Originally published in English in:


Magazine: Manufacturing Chemist



Issue:  September 2014



Title: Analytical strategies for the de-formulation of dry powder inhalers



Author:   Dr Paul Kippax, Pharmaceutical Portfolio Manager,

               Malvern Instruments



Our ref:  MAL/JOB/2876a



Words:  1709


Figures x3 (Thumbnails only at end of document – high res files supplied separately)



Figure 1: Powder concentration (Cv) versus Time profiles recorded for the emptying of a dry powder inhaler using laser diffraction. The data show that a higher  concentration is delivered at high flow rates. This relates to the energy available for aerosolisation of this cohesive formulation.




Figure 2: Comparing the Raman spectrum of individual particles with reference spectra for the APIs in the formulation enables secure chemical classification of the sample.




Figure 3: MDRS enables the precise classification of particles in an NGI sized fraction as being API, lactose or multicomponent agglomerates


 Analytical strategies for the de-formulation of dry powder inhalers



By DrPaul Kippax, Pharmaceutical Portfolio Manager, Malvern Instruments

作者:马尔文仪器制药产品经理Paul Kippax博士


The complexity of dry powder inhalers (DPIs) makes them one of the toughest generic development targets. In this article Dr Paul Kippax, Pharmaceutical Portfolio Manager for Malvern Instruments,reviews the guidance available from the FDA for DPI formulation development and discusses analytical strategies that can be employed for the process ofde-formulation of a complex DPI formulation.



The performance of aDPIis affected bycomplex and subtle interactions between the drug substance, excipients and the device used.Arguably this makes DPIs a more challenging generic target than any other pharmaceutical product, a conclusion evident from the current commercial landscape.Advair, for example, one of the most successful DPIs to come to market, is now off-patent but relatively free of generic competition,despite well-funded efforts by a number of generic companies to access market share [1].



The FDA’s guidance for industry for the development of DPIs [2]emphasisesthe importance ofparticlecharacterisation. Successful drug delivery with a DPI relies on dispersing particles to a suitable size for deposition in the lung. Particle characterisation helps to elucidate and control the dispersion mechanisms associated with drug delivery, and supports the development of a product with robust aerosolisation behavior, controlled bioavailability and consistent clinical efficacy. Relevant and efficient particle characterisation strategies aretherefore critical in the development of both new and generic products.



This article assessesthe potential of a number of analytical techniques to accelerate thede-formulation of DPIs, focusingon the use of Morphologically Directed Raman Spectroscopy (MDRS)in conjunction with cascade impaction to probe the nature of large particles and agglomerates.



Understanding DPIs


The majority ofDPIsare classified as passive, with delivery of the dose beingdriven solely by the inhalation manoeuvre of the patient. As the patientinhales, air is drawn through the device, aerosolizing the powder doseand dispersing it to form drug particles of a respirable size. The size of the particles delivered to the patient, and the total emitted dose, depend on theenergy supplied by inhalation being sufficient todisperse the formulation with the chosen device.The FDA draft guidance emphasises how all aspects of the DPI -device configuration, packaging (which can be blister, capsule or bulk reservoir) and the physical properties of the formulation - impact this dispersion process.



The size of the active pharmaceutical ingredient (API) particles received by the patientdirectly correlates with bioavailability in the lung. This is why delivered particle size, along with dose uniformity, is a critical parameter for DPI performance.A size range of 1-5 µm has been established as optimal for pulmonary delivery butparticles in this range tend to be highly cohesive.One of the central challenges of DPI development is to engineer a device/formulation combination that ensures successful dispersion of a cohesive dose using only the energy supplied by the inhaling patient.

患者摄入的活性药物成分(API),其颗粒大小直接关系到其在肺部的吸收程度。因此,干粉吸入剂的性能表现需要通过药物颗粒大小以及剂量均匀性来衡量。当前肺部给药的最优粒径范围为1- 5微米,但该范围内的颗粒往往具有较高的粘性。对干粉吸入剂进行开发时,研究人员面临的一个核心挑战就是要设计出一种设备/配方组合,以确保患者在吸气时就能均匀分散粘性药剂。


In many DPIs,efficient drug delivery is achieved by blending the API with a carrier particle. An alternative strategy is to engineer the particle properties of the API, such as shape, surface charge and roughness, to enable carrier-free delivery. De-formulation of a reference labeled drug (RLD)may therefore call for detailed analysis of a multi-component blend, and/or close scrutiny of the characteristics of the API.



