Hier “rcpe ppt header Large 01.

emf” platzieren

Qualiy-by-Design

CHE 833

Univ.-Prof. Dr. Sven Stegemann, Technische

Universität Graz

K1 Competence Center - Initiated by the Federal Ministry of Transport, Innovation and Technology (BMVIT)
and the Federal Ministry of Science, Research and Economy (BMWFW).
Funded by the Austrian Research Promotion Agency (FFG), Land Steiermark and the Styrian Business
Promotion Agency (SFG).
Table of content

▪ Introduction ▪ Lipid nanoparticles & mRNA

▪ Future of manufacturing ▪ Purification and franctionation in up-
▪ Quality-by-design and economy and downstream processing

▪ Quality-by-Design (QbD) in ▪ Parenteral dosage forms

pharmaceutical sciences ▪ Lyophilized dosage forms
▪ Statistical tools used in QbD ▪ Cell and Gene Therapy (CGT)
▪ Design of Experiments (DoE) ▪ Medical devices and combination
▪ Process Analytical Technology (PAT) products

▪ QbD for special dosage forms ▪ Conclusion

Introduction
Introduction

“It is important to recognize that quality cannot be tested into products, i.e.,
quality should be built in by design.”
Introduction
Product development cycle
Also Biotech is not performing well

Drug characterization
& optimization
Preformulation
(enabling F, MST, PIC)
Formulation and process
development Manufacturing transfer
Introduction

Drug
Development
path
FDA approvals 1998 - 2022
Introduction
2022 figures of FDA approvals
▪ The traditional drug development and product portfolio continues to change towards smaller
patient populations with unmet medical needs

https://www.fda.gov/media/164429/download?attachment
Introduction
▪ Newly approved drugs by the FDA 2015-2019
Introduction
Changing drug research and approval portefolio

FDA approvals 2022 by

therapeutic class

Senior et al Nature 41:174-82 (2023)

BCG & VFA: Medizinische Biotechnologie in Deutschland 2023

Introduction
Changing drug research and approval portefolio

BCG & VFA: Medizinische Biotechnologie in Deutschland 2023

Introduction

R&D spendings
per phase

The cost of
researching and
developing a new
chemical or
biological entity was
estimated at €1,926
million ($2,558
million in year 2013
dollars) in 2014

DiMasi et al, Journal of Health Economics, January 2016

Introduction
What is quality?

▪ The standard of something as measured against other things of a similar kind; the degree of
excellence of something.
▪ General excellence of standard or level.

What is design?

▪ A plan or drawing produced to show the look of function or workings of a building, garment, or
other object before it is made.

▪ The art or action of conceiving of and producing a plan or drawing of something before it is made.
Introduction

What is Quality by Design (QbD)?

▪ A concept introduced to the pharmaceutical

industry about 20 years ago
Introduction Initiate
Quality Risk Management Process

Risk Assessment

Risk Identification

Risk Analysis

Risk Evaluation
“Quality by design is an approach that aims to unacceptable

Risk Management tools

Risk Communication
ensure the quality of medicines by employing Risk Control

Risk Reduction
statistical, analytical and risk-management
Risk Acceptance
methodology in the design, development and
manufacturing of medicines” Output / Result of the
Quality Risk Management Process

Risk Review

Review Events
Introduction

▪ “One of the goals of quality by design is to ensure that all sources of variability affecting a

process are identified, explained and managed by appropriate measures”

▪ “Quality by design centers on the use of multivariate analysis, often in combination with modern

process-analytical chemistry methods and knowledge-management tools to enhance the

identification and understanding of critical attributes of materials and critical parameters of the
manufacturing process”
Introduction

▪ In March 2011, the European Medicines Agency (EMA) and the United States Food and Drug
Administration (US FDA) launched, under US-EU Confidentiality Arrangements, a joint pilot
program for the parallel assessment of applications containing Quality by Design (QbD)
elements
▪ The Japanese Pharmaceuticals and Medical Devices Agency (PMDA) participates as an
observer.
Future of manufacturing
Future of manufacturing
American Pharmaceutical Review · November 2004

▪ ...a journey ongoing for > 20 years

▪ ...progressing faster than ever
Future of manufacturing

▪ Starting more than 100 years ago...

Traditional manufacturing
was dedicated to achieve the
desired high pharmaceutical
product quality in terms
reproducible performance
and stability. It was assured
by 3 consecutive batches
meeting the set specification
before “freezing” the process
Future of manufacturing

▪ With regard to efficiency the St. Gallen Report concluded in 2006

• There is no clear understanding in the industry about their cost structure in
manufacturing
• 83 % of the companies expect increasing heterogeneity in customer demand
• Low Overall Equipment Efficacy (OEE) correlates with increasing number of unit
operations (much higher efficiency with standard equipment)
• Importance of employee training in cost efficiency
• Short cycle times are of critical importance for manufacturing flexibility
• High quality today achieved by “good inspection” rather than a “good process”
• High inter-correlation of each element in integrated systems causes unplanned
maintenance, work-in-progress stocks and rejections
• Operational improvement programs must focus on the entire manufacturing process
• R&D is key in designing the efficient product and process

Operational Excellence in the Pharmaceutical Industry – St Gallen Report (2006)

Future of manufacturing

▪ With regard to efficiency the St. Gallen Report concluded in 2006

• There is no clear understanding in the industry about their cost structure in
manufacturing
• 83 % of the companies expect increasing heterogeneity in customer demand
• Low Overall Equipment Efficacy (OEE) correlates with increasing number of unit
operations There
(much higher
is noefficiency with standard equipment)
clear understanding in the industry
• Importance ofabout
employee training
their costin structure
cost efficiency
in manufacturing
• Short cycle times are of critical importance for manufacturing flexibility
• High quality today achieved by “good inspection” rather than a “good process”
• High inter-correlation of each element in integrated systems causes unplanned
maintenance, work-in-progress stocks and rejections
• Operational improvement programs must focus on the entire manufacturing process
• R&D is key in designing the efficient product and process

Operational Excellence in the Pharmaceutical Industry – St Gallen Report (2006)

Future of manufacturing

The model to estimate Cost of Good Sold (COGs)

COGs = (Daily Dose x Bulk Cost) + Formulation cost x 100

Daily Selling Price

Formulation costs: Excipients + drug

Bulk cost: Overheads

This assumes that overheads are shared equally across all products
manufactured in one plant.
Future of manufacturing
▪ Cost structure of pharmaceutical manufacturing per unit operation
• GMP space
• Number of surrounded equipment
• Energy
• Solvent/water/detergence
• HVAC
• Validation/qualification
• QA/QC/QbD Mainly related
• Operators/planning to drug substance

• Cleaning/waste
• Maintenance
• Yield losses
• Inventory/JIT implication
• Depreciation
• Equipment effectiveness (utilization time)
• Safety stocks
Operational Excellence in the Pharmaceutical Industry – St Gallen Report (2006)
Future of manufacturing
▪ Cost structure of pharmaceutical manufacturing per unit operation
• GMP space
• Number of surrounded equipment
• Energy
• Solvent/water/detergence
Each unit operation adds significant
• HVAC ‘overhead’ costs to manufacturing.
• Validation/qualification
• QA/QC/QbD Mainly related
to drug substance Mainly related
• Operators/planning
Sharing the total plant overheads equally to drug substance

• Cleaning/wasteacross all products is highly misleading.

• Maintenance
• Yield losses
• Inventory/JIT implication
• Depreciation
• Equipment effectiveness (utilization time)
• Safety stocks
Operational Excellence in the Pharmaceutical Industry – St Gallen Report (2006)
Future of manufacturing

▪ There is an increasing awareness of the cost drivers in pharmaceutical manufacturing

• Energy costs
• Cycle times (Work in progress, inventory, stock,...)
• Cost for capital
• GMP space (4.000 – 12.000 $ / m2)
• Ecological costs (waste, CO2, etc)
Distribution of energy use in the pharmaceutical
industry
Overall Plug loads and Lighting Heating, Ventilation
(100 %) processes and air
conditioning
(HVAC)
Total 100 % 25 % 10 % 65 %
R&D 30 % Equipment GLP
Manufacturin 50 % Processes GMP
g

Galitsky et al 2006
Future of manufacturing
Energy Utilization index across 3 different pharmaceutical manufacturing
plants [kWh/m2/y]

Formulation unit 1 Formulation unit 2 Formulation unit 3

White Gray Black White Gray Black White Gray Black

6513 4032 69 4050 1526 122 8759 1490 225

Witty said in the interview with the Times he expects the GSK energy bill
to rise 20% to 30% next year.

Probing every possible cost-savings opportunity, GSK is considering

removing its factories from the local power Grid to supply its own
energy more cheaply.
Andrew Witty, September 2011
Zhang 2009
Future of manufacturing
Commitments of the pharmaceutical industry in times of climate change

Commitments to percentage greenhouse gas (GHG) emission reduction in scope 1, 2, and 3 emissions
by a range of target years by 16 pharmaceutical companies that made them
Booth et al Int. J. Environ. Res. Public Health 2023, 20, 3206.
Future of manufacturing
▪ Work in Progress (WIP) depends on
• the number of unit operation
• the number of components

▪ Safety stocks depends on

• Complexity of manufacturing
• the supply chain
• 3rd party involvement

▪ Cycle times
• 46 calendar days for raw materials
• 25 calendar days for WIP
• 91 calendar days Finished Goods Inventory
• 162 calendar days of total cycle time
• < 6 calendar days of this cycle time (3.7 %) is related to value
added
Cogdill et al J Pharm Innov (2007) 2:38–50
Future of manufacturing
▪ Overheads associated with inventory carry includes items such as
expiration loss, cost of facilities, insurance, paperwork, transportation,
physical handling, spillage/damage and theft/pilferage
▪ Applies to any incoming good including WIP and safety stocks
▪ Estimated at 20 – 25 % on the costs of the component
▪ Weighted average cost of capital (WACC) is a company- specific measure of the cost of internal
allocation of capital
▪ Estimated to be 7.5 %
▪ Cost for safety stocks (example) annual turn over $ 240 mio
▪ 40 batches à $ 6 mio
▪ Safety stock (10 batch)~ $ 60 mio
▪ $ 4.5 mio cost for safety stock/year

Cogdill et al J Pharm Innov (2007) 2:38–50; https://pwc-tools.de/kapitalkosten/en/healthcare-pharmaceuticals/

 Future of manufacturing
▪ Overall Equipment Effectiveness (OEE) depends on
• the number of unit operations Key Pharmaceutic A world-class
• their process uniqueness performance al industry manufacturing
• Low OEE has a low amortization and poor ROI indicators (2005) plant
OEE (%) 30 92
Cycle times (h) 720 8

▪ OEE is impacted by
• Stoppages and breakdowns
• Unplanned maintenance
• Down times and set up times
• Inter-correlation with other
process parts

IBM The metamorphosis of manufacturing 2005; Aigner et al 2017

Future of manufacturing

▪ Traditional manufacturing moved to lean manufacturing

Lean manufacturing
involves never ending efforts
to eliminate or reduce waste
or any activity that consumes
resources without adding
value in design,
manufacturing, distribution,
and customer service
processes.
Future of manufacturing
▪ Operational excellence drives lean manufacturing
• Lean manufacturing have been introduced into the pharmaceutical manufacturing over the
past decade to increase effectiveness and efficiency
• A review of the outcomes between 2004 – 2009 were disappointing

Total Productive Maintenance (TPM)

Total Quality Management (TQM)
Just-In-Time (JIT)

Friedli et al J Pharm Innov 2010

Future of manufacturing
Results of operational excellence and lean manufacturing efforts

• Unplanned maintenance increased from 25 % to 33 %

• Use of new technology remains very low 58 % vs 57 % considered new technology
• Rejection rate of finished products went down from 1 % to 0.74 %,
• Supplier complaints rise from 1 % to 2.4 % (raw material turn-over increased from 4 to 5
times/year), while pull production increased from 44 % to 59 %
• Set up time went up from 79 min to 93 min
• Cycle times from weighing to packaging increased from 22 days to 22.7 days
• Order fulfillment (right quantity & quality) remained at 95 %
• Lay-out optimization decreased from 68 % to 57 % impacting product flow and tact (complexity
of material handling & storage as well as process interruptions)
• Pharmaceutical manufacturing still struggles with effectiveness and can not focus on efficiency

Friedli et al J Pharm Innov 2010

Future of manufacturing

Conclusion

▪ “The US pharmaceutical industry could be wasting more than $50bn (€39bn) per year
in manufacturing costs due to inefficient processes.” (Macher and Nickerson 2006)

▪ “Manufacturing defects in fact account for almost three-quarters of all drug recalls.”
(Dean, Bruttin and van Dyck, April 2005 (Source: U.S. Food and Drug Administration
Enforcement Reports, 2000 –2004).
Future of manufacturing

▪ There is an emerging challenge of falsified medicines and drug shortage worldwide,

including Europe and USA

Prevalence of incidents involving health damage due

to falsified medicines by region (n, %). Total number
of incidents is, n = 48.
Rahman et al Trop Med Int Health 23: 1294–1303 (2018)
Future of manufacturing

▪ Benchmarking to other industries the pharmaceutical industry was definitive an

underperformer

Trout & Bisson 2009

Future of manufacturing

▪ Medical advances drive treatment towards personalized drug products and healthcare
interventions and hence towards (mass)customization

Dose by Different
number of mini-tableted
mini-tablets drugs

2 mg mini-
tablets

Berry et al J Manufact Systems 32: 404– 411 (2013)

Future of manufacturing

▪ The majority of future new drugs are orphan drugs (defined as: USA: <1/200.000 people;
EU: < 10.000 people in Europe)

Source: IQVIA analysis: FDA CDER, EMA NCEs 2019 data

Future of manufacturing

▪ The majority of pharmaceutical manufacturing is still mass manufacturing based on batch

processes, while other industries have moved to mass customization using continuous
processing
▪ „Regulatory agencies, including the FDA, EMA and PMDA support the adoption of CM for
pharmaceutical production based on science- and risk-based approaches.“ S. Lee, FDA
Deputy Director & Emerging Technology Team Chair, March 2017
▪ The selection of the manufacturing process is based on three major components that need to
be addressed in the pharmaceutical industry

Manufacturing costs
Development costs Lifecycle /service
• Component
• R&D expenses costs
costs
• Product and • Supply chain
• Plant &
process costs
operational
development • Market costs
costs
Future of manufacturing

▪ Product & process complexity is

Critical interface
already determined by the early Process
Material properties,
and late stage product characteristics,
characteristics,
development mechanistics,
functionalities
trajectory,
▪ QbD aims to understand the variability
variability
multifactorial interdependence
between materials and processes Formulation Process
to mitigate the risk development development
▪ Preferably, QbD should determine Product characteristics
the formulation and process with and performance
the least risk and prevent „over-
engineering“ consistent &
reproducible
Future of manufacturing
▪ FDA recognizes that CM is an emerging technology that
can enable pharmaceutical modernization and deliver
potential benefits to both industry and patients.
▪ CM can improve pharmaceutical manufacturing by, e.g.
an integrated process with fewer steps and shorter
processing times; smaller equipment footprint; enhanced
development approach (e.g., QbD) and use of PAT and
models; enabling real-time product quality monitoring;
and providing flexible operation to allow scale-up, scale-
down, and scale-out to accommodate changing supply
demands.
▪ FDA expects that adopting CM for pharmaceutical
production will reduce drug product quality issues, lower February 2019
manufacturing costs, and improve availability of quality
medicines to patients
Future of manufacturing
Definitions
▪ Continuous Manufacturing: An integrated process that consists of a series of two or more unit-
operations. In such a process, the input material(s) are continuously fed into and transformed
within the process, and the output materials are continuously removed from the system
▪ Advanced manufacturing: is a systematic predictive framework of methodologies than enable the
optimal implementation and control of a manufacturing process
▪ Batch: A specific quantity of a drug or other material that is intended to have uniform character
and quality, within specified limits and is produced according to a single manufacturing order
during the same cycle of manufacture
▪ Residence Time Distribution (RTD): A probability distribution that describes the amount of time a
mass or fluid element remains in a process
▪ Control Strategy: A planned set of controls, derived from current product and process
understanding that assures process performance and product quality
▪ Continued Process Verification: Assurance that during routine production the process remains in
a state of control
Future of manufacturing

▪ Continuous manufacturing have been installed in nearly all major pharmaceutical

companies
▪ Main continuous manufacturing line suppliers are GEA and Bohle

https://www.continuous-
production.com/continuous-manufacturing

https://www.gea.com/en/news/insights/2
018/from-batch-based-to-continuous-
manufacturing.jsp
Future of manufacturing
Future of manufacturing

