Themes

Viral load Tropism Drug resistance

HIV Virology & Resistance

Anna Maria Geretti Institute of Infection & Global Health The HIV Virology Timeline

Is HIV HIV replicates eradication HAART at high levels possible?

throughout Viral load testing HIV-1 isolated the infection Virus Virus replication HIV-1 genome replication drives disease sequenced causes pathogenesis immunological through Immune compromise activation & inflammation

1982 1985 1991 1995 1996 2009 2010  Fusion inhibitors CCR5 antagonists

Attachment Fusion Release of RNA

Assembly Reverse transcription Integration Transcription Nucleos(t)ide and Integrase Non-nucleoside RT inhibitors Maturation & budding inhibitors Protease inhibitors Key virological characteristics of HIV infection

 High virus replication rate 109-1010 virus particles produced each day  Rapid virus clearance

T½ virus producing cells: <1 day T½ plasma free: a few hours  Genetic evolution All possible point mutations in the viral genome can be generated daily  Virus latency – integration into host DNA ~ 1:106 resting CD4 T cells

Wong et al. PNAS 1997; Wong et al. Science 1997; Chun et al. Nature 1997; Chun et al. PNAS 1997; Siliciano et al. Nat Med 2003; Strain et al. PNAS 2003 Obstacles to HIV eradication with ART

Sanctuary sites

Cell-to-cell virus spread Integration and latency

Sigal et al. Nature 2011 HIV-1 DNA detection during suppressive ART

 HIV-1 DNA quantified in PBMC from 104 patients receiving suppressive ART for 1 to 15 years

HIV DNA CD4 10^6 per DNA HIV 1 3 5 7 9 11 13 15 Years of ART

PBMC = Peripheral blood mononuclear cells

Geretti et al. International Workshop on HIV & Hepatitis Viruses Drug Resistance 2013 HIV viral load predicts the rate of CD4 cell loss and disease progression

40 80 Progression 20 70 Death 60 0 50 -20 40

-40 30

per year per 20 -60

Risk (%) over 6 yrs 6 over (%) Risk 10 CD4 cell change change cell CD4 -80 0 <500 501- 3001- 10001- >30000 <500 501- 3001- 10001- >30000 3000 10000 30000 3000 10000 30000 Viral load Viral load

40

20

0

-20

-40

-60 Decrease in risk of of risk in Decrease

progression and death (%) death and progression -80 0.3 0.6 1 Viral load reduction with ART (log ) 10 Mellors et al. Science 1996 SMART Study: Stopping ART associated

with increased morbidity and mortality

15 Arm that stopped therapy at CD4 >350 and restarted at CD4 <250 (n= 2495)

10 p <0.0001

5 Arm that continued therapy (n=2512) % cumulative clinical events clinical % cumulative 0

0 4 8 12 16 20 24 28 32 36 40 44 Months after discontinuation

No. DC 2495 1839 1441 1112 915 733 569 449 375 297 178 34 VS 2512 1848 1434 1122 931 754 601 490 398 310 184 41

El-Sadr et al. NEJM 2006 Starting first-line ART DHSS guidelines 2013

Third agent NRTIs Evidence

EFV TDF/FTC AI EFV = consider avoiding if woman ATV/r TDF/FTC AI of child-bearing age DRV/r TDF/FTC AI TDF = caution if renal insufficiency RAL TDF/FTC AI ATV = do not use or caution with Recommended acid-lowering agents EFV ABC/3TC BI RPV = not if VL >100,000 cps; RPV TDF/FTC BI caution if CD4 <200; not with PPIs

RPV ABC/3TC BIII ABC = caution if VL >100,000 cps or ATV/r ABC/3TC BI high CVD risk; use only if HLA- DRV/r ABC/3TC BII B57.01 negative FPV/r or LPV/r ABC/3TC or BI Alternative EVG/c = only if eGFR >70 ml/min; (QD or BID) TDF/FTC potential DDIs with COBI; not with RAL ABC/3TC BIII other nephrotoxic drugs EVG/c TDF/FTC BI DHSS Guidelines Feb 2013 Mortality according to frequency of viral load measurements >400 cps in first-line ART

Cumulative mortality stratified by % of VL measurements ≥400 cps over first 18 months of ART

0.25 N=2046 100% Started ART before 2002 0.2 Follow-up: 8898 patient-yrs 51-75% 76-99% 0.15 26-50% 1-25% 0.1

0% Cumulative mortality Cumulative 0.05

0 0 18 36 54 72 Months after starting ART Lohse et al. Clin Infect Dis 2006 The goals of ART

. Restore and preserve immune function . Reduce HIV-related morbidity and mortality . Improve quality of life . Provide maximal and durable VL suppression

