When to StartWhen to Start

When to StartWhen to StartData PresentationData Presentation

DHHS and IAS-USA: Recommendations for Initiation of ART in Naïve Patients

Available at: http://www.aidsinfo.nih.gov/ContentFiles/AdultandAdolescentGL.pdf. Revision November 3, 2008; Hammer SM, et al. JAMA 2008;300:555-570.

Clinical CategoryCD4 Cell

Count (cells/mm3)

Viral Load(copies/mL)

2008 DHHSGuidelines

2008 IAS-USA

Guidelines

AIDS-defining illness or severe symptoms

Any value Any value Treat

Asymptomatic <200 Any value Treat

200 to 350 Any value Treat

>350 >100,000Generally defer

treatment

>350 <100,000Generally defer

treatment

Pregnant women Any value Any value TreatHIV-associated

nephropathy Any value Any value Treat

HIV/HBV coinfection when HBV treatment is indicated

Any value Any value TreatConsider treatmen

t

ACTG 5164: Early vs. Deferred ART with Acute OIs

• Assessment of optimal timing of ART in pts with acute OI (N=285)• 92% Tx-naïve; median CD4+ 29 cells/mm3 and HIV RNA 5.07 log10 c/ml

• OIs with effective antimicrobial therapy only• PCP (63%), bacterial infections, cryptococcal disease, MAC, toxoplasmosis• TB excluded

Zolopa A, et al. 15th CROI; Boston, MA (2008); Abst. 142.

-14

Study day

0 2 28 42 84 224

48wks

48wks

Enrollment

Deferred ArmStart ART

Opportunistic Infection

TreatmentStarts

ImmediateArm

Start ART

RecommendedStart window

A5164: Results Through 48 Weeks

Number of Patients with Death or AIDS Progression By Time of ART Start

• At 48 weeks, no difference in virologic suppression, IRIS (7.6%) or need for ART changes

Zolopa A, et al. 15th CROI; Boston, MA (2008); Abst. 142; Grant P, et al. 16th CROI; Montreal, Canada; February 8-11, 2009. Abst. 775.

20

34

DeferredImmediate

P=0.035

The SAPiT Trial: Starting Antiretroviral Therapy in TB

• Open-Label Randomized Trial; HIV+ pts with active TB (N=642)

• Randomized to one of 3 arms:• Arm 1: ART initiated during intensive phase of TB treatment• Arm 2: ART initiated after intensive phase of TB treatment• Arm 3: ART initiated after TB treatment completed - Sequential Tx

• IRIS: Integrated 12.1% vs. Sequential 3.8%

• Sept 2008: DSMB stopped sequential arm

Integrated Tx

Abdool Karim S, et al. 16th CROI; Montreal, Canada; February 8-11, 2009. Abst. 36a.

Months After Randomization

Su

rviv

alS

urv

ival

1.001.00

0.900.90

0.700.70

0.800.80

0.950.95

0.850.85

0.750.75

00 11 22 33 44 55 66 77 88 99 1010 1111 1212 1313 1414 1515 1616 1717 1818 1919 2020 2121 2222 2323 2424

Post –TB TreatmentContinuation Phase of TB treatment

IntensivePhaseof TBtreatment

Sequential Arm

Integrated Arm

Kaplan-Meier Survival Curve

P=0.003

Cautionary Note: ART and Cryptococcal Meningitis (Zimbabwe)

• Immediate vs. delayed (10 weeks) ART in Cryptococcal Meningitis (N=54)

• Tx: Fluconazole 800 mg daily and d4T/3TC/NVP

• No use of amphotericin nor management of raised intracranial pressure

• Mortality: 87% immediate vs. 37% delayed (P=0.002)

• Most deaths in immediate ART group occurred within the first month, possibly due to IRIS

• Fluconazole-NVP drug interaction postulated

Comparison of Kaplan-Meier Survival Estimatesby Treatment Group

Comparison of Kaplan-Meier Survival Estimatesby Treatment Group

1.00

1.00

0.75

0.75

0.00

0.00

0.25

0.25

00 200200 400400 600600 800800

0.50

0.50

Time to Death (days)Time to Death (days)

P=0.028P=0.028

Delayed

Early

Su

rviv

alS

urv

ival

Makadzange A, et al. 16th CROI; Montreal, Canada; February 8-11, 2009. Abst. 36cLB.

