For more information, visit hptn.org and follow us: Facebook: HIVptn | Twitter: @HIVptn | Youtube: HIVptn

U.S. NATIONAL INSTITUTES OF HEALTH:National Institute of Allergy and Infectious Diseases

National Institute of Mental HealthNational Institute of Drug Abuse

Female−to−maletransmissions

Male−to−femaletransmissions

Low Medium High Low Medium High

Low

Med

ium

Hig

hR

isk

grou

p of

reci

pien

t

0.00

0.05

0.10

0.15

0.20

0.25

Prop. of total transmissionsover trial period

Presented at the2020 Conference on Retroviruses and Opportunistic Infections (CROI)

Boston, Massachusetts, 8-11 March 2020

Poster Number: 1076

HIV-1 Dynamics Following Universal Testing-and-Treatment within HPTN 071 (PopART)

William J. M. Probert1, Rafael Sauter1, Michael Pickles2, Anne Cori2, Helen Ayles3, Peter Bock4, Deborah J. Donnell5, Sarah Fidler2, Richard J. Hayes6, Christophe Fraser1, for the HPTN 071 (PopART) study team1University of Oxford, Oxford, UK, 2Imperial College London, London, UK, 3Zambart, Lusaka, Zambia, 4Desmond Tutu TB Centre, Western Cape, South Africa,

5Fred Hutchinson Cancer Research Center, Seattle, WA, USA, 6London School of Hygiene & Tropical Medicine, London, UK

FIGURE 2. Predicted incidence (left-hand panels) and predicted probability oftransmitting HIV at least once (right-hand panels), over the trial period stratified bysex and sexual risk-taking behaviour. Distribution over 1000 parameter sets frommodel calibration.

RESULTS – PREDICTED PATTERNS DURING THE TRIALIncidence was predicted to be higher in the high-risk group but differences in the size of thesepopulations meant the largest proportion of transmissions was between those in the medium risk groups(Figs. 1,2,3). The overall probability of onwards transmission was similar between medium- and high-riskgroups (Fig. 2).

BACKGROUNDUniversal HIV testing-and-treatment (UTT) has been shownto be effective in high prevalence areas in sub-SaharanAfrica (SSA) to reduce HIV incidence. Community-wideinterventions may change the contribution of groups withdifferent sexual risk-taking behaviour to ongoing HIVincidence and transmission. Understanding these changeswill inform future policy towards achieving zero newinfections in the context of UTT.METHODSUsing an individual-based model (PopART-IBM), developedas part of the HPTN 071 (PopART) trial, we project theimpact of four scenarios of UTT to 2030, stratified bycategories of sexual risk-taking behaviour. The model wascalibrated to an intervention community in Zambia using trialdata. Categories of sexual risk-taking behavior were derivedfrom trial data from surveys of sexual behavior on reportednumber of lifetime sexual partners and age.

Projected scenarios were : 1) Simulation of the PopARTtrial and continuation of a CHiPs intervention after the trial;2) Discontinuation of a CHiPs intervention after the trial; 3)Simulation of nationwide CHiPs intervention in communitieswhere the trial did not take place; 4) No simulated trial andno CHiPs intervention (counterfactual).

Continuation of a UTT approach to2030 predicts both a substantial dropin incidence and an epidemic that willbecome more concentrated in thosewith the highest levels of sexual risk-taking behaviour.

FIGURE 3. Average proportion of total transmissions over trial period (2014.5-2018)between individuals in different groups of sexual risk-taking behavior and stratified bythe sex of the source of transmission. Denominator is total number of transmissionsover the trial period.

