Treatment as Prevention: HAART Expansion - A Powerful Strategy to Reduce AIDS Morbidity and Mortality and HIV Incidence

Julio Montaner MD, FRCPC, FCCP, FRSC Director, BC-Centre for Excellence on HIV/AIDS, Providence Health Care Professor of Medicine and Head, Division of AIDS, University of British Columbia President, International AIDS Society

UNAIDS, Geneva, March 24th 2010

British Columbia Centre for Excellence in HIV/AIDS USA - Trends in Annual Rates of Death Ages 25 to 44

40 Unintentional injury 35 Cancer 30 Population Heart 25 disease Suicide 20 15 HIV infection 10 Homicide Deaths per 100,000 5 Chronic liver disease 0 Stroke 82 84 86 88 90 92 94 Diabetes Year Harald zur Hausen Françoise Barré-Sinoussi Luc Montagnier

Announced Oct 6th 2008 Vancouver 1996 “One World One Hope” Vancouver 1996 “One World One Hope”

% Progression to AIDS in 3 yrs Plasma Viral Load, a strong CD4+ 100 cells/µL > 750 Predictor of outcome in 80 501-750 351-500 60 201-350 HIV Infected Individuals < 200 40 High Plasma Viral Load: Poor Prognosis 20 Low Plasma Viral Load: Good Prognosis 0 > 30 10-30 3-10 0.5-3 < 0.5 Plasma HIV RNA (thousand copies/mL) J Mellors et al. Annals 1997 Montaner et al JAMA, March 25th 1998 Vancouver 1996 “One World One Hope” AZT + NVP

AZT + ddI

% Progression to AIDS in 3 yrs

CD4+ 100 cells/µL Triple Therapy: AZT + ddI + NVP > 750 80 501-750 351-500 60 201-350 < 200 Gulick et al; JAMA, July 1, 1998 40 0 20

0 > 30 10-30 3-10 0.5-3 < 0.5 -1 Plasma HIV RNA (thousand copies/mL) Dual Therapy Regimens J Mellors et al. Annals 1997 -2

Change in Viral Load Triple Therapy: AZT+3TC+IDV -3 0 52 Study Weeks Impact of HAART in BC-CfE

140 Death Rate per 1000

120 Greater than 85% reduction

100 in death rate among those

80 engaged in care

60

40

20

0 1993-94 1995-96 1997-98 1999-00 2001-02 2003-04

Modified from Hogg et al, Lancet. 2009 Impact of HAART in BC-CfE

140 Death Rate per 1000 35 Life Expectancy at age 20

120 30

100 25

80 20

60 15

40 10

20 5

0 0 1993-94 1995-96 1997-98 1999-00 2001-02 2003-04 1993-94 1995-96 1997-98 1999-00 2001-02 2003-04

Modified from Hogg et al, Lancet. 2009

HAART Can Reduce HIV Transmission

HAART stops HIV replication ↓ HIV levels fall to undetectable in blood as well as in sexual fluids ↓ Sharp reduction in HIV transmission Prevention Strategies

- Education - Change in behavoir - - New strategies/technology - Vaccines

Existing strategies have failed to contain the global HIV pandemic Vertical Transmission

Canada, 1990 to 2008

Modified from Alimenti et al CAHR, 2009 Discordant Couples

S Attia, M Egger, M Muller, M Zwahlen and N Lowa. AIDS. 2009 Jul 17;23(11):1397-404 Discordant Couples

Studies of heterosexual discordant couples observed no transmission in patients treated with ART and with viral load below 400 c/ml, but data were compatible with one transmission per 79 person-years. S Attia, M Egger, M Muller, M Zwahlen and N Lowa. AIDS. 2009 Jul 17;23(11):1397-404 B&M Gates Fdn: HIV Transmission Risk in Heterosexual Serodiscordant Couples

. 3,400 couples In 7 African countries

. All counseled and given free condoms

. HAART initiated based on CD4 count eligibility

. Over the next 1 to 3 years, 103 new HIV infections

. All but 1 infection occurred in the untreated couples

. Estimated 92% reduction of HIV transmission by HAART

. Unadjusted relative risk: 0.17 (95% CI: 0.004-0.94; P = .037)

. Adjusted relative risk: 0.08 (95% CI: 0.002-0.57; P = .004)

. Adjusted for visit and CD4+ cell count at initiation

Donnell D, et al. CROI 2010. Abstract 136. B&M Gates Fdn: HIV Transmission Risk in Heterosexual Serodiscordant Couples

. 3,400More couples recently Inan 7 email African has countries been circulated saying that: . All “counseledThe single and case given of transmission free condoms involved a man who initiated ARVs 18 days before his 12-month . studyHAART visit. initiated At this basedvisit his on partner CD4 count tested eligibility positive for HIV, . Over the nexthaving 1 to been 3 years, negative 103 new at month HIV infections9.”

