Multivariate analysis in the pharmaceutical industry: enabling functional excipient parameters data into knowledge 01 Sunil Kumar N A century of product innovations Milled & sieved lactose Spray-dried lactose Anhydrous lactose Croscarmellose sodium Agglomerated anhydrous 1900 1960 1980 1990 2000 2010 2016 2017 Sodium starch glycolate Inhalation grade lactose Agglomerated lactose Partly pregelatinised starch Microcrystalline cellulose SuperTab 40LL 02
Quality is Multivariate 08 QbD: Increasing Focus on Excipient Variability Understanding the true variability of an excipient requires thinking beyond the Certificate of Analysis (C of A). What is impact on finished product quality? 04
QbD: More than a collection of TLAs! 03 Jennifer Maguire & Daniel Peng-presentation on How to Identify Critical Quality Attributes and Critical Process Parameters. FDA/PQRI 2nd Conference North Bethesda, Maryland October 6, 2015 Understanding excipient variability informs robust product design Regulators expect joint due-diligence with suppliers Justify reliance on pharmacopoeial specification Supplier collaboration to mitigate risk from excipients Consistency Historical C of A and in-process data Potential application-specific Critical Material Attributes (CMAs) Process Capability Unspecified attributes 05
Why are excipients important? Excipients: often constitute majority of formulation. enable safe efficacious dosage forms variability can impact finished product performance some more critical than others no such thing as non-critical excipient Ishikawa risk analysis model 06 Excipients are not like APIs Excipient companies are not Pharma companies and excipients are not like APIs. Description Pharma company Excipient company # of products Many Products Few Products Operations High volume Small volume batch continuous Composition Single synthetic Extracted multicomponent A good overview is given by the series of articles from Chris Moreton Functionality and Performance of Excipients in a Quality-by-Design World, 2009, Am. Pharm.Rev. Supplement 07
Univariate vs Multivariate Multivariate analysis (MVA) Explanation of PCA (Principal Component Analysis) Statistical tool to evaluate large data sets. Replaces multiple univariate control charts Two charts per product Score plot may show trends or clustering. Loading plot shows vectors driving differences in Score plot. 09 Example Excipient PCA score plot Hotellings T2 Plot Spray Dried Lactose (SuperTab 11SD) 200 batches 14 Parameters C of A plus in-house measurements PCA score plot highlights differences but not their magnitude or significance. 10
What drives Multivariate scatter. By utilizing loading scatter plots, the variation of individual parameters can be assessed. 11 Consistent year-on-year Production PCA plot coded by year of manufacture Absence of trend shows consistency of manufacture over several years Batches on within elipse represent extremes of multivariate quality More appropriate than requests for infeasible edge-of-specification batches 12
How your supplier can support QbD Data Samples Expertise 13 What Data can your Supplier provide? C of A Historical C of A data In process data Process Capability Other attributes Multivariate Analyses 14
What Qbd batches or Samples can your supplier provide? Adjacent grades for bracketing studies Material out of specification for parameters of interest Non- Pharmaceutical grades Experimental product 15 Does your supplier have In house Expertise? In-depth Material science Process knowledge High volume continuous manufacture is QbD Application support Regulatory Information Troubleshooting Unknown does not mean Unknowable! 16
Do not assume you know everything about excipients 17 https://ojs.abo.fi/ojs/index.php/jefc/article/view/190 Case studies of Excipient Variability in IR tablets 08
Impact of excipient variability in IR tablets Article 1: Kushnar et al supports examining excipient variability as part of the design and control strategy even for a IR tablet dosage form! Article 2: Kushnar et al suggests that results of manufacturability and performance of IR tablet is robust to a broad range of variation in drug properties, both within-grade and extra-grade excipient particle size variations, and the choice of manufacturing method. 18 Raw material property ranking not always seen in finished tablet MCC Raw Material MCC Tablets
Excipient Kano Analysis Satisfaction Non-critical becomes very critical if threshold crossed Non-Performance Performance Dissatisfaction Contact Information For more information please contact https://www.dfepharma.com/ Log in your future questions in Ask an expert section https://www.dfepharma.com/sitecore/co ntent/dfe-website/home/contact/ask-anexpert 20