1 . Nature 1996, 381:661-666. 3. Dragic T, L i twin V , Al laway GP , et al. HIV-1 entry into CD4+ cells is mediated by the chemokine receptor C C - C K R - 5 . Nature 1996, 381:667-673. 4. Huang Y , Paxton W A , Wolinsky S M , et al. The role of a mutant C C R 5 allele in H I V -1 transmission and disease progression. Nat Med 1996, 11:1240-1243. 5. L i u R, Paxton W A , Choe S, et al. Homozygous defect in HIV-1 coreceptor accounts for resistance of some multiple-exposed individuals to HIV-1 infection. Cel l 1996, 86:367-377. 6. Dean M , Carrington M , Winkler C, et al. Genetic restriction of HIV-1 infection and progression to A I D S by a deletion allele of the C K R 5 structural gene. Hemophilia Growth and Development Study, Multicenter A I D S Cohort Study, Multicenter Hemophilia Cohort Study, San Francisco City Cohort, A L I V E Study. Science 1996, 273:1856-1862. 7. Samson M , Libert F, Doranz B J , et al. Resistance to HIV-1 infection in Caucasian individuals bearing mutant alleles of the C C R 5 chemokine receptor gene. Nature 1996, 382:722-725. 96 8. Stewart G J , Ashton L J , B i t i R A , et al. Increased Frequency of C C R - 5 A32 heterozygotes among long-term non-progressors with HIV-1 infection. A I D S 1997, 11:1833-1838. 9. Combadiere C , Salzwedel K , Smith E D , Tiffany H L , Berger E A , Murphy P M . Identification of C X 3 C R 1 . A chemotactic receptor for the human C X 3 C chemokine fractalkine and a fusion coreceptor for H I V - 1 . J B i o l Chem 1998, 273: 23799-23804. 10. Rucker J, Edinger A , Sharron M . Utilization of Chemokine Receptors, Orphan Receptors, and Herpesvirus-Encoded Receptors by Diverse Human and Simian Immunodeficiency Viruses. J V i ro l 1997, 71:8999-9007. 11. Faure S, Meyer L , Costagliola D , , et al. Rapid Progression to A I D S in HIV+ individuals with a Structural Variant of the Chemokine Receptor C X 3 C R 1 . Science 2000,287: 2274-2277. 12. McDermott D H , Col la JS, Kleeberger C A , et al. Genetic Polymorphisms in C X 3 C R 1 and Risk of H I V Disease. Science 2000, 290: 2031a. 13. Brumme Z L , Chan K J , Dong W , et al. C C R 5 A32 and promoter polymorphisms are not correlated with initial virological and immunological treatment reponse. A I D S 2001, 15: 2259-2266. 14. Bratt G , Karlsson A , Leandersson A C , Albert J, Wahren B , Sandstrom E . Treatment history and baseline viral load, but not viral tropism or C C R 5 genotype, influence prolonged antiviral efficacy of highly active antiretroviral treatment. A I D S 1998, 12:2193-2202. 97 15. Guerin S, Meyer L , Theodorou I, et al. C C R 5 delta32 Deletion and Response to H A A R T in HIV-1-Infected Patients. S E R O C O / H E M O C O Study Group. A I D S 2000, 14:2788-2790. 16. O'Brien T R , McDermott D H , Ioannidis JP, et al. Effect of chemokine receptor gene polymorphisms on the response to potent antiretroviral therapy. A I D S 2000, 14:821-826. 17. Gottesman M M , Pastan I. Biochemistry of multidrug resistance mediated by the multidrug transporter. Annu Rev Biochem 1993, 62: 385-427. 18. Leveque D , Jehl F: P-glycoprotein and pharmacokinetics. Anticancer Res. 1995, 15: 331-336. 19. Sarkadi B , Mul ler M , Homolya L , , et al. Interaction of bioactive hydrophobic peptides with the human multidrug transporter. F A S E B J 1994, 8: 766-770. 20. Sharom FJ , Y u X , DiDiodato G , Chu JW. Synthetic hydrophobic peptides are substrates for P-glycoprotein and stimulate drug transport. Biochem J 1996, 320: 421-428. 21. K i m A E , Dintaman J M , Waddell DS , Silverman J A . Saquinavir, an H I V protease inhibitor, is transported by P-glycoprotein. J Pharmacol Exp Ther 1998, 286: 1439-1455. 22. Lee C G , Gottesman M M , Cardarelli C O , Ramachandra M , Jeang K T , Ambudkar S V , et al. HIV-1 protease inhibitors are substrates for the M D R - 1 multidrug transporter. Biochemistry 1998, 37: 3594-3601. 98 23. Washington C B , Curan G E , Man M C , Sikic BI , Blaschke T F . Interaction of anti-HIV protease inhibitors with the multidrug transporter P-glycoprotein in human cultured cells. J Acquir Immune Defic Syndr Humn Retrovirol 1998, 19: 203-209. 24. Jones K , Bray P G , Khoo S H , et al. P-Glycoprotein and transporter M R P - 1 reduce H I V protease inhibitor uptake in C D 4 cells: potential for accelerated viral drug resistance? A I D S 2001, 15: 1353-1358. 25. Huisman M T , Smit JW, Schinkel A H . Significance of P-glycoprotein for the pharmacology and clinical use of H I V protease inhibitors. A I D S 2000, 14: 237-242. 26. Hoffmeyer S, Burk O, von Richter O, et al. Functional polymorphisms of the human multidrug-resistance gene: Multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo. P N A S 2000, 97: 3473-3478. 27. Gottesman M M , Pastan I, Ambudkar V . P-glycoprotein and multidrug resistance. Curr Opin Genet Dev 1996, 6: 610-617. 28. Ambudkar S V , Dey S, Hrycyna C A , Ramachandra M , Pastan I, Gottesman M M . Biochemical, cellular, and pharmacological aspects of the multidrug transporter. Annu Rev Pharmacol Toxicol 1999, 39: 361-398. 29. Hi tz l M , Drescher S, van der Kuip H , et al. The C3435T mutation in the human M D R 1 gene is associated with altered efflux of the P-glycoprotein substrate rhodamine 123 from CD56+ natural killer cells. Pharmacogenetics 2001, 11: 293-298. 30. Schaeffeler E , Eichelbaum M , Brinkmann U , et al. Frequency of C3435T polymorphism of M D R 1 gene in African people. Lancet 2001, 358: 383-384. 99 31. Fellay J, Marzol ini C , Meaden E R , Back D J , Bucl in T, Chave JP, et al. Response to antiretroviral treatment in HIV-1 infected individuals with allelic variants of the multidrug resistance transporter 1: a pharmacogenetics study. Lancet 2002, 359:30-36. 32. Hogg R S , Y i p B , Chan K J , Wood E , et al. Rates of disease progression by baseline C D 4 cell count and viral load after initiating triple-drug therapy. J A M A 2001, 286:2568-2577. 33. Low-Beer S, Y i p B , O'Shaughnessy M V , Hogg R S , Montaner JS. Adherence to triple therapy and viral load response. J Acquir Immune Defic Syndr 2000, 23: 360-361. 34. Hogg R S , Heath K , Bangsberg D , et al. Intermittent use of triple-combination therapy is predictive of mortality at baseline and after 1 year of follow-up. A I D S 2002, 16(7):1051-1058 35. Hertogs K , de Bethune M P , Mi l le r V , et al. A Rapid Method for Simultaneous Detection of Phenotypic Resistance to Inhibitors of Protease and Reverse Transcriptase in Recombinant Human Immunodeficiency Virus Type 1 Isolates from Patients Treated with Antiretroviral Drugs. Antimicrob Agents Chemother 1998, 42: 269-276. 36. Harrigan PR, Montaner JS, Wegner S A , et al. World-wide variation in HIV-1 phenotypic susceptibility in untreated individuals: biologically relevant values for resistance testing. A I D S 2001, 15: 1671-1677. 37. Cascorbi I, Gerloff T, Johne A , et al. Frequency of single nucleotide polymorphisms in the P-glycoprotein drug transporter M D R - 1 gene in white subjects. C l i n Pharmacol Ther 2001, 69: 169-174. 100 CHAPTER 5 HIV GENETIC VARIATION: IMPLICATIONS OF ANTIRETROVIRAL RESISTANCE TO LONG-TERM HAART OUTCOMES 1 5.1 I N T R O D U C T I O N Morbidity and mortality associated with Human Immunodeficiency Virus (HIV) infection and Acquired Immunodeficiency Syndrome (AIDS) have declined significantly in North America and Western Europe as a result of the introduction of combination antiretroviral therapies [1-6]. Despite these advances in H I V treatment, HIV-related deaths continue to occur across the developed world [1-4, 7, 8]. Drug-resistant H I V variants are often cited as a major barrier to long-term antiretroviral efficacy and a major cause of treatment failure [9-16]. However, it remains to be conclusively determined what proportion of recent HIV-related deaths among those with access to antiretroviral therapy are due to diminished treatment options as a consequence of multi-drug resistance and what proportion may be due to other factors. Using data from a centralized H I V / A I D S Drug Treatment Program in British Columbia, Canada, the prevalence of antiretroviral resistance among HIV-infected persons who died between July 1997 and December 2001 was contrasted to the prevalence of resistance among living individuals experiencing virologic therapy failure over the same study period. Our objective was to determine the degree to which antiretroviral resistance may contribute to mortality among HIV-infected individuals in the era of Highly Active Antiretroviral Therapy ( H A A R T ) . 1 This chapter has been published as: Recsky M A , Brumme Z L , Chan KJ, Wynhoven B, Yip B, Dong WW, Heath K V , Montaner JS, Levy AR, Hogg RS, Harrigan PR. "Antiretroviral resistance among HIV-infected persons who have died in British Columbia, in the era of modern antiretroviral therapy". J Infect Dis. 2004; 190:285-92. 101 5.2 M A T E R I A L S A N D M E T H O D S British Columbia HIV/AIDS Drug Treatment Program In the province of British Columbia, Canada, antiretroviral agents have been distributed at no cost to all eligible HIV-infected individuals through a centrally administered H I V / A I D S Drug Treatment Program (DTP) since 1986. The D T P remains the only source of free-of-charge antiretroviral therapy in the province, and pharmaceutical sales suggest that <1% of HIV-infected British Columbians purchase antiretrovirals outside the D T P [17]. The D T P database contains sociodemographic data as well as comprehensive antiretroviral prescription and treatment data for all those accessing H I V care and treatment in British Columbia. Details on patient enrolment, policies and D T P treatment guidelines have been previously published [17]. A s a part of routine patient monitoring, plasma HIV-1 R N A levels (HIV viral loads) and C D 4 cell counts are generally measured at baseline, after one month, and at three month intervals thereafter. A portion of the plasma sample used for H I V viral load testing is also archived. Ethical approval for this study was obtained from the research ethics board at the University of British Columbia/St. Paul's Hospital, Vancouver. Study Sample and Variables of Interest A total of 637 deaths were recorded amongst persons enrolled in the D T P between July 1997 and December 2001. Deaths were identified through physician reports and through linkages with the B C Divis ion of Vi ta l Statistics. Deaths were classified according to ICD-9 (1995-1999) and ICD-10 (2000-2001) coding. Underlying causes of death coded under ICD-9 categories 042-044 and ICD-10 categories B20- B24 identified deaths directly caused by H I V infection. July 1997 was selected as the study start date because H A A R T became 102 universally available to all HIV-positive individuals seeking treatment in British Columbia at this time. Subjects included in this study form a part of a larger group of HIV-infected individuals who have died in British Columbia between 1995 and 2001, who have recently been described in detail with respect to sociodemographic characteristics, risk factors and access to antiretrovirals [17]. Time spent on antiretroviral treatment was estimated using a previously validated method based on analyzing prescription refill data [18]. Genotypic Resistance Testing Accidental deaths were assumed to be not directly related to antiretroviral resistance and these plasma samples were not genotyped. Samples with viral loads < 500 copies H I V R N A / m l were not genotyped due to an inadequate amount of plasma H I V . However, because these plasma samples originated from individuals with successful suppression of plasma H I V R N A , it is unlikely that these samples contained H I V harboring resistance mutations in sufficient quantities to compromise effectiveness of therapy. Genotypic drug resistance testing was restricted to those individuals who were prescribed antiretrovirals and who died of a non-accidental cause (Figure 1). For each subject meeting these requirements, the last on-therapy plasma sample was genotyped for drug resistance if the plasma viral load (pVL) was >500 copies H I V R N A / m l . Resistance genotyping was performed as previously described [19]. Samples were considered "resistant" if they displayed one or more major resistance mutations, based on the I A S - U S A list [20]: lamivudine resistance (184I/V); any other nucleos(t)ide Reverse Transcriptase Inhibitor (nRTI) resistance (41L, 62V, 65R, 67N, 69A/N/D or insertion, 70R, 74V, 751, 151M, 210W, 215F/Y or 219E/Q); any Non-Nucleoside Reverse Transcriptase Inhibitor (NNRTI) resistance (1001, 103N, 106A, 1081, 103 181C/I, 188C/H/L/Y, 190A/S, or 236L); or any Protease Inhibitor (PI) resistance (30N, 46I/L, 48V, 5 0 L / V , 5 4 L / M , 82A/F/S/T, 84V, or 90M). Drug resistance among living HIV-infected persons with virologic therapy failure In the evaluation of H I V treatment outcomes, mortality is often used as an endpoint in the definition of "therapy failure". However, surrogate markers such as H I V plasma viral load and C D 4 cell count are also commonly used as definitions of "therapy failure" in clinical trials and population-based studies. In order to provide context, the prevalence of resistance in the study group (where "therapy failure" was defined as mortality) was contrasted to the prevalence of resistance in a large group of living individuals selected based on a plasma-viral load driven definition of "therapy failure". A "virologic failure" group of N=1220 subjects was therefore selected from the H I V / A I D S Drug Treatment Program database which represented all living HIV-infected individuals in British Columbia failing their therapy (as defined by a p V L > 500 copies H I V R N A / m l ) between the study dates of July 1997-December 2001, for whom a physician-requested H I V genotypic drug resistance test had been performed. In cases where an individual had multiple genotypes performed over the period of interest, the latest available on-therapy genotype was selected. Genotypic resistance testing and interpretation of resistance genotypes was performed as described above. Individuals included in the mortality group (N=637) and the virologic failure group (N=1220) represent nearly 40% of the total of -4600 patients who ever received therapy in B . C . in this time period. Statistical Analyses Categorical variables were compared between groups using Pearson's chi-squared test. Contingency tables that contained one or more expected counts of less than five were 104 analyzed by the Fisher's exact test. The Cochran-Armitage test was used to assess resistance trends over time. Comparisons of continuous variables were carried out using Wilcoxon's rank-sum test. A l l reported p-values are two-sided. 105 5.3 R E S U L T S Causes of mortality and antiretroviral therapy in the study group A total of 637 deaths were recorded among HIV-infected individuals enrolled in a province-wide H I V / A I D S Drug Treatment Program between July 1997 and December 2001 (Figure 1). O f these, 83 (13.0%) were attributed to accidental causes, the two most common causes (by ICD-9 or ICD-10 coding) being illicit drug overdoses (57.8%) and concussion (18.1%). The remaining accidental deaths included suicides, traffic accidents, assaults and other injuries. Although H I V disease may have played an indirect role in these deaths, accidental deaths were assumed not to be directly related to antiretroviral resistance and were not included in genotypic analyses. The remaining 554 deaths (87.0%) were attributed to non-accidental causes. Deaths included 383 (69.1%) directly related to H I V infection (ICD-9 categories 042-044 and ICD-10 categories B20-B24), 34 (6.1%) liver disease, 25 (4.5%) various cardiac conditions, 20 (3.6%) viral and/or bacterial infections, 18 (3.2%) malignant neoplasms, 44 (7.9%) other causes, and 31 (5.6%) unknown causes. O f the 554 non-accidental deaths, 58 (10.5%) occurred in individuals who had enrolled in the province's H I V / A I D S Drug Treatment Program, but who had never been dispensed any antiretrovirals. Plasma samples from these individuals were not genotyped, as resistance mutations would not likely have been present. The remaining 496 (89.5%) non-accidental deaths occurred in individuals with at least some exposure to antiretroviral agents. The number of non-accidental deaths per year among treated individuals increased slightly over the period of interest (Table 1). The median time elapsed between initiation of first antiretroviral therapy and date of death in this cohort increased over the study period, 106 from 28 months for deaths occurring in the last half of 1997, to 47 months for deaths occurring in 2001 (p<0.001). The median time actually spent on antiretroviral therapy increased from 11 months of antiretroviral exposure for deaths occurring in the last half of 1997, to 28 months of antiretroviral exposure for deaths occurring in 2001. Similarly, the median time spent on H A A R T increased from 4 to 15 months over the study period. The median time elapsed between stopping therapy and death remained constant over the study period, with death occurring a median of 2 months after discontinuation of therapy [IQR 0.5-7.4 months]. Although reasons for discontinuation of therapy are unknown, it is of interest to note that the majority of individuals (287 of 496, 58%) were not being prescribed antiretrovirals at time of death. O f the 496 non-accidental deaths among treated individuals, 99 (20.0%) occurred in individuals for whom no viral load data were available and for whom no plasma samples were available for resistance testing. The median duration between initiation of therapy and death in these individuals was 18 months [IQR 5-44 months], however, the median time actually spent on antiretrovirals was only 2 months [IQR 0-4 months]. It is therefore unlikely that broad drug resistance would have played a major role in their deaths, given their relatively short exposure to antiretroviral agents. The median C D 4 count before death in this group was 90 cells/mm 3 [IQR 20-250]. Genotypic Resistance Testing O f the 397 treated individuals for whom plasma samples were available (Table 2), 147 (37.0%) deaths occurred in individuals whose last on-therapy viral load was <500 copies H I V R N A / m l . Although it was not possible to genotype these samples, the fact that plasma H I V R N A was suppressed below 500 copies/ml suggests that drug resistance mutations were 107 not present in quantities sufficient to compromise therapy. O f interest, 60 (41%) of these 147 individuals discontinued antiretrovirals for unknown reasons shortly after this viral load measurement, and were therefore not receiving any antiretrovirals at time of death. The remaining 250 (63.0%) deaths occurred in individuals whose last on-therapy viral load was >500 copies/ml; drug resistance genotyping was successful in 232 of these 250 samples (92.8%). The majority of the last on-therapy plasma samples (212 of 379, 55.9%) represented individuals with either successful suppression of plasma H I V R N A <500 copies/ml (N=147) or contained H I V harboring no major antiretroviral resistance mutations (N=65). The proportion of isolates with and without resistance-associated mutations remained relatively stable over the course of the study period (Figure 2A). A total of 167 samples (44.1%) harbored H I V with at least one major resistance mutation (Figure 2A) , an observation strongly driven by resistance to nRTIs (142 samples, 37.4%), specifically, resistance to lamivudine arising from the M184I /V mutation (101 samples, 26.6%). A total of 83 (21.9%) and 53 (13.9%) samples had mutations conferring resistance to N N R T I s and Pis, respectively (data not shown). Isolates harboring H I V with mutations conferring resistance to drugs from at least 2 classes and drugs from all 3 classes were observed relatively infrequently, at 23.4% and 5.8% of isolates, respectively (Figure 2A) . Individuals who started on mono or dual therapy were significantly more likely to harbor resistant H I V than individuals who received H A A R T exclusively (Figure 2B). Resistance to at least one class of drugs was observed in 138 of 253 individuals (54.5%) who started on mono or dual therapy, compared to 29 of 126 individuals (23.0%) who received H A A R T exclusively (p<0.001). Similarly, resistance to at least two drug classes was 108 observed in 77 individuals (30.4%) who started on mono or dual therapy, compared to 12 of 126 individuals (9.5%) who received H A A R T exclusively (p<0.001). O f interest, not a single individual who received H A A R T as their first therapy harbored H I V with resistance to all 3 antiretroviral classes. A l l individuals with 3-class resistance had initially received mono or dual therapy before receiving H A A R T (22 of 210, 10.4%) (p<0.001). Individuals with confirmed resistance mutations had a significantly longer median duration of therapy (32 months [IQR 16-52]) compared with individuals confirmed to have no resistance mutations (15 months [IQR 7-62]) (p<0.001). There were no other significant differences between these two groups (Table 2). Antiretroviral Resistance in individuals living with HIV and experiencing virologic therapy failure To put these observations into perspective, a virologic failure group was selected which included all living D T P participants with a p V L > 500 copies H I V R N A / m l over the same study period, for whom a physician-requested genotypic resistance test had been performed (N=1220). We observed a dramatically higher prevalence of antiretroviral resistance in the virologic failure group than in the mortality group. O f the N=1220 individuals in the virologic failure group, 924 (75.7%) harbored H I V with resistance to at least one drug class, compared to only 167 of 379 (44.1%) in the mortality group (p<0.0001). Similarly the prevalence of resistance to >2- and 3 drug classes was also significantly higher in the virologic failure group (42.3% and 11.1% of isolates, respectively) than in the mortality group (23.4% and 5.8% of isolates, respectively) (p<0.001). Resistant isolates in the mortality and virologic failure groups were similar with regards to the number of resistance mutations detected. Isolates from the virologic failure group with single-class 109 resistance had a median of 2 major resistance mutations, compared to a median of only one major resistance mutation in single-class resistant isolates from the mortality group. The number of resistance-associated mutations in isolates with 2 and 3-class antiretroviral resistance was the same for both groups (median 4 and 8 mutations, respectively). In contrast to the relatively stable prevalence of antiretroviral resistance in the mortality group (Figure 2A) , we observed a dramatic increase in the proportion of isolates in the virologic failure group with resistance to >2 classes of drugs over the study period (Figure 3). This was primarily due to an increase in N N R T I resistance among this group: in 1997, only 5.9% of isolates displayed N N R T I resistance, a figure that increased to 38% by 2001 (data not shown). Reliability of using a single plasma sample for Resistance Testing For the mortality group, resistance data were obtained from the last on-therapy plasma sample prior to death. It is possible, however, that resistance-associated mutations may have been present in earlier samples. In an attempt to determine the reliability of using this single sample, we analyzed retrospective data to assess the presence of resistance in earlier samples, i f available. Earlier genotype data were available for approximately half of patients. Despite this nonrandom sample, patterns of previous antiretroviral resistance were largely consistent with results from the last on-therapy genotype, with resistance to two or three drug-classes observed at some point in the individual's history in only 7% of those that were classified as "not resistant" and 12% of those with pVL<500 copies/ml at the last on-therapy genotype. 