In February 2013, Glaxosmithkline (GSK) Announced a Commitment

In February 2013, GlaxoSmithKline (GSK) announced a commitment to further clinical transparency through the public disclosure of GSK Clinical Study Reports (CSRs) on the GSK Clinical Study Register.

The following guiding principles have been applied to the disclosure:  Information will be excluded in order to protect the privacy of patients and all named persons associated with the study  Patient data listings will be completely removed* to protect patient privacy. Anonymized data from each patient may be made available subject to an approved research proposal. For further information please see the Patient Level Data section of the GSK Clinical Study Register.  Aggregate data will be included; with any direct reference to individual patients excluded *Complete removal of patient data listings may mean that page numbers are no longer consecutively numbered CONFIDENTIAL 2019N423508_00 The GlaxoSmithKline group of companies 201097  Division: Worldwide Development Information Type: Clinical Study Report Control: Usual Care

Title: The effectiveness of Asthma Control Test guided treatment compared with usual care in China adult asthma patients

Phase: IV

Compound Number: CCI18781+GR33343

Effective Date: 8-JUN-2021

Subject: Asthma, Asthma Control Test guided

Author(s): PPD

Indication Studied: Asthma

Initiation Date: 26-AUG-2016

Completion Date: 09-AUG-2019

Early Termination Date: NA

Earlier CSRs NA

Sponsor Signatory: Dr. Pauline Ng (and Medical Officer) Country Medical Director GlaxoSmithKline

This study was performed in compliance with Good Clinical Practices and GlaxoSmithKline Standard Operating Procedures for all processes involved, including the archiving of essential documents. This study complies with US 21 CFR 312.120.

Copyright 2021 the GlaxoSmithKline group of companies. All rights reserved. Unauthorised copying or use of this information is prohibited.

   CONFIDENTIAL 2019N423508_00 201097

Table of Contents Page

TITLE PAGE ...... 1 ABBREVIATIONS ...... 8 ETHICS AND GOOD CLINICAL PRACTICE ...... 9 1. INTRODUCTION ...... 10 2. STUDY OBJECTIVE(S) ...... 10 2.1. Primary objective ...... 10 2.2. Secondary objectives ...... 10 2.3. Exploratory objective ...... 10 3. INVESTIGATORS AND STUDY ADMINISTRATIVE STRUCTURE ...... 10 4. INVESTIGATIONAL PLAN ...... 11 4.1. Study Design ...... 11 4.2. Discussion of Study Design ...... 12 4.3. Protocol Amendment(s) ...... 12 4.4. Selection of Study Population ...... 12 4.4.1. Inclusion/Exclusion Criteria ...... 12 4.4.2. Withdrawal Criteria ...... 13 4.5. Treatments ...... 13 4.5.1. Investigational Product(s) and Reference Therapy...... 13 4.5.2. Treatment Assignment ...... 13 4.5.3. Blinding ...... 14 4.5.4. Prior and Concomitant Medications and Non-Drug Therapies ...... 14 4.5.5. Compliance ...... 14 4.6. Study Assessments and Procedures ...... 14 4.6.1. Screening and Critical Baseline Assessments ...... 15 4.6.2. Efficacy Assessment ...... 16 4.6.3. Safety Assessments ...... 20 4.7. Data Quality Assurance ...... 21 4.8. Statistical Analyses ...... 22 4.8.1. Statistical Hypotheses ...... 22 4.8.2. Sample Size Considerations ...... 22 4.8.3. Analysis Populations ...... 23 4.8.4. Interim Analyses ...... 24 4.8.5. Final Analyses ...... 24 4.8.6. Changes in Conduct of the Study or Planned Analyses ...... 29 5. STUDY POPULATION RESULTS ...... 31 5.1. Subject Disposition ...... 31 5.2. Protocol Deviations ...... 32 5.3. Populations Analyzed ...... 32 5.4. Demographics and Baseline Characteristics ...... 33 5.5. Past and Current Illness ...... 34 5.6. Asthma Medications ...... 35 5.7. Concurrent Medications ...... 36 6. EFFICACY RESULTS ...... 38

2 CONFIDENTIAL 2019N423508_00 201097

6.1. Primary Efficacy Results ...... 38 6.1.1. Percentage of Subjects Who Had an ACT Total Score ≥ 20 or an Improvement >3 Points in ACT Total Score in at Least One Post- baseline Assessment ...... 38 6.1.2. Percentage of Subjects Who Have an ACT Total Score ≥ 20 and an Improvement >3 Points in ACT Total Score ...... 43 6.1.3. Change from Baseline in ACT Total Score ...... 47 6.2. Secondary Efficacy Results ...... 48 6.2.1. Mean daytime symptom score over the 24-week treatment period ...... 48 6.2.2. Mean night-time symptom score over the 24-week treatment period ...... 49 6.2.3. Mean change from baseline to the end of study in FEV1 ...... 50 6.2.4. Mean morning (AM) PEF over the 24-week treatment period ...... 50 6.2.5. Mean evening (PM) PEF over the 24-week treatment period ...... 51 6.2.6. Mean change from baseline to the end of study in AQLQ(S) score ...... 52 6.2.7. Time to first ACT total score ≥20 or an Improvement >3 Points in ACT total score over the 24-week treatment period ...... 52 6.2.8. Time to first ACT total score ≥20 and an Improvement >3 Points in ACT total score over the 24-week treatment period ...... 53 6.3. Exploratory Efficacy Results ...... 54 6.4. Other Post-hoc Efficacy Results ...... 56 6.4.1. Percentage of Subjects Who Had an ACT Total Score ≥ 20 or an Improvement ≥ 3 Points in ACT Total Score in at Least One Post- baseline Assessment ...... 56 6.4.2. Percentage of Subjects Who Had an ACT Total Score ≥ 20 and an Improvement ≥ 3 Points in ACT Total Score in at Least One Post- baseline Assessment ...... 56 6.4.3. Time to first ACT total score ≥20 or an Improvement ≥3 Points in ACT total score over the 24-week treatment period ...... 57 6.4.4. Time to first ACT total score ≥20 and an Improvement ≥3 Points in ACT total score over the 24-week treatment period ...... 58 6.5. Efficacy Conclusions...... 58 7. SAFETY RESULTS ...... 60 7.1. Adverse Events...... 60 7.1.1. Overview of AEs ...... 60 7.1.2. Adverse Events Related to Study Process or GSK’s Drugs ...... 61 7.2. Serious and Other Significant Adverse Events ...... 62 7.2.1. Deaths ...... 62 7.2.2. Other Serious Adverse Events ...... 62 7.2.3. Other Significant Adverse Events ...... 62 7.3. Non-Serious Adverse Events of the Subjects...... 62 7.4. Adverse Events Leading to Temporarily Suspend\Delay GSK’s drugs ...... 63 7.5. Adverse Events Leading to Withdrawal ...... 63 7.6. Other Safety Evaluations ...... 63 7.7. Safety Conclusions ...... 64 8. DISCUSSION AND CONCLUSIONS ...... 65 8.1. Discussion ...... 65

3 CONFIDENTIAL 2019N423508_00 201097

8.2. Conclusions ...... 66 9. REFERENCES ...... 67 10. POST-TEXT TABLES AND FIGURES ...... 68 11. CASE NARRATIVES ...... 74 STUDY POPULATION DATA SOURCE TABLES...... 75 Table 1.01 Subjects Enrollment And Study Completion ...... 75 Table 1.02 Summary of Number of Subjects by Centre ...... 76 Table 1.03 Analysis Population (Screening/Randomized Subject) ...... 77 Table 1.04 Subjects with Important Protocol Deviation(Screening/Randomized Subjects) ...... 78 Table 1.05 Subjects' Demographic And Other Baseline Characteristics(ITT) ...... 79 Table 1.06 Subjects' History of Smoking, Alcohol Abuse, Medication Abuse, Asthma, And Asthma Exacerbation (ITT) ...... 83 Table 1.07 Subjects' Past and Current Illness(ITT) ...... 86 Table 1.08 Subjects' Asthma Medication (ITT) ...... 101 Table 1.09 Subjects' Concurrent Medication (ITT) ...... 103 EFFICACY DATA SOURCE FIGURES...... 139 Figure 1 Forest Plot of Adjusted Odds Ratios and 95% CIs From GLMM, Comparing ACT Guided Treatment With Usual Care On Primary Analysis and Sensitivity Analysis of ACT Response (ITT) ...... 139 Figure 2 Forest Plot of Adjusted Odds Ratios and 95% CIs From GLMM, Comparing ACT Guided Treatment With Usual Care On Primary Analysis and Sensitivity Analysis of ACT Response (PP) ...... 140 Figure 3 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (ITT) ...... 141 Figure 4 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (PP) ...... 142 Figure 5 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (ITT) ...... 143 Figure 6 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (PP) ...... 144 Figure 7 Forest Plot of LS Mean change from baseline treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (ITT) ...... 145 Figure 8 Forest Plot of LS Mean change from baseline treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (PP) ...... 146

4 CONFIDENTIAL 2019N423508_00 201097

Figure 9 Forest Plot of LS Mean treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On Mean Daytime Symptom Score By Visits (ITT) ...... 147 Figure 10 Forest Plot of LS Mean treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On Mean Night-time Symptom Score By Visits (ITT) ...... 148 Figure 11 Forest Plot of LS Mean change from baseline treatment differenes and 95% CIs From Mixed Model, Comparing ACT Guided Treatment With Usual Care On FEV1 (ITT) ...... 149 Figure 12 Forest Plot of LS Mean treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On Mean Morning PEF By Visits (ITT) ...... 150 Figure 13 Forest Plot of LS Mean treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On Mean Evening PEF By Visits (ITT) ...... 151 Figure 14 Forest Plot of LS Mean change from baseline treatment differenes and 95% CIs From Mixed Model, Comparing ACT Guided Treatment With Usual Care On AQLQ(S) score (ITT) ...... 152 Figure 15 KM Figure of Time to Asthma Control (ACT Total Score >=20 Or Improvement of More Than 3 Points in ACT Total Score) For the First Time(ITT) ...... 153 Figure 16 KM Figure of Time to Asthma Control (ACT Total Score >=20 And Improvement of More Than 3 Points in ACT Total Score) For the First Time(ITT) ...... 154 Figure 17 Forest Plot of Adjusted Odds Ratios and 95% CIs From GLMM, Comparing ACT Guided Treatment With Usual Care On Primary Analysis and Sensitivity Analysis of ACT Response (ITT) ...... 155 Figure 18 KM Figure of Time to Asthma Control (ACT Total Score >=20 Or Improvement >= 3 Points in ACT Total Score) For the First Time(ITT) ... 156 Figure 19 KM Figure of Time to Asthma Control (ACT Total Score >=20 And Improvement >= 3 Points in ACT Total Score) For the First Time(ITT) ... 157 EFFICACY DATA SOURCE TABLES ...... 158 Table 2.01 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24- Weeks Treatment Period (ITT) ...... 158 Table 2.02 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During 24- Weeks Treatment Period (ITT) ...... 159 Table 2.03 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During 24- Weeks Treatment Period (ITT) ...... 160

5 CONFIDENTIAL 2019N423508_00 201097

Table 2.04 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24- Weeks Treatment Period by Subgroup (PP) ...... 166 Table 2.05 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or AnImprovement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (ITT) ...... 172 Table 2.06 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (PP) ...... 175 Table 2.07 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score (ITT) ...... 178 Table 2.08 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score (PP) ...... 180 Table 2.09 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (ITT) ...... 182 Table 2.10 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (PP) ...... 185 Table 2.11 Summary of the Statistical Analysis of Change from Baseline in ACT Total Score (ITT) ...... 188 Table 2.12 Summary of the Statistical Analysis of Change from Baseline in ACT Total Score (PP) ...... 191 Table 2.13 Summary Of Mean Daytime Symptom Score Over the 24-Week Treatment Period (ITT) ...... 194 Table 2.14 Analysis of Mean Daytime Symptom Score Over the 24-Week Treatment Period (ITT) ...... 196 Table 2.15 Summary Of Mean Night-time Symptom Score Over the 24-Week Treatment Period (ITT) ...... 199 Table 2.16 Analysis of Mean Night-time Symptom Score Over the 24-Week Treatment Period (ITT) ...... 201 Table 2.17 Summary of the Change From Baseline in Forced Expiratory Volume In One Second (ITT) ...... 204 Table 2.18 Analysis of the Change From Baseline in Forced Expiratory Volume In One Second (ITT) ...... 205 Table 2.19 Summary Of Mean Morning Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) ...... 206 Table 2.20 Analysis of Mean Morning Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) ...... 208

6 CONFIDENTIAL 2019N423508_00 201097

Table 2.21 Summary Of Mean Evening Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) ...... 211 Table 2.22 Analysis of Mean Evening Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) ...... 213 Table 2.23 Summary Of the Change From Baseline in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score (ITT) ...... 216 Table 2.24 Analysis Of the Change From Baseline in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score (ITT) ...... 219 Table 2.25 Time to First ACT Total Score >=20 or Improvement of More Than 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT) .... 222 Table 2.26 Time to First ACT Total Score >=20 and Improvement of More Than 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT) .... 223 Table 2.27 Summary of Moderate/Severe Asthma Exacerbations (ITT) ...... 224 Table 2.28 Analysis of Moderate/Severe Asthma Exacerbations (ITT) ...... 226 Table 2.29 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >=3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24- Weeks Treatment Period (ITT) ...... 227 Table 2.30 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >=3 Points in ACT Total Score (ITT) ...... 228 Table 2.31 Time to First ACT Total Score >=20 or Improvement >= 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT) ...... 230 Table 2.32 Time to First ACT Total Score >=20 and Improvement >= 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT) ...... 231 SAFETY DATA SOURCE TABLES ...... 232 Table 3.01 Summary of Adverse Event (Safety Population) ...... 232 Table 3.02 Adverse Events of the Subjects (Safety Population) ...... 233 Table 3.03 Adverse Events Related to Study Process of the Subjects (Safety Population) ...... 248 Table 3.04 Adverse Events Related To GSK's Drugs of the Subjects (Safety Population) ...... 249 Table 3.05 Serious Adverse Events of the Subjects (Safety Population) ...... 250 Table 3.06 Serious Adverse Events Related to Study Process of the Subjects (Safety Population) ...... 254 Table 3.07 Serious Adverse Events Related To GSK's Drugs of the Subjects (Safety Population) ...... 255 Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) ...... 256 Table 3.09 Adverse Events Leading To Temporarily Suspend\Delay GSK's drugs (Safety Population) ...... 269 Table 3.10 Adverse Events Leading To Withdrawal (Safety Population) ...... 270 Table 3.11 Physical Examination of the subjects (Safety Population) ...... 271

7 CONFIDENTIAL 2019N423508_00 201097

ABBREVIATIONS

ACT Asthma Control Test AE Adverse Event ANCOVA Analysis of Covariance AQLQ Asthma Quality of Life Questionnaire AQLQ(S) Standardised Asthma Quality of Life Questionnaire CFR Code of Federal Regulations CI Confidence Intervals CRF Case Report Form CV Cardiovascular DRC Daily Record Card FEV1 Forced Expiratory Volume in one second GCP Good Clinical Practice GINA International Guidelines for Asthma Management GSK GlaxoSmithKline HR Hazard Ratio ICC Intra-cluster Correlation Coefficient ICH International Council for Harmonisation ICS Inhaled Corticosteroid IEC Independent Ethics Committee IRB Institutional Review Board ITT Intention-to-Treat KM Kaplan-Meier LABA Long-acting β2 agonist MMRM Mixed Model Repeated Measure OR Odds Ratios PEF Peak Expiratory Flow PP Per Protocol RAP Reporting and Analysis Plan SAE Serious Adverse Event SD Standard Deviation SOP Standard Operating Procedure WHO World Health Organisation

Trademark Information

Trademarks of the GlaxoSmithKline Trademarks not owned by the group of companies GlaxoSmithKline group of companies ACT AQLQ(S) Mini-Wright

   8 CONFIDENTIAL 2019N423508_00 201097

ETHICS AND GOOD CLINICAL PRACTICE

The study protocol, any amendments, the informed consent, and other information that required pre-approval were reviewed and approved by a national, regional, or investigational center ethics committee or institutional review board, in accordance with the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) Good Clinical Practice (GCP) and applicable country-specific requirements, including US 21 Code of Federal Regulations (CFR) 312.3(b) for constitution of independent ethics committees. Ethics committee or institutional review board approvals are maintained in the Sponsor’s study file.

Investigators were trained to conduct the study in accordance with GCPs and the study protocol, as defined in ICH E3, Section 9.6. Written commitments were obtained from investigators to conduct the study in accordance with ICH GCP and all applicable subject privacy requirements, and the ethical principles that are outlined in the Declaration of Helsinki 2016 and to conduct the study in accordance with the protocol.

The study was monitored in accordance with ICH E6, Section 5.18. If significant findings (e.g., potential serious misconduct or noncompliance with GCP, including potential Serious Breaches) were identified during monitoring or auditing of a site, these are presented in Section 5.2, Protocol Deviations, of this report.

Written informed consent was obtained from each subject prior to the performance of any study-specific procedures. The investigator agreed to provide the subject as much time as necessary to review the document, to inquire about details of the trial, and to decide whether or not to participate in the study. The informed consent was signed and dated by the study subject and by the person who conducted the informed consent discussion. Case report forms were provided for each subject’s data to be recorded.

   9 CONFIDENTIAL 2019N423508_00 201097

1. INTRODUCTION

The asthma control test (ACT) is a validated, short, easy to use, and self-administered instrument used to assess asthma control [Nathan, 2004]. However, physicians do not use it, even though it was designed to assess the level of asthma control in patients and so guide physicians treatment decisions. This study is aimed to evaluate the effectiveness of ACT guided treatment compared with usual care in asthma patients in China. It is designed to assist Chinese patients and physicians improving adherence to the guidelines through the inclusion of the ACT in the patient’s asthma management plan.

2. STUDY OBJECTIVE(S)

2.1. Primary objective

The primary objective of the study was to compare the effectiveness of ACT guided treatment versus usual care in achieving asthma control.

2.2. Secondary objectives

The secondary objective of the study was to compare the effectiveness of ACT guided treatment versus usual care in asthma subjects with respect to the following parameters:

• Symptoms

• Forced Expiratory Volume in one second (FEV1)

• Peak Expiratory Flow(PEF)

• The quality of life

• Time to asthma control

2.3. Exploratory objective

The exploratory objective of the study was to compare the effectiveness of ACT guided treatment versus usual care in asthma exacerbation.

3. INVESTIGATORS AND STUDY ADMINISTRATIVE STRUCTURE

There were 12 centers in China.

   10 CONFIDENTIAL 2019N423508_00 201097

4. INVESTIGATIONAL PLAN

4.1. Study Design

This was a prospective, multicentre, cluster-randomized, open-label study. Protocol waivers or exemptions were not allowed with the exception of immediate safety concerns. Therefore, adherence to the study design requirements, including those specified in the Time and Events Table, were essential and required for study conduct.

This study included baseline (week 0, visit 0) and a 24-week treatment period (week 0- week 24) (Figure 4.1).

A total of 528 subjects were required including a total of 12 centres (6 centres per group) and around 44 asthma subjects per centre.

No interim analysis was planned.

Figure 4.1 Overall Design

The study was planned to be a 24-week study.

At baseline (week 0, visit 0), subjects eligible for inclusion criteria signed informed consent form at Visit 0. At this visit, investigators collected baseline data, including demographic data (gender, age, height, weight, education level and smoking history), history of asthma (disease duration, disease severity and exacerbation history), and past and concurrent medical conditions and medication (both asthma and non-asthma). ACT, AQLQ(S) and lung function test were completed as the baseline measurement in sequence.

The treatment period was a 24 week treatment period. Subsequent scheduled visits occurred as follows: Week 4 (Visit 1), Week 8 (Visit 2), Week 12 (Visit 3), Week 16 (Visit 4), Week 20 (Visit 5) and Week 24 (Visit 6).

In this cluster-randomization design, each study centre, considered as a cluster, was randomized to either ACT guided treatment group or control group (usual care group). The randomization was stratified according to the hospital level (Tier 3 vs. Tier 2). The hospital level was determined in accordance with the uniform requirements of the

   11 CONFIDENTIAL 2019N423508_00 201097

government. And a Tier 3 hospital was on a larger scale and provided superior medical care than a Tier 2 hospital.

For the subjects who were recruited in the ACT centres, they were treated based on the ACT score (Table 1). For subjects who were recruited in the control centres, they were treated based on doctor’s subjective judgment.

During the treatment period, subjects filled in the paper diary record card (DRC) every morning and evening, recorded their morning and evening PEF, symptom and medication. At each visit, subjects completed the ACT at clinic. For the ACT guided group, the ACT was completed prior to any other assessments conducted. For the control group, subjects completed the ACT after investigator making the treatment decision. This ensured the investigators made treatment decision based on his/her clinical judgement, not based on ACT. For the control group, all subsequent assessments were conducted in the same order as the ACT guided group. Investigators were responsible for checking the subjects’ DRC, including symptom score and evaluating therapeutic compliance.

Table 1 ACT-guided treatment

ACT score Treatment adjustment

ACT=25, ≥3 months Step-down treatment ACT≥20, <25 or ACT=25, <3 months No change ACT≤19 Step-up treatment 4.2. Discussion of Study Design

The study was designed to address the effectiveness of the intervention if implemented in a system without other enhanced healthcare options; therefore a usual care control group was used.

Randomization by individual subjects might result the contamination due to the same investigator looking after both intervention and control group. Randomization by clusters (i.e., study centre) was used to reduce such bias.

4.3. Protocol Amendment(s)

There were none protocol amendment.

4.4. Selection of Study Population

Deviations from inclusion and exclusion criteria were not allowed because they can potentially jeopardize the scientific integrity of the study, regulatory acceptability or subject safety. Therefore, adherence to the criteria as specified in the protocol was essential.

4.4.1. Inclusion/Exclusion Criteria

The inclusion criteria was for study centre (clusters) and individual subject at visit 0.

   12 CONFIDENTIAL 2019N423508_00 201097

For study centre (clusters), it should be general practice, located in Shanghai and provide asthma care.

For subjects, their legally authorized representative had to provide signed written informed consent prior to study participation, were male or female and ≥18 to ≤70 years of age and had an ACT score <20 at Visit 0, were documented clinical history of asthma for at least 6 months and using inhaled corticosteroids (ICS) alone or combined with an inhaled long-acting β2 agonist (LABA) treatment within 1 year prior to Visit 0, and at Visit 0, a demonstrable reversible increase in FEV1 of at least 12% (and ≥200 mL), 15 minutes after inhaling a short-acting bronchodilator; at any time in the last 2 years documentary evidence of a reversible increase in FEV1 of at least 12% (and ≥200 mL) 15 mins after inhaling a short-acting bronchodilator; or demonstrable reversible increase in morning PEF of at least 15% (and ≥200 mL) either spontaneously or after inhalation of a short-acting bronchodilator; subjects must have been able to read, comprehend, and record information in Chinese.

Exclusion criteria included historical or current evidence of clinically significant uncontrolled disease of asthma or other body system. A subject was not be eligible for inclusion in this study if the subject had life-threatening, severe and unstable asthma, or the subject had clinically significant uncontrolled medical condition or disease, or if the subject was heavy smoker or had the history of alcohol / medication abuse or upper / lower respiratory tract infection or adverse reaction, or if the subject had other uncontrolled reason which might lead not to complete the study on plan.

4.4.2. Withdrawal Criteria

A subject might withdraw from study treatment at any time at his/her own request, or might be withdrawn at any time at the discretion of the investigator for safety, behavioural or administrative reasons.

The reason(s) for subjects not completing the study was recorded in the Case Report Form (CRF), and that the investigator documented if applicable, the reason (if specified by the subject) for withdrawal of consent. The end of study/early discontinuation procedures were performed at the time of discontinuation from the clinical study.

4.5. Treatments

4.5.1. Investigational Product(s) and Reference Therapy

The investigational product was ACT guided treatment to achieve asthma control.

4.5.2. Treatment Assignment

The study was a trial with the study centre as the unit of randomization. Computer generated randomization schedule for centres was provided by GlaxoSmithKline (GSK) or its designee. Once a randomization number had been assigned, it was not reassigned. Each centre was randomized to either ACT guided treatment group or control group

   13 CONFIDENTIAL 2019N423508_00 201097

(usual care group). The randomization was stratified according to the Tier of the hospitals (Tier 3 vs. Tier 2). Assignment to a treatment didn’t occur, as no treatment intervention was provided for this study. Each centre was asked to recruit 44 consecutive patients with asthma attending the practice.

Subjects were assigned to study treatment in accordance with the randomization schedule. For the subjects who were recruited in the ACT centres, they were treated based on the ACT score. For subjects who were recruited in the control centres, they were treated based on doctor’s subjective judgment.

4.5.3. Blinding

This was an open-label study.

4.5.4. Prior and Concomitant Medications and Non-Drug Therapies

There wasn’t permitted medication or procedure in the study. The prior and concomitant medications and non-drug therapies were documented in the CRF.

4.5.5. Compliance

During the treatment period, subjects filled in the paper diary record card every morning and evening.

Subjects’ treatment compliance was calculated by the records of diary cards.

4.6. Study Assessments and Procedures

This section lists the procedures and parameters of each planned study assessment. The exact timing of each assessment is listed in Table 2.

Table 2 Time and Events Table

Visit 6 Visit Visit Visit Visit Visit Visit ACTIVITY ( or Early 0 1 2 3 4 5 Withdrawal) Study Week (Visit 0 4w±1 8w±1 12w±1 16w±1 20w±1 24w±1 Window) Informed consent x Randomization/Allocation x of clusters Subject Demography x Medical history x Verification of inclusion / x exclusion criteria Efficacy Evaluation Verification of withdrawal x x x x x x criteria

   14 CONFIDENTIAL 2019N423508_00 201097

Visit 6 Visit Visit Visit Visit Visit Visit ACTIVITY ( or Early 0 1 2 3 4 5 Withdrawal) Study Week (Visit 0 4w±1 8w±1 12w±1 16w±1 20w±1 24w±1 Window) Issue diary record cards x x x x x x Collect/Review diary x x x x x x record cards Compliance assessment x x x x x x Asthma control Test (ACT) x x x x x x x score AQLQ(S) x x Lung function Test x x Safety Evaluation Adverse event assessment1 x x x x x x x 1. Serious AEs were recorded from the time the consent form was signed until the follow-up visit. All AEs were recorded from the start of study treatment until the follow-up visit.

4.6.1. Screening and Critical Baseline Assessments

All demographic and baseline assessments were completed according to the Time and Events schedule in Table 2 and listed in Table 3.

Table 3 Screening and Critical Baseline Assessments

Baseline Assessments Demographic Gender information ethnic origin date of birth, height weight education level smoking history History of asthma including duration of asthma History of asthma exacerbations in the 12 months prior to Visit 0 Concurrent medical conditions and concurrent medication Medical history Use of corticosteroids for asthma in the 12 months prior to Visit 0 duration of inhaled corticosteroid treatment

The following demographic parameters were captured: year of birth, sex, race and ethnicity.

Medical/medication was assessed as related to the inclusion/exclusion criteria listed in Section 4.4.

   15 CONFIDENTIAL 2019N423508_00 201097

4.6.2. Efficacy Assessment

4.6.2.1. Primary Endpoint (s)

The primary endpoint for the study was the percentage of subjects who had an ACT score ≥20 or an improvement of more than 3 points in ACT during the 24-week treatment period.

Asthma control test (ACT)

ACT was a self-administered questionnaire comprising five questions. Each of the five items in the ACT was assessed on a five point scale and the scores were summed to give a total score ranging from 5 to 25, with a score of ≥ 20 denoting ‘well-controlled asthma’, a score of 16-19 denoting ‘not well-controlled asthma’, and a score of ≤ 15 denoting ‘very poorly controlled asthma’. The recall period of the questionnaire was four weeks. ACT might be completed by the subject at each visit. For the control group, subjects completed the ACT after investigator making the treatment decision. Differences of 3 points in mean ACT scores between 2 groups or over time in an individual patient were clinically significant. An improvement of more than 3 points in ACT from baseline also served as primary endpoint.

4.6.2.2. Secondary Endpoint (s)

• Mean daytime symptom score over the 24-week treatment period

• Mean night-time symptom score over the 24-week treatment period

• Mean change from baseline to the end of study in FEV1

• Mean morning(AM) PEF over the 24-week treatment period

• Mean evening(PM) PEF over the 24-week treatment period

• Mean change from baseline to the end of study in AQLQ(S) score • Time to first ACT score ≥20 or an improvement of more than 3 points in ACT over the 24-week treatment period • Time to first ACT total score ≥20 and an improvement of more than 3 points in ACT total score over the 24-week treatment period

   16 CONFIDENTIAL 2019N423508_00 201097

Asthma symptom score

Subjects were instructed to record daytime and night-time asthma symptom scores prior to taking the morning or evening dose of study medication or rescue medication. Any asthma-related symptoms, such as wheeze, shortness of breath, cough or chest tightness experienced during the previous 12 hours were rated as follows. The following was recorded on DRC.

Daytime symptom score

0 = No symptoms during the day

1 = Symptoms for one short period during the day

2 = Symptoms for two or more short periods during the day

3 = Symptoms for most of the day which did not affect my daily activities

4 = Symptoms for most of the day which did affect my normal daily activities

5 = Symptoms so severe that I could not go to work or perform normal daily activities

Night-time symptom score

0 = No symptoms during the night

1 = Symptoms causing me to wake once or wake early

2 = Symptoms causing me to wake twice or more (including waking early)

3 = Symptoms causing me to be awake for most of the night

4 = Symptoms so severe that I did not sleep at all

FEV1

FEV1 was measured at visit 0 and 6. Subjects refrained from using short-acting bronchodilators for at least 6 hours prior to performing FEV1 measurements. The highest of three technically acceptable measurements of FEV1 was taken.

Peak Expiratory Flow (PEF)

At Visit 0, the subject was given a mini-Wright Peak Flow Meter and taught how to measure and record their PEF. At each visit the subjects’ ability to measure their PEF was checked and, if the investigator had any concerns, the subject was retrained. The instructions for use was provided with each Peak Flow Meter. PEF was measured while subject was in the sitting position.

Subjects recorded on diary card the best of three PEF measurements, using a mini-Wright peak flow meter in the morning (7:00-10:00 AM) and evening (6:00-9:00 PM) before

   17 CONFIDENTIAL 2019N423508_00 201097

taking any asthma drug. Bronchodilator therapy (e.g. salbutamol) was withheld, where possible, for 4 hours before recording PEF. Otherwise, the use of salbutamol was recorded on diary card.

GSK provided each centre with a supply of standard mini-Wright peak flow meters.

   18 CONFIDENTIAL 2019N423508_00 201097

Standardised Asthma Quality of Life Questionnaire (AQLQ(S))

The Asthma Quality of Life Questionnaire (AQLQ) was a self-administered questionnaire consists of 32 questions to measure the functional problem that were most troublesome to patients with asthma. There were 32 questions in 4 domains (symptoms, activity limitation, emotional function and environmental stimuli). Patients were asked to think about how they had been during the previous two weeks and to respond to each of the 32 questions on a 7-point scale (7 = not impaired at all - 1 = severely impaired). A change in score of greater than 0.5 can be considered clinically important. In the present study, the AQLQ(S) was used with 5 standardised activities to replace the patient-specific ones in the AQLQ. These five generic activities incorporated the activities that were most frequently chosen by patients in studies in which the original AQLQ was used; Strenuous activities, Moderate activities, Social activities, Work-related activities, and Sleeping. At visit 1 and 7, the investigator provided each subject with the AQLQ(S).

ACT, AQLQ(S) and lung function test were completed in sequence at Visit 0 and 6.

4.6.2.3. Other Endpoint (s)

Rate of moderate/severe asthma exacerbation over the 24-week treatment period.

Asthma Exacerbations

Subjects who experienced worsening of symptoms should:

 increase relief medication usage for relief of symptoms

 contact the investigator or primary physician immediately and report to the clinic as soon as possible (ideally within 24-hours)

 record their symptoms, PEF and relief medication usage in their DRC, as previously instructed

 if intervention therapy was required during a treatment period, subjects might receive oral corticosteroids (40-60 mg prednisolone daily, or equivalent, for 10 days), over and above their usual asthma medication.

Exacerbations were assessed by the physician at each scheduled visit by reviewing the DRC, as well as specific questioning on AEs.

Exacerbations were defined based on one or more of the following characteristics:

Definition of asthma exacerbation(s)

Moderate

A moderate asthma exacerbation was defined as a deterioration in asthma requiring treatment with an oral corticosteroid.

   19 CONFIDENTIAL 2019N423508_00 201097

Individual courses of oral corticosteroids were classified as separate exacerbations only if they were administered >1 week apart. Any course started within one week of finishing the previous course was considered part of the previous exacerbation.

Severe

A severe asthma exacerbation was defined as a deterioration in asthma which required hospital admission.

Time/date of resolution

Time/date at which the exacerbation had resolved, in the opinion of the investigator and/or subject.

Details of exacerbations were collected in the Asthma Exacerbations page of the CRF and DRC as follows at the visit immediately following the exacerbation:

Date of onset and resolution, daily morning PEF during the exacerbation; details of management of the exacerbation (e.g. clinic or emergency room visit, self-managed, hospital admission etc.); restriction/prevention from continuing usual activities. An exacerbation resulting in hospitalization was recorded as a serious adverse event (SAE) in the Serious Adverse Event pages of the CRF. However, it was not only these exacerbations resulting in hospitalization which required collection as SAEs, but those that met any of the definitions of seriousness, listed in Protocol Section 7.4.

4.6.3. Safety Assessments

The safety assessments were the monitoring of adverse events (AEs), cardiovascular (CV) and death events, pregnancy, physical exams, vital signs.

4.6.3.1. Adverse Events (AE)

The investigator or site staff were responsible for detecting, documenting and reporting events that met the definition of an AE or SAE in Protocol Appendix 2. AE information volunteered by the subject, discovered by investigator questioning or detected by other means was collected from the start of study treatment until the follow-up contact. The following information on AEs was obtained:

 Duration (start and stop dates).  Severity (mild, moderate, severe).  Causality (reasonable possibility yes/no).  Actions taken and outcome.

4.6.3.2. Pregnancy

Any pregnancy exposure to GSK products that was notified during study participation was reported using a pregnancy notification form.

   20 CONFIDENTIAL 2019N423508_00 201097

4.6.3.3. Physical Exams

A brief physical examination included, at a minimum assessments of the lungs, cardiovascular system, and nose.

4.6.3.4. Vital Signs

Vital signs were measured in semi-supine position after 5 minutes rest and included temperature, systolic and diastolic blood pressure and pulse rate.

4.7. Data Quality Assurance

Subject data was entered into GSK-defined eCRFs, transmitted electronically to GSK and combined with data provided from other sources (e.g., laboratory data) in a validated data system.

Verification of data accuracy and adherence to protocol requirements was achieved through regular monitoring visits at each investigational site. Subsequent data handling and reporting processes were performed according to processes detailed in GSK’s Standard Operating Procedures (SOPs). AEs and concomitant medications were coded using company standard dictionaries, Medical Dictionary for Regulatory Activities (MedDRA) and GSKDrug.

SAE, consistent with the data collected for other AEs, was entered into the database and quality assured, including reconciliation with the GCSP database.

All investigators and responsible study site staff attended an investigator training meeting and/or separate study site initiation visit to review study protocol procedures, study requirements, and GCP responsibilities. Investigators and staff were given opportunity to discuss any aspect of the study protocol and GCP requirements. Training records were reviewed to ensure investigators and staff were qualified to conduct the study and to document training in GCP. Any staff lacking in GCP training were either sent to a GCP training course or provided an electronic GCP training module. Documentation of GCP training was confirmed prior to staff participation in the study.

Principal investigators signed the investigator page of the protocol to confirm their commitment to conduct the study in accord with the protocol and GCP. The signed documents have been archived within individual investigator study files.

In accordance with applicable regulations, GCP and GSK procedures, GSK monitors contacted the site prior to the start of the study to review with the site staff the protocol, study requirements, and their responsibilities to satisfy regulatory, ethical, and GSK requirements. When reviewing data collection procedures, the discussion included identification, agreement and documentation of data items for which the CRF served as the source document. GSK monitored the study and site activity to verify that: (1) the data are authentic, accurate, and complete; (2) the safety and rights of subjects were protected; (3) the study was conducted in accordance with the currently approved

   21 CONFIDENTIAL 2019N423508_00 201097

protocol and any other study agreements, GCP and all applicable regulatory requirements.

All protocol deviations collected during the study were reviewed by the GSK study team in order to identify important protocol deviations. Consistent with ICH E3 guidance, only protocol deviations identified as “important” are provided in this clinical study report. Important deviations are defined as deviations that were likely to affect the interpretation of the results and/or led to exclusion of any subject data from an analysis. Important deviations include, but are not limited to, those related to study inclusion or exclusion criteria, adherence to the protocol, conduct of the study, subject management or subject assessment.

4.8. Statistical Analyses

4.8.1. Statistical Hypotheses

The null hypothesis for this study was that there was no difference between ACT guided treatment and usual care in the percentage of subjects who had an ACT score ≥20 or an improvement of more than 3 points in ACT during the 24-week treatment period for the intention-to-treat (ITT) population.

The alternative hypothesis for this study was that there was a difference between ACT guided treatment and usual care in the percentage of subjects who had an ACT score ≥20 or an improvement of more than 3 points in ACT during the 24-week treatment period for the ITT population.

The study was designed to show superiority of ACT guided treatment over usual care.

4.8.2. Sample Size Considerations

The study was powered to demonstrate superiority of ACT guided treatment over usual care from baseline over Week 24 in the primary endpoints.

We assumed an intra-cluster correlation coefficient (ICC) of ρ=0.01, 29% well-control rate for usual care group, and 44% for ACT guided treatment group. It was required to have 6 centres for each treatment group and a minimum of 35 subjects for each centre (a total of 420 evaluable subjects) to achieve a power of 80% with two-sided α=0.05. We anticipated a dropout rate of 20% and therefore planned to recruit 44 subjects in each centre (a total of 528 randomized subjects). These sample size calculations were produced based on Donner [Chapman K.R., 1981].

An ACT score ≥20 indicated controlled asthma. The assumed rate of subjects with an ACT score ≥20 for usual care group was based on a nationwide epidemiologic survey in China [Nan Su, 2013], which reported a 29% asthma control rate. The rate for ACT guided treatment was assumed based on GOAL study [Bateman, 2004], which reported 80% rate for salmeterol/fluticasone group. An additional adjustment had been made to reduce it to 44% to account for real life practice. A 15% difference between ACT guided treatment group and usual care group was considered clinically meaningful.

   22 CONFIDENTIAL 2019N423508_00 201097

Study centres were randomly assigned to the 2 groups using a computer generated randomization list. The randomization was stratified according to the Tier of the hospitals (Tier 3 vs. Tier 2).

4.8.3. Analysis Populations

Three populations were defined for this study.

Intent-to-Treat (ITT) Population: all subjects who signed informed consent form, were randomised and who had at least one post-baseline assessment. The ITT population was the primary population of interest.

Per Protocol (PP) Population: all subjects in the ITT population with no major protocol deviations and with ≥80% treatment compliance and diary compliance. The PP population wasn’t analyzed if this population comprised more than 95% or less than 50% of the ITT population. Only the primary efficacy variables were analyzed using the PP population.

Safety Population: all subjects who were enrolled into the study and who had at least one DRC assessment.

4.8.3.1. Protocol Deviations

Important deviations, which resulted in exclusion from the analysis population, were summarized and listed. Protocol deviations were tracked by the study team throughout the conduct of the study in accordance with the Protocol Deviation Management Plan (refer to RAP Appendix 1).

 Data was reviewed prior to freezing the database to ensure all important deviations and deviations which may lead to exclusion from the analysis were captured and categorised on the protocol deviations dataset.  This dataset was the basis for the summaries and listings of protocol deviations. A separate summary and listing of important protocol deviations were provided.

4.8.3.2. Diary Compliance

Subjects’ diary compliance was assessed by the records of diary cards. For the subjects to be included in the PP population, there was at least ≥80% expected diary counts in the diary cards according to the visits they had completed.

4.8.3.3. Treatment Compliance

Subjects’ treatment compliance was calculated as follow:

Treatment compliance=

   23 CONFIDENTIAL 2019N423508_00 201097

The number of days that the subjects took asthma drugs following the doctor’s advice was captured in the diary cards. For the subjects to be included in the PP population, the treatment compliance was ≥80%.

4.8.4. Interim Analyses

There were no interim analyses planned for this study.

4.8.5. Final Analyses

The final planned analyses was performed after all required database cleaning activities had been completed and final database release (DBR) and database freeze (DBF) had been declared by Data Management.

Unspecified, the summary tables were displayed using the descriptors as specified: ACT, UC and Total.

4.8.5.1. Study Population Analyses

The study population analyses were based on the ITT population.

Study population analyses included analyses of subject’s disposition, population, protocol deviations, demographic and baseline characteristics, subject’s smoking history, alcoholism history, drug abuse history, asthma history, asthma exacerbation history, medical history (coded by MedDRA version 21.0 or above), asthma medications, and concomitant medications (coded by WHODrug version Global Mar 2018 or above).

Subject disposition, including the number of subjects who were screened, screening failures and screening succeed/randomization were summarised and listed. Reasons for early study withdrawal, were also summarized and listed by subject.

The number and percentage of subjects by centre and the number of subjects were summarized.

The number and percentage of subjects at each analysis population were summarized for all screening / randomized subjects. The distribution of each analysis population and the reason of being excluded from ITT and PP were listed by subject.

The number and percentage of subjects with important protocol deviations were summarized for all screening / randomized subjects. The category and description of important protocol deviation were listed by subject.

Demographics (age, gender, pregnancy possibility, height, weight, body mass index [BMI], and race) and other baseline characteristics (temperature, pulse rate, systolic blood pressure, diastolic blood pressure) were summarized and listed by subject.

The history of smoking, alcohol abuse, medication abuse, asthma, and asthma exacerbation were summarized and listed by subject.

   24 CONFIDENTIAL 2019N423508_00 201097

The past and current illness were summarized and listed by subject.

The asthma medication and concomitant medications were summarized separately and also listed separately by subject.

4.8.5.2. Efficacy Analyses

4.8.5.2.1. Primary Efficacy Analyses

The primary endpoint for the study was the percentage of subjects who had an ACT total score ≥20 or an improvement of more than 3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period.

The calculation of percentage was based on the subjects with at least one post-baseline ACT assessment.

Primary Analysis Methods

The primary endpoint was analysed using mixed effect logistic regression. In this study, generalized linear mixed model (GLMM) was used to control for centre as a random effect since each centre was considered as a cluster in this cluster-randomization study. The statistical model on which the inference was based included terms for treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age. Centre, as the unit of randomization, was treated as a random effect. Type of baseline controller (ICS alone vs. ICS/LABA) derived from the current asthma medication record in visit 0. The results of the primary analyses were presented as adjusted odds ratios, 2-sided 95% confidence intervals (CI), and associated p-values comparing ACT guided treatment with usual care. The adjusted odds ratios and 95% confidence intervals were presented by a forest plot.

Interactions between treatment and each of the covariates were investigated in turn, with all main effects in the model regardless of their statistical significance. Any interaction terms found to be statistically significant was explored and if necessary results were reported for each level of the covariate. Investigation of interactions was confined to the primary endpoints. The effect of interactions (e.g., treatment by centre) was assessed at the 10% level of significance.

Analyses of the primary endpoints was performed on the ITT population and the PP population; the ITT population was considered primary. The PP population wasn’t analysed if this population comprised more than 95% or less than 50% of the ITT population. Only the primary efficacy variables was analysed using the PP population.

Sensitivity and Supportive Analysis Methods

 The primary endpoint was repeatedly analysed in the following subgroups, using the same method used in primary analysis method:  Type of baseline controller (inhaled corticosteroids [ICS] alone vs. ICS/long- acting β2 agonist [LABA])

   25 CONFIDENTIAL 2019N423508_00 201097

 Gender (male vs. female)  Age (<50 years old vs. ≥50 years old)

For subgroup analyses, the model included all the covariates used in the primary analysis model except for the subgroup variable itself. Adjusted odds ratios and their 95% confidence intervals were presented by a forest plot. These analyses were performed for the ITT and PP populations.

 A supportive analysis of the primary endpoint was performed using Generalised Estimating Equations (GEEs) if the model can converge.

All non-missing data from visits with scheduled ACT assessments (Weeks 4, 8, 12, 16, 20 and 24) were included and an unstructured working correlation matrix was implemented. The model included terms for treatment, baseline ACT total score, baseline ACT total score squared, age, gender, visit and treatment-by-visit interaction. The number and percentage of subjects with a response within each treatment group were presented by visit, together with the adjusted odds ratios (OR), 2-sided 95% CI, and associated p-values between ACT guided treatment and usual care. The adjusted odds ratios and 95% CI were presented by a forest plot.

This analysis was performed for the ITT and PP populations.

 Percentage of subjects who had an ACT total score ≥20 and an improvement of more than 3 points in ACT total score.

The following endpoints were analysed as post hoc sensitivity analyses.

 Percentage of subjects who had an ACT total score ≥20 and an improvement of more than 3 points in ACT total score at Week 12. Subjects with missing data at Week 12 were set as non-responders.  Percentage of subjects who had an ACT total score ≥20 and an improvement of more than 3 points in ACT total score at Week 24. Subjects with missing data at Week 24 were set as non-responders.  Percentage of subjects who had an ACT total score ≥20 and an improvement of more than 3 points in ACT total score at Weeks 16 and 20 and 24. Subjects with missing data at Weeks 16, 20 or 24 were set as non-responders.

These endpoints were analysed using the same model as defined for primary endpoint (GLMM). A supportive analysis was conducted to analysis the percentage of subjects who had an ACT total score≥20 and an improvement of more than 3 points in ACT total score, using the same GEEs method as mentioned above. All non-missing data from visits with scheduled ACT assessments (Weeks 4, 8, 12, 16, 20 and 24) were used.

These analyses were performed for the ITT and PP populations.

 Change from baseline in ACT total score in post-baseline visits.

A Mixed Model Repeated Measures (MMRM) was used with covariates of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA),

   26 CONFIDENTIAL 2019N423508_00 201097

gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect. The least squares (LS) mean and LS mean change from baseline for each treatment and the estimated LS mean change from baseline treatment difference were presented at each visit together with the 95% CI for the mean difference and P-value.

This analysis was performed for the ITT and PP populations.

4.8.5.2.2. Secondary Efficacy Analyses

The following secondary efficacy endpoints were analysed.

 Mean daytime symptom score over the 24-week treatment period  Mean night-time symptom score over the 24-week treatment period  Mean change from baseline to the end of study in FEV1  Mean morning (AM) PEF over the 24-week treatment period  Mean evening (PM) PEF over the 24-week treatment period  Mean change from baseline to the end of study in AQLQ(S) score  Time to first ACT total score ≥20 or an improvement of more than 3 points in ACT total score over the 24-week treatment period  Time to first ACT total score ≥20 and an improvement of more than 3 points in ACT total score over the 24-week treatment period

Analysis Methods

All secondary endpoints were confined to the ITT population. The continuous efficacy endpoints were summarized by mean, standard deviation (SD), median, minimum and maximum according to scheduled visits. Time to event endpoints were summarized by median time, 95% confidence interval of median time, lower and upper quantiles and range.

Continuous efficacy endpoints were analysed using analysis of covariance (ANCOVA) and adjusted for centre as a random effect.

For mean daytime and night-time symptom score and mean morning and evening PEF, the mean values were based on the available data on the diary card over each visit assessment period. No imputations were performed on missing data. The mean values were considered missing if less than 14 days were recorded in each visit assessment period. A Mixed Model Repeated Measures (MMRM) was used with covariates of treatment, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor. The results were presented as LS mean (SE), 95% CI for LS mean, treatment difference, 95% CI for treatment difference and P-value by visit.

The AQLQ(S) contained 32 items in four domains: activity limitation (11 items), symptoms (12 items), emotional function (5 items) and environmental stimuli (4 items). The following items are included in each of the 4 domains:

 Symptoms: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 29, 30

   27 CONFIDENTIAL 2019N423508_00 201097

 Activity Limitation: 1, 2, 3, 4, 5, 11, 19, 25, 28, 31, 32  Emotional Function: 7, 13, 15, 21, 27  Environmental Stimuli: 9, 17, 23, 26

The response format consisted of a seven-point scale where a value of 1 indicated “total impairment” and a value of 7 indicates “no impairment”. The total AQLQ(S) score was the mean of all 32 items in the questionnaire and each individual domain score was calculated as the mean of the items within that domain. Hence, the total and domain scores were also each defined on a range from 1 to 7 with higher scores indicating a higher quality of life.

For the total AQLQ(S) score, the score for a subject at any time point was only calculated if at least 90% of the questions were answered (calculated as the mean of those non- missing questions). If fewer than 90% of the questions were answered then the mean score for that subject at that time point was considered missing.

For each individual domain of the AQLQ(S) score, the score for a subject at any time point was only calculated if at least 90% of the questions for that domain were answered (calculated as the mean of those non-missing questions). If fewer than 90% of the questions were answered for that domain then the mean score for that subject and domain at that time point was considered missing.

The analysis of changing from baseline to the end of study in FEV1 and AQLQ(S) score only included those who completed the study and change from baseline was calculated as the difference between baseline visit and Visit 6.

For FEV1 and AQLQ(S) changing from baseline, a mixed model was used with covariates of treatment, centre, baseline FEV1/AQLQ(s), type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor. The results were presented as LS mean (SE), 95% confidence interval for LS mean, treatment difference, 95% confidence interval for treatment difference.

For time to event endpoint, Cox regression analysis was performed with treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age as the adjusted factors. The results were presented as hazard ratio (HR), 95% CI and P-value. Log-rank test and Kaplan-Meier (KM) methodology were used to estimate median time for each treatment arm. Kaplan-Meier curves were constructed to provide a visual description of the difference between the two treatment arms.

4.8.5.2.3. Exploratory Efficacy Analyses

Moderate/severe asthma exacerbations were summarized, including number of subjects with moderate/severe exacerbations, number of moderate/severe exacerbations and characteristics of moderate/severe exacerbations. Exacerbations separated by less than 7 days were treated as a continuation of the same exacerbation.

The moderate/severe exacerbation rate was analyzed using a generalized linear model, assuming the number of exacerbations had a negative binomial probability distribution

   28 CONFIDENTIAL 2019N423508_00 201097

and that its mean was related to covariate factors with a ‘log link’ function. The logarithm of time (year) on study was used as an offset variable. All moderate/severe exacerbations observed during the study, with the onset date of event from date of baseline visit to the end of study visit date or early withdrawal (from study) visit date, were included in the analysis. Missing data due to early withdrawal from the study weren’t imputed. The model included covariates for treatment, baseline ACT total score, type of baseline controller (ICS alone vs. ICS/LABA), gender and age. The annual moderate/severe exacerbation rates and 95% CI for each treatment were presented. The rate ratio and 95% CI, percent reduction in rate and 95% CI and P-value were also presented.

4.8.5.3. Safety Analyses

The safety analyses were based on the Safety population, unless otherwise specified.

4.8.5.3.1. Adverse Events Analyses

The number (%) of subjects in each treatment group with treatment emergent AEs was produced. Counts and percentages were also presented of subjects with serious adverse events (SAEs), AEs leading to withdrawal, AEs by severity, AEs by relationship to study procedure/study drug, asthma exacerbation and death.

Adverse events were coded by MedDRA version 21.0 or above. Adverse events were summarized and presented by system organ class (SOC) and preferred term (PT) by treatment group. Adverse events analyses including the analysis of AEs, SAEs, non- serious AE and other significant AEs were summarized. All AEs and SAEs were listed.

4.8.5.3.2. Adverse Events of Special Interest Analyses

Except for study indication Asthma, there was no other adverse event of special interest.

4.8.5.3.3. Other Safety Analyses

The analyses of safety test results including vital signs and physical examination were summarized and listed by descriptive statistics.

4.8.6. Changes in Conduct of the Study or Planned Analyses

There were no major changes or deviations to the originally planned statistical analysis specified in the protocol, dated 27-Jun-2016.

After unblinded analyses had been performed on interim data, some post-hoc sensitivity and supportive analyses were added for the primary efficacy endpoint in the RAP (See details in RAP Section 7.1.3).

Some additional post-hoc efficacy analyses were added after DBF, the endpoints were listed following:

   29 CONFIDENTIAL 2019N423508_00 201097

 Percentage of subjects who had an ACT score ≥20 or an improvement ≥3 points in ACT during the 24-week treatment period.

 Percentage of subjects who had an ACT score ≥20 and an improvement ≥3 points in ACT during the 24-week treatment period.

 Time to first ACT score ≥20 or an improvement ≥ 3 points in ACT over the 24- week treatment period

 Time to first ACT total score ≥20 and an improvement ≥3 points in ACT total score over the 24-week treatment period

For the analysis of percentage of subjects who had an ACT score ≥20 or/and an improvement ≥3 points in ACT were same as percentage of subjects who had an ACT score ≥20 or/and an improvement >3 points in ACT.

For the analysis of time to first ACT score ≥20 or/and an improvement ≥3 points in ACT were same as these time to first ACT score ≥20 or/and an improvement >3 points in ACT.

For analysis of the change from baseline in FEV1 and the change from baseline in AQLQ(S), the covariates of baseline ACT total score was replaced by baseline FEV1 and baseline AQLQ(S) respectively in the mixed model after RAP was final.

