Innovative Pipeline Sir Andrew Witty 3 November 2015 Cautionary - PowerPoint PPT Presentation

GSK R&D: what is important to us Innovative Patient Quality science need GSK achieved highest number of FDA • • Average of 35 publications annually in 5 Breakthrough Designations since 2013 14 approvals, 2010-15 worlds-class journals (Nature, Cell, • 3 FDA Priority Reviews since 2010 GlaxoSmithKline Science) Company 1 • Focus on preventative and curative • In 2014 and 2015 to date, GSK Company 2 medicines scientists listed as co-authors in more Company 3 • Company 4 Strong focus on patient input than 1,600 publications • All first cycle approvals since 2012 • • 80% of pre-clinical to Phase II assets Quality of life study endpoints • 10% faster in time to file approval than have a novel mechanism of action industry average • Target sciences initiative with • Clinical study cycle times 20% faster than EBI/Sanger & Altius Institute in Seattle average • Cost per patient visit 30% less than 2008 • Molecule quality focus Partnership Collaborations with academia, biotechs, pharmaceutical companies and regulators 11

Recruiting and developing the best scientists We’re committed to ensuring GSK remains the best place to develop medicines World-class External talent Scientific Expert leaders sourcing career advisory pathways networks Academia Industry Biotech 12

HIV / Infectious Diseases

Infectious disease burden continues to grow and present public health challenges Acute HIV HBV HCV complicated infectious diseases • • • • 36.9m living with HIV Globally, 240m people Globally, 130-150m Globally, ~3.5m annual worldwide; 1.2m deaths & have Chronic Hepatitis B 1 people have Chronic deaths due to lower 2m new infections Hepatitis C 1 respiratory tract infections 2 • More than 780k people die annually 1 • • each year 350-500k people die each Increasing antimicrobial • Resistance, adherence year drug resistance (MDR) • HBV evades immune and addressing long-term • • system, with limited Need for a cure completed Hospitalised infections & toxicities remain areas of options for durable in a single visit complications have direct significant unmet medical remission costs >$35bn annually in need US 3 • The ultimate goal is • Pathophysiology & tissue remission and cure damage suggest aberrant host immune responses as key driver 1 WHO 2015; 2 WHO 2014; 3 J Med Econ 2013 14

Infectious Diseases strategy: from innovative treatment regimens to the pursuit of cure HIV Innovative Long-acting Treatment Treatment GSK Pipeline Regimens HCV/HBV Prevention Remission & Cure Acute complicated infectious disease 15

Dolutegravir set to be at the heart of future treatment regimens Dolutegravir profile DTG/3TC: Planned launch H1 2019 Efficacy • Rapid and sustained viral load drop 2-drug STR for HIV treatment in naïve and suppressed patients, QD Simplification - Potential benefit on tolerability and drug burden No food requirements Barrier to Resistance • No resistance mutations selected in DTG/RPV: Planned launch H1 2018 first line failures (one patient had 2-drug STR for HIV treatment in suppressed patients, QD E157Q/P mutation without decreased Simplification - Potential benefit on tolerability and drug burden susceptibility to dolutegravir) ( ViiV Healthcare - Janssen sponsored) • Limited resistance mutation evolution in experienced patients on failure Triumeq ™ (abacavir/dolutegravir/lamivudine): Launched 2014 • Distinct resistance profile compared to other INIs (RAL, EVG) 3-drug STR for HIV treatment, QD Only currently available DTG containing Single Tablet Regimen (STR) Favorable PK Profile • Booster free Tivicay ™ (dolutegravir): Launched 2013 • No food requirement for adequate exposure For HIV treatment in combination with other ART, QD Well tolerated Approved Investigational ViiV Healthcare is a specialist joint venture solely dedicated to HIV, owned by GSK, Pfizer and Shionogi 16

PADDLE (Pilot Antiretroviral Design with Dolutegravir and LamivudinE): Investigator sponsored study design • Investigator sponsored study Patient Base line viral # load Week 24 Week 8 < 50 < 50 1 10.909 • 2 tablet treatment < 50 < 50 2 10.233 < 50 < 50 3 151.569 < 50 < 50 4 148.370 • ARV naive patients < 50 < 50 5 20.544 6 14.499 < 50 < 50 • 2 cohort study 7 18.597 < 50 < 50 8 24.368 < 50 < 50 9 10.832 < 50 < 50 • Open label single arm 10 7.978 < 50 < 50 11 273.676 < 50 < 50 12 64.103 < 50 < 50 Phase IV, pilot, open-label, single arm exploratory trial 13 33.829 < 50 < 50 14 15.151 < 50 < 50 1 st cohort 2 nd cohort 15 23.500 < 50 < 50 16 3.910 < 50 < 50 (n= 10) (n= 10) 17 25.828 < 50 < 50 18 73.069 < 50 < 50 < 50 < 50 19 106.320 DTG 50 mg QD DTG 50 mg QD < 50 < 50 20 7.368 LMV 300 mg QD LMV 300 mg QD From week 8 onwards all patients VL was undetectable (pVL < 50 copies/mL) Adapted from Cahn et al , EACS 2015, LBPS4/1 17 17

Cabotegravir: Long-acting antiretroviral Long-acting HIV Treatment HIV Prevention Mean concentration/time profile following single injection 1 LATTE Week 96 Results 2 Pre-clinical data 100mg IM 100mg SC 200mg IM 200mg SC 100 400mg IM 400mg SC (split) 8/8 protected 800mg IM (split) cabotegravir LA 80 Plasma CAB (g/mL) 1 untreated 4*PA-IC90 Aviremic (%) 60 100 100% Protection in 80 PA-IC90 Patients (%) NHP Rectal Challenge 3 40 60 0.1 40 Cabotegravir 10 mg (n=60) 8/8 infected Cabotegravir 30 mg (n=60) 20 20 Cabotegravir 60 mg (n=61) Efavirenz 600 mg (n=62) p<0.0001 0 0 2 4 8 12 16 24 28 32 36 40 48 60 72 84 96 0 Visit (week) 0 4 8 12 16 20 24 28 32 36 40 44 48 Proportion of patients with HIV-1 RNA concentration of less than 50 copies per mL 0 2 4 6 8 10 12 14 16 18 20 by visit in the intention to-treat exposed population Time (weeks) Weeks post first challenge Error bars indicate 95% CI 2 Margolis et al , Lancet Inf Dis 2015;15(10):1145-1155 1 Spreen et al , JAIDS 2014;67(5):481-486 3 Andrews et al , Science 2014;343(6175):1151-4 18