110 5.4 D I S C U S S I O N A N D C O N C L U S I O N S In order to determine the degree to which antiretroviral resistance may contribute to mortality among HIV-infected individuals in a province where antiretroviral medications and H I V care are provided at no cost through a government-sponsored program, the prevalence of antiretroviral resistance in two groups was investigated. The two groups were selected based on two independent definitions of antiretroviral therapy failure: a group comprising all HIV-infected individuals who had died over the period 1997-2001 (where "therapy failure" was defined as mortality) and a group comprising all living individuals for whom a resistance genotype had been requested over the same study period (where "therapy failure" was defined as a p V L > 500 copies H I V R N A / m l ) . Specifically, we wished to determine whether mortality in this population was primarily due to an exhaustion of available treatment options arising as a direct result of multi-drug resistance. The majority (56%) of individuals in the mortality group who received antiretroviral therapy harbored H I V with either no major resistance mutations or had plasma viral load suppression <500 copies H I V R N A / m l , suggesting that drug resistance was not sufficient to compromise therapy. Less than 25% of the mortality group harbored H I V with mutations conferring resistance to antiretrovirals from two or more drug classes, while only a minority (-5%) harbored H I V with mutations conferring resistance to drugs from all three classes. Of interest, triple-class resistance was not observed in any individual who received H A A R T exclusively. In the larger population, regardless of therapy access, we estimate that resistance to all three classes was present in <4% of the 554 individuals who died of non-accidental causes. I l l These results clearly suggest that exhaustion of available and effective treatment options, arising as a direct result of multi-drug resistance, was not a driving factor in the vast majority of deaths in this cohort. These results are somewhat in contrast to those from a recent study of 29 HIV-related deaths, that reported mutations conferring resistance to both H I V reverse transcriptase and protease inhibitors in 10 (34%) of the isolates [21]. The relatively low level of antiretroviral resistance observed in the mortality group contrasts to the high level of resistance observed in the virologic failure group. Consistent with the increased use of triple combination therapies over the period of interest, rates of two-and three-class resistance increased dramatically in the virologic failure group, while levels of multidrug resistance remained relatively stable in the mortality group over the same period. These results represent an interesting paradox: H I V drug resistance, and especially multi-class resistance, is common and increasingly prevalent over time in individuals experiencing virologic therapy failure who remain alive, but remains relatively uncommon in those who have died. Although the inverse association between antiretroviral resistance and mortality may seem counter-intuitive, the duration of antiretroviral therapy in the mortality group suggests that the low level of drug resistance was mainly driven by insufficient antiretroviral exposure. Among individuals who died, the median time elapsed between initiation of therapy and death was over three years; however the time actually spent on antiretrovirals was only 19 months. Given the large number of treatment possibilities today, 19 months is not long enough to have attempted and failed all possible therapy combinations. The discrepancy between these times indicates that in many individuals, antiretroviral therapies were used only intermittently or were discontinued. In fact, the majority of the 112 study population (including -40% of individuals whose last on-therapy viral load was <500 copies/ml) were not receiving any antiretrovirals at the time of death, having discontinued therapy for unknown reasons a median of two months previously. In the case of 10% of the study population, there was no antiretroviral exposure at all. Consistent with these observations are a recent study of HIV-infected individuals in British Columbia which reported a positive association between intermittent use of antiretroviral therapy and mortality [22], and a recent study of 88 HIV-related deaths occurring in 1999-2000 which reported that the majority of patients who died were not receiving H A A R T despite access to care [8]. In that study [8], an inability to tolerate H A A R T due to side effects or therapy toxicity, as well as non-adherence, were cited as the major reasons why patients were not receiving antiretrovirals at time of death. In addition, a recent study investigating barriers to antiretroviral access among 1239 HIV-infected British Columbians who died between 1995-2001, of which our cohort was a part, identified female gender, aboriginal ethnicity and low socioeconomic status as the strongest predictors of lack of therapy access, intermittent therapy use and/or discontinuation of therapy despite availability of antiretrovirals at no cost through a government-funded treatment program [17]. Although reasons for intermittent use of therapy are not known for individuals in our study, it is highly likely that side effects, illness, toxicity and other factors would have played a large role. Finally, a large study estimating the prevalence of antiretroviral resistance in the United States reported that -78% of HIV-infected individuals with viral loads >500 copies/ml harbored resistant H I V [23], a result which is entirely consistent with the observed prevalence of antiretroviral resistance in our virologic failure group. Most notably, these 113 investigators also reported a positive correlation between drug resistance and antiretroviral exposure; in this study the prevalence of drug resistance was significantly associated with greater access to treatment [23]. A limitation of this study is the fact that no viral load data or plasma samples were available for approximately 20% of deaths occurring in treated individuals. Although we cannot exclude the possibility that these individuals may have harbored drug-resistant H I V , the fact that these individuals only received a median of 2 months of therapy strongly suggests that drug resistance did not play a large role in their deaths. Additional limitations of this study include its retrospective nature, and the fact that reasons for interruption and/or discontinuation of therapy are largely unknown. Finally, it is important to note that the two different endpoints used (clinical failure as defined by mortality and virologic failure as determined by p V L > 500 copies/ml) results in groups which are not directly comparable. The mortality group includes all individuals who died, regardless of whether or not they received antiretrovirals, while the "virologic failure" group includes all individuals who have received therapy and maintained a p V L > 500 copies/ml. The total prevalence of resistance in all l iving HIV-infected individuals in British Columbia is likely much lower than in the "virologic failure" group presented here, as this group excludes those who consistently maintained pVL<500 copies/ml (approximately 1000 individuals), those with at least one pVL>500 copies/ml but for whom a genotypic resistance test was not requested by their physician (~1700 individuals), those lost to follow-up (~30 individuals), and those who have never received antiretrovirals (number unknown). The data do, however, show the tremendous difference in the prevalence and trends in resistance using these two different definitions of treatment failure. 114 In conclusion, results of this study strongly indicate that treatment failure due to antiretroviral resistance is not a major factor influencing mortality in this cohort for whom antiretrovirals were available at no cost. For most of the individuals studied, insufficient and/or intermittent exposure to antiretrovirals, comorbidities and other factors likely played a larger role. 115 Table 5.1: Number of deaths and antiretroviral treatment history in this cohort, by year. The considerable discrepancy between the median time elapsed between initiation of therapy and death, and the median time actually spent on antiretrovirals suggests that for many individuals, antiretroviral use was intermittent. Year of Death Number of Median months Median months spent Median months spent Median months Deaths elapsed between start on any antiretrovirals on HAART between stopping of therapy and death [IQR] [IQR] therapy and death [IQR] [IQR] 1997 (July-Dec) 45 28 [8-54] 11 [4-33] 4 [0-8] 2.5 [0.6-4.4] 1998 90 32 [19-60] 18 [7-33] 6 [0-13] 2.8 [0.6-8.0] 1999 105 35 [17-56] 15 [6-34] 8 [3-14] 1.8 [0.6-8.0] 2000 135 44 [28-71] 20 [8-44] 10 [4-22] 2.0 [0.5-8.5] 2001 121 47 [20-68] 28 [10-48] 15 [5-33] 1.6 [0.5-6.0] TOTAL 496 40 [19-62] 19 [7-41] 8 [2-18] 2.0 [0.5-7.4] OVERALL p<0.001 p=0.002 p<0.001 p=0.8 [IQR]: Interquartile range 116 Table 5.2: Comparison between those who died without any antiretroviral resistance compared to those who showed any antiretroviral resistance. No antiretroviral Virologic suppression Any antiretroviral p p p resistance (N=65) (N=147) resistance (N=167) (A) (B) (C) (A vs. B) (A vs. C) (B vs. C) Age at start of therapy 39.6(9.3) 42.4(10.3) 39.2(9.1) 0.039 0.767 0.002 (std dev) Age at death (std dev) 42.3(8.8) 45.8(10.6) 43.2(8.9) 0.008 0.247 0.031 %Male 81.2% 87.2% 87.4% 0.263 0.230 0.944 %IDU 31.3% 30.4% 29.3% 0.903 0.777 0.837 Median CD4 count 30 [10-160] 200 [70-370] 50 [10-200] <0.001 0.207 <0.001 (cells/mm3) [IQR] Median plasma viral 109,000 <500 49,000 <0.001 0.250 <0.001 load [3185-416,000] [5100-272,000] (copies HIV RNA/ml) [IQR] Median length of time 15 [7-62] 31 [13-47] 32 [16-52] <0.001 <0.001 0.153 on therapy (months) Adherence [IQR]* 78 [33-100] 100(73-100] 83 [58-100] <0.001 0.084 0.006 I D U : Injection Drug User IQR: Interquartile Range * Adherence defined as % of prescriptions refilled during first 12 months of therapy 117 Figure 5.1 (Next Page): Flowchart summarizing number of subjects in mortality group. Dotted boxes contain subsets of individuals whom antiretroviral resistance was either not detected in a genotypic resistance test, or for whom antiretroviral resistance was not likely to have contributed to death. Individuals in the "no follow-up p V L " group had only a very short duration of antiretroviral exposure; drug resistance therefore was not likely a relevant factor in their deaths. The thick solid box indicates individuals for whom antiretroviral resistance was confirmed in a genotypic test. 118 Deaths from all causes July 1997- December 2001 N=637 Accidental Death N=83 Non-Accidental Death N=554 Never received therapy N=58 Received therapy N=496 Last pVL<500 HIV RNA/ml N=147 i — J Last pVL>500 HIV RNA/ml N=250 I successfully genotyped N=232 No Antiretroviral Resistance N=65 _ _ _ _ _ _ . . J Antiretroviral Resistance N=167 119 « Figure 5.2A: Prevalence of antiretroviral resistance in the last on-therapy plasma sample, by year of death. Gray bars indicate isolates which contained no major resistance mutations or that represented samples from individuals with p V L <500 copies/mL. Hatched bars represent isolates with resistance to at least 1 drug class. White bars represent isolates with resistance to at least 2 drug classes, and black bars represent isolates with resistance to all three drug classes. Overall, the majority of isolates either contained no major resistance mutations or represented individuals with successful p V L suppression <500 copies/mL. Resistance to drugs from all three classes was rarely observed. 0.9 i 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 N=27 N=62 N=76 N=109 N=105 I—I 1997 (July-Dec) 1998 1999 2000 2001 120 Figure 5.2B: Proportion of resistant isolates, stratified by type of therapy received. Gray bars indicate isolates which contained no major resistance mutations or that represented samples from individuals with pVL<500 copies/mL. Hatched bars represent isolates with resistance to at least 1 drug class. White bars represent isolates with resistance to at least 2 drug classes, and black bars represent isolates with resistance to all three drug classes. Individuals who received mono or dual therapy at some point in their treatment history were significantly more l ikely to harbor H I V with resistance mutations than individuals who received H A A R T exclusively. None of the patients who received H A A R T exclusively had mutations conferring resistance to all three drug classes. 0.9 0.8 N=43 N=210 N=126 | 0.6 (A i 0.5 -I 0.4 -I 0.3 o * 0.2 0.1 0 mono/dual only mono/dual then HAART HAART only 121 Figure 5.3: Prevalence of antiretroviral resistance mutations, by year, in a group of living individuals enrolled in the Province's HIV/AIDS Drug Treatment program failing their current therapy (N=1220). Gray bars indicate isolates which contained no major resistance mutations or that represented samples from individuals failing therapy with plasma viral loads > 500 copies H I V R N A / m l . Hatched bars represent isolates with resistance to at least 1 drug class. White bars represent isolates with resistance to at least 2 drug classes, and black bars represent isolates with resistance to all three drug classes. Note the dramatic increase in the proportion of isolates with resistance to >2 classes of drugs over the study period. 122 5.5 R E F E R E N C E S 1. Palella FJ , Jr., Delaney K M , Moorman A C , et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. H I V Outpatient Study Investigators. N Engl J Med. 1998; 338:853-60. 2. Hogg R S , O'Shaughnessy M V , Gataric N , et al. Decline in deaths from A I D S due to new antiretrovirals. Lancet 1997; 349:1294. 3. Mocroft A , Ve l l a S, Benfield T L , et al. Changing patterns of mortality across Europe in patients infected with H I V - 1 . EuroSIDA Study Group. Lancet. 1998; 352:1725-30. 4. Lee L M , Karon J M , Selik R, Neal JJ and Fleming P L . Survival after A I D S diagnosis in adolescents and adults during the treatment era, United States, 1984-1997. J A M A 2001;285:1308-15. 5. Messeri P, Lee G , Abramson D M , Aidala A , Chiasson M A and Jessop DJ . Antiretroviral therapy and declining A I D S mortality in New York City. M e d Care. 2003;41:512-21. 6. Morris A , Wachter R M , Luce J, Turner J and Huang L . Improved survival with highly active antiretroviral therapy in HIV-infected patients with severe Pneumocystis carinii pneumonia. A I D S . 2003; 17:73-80. 7. Valdez H , Chowdhry T K , Asaad R, et al. Changing spectrum of mortality due to human immunodeficiency virus: analysis of 260 deaths during 1995-1999. C l i n Infec Dis. 2001;32:1487-93. 123 8. Jain M K , Skiest D J , Cloud JW, Jain C L , Burns D and Berggren R E . Changes in mortality related to human immunodeficiency virus infection: comparative analysis of inpatient deaths in 1995 and in 1999-2000. C l i n Infec Dis . 2003; 36:1030-8. 9. Kuritzkes D R . Clinical significance of drug resistance in HIV-1 infection. A I D S . 1996; 10:S27-31. 10. Mi l le r V , Phillips A , Rottmann C , et al. Dual resistance to zidovudine and lamivudine in patients treated with zidovudine-lamivudine combination therapy: association with therapy failure. J Infec Dis . 1998; 177:1521-32. 11. Wainberg M A , Friedland G . Public health implications of antiretroviral therapy and H I V drug resistance. J A M A . 1998; 279:1977-83. 12. Petrella M , Brenner B , Loemba H and Wainberg M A . H I V drug resistance and implications for the introduction of antiretroviral therapy in resource-poor countries. Drug Resist Updat. 2001; 4:339-46. 13. Pillay D , Taylor S and Richman D D . Incidence and impact of resistance against approved antiretroviral drugs. Rev Med Vi ro l . 2000; 10:231-53. 14. Pil lay D . The emergence and epidemiology of resistance in the nucleoside-experienced HIV-infected population. Antivir Ther. 2001; 6:15-24. 15. Harrigan PR, Cote H C . Clinical utility of testing human immunodeficiency virus for drug resistance. C l i n Infec Dis. 2000; 30:S 117-22. 16. DeGruttola V , D i x L , D 'Aqui la R, et al. The relation between baseline H I V drug resistance and response to antiretroviral therapy: re-analysis of retrospective and prospective studies using a standardized data analysis plan. Antivir Ther. 2000; 5:41-8. 124 17. Wood E , Montaner JS, Tyndall M W , et al. Prevalence and correlates of untreated human immunodeficiency virus type 1 infection among persons who have died in the era of modern antiretroviral therapy. J Infect Dis 2003; 188(8): 1164-70. 18. Low-Beer S, Y i p B , O'Shaughnessy M V , Hogg RS and Montaner JS. Adherence to triple therapy and viral load response. J Acquir Immune Defic Syndr. 2000; 23:360-1. 19. Alexander C S , Dong W , Chan K , et al. H I V protease and reverse transcriptase variation and therapy outcome in antiretroviral-naive individuals from a large North American cohort. A I D S . 2001; 15:601 -7. 20. D 'Aqui la R T , Schapiro J M , Brun-Vezinet F, Clotet B , Conway B , Demeter L M , et al. Drug resistance mutations in H I V - 1 . Topics in H I V Medicine 2002; 10:21-25. 21. Beinker N K , Mayers D L , Lange C G , et al. Genotypic drug resistance and cause of death in HIV-infected persons who died in 1999. J Acquir Immune Defic Syndr. 2001; 28:250-3. 22. Hogg R S , Heath K , Bangsberg D , et al. Intermittent use of triple-combination therapy is predictive of mortality at baseline and after 1 year of follow-up. A I D S . 2002; 16:1051-8. 23. Richman D D , Bozzette S, Morton S, et al. The prevalence of antiretroviral drug resistance in the U . S . 41st International Conference on Antimicrobial Agents and Chemotherapy, Chicago, December 2001. [Abstract LB-17] 125 CHAPTER 6 HIV GENETIC VARIATION: IMPLICATIONS OF HIV p6 G a g INSERTIONS ON RESPONSE TO HAART 1 6.1 I N T R O D U C T I O N The emergence of antiretroviral-resistant strains of Human Immunodeficiency Virus Type-1 (HIV-1) is a major factor influencing the effectiveness of antiretroviral therapy. Historically, antiretroviral resistance has been associated with mutations in the protease and reverse transcriptase genes, whose protein products are the targets of currently available antiretroviral agents. However, genetic alterations outside these regions may also contribute to drug resistance. For example, mutations at the H I V gag protease cleavage sites have been associated with reduced susceptibility to protease inhibitors [1,2]. More recently, insertion mutations within the H I V p 6 G a g gene, located at the C-terminal end of the gag immediately upstream of pol, have been implicated in H I V antiretroviral resistance [3]. The p 6 G a g protein, involved in the process of H I V budding from infected cells [4-7], contains a highly conserved, amino-terminal proline-rich region referred to as the L-domain or " P T A P " motif. In the late stages of the viral life cycle, the P T A P motif appears to recruit the cellular protein tumor susceptibility gene (Tsg)- lOl , involved in the regulation of intracellular protein trafficking, to sites of H I V assembly at the plasma membrane [8-10]. The interaction of H I V p 6 G a g with Tsg-101 is believed to be required for the release of new infectious viral particles from the host cell surface [8,9], in a cell-type-dependent manner [4,5]. This chapter has been published as: Brumme Z L , Chan K J , Dong W W , Wynhoven B , M o T , Hogg R S , Montaner JS , O'Shaughnessy M V , Harrigan P R . "Prevalence and clinical implications of insertions in the HIV-1 p6Gag N-terminal region in drug-naive individuals initiating antiretroviral therapy". Ant iv i ral Therapy, 2003; 8:91-6. 126 A recent study of the p 6 G a g region in 296 HIV-infected individuals reported that a duplication of the proline-rich P T A P motif occurred relatively frequently in individuals treated with nucleoside analog reverse transcriptase inhibitors (NRTIs), but only rarely in drug naive individuals without associated resistance mutations, suggesting a link between the P T A P motif and resistance to NRTIs [3]. Furthermore, clones generated with the P T A P duplication exhibited increased infectivity and a competitive growth advantage in the presence of NRTIs in vitro [3]. However, an independent study of a small cohort of treatment-experienced individuals reported only a weak correlation between insertions in the proline-rich p 6 G a g motif and a lower risk of virological failure, which was not statistically significant [11]. The objective of this study was to assess the prevalence and influence of amino acid insertions within the H I V p 6 G a g N-terminal proline-rich region on initial virological and immunological response to antiretroviral therapy in a large, population-based cohort of H I V -infected, drug naive individuals initiating their first antiretroviral therapy. 127 6.2 M A T E R I A L S A N D M E T H O D S Study Design and Patients In the province of British Columbia, Canada, antiretroviral drugs are distributed free of charge to eligible HIV-infected individuals through a centralized drug treatment program based at the British Columbia Centre for Excellence in H I V / A I D S ("the B . C . Centre") [see reference 12 for details of policies and treatment guidelines]. A l l antiretroviral naive individuals first seeking treatment at the B . C . Centre between June 1, 1996 and August 31, 1998, were eligible for analysis (N=479). O f these, 461 (96.2%) had a pre-therapy plasma sample available for testing. Greater than 80% of patients were prescribed triple therapy including two nucleoside analogues and one protease inhibitor, consistent with treatment guidelines at the time. Ethical approval for this study was obtained from the institutional ethics board at St. Paul's Hospital, Vancouver. Samples were assigned anonymous identifiers prior to genotypic analyses. Plasma Viral Load and CD4 Cell Count Plasma HIV-1 R N A levels (viral loads) were determined using the Roche Amplicor Monitor assay (Roche Diagnostics, Laval , Quebec, Canada) using either the standard method or the ultrasensitive adaptation. The ultrasensitive adaptation to this assay was not routinely available over the time period of this study; for this reason a constant cut-off of 500 H I V R N A copies/ml was used throughout the analysis. C D 4 cell counts were measured by flow cytometry, followed by fluorescent monoclonal antibody analysis (Beckman Coulter, Inc., Mississauga, Ontario, Canada). 128 Detection of HIV pfj0"8 PTAP insertions, HIV envelope V3 loop sequences, and HIV-RNA genotyping for drug resistance. Plasma nucleic acids were extracted and precipitated with guanidine thiocyanate and isopropanol, followed by two ethanol washes. The H I V p 6 G a g amino-terminal region (located in the C-terminal region of gag, directly upstream of pol) was amplified by nested R T - P C R , using forward primer 5' G C C C C T A G G A A A A A G G G C T G T T G G 3' and reverse primer 5' T T C T G T C A A T G G C C A T T G T T T A A C 3' in the first round, and forward primer 5' C C T A G G A A A A A G G G C T G T T G A A A T G T G 3' and reverse primer 5' C C T G G C T T T A A T T T T A C T G G T A C A G 3' in the second round, to generate a final 0.6 Kb product. Note that mutations and/or insertions in the C-terminal region of the H I V p 6 G a g were not included in the analysis presented here. The region encoding the H I V V 3 loop within the env gene was amplified by nested R T - P C R , using forward primer 5' C A G A A A A A T T G T G G G T C A C A G T C T A T T A T G G G G T A C C T 3' and reverse primer 5' T A T A A T T C A C T T C T C C A A T T G T C C 3' in the first round, and forward primer 5' A C A C A T G C C T G T G T A C C C A C A G A 3' and reverse primer 5' A T T A C A G T A G A A A A A T T C C C C T C C A C 3' in the second round, to generate a 0.95 Kb product. P C R amplification products were sequenced directly in both the 5' and 3' directions using the BigDye dye terminator cycle sequencing kit (Applied Biosystems, Foster City, C A ) and resolved on an A B I 3700 automated sequencer. H I V drug resistance genotyping was performed as previously described [13]. Statistical Analysis Time to virological success was defined as the time to achieve a plasma viral load (pVL) < 500 copies HIV-1 R N A / m l from baseline. Time to virological failure was defined as 129 the time elapsed between virological success and the second of two consecutive p V L measurements > 500 copies/ml. If a person never achieved virological success, they were excluded from the analysis of time to virological failure. If a patient reached therapeutic success and did not subsequently fail then they were censored at the time of their last viral load date (occurring prior to December 1, 2000). Time to immunological failure was defined as the time to the second of two successive C D 4 cell counts below baseline. Event-free subjects were right censored as of December 1, 2000, or at the time of their last viral load or C D 4 cell test. The influence of H I V p 6 G a g N-terminal insertions on time to virological success and virological and immunological failure were assessed by Kaplan-Meier survival analysis, over a median follow-up time of 40 months. Cox proportional hazard regression was used to calculate univariate and adjusted risk ratios and 95% confidence intervals (CIs). In this analysis, gender, prior diagnosis of A I D S , initiation of triple therapy, detection of H I V p 6 G a g insertion, 100% adherence (evaluated using a previously validated approach based on the frequency of prescriptions filled on time during the first year of therapy, where an adherence level of 100% was defined as fill ing all prescriptions on schedule [14]) and history of injection drug use were tested as binary variables. Age, p V L , and C D 4 cell count were treated as either ordinate or continuous variables. A l l tests of significance were two-sided, with a p-value of less than 0.05 indicating statistical significance. 130 6.3 R E S U L T S Prevalence of HIV pCo0"8 insertions and association with baseline parameters O f the 423 of 461 drug-naive individuals (91.8%) who were successfully genotyped, 70 (16.5%) contained a p 6 G a g N-terminal amino acid insertion (Table 1). N-terminal p 6 G a g sequences have been deposited in GenBank (Accession numbers: AF538974-AF539396). The most commonly observed insertion was " P T A " (34/70, 48.5%). Note that because different equivalent alignments are possible in some cases, the classification of insertions is somewhat arbitrary. For example, " P T A " may be designated " T A P " or " A P P " , depending on the alignment chosen. Baseline demographic characteristics of the sample population were generally evenly matched between those with or without H I V p 6 G a g insertions (Table 2). H I V p 6 G a g insertions were significantly associated with a lower median baseline C D 4 count (p=0.002). Association between HIV p(Fas insertions and HIV envelope The majority of the study group (382 of 423 patients, or 90.3%) were additionally genotyped for their H I V envelope V 3 loop sequence. Briefly, nested R T - P C R and direct D N A sequencing were used to identify basic amino acid residues at positions 11 and 25 of the V3 loop, which have been associated with an H I V syncytium-inducing (SI) phenotype [15,16]. A significant association between the presence of a p 6 G a g insertion and the presence of basic amino acid residues in the H I V V3 loop was observed (p6 G a g insertion observed in 50 of 335 inferred non-syncytium-inducing [NSI] viruses vs. 17 of 47 inferred SI viruses, p=0.0003). O f interest, although the presence of an H I V p 6 G a g insertion was significantly associated with a lower baseline C D 4 count (p=0.002, table 2), there was no association between H I V V 3 sequence and baseline C D 4 count (p=0.2, data not shown). 131 A search of the HIV-1 complete genome sequences available from the Los Alamos H I V sequence database indicated a prevalence of H I V p 6 G a g insertions of approximately 10.5%, however no significant association was observed between presence of p 6 G a g insertions and inferred H I V tropism in this data set (p>0.1, data not shown). Association between HIV p6°a& insertion and baseline antiretroviral resistance The relationship between H I V p 6 G a g insertions and antiretroviral resistance [3] was also investigated. Baseline genotypic resistance analysis (performed as described in [13] using the same pre-therapy plasma sample) was completed for all 423 samples. Baseline antiretroviral resistance was defined as the presence of one or more of the following key mutations: D 3 0 N , M46I /L , G 4 8 V , I50V, V 8 2 A / F / T / S , I84V, L 9 0 M in the H I V protease, or M 4 1 L , K65R, T69D, K 7 0 R , L 7 4 V , K103N, Q151M, Y181C/I , M 1 8 4 V / I or T215Y/F in the H I V reverse transcriptase (adapted from [17]). According to this definition, baseline resistance mutations were observed in 15 of 423 samples (3.5%): 3 M 4 1 L , 4 M 1 8 4 V (including one sample with both M l 8 4 V and L 9 0 M ) , 4 M 4 6 L 1 M46I , 1 I50V, 1 T69D and 1 V 8 2 A . There was no association between the presence of H I V p 6 G a g insertions and the presence of resistance-associated mutations at baseline (p=0.7, data not shown). Association ofHlVp6°ag insertion with initial virological and immunological treatment response In this study, individuals were initially therapy naive, but generally the time since seroconversion was not known. Some individuals may have been infected with drug-resistant virus, and it is conceivable that over time, drug resistance mutations in the protease and/or R T may have reverted to wild-type, while the p 6 G a g insertion did not. For this reason, 132 we investigated the effect of insertions in the H I V p 6 G a g on initial virological and immunological therapy response (Figure 1). There was no influence of H I V p 6 G a g insertions on time to virological success (p=0.8, data not shown), virological failure (p=0.5), or immunological failure (p=0.7). The proportion of individuals never achieving virological success was 17.1% and 20.4% for the groups with and without p 6 G a g insertions, respectively (p=0.5). Cox proportional hazard regression was used to calculate univariate and multivariate risk ratios (R.R.) and 95% confidence intervals (CIs). In multivariate analyses, time to virological and immunological failure were significantly associated with initiating triple therapy [virological R .R . 