   30 CONFIDENTIAL 2019N423508_00 201097

5. STUDY POPULATION RESULTS

5.1. Subject Disposition

A total of 12 centres (Source Table 1.02) in China enrolled subjects for this multicenter study.

A total of 583 subjects were screened for this study; 53 subjects were screen failure (Source Table 1.01).

Five hundred thirty subjects were screen succeed and randomized of which 83.6% of subjects completed the study (Table 4) with 209 (82.9%) subjects in the ACT group and 234 subjects (84.2%) in the UC group. The primary reasons for subject withdrawal were subject’s decision (26 subjects, 29.9%), significantly non-compliant (23 subjects, 26.4%) and experienced the protocol-defined asthma exacerbation (16 subjects, 18.4%). More subjects in the ACT group (13 subjects, 30.2%) withdrew due to significantly non- compliant than that in UC group (10 subjects, 22.7%). Listing 1.01 provided the listing of subjects’ distributions.

Table 4 Subject Enrollment and Study Completion

ACT UC Total Screening 583 Screening failure 53 Screening succeed/randomization 252 278 530

Complete the study, n (%) 209 (82.9%) 234 (84.2%) 443 (83.6%) Withdrawal, the reasons are as 43 (17.1%) 44 (15.8%) 87 (16.4%) follows, n (%) [1] A subject is significantly non-compliant 13 (30.2%) 10 (22.7%) 23 (26.4%) with treatment or the requirments of the protocol Lost to follow up 3 (7.0%) 6 (13.6%) 9 (10.3%) Subject's decision not to participate any 11 (25.6%) 15 (34.1%) 26 (29.9%) further (withdrawal of consent) In the investigator's opinion, it is in the 2 (4.7%) 0 2 (2.3%) subject's best interest Subjects experience a protocol-defined 8 (18.6%) 8 (18.2%) 16 (18.4%) asthma exacerbation The study is terminated by the Sponsor 0 0 0 or designee Pregnancy 0 0 0 Others 6 (14.0%) 5 (11.4%) 11 (12.6%) Source: Table 1.01 [1] Percentages for withdrawal reasons are based on subjects withdrawal.

   31 CONFIDENTIAL 2019N423508_00 201097

5.2. Protocol Deviations

At the time of this report, no GCP noncompliance issues were identified by monitoring or audit.

At least one important protocol deviation was reported for 117 subjects (ACT: 27.0%; UC: 17.6%). These reported deviations included eligibility criteria (34 subjects), visit completion (38 subjects), assessment or time point completion (26 subjects), wrong study treatment / administration / dose (41 subjects), and others.

The most common important protocol deviations were wrong study treatment / administration / dose (ACT: 13.9%; UC: 2.2%).

The important protocol deviations in ACT group were more than those in UC group in this study. Table 5 shows the incidence of important protocol deviations for screening/randomized subjects. Source Listing 1.03 displays the listing of important protocol deviation.

Table 5 Summary of Important Protocol Deviations (Screening/Randomized Subjects)

Number (%) of Subjects Protocol Deviation ACT UC Total N=252 N=278 N=530 At least one Important Protocol deviation 68 (27.0%) 49 (17.6%) 117 (22.1%) 1 Informed Consent 2 (0.8%) 0 2 (0.4%) 2 Eligibility Criteria 15 (6.0%) 19 (6.8%) 34 (6.4%) 3 Withdrawal Criteria 2 (0.8%) 4 (1.4%) 6 (1.1%) 5 Visit Completion 24 (9.5%) 14 (5.0%) 38 (7.2%) 6 SAE/Pregnancy/Liver Events 1 (0.4%) 1 (0.4%) 2 (0.4%) 7 Assessment or Time Point completion 15 (6.0%) 11 (4.0%) 26 (4.9%) 8 Wrong Study Treatment/Administration/Dose 35 (13.9%) 6 (2.2%) 41 (7.7%) 9 Study Procedure 4 (1.6%) 0 4 (0.8%) Source: Table 1.04 Subjects can have more than one important protocol deviation.

5.3. Populations Analyzed

A total of 507 subjects had at least one post-baseline assessment on primary endpoints, second endpoints, and/or exploratory endpoints, and comprised the ITT Population. The ITT Population was the primary population used for study population and efficacy summaries. 439 subjects in the ITT population with no major protocol deviations and with ≥80% treatment compliance and diary compliance comprised Per-Protocol (PP) population. 506 subjects comprised safety population who were enrolled into the study and had at least one DRC assessment. A summary of all populations analyzed in this study is shown in Table 6. Source Listing 1.02 displays the listing of subjects’ analysis population.

   32 CONFIDENTIAL 2019N423508_00 201097

Table 6 Populations Analyzed (Screening/Randomized Subject)

Number (%) of Subjects Population ACT UC Total N=252 N=278 N=530 Intent-to-Treat (ITT) Population 242 (96.0%) 265 (95.3%) 507 (95.7%) Per Protocol (PP) Population 208 (82.5%) 231 (83.1%) 439 (82.8%) Safety Population 241 (95.6%) 265 (95.3%) 506 (95.5%) Source: Table 1.03 Note: Percentage is calculated based on all screening/randomized subjects.

5.4. Demographics and Baseline Characteristics

Demographic characteristics for the ITT Population were all similar between the two treatment groups (Table 7).

Table 7 Demographic and Other Baseline Characteristics (ITT Population)

ACT UC Total Demographic N=242 N=265 N=507 Age (years) Mean (SD) 48.0 (12.81) 48.0 (13.43) 48.0 (13.13) Min-Max 18 - 73 18 - 70 18 - 73 Gender, n (%) Female 115 (47.5%) 135 (50.9%) 250 (49.3%) Male 127 (52.5%) 130 (49.1%) 257 (50.7%) Race, n (%) Han nationality 237 (97.9%) 261 (98.5%) 498 (98.2%) Non-Han nationality 5 (2.1%) 4 (1.5%) 9 (1.8%) Height (cm) Mean (SD) 165.24 (7.820) 165.60 (8.238) 165.43 (8.035) Min-Max 145.0 - 186.0 145.0 - 187.0 145.0 - 187.0 Weight (kg) Mean (SD) 65.89 (11.074) 64.90 (11.694) 65.37 (11.402) Min-Max 33.0 - 105.0 41.0 - 100.0 33.0 - 105.0 BMI Mean (SD) 24.06 (3.246) 23.56 (3.141) 23.80 (3.199) Min-Max 15.7 - 36.3 16.0 - 34.3 15.7 - 36.3 Temperature (°C) Mean (SD) 36.60 (0.239) 36.57 (0.277) 36.58 (0.260) Min-Max 36.0 - 37.2 36.0 - 37.2 36.0 - 37.2 Pulse Rate (beats/min) Mean (SD) 78.2 (7.37) 78.5 (7.38) 78.3 (7.37) Min-Max 57 - 104 59 - 112 57 - 112

   33 CONFIDENTIAL 2019N423508_00 201097

ACT UC Total Demographic N=242 N=265 N=507 Systolic Blood Pressure (mmHg) Mean (SD) 122.4 (9.86) 120.4 (11.68) 121.4 (10.89) Min-Max 90 - 150 90 - 180 90 - 180 Diastolic Blood Pressure (mmHg) Mean (SD) 77.2 (7.39) 76.6 (7.72) 76.9 (7.56) Min-Max 53 - 98 50 - 100 50 - 100 Source: Table 1.05

The mean age was similar across the two treatment groups, with a range of 18 to 73 years across the groups. There was similar proportion of males in the two treatment groups and most subjects in the two treatment groups were Han nationality. Other baseline characteristics were also similar among the two treatment groups (Table 7). A listing of subject demographic and other baseline characteristics is provided in Listing 1.04.

Most of subjects in the ITT Population were never smoking, and only one subject was smoking in ACT group, and others quitted smoking. None subject in the ITT Population had alcohol abuse history or medication abuse history. Most subjects in the ITT Population had mild or moderate asthma history, and most subjects had no asthma exacerbation history. More details information were summarized in Source Table 1.06.

Source Listing 1.04 displays the listing of subjects’ demographic and other baseline characteristics.

Source Listing 1.05 displays the listing of subjects’ history of smoking, alcohol abuse and drug abuse.

Source Listing 1.06 displays the listing of subjects’ history of asthma and asthma exacerbation.

5.5. Past and Current Illness

Past illness for the ITT Population were similar among the two treatment groups. The most common past illness was Rhinitis (ACT: 15.3%; UC: 9.8%) and Hypertension (ACT: 1.2%; UC: 3.8%), and the incidence of other past illness were all lower than 3% in each treatment group.

The incidence of current illness in ACT group were lower than that in UC group for the ITT Population. The most common current illness was Rhinitis (ACT: 22.3%; UC: 55.1%), Hypertension (ACT: 22.3%; UC: 12.5%), Sinusitis (ACT: 2.9%; UC: 3.8%), Coronary artery disease (ACT: 2.5%; UC: 4.2%), Rhinitis allergic (ACT: 5.0%; UC: 0.4%), Chronic gastritis (ACT: 1.2%; UC: 4.2%) and Pharyngitis (ACT: 0.4%; UC: 3.0%), and the incidence of other current illness were all lower than 3% in each treatment group.

The past and current illness were summarized in Source Table 1.07. Source Listing 1.07 displays the listing of subjects’ past and current illness.

   34 CONFIDENTIAL 2019N423508_00 201097

5.6. Asthma Medications

Asthma Medications for the ITT Population were similar among the two treatment groups at visit 0 and during the study period. At visit 0, the most common asthma medications were BUDESONIDE;FORMOTEROL FUMARATE or BUDESONIDE; FORMOTEROL (ACT: 38.0%; UC: 64.2%) and FLUTICASONE PROPIONATE; SALMETEROL XINAFOATE or FLUTICASONE; SALMETEROL (ACT: 64.0%; UC: 30.9%), and during the study period the most common asthma medications were also BUDESONIDE;FORMOTEROL FUMARATE or BUDESONIDE; FORMOTEROL (ACT: 29.8%; UC: 56.6%) and FLUTICASONE PROPIONATE; SALMETEROL XINAFOATE or FLUTICASONE; SALMETEROL (ACT: 54.5%; UC: 28.3%).

The asthma medications with incidence ≥3% in any treatment group were summarized in Table 8.

Source Listing 1.08 displays the listing of subjects' current asthma medication record.

Source Listing 1.09 displays the listing of subjects' asthma treatment.

Table 8 Asthma Medications with Incidence ≥3% in Any Treatment Group (ITT Population)

Number (%) of Subjects Asthma Medications ACT UC Total N=242 N=265 N=507 At least one asthma medication at visit 0 242 (100.0%) 264 (99.6%) 506 (99.8%) FLUTICASONE PROPIONATE; SALMETEROL XINAFOATE or 155 (64.0%) 82 (30.9%) 237 (46.7%) FLUTICASONE; SALMETEROL BUDESONIDE;FORMOTEROL FUMARATE or BUDESONIDE; 92 (38.0%) 170 (64.2%) 262 (51.7%) FORMOTEROL SALBUTAMOL 50 (20.7%) 58 (21.9%) 108 (21.3%) MONTELUKAST SODIUM 41 (16.9%) 64 (24.2%) 105 (20.7%) SALBUTAMOL SULFATE 22 (9.1%) 35 (13.2%) 57 (11.2%) MONTELUKAST 17 (7.0%) 5 (1.9%) 22 (4.3%) BUDESONIDE 7 (2.9%) 8 (3.0%) 15 (3.0%) AMINOPHYLLINE;CHLORPHENAMINE MALEATE; METHOXYPHENAMINE 5 (2.1%) 12 (4.5%) 17 (3.4%) HYDROCHLORIDE; NOSCAPINE At least one asthma medication during 238 (98.3%) 265 (100.0%) 503 (99.2%) the study period FLUTICASONE PROPIONATE; SALMETEROL XINAFOATE or 132 (54.5%) 75 (28.3%) 207 (40.8%) FLUTICASONE; SALMETEROL BUDESONIDE;FORMOTEROL FUMARATE or BUDESONIDE; 72 (29.8%) 150 (56.6%) 222 (43.8%) FORMOTEROL SALBUTAMOL 18 (7.4%) 10 (3.8%) 28 (5.5%) SALBUTAMOL SULFATE 9 (3.7%) 10 (3.8%) 19 (3.7%)

   35 CONFIDENTIAL 2019N423508_00 201097

Number (%) of Subjects Asthma Medications ACT UC Total N=242 N=265 N=507 MONTELUKAST SODIUM 1 (0.4%) 8 (3.0%) 9 (1.8%) FLUTICASONE PROPIONATE 0 8 (3.0%) 8 (1.6%) Source: Table 1.08 NOTE: FLUTICASONE PROPIONATE; SALMETEROL XINAFOATE and FLUTICASONE; SALMETEROL are duplicate reports for the same drug. BUDESONIDE; FORMOTEROL FUMARATE and BUDESONIDE; FORMOTEROL are duplicate reports for the same drug.

5.7. Concurrent Medications

The incidence of concurrent medications in ACT group was lower than that in UC group for the ITT Population. The most common concurrent medication was UNSPECIFIED HERBAL AND TRADITIONAL MEDICINE (ACT: 35.5%; UC: 43.0%).

The concurrent medications with incidence ≥3% in any treatment group were summarized in Table 9.

Source Listing 1.10 displays the listing of subjects’ current medications.

Table 9 Concurrent Medications with Incidence ≥3% in Any Treatment Group (ITT Population)

Number (%) of Subjects Concurrent Medications ACT UC Total N=242 N=265 N=507 At least one Concurrent medication 131 (54.1%) 167 (63.0%) 298 (58.8%) UNSPECIFIED HERBAL AND 86 (35.5%) 114 (43.0%) 200 (39.4%) TRADITIONAL MEDICINE CINEOLE;DIPENTEN;PINENE 16 (6.6%) 17 (6.4%) 33 (6.5%) CEFACLOR 12 (5.0%) 10 (3.8%) 22 (4.3%) AMINOPHYLLINE;CHLORPHENAMINE MALEATE;METHOXYPHENAMINE 9 (3.7%) 18 (6.8%) 27 (5.3%) HYDROCHLORIDE;NOSCAPINE LEVOFLOXACIN 9 (3.7%) 18 (6.8%) 27 (5.3%) MONTELUKAST SODIUM 9 (3.7%) 16 (6.0%) 25 (4.9%) ACETYLSALICYLIC ACID 8 (3.3%) 8 (3.0%) 16 (3.2%) CEFIXIME 8 (3.3%) 14 (5.3%) 22 (4.3%) DOXOFYLLINE 8 (3.3%) 0 8 (1.6%) TRIAMCINOLONE ACETONIDE 8 (3.3%) 3 (1.1%) 11 (2.2%) DEXAMETHASONE 7 (2.9%) 9 (3.4%) 16 (3.2%) DEXTROMETHORPHAN HYDROBROMIDE;PARACETAMOL; 7 (2.9%) 9 (3.4%) 16 (3.2%) PSEUDOEPHEDRINE HYDROCHLORIDE AZITHROMYCIN 6 (2.5%) 17 (6.4%) 23 (4.5%) BUDESONIDE 6 (2.5%) 21 (7.9%) 27 (5.3%) RABEPRAZOLE 6 (2.5%) 9 (3.4%) 15 (3.0%) VALSARTAN 5 (2.1%) 8 (3.0%) 13 (2.6%)

   36 CONFIDENTIAL 2019N423508_00 201097

Number (%) of Subjects Concurrent Medications ACT UC Total N=242 N=265 N=507 CHLORPHENAMINE MALEATE; DEXTROMETHORPHAN 3 (1.2%) 16 (6.0%) 19 (3.7%) HYDROBROMIDE;PARACETAMOL; PSEUDOEPHEDRINE HYDROCHLORIDE OMEPRAZOLE 3 (1.2%) 9 (3.4%) 12 (2.4%) CEFPROZIL 2 (0.8%) 13 (4.9%) 15 (3.0%) Source: Table 1.09

   37 CONFIDENTIAL 2019N423508_00 201097

6. EFFICACY RESULTS

6.1. Primary Efficacy Results

6.1.1. Percentage of Subjects Who Had an ACT Total Score ≥ 20 or an Improvement >3 Points in ACT Total Score in at Least One Post- baseline Assessment

The primary endpoint for the study was the percentage of subjects who had an ACT total score ≥20 or an improvement of more than 3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period.

Primary Analysis Results

The primary endpoint was summarized using generalized linear mixed model (GLMM) included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor for ITT and PP population.

For ITT population, the incidence of response for well-controlled asthma in ACT group (99.2%) was higher than that in UC group (94.3%). The adjusted odds ratio (OR) was 7.87 with 95% CI (1.29, 48.11) and P-value was 0.027. Therefore, for ACT Total Score, the difference between ACT guided treatment and UC was statistically significant at the 5% level of significance. For all interactions between treatment and each covariate including type of baseline controller, gender and age, the interaction effect were all not statistically significant (P-value with type of baseline controller = 0.982; P-value with gender = 0.975; P-value with age = 0.891) at the 10% level of significance (Table 10). The adjusted OR and 95% CIs were also presented in Figure 6.1.

Table 10 Summary of the Statistical Analysis of Percentage of Subjects Who Had an ACT Total Score ≥ 20 or an Improvement >3 Points in ACT Total Score in at Least One Post-baseline Assessment during the 24-Weeks Treatment Period (ITT Population)

ACT UC N=242 N=265 n 240 262 Responder, n (%) [1] 238 (99.2%) 247 (94.3%) Non-Responder, n (%) 2 (0.8%) 15 (5.7%)

ACT vs. UC [2] Adjusted Odds Ratio 7.87 95% CI (1.29, 48.11) P-value 0.027

Interaction between treatment and covariates, P-value [3] With type of baseline controller 0.982 With gender 0.975 With age 0.891

   38 CONFIDENTIAL 2019N423508_00 201097

Source: Table 2.01 [1] Subjects who had an ACT total score ≥ 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor. [3] The interaction between treatment and each covariate was studied sequentially, and significant effects were considered at a 10% significance level.

Figure 6.1 Forest Plot of Adjusted Odds Ratios and 95% CIs from GLMM, Comparing ACT Guided Treatment with Usual Care on Primary Analysis and Sensitivity Analysis of ACT Response (ITT)

Source: Figure 1 [1] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

For PP population, the results were similar to those of ITT population. The difference between ACT guided treatment and UC was statistically significant in ACT Total Score (adjusted OR=12.10 with 95% CI (1.16, 126.19); P-value = 0.038) at the 5% level of significance and the interaction effect were all not statistically significant (P-values with type of baseline controller =0.976; P-value with gender = 0.978; P-value with age = 0.226) at the 10% level of significance (Source Table 2.02). The adjusted OR and 95% CIs were presented in Figure 6.2.

Source Listing 2.01 provided the listing of subjects’ ACT score.

Figure 6.2 Forest Plot of Adjusted Odds Ratios and 95% CIs from GLMM, Comparing ACT Guided Treatment with Usual Care on Primary Analysis and Sensitivity Analysis of ACT Response (PP)

Source: Figure 2 [1] Subjects who have an ACT total score ≥ 20 or an improvement > 3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period.

   39 CONFIDENTIAL 2019N423508_00 201097

[2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

Sensitivity and Supportive Analysis Results

After unblinded analyses had been performed on interim data, some post-hoc sensitivity and supportive analyses were added for the primary efficacy analysis endpoint. The primary endpoint was repeatedly analysed in subgroups, using the same method used in primary analysis method in ITT and PP populations.

For ITT population, the incidence of response for well-controlled asthma in ACT group was all higher than that in UC group within each subgroup. For male subgroup, the adjusted OR was 7.06 with 95% CI (1.16, 42.78) and P-value was 0.034. However, for any other subgroup, the adjusted OR was not applicable, or the lower limit of 95% CI were lower than 1 and P-values was higher than 0.05 (Table 11). Therefore, only for male subgroup, the difference between ACT guided treatment and UC was statistically significant in ACT Total Score at the 5% level of significance.

Table 11 Summary of the Statistical Analysis of Percentage of Subjects Who Had an ACT Total Score ≥ 20 or an Improvement >3 Points in ACT Total Score in at Least One Post-baseline Assessment during the 24-Weeks Treatment Period by Subgroup (ITT Population)

Subgroup ACT UC N=242 N=265 Type of Baseline Controller: ICS Alone n 8 12 Responder, n (%) [1] 8 (100.0%) 11 (91.7%) Non-Responder, n (%) 0 1 (8.3%)

ACT vs. UC [2] Adjusted Odds Ratio NA 95% CI NA P-value NA Type of Baseline Controller: ICS/LABA n 232 250 Responder, n (%) [1] 230 (99.1%) 236 (94.4%) Non-Responder, n (%) 2 (0.9%) 14 (5.6%)

ACT vs. UC [2] Adjusted Odds Ratio 6.43 95% CI (0.99, 41.59) P-value 0.051 Gender: Male n 126 129 Responder, n (%) [1] 124 (98.4%) 119 (92.2%) Non-Responder, n (%) 2 (1.6%) 10 (7.8%)

ACT vs. UC [3] Adjusted Odds Ratio 7.06 95% CI (1.16, 42.78) P-value 0.034 Gender: Female n 114 133

   40 CONFIDENTIAL 2019N423508_00 201097

Subgroup ACT UC N=242 N=265 Responder, n (%) [1] 114 (100.0%) 128 (96.2%) Non-Responder, n (%) 0 5 (3.8%)

ACT vs. UC [3] Adjusted Odds Ratio NA 95% CI NA P-value NA Age: <50 Years Old n 123 131 Responder, n (%) [1] 122 (99.2%) 123 (93.9%) Non-Responder, n (%) 1 (0.8%) 8 (6.1%)

ACT vs. UC [4] Adjusted Odds Ratio NA 95% CI NA P-value NA Age: ≥50 Years Old n 117 131 Responder, n (%) [1] 116 (99.1%) 124 (94.7%) Non-Responder, n (%) 1 (0.9%) 7 (5.3%)

ACT vs. UC [4] Adjusted Odds Ratio 8.45 95% CI (0.37, 195.56) P-value 0.169 Source: Table 2.03 [1] Subjects who had an ACT total score ≥ 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24 - Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

For PP population, the results were similar to those of ITT population that only for male subgroup the difference between ACT guided treatment and UC was statistically significant in ACT Total Score (adjusted OR = 12.77 with 95% CI (1.20, 136.02); P- value = 0.035) at the 5% level of significance (Source Table 2.04).

The primary endpoint was also performed using Generalised Estimating Equations (GEEs) for all non-missing data from visits with scheduled ACT assessments (Weeks 4, 8, 12, 16, 20 and 24) in ITT and PP population.

For ITT population, the incidence of response for well-controlled asthma in ACT group was all higher than that in UC group for well-controlled asthma at each visit, but the adjusted OR and P-values were not applicable (Table 12).

   41 CONFIDENTIAL 2019N423508_00 201097

Table 12 Summary of the Statistical Analysis of Percentage of Subjects Who Had an ACT Total Score ≥ 20 or an Improvement >3 Points in ACT Total Score in at Least One Post-baseline Assessment During the 24-Weeks Treatment Period at Weeks 4, 8, 12, 16, 20 And 24 (ITT Population)

Visit ACT UC N=242 N=265 Week 4 n 236 261 Responder, n (%) [1] 208 (88.1%) 177 (67.8%) Non-Responder, n (%) 28 (11.9%) 84 (32.2%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA Week 8 n 226 248 Responder, n (%) [1] 219 (96.9%) 197 (79.4%) Non-Responder, n (%) 7 (3.1%) 51 (20.6%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA Week 12 n 214 244 Responder, n (%) [1] 209 (97.7%) 207 (84.8%) Non-Responder, n (%) 5 (2.3%) 37 (15.2%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA Week 16 n 215 236 Responder, n (%) [1] 213 (99.1%) 196 (83.1%) Non-Responder, n (%) 2 (0.9%) 40 (16.9%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA Week 20 n 207 230 Responder, n (%) [1] 203 (98.1%) 199 (86.5%) Non-Responder, n (%) 4 (1.9%) 31 (13.5%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA Week 24 n 208 233 Responder, n (%) [1] 208 (100.0%) 209 (89.7%) Non-Responder, n (%) 0 24 (10.3%)

ACT vs. UC

   42 CONFIDENTIAL 2019N423508_00 201097

Visit ACT UC N=242 N=265 Adjusted Odds Ratio NA 95% CI NA P-value NA Source: Table 2.05 [1] Subjects who had an ACT total score ≥ 20 or an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment- by-visit interaction.

For PP population, the results were similar to those of ITT population. The incidence of response for well-controlled asthma in ACT group was all higher than that in UC group at each visit, but the adjusted OR and P-values were not applicable (Source Table 2.06).

6.1.2. Percentage of Subjects Who Have an ACT Total Score ≥ 20 and an Improvement >3 Points in ACT Total Score

After unblinded analyses had been performed on interim data, the post-hoc sensitivity analyses were added for percentage of subjects who had an ACT total score ≥20 and an improvement of more than 3 points in ACT total score.

The percentage of subjects who had an ACT total score ≥20 and an improvement of more than 3 points in ACT total score was summarized for Week 12, Week 24, and Week 16 and 20 and 24, using primary analysis method in ITT and PP population. Subjects without baseline ACT total score were excluded from the analysis. Subjects with baseline ACT total score but with missing data at post-base visit were set as non-responders.

For ITT population, at week 12 the incidence of responder in ACT group (75.9%) was higher than that in UC group (61.2%), and the adjusted OR was 2.10 with 95% CI (0.82, 5.34) and P-value was 0.105. At week 24, the incidence of responder in ACT group (83.0%) was also higher than that in UC group (67.7%), and the adjusted OR was 2.28 with 95% CI (1.07, 4.85) and P-value was 0.036. At weeks 16 and 20 and 24, the incidence of responder in ACT group (69.7%) was also higher than that in UC group (47.1%), and the adjusted OR was 2.52 with 95% CI (1.25, 5.07) and P-value was 0.016 (Table 13). Therefore, at week 24 and weeks 16 and 20 and 24, the difference between ACT guided treatment and UC were all statistically significant in ACT Total Score, but at week 12, the difference between ACT guided treatment and UC was not statistically significant in ACT Total Score at the 5% level of significance. The adjusted odds ratios and 95% CI were presented in Figure 6.1.

   43 CONFIDENTIAL 2019N423508_00 201097

Table 13 Summary of the Statistical Analysis of Percentage of Subjects Who Had an ACT Total Score ≥ 20 and an Improvement >3 Points in ACT Total Score (ITT Population)

Visit ACT UC N=242 N=265 Week 12 n 241 263 Responder, n (%) [1] 183 (75.9%) 161 (61.2%) Non-Responder, n (%) 58 (24.1%) 102 (38.8%)

ACT vs. UC [4] Adjusted Odds Ratio 2.10 95% CI (0.82, 5.34) P-value 0.105 Week 24 n 241 263 Responder, n (%) [2] 200 (83.0%) 178 (67.7%) Non-Responder, n (%) 41 (17.0%) 85 (32.3%)

ACT vs. UC [4] Adjusted Odds Ratio 2.28 95% CI (1.07, 4.85) P-value 0.036 Weeks 16 and 20 and 24 n 241 263 Responder, n (%) [3] 168 (69.7%) 124 (47.1%) Non-Responder, n (%) 73 (30.3%) 139 (52.9%)

ACT vs. UC [4] Adjusted Odds Ratio 2.52 95% CI (1.25, 5.07) P-value 0.016 Source: Table 2.07 [1] Subjects who had an ACT total score ≥ 20 and an improvement >3 points in ACT total score at Week 12. Subjects with missing data were set as non-responders. [2] Subjects who had an ACT total score ≥ 20 and an improvement >3 points in ACT total score at Week 24. Subjects with missing data were set as non-responders. [3] Subjects who had an ACT total score ≥ 20 and an improvement >3 points in ACT total score at Weeks 16 and 20 and 24. Subjects with missing data were set as non-responders. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

For PP population, the results were similar to those of ITT population. At week 12, the adjusted OR was 2.29 with 95% CI (0.75, 6.97) and P-value was 0.127. At week 24 the adjusted OR was 2.39 with 95% CI (1.08, 5.29) and P-value was 0.035. At weeks 16 and 20 and 24 the adjusted OR was 2.52 with 95% CI (1.15, 5.53) and P-value was 0.027. Therefore, at week 24 and weeks 16 and 20 and 24, the difference between ACT guided treatment and UC were all statistically significant in ACT Total Score, but at week 12, the difference between ACT guided treatment and UC was not statistically significant in ACT Total Score at the 5% level of significance (Source Table 2.08). The adjusted OR and 95% CIs was presented in Figure 6.2.

   44 CONFIDENTIAL 2019N423508_00 201097

The percentage of subjects who had an ACT total score ≥20 and an improvement of more than 3 points in ACT total score was also performed using GEEs for all non-missing data from visits with scheduled ACT assessments (Weeks 4, 8, 12, 16, 20 and 24) in ITT and PP population.

For ITT population, at each visit the incidence of responder in ACT group was all higher than that in UC group, and the lower limits of 95% CI for the adjusted OR were all higher than 1 and P-values were all lower than 0.001, therefore the difference was statistically significant at the 5% level of significance (Table 14). The adjusted OR and 95% CIs were presented in Figure 6.3.

Table 14 Summary of the Statistical Analysis of Percentage of Subjects Who Had an ACT Total Score ≥ 20 and an Improvement >3 Points in ACT Total Score in at Least One Post-baseline Assessment during the 24-Weeks Treatment Period at Weeks 4, 8, 12, 16, 20 And 24 (ITT Population)

Visit ACT UC N=242 N=265 Week 4 n 236 261 Responder, n (%) [1] 157 (66.5%) 122 (46.7%) Non-Responder, n (%) 79 (33.5%) 139 (53.3%)

ACT vs. UC Adjusted Odds Ratio 2.04 95% CI (1.41, 2.95) P-value <0.001 Week 8 n 226 248 Responder, n (%) [1] 184 (81.4%) 151 (60.9%) Non-Responder, n (%) 42 (18.6%) 97 (39.1%)

ACT vs. UC Adjusted Odds Ratio 2.63 95% CI (1.72, 4.03) P-value <0.001 Week 12 n 214 244 Responder, n (%) [1] 183 (85.5%) 161 (66.0%) Non-Responder, n (%) 31 (14.5%) 83 (34.0%)

ACT vs. UC Adjusted Odds Ratio 2.71 95% CI (1.73, 4.26) P-value <0.001 Week 16 n 215 236 Responder, n (%) [1] 197 (91.6%) 156 (66.1%) Non-Responder, n (%) 18 (8.4%) 80 (33.9%)

ACT vs. UC Adjusted Odds Ratio 4.78 95% CI (2.81, 8.13) P-value <0.001 Week 20 n 207 230

   45 CONFIDENTIAL 2019N423508_00 201097

Visit ACT UC N=242 N=265 Responder, n (%) [1] 188 (90.8%) 161 (70.0%) Non-Responder, n (%) 19 (9.2%) 69 (30.0%)

ACT vs. UC Adjusted Odds Ratio 3.81 95% CI (2.22, 6.53) P-value <0.001 Week 24 n 208 233 Responder, n (%) [1] 200 (96.2%) 178 (76.4%) Non-Responder, n (%) 8 (3.8%) 55 (23.6%)

ACT vs. UC Adjusted Odds Ratio 6.51 95% CI (3.20, 13.25) P-value <0.001 Source: Table 2.09 [1] Subjects who had an ACT total score ≥ 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment- by-visit interaction.

Figure 6.3 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment with Usual Care On Percentage of Subjects Who Had an ACT Total Score ≥ 20 and an Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (ITT)

Source: Figure 5 [1] Subjects who have an ACT total score ≥ 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, gender, visit and treatment-by-visit interaction.

For PP population, the results were similar to those of ITT population. At each visit the incidence of responder in ACT group was all higher than that in UC group and the lower limits of 95% CI for the adjusted OR were all higher than 1 and P-values were all lower than 0.001, therefore the difference was statistically significant at the 5% level of significance (Source Table 2.10). The adjusted odds ratios and 95% confidence intervals were presented in Figure 6.4.

   46 CONFIDENTIAL 2019N423508_00 201097

Figure 6.4 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment with Usual Care On Percentage of Subjects Who Had an ACT Total Score >= 20 and an Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (PP)

Source: Figure 6 [1] Subjects who have an ACT total score ≥ 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, gender, visit and treatment-by-visit interaction.

6.1.3. Change from Baseline in ACT Total Score

After unblinded analyses had been performed on interim data, the post-hoc sensitivity analyses were added for change from baseline in ACT total score in post-baseline visits.

Change from baseline in ACT total score in post-baseline visits was summarized using MMRM with covariates of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect for ITT and PP population.

For ITT population, at any post-baseline visit the LS mean change from baseline in ACT total score in ACT group was higher than that in UC group, and for the difference, the lower limit of 95% CI was higher than 0 and P-values were all lower than 0.05, therefore the difference was statistically significant at the 5% level of significance (Source Table 2.11). The LS mean change from baseline and 95% CIs were presented by Figure 6.5.

Figure 6.5 Forest Plot of LS Mean change from baseline treatment difference and 95% CIs from MMRM, Comparing ACT Guided Treatment with Usual Care on ACT Total Score at Weeks 4, 8, 12, 16, 20 and 24 (ITT)

Source: Figure 7

   47 CONFIDENTIAL 2019N423508_00 201097

Note: The analysis method was a MMRM model adjusted for of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect.

For PP population, the results were similar to those of ITT population. At any post- baseline visit, the LS mean change from baseline in ACT total score in ACT group was also higher than that in UC group and P-values were all lower than 0.05, therefore the difference was statistically significant at the 5% level of significance (Source Table 2.12). The LS mean change from baseline and 95% CIs were presented by Figure 6.6.

Figure 6.6 Forest Plot of LS Mean change from baseline treatment difference and 95% CIs from MMRM, Comparing ACT Guided Treatment with Usual Care on ACT Total Score at Weeks 4, 8, 12, 16, 20 and 24 (PP)

Source: Figure 8 Note: The analysis method was a MMRM model adjusted for of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect.

6.2. Secondary Efficacy Results

The following secondary efficacy endpoints were only analysed in ITT population.

6.2.1. Mean daytime symptom score over the 24-week treatment period

Mean daytime symptom score over the 24-week treatment period were summarized in Source Table 2.13 and Source Table 2.14. Source Listing 2.02 provided the listing of subjects’ daytime symptom score.

At each 4-week treatment interval, the LS mean of mean daytime symptom score in ACT group was lower than that in UC group, but the upper limit of 95% CI were all higher than 0 and P-values were all higher than 0.05, therefore the difference wasn’t statistically significant at the 5% level of significance. The LS mean treatment differences and 95% CIs were presented by Figure 6.7.

   48 CONFIDENTIAL 2019N423508_00 201097

Figure 6.7 Forest Plot of LS Mean treatment differences and 95% CIs from MMRM, Comparing ACT Guided Treatment with Usual Care on Mean Daytime Symptom Score by Visits (ITT)

Source: Figure 9 Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

6.2.2. Mean night-time symptom score over the 24-week treatment period

Mean nigh-time symptom score over the 24-week treatment period were summarized in Source Table 2.15 and Source Table 2.16. Source Listing 2.02 provided the listing of subjects’ night-time symptom score.

At each 4-week treatment interval, the LS mean of mean night-time symptom score in ACT group was lower than that in UC group, but the upper limit of 95% CI were all higher than 0 and P-values were all higher than 0.05, therefore the difference wasn’t statistically significant at the 5% level of significance. The LS mean treatment differences and 95% CIs were presented by Figure 6.8.

Figure 6.8 Forest Plot of LS Mean treatment differences and 95% CIs from MMRM, Comparing ACT Guided Treatment with Usual Care on Mean Night-time Symptom Score by Visits (ITT)

Source: Figure 10 Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

   49 CONFIDENTIAL 2019N423508_00 201097

6.2.3. Mean change from baseline to the end of study in FEV1

For the all subjects who completed the study, mean change from baseline to the end of study was summarized in Source Table 2.17 and Source Table 2.18. Source Listing 2.03 provided the listing of subjects’ lung function examination.

At week 24, the LS means of mean change form baseline in ACT group were all higher than that in UC group for FEV1(L) and FEV1(%), but the lower limit of 95% CI were all lower than 0 and P-values were all higher than 0.05, therefore the differences weren’t statistically significant at the 5% level of significance. The LS mean treatment differences and 95% CIs were presented by Figure 6.9.

Figure 6.9 Forest Plot of LS Mean change from baseline treatment differences and 95% CIs from Mixed Model, Comparing ACT Guided Treatment with Usual Care on FEV1 (ITT)

Source: Figure 11 Note: Analysis performed using a mixed model with covariates of treatment, center, baseline FEV1, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the center as a random factor. Only include those who complete the study and exclude those who withdraw the study. FEV1 stands for Forced Expiratory Volume in One Second.

6.2.4. Mean morning (AM) PEF over the 24-week treatment period

Mean morning (AM) PEF over the 24-week treatment period were summarized in Source Table 2.19 and Source Table 2.20. Source Listing 2.02 provided the listing of subject morning PEF.

At each 4-week treatment interval, the LS mean of mean AM PEF in ACT group was lower than that in UC group, but the upper limit of 95% CI were all higher than 0 and P- values were all higher than 0.05, therefore the difference wasn’t statistically significant at the 5% level of significance. The LS mean treatment differences and 95% CIs were presented by Figure 6.10.

   50 CONFIDENTIAL 2019N423508_00 201097

Figure 6.10 Forest Plot of LS Mean treatment differences and 95% CIs from MMRM, Comparing ACT Guided Treatment with Usual Care on Mean Morning PEF by Visits (ITT)

Source: Figure 12 Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor. PEF stands for Peak Expiratory Flow.

6.2.5. Mean evening (PM) PEF over the 24-week treatment period

Mean evening (PM) PEF over the 24-week treatment period were summarized in Source Table 2.21 and Source Table 2.22. Source Listing 2.02 provided the listing of subject evening PEF.

At each 4-week treatment interval, the LS mean of mean PM PEF in ACT group was lower than that in UC group, but the upper limit of 95% CI were all higher than 0 and P- values were all higher than 0.05, therefore the difference wasn’t statistically significant at the 5% level of significance. The LS mean treatment differences and 95% CIs were presented by Figure 6.11.

Figure 6.11 Forest Plot of LS Mean treatment differences and 95% CIs from MMRM, Comparing ACT Guided Treatment with Usual Care on Mean Evening PEF by Visits (ITT)

Source: Figure 13 Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor. PEF stands for Peak Expiratory Flow.

   51 CONFIDENTIAL 2019N423508_00 201097

6.2.6. Mean change from baseline to the end of study in AQLQ(S) score

For the all subjects who completed the study, mean change from baseline to the end of study was summarized in Source Table 2.23 and Source Table 2.24 for AQLQ(S) score which included 4 domains (symptoms and activity limitation and emotional function and environmental stimuli) and total score. Source Listing 2.04 provided the listing of subject AQLQ(S) score.

The LS means of mean change form baseline in 4 domains and total score were all higher than 0.5 which were considered clinically important, and the LS mean in ACT group were all higher than those in UC group, and the differences were statistically significant only for symptoms (Difference [95% CI] = 0.3(0.0, 0.6); P-value=0.030) and emotional function (Difference [95% CI] = 0.3(0.0, 0.6); P-value=0.042) at the 5% level of significance. The LS mean treatment differences and 95% CIs were presented by Figure 6.12.

Figure 6.12 Forest Plot of LS Mean change from baseline treatment differences and 95% CIs from Mixed Model, Comparing ACT Guided Treatment with Usual Care on AQLQ(S) score (ITT)

Source: Figure 14 Note: Analysis performed using a mixed model with covariates of treatment, center, baseline AQLQ score, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the center as a random factor. Only include those who complete the study and exclude those who withdraw the study. AQLQ(S) stands for Asthma Quality of Life Questionnaire.

6.2.7. Time to first ACT total score ≥20 or an Improvement >3 Points in ACT total score over the 24-week treatment period

The time to first ACT total score ≥20 or an Improvement >3 Points in ACT total score over the 24-week treatment period was summarized in Source Table 2.25 using cox regression analysis with treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age as the adjusted factors. The median of KM estimates were 31.0 and 30.0 in ACT group and UC group, respectively. Compared with UC group, HR was 1.333 with 95% CI (0.844, 2.107) and P-value was 0.218, the difference wasn’t statistically significant at the 5% level of significance. KM curves was constructed in Figure 6.13 to provide a visual description of the difference between the two treatment arms.

   52 CONFIDENTIAL 2019N423508_00 201097

Figure 6.13 KM Figure of Time to Asthma Control (ACT Total Score >=20 or Improvement of More than 3 Points in ACT Total Score)

For the First Time (ITT)

100 90 80 70 60 50 40 30 20 ACT 10 Estimated Probability of Having an Event(%) an Having of Probability Estimated UC 0 Total 0 28 56 84 112 140 168 Time to asthma control for the first time (days) Number of subjects at Risk

ACT 241 208 22 3 1 0 0

UC 264 224 73 29 20 16 12

Total 505 432 95 32 21 16 12

Source: Figure 15

6.2.8. Time to first ACT total score ≥20 and an Improvement >3 Points in ACT total score over the 24-week treatment period

The time to first ACT total score ≥20 and an Improvement >3 Points in ACT total score over the 24-week treatment period was summarized in Source Table 2.26 using the same analysis method in Section 6.2.7. The median of KM estimates were 33.0 and 55.0 in ACT group and UC group, respectively. Compared with UC group, HR was 1.843 with 95% CI (1.160, 2.926) and P-value was 0.010, the difference was statistically significant at the 5% level of significance. KM curves was constructed in Figure 6.14 to provide a visual description of the difference between the two treatment arms.

   53 CONFIDENTIAL 2019N423508_00 201097

Figure 6.14 KM Figure of Time to Asthma Control (ACT Total Score >=20 and Improvement of More than 3 Points in ACT Total Score)

For the First Time (ITT)

100 90 80 70 60 50 40 30 20 ACT 10 Estimated Probability of Having an Event(%) an Having of Probability Estimated UC 0 Total 0 28 56 84 112 140 168 Time to asthma control for the first time (days) Number of subjects at Risk

ACT 241 215 70 31 15 8 4

UC 264 236 128 80 59 46 29

Total 505 451 198 111 74 54 33

Source: Figure 16

6.3. Exploratory Efficacy Results

For the subjects who experienced moderate/severe asthma exacerbations during the study period, there were was 8 (3.3%) subjects with 10 moderate/severe asthma exacerbations in ACT group and 10 (3.8%) with 10 moderate/severe asthma exacerbations in UC group. Most of the moderate/severe asthma exacerbations were moderate (ACT: 60.0%; UC: 80.0%). The means (SD) of exacerbation (days) were 12.3 (15.72) and 6.8 (4.98) in ACT group and UC group, respectively.

For the annual exacerbation rate, it was analysed using a generalised linear model assuming a negative binominal distribution and covariates of treatment, baseline ACT total score, type of baseline controller (ICS alone vs. ICS/LABA), gender and age. The means of annual exacerbation rate were 0.10 with 95% CI (0.04, 0.25) and 0.09 with 95% CI (0.04, 0.22) in ACT group and UC group, respectively. Compared with UC group, the Rate Ratio was 1.09 with 95% CI (0.32, 3.73) and P-value was 0.897, the percent reduction in annual rate was -0.09 with 95% CI (-2.73, 0.68), the difference wasn’t statistically significant at the 5% level of significance.

   54 CONFIDENTIAL 2019N423508_00 201097

The details results of moderate/severe asthma exacerbations were displayed in Table 15.

Table 15 Summary of Moderate/Severe Asthma Exacerbations (ITT Population)

ACT UC N=242 N=265 Number of Subjects Who Experienced Moderate/Severe 8 (3.3%) 10 (3.8%) Asthma Exacerbations During the Study Period, n (%) Total Number of Moderate/Severe Asthma Exacerbations per Subject, n (%) 0 234 (96.7%) 255 (96.2%) 1 7 (2.9%) 10 (3.8%) 2 0 0 >2 1 (0.4%) 0 Number of Moderate/Severe Asthma Exacerbations 10 10 Outcome, n (%) Cured 8 (80.0%) 10 (100.0%) Unhealed 2 (20.0%) 0 Death 0 0 Measures, n (%) Clinic or Emergency Room Visit 7 (70.0%) 8 (80.0%) Self-managed 0 0 Hospital Admission 3 (30.0%) 2 (20.0%) Others 0 0 Severity, n (%) Moderate 6 (60.0%) 8 (80.0%) Severe 4 (40.0%) 2 (20.0%) Restrictions on Subsequent Daily Activities, n (%) Affecting Heavy Physical Activity 3 (30.0%) 4 (40.0%) Affecting Moderate Physical Activity 2 (20.0%) 3 (30.0%) Affecting Social Activities 1 (10.0%) 1 (10.0%) Affecting Daily Work 4 (40.0%) 2 (20.0%) Affecting Sleeping 0 0 Duration of Exacerbation (Days) n 8 10 Mean 12.3 6.8 SD 15.72 4.98 Median 7.5 6.0 Min. 3 1 Max. 50 14 Analysis performed using a generalised linear model n 241 263 Mean Annual Exacerbation Rate 0.10 0.09 95% CI (0.04, 0.25) (0.04, 0.22)

ACT vs. UC Rate Ratio 1.09 95% CI (0.32, 3.73) P-value 0.897

Percent Reduction in Annual Rate (%) [1] -0.09 95% CI (-2.73, 0.68)

   55 CONFIDENTIAL 2019N423508_00 201097

Source: Table 2.27 and Table 2.28 Note: Analysis performed using a generalized linear model assuming a negative binominal distribution and covariates of treatment, baseline ACT total score, type of baseline controller (ICS alone vs. ICS/LABA), gender and age. Note: n = the number of the events per treatment group. [1] Percent reduction in annual exacerbation rate was calculated as (1-rate ratio) ×100%.

6.4. Other Post-hoc Efficacy Results

6.4.1. Percentage of Subjects Who Had an ACT Total Score ≥ 20 or an Improvement ≥ 3 Points in ACT Total Score in at Least One Post- baseline Assessment

Percentage of subjects who had an ACT total score ≥ 20 or an improvement ≥3 Points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period was summarized in Source Table 2.29 only for ITT population. The results were similar to those of the primary endpoint result using primary analysis method. The difference between ACT guided treatment and UC was statistically significant in ACT Total Score (OR [95% CI] = 6.89 (1.28, 37.02); P-value=0.025) at the 5% level of significance and the interaction effect were all not statistically significant at the 10% level of significance. The adjusted odds ratios and 95% confidence intervals were presented in Figure 6.15.

Figure 6.15 Forest Plot of Adjusted Odds Ratios and 95% CIs from GLMM, Comparing ACT Guided Treatment with Usual Care on Primary Analysis and Sensitivity Analysis of ACT Response (ITT)

Source: Figure 17 [1] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

6.4.2. Percentage of Subjects Who Had an ACT Total Score ≥ 20 and an Improvement ≥ 3 Points in ACT Total Score in at Least One Post- baseline Assessment

Percentage of subjects who had an ACT total score ≥ 20 and an improvement ≥3 Points in ACT total score was summarized at week 12 and week 24 and Weeks 16 and 20 and 24 in Source Table 2.30 only for ITT population. Only at weeks 16 and 20 and 24, the difference between ACT guided treatment and UC was statistically significant in ACT

   56 CONFIDENTIAL 2019N423508_00 201097

Total Score (OR [95% CI] = 2.65 (1.22, 5.77); P-value = 0.020), but at week 12 and week 24, the difference between ACT guided treatment and UC was not statistically significant in ACT Total Score at the 5% level of significance. The adjusted odds ratios and 95% confidence intervals was presented in Figure 6.15.

6.4.3. Time to first ACT total score ≥20 or an Improvement ≥3 Points in ACT total score over the 24-week treatment period

The time to first ACT total score ≥20 or an Improvement ≥3 Points in ACT total score was summarized in Source Table 2.31 only for ITT population. Compared with UC group, the Hazard Ratio (HR) was 1.475 with 95% CI (0.933, 2.332) and P-value was 0.097, the difference wasn’t statistically significant at the 5% level of significance. KM curves was constructed in Figure 6.16. The trend is similar to that of time to first ACT total score ≥20 or an Improvement >3 Points in ACT total score in Section 6.2.7.

Figure 6.16 KM Figure of Time to Asthma Control (ACT Total Score ≥20 or Improvement ≥ 3 Points in ACT Total Score)

For the First Time (ITT)

100 90 80 70 60 50 40 30 20 ACT 10 Estimated Probability of Having an Event(%) an Having of Probability Estimated UC 0 Total 0 28 56 84 112 140 168 Time to asthma control for the first time (days) Number of subjects at Risk

ACT 241 205 14 3 1 0 0

UC 264 224 66 27 18 14 10

Total 505 429 80 30 19 14 10

Source: Figure 18

   57 CONFIDENTIAL 2019N423508_00 201097

6.4.4. Time to first ACT total score ≥20 and an Improvement ≥3 Points in ACT total score over the 24-week treatment period

The time to first ACT total score ≥20 and an Improvement ≥3 Points in ACT total score was summarized in Source Table 2.32 only for ITT population. Compared with UC group, HR was 1.708 with 95% CI (1.078, 2.707) and P-value was 0.023, the difference was statistically significant at the 5% level of significance. KM curves was constructed in Figure 6.17. The trend is similar to that of time to first ACT total score ≥20 and an Improvement >3 Points in ACT total score in Section 6.2.8.

Figure 6.17 KM Figure of Time to Asthma Control (ACT Total Score ≥20 and Improvement ≥ 3 Points in ACT Total Score)

For the First Time (ITT)

100 90 80 70 60 50 40 30 20 ACT 10 Estimated Probability of Having an Event(%) an Having of Probability Estimated UC 0 Total 0 28 56 84 112 140 168 Time to asthma control for the first time (days) Number of subjects at Risk

ACT 241 210 55 18 7 4 1

UC 264 231 105 56 39 30 21

Total 505 441 160 74 46 34 22

Source: Figure 19

6.5. Efficacy Conclusions

 For primary efficacy endpoint that the percentage of subjects who had an ACT total score ≥20 or an improvement > 3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period, ACT guided treatment well controlled asthma compared with usual care in China adult asthma patients, and the difference of treatments was statistically significant (OR [95% CI] = 7.87 (1.29, 48.11); P-values = 0.027) at the 5% level of significance, but the

   58 CONFIDENTIAL 2019N423508_00 201097

interaction effects were all not statistically significant at the 10% level of significance with type of baseline controller, gender and age.

 For the secondary efficacy endpoints, the difference between ACT guided treatment usual care was statistically significant (HR [95% CI] = 1.843 (1.160, 2.926); P-value = 0.010) only for the time to first ACT total score ≥20 and an Improvement >3 Points in ACT total score over the 24-week treatment period.

 For exploratory efficacy endpoint, the difference between ACT guided treatment usual care wasn’t statistically significant (P-value = 0.897) in the annual moderate/severe asthma exacerbations rate.

 For post-hoc efficacy endpoints, the difference of treatments were statistically significant for the percentage of subjects who had an ACT total score ≥20 or an improvement ≥3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period (OR [95% CI] = 6.89 (1.28, 37.02); P-value = 0.025) and the time to first ACT total score ≥20 and an improvement ≥3 Points in ACT total score over the 24-week treatment period (HR [95% CI] = 1.708 (1.078, 2.707); P-value = 0.023).

   59 CONFIDENTIAL 2019N423508_00 201097

7. SAFETY RESULTS

7.1. Adverse Events

7.1.1. Overview of AEs

The overview of AEs was displayed in Table 16.

A total of 215 (42.5%) subjects experienced at least one AE in which the incidence of ACT group (86 subjects, 35.7%) was lower than that in UC group (129 subjects, 48.7%). There was 3 (0.6%) subjects experienced the adverse event related to study process, and the incidence of ACT group (2 subjects, 0.8%) was little higher than that in UC group (1 subject, 0.4%). None of subjects in ACT group experienced the adverse event related to GSK’s drug, but one (0.4%) subject in UC group experienced this kind of adverse event.

A total of 205 (40.5%) subjects experienced at least one non-serious AE, and the incidence of ACT group (82 subjects, 34.0%) was lower than that in UC group (123 subjects, 46.4%).

A total of 22 (4.3%) subjects experienced at least one SAE in which the incidence of ACT group (9 subjects, 3.7%) was lower than that in UC group (13 subjects, 4.9%). None of SAE was related to study process or GSK’s drug.

Only one subject experienced at least one AE leading to temporarily suspend or delay GSK's drugs which was in UC group, and a total of 8 (1.6%) subjects experienced at least one AE leading to withdrawal in which the incidence of ACT group (5 subjects, 2.1%) was higher than that in UC group (3 subjects, 1.1%).

A total of 17 (3.4%) subjects experienced asthma exacerbations in which the incidence of ACT group (8 subjects, 3.3%) was lower than that in UC group (9 subjects, 3.4%).

A total of one subject wad dead which was in ACT group.

Source Listing 3.01 displays the listing of subjects’ adverse events.