Cabotegravir long-acting clinical studies Potential for better adherence 4Q2015 LATTE 2 results HIV TREATMENT Key Phase III-enabling data: combination CAB LA + RPV LA CAB LA + RPV LA as maintenance therapy (ViiV Healthcare - Janssen sponsored) Planned launch: Mid-2016 HIV Treatment Phase III start 2019/2020 CAB LA + RPV LA switch studies (transition from oral therapy to long-acting) Mid-2016 PrEP Phase III start (men) HIV PREVENTION CAB LA monotherapy vs. TDF/FTC (Truvada) in at-risk men who have sex with men/transgender women CAB LA monotherapy (Collaboration with third party being considered) Planned launch: End-2016 PrEP Phase III start (women) 2020+ CAB LA monotherapy vs comparator in at-risk women (Collaboration with third party being considered) 19

LATTE 2 - cabotegravir LA + rilpivirine LA for treatment of HIV Headline data – path to Phase III • Phase IIb trial examining long-acting (LA) cabotegravir (CAB) in combination with LA rilpivirine (RPV). 309 treatment naïve subjects initially treated with QD oral CAB 30mg + 2 NRTIs • Following virologic suppression 286 subjects qualified for entry into maintenance phase and were randomised 2:2:1 onto: 4 week injections with CAB LA + RPV LA (Q4W); 8 week injections with CAB LA + RPV LA (Q8W) or continuation of oral CAB + NRTIs • Through 32 weeks on 2-drug maintenance therapy with CAB LA and RPV LA, 95% (Q8W) and 94% (Q4W) of subjects were virologic successes (VL<50) compared to 91% of subjects continuing three drug oral CAB + NRTIs • Adverse events (AEs) leading to withdrawal were 5% (n=6) for Q4W, 2% (n=2) for Q8W, and 2% (n=1) for oral CAB + NRTIs. The most common AE was injection site pain (93% of injection recipients) • Detailed analyses just starting 20

Next wave cabotegravir long-acting combinations Opportunities with broadly neutralising antibodies • Cabotegravir long-acting • Broadly neutralising antibodies (bnAbs) • Every 2 or 3 months • GSK and the National Institute of Allergy and Infectious Diseases/National Institutes of Health collaboration to be announced later this week Nano-formulation Potential targets for neutralisation V1V2 Glycan: N332 Glycan supersite: PG9, PG16 PGT121, PGT128 PGT141-145 + 10-1074 CAP256-VRC26.25 PGDM1400 CD4 Binding site: VRC01, PG04, CH31, 3BNC117, Trimer (gp120/41): 12A12, VRC13, 8ANC195 VRC01-LS, PGT151 VRC07-523-LS, 35022 Z258-N6 gp41 MPER: 2FS, 4E10 10e8 Viral membrane Huang et al . Nature 2014;515(7525):138-42 A pilot clinical combination study of VRC01 and cabotegravir is planned for 2016 start 21

GSK & Regulus combination offers potential for a single administration treatment for HCV Single dose RG-101 • RG101 lowers viral load Prolonged VL declines in HCV patients Log10 reduction of HCV • GSK2878175 lowers viral viral load decline RG-101 2 mg/kg -1 n=14 load RG-101 4 mg/kg n=14 placebo n=4 -3 • Both molecules have potential for prolonged -5 0 1 2 3 4 5 6 7 8 PK/PD activity Weeks EASL 2015. Van Der Ree et al. J. Hepatology . 2015;62: S261 Single dose GSK175-LA • Prolonged pan-genotype Potent anti-HCV activity Prolonged PK in animals and anti-HCV activity (2 oral doses) GSK175-LA log 10 HCV RNA Pbo Relative Log10 0 concentration GSK744-LA reduction • Potential single -1 30 mg administration option -2 60 mg -3 • Clinical combination study -4 0 2 4 6 8 10 12 14 starts 2016 0 5 10 15 Days Weeks AASLD 2014. Gardner et al. Hepatology .2014;60:1164A) GSK: data on file 22

GSK & Isis collaboration targeting next generation of HBV medicines: functional cure • Antisense approach taken Reduction of HBV antigen by anti-HBV ASO in mice to knock down immune suppressive antigens Serum HBV Surface Antigen Serum HBV e Antigen Serum HBsAg (% before treatment) Serum HBeAg (% before treatment) • Entered collaboration with 100 saline 100 Isis Pharmaceuticals in saline 2010 10 – GSK contributed target, 10 mg/kg Isis provided platform 1 10 10 mg/kg & discovery 100 mg/kg 0.1 • Lead compound 100 mg/kg GSK3228836 1 0.01 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 – Phase II start planned Week Week 2016 GSK, data on file. Note: GSK3228836 subject to exercise of option by GSK 23

Infectious Diseases strategy: from innovative treatment regimens to the pursuit of cure HIV Innovative Long-acting Treatment Treatment GSK Pipeline Regimens HCV/HBV Prevention Remission & Cure Acute complicated infectious disease 24

First in a new class of antibacterials: gepotidacin (GSK2140944) – a topoisomerase inhibitor • Novel mechanism with bactericidal 100 activity against MDR pathogens Percent survival 80 All gepotidacin treated survived • (all organs culture negative) Promising safety & efficacy profiles 60 in Phase II studies 40 All untreated died • Effective against key resistant Only novel class antibacterial currently in clinical development 20 to demonstrate activity in FDA accepted model of plague strains: 0 – MDR MRSA, MDR E. coli & Drug 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 Days resistant N. gonorrhoeae 10 day treatment • Potential to address multiple Quinolone conventional & bio-threat indications Gepotidacin ‘class’ DNA • Progressed via successful partnerships with BARDA & DTRA Enzyme Planned Filing: 2019 for resistant infections. Gepotidacin binds to a novel site on gryase Discussions with FDA on Nature Struc Mol Biol, 2010,17;465 plague indication. MDR: multi-drug resistant; DTRA: Defense Threat Reduction Agency (US DoD); BARDA: Biomedical Advanced 25 Research & Development Authority (US HHS)

Infectious Diseases strategy: from innovative regimens to treatment and the pursuit of cure − dolutegravir based regimens − cabotegravir LA − cabotegravir LA + rilpivirine LA − Next generation agents and combinations Innovative Long-acting Treatment Treatment GSK Pipeline Regimens − HCV − HBV HIV, HCV, HBV HIV, HCV, HBV Prevention Remission & Cure HIV HIV, HCV, HBV (Qura) − gepotidacin − tafenoquine − i.v. danirixin 26

Respiratory

Respiratory diseases: still significant unmet need Asthma COPD Lung Fibrosis & Acute Lung Injury • Globally 242m people have • 329m people worldwide have • Each affects ~5m patients asthma (32% increase since 1990) COPD worldwide • Gold-standard options delivered • 3rd leading cause of death by • Idiopathic Pulmonary Fibrosis for mild/moderate asthma 2030 (IPF): median survival of just 2-5 • Major unmet medical need in years, 2 IPF products approved • Longitudinal studies (e.g. • Urgent need to improve symptoms severe asthma ECLIPSE) helping to identify − 5-10% of asthma patients prognostic biomarkers (e.g. and delay disease progression fibrinogen) − 60% of cost burden • Acute Lung Injury (ALI): hospital • Targeting underlying drivers of • Immune modulation offers mortality rates of up to 50% disease progression is key potential for better disease control • Need to identify better clinical path and even remission for drug development 28