0.41; 95%CI 0.27-0.63, immunological R .R . 0.61; 95%CI 0.39-0.95] and adherence [virological R .R. 0.28; 95%CI 0.20-0.41, immunological R.R. 0.30; 95%CI 0.21-0.45]. Consistent with the Kaplan-Meier analyses, there was no influence of N -terminal H I V p 6 G a g insertions on time to virological failure [R.R. 1.08; 95%CI 0.70-1.67], time to immunological failure [R.R. 1.19; 95%CI 0.75-1.89] or time to virological success [RR 1.06; 95%CI 0.78-1.45] (p>0.1, data not shown). 133 6.4 D I S C U S S I O N A N D C O N C L U S I O N S This study suggests that N-terminal insertions in H I V p 6 G a g are much more common in drug naive individuals than previously reported [3]. Almost 17% of this study population had p 6 G a g insertions, compared to a previous study that identified p 6 G a g insertions in 5.4% of drug naive individuals [3]. These results are consistent with a recent study that reported a p 6 G a g insertion prevalence of 17.6% in an antiretroviral naive population [18]. In contrast to a previous study suggesting a link between p 6 G a g P T A P insertions and transmission of drug-resistant H I V [3], we observed no association between p 6 G a g insertions and the presence of antiretroviral resistance mutations at baseline. These data are consistent with a recent study of 100 individuals examined during primary H I V infection, which reported that despite an increase in frequency of an H I V p 6 G a g motif duplication, which paralleled an increase in incidence of infection with drug resistant strains over the course of study follow-up, there was no direct association between p 6 G a g insertions and primary-resistance-associated mutations [19]. Importantly, after adjustment for baseline factors, there was no significant influence of N-terminal H I V p 6 G a g insertions on virological response to antiretroviral therapy, suggesting that there may be relatively little effect of these insertions on H I V drug resistance per se. Similarly, we were unable to demonstrate any correlation between H I V p 6 G a g insertions and immunological therapy response. It must be noted, however, that these insertions could have had a small impact on resistance that did not result in significant differences in virological and immunological response to therapy. Interestingly, a previous study of 42 highly treatment-experienced individuals reported a trend, although not statistically significant, between p 6 G a g insertions and improved virological response to therapy [11]. 134 The observation that p 6 G a g insertions were significantly associated with a lower baseline C D 4 cell count suggests that p 6 G a g insertions may indeed be associated with a poorer prognosis and/or more rapid disease progression in untreated patients. The mechanism for this association may be direct, through an increase in viral packaging efficiency in H I V containing a p 6 G a g insert [3], or as a result of variable interactions of p 6 G a g with host cell factors. It has been recently suggested that Tsg-101, a protein normally involved in intracellular trafficking, interacts with the H I V p 6 G a g P T A P motif at a late stage in the H I V life cycle in an event essential to viral budding and egress [8-10]. Since it has been shown that mutations within the conserved p 6 G a g P T A P motif inhibit both Tsg-101 binding and H I V particle release from the cell surface [9,10], it is conceivable that duplication of the P T A P motif may increase H I V infectivity by enhancing interactions with Tsg-101. Alternatively, the mechanism for this association may be indirect, through association with other naturally occurring mutations at various sites within the H I V genome. Indeed, in this study, basic amino acid residues at key positions in the H I V V 3 loop sequence, indicative of an H I V syncytium-inducing phenotype which is associated with entry into C X C R 4 co-receptor-expressing cells [20], and is associated with depletion of CD4-positive cells and more rapid disease progression [15, 16, 21, 22], were 2.4-fold more likely to be observed in individuals with p 6 G a g insertions, although analysis of the available HIV-1 genome sequences from the Los Alamos Sequence database did not reveal similar correlations between p 6 G a g insertions and inferred viral phenotype. It is possible that H I V p 6 G a g insertions may arise naturally, as a result of the increasing diversity of HIV-1 sequences over time, in a manner analogous to the shift from the non-syncytium-inducing (NSI) to the syncytium-inducing (SI) phenotype in later stages 135 of H I V disease progression. Since neither the date of seroconversion nor reasons for starting antiretroviral therapy are known for most individuals, it is conceivable that the association of H I V p 6 G a g insertions with SI virus and a poorer baseline prognosis may simply be attributed to increased time since seroconversion and/or natural disease progression and viral evolution in these individuals. In conclusion, results from this study suggest that H I V p 6 G a g insertions may arise naturally in drug naive individuals, rather than being directly associated with resistance to antiretrovirals, and are not predictive of antiretroviral therapy response. The association with V3 loop sequences merits further investigation. 136 Table 6.1: Distribution of N-terminal HIV p6G a g insertions observed in study group of antiretroviral naive individuals H I V p 6 G a g : N-terminus - L Q S R P E P T A P P E E S F - C-terminus Amino A c i d Insertion N P T A 34 P T A P P E 6 S R P E 4 L E P T A P 3 P T A P P A 3 Q S R P E 2 P R P E 2 P P A 2 Q S R L E 1 Q S R T E 1 Q N R P E P T A P 1 P R P E P T A 1 R P E P T A 1 L E P T A 1 A P V P T A P P 1 P T G P E E 1 P T G P P E 1 P S A 1 P P 1 V P T 1 ( V / A ) P T 1 E E S R 1 Total 70 The first 15 N-terminal amino acids in the wild-type H I V p 6 G a g region are indicated above the table. Insertions are listed according to their observed frequency. In some cases, insertions are somewhat arbitrarily named, depending on the choice of different equivalent alignments. 137 Table 6.2: Patient Characteristics at baseline Patient Characteristics All patients (N=461) Stratified by presence or absence of HIV p6 G a g insertions (N=423 patients genotyped successfully) insert no insert (N=70) (N=353) Gender: % Male 84.5% 88.6% 83.7% 0.31 Median Age, years 36 34 36 0.31 [IQR] [31-42] [30-41] [31-42] HIVpVL (x103 copies/ml) 89 130 90 0.13 [IQR] [30-280] [47-420] [33-280] CD4 count (x106 cells/L) 310 195 320 0.002 [IQR] [140-460] [100-370] [170-470] baseline AIDS diagnosis 49 9 34 0.42 (%) (10.6%) (12.9%) (9.6%) triple therapy 385 58 293 0.98 (%) (83.5%) (82.9%) (83.0%) 100% adherent 240 32 183 0.35 (%) (52.1%) (45.7%) (51.8%) [IQR= Interquartile Range] 138 Figure 6.1: Kaplan-Meier survival analyses, showing the influence of insertions within the N-terminal region of the HIV p6G a g on time to virological (a) and immunological (b) failure. To be included in the "time to virological failure" and "time to immunological failure" analyses, patients had to register at least one post-baseline H I V p V L < 500 copies/ml, and at least one post-baseline C D 4 cell count, respectively. The univariate p-values for the virological and immunological analyses were p=0.5 and p=0.7, respectively. The number of patients in each group (insertions vs. no insertions) still experiencing virological and immunological success at 24 months were [28 (insertion), 148 (no insertion)] and [28 (insertion), 167 (no insertion)], respectively. 139 6.5 R E F E R E N C E S 1. Doyon L , Croteau G , Thibeault D , Poulin F, Pilote L , Lamarre D . Second locus involved in human immunodeficiency virus type 1 resistance to protease inhibitors. Journal of Virology 1996; 70:3763-3769. 2. Zhang Y M , Imamichi H , Imamichi T, et al. Drug resistance during indinavir therapy is caused by mutations in the protease gene and in its Gag substrate cleavage sites. Journal of Virology 1997; 71:6662-6670. 3. Peters S, Munoz M , Yer ly S, et al. Resistance to nucleoside analog reverse transcriptase inhibitors mediated by human immunodeficiency virus type 1 p6 protein. Journal of Virology 2001; 75:9644-9653. 4. Demirov D G , Orenstein J M , Freed E O . The late domain of human immunodeficiency virus type 1 p6 promotes virus release in a cell type-dependent manner. Journal of Virology 2002; 76:105-117. 5. Dettenhofer M , Y u X F . Proline residues in human immunodeficiency virus type 1 p6(Gag) exert a cell type-dependent effect on viral replication and virion incorporation of Pol proteins. Journal of Virology 1999; 73:4696-4704. 6. Gottlinger H G , Dorfman T, Sodroski J G , Haseltine W A . Effect of mutations affecting the p6 gag protein on human immunodeficiency virus particle release. Proceedings of the National Academy of Sciences - U S A 1991; 88:3195-3199. 7. Y u X F , Dawson L , Tian C J , Flexner C, Dettenhofer M . Mutations of the human immunodeficiency virus type 1 p6Gag domain result in reduced retention of Pol proteins during virus assembly. Journal of Virology 1998; 72:3412-3417. 140 8. Garrus JE , von Schwedler U K , Pornillos O W , et al. T s g l O l and the vacuolar protein sorting pathway are essential for HIV-1 budding. Cel l 2001; 107:55-65. 9. Martin-Serrano J, Zang T & Bieniasz P D . HIV-1 and Ebola virus encode small peptide motifs that recruit T s g l O l to sites of particle assembly to facilitate egress. Nature Medicine 2001; 7:1313-1319. 10. VerPlank L , Bouamr F, LaGrassa, TJ , et al.. Tsg 101, a homologue of ubiquitin-conjugating (E2) enzymes, binds the L domain in H I V type 1 Pr55(Gag). Proceedings of the National Academy of Sciences - U S A 2001; 98:7724-7729. 11. Kaufmann G R , Suzuki K , Cunningham P, et al. Impact of H I V type 1 protease, reverse transcriptase, cleavage site, and p6 mutations on the virological response to quadruple therapy with saquinavir, ritonavir, and two nucleoside analogs. A I D S Research and Human Retroviruses 2001; 17:487-497. 12. Hogg R S , Y i p B , Chan K J , Wood E , et al. Rates of disease progression by baseline C D 4 cell count and viral load after initiating triple-drug therapy. Journal of the American Medical Association 2001; 286:2568-2577. 13. Hertogs, K , de Bethune M P , Mi l le r V , et al. A rapid method for simultaneous detection of phenotypic resistance to inhibitors of protease and reverse transcriptase in recombinant human immunodeficiency virus type 1 isolates from patients treated with antiretroviral drugs. Antimicrobial Agents and Chemotherapy 1998; 42:269-276. 14. Low-Beer S, Y i p B , O'Shaughnessy M V , Hogg R S , Montaner JS. Adherence to triple therapy and viral load response. Journal of Acquired Immune Deficiency Syndrome 2000; 23:360-361. 141 15. Fouchier R A , Groenink M , Kootstra N A , et al. Phenotype-associated sequence variation in the third variable domain of the human immunodeficiency virus type 1 gp 120 molecule. Journal of Virology 1992; 66:3183-3187. 16. Fouchier R A , Brouwer M , Broersen S M , Schuitemaker H . Simple determination of human immunodeficiency virus type 1 syncytium-inducing V 3 genotype by P C R . Journal of Clinical Microbiology 1995; 33:906-911. 17. D 'Aqui la R T , Schapiro J M , Brun-Vezinet F, et al. Drug Resistance Mutations in H I V -1. Topics in H I V Medicine 2002; 10:21-25. 18. Gallego O, De Mendoza C, Corral Z , Soriano V . Changes at the HIV-1 p6 Protein among Antiretroviral-Naive and Pre-Treated patients. 4 2 n d Interscience Conference on Antimicrobial Agents and Chomotherapy, San Diego, California, September 27-30/2002. Abstract H-2053. 19. Smith D , Little SJ, Daar E S , et al. Genetic Variation in p6gag in Primary H I V infection. 9th Conference on Retroviruses and Opportunistic Infections, Seattle, Washington, February 24-28, 2002. Abstract 357-M. 20. Bjorndal A , Deng H , Jansson M , et al. Coreceptor usage of primary human immunodeficiency virus type 1 isolates varies according to biological phenotype. J V i ro l 1997;71:7478-87. 21. Koot M , Keet IP, Vos A H , et al. Prognostic value of HIV-1 syncytium-inducing phenotype for rate of CD4+ cell depletion and progression to A I D S . Annals of Internal Medicine 1993; 118:681-688. 22. Spijkerman I, de W o l f F, Langendam M , Schuitemaker H , Coutinho R. Emergence of syncytium-inducing human immunodeficiency virus type 1 variants coincides with a 142 transient increase in viral R N A level and is an independent predictor for progression to A I D S . Journal of Infectious Diseases 1998; 2:397-403. 143 CHAPTER 7 HIV GENETIC VARIATION: RELEVANCE OF HIV V3 ENVELOPE GENOTYPE TO HAART OUTCOMES 1 7.1 I N T R O D U C T I O N The introduction of Highly Active Antiretroviral Therapy ( H A A R T ) has resulted in dramatic decreases in morbidity and mortality in HIV-infected individuals [1-3]. However, due to concerns about long-term tolerability, drug resistance, cost and other factors, the optimal time to initiate H A A R T remains controversial [4-6]. A dominant prognostic factor influencing H I V disease progression and mortality in antiretroviral-nai've individuals initiating H A A R T is the baseline C D 4 cell count [5-7]. However, evidence suggests that the baseline H I V plasma viral load is a weaker independent predictor of clinical outcomes after starting H A A R T [5,6]. In order to continue to develop clinically relevant tools for H I V / A I D S management in the H A A R T era, it is important to identify additional predictors of treatment outcome. The emergence of syncytium-inducing (SI) H I V , which arises in approximately 50% of untreated, H I V subtype B-infected individuals with late-stage disease, is associated with a rapid decline in C D 4 T-cells and reduced survival time [8-11]. SI phenotype has also been associated with poorer response to mono or dual therapy [12-14] and poorer short-term responses to triple therapy in clinical trials [15,16], though some studies have reported no effect [17]. It has become clear that SI phenotype change is generally associated with a change in the preferred ' This chapter has been published as: Brumme Z L , Dong WW, Yip B, Wynhoven B, Hoffman N G , Swanstrom R, Jensen M A , Mullins JI, Hogg RS, Montaner JS, Harrigan PR. "Clinical and immunological impact of HIV envelope V3 sequence variation after starting initial triple antiretroviral therapy". AIDS. 2004; 18:F1 -9. 144 virus co-receptor, from C C R 5 to C X C R 4 [18]. Simple nucleic acid sequence-based methods exist to predict H I V phenotype or co-receptor usage from the H I V envelope third variable loop ("V3") sequence [19-23]. In general, SI variants (and CXCR4-coreceptor-using, or X 4 variants) display positively charged amino acid residues at codons 11 and/or 25 within the V 3 loop, while NSI variants (and CCR5-coreceptor-using, or R5 variants) generally do not [19, 22-24]. This "11/25" rule is one of the most reliable methods for sequence-based predictions of viral phenotype [25]. Other bioinformatic approaches for predicting phenotype from sequence data have been developed, including a published neural network approach (NN) [25], and a Position Specific Scoring Matrix approach (PSSM) [26]. Unfortunately, directly determining H I V phenotype and/or co-receptor usage remains too time consuming and costly for routine clinical testing, so these measures are rarely employed outside of clinical trials. In contrast, nucleic acid sequence-based assays are more practical and widely employed in clinical monitoring of H I V [27]. The long-term impact of V3 sequence variation on plasma viral load suppression, drug resistance, immunological response and mortality after initiation of H A A R T is not known. We therefore conducted the present study to evaluate directly the impact of H I V V3 sequence variation on therapy outcomes in the H O M E R cohort, a population-based cohort of 1191 HIV-infected individuals initiating H A A R T . 145 7.2 M A T E R I A L S A N D M E T H O D S British Columbia HIV/AIDS Drug Treatment Program In the province of British Columbia, Canada, antiretrovirals are distributed free of charge to HIV-infected individuals through a centralised drug treatment program. Antiretrovirals are prescribed according to specific guidelines set by the B . C . Therapeutic Guidelines Committee, which are revised regularly and are in accordance with international guidelines [5]. Patient selection The study population is very similar to a cohort that has been described in detail previously [5]. Briefly, all HIV-positive, antiretroviral naive adults who started triple H A A R T (consisting of 2 nRTI and either a PI or an N N R T I ) in B . C . between 08/1996 and 09/1999 were eligible (N=1312). O f these, 121 (9.2%) were excluded for not having baseline C D 4 and plasma H I V R N A data available within 6 months prior to the start of triple H A A R T ; the study sample was therefore made up of the remaining 1191 subjects. These 1191 subjects form an open treatment-based cohort known as the H O M E R ( H A A R T Observational Medical Evaluation and Research) cohort. Ethical approval was obtained from the institutional ethics board. Samples were assigned anonymous code identifiers before analysis. Laboratory Methods Plasma HIV-1 R N A loads were determined using the Roche Amplicor Monitor assay using either the standard or ultrasensitive methods. The ultrasensitive assay was not routinely available over the study period; a constant cut-off of 500 H I V R N A copies/ml was therefore used in outcome analyses. C D 4 counts were measured by standard techniques. 146 Plasma nucleic acids from each pre-therapy sample were extracted using standard methods and a portion of the HIV-1 envelope containing the V 3 loop was amplified by nested R T - P C R , using first round primers 5' C A G A A A A A T T G T G G G T C A C A G T C T A T T A T G G G G T A C C T 3' (HIV-1 H X B 2 acc# K03455, nt 6316-6353) and 5' T A T A A T T C A C T T C T C C A A T T G T C C 3' (nt 7675-7652), and second-round primers 5' A C A C A T G C C T G T G T A C C C A C A G A 3' (nt 6435-6457) and 5' A T T A C A G T A G A A A A A T T C C C C T C C A C 3' (nt 7382-7357). If no product was amplified, an alternative second-round forward primer was used (5' A A T G T C A G C A C A G T A C A A T G T A C A C 3') (nt 6945-6969). P C R products were sequenced in both 5' and 3' directions using the BigDye dye terminator cycle sequencing kit (Applied Biosystems) and run on an A B I 3100 or 3700 automated sequencer. Data Analysis: a) Alignment of V3 sequences The majority of sequences in our study encoded 35 amino acid residues. However to maintain alignment, gaps needed to be inserted in approximately 10% of sequences. The final alignment was performed manually, based on the amino acid profile of a large number (>400) of independent V 3 loops [25]. Sequences were aligned to the following motifs: (T|A)(T|A) (that is, either threonine or alanine) at each of codons 22 & 23, (G) at codon 24, and (I|V)(I|V|T) at codons 26 and 27. A n y charged residue following (G)24 or preceding (I|V)(I|V|T) 26-27 was assigned to site 25. If no charged residue intervened between (G)24 and (I|V)(I|V|T) 26-27, site 25 was gapped. 147 b) Interpretation ofV3 sequences In the primary analysis, H I V envelope V3 loop sequences were dichotomized by the presence of positively charged residues at codons 11 and/or 25 of the V 3 loop [19]. Because we were interested in evaluating the predictive value of V3 sequence (rather than in-vitro phenotypes or cp-receptor usage) on therapy outcome, we elected to use the following nomenclature: H I V isolates who displayed positively charged residues at codons 11 and/or 25 of V 3 were classified as having an "11/25" genotype. In addition, we employed a neural network (NN) [25] as well as a Position-Specific Scoring Matrix (PSSM) [26] to categorize V3 sequences. V 3 sequence predictions inferred using N N and P S S M are indicated in lowercase letters ("si", "nsi") [24] to distinguish them from experimentally determined phenotypes. Nucleotide mixtures are sometimes observed upon direct D N A sequencing of clinical isolates. In these cases, nucleic acid sequences from each patient were translated into amino acid (aa) sequences containing all possible aa combinations. However, the translation of H I V sequences into all possible amino acid sequence combinations could potentially have resulted in the artificial disruption of natural sequence linkage patterns and the subsequent interpretation of amino acid combinations that would not naturally exist in vivo. We applied an arbitrary requirement that i f >25% of all possible sequence combinations from a given patient had the "11/25" genotype (or were classified as "si" by N N or P S S M ) , then the sample was classified as having an "11/25" genotype (or was classified as "si"). This definition affected only one, 15 or 12 samples for the "11/25", N N or P S S M approaches, respectively. Although somewhat arbitrary, this definition corresponds to the approximate 25% minimum threshold at which minority nucleotide quasi-species are detectable by RT-148 P C R and sequencing methods. Samples with poor sequence data and/or that translated into >128 possible amino acid sequence combinations were not interpreted. Clinical Endpoints a) Mortality The primary endpoint in this analysis was non-accidental death. Deaths that occurred during the study period were identified through physician reports and linkage with the B . C . Department of Vi ta l Statistics. Deaths were classified according to ICD-9 and ICD-10 coding. A l l non-accidental deaths were considered potentially HIV-related and classified as events. Time to death was the time elapsed between the antiretroviral start date and date of non-accidental death. Accidental deaths were considered non-events and subjects censored at time of death. b) Other endpoints Additional outcomes in this study were viral suppression and rebound, C D 4 response and development of antiretroviral resistance. Time to viral suppression was defined as the time elapsed between the antiretroviral start date and the first of at least two consecutive plasma viral loads (pVL) < 500 copies HIV-1 R N A / m l . In the primary analysis of time to viral suppression, suppression could occur before or after switching therapies, while in a secondary analysis, individuals who had not achieved viral suppression at the time of switching therapy were censored on the switch date as non-events. This was done in order to take into consideration the possibility that some individuals may have required some optimisation of therapy before achieving plasma viral suppression. The subsequent time to plasma viral rebound was also evaluated. The time to rebound analysis was restricted to individuals who had achieved plasma viral suppression < 500 149 copies HIV-1 R N A / m l , and was defined as the time elapsed between initial viral suppression (the first of at least two consecutive p V L < 500 copies HIV-1 R N A / m l ) and the halfway point between the last p V L < 500 copies/ml and the first of at least two consecutive p V L > 500 copies/ml. If an individual registered only one p V L > 500 copies/ml but then subsequently died of a non-accidental cause, this was also considered an event. Time to C D 4 decline below baseline was defined as the time elapsed between the antiretroviral start date and the first C D 4 count below the baseline C D 4 count. The relationship between V 3 sequence and time to development of antiretroviral resistance was investigated. For each subject, all plasma samples with p V L >1000 H I V R N A / m l collected in the 2.5 years following initiation of H A A R T were genotyped for antiretroviral resistance [28]. Samples with p V L <1000 H I V R N A / m l were not genotyped and were assumed to carry no resistance mutations. Time to resistance was defined as the time elapsed between starting therapy and the first sample containing at least one major resistance mutation, defined as any nRTI -associated mutation (184I/V, 41L, 62V, 65R, 67N, 69D or insertion, 70R, 74V, 751, 151M, 210W, 215F/Y or 219E/Q); any NNRTI-associated mutation (1001, 103N, 1 0 6 A / M , 1081, 181C/I, 188C/H/L, 190A/S, 230L or 236L) or any PI-associated mutation (30N, 46I/L, 48V, 50L/V, 5 4 V / I 7 M , 82A/F/S/T, 84V, or 90M) [29]. The study end date was March 31, 2002. For mortality analysis, event-free subjects were censored at the date of last known contact with the H I V / A I D S Drug Treatment program. For outcomes involving C D 4 and viral load, event-free subjects were censored on the last test date available until March 31, 2002. For antiretroviral resistance, event-free subjects were censored at the last test date available up until 2.5 years from initiation of H A A R T on or before March 31, 2002 150 Statistical Analyses The influence of H I V V 3 sequence on clinical outcomes was assessed by Kaplan-Meier methods. Cox proportional hazards regression was used to calculate univariate and multivariate Risk Ratios (RR) and 95% confidence intervals (CIs). Baseline variables examined included plasma H I V R N A (per log increment), C D 4 count (per 100 cell decrement), age (per 10 year increment), proportion of total time spent on antiretroviral therapy in the first year of follow-up (per 10% decrement) [30], type of therapy at initiation (NNRTI-containing vs. Pi-containing), A I D S diagnosis (yes vs. no), history of injection drug use (yes vs. no) and gender (male vs. female). A l l tests of significance were 2-sided, with a p-value < 0.05 indicating statistical significance. 