Table 16 Overview of Adverse Events (Safety Population)

Number (%) of Subjects ACT UC Total N=241 N=265 N=506 AEs 86 (35.7%) 129 (48.7%) 215 (42.5%) Any AEs related to study process 2 (0.8%) 1 (0.4%) 3 (0.6%) Any AEs related to GSK's drug 0 1 (0.4%) 1 (0.2%) Non-serious AEs 82 (34.0%) 123 (46.4%) 205 (40.5%) SAEs 9 (3.7%) 13 (4.9%) 22 (4.3%) SAEs related to study process 0 0 0 SAEs related to GSK's drugs 0 0 0

   60 CONFIDENTIAL 2019N423508_00 201097

Number (%) of Subjects ACT UC Total N=241 N=265 N=506 Any AEs leading to temporarily 0 1 (0.4%) 1 (0.2%) suspend\delay GSK's drugs Any AEs leading to withdrawal 5 (2.1%) 3 (1.1%) 8 (1.6%) Asthma exacerbations 8 (3.3%) 9 (3.4%) 17 (3.4%) Death 1 (0.4%) 0 1 (0.2%) Source: Table 3.01

Of all AEs, the most common reported AEs was Upper respiratory tract infection (ACT: 7.5%; UC: 17.7%) and Bronchitis (ACT: 1.2%; UC: 6.0%) and Cough (ACT: 4.1%; UC: 3.0%) and Oropharyngeal pain (ACT: 2.5%; UC: 3.4%), and for other AEs, the incidences were all lower than 3% (Source Table 3.02). The 4 most common AEs in each treatment group are shown in Table 17.

Table 17 Adverse Events with Incidence ≥3% in Any Treatment Group (Safety Population)

Number (%) of Subjects ACT UC Total N=241 N=265 N=506 Upper respiratory tract infection 18 (7.5%) 47 (17.7%) 65 (12.8%) Bronchitis 3 (1.2%) 16 (6.0%) 19 (3.8%) Cough 10 (4.1%) 8 (3.0%) 18 (3.6%) Oropharyngeal pain 6 (2.5%) 9 (3.4%) 15 (3.0%) Source: Table 3.02

7.1.2. Adverse Events Related to Study Process or GSK’s Drugs

The AEs related to study process or GSK’s drugs was displayed in Table 18.

The AEs related to study process reported in ACT group were Asthma (0.4%) and Cough (0.4%) and Throat irritation (0.4%), and those reported in UC group was Oropharyngeal pain (0.4%).

The AEs related to GSK’s drugs was only reported in UC group which was Rash (0.4%).

Table 18 Adverse Events Related to Study Process or GSK’s Drugs of the Subjects (Safety Population)

Number (%) of Subjects ACT UC Total N=241 N=265 N=506 Adverse Events Related to Study Process 2 (0.8%) 1 (0.4%) 3 (0.6%) Asthma 1 (0.4%) 0 1 (0.2%) Cough 1 (0.4%) 0 1 (0.2%)

   61 CONFIDENTIAL 2019N423508_00 201097

Number (%) of Subjects ACT UC Total N=241 N=265 N=506 Throat irritation 1 (0.4%) 0 1 (0.2%) Oropharyngeal pain 0 1 (0.4%) 1 (0.2%) Adverse Events Related to GSK’s Drugs 0 1 (0.4%) 1 (0.2%) Rash 0 1 (0.4%) 1 (0.2%) Source: Table 3.03 and Table 3.04

7.2. Serious and Other Significant Adverse Events

7.2.1. Deaths

There was one subject which experienced Epilepsy and dead which wasn’t related to study process or GSK’s drugs, and the subject received the ACT guided treatment (Source Table 3.05).

7.2.2. Other Serious Adverse Events

The non-fatal SAEs were also summarized in Source Table 3.05. The incidence of non- fatal SAEs were lower than 1%. None of SAE was related to study process (Source Table 3.06) or GSK’s drug (Source Table 3.07).

Source Listing 3.02 displays the listing of subjects’ SAEs.

7.2.3. Other Significant Adverse Events

There were no other significant adverse events.

7.3. Non-Serious Adverse Events of the Subjects.

Of all non-SAEs, the most common reported non-SAEs was Upper respiratory tract infection (ACT: 7.5%; UC: 17.7%) and Bronchitis (ACT: 1.2%; UC: 6.0%) and Cough (ACT: 4.1%; UC: 3.0%) and Oropharyngeal pain (ACT: 2.5%; UC: 3.4%), and for other AEs, the incidences were all lower than 3% (Source Table 3.02). The 4 most common AEs in each treatment group are shown in Table 19.

Table 19 Non-SAEs with Incidence ≥3% in Any Treatment Group (Safety Population)

Number (%) of Subjects ACT UC Total N=241 N=265 N=506 Upper respiratory tract infection 18 (7.5%) 47 (17.7%) 65 (12.8%) Bronchitis 3 (1.2%) 16 (6.0%) 19 (3.8%) Cough 10 (4.1%) 8 (3.0%) 18 (3.6%)

   62 CONFIDENTIAL 2019N423508_00 201097

Number (%) of Subjects ACT UC Total N=241 N=265 N=506 Oropharyngeal pain 6 (2.5%) 9 (3.4%) 15 (3.0%) Source: Table 3.08

7.4. Adverse Events Leading to Temporarily Suspend\Delay GSK’s drugs

Only one subject experienced AE leading to temporarily suspend or delay GSK's drugs which was in UC group and the AE was reported Urticaria (0.4%) (Source Table 3.09).

7.5. Adverse Events Leading to Withdrawal

The AEs leading to withdrawal was summarized in Source Table 3.10. The incidence of Asthma was 1.4% and the incidence of other AEs were all lower than 1%.

7.6. Other Safety Evaluations

Some subjects were reported abnormal and clinical significance physical examinations which were summarized in Table 20.

At baseline, the incidence in ACT group (5.0%) was higher than that in UC group (1.1%) for abnormal and clinical significance lung examination, and the incidence in ACT group was lower than that in UC group for abnormal and clinical significance cardiovascular system and nose examinations. The balance between ACT group and UC group wasn’t well controlled.

At post-baseline visits exclude week 8, the incidences in ACT group (5.0%) were all lower than that in UC group for abnormal and clinical significance change or new abnormity since last visit.

Source Listing 3.03 displays the listing of subjects’ physical examinations.

Table 20 Abnormal and clinical significance Physical Examination (Safety Population)

Number (%) of Subjects ACT UC Total N=241 N=265 N=506 Baseline Lung N-observed 241 264 505 n (%) 12 (5.0%) 3 (1.1%) 15 (3.0%) Cardiovascular System N-observed 241 264 505 n (%) 0 2 (0.8%) 2 (0.4%) Nose N-observed 241 264 505

   63 CONFIDENTIAL 2019N423508_00 201097

Number (%) of Subjects ACT UC Total N=241 N=265 N=506 n (%) 1 (0.4%) 4 (1.5%) 5 (1.0%) Change or new abnormity since last visit Week 4 N-observed 236 262 498 n (%) 0 1 (0.4%) 1 (0.2%) Week 8 N-observed 226 249 475 n (%) 1 (0.4%) 0 1 (0.2%) Week 12 N-observed 215 245 460 n (%) 0 1 (0.4%) 1 (0.2%) Week 16 N-observed 215 240 455 n (%) 0 4 (1.7%) 4 (0.9%) Week 20 N-observed 205 235 440 n (%) 0 7 (3.0%) 7 (1.6%) Week 24/ Withdrawal N-observed 223 247 470 n (%) 2 (0.9%) 4 (1.6%) 6 (1.3%) Source: Table 3.11

7.7. Safety Conclusions

 Overall, ACT guided treatment were well tolerated in which the incidences of all AEs and the incidence of asthma exacerbations was all lower than that with usual care. None of subjects with ACT guided treatment experienced any SAEs related to study process or GSK’s drug. None of subjects with ACT guided treatment experienced any AEs leading to temporarily suspend \ delay GSK's drugs. The most common AEs with Incidence ≥3% were Upper respiratory tract infection and Bronchitis and Cough and Oropharyngeal pain.

 The incidence of AEs related to study process or GSK’s drugs was low and similar between ACT guided treatment and usual care.

 Although there was reported death for one subject with ACT guided treatment, the event wasn’t related to study process or GSK’s drug.

 The change or new abnormity since last visit of abnormal and clinical significance were minor and the incidences of subjects with ACT guided treatment were lower than that with usual care.

   64 CONFIDENTIAL 2019N423508_00 201097

8. DISCUSSION AND CONCLUSIONS

8.1. Discussion

Current levels of asthma control in China fall short of the goals specified in international guidelines for asthma management (GINA). There is a large gap between what can be achieved with modern asthma management and what is currently being achieved. Now a very simple method of asthma control have been developed which is Asthma Control Test (ACT) aiming to improve the assessment quickly. ACT was designed to assess the level of asthma control in patients and so guide physicians treatment decisions. This study is aimed at evaluating the effectiveness of ACT guided treatment compared with usual care in asthma patients in China. It is designed to assist Chinese patients and physicians improving adherence to the guidelines through the inclusion of the ACT in the patient’s asthma management plan.

This study was conducted from 26Aug2016 to 09Aug2019 in 12 centers in Chinese with asthma.

The primary efficacy endpoint for this study was the percentage of subjects who had an ACT total score ≥20 or an improvement of more than 3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period. Compared with usual care, the adjusted odds ratio (OR) of response for ACT total score ≥20 or an improvement > 3 points in ACT total score was statistically significant higher using primary analysis method, which is in favor of ACT guided treatment.The supportive analysis population (PP) also provided the statistically significant difference between ACT guided treatment and usual care. Therefore, the planned primary efficiency objectives were met. This study has demonstrated the effectiveness of ACT guided treatment for China adult asthma patients for the primary efficacy endpoint.

Additionally, there wasn’t statistically significant difference between treatment and covariates including the type of baseline controller and gender and age. Sensitivity and supportive analysis results also didn’t provide the statistically significant difference between ACT guided treatment and usual care using GEEs method or using subgroup analysis method exclude male group in which difference was statistically significant. Therefore, no enough results can demonstrate that ACT guided treatment would be deviated by other factors.

Although most of the differences weren’t statistically significant between ACT guided treatment and usual care for the time to first ACT total score ≥20 or (and) an improvement >3 (≥3) Points in ACT total score over the 24-week treatment period and for other non-primary efficacy endpoints, the differences were statistically significant at each visit for the change from baseline in ACT total score.

Results of the study also showed that both ACT guided treatment and usual care were well tolerated over the 24-week treatment period. None of subjects experienced any SAEs related to study process or GSK’s drug. Death were reported during the treatment period, but it wasn’t related to study process or GSK’s drug. Therefore, the potential risk of clinical significance was very low.

   65 CONFIDENTIAL 2019N423508_00 201097

8.2. Conclusions

This study has demonstrated the superiority of ACT guided treatment over usual care for the ACT total score ≥20 or an improvement of more than 3 points in ACT total score for China adult asthma patients and ACT guided treatment was well tolerated in this study with no safety concerns identified.

   66 CONFIDENTIAL 2019N423508_00 201097

9. REFERENCES

Bateman ED, Boushey HA, Bousquet J, Busse WW, Clark TJH, Pauwels RA, et al. Can guideline defined asthma control be achieved? Am J Respir Crit Care Med 2004; 170:836–844.

Chapman K.R., Boulet L.P., Rea R.M., Franssen E.. Suboptimal asthma control: prevalence, detection and consequences in general practice. Eur Respir J 2008; 31: 320– 325.

Nan Su, Jiangtao Lin, Ping Chen, Jing Li, Changgui Wu, Kaisheng Yin, et al. Evaluation of asthma control and patient’s perception of asthma: findings and analysis of a nationwide questionnaire-based survey in China. J Asthma, 2013; 50(8): 861–870.

Nathan RA, Sorkness CA, Kosinski M, Schatz M, Li JT, Marcus P, et al. Development of the asthma control test: a survey for assessing asthma control. J Allergy Clin Immunol 2004;113(1):59-65.

   67 CONFIDENTIAL 2019N423508_00 201097

10. POST-TEXT TABLES AND FIGURES

Table 1.01 Subjects Enrollment And Study Completion

Table 1.02 Summary of Number of Subjects by Centre

Table 1.03 Analysis Population (Screening/Randomized Subject)

Table 1.04 Subjects with Important Protocol Deviation (Screening/Randomized Subjects)

Table 1.05 Subjects’ Demographic and Other Baseline Characteristics (ITT)

Table 1.06 Subjects’ History of Smoking, Alcohol Abuse, Medication Abuse, Asthma, and Asthma Exacerbation (ITT)

Table 1.07 Subjects’ Past and Current Illness (ITT)

Table 1.08 Subjects’ Asthma Medication (ITT)

Table 1.09 Subjects’ Concurrent Medication (ITT)

Table 2.01 Summary of the Statistical Analysis of Percentage of Subjects Who Have an ACT Total Score >= 20 or an Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment during the 24-Weeks Treatment Period (ITT)

Table 2.02 Summary of the Statistical Analysis of Percentage of Subjects Who Have an ACT Total Score >= 20 or an Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment during the 24-Weeks Treatment Period (PP)

Table 2.03 Summary of the Statistical Analysis of Percentage of Subjects Who Have an ACT Total Score >= 20 or an Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment during the 24-Weeks Treatment Period by Subgroup (ITT)

Table 2.04 Summary of the Statistical Analysis of Percentage of Subjects Who Have an ACT Total Score >= 20 or an Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment during the 24-Weeks Treatment Period by Subgroup (PP)

Table 2.05 Summary of the Statistical Analysis of Percentage of Subjects Who Have an ACT Total Score >= 20 or an Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (ITT)

Table 2.06 Summary of the Statistical Analysis of Percentage of Subjects Who Have an ACT Total Score >= 20 or an Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (PP)

Table 2.07 Summary of the Statistical Analysis of Percentage of Subjects Who Have an ACT Total Score >= 20 and an Improvement >3 Points in ACT Total Score (ITT)

   68 CONFIDENTIAL 2019N423508_00 201097

Table 2.08 Summary of the Statistical Analysis of Percentage of Subjects Who Have an ACT Total Score >= 20 and an Improvement >3 Points in ACT Total Score (PP)

Table 2.09 Summary of the Statistical Analysis of Percentage of Subjects Who Have an ACT Total Score >= 20 and an Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (ITT)

Table 2.10 Summary of the Statistical Analysis of Percentage of Subjects Who Have an ACT Total Score >= 20 and an Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (PP)

Table 2.11 Summary of the Statistical Analysis of Change from Baseline in ACT Total Score (ITT)

Table 2.12 Summary of the Statistical Analysis of Change from Baseline in ACT Total Score (PP)

Table 2.13 Summary of Mean Daytime Symptom Score over the 24-Week Treatment Period (ITT)

Table 2.14 Analysis of Mean Daytime Symptom Score over the 24-Week Treatment Period (ITT)

Table 2.15 Summary of Mean Night-time Symptom Score over the 24-Week Treatment Period (ITT)

Table 2.16 Analysis of Mean Night-time Symptom Score Over the 24-Week Treatment Period (ITT)

Table 2.17 Summary of the Change from Baseline in Forced Expiratory Volume in One Second (ITT)

Table 2.18 Analysis of the Change from Baseline in Forced Expiratory Volume In one Second (ITT)

Table 2.19 Summary of Mean Morning Peak Expiratory Flow over the 24-Week Treatment Period (ITT)

Table 2.20 Analysis of Mean Morning Peak Expiratory Flow over the 24-Week Treatment Period (ITT)

Table 2.21 Summary of Mean Evening Peak Expiratory Flow over the 24-Week Treatment Period (ITT)

Table 2.22 Analysis of Mean Evening Peak Expiratory Flow over the 24-Week Treatment Period (ITT)

Table 2.23 Summary of the Change from Baseline in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score (ITT)

   69 CONFIDENTIAL 2019N423508_00 201097

Table 2.24 Analysis of the Change from Baseline in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score (ITT)

Table 2.25 Time to First ACT Total Score ≥20 or Improvement of More Than 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT)

Table 2.26 Time to First ACT Total Score ≥20 and Improvement of More Than 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT)

Table 2.27 Summary of Moderate/Severe Asthma Exacerbations (ITT)

Table 2.28 Analysis of Moderate/Severe Asthma Exacerbations (ITT)

Table 2.29 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >=3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period (ITT)

Table 2.30 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >=3 Points in ACT Total Score (ITT)

Table 2.31 Time to First ACT Total Score ≥20 or Improvement >= 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT)

Table 2.32 Time to First ACT Total Score ≥20 and Improvement >= 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT)

Table 3.01 Summary of Adverse Event (Safety Population)

Table 3.02 Adverse Events of the Subjects (Safety Population)

Table 3.03 Adverse Events Related to Study Process of the Subjects (Safety Population)

Table 3.04 Adverse Events Related to GSK’s Drugs of the Subjects (Safety Population)

Table 3.05 Serious Adverse Events of the Subjects (Safety Population)

Table 3.06 Serious Adverse Events Related to Study Process of the Subjects (Safety Population)

Table 3.07 Serious Adverse Events Related to GSK’s Drugs of the Subjects (Safety Population)

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population)

Table 3.09 Adverse Events Leading to Temporarily Suspend\Delay GSK’s drugs (Safety Population)

Table 3.10 Adverse Events Leading To Withdrawal (Safety Population)

Table 3.11 Physical Examination of the subjects (Safety Population)

   70 CONFIDENTIAL 2019N423508_00 201097

Listing 1.01 Subjects Distribution

Listing 1.02 Analysis Population (Subjects Who Signed the First Informed Consent Form)

Listing 1.03 Protocol Deviation (ITT)

Listing 1.04 Subjects’ Demographic and Other Baseline Characteristics (ITT)

Listing 1.05 Subjects’ History of Smoking, Alcohol Abuse and Drug Abuse (ITT)

Listing 1.06 Subjects’ History of Asthma and Asthma exacerbation (ITT)

Listing 1.07 Subjects’ Past and Current Illness (ITT)

Listing 1.08 Subjects’ Current Asthma Medication Record (ITT)

Listing 1.09 Asthma Treatment (ITT)

Listing 1.10 Subjects’ Concurrent Medication (ITT)

Listing 2.01 ACT Score (ITT)

Listing 2.02 Subjects’ Log (ITT)

Listing 2.03 Subjects’ Lung Function Examination (ITT)

Listing 2.04 Subjects’ Asthma Quality Of Life Questionnaire (AQLQs) (ITT)

Listing 3.01 Subjects’ Adverse Events (SA)

Listing 3.02 Subjects’ Serious Adverse Event (SA)

Listing 3.03 Subjects’ Physical Examination (SA)

Figure 1 Forest Plot of Adjusted Odds Ratios and 95% CIs from GLMM, Comparing ACT Guided Treatment with Usual Care on Primary Analysis and Sensitivity Analysis of ACT Response (ITT)

Figure 2 Forest Plot of Adjusted Odds Ratios and 95% CIs from GLMM, Comparing ACT Guided Treatment with Usual Care on Primary Analysis and Sensitivity Analysis of ACT Response (PP)

Figure 3 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (ITT)

Figure 4 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total

   71 CONFIDENTIAL 2019N423508_00 201097

Score >= 20 Or An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (PP)

Figure 5 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (ITT)

Figure 6 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (PP)

Figure 7 Forest Plot of LS Mean change from baseline treatment differences and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (ITT)

Figure 8 Forest Plot of LS Mean change from baseline treatment differences and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (PP)

Figure 9 Forest Plot of LS Mean treatment differences and 95% CIs from MMRM, Comparing ACT Guided Treatment with Usual Care on Mean Daytime Symptom Score by Visits (ITT)

Figure 10 Forest Plot of LS Mean treatment differences and 95% CIs from MMRM, Comparing ACT Guided Treatment with Usual Care on Mean Night-time Symptom Score by Visits (ITT)

Figure 11 Forest Plot of LS Mean change from baseline treatment differences and 95% CIs from Mixed Model, Comparing ACT Guided Treatment with Usual Care on FEV1 (ITT)

Figure 12 Forest Plot of LS Mean treatment differences and 95% CIs from MMRM, Comparing ACT Guided Treatment with Usual Care on Mean Morning PEF by Visits (ITT)

Figure 13 Forest Plot of LS Mean treatment differences and 95% CIs from MMRM, Comparing ACT Guided Treatment with Usual Care on Mean Evening PEF by Visits (ITT)

Figure 14 Forest Plot of LS Mean change from baseline treatment differences and 95% CIs from Mixed Model, Comparing ACT Guided Treatment with Usual Care on AQLQ(S) score (ITT)

Figure 15 KM Figure of Time to Asthma Control (ACT Total Score ≥20 Or Improvement of More Than 3 Points in ACT Total Score) For the First Time (ITT)

   72 CONFIDENTIAL 2019N423508_00 201097

Figure 16 KM Figure of Time to Asthma Control (ACT Total Score ≥20 And Improvement of More Than 3 Points in ACT Total Score) For the First Time (ITT)

Figure 17 Forest Plot of Adjusted Odds Ratios and 95% CIs from GLMM, Comparing ACT Guided Treatment with Usual Care on Primary Analysis and Sensitivity Analysis of ACT Response (ITT)

Figure 18 KM Figure of Time to Asthma Control (ACT Total Score ≥20 Or Improvement >= 3 Points in ACT Total Score) For the First Time (ITT)

Figure 19 KM Figure of Time to Asthma Control (ACT Total Score ≥20 And Improvement >= 3 Points in ACT Total Score) For the First Time (ITT)

   73 CONFIDENTIAL 2019N423508_00 201097

11. CASE NARRATIVES

Case narratives are presented for deaths (Section 7.2.1).

There may be minor discrepancies in the details of the SAEs included in the clinical narratives compared with the safety tabulations. This is because the data comes from 2 different databases (i.e., locked clinical trials database and dynamic SAE database) and has been collected at different points in time. However, all key data points are reconciled. It is considered that these minor discrepancies do not change the overall clinical significance or understanding of the SAE.

Protocol ID: 201097

Subject number: PPD

Treatment number:PPD

Suspect drugs: None

Serious Events: Epilepsy

This 38-year-old female subject (Subject PPD signed the first informed consent and completed the screening successfully and was enrolled in the ACT guided treatment group on 16Jan2017.

The subject received FLUTICASONE PROPIONATE &SALMETEROL XINAFOATE twice one day and SALBUTAMOL as needed from 16th Jan 2017 to PPD 2017, for asthma treatment.

On PPD 2017, the subject experienced Fever which developed mild.

On PPD 2017, the subject went to out-patient department in PPD None clinically significant abnormal results were observed for hematologic and C-reactive protein examinations, and then was treated with SULBENICILLIN and UNSPECIFIED HERBAL AND TRADITIONAL MEDICINE by intravenous injection, and then went back home in the same day. The adverse event wasn’t related to study process or GSK’s drug by investigator. Therefore, the dose of GSK’s drug wasn’t changed.

On the early morning of PPD 2017, the subject experienced Epilepsy and then went to emergency department in PPD but finally she was reported death at PPD

The primary reason leading to death might be Epilepsy.

   74 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol: 201097 Date： 29SEP2019

Table 1.01 Subjects Enrollment And Study Completion (Page 1 of 1) ______ACT UC Total

Screening 583 Screening failure 53 Screening succeed/randomization 252 278 530

Complete the study 209 ( 82.9%) 234 ( 84.2%) 443 ( 83.6%) Withdrawal, the reasons are as follows [1] 43 ( 17.1%) 44 ( 15.8%) 87 ( 16.4%) A subject is significantly non-compliant with treatment 13 ( 30.2%) 10 ( 22.7%) 23 ( 26.4%) or the requirments of the protocol Lost to follow up 3 ( 7.0%) 6 ( 13.6%) 9 ( 10.3%) Subject's decision not to participate any further 11 ( 25.6%) 15 ( 34.1%) 26 ( 29.9%) (withdrawal of consent) In the investigator's opinion, it is in the subject's best 2 ( 4.7%) 0 2 ( 2.3%) interest Subjects experience a protocol-defined asthma 8 ( 18.6%) 8 ( 18.2%) 16 ( 18.4%) exacerbation The study is terminated by the Sponsor or designee 0 0 0 Pregnancy 0 0 0 Others 6 ( 14.0%) 5 ( 11.4%) 11 ( 12.6%)

[1] Percentages for withdrawal reasons are based on subjects withdrawal.

PPD

75 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 1.02 Summary of Number of Subjects by Centre (Page 1 of 1) ______ACT UC Total Centre ID N=252 N=278 N=530 PPD 0 52 ( 18.7%) 52 ( 9.8%) 46 ( 18.3%) 0 46 ( 8.7%) 0 52 ( 18.7%) 52 ( 9.8%) 0 44 ( 15.8%) 44 ( 8.3%) 0 43 ( 15.5%) 43 ( 8.1%) 37 ( 14.7%) 0 37 ( 7.0%) 0 43 ( 15.5%) 43 ( 8.1%) 0 44 ( 15.8%) 44 ( 8.3%) 38 ( 15.1%) 0 38 ( 7.2%) 45 ( 17.9%) 0 45 ( 8.5%) 48 ( 19.0%) 0 48 ( 9.1%) 38 ( 15.1%) 0 38 ( 7.2%)

PPD

76 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol: 201097 Date： 29SEP2019

Table 1.03 Analysis Population (Screening/Randomized Subject) (Page 1 of 1) ______ACT UC Total N=252 N=278 N=530

Intent-to-Treat (ITT) Population 242( 96.0%) 265( 95.3%) 507( 95.7%) Per Protocol (PP) Population 208( 82.5%) 231( 83.1%) 439( 82.8%) Safety Population 241( 95.6%) 265( 95.3%) 506( 95.5%)

PPD

77 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 30SEP2019

Table 1.04 Subjects with Important Protocol Deviation(Screening/Randomized Subjects) (Page 1 of 1) ______ACT UC Total N=252 N=278 N=530 At least one Protocol deviation 68 ( 27.0%) 49 ( 17.6%) 117 ( 22.1%) 1 Informed Consent 2 ( 0.8%) 0 2 ( 0.4%) 2 Eligibility Criteria 15 ( 6.0%) 19 ( 6.8%) 34 ( 6.4%) 3 Withdrawal Criteria 2 ( 0.8%) 4 ( 1.4%) 6 ( 1.1%) 5 Visit Completion 24 ( 9.5%) 14 ( 5.0%) 38 ( 7.2%) 6 SAE/Pregnancy/Liver Events 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) 7 Assessment or Time Point completion 15 ( 6.0%) 11 ( 4.0%) 26 ( 4.9%) 8 Wrong Study Treatment/Administration/Dose 35 ( 13.9%) 6 ( 2.2%) 41 ( 7.7%) 9 Study Procedure 4 ( 1.6%) 0 4 ( 0.8%)

PPD

78 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol: 201097 Date： 29SEP2019

Table 1.05 Subjects' Demographic And Other Baseline Characteristics(ITT) (Page 1 of 4) ______ACT UC Total N=242 N=265 N=507 Age (years) n 242 265 507 Mean 48.0 48.0 48.0 SD 12.81 13.43 13.13 Median 49.0 50.0 49.0 Min 18 18 18 Max 73 70 73

Gender n 242 265 507 Male 127 ( 52.5%) 130 ( 49.1%) 257 ( 50.7%) Female 115 ( 47.5%) 135 ( 50.9%) 250 ( 49.3%)

Pregnancy possibility n 115 135 250 Postmenopausal 56 ( 48.7%) 72 ( 53.3%) 128 ( 51.2%) Sterilization 14 ( 12.2%) 2 ( 1.5%) 16 ( 6.4%) (childbearing age) Pregnancy possible 45 ( 39.1%) 61 ( 45.2%) 106 ( 42.4%)

Race n 242 265 507 Han nationality 237 ( 97.9%) 261 ( 98.5%) 498 ( 98.2%) Non-Han nationality 5 ( 2.1%) 4 ( 1.5%) 9 ( 1.8%)

PPD

79 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol: 201097 Date： 29SEP2019

Table 1.05 Subjects' Demographic And Other Baseline Characteristics(ITT) (Page 2 of 4) ______ACT UC Total N=242 N=265 N=507 Height (cm) n 242 265 507 Mean 165.24 165.60 165.43 SD 7.820 8.238 8.035 Median 165.00 165.00 165.00 Min 145.0 145.0 145.0 Max 186.0 187.0 187.0

Weight (kg) n 242 265 507 Mean 65.89 64.90 65.37 SD 11.074 11.694 11.402 Median 65.00 64.00 65.00 Min 33.0 41.0 33.0 Max 105.0 100.0 105.0

BMI (kg/m2) n 242 265 507 Mean 24.06 23.56 23.80 SD 3.246 3.141 3.199 Median 24.00 23.30 23.70 Min 15.7 16.0 15.7 Max 36.3 34.3 36.3

PPD

80 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol: 201097 Date： 29SEP2019

Table 1.05 Subjects' Demographic And Other Baseline Characteristics(ITT) (Page 3 of 4) ______ACT UC Total N=242 N=265 N=507 Temperature (C) n 242 265 507 Mean 36.60 36.57 36.58 SD 0.239 0.277 0.260 Median 36.60 36.50 36.60 Min 36.0 36.0 36.0 Max 37.2 37.2 37.2

Pulse Rate (beats/min) n 242 265 507 Mean 78.2 78.5 78.3 SD 7.37 7.38 7.37 Median 78.0 78.0 78.0 Min 57 59 57 Max 104 112 112

Systolic Blood Pressure n 242 265 507 （mmHg） Mean 122.4 120.4 121.4 SD 9.86 11.68 10.89 Median 120.0 120.0 120.0 Min 90 90 90 Max 150 180 180

PPD

81 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol: 201097 Date： 29SEP2019

Table 1.05 Subjects' Demographic And Other Baseline Characteristics(ITT) (Page 4 of 4) ______ACT UC Total N=242 N=265 N=507 Diastolic Blood Pressure n 242 265 507 （mmHg） Mean 77.2 76.6 76.9 SD 7.39 7.72 7.56 Median 80.0 78.0 79.0 Min 53 50 50 Max 98 100 100

PPD

82 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 29SEP2019

Table 1.06 Subjects' History of Smoking, Alcohol Abuse, Medication Abuse, Asthma, And Asthma Exacerbation (ITT) (Page 1 of 3) ______ACT UC Total N=242 N=265 N=507 Smoking history n 242 265 507 Never smoking 208 ( 86.0%) 227 ( 85.7%) 435 ( 85.8%) Smoking 1 ( 0.4%) 0 1 ( 0.2%) Quit smoking 33 ( 13.6%) 38 ( 14.3%) 71 ( 14.0%)

Packs/Year n 32 34 66 Mean 4.9 5.6 5.3 SD 5.23 6.94 6.14 Median 4.5 3.5 4.0 Min 1.0 1.0 1.0 Max 30.0 40.0 40.0

Alcohol Abuse history n 242 265 507 Yes 0 0 0 No 242 (100.0%) 265 (100.0%) 507 (100.0%)

Medication Abuse history n 242 265 507 Yes 0 0 0 No 242 (100.0%) 265 (100.0%) 507 (100.0%)

PPD

83 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 29SEP2019

Table 1.06 Subjects' History of Smoking, Alcohol Abuse, Medication Abuse, Asthma, And Asthma Exacerbation (ITT) (Page 2 of 3) ______ACT UC Total N=242 N=265 N=507 Asthma history n 242 265 507 Severity Mild 117 ( 48.3%) 159 ( 60.0%) 276 ( 54.4%) Moderate 104 ( 43.0%) 75 ( 28.3%) 179 ( 35.3%) Severe 21 ( 8.7%) 31 ( 11.7%) 52 ( 10.3%)

ICS continuous treatment time (months) n 242 265 507 Mean 19.3 25.0 22.3 SD 29.10 40.33 35.50 Median 7.0 7.0 7.0 Min 0.0 0.0 0.0 Max 216.0 288.0 288.0

PPD

84 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 29SEP2019

Table 1.06 Subjects' History of Smoking, Alcohol Abuse, Medication Abuse, Asthma, And Asthma Exacerbation (ITT) (Page 3 of 3) ______ACT UC Total N=242 N=265 N=507 Asthma Exacerbation history n 242 265 507 No 205 ( 84.7%) 209 ( 78.9%) 414 ( 81.7%) Yes 37 ( 15.3%) 56 ( 21.1%) 93 ( 18.3%)

Disease attack number n 37 56 93 0 0 0 0 1 24 ( 64.9%) 38 ( 67.9%) 62 ( 66.7%) 2 7 ( 18.9%) 13 ( 23.2%) 20 ( 21.5%) >=3 6 ( 16.2%) 5 ( 8.9%) 11 ( 11.8%)

Hospitalization number n 37 56 93 0 20 ( 54.1%) 43 ( 76.8%) 63 ( 67.7%) 1 16 ( 43.2%) 9 ( 16.1%) 25 ( 26.9%) 2 1 ( 2.7%) 4 ( 7.1%) 5 ( 5.4%) >=3 0 0 0

PPD

85 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 1 of 15) ______ACT UC Total N=242 N=265 N=507

Past Illness 99 ( 40.9%) 108 ( 40.8%) 207 ( 40.8%) Rhinitis 37 ( 15.3%) 26 ( 9.8%) 63 ( 12.4%) Appendicitis 6 ( 2.5%) 3 ( 1.1%) 9 ( 1.8%) Pharyngitis 5 ( 2.1%) 2 ( 0.8%) 7 ( 1.4%) Arthritis 4 ( 1.7%) 0 4 ( 0.8%) Cholelithiasis 4 ( 1.7%) 3 ( 1.1%) 7 ( 1.4%) Rhinitis allergic 4 ( 1.7%) 3 ( 1.1%) 7 ( 1.4%) Sinusitis 4 ( 1.7%) 2 ( 0.8%) 6 ( 1.2%) Cerebral infarction 3 ( 1.2%) 1 ( 0.4%) 4 ( 0.8%) Cholecystitis 3 ( 1.2%) 1 ( 0.4%) 4 ( 0.8%) Chronic gastritis 3 ( 1.2%) 1 ( 0.4%) 4 ( 0.8%) Gastritis 3 ( 1.2%) 1 ( 0.4%) 4 ( 0.8%) Hypertension 3 ( 1.2%) 10 ( 3.8%) 13 ( 2.6%) Urticaria 3 ( 1.2%) 3 ( 1.1%) 6 ( 1.2%) Uterine leiomyoma 3 ( 1.2%) 2 ( 0.8%) 5 ( 1.0%) Vertigo 3 ( 1.2%) 2 ( 0.8%) 5 ( 1.0%) Bronchitis 2 ( 0.8%) 0 2 ( 0.4%) Cervicitis 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Conjunctivitis 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) Fibroadenoma of breast 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%)

PPD

86 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 2 of 15) ______ACT UC Total N=242 N=265 N=507

Functional gastrointestinal disorder 2 ( 0.8%) 0 2 ( 0.4%) Gastrooesophageal reflux disease 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) Gout 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Intervertebral disc protrusion 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Iron deficiency anaemia 2 ( 0.8%) 0 2 ( 0.4%) Nasal polyps 2 ( 0.8%) 4 ( 1.5%) 6 ( 1.2%) Nasal septum deviation 2 ( 0.8%) 0 2 ( 0.4%) Nephrolithiasis 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Otitis media 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Thyroid mass 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) Thyroiditis 2 ( 0.8%) 0 2 ( 0.4%) Urinary tract infection 2 ( 0.8%) 0 2 ( 0.4%) Vaginal infection 2 ( 0.8%) 0 2 ( 0.4%) Abortion 1 ( 0.4%) 0 1 ( 0.2%) Acne 1 ( 0.4%) 0 1 ( 0.2%) Arrhythmia 1 ( 0.4%) 0 1 ( 0.2%) Atrial septal defect 1 ( 0.4%) 0 1 ( 0.2%) Back pain 1 ( 0.4%) 0 1 ( 0.2%) Blepharitis 1 ( 0.4%) 0 1 ( 0.2%) Breast cancer 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%)

PPD

87 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 3 of 15) ______ACT UC Total N=242 N=265 N=507

Bronchitis chronic 1 ( 0.4%) 0 1 ( 0.2%) Caesarean section 1 ( 0.4%) 0 1 ( 0.2%) Cataract 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Cholecystectomy 1 ( 0.4%) 0 1 ( 0.2%) Cleft lip 1 ( 0.4%) 0 1 ( 0.2%) Colitis 1 ( 0.4%) 0 1 ( 0.2%) Conjunctivitis allergic 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Constipation 1 ( 0.4%) 0 1 ( 0.2%) Coronary artery disease 1 ( 0.4%) 0 1 ( 0.2%) Cubital tunnel syndrome 1 ( 0.4%) 0 1 ( 0.2%) Dental caries 1 ( 0.4%) 0 1 ( 0.2%) Dermatitis allergic 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Duodenal ulcer 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Eczema 1 ( 0.4%) 4 ( 1.5%) 5 ( 1.0%) Epiglottic cyst 1 ( 0.4%) 0 1 ( 0.2%) Epilepsy 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Gallbladder disorder 1 ( 0.4%) 0 1 ( 0.2%) Gastric ulcer 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Gastric ulcer haemorrhage 1 ( 0.4%) 0 1 ( 0.2%) Genital infection female 1 ( 0.4%) 0 1 ( 0.2%)

PPD

88 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 4 of 15) ______ACT UC Total N=242 N=265 N=507

Haemangioma 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Haemangioma of liver 1 ( 0.4%) 0 1 ( 0.2%) Haemorrhoids 1 ( 0.4%) 0 1 ( 0.2%) Hepatic function abnormal 1 ( 0.4%) 0 1 ( 0.2%) Hernia 1 ( 0.4%) 0 1 ( 0.2%) Hirsutism 1 ( 0.4%) 0 1 ( 0.2%) Hyperlipidaemia 1 ( 0.4%) 4 ( 1.5%) 5 ( 1.0%) Hypothyroidism 1 ( 0.4%) 0 1 ( 0.2%) Large intestine polyp 1 ( 0.4%) 0 1 ( 0.2%) Lower limb fracture 1 ( 0.4%) 0 1 ( 0.2%) Menstrual disorder 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Neurodermatitis 1 ( 0.4%) 0 1 ( 0.2%) Neuropathy peripheral 1 ( 0.4%) 0 1 ( 0.2%) Nodule 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Noninfective gingivitis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Oesophageal carcinoma 1 ( 0.4%) 0 1 ( 0.2%) Otitis externa 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Ovarian operation 1 ( 0.4%) 0 1 ( 0.2%) Pancreatitis 1 ( 0.4%) 0 1 ( 0.2%) Pulpitis dental 1 ( 0.4%) 0 1 ( 0.2%)

PPD

89 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 5 of 15) ______ACT UC Total N=242 N=265 N=507

Refraction disorder 1 ( 0.4%) 0 1 ( 0.2%) Salivary gland neoplasm 1 ( 0.4%) 0 1 ( 0.2%) Sleep disorder 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Spinal osteoarthritis 1 ( 0.4%) 0 1 ( 0.2%) Tinea pedis 1 ( 0.4%) 0 1 ( 0.2%) Tinnitus 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Uterine cervical erosion 1 ( 0.4%) 0 1 ( 0.2%) Uterine polyp 1 ( 0.4%) 0 1 ( 0.2%) Uterine prolapse 1 ( 0.4%) 0 1 ( 0.2%) Abdominal wall cyst 0 1 ( 0.4%) 1 ( 0.2%) Acrochordon 0 1 ( 0.4%) 1 ( 0.2%) Anal fissure 0 1 ( 0.4%) 1 ( 0.2%) Anal fistula 0 1 ( 0.4%) 1 ( 0.2%) Anxiety 0 1 ( 0.4%) 1 ( 0.2%) Appendicectomy 0 4 ( 1.5%) 4 ( 0.8%) Bronchiectasis 0 1 ( 0.4%) 1 ( 0.2%) Cerumen impaction 0 1 ( 0.4%) 1 ( 0.2%) Dermatitis 0 3 ( 1.1%) 3 ( 0.6%) Diabetes mellitus 0 1 ( 0.4%) 1 ( 0.2%) Dizziness 0 1 ( 0.4%) 1 ( 0.2%)

PPD

90 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 6 of 15) ______ACT UC Total N=242 N=265 N=507

Female sterilisation 0 1 ( 0.4%) 1 ( 0.2%) Gallbladder polyp 0 1 ( 0.4%) 1 ( 0.2%) Gouty arthritis 0 1 ( 0.4%) 1 ( 0.2%) Heart disease congenital 0 1 ( 0.4%) 1 ( 0.2%) Hepatic steatosis 0 2 ( 0.8%) 2 ( 0.4%) Hepatitis B virus test positive 0 1 ( 0.4%) 1 ( 0.2%) Hepatitis C 0 1 ( 0.4%) 1 ( 0.2%) Herpes zoster 0 1 ( 0.4%) 1 ( 0.2%) Hydrosalpinx 0 1 ( 0.4%) 1 ( 0.2%) Hypersensitivity 0 1 ( 0.4%) 1 ( 0.2%) Insomnia 0 3 ( 1.1%) 3 ( 0.6%) Intestinal obstruction 0 1 ( 0.4%) 1 ( 0.2%) Intestinal polyp 0 1 ( 0.4%) 1 ( 0.2%) Ligament injury 0 1 ( 0.4%) 1 ( 0.2%) Lymphadenitis 0 1 ( 0.4%) 1 ( 0.2%) Mass excision 0 1 ( 0.4%) 1 ( 0.2%) Melanocytic naevus 0 1 ( 0.4%) 1 ( 0.2%) Nasal operation 0 1 ( 0.4%) 1 ( 0.2%) Nasal polypectomy 0 1 ( 0.4%) 1 ( 0.2%) Osteoporosis 0 1 ( 0.4%) 1 ( 0.2%)

PPD

91 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 7 of 15) ______ACT UC Total N=242 N=265 N=507

Ovarian cyst 0 1 ( 0.4%) 1 ( 0.2%) Pancreatitis acute 0 1 ( 0.4%) 1 ( 0.2%) Papilloma viral infection 0 1 ( 0.4%) 1 ( 0.2%) Peptic ulcer 0 1 ( 0.4%) 1 ( 0.2%) Percutaneous coronary intervention 0 1 ( 0.4%) 1 ( 0.2%) Pericoronitis 0 1 ( 0.4%) 1 ( 0.2%) Pneumonitis 0 1 ( 0.4%) 1 ( 0.2%) Polycystic ovaries 0 1 ( 0.4%) 1 ( 0.2%) Prepuce redundant 0 1 ( 0.4%) 1 ( 0.2%) Proteinuria 0 1 ( 0.4%) 1 ( 0.2%) Pulmonary mass 0 1 ( 0.4%) 1 ( 0.2%) Pulmonary tuberculosis 0 1 ( 0.4%) 1 ( 0.2%) Rheumatic heart disease 0 1 ( 0.4%) 1 ( 0.2%) Spinal pain 0 1 ( 0.4%) 1 ( 0.2%) Steatohepatitis 0 1 ( 0.4%) 1 ( 0.2%) Stent placement 0 1 ( 0.4%) 1 ( 0.2%) Sudden hearing loss 0 1 ( 0.4%) 1 ( 0.2%) Tenosynovitis 0 1 ( 0.4%) 1 ( 0.2%) Teratoma 0 1 ( 0.4%) 1 ( 0.2%) Thyroglossal fistula 0 1 ( 0.4%) 1 ( 0.2%)

PPD

92 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 8 of 15) ______ACT UC Total N=242 N=265 N=507

Thyroid cancer 0 2 ( 0.8%) 2 ( 0.4%) Thyroid operation 0 1 ( 0.4%) 1 ( 0.2%) Type 2 diabetes mellitus 0 3 ( 1.1%) 3 ( 0.6%) Urethritis 0 4 ( 1.5%) 4 ( 0.8%) Uterine cervix hyperplasia 0 1 ( 0.4%) 1 ( 0.2%) Vaginal haemorrhage 0 1 ( 0.4%) 1 ( 0.2%) Wolff-Parkinson-White syndrome 0 1 ( 0.4%) 1 ( 0.2%) Xerophthalmia 0 1 ( 0.4%) 1 ( 0.2%)

Current Illness 132 ( 54.5%) 189 ( 71.3%) 321 ( 63.3%) Hypertension 54 ( 22.3%) 33 ( 12.5%) 87 ( 17.2%) Rhinitis 54 ( 22.3%) 146 ( 55.1%) 200 ( 39.4%) Rhinitis allergic 12 ( 5.0%) 1 ( 0.4%) 13 ( 2.6%) Sinusitis 7 ( 2.9%) 10 ( 3.8%) 17 ( 3.4%) Coronary artery disease 6 ( 2.5%) 11 ( 4.2%) 17 ( 3.4%) Hyperlipidaemia 6 ( 2.5%) 7 ( 2.6%) 13 ( 2.6%) Gastrooesophageal reflux disease 4 ( 1.7%) 7 ( 2.6%) 11 ( 2.2%) Gout 4 ( 1.7%) 0 4 ( 0.8%) Type 2 diabetes mellitus 4 ( 1.7%) 2 ( 0.8%) 6 ( 1.2%) Benign prostatic hyperplasia 3 ( 1.2%) 3 ( 1.1%) 6 ( 1.2%) Chronic gastritis 3 ( 1.2%) 11 ( 4.2%) 14 ( 2.8%)

PPD

93 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 9 of 15) ______ACT UC Total N=242 N=265 N=507

Hepatic steatosis 3 ( 1.2%) 1 ( 0.4%) 4 ( 0.8%) Thyroid mass 3 ( 1.2%) 5 ( 1.9%) 8 ( 1.6%) Uterine leiomyoma 3 ( 1.2%) 2 ( 0.8%) 5 ( 1.0%) Arrhythmia 2 ( 0.8%) 4 ( 1.5%) 6 ( 1.2%) Arthritis 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Constipation 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Dermatitis 2 ( 0.8%) 4 ( 1.5%) 6 ( 1.2%) Diabetes mellitus 2 ( 0.8%) 7 ( 2.6%) 9 ( 1.8%) Intervertebral disc protrusion 2 ( 0.8%) 6 ( 2.3%) 8 ( 1.6%) Lacunar infarction 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Polycystic ovaries 2 ( 0.8%) 0 2 ( 0.4%) Refraction disorder 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Anal skin tags 1 ( 0.4%) 0 1 ( 0.2%) Arteriosclerosis 1 ( 0.4%) 0 1 ( 0.2%) Arteriosclerosis coronary artery 1 ( 0.4%) 0 1 ( 0.2%) Arthropathy 1 ( 0.4%) 0 1 ( 0.2%) Atrial fibrillation 1 ( 0.4%) 0 1 ( 0.2%) Autoimmune thyroiditis 1 ( 0.4%) 0 1 ( 0.2%) Bone formation increased 1 ( 0.4%) 0 1 ( 0.2%) Bronchiectasis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%)

PPD

94 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 10 of 15) ______ACT UC Total N=242 N=265 N=507

Calculus urethral 1 ( 0.4%) 0 1 ( 0.2%) Cerebral atrophy 1 ( 0.4%) 0 1 ( 0.2%) Cerebral ischaemia 1 ( 0.4%) 0 1 ( 0.2%) Cholelithiasis 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Chronic hepatitis 1 ( 0.4%) 0 1 ( 0.2%) Colitis 1 ( 0.4%) 0 1 ( 0.2%) Conjunctivitis 1 ( 0.4%) 0 1 ( 0.2%) Deep vein thrombosis 1 ( 0.4%) 0 1 ( 0.2%) Deficiency anaemia 1 ( 0.4%) 0 1 ( 0.2%) Depression 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Dermatitis allergic 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%) Dysmenorrhoea 1 ( 0.4%) 0 1 ( 0.2%) Gallbladder polyp 1 ( 0.4%) 0 1 ( 0.2%) Gastritis 1 ( 0.4%) 6 ( 2.3%) 7 ( 1.4%) Glaucoma 1 ( 0.4%) 0 1 ( 0.2%) Glomerulonephritis chronic 1 ( 0.4%) 0 1 ( 0.2%) Glucose tolerance impaired 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Haemangioma of liver 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Hashitoxicosis 1 ( 0.4%) 0 1 ( 0.2%) Headache 1 ( 0.4%) 0 1 ( 0.2%)

PPD

95 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 11 of 15) ______ACT UC Total N=242 N=265 N=507

Hepatic cyst 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%) Hepatic function abnormal 1 ( 0.4%) 0 1 ( 0.2%) Hepatitis B 1 ( 0.4%) 0 1 ( 0.2%) Hepatitis B virus test positive 1 ( 0.4%) 0 1 ( 0.2%) Hyperthyroidism 1 ( 0.4%) 0 1 ( 0.2%) Hypothyroidism 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Iron deficiency anaemia 1 ( 0.4%) 0 1 ( 0.2%) Maculopathy 1 ( 0.4%) 0 1 ( 0.2%) Menstrual disorder 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Nephrolithiasis 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Osteoarthritis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Otitis media 1 ( 0.4%) 0 1 ( 0.2%) Pain in extremity 1 ( 0.4%) 0 1 ( 0.2%) Pharyngitis 1 ( 0.4%) 8 ( 3.0%) 9 ( 1.8%) Plantar fasciitis 1 ( 0.4%) 0 1 ( 0.2%) Psoriasis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Pulmonary mass 1 ( 0.4%) 0 1 ( 0.2%) Pulmonary sarcoidosis 1 ( 0.4%) 0 1 ( 0.2%) Remnant gastritis 1 ( 0.4%) 0 1 ( 0.2%) Renal cyst 1 ( 0.4%) 0 1 ( 0.2%)

PPD

96 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 12 of 15) ______ACT UC Total N=242 N=265 N=507

Rheumatoid arthritis 1 ( 0.4%) 0 1 ( 0.2%) Sleep disorder 1 ( 0.4%) 0 1 ( 0.2%) Stent placement 1 ( 0.4%) 0 1 ( 0.2%) Supraventricular extrasystoles 1 ( 0.4%) 0 1 ( 0.2%) Tachycardia 1 ( 0.4%) 0 1 ( 0.2%) Tendonitis 1 ( 0.4%) 0 1 ( 0.2%) Uterine cervical erosion 1 ( 0.4%) 0 1 ( 0.2%) Vaginal infection 1 ( 0.4%) 0 1 ( 0.2%) Vitreous opacities 1 ( 0.4%) 0 1 ( 0.2%) Vulvovaginal injury 1 ( 0.4%) 0 1 ( 0.2%) Abdominal discomfort 0 1 ( 0.4%) 1 ( 0.2%) Atrophic vulvovaginitis 0 1 ( 0.4%) 1 ( 0.2%) Breast disorder 0 1 ( 0.4%) 1 ( 0.2%) Breast hyperplasia 0 1 ( 0.4%) 1 ( 0.2%) Cataract 0 2 ( 0.8%) 2 ( 0.4%) Cervicitis 0 1 ( 0.4%) 1 ( 0.2%) Chest X-ray abnormal 0 1 ( 0.4%) 1 ( 0.2%) Cholecystitis chronic 0 2 ( 0.8%) 2 ( 0.4%) Chronic kidney disease 0 1 ( 0.4%) 1 ( 0.2%) Chronic sinusitis 0 1 ( 0.4%) 1 ( 0.2%)

PPD

97 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 13 of 15) ______ACT UC Total N=242 N=265 N=507

Colon cancer 0 1 ( 0.4%) 1 ( 0.2%) Dermatitis atopic 0 1 ( 0.4%) 1 ( 0.2%) Diabetic retinopathy 0 1 ( 0.4%) 1 ( 0.2%) Dizziness 0 4 ( 1.5%) 4 ( 0.8%) Duodenal ulcer 0 1 ( 0.4%) 1 ( 0.2%) Dyspepsia 0 1 ( 0.4%) 1 ( 0.2%) Eczema 0 4 ( 1.5%) 4 ( 0.8%) Extrasystoles 0 1 ( 0.4%) 1 ( 0.2%) Foreign body in ear 0 1 ( 0.4%) 1 ( 0.2%) Functional gastrointestinal disorder 0 2 ( 0.8%) 2 ( 0.4%) Gastric disorder 0 1 ( 0.4%) 1 ( 0.2%) Gastritis erosive 0 2 ( 0.8%) 2 ( 0.4%) Haematochezia 0 1 ( 0.4%) 1 ( 0.2%) Haemorrhoids 0 3 ( 1.1%) 3 ( 0.6%) Helicobacter gastritis 0 1 ( 0.4%) 1 ( 0.2%) Helicobacter infection 0 3 ( 1.1%) 3 ( 0.6%) Helicobacter test positive 0 1 ( 0.4%) 1 ( 0.2%) Hiatus hernia 0 1 ( 0.4%) 1 ( 0.2%) Hyperhidrosis 0 1 ( 0.4%) 1 ( 0.2%) Hyperkeratosis 0 1 ( 0.4%) 1 ( 0.2%)

PPD

98 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 14 of 15) ______ACT UC Total N=242 N=265 N=507

Hyperprolactinaemia 0 1 ( 0.4%) 1 ( 0.2%) Hypersensitivity 0 1 ( 0.4%) 1 ( 0.2%) Hyperuricaemia 0 1 ( 0.4%) 1 ( 0.2%) Hypoxic-ischaemic encephalopathy 0 1 ( 0.4%) 1 ( 0.2%) Insomnia 0 3 ( 1.1%) 3 ( 0.6%) Laryngeal obstruction 0 1 ( 0.4%) 1 ( 0.2%) Lymphadenitis 0 1 ( 0.4%) 1 ( 0.2%) Macular oedema 0 1 ( 0.4%) 1 ( 0.2%) Musculoskeletal discomfort 0 1 ( 0.4%) 1 ( 0.2%) Myocardial ischaemia 0 1 ( 0.4%) 1 ( 0.2%) Nephroptosis 0 1 ( 0.4%) 1 ( 0.2%) Oesophagitis 0 1 ( 0.4%) 1 ( 0.2%) Osteoporosis 0 3 ( 1.1%) 3 ( 0.6%) Periodontal disease 0 1 ( 0.4%) 1 ( 0.2%) Periodontitis 0 2 ( 0.8%) 2 ( 0.4%) Pruritus 0 1 ( 0.4%) 1 ( 0.2%) Pulpitis dental 0 1 ( 0.4%) 1 ( 0.2%) Sinus tachycardia 0 1 ( 0.4%) 1 ( 0.2%) Spinal column stenosis 0 1 ( 0.4%) 1 ( 0.2%) Spinal osteoarthritis 0 4 ( 1.5%) 4 ( 0.8%)

PPD

99 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol： 201097 Date： 29SEP2019

Table 1.07 Subjects' Past and Current Illness(ITT) (Page 15 of 15) ______ACT UC Total N=242 N=265 N=507

Tonsillar hypertrophy 0 1 ( 0.4%) 1 ( 0.2%) Urticaria 0 3 ( 1.1%) 3 ( 0.6%) Uterine cervix hyperplasia 0 1 ( 0.4%) 1 ( 0.2%) Vertigo 0 2 ( 0.8%) 2 ( 0.4%) Xerophthalmia 0 1 ( 0.4%) 1 ( 0.2%)

PPD

100 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 21NOV2019

Table1.08 Subjects' Asthma Medication (ITT) (Page 1 of 2)

ACT UC Total N=242 N=265 N=507 At least one asthma medication at visit 0 242 (100.0%) 264 ( 99.6%) 506 ( 99.8%) FLUTICASONE PROPIONATE;SALMETEROL XINAFOATE 154 ( 63.6%) 82 ( 30.9%) 236 ( 46.5%) BUDESONIDE;FORMOTEROL FUMARATE 88 ( 36.4%) 169 ( 63.8%) 257 ( 50.7%) SALBUTAMOL 50 ( 20.7%) 58 ( 21.9%) 108 ( 21.3%) MONTELUKAST SODIUM 41 ( 16.9%) 64 ( 24.2%) 105 ( 20.7%) SALBUTAMOL SULFATE 22 ( 9.1%) 35 ( 13.2%) 57 ( 11.2%) MONTELUKAST 17 ( 7.0%) 5 ( 1.9%) 22 ( 4.3%) BUDESONIDE 7 ( 2.9%) 8 ( 3.0%) 15 ( 3.0%) UNSPECIFIED HERBAL AND TRADITIONAL MEDICINE 6 ( 2.5%) 0 6 ( 1.2%) AMINOPHYLLINE;CHLORPHENAMINE MALEATE;METHOXYPHENAMINE 5 ( 2.1%) 12 ( 4.5%) 17 ( 3.4%) HYDROCHLORIDE;NOSCAPINE BUDESONIDE;FORMOTEROL 4 ( 1.7%) 1 ( 0.4%) 5 ( 1.0%) TIOTROPIUM BROMIDE 4 ( 1.7%) 2 ( 0.8%) 6 ( 1.2%) THEOPHYLLINE 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) BECLOMETASONE DIPROPIONATE 1 ( 0.4%) 0 1 ( 0.2%) FLUTICASONE;SALMETEROL 1 ( 0.4%) 0 1 ( 0.2%) FORMOTEROL FUMARATE 1 ( 0.4%) 0 1 ( 0.2%) LORATADINE 1 ( 0.4%) 0 1 ( 0.2%) METHOXYPHENAMINE HYDROCHLORIDE 1 ( 0.4%) 6 ( 2.3%) 7 ( 1.4%) PROCATEROL HYDROCHLORIDE 1 ( 0.4%) 0 1 ( 0.2%) AMBROXOL 0 1 ( 0.4%) 1 ( 0.2%)

NOTE：FLUTICASONE PROPIONATE; SALMETEROL XINAFOATE and FLUTICASONE; SALMETEROL are duplicate reports for the same drug. BUDESONIDE; FORMOTEROL FUMARATE and BUDESONIDE; FORMOTEROL are duplicate reports for the same drug.