Asthma R&D strategy: from secondary prevention to primary disease modification eosinophil Targeted Biologicals Extended Duration Biologicals GSK Pipeline neutrophil Once Daily Inhaled Remission-inducing dendritic cell 29

Nucala ™* (mepolizumab) demonstrates significant reduction in exacerbations Nucala (subcutaneous anti-IL-5 mAb): mepolizumab (s.c. or i.v.) reduced the number of asthma exacerbations in patients with severe eosinophilic • Straightforward patient selection & asthma biomarker 250 placebo (n=191) • 53% reduction in exacerbations Cumulative exacerbations • 61% reduction in ER visits/ hospitalisations 200 • Improvement in health status by 7 points (SGRQ) 150 mepolizumab 75mg, • Significant reduction in daily oral intravenously (n=191) corticosteroid dose while maintaining 100 control seen in trials mepolizumab 100mg, • Dosing every 4 weeks, no weight 50 subcutaneously (n=194) adjustment required • Well tolerated 0 0 4 8 12 16 20 24 28 32 Indication: Severe refractory eosinophilic asthma Positive CHMP: 24 Sep 2015 Week PDUFA : 4 Nov 2015 Adapted from MENSA study, Ortega et al . NEJM 2014; 371:1198-207 30 *The name Nucala is not approved for use by the FDA or EMA.

Nucala will be first in class with a strong profile XOLAIR dupilumab reslizumab benralizumab lebrikizumab tralokinumab Sanofi/ Nucala Novartis/ AstraZeneca Teva AstraZeneca Roche Regeneron Genentech Submitted Launched Submitted Ph III ongoing Ph III ongoing Ph III ongoing Ph III ongoing Phase Q4 2015/ Earliest launch Q4 2015 Launched 2017 2017 2019 2019 Q1 2016 assumption* Anti-IL-4R α Anti-IL-5 Anti-IgE Anti-IL-5 Anti-IL-5R Anti-IL-13 Anti-IL-13 Mechanism SC SC IV SC SC SC SC Delivery mechanism    Efficacy data Ph III Phase III ongoing    Safety data Ph III *Based on published filing date plus average review times 31

Nucala* has potential in other indications Anticipated file timelines 2014 2016 2017 2018 2019 2020+ to 2015 – Evidence Generation Asthma – Evidence Generation COPD Eosinophilic granulomatosis with polyangiitis (EGPA) Hyper-eosinophilic syndrome (HES) Nasal polyposis / Chronic rhinosinusitis Atopic dermatitis 32 *The name Nucala is not approved for use by the FDA or EMA and may not be approved for additional indications.

Two novel biologicals Targeted approaches for uncontrolled asthma patients sirukumab* (IL-6 mAb): Non-Th2 asthma TSLP dAb: Inhaled biologic • • Targets severe disease ineligible for Th2/eosinophilic directed mAbs Thymic Stromal Lymphopoietin (TSLP): key cytokine in (40% of severe asthma patients) epithelial immune response in asthma • • IL-6: key inflammatory driver and genetic association of this Inhaled domain antibody (dAb) directly targets site of action pathway in asthma and reduces systemic exposure to improve risk:benefit profile • • Expected to improve symptoms and exacerbations Clinical proof of concept demonstrated for anti-TSLP approach • • Phase II study start in 2016 Phase I start in 2016 Elevated IL-6 associated with eosinophilic-neutrophilic Target engagement after inhaled delivery of dAb: exemplar inflammation and decreased pulmonary function (FEV 1 ) Inhaled TNFR1 dAb reduced endotoxin (LPS) induced in asthma patients inflammation in healthy volunteers 10 Lung cells 10 4 /mL 500 FEV 1 (% predicted) Lung cytokines 400 120 IL-6 FEV 1 8 400 * 100 pg/mL 300 (pg/mL) 6 300 80 * * 200 60 * 4 # 200 * 40 100 2 100 20 * 0 0 0 0 Normal Eosinophilic Neutrophilic Eosinophilic + Normal Eosinophilic Neutrophilic Eosinophilic + IL-6 IL-8 MIP 1a Lung cell count bronchitis bronchitis neutrophilic cell count bronchitis bronchitis neutrophilic (n=29) (n=30) (n=11) bronchitis (n=29) (n=30) (n=11) bronchitis neutrophils (n=6) (n=6) *p < 0.05 vs neutrophilic bronchitis and eosinophilic bronchitis groups * p<0.05 t=test #p < 0.05 vs eosinophilic bronchitis group placebo inhaled TNFR1 dAb (26mg) n=18 subjects per group Chu, Allergy Asthma & Clinical Immunology.2015;11:14 Data on file (study TFR116236) 33 * sirukumab is part of a GSK Janssen Biologics (Ireland) collaboration

Nucala is at forefront of a diverse asthma biologic pipeline Nucala sirukumab Long acting Anti-TSLP dAb Anti-IL-5/13 Anti-IL-5 Anti-IL-6 Anti-IL-5 (NBE) Bispecific dAb-mAb Modality mAb mAb Extended pharmacology mAb Inhaled dAb in Ellipta extended pharmacology Delivery SC SC SC Inhaled SC mechanism Expected file 2014 2021-25 2021-25 2021-25 2021-25 Status Filed Phase II start 2016 Phase I/II start 2017 Phase I start 2016 Preclinical Moderate-severe Asthma Severe without elevated Moderate-severe Severe eosinophilic Moderate-severe eosinophilic eosinophilic and eosinophils eosinophilic segment neutrophilic Key cytokine in epithelial Additive efficacy of two Reason to Clinical data and strong IL-6 is key driver of non- Extended pharmacology allows immune response; complimentary mechanism rationale eosinophilic inflammation six monthly dosing Inhaled - directly targets mechanisms, in six believe site of action monthly dosing 34

GSK2245035 intranasal TLR7 agonist Demonstrates prolonged suppression of allergic response • Activates immune pathways that suppress Weekly dosing with intranasal GSK2245035 for 8 weeks in allergic rhinitis patients exaggerated Th2 response in asthma Mean IP -10 and 95% CI (pg/mL) • Allergen-independent immune modulation 10000 Week 8 engagement Increase in IP-10 • Clinical data demonstrate target engagement 20 ng Target levels 24 hours after 1000 (IP-10) with no tachyphylaxis last dose of 8 weekly treatments Pbo • Protection from nasal allergen challenge up to 100 3 weeks after last dose Pre-dose 24hrs • Weekly treatment may induce remission from 1 week post treatment 3 weeks post treatment Total nasal 9 asthma 9 allergen challenge Mean TNSS and 95% CI symptom score Protection from (TNSS) reduced 7 7 • Phase II asthma study 2016 after 8 weekly 5 5 treatments and maintained 3 weeks 3 3 after last dose n=6 n=6 n=13 Status: Phase IIa n=13 pbo pbo 20 ng 20 ng Indication: Asthma remission Probability of any reduction 0.92 0.84 Planned Filing: 2021-2025 GSK, data on file (study TL7116958) 35