151 7.3 R E S U L T S Study subjects and V3 genotypes The effect of H I V V 3 genotype on therapy response and mortality was investigated in the H O M E R cohort of 1191 individuals initiating first triple H A A R T . Three methods were used to categorize V 3 genotype data: the presence of positively charged amino acids at codons 11 and/or 25 of the V 3 loop (the "11/25" genotype) [19], a Neural Network approach [25] and a Position-Specific Scoring Matrix method [26]. Baseline V 3 genotypes were successfully obtained for 1085 of 1191 subjects (91.1%). In general, unsuccessfully genotyped samples had significantly lower p V L than successfully genotyped samples (p<0.001, Table 1), but were otherwise comparable in their baseline characteristics. O f the 1085 successfully genotyped samples, 118 (10.9%) had an "11/25" genotype as determined by V 3 amino acid analysis. Again, baseline characteristics were generally similar for the "11/25" and non-"11/25" genotype groups, although a lower median baseline p V L was observed in the "11/25" genotype group (p=0.008, Table 1). Influence of V3 sequence on mortality A total of 159 deaths were observed among the 1085 successfully genotyped H O M E R subjects. O f these, 32 (20.1%) were accidental and subjects censored at date of death. Among the 127 non-accidental deaths, 22 (17.3%) occurred in subjects with the "11/25" genotype, while the remaining 105 (82.7%) occurred in subjects with the "non-11/25" genotype. Crude mortality rates were 18.6% and 10.9% in the "11/25" and "non-11/25" groups, respectively. In an analysis of the influence of V 3 sequence on time to death after initiation of H A A R T , we observed a statistically significant association between the "11/25" genotype 152 and more rapid mortality (Risk Ratio 1.77, 95%CI 1.12-2.80, p=0.015) (Figure 1 and Table 2a). This association remained significant in multivariate analyses (Risk Ratio 1.70, 95% CI 1.06-2.71, p=0.027) controlling for baseline factors including age, plasma viral load, C D 4 count, the proportion of time spent on antiretrovirals in the first year of follow up [30], and A I D S diagnosis (Table 2a). Influence of V3 genotype on viral and immunological endpoints In the pre-defined primary analysis of time to achieving suppression of p V L below 500 H I V R N A copies/ml, we observed no significant influence of the V 3 sequence on time to viral suppression (univariate p>0.1, data not shown). Nevertheless, two lines of evidence from post-hoc data exploration suggested the possibility that viral suppression was actually somewhat more difficult to achieve and/or maintain in patients with the "11/25" genotype. There was a significant association between the "11/25" genotype and an increased time to viral suppression < 500 copies/ml if individuals were censored at time of switching therapy (multivariate Risk Ratio 0.63, 95% CI 0.47-0.84, p=0.002), suggesting that patients with the "11/25" genotype were more likely to require a therapy change to achieve plasma viral suppression (data not shown). Furthermore, in patients with the "11/25" genotype, the median p V L was less consistently between 50 and 500 copies H I V R N A / m l over the course of study follow-up (Figure 2a). Although the ultrasensitive p V L assay was not consistently available over the course of the study, the proportion of "11/25" and "non-11/25" samples genotyped using the ultrasensitive assay was comparable over the course of study follow-up (data not shown). In individuals who achieved viral suppression to below 500 H I V R N A copies/ml, we observed no significant association between V 3 sequence and time to subsequent viral rebound (univariate p>0.1, data not shown). 153 The pre-defined primary analysis of time to C D 4 cell decline below baseline was shorter in individuals with the "11/25" genotype (univariate p=0.009, Table 2b). This trend remained significant in multivariate analyses controlling for baseline factors and prescription refills in the first year of follow-up [30] (Risk Ratio 1.38, 95% C I 1.07-1.77, p=0.012, Table 2b). Over the course of study follow-up, the median C D 4 response was consistently nearly 100 cells lower in subjects with the "11/25" genotype (Figure 2b). O f interest, one of the strongest predictors of both C D 4 decline and mortality after starting H A A R T was the proportion of prescriptions filled in the first year of follow-up, a surrogate of adherence [30]. This association had been demonstrated in a previous study of the H O M E R cohort [5], and emphasizes the importance of obtaining and taking prescribed antiretrovirals. Influence of V3 genotype on time to antiretroviral resistance The influence of V 3 sequence on the time to the development of antiretroviral resistance was also investigated. For each study subject, all plasma samples with p V L > 1000 H I V R N A / m l collected in the 2.5 years following initiation of H A A R T (N>2600) were genotyped . O f the 1085 individuals for which V3 genotyping was successful, 33 (3.0%) had no follow-up plasma samples available for testing, and resistance testing failed in a further 26 (2.4%); these individuals were censored at time=0. A further 306 individuals (28.2%) had follow-up p V L measurements < 1000 H I V R N A / m l ; these subjects were censored at the last p V L < 1000 and assumed to carry no resistance mutations. O f the remaining subjects with follow-up p V L > 1000, 450 (41.5%) displayed no resistance mutations after genotyping, and 270 (24.9%) subjects displayed at least one major resistance mutation. O f the 270 subjects with at least one resistance mutation, 238 were in the "non-11/25" genotype group (24.6% of this group) while the remaining 32 were in the "11/25" genotype group (27.1% of this 154 group). No association between V 3 sequence and time to development of drug resistance mutations was observed (univariate p>0.1, data not shown). Evaluation of clinical endpoints using PSSM-based and neural-network phenotype predictions Both Position-Specific Scoring Matrix (PSSM) [26] and neural network (NN) approaches [25] were used as alternative methods for categorizing H I V genotypes. P S S M predicted "si" virus in 108 (10.0%) of samples, close to the 10.9% with the "11/25" genotype. Concordant values (cases where "11/25" genotypes were called "si" by P S S M or where "non-11/25" genotypes were called "nsi" by P S S M ) were obtained in 96% of cases. P S S M "si" predictions were associated with earlier mortality in univariate analyses (p<0.05) and a trend remained in multivariate analyses (Risk Ratio 1.53, 95% CI 0.95-2.46, p=0.081, data not shown). P S S M "si" predictions were not associated with time to viral suppression below 500 copies/ml in the primary analysis, however in a secondary analysis where individuals' data were censored at time of therapy switch, P S S M "si" predictions were significantly associated with a longer time to viral suppression in both univariate and multivariate analyses (multivariate Risk Ratio 0.68, 95% CI 0.50-0.93, p=0.015, data not shown). P S S M predictions were not associated with time to virologic failure, development of drug resistance, or C D 4 decline below baseline (p>0.1, data not shown). The neural network (NN) predictions diverged more strongly from those of the "11/25" and P S S M methods, with approximately 18% of viral isolates being classified as "si". Concordance was relatively low between N N sequence categories and both the "11/25" genotype (87%) and the P S S M sequence categories (87%). N N sequence categories were not significantly associated with any endpoints investigated (p>0.1, data not shown). 155 7.4 D I S C U S S I O N A N D C O N C L U S I O N S We assessed the ability of three sequence interpretation methods to predict clinical outcomes from baseline H I V envelope V3 sequences in a large cohort of antiretroviral naive individuals initiating first triple H A A R T : "the 11/25 rule", a method based on examining amino acid variation at two critical V 3 sites, as well as two bioinformatics-based approaches, a Neural Network (NN) and a Position Specific Scoring Matrix (PSSM) method. Of the three sequence interpretation methods, the "11/25 rule" was the strongest predictor of clinical parameters, including mortality, in this cohort. The data indicate a significant correlation between the presence of positively charged amino acids at codons 11 and 25 of the H I V envelope V3 loop (the "11/25" genotype) and inferior C D 4 cell response and reduced survival time following initiation of H A A R T . These poorer prognoses were maintained despite the fact that the time to achieve viral suppression to below 500 H I V R N A copies/ml was similar for the two groups in a primary analysis allowing for regimen changes. Although this association is most likely driven by the association of the "11/25" genotype with SI phenotype and/or X 4 co-receptor usage, it is important to stress that we were interested in the predictive value of H I V V 3 loop sequence data for clinical outcomes rather than the predictive values of in vitro phenotype, cellular tropism or co-receptor usage. The relative ease and lower costs associated with genotyping (when compared to phenotyping assays) may make this test of practical clinical relevance in the management of HIV-infected individuals. Presuming V 3 loop sequence is indeed predictive of viral phenotype and/or co-receptor usage [19-23], these observations are consistent with previous studies linking the 156 emergence of SI virus to a rapid loss of CD4-positive T-cells and accelerated disease progression in untreated individuals [8-11], and a poorer response to therapy in a limited number of studies of treated individuals [12-17]. O f interest, we observed that the "11/25" genotype was not associated with an increased time to plasma viral suppression below 500 copies/ml unless individuals were censored at the time of any switch in antiretroviral therapy. These results do not necessarily indicate that individuals harboring H I V with the "11/25" genotype are less likely to suppress plasma viral load, but rather that it may require some optimization of therapy to find a combination that is effective. Importantly, even though antiretroviral therapies led to suppression of viral load to a median below 500 H I V R N A copies/ml, immunological and clinical response was still poorer in the "11/25" genotype group. In contrast, we observed no measurable effects of the "11/25" genotype on time to viral rebound (after initial suppression) or the development of drug resistance mutations, suggesting that the "11/25" genotype does not drive virological failure. Sequence categories using the Position-Specific Scoring Matrix (PSSM) were largely consistent with sequence.interpretations using the "11/25" rule, and were similarly predictive of time to viral suppression and mortality, but not C D 4 response. Note that P S S M has a potential advantage over the simple dichotomization used here, as there is likely a continuum of H I V genetic sequences representing evolution to use the C X C R 4 co-receptor with increasing efficiency. In contrast to the "11/25" rule, P S S M takes this into consideration by evaluating the strength of the association between V 3 sequence and inferred viral phenotype. However for reasons of simplicity and consistency, we dichotomized P S S M sequence interpretations into either "si" or "nsi". It is possible, however, that sequence-based interpretation methods that take into consideration factors such as the efficiency of co-157 receptor use and/or the strength of the association between genotype and inferred phenotype may prove to be more useful than simple sequence-based prediction methods in clinically relevant settings. Rather surprisingly, H I V phenotypes predicted using the neural network (NN) differed substantially from "11/25" and P S S M predictions, and were not associated with clinical outcomes. The N N was originally trained to interpret collected datasets of viruses that were both genotyped and phenotyped (including both lab strains and primary isolates); its potential to predict outcomes from population-based sequence data from plasma samples therefore remains to be determined. In addition, the translation of nucleotide mixtures may have potentially disrupted natural patterns of sequence linkage, which could have most strongly affected the N N predictions. Since these analyses were performed, an additional approach to V 3 sequence interpretation has been published, which has not yet been investigated [31]. We are unable to investigate the possibility that the "11/25" genotype was associated with different lengths of time since seroconversion or more rapid disease progression before therapy, as seroconversion dates are largely unknown in this cohort. In addition, as our observations are based on a single V 3 genotype at baseline, we cannot rule out evolution of V3 sequences during study follow-up. Finally, as R T - P C R and D N A sequencing techniques were used to evaluate V 3 sequences, only the most predominant species are detected; both types of virus may co-exist within any individual representing minority quasispecies. In conclusion, results from this study indicate a significant correlation between positively charged residues at positions 11 and 25 of the HIV-1 envelope V 3 loop, and both reduced C D 4 response and decreased survival time following initiation of H A A R T . Baseline 158 V 3 loop sequence data may therefore provide additional information regarding therapy response and could potentially be incorporated into standard clinical genotyping procedures. 159 Table 7.1: Patient Characteristics at baseline. [IQR]: Interquartile range Baseline Characteristics Total study group (N=1191) Stratified by success of V3 genotyping Stratified by V3 genotype (N=1085 successfully genotyped samples) Genotyping not successful (N=106) Genotyping successful (N=1085) P "non-11/25" genotype (N=967) "11/25" genotype (N=118) P Male Gender 1004 84 920 0.134 823 97 0.407 (%) (84.3%) (79.3%) (84.8%) (85.1%) (82.2%) Median Age, years 37 38 37 0.802 37 37 0.782 [IQR] [32-44] [32-42] [32-44] [32-44] [32-44] Median HIV pVL 120 77 130 <0.001 130 87 0.008 (x103 copies/ml) [IQR] [42-310] [12-220] [47-320] [49-330] [23-250] Median CD4 count 280 260 280 0.328 280 280 0.796 (cells/mm3) [IQR] [130-420] [100-440] [130-420] [130-420] [120-430] Baseline AIDS 154 15 139 0.695 125 14 0.744 diagnosis (%) (12.9%) (14.2%) (12.8%) (12.9%) (11.9%) NNRTI-containing 306 20 286 0.092 259 27 0.364 therapy (%) (25.7%) (18.9%) (26.4%) (26.8%) (22.9%) History of injection 321 23 298 0.202 265 33 0.897 drug use (%) (27.0%) (21.7%) (27.5%) (27.4%) (28.0%) 160 Table 7.2a: Influence of baseline factors on time to death after initiation of HAART Baseline Risk Factor Univariate Multivariate Risk Ratio 95% CI P Risk Ratio 95% CI P "11/25" genotype 1.77 1.12-2.80 0.015 1.70 1.06-2.71 0.027 male gender 0.94 0.58-1.53 0.808 - - -Age (per 10 year increment) 1.45 1.21-1.73 <0.001 1.49 1.24-1.78 <0.001 baseline HIV RNA (per log increment) 1.78 1.28-2.48 0.001 1.44 1.03-2.00 0.032 baseline CD4 cell count (per 100 cell decrement) 1.31 1.19-1.45 <0.001 1.31 1.17-1.47 <0.001 % of time on antiretrovirals (per 10% decrement) 1.16 1.10-1.23 <0.001 1.25 1.18-1.32 <0.001 baseline AIDS diagnosis 1.70 1.10-2.64 0.017 1.30 0.81-2.11 0.280 History of injection drug use 1.14 0.79-1.66 0.481 - - -NNRTI vs. Pl-containing therapy 0.90 0.58-1.40 0.642 - - -161 Table 7.2b: Influence of baseline factors on time to CD4 decline below baseline after initiation of HAART Baseline Risk Factor Univariate Multivariate Risk Ratio 95% CI P Risk Ratio 95% CI P "11/25" genotype 1.40 1.09-1.79 0.009 1.38 1.07-1.77 0.012 male gender 0.68 0.55-0.85 <0.001 0.92 0.74-1.15 0.465 age (per 10 year increment) 0.95 0.87-1.05 0.325 - - -baseline HIV RNA (per log increment) 0.71 0.62-0.80 <0.001 0.99 0.86-1.14 0.877 baseline CD4 cell count (per 100 cell decrement) 0.85 0.83-0.87 <0.001 0.87 0.84-0.89 <0.001 % of time on antiretrovirals (per 10% decrement) 1.22 1.18-1.25 <0.001 1.19 1.16-1.22 <0.001 baseline AIDS diagnosis 0.38 0.27-0.53 <0.001 0.68 0.48-0.96 0.029 History of injection drug use 1.54 1.30-1.83 <0.001 1.21 1.01-1.44 0.042 NNRTI vs. Pl-containing therapy 1.33 1.11-1.61 0.002 1.38 1.13-1.67 0.001 162 Figure 7.1: Association between "11/25" genotype and time to death after initiating HAART. 70 I i i i i i • i i • • i i i i i • . • • i 0 6 12 18 24 30 36 42 48 Time from initiation of HAART (months) non-11/25 X (n) 967 898 834 660 388 11/25 O (n) 118 110 100 81 41 H I V isolates were classified into two groups based on amino acid analysis of the H I V V3 loop region. Isolates with positively charged amino acids at positions 11 and/or 25 of the V3 loop (a genotype previously associated with H I V SI phenotype [19]) were classified as having an "11/25" genotype. Individual events (non-accidental deaths) are indicated by circles ("11/25" genotype) or crosses ("non-11/25" genotype). There was a statistically significant association between the "11/25" genotype at baseline and an earlier time to death after initiation of H A A R T (log rank p=0.015). The number of subjects remaining in the analysis is indicated below the graphs. 163 Figure 7.2: Longitudinal analysis of population-based HIV viral loads (panel A) and CD4 counts (panel B), stratified by HIV V3 loop genotype. The median post-therapy plasma viral load (A) or C D 4 cell count (B) after starting triple drug therapy are indicated for individuals with the "11/25" genotype (solid lines with circles) or "non-11/25" genotype (dotted lines with crosses). Error bars represent the Interquartile Range associated with each point. If an individual had more than one C D 4 or viral load test performed during any given time interval, only the latest measurement was included. The gray horizontal line in Panel A indicates the 500 copies H I V R N A / m l (2.70 log) cutoff used in the primary analyses. The ultrasensitive adaptation of the plasma viral load assay was not available over the entire study period; data are presented where this assay was available. A Time from initiation of HAART (months) non-11/25-X- (n) 967 794 722 693 678 665 661 658 642 608 532 454 393 345 11/25 —O— (n) 118 93 81 71 71 73 71 75 68 65 67 50 36 37 164 B 700 -j 600 -Time from Initiation of HAART (months) non-11/25-X-(n) 967 676 626 606 614 565 568 567 571 526 466 382 342 305 11/25 —O—(n)118 73 72 62 59 61 61 63 53 55 56 47 28 32 165 7.5 R E F E R E N C E S 1. Palella FJ Jr, Delaney K M , Moorman A C , Loveless M O , Fuhrer J, Satten G A , et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 1998; 338:853-60. 2. Hogg R S , Y i p B , K u l l y C , Craib K J , O'Shaughnessy M V , Schechter M T , et al. Improved survival among HIV-infected patients after initiation of triple-drug antiretroviral regimens. Can Med Assoc J 1999; 160:659-65. 3. Wainberg M A , Friedland G . Public Health implications of anti-retroviral therapy and H I V drug resistance. J A M A 1999; 279:1977-83. 4. Yeni P G , Hammer S M , Carpenter C C , Cooper D A , Fischl M A , Gatell J M , et al. Antiretroviral treatment for adult H I V infection in 2002: updated recommendations of the International A I D S Society-USA Panel. J A M A 2002; 288:222-35. 5. Hogg R S , Y i p B , Chan K J , Wood E , Craib K J , O'Shaughnessy M V , et al. Rates of disease progression by baseline C D 4 cell count and viral load after initiating triple-drug therapy. J A M A 2001; 286:2568-77. 6. Egger M , M a y M , Chene G , Phillips A N , Ledergerber B , Dabis F, et al. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet 2002; 360:119-29. 7. Wood E , Hogg R S , Y i p B , Harrigan PR, O'Shaughnessy M V , Montaner JS. Is there a baseline C D 4 cell count that precludes a survival response to modern antiretroviral therapy? A I D S 2003; 17:711-20. 166 8. Tersmette M , Gruters R A , de W o l f F, de Goede R E , Lange J M , Schellekens FT, et al. Evidence for a role of virulent human immunodeficiency virus (HIV) variants in the pathogenesis of acquired immunodeficiency syndrome: studies on sequential H I V isolates. J V i r o l 1989; 63:2118-25. 9. Schuitemaker H , Koot M , Kootstra N A , Dercksen M W , de Goede R E , van Steenwijk RP , et al. Biological phenotype of human immunodeficiency virus type 1 clones at different stages of infection: progression of disease is associated with a shift from monocytotropic to T-cell tropic virus populations. J V i r o l 1992; 66:1354-60. 10. Koot M , Keet IP, Vos A H . Prognostic value of HIV-1 syncytium-inducing phenotype for rate of CD4+ cell depletion and progression to A I D S . A n n Int Med 1993;118:681-88. 11. Richman D D , Bozzette S A . The impact of the syncytium-inducing phenotype of human immunodeficiency virus on disease progression. J Infect Dis 1994; 169:968-74. 12. Rusconi S, De Pasquale M P , Mainini F, Bulgheroni E , Kurtagic S, Gori A , et al. Vira l load, viral phenotype modification, zidovudine susceptibility and reverse transcriptase mutations during the first 6 months of zidovudine monotherapy in HIV-1-infected people. Antivir Ther 1996; 1:211-19. 13. Katzenstein T L , Nielsen C, Bruun L , Mathiesen L R , Pedersen C , Gerstoft J. Quantification of HIV-1 R N A during antiretroviral therapy: association with viral phenotype and development of resistance. Antivir Ther 1996; 1:246-54. 167 14. Vida l C , Garcia F, Gatell J M , Leal M , Clotet B , Pumarola T, et al. Predictive factors influencing peak viral load drop in response to nucleoside reverse transcriptase inhibitors in antiretroviral-naive HIV-1-infected patients. J Acquir Immune Defic Syndr Hum Retrovirol 1998; 19:55-60. 15. Davey R T Jr, Chaitt D G , Reed G F , Freimuth W W , Herpin B R , Metcalf J A , et al. Randomized, controlled phase I/II, trial of combination therapy with delavirdine (U-90152S) and conventional nucleosides in human immunodeficiency virus type 1-infected patients. Antimicrob Agents Chemother 1996; 40:1657-64. 16. Hughes M D , Johnson V A , Hirsch M S , Bremer J W , Elbeik T, Erice A , et al. Monitoring plasma HIV-1 R N A levels in addition to CD4+ lymphocyte count improves assessment of antiretroviral therapeutic response. A C T G 241 Protocol Virology Substudy Team. A n n Intern Med 1997; 126:929-38. 17. Bratt G , Karlsson A , Leandersson A C , Albert J, Wahren B , Sandstrom E . Treatment history and baseline viral load, but not viral tropism or C C R - 5 genotype, influence prolonged antiviral efficacy of highly active antiretroviral treatment. A I D S 1998; 12:2193-202. 18. Bjorndal A , Deng H , Jansson M , Fiore JR, Colognesi C , Karlsson A , et al. Coreceptor usage of primary human immunodeficiency virus type 1 isolates varies according to biological phenotype. J V i ro l 1997; 71:7478-87. 19. Fouchier R A , Brouwer M , Broersen S M , Schuitemaker H . Simple determination of Human Immunodeficiency Virus Type 1 Syncytium-inducing V 3 Genotype by P C R . J C l i n M i c b 1995;33:906-11. 168 20. O'Brien W A , Koyanagi Y , Namazie A , Zhao JQ, Diagne A , Idler K , et al. HIV-1 tropism for mononuclear phagocytes can be determined by regions of gpl20 outside the CD4-binding domain. Nature 1990; 348:69-73. 21. Shioda T, Levy J A , Cheng-Mayer C. Macrophage and T cell line-tropisms of HIV-1 are determined by specific regions of the envelope gp l20 gene. Nature 1991; 349:167-69. 22. De Jong JJ, De Ronde A , Keulen W , Tersmette M , Goudsmit J. Min imal requirements for the human immunodeficiency virus type 1 V 3 domain to support the syncytium-inducing phenotype: analysis by single amino acid substitution. J Vi ro l 1992; 66:6777-80. 23. Fouchier R A , Groenink M , Kootstra N A , Tersmette M , Huisman H G , Miedema F, et al. Phenotype-associated sequence variation of the third variable domain of the human immunodeficiency virus type 1 gpl20 molecule. J V i r o l 1992; 66:3183-87. 24. Hoffman N G , Seillier-Moiseiwitsch F, A h n J, Walker J M , Swanstrom R. Variability in the human immunodeficiency virus type 1 gp 120 Env protein linked to phenotype-associated changes in the V3 loop. J Vi ro l 2002; 76:3852-64. 25. Resch W , Hoffman N , Swanstrom R. Improved success of phenotype prediction of the human immunodeficiency virus type 1 from envelope variable loop 3 sequence using neural networks. Virology 2001; 288:51-62. 26. Jensen M A , van't Wout A . Predicting HIV-1 co-receptor usage using sequence analysis. A I D S rev 2003; 5:104-12. 27. Hirsch M S , Brun-Vezinet F, D 'Aqui la R T , Hammer S M , Johnson V A , Kuritzkes DR, 169 et al. Antiretroviral drug resistance testing in adult HIV-1 infection: recommendations of an International A I D S Society-USA Panel. J A M A 2000; 283:2417-26. 28. Alexander C S , Dong W , Chan K , Jahnke N , O'Shaughnessy M V , M o T, et al. H I V protease and reverse transcriptase variation and therapy outcome in antiretroviral-na'ive individuals from a large North American cohort. A I D S 2001; 15:601-607. 29. D 'Aqui la R T , Schapiro J M , Brun-Vezinet F, Clotet B , Conway B , Demeter L M , et al. Drug resistance mutations in H I V - 1 . Topics in H I V Medicine 2002; 10:21-25. 30. Hogg R S , Heath K , Bangsberg D , Y i p B , Press N , O'Shaughnessy M V , et al. Intermittent use of triple-combination therapy is predictive of mortality at baseline and after 1 year of follow-up. A I D S 2002; 16:1051-58. 31. Pi l la i S, Good B , Richman D , Corbeil J. A new perspective on V 3 phenotype prediction. A I D S Res H u m Retroviruses 2003; 19:145-149. 170 C H A P T E R 8 H I V G E N E T I C V A R I A T I O N : R E L E V A N C E O F H I V E N V E L O P E C O R E C E P T O R U S A G E O N R E S P O N S E T O H A A R T 1 8.1 I N T R O D U C T I O N Human Immunodeficiency Virus type 1 (HIV-1) uses host cell membrane chemokine receptors in combination with C D 4 in order to gain entry into host cells [1,2,3]. The most important co-receptors in H I V pathogenesis are the chemokine receptors C C R 5 [1,2,4] and C X C R 4 [5], although in rare cases other co-receptors have also been shown to mediate entry of HIV-1 into target cells, at least in vitro [6,7,8]. In general, most HIV-1 variants isolated from newly-infected individuals utilize C C R 5 in combination with C D 4 to gain entry to host cells. These R5 variants predominantly infect activated C D 4 + T-lymphocytes as well as macrophages. While R5 variants are generally detectable over the entire course of H I V infection [9], variants able to utilize C X C R 4 emerge in approximately 40-50% of infected persons over the course of disease [10]. These X 4 variants predominantly target naive and resting C D 4 + T-lymphocytes and display different biological properties than their R5 counterparts, including increased replication rate, pathogenicity and syncytium-inducing (SI) capacity in cell culture [9,11]. In addition, dual-tropic variants capable of using both C X C R 4 and C C R 5 may also arise over the course of disease [12]. The factors mediating the R5-to-X4 phenotype "switch" over the natural course of H I V infection remain incompletely understood. The emergence of CXCR4-us ing variants is associated with a rapid decline in C D 4 + T-lymphocyte counts, accelerated disease 1 This chapter has been published as: Brumme Z L , Goodrich J , Mayer H , Brumme C J , H e n r i c k B M , W y n h o v e n B , Asse l inJ , Cheung P K , H o g g R S , Montaner JS, and Harrigan P R . Molecular and Clinical Epidemiology of CXCR4-using HIV-1 in a large population of therapy-naive individuals. J Infect Dis 2005; 192:466-74. 