PPD

101 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 21NOV2019

Table1.08 Subjects' Asthma Medication (ITT) (Page 2 of 2)

ACT UC Total N=242 N=265 N=507 BECLOMETASONE DIPROPIONATE;FORMOTEROL FUMARATE 0 2 ( 0.8%) 2 ( 0.4%) CETIRIZINE 0 2 ( 0.8%) 2 ( 0.4%) CINEOLE;DIPENTEN;PINENE 0 1 ( 0.4%) 1 ( 0.2%) FLUTICASONE PROPIONATE 0 7 ( 2.6%) 7 ( 1.4%) FORMOTEROL 0 1 ( 0.4%) 1 ( 0.2%)

At least one asthma medication during the study period 238 ( 98.3%) 265 (100.0%) 503 ( 99.2%) FLUTICASONE PROPIONATE;SALMETEROL XINAFOATE 132 ( 54.5%) 75 ( 28.3%) 207 ( 40.8%) BUDESONIDE;FORMOTEROL FUMARATE 72 ( 29.8%) 150 ( 56.6%) 222 ( 43.8%) SALBUTAMOL 18 ( 7.4%) 10 ( 3.8%) 28 ( 5.5%) SALBUTAMOL SULFATE 9 ( 3.7%) 10 ( 3.8%) 19 ( 3.7%) BUDESONIDE 5 ( 2.1%) 1 ( 0.4%) 6 ( 1.2%) FORMOTEROL FUMARATE 1 ( 0.4%) 0 1 ( 0.2%) MONTELUKAST SODIUM 1 ( 0.4%) 8 ( 3.0%) 9 ( 1.8%) AMINOPHYLLINE;CHLORPHENAMINE MALEATE;METHOXYPHENAMINE 0 2 ( 0.8%) 2 ( 0.4%) HYDROCHLORIDE;NOSCAPINE BECLOMETASONE DIPROPIONATE;FORMOTEROL FUMARATE 0 1 ( 0.4%) 1 ( 0.2%) FLUTICASONE PROPIONATE 0 8 ( 3.0%) 8 ( 1.6%) FORMOTEROL 0 1 ( 0.4%) 1 ( 0.2%) LORATADINE 0 1 ( 0.4%) 1 ( 0.2%) METHYLPREDNISOLONE 0 1 ( 0.4%) 1 ( 0.2%) THEOPHYLLINE 0 1 ( 0.4%) 1 ( 0.2%)

NOTE：FLUTICASONE PROPIONATE; SALMETEROL XINAFOATE and FLUTICASONE; SALMETEROL are duplicate reports for the same drug. BUDESONIDE; FORMOTEROL FUMARATE and BUDESONIDE; FORMOTEROL are duplicate reports for the same drug.

PPD

102 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 1 of 36)

ACT UC Total N=242 N=265 N=507 At least one Concurrent medication 131 ( 54.1%) 167 ( 63.0%) 298 ( 58.8%) UNSPECIFIED HERBAL AND TRADITIONAL MEDICINE 86 ( 35.5%) 114 ( 43.0%) 200 ( 39.4%) CINEOLE;DIPENTEN;PINENE 16 ( 6.6%) 17 ( 6.4%) 33 ( 6.5%) CEFACLOR 12 ( 5.0%) 10 ( 3.8%) 22 ( 4.3%) AMINOPHYLLINE;CHLORPHENAMINE MALEATE;METHOXYPHENAMINE 9 ( 3.7%) 18 ( 6.8%) 27 ( 5.3%) HYDROCHLORIDE;NOSCAPINE LEVOFLOXACIN 9 ( 3.7%) 18 ( 6.8%) 27 ( 5.3%) MONTELUKAST SODIUM 9 ( 3.7%) 16 ( 6.0%) 25 ( 4.9%) ACETYLSALICYLIC ACID 8 ( 3.3%) 8 ( 3.0%) 16 ( 3.2%) CEFIXIME 8 ( 3.3%) 14 ( 5.3%) 22 ( 4.3%) DOXOFYLLINE 8 ( 3.3%) 0 8 ( 1.6%) TRIAMCINOLONE ACETONIDE 8 ( 3.3%) 3 ( 1.1%) 11 ( 2.2%) CETIRIZINE 7 ( 2.9%) 7 ( 2.6%) 14 ( 2.8%)

PPD

103 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 2 of 36)

ACT UC Total N=242 N=265 N=507 DEXAMETHASONE 7 ( 2.9%) 9 ( 3.4%) 16 ( 3.2%) DEXTROMETHORPHAN HYDROBROMIDE;PARACETAMOL;PSEUDOEPHEDRINE 7 ( 2.9%) 9 ( 3.4%) 16 ( 3.2%) HYDROCHLORIDE AMLODIPINE BESILATE 6 ( 2.5%) 4 ( 1.5%) 10 ( 2.0%) AZITHROMYCIN 6 ( 2.5%) 17 ( 6.4%) 23 ( 4.5%) BUDESONIDE 6 ( 2.5%) 21 ( 7.9%) 27 ( 5.3%) CANDESARTAN 6 ( 2.5%) 1 ( 0.4%) 7 ( 1.4%) EBASTINE 6 ( 2.5%) 4 ( 1.5%) 10 ( 2.0%) RABEPRAZOLE 6 ( 2.5%) 9 ( 3.4%) 15 ( 3.0%) ATORVASTATIN 5 ( 2.1%) 3 ( 1.1%) 8 ( 1.6%) BROMHEXINE 5 ( 2.1%) 1 ( 0.4%) 6 ( 1.2%) VALSARTAN 5 ( 2.1%) 8 ( 3.0%) 13 ( 2.6%) ACHYRANTHES BIDENTATA ROOT;CARTHAMUS TINCTORIUS FLOWER;CITRUS 4 ( 1.7%) 0 4 ( 0.8%) RETICULATA FRUIT PEEL;COPTIS CHINENSIS RHIZOME;GLYCYRRHIZA URALENSIS ROOT;PINELLIA TERNATA RHIZOME;PORIA COCOS; PSEUDOSTELLARI

PPD

104 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 3 of 36)

ACT UC Total N=242 N=265 N=507 AMINOPHENAZONE;CAFFEINE;CHLORPHENAMINE MALEATE;PARACETAMOL 4 ( 1.7%) 3 ( 1.1%) 7 ( 1.4%) BUPLEURUM CHINENSE, ROOT;FALLOPIA JAPONICA ROOT;FORSYTHIA 4 ( 1.7%) 4 ( 1.5%) 8 ( 1.6%) SUSPENSA FRUIT;GLYCYRRHIZA URALENSIS ROOT;ISATIS INDIGOTICA ROOT; PATRINIA SCABIOSIFOLIA;PHRAGMITES COMMUNIS RHIZOME;VERBENA OFFIC CEFDINIR 4 ( 1.7%) 4 ( 1.5%) 8 ( 1.6%) CEFOTIAM 4 ( 1.7%) 0 4 ( 0.8%) CLOSTRIDIUM BUTYRICUM 4 ( 1.7%) 1 ( 0.4%) 5 ( 1.0%) FELODIPINE 4 ( 1.7%) 2 ( 0.8%) 6 ( 1.2%) FLUTICASONE PROPIONATE 4 ( 1.7%) 3 ( 1.1%) 7 ( 1.4%) FUDOSTEINE 4 ( 1.7%) 5 ( 1.9%) 9 ( 1.8%) GLIMEPIRIDE 4 ( 1.7%) 3 ( 1.1%) 7 ( 1.4%) GLYCYRRHIZIC ACID 4 ( 1.7%) 2 ( 0.8%) 6 ( 1.2%) LEVAMLODIPINE MALEATE 4 ( 1.7%) 1 ( 0.4%) 5 ( 1.0%) MECOBALAMIN 4 ( 1.7%) 4 ( 1.5%) 8 ( 1.6%)

PPD

105 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 4 of 36)

ACT UC Total N=242 N=265 N=507 MONTELUKAST 4 ( 1.7%) 0 4 ( 0.8%) ROSUVASTATIN CALCIUM 4 ( 1.7%) 3 ( 1.1%) 7 ( 1.4%) TELMISARTAN 4 ( 1.7%) 3 ( 1.1%) 7 ( 1.4%) AMANTADINE HYDROCHLORIDE;CAFFEINE;CHLORPHENAMINE MALEATE;COW 3 ( 1.2%) 1 ( 0.4%) 4 ( 0.8%) BEZOAR;PARACETAMOL AMLODIPINE 3 ( 1.2%) 3 ( 1.1%) 6 ( 1.2%) AZELASTINE HYDROCHLORIDE 3 ( 1.2%) 2 ( 0.8%) 5 ( 1.0%) CHLORPHENAMINE MALEATE;DEXTROMETHORPHAN HYDROBROMIDE; 3 ( 1.2%) 16 ( 6.0%) 19 ( 3.7%) PARACETAMOL;PSEUDOEPHEDRINE HYDROCHLORIDE CHLORPHENAMINE;DEXTROMETHORPHAN;PARACETAMOL;PSEUDOEPHEDRINE 3 ( 1.2%) 2 ( 0.8%) 5 ( 1.0%) CLARITHROMYCIN 3 ( 1.2%) 6 ( 2.3%) 9 ( 1.8%) COLCHICINE 3 ( 1.2%) 0 3 ( 0.6%) CYNANCHUM STAUNTONII ROOT WITH RHIZOME;ERIOBOTRYA JAPONICA LEAF; 3 ( 1.2%) 6 ( 2.3%) 9 ( 1.8%) MENTHOL;MORUS ALBA ROOT BARK;PAPAVER SOMNIFERUM PEEL;PLATYCODON GRANDIFLORUS ROOT;STEMONA SESSILIFOLIA ROOT TUBER HYDROCHLOROTHIAZIDE;IRBESARTAN 3 ( 1.2%) 3 ( 1.1%) 6 ( 1.2%)

PPD

106 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 5 of 36)

ACT UC Total N=242 N=265 N=507 LEVOTHYROXINE SODIUM 3 ( 1.2%) 3 ( 1.1%) 6 ( 1.2%) METFORMIN 3 ( 1.2%) 7 ( 2.6%) 10 ( 2.0%) METRONIDAZOLE 3 ( 1.2%) 2 ( 0.8%) 5 ( 1.0%) MEZLOCILLIN 3 ( 1.2%) 0 3 ( 0.6%) MOMETASONE FUROATE 3 ( 1.2%) 5 ( 1.9%) 8 ( 1.6%) NIFEDIPINE 3 ( 1.2%) 2 ( 0.8%) 5 ( 1.0%) OLMESARTAN 3 ( 1.2%) 1 ( 0.4%) 4 ( 0.8%) OMEPRAZOLE 3 ( 1.2%) 9 ( 3.4%) 12 ( 2.4%) PHLOROGLUCINOL 3 ( 1.2%) 0 3 ( 0.6%) PRANOPROFEN 3 ( 1.2%) 1 ( 0.4%) 4 ( 0.8%) SODIUM BICARBONATE 3 ( 1.2%) 0 3 ( 0.6%) AESCULUS HIPPOCASTANUM EXTRACT 2 ( 0.8%) 0 2 ( 0.4%)

PPD

107 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 6 of 36)

ACT UC Total N=242 N=265 N=507 ALPHA-AMYLASE SWINE PANCREAS;CELLULASE;LIPASE;PROTEASE 2 ( 0.8%) 0 2 ( 0.4%) AMBROXOL 2 ( 0.8%) 5 ( 1.9%) 7 ( 1.4%) AMLODIPINE L-ASPARTATE 2 ( 0.8%) 0 2 ( 0.4%) ASCORBIC ACID 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) ASTER TATARICUS;CITRUS AURANTIUM;EPHEDRA SINICA;FAGOPYRUM 2 ( 0.8%) 0 2 ( 0.4%) DIBOTRYS;GLYCYRRHIZA URALENSIS;HOUTTUYNIA CORDATA;ILEX CHINENSIS; PEUCEDANUM PRAERUPTORUM BERBERINE HYDROCHLORIDE 2 ( 0.8%) 0 2 ( 0.4%) BISOPROLOL FUMARATE 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) BORNEOL;COW BEZOAR;GLYCYRRHIZA SPP.;INDIGO;MENTHOL;TERMINALIA 2 ( 0.8%) 6 ( 2.3%) 8 ( 1.6%) CHEBULA CARBOCISTEINE 2 ( 0.8%) 0 2 ( 0.4%) CEFPROZIL 2 ( 0.8%) 13 ( 4.9%) 15 ( 3.0%) CEFTAZIDIME PENTAHYDRATE 2 ( 0.8%) 0 2 ( 0.4%) CEFUROXIME 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%)

PPD

108 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 7 of 36)

ACT UC Total N=242 N=265 N=507 CETIRIZINE HYDROCHLORIDE 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) CLOBETASOL PROPIONATE;MICONAZOLE NITRATE 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) CLOPIDOGREL 2 ( 0.8%) 0 2 ( 0.4%) CYPROTERONE ACETATE;ETHINYLESTRADIOL 2 ( 0.8%) 0 2 ( 0.4%) ENALAPRIL 2 ( 0.8%) 0 2 ( 0.4%) ERDOSTEINE 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) IBUPROFEN 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) IBUPROFEN ARGININE 2 ( 0.8%) 0 2 ( 0.4%) IPRATROPIUM BROMIDE;SALBUTAMOL SULFATE 2 ( 0.8%) 0 2 ( 0.4%) IRBESARTAN 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) ISOSORBIDE MONONITRATE 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) LEVAMLODIPINE 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%)

PPD

109 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 8 of 36)

ACT UC Total N=242 N=265 N=507 LIDOCAINE 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) LOSARTAN 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) METHYLPREDNISOLONE 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) METHYLPREDNISOLONE SODIUM SUCCINATE 2 ( 0.8%) 7 ( 2.6%) 9 ( 1.8%) MOSAPRIDE 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) OFLOXACIN 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) OTHER NASAL PREPARATIONS 2 ( 0.8%) 0 2 ( 0.4%) PANTOPRAZOLE 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) SALBUTAMOL SULFATE 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) SULBENICILLIN 2 ( 0.8%) 0 2 ( 0.4%) TAMSULOSIN 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) TRIPOTASSIUM DICITRATOBISMUTHATE 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%)

PPD

110 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 9 of 36)

ACT UC Total N=242 N=265 N=507 WARFARIN 2 ( 0.8%) 0 2 ( 0.4%) ACARBOSE 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) ACECLOFENAC 1 ( 0.4%) 0 1 ( 0.2%) ACHYRANTHES BIDENTATA ROOT;ANGELICA SINENSIS ROOT;ASTRAGALUS 1 ( 0.4%) 0 1 ( 0.2%) SPP. ROOT;BOSWELLIA SACRA;CARTHAMUS TINCTORIUS FLOWER;CINNAMOMUM CASSIA TWIG;COMMIPHORA MYRRHA RESIN;LEECH EXTRACT;LIGUSTICUM C ACICLOVIR 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) ACORUS CALAMUS VAR. ANGUSTATUS RHIZOME;ANEMARRHENA 1 ( 0.4%) 0 1 ( 0.2%) ASPHODELOIDES RHIZOME;CALCIUM SULFATE DIHYDRATE;CURCUMA SPP. ROOT TUBER;FORSYTHIA SUSPENSA FRUIT;ISATIS TINCTORIA ROOT; PHRAGMITES COMMUN ACORUS GRAMINEUS RHIZOME;ALPINIA OXYPHYLLA FRUIT;GLYCYRRHIZA 1 ( 0.4%) 0 1 ( 0.2%) GLABRA ROOT WITH RHIZOME;LINDERA AGGREGATA ROOT;PORIA COCOS SCLEROTIUM ALANYL GLUTAMINE 1 ( 0.4%) 0 1 ( 0.2%) ALISMA PLANTAGO-AQUATICA VAR. ORIENTALE RHIZOME;ANGELICA 1 ( 0.4%) 0 1 ( 0.2%) SINENSIS ROOT;ATRACTYLODES LANCEA RHIZOME;COIX LACRYMA-JOBI SUBSP. MA-YUEN KERNEL;CORYDALIS YANHUSUO TUBE;CYPERUS ROTUNDUS RHIZOME; AMBROXOL HYDROCHLORIDE 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) AMINOMETHYLBENZOIC ACID 1 ( 0.4%) 0 1 ( 0.2%) AMINOPHENAZONE;CAFFEINE;PHENACETIN;PHENOBARBITAL 1 ( 0.4%) 0 1 ( 0.2%)

PPD

111 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 10 of 36)

ACT UC Total N=242 N=265 N=507 AMLODIPINE BESILATE;BENAZEPRIL HYDROCHLORIDE 1 ( 0.4%) 0 1 ( 0.2%) AMMONIA;CAMPHOR;GLYCEROL;GLYCYRRHIZA GLABRA;GUAIFENESIN 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) AMOXICILLIN 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) ANDROGRAPHIS PANICULATA HERB;ANGELICA SINENSIS ROOT;CODONOPSIS 1 ( 0.4%) 0 1 ( 0.2%) PILOSULA ROOT;FLEMINGIA PROSTRATA ROOT;MAHONIA BEALEI STEM;ROSA LAEVIGATA ROOT;SPATHOLOBUS SUBERECTUS STEM;ZANTHOXYLUM DISSITUM

ANGELICA DAHURICA VAR. FORMOSANA ROOT;ASARUM HETEROTROPOIDES 1 ( 0.4%) 4 ( 1.5%) 5 ( 1.0%) VAR. MANDSHURICUM HERB;ASTRAGALUS MONGHOLICUS ROOT;BUPLEURUM CHINENSE, ROOT;CLEMATIS ARMANDII STEM;GARDENIA JASMINOIDES FRUIT;

ANGELICA SINENSIS ROOT;CENTIPEDA MINIMA HERB;CHRYSANTHEMUM 1 ( 0.4%) 0 1 ( 0.2%) INDICUM INFLORESCENCE;EPHEDRA SINICA HERB;MENTHA CANADENSIS OIL; POGOSTEMON CABLIN HERB;SCUTELLARIA BAICALENSIS ROOT;XANTHIUM SIB ARCTIUM LAPPA FRUIT;FORSYTHIA SUSPENSA FRUIT;GLYCINE MAX 1 ( 0.4%) 0 1 ( 0.2%) FERMENTED SEED;GLYCYRRHIZA URALENSIS ROOT WITH RHIZOME;LONICERA JAPONICA FLOWER BUD;LOPHATHERUM GRACILE HERB;MENTHA CANADENSIS HERB ASTRAGALUS MONGHOLICUS ROOT;CAESALPINIA SAPPAN WOOD;CARTHAMUS 1 ( 0.4%) 0 1 ( 0.2%) TINCTORIUS FLOWER;CORYDALIS YANHUSUO TUBE;LIGUSTICUM CHUANXIONG RHIZOME;PAEONIA LACTIFLORA SUN DRIED ROOT;PANAX NOTOGINSENG ROOT;@ ASTRAGALUS MONGHOLICUS ROOT;LESPEDEZA BICOLOR ROOT;PYRROSIA 1 ( 0.4%) 0 1 ( 0.2%) LINGUA LEAF;SCUTELLARIA BAICALENSIS ROOT ATENOLOL;NITRENDIPINE 1 ( 0.4%) 0 1 ( 0.2%) ATORVASTATIN CALCIUM 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) AZINTAMIDE;CELLULASE;PANCREATIN;SIMETICONE 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%)

PPD

112 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 11 of 36)

ACT UC Total N=242 N=265 N=507 BAMBUSA SPP. 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) BENZALKONIUM BROMIDE 1 ( 0.4%) 0 1 ( 0.2%) BENZBROMARONE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) BENZYLPENICILLIN SODIUM 1 ( 0.4%) 0 1 ( 0.2%) BERAPROST SODIUM 1 ( 0.4%) 0 1 ( 0.2%) BISACODYL 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) BISOPROLOL 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) BORNEOL;GINKGO BILOBA LEAF;GYNOSTEMMA PENTAPHYLLUM WHOLE PLANT; 1 ( 0.4%) 0 1 ( 0.2%) SALVIA YUNNANENSIS ROOT BROMPHENIRAMINE MALEATE;CODEINE PHOSPHATE;EPHEDRINE 1 ( 0.4%) 0 1 ( 0.2%) HYDROCHLORIDE;GUAIFENESIN BUPLEURUM CHINENSE, ROOT;CITRUS RETICULATA PEEL;GLYCYRRHIZA 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) URALENSIS ROOT WITH RHIZOME;PAEONIA LACTIFLORA SUN DRIED ROOT; SAPOSHNIKOVIA DIVARICATA ROOT;ZINGIBER OFFICINALE RHIZOME CAFFEINE;CHLORPHENAMINE MALEATE;PARACETAMOL 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) CALCIPOTRIOL 1 ( 0.4%) 0 1 ( 0.2%)

PPD

113 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 12 of 36)

ACT UC Total N=242 N=265 N=507 CALCIUM DOBESILATE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) CALCIUM GLUCONATE 1 ( 0.4%) 0 1 ( 0.2%) CARBAZOCHROME SODIUM SULFONATE;SODIUM CHLORIDE 1 ( 0.4%) 0 1 ( 0.2%) CEFADROXIL 1 ( 0.4%) 0 1 ( 0.2%) CEFALEXIN 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) CEFALOTIN 1 ( 0.4%) 0 1 ( 0.2%) CEFOXITIN SODIUM 1 ( 0.4%) 0 1 ( 0.2%) CEFRADINE 1 ( 0.4%) 5 ( 1.9%) 6 ( 1.2%) CEFTIZOXIME 1 ( 0.4%) 0 1 ( 0.2%) CEFUROXIME AXETIL 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) CEFUROXIME SODIUM 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%) CEPHALOSPORIN NOS 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%)

PPD

114 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 13 of 36)

ACT UC Total N=242 N=265 N=507 CHLORPHENAMINE MALEATE;PARACETAMOL 1 ( 0.4%) 0 1 ( 0.2%) CHLORPHENAMINE;PSEUDOEPHEDRINE 1 ( 0.4%) 0 1 ( 0.2%) CINEPAZIDE MALEATE 1 ( 0.4%) 0 1 ( 0.2%) CLINDAMYCIN PALMITATE HYDROCHLORIDE 1 ( 0.4%) 0 1 ( 0.2%) COBAMAMIDE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) CODEINE PHOSPHATE;IBUPROFEN 1 ( 0.4%) 0 1 ( 0.2%) COMBINATIONS OF VITAMINS 1 ( 0.4%) 0 1 ( 0.2%) COPTIS CHINENSIS RHIZOME;SAUSSUREA COSTUS ROOT 1 ( 0.4%) 0 1 ( 0.2%) CORYDALIS BUNGEANA HERB;ISATIS INDIGOTICA ROOT;SCUTELLARIA 1 ( 0.4%) 0 1 ( 0.2%) BAICALENSIS ROOT;TARAXACUM MONGOLICUM HERB CYANOCOBALAMIN 1 ( 0.4%) 0 1 ( 0.2%) DECONGESTANTS AND OTHER NASAL PREPARATIONS FOR TOPICAL USE 1 ( 0.4%) 0 1 ( 0.2%) DEXTRAN 40 1 ( 0.4%) 0 1 ( 0.2%)

PPD

115 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 14 of 36)

ACT UC Total N=242 N=265 N=507 DICLOFENAC DIETHYLAMINE 1 ( 0.4%) 0 1 ( 0.2%) DILTIAZEM 1 ( 0.4%) 0 1 ( 0.2%) DIOSMIN 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) DOXAZOSIN 1 ( 0.4%) 0 1 ( 0.2%) DOXYCYCLINE 1 ( 0.4%) 0 1 ( 0.2%) ENALAPRIL MALEATE 1 ( 0.4%) 0 1 ( 0.2%) ENTECAVIR 1 ( 0.4%) 0 1 ( 0.2%) EPERISONE 1 ( 0.4%) 0 1 ( 0.2%) EPHEDRINE HYDROCHLORIDE;SULFADIAZINE 1 ( 0.4%) 0 1 ( 0.2%) EPINEPHRINE HYDROCHLORIDE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) ERIGERON BREVISCAPUS HERB;OPHIOPOGON JAPONICUS;PANAX GINSENG 1 ( 0.4%) 0 1 ( 0.2%) ROOT;SCHISANDRA CHINENSIS FRUIT FAGOPYRUM DIBOTRYS RHIZOME 1 ( 0.4%) 0 1 ( 0.2%)

PPD

116 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 15 of 36)

ACT UC Total N=242 N=265 N=507 FEBUXOSTAT 1 ( 0.4%) 0 1 ( 0.2%) FERROUS SUCCINATE 1 ( 0.4%) 0 1 ( 0.2%) FINASTERIDE 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) FLURBIPROFEN AXETIL 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) FLUTICASONE PROPIONATE;SALMETEROL XINAFOATE 1 ( 0.4%) 0 1 ( 0.2%) FLUVASTATIN 1 ( 0.4%) 0 1 ( 0.2%) FOLIC ACID 1 ( 0.4%) 0 1 ( 0.2%) FORSYTHIA SUSPENSA FRUIT;LONICERA JAPONICA FLOWER BUD; 1 ( 0.4%) 0 1 ( 0.2%) SCUTELLARIA BAICALENSIS ROOT FOSFOMYCIN TROMETAMOL 1 ( 0.4%) 0 1 ( 0.2%) FOSINOPRIL 1 ( 0.4%) 0 1 ( 0.2%) GABAPENTIN 1 ( 0.4%) 0 1 ( 0.2%) GINKGO BILOBA EXTRACT 1 ( 0.4%) 4 ( 1.5%) 5 ( 1.0%)

PPD

117 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 16 of 36)

ACT UC Total N=242 N=265 N=507 GLICLAZIDE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) GLYCEROL 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) HALOMETASONE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) HYALURONATE SODIUM 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) HYDROCHLOROTHIAZIDE;LOSARTAN POTASSIUM 1 ( 0.4%) 0 1 ( 0.2%) HYDROTALCITE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) HYDROXYCHLOROQUINE SULFATE 1 ( 0.4%) 0 1 ( 0.2%) IMIDAPRIL 1 ( 0.4%) 0 1 ( 0.2%) ISEPAMICIN SULFATE 1 ( 0.4%) 0 1 ( 0.2%) KETOTIFEN FUMARATE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) LACTOBACILLUS ACIDOPHILUS 1 ( 0.4%) 0 1 ( 0.2%) LAFUTIDINE 1 ( 0.4%) 0 1 ( 0.2%)

PPD

118 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 17 of 36)

ACT UC Total N=242 N=265 N=507 LERCANIDIPINE 1 ( 0.4%) 0 1 ( 0.2%) LEVOFLOXACIN LACTATE 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) LEVOTHYROXINE 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) LONICERA JAPONICA FLOWER BUD;RHEUM PALMATUM ROOT WITH RHIZOME; 1 ( 0.4%) 0 1 ( 0.2%) SANGUISORBA OFFICINALIS ROOT;SCUTELLARIA BAICALENSIS ROOT; SIEGESBECKIA ORIENTALIS HERB;STYPHNOLOBIUM JAPONICUM FLOWER LORATADINE 1 ( 0.4%) 7 ( 2.6%) 8 ( 1.6%) METFORMIN HYDROCHLORIDE 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) METOCLOPRAMIDE DIHYDROCHLORIDE 1 ( 0.4%) 0 1 ( 0.2%) METOPROLOL 1 ( 0.4%) 0 1 ( 0.2%) MEXILETINE 1 ( 0.4%) 0 1 ( 0.2%) MEZLOCILLIN SODIUM;SULBACTAM SODIUM 1 ( 0.4%) 0 1 ( 0.2%) MIFEPRISTONE 1 ( 0.4%) 0 1 ( 0.2%) MOMETASONE 1 ( 0.4%) 4 ( 1.5%) 5 ( 1.0%)

PPD

119 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 18 of 36)

ACT UC Total N=242 N=265 N=507 MONTMORILLONITE 1 ( 0.4%) 0 1 ( 0.2%) MOXIFLOXACIN HYDROCHLORIDE 1 ( 0.4%) 6 ( 2.3%) 7 ( 1.4%) MUPIROCIN 1 ( 0.4%) 0 1 ( 0.2%) MYRTOL 1 ( 0.4%) 0 1 ( 0.2%) NORFLOXACIN 1 ( 0.4%) 0 1 ( 0.2%) ORNIDAZOLE;SODIUM CHLORIDE 1 ( 0.4%) 0 1 ( 0.2%) OSSOTIDE 1 ( 0.4%) 0 1 ( 0.2%) OTHER DERMATOLOGICALS 1 ( 0.4%) 0 1 ( 0.2%) OXYMETAZOLINE HYDROCHLORIDE 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) OXYTOCIN 1 ( 0.4%) 0 1 ( 0.2%) PANTOPRAZOLE SODIUM SESQUIHYDRATE 1 ( 0.4%) 0 1 ( 0.2%) PERINDOPRIL ERBUMINE 1 ( 0.4%) 0 1 ( 0.2%)

PPD

120 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 19 of 36)

ACT UC Total N=242 N=265 N=507 PERIPLANETA AMERICANA 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) PERSICARIA CAPITATA 1 ( 0.4%) 0 1 ( 0.2%) PIOGLITAZONE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) POLYMYXIN 1 ( 0.4%) 0 1 ( 0.2%) POTASSIUM CHLORIDE 1 ( 0.4%) 0 1 ( 0.2%) PRAVASTATIN SODIUM 1 ( 0.4%) 0 1 ( 0.2%) PROGESTERONE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) PYRIDOXINE HYDROCHLORIDE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) RABEPRAZOLE SODIUM 1 ( 0.4%) 5 ( 1.9%) 6 ( 1.2%) REBAMIPIDE 1 ( 0.4%) 0 1 ( 0.2%) ROSUVASTATIN 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) SACCHAROMYCES CEREVISIAE 1 ( 0.4%) 0 1 ( 0.2%)

PPD

121 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 20 of 36)

ACT UC Total N=242 N=265 N=507 SALVIA MILTIORRHIZA ROOT 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%) SERRAPEPTASE 1 ( 0.4%) 0 1 ( 0.2%) SERTRALINE 1 ( 0.4%) 0 1 ( 0.2%) SIMVASTATIN 1 ( 0.4%) 0 1 ( 0.2%) SOTALOL HYDROCHLORIDE 1 ( 0.4%) 0 1 ( 0.2%) TANSHINONE IIA SODIUM SULFONATE 1 ( 0.4%) 0 1 ( 0.2%) TENOFOVIR DISOPROXIL FUMARATE 1 ( 0.4%) 0 1 ( 0.2%) TERBINAFINE 1 ( 0.4%) 0 1 ( 0.2%) TERBUTALINE 1 ( 0.4%) 0 1 ( 0.2%) THYMALFASIN 1 ( 0.4%) 0 1 ( 0.2%) TRIAMCINOLONE ACETONIDE ACETATE 1 ( 0.4%) 0 1 ( 0.2%) TRIMETAZIDINE 1 ( 0.4%) 0 1 ( 0.2%)

PPD

122 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 21 of 36)

ACT UC Total N=242 N=265 N=507 VALPROATE SODIUM 1 ( 0.4%) 0 1 ( 0.2%) VITAMIN B1 NOS 1 ( 0.4%) 0 1 ( 0.2%) VITAMINS NOS 1 ( 0.4%) 0 1 ( 0.2%) VOGLIBOSE 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) ACETYLCYSTEINE 0 1 ( 0.4%) 1 ( 0.2%) ACHYRANTHES BIDENTATA ROOT;ACONITUM CARMICHAELII ROOT;ALISMA 0 1 ( 0.4%) 1 ( 0.2%) PLANTAGO-AQUATICA VAR. ORIENTALE RHIZOME;CINNAMOMUM CASSIA TINCTURE;CORNUS OFFICINALIS FRUIT;DIOSCOREA POLYSTACHYA RHIZOME; PAE ACHYRANTHES BIDENTATA ROOT;ATRACTYLODES LANCEA RHIZOME;COIX 0 1 ( 0.4%) 1 ( 0.2%) LACRYMA-JOBI SEED;PHELLODENDRON AMURENSE BARK ACITRETIN 0 1 ( 0.4%) 1 ( 0.2%) ACONITUM CARMICHAELII ROOT;ACONITUM COREANUM TUBE;ASARUM 0 2 ( 0.8%) 2 ( 0.4%) SIEBOLDII ROOT;BENZYL ALCOHOL;BORNEOL;BOSWELLIA SACRA RESIN; CAMPHOR;CARTHAMUS TINCTORIUS FLOWER;CINNAMOMUM CASSIA;COMMIPHORA MYRRH ACONITUM CARMICHAELII ROOT;ACONITUM KUSNEZOFFII ROOT TUBER; 0 1 ( 0.4%) 1 ( 0.2%) CARTHAMUS TINCTORIUS FLOWER;CHAENOMELES SPECIOSA FRUIT;EPHEDRA SINICA HERB;GLYCYRRHIZA URALENSIS ROOT WITH RHIZOME;PRUNUS MUME S ACRIVASTINE 0 1 ( 0.4%) 1 ( 0.2%) ALENDRONATE SODIUM 0 1 ( 0.4%) 1 ( 0.2%)

PPD

123 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 22 of 36)

ACT UC Total N=242 N=265 N=507 ALENDRONATE SODIUM;COLECALCIFEROL 0 1 ( 0.4%) 1 ( 0.2%) ALPINIA OFFICINARUM RHIZOME;AMOMUM VILLOSUM FRUIT;CINNAMOMUM 0 1 ( 0.4%) 1 ( 0.2%) CASSIA TWIG;CORYDALIS YANHUSUO TUBER;FOENICULUM VULGARE FRUIT; GLYCYRRHIZA SPP. ROOT;OYSTER SHELL ALPRAZOLAM 0 1 ( 0.4%) 1 ( 0.2%) AMANTADINE HYDROCHLORIDE;CHLORPHENAMINE MALEATE;PARACETAMOL 0 1 ( 0.4%) 1 ( 0.2%) AMANTADINE HYDROCHLORIDE;PARACETAMOL 0 1 ( 0.4%) 1 ( 0.2%) AMIDES 0 1 ( 0.4%) 1 ( 0.2%) AMIODARONE HYDROCHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) AMLODIPINE;VALSARTAN 0 1 ( 0.4%) 1 ( 0.2%) AMOXICILLIN SODIUM;CLAVULANATE POTASSIUM 0 1 ( 0.4%) 1 ( 0.2%) AMOXICILLIN TRIHYDRATE;CLAVULANATE POTASSIUM 0 2 ( 0.8%) 2 ( 0.4%) ANEMARRHENA ASPHODELOIDES RHIZOME;ANGELICA DAHURICA VAR. 0 5 ( 1.9%) 5 ( 1.0%) FORMOSANA ROOT;ASARUM HETEROTROPOIDES VAR. MANDSHURICUM HERB; CHRYSANTHEMUM INDICUM INFLORESCENCE;EPHEDRA SINICA HERB; FORSYTHIA SU ANGELICA DAHURICA ROOT;ARECA CATECHU PEEL;ATRACTYLODES SPP. 0 1 ( 0.4%) 1 ( 0.2%) RHIZOME;CITRUS RETICULATA FRUIT PEEL;GLYCYRRHIZA URALENSIS; MAGNOLIA OFFICINALIS BARK;PERILLA FRUTESCENS;PINELLIA TERNATA RHIZOM

PPD

124 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 23 of 36)

ACT UC Total N=242 N=265 N=507 ANGELICA DAHURICA VAR. FORMOSANA ROOT;ASTRAGALUS MONGHOLICUS 0 2 ( 0.8%) 2 ( 0.4%) ROOT;BUPLEURUM CHINENSE, ROOT;CLEMATIS ARMANDII STEM;GARDENIA JASMINOIDES FRUIT;GENTIANA SCABRA ROOT;LIGUSTICUM CHUANXIONG RHIZ ANGELICA DAHURICA VAR. FORMOSANA ROOT;CENTIPEDA MINIMA HERB; 0 1 ( 0.4%) 1 ( 0.2%) GLYCYRRHIZA URALENSIS ROOT WITH RHIZOME;MAGNOLIA BIONDII FLOWER BUD;MENTHA CANADENSIS HERB;SCUTELLARIA BAICALENSIS ROOT;XANTHIUM

ANGELICA SINENSIS ROOT;CANNABIS SATIVA FRUIT;NOTOPTERYGIUM 0 1 ( 0.4%) 1 ( 0.2%) INCISUM ROOT WITH RHIZOME;PRUNUS PERSICA SEED;RHEUM PALMATUM ROOT WITH RHIZOME ARCHIDENDRON CLYPEARIA FRUIT DRY EXTRACT 0 1 ( 0.4%) 1 ( 0.2%) ARISAEMA SPP. TUBER;COPTIS SPP. RHIZOME;FRITILLARIA CIRRHOSA 0 3 ( 1.1%) 3 ( 0.6%) BULB;HERBAL NOS;RHEUM SPP. ROOT WITH RHIZOME ASARONE 0 1 ( 0.4%) 1 ( 0.2%) ASPARAGUS COCHINCHINENSIS ROOT TUBER;GLYCYRRHIZA URALENSIS ROOT 0 1 ( 0.4%) 1 ( 0.2%) WITH RHIZOME;LONICERA CONFUSA FLOWER BUD;OPHIOPOGON JAPONICUS ROOT TUBER;SCROPHULARIA NINGPOENSIS ROOT ASTRAGALUS MONGHOLICUS ROOT 0 1 ( 0.4%) 1 ( 0.2%) ASTRAGALUS MONGHOLICUS ROOT;ATRACTYLODES MACROCEPHALA, RHIZOMA; 0 2 ( 0.8%) 2 ( 0.4%) SAPOSHNIKOVIA DIVARICATA ROOT ASTRAGALUS MONGHOLICUS ROOT;CRATAEGUS PINNATIFIDA FRUIT; 0 1 ( 0.4%) 1 ( 0.2%) LIGUSTICUM CHUANXIONG RHIZOME;OPHIOPOGON JAPONICUS ROOT TUBER; PANAX GINSENG ROOT;SALVIA MILTIORRHIZA ROOT;SCHISANDRA CHINENSIS FRU ATROPINE 0 1 ( 0.4%) 1 ( 0.2%) AZELASTINE 0 1 ( 0.4%) 1 ( 0.2%)

PPD

125 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 24 of 36)

ACT UC Total N=242 N=265 N=507 BACILLUS LICHENFORMIS 0 1 ( 0.4%) 1 ( 0.2%) BACITRACIN;LIDOCAINE HYDROCHLORIDE;NEOMYCIN SULFATE;POLYMYXIN B 0 2 ( 0.8%) 2 ( 0.4%) SULFATE BAMBUSA SPP.;ERIOBOTRYA JAPONICA LEAF;HOUTTUYNIA CORDATA;MENTHA 0 4 ( 1.5%) 4 ( 0.8%) CANADENSIS;PINELLIA TERNATA TUBER;PLATYCODON GRANDIFLORUS ROOT; ZINGIBER OFFICINALE RHIZOME BASIC FIBROBLAST GROWTH FACTOR 0 1 ( 0.4%) 1 ( 0.2%) BETAHISTINE MESILATE 0 1 ( 0.4%) 1 ( 0.2%) BIFIDOBACTERIUM BIFIDUM;ESCHERICHIA COLI 0 1 ( 0.4%) 1 ( 0.2%) BISMUTH PECTIN 0 1 ( 0.4%) 1 ( 0.2%) BLOOD, CALF, DEPROT., LMW PORTION 0 1 ( 0.4%) 1 ( 0.2%) BULLEYACONITINE A 0 1 ( 0.4%) 1 ( 0.2%) BUPIVACAINE HYDROCHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) CAFFEINE;CHLORPHENAMINE MALEATE;COW BEZOAR;PARACETAMOL 0 1 ( 0.4%) 1 ( 0.2%) CAFFEINE;PARACETAMOL;PROPYPHENAZONE 0 2 ( 0.8%) 2 ( 0.4%)

PPD

126 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 25 of 36)

ACT UC Total N=242 N=265 N=507 CALAMINE 0 1 ( 0.4%) 1 ( 0.2%) CALAMINE;SULFUR 0 1 ( 0.4%) 1 ( 0.2%) CALCITONIN, SALMON 0 1 ( 0.4%) 1 ( 0.2%) CALCITRIOL 0 2 ( 0.8%) 2 ( 0.4%) CALCIUM CARBONATE;COLECALCIFEROL 0 1 ( 0.4%) 1 ( 0.2%) CALCIUM CHLORIDE DIHYDRATE;POTASSIUM CHLORIDE;SODIUM CHLORIDE; 0 2 ( 0.8%) 2 ( 0.4%) SODIUM LACTATE CALCIUM FOLINATE 0 1 ( 0.4%) 1 ( 0.2%) CALCIUM SULFATE DIHYDRATE;DRYOPTERIS CRASSIRHIZOMA RHIZOME; 0 1 ( 0.4%) 1 ( 0.2%) EPHEDRA SPP. HERB;FORSYTHIA SUSPENSA;GLYCYRRHIZA GLABRA ROOT; HOUTTUYNIA CORDATA HERB;ISATIS TINCTORIA ROOT;LONICERA JAPONICA FL CAMPHOR;MENTHOL 0 1 ( 0.4%) 1 ( 0.2%) CARTHAMUS TINCTORIUS FLOWER;SALVIA MILTIORRHIZA ROOT 0 2 ( 0.8%) 2 ( 0.4%) CEFMETAZOLE 0 1 ( 0.4%) 1 ( 0.2%) CEFMETAZOLE SODIUM 0 1 ( 0.4%) 1 ( 0.2%)

PPD

127 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 26 of 36)

ACT UC Total N=242 N=265 N=507 CEFOTAXIME SODIUM 0 2 ( 0.8%) 2 ( 0.4%) CEFOXITIN 0 1 ( 0.4%) 1 ( 0.2%) CHLORAMPHENICOL 0 4 ( 1.5%) 4 ( 0.8%) CHLORPHENAMINE MALEATE;DEXTROMETHORPHAN HYDROBROMIDE;EPHEDRINE 0 1 ( 0.4%) 1 ( 0.2%) HYDROCHLORIDE;GUAIFENESIN;PARACETAMOL CHLORPHENAMINE MALEATE;PARACETAMOL;PSEUDOEPHEDRINE HYDROCHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) CHLORTETRACYCLINE HYDROCHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) CLOBETASOL 0 1 ( 0.4%) 1 ( 0.2%) CLOBETASOL PROPIONATE;KETOCONAZOLE;NEOMYCIN SULFATE 0 1 ( 0.4%) 1 ( 0.2%) CLOPIDOGREL BISULFATE 0 2 ( 0.8%) 2 ( 0.4%) COLCHICUM AUTUMNALE BULB;CONVOLVULUS SCAMMONIA RESIN;CROCUS 0 2 ( 0.8%) 2 ( 0.4%) SATIVUS STYLE;OPERCULINA TURPETHUM HERB;SENNA ALEXANDRINA LEAF; TERMINALIA CHEBULA FRUIT CONBERCEPT 0 1 ( 0.4%) 1 ( 0.2%) CORDYCEPS SINENSIS 0 1 ( 0.4%) 1 ( 0.2%)

PPD

128 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 27 of 36)

ACT UC Total N=242 N=265 N=507 CYSTEINE HYDROCHLORIDE;GLYCINE;GLYCYRRHIZIC ACID, AMMONIUM SALT 0 1 ( 0.4%) 1 ( 0.2%) DESONIDE 0 3 ( 1.1%) 3 ( 0.6%) DEXAMETHASONE SODIUM PHOSPHATE 0 2 ( 0.8%) 2 ( 0.4%) DEXTRAN 70;GLUCOSE 0 1 ( 0.4%) 1 ( 0.2%) DEXTROMETHORPHAN HYDROBROMIDE;GUAIFENESIN 0 1 ( 0.4%) 1 ( 0.2%) DICLOFENAC SODIUM 0 4 ( 1.5%) 4 ( 0.8%) DIHYDROERGOTOXINE 0 1 ( 0.4%) 1 ( 0.2%) DIPROPHYLLINE 0 7 ( 2.6%) 7 ( 1.4%) DIRITHROMYCIN 0 2 ( 0.8%) 2 ( 0.4%) DOMPERIDONE 0 1 ( 0.4%) 1 ( 0.2%) DOXOFYLLINE;GLUCOSE 0 1 ( 0.4%) 1 ( 0.2%) EDARAVONE 0 2 ( 0.8%) 2 ( 0.4%)

PPD

129 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 28 of 36)

ACT UC Total N=242 N=265 N=507 ELECTROLYTES NOS;GLUCOSE 0 1 ( 0.4%) 1 ( 0.2%) ELEUTHEROCOCCUS SENTICOSUS ROOT WITH RHIZOME;HYPERICUM 0 1 ( 0.4%) 1 ( 0.2%) PERFORATUM HERB ERGOCALCIFEROL 0 1 ( 0.4%) 1 ( 0.2%) ERYTHROMYCIN 0 1 ( 0.4%) 1 ( 0.2%) ESCIN 0 2 ( 0.8%) 2 ( 0.4%) ESOMEPRAZOLE 0 1 ( 0.4%) 1 ( 0.2%) ESOMEPRAZOLE MAGNESIUM 0 2 ( 0.8%) 2 ( 0.4%) ESTAZOLAM 0 3 ( 1.1%) 3 ( 0.6%) ETORICOXIB 0 1 ( 0.4%) 1 ( 0.2%) FAMOTIDINE 0 1 ( 0.4%) 1 ( 0.2%) FLUOROMETHOLONE 0 1 ( 0.4%) 1 ( 0.2%) FLUOROMETHOLONE;GENTAMICIN 0 1 ( 0.4%) 1 ( 0.2%)

PPD

130 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 29 of 36)

ACT UC Total N=242 N=265 N=507 FLUOROURACIL 0 2 ( 0.8%) 2 ( 0.4%) FLUPENTIXOL DIHYDROCHLORIDE;MELITRACEN HYDROCHLORIDE 0 2 ( 0.8%) 2 ( 0.4%) FLUPENTIXOL;MELITRACEN 0 2 ( 0.8%) 2 ( 0.4%) FLURBIPROFEN 0 1 ( 0.4%) 1 ( 0.2%) FOSFOMYCIN SODIUM 0 1 ( 0.4%) 1 ( 0.2%) FRUCTOSE;GLUCOSE;MAGNESIUM CHLORIDE;POTASSIUM CHLORIDE;SODIUM 0 1 ( 0.4%) 1 ( 0.2%) CHLORIDE;SODIUM LACTATE;SODIUM PHOSPHATE MONOBASIC FURAZOLIDONE;INDOMETACIN 0 1 ( 0.4%) 1 ( 0.2%) GENTAMICIN 0 1 ( 0.4%) 1 ( 0.2%) GLYCERYL TRINITRATE 0 1 ( 0.4%) 1 ( 0.2%) GRAMICIDIN;NEOMYCIN SULFATE;NYSTATIN;TRIAMCINOLONE ACETONIDE 0 1 ( 0.4%) 1 ( 0.2%) HAEMOCOAGULASE 0 1 ( 0.4%) 1 ( 0.2%) HOUTTUYFONATE 0 1 ( 0.4%) 1 ( 0.2%)

PPD

131 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 30 of 36)

ACT UC Total N=242 N=265 N=507 HYDROCHLOROTHIAZIDE 0 1 ( 0.4%) 1 ( 0.2%) HYDROCHLOROTHIAZIDE;LOSARTAN 0 2 ( 0.8%) 2 ( 0.4%) INDAPAMIDE;PERINDOPRIL 0 1 ( 0.4%) 1 ( 0.2%) INSULIN HUMAN;INSULIN HUMAN INJECTION, ISOPHANE 0 1 ( 0.4%) 1 ( 0.2%) ISOSORBIDE DINITRATE 0 1 ( 0.4%) 1 ( 0.2%) ITRACONAZOLE 0 1 ( 0.4%) 1 ( 0.2%) IXERIS SONCHIFOLIA HERB 0 1 ( 0.4%) 1 ( 0.2%) KETOTIFEN 0 1 ( 0.4%) 1 ( 0.2%) LACIDIPINE 0 1 ( 0.4%) 1 ( 0.2%) LACTOBACILLUS ACIDOPHILUS COMPOUND 0 1 ( 0.4%) 1 ( 0.2%) LAMIOPHLOMIS ROTATA ROOT WITH RHIZOME 0 1 ( 0.4%) 1 ( 0.2%) LANSOPRAZOLE 0 3 ( 1.1%) 3 ( 0.6%)

PPD

132 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 31 of 36)

ACT UC Total N=242 N=265 N=507 LEVAMLODIPINE BESILATE 0 1 ( 0.4%) 1 ( 0.2%) LEVOCETIRIZINE 0 5 ( 1.9%) 5 ( 1.0%) LEVOCETIRIZINE DIHYDROCHLORIDE 0 3 ( 1.1%) 3 ( 0.6%) LEVOFLOXACIN HYDROCHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) LIDOCAINE HYDROCHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) LINCOMYCIN 0 1 ( 0.4%) 1 ( 0.2%) LONICERA JAPONICA FLOWER BUD;SCUTELLARIA BAICALENSIS ROOT 0 1 ( 0.4%) 1 ( 0.2%) EXTRACT LOSARTAN POTASSIUM 0 2 ( 0.8%) 2 ( 0.4%) LOXOPROFEN SODIUM 0 3 ( 1.1%) 3 ( 0.6%) LUMBROKINASE 0 1 ( 0.4%) 1 ( 0.2%) MAGNESIUM CARBONATE 0 1 ( 0.4%) 1 ( 0.2%) MANNITOL 0 1 ( 0.4%) 1 ( 0.2%)

PPD

133 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 32 of 36)

ACT UC Total N=242 N=265 N=507 MEDICATED DRESSINGS 0 1 ( 0.4%) 1 ( 0.2%) METHYLTHIONINIUM CHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) METOCLOPRAMIDE 0 1 ( 0.4%) 1 ( 0.2%) METOPROLOL TARTRATE 0 2 ( 0.8%) 2 ( 0.4%) MICONAZOLE 0 1 ( 0.4%) 1 ( 0.2%) MIDAZOLAM 0 1 ( 0.4%) 1 ( 0.2%) MIZOLASTINE 0 1 ( 0.4%) 1 ( 0.2%) MOSAPRIDE CITRATE 0 2 ( 0.8%) 2 ( 0.4%) MOXIFLOXACIN 0 1 ( 0.4%) 1 ( 0.2%) NICERGOLINE 0 1 ( 0.4%) 1 ( 0.2%) OLOPATADINE HYDROCHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) OMEPRAZOLE SODIUM 0 1 ( 0.4%) 1 ( 0.2%)

PPD

134 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 33 of 36)

ACT UC Total N=242 N=265 N=507 OPHIOPOGON JAPONICUS ROOT TUBER;PANAX GINSENG 0 1 ( 0.4%) 1 ( 0.2%) ORNIDAZOLE 0 1 ( 0.4%) 1 ( 0.2%) OTHER COLD PREPARATIONS 0 1 ( 0.4%) 1 ( 0.2%) OTHER RESPIRATORY SYSTEM PRODUCTS 0 1 ( 0.4%) 1 ( 0.2%) OXALIPLATIN 0 1 ( 0.4%) 1 ( 0.2%) PANAX NOTOGINSENG ROOT TOTAL SAPONIN EXTRACT 0 1 ( 0.4%) 1 ( 0.2%) PANCREATIN 0 1 ( 0.4%) 1 ( 0.2%) PARACETAMOL;PHENYLEPHRINE 0 1 ( 0.4%) 1 ( 0.2%) PARACETAMOL;TRAMADOL HYDROCHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) PENICILLIN NOS 0 1 ( 0.4%) 1 ( 0.2%) PINAVERIUM BROMIDE 0 2 ( 0.8%) 2 ( 0.4%) PREDNISOLONE 0 1 ( 0.4%) 1 ( 0.2%)

PPD

135 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 34 of 36)

ACT UC Total N=242 N=265 N=507 PREDNISOLONE ACETATE 0 1 ( 0.4%) 1 ( 0.2%) PREDNISONE 0 2 ( 0.8%) 2 ( 0.4%) PROMETHAZINE 0 1 ( 0.4%) 1 ( 0.2%) REPAGLINIDE 0 2 ( 0.8%) 2 ( 0.4%) RETINOL 0 1 ( 0.4%) 1 ( 0.2%) RIBAVIRIN 0 1 ( 0.4%) 1 ( 0.2%) RIBOFLAVIN 0 2 ( 0.8%) 2 ( 0.4%) ROXITHROMYCIN 0 1 ( 0.4%) 1 ( 0.2%) SALVIA MILTIORRHIZA 0 1 ( 0.4%) 1 ( 0.2%) SAXAGLIPTIN 0 1 ( 0.4%) 1 ( 0.2%) SELENIUM 0 2 ( 0.8%) 2 ( 0.4%) SODIUM CHLORIDE;TINIDAZOLE 0 1 ( 0.4%) 1 ( 0.2%)

PPD

136 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 35 of 36)

ACT UC Total N=242 N=265 N=507 SODIUM FERULATE 0 1 ( 0.4%) 1 ( 0.2%) SPIRONOLACTONE 0 1 ( 0.4%) 1 ( 0.2%) SUCCINYLATED GELATIN 0 1 ( 0.4%) 1 ( 0.2%) SULPIRIDE 0 1 ( 0.4%) 1 ( 0.2%) TACROLIMUS 0 1 ( 0.4%) 1 ( 0.2%) TAMSULOSIN HYDROCHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) TERBUTALINE SULFATE 0 1 ( 0.4%) 1 ( 0.2%) TETANUS ANTITOXIN 0 1 ( 0.4%) 1 ( 0.2%) THEOPHYLLINE 0 1 ( 0.4%) 1 ( 0.2%) THESIUM CHINENSE HERB 0 2 ( 0.8%) 2 ( 0.4%) THIOCTIC ACID 0 2 ( 0.8%) 2 ( 0.4%) TICAGRELOR 0 1 ( 0.4%) 1 ( 0.2%)

PPD

137 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 1.09 Subjects' Concurrent Medication (ITT) (Page 36 of 36)

ACT UC Total N=242 N=265 N=507 TRANEXAMIC ACID 0 1 ( 0.4%) 1 ( 0.2%) TRIMEBUTINE MALEATE 0 5 ( 1.9%) 5 ( 1.0%) UREA 0 1 ( 0.4%) 1 ( 0.2%) VERAPAMIL HYDROCHLORIDE 0 1 ( 0.4%) 1 ( 0.2%) VINPOCETINE 0 1 ( 0.4%) 1 ( 0.2%) WARFARIN SODIUM 0 2 ( 0.8%) 2 ( 0.4%)

PPD

138 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 1 Forest Plot of Adjusted Odds Ratios and 95% CIs From GLMM, Comparing ACT Guided Treatment With Usual Care On Primary Analysis and Sensitivity Analysis of ACT Response (ITT) (Page 1 of 1)

[1] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

PPD

139 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 2 Forest Plot of Adjusted Odds Ratios and 95% CIs From GLMM, Comparing ACT Guided Treatment With Usual Care On Primary Analysis and Sensitivity Analysis of ACT Response (PP) (Page 1 of 1)

[1] Subjects who have an ACT total score >= 20 or an improvement > 3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

PPD

140 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 3 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (ITT) (Page 1 of 1)

[1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, gender, visit and treatment-by-visit interaction.

