Exhibit 99.2

POSTER NO. PS272

Ethnic sensitivity assessment of fluticasone furoate (FF)/vilanterol (VI) in asthma patients in Japan and Korea: a pre-specified subgroup analysis

Gross AS(1), Goldfrad C(2), Hozawa S(3), James M(4), Clifton CS(1), Sugiyama Y(5), Jacques L(6)

(1)Clinical Pharmacology Modelling & Simulation, GlaxoSmithKline, Sydney, Australia; (2)Quantitative Sciences Division, GlaxoSmithKline, Uxbridge, UK; (3)Hiroshima Allergy and Respiratory Clinic, Hiroshima, Japan; (4)Medical Affairs, GlaxoSmithKline K.K., Tokyo, Japan; (5)Medicines Development Respiratory, GlaxoSmithKline K.K., Tokyo, Japan; (6)Respiratory Medicines Discovery and Development, GlaxoSmithKline, Uxbridge, UK

INTRODUCTION

·                  Inhaled corticosteroid (ICS)/long-acting beta2-agonist (LABA) combinations are recommended by Japanese(1) and global asthma guidelines(2) for asthma patients uncontrolled on ICS alone.

·                  FF/VI is a once-daily ICS/LABA combination therapy delivered via the ELLIPTATM dry powder inhaler, which is effective for 24h.

·                  Responses to pharmacotherapy can vary across ethnic groups(3),(4) including in Japanese patients.

OBJECTIVES

·                  To establish whether doses of FF/VI recommended from multinational studies are relevant to asthma patients in Japan.

·                    To compare the efficacy, safety, PK and PD data for FF/VI in patients from Japan and/or Korea with data from patients not from Japan or Korea

·                  Patients from Korea were included due to similarities in a range of intrinsic and extrinsic ethnic factors between the populations of Japan and Korea.(5)–(7)

METHODS

·                  A pre-specified subgroup analysis of multicentre, randomised, double-blind, parallel-group international studies that included asthma patients from Japan and/or Korea.

·                  Inclusion criteria: >12 years of age, pre-bronchodilator % predicted forced expiratory volume in one second (FEV1) of 40–90%, FEV1 reversibility of >12% and >200mL.

·           Efficacy results were pooled from three Phase III studies, ranging from 12 to 76 weeks duration, that included patients from Japan (GSK study numbers: HZA106827; HZA106829; HZA106837).

·           Change from baseline in trough FEV1 after 12 weeks with once-daily FF/VI 100/25mcg, once-daily FF 100mcg or placebo was analysed across two studies (HZA106827; HZA106837) and data for once-daily FF/VI 200/25mcg, once-daily FF 200mcg and twice-daily fluticasone propionate (FP) 500mcg were from one study (HZA106829).

·           Safety data were pooled from the three studies noted above and an additional three placebo-controlled Phase IIb studies, ranging from 28 days to 8 weeks duration, that included patients from Korea (B2C109575; FFA109685; FFA109687).

·           PK data were estimated based on a post-hoc analysis of population PK data for FF(8) and of study DB111207 data for VI(9); PD data, including 24h urinary cortisol excretion, were assessed.

RESULTS

Table 1. Demographic and baseline characteristics

Demographic and baseline characteristics (Efficacy population)

Demographic characteristics (Safety population)

(1)Mean (SD); (2)n=147; (3)n=3059; (4)n=3206; (5)n=4036; (6)n=4230

Efficacy

·                    Efficacy data were compared between patients (N=3214) from Japan and ‘Not-Japan’; 85% (N=2739) completed the studies.

·                    Improvements in trough FEV1 were reported in all populations for FF/VI 100/25mcg and FF 100mcg versus placebo and for FF/VI 100/25mcg versus FF 100mcg (Table 2)

·                  There was no evidence of a statistically significant difference in treatment effect between patients from Japan and Not-Japan (p=0.403).

·                    Changes from baseline in FEV1 were similar with FF/VI 200/25mcg and FF 200mcg in patients from Japan and the overall population (Figure 1)

·                  Improvements were greater than with twice-daily FP 500mcg.