Reviewing the analytical toolkit for DPI de-formulation


FDA draft guidance highlights size, size distribution, morphological features and the crystal habit of an API as being among the Critical Quality Attributes (CQAs) for DPIs - those parameters that define clinical performance. An important aspect of demonstrating bioequivalence (BE)is having a detailed understanding of the way in which Critical Material Attributes (CMAs) of the formulation interact with device parameters to affect these CQAs.



Cascade impaction is the technique specified in both the European and US Pharmacopoeias for measuring the aerodynamic particle size distribution of DPIs. This method size fractionatesthe dose delivered by a DPI on the basis of particle inertia, which is a function of aerodynamic particle size and particle velocity [3].The value of cascade impaction derives largely from its ability to deliver component specificAerodynamic Particle Size Distribution (APSD), which is obtained through the chemical characterisation of each individual size fraction. High Performance Liquid Chromatography (HPLC) is most usually employed for this purpose. This necessitates dissolution ofthe sample, meaning thatany information about the physical characteristics of particles within a specific size fraction is lost.



The technique highlighted by the FDA guidance for morphological studies within DPI development.is microscopy. This is an established technique butit too has limitations when it comes to efficientde-formulation.Conventional manual microscopy is labour intensive, timeconsuming anddoes not usually distinguish between the different components of a formulation.[2]



Complementary and alternative analytical techniques that can help to address these issues include laser diffraction, automated image analysis andMDRS, all of which can support and accelerate generic DPI development.



Laser diffraction – real-time monitoring of dispersion behavior

激光衍射 - 实时监控分散行为

Laser diffraction is an ensemble particle sizing technique that non-destructively delivers volume-based size distribution measurements across a particle size range of 0.01 – 3500µm. It is a highly automated and rapid method capable of measuring multiple particle size distributions in seconds. These attributes make laser diffraction highly complementary to cascade impaction.

激光衍射法是一种整体粒径测量技术,可对0.01– 3500微米粒径范围内的颗粒进行基于体积的粒径分布测量,且不会造成任何破坏。作为一种高度自动化的快速测量方法,它能够在数秒内测量多种粒度分布的情况。这些优点都足以让激光衍射法成为级联撞击技术的有效补充手段。


Laser diffraction measures the particle size of the whole formulation. It does notdeliver the component-specific information that cascade impaction provides, but it is far faster and provides valuable particle size data for the various stages of the actuation process.The capacity of laser diffraction for real-time analysis enables the measurement of particle size and concentration profiles for hundreds of device actuations in a single day.The technique can therefore be used to rapidly explore the dispersion behaviourof a DPI formulation and to correlate the effects of CMAs and device parameters with the efficiency of drug delivery [3]. Laser diffraction particle sizing provides detailed insight into dispersion dynamics, the efficiency of device emptying and the impact of flow rate on product performance (Figure 1). This information helps with scoping the design space for the product and supports therobust demonstration of BE.



Boosting the informational productivity of cascade impaction


Automated image analysis enables statistically relevant morphological characterisation of a sample in a fraction of the time required to gather far fewer data by microscopy. With automated image analysis, tens of thousands of individual particle images can be captured in a matter of minutes. These are then used to determine morphological parameters for each individual particle to generate size and shape distributions for a sample. Such data support the detailed morphological investigation of a DPI dose, enabling a thorough study of large particles and agglomerates as advocated by the regulatory guidance. Increasingly automated image analysis is therefore replacing microscopy for this aspect of DPI characterization.



In addition, because automated image analysis is both relatively fast and non-destructive, it can provide important morphological information which is beyond the scope ofHPLC analysis of the size fractionated samples produced by cascade impaction. This strategy enables particle size and shape information to be gathered from various stages of a cascade impactor to develop a deeper understanding of how dispersion proceeds [4].



Combining automatedimage analysiswithRaman spectroscopy enablesMorphologically Directed Raman Spectroscopy (MDRS), further extending this capability. With MDRS it becomes feasible to chemically identify individual particles and gather component specific morphological information for specific size fractions, within a DPI dose.