PPD

141 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 4 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (PP) (Page 1 of 1)

[1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, gender, visit and treatment-by-visit interaction.

PPD

142 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 5 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (ITT) (Page 1 of 1)

[1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, gender, visit and treatment-by-visit interaction.

PPD

143 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 6 Forest Plot of Adjusted Odds Ratios and 95% CIs From GEEs, Comparing ACT Guided Treatment With Usual Care On Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (PP) (Page 1 of 1)

[1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, gender, visit and treatment-by-visit interaction.

PPD

144 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 7 Forest Plot of LS Mean change from baseline treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (ITT) (Page 1 of 1)

Note: The analysis method was a MMRM model adjusted for of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect.

PPD

145 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 8 Forest Plot of LS Mean change from baseline treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On ACT Total Score At Weeks 4, 8, 12, 16, 20 and 24 (PP) (Page 1 of 1)

Note: The analysis method was a MMRM model adjusted for of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect.

PPD

146 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 9 Forest Plot of LS Mean treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On Mean Daytime Symptom Score By Visits (ITT) (Page 1 of 1)

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

147 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 10 Forest Plot of LS Mean treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On Mean Night-time Symptom Score By Visits (ITT) (Page 1 of 1)

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

148 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 11 Forest Plot of LS Mean change from baseline treatment differenes and 95% CIs From Mixed Model, Comparing ACT Guided Treatment With Usual Care On FEV1 (ITT) (Page 1 of 1)

Note: Analysis performed using a mixed model with covariates of treatment, center, baseline FEV1, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the center as a random factor. Only include those who complete the study and exclude those who withdraw the study. FEV1 stands for Forced Expiratory Volume in One Second.

PPD

149 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 12 Forest Plot of LS Mean treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On Mean Morning PEF By Visits (ITT) (Page 1 of 1)

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor. PEF stands for Peak Expiratory Flow.

PPD

150 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 13 Forest Plot of LS Mean treatment differenes and 95% CIs From MMRM, Comparing ACT Guided Treatment With Usual Care On Mean Evening PEF By Visits (ITT) (Page 1 of 1)

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor. PEF stands for Peak Expiratory Flow.

PPD

151 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 14 Forest Plot of LS Mean change from baseline treatment differenes and 95% CIs From Mixed Model, Comparing ACT Guided Treatment With Usual Care On AQLQ(S) score (ITT) (Page 1 of 1)

Note: Analysis performed using a mixed model with covariates of treatment, center, baseline AQLQ score, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the center as a random factor. Only include those who complete the study and exclude those who withdraw the study. AQLQ(S) stands for Asthma Quality of Life Questionnaire.

PPD

152 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date： 29SEP2019 Figure 15 KM Figure of Time to Asthma Control (ACT Total Score >=20 Or Improvement of More Than 3 Points in ACT Total Score) For the First Time(ITT) (Page 1 of 1)

PPD

153 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date： 29SEP2019 Figure 16 KM Figure of Time to Asthma Control (ACT Total Score >=20 And Improvement of More Than 3 Points in ACT Total Score) For the First Time(ITT) (Page 1 of 1)

PPD

154 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date: 03JAN2020

Figure 17 Forest Plot of Adjusted Odds Ratios and 95% CIs From GLMM, Comparing ACT Guided Treatment With Usual Care On Primary Analysis and Sensitivity Analysis of ACT Response (ITT) (Page 1 of 1)

[1] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

PPD

155 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date： 20NOV2019 Figure 18 KM Figure of Time to Asthma Control (ACT Total Score >=20 Or Improvement >= 3 Points in ACT Total Score) For the First Time(ITT) (Page 1 of 1)

100 90 80 70 60 50 40 30 20 ACT

10 Estimated Probability of Having an Event(%) an Having of Probability Estimated Estimated Probability of Having an Event(%) an Having of Probability Estimated UC 0 Total 0 28 56 84 112 140 168 Time to asthma control for the first time (days) Number of subjects at Risk

ACT 241 205 14 3 1 0 0

UC 264 224 66 27 18 14 10

Total 505 429 80 30 19 14 10

PPD

156 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version 1.0 Protocol: 201097 Date： 20NOV2019 Figure 19 KM Figure of Time to Asthma Control (ACT Total Score >=20 And Improvement >= 3 Points in ACT Total Score) For the First Time(ITT) (Page 1 of 1)

100 90 80 70 60 50 40 30 20 ACT

10 Estimated Probability of Having an Event(%) an Having of Probability Estimated Estimated Probability of Having an Event(%) an Having of Probability Estimated UC 0 Total 0 28 56 84 112 140 168 Time to asthma control for the first time (days) Number of subjects at Risk

ACT 241 210 55 18 7 4 1

UC 264 231 105 56 39 30 21

Total 505 441 160 74 46 34 22

PPD

157 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29OCT2019

Table 2.01 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period (ITT) (Page 1 of 1)

ACT UC N=242 N=265 n 240 262 Responder (%) [1] 238 ( 99.2%) 247 ( 94.3%) Non-Responder (%) 2 ( 0.8%) 15 ( 5.7%)

ACT vs. UC [2] Adjusted Odds Ratio 7.87 95% CI (1.29, 48.11) P-value 0.027

Interaction between treatment and covariates, P-value [3] With type of baseline controller 0.982 With gender 0.975 With age 0.891 [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor. [3] The interaction between treatment and each covariate was studied sequentially, and significant effects were considered at a 10% significance level.

PPD

158 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29OCT2019

Table 2.02 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period (PP) (Page 1 of 1)

ACT UC N=208 N=231 n 208 230 Responder (%) [1] 207 ( 99.5%) 218 ( 94.8%) Non-Responder (%) 1 ( 0.5%) 12 ( 5.2%)

ACT vs. UC [2] Adjusted Odds Ratio 12.10 95% CI (1.16, 126.19) P-value 0.038

Interaction between treatment and covariates, P-value [3] With type of baseline controller 0.976 With gender 0.978 With age 0.226 [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor. [3] The interaction between treatment and each covariate was studied sequentially, and significant effects were considered at a 10% significance level.

PPD

159 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.03 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (ITT) (Page 1 of 6)

ACT UC Subgroup N=242 N=265

Type of Baseline Controller: ICS Alone n 8 12 Responder (%) [1] 8 (100.0%) 11 ( 91.7%) Non-Responder (%) 0 1 ( 8.3%)

ACT vs. UC [2] Adjusted Odds Ratio NA 95% CI NA P-value NA [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

160 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.03 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (ITT) (Page 2 of 6)

ACT UC Subgroup N=242 N=265

Type of Baseline Controller: ICS/LABA n 232 250 Responder (%) [1] 230 ( 99.1%) 236 ( 94.4%) Non-Responder (%) 2 ( 0.9%) 14 ( 5.6%)

ACT vs. UC [2] Adjusted Odds Ratio 6.43 95% CI (0.99, 41.59) P-value 0.051 [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

161 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.03 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (ITT) (Page 3 of 6)

ACT UC Subgroup N=242 N=265

Gender: Male n 126 129 Responder (%) [1] 124 ( 98.4%) 119 ( 92.2%) Non-Responder (%) 2 ( 1.6%) 10 ( 7.8%)

ACT vs. UC [3] Adjusted Odds Ratio 7.06 95% CI (1.16, 42.78) P-value 0.034 [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

162 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.03 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (ITT) (Page 4 of 6)

ACT UC Subgroup N=242 N=265

Gender: Female n 114 133 Responder (%) [1] 114 (100.0%) 128 ( 96.2%) Non-Responder (%) 0 5 ( 3.8%)

ACT vs. UC [3] Adjusted Odds Ratio NA 95% CI NA P-value NA [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

163 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.03 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (ITT) (Page 5 of 6)

ACT UC Subgroup N=242 N=265

Age: <50 Years Old n 123 131 Responder (%) [1] 122 ( 99.2%) 123 ( 93.9%) Non-Responder (%) 1 ( 0.8%) 8 ( 6.1%)

ACT vs. UC [4] Adjusted Odds Ratio NA 95% CI NA P-value NA [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

164 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.03 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (ITT) (Page 6 of 6)

ACT UC Subgroup N=242 N=265

Age: >=50 Years Old n 117 131 Responder (%) [1] 116 ( 99.1%) 124 ( 94.7%) Non-Responder (%) 1 ( 0.9%) 7 ( 5.3%)

ACT vs. UC [4] Adjusted Odds Ratio 8.45 95% CI (0.37, 195.56) P-value 0.169 [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

165 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.04 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (PP) (Page 1 of 6)

ACT UC Subgroup N=208 N=231

Type of Baseline Controller: ICS Alone n 7 10 Responder (%) [1] 7 (100.0%) 9 ( 90.0%) Non-Responder (%) 0 1 ( 10.0%)

ACT vs. UC [2] Adjusted Odds Ratio NA 95% CI NA P-value NA [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

166 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.04 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (PP) (Page 2 of 6)

ACT UC Subgroup N=208 N=231

Type of Baseline Controller: ICS/LABA n 201 220 Responder (%) [1] 200 ( 99.5%) 209 ( 95.0%) Non-Responder (%) 1 ( 0.5%) 11 ( 5.0%)

ACT vs. UC [2] Adjusted Odds Ratio 9.80 95% CI (0.91, 105.26) P-value 0.059 [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

167 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.04 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (PP) (Page 3 of 6)

ACT UC Subgroup N=208 N=231

Gender: Male n 107 117 Responder (%) [1] 106 ( 99.1%) 108 ( 92.3%) Non-Responder (%) 1 ( 0.9%) 9 ( 7.7%)

ACT vs. UC [3] Adjusted Odds Ratio 12.77 95% CI (1.20, 136.02) P-value 0.035 [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

168 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.04 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (PP) (Page 4 of 6)

ACT UC Subgroup N=208 N=231

Gender: Female n 101 113 Responder (%) [1] 101 (100.0%) 110 ( 97.3%) Non-Responder (%) 0 3 ( 2.7%)

ACT vs. UC [3] Adjusted Odds Ratio NA 95% CI NA P-value NA [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

169 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.04 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (PP) (Page 5 of 6)

ACT UC Subgroup N=208 N=231

Age: <50 Years Old n 102 115 Responder (%) [1] 101 ( 99.0%) 108 ( 93.9%) Non-Responder (%) 1 ( 1.0%) 7 ( 6.1%)

ACT vs. UC [4] Adjusted Odds Ratio NA 95% CI NA P-value NA [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

170 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.04 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period by Subgroup (PP) (Page 6 of 6)

ACT UC Subgroup N=208 N=231

Age: >=50 Years Old n 106 115 Responder (%) [1] 106 (100.0%) 110 ( 95.7%) Non-Responder (%) 0 5 ( 4.3%)

ACT vs. UC [4] Adjusted Odds Ratio 8.45 95% CI (0.37, 195.56) P-value 0.169 [1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, gender and age, with the centre as a random factor. [3] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and age, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre and gender, type of baseline controller (ICS vs. ICS/LABA), with the centre as a random factor.

PPD

171 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.05 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or AnImprovement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (ITT) (Page 1 of 3)

ACT UC Visit (N=242) (N=265) Week 4 n 236 261 Responder (%) [1] 208( 88.1%) 177( 67.8%) Non-Responder (%) 28( 11.9%) 84( 32.2%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

Week 8 n 226 248 Responder (%) [1] 219( 96.9%) 197( 79.4%) Non-Responder (%) 7( 3.1%) 51( 20.6%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

[1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

172 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.05 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or AnImprovement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (ITT) (Page 2 of 3)

ACT UC Visit (N=242) (N=265) Week 12 n 214 244 Responder (%) [1] 209( 97.7%) 207( 84.8%) Non-Responder (%) 5( 2.3%) 37( 15.2%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

Week 16 n 215 236 Responder (%) [1] 213( 99.1%) 196( 83.1%) Non-Responder (%) 2( 0.9%) 40( 16.9%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

[1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

173 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.05 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or AnImprovement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (ITT) (Page 3 of 3)

ACT UC Visit (N=242) (N=265) Week 20 n 207 230 Responder (%) [1] 203( 98.1%) 199( 86.5%) Non-Responder (%) 4( 1.9%) 31( 13.5%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

Week 24 n 208 233 Responder (%) [1] 208(100.0%) 209( 89.7%) Non-Responder (%) 0 24( 10.3%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

[1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

174 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.06 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (PP) (Page 1 of 3)

ACT UC Visit (N=208) (N=231) Week 4 n 205 230 Responder (%) [1] 184( 89.8%) 158( 68.7%) Non-Responder (%) 21( 10.2%) 72( 31.3%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

Week 8 n 203 223 Responder (%) [1] 200( 98.5%) 177( 79.4%) Non-Responder (%) 3( 1.5%) 46( 20.6%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

[1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

175 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.06 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (PP) (Page 2 of 3)

ACT UC Visit (N=208) (N=231) Week 12 n 190 218 Responder (%) [1] 186( 97.9%) 185( 84.9%) Non-Responder (%) 4( 2.1%) 33( 15.1%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

Week 16 n 192 216 Responder (%) [1] 191( 99.5%) 179( 82.9%) Non-Responder (%) 1( 0.5%) 37( 17.1%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

[1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

176 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.06 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (PP) (Page 3 of 3)

ACT UC Visit (N=208) (N=231) Week 20 n 185 212 Responder (%) [1] 182( 98.4%) 183( 86.3%) Non-Responder (%) 3( 1.6%) 29( 13.7%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

Week 24 n 184 210 Responder (%) [1] 184(100.0%) 189( 90.0%) Non-Responder (%) 0 21( 10.0%)

ACT vs. UC Adjusted Odds Ratio NA 95% CI NA P-value NA

[1] Subjects who have an ACT total score >= 20 or an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

177 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.07 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score (ITT) (Page 1 of 2)

ACT UC Visit (N=242) (N=265)

Week 12 n 241 263 Responder (%) [1] 183( 75.9%) 161( 61.2%) Non-Responder (%) 58( 24.1%) 102( 38.8%)

ACT vs. UC [4] Adjusted Odds Ratio 2.10 95% CI (0.82, 5.34) P-value 0.105

Week 24 n 241 263 Responder (%) [2] 200( 83.0%) 178( 67.7%) Non-Responder (%) 41( 17.0%) 85( 32.3%)

ACT vs. UC [4] Adjusted Odds Ratio 2.28 [1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Week 12. Subjects with missing data will be set as non-responders. [2] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Week 24. Subjects with missing data will be set as non-responders. [3] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Weeks 16 and 20 and 24. Subjects with missing data will be set as non-responders. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

PPD

178 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.07 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score (ITT) (Page 2 of 2)

ACT UC Visit (N=242) (N=265)

Week 24 95% CI (1.07, 4.85) P-value 0.036

Weeks 16 and 20 and 24 n 241 263 Responder (%) [3] 168( 69.7%) 124( 47.1%) Non-Responder (%) 73( 30.3%) 139( 52.9%)

ACT vs. UC [4] Adjusted Odds Ratio 2.52 95% CI (1.25, 5.07) P-value 0.016 [1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Week 12. Subjects with missing data will be set as non-responders. [2] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Week 24. Subjects with missing data will be set as non-responders. [3] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Weeks 16 and 20 and 24. Subjects with missing data will be set as non-responders. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

PPD

179 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.08 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score (PP) (Page 1 of 2)

ACT UC Visit (N=208) (N=231)

Week 12 n 208 230 Responder (%) [1] 162( 77.9%) 141( 61.3%) Non-Responder (%) 46( 22.1%) 89( 38.7%)

ACT vs. UC [4] Adjusted Odds Ratio 2.29 95% CI (0.75, 6.97) P-value 0.127

Week 24 n 208 230 Responder (%) [2] 177( 85.1%) 160( 69.6%) Non-Responder (%) 31( 14.9%) 70( 30.4%)

ACT vs. UC [4] Adjusted Odds Ratio 2.39 [1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Week 12. Subjects with missing data will be set as non-responders. [2] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Week 24. Subjects with missing data will be set as non-responders. [3] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Weeks 16 and 20 and 24. Subjects with missing data will be set as non-responders. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

PPD

180 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.08 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score (PP) (Page 2 of 2)

ACT UC Visit (N=208) (N=231)

Week 24 95% CI (1.08, 5.29) P-value 0.035

Weeks 16 and 20 and 24 n 208 230 Responder (%) [3] 151( 72.6%) 116( 50.4%) Non-Responder (%) 57( 27.4%) 114( 49.6%)

ACT vs. UC [4] Adjusted Odds Ratio 2.52 95% CI (1.15, 5.53) P-value 0.027 [1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Week 12. Subjects with missing data will be set as non-responders. [2] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Week 24. Subjects with missing data will be set as non-responders. [3] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at Weeks 16 and 20 and 24. Subjects with missing data will be set as non-responders. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

PPD

181 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.09 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (ITT) (Page 1 of 3)

ACT UC Visit (N=242) (N=265) Week 4 n 236 261 Responder (%) [1] 157( 66.5%) 122( 46.7%) Non-Responder (%) 79( 33.5%) 139( 53.3%)

ACT vs. UC Adjusted Odds Ratio 2.04 95% CI (1.41, 2.95) P-value <0.001

Week 8 n 226 248 Responder (%) [1] 184( 81.4%) 151( 60.9%) Non-Responder (%) 42( 18.6%) 97( 39.1%)

ACT vs. UC Adjusted Odds Ratio 2.63 95% CI (1.72, 4.03) P-value <0.001

[1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

182 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.09 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (ITT) (Page 2 of 3)

ACT UC Visit (N=242) (N=265) Week 12 n 214 244 Responder (%) [1] 183( 85.5%) 161( 66.0%) Non-Responder (%) 31( 14.5%) 83( 34.0%)

ACT vs. UC Adjusted Odds Ratio 2.71 95% CI (1.73, 4.26) P-value <0.001

Week 16 n 215 236 Responder (%) [1] 197( 91.6%) 156( 66.1%) Non-Responder (%) 18( 8.4%) 80( 33.9%)

ACT vs. UC Adjusted Odds Ratio 4.78 95% CI (2.81, 8.13) P-value <0.001

[1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

183 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.09 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (ITT) (Page 3 of 3)

ACT UC Visit (N=242) (N=265) Week 20 n 207 230 Responder (%) [1] 188( 90.8%) 161( 70.0%) Non-Responder (%) 19( 9.2%) 69( 30.0%)

ACT vs. UC Adjusted Odds Ratio 3.81 95% CI (2.22, 6.53) P-value <0.001

Week 24 n 208 233 Responder (%) [1] 200( 96.2%) 178( 76.4%) Non-Responder (%) 8( 3.8%) 55( 23.6%)

ACT vs. UC Adjusted Odds Ratio 6.51 95% CI (3.20, 13.25) P-value <0.001

[1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

184 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.10 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (PP) (Page 1 of 3)

ACT UC Visit (N=208) (N=231) Week 4 n 205 230 Responder (%) [1] 139( 67.8%) 107( 46.5%) Non-Responder (%) 66( 32.2%) 123( 53.5%)

ACT vs. UC Adjusted Odds Ratio 2.23 95% CI (1.50, 3.33) P-value <0.001

Week 8 n 203 223 Responder (%) [1] 167( 82.3%) 132( 59.2%) Non-Responder (%) 36( 17.7%) 91( 40.8%)

ACT vs. UC Adjusted Odds Ratio 2.96 95% CI (1.88, 4.67) P-value <0.001

[1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

185 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.10 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (PP) (Page 2 of 3)

ACT UC Visit (N=208) (N=231) Week 12 n 190 218 Responder (%) [1] 162( 85.3%) 141( 64.7%) Non-Responder (%) 28( 14.7%) 77( 35.3%)

ACT vs. UC Adjusted Odds Ratio 2.83 95% CI (1.76, 4.55) P-value <0.001

Week 16 n 192 216 Responder (%) [1] 176( 91.7%) 141( 65.3%) Non-Responder (%) 16( 8.3%) 75( 34.7%)

ACT vs. UC Adjusted Odds Ratio 5.16 95% CI (2.93, 9.07) P-value <0.001

[1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

186 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date: 14OCT2019

Table 2.10 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >3 Points in ACT Total Score at Weeks 4, 8, 12, 16, 20 And 24 (Generalised Estimating Equations) (PP) (Page 3 of 3)

ACT UC Visit (N=208) (N=231) Week 20 n 185 212 Responder (%) [1] 167( 90.3%) 148( 69.8%) Non-Responder (%) 18( 9.7%) 64( 30.2%)

ACT vs. UC Adjusted Odds Ratio 3.61 95% CI (2.06, 6.33) P-value <0.001

Week 24 n 184 210 Responder (%) [1] 177( 96.2%) 160( 76.2%) Non-Responder (%) 7( 3.8%) 50( 23.8%)

ACT vs. UC Adjusted Odds Ratio 7.11 95% CI (3.26, 15.49) P-value <0.001

[1] Subjects who have an ACT total score >= 20 and an improvement >3 points in ACT total score at current visit. Note: The analysis method was Generalised Estimating Equations adjusted for treatment, baseline ACT total score, baseline ACT total score squared, age, type of baseline controller, gender, visit and treatment-by-visit interaction.

PPD

187 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.11 Summary of the Statistical Analysis of Change from Baseline in ACT Total Score (ITT) (Page 1 of 3)

ACT UC Visit N=242 N=265

Week 4 n 236 261 LS Mean 21.3 20.4 LS Mean Change (SE) 4.5 (0.36) 3.6 (0.35)

ACT vs. UC Difference 0.9 95% CI (0.0, 1.8) P-value 0.042

Week 8 n 226 248 LS Mean 22.4 21.4 LS Mean Change (SE) 5.7 (0.35) 4.7 (0.34)

ACT vs. UC Difference 1.0 95% CI (0.2, 1.8) P-value 0.022

Note: The analysis method was a MMRM model adjusted for of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect.

PPD

188 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.11 Summary of the Statistical Analysis of Change from Baseline in ACT Total Score (ITT) (Page 2 of 3)

ACT UC Visit N=242 N=265

Week 12 n 214 244 LS Mean 22.8 21.7 LS Mean Change (SE) 6.0 (0.34) 5.0 (0.34)

ACT vs. UC Difference 1.0 95% CI (0.2, 1.8) P-value 0.021

Week 16 n 215 236 LS Mean 23.1 21.8 LS Mean Change (SE) 6.4 (0.35) 5.1 (0.34)

ACT vs. UC Difference 1.3 95% CI (0.4, 2.1) P-value 0.006

Note: The analysis method was a MMRM model adjusted for of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect.

PPD

189 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.11 Summary of the Statistical Analysis of Change from Baseline in ACT Total Score (ITT) (Page 3 of 3)

ACT UC Visit N=242 N=265

Week 20 n 207 230 LS Mean 23.2 22.2 LS Mean Change (SE) 6.5 (0.34) 5.5 (0.33)

ACT vs. UC Difference 1.1 95% CI (0.2, 1.9) P-value 0.016

Week 24 n 208 233 LS Mean 23.6 22.3 LS Mean Change (SE) 6.9 (0.33) 5.6 (0.33)

ACT vs. UC Difference 1.3 95% CI (0.5, 2.1) P-value 0.005

Note: The analysis method was a MMRM model adjusted for of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect.

PPD

190 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.12 Summary of the Statistical Analysis of Change from Baseline in ACT Total Score (PP) (Page 1 of 3)

ACT UC Visit N=208 N=231

Week 4 n 205 230 LS Mean 21.3 20.4 LS Mean Change (SE) 4.5 (0.36) 3.5 (0.35)

ACT vs. UC Difference 1.0 95% CI (0.1, 1.8) P-value 0.028

Week 8 n 203 223 LS Mean 22.4 21.4 LS Mean Change (SE) 5.6 (0.35) 4.6 (0.34)

ACT vs. UC Difference 1.1 95% CI (0.3, 1.9) P-value 0.012

Note: The analysis method was a MMRM model adjusted for of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect.

PPD

191 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.12 Summary of the Statistical Analysis of Change from Baseline in ACT Total Score (PP) (Page 2 of 3)

ACT UC Visit N=208 N=231

Week 12 n 190 218 LS Mean 22.7 21.7 LS Mean Change (SE) 5.9 (0.34) 4.9 (0.34)

ACT vs. UC Difference 1.0 95% CI (0.2, 1.8) P-value 0.019

Week 16 n 192 216 LS Mean 23.0 21.7 LS Mean Change (SE) 6.2 (0.35) 4.9 (0.34)

ACT vs. UC Difference 1.3 95% CI (0.5, 2.1) P-value 0.005

Note: The analysis method was a MMRM model adjusted for of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect.

PPD

192 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.12 Summary of the Statistical Analysis of Change from Baseline in ACT Total Score (PP) (Page 3 of 3)

ACT UC Visit N=208 N=231

Week 20 n 185 212 LS Mean 23.1 22.2 LS Mean Change (SE) 6.3 (0.34) 5.4 (0.33)

ACT vs. UC Difference 0.9 95% CI (0.1, 1.7) P-value 0.025

Week 24 n 184 210 LS Mean 23.5 22.3 LS Mean Change (SE) 6.7 (0.33) 5.5 (0.33)

ACT vs. UC Difference 1.2 95% CI (0.4, 2.0) P-value 0.006

Note: The analysis method was a MMRM model adjusted for of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect.

PPD

193 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.13 Summary Of Mean Daytime Symptom Score Over the 24-Week Treatment Period (ITT) (Page 1 of 2) ______ACT UC Visit N=242 N=265

Week 0 to Week 4 n 237 261 Mean 0.3 0.3 SD 0.46 0.54 Median 0.1 0.0 Min. 0.0 0.0 Max. 3.0 2.5

Week 4 to Week 8 n 227 250 Mean 0.2 0.2 SD 0.41 0.44 Median 0.0 0.0 Min. 0.0 0.0 Max. 2.1 2.0

Week 8 to Week 12 n 217 241 Mean 0.1 0.2 SD 0.39 0.39 Median 0.0 0.0 Min. 0.0 0.0 Max. 2.0 2.0

PPD

194 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.13 Summary Of Mean Daytime Symptom Score Over the 24-Week Treatment Period (ITT) (Page 2 of 2) ______ACT UC Visit N=242 N=265

Week 12 to Week 16 n 210 239 Mean 0.1 0.2 SD 0.33 0.42 Median 0.0 0.0 Min. 0.0 0.0 Max. 2.0 2.4

Week 16 to Week 20 n 207 234 Mean 0.1 0.2 SD 0.34 0.38 Median 0.0 0.0 Min. 0.0 0.0 Max. 2.0 2.2

Week 20 to Week 24 n 205 228 Mean 0.1 0.2 SD 0.31 0.36 Median 0.0 0.0 Min. 0.0 0.0 Max. 2.0 2.0

PPD

195 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.14 Analysis of Mean Daytime Symptom Score Over the 24-Week Treatment Period (ITT) (Page 1 of 3)

ACT UC Visit N=242 N=265

Week 0 to Week 4 n 236 259 LS Mean (SE) 0.3 (0.05) 0.4 (0.05) 95% CI (0.2, 0.4) (0.2, 0.5)

ACT vs. UC Difference -0.1 95% CI (-0.2, 0.0) P-value 0.222

Week 4 to Week 8 n 227 249 LS Mean (SE) 0.2 (0.05) 0.3 (0.05) 95% CI (0.1, 0.3) (0.2, 0.4)

ACT vs. UC Difference -0.1 95% CI (-0.2, 0.0) P-value 0.156

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

196 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.14 Analysis of Mean Daytime Symptom Score Over the 24-Week Treatment Period (ITT) (Page 2 of 3)

ACT UC Visit N=242 N=265

Week 8 to Week 12 n 217 240 LS Mean (SE) 0.2 (0.05) 0.2 (0.05) 95% CI (0.1, 0.3) (0.1, 0.3)

ACT vs. UC Difference -0.1 95% CI (-0.2, 0.0) P-value 0.245

Week 12 to Week 16 n 210 238 LS Mean (SE) 0.1 (0.05) 0.2 (0.05) 95% CI (0.0, 0.2) (0.1, 0.3)

ACT vs. UC Difference -0.1 95% CI (-0.2, 0.0) P-value 0.057

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

197 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.14 Analysis of Mean Daytime Symptom Score Over the 24-Week Treatment Period (ITT) (Page 3 of 3)

ACT UC Visit N=242 N=265

Week 16 to Week 20 n 207 233 LS Mean (SE) 0.1 (0.05) 0.2 (0.05) 95% CI (0.0, 0.2) (0.1, 0.3)

ACT vs. UC Difference -0.1 95% CI (-0.2, 0.0) P-value 0.151

Week 20 to Week 24 n 205 227 LS Mean (SE) 0.1 (0.05) 0.2 (0.05) 95% CI (0.0, 0.2) (0.1, 0.3)

ACT vs. UC Difference -0.1 95% CI (-0.2, 0.0) P-value 0.114

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

198 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.15 Summary Of Mean Night-time Symptom Score Over the 24-Week Treatment Period (ITT) (Page 1 of 2) ______ACT UC Visit N=242 N=265

Week 0 to Week 4 n 237 260 Mean 0.4 0.5 SD 0.60 0.65 Median 0.1 0.1 Min. 0.0 0.0 Max. 3.0 3.0

Week 4 to Week 8 n 226 250 Mean 0.3 0.3 SD 0.53 0.51 Median 0.0 0.0 Min. 0.0 0.0 Max. 2.5 3.0

Week 8 to Week 12 n 217 241 Mean 0.2 0.3 SD 0.46 0.50 Median 0.0 0.0 Min. 0.0 0.0 Max. 2.3 3.0

PPD

199 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.15 Summary Of Mean Night-time Symptom Score Over the 24-Week Treatment Period (ITT) (Page 2 of 2) ______ACT UC Visit N=242 N=265

Week 12 to Week 16 n 209 239 Mean 0.2 0.2 SD 0.42 0.47 Median 0.0 0.0 Min. 0.0 0.0 Max. 2.3 2.4

Week 16 to Week 20 n 207 234 Mean 0.2 0.2 SD 0.40 0.44 Median 0.0 0.0 Min. 0.0 0.0 Max. 2.3 2.2

Week 20 to Week 24 n 205 228 Mean 0.2 0.2 SD 0.44 0.42 Median 0.0 0.0 Min. 0.0 0.0 Max. 3.0 2.0

PPD

200 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.16 Analysis of Mean Night-time Symptom Score Over the 24-Week Treatment Period (ITT) (Page 1 of 3)

ACT UC Visit N=242 N=265

Week 0 to Week 4 n 236 258 LS Mean (SE) 0.4 (0.07) 0.5 (0.06) 95% CI (0.3, 0.6) (0.4, 0.6)

ACT vs. UC Difference -0.05 95% CI (-0.18, 0.09) P-value 0.487

Week 4 to Week 8 n 226 249 LS Mean (SE) 0.3 (0.06) 0.3 (0.06) 95% CI (0.2, 0.4) (0.2, 0.4)

ACT vs. UC Difference -0.04 95% CI (-0.16, 0.09) P-value 0.546

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

201 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.16 Analysis of Mean Night-time Symptom Score Over the 24-Week Treatment Period (ITT) (Page 2 of 3)

ACT UC Visit N=242 N=265

Week 8 to Week 12 n 217 240 LS Mean (SE) 0.2 (0.06) 0.3 (0.06) 95% CI (0.1, 0.3) (0.2, 0.4)

ACT vs. UC Difference -0.06 95% CI (-0.17, 0.06) P-value 0.314

Week 12 to Week 16 n 209 238 LS Mean (SE) 0.2 (0.06) 0.3 (0.06) 95% CI (0.1, 0.3) (0.1, 0.4)

ACT vs. UC Difference -0.07 95% CI (-0.19, 0.04) P-value 0.197

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

202 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.16 Analysis of Mean Night-time Symptom Score Over the 24-Week Treatment Period (ITT) (Page 3 of 3)

ACT UC Visit N=242 N=265

Week 16 to Week 20 n 207 233 LS Mean (SE) 0.2 (0.06) 0.2 (0.06) 95% CI (0.1, 0.3) (0.1, 0.3)

ACT vs. UC Difference -0.06 95% CI (-0.17, 0.06) P-value 0.282

Week 20 to Week 24 n 205 227 LS Mean (SE) 0.2 (0.06) 0.2 (0.06) 95% CI (0.1, 0.3) (0.1, 0.3)

ACT vs. UC Difference -0.03 95% CI (-0.15, 0.08) P-value 0.543

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

203 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 20NOV2019

Table 2.17 Summary of the Change From Baseline in Forced Expiratory Volume In One Second (ITT) (Page 1 of 1)

ACT UC N=242 N=265

Change From Baseline in Absolute Value of Forced Expiratory Volume in One Second (FEV1) (L) n 208 228 Mean 0.147 0.086 SD 0.3758 0.4086 Median 0.105 0.040 Min. -0.660 -0.860 Max. 1.570 3.300

Change From Baseline in Relative Value of Forced Expiratory Volume in One Second (FEV1) (%) n 208 228 Mean 3.961 2.861 SD 15.1143 12.0354 Median 3.450 1.100 Min. -96.100 -32.170 Max. 43.600 51.000 Note: This analysis only included subjects who completed the study and change from baseline was calculated as the difference between baseline visit and Visit 6. Only include those who complete the study and exclude those who withdraw the study.

PPD

204 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 20NOV2019

Table 2.18 Analysis of the Change From Baseline in Forced Expiratory Volume In One Second (ITT) (Page 1 of 1)

ACT UC N=242 N=265 Change From Baseline in Absolute Value of Forced Expiratory Volume in One Second (FEV1) (L)

n 208 227 LS Mean (SE) 0.159 (0.0573) 0.098 (0.0559) 95% CI (0.043, 0.275) (-0.015, 0.212)

ACT vs. UC Difference 0.060 95% CI (-0.059, 0.180) P-value 0.273

Change From Baseline in Relative Value of Forced Expiratory Volume in One Second (FEV1) (%)

n 208 227 LS Mean (SE) 5.683 (1.8296) 3.828 (1.7815) 95% CI (2.012, 9.353) (0.249, 7.407)

ACT vs. UC Difference 1.855 95% CI (-1.569, 5.278) P-value 0.243

Note: Analysis performed using a mixed model with covariates of treatment, center, baseline FEV1, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the center as a random factor. Only include those who complete the study and exclude those who withdraw the study.

PPD

205 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.19 Summary Of Mean Morning Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) (Page 1 of 2) ______ACT UC Visit N=242 N=265

Week 0 to Week 4 n 240 263 Mean 353.0 361.7 SD 118.82 123.05 Median 338.9 349.2 Min. 93.2 103.7 Max. 696.0 731.7

Week 4 to Week 8 n 228 252 Mean 365.6 372.5 SD 122.13 122.94 Median 351.0 357.3 Min. 98.8 110.0 Max. 700.4 726.3

Week 8 to Week 12 n 220 242 Mean 372.3 376.7 SD 125.78 121.93 Median 361.7 371.8 Min. 107.3 120.9 Max. 693.6 718.7

PPD

206 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.19 Summary Of Mean Morning Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) (Page 2 of 2) ______ACT UC Visit N=242 N=265

Week 12 to Week 16 n 211 240 Mean 371.0 377.0 SD 121.85 119.90 Median 362.3 372.8 Min. 117.8 116.8 Max. 730.4 713.9

Week 16 to Week 20 n 207 236 Mean 375.4 379.9 SD 123.29 121.15 Median 375.2 372.8 Min. 128.9 110.7 Max. 733.7 718.7

Week 20 to Week 24 n 206 230 Mean 374.2 378.5 SD 123.10 124.07 Median 376.8 376.2 Min. 113.2 111.7 Max. 738.7 708.7

PPD

207 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.20 Analysis of Mean Morning Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) (Page 1 of 3)

ACT UC Visit N=242 N=265

Week 0 to Week 4 n 239 261 LS Mean (SE) 340.4 (13.37) 352.4 (13.03) 95% CI (313.4, 367.4) (326.1, 378.8)

ACT vs. UC Difference -12.0 95% CI (-40.3, 16.3) P-value 0.367

Week 4 to Week 8 n 228 251 LS Mean (SE) 355.5 (13.39) 362.7 (13.05) 95% CI (328.5, 382.5) (336.3, 389.1)

ACT vs. UC Difference -7.2 95% CI (-35.5, 21.2) P-value 0.587

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

208 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.20 Analysis of Mean Morning Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) (Page 2 of 3)

ACT UC Visit N=242 N=265

Week 8 to Week 12 n 220 241 LS Mean (SE) 362.2 (13.42) 366.9 (13.08) 95% CI (335.1, 389.3) (340.4, 393.3)

ACT vs. UC Difference -4.7 95% CI (-33.1, 23.7) P-value 0.721

Week 12 to Week 16 n 211 239 LS Mean (SE) 364.6 (13.40) 367.9 (13.06) 95% CI (337.6, 391.7) (341.5, 394.4)

ACT vs. UC Difference -3.3 95% CI (-31.7, 25.1) P-value 0.801

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

209 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.20 Analysis of Mean Morning Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) (Page 3 of 3)

ACT UC Visit N=242 N=265

Week 16 to Week 20 n 207 235 LS Mean (SE) 366.0 (13.38) 371.0 (13.04) 95% CI (339.0, 393.0) (344.6, 397.4)

ACT vs. UC Difference -5.0 95% CI (-33.3, 23.4) P-value 0.704

Week 20 to Week 24 n 206 229 LS Mean (SE) 367.0 (13.43) 371.0 (13.08) 95% CI (340.0, 394.1) (344.6, 397.5)

ACT vs. UC Difference -4.0 95% CI (-32.4, 24.5) P-value 0.762

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

210 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.21 Summary Of Mean Evening Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) (Page 1 of 2) ______ACT UC Visit N=242 N=265

Week 0 to Week 4 n 240 263 Mean 353.4 362.8 SD 119.43 121.43 Median 341.3 350.3 Min. 103.5 122.2 Max. 697.9 728.9

Week 4 to Week 8 n 228 252 Mean 364.8 373.1 SD 122.92 123.03 Median 356.5 361.3 Min. 107.9 117.2 Max. 700.7 728.8

Week 8 to Week 12 n 220 242 Mean 371.0 376.2 SD 125.91 122.34 Median 365.6 365.0 Min. 119.7 121.0 Max. 692.3 714.8

PPD

211 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.21 Summary Of Mean Evening Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) (Page 2 of 2) ______ACT UC Visit N=242 N=265

Week 12 to Week 16 n 211 240 Mean 370.5 376.6 SD 122.37 119.47 Median 368.8 373.9 Min. 112.9 120.4 Max. 729.2 702.1

Week 16 to Week 20 n 207 236 Mean 374.6 378.7 SD 124.14 120.62 Median 377.4 373.9 Min. 128.5 108.2 Max. 733.0 724.9

Week 20 to Week 24 n 206 230 Mean 373.7 378.1 SD 124.44 123.92 Median 376.5 375.4 Min. 121.8 111.8 Max. 737.7 720.0

PPD

212 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.22 Analysis of Mean Evening Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) (Page 1 of 3)

ACT UC Visit N=242 N=265

Week 0 to Week 4 n 239 261 LS Mean (SE) 342.2 (13.55) 354.9 (13.22) 95% CI (314.8, 369.6) (328.1, 381.7)

ACT vs. UC Difference -12.8 95% CI (-42.1, 16.5) P-value 0.354

Week 4 to Week 8 n 228 251 LS Mean (SE) 356.2 (13.60) 364.6 (13.27) 95% CI (328.7, 383.7) (337.7, 391.5)

ACT vs. UC Difference -8.4 95% CI (-37.8, 21.0) P-value 0.538

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

213 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.22 Analysis of Mean Evening Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) (Page 2 of 3)

ACT UC Visit N=242 N=265

Week 8 to Week 12 n 220 241 LS Mean (SE) 362.7 (13.63) 367.7 (13.29) 95% CI (335.2, 390.2) (340.8, 394.6)

ACT vs. UC Difference -5.0 95% CI (-34.4, 24.5) P-value 0.716

Week 12 to Week 16 n 211 239 LS Mean (SE) 365.8 (13.61) 368.9 (13.27) 95% CI (338.3, 393.3) (342.0, 395.8)

ACT vs. UC Difference -3.1 95% CI (-32.5, 26.3) P-value 0.822

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

214 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.22 Analysis of Mean Evening Peak Expiratory Flow Over the 24-Week Treatment Period (ITT) (Page 3 of 3)

ACT UC Visit N=242 N=265

Week 16 to Week 20 n 207 235 LS Mean (SE) 366.9 (13.60) 370.9 (13.26) 95% CI (339.4, 394.4) (344.0, 397.8)

ACT vs. UC Difference -4.0 95% CI (-33.4, 25.4) P-value 0.767

Week 20 to Week 24 n 206 229 LS Mean (SE) 368.1 (13.67) 372.0 (13.33) 95% CI (340.5, 395.7) (345.0, 398.9)

ACT vs. UC Difference -3.8 95% CI (-33.4, 25.7) P-value 0.779

Note: Analysis performed using a MMRM model with covariates of treatment, center, baseline ACT total score, type of baseline controller (ICS vs. ICS/LABA), gender, age and treatment-by-visit interaction, with the center as a random factor.

PPD

215 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.23 Summary Of the Change From Baseline in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score (ITT) (Page 1 of 3) ______ACT UC N=242 N=265

Total Score n 202 232 Mean 1.5 1.1 SD 1.06 0.95 Median 1.4 1.1 Min. -1.0 -1.3 Max. 4.4 4.4

Symptoms n 202 232 Mean 1.7 1.2 SD 1.08 1.02 Median 1.5 1.2 Min. -0.8 -1.2 Max. 4.6 4.4 Note: This analysis only included subjects who completed the study and change from baseline was calculated as the difference between baseline visit and Visit 6. Only include those who complete the study and exclude those who withdraw the study.

PPD

216 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.23 Summary Of the Change From Baseline in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score (ITT) (Page 2 of 3) ______ACT UC N=242 N=265

Activity Limitation n 202 231 Mean 1.4 1.0 SD 1.11 0.98 Median 1.3 0.9 Min. -1.2 -1.7 Max. 4.5 4.1

Emotional Function n 202 232 Mean 1.7 1.3 SD 1.30 1.13 Median 1.6 1.2 Min. -1.2 -1.2 Max. 5.0 5.2 Note: This analysis only included subjects who completed the study and change from baseline was calculated as the difference between baseline visit and Visit 6. Only include those who complete the study and exclude those who withdraw the study.

PPD

217 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.23 Summary Of the Change From Baseline in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score (ITT) (Page 3 of 3) ______ACT UC N=242 N=265

Environmental Stimuli n 203 232 Mean 1.5 1.1 SD 1.53 1.23 Median 1.5 1.0 Min. -2.3 -1.5 Max. 6.0 5.3 Note: This analysis only included subjects who completed the study and change from baseline was calculated as the difference between baseline visit and Visit 6. Only include those who complete the study and exclude those who withdraw the study.

PPD

218 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 20NOV2019

Table 2.24 Analysis Of the Change From Baseline in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score (ITT) (Page 1 of 3)

ACT UC N=242 N=265

Total Score

n 202 231 LS Mean (SE) 1.5 (0.13) 1.2 (0.13) 95% CI (1.3, 1.8) (1.0, 1.5)

ACT vs. UC Difference 0.3 95% CI (-0.0, 0.6) P-value 0.059

Symptoms

n 202 231 LS Mean (SE) 1.6 (0.12) 1.3 (0.12) 95% CI (1.4, 1.9) (1.1, 1.6)

ACT vs. UC Difference 0.3 95% CI (0.0, 0.6) P-value 0.030

Note: Analysis performed using a mixed model with covariates of treatment, center, baseline AQLQ score, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the center as a random factor. Only include those who complete the study and exclude those who withdraw the study.

PPD

219 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 20NOV2019

Table 2.24 Analysis Of the Change From Baseline in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score (ITT) (Page 2 of 3)

ACT UC N=242 N=265

Activity Limitation

n 202 230 LS Mean (SE) 1.3 (0.14) 1.1 (0.13) 95% CI (1.1, 1.6) (0.8, 1.3)

ACT vs. UC Difference 0.3 95% CI (-0.0, 0.6) P-value 0.078

Emotional Function

n 202 231 LS Mean (SE) 1.7 (0.14) 1.4 (0.14) 95% CI (1.4, 2.0) (1.1, 1.7)

ACT vs. UC Difference 0.3 95% CI (0.0, 0.6) P-value 0.042

Note: Analysis performed using a mixed model with covariates of treatment, center, baseline AQLQ score, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the center as a random factor. Only include those who complete the study and exclude those who withdraw the study.

PPD

220 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 20NOV2019

Table 2.24 Analysis Of the Change From Baseline in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score (ITT) (Page 3 of 3)

ACT UC N=242 N=265

Environmental Stimuli

n 203 231 LS Mean (SE) 1.5 (0.20) 1.2 (0.20) 95% CI (1.1, 1.9) (0.8, 1.6)

ACT vs. UC Difference 0.3 95% CI (-0.2, 0.8) P-value 0.234

Note: Analysis performed using a mixed model with covariates of treatment, center, baseline AQLQ score, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the center as a random factor. Only include those who complete the study and exclude those who withdraw the study.

PPD

221 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.25 Time to First ACT Total Score >=20 or Improvement of More Than 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT) (Page 1 of 1)

ACT UC N=242 N=265

Nmber of Subjects With Event (%) 238 ( 98.3%) 249 ( 94.0%) Number of Subjects Censored (%) 3 ( 1.2%) 15 ( 5.7%) KM Estimate of Time to Event (Days) Median (50% of the Quantile) [1] 31.0 30.0 95% Confidence Interval [30.0, 31.0] [30.0, 32.0] 25% Quantile (Lower Quartile), 75% Quantile 29.0, 35.0 29.0, 57.0 (Upper Quartile)

ACT vs. UC Hazard Ratio [2] 1.333 95% CI (0.844, 2.107) P-value 0.218

[1] Kaplan-Meier estimates. [2] Hazard ratos, 95% CI and P-value are from a Cox proportional hazards model with treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age as covariates.

PPD

222 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 14OCT2019

Table 2.26 Time to First ACT Total Score >=20 and Improvement of More Than 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT) (Page 1 of 1)

ACT UC N=242 N=265

Nmber of Subjects With Event (%) 229 ( 94.6%) 224 ( 84.5%) Number of Subjects Censored (%) 12 ( 5.0%) 40 ( 15.1%) KM Estimate of Time to Event (Days) Median (50% of the Quantile) [1] 33.0 55.0 95% Confidence Interval [31.0, 35.0] [36.0, 58.0] 25% Quantile (Lower Quartile), 75% Quantile 29.0, 57.0 29.0, 98.0 (Upper Quartile)

ACT vs. UC Hazard Ratio [2] 1.843 95% CI (1.160, 2.926) P-value 0.010

[1] Kaplan-Meier estimates. [2] Hazard ratos, 95% CI and P-value are from a Cox proportional hazards model with treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age as covariates.

PPD

223 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.27 Summary of Moderate/Severe Asthma Exacerbations (ITT) (Page 1 of 2) ______ACT UC N=242 N=265 Number of Subjects Who Experienced 8 ( 3.3%) 10 ( 3.8%) Moderate/Severe Asthma Exacerbations During the Study Period (%)

Total Number of Moderate/Severe Asthma Exacerbations per Subject (%) 0 234 ( 96.7%) 255 ( 96.2%) 1 7 ( 2.9%) 10 ( 3.8%) 2 0 0 >2 1 ( 0.4%) 0

Number of Moderate/Severe Asthma 10 10 Exacerbations Outcome Cured 8 ( 80.0%) 10 (100.0%) Unhealed 2 ( 20.0%) 0 Death 0 0

Measures Clinic or Emergency Room Visit 7 ( 70.0%) 8 ( 80.0%) Self-managed 0 0 Hospital Admission 3 ( 30.0%) 2 ( 20.0%) Others 0 0

PPD

224 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 2.27 Summary of Moderate/Severe Asthma Exacerbations (ITT) (Page 2 of 2) ______ACT UC N=242 N=265 Severity Moderate 6 ( 60.0%) 8 ( 80.0%) Severe 4 ( 40.0%) 2 ( 20.0%)

Restrictions on Subsequent Daily Activities Affecting Heavy Physical Activity 3 ( 30.0%) 4 ( 40.0%) Affecting Moderate Physical Activity 2 ( 20.0%) 3 ( 30.0%) Affecting Social Activities 1 ( 10.0%) 1 ( 10.0%) Affecting Daily Work 4 ( 40.0%) 2 ( 20.0%) Affecting Sleeping 0 0

Duration of Exacerbation (Days) n 8 10 Mean 12.3 6.8 SD 15.72 4.98 Median 7.5 6.0 Min. 3 1 Max. 50 14

PPD

225 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 14OCT2019

Table 2.28 Analysis of Moderate/Severe Asthma Exacerbations (ITT) (Page 1 of 1)

ACT UC N=242 N=265 n 241 263 Mean Annual Exacerbation Rate 0.10 0.09 95% CI (0.04, 0.25) (0.04, 0.22)

ACT vs. UC Rate Ratio 1.09 95% CI (0.32, 3.73) P-value 0.897

Percent Reduction in Annual Rate (%) [1] -0.09 95% CI (-2.73, 0.68)

Note: Analysis performed using a generalised linear model assuming a negative binominal distribution and covariates of treatment,baseline ACT total score, type of baseline controller (ICS alone vs. ICS/LABA), gender and age. Note: n = the number of the events per treatment group. [1] Percent reduction in annual exacerbation rate was calculated as (1-rate ratio)×100%.

