7th IAS Conference on HIV Pathogenesis, Treatment and Prevention June 30-July 3, 2013; Kuala Lumpur, Malaysia Dolutegravir (DTG) is Superior to Raltegravir (RAL) in ART-Experienced, Integrase- Naive Subjects: Week 48 Results From SAILING (ING111762) Pedro Cahn, 1 Anton Pozniak, 2 Horacio Mingrone, 3 Carlos Brites, 4 Jaime Federico Andrade-Villanueva, 5 Jan Fourie, 6 Moti Ramgopal, 7 Debbie Hagins, 8 Jose Madruga, 9 Tamara Newman, 10 John Lombaard, 11 David Dorey, 12 Mark Underwood, 13 Sandy Griffith, 13 Sherene Min, 13 on behalf of the extended SAILING study team 1 Fundación Huésped, Buenos Aires, Argentina; 2 Chelsea and Westminster Hospital NHS Foundation Trust, London, UK; 3 Fundación IDEAA, Ciudad Autónoma de Buenos Aires, Buenos Aires, Argentina; 4 Complexo Hospitalar Prof. Edgard Santos, Universidade Federal da Bahia, Salvador, Brazil; 5 Hospital Civil de Guadalajara “Fray Antonio Alcalde,” CUCS, Universidad de Guadalajara, Guadalajara, Mexico; 6 Fourie Medical Centre, Dundee, South Africa; 7 Midway Immunology and Research Center, Fort Pierce, FL, USA; 8 Chatham CARE Center, Savannah, GA, USA; 9 Centro de Referencia e Treinamento DST/AIDS, São Paulo, Brazil; 10 Instituto de Infectologia Emílio Ribas, São Paulo, Brazil; 11 JOSHA Research, Bloemfontein, South Africa; GlaxoSmithKline, 12 Mississauga, ON, Canada; 13 Research Triangle Park, NC, USA

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. Study Rationale Dolutegravir has been shown to be effective in antiretroviral (ART)-naive (SPRING-2 and SINGLE) and integrase inhibitor (INI)-resistant subjects (VIKING-3). SAILING was designed to test the efficacy and safety of DTG versus RAL when used with a background regimen of 2 antiretrovirals (1 of which must have been fully active) in ART- experienced, INI-naive subjects with at least 2-class drug resistance.

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. SAILING (ING111762) Study Design Week 48 primary analysis Randomization Week 24 planned interim a At Screening and a second consecutive test >400 c/mL within 4 months prior to Screening (if Screening HIV-1 RNA >1000 c/mL, no additional HIV-1 RNA assessment was needed). PBO, placebo; BR, background regimen comprising at least 1 and no more than 2 active agents. HIV ART-experienced, INI-naive HIV-1 RNA >400 c/mL a 1:1 Randomization stratified by HIV-1 RNA (≤ or >50,000), DRV/r use and # of fully active drugs HIV ART-experienced, INI-naive HIV-1 RNA >400 c/mL a 1:1 Randomization stratified by HIV-1 RNA (≤ or >50,000), DRV/r use and # of fully active drugs DTG 50 mg QD + RAL PBO + BR DTG 50 mg QD + RAL PBO + BR RAL 400 mg BID + DTG PBO + BR RAL 400 mg BID + DTG PBO + BR

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. DTG 50 mg QD (n=354) RAL 400 mg BID (n=361) Age, median (y)4243 Gender, female30%34% Race, white50%49% African American/African heritage41%44% HIV-1 RNA, median (log 10 c/mL) >50,000 c/mL30%29% CD4+ count, median (cells/mm 3 ) <200 cells/mm 3 49%51% HBV/HCV coinfection14%18% Duration prior ART, median (y) ≥3 Class resistance47%51% DRV/r in background regimen DRV/r use without primary PI mutations72 (20%)77 (21%) No DRV/r use or DRV/r use with primary PI mutations 282 (80%)284 (79%) Baseline Characteristics

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. Subject Accountability Patients screened N=1441 Not randomized N=717 Randomized phase Patients randomized N=724 Patients randomized to DTG 50 mg QD N=360 Patients randomized to RAL 400 mg BID N=364 Not treated N=3 Treated (ITT-E) N=357 Patients excluded at site N=3 Treated (ITT-E) N=362 Not treated N=2 Patients excluded at site N=1 Completion status at Week (84%) completed 55 (16%) withdrew 4 adverse event 20 lack of efficacy 9 protocol deviation 5 stopping criteria 5 lost to follow-up 1 investigator discretion 11 withdrew consent Completion status at Week (78%) completed 78 (22%) withdrew 11 adverse event 42 lack of efficacy 6 protocol deviation 3 stopping criteria 10 lost to follow-up 1 investigator discretion 5 withdrew consent mITT-E N=354 mITT-E N=361