 EACS 2012: <50 cps  BHIVA 2012: <50 cps  IAS-USA 2012 : <50 cps  DHHS 2013:

Defining cut-offs: Viral load assays

The recommended target for defining ART success dictated by the technical properties of the viral load assay, rather than selected a priori based upon clinical significance . First-generation assays <400 cps . Second-generation assays <50 cps (e.g. Roche Amplicor v1.5)

However Patients who achieve and maintain a viral load <50 cps have a small risk of rebound >50 cps during follow-up, and the risk declines further the longer the viral load is <50 cps Assay discorcordance at the 50 cps cut-off

>50 cps 2 Amplicor 1.5 RealTime TaqMan v2 ArtusHIV Amplicor 1.5 - NA 6% - <50 RealTime NA - 13% 5% cps TaqMan v2 5% 7% - -

ArtusHIV - 5% - -

1. The International Viral Load Assay Collaboration (unpublished); 2. Garcia et al. JCV 2013 Defining virological failure

EACS 2012: Confirmed VL >50 cps 6 months after initiation or modification of ART

BHIVA 2012: Failure to achieve VL <50 cps 6 months after commencing ART or following suppression <50 cps, confirmed VL rebound >400 cps

DHHS 2013: Inability to achieve or maintain VL <200 cps

IAS-USA 2012: Sustained VL elevation between 50 cps and 200 cps should prompt evaluation of factors leading to failure and consideration of changing ART

VL = Viral load Third-generation viral load assays

 Lower limit of quantification (LLQ) 40 cps (RealTime; Taqman v1) or 20 cps (Taqman v2)  Report qualitative RNA detection below the LLQ as “target detected”  Patients on ART can show one of three results: . VL quantified above the LLQ . RNA detected below the LLQ (RNA+) . RNA not detected (RNA-) Plasma HIV-1 RNA detection below 50 cps predicts viral load rebound

Time to rebound 1 40 0.9 Confirmed (or last available) VL >50 cps 34.2% 0.8 35 0.7 log rank test p<0.0001 0.6 30 0.5 Rebound >50 cps/ml 0.4 25 0.3 Rebound >400 cps/ml 0.2 20

0.1 Proportionwith rebound VL 0 15 13% 0 3 6 9 12 15 18 11.3% Time since T0 (months)

1 10

0.9 4%

Proportion with VL rebound rebound VL with Proportion 3.8% 0.8 Confirmed (or last available) VL >400 cps 5

0.7 1.2%

0.6 0 VL 40-49 cps 0.5 T0 T0 VL 40-49 T0 RNA+ T0 RNA- + 0.4 TOVL RNA (~10-49 cps) cps/ml 0.3 - TOVL RNA 0.2

Proportionwith rebound VL 0.1

0 0 3 6 9 12 15 18 Time since T0 (months) Doyle et al. Clin Infect Dis 2012 Take-away points

. Viral load informs prognosis, guides ART initiation and is the key surrogate marker of ART efficacy . The optimal level of viral load suppression with ART is <50 cps and probably <10 cps . There remain uncertainties about the optimal management of low-level viraemia leading to discrepancies within guidelines . Viral load assay performance differs . Importance of clinic-lab dialogue . Importance of using one assay for monitoring HIV tropism

 Defined by the differential use of co-receptors and by the cellular distribution of co-receptors

X4 D R5

CD4 CXCR4 CCR5

Naive CD4+ cells Memory CD4+ cells Must be activated to memory phenotype to become target of R5 Esté et al. Lancet 2007 Genotypic Tropism Testing

1. Gene sequencing 2. Sequence determination 3. Results interpretation Envelope via algorithms gene from env patients virus

gp120

Cell membrane V3 Sequence CTRPNNNT-RKSIPMGPG--QAIYATGAIIGDIRQAHC R5 CCR5 ...... R..-.HI..RH..VM...E-...N...... X4 Mechanisms of viral escape during therapy with CCR5 antagonists

Resistant virus Quasispecies R5 virus CXCR4-using virus bound to CCR5

Assay’s X4 detection limit

Van der Ryst et al. ICAAC 2007 Mechanisms of HIV genetic evolution

1. Errors by viral reverse transcriptase ~1 mis-incorporation per genome round 2. Errors by cellular RNA polymerase II 3. APOBEC-driven GA hypermutation Deamination of cytosine residues in nascent viral DNA 4. Recombination between HIV strains ~7-30 events per genome round Consequences of HIV genetic variability

. Challenge for diagnostic assays . Escape from immune and drug pressure . Variations in drug susceptibility and resistance pathways . Viral fitness, tropism and disease pathogenesis

Dominant quasispecies  rapid turnover Escape  rapid adaptation Fitness Take-away points

