Title: Projeto Praa Onze Universidade Federal do Rio de Janeiro
1Projeto Praça OnzeUniversidade Federal do Rio de
Janeiro
When to Start Antiretroviral Therapy Where is
the Pendulum Now?
Mauro Schechter, MD PhD Professor of Infectious
Diseases Principal Investigator, Projeto Praça
Onze Universidade Federal do Rio de Janeiro
IAS Conference, Sydney, July 2007
2Case 1
- Asymptomatic 35 year old male patient, known to
be HIV since 1999, with confirmed CD4 count of
280 cells/?L and plasma viral load of 10,000
copies/mL. Would you - Start treatment immediately
- Start treatment only if the patient requested
- Wait until the CD4 count drops below 200 cells/?L
3Case 2
- Asymptomatic 35 year old male patient, known to
be HIV since 1999, with confirmed CD4 count of
280 cells/?L and plasma viral load of 150,000
copies/mL. Would you - Start treatment immediately
- Start treatment only if the patient requested
- Wait until the CD4 count drops below 200 cells/?L
4Case 3
- Asymptomatic 35 year old male patient, known to
be HIV since 1999, with confirmed CD4 count of
320 cells/?L and plasma viral load of 10,000
copies/mL. Would you - Start treatment immediately
- Start treatment only if the patient requested
- Wait until the CD4 count drops below 200 cells/?L
5Case 4
- Asymptomatic 35 year old male patient, known to
be HIV since 1999, with confirmed CD4 count of
320 cells/?L and plasma viral load of 10,000
copies/mL. Six months before his CD4 count was
450 cells/?L and one year before 630 cells/?L.
Would you - Start treatment immediately
- Start treatment only if the patient requested
- Wait until the CD4 count drops below 200 cells/?L
6When to Start Antiretroviral Therapy
Late clinical stages
Early Clinical Stages
gt 500
lt 200
200
350
Any viral load
High Viral load
CD4
Schechter, 2004 (JID 20041901043-1045)
7IASUSA Guidelines When to Start Antiretroviral
Therapy
8Reasons for Deferral
- Risk of progression to AIDS or death is
relatively small in comparison with those at
lower CD4 counts - Concerns about complications of ART
- Potential future availability of more potent and
tolerable drugs - Fear of waning adherence, resistance
accumulation, and exhaustion of drug options
9Rate of AIDS or death by current CD4 count in
ART-naïve patients
Current CD4 count Person-years Events
Rate (/100 pys) (95 CI) gt
650 5464 53 0.97 (0.70
1.24) 500-649 5740 89
1.55 (1.22 1.88) 350-499
8643 212 2.45 (2.11
2.79) 200-349 7062 349
4.94 (4.41 5.47)
CHIC cohort
10Response to Treatment in the First Year of
HAART
- 5 cohorts in Europe and Canada, N4,143
- ART-naïve, started HAART 1996-2002
with CV gt 500 c/mL
100
90
80
70
60
with VL gt 500
50
40
30
24.8
23.0
17.3
20
12.4
10
8.4
8
10
0
1996
1997
1998
1999
2000
2001
2002
Lampe S, et al. 12th CROI, 2005
11HOPS Cohort Early, Uninterrupted ART Associated
with Improved Outcomes
Incidence per 1,000 patient years by pre-HAART
CD4 count and time on HAART (n 4,421)
Mortality
Opportunistic Infections
71.5
HAART lt 95 of time HAART gt 95 of time
HAART lt 95 of time HAART gt 95 of time
55.9
P 0.05 for difference by HAART use
P 0.05 for difference by HAART use
47.8
37.8
38.5
Incidence per 1,000 Person-Years
Incidence per 1,000 Person-Years
26.1
25.5
22.3
21.4
20.1
16.2
15.9
14.2
11.5
10.4
7.5
7.2
7.3
5.4
2.4
0-49 50-199 200-349
350-499 500
0-49 50-199 200-349 350-499
500
CD4 category
CD4 category
CD4 Cell Count Category
Lichtenstein K, et al. CROI 2006.
12Higher CD4 count is associated with decreased
risk of death from non-AIDS malignancy and
liver-related causes
lt50
50-
100-
200-
Latest CD4 count
350-
Non-AIDS malignancy
gt500
lt50
50-
100-
200-
Liver-related
350-
gt500
1
10
100
Hazard ratio (95 CI)
Arch Intern Med 2006 and CROI 2007
13Cardiovascular Events in SMART
Phillips A, et al. 14th CROI, 2007
14FIRST Study Risk of Non-Opportunistic Diseases
(non-OD)
- 1397 pts starting 1 of 3 ART strategies
- Incidence of non-ODs (non-AIDS-defining
malignancies hepatic, CV and renal disease)
declined with higher CD4 count - At CD4 gt200, non-OD events accounted for more
morbidity and mortality than OD - Most recent CD4 count independently predicted
risk of non-OD
Event rate by latest CD4
- Rates decline at higher CD4 counts
- Non-OD gt OD at CD4 gt200
Baker J, et al. CROI 2007
15Cost-Effectiveness of Early vs. Deferred HAART
- Markov modeling approach to compare HAART started
at CD4 count gt 350 CD4 vs 200-350 - Johns Hopkins HIV clinic database
- Starting HAART earlier rather than later is
a cost-effective strategy (by the generally
accepted benchmark in the US).
Mauskopf JA, et al. JAIDS 2005
16Should the Pendulum Swing Back?
- The Case for Earlier Therapy
- Easier, more potent, and less toxic therapy
- Better response to therapy
- Fear of exhaustion of drug options due to
resistance has lessened - Cohort studies suggest benefit with earlier
therapy - Prevention of specific complications of HIV
infection - Prevention of non-opportunistic complications
- Cost-effectiveness
17Should the Pendulum Swing Back?
- But
- Absolute risk differences between groups starting
ART between 200-349 versus gt 350 are small and
may not reflect all of the benefits of ART or all
of the risks (e.g., there is no control group
with the same CD4 count who did not start ART) - Bias possible due to confounding by indication
for initiation of early ART (i.e., what type of
patient has been started at ART gt 350?)
18Should the Pendulum Swing Back?
- Is a controlled, randomized trial necessary to
answer the when to start question? - Yes
- Yes, but it would be almost impossible to recruit
- Yes, but it would take too long
- No, observational data would be enough to modify
present recommendations