Modelling HIV and HCV treatment as prevention amongst

Treatment as prevention (TASP) for HIV
and HCV: The evidence and modelling
Peter Vickerman
HIV treatment as prevention
• Discussion started in late 80s/early 90s, because treatment
reduced viral load in plasma and semen
• Ever since HIV treatment has been available, modelling
has been used to argue whether it could decrease or
increase transmission (Anderson Nature 1991 and others, Blower
Science 2000)
• Early cohort studies suggesting it may reduce transmission
in sero-discordant couples (Musicco Arch Int med 1994)
• Quinn NEJM 2000 was first to show viral load related to
HIV infectivity
Hypothesis proven by HPTN 0052
(Cohen NEJM 2011)
• Sero-discordant heterosexual couples
• 96% reduction for linked HIV transmissions and 89%
reduction for all transmissions
However, results vary
• Some cohort studies have suggested lower effectiveness:
• Pooled estimate gives 42% reduction
• Is this due to bias or lower adherence in real life settings
• All studies from sero-discordant couples – generalisable?
Risk of HIV Transmission in Serodiscordant Couples Treated vs Untreated
With Antiretroviral Therapy in Observational StudiesSource: Figure adapted
with permission from Cochrane HIV/AIDS Group. (Anglemyer JAMA 2013)
Population impact amongst heterosexuals in
real world setting
• Two innovative cohort studies from Kwa Zulu Natal, South Africa
1. Is risk of HIV transmission related to coverage of ART in surrounding
local community (Tanser Science 2013)?
2. Is risk of HIV transmission related to coverage of ART amongst
opposite sex household members (Vandormael, Lancet Global Health 2014)?
• Study 1 showed that 38% less likely to become HIV-infected if 30-40%
coverage of ART in community instead of <10%
• Study 2 showed transmission risk reduced by 5-6% for each 10% increase
in ART coverage amongst HIV+ household members:
• If 100% coverage then 45% reduction in incidence
Tanser, Science 2013
Vandormael, Lancet Global Health 2014
Modelling impact of
scaling up ART
• Model projections from 12
models for South Africa
• Broadly agree with Tanser
• Impact of 80% ART access
to all HIV +ves by 2020:
• 50-70% ↓ HIV incidence
if 15% LTFU 3 years
• 60-80% ↓ if no LTFU
Eaton Plos med 2012
HIV TASP IN HIGH RISK
GROUPS - MSM AND PWID
Evidence of efficacy in MSM
• Until recently, little data on effect of ART on MSM HIV transmission
• New data from European PARTNER study found NO transmissions
through condomless sex when HIV+ partner on ART and virally
suppressed:
• But only 330 years of follow up so far
Rodger, A. et al. HIV transmission risk through condomless sex if the
HIV positive partner is on suppressive ART: PARTNER study. CROI
2014
What about at population level for MSM –
example of UK for 2001 to 2010
Improvements in cascade of care from 2001 to 2010
• 4 fold increase in frequency of HIV testing of MSM
• Time to diagnosis decreased from 4 to 3.2 years
• Proportion diagnosed at CD4>350 increased from 48% to 65%
• ART coverage in diagnosed MSM rose from 69% to 80%
BUT:
• Estimated number undiagnosed has not changed at about 8000
• Number new HIV infections each year remained stable at ~2500
• Similar situation of increasing or stable HIV diagnoses in MSM
populations in other western settings with high coverage of ART
Birrell. Lancet 13: 313–18 2013
Griensven curr op hiv aids:4: 300-307 2009
Muessig AIDS 2012
What does modelling suggest for UK
If no ART
% condomless sex
HIV incidence
• Modelling suggests that increases in condomless sex acts maintained HIV
incidence with ART – agrees with increase in STI prevalence over period
• Model projects much higher HIV incidence without ART
• Undiagnosed cases contribute 82% of new HIV infections and mostly acute
• To reduce HIV incidence need to reduce both undiagnosed fraction and
provide ART at diagnosis
Phillips, A. et al. Plos one 2013
Reduce % undiagnosed
and ART at diagnosis
In People who inject drugs (PWID)
HIV or HCV incidence (per 100 person years)
• Although biologically plausible, NO direct evidence for efficacy or
effectiveness
• Some studies used ecological correlations between community
measures of HIV viral load and HIV incidence to postulate that
scaled-up ART has decreased HIV transmission amongst IDUs
30
HCV incidence
25
HIV incidence
20
15
10
5
0
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Year
Wood BMJ 2009 for Vancouver; Kirk CROI 2011 for Baltimore
What has modelling shown?
• If protective then ART is:
• Likely to be effective and cost-effective
• Impact limited by role of HIV acute
phase
• Unlikely to reduce HIV transmission to
low levels unless combined with other
interventions
Strathdee Lancet 2010, Degenhardt Lancet 2010,
Kato JAIDS 2013, Alistar Plos Med 2011
TASP FOR HCV IN PWID
HCV treatment as prevention for PWID
•
•
•
•
Highly effective curative treatment exists
No evidence that treatment can reduce transmission at population level
However, modelling suggests could have strong impact
And could be more cost-effective than treating non-IDUs
Zeiler DAD 2010, Vickerman DAD 2010,
Martin J. Hepatology 2011, Martin Hepatology 2012
New DAA THERAPY COULD dramatically reduce
HCV PREVALENCE over 15 YEARS , BUT…
Treatment rates required to halve
chronic prevalence within 15 years:
Edinburgh: 15/1000 PWID annually
(2-fold increase)
Melbourne: 40/1000 PWID annually
(13-fold increase)
Vancouver: 76/1000 PWID annually
(15-fold increase)
If future treatments cost $50,000
USD per course, halving prevalence
within 15 years would require:
• Edinburgh: $3.2 million USD
annually
• Melbourne & Vancouver:
~$50 million USD annually
Martin NK, Vickerman P, Grebely J, Hellard M, Hutchinson SJ, Lima VD, Foster
GR, Dillon J, Goldberg DJ, Dore G, and Hickman M. HCV treatment for prevention
among people who inject drugs: modeling treatment scale-up in the age of direct
acting antivirals. Hepatology 2013
Summary
• HIV treatment can be highly effective for reducing infection risk in serodiscordant couples and MSM:
• Impact at population level is more uncertain although evidence emerging
• However, in MSM HIV incidence has increased in many settings as ART has
been scaled up massively
• Evidence for impact in PWID is weak although biologically plausible, but if
effective:
• Could reduce HIV transmission dramatically, be cost-effective, but unlikely to
lead to elimination
• HCV treatment works, but no evidence that works as a prevention strategy,
although biologically plausible
• Modelling suggests could have large impact and be cost-effective, but
• Evidence is needed at population level, costs need to be reduced and
cascade of care needs improving to enable scale up