Asthma R&D strategy: From secondary prevention to primary disease modification Targeted Biologicals Extended Duration Biologicals • Nucala • sirukumab GSK Pipeline • Long-acting anti-IL-5 • Anti-TSLP dAb • Anti-IL-5/13 • Anti-IL-5/13 Once Daily Inhaled Remission-inducing • Relvar/Breo Ellipta ™ • TLR7 agonist • Arnuity Ellipta ™ • Incruse Ellipta™ • Anti-TSLP dAb 36

COPD R&D strategy: Targeting the fundamental drivers of disease eosinophil Targeted Biologicals Infection Driven Exacerbations GSK Pipeline neutrophil Once Daily Inhaled Preserve Lung Function epithelial cell 37

Closed Triple: once daily triple therapy in established Ellipta inhaler • Collaboration with Theravance Consistent improvement in lung function with UMEC plus • Open triple filed with FDA ICS/LABA vs. ICS/LABA • Phase IIIa lung function study fully recruited (FULFIL) UMEC add on vs. ICS/LABA (Study 201314, ITT pop n=236) • EU Closed Triple filing: end 2016 (lung function) • US Closed Triple filing: H1 2018 (exacerbations) UMEC add on vs. Advair (Study 116136, ITT pop n=409) • Triple therapy already part of some clinical practice 1 1 Global Initiative for Chronic Obstructive Lung Disease (GOLD), 2015 UMEC add on vs. Advair (Study 116135, ITT pop n=404) Eosinophil signature being evaluated prospectively in IMPACT study UMEC add on vs. Breo (Study 200109, ITT pop n=412) FF/UMEC/VI (n=4000) UMEC add on vs. Breo (Study 200110, ITT pop n=412) Current COPD FF/VI Follow-up meds (n=4000) 0 20 40 60 80 100 120 140 160 180 200 UMEC/VI Difference in trough FEV 1 24 hours after last dose (mL, 95% CI) (n=2000) Improvement 38

GSK2269557, inhaled PI3K δ inhibitor targets neutrophil- mediated lung damage in COPD • PI3K δ over-activation causes human rare disease Activating mutations in PI3K δ in APDS drive lung infections activated PI3K δ syndrome (APDS) • APDS patients display severe recurrent COPD-like bacterial infections • Inhaled delivery offers potential efficacy/safety advantage and opportunity for combination therapy Angulo et al . Science 2013; 342: 866 • Target engagement demonstrated in healthy Directionality of neutrophil migration is aberrant in COPD smokers (PIP3) patients and corrected by PI3K δ inhibition - in vitro • GSK2269557 on top of standard of care in COPD shows decreased markers of inflammation • Currently testing in exacerbating COPD patients and Phase IIb studies to start 2016/17 Status: Phase IIa Indication: COPD exacerbation Planned Filing: 2021-2025 Healthy control COPD Sapey et al . AJRCCM 2011;183:1176 39

Danirixin (GSK1325756): an oral CXCR2 antagonist Demonstrates potential to reduce lung damage in COPD • Blocks chemokine receptor on neutrophils and Real-time data demonstrate improvement of symptoms with other cell types (CXCR2) danirixin in symptomatic COPD (frequent exacerbators) • Target engagement demonstrated with danirixin 25 Total Symptom Score (EXACT-RS) (neutrophil activation biomarker, CD11b) • Competitor compounds produced clinical 20 effects, but with reduction in blood neutrophils 1 15 • In the clinic, danirixin has efficacy at a dose not associated with reduced blood neutrophils 10 • COPD Phase IIb start 2016 5 • Influenza infection Phase IIa study ongoing 0 1 Am J Respir Crit Care Med 2015;191:1001 – 1011 0 20 40 60 80 100 120 140 160 180 Day Status: Phase IIa placebo (n=35 in study) danirixin 75mg (n=35 in study) Indication: Symptomatic COPD Planned Filing: 2021-2025 GSK, data on file (study 200163) 40

COPD R&D strategy: pipeline Targeting the fundamental drivers of disease Targeted Biologicals Infection Driven Exacerbations • • PI3K δ Nucala • danirixin GSK Pipeline Once Daily Inhaled Preserve Lung Function • Anoro ™ Ellipta • PI3K δ • • Relvar/Breo Ellipta danirixin • Incruse Ellipta • Closed Triple (Ellipta device) • GSK961081 +FF • PI3K δ 41

Drivers of our long-term leadership in asthma and COPD • Excellence in inhaler / delivery technologies neutrophil eosinophil • Targeted biological know-how • Deep understanding of novel respiratory targets • Understanding of patient phenotypes • Expertise in trial design and delivery epithelial cell dendritic cell 42

Respiratory R&D beyond Asthma and COPD Taking our respiratory know-how into new diseases Platform for clinical development of IPF (GSK3008348)  v β 6 PET ligand binds to fibrotic regions in IPF lungs • αvβ6 expression in IPF lung biopsies predicts mortality • Small molecule inhaled α vb6 inhibitor (deposition of Tc - labelled salbutamol in lungs of IPF patients supports inhaled approach) • Displacement of α vb6 PET ligand allows dose ranging in patients Healthy volunteer IPF  v β 6 PET ligand FDIH An inhaled dAb platform for acute lung injury (GSK2862277) • High sTNFR1 levels associated with high mortality • dAb blocks TNFR1 signalling without impacting beneficial TNFR2 signalling • Inhaled TNFR1 dAb reduced endotoxin (LPS) induced inflammation in healthy volunteers • Now in Phase II study mAb (grey, blue, red) vs. dAb (green) 43

PHI and Oxygen Sensing

Daprodustat 1 (GSK1278863) low dose PHI for treatment of anaemia of CKD: New Phase IIb data • Standard of care (rhEPO) limited by increased CV risk and IV/SQ administration daprodustat Phase IIb 3 : • PHI oral tablet to replace injectable rhEPO: low dose, Pre-dialysis subjects naive to rhEPO; target Hgb 10.0-11.5 g/dL (n=96) convenient titration, potential for improved CV safety 12.0 Pre-treatment phase Treatment phase 11.5 Phase II summary (IIa and new IIb) Hgb (g/dL) C 11.0 C • Phase IIa data recently published 2 C C G G G G C C 10.5 G • Raises Hgb in dialysis and non-dialysis subjects, G C 10.0 C C G G either naïve to or switching from rhEPO G Low dose (most subjects  10mg); Simple titration • 9.5 daprodustat rhEPO regimen 0.0 • Durable effect (up to 6 months in Phase IIb) -4 Day 0 4 8 12 16 20 24 • Minimal elevation in EPO levels; No BP increase 1 BL Visit (weeks) • Safety profile consistent with CKD • Phase III start 2016 Status: Phase II Indication: Treatment of anaemia of CKD in subjects 1 USAN, INN approval pending on dialysis and not yet on dialysis 2 J Am Soc Nephrol Oct 22, 2015 (epub) Planned Filing: 2019 Japan, 2021 US/ROW 3 GSK, data on file (Study PHI113737) 45