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. Mean CD4+ change from Baseline was similar between arms: DTG: cells/mm 3 (n=294); RAL: cells/mm 3 (n=283). 71% 64% BL Week Proportion (%) DTG 50 mg QD was statistically superior to RAL 400 mg BID at Week 48. *Adjusted treatment difference (95% CI): 7.4% (0.7%, 14.2%); P=0.03 *Adjusted difference based on stratified analysis adjusting for Baseline HIV-1 RNA (≤50,000 c/mL vs >50,000 c/mL), DRV/r use without primary PI mutations and Baseline PSS (2 vs <2). Proportion (95% CI) With HIV-1 RNA <50 c/mL (Snapshot, mITT-E)

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. Primary Endpoint: HIV-1 RNA <50 c/mL at Week 48 95% CI for difference Favors RAL Favors DTG -20%020% %

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. DTG RAL no. with response/total no. (%) Overall251 (71)230 (64) Baseline plasma HIV-1 RNA ≤50,000 c/mL >50,000 c/mL 186/249 (75) 65/105 (62) 180/254 (71) 50/107 (47) Baseline CD4+ cell count <200 cells/mm 3 ≥200 cells/mm 3 110/173 (64) 141/181 (78) 106/184 (58) 124/177 (70) DRV/r with no primary PI mutations DRV/r use without primary PI mutations No DRV/r use or DRV/r use with primary PI mutations 50/72 (69) 201/282 (71) 54/77 (70) 176/284 (62) Background regimen PSS = 2* PSS <2 181/250 (72) 70/104 (67) 169/267 (63) 61/94 (65) *PSS based on full susceptibility, reported category “2” includes 2 subjects with PSS = 3. SubgroupResponse rates Difference (DTG-RAL) and 95% CI In favor of RALIn favor of DTG Percentage of Subjects With HIV-1 RNA <50 c/mL by Subgroup

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. DTG 50 mg QD (n=354) RAL 400 mg BID (n=361) Difference (%, 95% CI) Per protocol, HIV-1 RNA <50 c/mL238/325 (73%) 225/340 (66%)7.5% (0.6%, 14.3%) Kaplan-Meier proportion without failure, % (95% CI) Treatment-related discontinuation = failure a 92.0 (88.5, 94.4)85.1 (80.8, 88.4)6.9 (2.1, 11.7) Efficacy-related discontinuation = failure b 93.7 (90.4, 95.8)86.7 (82.6, 89.9)6.9 (2.5, 11.4) a Protocol-defined virologic failure or withdrawal because of drug-related adverse events, safety stopping criteria or lack of efficacy. b Protocol-defined virologic failure or withdrawal due to lack of efficacy. Sensitivity Analyses

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. Non-response HIV-1 RNA <1 log 10 c/mL decrease by Week 16, unless <400 c/mL HIV-1 RNA ≥400 c/mL on or after Week 24, through Week 48 Rebound HIV-1 RNA ≥400 c/mL after confirmed <400 c/mL HIV-1 RNA >1 log 10 c/mL above nadir (lowest prior value ≥400 c/mL) Protocol-Defined Virologic Failure

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. Fewer PDVFs for DTG versus RAL by Week 48 DTG 50 mg QD (N=354) RAL 400 mg BID (N=361) Week 1610 (3%)21 (6%) Virologic non-response013 (4%) Rebound10 (3%) 8 (2%) Week 2415 (4%)35 (10%) Virologic non-response 1 (<1%)19 (5%) Rebound14 (4%)16 (4%) Week 4821 (6%)45 (12%) Virologic non-response 2 (<1%)19 (5%) Rebound19 (5%)26 (7%) Protocol-Defined Virologic Failure

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. Supportive Analysis Treatment-emergent resistance to background regimen was also statistically significant.** DTG 1% vs RAL 3%, adjusted difference (95% CI) of -2.2% (-4.3%, -0.1%) Proportion of Subjects With INI Treatment-Emergent Genotypic/Phenotypic Resistance Treatment Failure with INI-r (n)/ mITT-E population (N) Adjusted difference in proportion (95% CI) (DTG-RAL) DTG 50 mg QD4/354 (1%) RAL 400 mg BID17/361 (5%)-3.7% (-6.1%, -1.2%)* *DTG was superior vs RAL at Week 48 (P=0.003), pre-specified and adjusted for multiple testing. Treatment-Emergent Resistance to INI and to Background Regimen Key Secondary Endpoint **This analysis was pre-specified but was unadjusted for multiple testing.

