Switch from TDF to TAF GSUS292 0109 Study

Switch from TDF to TAF § GS-US-292 -0109 Study

GS-US-292 -0109 Study: Switch TDF to TAF § Design Randomisation 2: 1 Open-label HIV+ ≥ 18 years On E/C/F/TDF or EFV/FTC/TDF or boosted ATV + FTC/TDF > 96 weeks HIV RNA < 50 c/m. L e. GFR (Cockroft-Gault) > 50 m. L/min N = 959 N = 477 W 48 W 96 Switch to E/C/F/TAF Continue TDF-based regimen *Randomisation was stratified by previous treatment regimen § Endpoints – Primary: proportion of patients maintaining HIV RNA < 50 c/m. L at W 48 (ITT, snapshot) ; non-inferiority if lower margin of a two-sided 95% CI for the difference = -12%, 99% power – Secondary: percentage change in hip and spine bone mineral density, change in serum creatinine, and change in EFV-related symptom score at W 48 GS-US-292 -0109 Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Baseline characteristics and patient disposition E/C/F/TAF N = 959 TDF-based regimens N = 477 41 40 11% 10% 68% / 18% / 6% 66% / 21% / 7% HIV-1 RNA < 50 c/m. L 98% CD 4 cell count (/mm 3), median 675 662 105. 7 92. 8 107. 7 93. 9 Dipstick proteinuria: grade 1 / grade 2 9% / < 1% Discontinuation by W 48, N (%) For adverse event For lack of efficacy Investigator discretion Consent withdrawal Loss to follow-up Non-compliance Death 32 (3. 3%) 9 1 2 8 6 2 4 40 (8. 4%) 12 0 3 16 7 2 0 Median age, years Female Race: white / black / asian e. GFR, m. L/min, median Cockroft-Gault CKD-EPI GS-US-292 -0109 Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Virologic outcome at W 48 and W 96 (ITT, snapshot) TDF-based regimens (N = 477) E/C/F/TAF (N = 959) % 100 HIV RNA < 50 c/m. L 97. 2 93. 1 93 Virologic failure No virologic data 89 80 60 40 20 1* 0 Week 48 Week 96 Difference (95% CI) ≠ (95% CI) = 3. 7% (0. 4 to 7. 0) = 4. 1% (1. 6 to 6. 7) Superiority of E/C/F/TAF 1. 3 Week 48 2 2 Week 96 1. 8 5. 7 Week 48 5 9 Week 96 N=10 N=6 * One patient in the TAF group with virologic failure (W 8) had genotypic resistance : M 184 I/M. The patient resuppressed 4 weeks later without a change of regimen GS-US-292 -0109 Mills A. Lancet Infec Dis 2016; 16: 43 -52 ; De. Jesus E. ASM Microbe 2016, Abs. LB-087

GS-US-292 -0109 Study: Switch TDF to TAF HIV-1 RNA < 50 c/m. L at W 48 by prior treatment regimen, % Primary Endpoint E/C/F/TAF p < 0. 001 97 100 93 TDF-based regimens p = 0. 02 96 90 p = 0. 02 97 92 p = NS 98 97 306 153 80 60 40 20 0 959 All prior regimens 95% CI = GS-US-292 -0109 477 1. 6 to 6. 7 251 125 Prior EFV/FTC/TDF 0. 5 to 12. 3 402 199 Prior boosted ATV + FTC/TDF 0. 9 to 9. 2 Prior E/C/F/TDF -1. 9 to 3. 9 Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Differences in patients with HIV-1 RNA < 50 c/m. L at W 96 by subgroup, % (95% CI) TDF-based regimens E/C/F/TAF Overall Age, years Sex Race Adherence < 50 > 50 Male Female Black Nonblack < 95% > 95% -12 GS-US-292 -0109 -6 0 -6 12 De. Jesus E. ASM Microbe 2016, Abs. LB-087