171 progression and reduced survival time in untreated individuals [13,14,15,16], as well as poorer response to treatment in the pre-Highly Active Antiretroviral Therapy ( H A A R T ) era [17,18,19,20,21]. However, it is not known whether X 4 variants are inherently more pathogenic and are directly responsible for more rapid disease progression, or whether CXCR4-us ing H I V variants may emerge as a consequence of progressive immune dysfunction [22]. Regardless of the direction of causation, the association of X 4 H I V with poorer prognosis and inferior therapy response remains an important issue in clinical practice, and it is important that the prognostic implications of H I V co-receptor usage be re-evaluated in the H A A R T era. At present, the epidemiology of R5 and X 4 H I V is also of particular relevance due to the development of H I V co-receptor inhibitors. This new class of antiretroviral agents is designed to specifically inhibit H I V binding to either C C R 5 and/or C X C R 4 , thereby preventing H I V entry into target cells. Although co-receptor inhibitors currently show promising effects in early clinical trials [22,23], it is not known on a population basis what proportion of HIV-infected individuals may potentially benefit most from antiretroviral agents targeting C C R 5 and/or C X C R 4 . Using the newly-developed Virologic Phenosense assay, we investigated the epidemiology of H I V co-receptor usage in the H O M E R cohort in British Columbia, Canada, consisting of 1191 antiretroviral-nai've individuals initiating their first triple combination therapy between August 1996 and September 1999 [24,25]. In this cross-sectional study, we wished to characterize the prevalence of X 4 H I V at therapy initiation and to identify sociodemographic, clinical and genetic risk factors associated with phenotypic C X C R 4 usage. Finally, using longitudinal clinical data collected over the course of study follow-up, 172 we wished to establish the impact of C X C R 4 co-receptor usage on clinical and virological outcomes after initiation of triple combination therapy. 173 8.2 M A T E R I A L S A N D M E T H O D S The B.C. Centre for Excellence in HIV/AIDS Drug Treatment Program In the province of British Columbia, Canada, antiretrovirals are distributed free of charge to HIV-infected individuals through a centralized drug treatment program based at the B . C . Centre for Excellence in H I V / A I D S (the Centre). Antiretrovirals are prescribed according to specific guidelines set by the B . C . Therapeutic Guidelines Committee, which are revised regularly and are in accordance with international guidelines [24]. Patients who enroll in the program may provide informed consent and participate in a survey that collects sociodemographic data. Routine clinical monitoring of patients takes place at approximately 3-month intervals at which time plasma viral load testing (Roche Amplicor Monitor Assay) and C D 4 counts are performed. These data are stored in the Centre's Drug Treatment Program database. Ethical approval for this study was obtained from the research ethics board at the University of British Columbia, St. Paul's Hospital site. The HOMER cohort The H A A R T Observational Medical Evaluation and Research ( H O M E R ) cohort includes all HIV-positive, antiretroviral naive adults who started triple antiretroviral therapy (consisting of two nucleoside reverse transcriptase inhibitors [nRTIs] and either a protease inhibitor [PI] or a non-nucleoside reverse transcriptase inhibitor [NNRTI]) through the B . C . Drug Treatment Program between August 1996 and September 1999 (N=1191). This cohort has been the focus of a number of population-based studies and has been described in detail previously [24,25]. Subjects were followed for a median of 4.7 years (56 months) after initiation of antiretroviral therapy. 174 Determination of baseline HIV co-receptor phenotype and envelope V3 genotype For each study subject, a single pre-therapy (baseline) plasma sample collected < 6 months prior to initiation of therapy was assayed using the Virologic Phenosense Assay (Virologic, San Francisco) in order to determine H I V co-receptor usage. Vi ra l R N A was extracted using oligo(dT) columns and HIV-1 env-specific primers were used to amplify a 2.5-kb R T - P C R product that spanned the entire gp l60 open reading frame. R T - P C R products were digested, purified and ligated into an E. Coli expression vector, and gene libraries constructed. A replication-defective retroviral vector (pHIVluc) containing a luciferase expression cassette inserted within the env gene was used to co-transfect human embryonic kidney cell cultures with the sample plasmid D N A . Recombinant viruses were harvested after 48 hours and assessed for their ability to infect cells expressing C C R 5 or C X C R 4 . The Phenosense assay classifies isolates as R5 , X 4 or R 5 / X 4 (indicating dual and/or mixed-tropic virus). The same baseline plasma samples were used to determine H I V V 3 envelope sequence as described previously [25]. Isolates displaying positively-charged amino acids at codons 11 and/or 25 of the V 3 loop [26], associated with an H I V syncytium-inducing (SI) phenotype [26], were classified as having an "11/25 genotype" [25]. Determination of CCR5A32 genotype Blood (n=796, 66.8%) or plasma (n=395, 33.2%) samples were available from all 1191 study subjects. D N A was extracted from blood or plasma using the Qiagen D N A kit adapted for use on the Qiagen Biorobot 9604. Extracted D N A was amplified in a single round of P C R using primers flanking the A32 region. P C R products were visualised by 175 electrophoresis on a 2% agarose gel and confirmed by automated D N A sequencing in both 5' and 3' directions on an A B I 3700 D N A sequencer. Statistical Methods Association of baseline HIV co-receptor use with sociodemographic, clinical and genetic variables Associations between baseline H I V co-receptor usage and dichotomous baseline parameters including gender, baseline A I D S diagnosis, history of injection drug use, human C C R 5 genotype (CCR5wt/A32 or wt/wt) and H I V V3 genotype ("11/25" genotype or not) [25] were determined using the Chi-squared test. Associations between baseline H I V co-receptor usage and continuous variables including age, plasma viral load ( p V L ) and C D 4 count were determined using the Wilcoxon Rank-sum test. Baseline predictors of X 4 usage and associated odds ratios were calculated using univariate and multivariate logistic regression. Association between baseline HIV co-receptor usage and response to antiretroviral therapy The primary endpoint investigated in the outcome analyses was time to non-accidental death, defined as the time from therapy initiation to non-accidental death (occurring on or before June 30, 2003, the date of latest linkage with mortality statistics from the British Columbia Vi ta l Statistics Agency). Deaths were classified according to ICD-10 coding. Accidental deaths were not considered events and subjects censored at date of death. Additional clinical outcomes included time to suppression of plasma H I V R N A (time from therapy initiation to first of two consecutive viral loads <500 copies/ml), time to plasma H I V R N A rebound (subsequent time to the first of two consecutive viral loads >500 copies/ml), and time to C D 4 decline (time to first decline of C D 4 count below baseline). For the clinical 176 outcome analyses, event-free subjects were censored at the collection date of the last tested sample up to and including June 30, 2003. The influence of H I V co-receptor usage on clinical outcomes was assessed by Kaplan-Meier methods. Cox proportional hazards regression was used to calculate univariate and multivariate Hazard Ratios (HR) and 95% confidence intervals (CI). Baseline variables included in the model were gender (male vs. female [reference group]), age (per 10 year increment), A I D S diagnosis (yes vs. no), plasma H I V R N A (per logio increment), C D 4 count (per 100 cell decrement), proportion of total time spent on antiretroviral therapy in the first year of follow-up (per 10% increment) [27], type of therapy at initiation (Pi-containing vs. NNRTI-'containing), history of injection drug use (yes vs. no), C C R 5 genotype (wt/A32 vs. wtAvt) and H I V V 3 genotype ('11/25' vs. not) [25]. In both the logistic regression and Cox proportional hazard regression analyses, all factors significant in univariate analyses were included in multivariate analyses. A l l tests of significance were two-sided, with a p-value < 0.05 indicating statistical significance. 177 8.3 R E S U L T S Results of HIV co-receptor phenotyping The prevalence and determinants of HIV-1 co-receptor usage were investigated in the H O M E R [24,25] cohort of 1191 antiretroviral naive individuals initiating therapy. H I V co-receptor usage data were obtained for 979 of 1191 subjects (82.2%). Phenotype data were more likely to be available for male subjects (p=0.04), and for subjects with higher plasma viral loads (p<0.0001), lower C D 4 counts (p=0.007) and the human C C R 5 wt/wt genotype (p=0.007) (Table la). O f the 979 phenotyped subjects, 801 (81.8%) harbored R5 variants, 177 (18.1%) harbored R 5 / X 4 variants, and only one (0.1%) harbored exclusively X 4 variants. For the remainder of the analyses, this individual was included in the R 5 / X 4 group. Baseline characteristics stratified by HIV co-receptor usage in a therapy-naive population Comparison of baseline parameters from subjects harboring R 5 / X 4 variants with those harboring exclusively R5 variants revealed several important differences in clinical characteristics (Table la). Subjects with R 5 / X 4 variants had significantly higher H I V plasma viral loads (p=0.0006), lower C D 4 counts (p<0.0001), and were significantly more likely to be diagnosed with an AIDS-defining illness prior to therapy initiation (p<0.0001). The distribution of H I V co-receptor usage at clinically useful C D 4 strata highlights the strong association between the detection of X 4 variants and absolute C D 4 count at baseline, ranging from <10% for C D 4 counts above 200 cells/mm 3 to >50% for C D 4 counts below 25 cells/mm 3 (Table lb) . As expected, subjects with basic amino acids at codons 11 and/or 25 of the H I V V3 loop, (an '11/25' genotype [25,26]), were significantly more likely to harbor R 5 / X 4 H I V (p<0.0001, Table la) than subjects with neutral or negatively-charged residues at these sites. 178 The '11/25 genotype', in this case evaluated using direct P C R and sequencing of patient-derived isolates, displayed a specificity of 93.1% and a sensitivity of 32.7% for predicting the presence of X 4 variants. In addition, individuals heterozygous for the C C R 5 A 3 2 deletion (CCR5 wt/A32) were also significantly more likely to harbor R 5 / X 4 H I V variants than individuals with the C C R 5 wt/wt genotype (p=0.0085) (Table la). Baseline characteristics stratified by combined CCR5 genotype and HIV phenotype data In order to further investigate the association between C C R 5 A 3 2 genotype and R5 /X4 H I V , we combined the human C C R 5 genotype and H I V co-receptor usage data. Combined H I V phenotype and human C C R 5 genotype data were available for 967 subjects (Table 2). The four categories [CCR5 wt/wt, R5 phenotype (N=697); C C R 5 wt/wt, R 5 / X 4 phenotype (N=142); C C R 5 wt/A32, R5 phenotype (N=94); C C R 5 wt/A32, R 5 / X 4 phenotype (N=34)] were compared with respect to baseline sociodemographic and clinical characteristics. Individuals heterozygous for the C C R 5 A 3 2 deletion and who harbored exclusively R5 H I V displayed the most favourable clinical profile; of the four groups, these subjects had the lowest plasma viral loads and the highest C D 4 counts at baseline (Table 2). In contrast, individuals with R 5 / X 4 H I V , regardless ofCCR5 genotype, had the poorest clinical profile; these individuals had the highest plasma viral loads, the lowest C D 4 counts, and were more likely to be diagnosed with an AIDS-defining illness at time of therapy initiation than subjects harboring only R5 H I V (Table 2). In addition, we wished to investigate whether R 5 / X 4 H I V variants from C C R 5 wt/wt or C C R 5 wt/A32 individuals might differ in their predominant H I V V 3 amino acid sequence. Although, as mentioned previously, R 5 / X 4 H I V was significantly associated with the '11/25 genotype' [25] (p<0.0001, Table 2), we observed no significant difference in the ability of the 179 '11/25' genotype to predict H I V co-receptor usage in either C C R 5 wt/wt or C C R 5 wt/A32 individuals (sensitivity 30.1% vs. 39.4% respectively, p=0.3, data not shown). Predictors ofR5/X4 HIV in a therapy-naive population Univariate and multivariate logistic regression were used to identify significant baseline predictors of R 5 / X 4 H I V and to calculate univariate and multivariate odds ratios associated with these parameters (Table 3). In multivariate analyses, the strongest predictors of R 5 / X 4 H I V were low C D 4 count (Odds Ratio [OR] 1.53 per 100-cell decrement in CD4 count, p<0.0001) and the H I V V 3 loop '11/25' genotype (OR 9.11, p<0.0001). Additional predictors of R 5 / X 4 H I V were high baseline plasma viral load (OR 1.46 per login increment in H I V R N A , p=0.04) and the heterozygous C C R 5 wt/A32 genotype (OR 2.48, p=0.0005) (Table 3). Association of baseline R5/X4 HIV with mortality and other clinical outcomes In univariate analyses, the presence of R 5 / X 4 H I V at baseline was significantly associated with decreased survival after initiating antiretroviral therapy (univariate p=0.05, Table 4a and Figure la). However, after adjusting for baseline factors including C D 4 count, age and plasma viral load, this association was no longer significant (multivariate p=0.45, Table 4a). We observed no significant associations between baseline H I V co-receptor usage and either C D 4 response after initiation of therapy (p=0.07, Figure l b and Table 4b) or the time to achieve suppression of plasma H I V R N A <500 copies/ml (p=0.98, HR=1.00). In univariate analyses, subjects harboring R 5 / X 4 H I V appeared to be less likely to experience virologic failure (Hazard Ratio [HR] 0.7, p=0.012) after initial suppression (Figure l c and Table 4c). However, in multivariate analyses adjusting for baseline parameters, H I V co-180 receptor usage was not significantly associated with either virologic or immunologic response after initiation of H A A R T (p>0.1, Tables 4b and 4c). Analysis of combined CCR5 genotype and HIV co-receptor phenotype to predict therapy outcome Combining data from both C C R 5 genotype and H I V co-receptor phenotype further clarified the association between these parameters and survival following therapy initiation (Figure Id). In these analyses, the reference group was defined as subjects with the human C C R 5 wt/wt genotype harboring exclusively R5 H I V . In univariate analyses, subjects with the C C R 5 wt/A32 genotype harboring exclusively R5 H I V were at decreased risk of death (HR=0.54, p=0.09) compared with the reference group. In contrast, subjects with the C C R 5 wt/wt genotype harboring R 5 / X 4 H I V were at significantly increased risk of death when compared to the reference group (HR 1.54, p=0.04). Subjects with the C C R 5 wt/A32 genotype harboring R 5 / X 4 H I V were at comparable risk of mortality with respect to the reference group (HR 1.09, p=0.8). However, in multivariate analyses adjusting for baseline parameters including age, plasma viral load and C D 4 count, these associations were not significant (p>0.1, data not shown). 181 8.4 D I S C U S S I O N A N D C O N C L U S I O N We investigated the prevalence and clinical correlates of X 4 H I V in a large population of antiretroviral naive individuals initiating first H A A R T . Consistent with results from other cross-sectional studies [28,29,30,31] we observed a strong correlation between the presence of CXCR4-us ing H I V and clinical parameters including viral load, diagnosis of A I D S defining-illnesses, and most strikingly, C D 4 count. In this study, C D 4 counts from subjects harboring X 4 variants were on average three times lower than those from subjects harboring exclusively R5 variants. Genotypic (sequence-based) and phenotypic HW co-receptor usage assays as predictors of therapy response We previously undertook a study on the same cohort that identified the H I V V3 loop sequence as a significant predictor of survival and C D 4 response after initiating antiretroviral therapy [25]. A s expected, we observed a highly significant correlation between positively-charged amino acids at key residues of the H I V V 3 loop (an T1/25' genotype), and the presence of X 4 H I V at baseline [25,26]. In the present study, the T1/25' genotype predicted the presence of C X C R 4 - u s i n g virus with a specificity of 93% and a sensitivity of 33%. This value appears to be considerably lower than other studies employing V 3 sequence-based interpretation methods [32,33], however this may be explained by the fact that our V 3 genotypes were obtained using bulk P C R and sequencing of nucleic material isolated directly from clinical isolates, rather than analysis of cloned P C R products. A s population-based sequencing methods detect only the predominant circulating species, the prevalence of X 4 variants may be underestimated i f they are present as minority variants. H I V Genotype-Phenotype correlations wi l l be examined in detail in a further study. 182 However, somewhat in contrast to the previous study linking baseline V 3 sequence to therapy outcome [25], the present study suggests that H I V co-receptor use, although strongly associated with baseline clinical parameters, may not be independently associated with response to treatment. H I V co-receptor use at baseline did not remain an independent predictor of therapy response after adjusting for the strong association with baseline clinical parameters including C D 4 count and viral load. Association between CCR5A32 genotype and HIV co-receptor usage One hypothetical but potentially important consequence of the administration of C C R 5 antagonists is the selection of CXCR4-us ing H I V variants [34]. The naturally occurring C C R 5 A 3 2 deletion, associated with improved prognosis and slower H I V disease progression [35,36], may provide useful insights into this issue. C C R 5 wt/A32 individuals have only one functioning C C R 5 allele and express C C R 5 at much reduced levels [37], and therefore if reduced cell-surface C C R 5 availability indeed selects for CXCR4-us ing variants, one would expect to observe an increased prevalence of X 4 H I V in these individuals [38]. Indeed, our data indicate that after accounting for other baseline parameters, individuals with the C C R 5 wt/A32 genotype are nearly twice as likely to harbor R 5 / X 4 H I V than C C R 5 wt/wt individuals, although other studies have reported no correlation [39]. Although our results suggest that limited cell-surface C C R 5 may lead to selective pressure favoring the selection of CXCR4-us ing variants, one must be cautious in drawing conclusions from cross-sectional data. In addition we must acknowledge the potential bias resulting from the possibility that C C R 5 wt/A32 individuals harboring CXCR4-us ing variants may be over-represented in our cohort due to poor survival of X4-carriers with a C C R 5 wt/wt genotype. 183 Whether this observation may be relevant to the administration of C C R 5 co-receptor antagonists remains to be determined. When taking into consideration previous data linking the C C R 5 A 3 2 deletion with slower H I V disease progression [35,36] and the H I V X 4 phenotype with faster disease progression [13,14,15,16], the possibility that individuals with the C C R 5 A 3 2 genotype are more likely to harbor X 4 H I V appears contradictory. However, an analysis of the combined C C R 5 genotype and H I V co-receptor phenotype data indicate that the association between C C R 5 wt/A32 genotype and improved baseline clinical profile as well as survival after initiation of antiretroviral therapy is observed only when the predominant circulating virus at baseline is exclusively R5 (Table 2). The protective effects of the C C R 5 A 3 2 deletion appear to be largely lost when X 4 variants are present at baseline, consistent with previous reports [39,40]. However it is interesting to note that individuals with a C C R 5 wt/A32 genotype harboring R 5 / X 4 variants have a higher baseline viral load than C C R 5 wt/wt individuals with R 5 / X 4 variants (Table 2), but slightly better survival following initiation of H A A R T (figure Id). A potential explanation is that the C C R 5 wt/A32 genotype may still confer a slight survival advantage in the presence of X4-containing variants, however we must caution that the C C R 5 wt/A32 R 5 / X 4 group is relatively small and again the potential biases resulting from cross-sectional data and potential under-representation of X4-carriers with a C C R 5 wt/wt genotype in this cohort must be acknowledged. Limitations of this study include the fact that associations between H I V co-receptor usage and clinical parameters are based on cross-sectional data from a population of therapy-naive individuals initiating their first antiretroviral treatment. A s the contemporary guidelines for when to initiate therapy were predominantly based on the presence of A I D S -184 defining illnesses or C D 4 - and viral load thresholds, our study population may not be representative of the HIV-infected, antiretroviral naive population in general. In addition, because of the cross-sectional nature of this study, we are unable to comment upon the incidence of X 4 variants, the length of time these variants have been present, or on the evolution of H I V co-receptor usage over the natural history of infection. Similarly as seroconversion dates are not known and historical plasma samples not generally available we also are unable to investigate what proportion of our study group may have acquired X 4 variants at transmission. Finally, because population-based genotyping and phenotyping approaches were utilized, we are unable to distinguish whether subjects with R 5 / X 4 H I V harbor dual-tropic variants, a mixture of R5 and X 4 variants, or both. Despite these limitations, our study is one of the first to characterize the epidemiology and clinical predictors of H I V co-receptor usage in a large therapy naive population and present longitudinal clinical outcome data after initiation of therapy. In addition it is one of the largest studies evaluating associations between H I V V 3 genotype, co-receptor phenotype, human C C R 5 A 3 2 genotype and other clinical parameters. Strong associations between R 5 / X 4 H I V and clinical markers of H I V disease progression including higher plasma viral load and significantly lower C D 4 counts confirm the well-characterized link between X 4 H I V and poorer prognosis [13,14,15,16], although the direction of causation still remains unclear. Despite the association of X 4 H I V with poorer pre-therapy prognosis, however, the presence of CXCR4-us ing H I V at baseline did not remain an independent predictor of survival or clinical response to H A A R T after adjustment for baseline parameters including C D 4 count and plasma viral load. 185 Table 8.1a: Patient Characteristics at baseline, stratified by HIV co-receptor usage. [IQR]: Interquartile range Baseline Characteristics Total study group Stratified by co-receptor usage data available Stratified by HIV co-receptor use (N=979 samples for which co-receptor data available) (N=1191) Co-receptor data not available (N=212) Co-receptor data available (N=979) P R5 (N=801) R5/X4 (N=178) P Male Gender 1004 169 835 0.043 678 157 0.225 (%) (84.3%) (79.7%) (85.3%) (84.6%) (88.2%) Median Age, years 37 36 37 0.076 37 38 0.632 [IQR] [32-44] [31-43] [32-44] [32-44] [33-43] Median HIV pVL (x103 copies/ml) [IQR] 120 [42-310] 61 [13-190] 130 [55-340] <0.0001 120 [49-310] 175 [84-420] 0.0006 Median CD4 count (cells/mm3) [IQR] 280 [130-420] 330 [155-455] 260 [120-420] 0.007 290 [160-430] 110 [30-260] <0.0001 Baseline AIDS diagnosis (%) 154 (12.9%) 24 (11.3%) 131 (13.4%) 0.419 91 (11.4%) 40 (22.5%) <0.0001 History of injection drug use (%) 321 (27.0%) 67 (31.6%) 268 (27.4%) 0.214 229 (28.6%) 39 (21.9%) 0.071 CCR5 genetics 171 of 1177 43 of 210 128 of 967 0.007 94 of 791 34 of 176 0.0085 (% CCR5 wt/A32) (14.5%) (20.5%) (13.2%) (11.9%) (19.3%) Note: N=967 with CCR5 and phenotype data available V3 '11/25' genotype 130 of 1117 21 Of 162 109 of 955 0.570 54 of 787 55 Of 168 <0.0001 (% 11/25 genotype) (11.6%) (13.0%) (11.4%) (6.9%) (32.7%) Note: N=955 with V3 geno and pheno data available 186 Table 8.1b: Distribution of R5 /X4 HIV by baseline (pre-therapy) C D 4 count strata Baseline co-receptor usage Baseline CD4 count strata (cells/mm3) <25 25-49 50-99 100-199 200-349 350-499 500+ TOTAL R5 HIV 36 31 61 120 236 175 142 801 (% of CD4 stratum) 45.6% 68.9% 74.4% 71.9% . 