PPD

226 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 20NOV2019

Table 2.29 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 Or An Improvement >=3 Points in ACT Total Score in At Least One Post-baseline Assessment During the 24-Weeks Treatment Period (ITT) (Page 1 of 1)

ACT UC N=242 N=265 n 240 262 Responder (%) [1] 238 ( 99.2%) 249 ( 95.0%) Non-Responder (%) 2 ( 0.8%) 13 ( 5.0%)

ACT vs. UC [2] Adjusted Odds Ratio 6.89 95% CI (1.28, 37.02) P-value 0.025

Interaction between treatment and covariates, P-value [3] With type of baseline controller 0.982 With gender 0.974 With age 0.912 [1] Subjects who have an ACT total score >= 20 or an improvement >=3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor. [3] The interaction between treatment and each covariate was studied sequentially, and significant effects were considered at a 10% significance level.

PPD

227 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 20NOV2019

Table 2.30 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >=3 Points in ACT Total Score (ITT) (Page 1 of 2)

ACT UC Visit (N=242) (N=265)

Week 12 n 241 263 Responder (%) [1] 200( 83.0%) 183( 69.6%) Non-Responder (%) 41( 17.0%) 80( 30.4%)

ACT vs. UC [4] Adjusted Odds Ratio 2.36 95% CI (0.94, 5.96) P-value 0.065

Week 24 n 241 263 Responder (%) [2] 204( 84.6%) 192( 73.0%) Non-Responder (%) 37( 15.4%) 71( 27.0%)

ACT vs. UC [4] Adjusted Odds Ratio 2.08 [1] Subjects who have an ACT total score >= 20 and an improvement >=3 points in ACT total score at Week 12. Subjects with missing data will be set as non-responders. [2] Subjects who have an ACT total score >= 20 and an improvement >=3 points in ACT total score at Week 24. Subjects with missing data will be set as non-responders. [3] Subjects who have an ACT total score >= 20 and an improvement >=3 points in ACT total score at Weeks 16 and 20 and 24. Subjects with missing data will be set as non-responders. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

PPD

228 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 20NOV2019

Table 2.30 Summary of the Statistical Analysis of Percentage of Subjects Who Have An ACT Total Score >= 20 And An Improvement >=3 Points in ACT Total Score (ITT) (Page 2 of 2)

ACT UC Visit (N=242) (N=265)

Week 24 95% CI (0.84, 5.18) P-value 0.103

Weeks 16 and 20 and 24 n 241 263 Responder (%) [3] 183( 75.9%) 147( 55.9%) Non-Responder (%) 58( 24.1%) 116( 44.1%)

ACT vs. UC [4] Adjusted Odds Ratio 2.65 95% CI (1.22, 5.77) P-value 0.020 [1] Subjects who have an ACT total score >= 20 and an improvement >=3 points in ACT total score at Week 12. Subjects with missing data will be set as non-responders. [2] Subjects who have an ACT total score >= 20 and an improvement >=3 points in ACT total score at Week 24. Subjects with missing data will be set as non-responders. [3] Subjects who have an ACT total score >= 20 and an improvement >=3 points in ACT total score at Weeks 16 and 20 and 24. Subjects with missing data will be set as non-responders. [4] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor.

PPD

229 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 20NOV2019

Table 2.31 Time to First ACT Total Score >=20 or Improvement >= 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT) (Page 1 of 1)

ACT UC N=242 N=265

Nmber of Subjects With Event (%) 238 ( 98.3%) 251 ( 94.7%) Number of Subjects Censored (%) 3 ( 1.2%) 13 ( 4.9%) KM Estimate of Time to Event (Days) Median (50% of the Quantile) [1] 30.0 30.0 95% Confidence Interval [30.0, 31.0] [30.0, 31.0] 25% Quantile (Lower Quartile), 75% Quantile 29.0, 34.0 29.0, 56.0 (Upper Quartile)

ACT vs. UC Hazard Ratio [2] 1.475 95% CI (0.933, 2.332) P-value 0.097

[1] Kaplan-Meier estimates. [2] Hazard ratos, 95% CI and P-value are from a Cox proportional hazards model with treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age as covariates.

PPD

230 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version: 1.0 Protocol: 201097 Date: 20NOV2019

Table 2.32 Time to First ACT Total Score >=20 and Improvement >= 3 Points in ACT Total Score Over the 24-week Treatment Period (ITT) (Page 1 of 1)

ACT UC N=242 N=265

Nmber of Subjects With Event (%) 235 ( 97.1%) 236 ( 89.1%) Number of Subjects Censored (%) 6 ( 2.5%) 28 ( 10.6%) KM Estimate of Time to Event (Days) Median (50% of the Quantile) [1] 31.0 35.0 95% Confidence Interval [31.0, 33.0] [31.0, 40.0] 25% Quantile (Lower Quartile), 75% Quantile 29.0, 53.0 29.0, 64.0 (Upper Quartile)

ACT vs. UC Hazard Ratio [2] 1.708 95% CI (1.078, 2.707) P-value 0.023

[1] Kaplan-Meier estimates. [2] Hazard ratos, 95% CI and P-value are from a Cox proportional hazards model with treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age as covariates.

PPD

231 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 26SEP2019

Table 3.01 Summary of Adverse Event (Safety Population) (Page 1 of 1) ______ACT UC Total N=241 N=265 N=506

AEs 86 ( 35.7%) 129 ( 48.7%) 215 ( 42.5%) Any AEs related to study process 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Any AEs related to GSK's drug 0 1 ( 0.4%) 1 ( 0.2%) Non-serious AEs 82 ( 34.0%) 123 ( 46.4%) 205 ( 40.5%) SAEs 9 ( 3.7%) 13 ( 4.9%) 22 ( 4.3%) SAEs related to study process 0 0 0 SAEs related to GSK's drugs 0 0 0 Any AEs leading to temporarily suspend\delay GSK's 0 1 ( 0.4%) 1 ( 0.2%) drugs Any AEs leading to withdrawal 5 ( 2.1%) 3 ( 1.1%) 8 ( 1.6%) Asthma exacerbations 8 ( 3.3%) 9 ( 3.4%) 17 ( 3.4%) Death 1 ( 0.4%) 0 1 ( 0.2%)

PPD

232 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 1 of 15) ______ACT UC Total N=241 N=265 N=506

Number of subjects with at least one AE 86 ( 35.7%) 129 ( 48.7%) 215 ( 42.5%)

Infections and infestations 44 ( 18.3%) 72 ( 27.2%) 116 ( 22.9%) Upper respiratory tract infection 18 ( 7.5%) 47 ( 17.7%) 65 ( 12.8%) Rhinitis 6 ( 2.5%) 3 ( 1.1%) 9 ( 1.8%) Nasopharyngitis 4 ( 1.7%) 4 ( 1.5%) 8 ( 1.6%) Bronchitis 3 ( 1.2%) 16 ( 6.0%) 19 ( 3.8%) Conjunctivitis 3 ( 1.2%) 3 ( 1.1%) 6 ( 1.2%) Respiratory tract infection 3 ( 1.2%) 6 ( 2.3%) 9 ( 1.8%) Urinary tract infection 3 ( 1.2%) 0 3 ( 0.6%) Gastroenteritis 2 ( 0.8%) 0 2 ( 0.4%) Otitis media 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Pelvic inflammatory disease 2 ( 0.8%) 0 2 ( 0.4%) Vaginal infection 2 ( 0.8%) 0 2 ( 0.4%)

PPD

233 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 2 of 15) ______ACT UC Total N=241 N=265 N=506

Chronic sinusitis 1 ( 0.4%) 0 1 ( 0.2%) Herpes zoster 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Otitis media chronic 1 ( 0.4%) 0 1 ( 0.2%) Pharyngitis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Pulpitis dental 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Salpingitis 1 ( 0.4%) 0 1 ( 0.2%) Sinusitis 1 ( 0.4%) 0 1 ( 0.2%) Tinea cruris 1 ( 0.4%) 0 1 ( 0.2%) Tinea pedis 1 ( 0.4%) 0 1 ( 0.2%) Conjunctivitis viral 0 1 ( 0.4%) 1 ( 0.2%) Lung infection 0 1 ( 0.4%) 1 ( 0.2%) Otitis externa 0 1 ( 0.4%) 1 ( 0.2%) Periodontitis 0 2 ( 0.8%) 2 ( 0.4%)

PPD

234 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 3 of 15) ______ACT UC Total N=241 N=265 N=506

Tonsillitis 0 1 ( 0.4%) 1 ( 0.2%) Tooth abscess 0 1 ( 0.4%) 1 ( 0.2%) Urethritis 0 1 ( 0.4%) 1 ( 0.2%)

Respiratory, thoracic and mediastinal 32 ( 13.3%) 30 ( 11.3%) 62 ( 12.3%) disorders Cough 10 ( 4.1%) 8 ( 3.0%) 18 ( 3.6%) Oropharyngeal pain 6 ( 2.5%) 9 ( 3.4%) 15 ( 3.0%) Throat irritation 5 ( 2.1%) 1 ( 0.4%) 6 ( 1.2%) Asthma 3 ( 1.2%) 3 ( 1.1%) 6 ( 1.2%) Rhinitis allergic 3 ( 1.2%) 3 ( 1.1%) 6 ( 1.2%) Rhinorrhoea 3 ( 1.2%) 0 3 ( 0.6%) Nasal obstruction 2 ( 0.8%) 0 2 ( 0.4%) Pneumonitis 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Bronchiectasis 1 ( 0.4%) 0 1 ( 0.2%)

PPD

235 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 4 of 15) ______ACT UC Total N=241 N=265 N=506

Dysphonia 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Nasal polyps 1 ( 0.4%) 0 1 ( 0.2%) Oropharyngeal discomfort 1 ( 0.4%) 0 1 ( 0.2%) Pharyngeal oedema 1 ( 0.4%) 0 1 ( 0.2%) Productive cough 1 ( 0.4%) 4 ( 1.5%) 5 ( 1.0%) Tachypnoea 1 ( 0.4%) 0 1 ( 0.2%) Bronchitis chronic 0 1 ( 0.4%) 1 ( 0.2%) Dyspnoea 0 1 ( 0.4%) 1 ( 0.2%) Laryngeal pain 0 1 ( 0.4%) 1 ( 0.2%) Sneezing 0 1 ( 0.4%) 1 ( 0.2%) Sputum increased 0 1 ( 0.4%) 1 ( 0.2%)

Gastrointestinal disorders 14 ( 5.8%) 12 ( 4.5%) 26 ( 5.1%) Abdominal pain 3 ( 1.2%) 0 3 ( 0.6%)

PPD

236 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 5 of 15) ______ACT UC Total N=241 N=265 N=506

Diarrhoea 3 ( 1.2%) 0 3 ( 0.6%) Dyspepsia 2 ( 0.8%) 0 2 ( 0.4%) Gastritis 2 ( 0.8%) 0 2 ( 0.4%) Gastrooesophageal reflux disease 2 ( 0.8%) 0 2 ( 0.4%) Chronic gastritis 1 ( 0.4%) 0 1 ( 0.2%) Dysbacteriosis 1 ( 0.4%) 0 1 ( 0.2%) Enteritis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Gastrointestinal hypomotility 1 ( 0.4%) 0 1 ( 0.2%) Gingival pain 1 ( 0.4%) 0 1 ( 0.2%) Gingival swelling 1 ( 0.4%) 0 1 ( 0.2%) Haemorrhoids 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Large intestine polyp 1 ( 0.4%) 0 1 ( 0.2%) Toothache 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%)

PPD

237 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 6 of 15) ______ACT UC Total N=241 N=265 N=506

Upper gastrointestinal haemorrhage 1 ( 0.4%) 0 1 ( 0.2%) Vomiting 1 ( 0.4%) 0 1 ( 0.2%) Abdominal discomfort 0 1 ( 0.4%) 1 ( 0.2%) Breath odour 0 1 ( 0.4%) 1 ( 0.2%) Flatulence 0 2 ( 0.8%) 2 ( 0.4%) Gastrointestinal disorder 0 1 ( 0.4%) 1 ( 0.2%) Haematochezia 0 1 ( 0.4%) 1 ( 0.2%) Hypertrophic anal papilla 0 1 ( 0.4%) 1 ( 0.2%) Mouth ulceration 0 1 ( 0.4%) 1 ( 0.2%) Nausea 0 1 ( 0.4%) 1 ( 0.2%) Stomatitis 0 1 ( 0.4%) 1 ( 0.2%) Tooth impacted 0 1 ( 0.4%) 1 ( 0.2%)

PPD

238 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 7 of 15) ______ACT UC Total N=241 N=265 N=506

General disorders and administration site 10 ( 4.1%) 10 ( 3.8%) 20 ( 4.0%) conditions Pyrexia 5 ( 2.1%) 4 ( 1.5%) 9 ( 1.8%) Chest discomfort 4 ( 1.7%) 1 ( 0.4%) 5 ( 1.0%) Chest pain 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) Discomfort 0 1 ( 0.4%) 1 ( 0.2%) Fatigue 0 1 ( 0.4%) 1 ( 0.2%)

Skin and subcutaneous tissue disorders 6 ( 2.5%) 11 ( 4.2%) 17 ( 3.4%) Acne 1 ( 0.4%) 0 1 ( 0.2%) Dermatitis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Dermatitis allergic 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Drug eruption 1 ( 0.4%) 0 1 ( 0.2%) Eczema 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Urticaria 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%) Hyperhidrosis 0 1 ( 0.4%) 1 ( 0.2%)

PPD

239 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 8 of 15) ______ACT UC Total N=241 N=265 N=506

Rash 0 4 ( 1.5%) 4 ( 0.8%)

Reproductive system and breast disorders 5 ( 2.1%) 2 ( 0.8%) 7 ( 1.4%) Balanoposthitis 1 ( 0.4%) 0 1 ( 0.2%) Dysmenorrhoea 1 ( 0.4%) 0 1 ( 0.2%) Endometrial thickening 1 ( 0.4%) 0 1 ( 0.2%) Menstrual disorder 1 ( 0.4%) 0 1 ( 0.2%) Pelvic adhesions 1 ( 0.4%) 0 1 ( 0.2%) Pelvic fluid collection 1 ( 0.4%) 0 1 ( 0.2%) Uterine polyp 1 ( 0.4%) 0 1 ( 0.2%) Breast hyperplasia 0 2 ( 0.8%) 2 ( 0.4%)

Injury, poisoning and procedural 4 ( 1.7%) 7 ( 2.6%) 11 ( 2.2%) complications Foot fracture 1 ( 0.4%) 0 1 ( 0.2%) Injury 1 ( 0.4%) 0 1 ( 0.2%)

PPD

240 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 9 of 15) ______ACT UC Total N=241 N=265 N=506

Meniscus injury 1 ( 0.4%) 0 1 ( 0.2%) Rib fracture 1 ( 0.4%) 0 1 ( 0.2%) Humerus fracture 0 1 ( 0.4%) 1 ( 0.2%) Ligament sprain 0 1 ( 0.4%) 1 ( 0.2%) Limb injury 0 1 ( 0.4%) 1 ( 0.2%) Muscle injury 0 1 ( 0.4%) 1 ( 0.2%) Skin injury 0 1 ( 0.4%) 1 ( 0.2%) Thoracic vertebral fracture 0 1 ( 0.4%) 1 ( 0.2%) Tooth injury 0 1 ( 0.4%) 1 ( 0.2%)

Musculoskeletal and connective tissue 3 ( 1.2%) 10 ( 3.8%) 13 ( 2.6%) disorders Osteoarthritis 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Tenosynovitis 1 ( 0.4%) 0 1 ( 0.2%) Back pain 0 1 ( 0.4%) 1 ( 0.2%)

PPD

241 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 10 of 15) ______ACT UC Total N=241 N=265 N=506

Fistula 0 1 ( 0.4%) 1 ( 0.2%) Intervertebral disc protrusion 0 2 ( 0.8%) 2 ( 0.4%) Lumbar spinal stenosis 0 1 ( 0.4%) 1 ( 0.2%) Myofascitis 0 1 ( 0.4%) 1 ( 0.2%) Neck pain 0 1 ( 0.4%) 1 ( 0.2%) Periarthritis 0 1 ( 0.4%) 1 ( 0.2%) Soft tissue disorder 0 1 ( 0.4%) 1 ( 0.2%) Spinal osteoarthritis 0 2 ( 0.8%) 2 ( 0.4%)

Nervous system disorders 3 ( 1.2%) 9 ( 3.4%) 12 ( 2.4%) Headache 2 ( 0.8%) 4 ( 1.5%) 6 ( 1.2%) Dizziness 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Epilepsy 1 ( 0.4%) 0 1 ( 0.2%) Cerebral arteriosclerosis 0 1 ( 0.4%) 1 ( 0.2%)

PPD

242 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 11 of 15) ______ACT UC Total N=241 N=265 N=506

Poor quality sleep 0 1 ( 0.4%) 1 ( 0.2%) Vascular headache 0 1 ( 0.4%) 1 ( 0.2%)

Cardiac disorders 2 ( 0.8%) 5 ( 1.9%) 7 ( 1.4%) Arrhythmia 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Palpitations 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Ventricular extrasystoles 1 ( 0.4%) 0 1 ( 0.2%) Angina pectoris 0 1 ( 0.4%) 1 ( 0.2%) Angina unstable 0 1 ( 0.4%) 1 ( 0.2%) Coronary artery disease 0 2 ( 0.8%) 2 ( 0.4%)

Endocrine disorders 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Adrenal mass 1 ( 0.4%) 0 1 ( 0.2%) Thyroid mass 1 ( 0.4%) 0 1 ( 0.2%) Hyperthyroidism 0 1 ( 0.4%) 1 ( 0.2%)

PPD

243 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 12 of 15) ______ACT UC Total N=241 N=265 N=506

Eye disorders 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) Ocular hyperaemia 1 ( 0.4%) 0 1 ( 0.2%) Scleritis 1 ( 0.4%) 0 1 ( 0.2%) Conjunctival haemorrhage 0 1 ( 0.4%) 1 ( 0.2%) Macular oedema 0 1 ( 0.4%) 1 ( 0.2%) Vitreous opacities 0 1 ( 0.4%) 1 ( 0.2%)

Neoplasms benign, malignant and 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) unspecified (incl cysts and polyps) Uterine leiomyoma 2 ( 0.8%) 0 2 ( 0.4%) Gastric cancer 0 1 ( 0.4%) 1 ( 0.2%) Melanocytic naevus 0 1 ( 0.4%) 1 ( 0.2%) Skin papilloma 0 1 ( 0.4%) 1 ( 0.2%)

Renal and urinary disorders 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) Nephrolithiasis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%)

PPD

244 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 13 of 15) ______ACT UC Total N=241 N=265 N=506

Urate nephropathy 1 ( 0.4%) 0 1 ( 0.2%) Glomerulonephritis chronic 0 1 ( 0.4%) 1 ( 0.2%) Proteinuria 0 1 ( 0.4%) 1 ( 0.2%) Ureterolithiasis 0 1 ( 0.4%) 1 ( 0.2%)

Blood and lymphatic system disorders 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Lymphadenitis 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%)

Ear and labyrinth disorders 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Deafness neurosensory 1 ( 0.4%) 0 1 ( 0.2%) Tinnitus 0 1 ( 0.4%) 1 ( 0.2%) Vertigo 0 1 ( 0.4%) 1 ( 0.2%)

Hepatobiliary disorders 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Hepatic steatosis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%)

PPD

245 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 14 of 15) ______ACT UC Total N=241 N=265 N=506

Immune system disorders 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%) Hypersensitivity 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%)

Metabolism and nutrition disorders 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Vitamin B1 deficiency 1 ( 0.4%) 0 1 ( 0.2%) Hyperlipidaemia 0 1 ( 0.4%) 1 ( 0.2%) Hyperuricaemia 0 1 ( 0.4%) 1 ( 0.2%)

Surgical and medical procedures 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Tooth extraction 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%)

Vascular disorders 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Varicose vein 1 ( 0.4%) 0 1 ( 0.2%) Hypertension 0 1 ( 0.4%) 1 ( 0.2%) Subclavian artery occlusion 0 1 ( 0.4%) 1 ( 0.2%)

Congenital, familial and genetic disorders 0 1 ( 0.4%) 1 ( 0.2%) Branchial cleft sinus 0 1 ( 0.4%) 1 ( 0.2%)

PPD

246 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.02 Adverse Events of the Subjects (Safety Population) (Page 15 of 15) ______ACT UC Total N=241 N=265 N=506

Psychiatric disorders 0 2 ( 0.8%) 2 ( 0.4%) Insomnia 0 1 ( 0.4%) 1 ( 0.2%) Sleep disorder 0 1 ( 0.4%) 1 ( 0.2%)

PPD

247 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.03 Adverse Events Related to Study Process of the Subjects (Safety Population) (Page 1 of 1) ______ACT UC Total N=241 N=265 N=506

Number of subjects with at least one AE 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%)

Respiratory, thoracic and mediastinal 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) disorders Asthma 1 ( 0.4%) 0 1 ( 0.2%) Cough 1 ( 0.4%) 0 1 ( 0.2%) Throat irritation 1 ( 0.4%) 0 1 ( 0.2%) Oropharyngeal pain 0 1 ( 0.4%) 1 ( 0.2%)

PPD

248 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.04 Adverse Events Related To GSK's Drugs of the Subjects (Safety Population) (Page 1 of 1) ______ACT UC Total N=241 N=265 N=506

Number of subjects with at least one AE 0 1 ( 0.4%) 1 ( 0.2%)

Skin and subcutaneous tissue disorders 0 1 ( 0.4%) 1 ( 0.2%) Rash 0 1 ( 0.4%) 1 ( 0.2%)

PPD

249 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.05 Serious Adverse Events of the Subjects (Safety Population) (Page 1 of 4) ______ACT UC Total N=241 N=265 N=506

Number of subjects with at least one AE 9 ( 3.7%) 13 ( 4.9%) 22 ( 4.3%)

Respiratory, thoracic and mediastinal 3 ( 1.2%) 1 ( 0.4%) 4 ( 0.8%) disorders Asthma 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Nasal polyps 1 ( 0.4%) 0 1 ( 0.2%)

Infections and infestations 2 ( 0.8%) 0 2 ( 0.4%) Chronic sinusitis 1 ( 0.4%) 0 1 ( 0.2%) Otitis media chronic 1 ( 0.4%) 0 1 ( 0.2%) Salpingitis 1 ( 0.4%) 0 1 ( 0.2%)

Neoplasms benign, malignant and 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) unspecified (incl cysts and polyps) Uterine leiomyoma 2 ( 0.8%) 0 2 ( 0.4%) Gastric cancer 0 1 ( 0.4%) 1 ( 0.2%) Melanocytic naevus 0 1 ( 0.4%) 1 ( 0.2%)

PPD

250 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.05 Serious Adverse Events of the Subjects (Safety Population) (Page 2 of 4) ______ACT UC Total N=241 N=265 N=506

Reproductive system and breast disorders 2 ( 0.8%) 0 2 ( 0.4%) Pelvic adhesions 1 ( 0.4%) 0 1 ( 0.2%) Uterine polyp 1 ( 0.4%) 0 1 ( 0.2%)

Cardiac disorders 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Arrhythmia 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Ventricular extrasystoles 1 ( 0.4%) 0 1 ( 0.2%) Angina unstable 0 1 ( 0.4%) 1 ( 0.2%) Coronary artery disease 0 2 ( 0.8%) 2 ( 0.4%)

Gastrointestinal disorders 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Large intestine polyp 1 ( 0.4%) 0 1 ( 0.2%) Upper gastrointestinal haemorrhage 1 ( 0.4%) 0 1 ( 0.2%) Haematochezia 0 1 ( 0.4%) 1 ( 0.2%) Haemorrhoids 0 1 ( 0.4%) 1 ( 0.2%) Hypertrophic anal papilla 0 1 ( 0.4%) 1 ( 0.2%)

PPD

251 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.05 Serious Adverse Events of the Subjects (Safety Population) (Page 3 of 4) ______ACT UC Total N=241 N=265 N=506

Injury, poisoning and procedural 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) complications Meniscus injury 1 ( 0.4%) 0 1 ( 0.2%) Humerus fracture 0 1 ( 0.4%) 1 ( 0.2%)

Musculoskeletal and connective tissue 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%) disorders Osteoarthritis 1 ( 0.4%) 0 1 ( 0.2%) Fistula 0 1 ( 0.4%) 1 ( 0.2%) Intervertebral disc protrusion 0 2 ( 0.8%) 2 ( 0.4%) Lumbar spinal stenosis 0 1 ( 0.4%) 1 ( 0.2%)

Nervous system disorders 1 ( 0.4%) 0 1 ( 0.2%) Epilepsy 1 ( 0.4%) 0 1 ( 0.2%)

Vascular disorders 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Varicose vein 1 ( 0.4%) 0 1 ( 0.2%) Subclavian artery occlusion 0 1 ( 0.4%) 1 ( 0.2%)

PPD

252 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.05 Serious Adverse Events of the Subjects (Safety Population) (Page 4 of 4) ______ACT UC Total N=241 N=265 N=506

Congenital, familial and genetic disorders 0 1 ( 0.4%) 1 ( 0.2%) Branchial cleft sinus 0 1 ( 0.4%) 1 ( 0.2%)

Eye disorders 0 1 ( 0.4%) 1 ( 0.2%) Macular oedema 0 1 ( 0.4%) 1 ( 0.2%)

General disorders and administration site 0 1 ( 0.4%) 1 ( 0.2%) conditions Chest pain 0 1 ( 0.4%) 1 ( 0.2%)

Renal and urinary disorders 0 1 ( 0.4%) 1 ( 0.2%) Nephrolithiasis 0 1 ( 0.4%) 1 ( 0.2%) Ureterolithiasis 0 1 ( 0.4%) 1 ( 0.2%)

PPD

253 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.06 Serious Adverse Events Related to Study Process of the Subjects (Safety Population) (Page 1 of 1) ______ACT UC Total N=241 N=265 N=506

Number of subjects with at least one AE 0 0 0

PPD

254 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.07 Serious Adverse Events Related To GSK's Drugs of the Subjects (Safety Population) (Page 1 of 1) ______ACT UC Total N=241 N=265 N=506

Number of subjects with at least one AE 0 0 0

PPD

255 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 1 of 13) ______ACT UC Total N=241 N=265 N=506

Number of subjects with at least one AE 82 ( 34.0%) 123 ( 46.4%) 205 ( 40.5%)

Infections and infestations 43 ( 17.8%) 72 ( 27.2%) 115 ( 22.7%) Upper respiratory tract infection 18 ( 7.5%) 47 ( 17.7%) 65 ( 12.8%) Rhinitis 6 ( 2.5%) 3 ( 1.1%) 9 ( 1.8%) Nasopharyngitis 4 ( 1.7%) 4 ( 1.5%) 8 ( 1.6%) Bronchitis 3 ( 1.2%) 16 ( 6.0%) 19 ( 3.8%) Conjunctivitis 3 ( 1.2%) 3 ( 1.1%) 6 ( 1.2%) Respiratory tract infection 3 ( 1.2%) 6 ( 2.3%) 9 ( 1.8%) Urinary tract infection 3 ( 1.2%) 0 3 ( 0.6%) Gastroenteritis 2 ( 0.8%) 0 2 ( 0.4%) Otitis media 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Pelvic inflammatory disease 2 ( 0.8%) 0 2 ( 0.4%) Vaginal infection 2 ( 0.8%) 0 2 ( 0.4%)

PPD

256 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 2 of 13) ______ACT UC Total N=241 N=265 N=506

Herpes zoster 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Pharyngitis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Pulpitis dental 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Sinusitis 1 ( 0.4%) 0 1 ( 0.2%) Tinea cruris 1 ( 0.4%) 0 1 ( 0.2%) Tinea pedis 1 ( 0.4%) 0 1 ( 0.2%) Conjunctivitis viral 0 1 ( 0.4%) 1 ( 0.2%) Lung infection 0 1 ( 0.4%) 1 ( 0.2%) Otitis externa 0 1 ( 0.4%) 1 ( 0.2%) Periodontitis 0 2 ( 0.8%) 2 ( 0.4%) Tonsillitis 0 1 ( 0.4%) 1 ( 0.2%) Tooth abscess 0 1 ( 0.4%) 1 ( 0.2%) Urethritis 0 1 ( 0.4%) 1 ( 0.2%)

PPD

257 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 3 of 13) ______ACT UC Total N=241 N=265 N=506

Respiratory, thoracic and mediastinal 30 ( 12.4%) 29 ( 10.9%) 59 ( 11.7%) disorders Cough 10 ( 4.1%) 8 ( 3.0%) 18 ( 3.6%) Oropharyngeal pain 6 ( 2.5%) 9 ( 3.4%) 15 ( 3.0%) Throat irritation 5 ( 2.1%) 1 ( 0.4%) 6 ( 1.2%) Rhinitis allergic 3 ( 1.2%) 3 ( 1.1%) 6 ( 1.2%) Rhinorrhoea 3 ( 1.2%) 0 3 ( 0.6%) Nasal obstruction 2 ( 0.8%) 0 2 ( 0.4%) Pneumonitis 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Asthma 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Bronchiectasis 1 ( 0.4%) 0 1 ( 0.2%) Dysphonia 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Oropharyngeal discomfort 1 ( 0.4%) 0 1 ( 0.2%) Pharyngeal oedema 1 ( 0.4%) 0 1 ( 0.2%)

PPD

258 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 4 of 13) ______ACT UC Total N=241 N=265 N=506

Productive cough 1 ( 0.4%) 4 ( 1.5%) 5 ( 1.0%) Tachypnoea 1 ( 0.4%) 0 1 ( 0.2%) Bronchitis chronic 0 1 ( 0.4%) 1 ( 0.2%) Dyspnoea 0 1 ( 0.4%) 1 ( 0.2%) Laryngeal pain 0 1 ( 0.4%) 1 ( 0.2%) Sneezing 0 1 ( 0.4%) 1 ( 0.2%) Sputum increased 0 1 ( 0.4%) 1 ( 0.2%)

Gastrointestinal disorders 14 ( 5.8%) 11 ( 4.2%) 25 ( 4.9%) Abdominal pain 3 ( 1.2%) 0 3 ( 0.6%) Diarrhoea 3 ( 1.2%) 0 3 ( 0.6%) Dyspepsia 2 ( 0.8%) 0 2 ( 0.4%) Gastritis 2 ( 0.8%) 0 2 ( 0.4%) Gastrooesophageal reflux disease 2 ( 0.8%) 0 2 ( 0.4%)

PPD

259 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 5 of 13) ______ACT UC Total N=241 N=265 N=506

Chronic gastritis 1 ( 0.4%) 0 1 ( 0.2%) Dysbacteriosis 1 ( 0.4%) 0 1 ( 0.2%) Enteritis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Gastrointestinal hypomotility 1 ( 0.4%) 0 1 ( 0.2%) Gingival pain 1 ( 0.4%) 0 1 ( 0.2%) Gingival swelling 1 ( 0.4%) 0 1 ( 0.2%) Haemorrhoids 1 ( 0.4%) 0 1 ( 0.2%) Toothache 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Vomiting 1 ( 0.4%) 0 1 ( 0.2%) Abdominal discomfort 0 1 ( 0.4%) 1 ( 0.2%) Breath odour 0 1 ( 0.4%) 1 ( 0.2%) Flatulence 0 2 ( 0.8%) 2 ( 0.4%) Gastrointestinal disorder 0 1 ( 0.4%) 1 ( 0.2%)

PPD

260 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 6 of 13) ______ACT UC Total N=241 N=265 N=506

Mouth ulceration 0 1 ( 0.4%) 1 ( 0.2%) Nausea 0 1 ( 0.4%) 1 ( 0.2%) Stomatitis 0 1 ( 0.4%) 1 ( 0.2%) Tooth impacted 0 1 ( 0.4%) 1 ( 0.2%)

General disorders and administration site 10 ( 4.1%) 9 ( 3.4%) 19 ( 3.8%) conditions Pyrexia 5 ( 2.1%) 4 ( 1.5%) 9 ( 1.8%) Chest discomfort 4 ( 1.7%) 1 ( 0.4%) 5 ( 1.0%) Chest pain 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) Discomfort 0 1 ( 0.4%) 1 ( 0.2%) Fatigue 0 1 ( 0.4%) 1 ( 0.2%)

Skin and subcutaneous tissue disorders 6 ( 2.5%) 11 ( 4.2%) 17 ( 3.4%) Acne 1 ( 0.4%) 0 1 ( 0.2%) Dermatitis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%)

PPD

261 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 7 of 13) ______ACT UC Total N=241 N=265 N=506

Dermatitis allergic 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Drug eruption 1 ( 0.4%) 0 1 ( 0.2%) Eczema 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Urticaria 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%) Hyperhidrosis 0 1 ( 0.4%) 1 ( 0.2%) Rash 0 4 ( 1.5%) 4 ( 0.8%)

Injury, poisoning and procedural 3 ( 1.2%) 6 ( 2.3%) 9 ( 1.8%) complications Foot fracture 1 ( 0.4%) 0 1 ( 0.2%) Injury 1 ( 0.4%) 0 1 ( 0.2%) Rib fracture 1 ( 0.4%) 0 1 ( 0.2%) Ligament sprain 0 1 ( 0.4%) 1 ( 0.2%) Limb injury 0 1 ( 0.4%) 1 ( 0.2%) Muscle injury 0 1 ( 0.4%) 1 ( 0.2%)

PPD

262 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 8 of 13) ______ACT UC Total N=241 N=265 N=506

Skin injury 0 1 ( 0.4%) 1 ( 0.2%) Thoracic vertebral fracture 0 1 ( 0.4%) 1 ( 0.2%) Tooth injury 0 1 ( 0.4%) 1 ( 0.2%)

Reproductive system and breast disorders 3 ( 1.2%) 2 ( 0.8%) 5 ( 1.0%) Balanoposthitis 1 ( 0.4%) 0 1 ( 0.2%) Dysmenorrhoea 1 ( 0.4%) 0 1 ( 0.2%) Endometrial thickening 1 ( 0.4%) 0 1 ( 0.2%) Menstrual disorder 1 ( 0.4%) 0 1 ( 0.2%) Pelvic fluid collection 1 ( 0.4%) 0 1 ( 0.2%) Breast hyperplasia 0 2 ( 0.8%) 2 ( 0.4%)

Cardiac disorders 2 ( 0.8%) 3 ( 1.1%) 5 ( 1.0%) Palpitations 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Ventricular extrasystoles 1 ( 0.4%) 0 1 ( 0.2%)

PPD

263 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 9 of 13) ______ACT UC Total N=241 N=265 N=506

Angina pectoris 0 1 ( 0.4%) 1 ( 0.2%)

Endocrine disorders 2 ( 0.8%) 1 ( 0.4%) 3 ( 0.6%) Adrenal mass 1 ( 0.4%) 0 1 ( 0.2%) Thyroid mass 1 ( 0.4%) 0 1 ( 0.2%) Hyperthyroidism 0 1 ( 0.4%) 1 ( 0.2%)

Eye disorders 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) Ocular hyperaemia 1 ( 0.4%) 0 1 ( 0.2%) Scleritis 1 ( 0.4%) 0 1 ( 0.2%) Conjunctival haemorrhage 0 1 ( 0.4%) 1 ( 0.2%) Vitreous opacities 0 1 ( 0.4%) 1 ( 0.2%)

Musculoskeletal and connective tissue 2 ( 0.8%) 8 ( 3.0%) 10 ( 2.0%) disorders Osteoarthritis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Tenosynovitis 1 ( 0.4%) 0 1 ( 0.2%)

PPD

264 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 10 of 13) ______ACT UC Total N=241 N=265 N=506

Back pain 0 1 ( 0.4%) 1 ( 0.2%) Myofascitis 0 1 ( 0.4%) 1 ( 0.2%) Neck pain 0 1 ( 0.4%) 1 ( 0.2%) Periarthritis 0 1 ( 0.4%) 1 ( 0.2%) Soft tissue disorder 0 1 ( 0.4%) 1 ( 0.2%) Spinal osteoarthritis 0 2 ( 0.8%) 2 ( 0.4%)

Nervous system disorders 2 ( 0.8%) 9 ( 3.4%) 11 ( 2.2%) Headache 2 ( 0.8%) 4 ( 1.5%) 6 ( 1.2%) Dizziness 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Cerebral arteriosclerosis 0 1 ( 0.4%) 1 ( 0.2%) Poor quality sleep 0 1 ( 0.4%) 1 ( 0.2%) Vascular headache 0 1 ( 0.4%) 1 ( 0.2%)

PPD

265 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 11 of 13) ______ACT UC Total N=241 N=265 N=506

Renal and urinary disorders 2 ( 0.8%) 2 ( 0.8%) 4 ( 0.8%) Nephrolithiasis 1 ( 0.4%) 0 1 ( 0.2%) Urate nephropathy 1 ( 0.4%) 0 1 ( 0.2%) Glomerulonephritis chronic 0 1 ( 0.4%) 1 ( 0.2%) Proteinuria 0 1 ( 0.4%) 1 ( 0.2%)

Blood and lymphatic system disorders 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Lymphadenitis 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%)

Ear and labyrinth disorders 1 ( 0.4%) 2 ( 0.8%) 3 ( 0.6%) Deafness neurosensory 1 ( 0.4%) 0 1 ( 0.2%) Tinnitus 0 1 ( 0.4%) 1 ( 0.2%) Vertigo 0 1 ( 0.4%) 1 ( 0.2%)

Hepatobiliary disorders 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Hepatic steatosis 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%)

PPD

266 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 12 of 13) ______ACT UC Total N=241 N=265 N=506

Immune system disorders 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%) Hypersensitivity 1 ( 0.4%) 3 ( 1.1%) 4 ( 0.8%)

Metabolism and nutrition disorders 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Vitamin B1 deficiency 1 ( 0.4%) 0 1 ( 0.2%) Hyperlipidaemia 0 1 ( 0.4%) 1 ( 0.2%) Hyperuricaemia 0 1 ( 0.4%) 1 ( 0.2%)

Surgical and medical procedures 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) Tooth extraction 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%)

Neoplasms benign, malignant and 0 1 ( 0.4%) 1 ( 0.2%) unspecified (incl cysts and polyps) Skin papilloma 0 1 ( 0.4%) 1 ( 0.2%)

Psychiatric disorders 0 2 ( 0.8%) 2 ( 0.4%) Insomnia 0 1 ( 0.4%) 1 ( 0.2%) Sleep disorder 0 1 ( 0.4%) 1 ( 0.2%)

PPD

267 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.08 Non-Serious Adverse Events of the Subjects (Safety Population) (Page 13 of 13) ______ACT UC Total N=241 N=265 N=506

Vascular disorders 0 1 ( 0.4%) 1 ( 0.2%) Hypertension 0 1 ( 0.4%) 1 ( 0.2%)

PPD

268 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.09 Adverse Events Leading To Temporarily Suspend\Delay GSK's drugs (Safety Population) (Page 1 of 1) ______ACT UC Total N=241 N=265 N=506

Number of subjects with at least one AE 0 1 ( 0.4%) 1 ( 0.2%)

Skin and subcutaneous tissue disorders 0 1 ( 0.4%) 1 ( 0.2%) Urticaria 0 1 ( 0.4%) 1 ( 0.2%)

PPD

269 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 Date： 29SEP2019

Table 3.10 Adverse Events Leading To Withdrawal (Safety Population) (Page 1 of 1) ______ACT UC Total N=241 N=265 N=506

Number of subjects with at least one AE 5 ( 2.1%) 3 ( 1.1%) 8 ( 1.6%)

Respiratory, thoracic and mediastinal 3 ( 1.2%) 2 ( 0.8%) 5 ( 1.0%) disorders Asthma 3 ( 1.2%) 1 ( 0.4%) 4 ( 0.8%) Oropharyngeal pain 0 1 ( 0.4%) 1 ( 0.2%)

Cardiac disorders 1 ( 0.4%) 0 1 ( 0.2%) Arrhythmia 1 ( 0.4%) 0 1 ( 0.2%) Ventricular extrasystoles 1 ( 0.4%) 0 1 ( 0.2%)

Nervous system disorders 1 ( 0.4%) 0 1 ( 0.2%) Epilepsy 1 ( 0.4%) 0 1 ( 0.2%)

Skin and subcutaneous tissue disorders 1 ( 0.4%) 0 1 ( 0.2%) Dermatitis allergic 1 ( 0.4%) 0 1 ( 0.2%)

Injury, poisoning and procedural 0 1 ( 0.4%) 1 ( 0.2%) complications Thoracic vertebral fracture 0 1 ( 0.4%) 1 ( 0.2%)

PPD

270 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 日期： 30SEP2019

Table 3.11 Physical Examination of the subjects (Safety Population) (Page 1 of 3) ______ACT UC Total N=241 N=265 N=506 Baseline Lung N 241 264 505 Normal 223 ( 92.5%) 260 ( 98.5%) 483 ( 95.6%) Abnormal, no clinical 6 ( 2.5%) 1 ( 0.4%) 7 ( 1.4%) significance Abnormal, clinical 12 ( 5.0%) 3 ( 1.1%) 15 ( 3.0%) significance

Cardiovascular System N 241 264 505 Normal 241 (100.0%) 261 ( 98.9%) 502 ( 99.4%) Abnormal, no clinical 0 1 ( 0.4%) 1 ( 0.2%) significance Abnormal, clinical 0 2 ( 0.8%) 2 ( 0.4%) significance

Nose N 241 264 505 Normal 230 ( 95.4%) 257 ( 97.3%) 487 ( 96.4%) Abnormal, no clinical 10 ( 4.1%) 3 ( 1.1%) 13 ( 2.6%) significance Abnormal, clinical 1 ( 0.4%) 4 ( 1.5%) 5 ( 1.0%) significance

PPD

271 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 日期： 30SEP2019

Table 3.11 Physical Examination of the subjects (Safety Population) (Page 2 of 3) ______ACT UC Total N=241 N=265 N=506 Week 4 Change or new abnormity since last visit N 236 262 498 No 235 ( 99.6%) 260 ( 99.2%) 495 ( 99.4%) Yes, abnormal clinical 0 1 ( 0.4%) 1 ( 0.2%) significance Yes, abnormal no clinical 1 ( 0.4%) 1 ( 0.4%) 2 ( 0.4%) significance

Week 8 Change or new abnormity since last visit N 226 249 475 No 225 ( 99.6%) 247 ( 99.2%) 472 ( 99.4%) Yes, abnormal clinical 1 ( 0.4%) 0 1 ( 0.2%) significance Yes, abnormal no clinical 0 2 ( 0.8%) 2 ( 0.4%) significance

Week 12 Change or new abnormity since last visit N 215 245 460 No 214 ( 99.5%) 244 ( 99.6%) 458 ( 99.6%) Yes, abnormal clinical 0 1 ( 0.4%) 1 ( 0.2%) significance Yes, abnormal no clinical 1 ( 0.5%) 0 1 ( 0.2%) significance

PPD

272 CONFIDENTIAL 2019N423508_00 201097

GlaxoSmithKline group of companies Version：1.0 Protocol：201097 日期： 30SEP2019

Table 3.11 Physical Examination of the subjects (Safety Population) (Page 3 of 3) ______ACT UC Total N=241 N=265 N=506 Week 16 Change or new abnormity since last visit N 215 240 455 No 214 ( 99.5%) 236 ( 98.3%) 450 ( 98.9%) Yes, abnormal clinical 0 4 ( 1.7%) 4 ( 0.9%) significance Yes, abnormal no clinical 1 ( 0.5%) 0 1 ( 0.2%) significance

Week 20 Change or new abnormity since last visit N 205 235 440 No 204 ( 99.5%) 228 ( 97.0%) 432 ( 98.2%) Yes, abnormal clinical 0 7 ( 3.0%) 7 ( 1.6%) significance Yes, abnormal no clinical 1 ( 0.5%) 0 1 ( 0.2%) significance

Week 24 / Withdrawal Change or new abnormity since last visit N 223 247 470 No 221 ( 99.1%) 240 ( 97.2%) 461 ( 98.1%) Yes, abnormal clinical 2 ( 0.9%) 4 ( 1.6%) 6 ( 1.3%) significance Yes, abnormal no clinical 0 3 ( 1.2%) 3 ( 0.6%) significance

PPD

273 CONFIDENTIAL 2019N423508_00 201097

Synopsis

Name of company: GlaxoSmithKline Research & Development Limited

Name of finished product: Not available Name of active substance: None at the time of this report

Study Number: 201097

Title: The effectiveness of Asthma Control Test guided treatment compared with usual care in China adult asthma patients

Investigator(s): 12 investigators in China

Study center(s): 12 centers in China

Publication(s): None at the time of this report

Study Period: 26Aug2016- 09Aug2019

Phase of Development: IV

Objectives:

The primary objective of the study was:

• To compare the effectiveness of ACT guided treatment versus usual care in achieving asthma control The secondary objective of the study was:

• To compare the effectiveness of ACT guided treatment versus usual care in asthma subjects with respect to the following parameters:

• Symptoms

• Forced Expiratory Volume in one second (FEV1)

• Peak Expiratory Flow (PEF)

• The quality of life

• Time to asthma control The other objectives of the study were:

• To compare the effectiveness of ACT guided treatment versus usual care in asthma exacerbation

   CONFIDENTIAL 2019N423508_00 201097

Methodology:

This was a prospective, multicentre, cluster-randomised, open-label study. The study was a 24-week study which subsequent scheduled visits occurred as follows: Week 0 (Visit 0), Week 4 (Visit 1), Week 8 (Visit 2), Week 12 (Visit 3), Week 16 (Visit 4), Week 20 (Visit 5) and Week 24 (Visit 6). In this cluster-randomization design, each study centre, considered as a cluster, was randomized to either ACT guided treatment group or control group (usual care group).

Number of subjects:

A total of 583 subjects were screened, 530 subjects were randomized, and 443 subjects completed the study (209 ACT, 234 UC).

Diagnosis and main criteria for inclusion:

For the study centres (cluster) who were eligible for enrolment in the study, were general practice, located in Shanghai and provided asthma care.

For subjects who were eligible for enrolment in the study, their legally authorized representative had to provide signed written informed consent, were male or female, must be ≥18 to ≤70 years of age and had an ACT score <20 at Visit 0, be documented clinical history of asthma for at least 6 months and using inhaled corticosteroids (ICS) alone or combined with an inhaled long-acting β2 agonist (LABA) treatment within 1 year prior to Visit 0, and at Visit 0, a demonstrable reversible increased in FEV1 of at least 12% (and ≥200 mL), 15 minutes after inhaling a short-acting bronchodilator or; at any time in the last 2 years documentary evidence of a reversible increase in FEV1 of at least 12% (and ≥200 mL) 15 mins after inhaling a short-acting bronchodilator; or demonstrable reversible increase in morning PEF of at least 15% (and ≥200 mL) either spontaneously or after inhalation of a short-acting bronchodilator; subjects must have been able to read, comprehend, and record information in Chinese.

Treatment administration:

For the subjects who were recruited in the ACT centres, they were treated based on the ACT score. For subjects who were recruited in the control centres, they were treated based on doctor’s subjective judgment.

Criteria for evaluation:

Primary Endpoint

• The percentage of subjects who had an ACT total score ≥20 or an improvement of more than 3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period

   CONFIDENTIAL 2019N423508_00 201097

Secondary Endpoints

• Mean daytime symptom score over the 24-week treatment period

• Mean night-time symptom score over the 24-week treatment period

• Mean change from baseline to the end of study in FEV1

• Mean morning(AM) PEF over the 24-week treatment period

• Mean evening(PM) PEF over the 24-week treatment period

• Mean change from baseline to the end of study in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score

• Time to first ACT total score ≥20 or improvement of more than 3 points in ACT over the 24-week treatment period

• Time to first ACT total score ≥20 and improvement of more than 3 points in ACT total score over the 24-week treatment period

Exploratory endpoints

• Rate of moderate/severe asthma exacerbation over the 24-week treatment period Statistical methods:

The null hypothesis for this study was that there was no difference between ACT guided treatment and usual care in the percentage of subjects who had an ACT total score ≥20 or an improvement of more than 3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period for the intention-to-treat (ITT) population.

The alternative hypothesis for this study was that there was a difference between ACT guided treatment and usual care in the percentage of subjects who had an ACT total score ≥20 or an improvement of more than 3 points in ACT total score in at least one post- baseline assessment during the 24-week treatment period for the ITT population.

The study was designed to show superiority of ACT guided treatment over usual care.

The primary efficacy endpoint was statistically analyzed using mixed effect logistic regression model (generalized linear mixed model, GLMM) which included terms for treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age. Centre, as the unit of randomization, was treated as a random effect.

The continuous secondary efficacy endpoints were statistically analyzed using analysis of covariance (ANCOVA) and adjusted for centre as a random effect.

   CONFIDENTIAL 2019N423508_00 201097

The time to event secondary efficacy endpoints were statistically analyzed using Cox regression model which was performed with treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age as the adjusted factors.

The exploratory endpoint, i.e., moderate/severe exacerbation rate was analyzed using a generalized linear model, assuming the number of exacerbations had a negative binomial probability distribution and that its mean was related to covariate factors with a ‘log link’ function.

Summary:

Study Population and Demographics

A total of 583 subjects were screened for this study; and 53 subjects were screen failure, and 530 subjects were screening succeed and randomized in which 443 (83.6%) subjects completed the study with 209 (82.9%) subjects in the ACT group and 234 subjects (84.2%) in the UC group . The primary reasons for subject withdrawal were subject’s decision (26 subjects, 29.9%), significantly non-compliant (23 subjects, 26.4%) and experienced the protocol-defined asthma exacerbation (16 subjects, 18.4%).

A total of 507 subjects had at least one post-baseline assessment on primary endpoints, second endpoints, and/or exploratory endpoints, and comprised the ITT Population. 439 subjects in the ITT population had no major protocol deviations and ≥80% treatment compliance and diary compliance, and comprised Per-Protocol (PP) population. 506 subjects comprised safety population who were enrolled into the study and had at least one DRC assessment.

Demographic characteristics for the ITT Population were all similar between the two treatment groups.

Demographic ACT UC Total N=242 N=265 N=507 Age (years) Mean (SD) 48.0 (12.81) 48.0 (13.43) 48.0 (13.13) Min-Max 18 - 73 18 - 70 18 - 73 Gender, n (%) Female 115 (47.5%) 135 (50.9%) 250 (49.3%) Male 127 (52.5%) 130 (49.1%) 257 (50.7%) Race, n (%) Han nationality 237 (97.9%) 261 (98.5%) 498 (98.2%) Non-Han nationality 5 (2.1%) 4 (1.5%) 9 (1.8%) Height (cm) Mean (SD) 165.24 (7.820) 165.60 (8.238) 165.43 (8.035) Min-Max 145.0 - 186.0 145.0 - 187.0 145.0 - 187.0 Weight (kg) Mean (SD) 65.89 (11.074) 64.90 (11.694) 65.37 (11.402)

   CONFIDENTIAL 2019N423508_00 201097

Demographic ACT UC Total N=242 N=265 N=507 Min-Max 33.0 - 105.0 41.0 - 100.0 33.0 - 105.0 BMI Mean (SD) 24.06 (3.246) 23.56 (3.141) 23.80 (3.199) Min-Max 15.7 - 36.3 16.0 - 34.3 15.7 - 36.3 Primary Efficacy Results

The primary efficacy endpoint for this study was the percentage of subjects who had an ACT total score ≥20 or an improvement of more than 3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period.

Compared with usual care, the odds ratio of well controlled asthma with ACT guided treatment was higher than 1 for China adult asthma patients, and the difference of treatments was statistically significant (P-value=0.027) at the 5% level of significance, but the interaction effect were all not statistically significant (P-values>0.1) at the 10% level of significance.

ACT UC N=242 N=265 n 240 262 Responder, n (%) [1] 238 (99.2%) 247 (94.3%) Non-Responder, n (%) 2 (0.8%) 15 (5.7%)

ACT vs. UC [2] Adjusted Odds Ratio 7.87 95% CI (1.29, 48.11) P-value 0.027

Interaction between treatment and covariates, P-value [3] With type of baseline controller 0.982 With gender 0.975 With age 0.891 [1] Subjects who had an ACT total score ≥ 20 or an improvement >3 points in ACT total score in at least one post-baseline assessment during the 24-Weeks treatment period. [2] Generalized linear mixed model (GLMM) was used for statistical analysis. The model included treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor. [3] The interaction between treatment and each covariate was studied sequentially, and significant effects were considered at a 10% significance level.

Secondary Efficacy Results

For mean daytime symptom score over the 24-week treatment period, at each 4-week treatment interval, the difference between ACT guided treatment usual care wasn’t statistically significant (P-value of Week0-Week4 = 0.222; P-value of Week4-Week8 = 0.156; P-value of Week8-Week12 = 0.245; P-value of Week12-Week16 = 0.057; P-value of Week16-Week20 = 0.151; P-value of Week20-Week24 = 0.114).

