Slide 6
Crown copyright
3
Why is demonstration of BE on PK difficult
1) Variability of inhalation manoeuvres Reported ISV for salbutamol (2 to 5 periods in each study)
Device Intrasubject coefficient of variation (%)
Evohaler MDI1 Evohaler MDI2 Evohaler MDI
3
8.2% 6.1% 10.1% 7.1% 35.9% 40.4% 42.4% 52.0% 31.8% Subjects trained only at start of period 1 (of 5 periods) University of Bradford, UK – likely highly trained volunteers, experienced with inhalation studies
Easibreathe breath actuated MDI3 Easibreathe breath-actuated MDI4 Accuhaler DPI4 Turbuhaler DPI4 Evohaler MDI4 Evohaler MDI + Volumatic4
1. Hindle, Br J Clin Pharm 1992 2. Hindle, Thorax 1994 3. Tomlinson, Br J Clin Pharm 2003 4. Aswania, J Aerosol Med 2004
Slide 7
Crown copyright
Why is demonstration of BE on PK difficult
Generic vs Reference salmeterol-fluticasone MDI BE study data Individual data for 8 subjects (of 52) with highest and lowest salmeterol T/R AUC0-t ratios Individual T/R AUC0-t ratio Subject 1 Highest 2 3 4 5 Lowest 6 7 8 Salmeterol 390% 374% Fluticasone 368% 194%
342% 109% Correlation coefficient = 0.89 285% 153% 53% 51% 49% 41% 31% 34% 44% 32%
Slide 8
Crown copyright
4
How to address inhalation variability
↑ sample size addresses high intrinsic ISV but not differences in amount of drug inhaled due to unreliable inhalation Use method to standardise / monitor inhalation manoeuvre, e.g., inhalation profile recorder. Should: Variability Reduce type 2 error Not routinely seen as yet in registration studies
Slide 9
Crown copyright
Why is demonstration of BE on PK difficult
Variability of reference (and test) OIP Specifications for R product allow variations in delivered dose of e.g., 20 - 35% Given, inter alia, methodological issues related to quantification, specifications for FPD allow variation between R batches of, e.g., 25 – 45% Such variability / specification range is not a standard feature of medicinal product delivered e.g., by oral route Clearly variations in R FPD have huge implications for likelihood of demonstrating BE in a PK study In vitro comparisons may be less affected by such variability as mean in vitro parameters typically derived from several T & R batches – vs. 1 batch of T & R in BE

上一页下一页