. All but 1 infection occurred in the untreated couples

. Estimated 92% reduction of HIV transmission by HAART

. Unadjusted relative risk: 0.17 (95% CI: 0.004-0.94; P = .037)

. Adjusted relative risk: 0.08 (95% CI: 0.002-0.57; P = .004)

. Adjusted for visit and CD4+ cell count at initiation

Donnell D, et al. CROI 2010. Abstract 136.

Wood et al, BMJ, 2009 Wood et al, BMJ, 2009

Whiskers represent 95% confidence intervals. Wood et al, BMJ, 2009

Whiskers represent 95% confidence intervals. Wood et al, BMJ, May 16, 2009 Impact of HAART in BC-CfE

6 months Baseline Frequency of people of Frequency

Plasma Viral Load (log10 copies/mL) Distribution

Modified from Anema et al. EIDJ 2009 Impact of HAART in BC-CfE

Baseline 24 months Frequency of people of Frequency

Plasma Viral Load (log10 copies/mL) Distribution

Modified from Anema et al. EIDJ 2009 New HIV and Syphilis in BC Rate per 100,000 population

M REKART, BC-CDC, 2006 New HIV and Syphilis in BC Rate per 100,000 population

M REKART, BC-CDC, 2006 Cost-Effectiveness of HAART BC-DTP

“HIV deficit” in BC in 2005: 400

Cost of Medical Management of 1 HIV infection over a lifetime = $250,000 800 cases per year Averted lifetime Rx cost up to 2001 US $96.4M A total of 3,963 pts were on HAART in BC in 2005 400 cases per year Total actual drug cost (using patented drugs) in 2005 $49 million US Cost-Effectiveness of HAART BC-DTP

“HIV deficit” in BC in 2005: 400

Cost of Medical Management of 1 HIV infection over a lifetime = $250,000

Averted lifetime Rx cost up to U$A 100M A total of 3,963 pts were on HAART in BC in 2005 Total actual drug cost (using patented drugs) in 2005

U$A 50M The Bc-CfE Mathematical Model Viviane D. Lima et al JID 2008

Adherence: 0% - <40% Adherence: 0% - <40% Guideline: !200 cells/mm 3 Guideline: !350 cells/mm 3 550 500

500 50% 50% 450 450 400 400 350 75% Infections Infections 350 75% CD4 350/mm3 300 New CD4 200/mm3 New 300 of of 250 90% 250 90% Adh <40% 200

200 Number Adh <40% Number 150 150 100% 100% 100 100 1995 2000 2005 2010 2015 2020 2025 2030 2035 1995 2000 2005 2010 2015 2020 2025 2030 2035

Year Year

Coverage: Coverage: 50% 75% 90% 100% 50% 75% 90% 100%

Current Adherence: 78.5% Current Adherence: 78.5% Guideline: !200 cells/mm 3 Guideline: !350 cells/mm 3

550 550

500 500 450 50% 450 50% 400 400 CD4 350/mm3 CD4 200/mm3 Infections 350 75% Infections 350 New 300 New 300 of of 75% Adh 40 - 80% Adh 40 - 80% 250 90% 250 Number 200 Number 200 90% 150 150

100 100% 100 100% 1995 2000 2005 2010 2015 2020 2025 2030 2035 1995 2000 2005 2010 2015 2020 2025 2030 2035

Year Year

Coverage: 50% 75% 90% 100% Coverage: 50% 75% 90% 100%

Adherence: 80% - <95% Adherence: 80% - <95% Guideline: !200 cells/mm 3 Guideline: !350 cells/mm 3 550 550

500 500 450 50% 450 400 400 50% CD4 350/mm3 CD4 200/mm3 Infections 350 75% Infections 350 New 300 New 300 of of 75% Adh 80 - 95% Adh 80 - 95% 250 90% 250 Number 200 Number 200 90% 150 150 100 100% 100 100% 1995 2000 2005 2010 2015 2020 2025 2030 2035 1995 2000 2005 2010 2015 2020 2025 2030 2035

Year Year

Coverage: 50% 75% 90% 100% Coverage: 50% 75% 90% 100%

Adherence: 95% - 100% Adherence: 95% - 100% Guideline: !200 cells/mm 3 Guideline: !350 cells/mm 3 550 550

500 500

450 50% 450 400 400 50% CD4 350/mm3 CD4 200/mm3 Infections 350 75% Infections 350

New 300 New 300

of of 75% Adh 95 - 100% Adh 95 - 100% 250 90% 250 Number 200 Number 200 90% 150 100% 150 100 100 100% 1995 2000 2005 2010 2015 2020 2025 2030 2035 1995 2000 2005 2010 2015 2020 2025 2030 2035 V D Lima et al Year Year JID 2008 Coverage: 50% 75% 90% 100% Coverage: 50% 75% 90% 100% Incremental net benefit (Millions of CDN $) over 30 years Overall population and patient-centered 800

Overall Population Patient-Centered incremental net benefit of

600 increasing uptake of Net Benefit (million $ Can 2005)

400 HAART from 50% to 75% over 30 years, based on a 200 willingness-to-pay

0 thresholds of $50,000 per

0 5 10 15 20 25 30 quality-adjusted life year.