. Drug-resistant mutants emerge ”spontaneously “ during HIV replication . Due to impaired fitness, these “spontaneous” drug-resistant mutants exists only at very low level (in the absence of drug pressure) . Single mutants > double mutants >> triple mutants PCR

Viral gene (e.g., RT) HIV RNA Plasma

How we detect resistance in routine practice

Sequencing Mutations

RT M184V Methionine Valine @ codon 184 of RT

ATG / AUG GTG / GUG Recommendations for drug resistance testing1

Time Comment Method Evidence Diagnosis Recommended Genotypic Ia Starting ART Recommended if not already carried out* Genotypic Ia After Consider if <1log VL drop after 4 wks Genotypic IV starting ART Consider if VL >50 cps/ml after 12-16 wks Genotypic III Recommended if VL >50 cps/ml at 24 wks Genotypic Ia ART failure Recommended** Genotypic Ia

*Repeat testing to detect superinfection not routinely recommended but may be considered in selected cases (Iib)2 **Consider phenotypic or virtual phenotypic testing if interpretation is uncertain

VL = Viral load

1. BHIVA guidelines for the routine investigation & monitoring of adult HIV-1 infected individuals 2011; 2. Doyle, AIDS 2011 Transmitted drug resistance

Stable after transmission Drug Gradual reversion over time pressure Persistence at low frequency in plasma Persistence in latently infected cells

Transmission 15-30% Mutation Frequency 0.001 0.01 100 0.1 10

1

Detected Detected by SS methods ultrasensitive Detected by background Natural

(SS) Sangersequencing Limit of

detection of

Impact of transmitted drug resistance Resistance in ART-naïve patients (FIRST Study, n=258) Resistance test RAMs SS UDS NNRTI 7% 15% NRTI 6% 14% PI 2% 5% Any 14% 28%

Risk of failure of first-line NNRTI-based ART  SS RAMs: HR 12.4 [3.4-45.1]  UDS RAMs: HR 2.5 [1.2-5.4]

SS = Sanger sequencing; UDS = Ultra-deep sequencing; RAMs = Resistance-associated mutations

Simen et al. JID 2009 Take away points

. Transmitted drug resistant mutants persist as dominant species at variable rates in the absence of drug pressure . Resistant mutants that have apparently disappeared persist at low frequency in plasma and are “archived” in latently infected cells . Transmitted NNRTI and NRTI mutants reduce responses to NNRTI-based ART with a dose-dependent effect Emergence and evolution of resistance

Emergence Evolution

Single mutant Double mutant Triple mutant

. Increasing number of mutations . Accumulation of mutations on the same viral genome . Initially reduced viral fitness . Compensatory changes restore fitness over time Genetic barrier & Cross-resistance

Class ARVs Genetic barrier X-resistance ZDV/3TC, d4T/3TC +/++ +++ NRTIs ABC/3TC, TDF/3TC + +++ TDF/FTC +/++ +++ EFV, NVP, RPV + +++ NNRTIs ETV +/++ +++ PIs Unboosted +/++ ++/+++ Boosted +++/++++ +/++ Fusion inhibitors T20 + - CCR5 antagonists MVC +/++ - RAL, EVG + +++ Integrase inhibitors DTG ++/+++ ++ Resistance after failure of first-line NNRTI-based ART  Cohort from 6 sub-Saharan African countries (Kenya, Nigeria, South Africa, Uganda, Zambia, and Zimbabwe)  142 patients with viral load >1000 cps after 12 months of ART assessed for resistance

70 57,8 60,6 60 53,5 48,6 50 40 30 20 12 8,5 10 6,3 2,1 % with with %resistance 0

Hamers et al, Lancet Infect Dis 2011

EFV, RAL, EVG/c, or ATV/r in first-line ART

100 100 mL 86 81 89% 75 76 90 86% ps/ 80 80 82 79 86% 83% 70 60 69 67 (P=0.19) (P=0.27)

60

1 RNA <50<50 (%) (%) c/mL c/mL RNA RNA 11 - RaltegravirTDF/FTC/RAL group (n=281) (n=281) - 40 50

RNA RNA c <50 EfavirenzTDF/FTC/EFV group (n=282) (n=282)

1 1 40 - 20 30 0 20 TDF/FTC/EVG/cSTB (n=348) (n=348) % % HIV • 0 16 48 72 96 120 156 192

10 • TDF/FTC/EFVATR (n=352) (n=352) Subjectswith HIV Week Subjectswith HIV 0 BL 24 48 72 96 Week RAMs STARTMRK 236-102 236-103 wk 192 wk 96 wk 96 100 92% 87% RAL EFV EVG EFV EVG ATV 90 INSTI 4/16 - 9/17 - 5/17 - 80 88% 85% 70 (P=0.15) (P=0.60)

