Exhibit 99.2

POSTER P2113

Safety of fluticasone furoate (FF), an inhaled corticosteroid in combination with vilanterol (VI), a long-acting beta agonist in management of COPD exacerbations

Calverley PMA(1), Dransfield MT(2), Bourbeau J(3), Jones P(4), Hanania NA(5), Mahler DA(6), Vestbo J(7), Wachtel A(8), Martinez F(9), Barnhart F(10), Sanford L(11), Lettis S(11), Crim C(10)

(1)University of Liverpool, Liverpool, UK; (2)University of Alabama at Birmingham, Birmingham, AL, USA; (3)McGill University, Montreal, Canada; (4)St George’s University of London, London, UK; (5)Baylor College of Medicine, Houston, TX, USA; (6)Dartmouth Medical School, Hanover, NH, USA; (7)Manchester Academic Health Sciences Centre, Manchester, UK; (8)UCLA School of Medicine, Los Angeles, CA, USA; (9)University of Michigan, Ann Arbor, MI, USA; (10)Respiratory Medicine Development Center, GlaxoSmithKline, RTP, NC, USA; (11)Quantitative Sciences Division, GlaxoSmithKline, Uxbridge, UK

INTRODUCTION

·                  The combination of inhaled corticosteroids (ICS) with a long-acting beta2 agonist (LABA) is an established treatment option for COPD patients who experience acute exacerbations.(1)

·                  FF and VI are, respectively, a novel ICS and LABA in development as a once-daily combination therapy (FF/VI) for COPD and asthma.

·                  Although the benefits of ICS in COPD are established, an association of ICS use with an increased risk of pneumonia(2) and bone disorders(3) has been reported.

OBJECTIVES

·                  To assess the safety and tolerability of three once-daily dosing regimens of FF/VI and one of VI in patients >40 years of age with moderate-to-severe COPD. Efficacy endpoints are described separately (poster no. P2888).

METHODS

·                  Two phase III, multi-centre, randomised, double-blind, parallel-group studies (HZC871 and HZC970) identical in design, conduct and analysis. Data from these studies were pooled.

·                  Patients (post-bronchodilator FEV1 <70%; FEV1/FVC ratio <70%; smoking history >10 pack years; having had >1 documented COPD exacerbation in the year prior to screening) completed a 4-week run-in period during which they received fluticasone propionate/salmeterol 250/50mcg twice daily via DISKUS™/ACCUHALER™ to determine adherence.

·                  Patients were randomised to receive one of the following, once daily in the morning via a novel dry powder inhaler for 52 weeks

·                  FF/VI 50/25mcg

·                  FF/VI 100/25mcg

·                  FF/VI 200/25mcg

·                  VI 25mcg.

RESULTS

Study population and demographics

·                  Pooled patient demographics and baseline characteristics are provided in Table 1.

·                  3255 patients were randomised and received at least one dose of study medication (ITT population); 2406 completed the studies.

Table 1. Patient demographics and baseline characteristics (pooled HZC871 & HZC970, ITT population)

Values are mean (SD) unless otherwise stated; ITT = intent-to-treat

·                  On-treatment withdrawal rates (FF/VI 50/25, 100/25, 200/25mcg, VI 25mcg) were, respectively: 25%, 25%, 25%, 29%. The incidence of withdrawal due to adverse events (AEs) was 7–8% in the FF/VI groups and 6% in the VI group.

Safety

·                  The most common on-treatment AEs were nasopharyngitis (14–19%), candidiasis (7–13%) and upper respiratory tract infection (9–11%) (Table 2).

·                  More patients receiving FF/VI (17–21%) than VI (14%) experienced AEs that were deemed treatment related by the investigator.

·                  Percent of patients with events in the Cardiac disorders system order class: VI 6%, FF/VI arms 4–5%.

·                  No fatal AEs were considered by the investigator to be treatment related.

·                  The incidence of local steroid effects and bone disorders, including fractures, was higher in patients receiving FF/VI than VI (Table 3). Bone fractures were the majority of events in the bone disorders category. The incidence of bone fractures overall was low in all treatment groups, with a higher incidence in all FF/VI groups (2%) compared with the VI 25mcg group (<1%)

·                  the proportion of traumatic fractures was 75%, 73%, 68% and 38%, respectively for VI 25mcg, FF/VI 50/25, 100/25 and 200/25mcg. Fractures customarily associated with corticosteroid use (e.g. spinal compression/thoracolumbar vertebral fractures, hip and acetabular fractures) occurred in <1% across all treatment arms.

Table 2. Summary of AEs (pooled HZC871 & HZC970, ITT population)

Values are n (%) unless otherwise stated;

(a)3 patients had an event whilst on-treatment and an event post-treatment, both of which were recorded as leading to death; (b)candidiasis, oral candidiasis, oropharyngeal candidiasis and oesophageal candidiasis

Table 3. AE groupings of pre-defined special interest (pooled HZC871 & HZC970, ITT population)

(a) A hazard ratio less than unity favours VI

Safety (continued)

·                  Pneumonia occurred approximately twice as often in the FF/VI groups than in the VI group (Table 3).

·                  There were six cases of fatal pneumonia and one case of fatal COPD exacerbation with concurrent pneumonia in the FF/VI 200/25mcg group, all of which occurred in the HZC871 study and the majority of these cases were reported from one site; one case of fatal pneumonia occurred in the FF/VI 100/25 group of the HZC970 study

·                  the number of fatal cases was too small to allow investigation of risk factors for fatal pneumonia.

·                  No clinically relevant changes in clinical chemistry or haematology values, or in ECG-measured heart rate or QTcF, were observed.

CONCLUSIONS

·                  The use of FF/VI was associated with the expected increase in known steroid-associated AEs and bone disorders; correspondingly, the incidence of AEs and treatment-related AEs was higher in patients receiving FF/VI than VI.

·                  Overall rates of serious and fatal AEs were similar across the four treatment groups.

·                  The increase in risk of pneumonia with FF is consistent with the findings of previous studies of ICS in COPD.(2)

REFERENCES

(1)         GOLD 2011. Available at: http://www.goldcopd.org/uploads/users/files/GOLD_Report_2011_ Feb21.pdf [Last accessed 26 June 2012].

(2)         Crim, C et al. Eur Respir J 2009;34:641–647.

(3)         Loke YC, et al. Thorax 2011;66:699-708

ACKNOWLEDGEMENTS

·                  The presenting author, PMA Calverley, declares the following real or perceived conflicts of interest during the last 3 years in relation to this presentation: received funds for advising the sponsors of this study on its conduct and analysis; performed similar functions with other pharmaceutical companies including AstraZeneca, Boehringer Ingelheim, Novartis, Nycomed and Takeda.

·                  These studies were funded by GlaxoSmithKline; GSK Study Codes HZC102871 & HZC102970, Clinicaltrials.gov NCT01009463 & NCT01017952.

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

Presented at the European Respiratory Society Annual Congress 2012 Vienna, Austria, 1–5 September, 2012