Daprodustat: success factors for development Key success factors • Low dose Large experience in CKD subjects 659 (up to 6 months) • No inhibition of collagen- Yes (rhEPO) Active comparator for CV safety assessment 4-hydroxylase • Single Phase III CV  10mg QD in most Low dose outcomes studies for subjects non-dialysis and dialysis Flexible dose regimen: Non-Dialysis QD Dialysis QD / TIW Phase III designed for clear assessment of CV risk Single CV outcome trials for ND and HD Inhibition of collagen-4-hydroxylase (cardiac tox risk) No Concern for hepatotoxicity (e.g. exclusion of No acetaminophen in phase III trials) 46

Daprodustat Indication expansion to maximise value of HIF-activating mechanism Muscle injury from repetitive arm Diabetic Foot Ulcer motion in healthy volunteers – Preclinical data demonstrate benefit of HIF induction in diabetic skin – Topical daprodustat formulation in ongoing Phase Ib study – No systemic exposure and no Hgb elevation – Efficacy data on wound healing in 2016 Muscle Injury – daprodustat reduces total CPK release Novel muscle repair activity discovered in pre-clinical injury model daprodustat reduces CPK release over 72 hours at 48 hours – Phase I: Reduction in muscle injury in healthy volunteers Total CK release (IU/L) 16000 14000 12000 Future potential expansion into other anaemia indications 10000 8000 – Myelodysplastic Syndrome (MDS) 6000 4000 2000 – Peri-surgical anaemia (ortho, GI, CV) n=15 n=15 0 daprodustat placebo PLACEBO GSK1278863 50 50mg MG GSK, data on file (Study PHI20084) 47

Introducing our experts GSK’s leading scientists in infectious disease, respiratory medicine and CV John Pottage Zhi Hong Senior Vice President, Senior Vice President, Chief Scientific and Medical Head Infectious Diseases Officer for ViiV Healthcare TAU Steve Pascoe Edith Hessel Dave Allen Vice President, Vice President, Head Senior Vice President, Head Unit Physician Refractory Respiratory Head Respiratory TAU Respiratory Inflammation DPU Ruchira Glaser John Lepore Clinical Development Director, Senior Vice President, Metabolic Pathways and Head Metabolic Pathways Cardiovascular and Cardiovascular 48

Q&A

Moncef Slaoui Chairman of Vaccines

R&D programmes to deliver near-term growth with significant future opportunities and novel immunisation platforms 3 A new vaccine concept − COPD 2 Longer term R&D Focus − RSV − GBS Near/mid term key R&D focus: 1 − Shingrix TM − Meningitis − Lifecycle management 51 RSV=Respiratory Syncytial Virus; GBS=Group B Streptococcus; COPD=Chronic Obstructive Pulmonary Disease

Shingrix TM Shingrix ™ is not approved for use by the FDA or EMA

Existing zoster vaccine One dose, live attenuated vaccine Efficacy: 51% against shingles in ages 60+ – Inverse correlation between age at vaccination and protection – Limited persistence of protection Indication for ages 50+ US ACIP recommendation for ages 60+ Contraindicated in immunocompromised individuals Estimated to have <25% coverage in US* 2014 reported sales of $868m (>$600m in US) *CDC: MMWR, February 2015; http://www.cdc.gov/Mmwr/preview/mmwrhtml/mm6404a6.htm; Zostavax™ US PI Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Zostavax data based on US PI. 53

Shingrix candidate vaccine developed to differentiate Two doses, sub-unit (non-live) vaccine, novel adjuvant Efficacy: 91% - 97% against shingles – High efficacy across identified age groups – Persistence over time Targeting indication and recommendation in ages 50+ Data on immunocompromised individuals in 2017 Expect US, EU, Japan filings in 2H 2016 Expected to contribute ~1/3 of 2020 sales growth targets for GSK vaccines Lal et al . N Engl J Med 2015; ZOE-50 and ZOE-70 pooled analysis – unpublished data Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Shingrix data based on ph III clinical trials. 54

Shingrix - Efficacy against shingles 100 90 Efficacy against shingles - % 80 70 60 50 40 30 20 10 0 70+ yrs 50-59 yrs 60-69 yrs Lal et al . N Engl J Med 2015; ZOE-50 and ZOE-70 pooled analysis – unpublished data Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Shingrix data based on ph III clinical trials. 55

Existing vaccine - Efficacy against shingles 100 90 Efficacy against shingles - % 80 70 60 50 40 30 20 10 0 50-59 yrs 60-69 yrs 70+ yrs Oxman et al . N Engl J Med 2005; 352: 2271 – 84; Schmader et al . Clinical Infectious Diseases 2012;54(7):922 – 8; Zostavax ™ US PI 56 Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Zostavax data based on US PI.

Shingrix - Immune response across age segments Frequency of gE-specific CD4+ T cells by age groups Number of CD4+ T cells per million 3500 3000 2500 2000 1500 1000 500 0 ≥ 80 yrs ≥ 80 50-59 yrs 60-69 yrs 70-79 yrs 50-59 60-69 70-79 saline gE/saline gE/AS01B Chlibek et al . J Infect Dis 2013; 208:1953-61 Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Shingrix data based on clinical trials. 57

Existing vaccine - Immune response across age segments vaccine placebo Responder cell frequency value 12 9 6 3 0 N =V/P 204/240 186/188 162/149 103/94 31/31 60-64 yrs 65-69 yrs 70-74 yrs 75-79 yrs >79 yrs Levin et al . J Infect Dis 2008; 197:825 – 35 Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Zostavax data based on published data. 58

Shingrix - Efficacy against PHN PHN: post herpetic neuralgia, a severe complication of zoster 100 90 80 Efficacy against PHN - % 70 60 50 40 30 20 10 0 Over 50 yrs Over 70 yrs Over 60 yrs ZOE-50 and ZOE-70 pooled analysis – unpublished data Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Shingrix data based on ph III clinical trials. 59

Existing vaccine - Efficacy against PHN PHN: post herpetic neuralgia, a severe complication of zoster 100 90 80 Efficacy against PHN - % 70 60 50 40 30 20 10 0 50-59 yrs 60-69 yrs 70+ yrs Zostavax US PI; Oxman et al . N Engl J Med 2005 Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Zostavax data based on published data. 60

Shingrix - Duration of protection against shingles 100 Efficacy against shingles - % 80 60 40 20 0 Yr1 Yr2 Yr3 Yr4 Yr5 Yr6 ZOE-50 statistical report – unpublished data Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Shingrix data based on ph III clinical trials. 61

Existing vaccine - Duration of protection against shingles 100 Efficacy against shingles - % 80 60 40 20 0 Yr1 Yr2 Yr3 Yr4 Yr5 Yr6 Morrison et al . Clin Infect Dis 2015; 60: 900-909 Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Zostavax data based on published data. 62

Immune response persistency is a good predictor of duration of efficacy Shingrix immune response 10000 gE-specific CD4+ T-cells / million cells (geometric mean frequency (GMF) 1000 100 10 0 1 2 3 4 5 6 Years of follow up Chlibek et al . Vaccine 2015 doi:10.1016/j.vaccine.2015.09.073 63 Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Shingrix data based on clinical trials.

