21
NDA 20-833, SE1-004
Flovent Diskus, COPD indication, GlaxoSmithKline
BackgrounderPulmonary and Allergy Drugs Advisory Committee
Application Number: 20-833, SE1-004 Application Type: NDA
Sponsor: GlaxoSmithKline Proprietary Name: Flovent® Diskus®
Category of Drug: Corticosteroid USAN/Established Name: Fluticasone propionate
Route of Administration: Oral inhalation
dry powder
Medical Reviewer: Charles E. Lee, M.D. Review Date: 12/14/01
REVIEW SUMMARY:
This is an NDA review for a chronic obstructive pulmonary disease (COPD) indication for Flovent® Diskus®. The sponsor is GlaxoSmithKline. The proposed indication is the long-term maintenance treatment of COPD. The proposed starting dose for adults is one inhalation (250 mcg) twice daily. The labeling allows for increasing the dose to 2 inhalations (500 mcg) twice daily in patients with an inadequate response. There were three pivotal studies submitted in this application, Studies FLTA3025, SCFA3006, and SFCA3007. Although statistical significance was shown for FP 500, the effect size was relatively modest, particularly for FLTA3025, where 24 weeks of treatment resulted in a net improvement in FEV1 of only 50 mL. These studies do not appear to support the efficacy of FP 250 in the population studied, as a statistically significant difference from placebo was not noted in replicate studies. Most secondary efficacy endpoint variables did not support the efficacy of FP 500 or FP 250. The health-related quality of life instrument does not support the efficacy of FP 500 or FP 250. The sponsor does not appear to have established the safety of FP for this indication in the proposed population. The sponsor has not adequately studied the safety of FP in non-Caucasian patients with COPD. The pivotal studies were likely to be of insufficient duration to detect differences between treatment groups for time and exposure-related events such as fractures, cataracts, ocular pressure disorders, and disorders of glucose metabolism. The pivotal studies were not designed to specifically look for cataracts or systemic bone effects. Safety data from supportive studies raise concerns about a higher incidence of respiratory infections, pneumonia, HPA axis effects, among others, in FP-treated patients, and do not adequately address bone effects in the proposed population. The sponsor does not appear to have sufficiently established that the proposed dose of FP is effective or safe in the treatment of COPD.
SIGNED:
Medical Reviewer: Date:
Medical Team Leader: Date:
21
NDA 20-833, SE1-004
Flovent Diskus, COPD indication, GlaxoSmithKline
21
NDA 20-833, SE1-004
Flovent Diskus, COPD indication, GlaxoSmithKline
Table of contents
Table of Contents 2
Executive Summary 5
1. Recommendations 5
2. Summary of clinical findings 5
2.1. Brief overview of clinical program 5
2.2. Efficacy 5
2.3. Safety 7
2.4. Dosing 10
2.5. Special populations 10
Clinical Review 12
1. Introduction and background 12
1.1. Introduction 12
1.2. Foreign Marketing and Regulatory History 13
2. Clinically relevant findings from chemistry, toxicology, microbiology, biopharmaceutics, statistics, and/or other consultant reviews 14
3. Human pharmacokinetics and pharmacodynamics 16
3.1. Previously established data 16
3.2. Pharmacokinetic and pharmacodynamic data in this application 17
3.2.1. FLTA3025 17
3.2.2. FMS40243 18
3.2.3. FLTA1003 18
3.3. Comments from Clinical Pharmacology and Biopharmaceutics review 19
4. Description of clinical data and sources 20
5. Clinical review methods 23
5.1. Conduct of the review 23
5.2. Data quality 24
5.2.1. Ethical standards and financial disclosure 25
6. Integrated review of efficacy 27
6.1. Summary and conclusions 27
6.2. Content 29
6.3. Description of pivotal studies 29
6.3.1. Primary efficacy endpoint 30
6.3.2. Secondary efficacy endpoints 30
6.3.3. Disposition, demographics, and baseline characteristics 31
6.3.4. Primary efficacy endpoint 35
6.3.5. Secondary efficacy endpoints 37
6.3.6. COPD exacerbations 39
6.3.7. Health-related quality of life instrument 41
6.4. Subgroup analyses, smoking status and reversibility 42
6.4.1. Smoking status, subgroup analysis 42
6.4.2. “Non-reversible” population, subgroup analysis 43
6.5. Subgroup analyses, demographics 45
6.6. References 46
7. Integrated review of safety 47
7.1. Summary and conclusions 47
7.1.1. Content 50
7.1.2. Safety data from pivotal studies 50
7.1.3. Description of pivotal studies 50
7.1.4. Subgroup analyses of safety endpoints in pivotal studies 65
7.1.5. Supportive safety data from other studies 67
7.1.6. Spontaneous postmarketing reports 79
7.1.7. Safety update 79
7.1.8. References 79
8. Dosing, regimen, and administration issues 80
9. Use in special populations 80
9.1. Efficacy in special populations 80