NA-ACCORD: Improved Survival When ART is Started with ≥350 CD4

• North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD)

• Regional collaboration of 22 HIV research cohorts from United States and Canada

• Study of HIV+ patients with: • CD4 count 351-500 cells/mm3 • Active follow-up between 1996 and 2006

• Outcome: All-cause mortality

• Groups compared from same CD4 count level:• Immediate treatment: Initiate ART within 1.5 yrs after 1st CD4 count

between 351-500 cells/mm3

• Deferred treatment: Do not initiate ART in this time frame

• Patient data censored if treatment not initiated within the 1.5 year interval after the target CD4 count for ART initiation

Kitahata M, et al. 48th ICAAC/46th IDSA; Washington, DC; October 25-28. Abst. H-896b

NA-ACCORD: Baseline Characteristics

Initiate HAART (n=2,473)

Defer HAART (n=5,901)

Follow up person-years 8,358 16,636

Males (%) 83 75

Median Age (years) 40 38

White (%) 39 38

Median CD4 count cells/mm3 421 432

Median log10 HIV RNA copies/mL 4.3 4.1

Hepatitis C virus infection (%) 27 34

History of Injection Drug Use (%) 16 21

Kitahata M, et al. 48th ICAAC/46th IDSA; Washington, DC; October 25-28. Abst. H-896b

NA-ACCORD: Results

Relative Hazard (RH)*

95% Confidence

IntervalP-value

Deferral of HAART at 351-500 cells/mm3 1.7 1.4, 2.1 <0.001

Female Sex 1.1 0.9, 1.5 0.290

Older Age (per 10 years) 1.6 1.5, 1.8 <0.001

Baseline CD4 count (per 100 cells/mm3) 0.9 0.7, 1.0 0.083

• HIV RNA was not an independent predictor of mortality• Rate of virologic suppression (<500 c/ml) similar between groups

Kitahata M, et al. 48th ICAAC/46th IDSA; Washington, DC; October 25-28. Abst. H-896b

* Stratified by Cohort and Year

NA-ACCORD: Improved Survival When ART is Started When CD4 Count ≥500 cells/mm3

• ARV-naïve; CD4 count >500 cells/mm3, no prior H/O AIDS-defining illness, follow-up between 1996 and 2006

• All-cause mortality compared between immediate vs. deferred ART• Immediate Group: Start ART with > 500 cells/mm3

• Deferred Group: Start ART within 1.5 years of CD4 <500

• Statistical analysis adjusted for baseline population differences

Kitahata M, et al. 16th CROI, Montreal, Canada, 2009. Abst. 71.

0.00

0.05

0.10

0.15

0.20

0 2 4 6 8 10Years after 1996

Relative Hazard of Deferral 1.6 (1.3,1.9; p<0.001)

Defer HAART > 500 CD4 cells (N=6,539)Initiate HAART > 500 CD4 cells (N= 2,616)

Mo

rtal

ity

SMART Naïve/Off Treatment Sub-study:Clinical Outcomes of (re-)Starting ARVs

Emery S, et al. JID 2008; 197: 1133–1144.