   CONFIDENTIAL 2019N423508_00 201097

For mean night-time symptom score over the 24-week treatment period, at each 4-week treatment interval, the difference between ACT guided treatment usual care wasn’t statistically significant (P-value of Week0-Week4 = 0.487; P-value of Week4-Week8 = 0.546; P-value of Week8-Week12 = 0.314; P-value of Week12-Week16 = 0.197; P-value of Week16-Week20 = 0.282; P-value of Week20-Week24 = 0.543).

For mean change from baseline to the end of study in FEV1, the difference between ACT guided treatment usual care wasn’t statistically significant (P-value of FEV1 (L) = 0.273; P-value of FEV1 (%) = 0.243).

For mean morning (AM) PEF over the 24-week treatment period, at each 4-week treatment interval, the difference between ACT guided treatment usual care wasn’t statistically significant (P-value of Week0-Week4 = 0.367; P-value of Week4-Week8 = 0.587; P-value of Week8-Week12 = 0.721; P-value of Week12-Week16 = 0.801; P-value of Week16-Week20 = 0.704; P-value of Week20-Week24 = 0.762).

For mean evening (PM) PEF over the 24-week treatment period, at each 4-week treatment interval, the difference between ACT guided treatment usual care wasn’t statistically significant (P-value of Week0-Week4 = 0.354; P-value of Week4-Week8 = 0.538; P-value of Week8-Week12 = 0.716; P-value of Week12-Week16 = 0.822; P-value of Week16-Week20 = 0.767; P-value of Week20-Week24 = 0.779).

For mean change from baseline to the end of study in total AQLQ(S) score, the difference between ACT guided treatment usual care wasn’t statistically significant (P-values = 0.059).

For the time to first ACT total score ≥20 or an Improvement >3 Points in ACT total score over the 24-week treatment period, the difference between ACT guided treatment usual care wasn’t statistically significant (P-value = 0.218).

For the time to first ACT total score ≥20 and an Improvement >3 Points in ACT total score over the 24-week treatment period, the difference between ACT guided treatment usual care was statistically significant (P-value = 0.010).

Exploratory Results

For moderate/severe asthma exacerbations, the difference between ACT guided treatment usual care wasn’t statistically significant (P-value = 0.897).

Safety Results

Overall, ACT guided treatment were well tolerated in which the incidences of all AEs and asthma exacerbations were all lower than that with usual care. None of subjects with ACT guided treatment experienced any SAEs related to study process or GSK’s drug. None of subjects with ACT guided treatment experienced any AEs leading to temporarily suspend \ delay GSK's drugs. Only cough was reported with incidence ≥3% and more common than usual care.

   CONFIDENTIAL 2019N423508_00 201097

The incidence of AEs related to study process or GSK’s drugs was low and similar between ACT guided treatment and usual care.

The AEs related to study process reported in ACT group were Asthma (0.4%) and Cough (0.4%) and Throat irritation (0.4%), and those reported in UC group was Oropharyngeal pain (0.4%).

The AEs related to GSK’s drugs was only reported in UC group which was Rash (0.4%).

Although there was reported death for one subject with ACT guided treatment, the event wasn’t related to study process or GSK’s drug.

Number (%) of Subjects ACT UC Total N=241 N=265 N=506 At least one AE 86 (35.7%) 129 (48.7%) 215 (42.5%) Asthma exacerbations 8 (3.3%) 9 (3.4%) 17 (3.4%) AEs with Incidence Rate ≥3% in Any Treatment Group and More Common than UC group Cough 10 (4.1%) 8 (3.0%) 18 (3.6%) Adverse Events Related to Study Process or GSK’s Drugs Adverse Events Related to Study Process 2 (0.8%) 1 (0.4%) 3 (0.6%) Asthma 1 (0.4%) 0 1 (0.2%) Cough 1 (0.4%) 0 1 (0.2%) Throat irritation 1 (0.4%) 0 1 (0.2%) Oropharyngeal pain 0 1 (0.4%) 1 (0.2%) Adverse Events Related to GSK’s Drugs 0 1 (0.4%) 1 (0.2%) Rash 0 1 (0.4%) 1 (0.2%) Death 1 (0.4%) 0 1 (0.2%)

Conclusions:

This study has demonstrated the superiority of ACT guided treatment over usual care for the ACT total score ≥20 or an improvement of more than 3 points in ACT total score for China adult asthma patients and ACT guided treatment was well tolerated in this study with no safety concerns identified.

Effective Date: 08-JUN-2021

   2013N186640_00 CONFIDENTIAL 2019N423508_00 GlaxoSmithKline group of companies 201097

TITLE PAGE

Division: Worldwide Development

Information Type: Clinical Protocol

Title: The effectiveness of Asthma Control Test guided treatment compared with usual care in China adult asthma patients

Compound Number: None

Development Phase IV

Effective Date: 27-JUN-2016

Author(s): PPD

Copyright 2016 the GlaxoSmithKline group of companies. All rights reserved. Unauthorised copying or use of this information is prohibited.

1 2019N423508_00

PPD

PPD

PPD 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

MEDICAL MONITOR/SPONSOR INFORMATION PAGE

Medical Monitor/SAE Contact Information:

Role Name Day Time Phone After-hours Fax Site Address Number and email Phone/Cell/ Number address Pager Number

Primary PPD 5F The Medical Headquarters Monitor Building No 168, Middle Tibet Road Shanghai, China Secondary 5F The Medical Headquarters Monitor Building No 168, Middle Tibet Road Shanghai, China SAE contact NA NA NA PPD Building 1, 917 information Halei Road Zhangjiang Hi- tech Park, Pudong Shanghai 201203, China

Sponsor Legal Registered Address:

GlaxoSmithKline(China) Investment Co., Ltd. 9F Tower A, Ocean International Center No.56 Mid 4th East Ring Road, Chaoyang District Beijing, 100025, China Telephone: PPD

In some countries, the clinical trial sponsor may be the local GlaxoSmithKline Affiliate Company (or designee). If applicable, the details of the alternative Sponsor and contact person in the territory will be provided to the relevant regulatory authority as part of the clinical trial submission.

3 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

INVESTIGATOR PROTOCOL AGREEMENT PAGE

 I confirm agreement to conduct the study in compliance with the protocol.  I acknowledge that I am responsible for overall study conduct. I agree to personally conduct or supervise the described study.  I agree to ensure that all associates, colleagues and employees assisting in the conduct of the study are informed about their obligations. Mechanisms are in place to ensure that site staff receives the appropriate information throughout the study.

Investigator Name:

Investigator Address:

Investigator Phone Number:

Investigator Signature Date

4 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

TABLE OF CONTENTS

PAGE

1. PROTOCOL SYNOPSIS FOR STUDY 201097...... 7

2. INTRODUCTION...... 11 2.1. Study Rationale ...... 11 2.2. Brief Background ...... 11

3. OBJECTIVE(S) AND ENDPOINT(S) ...... 12

4. STUDY DESIGN ...... 13 4.1. Overall Design ...... 13 4.2. Treatment Arms and Duration...... 14 4.3. Type and Number of Subjects...... 15 4.4. Design Justification...... 15 4.5. Benefit:Risk Assessment ...... 15 4.5.1. Risk Assessment ...... 16 4.5.2. Benefit Assessment ...... 16 4.5.3. Overall Benefit: Risk Conclusion...... 16

5. SELECTION OF STUDY POPULATION AND WITHDRAWAL CRITERIA ...... 16 5.1. Inclusion Criteria...... 16 5.2. Exclusion Criteria...... 17 5.3. Withdrawal/Stopping Criteria...... 19 5.4. Subject and Study Completion...... 21

6. STUDY TREATMENT ...... 21 6.1. Treatment Assignment...... 21 6.2. Blinding...... 21 6.3. Treatment after the End of the Study ...... 21

7. STUDY ASSESSMENTS AND PROCEDURES ...... 22 7.1. Time and Events Table...... 22 7.2. Screening and Critical Baseline Assessments ...... 23 7.3. Efficacy...... 23 7.3.1. Primary Endpoint (s) ...... 23 7.3.2. Secondary Endpoint (s) ...... 24 7.3.3. Other Endpoint (s) ...... 26 7.3.4. Guidance for Questionnaires Administration ...... 28 7.4. Safety ...... 29 7.4.1. Adverse Events (AE) and Serious Adverse Events (SAEs)...... 29 7.4.1.1. Time period and Frequency for collecting AE and SAE information...... 29 7.4.1.2. Method of Detecting AEs and SAEs ...... 30 7.4.1.3. Follow-up of AEs and SAEs...... 30 7.4.1.4. Cardiovascular and Death Events ...... 31 7.4.1.5. Regulatory Reporting Requirements for SAEs and ADRs...... 31 7.4.2. Pregnancy ...... 32 7.4.3. Physical Exams ...... 32 7.4.4. Vital Signs...... 32

5 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

8. DATA MANAGEMENT ...... 32

9. STATISTICAL CONSIDERATIONS AND DATA ANALYSES ...... 33 9.1. Hypotheses...... 33 9.2. Sample Size Considerations...... 33 9.2.1. Sample Size Assumptions ...... 33 9.2.2. Sample Size Sensitivity...... 34 9.2.3. Sample Size Re-estimation or Adjustment...... 35 9.3. Data Analysis Considerations ...... 35 9.3.1. Analysis Populations...... 35 9.3.2. Consideration on Missing Data Handling ...... 36 9.3.3. Interim Analysis ...... 36 9.4. Key Elements of Analysis Plan ...... 36 9.4.1. Primary Analyses...... 37 9.4.2. Secondary Analyses ...... 37 9.4.3. Other Analyses ...... 37

10. STUDY GOVERNANCE CONSIDERATIONS...... 38 10.1. Posting of Information on Publicly Available Clinical Trial Registers...... 38 10.2. Regulatory and Ethical Considerations, Including the Informed Consent Process ...... 38 10.3. Quality Control (Study Monitoring) ...... 39 10.4. Quality Assurance...... 40 10.5. Study and Site Closure ...... 40 10.6. Records Retention ...... 41 10.7. Provision of Study Results to Investigators, Posting of Information on Publically Available Clinical Trials Registers and Publication ...... 42

11. REFERENCES...... 43

12. APPENDICES ...... 44 12.1. Appendix 1 – Abbreviations and Trademarks...... 44 12.2. Appendix 2 : Definition of and Procedures for Recording, Evaluating, Follow-Up and Reporting of Adverse Events ...... 46 12.2.1. Definition of Adverse Events...... 46 12.2.2. Definition of Serious Adverse Events...... 47 12.2.3. Definition of Cardiovascular Events ...... 48 12.2.4. Recording of AEs and SAEs ...... 49 12.2.5. Evaluating AEs and SAEs...... 49 12.2.6. Reporting of SAEs/ADR to GSK ...... 51 12.3. Appendix 3 - Country Specific Requirements...... 52 12.4. Appendix 4: Asthma Control Test ...... 53 12.5. Appendix 5: Standardised Asthma Quality of Life Questionnaire (AQLQ(S)) ...... 54

6 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

1. PROTOCOL SYNOPSIS FOR STUDY 201097

Rationale

Current levels of asthma control in China fall short of the goals specified in international guidelines for asthma management (GINA). A nationwide multi-centre epidemiologic survey performed by China Asthma Alliance also showed that only 28.7% of the patients with asthma in China achieved complete asthma control. There is a large gap between what can be achieved with modern asthma management and what is currently being achieved. One of the main reasons for this is a lack of recognition of asthma control and the requirement for more effective treatment. Physicians can select appropriate treatment regimen only after they identify those patients with uncontrolled asthma. Simple methods aiming to improve the assessment of asthma control have been developed. The Asthma Control Test (ACT) is a validated, short, easy to use, and self-administered instrument used to assess asthma control. However, physicians do not use it, even though it was designed to assess the level of asthma control in patients and so guide physicians treatment decisions. This study is aimed at evaluating the effectiveness of ACT guided treatment compared with usual care in asthma patients in China. It is designed to assist Chinese patients and physicians improving adherence to the guidelines through the inclusion of the ACT in the patient’s asthma management plan.

7 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

Objective(s)/Endpoint(s)

Objectives Endpoints Primary  the percentage of subjects who have an  To compare the effectiveness of ACT ACT score ≥20 or an improvement of guided treatment versus usual care in more than 3 points in ACT during the 24- achieving asthma control week treatment period Secondary  To compare the effectiveness of ACT  Mean daytime symptom score over the guided treatment versus usual care in 24-week treatment period asthma subjects with respect to the  Mean night-time symptom score over the following parameters: 24-week treatment period • Symptoms  Mean change from baseline to the end of • Forced Expiratory Volume in study in FEV1 one second (FEV1)  Mean morning(AM) PEF over the 24- • Peak Expiratory Flow(PEF) week treatment period • The quality of life  Mean evening(PM) PEF over the 24- • Time to asthma control week treatment period  Mean change from baseline to the end of study in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score  Time to first ACT score ≥20 or improvement of more than 3 points in ACT over the 24-week treatment period Others  To compare the effectiveness of ACT  Rate of moderate/severe asthma guided treatment versus usual care in exacerbation over the 24-week treatment asthma exacerbation period Overall Design

This is a prospective, multicentre, cluster-randomised, open-label study.

Treatment Arms and Duration

The study is planned to be a 24-week study.

Baseline (week 0, visit 0): ACT, AQLQ(S) and lung function test will be completed as the baseline measurement in sequence.

Treatment period (week 0-week 24): a 24 week treatment period. Subsequent scheduled visits occur as follows: Week 4 (Visit 1), Week 8 (Visit 2), Week 12 (Visit 3), Week 16 (Visit 4), Week 20 (Visit 5) and Week 24 (Visit 6).

8 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

In this cluster-randomization design, each study centre, considered as a cluster, will be randomized to either ACT guided treatment group or control group (usual care group). The randomization will be stratified according to the hospital level (Tier 3 vs. Tier 2). The hospital level is determined in accordance with the uniform requirements of the government. A Tier 3 hospital is on a larger scale and provides superior medical care than a Tier 2 hospital.

For the subjects who are recruited in the ACT centres, they will be treated based on the ACT score. For subjects who are recruited in the control centres, they will be treated based on doctor’s subjective judgment.

ACT score Treatment adjustment

ACT=25, ≥3 months Step-down treatment

ACT20, <25 or ACT=25, <3 No change months

ACT≤19 Step-up treatment

Type and Number of Subjects

A total of 528 subjects are required in a total of 12 centres (6 centres per group) and around 44 asthma subjects per centre.

Analysis

The null hypothesis for this study is that there is no difference between ACT guided treatment and usual care in the percentage of subjects who have an ACT score ≥20 or an improvement of more than 3 points in ACT during the 24-week treatment period for the intention-to-treat (ITT) population.

The alternative hypothesis for this study is that there is a difference between ACT guided treatment and usual care in the percentage of subjects who have an ACT score ≥20 or an improvement of more than 3 points in ACT during the 24-week treatment period for the ITT population.

9 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

The primary endpoints will be analyzed using logistic regression. The statistical model on which the inference will be based will include terms for treatment, centre, type of baseline controller (inhaled corticosteroids [ICS] alone vs. ICS/long-acting β2 agonist [LABA]), gender and age. Centre, as the unit of randomization, will be treated as a random effect. The results of the primary analyses will be presented as point estimates, 95% confidence intervals and associated p-values for the adjusted mean differences between ACT guided treatment and usual care.

10 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

2. INTRODUCTION

2.1. Study Rationale

Current levels of asthma control in China fall short of the goals specified in international guidelines for asthma management (GINA). A nationwide multi-centre epidemiologic survey performed by China Asthma Alliance also showed that only 28.7% of the patients with asthma in China achieved complete asthma control [Nan, 2013]. There is a major gap between what can be achieved with modern asthma management and what is currently being achieved. One of the main reasons for this is a lack of recognition of asthma control and the requirement for more effective treatment. Physicians can select appropriate treatment regimen only after they identify those patients with uncontrolled asthma. Simple methods aiming to improve the assessment of asthma control have been developed. The asthma control test (ACT) is a validated, short, easy to use, and self- administered instrument used to assess asthma control [Nathan, 2004]. However, physicians do not use it, even though it was designed to assess the level of asthma control in patients and so guide physicians treatment decisions. This study is aimed to evaluate the effectiveness of ACT guided treatment compared with usual care in asthma patients in China. It is designed to assist Chinese patients and physicians improving adherence to the guidelines through the inclusion of the ACT in the patient’s asthma management plan.

2.2. Brief Background

GINA has adopted a five-step approach to control asthma, where each step represents a different treatment option with increasing efficacy [GINA, 2013]. The five-step approach provides treatment options to achieve and maintain control, with the least amount of medication. Asthma control assessment is very important to assess what is the appropriate treatment for the patient. In a questionnaire survey in Canada, primary care physicians were given questionnaires to guide them through control assessments. They were also asked to indicate whether or not they considered the patient’s asthma to be under control. It was found that physicians significantly overestimated control among their patients and were discordant with guideline classification of control as 31% of the patients were assessed as uncontrolled patients, 13% were well-controlled and 2% were totally controlled [Chapman, 2008]. In China, poor physician:patient dialogue has made the situation worse. During busy clinical practice, physicians often make management

11 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097 decisions based on their subjective perceived level of asthma control. This potentially leads to under-treatment and consequently sub-optimal asthma control.

3. OBJECTIVE(S) AND ENDPOINT(S)

Objectives Endpoints Primary  the percentage of subjects who have an  To compare the effectiveness of ACT ACT score ≥20 or an improvement of guided treatment versus usual care in more than 3 points in ACT during the 24- achieving asthma control week treatment period Secondary  To compare the effectiveness of ACT  Mean daytime symptom score over the guided treatment versus usual care in 24-week treatment period asthma subjects with respect to the  Mean night-time symptom score over the following parameters: 24-week treatment period • Symptoms  Mean change from baseline to the end of • Forced Expiratory Volume in study in FEV1 one second (FEV1)  Mean morning (AM) PEF over the 24- • Peak Expiratory Flow(PEF) week treatment period • The quality of life  Mean evening (PM) PEF over the 24- • Time to asthma control week treatment period  Mean change from baseline to the end of study in Standardised Asthma Quality of Life Questionnaire (AQLQ(S)) score  Time to first ACT score ≥20 or improvement of more than 3 points in ACT over the 24-week treatment period Others  To compare the effectiveness of ACT  Rate of moderate/severe asthma guided treatment versus usual care in exacerbation over the 24-week treatment asthma exacerbation period

12 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

4. STUDY DESIGN

4.1. Overall Design

This is a prospective, multicentre, cluster-randomized, open-label study.

Protocol waivers or exemptions are not allowed with the exception of immediate safety concerns. Therefore, adherence to the study design requirements, including those specified in the Time and Events Table, are essential and required for study conduct.

13 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

4.2. Treatment Arms and Duration

 The study is planned to be a 24-week study.

 Baseline (week 0, visit 0): Subjects eligible for inclusion criteria will sign informed consent form at Visit 0. At this visit, investigators should collect baseline data, including demographic data (gender, age, height, weight, education level and smoking history), history of asthma (disease duration, disease severity and exacerbation history), and past and concurrent medical conditions and medication (both asthma and non-asthma). ACT, AQLQ(S) and lung function test will be completed as the baseline measurement in sequence.

 Treatment period (week 0-week 24): a 24 week treatment period. Subsequent scheduled visits occur as follows: Week 4 (Visit 1), Week 8 (Visit 2), Week 12 (Visit 3), Week 16 (Visit 4), Week 20 (Visit 5) and Week 24 (Visit 6).

 In this cluster-randomization design, each study centre, considered as a cluster, will be randomized to either ACT guided treatment group or control group (usual care group). The randomization will be stratified according to the hospital level (Tier 3 vs. Tier 2). The hospital level is determined in accordance with the uniform requirements of the government. And a Tier 3 hospital is on a larger scale and provides superior medical care than a Tier 2 hospital.

 For the subjects who are recruited in the ACT centres, they will be treated based on the ACT score. For subjects who are recruited in the control centres, they will be treated based on doctor’s subjective judgment.

Table 1 ACT-guided treatment

ACT score Treatment adjustment

ACT=25, ≥3 months Step-down treatment

ACT20, <25 or ACT=25, <3 No change months

14 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

ACT score Treatment adjustment

ACT≤19 Step-up treatment

 During the treatment period, subjects should fill in the paper diary record card (DRC) every morning and evening, record their morning and evening PEF, symptom and medication. At each visit, subjects should complete the ACT at clinic. For the ACT guided group, the ACT must be completed prior to any other assessments are conducted. For the control group, subjects should complete the ACT after investigator making the treatment decision. This will ensure the investigators make treatment decision based on his/her clinical judgement, not based on ACT. For the control group, all subsequent assessments should be conducted in the same order as the ACT guided group. Investigators will be responsible for checking the subjects’ DRC, including symptom score and evaluating therapeutic compliance.

4.3. Type and Number of Subjects

A total of 528 subjects are required including a total of 12 centres (6 centres per group) and around 44 asthma subjects per centre.

4.4. Design Justification

 We design our study to address the effectiveness of the intervention if implemented in a system without other enhanced healthcare options; we therefore use a usual care control group.

 Randomization by individual subjects may result the contamination due to the same investigator looking after both intervention and control group. Randomization by clusters (i.e., study centre) will reduce such bias.

4.5. Benefit:Risk Assessment

The following section outlines the risk assessment and mitigation strategy for this protocol:

15 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

4.5.1. Risk Assessment

Potential Risk of Clinical Summary of Data/Rationale Mitigation Strategy Significance for Risk Study Procedures

Patients’ inappropriate ACT is consisted of five Once a subject is enrolled understanding on ACT questions, which each have a into the study, the questionnaire five-point rating scale and investigator is responsible for relate to the past month. It is training the subject to not an objective tool. understand and use the ACT.

4.5.2. Benefit Assessment

Usage of the ACT is beneficial for patients. The ACT as an easy and patient-friendly tool for asthma control assessment has been fully validated and used in interventional studies and routine clinical practice in many countries. The ACT empowers patients to take charge of their condition and work in partnership with their physicians, which could facilitate the achievement of guideline-defined asthma control.

4.5.3. Overall Benefit: Risk Conclusion

Taking into account the measures taken to minimize risk to subjects participating in this study, the potential risks identified in association with inappropriate understanding on ACT questionnaire are justified by the anticipated benefits that may be afforded to subjects with asthma.

5. SELECTION OF STUDY POPULATION AND WITHDRAWAL CRITERIA

Deviations from inclusion and exclusion criteria are not allowed because they can potentially jeopardize the scientific integrity of the study, regulatory acceptability or subject safety. Therefore, adherence to the criteria as specified in the protocol is essential.

5.1. Inclusion Criteria

Inclusion criteria for study centre (clusters):

A study centre (cluster) will be eligible for inclusion in this study only if all of the following criteria apply:

16 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

1. General practice;

2. Located in Shanghai;

3. Providing asthma care.

Inclusion criteria for subjects at Visit 0:

A subject will be eligible for inclusion in this study only if all of the following criteria apply:

4. Age:18 to 70 years inclusive;

5. Gender: Male or Female;

6. Documented clinical history of asthma for at least 6 months prior to Visit 0;

7. At Visit 0, a demonstrable reversible increase in FEV1 of at least 12% (and ≥200 mL), 15 minutes after inhaling a short-acting bronchodilator or; at any time in the last 2 years documentary evidence of a reversible increase in FEV1 of at least 12% (and ≥200 mL) 15 mins after inhaling a short-acting bronchodilator; or demonstrable reversible increase in morning PEF of at least 15% (and ≥200 mL) either spontaneously or after inhalation of a short-acting bronchodilator;

8. History of using inhaled corticosteroids (ICS) alone or combined with an inhaled long-acting β2 agonist (LABA) treatment within 1 year prior to Visit 0;

9. Subjects must have an ACT score <20 at Visit 0;

10. Subject must have been able to read, comprehend, and record information in Chinese;

11. A signed and dated written informed consent must be obtained from the subject prior to study participation.

5.2. Exclusion Criteria

A subject will not be eligible for inclusion in this study if any of the following criteria apply:

17 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

1. History of Life-threatening asthma: Defined for this protocol as an asthma episode that required intubation and/or was associated with hypercapnea, respiratory arrest or hypoxic seizures within the last 6 months before Visit 0.

2. Subjects having severe and unstable asthma, with ACT score <12 at Visit 0, history of repeated severe exacerbations (3/year) and/or a severe exacerbation in the previous 6 weeks before Visit 0.

3. A current evidence of clinically significant uncontrolled medical condition or disease (e.g., psychological disorders, mental deficiency, severe hepatic and renal dysfunction, malignancy);

4. Current smoker or ex-smoker with a more than 10 pack-year history of smoking;

5. Current clinically significant respiratory diseases other than asthma, (e.g lung cancer, lung fibrosis, sarcoidosis, tuberculosis, chronic obstructive pulmonary disease [COPD]);

6. History of alcohol or medication abuse;

7. History of upper or lower respiratory tract infection within 4 weeks prior to Visit 0;

8. Enrolled in an asthma clinic or outpatient service in the past 12 months that provides comprehensive asthma management;

9. Subjects with a history of adverse reaction including immediate or delayed hypersensitivity to any intranasal, inhaled, or systemic corticosteroid and LABA therapy and to components of the inhalation powder (e.g., lactose) at Visit 0. In addition, subjects with a history of severe milk protein allergy that, in the opinion of the Investigator, contraindicates the subject’s participation will also be excluded.

10. Females who are currently pregnant and lactating;

11. Subjects who have received any of the following medications in the 6 weeks preceding visit 0: oral/parenteral corticosteroids, oral β2-agonists or slow-release

18 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

bronchodilators, sodium cromoglycate or nedocromil sodium, ketotifen, anticholinergics, and anti-IgE treatment;

12. Subjects who comply poorly with asthma treatment in the opinion of the investigator/inability or unwillingness to take asthma medication (non-compliance), follow directions or unable to complete a written paper daily record card and self- rating questionnaires;

13. Concurrently participating in another clinical study in which the subject is or will be exposed to an investigational or a non-investigational medication or device, or has participated in a clinical trial and has received an investigational product within 30 days, 5 half-lives or twice the duration of the biological effect of the investigational product (whichever is longer).

14. Subjects with contraindications to any asthma medications they will be taking during the study period, or whom should be excluded on account of the special warnings and precautions within the label of the asthma medication they are to be treated with during the study period.

15. Affiliation with Investigator Site: Is an investigator, sub-investigator, study coordinator, employee of a participating investigator or study site, or immediate family member of the afore mentioned that is involved in this study.

16. Subjects who plan to move away from the geographical area where the study is being conducted during the study.

5.3. Withdrawal/Stopping Criteria

The primary reason for the subject discontinuing from the study will be recorded in the Case Record Form. The investigator must determine the primary reason for discontinuation. The end of study/early discontinuation procedures must be performed at the time of discontinuation from the clinical study.

The subject must be discontinued from the study for any of the following reasons:

19 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

1. A subject is significantly non-compliant with treatment or the requirements of the protocol;

2. Lost to follow up;

3. Subject’s decision not to participate any further (withdrawal of consent);

4. In the investigator’s opinion, it is in the subject’s best interest;

5. Subjects experience a protocol-defined asthma exacerbation;

6. The study is terminated by the Sponsor or designee;

7. Pregnancy (Subject should notify investigator immediately on becoming pregnant).

A subject may withdraw from study treatment at any time at his/her own request, or may be withdrawn at any time at the discretion of the investigator for safety, behavioural or administrative reasons.

The reason(s) for subjects not completing the study will be recorded in the Case Report Form (CRF), and that the investigator must document, if applicable, the reason (if specified by the subject) for withdrawal of consent. The end of study/early discontinuation procedures must be performed at the time of discontinuation from the clinical study.

The following actions must be taken in relation to a subject who fails to attend the clinic for a required study visit:  The site must attempt to contact the subject and re-schedule the missed visit as soon as possible.  The site must counsel the subject on the importance of maintaining the assigned visit schedule and ascertain whether or not the subject wishes to and/or should continue in the study based on previous non-compliance.  In cases where the subject is deemed ‘lost to follow up’, the investigator or designee must make every effort to regain contact with the subject (where possible, 3 telephone calls and if necessary a certified letter to the subject’s last

20 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

known mailing address or local equivalent methods). These contact attempts should be documented in the subject’s medical record.  Should the subject continue to be unreachable, only then will he/she be considered to have withdrawn from the study with a primary reason of “Lost to Follow-up”.

5.4. Subject and Study Completion

A completed subject is one who has completed all visits of the study.

The end of the study is defined as the last subject’s last visit.

6. STUDY TREATMENT

6.1. Treatment Assignment

The study is a trial with the study centre as the unit of randomization. Computer generated randomization schedule for centres will be provided by GlaxoSmithKline (GSK) or its designee. Once a randomization number has been assigned, it is not re- assigned. Each centre will be randomized to either ACT guided treatment group or control group (usual care group). The randomization will be stratified according to the Tier of the hospitals (Tier 3 vs. Tier 2). Assignment to a treatment will not occur, as no treatment intervention is provided for this study.Each centre will be asked to recruit 44 consecutive patients with asthma attending the practice. For the subjects who are recruited in the ACT centres, they will be treated based on the ACT score. For subjects who are recruited in the control centres, they will be treated based on doctor’s subjective judgment.

The subject number will be documented in the subject’s clinic notes and in the CRF.

6.2. Blinding

This will be an open-label study.

6.3. Treatment after the End of the Study

Subjects will not receive any additional treatment from GSK after completion of the study.

21 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

The investigator is responsible for ensuring that consideration has been given to the post- study care of the subject’s medical condition, whether or not GSK is providing specific post-study treatment.

7. STUDY ASSESSMENTS AND PROCEDURES

This section lists the procedures and parameters of each planned study assessment. The exact timing of each assessment is listed in the Time and Events Table Section 7.1.

7.1. Time and Events Table

ACTIVITY Visit Visit Visit Visit 3 Visit 4 Visit 5 Visit 6 0 1 2 ( or Early Withdrawal) Study Week (Visit Window） 0 4w±1 8w±1 12w±1 16w±1 20w±1 24w±1 Informed consent x Randomization/Allocation of clusters x Subject Demography x Medical history x Verification of inclusion/exclusion x criteria Efficacy Evaluation Verification of withdrawal criteria x x x x x x Issue diary record cards x x x x x x Collect/Review diary record cards x x x x x x Compliance assessment x x x x x x Asthma control Test (ACT) score x x x x x x x AQLQ(S) x x Lung function Test x x Safety Evaluation Adverse event assessment1 x x x x x x x

1. Serious AEs will be recorded from the time the consent form is signed until the follow-up visit. All AEs will be recorded from the start of study treatment until the follow-up visit.

22 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

7.2. Screening and Critical Baseline Assessments

Baseline Assessments

Demographic information Gender

ethnic origin

date of birth, height

weight

education level

smoking history

History of asthma including duration of asthma

History of asthma exacerbations in the 12 months prior to Visit 0

Concurrent medical conditions and concurrent medication

Medical history Use of corticosteroids for asthma in the 12 months prior to Visit 0

duration of inhaled corticosteroid treatment

The following demographic parameters will be captured: year of birth, sex, race and ethnicity.

Medical/medication will be assessed as related to the inclusion/exclusion criteria listed in Section 5.

7.3. Efficacy

7.3.1. Primary Endpoint (s)

 The primary endpoint for the study is the percentage of subjects who have an ACT score ≥20 or an improvement of more than 3 points in ACT during the 24-week treatment period.

Asthma control test (ACT)

ACT is a self-administered questionnaire comprising five questions [Nathan, 2004]. Each of the five items in the ACT is assessed on a five point scale and the scores are summed

23 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

to give a total score ranging from 5 to 25, with a score of ≥20 denoting ‘well-controlled asthma’, a score of 16-19 denoting ‘not well-controlled asthma’, and a score of ≤ 15 denoting ‘very poorly controlled asthma’. The recall period of the questionnaire is four weeks. ACT must be completed by the subject at each visit. For the control group, subjects should complete the ACT after investigator making the treatment decision. Differences of 3 points in mean ACT scores between 2 groups or over time in an individual patient are clinically significant [Schatz, 2009]. An improvement of more than 3 points in ACT from baseline will also serve as primary endpoint. The questionnaire will be attached separately.

7.3.2. Secondary Endpoint (s)

 Mean daytime symptom score over the 24-week treatment period

 Mean night-time symptom score over the 24-week treatment period

 Mean change from baseline to the end of study in FEV1

 Mean morning(AM) PEF over the 24-week treatment period

 Mean evening(PM) PEF over the 24-week treatment period

 Mean change from baseline to the end of study in AQLQ(S) score

 Time to first ACT score ≥20 or improvement of more than 3 points in ACT over the 24-week treatment period

Asthma symptom score

Subjects will be instructed to record daytime and night-time asthma symptom scores prior to taking the morning or evening dose of study medication or rescue medication. Any asthma-related symptoms, such as wheeze, shortness of breath, cough or chest tightness experienced during the previous 12 hours will be rated as follows. The following will be recorded on DRC. Daytime symptom score

0 = No symptoms during the day

24 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

1 = Symptoms for one short period during the day

2 = Symptoms for two or more short periods during the day

3 = Symptoms for most of the day which did not affect my daily activities

4 = Symptoms for most of the day which did affect my normal daily activities

5 = Symptoms so severe that I could not go to work or perform normal daily activities

Night-time symptom score

 0 = No symptoms during the night  1 = Symptoms causing me to wake once or wake early  2 = Symptoms causing me to wake twice or more (including waking early)  3 = Symptoms causing me to be awake for most of the night  4 = Symptoms so severe that I did not sleep at all

FEV1

FEV1 will be measured at visit 0 and 6. Subjects should refrain from using short-acting bronchodilators for at least 6 hours prior to performing FEV1 measurements. The highest of three technically acceptable measurements of FEV1 will be taken.

Peak Expiratory Flow (PEF)

At Visit 0, the subject will be given a Mini-Wright Peak Flow Meter and taught how to measure and record their PEF. At each visit, the subjects’ ability to measure their PEF will be checked and, if the investigator has any concerns, the subject will be retrained. The instructions for use will be provided with each Peak Flow Meter. PEF should be measured while subject is in the sitting position. Subjects should record on diary card the best of three PEF measurements, using a mini- Wright peak flow meter in the morning (7:00-10:00 AM) and evening (6:00-9:00 PM) before taking any asthma drug. Bronchodilator therapy (e.g. salbutamol) is withheld, where possible, for 4 hours before recording PEF. Otherwise, the use of salbutamol will be recorded on diary card. GSK will provide each centre with a supply of standard mini-Wright peak flow meters.

25 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

Standardised Asthma Quality of Life Questionnaire (AQLQ(S))

The Asthma Quality of Life Questionnaire (AQLQ) is a self-administered questionnaire consisting of 32 questions to measure the functional problem that are most troublesome to patients with asthma [Juniper, 1992]. There are 32 questions in 4 domains (symptoms, activity limitation, emotional function and environmental stimuli). Patients are asked to think about how they have been during the previous two weeks and to respond to each of the 32 questions on a 7-point scale (7 = not impaired at all - 1 = severely impaired). A change in score of greater than 0.5 can be considered clinically important [Juniper, 1994]. In the present study, the AQLQ(S) will be used with 5 standardised activities to replace the patient-specific ones in the AQLQ. These five generic activities incorporate the activities that were most frequently chosen by patients in studies in which the original AQLQ was used; Strenuous activities, Moderate activities, Social activities, Work-related activities, and Sleeping [Juniper, 1999]. At visit 0 and 6, the investigator will provide each subject with the AQLQ(S).

The questionnaire will be attached separately.

ACT, AQLQ(S) and lung function test should be completed in sequence at Visit 0 and 6.

7.3.3. Other Endpoint (s)

 Rate of moderate/severe asthma exacerbation over the 24-week treatment period

Asthma Exacerbations

Subjects who experience worsening of symptoms should:  increase relief medication usage for relief of symptoms  contact the investigator or primary physician immediately and report to the clinic as soon as possible (ideally within 24 hours)  record their symptoms, PEF and relief medication usage in their DRC, as previously instructed

26 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

 if intervention therapy is required during a treatment period, subjects may receive oral corticosteroids (40-60 mg prednisolone daily, or equivalent, for 10 days), over and above their usual asthma medication. Exacerbations will be assessed by the physician at each scheduled visit by reviewing the DRC, as well as specific questioning on adverse events (AEs).

Exacerbations will be defined based on one or more of the following characteristics: Definition of asthma exacerbation(s) Moderate A moderate asthma exacerbation is defined as a deterioration in asthma requiring treatment with an oral corticosteroid. Individual courses of oral corticosteroids are classified as separate exacerbations only if they are administered >1 week apart. Any course started within one week of finishing the previous course is considered part of the previous exacerbation. Severe A severe asthma exacerbation is defined as a deterioration in asthma which requires hospital admission.

Time/date of resolution Time/date at which the exacerbation has resolved, in the opinion of the investigator and/or subject.

Details of exacerbations must be collected in the Asthma Exacerbations page of the CRF and DRC as follows at the Visit immediately following the exacerbation: Date of onset and resolution, daily morning PEF during the exacerbation; details of management of the exacerbation (e.g. clinic or emergency room visit, self-managed, hospital admission etc.); restriction/prevention from continuing usual activities. An exacerbation resulting in hospitalization will be recorded as a serious adverse event (SAE) in the Serious Adverse Event pages of the CRF. However, it is not only these exacerbations resulting in hospitalization which require collection as SAEs, but those that meet any of the definitions of Seriousness, listed in Section 7.4.

27 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

7.3.4. Guidance for Questionnaires Administration

All questionnaires in the study are self-administered. It is recommended that a subject completes all questionnaires before the physical examination and lung function. For the ACT guided group, the ACT should be completed first followed by the AQLQ (at visits during which the AQLQ is measured) and the treatment decision. The remaining assessments should be conducted after the AQLQ has been completed. This order should be used throughout the study. For the control group, the ACT must be completed after the treatment decision has been made, but the remaining assessments should be conducted in the same order as for the ACT guided group.

It is important that in the ACT guided group, the patients are blinded to the treatment decision to prevent any potential bias when the patients complete the AQLQ.

The following guidance is applicable to all questionnaires for the present study.

Subjects should be requested to perform the questionnaire as completely and accurately as possible. Regardless of when the questionnaire is completed, the subject should be given adequate time to complete all questions. No stated or implied time limit for completing the questionnaires should be given. The subject should be given a quiet area in which to complete the questionnaire.

If a subject has difficulty in completing the questionnaire, a third party (an investigator or study personnel) can help the subject to understand and answer the questions. If the subject should request help or clarification of any question in the questionnaire, the investigator or the study personnel is to instruct the subject to reread the instructions and to give the best answer possible to each question. The investigator or study personnel will not provide the subject with any possible answer to any question.

A subject may use a black or blue pen to complete the questionnaire. If the subject wishes to change a response, the original response should be crossed out with a single line, and then dated and initiated by the subject, as applicable and the alternative response should be marked. Once the subject returns the questionnaire to the investigator or the study personnel, no changes will be allowed except for incomplete items.

28 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

After a subject has completed the questionnaire, the investigator may review the questionnaire for completeness. If any portion of the questionnaire is not complete, the subject should be given the opportunity to complete any missing items prior to the physician evaluation. The investigator or the study personnel is under no obligation to review or to validate the accuracy of the completed questionnaire.

7.4. Safety

Planned time points for all safety assessments are listed in the Time and Events Table (Section 7.1).

7.4.1. Adverse Events (AE) and Serious Adverse Events (SAEs)

The definitions of an AE or SAE can be found in Appendix 2.

The investigator and their designees are responsible for detecting, documenting and reporting events that meet the definition of an AE or SAE.

7.4.1.1. Time period and Frequency for collecting AE and SAE information

 The investigator or site staff is responsible for detecting, documenting and reporting events that meet the definition of an AE or SAE.

 Serious AEs will be recorded from the time the consent form is signed until the follow-up contact. All AEs will be recorded from the start of study treatment until the follow-up contact.

 Medical occurrences that begin prior to the start of study treatment but after obtaining informed consent may be recorded on the Medical History/Current Medical Conditions section of the CRF.

 Any SAEs assessed as related to study participation (e.g., protocol-mandated procedures, invasive tests, or change in existing therapy) or related to a GSK product will be recorded from the time a subject consents to participate in the study up to and including any follow-up contact.

 All SAEs will be recorded and reported to GSK within 24 hours, as indicated in Appendix 2.

29 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

 Once the investigator determines that an AE is identified as explicitly attributed to any GSK product (including products not covered in the specific study objective), the adverse drug reactions (ADR) page should be completed and reported to GSK China within 5 calendar days. Any follow-up information on a previously reported ADR will also be reported to GSK China within 5 calendar days.

 Investigators are not obligated to actively seek AEs or SAEs in former study subjects. However, if the investigator learns of any SAE, including a death, at any time after a subject has been discharged from the study, and he/she considers the event reasonably related to the study treatment or study participation, the investigator must promptly notify GSK.

NOTE: The method of recording, evaluating and assessing causality of AEs and SAEs plus procedures for completing and transmitting SAE reports to GSK are provided in Appendix 2.

7.4.1.2. Method of Detecting AEs and SAEs

Care will be taken not to introduce bias when detecting AEs and/or SAEs. Open-ended and non-leading verbal questioning of the subject is the preferred method to inquire about AE occurrence. Appropriate questions include:

 “How are you feeling?”

 “Have you had any (other) medical problems since your last visit/contact?”

 “Have you taken any new medicines, other than those provided in this study, since your last visit/contact?”

7.4.1.3. Follow-up of AEs and SAEs

After the initial AE/SAE report, the investigator is required to proactively follow each subject at subsequent visits/contacts. All SAE will be followed until resolution, until the condition stabilizes, until the event is otherwise explained, or until the subject is lost to follow-up (as defined in Section 5.3).

30 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

7.4.1.4. Cardiovascular and Death Events

For any cardiovascular events detailed in Appendix 2 and all deaths, whether or not they are considered SAEs, specific Cardiovascular (CV) and Death sections of the CRF will be required to be completed. These sections include questions regarding cardiovascular (including sudden cardiac death) and non-cardiovascular death.

The CV CRFs are presented as queries in response to reporting of certain CV medical dictionary for regulatory activities (MedDRA) terms. The CV information should be recorded in the specific cardiovascular section of the CRF within one week of receipt of a CV Event data query prompting its completion.

The Death CRF is provided immediately after the occurrence or outcome of death is reported. Initial and follow-up reports regarding death must be completed within one week of when the death is reported.

7.4.1.5. Regulatory Reporting Requirements for SAEs and ADRs

Prompt notification by the investigator to GSK of SAEs and non-serious AEs related to the use of any GSK products (even for non- interventional post-marketing studies) is essential so that legal obligations and ethical responsibilities towards the safety of subjects and the safety of a product under clinical investigation are met.

GSK has a legal responsibility to notify both the local regulatory authority and other regulatory agencies about the safety of a product under clinical investigation. GSK will comply with country specific regulatory requirements relating to safety reporting to the regulatory authority, Institutional Review Board (IRB)/Independent Ethics Committee (IEC) and investigators.

Investigator safety reports are prepared for suspected unexpected serious adverse reactions according to local regulatory requirements and GSK policy and are forwarded to investigators as necessary.

An investigator who receives an investigator safety report describing a SAE(s) or other specific safety information (e.g., summary or listing of SAEs) from GSK will file it with the IB and will notify the IRB/IEC, if appropriate according to local requirements.

31 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

7.4.2. Pregnancy

 Any pregnancy exposure to GSK products that is notified during study participation must be reported using a pregnancy notification form. To ensure subject safety, each pregnancy must be reported to GSK within 2 weeks of learning of its occurrence. The pregnancy must be followed up to determine outcome (including premature termination) and status of mother and child. Pregnancy complications and elective terminations for medical reasons must be reported as an AE or SAE. Spontaneous abortions must be reported as an SAE.

 In addition, the investigator must attempt to collect pregnancy information on any female partners of male study subjects who become pregnant while the subject is enrolled in the study. Pregnancy information must be reported to GSK as described above.

7.4.3. Physical Exams

 A brief physical examination will include, at a minimum assessments of the lungs, cardiovascular system, and nose.

7.4.4. Vital Signs

 Vital signs will be measured in semi-supine position after 5 minutes rest and will include temperature, systolic and diastolic blood pressure and pulse rate.

8. DATA MANAGEMENT

 For this study subject data will be collected using GSK defined CRFs and combined with data provided from other sources in a validated data system.

 Management of clinical data will be performed in accordance with applicable GSK standards and data cleaning procedures to ensure the integrity of the data, e.g., removing errors and inconsistencies in the data.

 Adverse events will be coded using MedDRA and concomitant medications terms will be coded using the World Health Organisation (WHO) Drug system.

32 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

 CRFs (including queries and audit trails) will be retained by GSK, and copies will be sent to the investigator to maintain as the investigator copy. Subject initials will not be collected or transmitted to GSK according to GSK policy.

9. STATISTICAL CONSIDERATIONS AND DATA ANALYSES

9.1. Hypotheses

The null hypothesis for this study is that there is no difference between ACT guided treatment and usual care in the percentage of subjects who have an ACT score ≥20 or an improvement of more than 3 points in ACT during the 24-week treatment period for the intention-to-treat (ITT) population.

The alternative hypothesis for this study is that there is a difference between ACT guided treatment and usual care in the percentage of subjects who have an ACT score ≥20 or an improvement of more than 3 points in ACT during the 24-week treatment period for the ITT population.

The study is designed to show superiority of ACT guided treatment over usual care.

9.2. Sample Size Considerations

9.2.1. Sample Size Assumptions

The study will be powered to demonstrate superiority of ACT guided treatment over usual care from baseline over Week 24 in the primary endpoints.

We assume an intra-cluster correlation coefficient (ICC) of ρ=0.01, 29% well-control rate for usual care group, and 44% for ACT guided treatment group. It is required to have 6 centres for each treatment group and a minimum of 35 subjects for each centre (a total of 420 evaluable subjects) to achieve a power of 80% with two-sided α=0.05. We anticipate a dropout rate of 20% and therefore planned to recruit 44 subjects in each centre (a total of 528 randomized subjects). These sample size calculations are produced based on Donner [Donner, 1981].

An ACT score ≥20 indicates controlled asthma. The assumed rate of subjects with an ACT score ≥20 for usual care group is based on a nationwide epidemiologic survey in

33 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

China [Nan, 2013], which reported a 29% asthma control rate. The rate for ACT guided treatment is assumed based on GOAL study [Bateman, 2004], which reported a 80% rate for salmeterol/fluticasone group. An additional adjustment has been made to reduce it to 44% to account for real life practice. A 15% difference between ACT guided treatment group and usual care group is considered clinically meaningful.

Study centres will be randomly assigned to the 2 groups using a computer generated randomization list. The randomization will be stratified according to the Tier of the hospitals (Tier 3 vs. Tier 2).

9.2.2. Sample Size Sensitivity

The robustness and sensitivity of the sample size calculation has been considered in order to assess the impact on the power of the study should the observed well-control rates for the primary endpoints be higher or lower than expected. Table 2 displays the estimated sample size and power for a variety of rates with other assumptions unchanged.

Table 2 Sample Size and Power Calculation by Efficacy Rates

Well-control rate Well-control rate Total sample size Power if 528 ACT guided usual care to achieve 80% subjects are treatment power enrolled

40% 27% 680 69%

40% 29% 970 54%

40% 31% 1474 39%

44% 27% 404 89%

44% 29% 528 80%

44% 31% 718 67%

48% 27% 268 98%

48% 29% 332 94%

48% 31% 422 88%

34 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

9.2.3. Sample Size Re-estimation or Adjustment

No sample size re-estimation is currently planned for this study. However, if during the course of the study, new information becomes available about clinically meaningful differences for the primary endpoints, a sample size re-estimation may be conducted. Full details of the procedure used would be specified in the reporting and analysis plan (RAP) and any subsequent change to the target sample size would be documented in a protocol amendment.

9.3. Data Analysis Considerations

All statistical analyses will be conducted by GSK or its designee.

In general, descriptive statistics including the number of observations, mean, standard deviation (SD), median, minimum and maximum will be provided for continuous data, frequency counts and percentages will be used in tabulation for categorical variables. Unless specified otherwise, the percentage calculation will not take the missing values into account.

If not specified otherwise, all data will be summarized/listed as observed. Missing values will not be imputed.

All tests will be two-tailed, the general two-sided significance level will be α= 0.05.

A complete description of data handling rules and planned statistical analyses are provided in a separate RAP. The RAP will be finalized before the database lock.

9.3.1. Analysis Populations

Three populations are defined for this study.

 Intent-to-Treat (ITT) Population: all subjects who signed informed consent form, were randomised and who had at least one post-baseline assessment. The ITT population will be the primary population of interest.  Per Protocol (PP) Population: all subjects in the ITT population with no major protocol deviations and with ≥80% treatment compliance and diary compliance. The PP population will not be analyzed if this population comprises more than 95% or

35 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

less than 50% of the ITT population. Only the primary efficacy variables will be analyzed using the PP population.  Safety Population: all subjects who are enrolled into the study and who have at least one DRC assessment.

9.3.2. Consideration on Missing Data Handling

Missing values will not be considered when calculating the ACT score for primary endpoint, symptom score and/or PEF over 24-week treatment period. If not specified otherwise, missing values will not be imputed.

9.3.3. Interim Analysis

There are no interim analyses planned for this study. 9.4. Key Elements of Analysis Plan

All hypothesis tests and confidence intervals will be two-sided. Except where otherwise stated, the estimation of treatment effect will be adjusted for the effects of centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, and age. Centre, as the unit of randomization, will be treated as a random effect.

All efficacy measures over the course of the study will be presented and summarised in graphs and tables. Continuous data will be summarised by means, standard deviations, medians, minimum and maximum; categorical data will be summarised by counts and percentages.

All outcomes will be pertained to the individual participant’s level and the analysis model will be adjusted for the effect of centre (cluster).

36 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

9.4.1. Primary Analyses

The primary endpoints will be analyzed using logistic regression. The statistical model on which the inference will be based will include terms for treatment, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age. Centre, as the unit of randomization, will be treated as a random effect. The results of the primary analyses will be presented as point estimates, 95% confidence intervals and associated p-values for the adjusted mean differences between ACT guided treatment and usual care.

Analyses of the primary endpoints will be performed on the ITT population and the PP population; the ITT population is considered primary.

9.4.2. Secondary Analyses

All secondary endpoints will be confined to the ITT population.

Continuous efficacy endpoints will be analyzed using analysis of covariance (ANCOVA) and adjusted for centre as a random effect. For daytime and night-time symptom score, the same method (ANCOVA adjusted by centre) will be used if the parametric model assumptions hold. The assumptions of normality and homogeneity of variance will be assessed by inspection of normal probability plots and residual plots. If these assumptions are not met, alternative methods (e.g., nonparametric methods) may be performed in order to assess the robustness of the conclusions from the primary analysis.

For time to event endpoint, stratified log-rank test and Cox regression analysis will be performed with centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age as the stratification factors. Kaplan-Meier methodology will be used to estimate median time for each treatment arm. Kaplan-Meier curves will be constructed to provide a visual description of the difference between the two treatment arms.

9.4.3. Other Analyses

Asthma exacerbation rate will be analyzed using logistic regression as described for the primary endpoints.

Interactions between treatment and each of the covariates will be investigated in turn, with all main effects in the model regardless of their statistical significance. Any

37 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

interaction terms found to be statistically significant will be explored and if necessary results will be reported for each level of the covariate. Investigation of interactions will be confined to the primary endpoints. The effect of interactions (e.g., treatment by centre) will be assessed at the 10% level of significance.

Adverse events

The number (%) of subjects in each treatment group with treatment emergent AEs will be produced. Counts and percentages will also be presented of subjects with serious adverse events (SAEs), AEs leading to withdrawal, AEs by severity and AEs by relationship to study drug.

Other safety measures

Reason for withdrawal will be summarised.

Summary statistics for vital signs evaluations at each visit will be presented by treatment group. In addition, summary statistics for change from baseline for vital signs evaluations will be presented by treatment group.

10. STUDY GOVERNANCE CONSIDERATIONS

10.1. Posting of Information on Publicly Available Clinical Trial Registers

Study information from this protocol will be posted on publicly available clinical trial registers before enrollment of subjects begins.

10.2. Regulatory and Ethical Considerations, Including the Informed Consent Process

The study will be conducted in accordance with all applicable regulatory requirements, and with GSK policy.

The study will also be conducted in accordance with International Council for Harmonisation (ICH) Good Clinical Practice (GCP), all applicable subject privacy requirements, and the guiding principles of the current version of the Declaration of Helsinki. This includes, but is not limited to, the following:

38 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

 IRB/IEC review and favorable opinion/approval of the study protocol and amendments as applicable

 Signed informed consent to be obtained for each subject before participation in the study (and for amendments as applicable)

 Investigator reporting requirements (e.g. reporting of AEs/SAEs/protocol deviations to IRB/IEC)

 GSK will provide full details of the above procedures, either verbally, in writing, or both.

 Signed informed consent must be obtained for each subject prior to participation in the study

 The IEC/IRB, and where applicable the regulatory authority, approve the clinical protocol and all optional assessments, including genetic research.

 Optional assessments (including those in a separate protocol and/or under separate informed consent) and the clinical protocol should be concurrently submitted for approval unless regulation requires separate submission.

 Approval of the optional assessments may occur after approval is granted for the clinical protocol where required by regulatory authorities. In this situation, written approval of the clinical protocol should state that approval of optional assessments is being deferred and the study, with the exception of the optional assessments, can be initiated.