Case study: Using MDRS to gain insight into the dispersion behaviour of a DPI formulation containing two APIs



A sample of a commercially available dry powder inhaler containing two APIs was actuated into a Next Generation Impactor (Copley Scientific, UK) to disperse and fractionate the dose. A collection disk was placed in the collection cup of stage 3 of the impactor to capture particles on a suitable surface for MDRS. The collected dose was then transferred to a Morphologi G3-ID (Malvern Instruments Ltd, UK) for analysis.

研究人员将含有两种API的市售干粉吸入剂样本加入新一代撞击器(英国科普利科技有限公司产品)中进行分散和分级,并在撞击器中放置样品收集盘以捕获颗粒进行MDRS分析。随后,研究人员把收集到的颗粒转移至Morphologi G3-ID设备上(英国马尔文仪器有限公司产品)进行分析。


The deposited particles were initially characterised by automated image analysis to gather morphological information. From this analysis more than 1500 particles were then selected for chemical identification on the basis of their size and shape. Raman spectra were gathered for each of the selected particles and compared with a reference library containing spectra of the pure APIs and of lactose, an additional component in this formulation. This enabled secure chemical classification of the particles in the population of interest. A correlation score of 1 indicates that the particle has a spectrum closely matched to one of the pure components, while a lower score closer to zero, suggests an absence of that particular component.



Figure 2 shows an example spectrum for a specific particle, alongside reference spectra for the two active ingredients, and associated correlation scores. These scores suggest that this particle is amulti-component agglomerate (MCA), with clear spectral features associated with both APIs being observed within the particle’s spectrum. In a similar way, MCAs containing one or both APIs with lactose can also be identified.



Figure 3 shows chemical identification data for the entire particle population of interest. These data show that of the particles captured on stage 3, and subsequentlyclassified as of interest, only around 1% are API 1 alone. The majorityare API2 but there are appreciable numbers of discrete lactose particles and multicomponent agglomerates. Example particle images are shown for each type of identified particle.


These data provide far more insight into how the formulation is dispersed than does cascade impactionwith HPLC, which simply generates averaged chemical compositions for each size fraction. For example, they show whether both of the APIs detach from the lactose with equal ease and whether or not the APIs are likely to co-locate in the lung, on the basis of their size. This information is extremely helpful for understanding how a DPI delivers clinical efficacy and in demonstrating bioequivalence.



Working towards efficient de-formulation


De-formulation arguably presents an even greater challenge than formulation, most especially for complex products such as DPIs. Honing the analytical toolkit for DPI de-formulation is therefore essential. The traditional tools for inhaled product characterisation – cascade impaction and microscopy – have established benefits, but alternative techniques also have much to offer. Bringing speedand/or greater informational productivity, laser diffraction particle sizing, automated imaging and Morphologically Directed Raman Spectroscopy support the application of QbD and the efficient development of ANDAs.






[2]Guidance for industry. Metered Dose Inhaler and Dry Powder Inhaler Drug Products. Chemistry, Manufacturing and Control Documentation. Available for download at:http://www.fda.gov/downloads/Drugs/Guidances/ucm070573.pdf

[3] Kippax et al, “Unlocking the Secrets of the Dry Powder Inhaler Plume”, Proc. Drug Delivery to the Lungs 17, Edinburgh, 2006

[4] Huck D., Kippax P., Virden A., Kinnuen H., Shur J., Price R: “Improved understanding of the physical properties of DPI formulations by combining NGI size classification with Automated Image Analysis.” Drug Delivery to the Lungs 22, 2011, Conference Proceedings, 165-168.



Figure 1: Powder concentration (Cv) versus Time profiles recorded for the emptying of a dry powder inhaler using laser diffraction. The data show that a higher concentration is delivered at high flow rates. This relates to the energy available for aerosolisation of this cohesive formulation.



 Figure 2: Comparing the Raman spectrum of individual particles with reference spectra for the APIs in the formulation enables secure chemical classification of the sample.




Figure 3: MDRS enables the precise classification of particles in an NGI sized fraction as being API, lactose or multicomponent agglomerates.



中国国际粉体加工/散料输送展览会(IPB 2015)

展位号:A 1508

地点:上海跨国采购会展中心 (上海市光复西路2739号)