Shingrix: a potentially significant advance in vaccination to prevent shingles High overall vaccine efficacy across 2016 2017 2018 2019 identified age groups, including oldest persons Persistence of vaccine efficacy up to 4 years H2: US, EU, Japan across all ages Filings Six-year persistence of immune response, Planned/ongoing studies: modeled to persist above baseline for at least 15 years (based on 6 year data) Vaccine co-administration Revaccination of Zostavax* population Clinically acceptable reactogenicity Comparative tolerability AS01 adjuvant = new platform for elderly vaccines Phase III efficacy study Annual capacity of ~25-30m doses by 2020 in immuno-compromised *Zostavax is a trademark of Merck & Co 64 Not based on head to head data; Shingrix ™ and Zostavax ™ have not been tested head to head. Shingrix data based on clinical trials.

Meningococcal Meningitis

Meningococcal disease: evolving and unpredictable epidemiology – requires combination vaccine Disease incidence by age (2012) ~139 million annual global birth cohort 1 1 US 3 ~4m US, ~5m EU, ~130m ROW 0.8 0.6 Changes in serogroup distribution in US over time 2 Number of cases per 100,000 0.4 0.2 0 5-14 15-24 25-39 40-64 ≥65 <5 1989- 12 2005 2013 UK 4 1991 10 8 6 4 B C Y W, A & Other 2 0 5-14 15-24 25-44 45-64 ≥65 <5 Sources: 1) CDC http://goo.gl/RykLaI; Eurostat http://goo.gl/3wp9pp; UNICEF http://goo.gl/DD8pXp 2) Jackson & Wenger MMWR 1993 http://goo.gl/fsbbBz; Active Bacterial Core Surveillance: http://goo.gl/riji5X 3) CDC 66 http://goo.gl/PPtAEj; US Census Bureau https://goo.gl/liNpPU 4) UK.gov https://goo.gl/NxThrj Office for National Statistics http://goo.gl/GJLRpX

Most advanced meningitis vaccines portfolio, including candidate pentavalent Menveo TM Bexsero TM MenABCWY – Candidate pentavalent – MenACWY tetravalent vaccine – MenB vaccine combination vaccine for – Approved in US and EU (2010) – Approved in US in 2015 adolescent in US (adolescents) and EU (2 months – ACIP recommendation for – Most advanced in development old and above) adolescents – Phase III start in 2017 – ACIP category B (permissive) – Approved in 64 countries recommendation – US filing expected in 2020 – 2015 sales (Mar – Sept): £135m – Approved in 38 countries – 2015 sales (Mar – Sept): £78m Meningitis portfolio expected to contribute ~1/3 of 2020 sales growth targets for GSK vaccines 67

Bexsero: multi-component antigen composition adds value, differentiation Bexsero – 4 antigens composition – 2 dose regimen % subjects with bactericidal activity 100 80 60 40 20 0 Strain 1 Strain 2 Strain 3 Strain 4* Baseline 1m Post Dose 1 1m Post Dose 2 68 Sources: Santolaya et al . Hum Vac & Imm 2013 http://goo.gl/8oWB4P; * Strain 4 GSK data on file. Post hoc assays on a subset

Competing vaccine for MenB Competing vaccine – 1 antigen composition with 2 variants – 3 dose regimen 100 % subjects with bactericidal activity 80 60 40 20 0 Strain 1 Strain 2 Strain 3 Strain 4 Baseline 1m Post Dose 1 1m Post Dose 2 1m Post Dose 3 69 ClinTrial.gov, study NCT01299480 https://goo.gl/Eqbmph

Slide intentionally blank 70

Slide intentionally blank 71

MenABCWY Phase III starts in 2017 – US focused development 11 years 17 years – 1 dose adolescent booster – Phase III programme start in 2017 MenACWY MenACWY MenACWY (~80% (~29% – Filing expected 2020 for adolescents schedule penetration) penetration) previously immunised for MenACWY Post dose 2 immune response rate MenB Bexsero or Bexsero or Other MenB Other MenB Other MenB additions 100 100 100 99 99 hSBA titer ≥threshold 97 93 % of subjects with 80 3-4 injections 60 64 62 40 MenABCWY 20 combination MenABCWY Bexsero MenACWY planned 0 MenB 2 injections 72 Saez-Llorens X et al . Hum Vacc Imm 2015 http://goo.gl/PXXRs6; GSK data on file. CDC MMWR: http://www.cdc.gov/mmwr/preview/mmwrhtml/mm6429a3.htm#tab1

Meningitis portfolio presents significant opportunity GSK has most advanced and comprehensive portfolio for meningitis vaccines 2017 2018 2019 2020 Bexsero demonstrated significant public health benefit, could drive further UMV recommendations Combination approach is MenABCWY MenABCWY Bexsero optimal option for prevention Phase III start US filing US/UK paediatric data Bexsero capacity ~25m doses in 2018 73

Respiratory Syncytial Virus (RSV)

Period of most severe RSV cases for young infants occurs from birth to 12 months 35,000 30,000 US Hospitalisations 25,000 20,000 15,000 10,000 5,000 0 0-2 3-5 6-8 9-11 12-14 15-17 18-20 21-23 mos mos mos mos mos mos mos mos 75 Paramore, Pharmacoeconomics 22:274-285, 2004

Period of most severe RSV cases for young infants occurs from birth to 12 months 35,000 Paediatric Vaccine Maternal Vaccine 30,000 US Hospitalisations 25,000 Week of pregnancy Age in months 28 30 32 34 36 38 40 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 20,000 Maternal Paediatric 15,000 10,000 Maternal IgG 5,000 Infant’s immune response Period of greatest Risk for severe 0 RSV disease 0-2 3-5 6-8 9-11 12-14 15-17 18-20 21-23 mos mos mos mos mos mos mos mos 76 Paramore, Pharmacoeconomics 22:274-285, 2004

