CASCADE IMPACTION METHODS FOR DRY POWDER INHALERS USING THE HIGH FLOW-RATE MARPLE-MILLER IMPACTOR

The magnitude and effects of stage overload and particle re-entrainmen t in the new, Marple-Miller cascade impactor (MMI) were evaluated at 6 0 liter/min by sampling and determining the aerodynamic size distribut ions from two, excipient-free, powder inhalers (Turbohaler(TM) and Spi nhaler(TM)) according to a variety of experimental protocols. Drug dis tributions were compared statistically, for both inhalers, following s ingle dose experiments in the presence and absence of silicone oil imp actor stage coating and between single dose and multiple dose experime nts in its presence. Stage coating was found to be essential to preven t re-entrainment of drug from both inhalers. One or less than or equal to 25 dose sampling was shown to produce valid results provided impac tion stages were coated for the 0.5 mg Bricanyl Turbohaler (44.7 +/- 9 .6% of emitted dose < 5 mu m; overload and re-entrainment was evident following sampling of 40 doses). One or 2 dose sampling was shown to p roduce valid results for 20 mg Spinhaler, provided drug capture was en hanced further by also coating the aerosol inlet port to MMI (10.7 +/- 1.3% of emitted dose < 5 mu m; overload of stage 1 and re-entrainment was evident following sampling of 2 doses in the absence of a coated aerosol inlet port). The absence of significant re-entrainment could b e shown most effectively by statistically comparing values for percent deposition, at different sites in the stage-coated impactor, between single and multiple dose (greater than or equal to 2) experiments. Suc h an experiment should be performed for each type of inhaler and formu lation to be tested, as a means of validating the exact impaction tech nique to be used for size distribution analysis of powder inhaler emis sions.