GS-US-292 -0109 Study: Switch TDF to TAF Changes in bone mineral density from baseline to W 48 Hip BMD change (N) T-score mean change from baseline Percent change from baseline 0 to 3% increase in BMD > 3% increase Spine BMD change (N) T-score mean change from baseline Percent change from baseline 0 to 3% increase in BMD > 3% increase GS-US-292 -0109 E/C/F/TAF TDF-based regimens Difference in least square means (95% CI) 869 428 0. 11 - 0. 02 0. 13 (0. 10 – 0. 15) ; p < 0. 0001 1. 47% - 0. 34% 1. 81 (1. 49 – 2. 13) ; p < 0. 0001 56% 21% 38% 8% p < 0. 0001 881 436 0. 17 - 0. 02 0. 19 (0. 16 – 0. 23) ; p < 0. 0001 1. 56% - 0. 44% 2. 00 (1. 55 – 2. 45) ; p < 0. 0001 41% 33% 34% 13% p < 0. 0001 Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Changes in bone mineral density from baseline to W 48 in patients previously on ATV + r + FTC/TDF Hip BMD mean change (N) E/C/F/TAF ATV + r + FTC/TDF 41 13 During 144 weeks of prior ATV + r + FTC/TDF Percent change from baseline to W 48 Spine BMD mean change (N) - 4. 58% + 1. 35% - 0. 77% 42 16 During 144 weeks of prior ATV + r + FTC/TDF Percent change from baseline to W 48 GS-US-292 -0109 - 3. 1% + 2. 83% - 0. 74% Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Median % change in BMD (Q 1, Q 3) on DXA Spine BMD All Patients (N=1369) 4. 5 E/C/F/TAF TDF-based regimens 3. 5 2. 0 1. 5 p < 0, 001 0. 5 -0. 3 -0. 5 -1. 5 -2. 5 Baseline Hip BMD All Patients (N=1354) Week 24 Week 48 Week 96 4. 5 3. 5 2. 1 1. 5 p < 0, 001 0. 5 -0. 6 -1. 5 -2. 5 Baseline § Week 24 Week 48 Week 96 Regardless of prior treatment regimen, differences between arms were statistically significant for spine and hip BMD GS-US-292 -0109 De. Jesus E. ASM Microbes 2016, Abs. LB-087