Continuous manufacturing advantages from early to late stage development & manufacturning

▪ Process development according to QbD principles enabled with limited drug substance
▪ Support fast track development programs through early finalization of formulation and
processing
▪ Product consistance throughout the clinical and commercial manufacturing
▪ Avoidance of product and process scaling up phase
▪ Highly consistant product quality and performance through continuous close loop monitoring
▪ High flexibility in manufacturing scale and „batch size“
▪ Ease of manufacturing transfer for commercial manufacturing or supply chain
▪ Smaller manufacturing footprint, equipment, capital cost and utility requirements
▪ Real-Time-Release Testing (RTRT)
Future of manufacturing

Continuous manufacturing disadvantages from early to late stage development & manufacturning

▪ Continuous manufacturing line accessability and capital investment

▪ Thorough product and process understanding (QbD) to be generated during early development
(incl. a number of DoEs)
▪ Identification and implementation of control strategies (temporary)
▪ Extensive cleaning and cleaning validation program
▪ Availability of the required skills (temporary)
▪ Limited regulatory guidance (temporary)
Future of manufacturing

Comparison between a batch and continuous manufacturing processing

 Future of manufacturing

Stages of data transformation on the path to

Data are transformed from raw
realizing Industry 4.0
signals captured from a system to full
digital maturity. Data are initially
collected from a manufacturing
process, then organized by data
digitization and analysis as Big Data
into information, then synthesized into
knowledge by the meaning discerned
via artificial intelligence, and finally to
actionable wisdom attained through
the combined insights of digital
maturity

Arden et al Int J Pharm 602 (2021) 120554

Future of manufacturing
A cyber-physical system (CPS) for pharmaceutical manufacturing in Industry 4.0

Arden et al Int J Pharm 602 (2021) 120554

Quality-by-Design (QbD) and economy
Future of manufacturing

Considerations for continuous manufacturing

▪ Evaluation of different manufacturing processes
▪ Factors to calculate process chain costs

Aigner et al Chapter 14 in: Continuous Manufacturing of Pharmaceuticals 2017

 Future of manufacturing
Evaluation of different manufacturing processes
▪ In the early stages
product and process
specific assumptions
have to be made,
which might require
qualitative and semi-
quantitative validation

▪ Aigner, I., Hsiao, W-K., Dujmovic, D.,

Stegemann, S. Khinast, J. Methodology
for economic and technical comparison of
continuous and batch processes to
enhance early stage decision-making. In:
Continuous Manufacturing of
Pharmaceuticals, Eds: Kleinebudde, P.,
Khinast, J., Rantanen, J. pp 485-505. John
Wiley & Sons Ltd. (2017)

Aigner et al Chapter 14 in: Continuous Manufacturing of Pharmaceuticals 2017

Future of manufacturing

Considerations for continuous manufacturing

Evaluation of the sytem boundaries and performance
targets (manufacturing efficiency) is specific for a
product and include:
• Chosen production capacity of the plant
• Analysis of the process in comparison to alternatives
• Long term production planning
• Supply chain forcast
• Number of shifts

Aigner et al Chapter 14 in: Continuous Manufacturing of Pharmaceuticals 2017

Future of manufacturing

Considerations for continuous manufacturing

Modeling of the process chain includes:
• Unit operations
• Main and auxiliar equipment for process Main and Overall
integration auxiliar Equipment
equipment effectiveness
• Batch size or throughput rate
• Overall equipment or processing times

Aigner et al Chapter 14 in: Continuous Manufacturing of Pharmaceuticals 2017

 Future of manufacturing

Considerations for continuous

manufacturing
Performing a technical feasibility and risk
assessment include
• Calculated risk associated with a product
or process to not meet desired targets
• Done based on criteria specific, predefined
risk scores for unknowns
• Weighing factors are being used to based
on the importance of a criteria
Major criteria for feasibility and risk assessment

Aigner et al Chapter 14 in: Continuous Manufacturing of Pharmaceuticals 2017

Future of manufacturing
Considerations for continuous manufacturing
Evaluation of the process options is derived from the
feasibility and risk assessment
• Establish the required equipment including auxiliar
and control (PAT) tools
• Calculate the direct and indirect costs for materials
and manufacturing
• Simulation and estimation of approximate direct and
indirect costs using e.g. flow sheet modeling
• Process simulation to based on prior knowledge
• Cost calculation based on different models (e.g.
Product costing with activity unit) and equipment cost
calculations based on exisiting knowledge and
factors (e.g. Total planned cost estimate, Lang factor)

Aigner et al Chapter 14 in: Continuous Manufacturing of Pharmaceuticals 2017

Future of manufacturing

Considerations for continuous manufacturing

Technology-Economic profiling for the
manufacturing process decison
• The highest scored process can be derived from
the process evaluation sheet summarzing all
results
• For further precision, Sensitivity and Uncertainty
analysis can be performed as well as laboratory
scale test runs to narrow down certain
assumptions and predictions

Aigner et al Chapter 14 in: Continuous Manufacturing of Pharmaceuticals 2017

Future of manufacturing
 Future of manufacturing
Portable
Skid-mounted equipment
High Small autonomous POD‘s
Modularity Rapid deployment & Re-deployment
Multiple configurations, Rapid Multi-company POD* farms
Flexibility & Adaptability

process/product changeover,
Miniature Accept ‚next-gen‘ technologies
Small footprint
Enable portable design
Reduced energy consumption
Enable modularity

Continuous manufacturing Transformational

Variable lot size for demand-based supply.
Same equipment at multiple scale development,
Enable miniaturization manufacturing &
distribution modeles
Batch manufacturing
Low
Established Current state of the art Future/Emerging
* POD: Point of Distribution

Phil Nixon (Pfizer) AAPS Ardenhouse Conference March 16-18, 2015

 Future of manufacturing
State of the art continuous manufacturing lines

e.g. ConsiGma 25 e.g. ConsiGma CDC

Modular manufacturing suite which can be Direct compression line equipped with Lighthouse
equipped with blending, milling, wet/dry Probe (Fibre optic probe (Reflectant (UV-VIS-NIR),
granulation, compression, coating scattering, raman, fluorescence, video imaging)
Future of manufacturing
The ConsiGma 25 is equipped with a variety of PAT tools for online control of the CQA and CPP

Continuous manufacturing lines are installed at Pfizer Groton, Vertex and others
Quality-by-Design (QbD) in pharmaceutical
sciences
QbD in pharmaceutical sciences

Foundation of the Quality-by-Design as manufacturing science

The basic concept starts with a definition of who the users will be and need, what
this means for the product design and how this can be converted into a reproducible
product quality

Juran JM (1992), Juran on Quality By Design, Bass J Ed.

QbD in pharmaceutical sciences

Foundation of the Quality-by-Design as manufacturing science

This translates into a Targeted Product Profile (TPP), the Critical Qualiy Attributes
(CQA), the Critical Process Parameter (CPP) and the Design Space in which product
quality is guaranteed
Design
Space

CPPs
CPPs

CQAs
CQAs
TPPs
Patients

TPPs

Juran JM (1992), Juran on Quality By Design, Bass J Ed.

 QbD in pharmaceutical sciences

Foundation of the
Quality-by-Design as
manufacturing science

ISPE PQLI Guide, 2011 (Pharm. Eng. (2010),30(2),1-6

QbD in pharmaceutical sciences

?
Formulation Transferred to
Development Manufacturing
Formulation, process,
Laboratory Pilot plant 1 kg
specification, QC frozen
vs
1000 kg
Formulation and (process)
already fixed Transferred to Transferred to
launch site manufacturing
site

Continuous
improvement
QbD in pharmaceutical sciences
Q8 Pharmaceutical Development May 2006
❖ Quality by Design
Q8 R1 Pharmaceutical Development Nov 2007
The regulatory Framework

❖ Annex to ICH Q8 (practical guidance)

Q8 R2 Pharmaceutical Development Nov 2009
❖ Annex to ICH Q8 (suggested contents for the 3.2.P.2)
Q9 Quality Risk Management June 2006
❖ Integrated Risk management system
Q10 Pharmaceutical Quality System May 2007
❖ Integrated quality management system
Q11 Development and manufacturing of drug substances May 2012
❖ Application of Q8-10 to drug substance
Q12 Technical and regulatory considerations for life cycle management Sept 2014
❖ Framework to facilitate life cycle management
Q13 Continuous Manufacturing of Drug Substances and Drug Products Nov 2022
 QbD in pharmaceutical sciences
Critical Quality Attribute (CQA)
A physical, chemical, biological or microbiological property or characteristic that should be within an appropriate
limit, range, or distribution to ensure the desired product quality for the finished product
Critical Material Attributes (CMA) often refers to CQA of excipients
Critical Process Parameter (CPP)
A process parameter whose variability has an impact on a critical quality attribute and therefore should be
monitored or controlled to ensure the process produces the desired quality
Quality by Design (QbD)
A systematic approach to development that begins with predefined objectives and emphasizes product and
process understanding and process control, based on sound science and quality risk management
Target Product Profile (TPP)
A prospective summary of the quality characteristics of a drug product that ideally will be achieved to ensure the
desired quality, taking into account safety and efficacy of the drug product
QbD in pharmaceutical sciences
ICH Q 8
“Critical sources of variability that can lead to product failures should be identified, appropriately understood,
and managed or controlled.”

ICH Q 9
“Quality risk management is a systematic process for the assessment, control, communication and review of
risk to the quality of the drug product across the product life cycle.”

ICH Q 10
“Systematic approach to acquiring, analyzing, storing and disseminating information related to products,
processes and components across the product life-cycle.”

ICH Q 9 defines quality as:

“…the protection of the patient by managing the risk to quality should be considered of prime importance.”

ICH Q 8 R2 defines quality as:

“In all cases, the product should be designed to meet patients’ needs and the intended product
performance.”
QbD in pharmaceutical sciences

Opportunities to impact
risk using quality risk
Design management Q9
Process

Materials Manufacturing

Facilities
Distribution

Patient

Q8 Q10
G.- Claycamp, FDA, June 2006
QbD in pharmaceutical sciences
Target Product Profile

Drug substance properties; prior knowledge

Proposed formulation and manufacturing process

(Risk Identification)

Cause and effect process

(Risk Analysis)

Risk-based classification
(Risk Evaluation)

Proposed Parameters to investigate (e.g. by DOE)

(Risk Reduction)

FORMULATION PROCESS DESIGN

DESIGN SPACE CONTROL STRATEGY SPACE BY UNIT
(Risk Reduction)
OPERATION
QbD in pharmaceutical sciences

Relative prevalence of papers

according to the studied
factors:
▪ Formulation Attributes (FAs)
▪ Process Parameters (PPs)
▪ both (FAs + PPs)

[percentages are calculated with reference

to the total number of papers (n = 60)]

Grangeia et al EJPB 147 (2020) 19–37

QbD in pharmaceutical sciences

▪ Relative prevalence of the

Manufacturing Unit
Operations reported in the
research papers included in
this review

Grangeia et al EJPB 147 (2020) 19–37

QbD in pharmaceutical sciences
Risk Assessment

▪ Risk assessment consists of the identification of hazards and the analysis and evaluation of
risks associated with exposure to those hazards (as defined below).
▪ Quality risk assessments begin with a well-defined problem description or risk question.
▪ When the risk in question is well defined, an appropriate risk management tool and the types
of information needed to address the risk question will be more readily identifiable.
▪ As an aid to clearly defining the risk(s) for risk assessment purposes, three fundamental
questions are often helpful:
1. What might go wrong?
2. What is the likelihood (probability) it will go wrong?
3. What are the consequences (severity)?
QbD in pharmaceutical sciences
Initial Risk Assessment

▪ Basic risk management facilitation methods

▪ Process mapping
− provides a clear and simple visual
representation of involved steps
− facilitates understanding, explaining and
systematically analyzing complex processes
and associated risks
− a pre-requisite for the use of some other
methods (e.g. FMEA)
QbD in pharmaceutical sciences
Initial Risk Assessment

▪ Basic risk management facilitation

methods
▪ Cause and effect diagrams (Ishikawa/fish
bone)
▪ How to perform?
− Define and agree a precise problem
statement (put as “head” of fishbone)
− Evaluation “What could be the
problem’s causes?” in the main “bones
of the diagram”
− Evaluation of root causes of each bone
QbD in pharmaceutical sciences

IPO diagrams
input–process–output (IPO)
model, or input-process-
output pattern can be used to
structure the risk assessment
regarding CMA and CPP
QbD in pharmaceutical sciences

▪ Failure Modes and Effects Analysis (FMEA) is a systematic, proactive method for evaluating a
process to identify where and how it might fail and to assess the relative impact of different
failures, in order to identify the parts of the process that are most in need of change.
▪ There are Seven Steps to Developing an FMEA:

1. FMEA Pre-Work and Assemble the FMEA Team

2. Path 1 Development (Requirements through Severity Ranking)

3. Path 2 Development (Potential Causes and Prevention Controls through Occurrence Ranking)

4. Path 3 Development (Testing and Detection Controls through Detection Ranking)

5. Action Priority & Assignment

6. Actions Taken / Design Review

7. Re-ranking RPN & Closure

 QbD in pharmaceutical sciences

FMEA framework and structure to determine the risk

Function or
Failure Potential Potential Detection
Process SEV OCC DET RPN
Type Impact Causes Mode
Step

What are the

existing
Briefly What is the
How What How controls that
outline Describe impact on the How
severe is causes the frequently either prevent Risk
function, what has key output easy is it
the effect key input is this the failure priority
step or item gone variables or to
to the to go likely to from occurring number
being wrong internal detect?
customer? wrong? occur? or detect it
analyzed requirements?
should it
occur?
QbD in pharmaceutical sciences
QbD in pharmaceutical sciences
 QbD in pharmaceutical sciences

▪ Understanding the variability of materials & processes is key for consistent product quality

Raw Hpc_lf.M1
material quality history
(PLS), Untitled, Work set
Scores: t[1]/t[2]

0.30
Principal Component
47
90
147 146
Analysis (PCA) of NIR
66 62
0.20 81
86 105154
21
28
175 171
95
65 59
61
data for different batches
of the same “quality” of
183 176 174 143 99 100
63
142107 155189 104
185 12 1 64 96
179 82 144 102 60 97
20 126
0.10 182 89 186
45 168 41
134 137 85 23129 178 188
84
94
111 106
27 184
1355646167
93180
153
32 35 53
165
162 58
166
145
148
48
98 51
13
49 52
raw material supplied
112
127 36 169 42 101
103
0.00 150 113
132
26
33152116
67
108 24 160
38 177
76 181
110
17
5440
34 43 163
30
161
50 from two different
t[2]

92 87
88133 83 136172 55 57 157
194 117 170109
128
73
31
141
3
151 80
140 44 79 77
138
119
878 5
16415 29
16396 9
187
158
Vendor 2 vendors.
-0.10 130 195 14 74 22
118 173
139 37
190 7 475
192
10 68
156 124
14971 191
72
19
-0.20 Vendor 1 131
193 70
159
25
69 11 91
18 120
1252
-0.30 123 121

122 Vendor 1 OK
-2 -1 0 1 2
t[1]
Vendor 2 not OK
Note that all batches comply with pharmacopoeia standards!

• J. Vessman, Pharm. Anal., Overview, in Encyclopedia of Anal. Sciences, Academic Press Limited, 1995, p. 3807.
• O. Svensson, M. Josefson, F.W. Langkilde, Applied Spectroscopy, 51 (1997) 1826-1835.
QbD in pharmaceutical sciences

▪ The QbT knowledge space is very limited and excipients batches outside the tested space
might fail
Hpc_lf.M1 (PLS), Untitled, Work set
Raw material
Scores: t[1]/t[2] quality
QbT
0.30
90
47
The 3 excipient
147 146
66 62
0.20 81
21
28 95
61 batches tested in ICH
86 105154 175 171
143 65 59

20
183 176
142107 155189 104
179 82
126
185
174
144 102
99 100
12 1 64 9663
60 97 are good
0.10 182 89 186
45 168 41 98 51
137 85 23129 178 188 48
134 111 106 1355646167
93180 165
162 58 145
148 13
84
94 27 184
153 35 53 166 49 52
112
127 32
36 169 42 101
103
113
132 67 24 160
17
0.00 150 26 108
38 177 161
33152116
76 181
110 5440 163
30 50
34 43
t[2]

92 57 157
194 128
87
88133 83 136172
117 170109
31
141
3
79 77
55
878 5
187
158
Vendor 2 QbT
73151 80 44 138 16415 29 9
-0.10 130 195 14 14074 22
190 7
119
118
475
192
16396
139 173
37
10 68
Batches out of this
156 124
14971 191
72
-0.20 Vendor 1 131
193
19
70
159
range are unknown
25
69 11 91
18 120
1252
-0.30 123 121

