1. Comparison of NNRTI vs PI/r  EFV vs LPV/r vs EFV + LPV/r –A5142 –Mexican Study  NVP vs ATV/r –ARTEN  EFV vs ATV/r –A5202

2. ACTG A5202 Study: ABC/3TC vs TDF/FTC  Multicenter, randomized, blinded equivalence study in 1858 HIV-1 infected patients  Comparison of antiviral activity, safety and tolerability of ABC/3TC and TDF/FTC, given with EFV or ATV/r  Scheduled interim review by the DSMB of the NIAID: inferior virologic efficacy of ABC/3TC in patients with a screening HIV RNA > 100,000 c/mL  Report of data from the 797 patients with screening HIV RNA > 100,000 c/mL  Inclusion criteria: HIV-1 infection, > 16 years, < 7 days of prior antiretroviral therapy, acceptable laboratory values  Design: randomized, partially blinded study comparing 4 once-daily regimens for the initial treatment of HIV-1 infection: –EFV 600 mg or ATV/r 300/100 mg, in combination with ABC/3TC or TDF/FTC (double-blinding for the NRTIs) –Randomisation was stratified on screening HIV RNA (> or < 100,000 c/mL) –Planned study duration was 96 weeks after enrolment of the last patient –Genotypic resistance test was required in patients with recent HIV-1 acquisition –Testing for HLA-B*5701 was permitted but not required Sax PE. NEJM 2009;361:2230-40 A5202

3. ACTG A5202 Study: ABC/3TC vs TDF/FTC  Primary efficacy endpoint: time to virologic failure (confirmed HIV RNA > 1,000 c/mL at or after W16 and before W24, or > 200 c/mL at or after W24)  Primary hypotheses: –Equivalence of ABC/3TC and TDF/FTC (for each regimens with ATV/r and EFV) –Equivalence of ATV/r and EFV (for each NRTI regimen) –Equivalence if the two-sided 95% CI for the hazard ratio was between 0.71 and 1.40 (power of 89.8%) –Pre specified early-stopping rules for inferiority at annual efficacy review by DSMB  Analyses of efficacy by ITT, stratified according to the screening HIV RNA  Kaplan-Meier estimation of time-to-event, with comparison by two-sided log-rank tests. Hazard ratios estimated by Cox models  Primary safety endpoint: time to the first grade 3 or 4 sign, symptom, or laboratory abnormality at least 1 grade higher than at baseline (except isolated unconjugated bilirubin and creatine kinase) while on randomly assigned treatment Sax PE. NEJM 2009;361:2230-40 Statistical analysis A5202

4. ABC/3TC N = 398 TDF/FTC N = 399 Total N = 797 Median age, years384039 Female17%14%15% White/Black/Other43% / 28% / 30%51% / 24% / 25%47% / 26% / 27% History of AIDS26%22%24% HIV RNA, log 10 c/mL, median (IQR) *5.0 (4.7–5.6) HIV RNA > 100,000 c/mL HIV RNA 50,000–99,999 c/mL 49% 30% 50% 27% 50% 28% CD4 count/mm 3, median (IQR) **138 (36–282)146 (45–294)145 (41–285) CD4 < 200/mm 3 CD4 < 50/mm 3 61% 31% 58% 28% 59% 30% HBsAg+ or HCV Ab+ 9%8% Genotype tested at screening44%42%43% Baseline characteristics of the patients with screening HIV RNA > 100,000 c/mL * Geometric mean of screening and entry visits; ** mean of screening and entry visits ACTG A5202 Study: ABC/3TC vs TDF/FTC (screening HIV RNA > 100,000 c/mL) Sax PE. NEJM 2009;361:2230-40 A5202