90.8% 90.7% 92.8% (81.8%) R5/X4 HIV 43 14 21 47 24 18 11 178 (% of CD4 stratum) 54.4% 31.1% 25.6% 28.1% 9.2% 9.3% 7.2% (18.2%) Total for each CD4 stratum 79 45 82 167 260 193 153 979 187 Table 8.2: Patient Characteristics at baseline, stratified by HIV co-receptor usage and human CCR5A32 genotype. [IQR]: Interquartile range Baseline Characteristics Stratified by HIV co-receptor usage and CCR5A32 genotype (N=967) CCR5 wt/wt CCR5 wt/A32 R5 (N=697) R5/X4 (N=142) R5 (N=94) R5/X4 (N=34) P (any differences) Male Gender 584 125 85 30 0.23 . (%) (83.8%) (88.0%) (90.4%) (88.2%) Median Age, years 37 37 39 38 0.31 [IQR] [32-43] [32-44] [33-45] [34-47] Median HIV pVL (x103 copies/ml) [IQR] 130 [54-340] 165 [84-400] 78 [29-142] 295 [86-530] <0.0001 Median CD4 count (cells/mm3) [IQR] 280 [150-430] 120 [30-260] 360 [250-490] 105 [30-300] <0.0001 Baseline AIDS diagnosis 80 33 10 7 0.001 (%) (11.5%) (23.2%) (10.6%) (20.6%) History of injection drug use (%) 202 (29.0%) 30 (21.3%) 24 (25.5%) 8 (23.5%) 0.25 V3 genotype 50 of 685 40 of 133 4 of 92 13 of 33 <0.0001 (% '11/25' genotype) (7.3%) (30.1%) (4.4%) (39.4%) (N=943 with CCR5, phenotype and genotype data available) 188 Table 8.3: Univariate and Multivariate predictors of the presence of X4-using HIV variants at baseline Baseline Risk Factor Univariate Multivariate (N=943) Odds Ratio 95% CI P Odds Ratio 95% CI P Male gender 1.36 0.83-2.22 0.227 _ Age (per 10 year increment) 1.04 0.87-1.24 0.705 - - -Baseline HIV RNA (per log incr) 1.62 1.19-2.20 0.002 1.46 1.02-2.08 0.040 Baseline CD4 cell count (per 100 cell deer) 1.52 1.69-1.35 <0.0001 1.53 1.34-1.73 <0.0001 Baseline AIDS diagnosis 2.26 1.50-3.42 0.0001 1.17 0.72-1.91 0.531 History of injection drug use 0.70 0.48-1.03 0.072 - - -CCR5 wt/A32 genotype 1.78 1.15-2.73 0.009 2.48 1.48-4.15 0.0005 HIVV3'11/25' genotype 6.61 4.32-10.1 <0.0001 9.11 5.54-14.98 <0.0001 189 Table 8.4a: Influence of baseline factors on time to non-accidental death after initiation of HAART Baseline Risk Factor Univariate Multivariate (N=967) Hazard Ratio 95% CI P Hazard Ratio 95% CI P Presence of X4 HIV variants 1.45 1.00-2.13 0.053 1.14 0.76-1.70 0.525 male gender 0.91 0.60-1.38 0.661 - - -age (per 10 year increment) 1.45 1.24-1.69 <0.0001 1.48 1.25-1.76 <0.0001 baseline HIV RNA (per log increment) 1.67 1.28-2.19 0.0002 1.38 0.99-1.92 0.056 baseline CD4 cell count (per 100 cell deer) 1.27 1.16-1.39 <0.0001 1.32 1.18-1.48 <0.0001 % of time on antiretrovirals (per 10% incr) 0.86 0.83-0.91 <0.0001 0.80 0.76-0.85 <0.0001 baseline AIDS diagnosis 1.71 1.17-2.51 0.006 1.11 0.70-1.76 0.648 History of injection drug use 1.01 0.72-1.41 0.976 - - -PI vs. NNRTI -containing therapy 1.28 0.87-1.87 0.213 - - -CCR5 wt/A32 genotype 0.64 0.38-1.07 0.087 0.74 0.41-1.32 0.302 Table 8.4b: Influence of baseline factors on time to C D 4 decline below baseline after initiation of HAART Baseline Risk Factor Univariate Multivariate (N=967) Hazard Ratio 95% CI P Hazard Ratio 95% CI P Presence of X4 HIV variants 0.81 0.65-1.02 0.074 1.17 0.93-1.48 0.185 male gender 0.74 0.61-0.91 0.003 0.98 0.78-1.23 0.839 age (per 10 year increment) 0.95 0.87-1.03 0.196 - - -baseline HIV RNA (per log increment:) 0.76 0.68-0.84 <0.0001 0.91 0.78-1.06 0.236 baseline CD4 cell count (per 100 cell deer) 0.84 0.83-0.86 <0.0001 0.84 0.80-0.87 <0.0001 % of time on antiretrovirals (per 10% incr) 0.83 0.81-0.85 <0.0001 0.86 0.83-0.88 <0.0001 baseline AIDS diagnosis 0.39 0.29-0.53 <0.0001 0.64 0.45-0.91 0.013 History of injection drug use 1.63 1.39-1.91 <0.0001 1.42 1.18-1.70 0.0002 PI vs. NNRTI -containing therapy 0.80 0.67-0.95 0.009 0.76 0.62-0.93 0.0064 CCR5 wt/A32 genotype 1.08 0.87-1.33 0.505 1.02 0.80-1.30 0.870 190 Table 8.4c: Influence of baseline factors on time to pVL failure >500 copies/ml (after initial suppression of pVL <500 copies/ml) Baseline Risk Factor Univariate Multivariate (N=751)* Hazard Ratio 95% CI P Hazard Ratio 95% CI P Presence of X4 HIV variants 0.72 0.56-0.93 0.012 0.85 0.65-1.11 0.227 male gender 0.70 0.55-0.88 0.002 0.80 0.62-1.06 0.116 age (per 10 year increment) 0.88 0.80-0.97 0.008 0.94 0.84-1.04 0.227 baseline HIV RNA (per log increment) 0.98 0.87-1.11 0.775 - -baseline CD4 cell count (per 100 cell deer) 1.09 1.06-1.13 <0.0001 0.95 0.91-1.00 0.031 % of time on antiretrovirals (per 10% incr) 0.88 0.85-0.91 <0.0001 0.90 0.86-0.94 <0.0001 baseline AIDS diagnosis 0.56 0.43-0.74 <0.0001 0.67 0.48-0.93 0.016 History of injection drug use 1.44 1.20-1.74 0.0001 1.17 0.95-1.45 0.150 PI vs. NNRTI -containing therapy 0.91 0.75-1.10 0.320 - - -CCR5 wt/A32 genotype 0.81 0.64-1.03 0.085 0.89 0.68-1.16 0.387 * note that subjects who did not achieve initial suppression of p V L <500 copies H I V R N A / m l were censored at time of H A A R T initiation 191 Figure 8.1, panels A -D (Next page): Influence of HIV co-receptor usage on survival and clinical outcomes after initiation of first triple antiretroviral therapy. Kaplan-Meier analyses of the influence of baseline H I V co-receptor usage (R5 vs. R5 /X4 virus) on time to non-accidental death (panel A ) ; time to C D 4 decline below baseline (panel B) ; and time to virologic rebound >500 copies/ml (panel C) after initiation of first triple antiretroviral therapy, over a median follow-up time of 56 months. Individuals with R5 HIV are represented by the circles (o), while individuals with R 5 / X 4 H I V are indicated by the plus (+) signs. Panel D is a Kaplan-Meier analysis of the influence of combined human C C R 5 A 3 2 genotype and H I V co-receptor usage on time to non-accidental death following initiation of therapy. Categories are as follows: human C C R 5 wt/A32 genotype with R5 H I V (plus [+] signs), human C C R 5 wt/wt genotype with R5 H I V (stars [*]), human C C R 5 wt/A32 genotype with R 5 / X 4 H I V (circles [o]), human C C R 5 wt/wt genotype with R 5 / X 4 H I V (triangles [A]). 192 193 8.5 R E F E R E N C E S 1. Alkhatib G , Combadiere C , Broder C C , et al. C C C K R 5 : a R A N T E S , MlP- la lpha , M l P - l b e t a receptor as a fusion cofactor for macrophage-tropic H I V - 1 . Science 1996; 272: 1955-58. 2. Deng H , L i u R, Ellmeier W , et al. Identification of a major co-receptor for primary isolates of H I V - 1 . Nature 1996; 381: 661-66. 3. Berger E A , Doms R W , Fenyo E M , et al. A new classification for H I V - 1 . Nature 1998; 391: 240. 4. Dragic T, L i twin V , Al laway G P , et al. HIV-1 entry into CD4+ cells is mediated by the chemokine receptor C C - C K R - 5 . Nature 1996; 381: 667-73. 5. Feng Y , Broder C C , Kennedy P E , Berger E A . HIV-1 entry cofactor: functional c D N A cloning of a seven-transmembrane, G protein-coupled receptor. Science 1996; 272: 872-77. 6. Wil ley SJ, Reeves JD, Hudson R, et al. Identification of a subset of human immunodeficiency virus type 1 (HIV-1), H I V - 2 , and simian immunodeficiency virus strains able to exploit an alternative coreceptor on untransformed human brain and lymphoid cells. J V i r o l 2003; 77: 6138-52. 7. Zhang Y J , Dragic T, Cao Y , et al. Use of coreceptors other than C C R 5 by non-syncytium-inducing adult and pediatric isolates of human immunodeficiency virus type 1 is rare in vitro. J V i r o l 1998; 72: 9337-44. 8. Lee S, Tiffany H L , K i n g L , Murphy P M , Golding H , Zaitseva M B . C C R 8 on human thymocytes functions as a human immunodeficiency virus type 1 coreceptor. J Vi ro l 2000; 74: 6946-52. 194 9. Schuitemaker H , Kootstra N A , de Goede R E , de W o l f F , Miedema F, Tersmette M . Monocytotropic human immunodeficiency virus type 1 (HIV-1) variants detectable in all stages of HIV-1 infection lack T-cell line tropism and syncytium-inducing ability in primary T-cell culture. J V i r o l 1991; 65: 356-63. 10. Berger E A , Murphy P M , Farber J M . Chemokine receptors as HIV-1 coreceptors: roles in viral entry, tropism, and disease. Annu Rev Immunol 1999; 17: 657-700. 11. Connor RI , Ho D D . Human immunodeficiency virus type 1 variants with increased replicative capacity develop during the asymptomatic stage before disease progression. J V i r o l 1994; 68: 4400-08. 12. Glushakova S, Y i Y , Grivel JC , et al. Preferential coreceptor utilization and cytopathicity by dual-tropic HIV-1 in human lymphoid tissue ex vivo. J C l i n Invest 1999; 104: R7-11. 13. Tersmette M , Gruters R A , de Wol f F, et al. Evidence for a role of virulent human immunodeficiency virus (HIV) variants in the pathogenesis of acquired immunodeficiency syndrome: studies on sequential H I V isolates. J V i ro l 1989; 63: 2118-25. 14. Schuitemaker H , Koot M , Kootstra N A , et al. Biological phenotype of human immunodeficiency virus type 1 clones at different stages of infection: progression of disease is associated with a shift from monocytotropic to T-cell tropic virus populations. J V i r o l 1992; 66:1354-60. 195 15. Koot M , Keet IP, Vos A H . Prognostic value of HIV-1 syncytium-inducing phenotype for rate of CD4+ cell depletion and progression to A I D S . A n n Int M e d 1993; 118: 681-88. 16. Richman D D , Bozzette S A . The impact of the syncytium-inducing phenotype of human immunodeficiency virus on disease progression. J Infect Dis 1994; 169: 968-74. 17. Rusconi S, De Pasquale M P , Mainini F, et al. Vi ra l load, viral phenotype modification, zidovudine susceptibility and reverse transcriptase mutations during the first 6 months of zidovudine monotherapy in HIV-1-infected people. Antivir Ther 1996;1:211-19. 18. Katzenstein T L , Nielsen C , Bruun L , Mathiesen L R , Pedersen C , Gerstoft J. Quantification of HIV-1 R N A during antiretroviral therapy: association with viral phenotype and development of resistance. Antivir Ther 1996; 1: 246-54. 19. Vida l C , Garcia F, Gatell J M , et al. Predictive factors influencing peak viral load drop in response to nucleoside reverse transcriptase inhibitors in antiretroviral-naive H I V -1-infected patients. J Acquir Immune Defic Syndr Hum Retroviral 1998; 19: 55-60. 20. Davey R T Jr, Chaitt D G , Reed G F , et al. Randomized, controlled phase I/II, trial of combination therapy with delavirdine (U-90152S) and conventional nucleosides in human immunodeficiency virus type 1-infected patients. Antimicrob Agents Chemother 1996; 40:1657-64. 21. Hughes M D , Johnson V A , Hirsch M S , et al. Monitoring plasma HIV-1 R N A levels in addition to CD4+ lymphocyte count improves assessment of antiretroviral therapeutic 196 response. A C T G 241 Protocol Virology Substudy Team. A n n Intern Med 1997; 126: 929-38. 22. Shaheen F, Collman R G . Co-receptor antagonists as HIV-1 entry inhibitors. Curr Opin Infect Dis 2004; 17: 7-16. 23. Moore JP, Kitchen S G , Pugach P, Zack J A . The C C R 5 and C X C R 4 coreceptors-central to understanding the transmission and pathogenesis of human immunodeficiency virus type 1 infection. A I D S Res Hum Retroviruses 2004; 20:111-26. 24. Hogg R S , Y i p B , Chan K J , et al. Rates of disease progression by baseline C D 4 cell count and viral load after initiating triple-drug therapy. J A M A 2001; 286: 2568-77. 25. Brumme Z L , Dong W W , Y i p B , et al. Clinical and immunological impact of H I V envelope V 3 sequence variation after starting initial triple antiretroviral therapy. A I D S 2004; 18: F l - 9 . 26. Fouchier R A , Brouwer M , Broersen S M , Schuitemaker H . Simple determination of Human Immunodeficiency Virus Type 1 Syncytium-inducing V 3 Genotype by P C R . J C l i n M i c b 1995; 33: 906-11. 27. Hogg R S , Heath K , Bangsberg D , et al. Intermittent use of triple-combination therapy is predictive of mortality at baseline and after 1 year of follow-up. A I D S 2002; 16: 1051-58. 28. Bozzette S A , McCutchan J A , Spector S A , Wright B , Richman D D . A cross-sectional comparison of persons with syncytium- and non-syncytium-inducing human immunodeficiency virus. J Infect Dis 1993; 168: 1374-79. 197 29. Furrer H , Wendland T, Minder C, et al. Association of syncytium-inducing phenotype of HIV-1 with C D 4 cell count, viral load and sociodemographic characteristics. A I D S 1998; 12: 1341-46. 30. Moyle G , Wildfire A , Mandalia S, et al. Prevalence and predictive factors for C C R 5 and C X C R 4 co-receptor usage in a large cohort of H I V positive individuals. Abstract H-1135, 44 t h International Conference on Antimicrobial Agents and Chemotherapy, October 30-November 2, 2004, Washington, D C . 31. Demarest J, Bonny T, Vavro C, et al. HIV-1 co-receptor tropism in treatment naive and experienced subjects. Abstract H - l 136, 44 t h International Conference on Antimicrobial Agents and Chemotherapy, October 30-November 2, 2004, Washington, D C . 32. Jensen M A , L i FS , van't Wout A B , et al. Improved coreceptor usage prediction and genotypic monitoring of R5-to-X4 transition by motif analysis of human immunodeficiency virus type 1 env V 3 loop sequences. J V i r o l 2003; 77:13376-88. 33. Resch W , Hoffman N , Swanstrom R. Improved success of phenotype prediction of the human immunodeficiency virus type 1 from envelope variable loop 3 sequence using neural networks. Virology 2001; 288:51-62. 34. Mosier D E , Picchio G R , Gul iz ia RJ , et al. Highly potent R A N T E S analogues either prevent CCR5-using human immunodeficiency virus type 1 infection in vivo or rapidly select for CXCR4-us ing variants. J V i ro l 1999; 73:3544-50. 35. Dean M , Carrington M , Winkler C , et al. Genetic restriction of HIV-1 infection and progression to A I D S by a deletion allele of the C K R 5 structural gene. Science 1996; 273:1856-62. 198 36. Eugen-Olsen J, Iversen A K , Garred P, et al. Heterozygosity for a deletion in the C K R - 5 gene leads to prolonged AIDS-free survival and slower C D 4 T-cell decline in a cohort of HIV-seropositive individuals. A I D S 1997; 11:305-10. 37. W u L , Paxton W A , Kassam N , et al. C C R 5 levels and expression pattern correlate with infectability by macrophage-tropic H I V - 1 , in vitro. J Exp Med 1997; 185:1681-91. 38. D 'Aqui la R T , Sutton L , Savara A , Hughes M D , Johnson V A . CCR5/delta(ccr5) heterozygosity: a selective pressure for the syncytium-inducing human immunodeficiency virus type 1 phenotype. J Infect Dis 1998; 177: 1549-53. 39. de Roda Husman A M , Koot M , Cornelissen M , et al. Association between C C R 5 genotype and the clinical course of HIV-1 infection. A n n Intern M e d 1997; 127:882-90. 40. Michael N L , Chang G , Louie L G , et al. The role of viral phenotype and C C R - 5 gene defects in HIV-1 transmission and disease progression. Nat M e d 1997; 3: 338-40. 199 CHAPTER 9 GENERAL DISCUSSION AND CONCLUSION 9.1 Introduction Much attention has been focused on the elucidation of the role of host and viral genetic factors in the pathogenesis of H I V / A I D S . It is now appreciated that a wide variety of host genetic factors play a significant role in the natural course of H I V disease, and influence everything from the risk of infection upon exposure [1,2,3,4], the rate of disease progression once infected [5,6,7,8], and the strength and diversity of the immune response [8,9,10]. Similarly, H I V genetic diversity has also been shown to modulate viral infectivity [11], evasion of the host immune response [12,13], disease progression [14,15], and rate of T-cell death and subsequent decline of immune function [16]. The introduction of H A A R T in the mid-1990s revolutionized the treatment of H I V / A I D S , at least in areas of the world where access to antiretrovirals was available. Treatment with H A A R T resulted in reductions in plasma viral load to undetectable levels using sensitive assays [17,18,19], increases in C D 4 counts, and improvement of immune function. This resulted in dramatically decreased incidence of morbidity due to A I D S -defining illnesses and opportunistic infections, and most importantly, significantly prolonged survival [20,21,22,23]. The initial success of treatment with triple combination antiretroviral therapies was such that some postulated the eradication of H I V / A I D S after only a few years of treatment [24], although it is now clear that this was on overly optimistic prediction. Ten years later, it is evident that despite the successes seen as a result of H A A R T , many challenges and outstanding questions remain. For example, for reasons that are poorly understood, approximately 10% of individuals fail to respond to H A A R T , despite 200 apparently good adherence and no evidence of resistance mutations [25]. Among those who initially achieve treatment success under H A A R T , subsequent virologic failure rates approach 20% in previously treatment-naive individuals [25] and 30-50% in previously treated individuals [25,26]. Other individuals exhibit what is known as a "discordant" response - meaning that while improvements in C D 4 counts and/or immune function are achieved, virologic suppression is not achieved [27,28,29], or vice-versa [30]. Even among those individuals who maintain successful virologic and immune control of H I V under H A A R T , it is now clear that antiretroviral therapy is a lifelong commitment and that treatment success must be maintained. In addition, emerging challenges and toxicities associated with long-term antiretroviral therapy include l ipid abnormalities [31], mitochondrial toxicities [32], cardiovascular [33], renal [34] and hepatic [35] toxicities, antiretroviral resistance [36] as well as others [37]. Given that long-term treatment success is essential, it is of importance that we achieve a greater understanding of the factors that contribute to the success of antiretroviral treatment, including genetic factors which may influence response to H A A R T . The identification of factors which influence antiretroviral therapy response may be of direct relevance in improvement of the lives of individuals living with H I V / A I D S . For example, i f predictors of treatment failure are defined, then individuals at risk of treatment failure could be identified and interventions designed to optimize treatment. Furthermore, the identification of host and viral genetic factors that influence therapy response may lead to the development of novel, genotype-based tools for the clinical management of HIV-infected individuals. A s described earlier, the objectives of this thesis were to investigate whether previously-identified host and viral genetic factors exert an influence on the clinical course of 201 H I V / A I D S in the context of H A A R T , as well as to investigate additional genetic markers that may be of direct relevance to H A A R T outcomes. This chapter summarizes the major conclusions of each of the individual research papers, and discusses these conclusions in context of new studies that have been published since the undertaking of the studies presented here. This w i l l be followed by a discussion of the broader implications of this work, as well as the potential application of these results in modern clinical practice. 9.2 Human Genetic Variation: the chemokine coreceptors a) CCR5A32 As already discussed, a 32-base pair deletion in the human C C R 5 chemokine receptor gene is known to confer protection against H I V infection (in individuals homozygous for the deletion) [38,39,40] and contribute to slower H I V disease progression (in individuals heterozygous for the deletion) [39,40]. While some studies reported significantly improved short-term virologic and/or immunologic responses to H A A R T among CCR5wt /A32 individuals [41,42,43], others reported no association between C C R 5 genotype and treatment outcome [44,45]. In most cases, however, these studies were limited to the evaluation of the effect of C C R 5 A 3 2 on short-term clinical outcomes. In order to more conclusively investigate this issue, we undertook a large, population-based study of >1000 HIV-infected antiretroviral naive individuals initiating first H A A R T , and sought to evaluate what effect, if any, the C C R 5 A 3 2 genotype exerted on response to H A A R T over a longer-term follow-up [46, chapter 3]. After adjustment for baseline sociodemographic and clinical factors known to affect therapy response, individuals heterozygous for the C C R 5 A 3 2 deletion experienced significantly more rapid initial suppression of plasma H I V R N A below 400 copies/ml, however there was no association 202 between C C R 5 genotype and the subsequent duration of viral suppression. Similarly, we observed no association between C C R 5 genotype and initial immunologic response, as measured by the time to a decline of C D 4 count to below baseline. In an analysis of longer-term outcomes, we observed a weak trend, although not statistically significant, to improved survival in individuals heterozygous for the C C R 5 A 3 2 deletion, when compared to C C R 5 wild-type subjects, over a 5-year follow-up period after initiating H A A R T [46, chapter 3]. Although a single study can by no means be conclusive, results from this large population-based analysis strongly suggest that any protective effect conferred by CCR5A32 alone in the context of H A A R T may not be clinically significant on an individual basis. The results of this analysis alone, therefore, do not support the utility of the C C R 5 A 3 2 genotype as an independent clinical prognostic marker for H A A R T response. However, discussed later, additional evidence from our group as well as others suggests that the C C R 5 genotype may be of useful prognostic value when taken in context with other viral genetic markers, namely the coreceptor usage of the predominant circulating virus [47,48,49,50], or when evaluated in context with other host factors [51,52]. b) CX3CR1 Despite the correlations observed between C C R 5 A 3 2 and the natural course of untreated H I V disease, it was clear that C C R 5 genetics could at most account for only a small proportion of the variation in H I V disease progression observed in HIV-infected populations. A number of studies identified additional genetic polymorphisms among chemokine ligand and receptor genes relevant to H I V disease, including a study by Faure et al reporting rapid progression to A I D S in individuals homozygous for an I249/M280 amino acid haplotype in the human CX3CRI chemokine receptor gene, when compared to 203 individuals with other haplotypes [53]. Since the publication of the original study [53] (as well as the publication of our research study investigating the effects of the C X 3 C R I I249/M280 variation on response to H A A R T [54, chapter 4]), these researchers have subsequently confirmed their findings in a larger, population-based study [55], although another group, investigating the M280 polymorphism only, reported no significant correlation between C X 3 C R I M280 and severity of H I V disease [56]. The specific role that C X 3 C R I plays in the course of H I V disease, as well as the specific mechanisms by which the I249/M280 amino acid haplotype contributes to more rapid disease progression remain largely unknown. Although it is known that C X 3 C R I functions as a minor coreceptor for H I V entry into target cells [57], it is not known whether this is a major mechanism by which C X 3 C R I exerts its influence on H I V disease progression. It is more likely that the natural immune functions of C X 3 C R I (which include adhesion and migration of leukocytes [58] as well as regulation of CD8+ T-cell function [59]) play a larger role in defining the role of C X 3 C R I in H I V pathogenesis. It is now known that H I V infection induces a massive increase in the expression of C X 3 C R I on CD4+ and CD8+ T-cells [58, 59], and indeed that increased C X 3 C R I expression as a result of H I V infection is associated with declining C D 4 T-cell number and increasing plasma viral load [59]. This, combined with the finding that fractalkine, the natural ligand for C X 3 C R I , displays significantly reduced binding affinity to peripheral blood mononuclear cells from C X 3 C R I I249/M280 homozygotes when compared to cells expressing the "wild-type" V249/T280 receptor [55], suggests that the mechanism whereby the I249/M280 polymorphisms contribute to H I V pathogenesis may be due to alterations in receptor-ligand interactions and the immune consequences thereof. 204 Similarly, the mechanisms whereby the C X 3 C R I polymorphisms may contribute to virologic and immunologic control of H I V infection during treatment with H A A R T also remain poorly understood. Since the publication of our study reporting a statistically significant trend to earlier immunological failure following initiation of H A A R T in individuals with the C X 3 C R 1 1249 mutation [54, chapter 4], a recent study of HIV-infected children participating in randomized clinical trials of mono and dual therapy also reported a significant independent association of C X 3 C R I 1249 mutation with more rapid virologic and immunologic disease progression, even after adjustment for factors such as baseline C D 4 count and viral load [60]. To date, the data supporting the value of C X 3 C R I polymorphisms in predicting clinical prognosis and outcomes in both untreated and treated H I V infection is largely consistent and suggests that the C X 3 C R I 1249 polymorphism may be a useful prognostic indicator of more rapid H I V disease progression. Additional studies are needed in order to further elucidate the function of C X 3 C R I in H I V pathogenesis and evaluate the potential clinical utility of determining C X 3 C R I genotypes in treated and untreated HIV-infected individuals. 