122 Vendor 1 OK
-2 -1 0 1 2
t[1]
Vendor 2 not OK
Note that all batches comply with pharmacopoeia standards!

• J. Vessman, Pharm. Anal., Overview, in Encyclopedia of Anal. Sciences, Academic Press Limited, 1995, p. 3807.
• O. Svensson, M. Josefson, F.W. Langkilde, Applied Spectroscopy, 51 (1997) 1826-1835.
QbD in pharmaceutical sciences

▪ The QbD knowledge space is broad including the point of failure

Hpc_lf.M1 (PLS), Untitled, Work set
Raw material
Scores: t[1]/t[2] quality QbD
0.30
Measuring the breath
47
90
147 146
of specifications
66 62
0.20 81
86 105154
21
28
175
176
171
143
95
65 59
61
confirm suitable
183 174 99 100
63
12 1 64 96
0.10 20
142107 155189 104
179 82
126
182
185
89 186
144 102 60 97 range
137 85 23129 178 188 45 168 41 48
98 51
134 111 106 1355646167
93180 165
162 58 145
148 13
84
94 27 184
153 35 53 166 49 52
32
0.00 150
112
127
113
132
26
33152116
3667
108 24 160
38 177
76 181
110
169
17
5440
42
163
30
161
101
103
50
QbD
34 43
t[2]

92 87 172 55 57 157
194 88133 83 136
117 170109
128
73
31
141
3
151 80 44 79 77
138
878 5
16415 29 9
187
158
Vendor 2 The evaluation go
-0.10 140 119 16396
130 195 14 74 22
190 7
118

156
475
192
139
124
149
173
37
10
71 191
72
68 beyond the point of
19
-0.20 Vendor 1 131
193 70
159
25
failure
69 11 91
18 120
1252
-0.30 123 121

122 Vendor 1 OK
-2 -1 0 1 2
t[1]
Vendor 2 not OK
Note that all batches comply with pharmacopoeia standards!

• J. Vessman, Pharm. Anal., Overview, in Encyclopedia of Anal. Sciences, Academic Press Limited, 1995, p. 3807.
• O. Svensson, M. Josefson, F.W. Langkilde, Applied Spectroscopy, 51 (1997) 1826-1835.
 QbD in pharmaceutical sciences

▪ The significant difference between QbT and QbD

QbD
Hpc_lf.M1 (PLS), Untitled, Work set
Raw material quality
Scores: t[1]/t[2] Knowledge space
0.30 47
90
147 146
21 66 62 61
0.20 81 28 95
86 105154 175
176
171
143 65 59
183 174 99 100

0.10 20
142107 155189 104
179 82
126
185 12 1 64 96
144 102
63
60 97 QbD
182 89 186
137 85 23129 178 188 45 168 41 98 51
134
84
94
111 106
27
112
127
184
153
32
36
1355646167
93180
35 53
169
165
162 58
166
145
148
101
103
48
13
49 52
Design Space
113 67 24 160 42
150 132
26 108 17 161
0.00 116 38 177
76 181 5440
33152 110
34 43 163
30 50
t[2]

92 87
88133 83 136172 55 57 157
194 117 170109
128
73
31
141
3
151 80
140 44 79 77
138
119
878 5
16415 29
16396 9
187
158
Vendor 2
-0.10 130 195 14 74 22
118 173
139 37
190 7 475
192
10 68
156 124
14971 191
72
19
-0.20 Vendor 1 131
193 70
159
25 QbT
69 11 91
18 120
-0.30 123
1252
121
Tested Space
122 Vendor 1 OK
-2 -1 0 1 2
t[1]
Vendor 2 not OK
Note that all batches comply with pharmacopoeia standards!

• J. Vessman, Pharm. Anal., Overview, in Encyclopedia of Anal. Sciences, Academic Press Limited, 1995, p. 3807.
• O. Svensson, M. Josefson, F.W. Langkilde, Applied Spectroscopy, 51 (1997) 1826-1835.
QbD as an enabler of continuous manufacturing
▪ The difference between standard test data and QbD data based on capsule weight distribution

Size 1 Overall - WGT - TRF method Size 1 Overall - WGT

LSL Target USL LSL Target USL

P rocess D ata O v erall C apability P rocess D ata O v erall C apability
LS L 71 Pp 2.76 LS L 71 Pp 1.07
Target 76 PPL 2.53 Target 76 PPL 0.98
USL 81 P P U 2.99 USL 81 P P U 1.16
S ample M ean 75.5844 P pk 2.53 S ample M ean 75.5836 P pk 0.98
S ample N 84 C pm 2.27 S ample N 8404 C pm 1.04
S tD ev (O v erall) 0.604108 S tD ev (O v erall) 1.55376

72.0 73.5 75.0 76.5 78.0 79.5 81.0 70.4 72.0 73.6 75.2 76.8 78.4 80.0 81.6

Representative sample and determine Individual weight across 8500

the average weight of 100 capsules. capsules across a batch

Stegemann et al PharmSciTech 15(3) 542-549 (2014)

QbD in pharmaceutical sciences
Design Space concept

Knowledge Space
Flexible
Design Space
OPERATIONAL
FREEDOM

Control Space
Control Space
Limited
Old PARADIGM New
QbD in pharmaceutical sciences

Relative prevalence by RA Tools • Ishikawa Diagram

• Failure Mode & Effects Analysis (FMEA)
• Risk Estimation Matrix (REM)
• Preliminary Hazard Analysis (PHA)
• Input Process Output (IPO)
• Diagram and Critical Relationship (CR)
Matrices

Grangeia et al EJPB 147 (2020) 19–37

QbD in pharmaceutical sciences

Example: twin-screw wet-granulation and continuous fluid-bed drying

process
▪ Risk assessment of process-units twin-screw wet-granulation and continuous
fluid-bed drying
▪ subsequent design of experiment testing of factor-response relationships to
identify and quantify critical process parameters in regard to dried granules’
critical quality attributes moisture content (LOD) and particle size distribution.

Pauli et al J Pharm Innov 13:247-60 (2018)

QbD in pharmaceutical sciences

The drying models can be divided into two major

categories: and the Arrhenius concepts
▪ the equilibrium model assumes heat transfer limitation
between the drying air and the particles
▪ the Arrhenius considers that the drying process is
governed by the evaporation rate

There are channels in the air flow, reducing the overall solid
surface area and making the liquid diffusion from the pores
to the surface a macroscopically significant step
▪ above a critical moisture content (XC) the drying rate (RV)
is independent of the actual moisture (X) The main effects considered in the drying
▪ Below XC, the liquid is transferred to the surface of the model in macroscopic scale and the level of
particles from the inner pores first and then is evaporated, particles
which translates to a moisture content-dependent drying
rate
Domokos et al Powder Technol 388: 70-81 (2021)
QbD in pharmaceutical sciences

Fishbone Diagram for twin-

screw wet-granulation
listing all conceivable
process parameters and
material attributes that could
potentially vary over the
course of production

Pauli et al J Pharm Innov 13:247-60 (2018)

QbD in pharmaceutical sciences

Fishbone Diagram for

continuous fluid-bed drying
listing all conceivable process
parameters and material
attributes that could
potentially vary over the
course of production

Pauli et al J Pharm Innov 13:247-60 (2018)

QbD in pharmaceutical sciences

Definition of potentially critical material attributes (pCMA) and potentially

critical process parameter (pCPP) including the justification

Risk Assessment
outcomes TSWG-
FBD.docx

Pauli et al J Pharm Innov 13:247-60 (2018)

QbD in pharmaceutical sciences

▪ Summary of potential critical process parameters and potential critical material attributes for
continuous twin-screw wet-granulation and fluid bed drying as identified during risk
assessment
▪ Definition of the range to be investigated

Pauli et al J Pharm Innov 13:247-60 (2018)

QbD in pharmaceutical sciences

▪ What are these typical material attributes, process parameter and quality attributes to be
considered

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

Yu et al AAPS J 16 (4): 771 (2014)

QbD in pharmaceutical sciences

▪ Finally each material attribute and process

parameter must be seen in the entirety of the
manufacturing process

▪ This also include the interconnecting steps

and processes of unit operations
QbD in pharmaceutical sciences
The Target Product Profile (TPP) defines the:
• Physicochemical attributes of the desired drug product
✓ dosage form
✓ dosing regimen that achieves a predefined clinical result
✓ pharmacokinetics (PK) bioequivalence
✓ efficacy and safety profiles
✓ product identification (tablet size, shape, and color)
✓ product stability
✓ …

▪ Intended patient population

✓ “In all cases, the product should be designed to meet patients’ needs and the intended
product performance.” (Q8(R2))
✓ ...
QbD in pharmaceutical sciences
Target Product Profile (TPP)

• Describes all the desired product

attributes and performance criteria
• Set the objectives for the product
development scientists
• In the majority of cases the desired
dosage form is already mentioned
(Tablet)
• Only very few consider „special
populations“ in the TPP for NCE
• Required pediatric formulations
have to be derived from the
standard adult form
QbD in pharmaceutical sciences
Defining a Target Product Profile
▪ Example: Sakura Tablet – Japanese Mock File

Target Product Profile-Element Target

Strength and dosage form Immediate release tablet containing
30mg of active ingredient
Specifications to assure safety and Assay, Uniformity of Dosage Unit
Efficacy during shelf life (content uniformity) and dissolution
Description and hardness Robust tablet able to withstand
Transport and handling
Appearance Film-coated tablet with a suitable
Size to aid patient acceptability and
compliance.
Total tablet weight containing
QbD in pharmaceutical sciences

Defining a Target Product Profile

▪ Example: ACE-Tablets – Pharmaceutical Development Case Study, CMC-IM Working Group

TargetProductProfile-Element Target
Dosage form Tablet, facilitating patient‘s
compliance; appropriate size, single
Tablet dose
Specifications to assure safety and Identity, assay, appearance,
efficacy Chemical and microbiological purity,
dissolution, content uniformity
CCS Adequate protection from moisture
vapor, protection through
distribution
QbD in pharmaceutical sciences

Defining a Target Product Profile

▪ Example: Tablets – EFPIA Mock P2

Target Product Profile-Element Target

Description Round normal convex uncoated tablet
Identification Positive
Assay 20mg±5% active at time of manufacture
Degradation products Qualified meeting ICHQ3B and Q6A
criteria
Dissolution Immediate release
Uniformity of dosage units Meets pharmacopoeial acceptance
criteria
Microbiological limits Meets pharmacopoeial acceptance
criteria
QbD in pharmaceutical sciences

Quality Target Product Profile = quantitative surrogate of TPP parameters within a certain range
to achieve product quality as outlined in TPP (example)
Target Product Profile-Element Target Associated Quality Attribute

Strength and dosage form Immediate release tablet Containing 30mg of -In vivo performance
active ingredient -Dissolution
Specifications to assure Assay, Uniformity of Dosage Unit (content -Assay
Safety and efficacy during uniformity) and dissolution -Degradation
Shelf life -ContentUniformity
-Chemical/physical
stability
-Dissolution
Description and hardness Robust tablet able to withstand -Friability
Transport and handling
Appearance Film-coated tablet with a suitable size to aid -Appearance
patientsacceptability and compliance. Total
tablet weight containing 30mg of active
ingredients is 100mg with a diameter of 6mm
QbD in pharmaceutical sciences
Quality Attribute Target
Dosage form Tablet, maximum weight 200mg
Potency 20mg
Pharmacokinetics Immediate release (T in 2h)
Appearance Tablet conforming to description shape and size
Identity Positive for acetriptan
Assay 95–105%
Impurities ACE12345 NMT 0.5%; other Imp. NMT 0.2%, total NMT 1%
Water NMT 1%
Content Uniformity Meets USP
Resistance to Crushing (Hardness) 5-12 kP
Friability NMT1.0%
Dissolution Consistent with immediate release, e.g. NLT75% at 30mins
Disintegration NMT 15 mins
Microbiology If testing required, meets USP criteria
QbD in pharmaceutical sciences
Criticality Assessment of Quality Attributes
▪ It is the potential impact of product quality attributes on safety and efficacy that defines criticality
▪ Quality Attributes from the TPP have to undergo a risk/criticality assessment (acc. ICH Q9) to
define Critical Quality attributes (CQAs), i.e. elements of the TPP that may be important to
Safety & Efficacy of the product used to focus further product and process development
▪ Attributes already established as regulatory requirements (ICH Q6) can be regarded as CQAs
per se without prior criticality determination (i.e. testing (control) is a regulatory requirement and
thus no criticality determination has to be performed) standard quality attributes (SQAs)

Specification: Test Procedures and Acceptance Criteria for New Drug Substances and New Drug Products:
Chemical Substances Q6A
General: Appearance, Identification, Assay, Impurities
Tablets: Dissolution, Disintegration, Hardness/Friability, Uniformity of dosage form
Parenterals: Uniformity of dosage units, pH, Sterility, Endotoxins, Particulate, Matter, Particle size distribution
QbD in pharmaceutical sciences

Criticality Assessment of Quality Attributes

▪ Critical Quality Attributes as the Foundation of QbD

▪ Evaluation by Risk Assessment

Ranking Severity Probability

1 NON CRITICAL NO OCCURRENCE
No damage to patient/full therapeutic effect
2 MEDIUM CRITICAL POSSIBLE
Reversible damage/weakened therapeutic effect OCCURRENCE
3 HIGH CRITICAL DEFINITE
Dead or permanent injury/no therapeutic effect OCCURRENCE

RPN = Severity x Probability

Total RPN > 3 = critical
QbD in pharmaceutical sciences
Example
Quality PROPABILITY RPN
What might go wrong? SEVERITY
Attribute What is the Risk
(Failure mode) What are the consequences?
likelihood? Assessment
Identity API has not the required MEDIUM CRITICAL POSSIBLE 2x2=4
chemical structure or solid The desired Safety and Efficacy OCCURENCE CQA
State form profile is not delivered
Assay The wrong dose is delivered HIGH CRITICAL POSSIBLE 3x2=6
to the patient No or weak therapeutic effect OCCURENCE CQA
Dissolution Does not comply with the HIGH CRITICAL POSSIBLE 3x2=6
requirement for an Therapeutic effect OCCURENCE CQA
immediate release tablet
Assay Free Fatty FFA content too high due to HIGH CRITICAL POSSIBLE 3x2=6
Acids (of an iv stress during production Weak therapeutic effect OCCURENCE CQA
emulsion)

Peroxide level Higher peroxide level due to HIGH CRITICAL DEFINITE 3x3=9
stress during production OCCURENCE CQA
QbD in pharmaceutical sciences
QbD in pharmaceutical sciences

Criticality Assessment of Quality Attributes

The acceptable variability in critical product quality attributes must be established!