Serious non-AIDS

Hazard Ratio = 7.05 (95% CI: 1.58-31.5) p=0.01

Months

Cum

. Pro

babi

lity

(X10

0)

No. at RiskDeferred ART

Immediate ART

25

20

15

10

5

00 4 8 12 16 20 24 28 32 36

228249

189210

159180

128145

96125

73106

5980

3658

2744

2436

Opportunistic disease (fatal and non-fatal)

Hazard Ratio = 4.40 (95%.CI: 1.23-15.8) p=0.02

MonthsNo. at Risk

25

20

15

10

5

00 4 8 12 16 20 24 28 32 36

228249

192210

162179

130144

95124

73104

5880

3758

2644

2135

Opportunistic disease and death

Deferred ARTImmediate ART

Hazard Ratio = 4.38 (95%.CI: 1.45-13.2) p=0.009

Months

Cum

. Pro

babi

lity

(X10

0)

No. at RiskDeferred ART

Immediate ART

25

20

15

10

5

00 4 8 12 16 20 24 28 32 36

228249

192210

162179

130144

95124

73104

5880

3758

2644

2135

MonthsNo. at Risk

25

20

15

10

5

00 4 8 12 16 20 24 28 32 36

228249

188210

157179

125144

90124

69104

5579

3357

2443

2035

Hazard Ratio = 5.08 (95% CI: 1.91-13.5) p=0.001

Composite endpoint

SMART: Influence of CD4+ Count and Treatment on Clinical Event Rate

* per 100 person years

Last CD4+ cell count (cells/mL)

< 250 250-349 350-499 > 500 Overall

Continuous ART

Event Rate* 10.4 6.7 1.8 0.0 1.3

Intermittent ART

Event Rate* 16.0 9.2 7.6 3.1 7.0

Emery S, et al. JID 2008; 197: 1133–1144.

Antiretroviral Therapy Cohort Collaboration (ART-CC)

• Collaboration of HIV cohort studies to estimate risk of deferring ART at different CD4 Count levels

• ARV-naïve patients (N=24,444) starting ART after 1997 with <550 cells/mm3

• Patients with H/O AIDS or IDU excluded

• Rates of AIDS and death with immediate vs. deferred ART compared in adjacent CD4 ranges of 100 cells/mm3

• Adjusted for lead-time and unseen events in final analysis

Sterne J, et al. 16th CROI, Montreal, Canada, 2009. Abst. 72LB.

ART-CC: When Should ART be Started?

Hazard ratios for AIDS or death, adjusted for lead time/unseen events

• Delaying ART to <350 (but not <375) cells/mm3 is associated with an increased risk of AIDS or death

Comparison Hazard Ratio (95% CI)

276-375 vs 376-475 1.19 (0.96 to 1.47)

251-350 vs 351-450 1.28 (1.04 to 1.57)

226-325 vs 326-425 1.21 (1.01 to 1.46)

Sterne J, et al. 16th CROI, Montreal, Canada, 2009. Abst. 72LB.

44

Haz

ard

Rat

io f

or

AID

S o

r D

eath

Haz

ard

Rat

io f

or

AID

S o

r D

eath

22.5.5

11

500500 400400 300300 200200 100100 00CD4 Threshold (cells/mm3)CD4 Threshold (cells/mm3)

Observational Studies Can Be Wrong

• Numerous epidemiological studies demonstrated women on hormone replacement therapy (HRT) had decrease in risk of coronary heart disease (CHD)

• These studies led HCPs to recommend HRT as protection against CHD

• Controlled trials demonstrated HRT caused small, but significant, increase in CHD risk

• Re-analysis of epidemiological studies showed women on HRT more likely to be from socio-economic groups with better than average diet and exercise

• HRT use and decreased CHD risk were coincident effects of a common cause, rather than cause and effect1

Lawlor DA, et al. Intl J Epidemiol 2004;33:464-467.

Improved CD4 Recovery When Starting with Higher CD4 Count

• CD4-count increases on sustained suppressive (<400 c/mL) ARV treatment (n=655) by baseline count

• >350 cells/mm3: CD4 counts return to near-normal levels

• ≤350 cells/mm3: CD4 counts significantly increased but plateau after 4 years below normal range

• Differences in CD4 counts associated with differences in morbidity and mortality

Median CD4 Counts Over 6 YearsStratified by Baseline CD4 Count

Moore RD, Keruly JC. Clin Infect Dis 2007;44:441-446.