Candidate paediatric RSV vaccine, a novel approach Genetically engineered recombinant CHAd155 Same vector used in ebola vaccine Phase Phase Phase – I II III Non-alum composition Sero+ infants Sero- infants Sero- infants Healthy 6-18 mos 2-12 mos 2 mos men fibre Sero- infants Proof of Study start 6-11 mos principle 2021 2022 core Completed Planned Double stranded DNA Target RSV genes 77

Novel candidate RSV maternal vaccine approach Absorbtion with Pre-F but not Post-F depletes neutralising Site I Site I IgG from convalescent serum For RSV F protein, the Site IV Site IV Site Ø Site Ø % neutralising antibody absorbed 100 correct antigen structure Site II Site II 90 is critical Site II Site II 80 Pre-F absorbs out 70 neutralising RSV 60 antibodies more than 10x 50 Site IV better than Post-F and 40 30 induces potent antibody 20 responses in humans LOD 10 Pre-F Post-F Post-F Pre-F Graham B et al., Current Opinion in Immunology 35; 30-38, 2015 GSK preclinical data, unpublished 78

Stabilised Pre-F generated high titers by Day 7 and potent boost of PCA without adjuvant GSK Pre-F 2048 25 Neutralising Ab Titer Fold increase PCA 20 1024 15 512 10 256 5 128 0 60 plain 60 plain Day 30 Day 0 Day 7 Day 30 >20 fold PCA increase after single dose without adjuvant 79 GSK internal data, unpublished

Stabilised Pre-F generated high titers by Day 7 and potent boost of PCA without adjuvant GSK Pre-F Post-F 2048 25 2048 25 Neutralising Ab Titer Neutralising Ab Titer Fold increase PCA Fold increase PCA 20 20 1024 1024 15 15 512 512 10 10 256 256 5 5 128 0 128 0 60 plain 60 Al 120 Al 60 plain 60 plain Day 30 Day 30 Day 0 Day 7 Day 0 Day 7 Day 30 Day 30 >20 fold PCA increase after single dose without adjuvant >10 fold PCA increase requires 120 ug + adjuvant Glenn GM, J Infect Dis. 2015 Aug 10. pii: jiv406. [Epub ahead of print] (data on 60 ug with/without alum) 80 GSK internal data, unpublished Presentation by Novavax at World Vaccine Congress April 2015 (data on 120 ug/alum dose PCA)

Novel candidate RSV maternal vaccine approach Phase Phase Phase Phase I IIa IIb III – Pregnant VE in infants of Healthy men Non-pregnant women vaccinated women Dose selection women Proof of Study start principle 2019 2018 Completed Ongoing Planned 81

Group B Streptococcus (GBS)

Maternal immunisation for GBS No GBS disease GBS disease The leading 40 cause of pneumonia, meningitis and sepsis in 35 neonates 30 % infants 25 1 in 2500 of babies develop GBS disease 20 despite antibiotic 15 prophylaxis of colonised mothers 10 5 No vaccine is available 0 <0.5 <0.5-.99 1-1.99 2-2.99 3-3.99 4-4.99 5-5.99 6-6.99 7-7.99 8-8.99 9-9.99 10-14.99 15-19.99 <20 Maternal antibody concentration 83 Gibbs, Obstet Gynecol, 104;1062-1075, 2004 Lin FY et al . J Infect Dis. 2004;190:928-934

Maternal immunisation for GBS No GBS disease GBS disease The leading 40 cause of pneumonia, meningitis and sepsis in 35 neonates 30 % infants 25 1 in 2500 of babies develop GBS disease 20 Protected despite antibiotic 15 prophylaxis of colonised mothers 10 5 No vaccine is available 0 <0.5 <0.5-.99 1-1.99 2-2.99 3-3.99 4-4.99 5-5.99 6-6.99 7-7.99 8-8.99 9-9.99 10-14.99 15-19.99 <20 Maternal antibody concentration 84 Gibbs, Obstet Gynecol, 104;1062-1075, 2004 Lin FY et al . J Infect Dis. 2004;190:928-934

GBS maternal immunisation expanded programme Large Phase II trivalent completed Based on Capsular Decision to expand composition to pentavalent polysaccharide (CPS) from 5 dominant GBS Validate correlate of protection with FDA serotypes conjugated to Clinical development plan to be agreed with FDA a protein carrier Capsular 5 dominant polysaccharide serotypes from GBS Designed to help protect against >95% of 25 Vaccinated mothers 1gG against serotypes 1a (µg/mL) globally prevalent Infants of vaccinated mothers Unvaccinated mothers and their infants 20 serotypes 15 Phase II trivalent 2014 2015 2016 2017 2018 2019 10 vaccine antibody data shows response at 5 Phase II Validation of Trivalent Assay development protective Further development period of greatest risk 0 Completed antibody threshold Pre Delivery Birth Day 42 Day 90 EOD LOD Period of greatest risk disease for GBS 85 Le Doare, Vaccine 31(4) D7, 2013 ; GSK clinical data, unpublished

Maternal immunisation validated strategy to prevent diseases that afflict very young infants Infants protected by maternal flu vaccination 0.05 VE = 50.5% p = 0.01 – confirmed influenza – Proportion with Placebo Flu vaccine 0.00 0 2 4 6 Age (months) 86 Source: N Engl J Med. 2014 Sep 4;371(10):918-31

GSK potential maternal immunisation vaccine portfolio GBS Pertussis 50% 45% 40% 35% Percent of cases 30% 25% Influenza RSV 20% 15% 10% 5% Influenza-Associated Mortality (deaths per 100,000 children) 0% 0-7d 8-30d 30-90d >90d Age of onset Winter K, MMWR 63:1122- Melin, Clin Microbiol Inf, 1140,2014 17;1294-1303, 2011 <6 m 6-11m 1 yr 2 yr 3 yr 4yr 5-10yr 11-17yr Age Group Bhat N Engl J Med.353:2559-67, 2005 Paramore, Pharmacoeconomics 22:274-285, 2004 87

A new vaccine concept

Testing hypothesis for a COPD vaccine Epi studies show association between lung infections & COPD exacerbations 1,2 Development plan to support proof of concept NTHi and Mcat: 2 lung pathogens potentially Epi 001 associated with 30-50% of COPD Detailed molecular microbiology study of AE-COPD exacerbations 1,2 – NTHi 001 NTHiMcat NTHi 003 NTHi 004 Adults with Healthy Adults Older adult Adults with GOLD III/IV 75% effective vaccine could eliminate 20- smokers GOLD II/III COPD COPD 35% of exacerbations Proof of concept NTHiMcat 001 End 2019 Healthy Adults Proof of 3 antigen vaccine covering NTHi using AS01 Completed principle End 2017 Ongoing adjuvant in Phase II POC trial Safety and Planned immuno data end 2016 Key POC data in COPD patients = 2017 Phase III to be defined based on POC data 1GSK unpublished data EpiHip001. 2Sethi et al . N Engl J Med. 2002 Aug 15;347(7):465-71. 89