▪ This assessment would consider, e.g., previous clinical exposure of the product, nonclinical
(animal) studies, in vitro biological activity assay, an understanding of the molecule’s action
and manufacturing capability
▪ Set the acceptable specifications for raw materials and process variables that lead to the
desired finished product quality attributes
QbD in pharmaceutical sciences

• QbD is “the multidimensional combination

and interaction of input variables (e.g.
material attributes) and process
parameters that have been demonstrated
to provide assurance of quality.”
• The least components and unit operations
to achieve the desired TPP targets:
✓ the lower the risk
✓ the higher the manufacturing efficiency
✓ the less complex supply chain

Van Buskirk AAPS PharmSciTech 15:665 (2014)

QbD in pharmaceutical sciences

Enablers for continuous manufacturing

• Science and model based formulation and
process development (QbD)
• Capable to accomodate for the product and
process variables
• Closed loop PAT enabled control of the CQA
and CPP
• Data processing and process adjustment
within the design space
QbD in pharmaceutical sciences
Continuous manufacturing principles
Blending/mixing depends on various factors that
need to be investigated to establish the „design
spaces“:
▪ Material properties
▪ Process parameter
GEA continuous blender
• Flow rate
• Impella speed
• ...
▪ Blender design
• Impella design configuration (design of blades,
blade angle, number of blades)
• Blender/mixer angle
• Weir
• ...
Portillo et al Powder Technol 182, 368 (2008); Vanarase et al Powder Technol 208, 26 (2011); Vanarase et al Powder Technol 246, 63 (2013)
Statistical tools used in QbD
 Statistical tools used in QbD
Statistical tools in QbD
▪ Typically, only a few process disturbances or independent process changes routinely occur, and
the hundreds of measurements on the process variables are only different reflections of these few
underlying events

▪ Principle Component Analysis (PCA)

▪ Soft Independent Modeling of Class Analogy (SIMCA)
▪ Partial Least Squares Regression (PLS)
▪ Response Surface Methodology (RSM)
▪ Response Contour Plot
▪ Experimental Design (Design of Experiments (DoE)
▪ Regression Analysis
▪ Coefficient plot
Statistical tools used in QbD
Statistical analysis from a historical set of data or data from a Design of Experiments (DoE)
can be used for:
• Product development (product & process design)
• Generate process understanding (product improvement)
• Process monitoring and control

Tomba et al. Int J Pharm 457: 283–297 (2013)

Statistical tools used in QbD

Block diagram representation of the use of a latent variable regression model

(a) for product property prediction - product development (product & process design)

Tomba et al. Int J Pharm 457: 283–297 (2013)

Statistical tools used in QbD

Block diagram representation of the use of a latent variable regression model

(b) for product design (by model inversion) - process understanding, monitoring & control

Tomba et al. Int J Pharm 457: 283–297 (2013)

Statistical tools used in QbD

Block diagram representation of the use of a latent variable regression model

(c) for process design (by model inversion) - process understanding, monitoring & control

Tomba et al. Int J Pharm 457: 283–297 (2013)

Statistical tools used in QbD
Principle Component Analysis (PCA)
▪ PCA is a mathematical procedure that transforms a large set of variables into a lower
dimensional set of new variables designated as principal components
▪ PCA is an exploratory analysis technique which, depending of the analysis objective,
can be used on its own to look for features/patterns/clusters in the data
▪ The philosophy behind PCA: in any data set it is likely that the key information is
contained in some dominating sources of variability other than typical variability (e.g.
noise in the measurements)
▪ Each of these new variables, typically called principal components (PC) but also
sometimes referred to as latent variables (LV), is a linear combination of the original
variables
▪ The first principal component to be extracted is that which captures the highest
amount of variability in the data set and each subsequent component to be obtained
is that which captures the highest amount of the residual variance
Ferreira & Tobyn Pharm Dev Technol 2015;20(5):513-27
Statistical tools used in QbD

Summary and meaning of key PCA model outputs

Variance explained
This value tells how much of the total variance included in the data set (or in each individual
variable) is explained by the model (or each individual principal component). It is a measure of how
well the model fits the data. It can be expressed as a number between 0 and 1 or a percentage.

Scores plot
The scores of a sample in a principal component is the projection of that sample/observation in the
principal component. This plot will most commonly be shown as the scatter plot of the scores of the
samples on two components and is used to inspect the relationship between samples by looking at
their relative positions in the scatter plot. This type of plots is commonly used to identify clusters
(groupings), atypical observations and trends.

Ferreira & Tobyn Pharm Dev Technol 2015;20(5):513-27

Statistical tools used in QbD

Loadings plot
The loading of an original variable in a principal component is a measure of how much that variable
has contributed to that component. Variables with high (positive or negative) loadings have a strong
contribution for that component. As for the scores plot, it’s usual to see scatter plots of two
components. The relative position of each original variable in the loadings plot can tells us about
how they relate to one another. For a better understanding of which measurements are driving the
trends/groupings seen in the scores plot, the scores and loadings plot need to be examined in
conjunction (sometimes the scores and loadings plots are superimposed in a biplot).

Hotelling’s T2
Diagnostic statistics that measures the distance from the sample to the centre of the model.
Samples with high Hotelling’s T2 are different from all the other samples and can have a large
influence in the model obtained.

Ferreira & Tobyn Pharm Dev Technol 2015;20(5):513-27

Statistical tools used in QbD

Residuals
Diagnostic statistic that gives a measure of the lack of fit of the model to each sample. It measures
the distance between the sample and its projection on the model. Samples with high residuals are
poorly explained by the model

Ferreira & Tobyn Pharm Dev Technol 2015;20(5):513-27

 Statistical tools used in QbD
Principle Component Analysis (PCA)
▪ For a sample of mean centered and scaled measurements
with n observations on k variables, X, the principal components
are derived as linear combinations ti = Xpi in such a way that
subject to the eigenvector |pi| = 1, the first PC has the
maximum variance, the second PC has the next greatest
variance and is subject to the condition that it is uncorrelated
with (orthogonal to) the first PC, etc

(a) Simple interpretation of PCA and dimensionality reduction. The

principal components t1 and t2 use the correlation of five variables and
break the process in two orthogonal events. The first principal
component corresponds to the event that affects the largest number of
variables, the second to the event that affects the next number of
variables, and so on. (b) These components can be plotted against each
other. A five-variable system is projected onto a two-dimensional plane.
Statistical tools used in QbD
From data table to variable space One object in
Var. 3 3-variable space

Variables p (e.g., wavenumbers)

Objects n (e.g., samples)

Var.1 Var.2 Var.3 Var.4

Sample1 Abs.1.1 Abs1.2 Abs1.3 Abs1.4

Sample2 Abs.2.1 Xnp

= Var. 2
Sample3 Abs.3.1
Data + Noise
Sample4 Abs.4.1

Var. 1

The whole table yields a swarm of points in variable space.

77
Statistical tools used in QbD
The basic idea behind Multivariate Analysis is that the number of underlying factors acting
on a system is much smaller than the number of measurements taken on the system.
(1) Looking for the mean
mean

var.
var. 3 3 (2) Mean centering
Each of the points represents a
var. 3
sample that is measured e.g., at
three different wavenumbers.

var. 2
var. 2

var. 1
var. 1
(3) Unit variance scaling
Statistical tools used in QbD

The first principal component (PC1)

is set to describe the largest
variation in the data, which is the
same as the direction in which the
points spread most in the variable
space

The Score value (ti1) for the point i is

the distance from the projection of
the point on the 1st component to
the origin

PC1 hence is the first latent variable

in a new coordinate system that
describes the variation in the data
Statistical tools used in QbD

The second principal component

(PC2) is set to describe the largest
variation in the data, perpendicular
(orthogonal) to the 1st component

The Score value (ti2) for the point i is

the distance from the projection of
the point on the 2nd component to
the origin

PC2 hence is the second latent

variable in a new coordinate system
that describes the variation in the
data
Statistical tools used in QbD

The loading (p) describes the original variables importance

for respective PC. This is the same as the similarity in
direction between the original variable and the PC.

The loading (p) is described as the cosine of the angle

between the original variable and the PC.
Statistical tools used in QbD
Principal Components (PC)
Main data variations, also known as „latent variables“, „factors“ and „eigenvectors“

Scores (T)
Map of samples: Projected locations of objects onto the PCs

Loadings (P)
Map of variables: Correlation between variables (regression of X on T)

Residuals (E)
Error. The data can be devided into structure and residual (noise): X = Xstructure + E

Variance
Residual variance – variance remaining in E
Explained variance – The % variance explained by Xstructure

Model Equation: X = TPT + E = structure + noise

Statistical tools used in QbD
SIMCA : Soft Independent Modeling of Class Analogy:
▪ based on using separate bilinear modelling
▪ Individual data class models are often PCA models
▪ A complete SIMCA model usually consists of several PC-models one for each class

1.) Training stage

data mining, e.g. PCA

2.) Classification stage

Testing the model
Statistical tools used in QbD

Each group model is optimized

regarding:

→ number of PCs

→ data pretreatments (MSC,

normalisation, derivatives)

→ wavenumber range (p variables)

Statistical tools used in QbD

For future predictions a new set of X-measurements (e.g., spectra)

is used to predict Y-values (e.g. concentration)
Statistical tools used in QbD
Partial Least Squares regression (PLS)

▪ PLS also known as projection to latent structures, is the most commonly used multivariate regression
method
▪ PLS is particularly useful to generate calibration models that allow the prediction of outcomes of
expensive and/or time consuming measurements using other cheaper and/or faster techniques
▪ This technique is used to find relationships between two blocks of data (often referred to as independent
variables or X and dependent variables or Y)
▪ PLS regression is also a projection technique and as such is related to PCA in the sense that each block
of data is decomposed
▪ In PCA the principal components are sequentially extracted to maximize the amount of variance in
the data captured by each component, in the PLS regression the correlation between X and Y is
taken into account and the principal components will be selected as the directions that explain the larger
amount of variance in X which is directly related with variance in Y

Ferreira et al Pharm Dev Technol (2014)

Statistical tools used in QbD

Summary and meaning of key PLS model outputs

Variance explained (R2) in Y

This value tells how much of the total variance included in the dependent variable (Y) is explained by the
model. Provides an indication of how well the model fits the training set.

Variance predicted (Q2) in Y

This value tells how much of the total variance included in the dependent variable (Y) is predicted by the
model according to internal validation. Provides an indication of how well the model predicts new data and
thus gives a more realistic assessment of the model predictive ability when applied to new samples.

Ferreira & Tobyn Pharm Dev Technol 2015;20(5):513-27

Statistical tools used in QbD
Root mean square error (RMSE)
Compares predicted and measured values (for the dependent variable) and gives a measure of how
well the model is able to explain/predict the dependent variable (higher error ¼ poor model). The
RMSE can be calculated for the calibration (RMSEC), cross-validation/internal validation (RMSECV)
or external prediction (RMSEP).

Regression coefficients
The PLS model can be represented as a multiple regression equation (Y ¼ b0 + b1x1 + b2x2 + . . .)
where b1, b2. . . represent the regression coefficients for each variable over all model components.
The scaled regression coefficients reflect the relative importance of the variables to the model and
can be used to assess which are the original variables with a stronger relationship with the
dependent variable (Y).

Variable Importance for the Projection (VIP)

The VIP plot summarizes the importance (weight) of each original variable in the regression model

Ferreira & Tobyn Pharm Dev Technol 2015;20(5):513-27

Statistical tools used in QbD

▪ Depending on the intended use of the model, different effort will be applied to its validation but
it’s advisable as a minimum to perform an internal validation (also known as cross-validation)
▪ This procedure involves sequentially removing samples from the data set, calculating a model
with the remaining samples and applying that model to the samples set aside to obtain a
prediction – this provides an indication how robust the model is to predict new samples
▪ MVA methods are empirical or ‘‘data-driven’’ i.e. they make no use of previous information or a
priori system knowledge and instead focus on extracting the ‘‘hidden’’ information contained in
large/complex data sets
▪ The power of MVA resides in the fact that it enables understanding the relationships in the data
and generation of knowledge even in the absence of fundamental/mechanistic understanding
▪ it is important to ensure that the data included in the analysis provides a good representation of
the expected system variability and large amounts of data may be required to have confidence
that the relationships identified are representative

Ferreira et al Pharm Dev Technol (2014)

Statistical tools used in QbD
Partial Least Squares Regressions (PLS)
A connection between an independent variable matrix X, and a matrix Y
of M dependent variables y shall now be established, similar to PCA
Statistical tools used in QbD
The effect of introducing a second PLS component can also be illustrated by looking at
the residuals after PLS comp. 1 and PLS comp. 2 → the residuals f2 (after PLS comp. 2)
are smaller than f1 (after PLS comp. 1)
Statistical tools used in QbD
Soft Independent Modelling of Class Analogy
(SIMCA)
▪ SIMCA is a disjoint classification algorithm.
▪ In SIMCA each class of samples is modelled
separately using PCA.
▪ A PCA analysis is performed on each class,
represented by the samples in the training set
belonging to this class.
▪ Based on this analysis, boundaries around the
class are defined based on two spatial distances
(Euclidean and Mahalanobis distances).
▪ The selection of the number of PCs to be taken
into account to define the boundaries for each
class is based on a cross-validation procedure
Statistical tools used in QbD
▪ The data obtained from experiments designed to detect the relationships between operating
variables and CPPs of the process, in order to guarantee the CQAs are statistically modelled
▪ Response surface DoEs are empirical models for the approximation of the underlying unknown
physical mechanisms that are supposed to have generated the experimental data
▪ A statistical model fits the data, as shown:

▪ where y is a CPP (i.e., a product temperature), xi are the operating variables (i.e., impella
speed), ε is a random error assumed to follow a Gaussian distribution, and bi are model
parameters. The terms xi2 allow the evaluation of a quadratic impact of the operating variable on
the CPPs, and the term x1x2 allows the test for an interaction between operating variables
▪ The objective is to estimate the bi parameters from experimental data which best predict the
CPP values from observed operating variable values
Statistical tools used in QbD

Response Surface Methodology (RSM)

▪ The response surface applied for
optimisation of the experimental values of
previously identified significant experimental
3 D plot
factors can be visualised graphically.
▪ The function f (x1, x2) can be plotted against
the levels of x1 and x2. Three-dimensional
graphs show the response surface from the
side and are called response surface plots.
▪ Sometimes it is less complicated to view the
response surface in two-dimensional graphs
(contour plots) that show contour lines of x1
and x2 pairs which have the same response 2 D plot
value y.
Statistical tools used in QbD

▪ Data
preprocessing –
filtering of
irrelevant features
− Baseline
correction,
SNV, MSC,
Derivatives,
etc.
▪ Model Building:
Qualitative (e.g.,
PCA),
Quantitative (e.g.,
PLS, PLS2)
Statistical tools used in QbD
Spectral pretreatments are applied to reduce the offset effects due to e.g. scattering etc:
▪ increase the signal-to-noise ratio
▪ to correct for spectral interferences

Sources of spectral variations can be versatile:

▪ Interaction of different compounds
▪ Light scattering from solid samples or turbid solutions
▪ Pathlength variations through in-homogenously distributed or agglomerating particles

Distortions caused by the spectrometer hardware like:

▪ Baseline drift
▪ Wavelength shifts
▪ Effects from detector non-linearity or stray light
▪ Noise coming from the detector, amplifier or analogue-digital converter
Statistical tools used in QbD
The Savitzky-Golay smoothing The Savitzky-Golay derivatives
▪ Approach fits a low-degree polynomial ▪ The resolution enhancement makes it easier
through the data points within the local spectral to identify weak peaks in the spectra that are
window to derive the processed signal values not visible in the original spectra.
from the polynomial function.
▪ The main drawback of derivatives on
▪ A drawback is that suboptimal cut-off spectroscopic data is the loss of the original
frequency can influence useful spectral parts, shape of the spectral curve, but most notably,
so that the spectral resolution decreases. reduction of the signal-to noise ratio.