9.1.1. Analysis of efficacy by gender 81
9.1.2. Analysis of efficacy by age 81
9.1.3. Analysis of efficacy by race 81
9.2. Safety in special populations 81
9.2.1. Analysis of safety by gender 82
9.2.2. Analysis of safety by age 82
9.2.3. Analysis of safety by race 83
10. Conclusions and recommendations 83
11. APPENDIX: Clinical Studies 86
11.1. FLTA3025: A randomized, double-blind, parallel-group, comparative trial of inhaled fluticasone propionate 250 mcg BID, 500 mcg BID, and placebo BID via the Diskus in subjects with chronic obstructive pulmonary disease (COPD) 86
11.1.1. Summary and reviewer's conclusion of study results 86
11.1.2. Study design 88
11.1.3. Objectives 88
11.1.4. Inclusion criteria 89
11.1.5. Exclusion criteria 91
11.1.6. Protocol amendments 93
11.1.7. Study procedures 93
11.1.8. Allowable concurrent medications 97
11.1.9. Prohibited medications 97
11.1.10. Drug product and placebo 97
11.1.11. Assessment of compliance 98
11.1.12. Assessment of signs and symptoms 98
11.1.13. Health-related quality of life instrument 100
11.1.14. Efficacy variables 101
11.1.15. Pharmacokinetics and pharmacodynamics 102
11.1.16. Safety variables 102
11.1.17. Statistical considerations 102
11.1.18. Results 104
11.1.19. References 132
11.2. SFCA3006: A randomized, double-blind, parallel-group trial of evaluating the safety and efficacy of the Diskus formulations of Salmeterol 50 mcg BID and fluticasone propionate 500 mcg BID individually and in combination as compared to placebo in COPD subjects 133
11.2.1. Summary and reviewer's conclusion of study results 133
11.2.2. Study design 135
11.2.3. Objectives 135
11.2.4. Inclusion criteria 136
11.2.5. Exclusion criteria 138
11.2.6. Protocol amendments 140
11.2.7. Study procedures 140
11.2.8. Allowable concurrent medications 144
11.2.9. Prohibited medications 144
11.2.10. Drug product and placebo 144
11.2.11. Assessment of compliance 145
11.2.12. Assessment of signs and symptoms 145
11.2.13. Health-related quality of life instrument 147
11.2.14. Efficacy variables 148
11.2.15. Safety variables 149
11.2.16. Statistical considerations 150
11.2.17. Results 151
11.2.18. References 183
11.3. SFCA3007: A randomized, double-blind, parallel-group trial of evaluating the safety and efficacy of the Diskus formulations of Salmeterol 50 mcg BID and fluticasone propionate 250 mcg BID individually and in combination as compared to placebo in COPD subjects 185
11.3.1. Summary and reviewer's conclusion of study results 185
11.3.2. Study design 187
11.3.3. Objectives 187
11.3.4. Inclusion criteria 188
11.3.5. Exclusion criteria 190
11.3.6. Protocol amendments 192
11.3.7. Study procedures 192
11.3.8. Allowable concurrent medications 196
11.3.9. Prohibited medications 196
11.3.10. Drug product and placebo 196
11.3.11. Assessment of compliance 197
11.3.12. Assessment of signs and symptoms 197
11.3.13. Health-related quality of life instrument 199
11.3.14. Efficacy variables 200
11.3.15. Safety variables 201
11.3.16. Statistical considerations 202
11.3.17. Results 203
11.3.18. References 235
Executive summary
1. Recommendations
Recommendations will be added after the Pulmonary and Allergy Advisory Committee meeting, which is scheduled for January 17, 2002.
2. Summary of clinical findings
2.1. Brief overview of clinical program
Fluticasone propionate is approved in the form of a dry powder inhaler as Flovent® Rotadisk, NDA 20-549, November 7, 1997, and as Flovent® Diskus®, NDA 20-833, September 29, 2000. The dry powder inhaler products are approved for use in adults and children ages 4 years and older for the maintenance treatment of asthma as prophylactic therapy. Fluticasone propionate inhalation aerosol (Flovent®, NDA 20-548) was approved March 27, 1996 in three dosage strengths (44 mcg, 110 mcg, and 220 mcg) for the maintenance treatment of asthma as prophylactic therapy.
The sponsor, GlaxoSmithKline, has submitted this NDA supplement for a chronic obstructive pulmonary disease (COPD) indication for Flovent® Diskus®. The proposed indication is the long-term maintenance treatment of COPD. The proposed starting dose for adults is one inhalation (250 mcg) twice daily. The proposed labeling allows for increasing the dose to 2 inhalations (500 mcg) twice daily to provide additional control for patients who do not respond adequately to the starting dose. The clinical program was also designed to pursue COPD indications for Serevent Diskus (salmeterol xinafoate 50 mcg) and Advair Diskus (fluticasone propionate 250 mcg /salmeterol xinafoate 50 mcg and fluticasone propionate 500 mcg/salmeterol xinafoate 50 mcg). Submissions similar to the Flovent Diskus NDA supplement have been simultaneously submitted to the Serevent Diskus NDA (NDA 20-692) and to the Advair Diskus NDA (NDA 21-077).