9.3 Pharmacokinetics: drug metabolism and the P-gp membrane efflux transporter A s outlined in detail in chapter 4, the human multidrug resistance ( M D R ) - l gene encodes the ATP-dependent membrane efflux transporter P- glycoprotein (P-gp), whose substrates include H I V protease inhibitors. A single, synonymous C4345T polymorphism has been identified in the M D R - 1 gene which affects membrane expression of the P-gp transporter [61], a polymorphism which presumably wi l l have consequences for the absorption, bioavailability and metabolism of H I V protease inhibitors, as well as potential consequences for antiretroviral treatment response. The consequences of altered P-gp 205 expression levels to antiretroviral concentrations and bioavailability in different body compartments are complex: differential expression levels of P-gp in the gastrointestinal tract could affect the intestinal absorption and thus bioavailability of drugs (including protease inhibitors), expression of P-gp in the liver and kidney could affect the rate at which drugs are metabolized/eliminated, expression of P-gp in the epithelial cells forming the blood-brain barrier limit entry of drugs in to the brain, and expression of P-gp on lymphocytes affects intracellular drug concentrations in target cells [62]. Indeed it has been shown that individuals homozygous for 3435T display significantly reduced P-gp expression and activity in the gastrointestinal tract [61], reduced P-gp function and expression in cells of the immune system [63,64], significantly reduced plasma levels of the PI nelfinavir [64], and significantly increased intracellular concentrations of nelfinavir [65], although other studies report somewhat contrasting results [66,67]. The debate as to whether the C3435T M D R - 1 polymorphisms influence response to antiretroviral treatment, however, has been much more controversial. Since Fellay et al reported a significant association between the homozygous 3435T genotype and improved C D 4 response to Pi-containing-regimens in 2002 [64], there have been a number of conflicting studies attempting to address this issue, the majority of which have failed to observe a strong correlation between M D R - 1 C3435T and treatment response. Our laboratory, in a population-based study of -450 antiretroviral naive individuals initiating their first H A A R T reported a trend, although not statistically significant, to earlier virologic failure in individuals with the M D R - 1 3435C/C genotype, but no correlation between M D R -1 genotype on immunologic response [54]. Since then, numerous studies have reported no correlation between M D R - 1 C3435T genotype and initial virologic and/or immunologic 206 response to combination therapy among antiretroviral-nai've individuals [68,69,70] as well as individuals with previous antiretroviral exposure [65]. A recent study of HIV-1 infected children reported higher plasma nelfinavir levels and improved virologic responses to H A A R T among subjects with the M D R - 1 3435C/T genotype when compared to those with the 3435C/C genotype [71]. The small number of study subjects with the 3435T/T genotype (n=7) displayed similar pharmacokinetic and virologic response profiles to those with the 3435C/C genotype [71]. These contrasting reports are perhaps not surprising given the fact that a single nucleotide polymorphism in a single human gene, however relevant, can realistically only account for a very small portion of the inter-individual differences observed pharmacokinetic profiles and treatment responses. This is especially true in the case of a polymorphism such as the M D R - 1 C3435T, a synonymous substitution that does not affect the amino acid sequence of the protein and thus most likely confers a protective effect indirectly, possibly through linkage with polymorphisms at other sites. In fact, the M D R - 1 C3435T polymorphism is now known to be in linkage disequilibrium with the M D R - 1 G2677T substitution, among others [72], and it is suggested that an M D R - 1 haplotype analysis, taking into consideration multiple polymorphisms on a single allele, may be superior than analysis of single nucleotide polymorphisms (SNPs) in predicting pharmacokinetics [72] and treatment response. Currently, the available evidence is not sufficient to draw a firm association between M D R - 1 C3435T and treatment response, and does not support the use of M D R - 1 genotyping in clinical practice. A s genotyping technologies continue to improve [73], however, and more understanding of the relative contributions of SNPs in M D R - 1 and other genes relevant 207 to the pharmacokinetics of antiretroviral agents is achieved, it is possible that M D R - 1 polymorphisms may in the future be interpreted in a broader genetic context. 9.4 Relevance of HIV Genetic Variation in the era of HAART In addition to host genetic factors, this thesis sought to additionally investigate the effects of genetic variation in the H I V genome on the same parameters and outcomes. In the two decades since the complete sequencing of the H I V genome [74], major advances have been made in the understanding of the structure and function of all nine of H I V ' s protein-encoding genes, as well as in the understanding of the implications of viral genetic variation to the pathogenesis of H I V . This understanding was initially largely achieved through in vitro mutagenesis experiments where H I V proteins and/or whole viruses harboring specific amino acid changes were engineered, expressed and analyzed in order to draw conclusions about the structural and functional consequences of specific genetic changes. The relevance of these in vitro data to H I V infection and disease progression in vivo could then be evaluated by investigating the prevalence and clinical consequences of specific genetic changes in H I V isolates derived from individuals at different disease stages. The combination of improvements in modern genotyping technologies, and the linkage of these sequence data to large datasets of clinical outcomes, now confers the capability to directly assess the in vivo effects of specific mutations/polymorphisms on the response to antiretroviral therapy, on a population basis. Currently there is more evidence supporting the clinical utility of incorporating H I V sequence-based assays (as opposed to human genotypic assays) into clinical practice, especially in the context of modern antiretroviral therapies. Indeed, viral genetic data are already extensively used in the clinical management of HIV-infected individuals in the form 208 of genotypic and phenotypic antiretroviral resistance assays, in order to guide decisions regarding treatment [36]. Given the extensive research currently focused on the clinical consequences of H I V genetic variation and sequence evolution, it is likely that novel H I V genotypic assays wi l l continue to be incorporated into routine clinical practice. The following section wi l l focus on summarizing the current evidence supporting (or not supporting) the relevance and potential clinical utility of additional specific H I V genetic markers in the context of H A A R T . 9.5 Antiretroviral resistance testing As has been discussed previously, the most important example of viral genetic variation directly related to antiretroviral treatment is the emergence of drug-resistant H I V . Drug-resistant H I V variants are a major cause of treatment failure and remain a formidable barrier to long-term antiretroviral efficacy [36,75,76]. In fact, it is estimated that virologic failure occurs in up to 20% of previously antiretroviral naive individuals [25] and in >30% of previously treated individuals within two years of initiation of antiretroviral therapy [25,26]. In addition, a survey of antiretroviral resistance among HIV-infected individuals in the United States estimates that greater than 75% of individuals with detectable plasma viremia during antiretroviral therapy harboured H I V with resistance to one or more antiretroviral agents [77]. However, although the impact of antiretroviral resistance (and, especially, multidrug resistance) on the efficacy of H A A R T and/or the availability of appropriate treatment options is indeed a major concern, it is comforting to note that multidrug resistance is not yet a significant contributor to HIV-related deaths at least in British Columbia [78,chapter 6]. In our province, the majority of deaths among HIV-infected persons appear to be largely confined to cases where antiretrovirals were not accessed or unable to be consistently taken 209 [78]. Although these results clearly illustrate the need for programs to improve both antiretroviral access and adherence, these results also confirm the continuing efficacy of H A A R T . The management of H I V is somewhat unique in the field of medicine because it represents one of the few examples where the evolving genotype of the infectious agent is routinely characterized and incorporated in clinical decision-making. In fact, genotypic resistance tests haven been shown to improve virologic response to therapy in some randomized controlled trials, and are currently recommended for use in a variety of settings including acute or recent infection, prior to the initiation of therapy, in cases of virologic failure while receiving antiretroviral therapy, and during pregnancy [36]. In light of the continual development of novel antiretroviral agents, as well as increased research into the impact of H I V genetic variation on disease progression and response to therapy, there wi l l undoubtedly continue to be a significant role for H I V drug resistance genotyping in clinical practice. In addition, it is likely that additional H I V genotyping assays, focusing on other regions of the genome [79,80], w i l l be incorporated into routine clinical practice as well. 9.6 HIVp6Gag The first investigation of H I V genetic parameters and response to H A A R T undertaken as a part of this thesis work focused on proline-containing insertion mutations within H I V p 6 G a g . Peters et al. reported that insertion mutations within the p 6 G a g proline-rich " P T A P " motif were associated with resistance to nucleoside analog reverse transcriptase inhibitors (NRTIs) [81]. Briefly, p 6 G a g insertions appeared to occur more frequently among individuals treated with NRTIs when compared to antiretroviral naive individuals, and when analyzed in an in vitro system, clones generated with a P T A P duplication exhibited increased 210 infectivity and a competitive growth advantage in the presence of NRTIs [81]. In order to investigate whether p 6 G a g insertions were indeed associated with N R T I resistance and/or poorer response to antiretroviral therapy, we retrospectively genotyped a cohort of antiretroviral-nai've individuals for H I V p 6 G a g insertions and analyzed virologic and immunologic therapy outcomes in subjects with and without p 6 G a g insertions [82, chapter 5]. As discussed, we observed a relatively high prevalence of p 6 G a g insertions in this antiretroviral naive population (-17%), suggesting that p 6 G a g insertions may represent naturally-occurring genetic variation among circulating H I V strains in untreated populations [82]. Furthermore we did not observe any correlation between p 6 G a g P T A P insertions and treatment response, nor with the presence of baseline N R T I resistance mutations [82]. Since the publication of our study [82] the observation that p 6 G a g P T A P insertions most likely represent natural polymorphisms is supported by results from Gallego et ai, who observed similar rates of P T A P insertions (-15%) among drug-naive and pretreated patients [83], as well as Ibe et al,. who reported that p 6 G a g insertions "selected" under H A A R T were in all cases already present prior to initiation of therapy, and therefore most likely represent naturally-occurring polymorphisms which may confer some kind of replicative and/or survival advantage in the presence of antiretrovirals [84]. With respect to the potential associations between p 6 G a g insertions and response to therapy, the only other evidence available to date are contrasting studies by Kaufmann et al, who reported a trend, although not statistically significant, between the presence of p 6 G a g insertions and decreased likelihood of virologic failure [85], and Lastere et al, who reported a trend, again not statistically significant, towards earlier virologic failure among heavily experienced patients harbouring H I V p 6 G a g P T A P insertions receiving amprenavir [86]. 211 O f interest, in our study, we also observed a significant correlation between the presence of p 6 G a g insertions, and the "11/25" H I V envelope genotype [82], which, i f confirmed in other independent studies, may further help clarify the relationship between H I V p 6 G a g genetic variation and H I V clinical prognosis and treatment response. To our knowledge, however, there have not yet been any additional studies investigating this correlation. To date, therefore, there is not enough evidence to conclusively evaluate the relevance of H I V p 6 G a g insertions in the context of H A A R T , although the available evidence suggests these insertions are not clinically significant. The implications of these insertions for H I V drug resistance, antiretroviral treatment response, and H I V envelope sequence variation merit further investigation. 9.7 HIV envelope The H I V envelope may be the most genetically variable region in the viral genome and is known to play a significant role H I V pathogenesis, disease progression, and evasion of the immune response. In addition, the H I V envelope is largely responsible for determining H I V co-receptor usage. Briefly, the ability of the virus to use either the C C R 5 or C X C R 4 coreceptors lies at least partially in the amino acid sequence of the third variable ("V3") region of the H I V envelope [87,88]. In general, clonal isolates able to use the C X C R 4 coreceptor display positively-charged amino acids at codons 11 and/or 25 of V 3 (an "11/25" genotype [89]), while isolates able to use the C C R 5 coreceptor do not [90]. Although much remains to be elucidated with respect to coreceptor evolution over the course of H I V disease, it is known that the appearance of viruses with the capacity to use the C X C R 4 receptor is associated with rapid immune decline and accelerated progression to A I D S and death [14,15,16]. 212 Despite the clear evidence indicating an important role of H I V envelope variation and/or coreceptor usage during untreated H I V infection, the relevance of H I V envelope variation and/or coreceptor usage in the H A A R T era has not been conclusively evaluated. A s a part of this thesis, therefore, we wished to investigate whether H I V envelope variation/coreceptor usage were independent markers of response to H A A R T in antiretroviral naive individuals. This section wi l l summarize our findings and conclusions on this subject, as well as present any new data that have been published since the completion of our studies. a) V3 loop genotyping A large-scale investigation of the relevance of V3 sequence variation in the context of H A A R T was undertaken as part of this thesis [89, chapter 7]. A s V 3 loop genotyping is relatively rapid and low-cost compared to currently available phenotyping assays, we wished to evaluate the potential clinical utility of the simplified H I V V 3 "11/25 genotype" as an independent prognostic factor for response to H A A R T . A n investigation of the baseline prevalence of the H I V envelope "11/25" genotype in the H O M E R cohort, a large cohort of antiretroviral-nai've individuals in British Columbia initiating H A A R T , revealed an "11/25 genotype" prevalence o f - 1 0 % [89]. Subjects with and without the "11/25 genotype" were comparable with respect to their baseline clinical and sociodemographic characteristics. However, despite the fact that subjects with and without the "11/25 genotype" appeared to have similar baseline clinical prognoses, the "11/25 genotype" was significantly and independently associated with more rapid immunologic decline, as well as poorer survival in the five-year period following initiation of H A A R T . In addition, there was evidence suggesting that the "11/25 genotype" may also negatively affect virologic response. If, in the virologic outcome analyses, individuals were censored at time 213 of switching therapy, a significant association between the "11/25" genotype and an increased time to viral suppression was observed. In addition, a longitudinal analysis of the median p V L over the course of study follow-up also suggested that long-term virologic control was more difficult to maintain among subjects with the "11/25" genotype [89]. Implications of HIV V3 genotype in the context of HAART The potential utility of H I V envelope genotyping in routine clinical practice has not been extensively evaluated. To our knowledge, ours is the first study investigating the implications of V 3 genotype on response to H A A R T in previously antiretroviral-nai've individuals. Taken together with ample evidence indicating that the presence of positively-charged amino acids at key sites in the H I V V3 loop are predictive of H I V disease progression in untreated infection [88,90], this study strongly suggests that H I V envelope genotype has the potential to provide useful clinical information, in addition to plasma viral load and C D 4 count, to help guide therapy-related decisions. Based on results showing poorer immunologic response to treatment, more therapy switches and poorer survival among individuals with "11/25" H I V , it may be argued that the "11/25 genotype" could be used as an indicator to initiate H A A R T at the higher end of currently recommended guidelines [76]. Alternatively, or in addition, the presence of "11/25" H I V could translate into a decision to choose a more potent initial treatment regimen, or the recommendation to monitor patients' responses to treatment more closely. Further evidence for the potential application of H I V V 3 loop genotyping in the modern treatment era is demonstrated by a recent study by Toulson et al [91] investigating the effect of various clinical parameters on the success and length of treatment interruptions among individuals with nadir C D 4 counts >200 cells/mm 3 . In this study, the presence of 214 "11/25" H I V , along with other clinical markers, was significantly and independently predictive of a rapid C D 4 decline during the treatment interruption, and the need to re-initiate H A A R T earlier when compared to the "non-11/25" genotype [91]. Results from this study indicate that V 3 genotyping might be useful as an independent prognostic marker in a variety of settings relating to antiretroviral treatment. b) HIV coreceptor phenotyping In a related study, we also sought to evaluate whether baseline H I V coreceptor phenotype may be a useful prognostic clinical marker for therapy response in the current era of H A A R T [47, chapter 8]. In addition, a baseline cross-sectional characterization of the prevalence and clinical correlates of CCR5-and CXCR4-us ing H I V in this cohort would be of relevance in the context of H I V coreceptor inhibitors, a new class of antiretroviral agents currently in development. A cross-sectional analysis of the baseline H I V coreceptor usage in the H O M E R cohort revealed that approximately 20% of subjects harboured X 4 variants at study entry, most (with the exception of one individual with exclusively X 4 H I V ) as a mixture of either R5 and X 4 variants and/or dual-tropic variants [47]. The presence of X4-using H I V at baseline was significantly correlated with poorer clinical prognosis in this cohort: subjects harboring X 4 H I V had significantly lower C D 4 counts, higher viral loads and higher incidences of AIDS-defining illnesses than individuals with exclusively R5 variants [47]. These findings are consistent with results of other recent cross-sectional analyses attempting to identify clinical correlates of X 4 H I V [92,93], and are also consistent with the large body of historical research linking the presence of X4/syncytium-inducing viral variants to accelerated disease progression during untreated infection [14,15,16]. 215 Perhaps surprisingly, however, coreceptor phenotype appeared to be less useful as an independent prognostic marker than V3 genotype in the context of H A A R T , mainly due to the very strong correlations between baseline coreceptor phenotype and the other major clinical markers of H I V disease progression [47]. In univariate analyses, the presence of X4-using H I V prior to the initiation of therapy was indeed correlated with reduced survival following initiation of H A A R T , but this association did not remain significant after adjusting for other baseline clinical parameters, most notably C D 4 count. These data indicate that coreceptor usage may not provide useful additional prognostic information for H A A R T response above and beyond the routine clinical markers already used. Results, therefore, do not support the incorporation of H I V co-receptor phenotype into clinical practice at this point in time, at least in the context of currently-available H A A R T regimens. 9.8 Unresolved research questions surrounding HIV envelope genotyping/phenotyping and evidence for potential future clinical utility of coreceptor assays a) Evolution of coreceptor phenotypes during treatment with HAART The potential evolution of coreceptor usage and/or envelope sequence under antiretroviral therapy, as well as the relevance of this evolution to the success of H A A R T remains to be conclusively determined. To date, this issue has been investigated in a number of longitudinal studies, most of which have focused on a small number of subjects as the cost of H I V coreceptor phenotyping remains quite high. The majority of studies to date report that antiretroviral therapy suppresses CXCR4-us ing strains and/or induces a shift to predominantly R5 strains in individuals with X 4 or mixed R 5 / X 4 strains at baseline [94,95,96], although one study reports that R5 outgrowth may be only a temporary phenomenon [97]. In contrast, other studies report no effect of H A A R T on co-receptor 216 phenotype [98]. Yet others report the emergence of H I V envelope sequences with the "11/25" genotype under effective antiretroviral therapy, in subjects that predominantly harbored "non-11/25" H I V envelope genotypes at baseline [99]. Although we have not yet undertaken a longitudinal analysis of H I V envelope variation and/or coreceptor phenotype evolution during antiretroviral therapy, we believe that our studies investigating H I V envelope genotypes and coreceptor phenotypes as baseline prognostic variables [47,89] contribute to the debate on this issue. Further characterization of the evolution of coreceptor usage under antiretroviral therapy, especially among larger cohorts such as H O M E R , w i l l help evaluate the clinical relevance and potential clinical utility of H I V coreceptor phenotyping in the context of modern H A A R T . b) HIV envelope genotype/coreceptor phenotype correlations As already described, the ability of the virus to use either the C C R 5 or C X C R 4 coreceptors lies at least partially in the amino acid sequence of the envelope V 3 loop; X 4 variants generally display positively charged amino acids at codons 11 and/or 25 (an "11/25" genotype), while R5 variants do not [88,90]. The "11/25" rule, however, is by no means a perfect predictor of coreceptor usage: previous studies of cloned isolates report sensitivities and specificities of - 5 0 % and -85-95%, respectively, for the ability of the "11/25" genotype to predict C X C R 4 usage [100,101,102], while studies employing direct ('bulk') P C R and sequencing approaches report even lower sensitivities and specificities (33% and 93%) [47]. Despite this, in our study the presence of "11/25" H I V was associated with a >9-fold increased risk of harboring X4-using virus [47]. Although there exist other bioinformatics-based methods to predict coreceptor usage from V 3 genotype data (for example, [101,102]), 217 these do not appear to exceed the "11/25" rule with respect to accuracy of coreceptor prediction, at least in our data set of matched genotypes and phenotypes (unpublished data). Given the observation that the "11/25" rule is a relatively poor predictor of coreceptor usage, but nevertheless a very strong marker of clinical response to H A A R T [47], many questions remain outstanding. What, indeed, is the significance of the presence of positively-charged amino acids at key sites of V 3 , and why are the presence of these amino acids independently predictive of therapy outcomes? One potential explanation may lie in the continual sequence changes occurring in the H I V envelope sequence during the evolution towards C X C R 4 usage discussed in chapter 7: over the course of H I V disease progression "there is likely a continuum of H I V genetic sequences representing evolution to use the C X C R 4 co-receptor with increasing efficiency" [89]. The 11/25 rule, in other words, may only be representative of CXCR4-us ing strains at a certain stage of evolution - specifically, where the continuum of V 3 genotype evolution temporarily passes through a stage where positively-charged amino acids predominate at key sites. Therefore, one may hypothesize that perhaps in the early-to-middle stages of C X C R 4 evolution, the "11/25 rule" may display increased capability to predict coreceptor usage, whereas in the very last stages of C X C R 4 evolution (ie at very low C D 4 counts) this rule may no longer be relevant. Two observations from our research support this hypothesis. First, there were approximately double the number of H O M E R cohort participants harboring X4-using variants at baseline [47], than subjects harboring "11/25" genotype H I V at baseline [89], suggesting that the coreceptor phenotype assay was capturing an additional subset of the population. Secondly, X4-using H I V was significantly associated with poorer baseline clinical parameters (including very low C D 4 counts in some cases) [47], whereas "11/25 218 genotype" H I V was not [89], suggesting that the phenotyping assay may have been additionally capturing the subset of the population at the most advanced stage of C X C R 4 -coreceptor usage, the stage in which the "11/25" genotype may no longer apply. Indeed, the inclusion of this " low baseline-CD4" subset among the C X C R 4 phenotype group would explain why, after multivariate analyses controlling for baseline clinical parameters, the presence of CXCR4-us ing H I V was no longer predictive of survival. The "11/25 genotype" therefore perhaps remains independently predictive of therapy outcomes because these genotypes may represent early evolution towards C X C R 4 usage, before other markers of disease prognosis (ie C D 4 and p V L ) have been affected. Further studies analysing genotype/phenotype correlations, as well as combined effects of genotypes and phenotype on therapy outcomes are currently ongoing. c) Evidence for adaptation of HIV to human genotypes: the CCR5 mutation and HIV coreceptor usage The availability of both H I V coreceptor usage and human C C R 5 A 3 2 genotypes for the same patient population (again, the H O M E R cohort) provided the opportunity to observe the potential effects of human genetic variation on the course of H I V evolution, and to investigate the consequences that this may have with respect to H I V disease. As outlined in chapter 8 [47], we were interested in whether reduced cell-surface expression of C C R 5 , (and therefore, by inference, the CCR5wt /A32 heterozygous genotype), may potentially drive evolution towards use of C X C R 4 . Indeed, our baseline cross-sectional data indicated that CCR5wt /A32 individuals were nearly twice as likely to harbor X4-using H I V than C C R 5 wt/wt individuals, suggesting that reduced cell-surface C C R 5 expression might indeed select for X4-using variants, at least in 219 some individuals [47], although we do acknowledge the potential biases in these observations [47]. There is also supporting evidence from other independent studies that report a loss of protection by the C C R 5 A32 allele in cases where X 4 variants are present [103,104,105,106]. Indeed, in our study, CXCR4-us ing H I V appeared to be associated with poor prognosis independent of C C R 5 genotype [47]. The association between H I V coreceptor phenotype and human C C R 5 A 3 2 genotype, and the consequences of this association, may help explain some of the contrasting results obtained in various studies of the independent impact of C C R 5 A 3 2 on H I V disease progression and/or treatment response, where corresponding H I V coreceptor phenotype was not evaluated. If the A32 allele confers a significant protective effect only in the presence of CCR5-using strains, then C C R 5 A 3 2 genotypes, i f ever incorporated into clinical practice, should ideally be evaluated in combination with a current H I V coreceptor phenotype. Finally, the observation that limited cell-surface C C R 5 may lead to selective pressure favoring the emergence of CXCR4-us ing variants may be relevant for the use of C C R 5 coreceptor antagonists in the treatment of HIV-infected persons. However, the full implications of this have yet to be assessed. d) Use of HIV coreceptor phenotyping in the context of future antiretroviral agents The epidemiology of H I V coreceptor usage in both treated and untreated populations is of direct relevance in the context of future antiretroviral therapies, including coreceptor antagonists, agents aimed at inhibiting H I V entry into target cells through blockade of the C C R 5 and/or C X C R 4 coreceptors. On a population basis, studies investigating the prevalence of C C R 5 and CXCR4-us ing strains among HIV-infected populations wi l l assist in 220 estimating the proportion of patients that may benefit most from treatment with C C R 5 antagonists, C X C R 4 antagonists, or both. H I V coreceptor phenotyping may soon be recommended on an individual basis, in order to determine whether a patient should receive antiretroviral therapy including a C C R 5 antagonist. In a recent study, Moyle et al suggest that, due to concerns regarding the clinical implications of administering CCR5-antagonists to patients with mixed R 5 / X 4 populations, patients may have to be screened for H I V coreceptor phenotype prior to treatment with a coreceptor antagonist [92]. Because phenotyping remains a relatively time-consuming and costly process, it would be ideal if, through epidemiological characterizations of H I V coreceptor usage such as the one undertaken as a part of this thesis work [47], reliable predictive markers of co-receptor usage could be established [92]. However, although the risk of harboring X 4 variants indeed increases with declining C D 4 count and increasing viral load [47,92], a high C D 4 count and low viral load do not necessarily exclude the presence of X 4 variants. Based on the available data, therefore, these laboratory markers cannot be relied upon to accurately predict H I V coreceptor usage in the clinical setting [92]. It is likely, therefore, that H I V coreceptor usage assays may shortly be adapted for use in the clinical setting, in the context of H I V coreceptor antagonists as antiretrovirals. 