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. Adverse Events DTG 50 mg QD (N=357) RAL 400 mg BID (N=362) Discontinuations due to safety events 9 (3%)14 (4%) Most commonly reported (≥10%) AEs in either arm Diarrhea71 (20%)64 (18%) Upper respiratory tract infection38 (11%)29 (8%) Drug-related (≥2% in either arm) 73 (20%)85 (23%) Diarrhea29 (8%)21 (6%) Nausea13 (4%)16 (4%) Vomiting8 (2%)11 (3%) Headache7 (2%) Fatigue4 (1%)10 (3%) Rash5 (1%)6 (2%) Insomnia0 6 (2%) Abdominal pain upper6 (2%)0 Drug-related Grade (8%)32 (9%) Drug-related Grade 4 1 (<1%) Serious – any event 33 (9%)42 (12%) Serious drug-related – any event 2 (<1%) a 4 (1%) b Fatal AEs 03 (<1%) c a DTG: 1 hepatotoxicity, 1 myositis and acute renal failure b RAL: 1 oral mucosal blistering and rash pruritic, 1 pancreatitis, 1 hepatitis, 1 suicidal ideation c 1 adenocarcinoma, 1 acute hepatic and renal failure, 1 cervical carcinoma

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. DTG 50 mg QD RAL 400 mg BID Select treatment-emergent Grade 3-4 laboratory abnormalities N=357 n (%) N=362 n (%) Creatine phosphokinase (CPK)7 (2)4 (1) Alanine aminotransferase (ALT)9 (3)7 (2) Lipase4 (1)7 (2) Total bilirubin a 21 (6)14 (4) Creatinine1 (<1) Renal laboratory values Change from Baseline serum creatinine (μmol/L), mean (SD) 11.1 (15.53) b (n=291) 5.1 (12.23) (n=283) Change from Baseline urine albumin/creatinine ratio (mg/mmol), mean (SD) (27.51) (n= 260) (31.81) (n=253) a 16/21 subjects in the DTG arm and 11/14 subjects in the RAL arm were receiving atazanavir or atazanavir/ritonavir. b As previously described, small non-progressive increase in serum creatinine due to OCT2 inhibition Select Laboratory Abnormalities

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. DTG once daily has higher virologic efficacy when compared with RAL twice daily in a treatment-experienced, INI-naive population. 71% on DTG versus 64% on RAL had HIV-1 RNA <50 c/mL at Week 48. DTG 50 mg once daily had a similar tolerability and safety profile to RAL twice daily with a wide variety of background regimens in treatment-experienced subjects. DTG statistical superiority was driven by fewer withdrawals due to lack of efficacy, lower number of protocol defined virologic failures and lower treatment emergent resistance. Conclusions

Cahn et al. IAS 2013; Kuala Lumpur, Malaysia. Abstract WELBB03. We thank everyone who has contributed to the success of this study, including All study participants and their families The SAILING clinical investigators and their staff The GSK and ViiV Healthcare study team This study was funded by ViiV Healthcare Argentina:Chile:Mexico:South Africa:Taiwan:USA (cont): BellosoAfani SaudAndrade-VillanuevaFourieHis-Hsun LinGuptaRobbins BenetucciBeltran BuendiaGranados-ReyesLatiffYu-Hui LinHaginsRodriguez CesarLassoMosqueda-GómezLombaardHung-Chin TsaiHardyRuane CassettiPerezRosas-DossettiSpain:Jen-Hsien WangHenrySaag LupoWolffNetherlands:BarrosWing-Wai WongHicksSamuel MingroneFrance:HollanderBoix MartínezUK:HodderSandkovsky Australia:AjanaRomania:Cano SánchezHayHuhnScarsella BlochDe TruchisDuiculescuCastro IglesiasKeggJeffersonSchneider ElliottKatlamaRuginaClotetKhooM JohnsonSchrader Belgium:MolinaStreinu-Cercelde los Santos GilWilkinsP JohnsonScribner ClumeckPialouxRussia:DeigUSA:KozalShalit FlorencePoizot-MartinChernovaFloresAkilKumarSiraj LegrandQuertainmontKhafizovGalindoAnsteadMcCurdySlim MoutschenTeicherKozyrevGonzález GarcíaBartczakMcDonaldSloan Brazil:YeniKulaginGórgolas Hernández-MoraBlickMeierSmall Andrade NetoGreece:MoshkovichMasiaBrarMildvanTashima BahiaGargalianos-KakolyrisNagimovaRiberaBurackMurphyTebas CaseiroGogosPokrovskyRodríguez BañoCaseyNahassVanig DiazPaparizosProninVera MéndezEron, Jr.NewmanWard GrinsztejnItaly:RakhmanovaVon Wichmann De MiguelFeinbergO’KeefeWarner SchiavonDi PerriRyamovaCanada:FelizartaOsiyemiWheeler PereiraGoriShuldyakovConwayFileParksZingman ReuterLazzarinSoninLogueGatheJacobs TupinambasPigaHungary:SmaillGoulstonRichmond RusconiBánhegyiTremblayGulick Acknowledgments