Time (years) K Johnston et al, submitted, 2010 Summary

HAART is widely regarded as a cost effective, life- saving strategy

↓ Mortality of treated HIV/AIDS patients ↓ Morbidity of treated HIV/AIDS patients ↓ Health Resource utilization ↓ Vertical Transmission of HIV infection

Furthermore, when the impact of HAART on HIV transmission is considered, HAART expansion becomes a cost-averting strategy The third approach, though, is the most intriguing. This is to do nothing more than press ahead faster with the treatment program. Since treatment reduces viral load, it should, in theory, make those being treated less infectious. Of course, theory is one thing and practice another. But studies in Taiwan and British Columbia (the latter by Julio Montaner, the incoming president of the International AIDS Society, which organizes the conference) have shown big falls in transmission rates as ARVs have been rolled out. The Power of HAART: Demographic Model

Treat 30%

HIV prevalence Cost of treatment

Treat all Treat 30%

Treat all Number of infections prevented

Montaner et al, Lancet 2006 R Granich, C Gilks, C Dye, K De Cock, B Williams. The Lancet Nov 26th 2008 AIDS Nov 27th 2008, The Economist Deploying the drugs used to treat AIDS may be the way to limit its spread

Illustration by Peter Schrank

Thank you AIDS Nov 27th 2008, The Economist Deploying the drugs used to treat AIDS may be the way to limit its spread

Illustration by Peter Schrank

Thank you

An Alternative Approach

Preliminary Results Methods

 Prospective ecological study in BC, Canada

 Used administrative records to evaluate the association between expansion of HAART coverage, population level plasma HIV-1-viral load and new HIV diagnoses

 HIV testing, CD4 & viral load testing and HAART distribution are centralized and free in BC

 Data for second half of 2009 is preliminary due to delayed reporting, therefore only the first half of 2009 was used for statistical analyses

Montaner et al, CROI 2010 January 2004

Summer of 1996 The second expansion of HAART occurred prior to the new 2008 IAS-USA Guidelines, which were adopted in BC at the end of 2008

The first expansion of HAART occurred as a result of the new 1996 IAS-USA Guidelines, which were adopted in BC in the summer of 1996

Year Montaner et al, CROI 2010

1.00

0.20

0.10

0.04 Acquired resistance 1.00 falling 0.02 Incidence/yr 0.01 0.201995 2000 2005 2010 90 0.10 80 Plasma viral load 0.04 suppression rising

70 0.02 Viral load

< 50/mL (%) 0.6001 1995 2000 2005 2010 Number of New HIV+ Diagnoses Montaner et al, CROI 2010 Year New HIV+ Diagnoses (IDU) New HIV+ Diagnoses Active on HAART HIV+ Diagnoses per Year per Diagnoses HIV+ Number of Active HAART of Active Number New HIV+ Diagnoses (All) New HIV+ .015

Participants and Number of New Number and Participants p=0 Number of Active HAART Participants HAART Active of Number Number of Active HAART Participants and Number of New HIV+ Diagnoses per Year Number of New HIV+ Diagnoses p=0.015

New HIV+ Diagnoses (All)

Active on HAART p=0.085

New HIV+ Diagnoses (IDU) Number of Active HAART Participants

p=0.026

Year Montaner et al, CROI 2010 “True” New Yearly HIV Diagnoses in BC

BC-CDC Updated March 2010 HIV testing in BC, 1985 to 2008

Year # of HIV Tests

Jan 2004

BC-CDC Report, 2009 , 1999-2008 Infectious Syphilis, 1999-2008

• BC 2004 • BC 2004

x Canada x Canada

Genital Chlamydia, 1999-2008 Gonorrhea, 1999-2008

• BC x Canada

x Canada

2004 • BC 2004 Highest Non IDU HIV-1- Plasma Viral load per Year

IDU Ever on Treatment & Censoring at the time of Death or Move

The proportion of HIV infected IDUs engaged in care in BC with plasma viral load >1500 c/mL, as a surrogate for “high” community HIV-1-viral load, decreased from ~50% in 2000-04 to ~20% in 2009 (p<0.001) Montaner et al, CROI 2010 “Provincial Viral Load” All Patients Ever Tested for Plasma HIV-1-Viral Load in BC

