1. NRTI-sparing  SPARTAN  PROGRESS  NEAT001/ANRS 143  MODERN

2. Raffi F. Lancet 2014;384:1942-51 NEAT 001 / ANRS 143  Design  Objective –Non inferiority of RAL compared to TDF/FTC : % of participants reaching the primary endpoint by week 96 estimated by the Kaplan-Meier method in intention-to-treat analysis (upper margin of the 2-sided 95% CI for the difference = 9%, 85% power) DRV/r 800/100 mg QD + RAL bid DRV/r QD + TDF/FTC QD Randomisation* 1 : 1 Open label > 18 years ARV-naïve HIV RNA > 1,000 c/mL CD4 < 500/mm 3 HBs Ag negative Creatinine clearance > 60 mL/min No resistance mutations * Randomisation was stratified by country and participation in the viral and immunological dynamics and inflammation sub-study N = 404 N = 401 W96 Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC

3.  Primary endpoint : time to failure, as the first occurrence of any of the following components Virological –Change of treatment before W32 because of insufficient virologic response HIV-1 RNA reduction < 1 log 10 c/ml by W18* or HIV-1 RNA ≥ 400 c/ml at W24* –HIV-1 RNA ≥ 50 c/ml at W32* –HIV-1 RNA ≥ 50 c/ml at any time after W32* Clinical –Death due to any cause –Any new or recurrent AIDS defining event** –Any new serious non AIDS defining event**  All patients followed-up until last patient reached W96, events recorded until end of F-U  Major secondary endpoints : safety, changes in CD4 and HIV RNA, genotypic resistance * Confirmed by a subsequent measurement ; ** confirmed by the Endpoint Review Committee NEAT 001 / ANRS 143 Endpoints Raffi F. Lancet 2014;384:1942-51

4. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC DRV/r + RAL N = 401 DRV/r + TDF/FTC N = 404 Median age, years3739 Female12%11% HIV RNA (log 10 c/mL), median4.784.75 HIV RNA > 100,000 c/mL36%32% CD4 cell count (/mm 3 ), median340325 CD4 < 200 per mm 3 15%16% Hepatitis C coinfection4% Discontinuation by W96N = 38N = 22 Discontinuation by end of follow-upN = 53N = 38  Data were censored at the first date of meeting the primary endpoint, completion of 96 weeks of follow-up, last time the primary endpoint status was known, or date of withdrawal Baseline characteristics and patient disposition NEAT 001 / ANRS 143 Raffi F. Lancet 2014;384:1942-51

5. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC NEAT 001 / ANRS 143  If a patient reached more than one component, only the first was taken into account DRV/r + RAL N = 401 DRV/r + TDF/FTC N = 404 Total patients meeting primary endpoint during follow-up77 (19.2%)61 (15.3%) Change of regimen because of insufficient virologic response < 1 log 10 c/mL reduction in HIV RNA at W1810 HIV RNA ≥ 400 c/mL at W2400 HIV RNA ≥ 50 c/mL at W322728 HIV RNA ≥ 50 c/mL after W323322 Death31 AIDS event53 Serious non-AIDS event87 Patients who met the primary endpoint Raffi F. Lancet 2014;384:1942-51

6. NEAT 001 / ANRS 143 Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC Estimated proportion reaching primary endpoint at W96 RAL: 17.8% vs TDF/FTC: 13.8% Adjusted difference: 4% (95% CI: -0.8, 8.8%) Probability of reaching primary endpoint log rank p=0.12 0 0.25 0.50 0.75 1.00 4023953933613503403312159012 4003833753463273153062108911 08183248648096112128144 Time (weeks) Numbers of patients DRV/r + RAL DRV/r + TDF/FTC Raffi F. Lancet 2014;384:1942-51

7. NEAT 001 / ANRS 143 Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC Kaplan-Meier estimates of proportion of patients in each group reaching endpoints at W96 Primary and secondary analyses around the primary endpoint Adjusted difference in proportions of failure at W96 (%, 95% CI) DRV/r + RAL (%) DRV/r + TDF/FTC (%) -3-2012345678910 Adjusted difference in proportions of failure at W96 (%, 95% CI) Primary endpoint4.0 (-0.8 ; 8.8)13.817.8 Without SNAIDS endpoints3.9 (-0.7 ; 8.6)12.116.0 Without treatment change3.7 (-1.0 ; 8.5)14.017.7 > 200 copies per mL in place of > 50 copies per mL5.5 (2.0 ; 9,0)5.010.5 Virological components only3.6 (-0.9 ; 8.2)11.815.5 Clinical components only1.5 (-0.6 ; 3.7)2.43.9 Including treatment discontinuation for virological reason 4.0 (-0.8 ; 8.9)13.617.7 Per-protocol analysis3.2 (-1.3 ; 7.7)12.615.8 Including treatment discontinuation for any reason 3.3 (-1.9 ; 8.5)19.723.0 Raffi F. Lancet 2014;384:1942-51