The Normalization Multiplicative Signal Correction (MSC) and

Standard Normal Variate (SNV)
▪ Normalisation is used to compensate for
variations due to, for example, varying ▪ Show very similar results whereas especially
amount of analytes. The result is that all SNV shows good results in combination with
spectra display a common size, so they have derivative spectra to compensate for
been normalized. variations in particle size.
Statistical tools used in QbD

Normalization
Effect of normalization on
near-IR spectra of five
synthetic gluten and starch
mixtures.
Normalize to (divide each
variable by) the sum of the
absolute value of all
variables for the given
sample
Design of Experiments (DoE)
Design of Experiments (DoE)
What is DoE?
• Methodology to design (plan) and statistically evaluate experiments
• Set of representative experiments, in which all factors under investigation are varied
simultaneously and systematically
• From this set, a model is derived which captures the relation between factor settings and
experimental results (responses, e.g., CQAs, performance attributes)
Aim of DoE
• Maximizing the information content from experimental series (relationship between inputs
and output) while keeping the number of experiments low
Benefits
▪ Information on relationship between inputs and output are statistically significant and effects
of input variables on output are quantifiable
▪ Experimental designs for different objectives (e.g., number of variables, screening,
optimization, formulation development)
Design of Experiments (DoE)
▪ Low-resolution designs (III or IV) like
fractional factorial design (FFD),
Taguchi design, and Plackett-Burman
design are primarily used for the
purpose of screening
▪ FFDs are expressed using the notation
Xk-p, where X is the number of levels of
each factor investigated, k is the
number of factors investigated, and p
describes the size of the fraction of the
full factorial used. For example, a 25–2
design is 1/4 of a two level, five factor
factorial design, which requires only 8
runs instead of full 32 runs
Design of Experiments (DoE)
▪ The “Taguchi method” utilize two-, three-,
and mixed-level FFDs . This design
primarily relies on the use of orthogonal
arrays, which provide a set of well
balanced (minimum) experiments serve as
objective functions for optimization.
Taguchi design starts with minimum of
three factors and produces four
experimental runs at two levels without
requisite of any center point runs.
▪ This design has run numbers that are a
multiple of 4. Plackett-Burman design
starts with minimum of 11 factors and
produces 12 experimental runs without
requisite of any center point runs.
 Design of Experiments (DoE)
▪ Response surface designs are particularly
used for optimization of the factors identified
from the risk assessment and/or screening
study. The response surface designs include
full factorial design, central composite
design (CCD), Box-Behnken design (BBD),
optimal design, and mixture designs.
▪ CCD is considered as an augmented form of
three-level factorial design coupled with star
points or axial points.
▪ BBD is an independent quadratic design in
that it does not contain an embedded
factorial or FFD. In this design, the treatment
combinations are at the midpoints of edges
of the process space and at the center.
Design of Experiments (DoE)
DoE and Process Characterization

X1
X2
X3
Process Y1, Y2
X4

• Determine influential variables (factors)

• Determine the range influential factors to optimize response
• Determine the range of influential factors to minimize response variability
→ Improve process yield
→ Reduce variability
→ Reduce development time
→ Reduce overall costsProcess
Design of Experiments (DoE)

Problem definition – Steps in experimental design

1. Experimental objective: Why is an experiment done? For what purpose? What is the desired
result?
2. (Input) factors: Variables that are changed to give different results on the measured
responses
3. Responses: e.g., CQAs, performance attributes
4. Model: polynomial model corresponding to the objective
5. Design: supports the selected model, follows from the objective
6. Worksheet: review of the actual design
Design of Experiments (DoE)

Problem definition – step 1: experimental objective

i. Familiarization
ii. Screening
iii. Finding the optimal region
iv. Optimization
v. Robustness testing
vi. Mechanistic modeling
Design of Experiments (DoE)

Problem definition – step 2: definition of factors

▪ Variables that exert an influence on the system due to changes in their levels

Controlled Quantitative Controlled Qualitative

low center high
Bender type A B C
Temperature 35°C 40°C 45°C
Supplier DE SP BG
Amount 2g 4g 6g
Catalyst Na K Ca
Speed 200 rpm 250 rpm 300 rpm
Surface coating PVC PvDC PP
pH 4 6 8

Uncontrolled
Outside/Inside Temperature
Outside/Inside Moisture
Design of Experiments (DoE)

Problem definition – step 3: specification of responses

▪ Regular and derived responses
▪ Quantitative and qualitative responses

Problem definition – step 4: Selection of model

▪ Polynomial regression models

Three main types of polynomial models

x …….factors
Linear: y = β0 + β1x1 + β2x2 + … + ε y ........responses
Interaction: y = β0 + β1x1 + β2x2 + β12x1x2 + … + ε b.........regression coefficients
e.........residual
Quadratic: y = β0 + β1x1 + β2x2 + β11x1 2+ β22x2 2 + β12x1x2 + … + ε
 Design of Experiments (DoE)
2k-fractional factorial Response Surface
Full factorial designs
designs Designs
Overview Experimental Designs

Fractional factorial non-linear

2 levels per factor > 2 levels per
designs with 2 relationship, 5
levels per factor (2k –full factorial) factor
levels per factor
(examplary)

Main effects (and Main effects and all interactions

interactions) Detailed information Quadratic effects
on factors and
Screening interactions Optimization

Factors 3 - 47 Factors 2 - 4 Factors 2 - 3 Factors 2 - 7

Design of Experiments (DoE)

▪ Chosen model and design to be generated are intimately linked

▪ The number of factors, their levels and nature (quantitative,
qualitative,…), and the selected experimental objective.

Problem definition – step 6: creation of worksheet

▪ WS contains the selected design as work instruction for the
experimenter
Design of Experiments (DoE)
Introduction to Full Factorial Designs
▪ Basis for all classical experimental designs
▪ Require relatively few runs per investigated factor
▪ Upgrading to form composite designs used in optimization
▪ Basis for two-level fractional factorial designs (used in screening with many factors)
▪ 2 – 4 factors; with many factors switch to fractional factorial design

Design
22 two-level design in 2 factors
23 two-level design in 3 factors
32 three-level design in 2 factors
Design of Experiments (DoE)
Regression analysis – the Summary of Fit plot
▪ R2 goodness of fit: how well the regression model fits the raw data (0 – 1)
▪ Q2 goodness of prediction: estimation for the predictive power of a model (-∞ – 1)
▪ Model validity: right type of model (> 0.25)
▪ Reproducibility: replicate error (> 0.5)

Model to be judged as good:

• Difference R2 – Q2: < 0.2 – 0.3
• Q2 > 0.5
• Model validity: > 0.25
• Reproducibility: > 0.5
Design of Experiments (DoE)

Model interpretation – Coefficient plot

▪ Regression coefficients incl. confidence
intervals
▪ Size of confidence intervals depending on:
(i) experimental design,
(ii) goodness of the regression model
(iii) number of degrees of freedom
▪ Main effects and interaction terms

Coefficient plot for the six variables included in model 1

Design of Experiments (DoE)

Use of model – Response contour plot

• Making predictions
Design of Experiments (DoE)
Systematic characterization of raw materials, processes and the corresponding product quality
▪ Design of Experiments (DoE) (Umetrics MKS MODDE)
▪ Systematic parameter variation to develop a Design Space (DS)
▪ Estimation of the impact of external influences

Experimental Design Reproducibility Process Understanding

Design of Experiments (DoE)
▪ NIR spectroscopy for monitoring the powder-blend composition and identification of the mixing end-
point.
▪ Three different fill levels were studied with either acetyl salicylic acid (ASA) or agglomerated α-lactose
monohydrate (LM) on top

Experimental setup with a four-bladed mixer PLS Models for blend quality monitoring

PLS Models for nominal ASA content:

Preprocessings: SNV and 4th order baseline correction
Model1: 0-100% - RMSEP 2.97 % (11 fractions)
Static sample size ~ 6 mm³ (Penetration depth 0.3-0.5 mm, Spot Model1: 40-60% - RMSEP 2.08 % (13 fractions)
size 4 mm), Integration time = 0.3 s, Impeller speed = 4 rpm 10 spectra recorded from each fraction
Design of Experiments (DoE)

▪ Simultaneous NIR monitoring of powder blending dynamics at different positions

▪ In-line measurements at critical positions like corners
▪ Spatially-resolved investigation of mixing dynamics
 Design of Experiments (DoE)
▪ Experimental investigation for setting up powder mixing include different fill levels and
powder loading orders
▪ To investigate granular flow properties, compressed convective zones and dead zones
dependent on the fill level (effective powder homogeneity)
Formulation blend

Observed vs. predicted plot of

Schematic illustration of the blending vessel. Measurement ports the overall PLS regression model
1–10 for the optical fibers are indicated by the arrows. The predicting the API concentration.
dashed lines indicate the investigated fill levels.
Scheibelhofer et al AAPS PharmSciTech 14 (1): 234 (2013)
Design of Experiments (DoE)

Position 2

Experiments with a
different H/D ratio. The
starting composition is
always 80% LM on top of
20% ASA

Seventy grams of LM was deposited on 70 g of ASA (resulting

in H/D 0.35). The solid lines show the predicted API values,
and the transparent area represents the respective 95%
confidence interval of the prediction. Positions higher than 6
were above the powder fill level

Scheibelhofer et al AAPS PharmSciTech 14 (1): 234 (2013)

Design of Experiments (DoE)
solid feed liquid screw barrel Dryer
drying
Exp Run- rate feed speed temperature rotation drying air
temperature
# order rate TSG speed flow (DAV)
(SFR) (BT) (DT)
(LFR) (SS) (DRS)
Fractional factorial design matrix 1 4 - - + + + - -
of performed Screening-DoE, 2 6 - - - + - + +
3 9 - - + - + + +
based on quantitative pCPPs 4 12 - - - - - - -
5 14 - + - + + - +
identified during risk 6 15 - + + - - - +
7 17 - + - - + + -
assessment. Factor low is 8 18 - + + + - + -
denoted as “-“, factor high as “+”, 9 11 + + - + - - -
10 13 + - - - + - +
center point settings are “0” 11 19 + + - - - + +
12 1 + - + - - + -
13 2 + - + + - - +
14 3 + + + - + - -
15 7 + - - + + + -
16 8 + + + + + + +
17 5 0 0 0 0 0 0 0
18 10 0 0 0 0 0 0 0
19 16 0 0 0 0 0 0 0

Pauli et al J Pharm Innov 13:247-60 (2018)

Design of Experiments (DoE)

Summary of DoE results. pCPPs

and pCMAs that were initially
identified by risk analysis were
either confirmed to be critical
(CPPs) or downgraded to not
critical (nCPPs/nCMAs)

Pauli et al J Pharm Innov 13:247-60 (2018)

 Design of Experiments (DoE)
To keep the process within the specification (granule moisture) changing the total material
throughput ṁtot might be necessary
▪ The following trial was
conducted: Granulation was
initially started at a total material
mass flow ṁtot of 5.2 kg/h and
DRS of 17 rph and DAV of 120
m3/h. Then, ṁtot was first
decreased and later increased
by 50% from its initial setpoint
▪ It was demonstrated that model-
based adaption of dryer rotation
speed (DRS) and drying airflow
(DAV) can control dried granules’
LOD within acceptance limits at
varying ṁtot,
Pauli et al J Pharm Innov 13:247-60 (2018)
Design of Experiments (DoE)

Response surfaces of the split-plot

experimental study. The settings of the
other two factors are as follows:
(a) 1 kg/h mass flow rate – 30 °C
temperature
(b) 1 kg/h mass flow rate – 50 °C
temperature
(c) 2 kg/h mass flow rate – 30 °C
temperature
(d) 2 kg/h mass flow rate – 50 °C
temperature

L/S = Liquid to solid ratio

Domokos et al Powder Technol 388: 70-81 (2021)

Design of Experiments (DoE)
Modelling roller compaction
▪ Roller compaction is a dry granulation technology
▪ The aim of this work is to reduce the material
consumption in a systematic design of
experiment (DoE).
▪ A mechanistic approach based on previous
models has been developed in order to replace
the experiments with higher throughput and
material consumption (high roller and screw
speeds) with model predictions
▪ Two models exist: (1) Screw mass flow rate
controlled or (2) gap geometry and the roller Schematic of a roller compactor. (1) inlet funnel with
speed controlled agitator; (2) feed screw; (3) tamp screw; (4) small quantity
inlet funnel; (5) rollers; and (6) rotor miller. The mechanistic
model describes the region of between the rollers
Toson et al Int J Pharm X 1 (2019) 100005
Design of Experiments (DoE)
Schematic diagram of the compaction
process showing the feeding zone and the Screw mass flow rate controlled
nip and slip regions.
− The powder has the maximum density ƴG
at the gap.
− Past the roller gap, there is a region of
elastic recovery where ribbon thickness T
increases while reducing the ribbon density
to its final value ƴR.
− The mass flow can be predicted in the
feeding zone using either the screw mass
flow rate (ms ) or the gap geometry and
the roller speed (mroll). Gap geometry and the
− Experimentally, it is the throughput roller speed controlled
observed after the ribbon relaxation (mobs).
Toson et al Int J Pharm X 1 (2019) 100005
 Design of Experiments (DoE)
Full factorial DoE with 11 experimental runs
The critical process parameters are the roll speed NR,
the gap width S, and the specific compaction force
SCF. The simulated values used the experimental runs
at low throughput (NR level=–1) for calibration,
whereas the high the results for higher throughput NR
levels=0 and +1) were fully predicted

Process settings used in the DoE for IbuMCC

and IbuMannitol formulations (Ibu =
Ibuprofen)

Toson et al Int J Pharm X 1 (2019) 100005

Design of Experiments (DoE)

▪ Experimentally observed ribbon solid fraction and the prediction error for the IbuMCC and
IbuMannitol formulations. The predictions used the calibrated compression profile K, ƴ0. The
two methods to calculate the screw constant in screw-controlled mode are compared (cS1 and
cS2). Runs marked with an asterisk (*) were used in the calibration sequence.

Toson et al Int J Pharm X 1 (2019) 100005

Design of Experiments (DoE)
Graphical exploration around the optimal point (γR=0.725; mobs =10 kg/h) comparing the multiple
linear regression (MLR) model based on the experimental data with the MLR model based on the
simulated data for IbuMCC. The four experimental data points used to calibrate the mechanistic
model are marked ●

Toson et al Int J Pharm X 1 (2019) 100005

Design of Experiments (DoE)

Screening DoE methods used in development

Grangeia et al EJPB 147 (2020) 19–37

Design of Experiments (DoE)

Optimising DoE methods used in development

Grangeia et al EJPB 147 (2020) 19–37

QbD in pharmaceutical sciences

Conclusion

▪ An iteration process for calibrating the K and ƴ0 values, which can reduce the material
amount required, the number of RC experiments and the material characterization efforts at
a low software expense
▪ A simple method that uses the RC throughput to calculate the volume relaxation that the
ribbon experiences after exiting the rollers and the relaxation factor (β) can be used to
predict the throughput with an acceptable accuracy
▪ Based on a limited number of experiments, a design space for a broad range of throughputs
can be predicted with a high accuracy
▪ It may be used in the future as a soft sensor for RC processes operating in continuous
manufacturing lines for in-line adjustment of throughput while maintaining the desired ribbon
solid fraction

Toson et al Int J Pharm X 1 (2019) 100005

Process Analytical Technology (PAT)
 Process Analytical Technology (PAT)
Process Analytical Technology (PAT)

Design and optimization of drug formulations and manufacturing

processes within the PAT framework can include the following
steps (the sequence of steps can vary):
▪ Identify and measure critical material and process attributes
relating to product quality
▪ Design a process measurement system to allow real time or
near real time (e.g., on-, in-, or at-line) monitoring of all critical
attributes
▪ Design process controls that provide adjustments to ensure
control of all critical attributes
▪ Develop mathematical relationships between product quality
attributes and measurements of critical material and process
attributes
Process Analytical Technology (PAT)

▪ PAT is a broader field encompassing a set of Process

instrument
tools and principles to enhance
manufacturing process understanding and
control which includes process analysis, Data
Aquisition Implementa
chemical engineering, chemometrics, & -tion
Instrument approach
knowledge and risk management, and Control

process automation and control.

PAT

Hard (green) and soft (blue) Data

Processing Method
process analysis elements & develop-
Chemoma- ment
trics
Process Analytical Technology (PAT)

• Process Control (PC) has the goal of keeping a chemical or pharmaceutical

process within predefined boundaries with respect to predefined variables using
manipulation of certain control parameters.
• PC is not to be confused with process monitoring
• PC directly affects the safety and reliability of a process.
• PC determines the quality of the products produced by a process.
• PC can compensate for variations in input variables
• PC can affect how efficient a process is operated.
• PC has a major impact on the profitability of a company.
• PC control established over 80 years in the chemical industry.
• PC is new to the pharmaceutical industry
Process Analytical Technology (PAT)

Process Control is achieved by e.g. PAT

• PAT is a system for designing, analyzing, and controlling manufacturing through timely
measurements of critical quality and performance attributes of raw and in-process
materials and processes with the goal of ensuring final product quality
• A mechanism to design, analyze, and control pharmaceutical manufacturing processes
through the measurement of critical process parameters (CPP) which affect product
quality attributes (CQA)
• The real-time testing and adjustment based on the complete understanding of how the
components and related processes affect the final product
• Nothing less than a shift in process control philosophy – replacing end of unit
operation testing with continuous, or nearly so, on-line monitoring of product quality
Process Analytical Technology (PAT)
What are the requirements for PAT
• Robust process instrumentation with suitable analytical merits
• Optimized process integration and operation
• Robust chemometric models or data processing algorithms
• Representative process data over time and conditions
• Autonomous process instrument control and data acquisition
• Real - time measurement assurance (i.e., smart sensing)
• Suitable information technology and automation controls infrastructure for efficient data fusion
and archive management
• Sufficient method development, validation and ongoing compliance
• Suitable instrument validation and compliance
• A comprehensive onsite process instrument program (metrology, instrument maintenance,
training, sufficient on-site instrument specialist, etc
• Identified performance metrics and continuous improvement plans
Process Analytical Technology (PAT)
Statistical process control (SPC)

Often monitoring identified critical process parameters are considered to be assure together with critical quality
attributes by statistical process control (SPC).
Consider: The parameters that appear to be critical in the DOE are not necessarily the ones that will give
information about the “wellness of the process” in SPC charts.
Reason: Even though identified as important in DOE to be controlled, SPC charts on routine data are
noncausal!
Example: Suppose that temperature is important to the yield, as determined by DOE. This means that during
production the desired temperature profile will be regulated by the controllers. The SPC monitoring of
temperature will provide good results.
Whereby, what is really important is how much effort the controller is putting to maintain the temperature (how
much the valve to the cooling agent opened or closed during the reaction) .
Therefore, monitoring the controller action will provide much more information about abnormal situations than
monitoring the temperature, although temperature was identified as critical process parameter by DOE
Process Analytical Technology (PAT)
Process Model
▪ The choice of the process conditions of
the unit operation will be dictated by a
model that takes into account the value
input variable and calculates process
conditions such that quality is on
target. When the model is empirical,
multivariate analysis can be used.