Censoring at the time of Death or Move New Data: Pre HAART CD4 Count

Montaner et al, 2010, Preliminary Data Community pVL and New HIV Diagnoses San Francisco

) 30,000 P = .005 for Mean CVL association* Cases HIV Diagnosed Newly of Number 1200 Newly diagnosed and 25,000 reported HIV cases 1000 copies/mL 20,000 800 798 15,000 600 642 523 518 10,000 434 400

5000 200

Mean Community Viral Load ( 0 0 2004 2005 2006 2007 2008 *Data insufficient to prove significant association with reduced HIV incidence.

Das-Douglas M, et al. CROI 2010. Abstract 33. A Formidable Challenge

A Unique Opportunity When to Start HAART? A matter of Perspective

Viral Load

CD4 Count

years When to Start HAART? A matter of Perspective

Viral Load

CD4 Count

years

years When to Start HAART? A matter of Perspective

Viral Load

CD4 Count

years

years

years When to Start HAART? A matter of Perspective

Viral Load

CD4 Count

years

years

years Cost: 2010 to 2050 70 7

60 6

50 5 )

40 4 Bn$ 30 3

20 2 Cost ( Deaths (M)

10 1 Person years on ART (M) 0 0 0 Current1 2002 3503 5004 Immediate5 6 Economics of ART up to 2050 in South Africa Current policy vs. Universal Access at different CD4 counts

Granich. CROI 2010 STOPSTOP HIVHIV && AIDSAIDS: Seek and Treat to Optimally Prevent HIV & AIDS*

* Supported through a $2.5M five year Avant Garde Award by the National Institute for Drug Abuse (NIDA) at the NIH in 2008 and $48M (+ drugs) four year outreach grant by BC Govt in 2010 STOPSTOP HIVHIV && AIDSAIDS: Seek and Treat to Optimally Prevent HIV & AIDS*

Prospectively Evaluate the Impact of HAART Expansion on AIDS Morbidity and Mortality and HIV Incidence in BC

Intervention Primary Endpoint HAART Expansion HIV Incidence within medical guidelines at years 3 to 5

Secondary Endpoints: MORBIDITY AND MORTALITY, CD4 COUNTS, HIV-1-RNA LEVELS, RESISTANCE, ADVERSE EVENTS, SAFETY, ADHERENCE, HOSPITALIZATIONS, RESOURCE UTILIZATION

* Supported through a $2.5M five year Avant Garde Award by the National Institute for Drug Abuse (NIDA) at the NIH in 2008 and $48M (+ drugs) four year outreach grant by BC Govt in 2010 HAART Expansion to Reduce AIDS Morbidity & Mortality, and HIV Incidence

 HAART has a substantial added preventive value

 The magnitude of this effect is not yet fully characterized, and may well vary in different settings

 Seek and Treat among those who have a medical indication for HAART cannot wait for the above to be resolved

 Many lives will be saved and much insight will be gained from closely monitoring a more “aggressive”roll out of HAART within Rx Guidelines

 Seek and Treat outside the range where treatment is medically indicated remains a research question

 However, Rx Guidelines leave few outside the “treatment envelope”

 TAP should serve to re-energize Universal Access Combination prevention

Biomedical Interventions

HIV testing, Structural linkage to care Interventions and expanded HAART HIVHIV coverage PreventionPrevention

Individual Community and small Interventions group behavioral interventions

Modified from T. Coates A Statistician’s Opinion

All scientific work is incomplete - whether it be observational or experimental. All scientific work is liable to be upset or modified by advancing knowledge. That does not confer upon us a freedom to ignore the knowledge we already have, or to postpone the action that it appears to demand at a given time.

Bradford-Hill, A. 1965 The environment and disease: Association or Causation? President address at January 14 meeting. Proceedings of the Royal Society of Medicine 163 (seriesB): 295-300 Seek and Treat to Optimally Prevent HIV & AIDS STOP HIV & AIDS

R Hogg, E Wood, T Kerr, M Tyndall, A Levy, PR Harrigan, Viviane Lima, Aranka Anema, Stephen Smith, Warren O’Brien Pedro Cahn, Jose Esparza, Craig Mc Clure, Robin Gorna Jacques Normand, Nora Volkow IAS - USA ART Guidelines Panel, IAS, WHO and UNAIDS

BC-MoH and MHL&S SPH Foundation Merck, Gilead, ViiV MSHRF, CIHR, NIDA and NIH H&W, Ottawa

Research Staff and Study Participants

British Columbia Centre for Excellence in HIV/AIDS British Columbia Centre for Excellence in HIV/AIDS

Thank You