5. –Median follow-up = 60 weeks –Discontinuation: 10% = 41 patients on ABC/3TC and 38 on TDF/FTC –Risk of subsequent virologic failure among 448 patients with > 2 consecutive HIV RNA < 50 c/mL = 12 in ABC/3TC group vs 9 in TDF/FTC group (p = 0.25) –Median CD4/mm 3 increase at W48: 194 (ABC/3TC) vs 199 (TDF/FTC) Early and late virologic failure according to the protocol-defined criteria by treatment group ABC/3TC N = 57 TDF/FTC N = 26 HIV RNA > 1,000 c/mL at W16 to < W24 without previous level < 200 c/mL199 HIV RNA > 200 c/mL at or after W24 without previous level < 200 c/mL92 HIV RNA > 200 c/mL at or after W24 with previous level < 200 c/mL2915 Weeks since randomisation p < 0.001, log-rank test Hazard ratio : 2.33 (95% CI : 1.46-3.72) 041624364860728496108 0 20 40 60 80 100 TDF/FTC (26 events) ABC/3TC (57 events) 398363313267222188137874920 3993613212842362041601046523 ABC/3TC TDF/FTC No. at risk Probability of No Virologic Failure (%) Time to virologic failure ACTG A5202 Study: ABC/3TC vs TDF/FTC (screening HIV RNA > 100,000 c/mL) Sax PE. NEJM 2009;361:2230-40 A5202

6. Weeks since randomisation p < 0.001, log-rank test Hazard ratio: 1.87 (95% CI: 1.38-2.54) 041624364860728496108 0 20 40 60 80 100 TDF/FTC (68 events) ABC/3TC (114 events) 396341290247206173124784619 3973553162812352041581016321 ABC/3TC TDF/FTC No. at risk Probability of No Regimen Failure (%) Time to regimen failure* Weeks since treatment dispensation p < 0.001, log-rank test Hazard ratio: 1.89 (95% CI: 1.43-2.50) 041624364860728496108 0 20 40 60 80 100 TDF/FTC (78 events) ABC/3TC (130 events) 3972582191771481188249275 397299272233188156112713512 ABC/3TC TDF/FTC No. at risk Probability of No Primary Safety Event (%) Time to safety endpoint Sax PE. NEJM 2009;361:2230-40 A5202 * Regimen failure: virologic failure or NRTI modification (time to first event) ACTG A5202 Study: ABC/3TC vs TDF/FTC (screening HIV RNA > 100,000 c/mL)

7. 041624364860728496108 0 20 40 60 80 100 Weeks since randomisation p = 0.20, chi-square test at week 48 TDF/FTC ABC/3TC 38835732429324521216311459 39335232528524421116910969 ABC/3TC TDF/FTC No. with RNA value Sax PE. NEJM 2009;361:2230-40 A5202 % of patients with HIV-1 RNA < 50 c/mL * * ITT analysis involving all patients, regardless of prior NRTI discontinuation or virologic failure This analysis represents the aggregate success of both initial (randomly assigned) and subsequent therapy ACTG A5202 Study: ABC/3TC vs TDF/FTC (screening HIV RNA > 100,000 c/mL)

8. SubgroupEvents per 100 person-yearHazard Ratio (95% CI) p for interaction Overall57/398 (14.82)26/399 (6.34)2.33 (1.46–3.72)< 0.001 Sex0.04 Male50/331 (15.41)18/345 (5.24)3.00 (1.74–5.17) Female7/67 (11.63)8/54 (11.97)0.85 (0.30–2.39) Age0.07 30 yr3.24 (1.73–6.08) 40 yr2.08 (1.28–3.39) Race or ethnic group0.55 White18/170 (10.33)8/202 (3.63)2.82 (1.22–6.53) Black26/112 (26.71)12/94 (13.93)1.94 (0.96–3.90) Hispanic9/103 (8.87)6/93 (6.59)1.35 (0.48–3.83) HIV-1 RNA0.20 5.5 log 10 c/mL2.64 (1.58–4.40) 6.0 log 10 c/mL3.39 (1.60–7.22) CD4 count0.007 50 cells/mm 3 3.54 (1.97–6.36) 200 cells/mm 3 1.68 (0.98–2.88) Genotype tested at screening0.02 Yes22/175 (14.05)3/166 (1.99)7.21 (2.15–24.2) No35/223 (15.35)23/233 (8.88)1.71 (1.00–2.91) 0.04125 ABC/3TC betterTDF/FTC better Estimated effect of ABC/3TC (N=398) vs TDF/FTC (N=399) on the hazard of virologic failure A5202 Sax PE. NEJM 2009;361:2230-40 ACTG A5202 Study: ABC/3TC vs TDF/FTC (screening HIV RNA > 100,000 c/mL)