EFV - 7/14 - 10/23 - 0/15 60

1 RNA <50<50 (%) (%) c/mL c/mL RNA RNA 11

- - ATV - - - - - 0/15 50 40 NRTI 6/18 6/14 10/17 3/23 5/17 0/15 30 M184V/I 6/18 5/14 10/17 3/23 5/17 0/15 20 TDF/FTC/EVG/c (n=353) • STB (n=353)

10 • TDF/FTC/ATV/r (n=355) Subjects withSubjects HIV K65R 0/18 1/14 4/17 3/23 1/17 0/15 withSubjects HIV ATV/r + TVD (n=355) 0 0 24 48 72 96 Week De Jesus et al, HCT 2012; Zolopa et al, ICDTHI 2012; Rockstroh et al, ICDTHI 2012

EVG/r with a weak or a strong backbone

Study GS-US-183-105

Zolopa CROI 2007 Etravirine with a strong Etravirine with a weak backbone: DUET studies backbone: TMC125-C227

Patients with VL <50 cps/ml at wk 48 100 (ITT-TLOVR)

90 ETV + OBR (n=599)

80 Placebo + OBR (n=604) 95% CI 95% 70 61% ± 60 50 40% 40 30

Responders (%) (%) Responders p<0.0001* 20 10 0 0 2 4 8 12 16 20 24 32 40 48 Time (weeks) ART-experienced patients ART-experienced patients with NNRTI resistance with NNRTI and PI resistance

OBR = Optimized Background Regimen Katlama, AIDS 2009; Ruxrungtham, HIV Med 2008 Take away points

. The genetic barrier to resistance is defined by the number of mutations required to reduce susceptibility, fitness cost of mutations, interactions between mutations, drug levels, activity of other drugs in the regimen . Patients receiving NRTIs, NNRTIs or INSTIs usually show resistance at the time of virological failure . PI/r augment the genetic barrier of the entire regimen . PI/r protect the residual activity of drugs with partial resistance and low genetic barrier Drug pressure

Resistant strain

Wild-type strain

Limit of detection

30%

- 20 Resistant strain

Wild-type strain

Limit of detection

30%

- 20 SWITCHMRK: Replacing LPV/r with RAL

 Patients on ≥2NRTIs + LPV/r with VL <50 cps for ≥3 months randomised to continue LPV/r or switch to RAL

RAL LPV/r

(n=350) (n=352) 100 92 89 90

4 2

Mean CD4 count 436 454 77

k 80

w

LPV/r >1 year 83% 82% t

a

Prior ART duration, yrs 3.4 4.1 l 60 RAL

m

/ s

Previous ARVs, n 5 5 p 40 LPV/r

c

a 0

LPV/r as first regimen 37% 37% 5 <

20

Previous VF,b 32% 35% % 0 % <50 cps at wk 24 81-88% 87-94% Prior VF No prior VF aData obtained retrospectively bInvestigator-reported history  Non-inferiority of RAL not demonstrated

VF = Virological Failure  Less diarrhoea and better lipids on RAL

Eron et al, Lancet 2010 Take away points

. Once drug pressure is removed, resistant mutants are outgrown by fitter wild-type virus becoming undetectable by routine tests . Resistant mutants that have apparently disappeared persist at low frequency in plasma and are “archived” in latently infected cells . The memory of resistance is long-lived . Archived resistance can compromise a regimen with a low genetic barrier . When changing ART, consider the overall ART history and take into account past resistance( known or likely) Partial treatment interruption in patients with resistance reveals residual activity

1.0 Change in viral load 0.5 2 weeks after interruption 0.0 (log10 cps/ml) –0.5 NRTI PI NNRTI T20 Discontinued treatment class

Deeks, CROI 2005 Mechanisms of resistance: Primer unblocking in NRTI resistance

. T215Y-mediated resistance . Hydrolytic removal of the chain- terminating NRTI enables DNA synthesis to resume P P P

P P

P P . M184V antagonises the process delaying P the emergence of T215Y and increasing

susceptibility to ZDV, d4T and TDF P P P

P

P P P P

Gotte, J Viol 2000 Take away points

. Resistant mutants often display reduced fitness with a beneficial effect on viral load and CD4 counts . Compensatory changes emerge over time that partially restore virus fitness . Antagonistic effects between mutations can also have beneficial effects . Best evidence of residual activity despite resistance for the NRTIs

Clinical implications

. The likelihood of drug-resistance depends upon the individual drug, the overall regimen and the drug levels . Avoid functional monotherapy with drugs that have a low genetic barrier to resistance . In the presence of partial resistance ensure optimal activity of the overall regimen to prevent further resistance . Be mindful of pre-existing resistance when switching patients with suppressed viraemia . Preventing the accumulation of resistance remains a key goal of successfully managed ART

Thank you