This submission refers to three pivotal clinical studies, Studies FLTA3025, SFCA3006, and SFCA3007. These studies are submitted in support of the efficacy of FP in the treatment of COPD. The pivotal studies were randomized, double-blind, parallel group, placebo-controlled, multicenter studies. Each of the pivotal studies had a 24-week treatment duration. Study FLTA3025 evaluated the efficacy of FP 250 and FP 500 compared with placebo. SAL 50/FP 500 and FP 500 were compared with each other and with placebo in SFCA3006. SAL 50/FP 250 and FP 250 were compared with each other and with placebo in SFCA3007. All three studies were similar in design, subject entrance criteria, and overall conduct to provide replication of treatment arms across studies. The sponsor provided safety data from the pivotal studies, supportive safety data from other studies, and a safety update, which included postmarketing reports and a literature review, in support of the safety of FP in the treatment of COPD.
2.2. Efficacy
The primary efficacy variable for the evaluation of FP in the pivotal studies was the pre-dose FEV1. The primary analysis for FP was the comparison of the mean morning pre-dose FEV1 between treatment groups at endpoint. Comparisons of FP 500 with placebo showed statistically significant differences in mean change from baseline in pre-dose FEV1 in both FLTA3025 and SFCA3006, although the effect size observed in FLTA3025 was only 50 mL, half of the “clinically significant” 100 mL the sponsor had powered the study to detect. The effect size reported in SFCA3006 was 113 mL. Comparisons of FP 250 with placebo showed statistically significant differences in mean change from baseline in pre-dose FEV1 only for SFCA3006 with an effect size of 108 mL, but not for FLTA3025, which had an effect size of only 27 mL.
No active treatment group had a clinically significant difference from the placebo group in mean change from baseline in the Global Assessment Score (GAS) of the Chronic Bronchitis Symptom Questionnaire (CBSQ). FP 500 and FP 250 did not have a clinically significant difference from the placebo group in the mean Transition Dyspnea Index (TDI) score. In FLTA3025 there were slightly fewer COPD exacerbations in FP 500 and FP 250 groups than in placebo, with a dose-response effect noted. Otherwise, the incidence of COPD exacerbations in the three pivotal studies was fairly similar among treatment groups. The incidence of moderate or severe COPD exacerbations in the three pivotal studies was similar among treatment groups. There were no significant differences between the FP and placebo groups for time to withdrawal from the study, time to withdrawal due to COPD exacerbation, and time to withdrawal due to COPD-related condition. Small changes from baseline in AM PEFR, daily Ventolin use, and number of awakenings per night requiring Ventolin use were noted for FP 250 and FP 500 groups.
The Chronic Respiratory Disease Questionnaire (CRDQ) was used to compare changes in the COPD-related quality of life for treatment groups. No active treatment group had a clinically significant difference from the placebo group in mean change from baseline in the Overall score of the CRDQ.
Subgroup analysis by smoking status showed that former smokers had larger mean changes from baseline in FEV1 at endpoint for all active study treatments than did current smokers. Subgroup analysis for the non-reversible and reversible groups showed that the non-reversible group had smaller mean changes from baseline in FEV1 at endpoint for all active study treatments than did the reversible group.
These studies do not appear to sufficiently support the efficacy of FP 500 or FP 250 in the treatment of COPD. The main issues included the patient population studied, inconsistent findings in the three pivotal studies with the primary endpoint, the absence of strong support from the secondary endpoints, and the failure of the studies to fully support the clinical relevance of the primary endpoint, particularly with regard to the quality of life (QOL) instrument and COPD exacerbation.
With regard to the patient population, the sponsor has not adequately studied the efficacy of FP in non-Caucasian patients with COPD. Additional study of the efficacy of FP in a more racially diverse study population is indicated. There are serious questions about whether the patient population studied, of which 51% to 59% were highly reversible, is representative of the US COPD population as a whole, for which a broad indication is sought. Although statistical significance was shown for FP 500, the effect size was relatively modest, particularly for FLTA3025, where 24 weeks of treatment resulted in a net improvement in FEV1 of only 50 mL. It should be noted that this result was also heavily driven by the “reversible” subgroup and by subjects who had stopped smoking. These studies do not appear to sufficiently support the efficacy of FP 250 in the population studied, as a statistically significant difference from placebo was not noted in replicate studies. In general, most of the secondary efficacy variables did not support the efficacy of FP 500 or FP 250. Secondary efficacy variables that would be expected to be correlated to the primary endpoint, such as PEFR, showed modest treatment effects and therefore do not add substantially to the argument in favor of efficacy for this product.
The quality of life (QOL) instrument, the CRDQ, does not appear to support the efficacy of FP 500 or FP 250. This is particularly concerning because all three of the pivotal studies had evaluation of QOL as one of their primary objectives. This was included because relatively short-term (6 months) changes in FEV1 have uncertain clinical significance and do not have the same correlation with mortality that long-term (3 year) changes would have. The small effect size in FEV1 combined with the observation that the patients experienced no detectable benefit vs. placebo in a well-validated QOL instrument argues strongly against a conclusion of efficacy for FP for the COPD indication.