9. ABC/3TC N = 397 TDF/FTC N = 397 Any lab. abnormality or clinical adverse event, N (%)130 (33)78 (20) Metabolic abnormality, N (%)41 (10)11 (3) Triglycerides elevation, N153 Total cholesterol elevation, N202 LDL-cholesterol elevation, N134 ALT, N75 AST, N124 Diarrhoea or loose stool, N77 Nausea or vomiting, N33 General signs or symptoms, N (%)58 (15)38 (10) Pain or discomfort, N2414 Rash, N88 Pruritus, N92 Fever, N108 Asthenia, N610 Headache, N86 Grade 3 or 4 signs, symptoms or laboratory abnormalities at least 1 grade higher than the grade at baseline, during the initial regimen A5202 Sax PE. NEJM 2009;361:2230-40 ACTG A5202 Study: ABC/3TC vs TDF/FTC (screening HIV RNA > 100,000 c/mL)

10. ABC/3TCTDF/FTCp Median change in fasting lipids at W48 (mg/dL) Total cholesterol + 34+ 26< 0.001 HDL-cholesterol+ 9+ 70.05 Triglycerides+ 25+ 30.001 Ratio of total cholesterol to HDL-cholesterol- 0,2 0.5 Suspected study drug-related HSR, N (%)27 *, ** (7%)27 ** (7%) Renal failure, N22 Median change in calculated creatinine clearance at W48 (mL/min) + 4+ 20.10 Myocardial infarction, N00 Bone fractures, N710 AIDS events, N (%)26 (7%)17 (4%) HIV-related cancers, N84 * 1 death after restarting ABC-containing study medication ** subsequent viriologic failure among patients with suspected HSR = 4 in ABC/3TC group, 3 in TDF/FTC group Selected clinical and laboratory events A5202 Sax PE. NEJM 2009;361:2230-40 ACTG A5202 Study: ABC/3TC vs TDF/FTC (screening HIV RNA > 100,000 c/mL)

11.  Summary - Conclusion –For initial treatment of HIV-1 infection, patients with a screening HIV RNA > 100,000 c/mL whose regimen contains TDF/FTC as compared with ABC/3TC were significantly less likely to experience virologic failure: TDF/FTC superiority in virologic outcome was observed throughout the duration of the study and in multiple sensitivity analyses to experience tolerability failure –Possible explanation: ABC/3TC is less potent than TDF/FTC –Difference in virologic failure between NRTIs significantly increased with a lower CD4 count –Differences in virologic failure persisted after adjustement for multiple baseline covariates –Occurrence of suspected hypersensitivity reactions did not influence study outcomes: equal number in both groups, virologic failure infrequent –Important implications for clinical practice of this double-blind, randomized, prospective study Patients with high HIV RNA have a risk of virologic failure twice higher with ABC/3TC as compared with TDF/FTC Treatment guidelines recommend to consider results of this study when selecting NRTIs for first-line antiretroviral therapy in patients with high HIV RNA A5202 Sax PE. NEJM 2009;361:2230-40 ACTG A5202 Study: ABC/3TC vs TDF/FTC (screening HIV RNA > 100,000 c/mL)

12. ABC/3TCTDF/FTC Total n = 1,857 EFV n = 465 ATV/r n = 463 EFV n = 464 ATV/r n = 465 Median age, years373839 38 Female21%16 %15%17% History of AIDS19%18 %15% 17% HIV RNA (log 10 c/ml), median4.74.64.7 CD4 cell count (/mm 3 ), median225236234224230 HCV Ab+6%7%9%7% Genotype at screening43%48%47%40%45% Patients characteristics at screenning ACTG A5202 Study: ABC/3TC vs TDF/FTC, in combination with EFV vs ATV/r - Final results (all patients)  Results in the 797 patients with screening HIV RNA > 100,000 c/mL : –at DSMB action, time to virologic failure was significantly shorter with ABC/3TC as compared with TDF/FTC, independently of 3 rd drug [HR (95% CI)]: 2.33 (1.46-3.72) (Sax PE, NEJM 2009;361:2230-40) with EFV: 2.46 (1.20-5.05) with ATV/r: 2.22 (1.19-4.14) A5202 Sax PE, NEJM 2009;361:2230-40) ; Daar ES. CROI 2010; Abs. 59LB