▪ The example here illustrates a Control strategy

feedforward control scheme for unit N using projection
based on input information on the space
“state of the intermediate product” from
unit N1.
Process Analytical Technology (PAT)
Traditional Process Control compared to a PAT-based Process Control

Traditional approach PAT approach

Design of manufacturing Mainly empirical Systematic understanding of
process formulation and process factors
Process verification Based on initial full-scale Continuous real-time process
batches verification
Process analyzers Mostly univariate process and Real-time measurement of quality
off-line material measurements and performance attributes
Process controls Primarily go/no go decisions Process is adjusted to ensure
control of quality attributes
Quality assurance Intermediate and end product Ensured by the design of the
testing of collected samples manufacturing process
Innovation in life cycle Difficult; process is often frozen Continual improvement and
optimization possible & desired
Process Analytical Technology (PAT)
PAT experience from other industries
• PAT has been used by the chemical industry since mid 20 century
• Core concepts in use for decades in chemical, semi-conductor, petroleum and other
manufacturing industries
• Formalized with founding of CPAC in 1984
• Goal of PAC (process analytical chemistry) – supply quantitative and qualitative information
about a chemical process
• Initiated by the FDA as part of the 21st Century GMP initiative in 2001 with the goal of
increasing productivity of pharmaceutical manufacturing
− Analytical in PAT includes chemical, physical microbiological, and/or mathematical
analyses
• PAT is a key tool for real time process control in the QbD framework
Process Analytical Technology (PAT)
Theoretical Target Product Profile
▪ PAT starts with the TPP, risk assessment and
QbD process to identify the critical factors for
the finished product performance
▪ PAT is applied as the analytical controls to
monitor the product throughout the
manufacturing to stay within the defined
boundaries

Risk Assessment to Identify Variables Potentially Impacting

Product Quality
Process Analytical Technology (PAT)

▪ Based on the identified critical

factors, the respective PAT tools will
be defined and applied within the
relevant process steps
Process Analytical Technology (PAT)
Direct methods
▪ Near infrared (NIR)
− Spectroscopy
− Chemical Imaging
▪ Raman
− Spectroscopy
− Chemical Mapping
▪ UV-Vis spectroscopy
▪ Mass spectroscopy
Principles for process monitoring ▪ Gas chromatography
▪ SWAXS
Indirect methods ▪ Terahertz spectroscopy
▪ Optical coherence
Generic sensors
▪ Physical (size, shape, temperature, pressure, etc.)
▪ Chemical (pH value, etc.)
Process Analytical Technology (PAT)
▪ Spectroscopic techniques
− Infrared spectroscopy, Raman spectroscopy, nuclear magnetic resonance spectroscopy, ...
▪ Chemical imaging & real-time image analysis
▪ Thermoanalytical techniques
− Differential scanning calorimetry, differential thermal analysis, thermogravimetry, …
▪ X-ray diffraction techniques
− Single crystal diffraction, powder diffraction, small-angle, wide-angle, …
▪ Microscopy
− Optical microscopy, electron microscopy, scanning force microscopy, …
▪ Methods for particle analysis
− Laser diffraction, dynamic light scattering, coulter counter, …
Process Analytical Technology (PAT)

Overview of spectroscopic techniques applied and/or investigated for pharmaceutical

applications:
▪ Ultraviolet-Visible Spectroscopy (UV-Vis)
▪ Infrared Spectroscopy
▪ Raman Spectroscopy
▪ Solid-State Nuclear Magnetic Resonance
Spectroscopy (SSNMR)
▪ Time-of-Flight Secondary Ion Mass
Spectroscopy (TOF-SIMS)
▪ Terahertz Pulsed Spectroscopy (TPS)
Process Analytical Technology (PAT)
Detector Diffuse Reflectance Qualitative Analysis

Determination of Physicochemical Parameters

Infrared
spectroscopy
Monochromator

Chemometrics /
Multivariate Data Analysis
Sample Detector
Transmittance
Light Source

Physical Properties Chemical Properties

Quantitative Analysis
Particle
ParticleSize,
Size,Porosity,…… Derivatized
Porosity,…… DerivatizedSurface,……
Surface,……
Process Analytical Technology (PAT)
Process Analytical Technology (PAT)
Averaged NIR spectra

with increasing coating

Absorbance (a.u.)

thickness the absorbance Identification of

end coating
intensity increases. the qualitative
and quantitative
differentiating
wavelength

uncoated
begin coating

Wavenumber (cm-1)
Process Analytical Technology (PAT)
Attenuated Total Reflection (ATR) Spectroscopy
• Sample preparation is a challenge in many traditional IR spectroscopic techniques based on
the transmission of an IR beam through a sample. ATR is an interesting technique requiring
minimal or no sample preparation at all.
• Principle: an infrared beam is guided through a crystal by internal reflection. A sample placed
in the evanescent field (i.e. a couple of micrometers next to the crystal) contributes to the total
absorption of the IR beam.

Source: Solvias

Source: PerkinElmer
Source: HELLMA
Process Analytical Technology (PAT)
Raman spectroscopy ▪ Under constant experimental conditions, the
Benefits: number of Raman scattered photons is
proportional to analyte concentration
▪ Minimal or no sample preparation
▪ Quantitative methods can be developed with
▪ Less overlap between APIs and excipients
simple peak height measurements
than with IR spectroscopy
▪ Multiple components in complex mixtures can
▪ Complements diffraction techniques
be quantified if a distinct wavelength for each
▪ Often complements with IR spectroscopy; component can be identified.
IR-inactive vibrations can be strong in
▪ When isolated bands are not apparent, advanced
Raman spectra and vice versa
multivariate statistical tools (chemometrics) e.g.,
Shortcomings: partial least squares (PLS) can help
▪ Small sampling area (limited by laser ▪ Fluorescence is a common interference and its
beam size) severity tends to increase with shorter
▪ Sample heating (depends on laser beam) wavelengths. By using longer wavelength lasers,
many fluorescence problems can be avoided
Process Analytical Technology (PAT)

▪ Fiber optical cables between measurement point and instrument, with a dedicated path to
carry laser light and a separate to return the Raman signal to the spectrometer.
▪ Silica Raman scatter always is generated when the laser travels through fiber optics. It will
overwhelm any analyte signal if it is not removed before the laser reaches the sample; the
same is for case return fiber.
PhAT Probe Head for Solid Sampling

Source: Kaiser Optical Systems, Inc. www.kosi.com

Process Analytical Technology (PAT)
Laser excitation at 785nm

▪ API and excipients distribution

▪ Analysis of coating defects and homogeneity
▪ Spatial resolution for Raman images
− 100µm excitation spot size - typical spot power 100 mW
− Stage movement >1µm - typical 50 µm
− Raman auto-focus for bulged surfaces - typical +/- 1 mm
Process Analytical Technology (PAT)
Chemical images filtered at significant reference peaks of start
API, excipients or coatings

Reference Raman spectra of

pure substances of A, B and C

Coating process
end
Process Analytical Technology (PAT)
Xray diffraction
▪ Principle: Observing the scattering of an X-ray beam hitting a sample as a
function of incident and scattering angle. An analysis of angle vs. intensity
provides information about the crystal structure.
▪ The dimension of a structure is reciprocal to the scattering angle (Bragg‘s law)
− Large structures (e.g. particles, pores) scatter to small angles
− Small structures (crystal lattice) scatter to large angles
2dsin = n
N = order of the diffracted beam
Λ = wavelength of the X-rays
d = distance between the planes of the
crystal lattice
 = angle between incident ray and
scattering planes
Constructive interference Destructive interference
Process Analytical Technology (PAT)

Single-crystal X-ray diffraction

▪ Analysis of the evenly spaced planes of a (perfect) single crystal
X-ray powder diffraction (XRPD)
▪ Analysis of crystalline properties of powdered solid samples
Small-angle X-ray diffraction (SAXS)
▪ Analysis of large structures (e.g. particles or pores) at small scattering angles
Ultra-small-angle X-ray diffraction (USAXS)
▪ Analysis of even larger structures possible than with SAXS
Wide-angle X-ray diffraction (WAXS)
▪ Analysis of small (sub-nanometer) structures at large scattering angles
Small & Wide-angle X-ray diffraction (SWAXS)
▪ Analysis of large and small structures at large scattering angles
Process Analytical Technology (PAT)
Small/Wide Angle X-Ray Scattering – HECUS S3-Micro
▪ Crystallinity
▪ Polymorphism SAXS 15Å

▪ Internal surface

log I (a.u.)
q (1/Å)

3.79Å
WAXS Rmix
4.38Å

Tmix
I (a.u.)

3.34Å
4.77
Å
3.93Å

2 θ (°)
Process Analytical Technology (PAT)
Laser diffraction
▪ Principle: Laser light is scattered by particles suspended in a carrier gas. particle size
(distribution) can be inferred from the diffraction pattern (i.e. intensity versus diffraction angle)
Benefits
▪ Short analytical time
▪ High precision, robustness & reproducibility
▪ Wide measurement range
▪ Flexibility of operation using liquids
Shortcomings http://www.malvern.co.uk/
▪ Particles are assumed to be spherical
▪ Volume diameter of non-spherical parameters is typically overestimated
▪ Re-scattering by other particles is neglected
Process Analytical Technology (PAT)

Focus beam reflectant

▪ Principle: A laser beam rotates at high speed and propagates into a solution. When the laser
hits a particle, light is scattered back and measured to give a chord length distribution.
▪ The particle size distribution can be inferred from the cord-size distribution.

Source: Mettler-ToledoApplication_note_Crystallization_Studies_on_MultiMax_with_FBRM.pdf
Process Analytical Technology (PAT)

Optical Coherence Tomography Tab.1

Tab.2
Measurement principle:
Variation of thickness
Tab.3
within a batch

Tab.4

Tab.5

Side view

In cooperation with Determined coating film thickness correlates directly

to the optical refractive index of the material.
Process Analytical Technology (PAT)

3D Image reconstruction of several OCT images

Histogram

Propability Density

Coatingprocess
Coating Thickness (µm)

Scanning electron microscopy (SEM)

Process Analytical Technology (PAT)
Off-line At-line On-line In-line
Optical microscopy x x
Electron microscopy x
Infrared spectroscopy x x x x
Raman spectroscopy x x x x
Chemical imaging x x x x
X-ray diffraction x x x x
Nuclear Magnetic Resonance Spectroscopy x x x x
Differential Scanning Calorimetry x
Differential Thermal Analysis x
Thermogravimetry x
Focused Beam Reflectance Measurement x x x x
Laser Diffraction x x x
Dynamic Light Scattering x x x
Optical Coherence Tomography x x x
Process Analytical Technology (PAT)
Feedbackwards control Composition
analyzer/controller
AC

1- API / excipient premix

2- Excipient

Example: PAT-tool:
NIR HELIOS

AT
Composition
analyzer/
transmitter
Process Analytical Technology (PAT)

Feedforward control
Composition
analyzer/controller
AC
1- API / excipient premix Example: PAT-tool:
NIR HELIOS
2- Excipient
AT

Composition
analyzer/
transmitter
PAT used for endpoint detection

Objective of the method

▪ Online API prediction

▪ Online moving block standard
deviation of API prediction
▪ Endpoint decision not on time but
based on API concetration to be
between 90 and 110 and standard
deviation of API over the last 10
revolutions below 2,5
▪ Stopp at defined endpoint
Process Analytical Technology (PAT)
EudraLex The Rules Governing Medicinal Products in the European Union
Volume 4 EU
Guidelines for Good Manufacturing Practice for Medicinal Products for Human and Veterinary Use
Annex 17: Real Time Release Testing and Parametric Release

▪ When designing the Real Time Release Testing (RTRT) strategy, the following minimum
criteria are expected to be established and met:
(i) Real time measurement and control of relevant in-process material attributes and process
parameters should be accurate predictors of the corresponding finished product attributes.
(ii) The valid combination of relevant assessed material attributes and process controls to
replace finished product attributes should be established with scientific evidence based on
material, product and process knowledge.
(iii) The combined process measurements (process parameters and material attributes) and
any other test data generated during the manufacturing process should provide a robust
foundation for RTRT and the batch release decision.
Process Analytical Technology (PAT)
▪ A RTRT strategy should be integrated and controlled through the PQS. This should
include or reference information at least of the following:
- quality risk management, including a full process related risk assessment, in accordance
with the principles described in EudraLex, Volume 4, Part I Chapter 1 and Part II
Chapter 2
- change control program
- control strategy
- specific personnel training program
- qualification and validation policy
- deviation/CAPA system
- contingency procedure in case of a process sensor/equipment failure
- periodic review/assessment program to measure the effectiveness of the RTRT plan for
continued assurance of product quality.
QbD for special dosage forms
QbD for special dosage forms

▪ Additive manufacturing is considered to be the use of computer-aided-design (CAD)

software or 3D object scanners to direct hardware to deposit material, layer upon layer,
in precise geometric shapes. In additive manufacturing material is constantly added to
create an object e.g. tablet (3D printing)
▪ In a wider sense, 2D printing can also be considered additive manufacturing due to the
similarities by adding drug layers on substrates
▪ For 3D printing, different technologies are being used for the layering process
QbD for special dosage forms

▪ Vat photopolymerisation; which is a process that utilises a light source (e.g. laser) to
selectively cure a vat of liquid photopolymer, transforming it into a solid object. Examples of
such are the stereolithography (SLA), digital light processing (DLP) and continuous liquid
interface production (CLIP) technologies.
▪ Binder jetting (BJ); which revolves around the selective binding of solid powder particles by
spraying a liquid agent.
▪ Powder bed fusion; which is a selective thermal process that involves the fusion of powder
particles by the application of a laser or other heat source. It includes selective laser sintering
(SLS), multi jet fusion (MJF), direct metal laser sintering/selective laser melting (DMLS/SLM)
and electron beam melting (EBM).
▪ Material jetting; which is a selective technique in which liquid droplets of materials are
deposited on a surface. These droplets spontaneously solidify (known as drop-on-demand
(DOD)) or can be cured or fused using a UV light (known as material jetting (MJ)) or a heat
source (known as nanoparticle jetting (NPJ).

Awad et al Drug Discov Today 23(8):1547 (2018)

QbD for special dosage forms

▪ Direct energy deposition; which is a process that selectively deposits a form of focused
thermal energy (e.g. laser) directly onto powder particles, causing them to melt and fuse. It
involves two technologies; laser engineering net shape (LENS) and electron beam additive
manufacturing (EBAM).
▪ Sheet lamination; which compromises the bonding of materials in the form of sheets (e.g.
cut paper, plastic or metal) to fabricate 3D objects. It is often known as laminated object
manufacturing (LOM) or ultrasonic additive manufacturing (UAM).
▪ Material extrusion; which is a technology that involves the selective dispensing of material
in a semi-solid form. This technology is further subdivided into fused deposition modelling
(FDM), which utilises thermoplastics, and semi-solid extrusion (SSE), which utilises gels and
pastes.