Table 2. Comparison of change from baseline in trough FEV1 at Week 12 between treatment arms (Efficacy population)

Least squares mean change (95% confidence interval). Studies included: HZA106827, HZA106837. Data were analysed using a Last Observation Carried Forward Analysis of Covariance model, with terms for baseline FEV1, region, gender, age, treatment group, study & region by treatment interaction. 1n=193; 2n=18, 3n=175

Figure 1. Box plot of change from baseline in trough FEV1 at Week 12 by geographic region (Efficacy population)

Safety

·                    Safety data were compared between patients (N=4231) from Japan/Korea and Not-Japan/Korea; N=3584 (85%) completed the studies.

·                    In all active treatment groups, a greater proportion of patients from Japan/Korea versus Not-Japan/Korea reported on-treatment adverse events; this trend was also observed in the placebo group, suggesting that this is not related to FF or VI (Table 3).

Table 3. Summary of on-treatment adverse events and serious adverse events (Safety population)

Data presented as number of patients with an adverse event/number of patients in population group (%);

(1)Only subjects from Japan;

(2)One additional subject (South East Asian) died during follow-up

Safety (cont’d)

·                    Serious adverse events reported for patients from Japan+Korea (1 FF/VI 100/25mcg, 1 FF 100mcg: both subarachnoid haemorrhage) were not considered drug related.

PK/PD

·                    Estimated FF AUC(0-24) and VI Cmax were higher in Japanese versus White/Caucasian patients (Table 4).

PK/PD (cont’d)

·                    No clinically relevant effects on cortisol concentrations (Figure 2) or heart rate were observed.

·                    The PK profile of FF at clinical doses (<200mcg) did not differ when administered simultaneously with VI 25mcg compared with FF alone for either Japanese or White/Caucasian patients.

Table 4. Model predicted PK parameters

Geometric mean (95% confidence interval); FF data from post-hoc analysis of population PK data(8); VI data from post-hoc analysis of study DB111207(9)

Figure 2. Ratio of 24h urinary cortisol excretion to baseline at the end of treatment  (Urinary cortisol population(1))

(1)Urinary cortisol population: subset of patients from Safety population, that were not considered to have confounding factors, and for whom urine samples were available;

(2)Only subjects from Japan

CONCLUSIONS

·                    The efficacy and safety profile of FF/VI is similar in asthma patients from Japan+Korea and Not-Japan+Korea.

·                    The FF/VI clinical doses recommended based on global studies are also suitable for asthma patients in Japan.

REFERENCES

(1)         Ohta K, et al. Allergol Int. 2011;60:115–45.

(2)         Global Initiative for Asthma. Global Burden of Asthma. Available at: http://www.ginasthma.org/ local/uploads/files/GINABurdenReport_1.pdf. Last accessed: 10 Sept 2013.

(3)         Bjornsson TD, et al. J Clin Pharmacol 2003;43:943–67.

(4)         Huang SM, et al. Clin Pharmacol Ther 2008;84:287–94.

(5)         Jin HJ, et al. PLoS ONE 2009;4:1–10.

(6)        HUGO Pan-Asian SNP Consortium, et al. Science 2009;326:1541–45.

(7)         Chowbay B, et al. Drug Metab Rev 2005;37:327–78.

(8)         GSK data on file.

(9)         Nakahara N, et al. Int J Clin Pharmacol Ther 2013;51:660–72

ACKNOWLEDEGMENTS

·                    The presenting author, Yutaro Sugiyama, is employed by GlaxoSmithKline.

·                    The authors acknowledge the contributions of the following employees of GlaxoSmithKline: Dr Romina Nand, Dr Carol Lee and Dr Ann Allen.

·                    These studies were funded by GlaxoSmithKline (GSK study codes HZA106827 (cllinicaltrial.gov registration number: NCT01165138); HZA106829 (NCT01134042); HZA106837 (NCT01086384); B2C109575 (NCT00600171); FFA109685 (NCT00603278); FFA109687 (NCT00603382); DB111207 (NCT00964249).

·                    Editorial support (in the form of writing assistance, assembling tables and figures, collating author comments, grammatical editing and referencing) was provided by Laura Maguire, MChem, at Gardiner-Caldwell Communications (Macclesfield, UK) and was funded by GlaxoSmithKline.

ELLIPTATM is a trade mark of the GlaxoSmithKline group of companies

Presented at the 18th Congress of the Asian Pacific Society of Respirology, Yokohama, Japan, 11–14 November 2013