8. Kaplan-Meier estimates of proportion of patients in each group reaching endpoint at W96 – Analyses by baseline HIV RNA and CD4 cell count NEAT 001 / ANRS 143 Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC Adjusted difference in proportions of failure at W96 (%, 95% CI) DRV/r + RAL (%) DRV/r + TDF/FTC (%) -5010152025303540 Adjusted difference in proportions of failure at W96 (%, 95% CI) Baseline HIV RNA < 100 000 copies/mL0.1 (-3.8 ; 4.0)7.37.4 Baseline HIV RNA > 100 000 copies/mL9.6 (-0.1 ; 20.1)27.336.8 CD4 cell count < 200 cells/mm 3 22.3 (7.4 ; 37.1)20.943.2 CD4 cell count > 200 cells/mm 3 1.4 (-3.5 ; 6,3)12.313.7 5 Raffi F. Lancet 2014;384:1942-51

9. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC Endpoint N%≠ (95% CI) CD4 < 200/mm 3 and HIV RNA < 100,000 c/mL (N = 46) DRV/r + RAL3/239.4% 0.4% (-13.7 ; 14.6) DRV/r + TDF/FTC3/239.0% CD4 ≥ 200/mm 3 and HIV RNA < 100,000 c/mL (N = 484) DRV/r + RAL19/2327.1% 0% (-3.9 ; 3.9) DRV/r + TDF/FTC21/2527.1% CD4 < 200/mm 3 and HIV RNA ≥ 100,000 c/mL (N = 77) DRV/r + RAL23/3760.1% 30.3% (13.8 ; 46.8) DRV/r + TDF/FTC12/4029.9% CD4 ≥ 200/mm 3 and HIV RNA ≥ 100,000 c/mL (N = 198) DRV/r + RAL32/10926.5% -1.9% (-13.9 ; 10.0) DRV/r + TDF/FTC25/8928.4% Kaplan-Meier estimates of proportion of patients reaching primary endpoint at W96 according to combined baseline CD4 and HIV RNA NEAT 001 / ANRS 143 Raffi F. Lancet 2014;384:1942-51

10. HIV RNA < 50 c/mL NEAT 001 / ANRS 143 Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC Data are proportions (95% CI) based on available viral load data 20 40 100 60 % 80 0 0481218243248648096 DRV/r + RAL, 89.4% (95% CI : 85.7-92.4) DRV/r + TDF/FTC, 93.3% (95% CI : 90.3-95.6) Raffi F. Lancet 2014;384:1942-51

11. NEAT 001 / ANRS 143 Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC DRV/r + RAL N = 401 DRV/r + TDF/FTC N = 404 All PDVF6652 Total number of patients who met criteria for genotype testing* 3615 PDVF patients meeting criteria for genotype testing339 Patients without PDVF meeting criteria for genotype testing 36 Patients undergoing genotyping2913 Major resistance mutations Reverse transcriptase Protease Integrase 61†05‡61†05‡ 00000000 Virological failure and emerging resistance mutations per trial arm RAL = raltegravir. DRV/r = ritonavir-boosted darunavir. TDF-FTC = tenofovir-emtricitabine PDVF = protocol-defined virological failure *Genotypic testing carried out by local laboratories when patients had a single HIV RNA > 500 copies/mL at or after week 32 up to the end of follow up † K65R. ‡ N155H Raffi F. Lancet 2014;384:1942-51

12. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC DRV/r + RAL N = 401 DRV/r + TDF/FTC N = 404 Patients meeting criteria for genotypic testing6958 Patients with available genotype (≥ 1 amplified gene)6149 ≥ 1 major IAS resistance mutations Reverse transcriptase (NRTI only) Protease Integrase 17* 3/53 (5.7%) 1**/57 (1.8%) 14***/55 (25.5%) 00000000 Emergence of resistance mutations in full data set (confirmed HIV RNA > 50 c/mL or any single HIV RNA ≥ 500 c/ml at or after W32) * 1 patient had 2 major IAS resistance mutations (1 NRTI + 1 INI) ** L76V *** N155H = 12 ; N155H + Q148R = 1, Y143C = 1 ; HIV RNA at genotypic testing < 200 c/mL : 4/14 Predictor of integrase resistance mutations emergence : baseline HIV RNA, p=0.006, no association with baseline CD4 nor HIV RNA at time of testing Lambert-Niclot S, IHDRW 2015, Abs. 21 NEAT 001 / ANRS 143

13. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC Safety NEAT 001 / ANRS 143 DRV/r + RAL N = 401 DRV/r + TDF/FTC N = 404 Adjusted difference (95% CI) Discontinuation for any reason14.8% (W96)9.1% (W96) 5.8% (2.0 to 10.0) Discontinued after reaching the primary endpoint36/80 (44%)8/51 (16%) Discontinuation for treatment-limiting AE1.5%2.6% AE leading to treatment modification, all grade 3.9 /100 patient-year 4.2/100 patient-year AE leading to treatment modification, grade 3-4N = 7N = 6 Serious adverse events 10.2/100 patient-year 8.3/100 patient-year Death4*1** Grade 2-4 rash 1.3/100 patient-year 1.6/100 patient-year * melanoma, Burkitt’s lymphoma, severe sepsis with organ failure after DRESS, suicide ** morphine overdose Raffi F. Lancet 2014;384:1942-51

14. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC NEAT 001 / ANRS 143 Laboratory parameters DRV/r + RAL N = 401 DRV/r + TDF/FTC N = 404 p Mean increase from baseline to W96 Total cholesterol (mmol/L)+ 0.9+0.5< 0.001 HDL-cholesterol (mmol/L)+0.2+0.1< 0.001 LDL-cholesterol (mmol/L)+0.5+0.40.021 Estimated creatinine clearance (mL/min)+ 0.8- 4.6< 0.001 Graded abnormality over 96 weeks ≥ Grade 2 creatinine00 Grade 3-4 CK increase6%5%- Grade 3-4 ALAT increase5%3%0.036 Conversion mmol/L to g/L : x 0.387 Raffi F. Lancet 2014;384:1942-51

15. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC  Bone sub-study –146 patients : 70 DRV/r + RAL vs 76 DRV/r + TDF/FTC Randomised at same time as the main study Baseline demographic and HIV characteristics similar in the 2 sub-groups History of fractures : 12 % vs 20 % (p = 0.17) Osteopenia/osteoporosis at baseline : 24 in DRV/r + RAL vs 21 in DRV/r + TDF/FTC –Compare changes in bone mineral density (BMD) between treatment arms Whole body dual-energy x-ray absorptiometry (DXA) scans assessed BMD of total hip, lumbar spine and femoral neck, at baseline, W48 and W96 –Evaluate clinical factors associated with BMD loss –Primary endpoint Mean percentage change of BMD in lumbar spine and hip at 48 weeks –Secondary endpoints Mean percentage change at 96 weeks Proportion with WHO criteria for osteoporosis/osteopenia Proportion with a Z score < -2 Incidence of fractures –Analyses by intent-to-treat exposed approach (ITT-e) Bernardino JI, ICAAC 2014, Abs. H-1198 NEAT 001 / ANRS 143

16. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC Mean % Change in femoral neck BMDMean % change in total hip BMD Mean % change in lumbar spine BMD Bernardino JI, ICAAC 2014, Abs. H-1198 NEAT 001 / ANRS 143 DRV/r + RAL DRV/r + TDF/FTC -6 -5 -4 -3 -2 0 1 BaselineW48S96 55 67 49 59 -1.41 -3.45 -5.99 -1.74 p = 0.0084 p = 0.025 n -4 -3 -2 0 1 BaselineW48W96 55 67 49 58 -0.73 -1.57 -3.30 -3.86 p < 0.0001 p = 0.0032 n -2.49 -2.80 -0.43 p = 0.046 p = 0.0054 -3 -2 0 1 BaselineW48W96 51 63 48 57 n

17. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC Bernardino JI, ICAAC 2014, Abs. H-1198 Factors associated with % change in BMD at 96 weeks (multivariate analysis) β95% CIp Lumbar spine Treatment armTDF/FTC + DRV/r vs RAL + DRV/r-2.34-3.96 ; -0.730.005 VL at baselinePer log 10 c/mL higher-1.24-2.49 ; 0.010.054 Femoral neck Treatment armTDF/FTC + DRV/r vs RAL + DRV/r-3.48-7.44 ; 0.480.088 StageB/C vs A-4.01-9.44 ; 1.420.15 Prior history of bone fractureAny vs none-4.81-10.12 ;0.500.079 Sports (hours)Any vs none-5.50-10.65 ; -0.350.039 Total Hip Treatment armTDF/FTC + DRV/r vs RAL + DRV/r-2.32-3.79 ; -0.850.002 aOR of BMD loss > 5% for TDF/FTC vs RAL : 4.78 (95% CI : 1.63 to 14.03) ; p=0.0044 New fractures – 4 at W48 : 3 in TDF/FTC + DRV/r and 1 in RAL + DRV/r – 3 more at W96 : 2 in TDF/FTC + DRV/r and 1 in RAL + DRV/r New cases of osteopenia, osteoporosis or Z-score < - 2 : no difference between arms NEAT 001 / ANRS 143

18. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC  Biomarkers measured on paired samples D0, W48 –Bone formation: osteocalcin, bone specific alkaline phosphatase (BSAP), pro-collagen type 1 N pro-peptide (P1NP), osteoprotegerin –Bone resorption: serum C terminal collagen crosslinks (CTX-1), receptor activator of NF  (RANKL), osteopontin, urine CTX-1/creatinine ratio –25-OH vitamin D, intact PTH –Inflammatory markers : IL-1 , IL-6, TNF-   Compared to DRV/r + RAL, patients treated with DRV/r + TDF/FTC had significancly greater increases in bone turnover markers that were associated with a greater decrease in BMD at W96 especially at femoral neck and total hip  Independently of ART regimen used –Baseline P1NP and osteopontin levels were associated with greater BMD changes –Increases > 5% in serum iPTH and urine CTX-1/creatinine ratio at W48 were associated with a BMD loss ≥ 5 % at W96 Bernardino JI, CROI 2015, Abs. 766 NEAT 001 / ANRS 143 Bone biomarkers substudy

19. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC Bernardino JI, CROI 2015, Abs. 766 -100 Median percentage changes in biomarkers at W48 600 400 200 0 p=0.0082 N = 38N = 47 0 200 400 600 800 p=0.0008 N = 38N = 46 1 000 800 600 400 200 0 p=0.048 N = 38N = 46 6 000 4 000 2 000 0 p=0.27 N = 38N = 46 0 200 400 600 p=0.033 N = 38N = 45 600 400 200 0 p=0.013 N = 37N = 46 200 100 0 p=0.38 N = 12N = 13 0 200 400 600 800 p=0.90 N = 38N = 45 Osteocalcin RANK-L CTX-1 25-OH D3 P1NP Osteopontin 250 200 150 100 50 0 p<0.0001 N = 38N = 45 BSAP PTH Osteoprotegerin DRV/r + RAL DRV/r + TVD NEAT 001 / ANRS 143

20. Study NEAT 001/ANRS 143: DRV/r + RAL vs DRV/r + TDF/FTC  Conclusion –Overall, the NtRTI-sparing strategy of DRV/r + RAL, requiring twice daily medication intake, was well tolerated and had comparable efficacy to the once daily standard regimen of DRV/r + TDF/FTC at 96 weeks. DRV/r + RAL was non inferior to DRV/r + TDF/FTC for the composite primary endpoint based on clinical and virological failure Sensitivity analyses around the primary endpoint, as well as secondary analyses, including per-protocol analysis, supported non-inferiority –However, prespecified subgroup analyses showed that the NtRTI-sparing strategy was less efficacious than the standard regimen in patients with CD4 cell count < 200/mm 3 at treatment start. –Despite a low rate of virological failure in both arms, emergence of resistance mutations was higher in the raltegravir group. –The final results of this head-to-head phase III strategic study suggest that the dual therapy DRV/r + RAL represents an alternative option to the standard combination of DRV/r + TDF/FTC for first line therapy in patients with CD4 cell > 200/mm 3 NEAT 001 / ANRS 143 Raffi F. Lancet 2014;384:1942-51