9.9 Implications and Future Directions It is slowly being appreciated that the impact of human genetic variation on H I V pathogenesis (or indeed, the pathogenesis of any disease, whether acquired or inherited) is a complex phenomenon. To date, the majority of immunogenetic and pharmacogenetic studies of the influence of human genetic variation on H I V disease progression, including the ones highlighted in this thesis work, have focused on single nucleotide polymorphisms, or at the 221 most, combinations of polymorphisms at a small number of sites within a single gene. Although simple genetic markers of H I V infection and disease progression have indeed been identified (as evidenced by the discovery of the C C R 5 A32 deletion allele), it is more likely that the course of H I V disease reflects a multitude of interacting genotypes and phenotypes reflecting variation within genes scattered throughout the whole human genome [51,107,108]. Improvements in molecular typing technologies, computer processing and bioinformatics, as well as an increased understanding of the results of the human genome project [109] wi l l result in increasing potential to generate, store and manipulate large sets of D N A sequence data, as well to assess the impact of multiple interacting combinations of genotypes on clinical outcomes. In the future, therefore, not only wi l l we most likely observe the routine use of human genetic screening assays in the clinical management of infectious disease, including H I V , but we wi l l most likely also witness the expansion of genetic testing from single nucleotide polymorphisms to complete genetic haplotypes in order to provide a more inclusive assessment of the impact of human genetics on clinical disease prognosis. Given the extensive research focused on the clinical consequences of viral genetic variation and sequence evolution in the context of both untreated and treated infection, as well as the fact that H I V genotypic and phenotypic assays are already routinely used in the clinical management of infected individuals, it is certain that novel H I V genotypic and phenotypic assays w i l l continue to be incorporated into clinical practice. Based on the work included in this thesis, as well as work by others, there is strong evidence to support the incorporation of H I V envelope V 3 genotyping (along with already-established markers CD4 and plasma viral load) for use in helping guide decisions as to when to initiate therapy and/or 222 whether a therapy interruption may be safely recommended. Current data also supports a potential role for H I V coreceptor phenotyping as a clinical management tool, in the context of emerging antiretroviral therapies. 9.10 Concluding Remarks In conclusion, the future of H I V research wi l l undoubtedly continue to focus on the role of human and viral genetic variation on H I V disease progression as well as response to current and upcoming novel antiretroviral agents. A n d as the relationship between host and viral genetics continues to be explored and understood, and genetic markers relevant to the progress of H I V disease continue to be identified, it is likely that genotyping assays wi l l adopt an ever-increasing role in the management of H I V infection. 223 9.11 REFERENCES 1. Mazzoli S, Trabattoni D, Lo Caputo S, et al. HIV-specific mucosal and cellular immunity in HIV-seronegative partners of HIV-seropositive individuals. Nat Med. 1997; 3:1250-7. 2. Bernard NF, Yannakis CM, Lee JS, Tsoukas CM. Human immunodeficiency virus (HlV)-specific cytotoxic T lymphocyte activity in HIV-exposed seronegative persons. J Infect Dis 1999;179(3):538-47. 3. Fowke KR, Nagelkerke NJ, Kimani J, et al. Resistance to HIV-1 infection among persistently seronegative prostitutes in Nairobi, Kenya. Lancet 1996;348:1347-51. 4. Liu R, Paxton WA, Choe S, et al. Homozygous defect in HIV-1 coreceptor accounts for resistance of some multiple-exposed individuals to HIV-1 infection. Cell 1996, 86:367-377. 5. Hogan CM, Hammer SM. Host determinants in HIV infection and disease. Part 2: genetic factors and implications for antiretroviral therapeutics. Ann Intern Med. 2001;134:978-96. 6. O'Brien SJ, Nelson GW. Human genes that limit AIDS. Nat Genet. 2004; 36:565-74. 7. Kaslow RA, Dorak T, Tang JJ. Influence of host genetic variation on susceptibility to HIV type 1 infection. J Infect Dis 2005; 191 Suppl LS68-77. 8. Nolan D, Gaudieri S, John M, Mallal S. Impact of host genetics on HIV disease progression and treatment: new conflicts on an ancient battleground. AIDS. 2004 Jun 18;18(9):1231-40. 9. Klein J, Sato A. The HLA system. First of two parts. N Engl J Med. 2000; 343:702-9. 224 10. Kle in J, Sato A . The H L A system. Second of two parts. N Engl J Med 2000; 343:782-6. 11. M a l i m M H , Emerman M . HIV-1 sequence variation: drift, shift, and attenuation. Cel l . 2001 Feb 23;104(4):469-72. 12. Johnson W E , Desrosiers R C . Vira l persistance: HIV's strategies of immune system evasion. Annu Rev Med. 2002;53:499-518. 13. d a S i l v a J . The evolutionary adaptation of HIV-1 to specific immunity. Curr H I V Res. 2003; 1:363-71. 14. Tersmette M , Gruters R A , de Wol f F, et al. Evidence for a role of virulent human immunodeficiency virus (HIV) variants in the pathogenesis of acquired immunodeficiency syndrome: studies on sequential H I V isolates. J V i ro l 1989; 63: 2118-25. 15. Schuitemaker H , Koot M , Kootstra N A , et al. Biological phenotype of human immunodeficiency virus type 1 clones at different stages of infection: progression of disease is associated with a shift from monocytotropic to T-cell tropic virus populations. J V i r o l 1992; 66:1354-60. 16. Koot M , Keet IP, Vos A H . Prognostic value of HIV-1 syncytium-inducing phenotype for rate of CD4+ cell depletion and progression to A I D S . A n n Int M e d 1993; 118: 681-88. 17. Mulder J, Resnick R, Saget B , et al. A rapid and simple method for extracting human immunodeficiency virus type 1 R N A from plasma: enhanced sensitivity. J C l in Microbiol . 1997; 35:1278-80. 225 18. van Gemen B , Kievits T, Schukkink R, et al. Quantification of HIV-1 R N A in plasma using N A S B A during HIV-1 primary infection. J V i ro l Methods. 1993; 43:177-87. 19. Pachl C , Todd J A , Kern D G , et al. Rapid and precise quantification of HIV-1 R N A in plasma using a branched D N A signal amplification assay. J Acquir Immune Defic Syndr Hum Retrovirol. 1995; 8:446-54. 20. Palella F J Jr, Delaney K M , Moorman A C , et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J M e d 1998; 338:853-60. 21. Hogg R S , Y i p B , K u l l y C , et al. Improved survival among HIV-infected patients after initiation of triple-drug antiretroviral regimens. Can M e d Assoc J 1999; 160:659-65. 22. Mocroft A , Ve l l a S, Benfield T L , et al. Changing patterns of mortality across Europe in patients infected with H I V - 1 . EuroSIDA Study Group. Lancet. 1998; 352:1725-30. 23. Detels R, Munoz A , McFarlane G , et al. Effectiveness of potent antiretroviral therapy on time to A I D S and death in men with known H I V infection duration. Multicenter A I D S Cohort Study Investigators. J A M A . 1998; 280:1497-503. 24. Perelson A S , Essunger P, Cao Y , et al. Decay characteristics of HIV-1-infected compartments during combination therapy. Nature. 1997; 387:188-91. 25. Ledergerber B , Egger M , Opravil M , et al. Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Swiss H I V Cohort Study. Lancet 1999; 353:863-8. 26. Deeks S G , Hecht F M , Swanson M , et al. H I V R N A and C D 4 cell count response to protease inhibitor therapy in an urban A I D S clinic: response to both initial and salvage therapy. A I D S 1999; 13:F35-43. 226 27. Piketty C, Weiss L , Thomas F, Mohamed A S , Belec L , Kazatchkine M D . Long-term clinical outcome of human immunodeficiency virus-infected patients with discordant immunologic and virologic responses to a protease inhibitor-containing regimen. J Infect Dis 2001; 183:1328-35. 28. Price D A , Scullard G , Oxenius A , et al. Discordant outcomes following failure of antiretroviral therapy are associated with substantial differences in human immunodeficiency virus-specific cellular immunity. J V i r o l 2003;77:6041-9. 29. Kaplan SS, Ferrari G , W r i n T, et al. Longitudinal assessment of immune response and viral characteristics in HIV-infected patients with prolonged CD4(+)/viral load discordance. A I D S Res Hum Retroviruses. 2005; 21:13-6. 30. Grabar S, Le Moing V , Goujard C, et al. Cl inical outcome of patients with HIV-1 infection according to immunologic and virologic response after 6 months of highly active antiretroviral therapy. A n n Intern Med. 2000; 133:401-10. 31. Kravcik S. H I V lipodystrophy: a review. H I V C l i n Trials. 2000; 1:37-50. 32. Montaner JS, Cote H C , Harris M , et al. Mitochondrial toxicity in the era of H A A R T : evaluating venous lactate and peripheral blood mitochondrial D N A in HIV-infected patients taking antiretroviral therapy. J Acquir Immune Defic Syndr. 2003; 34 Suppl LS85-90. 33. Bozkurt B . Cardiovascular toxicity with highly active antiretroviral therapy: review of clinical studies. Cardiovasc Toxicol . 2004; 4:243-60. 34. Peyriere H , Reynes J, Rouanet I, et al. Renal tubular dysfunction associated with tenofovir therapy: report of 7 cases. J Acquir Immune Defic Syndr 2004; 35:269-73. 227 35. Kontorinis N , Dieterich D . Hepatotoxicity of antiretroviral therapy. A I D S Rev. 2003; 5:36-43. 36. Hirsch M S , Brun-Vezinet F, Clotet B , et al. Antiretroviral drug resistance testing in adults infected with human immunodeficiency virus type 1: 2003 recommendations of an International A I D S Society-USA Panel. C l i n Infect Dis 2003; 37:113-28. 37. Volberding P A . H I V therapy in 2003: consensus and controversy. A I D S . 2003; 17 Suppl 1:S4-11. 38. Samson M , Libert F, Doranz, B J , et al. Resistance to HIV-1 infection in Caucasian individuals bearing mutant alleles of the C C R 5 chemokine receptor gene. Nature 1996, 382:722-725. 39. Huang Y , Paxton W A , Wolinsky S M , et al. The role of a mutant C C R 5 allele in HIV-1 transmission and disease progression. Nat M e d 1996, 11:1240-1243. 40. Dean M , Carrington M , Winkler C , et al. Genetic restriction of HIV-1 infection and progression to A I D S by a deletion allele of the C K R 5 structural gene. Science 1996, 273:1856-1862. 41. Guerin S, Meyer L , Theodorou I, Boufassa F, Magierowska M , Goujard C, et al. C C R 5 delta32 deletion and response to highly active antiretroviral therapy in HIV-1 -infected patients. A I D S 2000; 14:2788-90. 42. Valdez H , Purvis SF, Lederman M M , Fillingame M , Zimmerman P A . Association of the CCR5delta32 mutation with improved response to antiretroviral therapy. J A M A 1999; 282:734. 228 43. Bogner JR, Lutz B , K le in H G , Pollerer C , Troendle U , Goebel F D . Association of highly active antiretroviral therapy failure with chemokine receptor 5 wi ld type. H I V Med. 2004; 5:264-72. 44. Wi t F W , van Rij R P , Weverling G J , et al. C C chemokine receptor 5 delta32 and C C chemokine receptor 2 641 polymorphisms do not influence the virologic and immunologic response to antiretroviral combination therapy in human immunodeficiency virus type 1-infected patients. J Infect Dis 2002; 186:1726-32. 45. Brumme Z L , Chan K J , Dong W , et al. CCR5Delta32 and promoter polymorphisms are not correlated with initial virological or immunological treatment response. AIDS 2001; 15:2259-66. 46. Brumme Z L , Henrick B M , Brumme C J , Hogg R S , Montaner JSG,Harrigan PR. Association of C C R 5 A 3 2 mutation with clinical response and >5-year survival following initiation of first triple antiretroviral regimen. Antiviral Therapy, in press. 47. Brumme Z L , Goodrich J, Mayer H B , et al. Molecular and Clinical Epidemiology of CXCR4-us ing HIV-1 in a large population of therapy-naive individuals. J Infect Dis 2005; 192:466-74. 48. Bratt G , Sandstrom E , Albert J, Samson M , Wahren B . The influence of M T - 2 tropism on the prognostic implications of the delta32 deletion in the C,CR-5 gene. A I D S 1997; 11:1415-9. 49. de Roda Husman A M , Koot M , Cornelissen M , et al. Association between C C R 5 genotype and the clinical course of HIV-1 infection. A n n Intern Med 1997; 127:882-90. 229 50. Michael N L , Chang G , Louie L G , et al. The role of viral phenotype and C C R - 5 gene defects in HIV-1 transmission and disease progression. Nat M e d 1997; 3: 338-40. 51. Tang J, Wi lson C M , Meleth S, et al. Host genetic profiles predict virological and immunological control of HIV-1 infection in adolescents. A I D S . 2002; 16(:2275-84. 52. Gonzalez E , Kulkarni H , Bolivar H , et al. The influence of C C L 3 L 1 gene-containing segmental duplications on HIV-1 /AIDS susceptibility. Science. 2005; 307:1434-40. 53. Faure S, Meyer L , Costagliola D , et al. Rapid progression to A I D S in HIV+ individuals with a structural variant of the chemokine receptor C X 3 C R 1 . Science 2000; 287:2274-7. 54. Brumme Z L , Dong W W , Chan K J , et al. Influence of polymorphisms within the C X 3 C R 1 and M D R - 1 genes on initial antiretroviral therapy response. A I D S 2003; 17:201-8. 55. Faure S, Meyer L , Genin E , et al. Deleterious genetic influence of C X 3 C R 1 genotypes on HIV-1 disease progression. J Acquir Immune Defic Syndr 2003; 32:335-7. 56. K w a D , Boeser-Nunnink B , Schuitemaker H . Lack of evidence for an association between a polymorphism in C X 3 C R 1 and the clinical course of H I V infection or virus phenotype evolution. A I D S 2003; 17:759-61. 57. Combadiere C , Salzwedel K , Smith E D , Tiffany H L , Berger E A , Murphy P M . Identification of C X 3 C R 1 . A chemotactic receptor for the human C X 3 C chemokine fractalkine and a fusion coreceptor for H I V - 1 . J B i o l Chem 1998;273:23799-804. 230 58. Foussat A , Bouchet-Delbos L , Berrebi D , et al. Deregulation of the expression of the fractalkine/fractalkine receptor complex in HIV-1-infected patients. Blood. 2001; 98:1678-86. 59. Combadiere B , Faure S, Autran B , Debre P, Combadiere C . The chemokine receptor C X 3 C R 1 controls homing and anti-viral potencies of C D 8 effector-memory T lymphocytes in HIV-infected patients. A I D S 2003; 17:1279-90. 60. Singh K K , Hughes M D , Chen J, Spector S A . Genetic Polymorphisms in C X 3 C R 1 Predict HIV-1 Disease Progression in Children Independently of CD4+ Lymphocyte Count and HIV-1 R N A Load. J Infect Dis 2005; 191:1971-80. 61. Hoffmeyer S, Burk O, von Richter O, et al. Functional polymorphisms of the human multidrug-resistance gene: Multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo. P N A S 2000; 97: 3473-3478. 62. K i m R B . Drug transporters in H I V Therapy. Top H I V M e d 2003; 11:136-9. 63. Hi tz l M , Drescher S, van der Kuip H , et al. The C3435T mutation in the human M D R 1 gene is associated with altered efflux of the P-glycoprotein substrate rhodamine 123 from CD56+ natural killer cells. Pharmacogenetics 2001; 11:293-8. 64. Fellay J, Marzol ini C , Meaden E R , et al. Response to antiretroviral treatment in HIV-1-infected individuals with allelic variants of the multidrug resistance transporter 1: a pharmacogenetics study. Lancet. 2002; 359:30-6. 65. Zhu D , Taguchi-Nakamura H , Goto M , et al. Influence of single-nucleotide polymorphisms in the multidrug resistance-1 gene on the cellular export of nelfinavir and its clinical implication for highly active antiretroviral therapy. Antivir Ther. 2004; 9:929-35. 231 66. Oselin K , Gerloff T, Mrozikiewicz P M , Pahkla R, Roots I. M D R 1 polymorphisms G2677T in exon 21 and C3435T in exon 26 fail to affect rhodamine 123 efflux in peripheral blood lymphocytes. Fundam C l i n Pharmacol 2003; 17:463-9. 67. K i m R B , Leake B F , Choo E F , et al. Identification of functionally variant M D R 1 alleles among European Americans and African Americans. C l i n Pharmacol Ther 2001;70:189-99. 68. Haas D W , W u H , L i H , et al. M D R 1 gene polymorphisms and phase 1 viral decay during HIV-1 infection: an adult A I D S Clinical Trials Group study. J Acquir Immune Defic Syndr 2003; 34:295-8. 69. Nasi M , Borghi V , Pinti M , et al. M D R 1 C3435T genetic polymorphism does not influence the response to antiretroviral therapy in drug-naive HIV-positive patients. A I D S 2003; 17:1696-8. 70. Winzer R, Langmann P, Z i l l y M , et al. No influence of the P-glycoprotein polymorphisms M D R 1 G2677T/A and C3435T on the virological and immunological response in treatment naive HIV-positive patients. A n n C l i n Microbiol Antimicrob. 2005; 4:3. 71. Saitoh A , Singh K K , Powell C A , et al. A n MDR1-3435 variant is associated with higher plasma nelfinavir levels and more rapid virologic response in HIV-1 infected children. A I D S 2005; 19:371-80. 72. Marzol ini C , Paus E , Bucl in T, K i m R B . Polymorphisms in human M D R 1 (P-glycoprotein): recent advances and clinical relevance. C l i n Pharmacol Ther 2004; 75:13-33. 73. Kl ing , J. Ultrafast D N A sequencing. Nature Biotechnology 2003; 21:1425 - 1427 232 74. Ratner L , Haseltine W , Patarca R, et al. Complete nucleotide sequence of the A I D S virus, H T L V - I I I . Nature. 1985; 313:277-84. 75. Pil lay D , Taylor S and Richman D D . Incidence and impact of resistance against approved antiretroviral drugs. Rev Med Vi ro l . 2000; 10:231-53. 76. Yen i P G , Hammer S M , Hirsch M S , et al. Treatment for adult H I V infection: 2004 recommendations of the International A I D S Society-USA Panel. J A M A . 2004;292:251-65. 77. Richman D D , Morton S C , Wrin T, et al. The prevalence of antiretroviral drug resistance in the United States. A I D S 2004; 18:1393-401. 78. Recsky M A , Brumme Z L , Chan K J , et al. Antiretroviral resistance among H I V -infected persons who have died in British Columbia, in the era of modern antiretroviral therapy. J Infect Dis. 2004; 190:285-92. 79. Baldwin C E , Sanders R W , Deng Y , et al. Emergence of a drug-dependent human immunodeficiency virus type 1 variant during therapy with the T20 fusion inhibitor. J V i ro l . 2004;78:12428-37. 80. Menzo S, Castagna A , Monachetti A , E T A L . Resistance and replicative capacity of HIV-1 strains selected in vivo by long-term enfuvirtide treatment. New Microbiol . 2004; 27(2 Suppl 1):51-61. 81. Peters S, Munoz M , Yer ly S, et al. Resistance to nucleoside analog reverse transcriptase inhibitors mediated by human immunodeficiency virus type 1 p6 protein. J V i r o l 2001; 75:9644-53. 233 82. Brumme Z L , Chan K J , Dong W W , et al. Prevalence and clinical implications of insertions in the HIV-1 p6Gag N-terfninal region in drug-naive individuals initiating antiretroviral therapy. Antivir Ther 2003; 8:91-6. 83. Gallego O, de Mendoza C, Corral A , Soriano V . Changes in the human immunodeficiency virus p7-pl-p6 gag gene in drug-naive and pretreated patients. J C l i n Microbiol 2003; 41:1245-7. 84. Ibe S, Shibata N , Utsumi M , et al. Selection of human immunodeficiency virus type 1 variants with an insertion mutation in the p6(gag) and p6(pol) genes under highly active antiretroviral therapy. Microbiol Immunol 2003; 47:71-9. 85. Kaufmann G R , Suzuki K , Cunningham P, et al. Impact of H I V type 1 protease, reverse transcriptase, cleavage site, and p6 mutations on the virological response to quadruple therapy with saquinavir, ritonavir, and two nucleoside analogs. A I D S Res Hum Retrov 2001; 17:487-497. 86. Lastere S, Dalban C , Co l l in G , et al. Impact of insertions in the HIV-1 p6 P T A P P region on the virological response to amprenavir. Antivir Ther 2004; 9:221-7. 87. De Jong JJ, De Ronde A , Keulen W , Tersmette M , Goudsmit J. Minimal requirements for the human immunodeficiency virus type 1 V 3 domain to support the syncytium-inducing phenotype: analysis by single amino acid substitution. J Virol 1992; 66:6777-80. 88. Fouchier R A , Groenink M , Kootstra N A , et al. Phenotype-associated sequence variation in the third variable domain of the human immunodeficiency virus type 1 gpl20 molecule. J Virol 1992; 66:3183-7. 234 89. Brumme Z L , Dong W W , Y i p B , et al. Clinical and immunological impact of H I V envelope V3 sequence variation after starting initial triple antiretroviral therapy. A I D S 2004; 18:Fl-9. 90. Fouchier R A , Brouwer M , Broersen S M , Schuitemaker H . Simple determination of human immunodeficiency virus type 1 syncytium-inducing V3 genotype by P C R . J C l i n Microbiol . 1995; 33:906-11. 91. Toulson A R , Harrigan PR, Heath K , et al. Treatment Interruption of antiretroviral therapy among patients with nadir C D 4 cell counts > 200 cells/mm3. J Infect Dis, in press 92. Moyle G J , Wildfire A , Mandalia S, et al. Epidemiology and predictive factors for chemokine receptor use in HIV-1 infection. J Infect Dis 2005; 191:866-72. 93. Furrer H , Wendland T, Minder C, et al. Association of syncytium-inducing phenotype of HIV-1 with C D 4 cell count, viral load and sociodemographic characteristics. A I D S 1998; 12:1341-6. 94. Philpott S, Weiser B , Anastos K , et al. Preferential suppression of CXCR4-specif ic strains of HIV-1 by antiviral therapy. J C l i n Invest. 2001; 107:431-8. 95. Galan I, Jimenez J L , Gonzalez-Rivera M , et al. Virological phenotype switches under salvage therapy with lopinavir-ritonavir in heavily pretreated HIV-1 vertically infected children. A I D S 2004; 18:247-55. 96. Zheng N N , McQueen P W , Hurren L , et al. Changes in biologic phenotype of human immunodeficiency virus during treatment of patients with didanosine. J Infect Dis 1996; 173:1092-6. 235 97. Skrabal K , Trouplin V , Labrosse B , et al. Impact of antiretroviral treatment on the tropism of HIV-1 plasma virus populations. A I D S . 2003 A p r 11;17(6):809-14. 98. van Rij R P , Visser J A , van Praag R M , et al. Both R5 and X 4 human immunodeficiency virus type 1 variants persist during prolonged therapy with five antiretroviral drugs. J V i r o l 2002; 76:3054-8. 99. Delobel P, Sandres-Saune K , Cazabat M , et al. R5 to X 4 switch of the predominant HIV-1 population in cellular reservoirs during effective highly active antiretroviral therapy. J Acquir Immune Defic Syndr 2005; 38:382-92. 100. McDonald R A , Chang G , Michael N L . Relationship between V 3 genotype, biologic phenotype, tropism, and coreceptor use for primary isolates of human immunodeficiency virus type 1. J Hum Vi ro l 2001; 4:179-87. 101. Jensen M A , L i FS , van't Wout A B , et al. Improved coreceptor usage prediction and genotypic monitoring of R5-to-X4 transition by motif analysis of human immunodeficiency virus type 1 env V 3 loop sequences. J V i r o l 2003; 77:13376-88. 102. Resch W , Hoffman N , Swanstrom R. Improved success of phenotype prediction of the human immunodeficiency virus type 1 from envelope variable loop 3 sequence using neural networks. Virology 2001; 288:51-62. 103. Bratt G , Leandersson A C , Albert J, Sandstrom E , Wahren B . M T - 2 tropism and C C R - 5 genotype strongly influence disease progression in HIV-1-infected individuals. A I D S . 1998; 12:729-36. 104. Bratt G , Sandstrom E , Albert J, Samson M , Wahren B . The influence of M T - 2 tropism on the prognostic implications of the delta32 deletion in the C C R - 5 gene. A I D S . 1997; 11:1415-9. 236 105. de Roda Husman A M , Koot M , Cornelissen M , et al. Association between C C R 5 genotype and the clinical course of HIV-1 infection. A n n Intern M e d 1997; 127:882-90. 106. Michael N L , Chang G , Louie L G , et al. The role of viral phenotype and C C R - 5 gene defects in HIV-1 transmission and disease progression. Nat M e d 1997; 3: 338-40. 107. Saah A J , Hoover D R , Weng S, et al. Association of H L A profiles with early plasma viral load, CD4+ cell count and rate of progression to A I D S following acute HIV-1 infection. Multicenter A I D S Cohort Study. A I D S . 1998; 12:2107-13. 108. Kaslow R A , Carrington M , Apple R, et al. Influence of combinations of human major histocompatibility complex genes on the course of HIV-1 infection. Nat M e d 1996; 2:405-11 109. Venter JC, Adams M D , Myers E W , et al. The sequence of the human genome. Science 2001;291:1304-51. 237 APPENDIX Certificates of Ethical Approval: Original, Renewals and Updates 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0092862/manifest

Comment

Related Items