900

800

700

600

500

400

300

200

100

00 1 2 3 4 5 6

Years After Starting HAART

CD

4 C

ou

nt

(cel

ls/m

m3 )

<200 201–350 >350

Lower Risk of Triple-Class Failure when Starting at Higher CD4

• Large collaborative European cohort assessed risk factors for triple-class virologic failure (TCVF):

• All pts starting ART with 2 NRTIs + NNRTI or boosted PI (N=45,977)• 980 developed TCVF

• Risk factors for TCVF:• Non-MSM• Younger age• Higher HIV RNA • Lower CD4

*Adjusted for all factors in the model

Lodwick R, et al. 16th CROI; Montreal, Canada; February 8-11, 2009. Abst. 585.

Adjusted* hazard ratios for TCVF by baseline CD4 and VL after starting ART Adjusted* hazard ratios for TCVF by baseline CD4 and VL after starting ART

0.10.1 11

Lower risk of TCVF Greater risk of TCVFLower risk of TCVF Greater risk of TCVF

Baseline CD4 count(cells/mm3)Baseline CD4 count(cells/mm3)

Baseline viral load(log10 copies/ml)Baseline viral load(log10 copies/ml)

0-4950-199200-349350-499500-Unknown

0-3.94.0-4.44.5-4.95.0-5.45.5-5.96.0-Unknown

0-4950-199200-349350-499500-Unknown

0-3.94.0-4.44.5-4.95.0-5.45.5-5.96.0-Unknown

Treatment as Prevention: Rwanda

• Study to evaluate effect of ART on HIV transmission among HIV serodiscordant couples (N=2,993)

• ART only if clinically indicated

• Negative partner tested q3 months

• Sexual risk assessed by• Self report• Sperm on vaginal smear• Pregnancy• Combined variable using any of above

Sullivan P, et al. 16th CROI, Montreal, Canada, 2009. Abst. 52bLB.

Treatment as Prevention: Results

• Sexual risk behaviors lower in those on ART (19% vs. 25%, P<0.05)

• Both ART and change in behavior independently reduced HIV transmission

ARV Status CY Observation

No. LinkedInfections

InfectionRate (C-Y)

Infection RateRatio (95% CI)

Not on ARV 5,062 171 3.4/100 ---

On ARV 547 4 0.7/100 0.21 (0.08, 0.59)

On ARV –conservative* 547 6 1.0/100 0.32 (0.14, 0.73)

* Includes 2 partners who seroconverted in the same 3-month interval when the HIV-infected partner initiated ARVs

2,993 couples were followed for a median of 512 days HIV-free Survival of HIV-negative partners,by ARV status of HIV+ Partner

0 500 1000 1500 2000 2500

0.0

0.2

0.4

0.6

0.8

1.0

Su

rviv

al P

rob

ab

ilit

y

Days

Off ARV On ARV

CensoredLogrank P<.0001

Sullivan P, et al. 16th CROI, Montreal, Canada, 2009. Abst. 52bLB.

Starting Earlier isCost-Effective

• Markov modeling using Johns Hopkins HIV database

• Beginning ARV treatment at CD4+ >350 cells/mm3 vs. 200-350 cells/mm3 results in an incremental cost per QALY gained of $31,226

• Starting ARV therapy at >350 cells/mm3 is more cost-effective than:

• Coronary bypass• Hemodialysis• Screening mammograms

Mauskopf J, et al. JAIDS. 2005;39:562-569.

HAART Initiation (cells/mm3)

Incremental Lifetime Cost

Cost per Life-year gained

Cost per QALY Gained

>350 vs. 200-350 $19,074 $25,567 $31,226

200-350 vs. <200 $28,066 $22,064 $25,806

QALY=Quality Adjusted Life Year

When to StartWhen to StartDebate/DiscussionDebate/Discussion