GS-US-292 -0109 Study: Switch TDF to TAF Change in Osteopenia/Osteoporosis Diagnosis (defined by T-Score) Normal % 100 80 Osteopenia Osteoporosis Spine Hip 5. 8 4. 8 7. 2 7. 6 36 32 35 37 60 100 80 % 0. 7 1. 3 2. 1 31 26 32 32 69 73 67 66 Baseline Week 48 60 59 40 64 57 56 40 20 20 0 0 Baseline Week 48 E/C/F/TAF N = 912 Baseline Week 48 TDF-Based Regimen N = 457 E/C/F/TAF N = 902 TDF-Based Regimen N = 452 Differences between E/C/F/TAF and TDF-based regimens were statistically significant (p < 0. 001) GS-US-292 -0109 Mills A. IAS 2015, Abs. TUAB 0102; Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Serum creatinine and e. GFR changes from baseline to W 48 Mean serum creatinine change, mmol/L Median e. GFR (Cockroft-Gault), m. L/min E/C/F/TAF EFV/FTC/TDF (unboosted regimen) Other TDF-based regimens - 0. 4 - + 2. 9 + 9. 2 + 1. 77 - + 1. 2 - - 3. 7 Median changes in markers of renal function from baseline to W 48 E/C/F/TAF TDF-Based regimens p Fractional excretion of phosphate + 0. 2 + 0. 7 0. 054 Fractional excretion of uric acid - 0. 8 + 0. 3 < 0. 001 Tm. P/GFR *, mg/d. L - 0. 1 0. 63 * Tm. P/GFR: ratio of tubular maximum reabsorption of phosphate (Tm. P) to GFR GS-US-292 -0109 Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Renal Safety Results (Median % change in proteinuria at W 48) E/C/F/TAF TDF-based regimens Tubular Proteinuria 30 UPCR UACR RBP: Cr β-2 -m: Cr 20 18. 1 10 -20 -30 -40 -50 -60 GS-US-292 -0109 8. 5 9. 6 0 -20. 9 18. 7 -17. 9 -33. 4 -52. 3 Each difference between treatment arms was statistically significant (p < 0. 001) UPCR: urine protein: creatinine ratio ; UACR: urine albumin: creatinine ratio ; RBP, retinol-binding protein ; β-2 -m: beta-2 microglobulin Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Adverse events, N (%) (W 48) E/C/F/TAF N = 959 TDF-based regimens N = 477 Study-drug related adverse event 21% 16% Grade 3 or 4 adverse event 9% 11% 65 (7%) 35 (7%) 0 2 (< 1%) 9 (0. 94%) * 2 0 12 (2. 52%) ** 5 (1 Fanconi) 3 (all on boosted ATV) Serious adverse event Study drug-related serious adverse event Premature study drug discontinuation for kidney-related adverse event for jaundice * Panic attack; apathy, amnesia, speech disorder; Reiter’s syndrome; nausea, vomiting, and headache; suicide attempt; leg swelling and impaired concentration; depression; acute renal failure (after cancer chemotherapy: sepsis and multisystem organ failure); tubulo-interstitial nephritis (recurrent hematuria with subsequent diagnosis of Hodgkin lymphoma) ** Abnormal dreams; depression, insomnia, and irritability; depression, insomnia, and nightmares; elevated bilirubin; jaundice (N = 2); memory impairment; chronic kidney disease; elevated serum creatinine (N = 2); Fanconi’s syndrome and mild jaundice; renal colic § No discontinuation for adverse event in both groups between W 48 and W 96 GS-US-292 -0109 Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Most-common drug-related adverse event, % (W 48) GS-US-292 -0109 E/C/F/TAF N = 959 TDF-based regimens N = 477 Upper respiratory tract infection 16 11 Diarrhea 10 9 Nasopharyngitis 9 8 Headache 7 4 Cough 7 5 Syphilis 5 6 Insomnia 5 6 Arthralgia 6 5 Bronchitis 6 5 Depression 4 6 Osteopenia 6 5 Back pain 5 5 Nausea 5 3 Sinusitis 5 5 Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Grade 2 -4 laboratory abnormalities (W 48) E/C/F/TAF N = 959 TDF-based regimens N = 477 25% 31% Creatine kinase 10% AST 5% 7% ALT 5% 5% Neutropenia 4% 3% Hypophosphatemia 2% 3% Hyperuricemia 2% 1% Alkaline phosphatase < 1% Leukopenia < 1% Platelets < 1% Total bilirubin < 1% 24% Hemoglobin 0 < 1% Creatinine 0 < 1% Any abnormality GS-US-292 -0109 Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF Fasting Lipids (mg/d. L, median), baseline and W 48 250 200 5 E/C/F/TAF Baseline Week 48 202 183 181 TDF-Based Regimens Baseline Week 48 4 181 150 125 116 100 131 116 120 114 52 3. 6 114 112 52 50 3 3. 8 50 49 2 1 0 0 Total Cholesterol p < 0. 001 LDL HDL Triglycerides p < 0. 001 p = 0. 003 p < 0. 001 TC: HDL Ratio p = 0. 004 § Participants initiating lipid-modifying medications: E/C/F/TAF: 8% ; TDF-based regimens: 6% GS-US-292 -0109 Mills A. Lancet Infec Dis 2016; 16: 43 -52

GS-US-292 -0109 Study: Switch TDF to TAF § Conclusion – Switching virologically suppressed patients to TAF regimen (EVG 150 mg, cobicistat 150 mg, emtricitabine 200 mg, and TAF 10 mg) was non inferior to the to 3 TDF-containing regimens at week 48 and week 96 – It was also superior to continuing with the TDF-containing regimens. – Switching to the TAF group showed several important safety advantages of TAF over any of the TDF-containing regimens : • Improvements in hip and spine bone mineral density • Improvements in renal function – Decreases in serum creatinine in patients switched from a boosted regimen – Decreases in dipstick proteinuria, in quantitative tests of total urine protein, and total urine albumin, in specific proximal tubular proteins – Improvements in proximal renal tubular function (fractional excretion of uric acid, fractional excretion of phosphate, and renal tubular maximum reabsorption rate of phosphate to the glomerular filtration rate – Slightly higher lipid concentrations with TAF than with TDF-based regimens GS-US-292 -0109 Mills A. Lancet Infec Dis 2016; 16: 43 -52