Awad et al Drug Discov Today 23(8):1547 (2018)

QbD for special dosage forms

Kjar et al Pharmaceutics 11: 390 (2019)

QbD for special dosage forms
Excipients most commonly used in
3D printing

Konta et al Bioengin 4:79 (2017 )

QbD for special dosage forms
Spritam® is marketed by Aprecia Pharma using
powder bed fusion by the layer-by-layer
production system. The first layer consists of
the active ingredient and all of the excipients
necessary to produce the matrix tablet.
Subsequently, a binder liquid is deposited for
perfect integration and aggregation between all
of the successive and identical layers. Spritam®
can be manufactured with up to 1000 mg of
levetiracetam.

https://www.you
tube.com/watch
?v=IQZ-1xc-
hGg
Kjar et al Pharmaceutics 11: 390 (2019)
 QbD for special dosage forms
Feedback from FDA – Considerations
▪ Raw material control: printability, physicochemical characteristics, thermal conductivity, viscoelastic
property, Print fluid characteristics (Viscosity, Surface tension, Density, Rheology/ Viscoelastic
property)
▪ Process control: Identify high risk steps which may impact identity, appearance, assay, content
uniformity, drug release, impurity level, hardness, friability, crystallinity, and API polymorphic form.
Example of such steps could be: printing, solidification, recycling, controlling mass & energy
transport. PAT can be used for control.
Feedback from FDA - 3DPrinted Drug Products: Typical Quality Defects
▪ Banding: ripples on a product’s sides caused by vibration in the x-y plane during printing
▪ Leaning: off-axis products caused by drift in the x-y plane during printing
▪ Warping: product distortion caused by thermal expansion or contraction
▪ Shifting: First few layers were shifted because of base line shift during printing
▪ Collapse: loss of porosity caused by sagging layers or excessive mass/energy input
▪ Residuals: unbound powder or uncrosslinked monomer caused by incomplete printing

Akm Khairuzzaman (Acting Branch Chief, OPQ/FDA), USP Meeting Api 29, 2016
QbD for special dosage forms
▪ Solvent-casting method. The ODF is preferably formulated using the solvent-casting method,
whereby the water-soluble ingredients are dissolved to form a clear, viscous solution.
▪ Hot-melt extrusion . In the HME process, the API and other excipients are mixed in a dry state, the
heating process is started, and the molten mass is extruded out of the hot-melt extruder
▪ Semisolid casting . In the semisolid-casting method, a solution of the water-soluble, film-forming
polymer and acid insoluble polymer (e.g., cellulose acetate phthalate and cellulose acetate butyrate)
and plasticizer is added. The prepared gel mass is cast into films or ribbons using a controlled heat
source.
▪ Solid-dispersion extrusion . The term solid dispersion refers to the dispersion of one or more APIs
in an inert carrier in a solid state in the presence of amorphous hydrophilic polymers using methods
such as HME.
▪ Rolling method . In the rolling method, a solution or suspension containing the drug is rolled on a
carrier. The solvent is mainly water and a mixture of water and alcohol. The film is dried on the rollers
and cut into desired size and shapes.
QbD for special dosage forms
Films are prepared by extruding or casting methods
▪ The films can:
− contain the drug
− be neutral carriers
− provide muco-
adhesive properties
− support rapid
disintegration
− serve taste masking
QbD for special dosage forms

2D printing is defined as a the printing of drugs on substrates e.g. on oral films

▪ SciFLEXARRAYER S3 printing
system from SCIENION
▪ sciDROP PICO: piezo electric
nozzle
▪ Borosilicate glass nozzle with 80
µm capillary orifice

Scarpa et al. Int J Pharm 523 (2017) 327–335

QbD for special dosage forms
▪ 2D printing is still in its infancy,
due to the complex nature
▪ It consists of preparation of a
printable drug solution (ink)
able to form µl drop sizes
▪ The maximum amount of drug
to be printed on a substrate is
< 25 mg
▪ Multiple printing cycles are
possible, but increase
complexity

Scarpa et al. Int J Pharm 523 (2017) 327–335

QbD for special dosage forms
Tesa Listerine
SEM pictures of
▪ 1st column blank substrates (top to
bottom): Tesa, HPMC clear, Gelatin
clear, yMCC; 2nd column after HPMC clear HPMC opaque
printing Sodium Picosulfate onto the
substrates (top to bottom) Tesa,
HPMC clear, Gelatin clear, yMCC
▪ 3rd column blank substrates (top to Gelatin clear Gelatin opaque
bottom): List, HPMC opaque, Gelatin
opaque, pMCC; 4th column after
printing Sodium Picosulfate onto the
substrates (top to bottom): List, yMCC pMCC
HPMC opaque, Gelatin opaque,
pMCC.

Wimmer-Teubenbacher et al Int J Pharm 547 (2018) 169–180

 QbD for special dosage forms
▪ Sodium Picosulfate drop deposition is highly contact angles of SP on substrates
dependent on the physicochemical and plotted against surface areas of
printed SP dots, Grubb’s outlier test
surface characteristics of the substrate
has been performed on the results
▪ There is an aging effect upon storage for the contact angles, error bars
are standard deviations (n=10)

SEM images of SP printed on (a) HPMC, (b) HPMCT, (c) TESA, (d) GE, (e)
GET, (f) LIST after storage at room conditions for a period of eight weeks

Wimmer-Teubenbacher et al Int J Pharm 547 (2018) 169–180

QbD for special dosage forms

Dissolution performance of Sodium

Picosulfate (SP) printed on different
substrates
▪ Comparison of drug release of
printed SP from the different
substrates; release from (a) Gelatin
clear and Gelatin opaque films, (b)
HPMC clear and HPMC opaque
films, (c) yMCC and pMCC films and
(d) Tesa and Listerine films; error
bars are standard deviation (n=3)

Wimmer-Teubenbacher et al Int J Pharm 547 (2018) 169–180

QbD for special dosage forms

Liquid dispensing technology on a tablet core (GSK)

▪ The drug solubilized in a liquid is dispensed on the tablet
QbD for special dosage forms

▪ The process steps and relevant controls

QbD for special dosage forms
Technology equiped with
multiple PAT tools:
▪ 100 % viusal control of tablet
before and after manufacturing
▪ Wet dose position verification
▪ UV solution analysis
▪ Droplet weight check & drop
volume/dose content
▪ Temperature control drying
tunnel
▪ NIR dose position inspection
▪ Coating inspection
QbD for special dosage forms

Laboratory (and commercial) scale equipment (e.g. Lab Machine OPTIMA TDC125)
▪ Transdermal patches and Oral film strips
https://www.optima-
packaging.com/en/machine-
solutions/lab-machine-optima-
tdc-125
QbD for special dosage forms
▪ Commercial manufacturing and Packaging Line (e.g. OPTIMA TDC300)
▪ Transdermal patches and oral film strips
QbD for special dosage forms
Laboratory and commercial packaging machines (e.g. Optima Life Science's rotary 4SS)
▪ Pouch Packaging Machine seals medical and diagnostic devices into pouches
QbD for special dosage forms
Soft gelatin capsule (SGC) manufacturing process
▪ The SGC process involves the preparation of a gelatin band that is passed through die rolls in
which the shells are formed, while pumps are filling at the same time.
▪ A SGC shell contains gelatin, plasticizer (e.g. glycerol, sorbitol) and colorants.
▪ The shells are soft and wet after manufacturing and are dried down in drying chambers for days
to weeks
▪ The formulations can be hydrophilic and hydrophobic and contain ethanol, LMW PEG (e.g.
PEG 200) and other solvents that are incompatible with liquid filled hard capsules
▪ Migration of shell/formulation components can occur as well as solvents might evaporate
during drying
QbD for special dosage forms
Shell formulation
▪ The formulation of gelatin ribbon contains about 50 % of water upon encapsulation
▪ After encapsulation the capsules are dried down to a certain „hardness“ reflecting about
8 % of water. Drying can take days or several weeks
▪ During drying shell and formulation components can migrate and volatile solvents from
the formulation evaporate
▪ Different shell composition are being used for different formulation characteritics (e.g.
hydrophilic, hydrophobic, etc)
Component Function Typical content (% w/w)
Gelatin Polymeric base 66.3
Glycerine/Sorbitol Plasticizer 33.0
Methylparaben + propylparaben (80/20) Preservative 0.1
Colour Colourant 0.1
Titanium dioxide Opacifier 0.5
Water Solvent/process aid q.s. (0.7–1.3 × of gelatin)
QbD for special dosage forms
Soft gelatin capsules can be filled with liquids, paste, and semi solids. Typical vehicles
used in soft capsule formulations are:
▪ Water-immiscible volatile and non-volatile liquids such as vegetable and aromatic oils,
aromatic and aliphatic hydrocarbons, chlorinated hydrocarbons, ethers, esters, alcohols, and
organic acids.
▪ Water-miscible non-volatile liquids, such as polyethylene glycols (LMW PEG), and nonionic
surface-active agents, such as polysorbate 80.
▪ Water-miscible and relatively non-volatile compounds such as propylene glycol (up to ~8%)
and isopropyl alcohol, depending on factors such as concentration used and packaging
conditions
▪ The processing conditions of > 35°C for soft gelatin capsules should be avoided as well as
compatibility studies with the final formulation is recommended
QbD for special dosage forms
▪ Soft capsule are manufactured within the finished doage form manufacturing by
simultaneous shell formation, filling, sealing and drying

1. Fill formulation
2. Shell formulation
3. Shell formation/filling
4. Fast drum drying
5. Final drying on trays
6. Counting
7. Packaging
8. Dispatching
QbD for special dosage forms

▪ The most common soft gelatin Soft capsule manufacturing videos

process is the rotary die
principle https://www.youtube.
https://www.youtube.
com/watch?v=pu9bv
com/watch?v=pu9bv
GlCxVc
GlCxVc
https://www.youtube.
https://www.youtube.
com/watch?v=JvOI3
com/watch?v=JvOI3
12TYmA
12TYmA
https://www.youtube.c
https://www.youtube.c
om/watch?v=Kht6L8v
om/watch?v=Kht6L8v
KA-U
KA-U
QbD for special dosage forms

In-process controls Final controls

1. Gel ribbon thickness and uniformity 1. Sealing & leakage
across the ribbon 2. Potency and impurity content
2. Softgels seal thickness at the time of
3. Weight variation
encapsulation
4. Microbiology
3. Weight of the capsule fill and its variation
from capsule-to-capsule 5. Disintegration/Dissolution

4. Weight of the capsule shell and its 6. Content uniformity

variation from capsule-to-capsule 7. Water content/hardness
5. Moisture level of the capsule shell before 8. …
and after drying
QbD for special dosage forms
Size and shape
▪ Determined by the die rolls
▪ Influenced by the fill formulation and drying
▪ Nearly unlimited in fill volume and shape options

Die rolls
QbD for special dosage forms
Oral liquid manufacturing includes:
▪ Dosing/feeding/metering
▪ De-aerating
▪ Heating/cooling
▪ Homogenizer/Mixer
▪ Filtration
▪ Cleaning in place
▪ Continuous or batch
QbD for special dosage forms
▪ Suppositories are manufactured by forming the blister with the upper end open, filling the
blister with the hot melt of the suppository formulation and sealing of the upper end of the
blister
Blister filling station

Blister formation from two foils

Blister sealing station

https://www.youtube.com/watch?v=Why4cN0OyC0
Lipid nanoparticles and mRNA
Lipid nanoparticles and mRNA

mRNA vaccines – the most important launch

2020/21
▪ mRNA vaccines have emerged and progressed
very fast due to the immediate need in the pandemic
situation
▪ mRNAs are a relatively new biotherapeutic approach
in immunology
▪ The formulation consist of new excipients and new
manufacturing technology
Lipid nanoparticles and mRNA

Moderna Science Day 2020

Lipid nanoparticles and mRNA
Chemical modification of IVT mRNA Untranslated regions

As native mRNAs, the inclusion of the three important domains is required:

1) the 7 methylguanosine cap (5ʹ cap)
2) 2) the 5ʹ and 3ʹ untranslated regions (UTRs)
3) 3) the poly(A) tail
The 5ʹ cap is a methylated guanosine appended at the 5ʹ end (5ʹ to 5ʹ triphosphate linkage). protects
the mRNA from degradation by exonucleases and facilitates ribosomal recruitment and mRNA
translation. mRNA stability enhanced by the poly(A) tail, a chain of 50–100 adenosines (3ʹ end.

Extracellular and intracellular hurdles including: i) serum ribonucleases, which may hydrolyze RNAs;
ii) non-specific interactions with serum proteins; iii) resident organ-specific macrophages, which may
sequester mRNAs from the bloodstream; iv) the vascular endothelium; v) the cell surface membrane;
vi) the intracellular endosomal barriers
Ibba et al Advanced Drug Delivery Reviews 177 (2021) 113930
Lipid nanoparticles and mRNA

Challenges in mRNA
delivery
▪ Degradation by
extracellular
ribosomes
▪ Transfection, cellular
up-take
▪ Endosomal
clearance
▪ Intracellular release
to the ribosomes
Lipid nanoparticles and mRNA

▪ Composition of mRNA vaccines

▪ Containing new excipients: (1) ionic lipid (2) DSPC (3) Pegylated lipids

Crommelin et al J Pharm Sci 110:997-1001 (2021)

Lipid nanoparticles and mRNA
Quantitative composition of
BionTech and Moderna SARS CoV
2 vaccine
▪ mRNA (5 cap, poly-A tail and
untranslated regions (UTRs)
modified)
▪ Ionic lipid
▪ DSPC
▪ PEG-lipid
▪ Cholesterol
▪ pH adjusted to pH 7-8
Lipid nanoparticles and mRNA

▪ The ionic lipids is the mRNA binding and protecting excipient

BionTech
Moderna
Buschmann et al Vaccines 9, 65 (2021)
Lipid nanoparticles and mRNA
▪ Functionality of the lipids
Lipid nanoparticles and mRNA
negatively charged backbone
▪ The ionic lipids are optimized of the mRNA
for the mRNA
▪ They stabilize the mRNA and
lipophilize them
cationic lipid

Moderna Science Day 2020

Lipid nanoparticles and mRNA

▪ The manufacturing process sequence

Moderna Science Day 2020

Lipid nanoparticles and mRNA

▪ Aqueous phase is mixed with an ethanolic

phase (missible, antisolvents)
▪ Nucleation occurs quickly during pH
increase (formation of mRNA-ionic lipid
complexes) => sharp drop of mRNA and
ionic lipid concentration
▪ DSPC, Cholesterol, PEG-lipids assemble
around the mRNA-ionic lipid complex to
form nanoparticles in the aqueous solvent
▪ The microfluidic system assure a precise
and reproducible microenvironment during
the nanoparticle formation process

Buschmann et al Vaccines 9, 65 (2021)

Lipid nanoparticles and mRNA
Micro-fluidic manufacturing of mRNA
vaccines

https://www.precisionnanosystems.com/platform-
technologies/product-comparison/spark#videos

Schematic illustration of microfluidic device with 69 cycle

numbers of staggered herringbone micromixers (SHM).

Maeki et al RSC Adv.,2015,5, 46181–46185

 LNP ready to
Lipid nanoparticles and mRNA burst formation of liposomal
structures containing
mRNA
▪ The stability of the mRNA lipid nanoparticles is
limited
aggregation
▪ After manufacturing their are frozen to -60 to –80°C liberated mRNA
(BionTech) and -20°C (Moderna) for 6 months (due
to limited data available at emergency authorization
date)
▪ The nanoparticles are sensitive to temperature and
mechanical stress in the liquid form
▪ Stability in syringes (lubricated withmedical silicon
oil, <0.25 mg/cm) confirmed for up to 8 h at ≤ 25°C

Brader et al Biophys J 120, 1–5 (2021)

Lipid nanoparticles and mRNA
▪ The particle size of the
nanoparticle matters for 3wp1 2wp2
the cellular up-take and
immune response
▪ The micro-fluidic settings
determine the particle
size of the nanoparticles

3wp1 = 3 weeks after 1st

2wp2 = 2 weeks after 2nd
Dosing schedule 3 weeks between
1st and 2nd dose
Balb/C mice IM administration 3 µg dose CMV mRNA
Moderna Science Day 2020
Lipid nanoparticles and mRNA
Conclusion
▪ A decade of mRNA nanoparticle science enabled the Covid
19 vaccine development in record time
▪ The nanoparticle internal structure is still under investigation
▪ The mRNA and the ionic lipid forms a distinct amphiphilic
complex
▪ The complex is embedded or surounded by a lipophilic
membrane
▪ The excipients, composition and processing is subject of
intensive research
▪ Further progress can be expected

Cryo-TEM image of mRNA-LNP showing

‘bleb・structures with distinctly
different electron density
Purification and fractionation in
up- and downstream processing
Up- and downstream Processing

Major need for downstream processing

▪ API synthesis and drug product (purification)
▪ Up-concentration
▪ Particle separation (fractionation)
▪ Buffer exchange

Hanke & Ottens Trends Biotechnol 32(4):210-20 (2014)

Up- and downstream Processing
Research Development Clinical Commercial
• Small scale • Fractionation and
• Impurity profiling • Fractionation and
fractionation and purification process
• Fractionation and
Objective

purification process
purification development and
purification tests profiling
• Quality by Design scaling up
• Compound & • Economical and
program • Process
impurity ecological
• Process optimization
characterization optimization
qualification • Process validation
• Process • Clinical scale
developemnt equipment trials
• Up-stream process • Commercial
design • Process
Approach

review process review

• Equipment qualification
• Compound profiling • Process validation
selection • Product conformity
• Heuristic or • Optimization
• Laboratory scale analysis
experimental strategy
process evaluation • CMC
design
• Product & Process Documentation
specification
Up- and downstream Processing

Filtration methods
surface filtration are defined by their pore size
depth filtration are defined by their pore size, depth filters are
porous matrices purifying effectively throughout the entire depth
cross-filtration like Tangential flow filtration (TFF) leverages a
high velocity flow tangentially across the membrane surface to
purify proteins or nanoparticles.
Diafiltration (DF) is a cross-flow filtration technique in which
fresh solvent is added to replace the volume loss due to the
filtration process
Ultra-filtration (UF) or nano-filtration (NF) are pressure-driven
membrane transport technologies passing mixtures through
hollow fibers of membrane materialultra-filtration technologies.