Data and planned filings support positive growth outlook 2016 2017 2018 2019 2020 2021-2025 Shingrix immuno- Q flu paediatric MMR US filing US/UK Bexsero MenABCWY RSV maternal compromised US filing paediatric data US filing filings efficacy filing Rotarix liquid Shingrix US, EU, MMR Japan GBS maternal US filing Japan filings filing filings RSV paediatric filings COPD vaccine Phase III 90

R&D programmes to deliver near-term growth with significant future opportunities and novel immunisation platforms 3 A new vaccine concept − COPD 2 Longer term R&D Focus − RSV − GBS Near/mid term key R&D focus: 1 − Shingrix − Meningitis − Lifecycle management 91

Introducing the Vaccines panel GSK’s leading scientists in vaccines Emmanuel Hanon Alain Brecx Senior Vice President, Vice President Head of Vaccines R&D Vaccine Development Lead - Zoster Rip Ballou Giovanni Della Cioppa Vice President Vice President, Head of Rockville R&D Centre Head of Siena R&D Centre 92

Q&A

Immuno-Inflammation

Immuno-Inflammation areas of focus Immune modulation to alter disease course, induce and sustain remission Rheumatoid Osteoarthritis Systemic Lupus Other immune- Arthritis (OA) Erythematosis mediated (RA) (SLE) diseases • Circa 5.3m RA patients in G7 • Circa 72m OA patients in G7 • Prevalence: 40 -100 out of • Mechanisms are relevant for countries 1 countries; largest proportion 100,000 4 ; 9/10 sufferers are mainstream diseases e.g psoriasis, Crohn’s disease & of musculoskeletal women in their 20s & 30s 4 • Aging demographics a major diseases 2,3 • Chronic disease with poor QoL, ulcerative colitis driver of market growth • Aging demographics a major involving musculoskeletal, • Opportunities exist to treat • Highly debilitating; associated haematological, cutaneous & driver of market growth less common disease e.g. with higher mortality & renal systems primary Sjögren's syndrome, • Major opportunity for a progression to other serious • Mortality rate 3x higher than the systemic sclerosis & conditions disease-modifying therapy general population, and 10x myasthenia gravis • Significant medical needs for • Immune modulation offers higher in under 40 5 remission-inducing therapies & opportunity to move from • Benlysta IV - 1 st drug approved for patients resistant to current only alleviating symptoms of for SLE in 50 years (2011) “wear and tear” standard of care 1 Decision Base Rheumatoid Arthritis 2015 ; 2 World Health Organisation 2010; 3 Decision Resources OA Pain 2012; 4 Danchenko N et al . Epidemiology of SLE: a comparison of worldwide disease burden. Lupus 2006; 15:308 – 318 5 Bernatsky S, Boivin JF, Joseph L, et al . Mortality in systemic lupus erythematosus. Arthritis Rheum 2006; 54:2550 – 2557.; * Decision resources 2013 estimate 95

Immuno-Inflammation R&D strategy: From symptomatic benefit to sustainable remission macrophage Targeted Biologicals Targeted Small neutrophil Molecules GSK Pipeline stromal cell Targeting Resistant Early Intervention & T cell Disease Remission Induction plasma cell B cell 96

sirukumab: rheumatoid arthritis The anti-IL-6 class is the fastest growing of the biologicals in RA • Collaboration with Janssen Biologics (Ireland) 12 wk data from Phase II study 1 • Low frequency sc dosing potential (monthly) Response rate (%) ** ** p < 0.05 80 63 57 60 • Targets the cytokine ** ** 30 40 27 27 • Efficacy demonstrated in Phase II; consistent 20 3 0 safety profile across doses ACR20 at wk 12 ACR50 at wk 12 (Primary EP) • >3000 patients in studies to date placebo 50mg Q4W 100mg Q2W 1 adapted from Smolen et al 2014 Ann Rheum Dis 73 (9) • Phase III interim read-out, full read out expected 24 w data from Japan monotherapy study 2 by year end 2015 82 Response rate (%) 74 80 64 • Indication expansion: Phase III in Giant Cell 60 49 Arteritis started screening. Phase II in asthma 36 40 25 start in 2016 20 0 Status: RA: Phase III ACR20 ACR50 ACR70 50mg Q4W 100mg Q2W Indications: RA (lead), GCA, asthma Planned Filing: RA 2016 2 ACR 2015 abstract #1672 97

Clinical improvement in RA is consistently associated with decreased macrophage infiltration • Activated macrophages are abundantly GM-CSF plays a key role in activation of expressed in early RA synovial tissue, macrophages at the site of injury or inflammation representing the predominant cell type • Reduction in macrophage infiltration correlates with improvement in disease activity scores 1,2 • Macrophage is a primary cause of tissue destruction and affects many other cell types • GM-CSF is important in every step of macrophage production and infiltration in the tissues 1 Boumans MJ, et al . Arthritis Rheum. 2011;63:3187-94. 2 Bresnihan B, et al . J Rheumatol 2009;36:1800-2. 98

GSK3196165 – aGM-CSF, targets key effector cells in RA Aiming to induce remission in early rheumatoid arthritis • In-licensed from MorphoSys AG % EULAR good/moderate response at 4 weeks: Rapid onset of action • Good magnitude of effect with fast onset of 80 Phase Ib/IIa study, N= 96 action and long duration post treatment % EULAR response 60 • Effect size appears similar or greater than anti-TNF 40 • Targeting the macrophage in early RA 20 • Potential for early use to induce remission 0 • BAROQUE (RA Phase IIb) ongoing. Placebo 0.3 mg/kg 1.0 mg/kg 1.5 mg/kg Initial clinical read-out 2016 Week 4 Week 6 Week 8 Status: Phase IIb Indication: Rheumatoid Arthritis Behrens, et al . Ann Rheum Dis. 2015;74:1058-64 Planned Filing: 2021-2025 99

GSK3196165: Potential for disease modification & analgesic activity in hand osteoarthritis (HOA) GM-CSF receptor expression on primary afferent nerve fibres • The macrophage is a mediator of tissue in mouse tibial bone and periosteal nerves destruction in OA • aGM-CSF is effective in animal models of OA • aGM-CSF rapidly reduces pain (through effect on nerves) in animal models of OA • Hand OA presents unique clinical development path M Schweizerhof et al . Nature Medicine 2009;15:802-807 Pain • Phase II to start in 2016 110 * Weight distribution * * * * # * * * * * * * 100 * p<0.05, anti-GM-CSF ** p<0.01, 90 *** p<0.001, control control vs. anti-GM CSF # p<0.0001, control and anti- 80 GM-CSF vs. t=0 readings Status: Phase II start 2016 70 Indication: Hand OA 0 5 10 15 20 25 30 35 40 45 Planned Filing: 2021-2025 Days post disease induction Cook et al . Arthritis Res Ther. 2012;14:R199 GSK3196165 in-licensed from MorphoSys AG 100