ACKNOWLEDGEMENTSHPTN 071 is sponsored by the National Institute of Allergy and Infectious Diseases (NIAID) underCooperative Agreements UM1-AI068619, UM1-AI068617, and UM1-AI068613, with funding from the U.S.President's Emergency Plan for AIDS Relief (PEPFAR). Additional funding is provided by the InternationalInitiative for Impact Evaluation (3ie) with support from the Bill & Melinda Gates Foundation, as well as byNIAID, the National Institute on Drug Abuse (NIDA) and the National Institute of Mental Health (NIMH), allpart of the U.S. National Institutes of Health (NIH). We also wish to acknowledge implementing partners inSouth Africa (City of Cape Town and Western Cape Government health departments, Kheth’ Impilo, ANOVAHealthcare, SACTWU Worker Health Programme and Supply Chain Management Services) and Zambia(Zambian Ministry of Health, CIDRZ, ZPCT II and JSI). The content is solely the responsibility of the authorsand does not necessarily represent the official views of the NIAID, NIMH, NIDA, PEPFAR, 3ie, or the Bill &Melinda Gates Foundation.

FIGURE 1. HIV prevalence at trial start (mid-2014) (left-hand panel) and proportionof the entire population (14+ years old; right-hand panel) in each subgroup.Distribution is over 1000 parameter sets from model calibration.

LowM

ediumH

igh

0 5 10 15 20 25

0

1

2

3

4

5

0.00

0.25

0.50

0.75

1.00

0.00

0.05

0.10

0.15

0.20

Incidence per 100 person−years

Den

sity

FemaleMale

Low risk

Medium risk

High risk

LowM

ediumH

igh

0.0 0.2 0.4

0

20

40

60

0

10

20

0

2

4

6

Probability of transmitting HIV over trial period among PLHIV at trial start

Den

sity

FemaleMale

LowM

ediumH

igh

0 5 10 15 20 25

0

1

2

3

4

5

0.00

0.25

0.50

0.75

1.00

0.00

0.05

0.10

0.15

0.20

Incidence per 100 person−years

Den

sity

FemaleMale

HIV acquisition HIV transmission

FIGURE 4. Projections of different scenarios of UTT from 2020-2030 stratified bysexual risk-taking behaviour. Shaded area is 95% quantiles of model output.

RESULTS - PROJECTIONS TO 2030Making ART universally accessible to all who are HIV-positive in the PopART community would lead to asubstantial decline in incidence in all risk groups but wouldconcentrate new cases in those with the highest levels ofrisk-taking behaviour (65% of incident cases vs 54% if noUTT was implemented; Fig. 4). While population HIVincidence to 2030 decreases substantially, the modelpredicts continued persistence of an HIV epidemic in thehigh-risk subpopulation in all scenarios unless nationwideUTT is adopted.

CONCLUSIONSOur results highlight that continuation of a UTT approach to2030 has the capacity to confer dramatic reductions in newHIV infections. Targeting of high-risk individuals for HIV andSTI prevention, testing, and treatment may be necessaryfollowing successful UTT interventions in order to eliminateHIV as a public health issue in SSA.

RESULTS – CHARACTERISTICS AT TRIAL STARTAt trial start (mid-2014.5) mean HIV prevalence in the risk groups was 6% (low), 22% (medium), and53% (high). The proportion of the 14+ year old population in each group was 52% (low), 38% (medium)10% (high). Predicted female-to-male HIV prevalence ratio decreased with increasing risk group (Fig. 1).

LowM

ediumH

igh

0.0 0.1 0.2 0.3

0.0

2.5

5.0

7.5

0.0

2.5

5.0

7.5

0

5

10

15

Proportion of population in subgroup (mid−2014)

Den

sity

FemaleMale

LowM

ediumH

igh

0.00 0.25 0.50 0.75

0

5

10

15

20

0

2

4

6

0

1

2

3

4

5

HIV prevalence (mid−2014)

Den

sity

FemaleMale

Low risk

Medium risk

High risk

LowM

ediumH

igh

0 5 10 15 20 25

0

1

2

3

4

5

0.00

0.25

0.50

0.75

1.00

0.00

0.05

0.10

0.15

0.20

Incidence per 100 person−years

Den

sity

FemaleMale

PopART/UTT continued(scenario 1)

No PopART/No UTT (scenario 4)

PrevalenceN

umber of

incident casesProportion of

incident cases

2010 2030 2010 20300.0

0.4

0.8

0.0

0.4

0.8

0

300

600

Risk group of source