Comparing PI vs. PI in HIV Treatment Studies

Comparison of PI vs PI ATV vs ATV/r LPV/r mono vs LPV/r + ZDV/3TC LPV/r QD vs BID BMS 089 MONARK M02-418 M05-730 A5073 GARDEL ALERT ATADAR KLEAN GEMINI CASTLE ARTEMIS LPV/r + 3TC vs LPV/r + 2 NRTI ATV/r vs FPV/r ATV/r vs DRV/r FPV/r vs LPV/r SQV/r vs LPV/r ATV/r vs LPV/r DRV/r vs LPV/r

Study BMS 089: ATV vs ATV/r QD, in combination with 3TC + d4T XR QD Design Randomisation* 1 : 1 Open-label W48 ATV/r 300/100 mg QD 200 adults N = 95 ARV-na ve or < 30 days of prior NRTI or < 7 days of prior NNRTI or PI HIV RNA > 2,000 c/mL Any CD4 cell count 3TC + d4T XR QD ATV 400 mg BID N = 105 3TC + d4T XR QD *Randomisation was stratified on HIV RNA < or > 100,000 c/mL Objective Non inferiority of ATV/r 300/100 vs ATV 400 at W48: % HIV RNA < 400 c/mL, ITT, TLOVR algorithm (lower margin of the 95% CI for the difference = -10%, 80% power) Malan DR. JAIDS 2008;47:161-7 BMS 089

Study BMS 089: ATV vs ATV/r QD, in combination with 3TC + d4T XR QD Baseline characteristics and patient disposition ATV/r ATV 400 mg 300/100 mg Randomized, N 95 105 Treated eligible patients, N 95 104 Median age, years 35 34 Female 27% 30% White/Black/Other 53% / 26% / 21% 57% / 26% / 17% HIV RNA (log10c/mL), median HIV RNA > 100,000 c/mL 4.8 5.1 42% 52% CD4 cell count (/mm3), median 201 194 CD4 < 200/mm3 48% 52% Hepatitis B and/or C positive 18% 19% Discontinuation before W48 12% 10% Malan DR. JAIDS 2008;47:161-7 BMS 089

Study BMS 089: ATV vs ATV/r QD, in combination with 3TC + d4T XR QD Response to treatment at week 48 (ITT, TLOVR) ATV/r (N = 95) Primary endpoint ATV (N = 105) % 100 HIV RNA < 400 c/mL HIV RNA < 50 c/mL HIV RNA < 50 c/mL 87 86 85 82 80 75 70 63 58 60 40 20 N = 46 50 49 55 0 All patients 95% CI for the difference = - 7; 17 Baseline HIV RNA < 100,000 c/mL Baseline HIV RNA > 100,000 c/mL 95% CI for the difference = - 8.2; 11.1 Median CD4 increase at W48: 174/mm3(ATV/r) vs 213/mm3(ATV) Malan DR. JAIDS 2008;47:161-7 BMS 089

Study BMS 089: ATV vs ATV/r QD, in combination with 3TC + d4T XR QD Treatment-emergent resistance Genotypic and phenotypic resistance in patients with virologic failure through W48 ATV/r N = 3 N = 2 ATV N = 10 N = 8 HIV RNA > 400 c/mL Resistance testing PI major substitutions 0 1 # I50L, N88S I50I/L 0 2 PI minor substitutions * 1 4 ATV phenotype: FC > 2.2 0 1 (FC = 26) # M184V 1 7 # Same patient; * Among 20I/T, 33F, 34Q, 36V, 64V, 71V, 73S, 74A, 83D Malan DR. JAIDS 2008;47:161-7 BMS 089

Study BMS 089: ATV vs ATV/r QD, in combination with 3TC + d4T XR QD Safety ATV/r ATV Serious adverse event 15% 16% Discontinuation for adverse event 8% < 1% Grade 2 to 4 treatment-related adverse event 43% 34% Jaundice 3% < 1% Grade 3-4 laboratory abnormality Total bilirubin > 2.5 x ULN 59% 20% ALT > 5 x ULN 6% 3% AST > 5 x ULN 3% 3% Total cholesterol > 300 mg/dL 1% < 1% Triglycerides > 751 mg/dL 2% < 1% Upward shift of > 1 NCEP category at W48 for ATV/r vs ATV: total cholesterol: 16% vs 11% LDL-cholesterol: 46% vs 48% triglycerides: 30% vs 18% Malan DR. JAIDS 2008;47:161-7 BMS 089

Study BMS 089: ATV vs ATV/r QD, in combination with 3TC + d4T XR QD Summary - Conclusion For first-line antiretroviral therapy, ATV/r 300/100 mg QD was virologically non inferior to ATV 400 mg QD, when combined with 3TC and d4T XR QD Response rate at an HIV RNA < 50 c/mL at W48 was 25% less in patients with baseline HIV RNA > 100,000 c/mL as compared with baseline HIV RNA < 100,000 c/mL Virologic failure occurred more frequently in the ATV group (9.5% vs 3.2%) with emergence of major atazanavir-associated mutations in 3/8 tested patients in the ATV group Results (HIV RNA < 50 c/mL at W48, virologic failure rate, resistance data) suggest ATV/r is more potent than ATV Discontinuation due to treatment-related adverse event occurred more frequently in the ATV/r group Increases in total cholesterol and triglycerides were greater in the ATV/r group Possible impact of stavudine on lipid increases Limitation: small size of the study Malan DR. JAIDS 2008;47:161-7 BMS 089