10.3. Quality Control (Study Monitoring)

 In accordance with applicable regulations including GCP, and GSK procedures, GSK monitors will contact the site prior to the start of the study to review with the site staff the protocol, study requirements, and their responsibilities to satisfy regulatory, ethical, and GSK requirements.

 When reviewing data collection procedures, the discussion will also include identification, agreement and documentation of data items for which the CRF will serve as the source document.

39 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

GSK will monitor the study and site activity to verify that the:

 Data are authentic, accurate, and complete.

 Safety and rights of subjects are being protected.

 Study is conducted in accordance with the currently approved protocol and any other study agreements, GCP, and all applicable regulatory requirements.

The investigator and the head of the medical institution (where applicable) agrees to allow the monitor direct access to all relevant documents

10.4. Quality Assurance

 To ensure compliance with GCP and all applicable regulatory requirements, GSK may conduct a quality assurance assessment and/or audit of the site records, and the regulatory agencies may conduct a regulatory inspection at any time during or after completion of the study.

 In the event of an assessment, audit or inspection, the investigator (and institution) must agree to grant the advisor(s), auditor(s) and inspector(s) direct access to all relevant documents and to allocate their time and the time of their staff to discuss the conduct of the study, any findings/relevant issues and to implement any corrective and/or preventative actions to address any findings/issues identified.

10.5. Study and Site Closure

 Upon completion or premature discontinuation of the study, the GSK monitor will conduct site closure activities with the investigator or site staff, as appropriate, in accordance with applicable regulations including GCP, and GSK Standard Operating Procedures.

 GSK reserves the right to temporarily suspend or prematurely discontinue this study at any time for reasons including, but not limited to, safety or ethical issues or severe non-compliance. For multicentre studies, this can occur at one or more or at all sites.

 If GSK determines such action is needed, GSK will discuss the reasons for taking such action with the investigator or the head of the medical institution (where applicable). When feasible, GSK will provide advance notification to the

40 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

investigator or the head of the medical institution, where applicable, of the impending action.

 If the study is suspended or prematurely discontinued for safety reasons, GSK will promptly inform all investigators, heads of the medical institutions (where applicable) and/or institution(s) conducting the study. GSK will also promptly inform the relevant regulatory authorities of the suspension or premature discontinuation of the study and the reason(s) for the action.

 If required by applicable regulations, the investigator or the head of the medical institution (where applicable) must inform the IRB/IEC promptly and provide the reason for the suspension or premature discontinuation.

10.6. Records Retention

 Following closure of the study, the investigator or the head of the medical institution (where applicable) must maintain all site study records (except for those required by local regulations to be maintained elsewhere), in a safe and secure location.

 The records must be maintained to allow easy and timely retrieval, when needed (e.g., for a GSK audit or regulatory inspection) and must be available for review in conjunction with assessment of the facility, supporting systems, and relevant site staff.

 Where permitted by local laws/regulations or institutional policy, some or all of these records can be maintained in a format other than hard copy (e.g., microfiche, scanned, electronic); however, caution needs to be exercised before such action is taken.

 The investigator must ensure that all reproductions are legible and are a true and accurate copy of the original and meet accessibility and retrieval standards, including re-generating a hard copy, if required. Furthermore, the investigator must ensure there is an acceptable back-up of these reproductions and that an acceptable quality control process exists for making these reproductions.

 GSK will inform the investigator of the time period for retaining these records to comply with all applicable regulatory requirements. The minimum retention time

41 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

will meet the strictest standard applicable to that site for the study, as dictated by any institutional requirements or local laws or regulations, GSK standards/procedures, and/or institutional requirements.

 The investigator must notify GSK of any changes in the archival arrangements, including, but not limited to, archival at an off-site facility or transfer of ownership of the records in the event the investigator is no longer associated with the site.

10.7. Provision of Study Results to Investigators, Posting of Information on Publically Available Clinical Trials Registers and Publication

Where required by applicable regulatory requirements, an investigator signatory will be identified for the approval of the clinical study report. The investigator will be provided reasonable access to statistical tables, figures, and relevant reports and will have the opportunity to review the complete study results at a GSK site or other mutually- agreeable location.

GSK will also provide the investigator with the full summary of the study results. The investigator is encouraged to share the summary results with the study subjects, as appropriate.

A manuscript will be progressed for publication in the scientific literature if the results provide important scientific or medical knowledge.

42 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

11. REFERENCES

Bateman ED, Boushey HA, Bousquet J, Busse WW, Clark TJH, Pauwels RA, et al. Can guideline defined asthma control be achieved? Am J Respir Crit Care Med 2004; 170:836– 844.

Chapman K.R., Boulet L.P., Rea R.M., Franssen E.. Suboptimal asthma control: prevalence, detection and consequences in general practice. Eur Respir J 2008; 31: 320–325

Donner A, Birkett N, Buck C. Randomization by cluster sample size requirements and analysis. Am. J. Epidemiol. 1981; 114 (6): 906-914.

Global Initiative for Asthma. GINA Report: global strategy for asthma management and prevention. www.ginasthma.com; 2013.

Juniper EF, Buist S, Cox F, Ferrie PJ, King DR. Validation of a standardized version of the Asthma Quality of Life Questionnaire. Chest 1999;115:1265-70.

Juniper EF, Guyatt GH, Epstein RS, Ferrie PJ, Jaeschke R, Hiller TK. Evaluation of impairment of health related quality of life in asthma: development of a questionnaire for use in clinical trials. Thorax 1992;47(2):76-83.

Juniper EF, Guyatt GH, Willan A, Griffith LE. Determining a minimal important change in a disease-specific Quality of Life Questionnaire. J Clin Epidemiol 1994;47(1):81-7.

Nan Su, Jiangtao Lin, Ping Chen, Jing Li, Changgui Wu, Kaisheng Yin, et al. Evaluation of asthma control and patient’s perception of asthma: findings and analysis of a nationwide questionnaire-based survey in China. J Asthma, 2013; 50(8): 861–870

Nathan RA, Sorkness CA, Kosinski M, Schatz M, Li JT, Marcus P, et al. Development of the asthma control test: a survey for assessing asthma control. J Allergy Clin Immunol 2004;113(1):59-65.

Schatz M, Kosinski M, Yarlas AS, et al. The minimally important difference of the Asthma Control Test. J Allergy Clin Immunol 2009;124(4):719-23.

43 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

12. APPENDICES

12.1. Appendix 1 – Abbreviations and Trademarks

Abbreviations

ACT Asthma Control Test ADR Adverse Drug Reactions AE Adverse Event ALT Alanine Transaminase ANCOVA Analysis of Covariance AQLQ Asthma Quality of Life Questionnaire AQLQ(S) Standardised Asthma Quality of Life Questionnaire COPD Chronic Obstructive Pulmonary Disease CRF Case Report Form CV Cardiovascular DRC Daily Record Card ECG Electrocardiogram FEV1 Forced Expiratory Volume in one second GCP Good Clinical Practice GINA International Guidelines for Asthma Management GSK GlaxoSmithKline IB Investigator Brochure ICC Intra-cluster Correlation Coefficient ICH International Council for Harmonisation ICS Inhaled Corticosteroid IEC Independent Ethics Committee INR International Normalized Ratio IRB Institutional Review Board ITT Intention-to-Treat LABA Long-acting β2 agonist Mg Milligram/s mL Millilitre/s PEF Peak Expiratory Flow PP Per Protocol RAP Reporting and Analysis Plan SAE Serious Adverse Event SD Standard Deviation ULN Upper Limit of Normal WHO World Health Organisation

44 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

Trademark Information

Trademarks of the GlaxoSmithKline Trademarks not owned by the group of companies GlaxoSmithKline group of companies ACT AQLQ(S)

Mini-Wright

45 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

12.2. Appendix 2 : Definition of and Procedures for Recording, Evaluating, Follow-Up and Reporting of Adverse Events

12.2.1. Definition of Adverse Events

Adverse Event Definition:

 An AE is any untoward medical occurrence in a patient or clinical investigation subject, temporally associated with the use of a medicinal product, whether or not considered related to the medicinal product.  NOTE: An AE can therefore be any unfavorable and unintended sign (including an abnormal laboratory finding), symptom, or disease (new or exacerbated) temporally associated with the use of a medicinal product.

Events meeting AE definition include:

 Any abnormal laboratory test results (hematology, clinical chemistry, or urinalysis) or other safety assessments (e.g., Electrocardiograms (ECGs), radiological scans, vital signs measurements), including those that worsen from baseline, and felt to be clinically significant in the medical and scientific judgement of the investigator.  Exacerbation of a chronic or intermittent pre-existing condition including either an increase in frequency and/or intensity of the condition.  New conditions detected or diagnosed after study treatment administration even though it may have been present prior to the start of the study.  Signs, symptoms, or the clinical sequelae of a suspected interaction.  Signs, symptoms, or the clinical sequelae of a suspected overdose of either study treatment or a concomitant medication (overdose per se will not be reported as an AE/SAE unless this is an intentional overdose taken with possible suicidal/self- harming intent. This should be reported regardless of sequelae).  The signs and symptoms and/or clinical sequelae resulting from lack of efficacy will be reported if they fulfil the definition of an AE or SAE. Also, "lack of efficacy" or "failure of expected pharmacological action" also constitutes an AE or SAE.

Events NOT meeting definition of an AE include:

 Any clinically significant abnormal laboratory findings or other abnormal safety assessments which are associated with the underlying disease, unless judged by the investigator to be more severe than expected for the subject’s condition.  The disease/disorder being studied or expected progression, signs, or symptoms of the disease/disorder being studied, unless more severe than expected for the subject’s condition.

46 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

 Medical or surgical procedure (e.g., endoscopy, appendectomy): the condition that leads to the procedure is an AE.  Situations where an untoward medical occurrence did not occur (social and/or convenience admission to a hospital).  Anticipated day-to-day fluctuations of pre-existing disease(s) or condition(s) present or detected at the start of the study that do not worsen.

12.2.2. Definition of Serious Adverse Events

If an event is not an AE per definition above, then it cannot be an SAE even if serious conditions are met (e.g., hospitalization for signs/symptoms of the disease under study, death due to progression of disease, etc).

Serious Adverse Event (SAE) is defined as any untoward medical occurrence that, at any dose:

a. Results in death b. Is life-threatening NOTE:

The term 'life-threatening' in the definition of 'serious' refers to an event in which the subject was at risk of death at the time of the event. It does not refer to an event, which hypothetically might have caused death, if it were more severe.

c. Requires hospitalization or prolongation of existing hospitalization NOTE:

 In general, hospitalization signifies that the subject has been detained (usually involving at least an overnight stay) at the hospital or emergency ward for observation and/or treatment that would not have been appropriate in the physician’s office or out-patient setting. Complications that occur during hospitalization are AEs. If a complication prolongs hospitalization or fulfills any other serious criteria, the event is serious. When in doubt as to whether “hospitalization” occurred or was necessary, the AE should be considered serious.  Hospitalization for elective treatment of a pre-existing condition that did not worsen from baseline is not considered an AE. d. Results in disability/incapacity NOTE:

 The term disability means a substantial disruption of a person’s ability to conduct normal life functions.  This definition is not intended to include experiences of relatively minor medical significance such as uncomplicated headache, nausea, vomiting, diarrhea, influenza,

47 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

and accidental trauma (e.g. sprained ankle) which may interfere or prevent everyday life functions but do not constitute a substantial disruption e. Is a congenital anomaly/birth defect f. Other situations:  Medical or scientific judgment should be exercised in deciding whether reporting is appropriate in other situations, such as important medical events that may not be immediately life-threatening or result in death or hospitalization but may jeopardize the subject or may require medical or surgical intervention to prevent one of the other outcomes listed in the above definition. These should also be considered serious.  Examples of such events are invasive or malignant cancers, intensive treatment in an emergency room or at home for allergic bronchospasm, blood dyscrasias or convulsions that do not result in hospitalization, or development of drug dependency or drug abuse g. Is associated with liver injury and impaired liver function defined as:  Alanine Transaminase (ALT) 3x upper limit of normal (ULN) and total bilirubin* 2xULN (>35% direct), or  ALT 3xULN and International normalized ratio (INR)** >1.5. * Serum bilirubin fractionation should be performed if testing is available; if unavailable, measure urinary bilirubin via dipstick. If fractionation is unavailable and ALT 3xULN and total bilirubin 2xULN, then the event is still to be reported as an SAE.

** INR testing not required per protocol and the threshold value does not apply to subjects receiving anticoagulants. If INR measurement is obtained, the value is to be recorded on the SAE form.

12.2.3. Definition of Cardiovascular Events

Cardiovascular Events (CV) Definition:

Investigators will be required to fill out the specific CV event page of the CRF for the following AEs and SAEs:

 Myocardial infarction/unstable angina  Congestive heart failure  Arrhythmias  Valvulopathy  Pulmonary hypertension  Cerebrovascular events/stroke and transient ischemic attack

48 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

 Peripheral arterial thromboembolism  Deep venous thrombosis/pulmonary embolism  Revascularization

12.2.4. Recording of AEs and SAEs

AEs and SAE Recording:

 When an AE/SAE occurs, it is the responsibility of the investigator to review all documentation (e.g., hospital progress notes, laboratory, and diagnostics reports) relative to the event.  The investigator will then record all relevant information regarding an AE/SAE in the CRF  It is not acceptable for the investigator to send photocopies of the subject’s medical records to GSK in lieu of completion of the GSK, AE/SAE CRF page.  There may be instances when copies of medical records for certain cases are requested by GSK. In this instance, all subject identifiers, with the exception of the subject number, will be blinded on the copies of the medical records prior to submission of to GSK.  The investigator will attempt to establish a diagnosis of the event based on signs, symptoms, and/or other clinical information. In such cases, the diagnosis will be documented as the AE/SAE and not the individual signs/symptoms.

12.2.5. Evaluating AEs and SAEs

Assessment of Intensity

The investigator will make an assessment of intensity for each AE and SAE reported during the study and will assign it to one of the following categories:

 Mild: An event that is easily tolerated by the subject, causing minimal discomfort and not interfering with everyday activities.  Moderate: An event that is sufficiently discomforting to interfere with normal everyday activities  Severe: An event that prevents normal everyday activities. - an AE that is assessed as severe will not be confused with an SAE. Severity is a category utilized for rating the intensity of an event; and both AEs and SAEs can be assessed as severe.  An event is defined as ‘serious’ when it meets at least one of the pre-defined outcomes as described in the definition of an SAE.

49 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

Assessment of Causality

 The investigator is obligated to assess the relationship between study treatment and the occurrence of each AE/SAE.  A "reasonable possibility" is meant to convey that there are facts/evidence or arguments to suggest a causal relationship, rather than a relationship cannot be ruled out.  The investigator will use clinical judgment to determine the relationship.  Alternative causes, such as natural history of the underlying diseases, concomitant therapy, other risk factors, and the temporal relationship of the event to the study treatment will be considered and investigated.  The investigator will also consult the Investigator Brochure (IB) and/or Product Information, for marketed products, in the determination of his/her assessment.  For each AE/SAE the investigator must document in the medical notes that he/she has reviewed the AE/SAE and has provided an assessment of causality.  There may be situations when an SAE has occurred and the investigator has minimal information to include in the initial report to GSK. However, it is very important that the investigator always make an assessment of causality for every event prior to the initial transmission of the SAE data to GSK.  The investigator may change his/her opinion of causality in light of follow-up information, amending the SAE data collection tool accordingly.  The causality assessment is one of the criteria used when determining regulatory reporting requirements.

Follow-up of AEs and SAEs

 The investigator is obligated to perform or arrange for the conduct of supplemental measurements and/or evaluations as may be indicated or as requested by GSK to elucidate as fully as possible the nature and/or causality of the AE or SAE.  The investigator is obligated to assist. This may include additional laboratory tests or investigations, histopathological examinations or consultation with other health care professionals.  If a subject dies during participation in the study or during a recognized follow-up period, the investigator will provide GSK with a copy of any post-mortem findings, including histopathology.  New or updated information will be recorded in the originally completed CRF.  The investigator will submit any updated SAE data to GSK within the designated reporting time frames.

50 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

12.2.6. Reporting of SAEs/ADR to GSK

SAE/ADR reporting to GSK via paper CRF

 Facsimile transmission of the SAE/ADR paper CRF is the preferred method to transmit this information to the SAE/ADR coordinator  In rare circumstances and in the absence of facsimile equipment, notification by telephone is acceptable, with a copy of the SAE/ADR data collection tool sent by overnight mail  Initial notification via the telephone does not replace the need for the investigator to complete and sign the SAE/ADR CRF pages within the designated reporting time frames.  Contacts for SAE/ADR receipt can be found at this beginning of the protocol on the Sponsor/Medical Monitor Contact Information page.

51 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

12.3. Appendix 3 - Country Specific Requirements

This requirement applies to China studies only.

52 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

12.4. Appendix 4: Asthma Control Test This section contained Clinical Outcome Assessment data collection questionnaires or indices, which are protected by third party copyright laws and therefore have been excluded.

53 2013N186640_00 CONFIDENTIAL 2019N423508_00 201097

12.5. Appendix 5: Standardised Asthma Quality of Life Questionnaire (AQLQ(S))

This section contained Clinical Outcome Assessment data collection questionnaires or indices, which are protected by third party copyright laws and therefore have been excluded.

54 2019N423508_00

PPD

PPD

PPD PPD

1

2019N423508_00

CONFIDENTIAL the GlaxoSmithKline group of companies 201097

Reporting and Analysis Plan (RAP)

The effectiveness of Asthma Control Test guided treatment compared with usual care in China adult asthma patients

Version: 1.0

Date: 11Sep2019

Copyright 2019 the GlaxoSmithKline group of companies. All rights reserved. Unauthorised copying or use of this information is prohibited.

1

2

2019N423508_00

CONFIDENTIAL the GlaxoSmithKline group of companies 201097

RAP SIGNATURE PAGE

Author Date PPD

Research Statistician 11Sep2019 MacroStat China

Approver Date PPD

Senior Statistician 11Sep2019 GSK, R&D, China

2

3

2019N423508_00

CONFIDENTIAL 201097

TABLE OF CONTENTS

TABLE OF CONTENTS ...... 3

1. INTRODUCTION...... 5

2. SUMMARY OF KEY PROTOCOL INFORMATION ...... 5 2.1. Changes to the Protocol Defined Statistical Analysis Plan ...... 5 2.2. Study Objective(s) and Endpoint(s)...... 5 2.3. Study Design ...... 6 2.4. Statistical Hypotheses / Statistical Analyses ...... 6

3. PLANNED ANALYSES ...... 7 3.1. Interim Analyses ...... 7 3.2. Final Analyses ...... 7

4. ANALYSIS POPULATIONS ...... 7 4.1. Protocol Deviations...... 7 4.2. Diary Compliance ...... 8 4.3. Treatment Compliance...... 8

5. CONSIDERATIONS FOR DATA ANALYSES AND DATA HANDLING CONVENTIONS...... 8 5.1. Study Treatment Display Descriptors...... 8 5.2. Baseline Definitions ...... 9 5.3. Multicentre Studies ...... 9 5.4. Examination of Covariates, Other Strata and Subgroups ...... 9 5.5. Multiple Comparisons and Multiplicity ...... 9 5.6. Other Considerations for Data Analyses and Data Handling Conventions...... 10

6. STUDY POPULATION ANALYSES ...... 11 6.1. Overview of Planned Study Population Analyses...... 11

7. EFFICACY ANALYSES...... 11 7.1. Primary Efficacy Analyses ...... 11 7.1.1. Endpoint / Variables...... 11 7.1.2. Primary Analysis Methods ...... 12 7.1.3. Sensitivity and Supportive Analysis Methods ...... 12 7.2. Secondary Efficacy Analyses...... 15 7.2.1. Endpoint / Variables...... 15 7.2.2. Analysis Methods...... 15 7.3. Exploratory Efficacy Analyses...... 17

8. SAFETY ANALYSES ...... 19 8.1. Adverse Events Analyses ...... 19 8.2. Adverse Events of Special Interest Analyses ...... 19 8.3. Clinical Laboratory Analyses...... 19 8.4. Other Safety Analyses ...... 19

9. REFERENCES...... 20

3

4

2019N423508_00

CONFIDENTIAL 201097

10. APPENDICES ...... 21 10.1. Appendix 1: Protocol Deviation Management and Definitions for Per Protocol Population...... 21 10.2. Appendix 2: Schedule of Activities ...... 22 10.2.1. Protocol Defined Schedule of Events...... 22 10.3. Appendix 3: Study Phases and Treatment Emergent Adverse Events ...... 23 10.4. Appendix 4: Derived and Transformed Data ...... 24 10.4.1. General...... 24 10.5. Appendix 5: Reporting Standards for Missing Data...... 25 10.5.1. Premature Withdrawals...... 25 10.5.2. Handling of Missing Data...... 25 10.5.2.1. Handling of Missing and Partial Dates ...... 25 10.6. Appendix 6: Values of Potential Clinical Importance ...... 26 10.6.1. Laboratory Values...... 26 10.6.2. ECG...... 26 10.6.3. Vital Signs...... 26 10.7. Appendix 7: Abbreviations & Trade Marks ...... 27 10.7.1. Abbreviations...... 27 10.7.2. Trademarks ...... 28 10.8. Appendix 8: List of Data Displays ...... 29 10.8.1. Data Display Numbering ...... 29

4

5

2019N423508_00

CONFIDENTIAL 201097

1. INTRODUCTION

The purpose of this reporting and analysis plan (RAP) is to describe the analyses to be included in the Clinical Study Report for Protocol 201097.

Revision Chronology: 201097 27-Jun-2016 Original

2. SUMMARY OF KEY PROTOCOL INFORMATION 2.1. Changes to the Protocol Defined Statistical Analysis Plan

There were no changes or deviations to the originally planned statistical analysis specified in the protocol, dated 27-Jun-2016. Some sensitivity and supportive analyses are added in the primary efficacy analysis in this RAP (See details in Section 7.1.3).

2.2. Study Objective(s) and Endpoint(s)

Objectives Endpoints Primary Objectives Primary Endpoints  To compare the effectiveness of ACT  The percentage of subjects who have an guided treatment versus usual care in ACT total score ≥20 or an improvement of achieving asthma control more than 3 points in ACT total score in at least one post- baseline assessment during the 24-week treatment period Secondary Objectives Secondary Endpoints  To compare the effectiveness of ACT  Mean daytime symptom score over the 24-week guided treatment versus usual care in treatment period asthma subjects with respect to the  Mean night-time symptom score over the following parameters: 24-week treatment period • Symptoms  Mean change from baseline to the end of • Forced Expiratory Volume in one study in FEV1 second (FEV1)  Mean morning(AM) PEF over the 24- • Peak Expiratory Flow(PEF) week treatment period • The quality of life  Mean evening(PM) PEF over the 24- • Time to asthma control week treatment period  Mean change from baseline to the end of study in Standardised Asthma Quality of Life Questionnaire (AQLQ[S]) score  Time to first ACT total score ≥20 or improvement of more than 3 points in ACT over the 24-week treatment period  Time to first ACT total score ≥20 and improvement of more than 3 points in ACT total score over the 24-week treatment period Exploratory Objectives Exploratory Endpoints  To compare the effectiveness of ACT  Rate of moderate/severe asthma exacerbation over the guided treatment versus usual care in 24-week treatment period asthma exacerbation

5

6

2019N423508_00

CONFIDENTIAL 201097

2.3. Study Design

Overview of Study Design and Key Features

Design  This is a prospective, multicentre, cluster-randomized, open-label study. Features Dosing  Not Applicable Time &  [Refer to Appendix 2: Schedule of Activities] Events Treatment  In this cluster-randomization design, each study centre, considered as a cluster, will Assignment be randomized to either ACT guided treatment group or control group (usual care group). Interim  No interim analysis is planned. Analysis

2.4. Statistical Hypotheses / Statistical Analyses

The null hypothesis for this study is that there is no difference between ACT guided treatment and usual care in the percentage of subjects who have an ACT total score ≥20 or an improvement of more than 3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period for the intention-to-treat (ITT) population.

The alternative hypothesis for this study is that there is a difference between ACT guided treatment and usual care in the percentage of subjects who have an ACT total score ≥20 or an improvement of more than 3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period for the ITT population.

The study is designed to show superiority of ACT guided treatment over usual care.

6

7

2019N423508_00

CONFIDENTIAL 201097

3. PLANNED ANALYSES

3.1. Interim Analyses

There are no interim analyses planned for this study.

3.2. Final Analyses

The final planned analyses will be performed after all required database cleaning activities have been completed and final database release (DBR) and database freeze (DBF) has been declared by Data Management.

4. ANALYSIS POPULATIONS

Population Definition / Criteria Analyses Evaluated Intent-To-Treat  All subjects who signed informed consent form, were Study Population (ITT) randomised and who had at least one post-baseline Efficacy assessment on primary endpoints, second endpoints, and/or exploratory endpoints. The ITT population will be the primary population of interest. Per-Protocol  All subjects in the ITT population with no major protocol Efficacy (PP) deviations and with ≥80% treatment compliance and diary compliance. The PP population will not be analyzed if this population comprises more than 95% or less than 50% of the ITT population. Only the primary efficacy variables will be analyzed using the PP population. Safety  All subjects who are enrolled into the study and who Safety have at least one DRC assessment.

Refer to Appendix 8: List of Data Displays which details the population used for each display.

4.1. Protocol Deviations

Important protocol deviations (including deviations related to study inclusion/exclusion criteria, conduct of the trial, patient management or patient assessment) will be summarised and listed.

Important deviations, which result in exclusion from the analysis population, will also be summarised and listed. Protocol deviations will be tracked by the study team throughout the conduct of the study in accordance with the Protocol Deviation Management Plan (refer to Appendix 1).

o Data will be reviewed prior to freezing the database to ensure all important deviations and deviations which may lead to exclusion from the analysis are captured and categorised on the protocol deviations dataset. o This dataset will be the basis for the summaries and listings of protocol deviations.

7

8

2019N423508_00

CONFIDENTIAL 201097

A separate summary and listing of all inclusion/exclusion criteria deviations will also be provided. This summary will be based on data as recorded on the inclusion/exclusion page of the eCRF.

4.2. Diary Compliance

Subjects’ diary compliance will be assessed by the records of diary cards. For the subjects to be included in the PP population, there should be at least ≥80% expected diary counts in the diary cards according to the visits they have completed.

4.3. Treatment Compliance

Subjects’ treatment compliance will be calculated as follow:

Treatment compliance=

The number of days that the subjects take asthma drugs following the doctor’s advice will be captured in the diary cards. For the subjects to be included in the PP population, the treatment compliance should ≥80%.

5. CONSIDERATIONS FOR DATA ANALYSES ANDDATA HANDLING CONVENTIONS

Continuous data will be summarized by treatment using descriptive statistics: number of subjects/observations (n), arithmetic mean values (Mean), median value (Median), standard deviation (SD), minimum value (Min), and maximum value (Max). Min and Max will be presented with the same number of decimal places as the raw data recorded in the database. Mean and Median will be presented using one more additional decimal place. SD will be presented with two more decimal places than the raw data recorded in the database.

Categorical data will be summarized by treatment using frequencies and percentages. Counts and percentages will be presented for each category. Percentages will be presented to one decimal place.

Statistical tests of treatment effects will be performed at a 2-sided significance level of 0.05, unless otherwise stated. And all p-values will be rounded up to 3 decimal places.

5.1. Study Treatment Display Descriptors

Code Description Order in TLF

ACT Asthma Control Test 1

UC Usual Care 2

8

9

2019N423508_00

CONFIDENTIAL 201097

Treatment comparisons will be displayed as follows using the descriptors as specified: ACT vs UC. 5.2. Baseline Definitions

For primary endpoint, baseline will be the measurements taken at visit 0. For secondary endpoints, baselines of FEV1 and AQLQ(S) will be the measurements taken at visit 0; for symptom score and PEF, baselines are unavailable since the data will be collected daily using diary cards from the first day of asthma treatment and baseline assessments before treatment will not be conducted.

Unless otherwise stated, if baseline data is missing, no derivation will be performed and baseline will be set to missing.

5.3. Multicentre Studies

In this cluster-randomization design, each study centre, considered as a cluster, will be randomized to either ACT guided treatment group or control group (usual care group). The randomization will be stratified according to the hospital level (Tier 3 vs. Tier 2). The hospital level is determined in accordance with the uniform requirements of the government. A Tier 3 hospital is on a larger scale and provides superior medical care than a Tier 2 hospital. Centre, as the unit of randomization, will be treated as a random effect in generalized linear mixed model (GLMM) used to analyse primary endpoints.

5.4. Examination of Covariates, Other Strata and Subgroups

The primary endpoints will be analysed using mixed effect logistic regression. In this study, generalized linear mixed model (GLMM) will be used to control for centre as a random effect since each centre is considered as a cluster in this cluster-randomization study. The statistical model on which the inference will be based will include terms for treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (inhaled corticosteroids [ICS] alone vs. ICS/long-acting β2 agonist [LABA]), gender and age. The primary endpoints will also be presented in the following subgroups: type of baseline controller (inhaled corticosteroids [ICS] alone vs. ICS/long- acting β2 agonist [LABA]), gender and age (<50 years old vs. >=50 years old).

5.5. Multiple Comparisons and Multiplicity

The primary comparison of interest is the comparison between ACT and usual care for the primary endpoint in the ITT population. Analyses of efficacy endpoints will not be subject to any multiplicity adjustment.

9

10

2019N423508_00

CONFIDENTIAL 201097

5.6. Other Considerations for Data Analyses and Data Handling Conventions

Other considerations for data analyses and data handling conventions are outlined in the appendices:

Section Component 10.3 Appendix 3: Study Phases and Treatment Emergent Adverse Events 10.4 Appendix 4: Derived and Transformed Data 10.5 Appendix 5: Reporting Standards for Missing Data 10.6 Appendix 6: Values of Potential Clinical Importance

10

11

2019N423508_00

CONFIDENTIAL 201097

6. STUDY POPULATION ANALYSES

6.1. Overview of Planned Study Population Analyses

The study population analyses will be based on the ITT population, unless otherwise specified.

Study population analyses including analyses of subject’s disposition, population, protocol deviations, demographic and baseline characteristics, subject’s smoking history, alcoholism history, drug abuse history, asthma history, asthma exacerbation history, medical history (coded by MedDRA version 21.0 or above), asthma medications, and concomitant medications (coded by WHODrug version Global Mar 2018 or above), will be summarized and listed. Details of the planned displays are presented in Appendix 8: List of Data Displays.

7. EFFICACY ANALYSES

All hypothesis tests and confidence intervals will be two-sided. All efficacy measures over the course of the study will be presented and summarised. Continuous data will be summarised by means, standard deviations (SD), medians, minimum and maximum; categorical data will be summarised by counts and percentages.

All outcomes will be pertained to the individual participant’s level and the analysis model will be adjusted for the effect of centre (cluster).

Unless otherwise specified, endpoints / variables defined in Section 7.1 and Section 7.2 will be summarised using descriptive statistics, graphically presented (where appropriate) and listed.

7.1. Primary Efficacy Analyses

7.1.1. Endpoint / Variables

The ACT is a validated self-administered questionnaire utilising 5 questions to assess asthma control during the past 4 weeks on a 5-point categorical scale (1 to 5).

The total score is calculated as the sum of the scores from all 5 questions, provided all scores are non-missing; if any individual scores are missing then the overall score will be set to missing.

The primary endpoint for the study is the percentage of subjects who have an ACT total score ≥20 or an improvement of more than 3 points in ACT total score in at least one post-baseline assessment during the 24-week treatment period.

The calculation of percentage will base on the subjects with at least one post-baseline ACT assessment.

11

12

2019N423508_00

CONFIDENTIAL 201097

7.1.2. Primary Analysis Methods

The primary endpoints will be analysed using mixed effect logistic regression. In this study, generalized linear mixed model (GLMM) will be used to control for centre as a random effect since each centre is considered as a cluster in this cluster-randomization study. The statistical model on which the inference will be based will include terms for treatment, baseline ACT total score, baseline ACT total score squared, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age. Centre, as the unit of randomization, will be treated as a random effect. Type of baseline controller (ICS alone vs. ICS/LABA) derives from the current asthma medication record in visit 0. The results of the primary analyses will be presented as adjusted odds ratios, 2-sided 95% confidence intervals, and associated p-values comparing ACT guided treatment with usual care. The adjusted odds ratios and 95% confidence intervals will be presented by a forest plot.

The example SAS codes are as below: proc glimmix data=input_data method=quad; class invsite rndgrp blcontrol gender; model response= rndgrp actbl actblsq blcontrol gender age /s dist=binomial oddsratio cl link=logit; random intercept/subject= invsite; lsmeans rndgrp /diff cl; run; where input_data is the name of the input dataset and response is the response variable. In terms of the covariates rndgrp, actbl, actblsq, blcontrol, gender and age are variables for randomised treatment code, baseline ACT total score, baseline ACT total score squared, type of baseline controller, gender and age respectively. For random effect, invsite is variable for centre.

Interactions between treatment and each of the covariates will be investigated in turn, with all main effects in the model regardless of their statistical significance. Any interaction terms found to be statistically significant will be explored and if necessary results will be reported for each level of the covariate. Investigation of interactions will be confined to the primary endpoints. The effect of interactions (e.g., treatment by centre) will be assessed at the 10% level of significance.

Analyses of the primary endpoints will be performed on the ITT population and the PP population; the ITT population is considered primary. The PP population will not be analysed if this population comprises more than 95% or less than 50% of the ITT population. Only the primary efficacy variables will be analysed using the PP population.

7.1.3. Sensitivity and Supportive Analysis Methods

 The primary endpoint will be repeatedly analysed in the following subgroups, using the same method used in primary analysis method:  Type of baseline controller (inhaled corticosteroids [ICS] alone vs. ICS/long- acting β2 agonist [LABA])  Gender (male vs. female)  Age (<50 years old vs. >=50 years old)

12

13

2019N423508_00

CONFIDENTIAL 201097

For subgroup analyses, the model will include all the covariates used in the primary analysis model except for the subgroup variable itself. Adjusted odds ratios and their 95% confidence intervals will be presented by a forest plot. These analyses will be performed for the ITT and PP populations.

 A supportive analysis of the primary endpoint will be performed using Generalised Estimating Equations (GEEs) if the model can converge.

All non-missing data from visits with scheduled ACT assessments (Weeks 4, 8, 12, 16, 20 and 24) will be included and an unstructured working correlation matrix will be implemented. The model will include terms for treatment, baseline ACT total score, baseline ACT total score squared, age, gender, visit and treatment-by-visit interaction. The number and percentage of subjects with a response within each treatment group will be presented by visit, together with the adjusted odds ratios, 2-sided 95% confidence intervals, and associated p-values between ACT guided treatment and usual care. The adjusted odds ratios and 95% confidence intervals will be presented by a forest plot.

The example SAS codes are as below: proc genmod data= input_data descending; class rndgrp gender visit subjid/ ref=first param=ref; model response=rndgrp actbl actblsq blcontrol age gender visit rndgrp*visit/dist=bin; repeated subject=subjid /within=visit type=un corrw; estimate "Treatment effect at Visit 1" rndgrp 1 visit 0 0 0 0 0 rndgrp*visit 0 0 0 0 0/exp; estimate "Treatment effect at Visit 2" rndgrp 1 visit 0 0 0 0 0 rndgrp*visit 1 0 0 0 0/exp; estimate "Treatment effect at Visit 3" rndgrp 1 visit 0 0 0 0 0 rndgrp*visit 0 1 0 0 0/exp; estimate "Treatment effect at Visit 4" rndgrp 1 visit 0 0 0 0 0 rndgrp*visit 0 0 1 0 0/exp; estimate "Treatment effect at Visit 5" rndgrp 1 visit 0 0 0 0 0 rndgrp*visit 0 0 0 1 0/exp; estimate "Treatment effect at Visit 6" rndgrp 1 visit 0 0 0 0 0 rndgrp*visit 0 0 0 0 1/exp; run; where input_data is the name of the input dataset and response is the response variable. In terms of the covariates rndgrp, actbl, actblsq, blcontrol, gender, age and visit are variables for randomised treatment code, baseline ACT total score, baseline ACT total score squared, type of baseline controller, gender, age and visit number, respectively. For repeated measures, visit is variable for visit number and subjid is variable for subject’s ID.

This analysis will be performed for the ITT and PP populations.

 Percentage of subjects who have an ACT total score ≥20 and an improvement of more than 3 points in ACT total score.

The following endpoints will be analysed as sensitivity analyses.

13

14

2019N423508_00

CONFIDENTIAL 201097

 Percentage of subjects who have an ACT total score ≥20 and an improvement of more than 3 points in ACT total score at Week 12. Subjects with missing data at Week 12 will be set as non-responders.  Percentage of subjects who have an ACT total score ≥20 and an improvement of more than 3 points in ACT total score at Week 24. Subjects with missing data at Week 24 will be set as non-responders.  Percentage of subjects who have an ACT total score ≥20 and an improvement of more than 3 points in ACT total score at Weeks 16 and 20 and 24. Subjects with missing data at Weeks 16, 20 or 24 will be set as non-responders.

These endpoints will be analysed using the same model as defined for primary endpoint (GLMM). A supportive analysis will be conducted to analysis the percentage of subjects who have an ACT total score≥20 and an improvement of more than 3 points in ACT total score, using the same GEEs method as mentioned above. All non-missing data from visits with scheduled ACT assessments (Weeks 4, 8, 12, 16, 20 and 24) will be used.

These analyses will be performed for the ITT and PP populations.

 Change from baseline in ACT total score in post-baseline visits.

A Mixed Model Repeated Measures (MMRM) will be used with covariates of treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender, age, visit, treatment-by-visit interaction and treatment-by-baseline ACT total score, with centre as a random effect. The least squares (LS) mean and LS mean change from baseline for each treatment and the estimated LS mean change from baseline treatment difference will be presented at each visit together with the 95% confidence interval for the mean difference and P-value.

The example SAS codes are as below: proc mixed data=input_data; class invsite rndgrp visit subjid blcontrol gender; model actchbl=rndgrp actbl blcontrol gender age visit rndgrp*visit rndgrp*actbl/ ddfm =kr; random intercept/subject= invsite; repeated visit / subject=subjid type=un; lsmeans rndgrp*visit / cl diff e; ods output lsmeans=lsmeans; ods output diffs=diffs; run; quit; where input_data is the name of the input dataset and actchbl is change from baseline in ACT score. In terms of the covariates rndgrp, actbl, blcontrol, gender, age and visit are variables for randomised treatment code, baseline ACT total score, type of baseline controller, gender, age and visit number, respectively. For random effect, invsite is variable for centre. For repeated measures, visit is variable for visit number and subjid is variable for subject’s ID.

14

15

2019N423508_00

CONFIDENTIAL 201097

This analysis will be performed for the ITT and PP populations.

7.2. Secondary Efficacy Analyses

7.2.1. Endpoint / Variables

 Mean daytime symptom score over the 24-week treatment period  Mean night-time symptom score over the 24-week treatment period  Mean change from baseline to the end of study in FEV1  Mean morning (AM) PEF over the 24-week treatment period  Mean evening (PM) PEF over the 24-week treatment period  Mean change from baseline to the end of study in AQLQ(S) score  Time to first ACT total score ≥20 or improvement of more than 3 points in ACT total score over the 24-week treatment period  Time to first ACT total score ≥20 and improvement of more than 3 points in ACT total score over the 24-week treatment period

7.2.2. Analysis Methods

All secondary endpoints will be confined to the ITT population. The continuous efficacy endpoints will be summarized by mean, standard deviation (SD), median, minimum and maximum according to scheduled visits. Time to event endpoint will be summarized by median time, 95% confidence interval of median time, lower and upper quantiles and range.

Continuous efficacy endpoints will be analysed using analysis of covariance (ANCOVA) and adjusted for centre as a random effect.

For mean daytime and night-time symptom score and mean morning and evening PEF, the mean values will be based on the available data on the diary card over each visit assessment period. No imputations will be performed on missing data. The mean values will be considered missing if less than 14 days are recorded in each visit assessment period. A Mixed Model Repeated Measures (MMRM) will be used with covariates of treatment, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor. The results will be presented as LS mean (SE), 95% confidence interval for LS mean, treatment difference, 95% confidence interval for treatment difference and P-value by visit.

The example SAS codes are as below: proc mixed data=input_data; class invsite rndgrp visit subjid blcontrol gender; model endpoint=rndgrp actbl blcontrol gender age/ ddfm =kr; random intercept/subject= invsite; repeated visit / subject=subjid type=un; lsmeans rndgrp*visit / cl diff e; ods output lsmeans=lsmeans; ods output diffs=diffs; run; quit;

15

16

2019N423508_00

CONFIDENTIAL 201097 where input_data is the name of the input dataset and endpoint is the endpoint variable. In terms of the covariates rndgrp, actbl, blcontrol, gender and age are variables for randomised treatment code, baseline ACT total score, type of baseline controller, gender and age respectively. For random effect, invsite is variable for centre. For repeated measures, visit is variable for visit number and subjid is variable for subject’s ID.

The AQLQ(S) contains 32 items in four domains: activity limitation (11 items), symptoms (12 items), emotional function (5 items) and environmental stimuli (4 items). The following items are included in each of the 4 domains:

• Symptoms: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 29, 30

• Activity Limitation: 1, 2, 3, 4, 5, 11, 19, 25, 28, 31, 32

• Emotional Function: 7, 13, 15, 21, 27

• Environmental Stimuli: 9, 17, 23, 26

The response format consists of a seven-point scale where a value of 1 indicates “total impairment” and a value of 7 indicates “no impairment”. The total AQLQ(S) score is the mean of all 32 items in the questionnaire and each individual domain score is calculated as the mean of the items within that domain. Hence, the total and domain scores are also each defined on a range from 1 to 7 with higher scores indicating a higher quality of life.

For the total AQLQ(S) score, the score for a subject at any time point will only be calculated if at least 90% of the questions were answered (calculated as the mean of those non-missing questions). If fewer than 90% of the questions were answered then the mean score for that subject at that time point will be considered missing.

For each individual domain of the AQLQ(S) score, the score for a subject at any time point will only be calculated if at least 90% of the questions for that domain were answered (calculated as the mean of those non-missing questions). If fewer than 90% of the questions were answered for that domain then the mean score for that subject and domain at that time point will be considered missing.

The analysis of changing from baseline to the end of study in FEV1 and AQLQ(S) score will only include those who complete the study and change from baseline will be calculated as the difference between baseline visit and Visit 6.

For FEV1 and AQLQ(S) changing from baseline, a mixed model will be used with covariates of treatment, centre, type of baseline controller (ICS vs. ICS/LABA), gender and age, with the centre as a random factor. The results will be presented as LS mean (SE), 95% confidence interval for LS mean, treatment difference, 95% confidence interval for treatment difference

The example SAS codes are as below: proc mixed data=input_data; class invsite rndgrp blcontrol gender;

16

17

2019N423508_00

CONFIDENTIAL 201097 model epchbl=rndgrp actbl blcontrol gender age/ ddfm =kr; random intercept/subject= invsite; lsmeans rndgrp/ cl diff e; ods output lsmeans=lsmeans; ods output diffs=diffs; run; quit; where input_data is the name of the input dataset and epchbl is the endpoint variable changing from baseline. In terms of the covariates rndgrp, actbl, blcontrol, gender and age are variables for randomised treatment code, baseline ACT total score, type of baseline controller, gender and age respectively. For random effect, invsite is variable for centre.

For time to event endpoint, Cox regression analysis will be performed with treatment, baseline ACT total score, centre, type of baseline controller (ICS alone vs. ICS/LABA), gender and age as the adjusted factors. The results will be presented as hazard ratio (SE), 95% confidence interval and P-value. Log-rank test and Kaplan-Meier methodology will be used to estimate median time for each treatment arm. Kaplan-Meier curves will be constructed to provide a visual description of the difference between the two treatment arms.

The example SAS codes are as below: proc phreg data=input_data; class rndgrp actbl invsite blcontrol gender; model timeto1*eventflag(0)= rndgrp actbl invsite blcontrol gender age / risklimits ties=exact; hazardratio rndgrp/ diff=all; run; proc lifetest data=input_data outsurv=survest; time timeto1*eventflag(0); strata rndgrp; run; where input_data is the name of the input dataset, timeto1 is the variable for time to response and eventflag is the variable for sensor. In terms of the covariates rndgrp, actbl, invsite, blcontrol, gender and age are variables for randomised treatment code, baseline ACT total score, centre, type of baseline controller, gender and age respectively.

7.3. Exploratory Efficacy Analyses

Moderate/severe asthma exacerbations will be summarized, including number of subjects with moderate/severe exacerbations, number of moderate/severe exacerbations and characteristics of moderate/severe exacerbations. Exacerbations separated by less than 7 days will be treated as a continuation of the same exacerbation

The moderate/severe exacerbation rate will be analyzed using a generalized linear model, assuming the number of exacerbations has a negative binomial probability distribution and that its mean is related to covariate factors with a ‘log link’ function. The logarithm of time (year) on study will be used as an offset variable. All moderate/severe

17

18

2019N423508_00

CONFIDENTIAL 201097 exacerbations observed during the study, with the onset date of event from date of baseline visit to the end of study visit date or early withdrawal (from study) visit date, will be included in the analysis. Missing data due to early withdrawal from the study will not be imputed. The model will include covariates for treatment, baseline ACT total score, type of baseline controller (ICS alone vs. ICS/LABA), gender and age. The annual moderate/severe exacerbation rates and 95% confidence interval for each treatment will be presented. The rate ratio and 95% confidence interval, percent reduction in rate and 95% confidence interval and P value will also be presented.

The example SAS codes are as below: proc genmod data= input_data; class rndgrp blcontrol gender; model no_exac = rndgrp actbl blcontrol gender age / dist=negbin link=log offset=logtime wald type3; lsmeans rndgrp / cl diff OM exp; run; where input_data is the name of the input dataset and no_exac is the number of moderate/severe asthma exacerbations per subject during the study. In terms of the covariates rndgrp, actbl, blcontrol, gender and age are variables for randomised treatment code, baseline ACT total score, type of baseline controller, gender and age respectively. For offset, logtime is logarithm of time (year) on study.

18

19

2019N423508_00

CONFIDENTIAL 201097

8. SAFETY ANALYSES

The safety analyses will be based on the Safety population, unless otherwise specified.

8.1. Adverse Events Analyses

The number (%) of subjects in each treatment group with treatment emergent AEs will be produced. Counts and percentages will also be presented of subjects with serious adverse events (SAEs), AEs leading to withdrawal, AEs by severity, AEs by relationship to study procedure/study drug, asthma exacerbation and death.

Adverse events will be coded by MedDRA version 21.0 or above. Adverse events will be summarized and presented by system organ class (SOC) and preferred term (PT) by treatment group. Adverse events analyses including the analysis of AEs, SAEs, non- serious AE and other significant AEs will be summarized. All AEs and SAEs will be listed. The details of the planned displays are provided in Appendix 8: List of Data Displays.

8.2. Adverse Events of Special Interest Analyses

Except for study indication Asthma, there is no other adverse event of special interest.

8.3. Clinical Laboratory Analyses

Clinical laboratory data is not collected and analysed in this study.

8.4. Other Safety Analyses

The analyses of non-laboratory safety test results including vital signs and physical examination will be summarized and listed by descriptive statistics. The details of the planned displays are presented in Appendix 8 List of Data Displays.

19

20

2019N423508_00

CONFIDENTIAL 201097

9. REFERENCES

PPD 3.1 Clinical Study Protocol_201097_20160627_ EN,2016

PPD 3.2 Clinical Study Protocol _201097_20160627_CN,2016

20

21

2019N423508_00

CONFIDENTIAL 201097

10. APPENDICES

10.1. Appendix 1: Protocol Deviation Management and Definitions for Per Protocol Population

Refer document “PDMP201097 final.docx”

21

22

2019N423508_00

CONFIDENTIAL 201097

10.2. Appendix 2: Schedule of Activities

10.2.1. Protocol Defined Schedule of Events

ACTIVITY Visit Visit Visit Visit 3 Visit 4 Visit 5 Visit 6 0 1 2 ( or Early Withdrawal) Study Week (Visit Window） 0 4w± 8w± 12w± 16w± 20w± 24w± 1 1 1 1 1 1 Informed consent x

Randomization/Allocation of x clusters

Subject Demography x

Medical history x

Verification of x inclusion/exclusion criteria Efficacy Evaluation Verification of withdrawal x x x x x criteria x Issue diary record cards x x x x x x

Collect/Review diary record x x x x x x cards Compliance assessment x x x x x x Asthma control Test (ACT) x x x x x x x score AQLQ(S) x x Lung function Test x x

Safety Evaluation Adverse event assessment1 x x x x x x x

1. Serious AEs will be recorded from the time the consent form is signed until the follow-up visit. All AEs will be recorded from the start of study treatment until the follow-up visit.

22

23

2019N423508_00

CONFIDENTIAL 201097

10.3. Appendix 3: Study Phases and Treatment Emergent Adverse Events

Assessments and events will not be classified by study phase. All the adverse events and concomitant medications are collected during the study as treatment emergent adverse events.

23

24

2019N423508_00

CONFIDENTIAL 201097

10.4. Appendix 4: Derived and Transformed Data

10.4.1. General

Multiple Measurements at One Analysis Time Point  Mean of the measurements will be calculated and used in any derivation of summary statistics but if listed, all data will be presented.  Participants having both High and Low values for Normal Ranges at any post-baseline visit for safety parameters will be counted in both the High and Low categories of “Any visit post-baseline” row of related summary tables. This will also be applicable to relevant Potential Clinical Importance summary tables. Study Day  Not Applicable.

24

25

2019N423508_00

CONFIDENTIAL 201097

10.5. Appendix 5: Reporting Standards for Missing Data

10.5.1. Premature Withdrawals

Element Reporting Detail General  Subject study completion (i.e. as specified in the protocol) was defined as who has completed all visits of the study.  Withdrawn subjects were not replaced in the study.  All available data from participants who were withdrawn from the study will be listed and all available planned data will be included in summary tables and figures, unless otherwise specified.  Withdrawal visits will be slotted as per Appendix 2: Schedule of Activities or will be summarised as withdrawal visits. 10.5.2. Handling of Missing Data

Element Reporting Detail General  Missing data occurs when any requested data is not provided, leading to blank fields on the collection instrument:  These data will be indicated by the use of a “blank” in subject listing displays. Unless all data for a specific visit are missing in which case the data is excluded from the table.  Answers such as “Not applicable” and “Not evaluable” are not considered to be missing data and should be displayed as such. Outliers  Any participants excluded from the summaries and/or statistical analyses will be documented along with the reason for exclusion in the clinical study report.

10.5.2.1. Handling of Missing and Partial Dates

Element Reporting Detail General  Partial dates will be displayed as captured in subject listing displays. Adverse  Completely missing start or end dates will remain missing, with no imputation applied. Events  Also no imputation applied for Partial dates.  The recorded partial date will be displayed in listings. Concomitant  Completely missing start or end dates will remain missing, with no imputation applied. Medications/  Also no imputation applied for Partial dates. Medical  The recorded partial date will be displayed in listings. History

25

26

2019N423508_00

CONFIDENTIAL 201097

10.6. Appendix 6: Values of Potential Clinical Importance

10.6.1. Laboratory Values

Laboratory values are not collected in this study.

10.6.2. ECG

ECG values are not collected in this study.

10.6.3. Vital Signs

Vital Signs Values are only collected at Visit 0 as baseline characteristics, will be listed and summarized by descriptive statistics, normal range and clinical concern range will not be collected and summarized.

26

27

2019N423508_00

CONFIDENTIAL 201097

10.7. Appendix 7: Abbreviations & Trade Marks

10.7.1. Abbreviations

Abbreviation Description ACT Asthma Control Test AE Adverse Event

ANCOVA Analysis of Covariance

AQLQ Asthma Quality of Life Questionnaire

AQLQ(S) Standardised Asthma Quality of Life Questionnaire

CRF Case Report Form

DRC Daily Record Card

ECG Electrocardiogram

FEV1 Forced Expiratory Volume in one second

GINA International Guidelines for Asthma Management

GSK GlaxoSmithKline

ITT Intention-to-Treat

LABA Long-acting β2 agonist

MedDRA Medical Dictionary for Regulatory Activities

Mg Milligram/s mL Millilitre/s

PEF Peak Expiratory Flow

PP Per Protocol

PT Preferred Term

RAP Reporting and Analysis Plan

SAE Serious Adverse Event

SOC System Organ Class

SD Standard Deviation

27

28

2019N423508_00

CONFIDENTIAL 201097

Abbreviation Description WHO World Health Organisation

10.7.2. Trademarks

Trademarks of the GlaxoSmithKline Trademarks not owned by the Group of Companies GlaxoSmithKline Group of Companies ACT AQLQ(S) MedDRA SAS WHODrug

28

29

2019N423508_00

CONFIDENTIAL 201097

10.8. Appendix 8: List of Data Displays

10.8.1. Data Display Numbering

The following numbering will be applied for RAP generated displays:

Section Tables Figures Listings Study Population 1.01 to 1.09 1.01 to 1.10 Efficacy 2.01 to 2.28 16 2.01 to 2.04 Safety 3.01 to 3.11 3.01 to 3.03

For Details of Mock Shells for Data Displays, please refer document “C_012_RESP_201097_TFL Template_Mock-ups.doc”.

29

30

2019N423508_00 CONFIDENTIAL 201097

Asthma Quality of Life Questionnaire for General Activities (AQLQ(S))

This section contained Clinical Outcome Assessment data collection questionnaires or indices, which are protected by third party copyright laws and therefore have been excluded.

1