13. HR (95% CI) Probability of absence of virologic failure at W96 Screening HIV RNA < 100,000 c/ml (n = 1,060) ABC/3TC vs TDF/FTC (+ ATV/r)1.26 (0.76-2.05)88.3% vs 90.3% ABC/3TC vs TDF/FTC (+ EFV)1.23 (0.77-1.96)87.4% vs 89.2% Overall, all patients (n = 1,857) ATV/r vs EFV (+ ABC/3TC)1.13 (0.82-1.56)83.4% vs 85.3% ATV/r vs EFV (+ TDF/FTC)1.01 (0.70-1.46)89% vs 89.8% ACTG A5202 Study: ABC/3TC vs TDF/FTC, in combination with EFV vs ATV/r - Final results (all patients)  In patients with screening HIV RNA < 100,000 c/ml, ABC/3TC and TDF/FTC have similar time to virologic failure, with ATV/r and EFV  Overall, ATV/r and EFV have similar time to virologic failure, with both NRTIs A5202 Time to virologic failure Sax PE, NEJM 2009;361:2230-40) ; Daar ES. CROI 2010; Abs. 59LB

14. Time to primary safety endpoint (Grade 3-4 event) Time to tolerability endpoint (modification of originally randomized regimen) HR (95% CI)p p Screening HIV RNA < 100,000 c/ml (n = 1,060) ABC/3TC vs TDF/FTC + ATV/r 1.13 (0.83 - 1.54) 0.44 1.43 (1.06 - 1.92) 0.018 ABC/3TC vs TDF/FTC + EFV 1.38 (1.03 - 1.85) 0.03 1.48 (1.12 - 1.95) 0.005 Overall, all patients (n = 1,857) ATV/r vs EFV + ABC/3TC 0.81 (0.66 - 1.00) 0.05 0.69 (0.55 - 0.86) 0.0008 ATV/r vs EFV + TDF/FTC 0.91 (0.72 - 1.15) 0.44 0.84 (0.66 - 1.07) 0.17 ACTG A5202 Study: ABC/3TC vs TDF/FTC, in combination with EFV vs ATV/r - Final results (all patients) A5202 Daar ES. CROI 2010; Abs. 59LB

15. ACTG A5202 Study: ABC/3TC vs TDF/FTC: final results (screening HIV RNA < 100,000 c/ml )  In patients with screening HIV RNA < 100,000 c/ml: ABC/3TC compared with TDF/FTC –Similar time to virologic failure with ATV/r and EFV –Shorter time to safety event with EFV –Shorter time to modification with ATV/r and EFV ( difference driven by suspected hypersensitivity reactions in ABC/3TC arms) –Greater increase in CD4 with EFV –Greater increase in total cholesterol, LDL- and HDL-cholesterol with both ATV/r and EFV; greater increase in triglycerides with ATV/r –Increase (ABC/3TC) versus modest decline (TDF/FTC) in creatinine clearance with ATV/r Daar ES, CROI 2010, Abs. 59LB A5202 Daar ES. CROI 2010; Abs. 59LB

16. ACTG A5202 Study: ABC/3TC vs TDF/FTC, in combination with EFV vs ATV/r - Final results (all patients)  ATV/r compared with EFV (all patients) –Similar time to virologic failure with both NRTIs Pre-specified equivalence boundary on HR was not met, as observed W96 event rate was lower than projected (~15% vs 32%) Difference and CIs for probability of being failure free at W96 were within + 10% criteria often used for defining equivalence (post hoc analysis) –Longer time to safety event and to 3 rd drug modification with ABC/3TC –Among virologic failures there was less resistance with both NRTIs –Greater increase in CD4 with TDF/FTC –Smaller increases in total cholesterol, LDL- and HDL-cholesterol with both NRTIs –Modest decline in creatinine clearance with TDF/FTC vs increase with ABC/3TC Daar ES, CROI 2010, Abs. 59LB A5202 Daar ES. CROI 2010; Abs. 59LB