Baseer Thesis Uni Saarbrücken (2019)

Up- and downstream Processing

Commerial scale Tangential Flow Filtration equipment

Sartoflow® Expert SU TFF System KrosFlo TFF

(Sartorius) (Repligen)
Up- and downstream Processing
Chromatograpic methods
▪ Consist of a stationary phase (e.g. resin) and a liquid phase of the mixture
▪ Chromatography is a versatile purification technique based on compound specific properties
▪ size, shape
▪ Charge
▪ isoelectric point
▪ charge distribution
▪ Hydrophobicity
▪ Solubility
▪ Density
▪ ligand-binding affinity
▪ metal binding
▪ reversible association
▪ posttranslational modifications
▪ specific sequences or structures
Up- and downstream Processing
Chromatographic methods
Chromatographic purification principles
Affinity (AFC) separation method based on a specific binding interaction between an
immobilized ligand and its binding partner
Ion exchange (CEX, AEX) Separation based on ions and charged molecules by affinity to the ion
exchanger
Hydrophobic interaction Separation of molecules according to differences in their surface
(HIC) hydrophobicity
Size exclusion (SEC) Separation of molecules based on their size by filtration through a gel

Reverse phase (RPC) Separation by a hydrophobic stationary phase and a polar mobile phase

Multimodal (MMC) Separation by more than one form of interaction between the stationary
phase and analytes
Up- and downstream Processing
CPP CQA
▪ Column capacity • Peak retention time
▪ Pressure • Peak area
▪ Solvent • Amount
▪ pH • Peak height
▪ temperature • Peak width at half height
▪ Speed (velocity) • Peak symmetry
▪ feeds (desity) • Peak tailing
▪ Mass transfer • Capacity factor (k)
▪ Voltage • Plate numbers
▪ Time • Resolution between peaks
▪ torque • Selectivity

DeLuca et al Trends in Analytical Chemistry 132 (2020) 116051

Up- and downstream Processing
Up- and downstream Processing
Purification of a triterpene glycoside saponin and determination
triterpene glycoside
saponin analyzed by
LC/Q-TOF MS
triterpene glycoside
saponin peak

Qi & Fox J Chromatograph A 1635 (2021) 461705

Up- and downstream Processing

Sustainability challenge HPLC = High-Performance Liquid

Chromatography
▪ There is a huge push from regulatory agencies on IEC = ion-exchange chromatography
sustainable manufacturing SEC = size-exclusion chromatography

▪ Actually for 1000 - 5000 Da peptides, the volume of

waste is estimated at 3000 - 15 000 kg/kg of API
including hazardous solvents and reagents
▪ Significant efforts in making peptide chemistry greener
by green solvents, reaction conditions, auxiliary
reagents, purification technologies in peptide and
organic synthesis and continuous processing

Distribution of purification technology used

Ferrazzano et al Green Chem., 2022, 24, 975

Parenteral dosage forms
Parenteral dosage forms
Manufacturing of aseptic and sterile products
▪ Aseptic and steril products consist of a tight packaging and or device component
▪ Aseptic products are always adminstered through a medical device component
▪ Prefilled syringe components
Parenteral dosage forms

Regulatory framework of drug-device combination products

▪ In Europe, if the device and the medicinal product form a single integral product which is
intended exclusively for
use in the given combination and which is not reusable is regulated as medicinal
products according to Directive 2001/83/EC. (including the essential requirements for safety-
and performance-of the device features)
▪ In contrast, devices with exchangeable cartridges, for examples, reusable pumps are
regulated in the EU as medical devices according to Directive 93/42/EEC
▪ In the USA, both device-drug combinations are regulated as combination products according
to 21 CFR Part 3.
Parenteral dosage forms
Why do we need parenteral drug delivery?

− Acute drug interventions – immediate pharmacodynamic activity (e.g. stroke, myocardial

infarction)
− Drug properties (e.g. poor solubility/permeability, first pass metabolism, inter - intra subject
variability)
− Presystemic metabolism/degradation (e.g. proteins)
− Unconscious or ananesthetized patients
− Drugs with high therapeutic doses
− Patients with swallowing issues
− Drug titration or constant plasma concentrations (e.g. short half life drugs)
− Tissue or organ targeted drug delivery
Parenteral dosage forms

▪ Buildings and facilities

Parenteral dosage forms

▪ Personnel training, Qualification & Monitoring

▪ Components and container/closure
▪ Endotoxin control
Parenteral dosage forms
▪ Validation of aseptic processing and sterilization
Parenteral dosage forms
▪ Laboratory controls & sterility testing
Parenteral dosage forms
Administration of aseptic Major requirements for Formulation and drug
product injectables delivery approaches

▪ Infusion (large volumes) – Safety/toxicity Aqueous solutions

injection (small volumes) Immunogeneity Emulsions
▪ Administration (intraveneous,
pH Active compound conjugates
intramascular, subcutaneous, Impurities Biodegradable particles
intra-occular (intravitreal, Leachables Liposomes
subretinal), intrethecal,... Particles (Lipid) nanoparticles
Sterility Thermo-hydrogeling
▪ Ampoules Absence of mirco-organisms systems
▪ Vials Free of pyrogens
Free from bacterial
▪ Pre-filled syringes endotoxins
▪ Injection-Pen Isotonicity
Biocompatibility
Parenteral dosage forms
▪ Packaging

Consider „extractable volume“ for ampoules and vials

(excess volume)
Parenteral dosage forms
Glass delamination
▪ Is the pH lower than 7 (acidic), an H3O+ exchange will occur with
the most mobile cations, i.e., typically with Na+ or other network-
modifying ions
▪ OH- ions in alkaline solutions (pH > 7) are able to split the structural
units of silicate glasses directly (i.e., the siloxane bridges). During a
basic attack, the backbone of a silicate glass is destroyed
▪ At neutral pH values (around pH = 7), a combined mechanism will be
observed due to the auto-dissociation of water. At first, ion-exchange
reactions occur similar like in acids. By that, the concentration of OH-
in the water raises leading to a start of a base attack
Regulatory guidance Most critical
▪ Pharmacopoeial monographs give detailed recommendations for the areas for
setup of respective studies (see, e.g., EP 3.2.1 and USP <1660>) delamination

Roehl et al Challenges in protein product development (2018) Ed. Mahler &Warne

Parenteral dosage forms

▪ Ingredients and potential leachables found in closure systems

Parenteral dosage forms

Stopper elastomer performance requirements

McAndrew et al Lyophilization of pharmaceuticals and biologics Ed Ward & Matechtschuk (2019)

Parenteral dosage forms

„Two-leg stopper“ for lyophilization products

Schematic of lyophilization stopper displaying

Stopper/vial system. Stopper positions of fluoropolymer film and silicone polymer
is partly inserted—this is the coating.
configuration during (A) neither film nor coating
lyophilization process (B) fluoropolymer film only
(C) fluoropolymer film and silicone polymer coating
(D) silicone polymer coating only
McAndrew et al Lyophilization of pharmaceuticals and biologics Ed Ward & Matechtschuk (2019)
Parenteral dosage forms
▪ CCI is a container system’s ability to prevent bacterial or chemical
The stopper seal the vial
ingress from exceeding the maximum allowable leak limit (MALL)
in 3 ways
▪ USP <1207> Package Integrity Evaluation—Sterile Products
1. land seal
▪ Deterministic: Leakage event detected/measured is based on a
2. transition seal predictable chain of events based on technologies that are readily
3. valve seal controlled/monitored, and yield quantitative data
▪ Probabilistic: Leakage event measurement relies on a series of
sequential and/or simultaneous events, random outcomes
described by probability distributions. associated with uncertainties
and require large sample sizes and rigorous test controls
USP <1207> methods
▪ laser-based headspace analysis (typically oxygen)
▪ high-voltage leak detection
▪ tracer gas leak detection (typically helium)
▪ mass extraction
▪ vacuum decay
▪ tracer liquids (probabilistic)
McAndrew et al Lyophilization of pharmaceuticals and biologics Ed Ward & Matechtschuk (2019)
Parenteral dosage forms
▪ Comparision of different
Container Closure Integrety
(CCI) test methods

Roehl et al Challenges in protein product development (2018) Ed. Mahler & Warne
 Parenteral dosage forms

▪ Needles are available

in different sizes
(length and diameter)
and tips (31G
smallest diameter,
14G largest diameter)
▪ Colors are used to
avoid errors
Parenteral dosage forms

▪ Drug formulation need to be tested

on compatibility with the packaging
component which is much more
critical then for oral forms

Protein
particle

Concentration of protein particles for (a) abatacept PBS (b) abatacept acetate
buffer (c) adalimumab PBS and adalimumab acetate buffer formulation
Krayukhina et al J Pharm Sci 2015
Parenteral dosage forms

Key attributes to build quality into the biologic drug product from manufacturing to patient

Mahler HC et al AAPS Newsmagazine 7:12-17 (2018)

Parenteral dosage forms

Key attributes to build quality into the biologic drug product from manufacturing to patient

Mahler HC et al AAPS Newsmagazine 7:12-17 (2018)

Parenteral dosage forms

Key attributes to build quality into the biologic drug product from manufacturing to patient

Mahler HC et al AAPS Newsmagazine 7:12-17 (2018)

Parenteral dosage forms

▪ Clean air in the manufacturing is key for sterilized products or aseptic preparations
▪ Clean rooms are defined e.g. in WHO guidelines by airborne particles
WHO Technical Report Series, No. 961, 2011
Annex 6
WHO good manufacturing practices for sterile
pharmaceutical products

Clean rooms grades required in operations

Parenteral dosage forms

▪ Clean rooms and clean-air devices should be routinely monitored while in operation and
the monitoring locations based on a formal risk analysis study and the results obtained
during the classification of rooms and/or clean-air devices.

WHO Technical Report Series,

No. 961, 2011
Annex 6
WHO good manufacturing
practices for sterile
pharmaceutical products
 Parenteral dosage forms
▪ Air treatment systems are composed of multiple different filter systems
▪ In duct UVC at outlet is the C band of UV light (UVC), the most effective germicidal light
wavelengths
Parenteral dosage forms
Production decontamination
▪ The industry standard for containment
decontamination is the treatment with
vapor phase hydrogen peroxide (VPHP)
▪ Critical factor to be considered in QbD:
residuals of the VPHP in the
containment atmosphere or absorbed to
the surfaces interacts with the product
▪ Assessment of the product sensitivity to
hydrogen peroxide by a spiking study
(Methionine + H2O2 => Methionine-
oxidized) Pseudo-first-order kinetic reactions of H2O2 consumption by
Met oxidation observed for various concentrations of H2O2
▪ The Spiking depends on the molecule spiked-in mAb1 formulation (10 mg/mL protein concentration)
and its methionine residues represented by solid lines; and in small protein formulation (5
mg/mL) by dashed lines
Grauschopf et al Challenges in protein product development (2018) Ed. Mahler &Warne
Parenteral dosage forms

▪ Early tests of materials of

construction (MOC) to be performed
▪ several materials used within the
containment show different patterns
of hydrogen peroxide uptake
▪ probability of slow decrease of VPHP
levels inside the containment
▪ Potential risk for interaction with the
product during production

VPHP uptake rate (ng/cm3) in various materials of

construction
Grauschopf et al Challenges in protein product development (2018) Ed. Mahler &Warne
Parenteral dosage forms
▪ Consideration of high hydrogen peroxide uptake by aqueous phases

Reverse

Meniscus

Droplet
Hydrogen peroxide uptake into the drug product solution at differently
shaped menisci of drug product solution at the tip of the filling nozzle
Grauschopf et al Challenges in protein product development (2018) Ed. Mahler &Warne
Parenteral dosage forms
▪ Bosch prefilled syringe line
Parenteral dosage forms

▪ Blow-fill-seal machine for prefilled syringes (Brevetti Angela)

Parenteral dosage forms
▪ Aseptic blow-fill-seal machine for bottles (GEA)
Lyophilized dosage forms
 Lyophilized dosage forms
Tg = Glas transition temperature
Lyophilization (freeze drying) TE = Eutectic temperature
Tf = freezing point
▪ the lyophilization process comprises three Freezing point curve
Vapor pressure
water
process steps: solution
Curve water
liquid
− Freezing (1)
Vapor pressure
− primary drying which is characterized Curve solution
solid
by the removal of crystallized water
Triple point water

Vapor pressure (hPa)

by sublimation under vacuum with Triple point solution
gentle heating (2a and 2 b)
− secondary drying in which unfrozen
water is desorbed under vacuum by gas

controlled heating to more elevated

temperatures (3)
Sublimation sphere of ice
Lyophilized dosage forms

▪ Formulations for lyophilization

requires some functional
excipients
▪ Inclusion of additional excipients
that reduce the overall glass
transition temperature of the matrix
(e.g. Sodium choride)
▪ additional tonicity by disaccharides
or bulking agents to enable more
efficient lyophilization
▪ …

Composition of marketed products

Dixon et al Challenges in protein product development (2018) Ed. Mahler &Warne

Lyophilized dosage forms
Successive stages in lyophilization: the evolution of the process parameters and the
corresponding state of the product to be lyophilized

Degobert et al Pharmaceutics 2021, 13, 1112

Lyophilized dosage forms
▪ Examples of freeze-dried nanocapsules and process conditions

Degobert et al Pharmaceutics 2021, 13, 1112

QbD for special dosage forms

▪ Ishikawa diagram of a risk assessment for a lyophilization process

Junckers et al Processes 2021, 9, 1600

Lyophilized dosage forms

▪ The results of the risk assessment revealed the potentially most critical factors

Junckers et al Processes 2021, 9, 1600

Lyophilized dosage forms

▪ The potentially most critical factors

displayed in Occurrence Impact diagram

▪ Design of experiments to evaluate the most critical

factors (temperature and pressure

Junckers et al Processes 2021, 9, 1600

Lyophilized dosage forms
Process consideration for the development of lyophilization
▪ Freezing ramp rate
▪ Freezing temperature
▪ Annealing ramp rate
▪ Annealing temperature
▪ Annealing time
▪ Refreezing rate
▪ Refreezing temperature
▪ Primary drying ramp
▪ Primary drying temperature
▪ Primary drying duration
▪ Chamber pressure
▪ Secondary drying ramp
▪ Secondary drying time
Typical lyophilization cycle
▪ Secondary drying pressure

Dixon et al Challenges in protein product development (2018) Ed. Mahler &Warne

Radhakrishnan et al Challenges in protein product development (2018) Ed. Mahler &Warne
Lyophilized dosage forms
▪ Process parameter to be considered for nanoparticle formulation

Degobert et al Pharmaceutics 2021, 13, 1112

Lyophilized dosage forms

▪ Additional
process
parameter to be
considered

Dixon et al Challenges in protein product development (2018) Ed. Mahler &Warne

Lyophilized dosage forms

▪ The model
concept was
build on the
determined
parameter
▪ MTM is able
to determine
the dry
layerstance
of the
lyophilized
cake

MTM: Manometric temperature measurement

Junckers et al Processes 2021, 9, 1600

Lyophilized dosage forms
Defining the CQA of lyophilised parenterals

▪ Appearance of cake and reconstituted solution

▪ pH
▪ Reconstitution time
▪ Residual moisture
▪ Potency
▪ Concentration or protein content for biotherapeutics
▪ Particulate matter
▪ Content uniformity
▪ Product purity- and impurity-related quality attributes
▪ Sterility
▪ Endotoxin
Lyophilized dosage forms

Other factors to be considered

Lyophilization chamber position

▪ The lyophilization is performed on trays with e.g. 580

vials
▪ The trays are introduced into the lyophilization chamber,
which theoretically should provide the same
temperature and pressure conditions to all vials
Lyophilized dosage forms
▪ 580 vials were arranged in hexagonal
clusters in a bottomless tray
▪ Vials were filled with 0.4 ± 0.02 mL with a
mAbs formulation
▪ The freezing step was performed by
decreasing the shelf temperature from 4 to
50 C at about 0.5 C min ; 3.75 or 7 C min
▪ The vials were then held at constant
temperature for 1 h
▪ Chamber pressure was decreased (3, 5,
and 9 Pa) and shelf inlet temperature was
raised (29, 32 or 35 C) to the target primary
drying temperature in 30 min.
▪ A total of 124 edge vials and 100 central
vials were individually assessed
The bottom product temperature Tb differs within the
lyophilization chamber
Cell- und Gene Therapy (CGT)
Cell- and Gene Therapy

▪ CGT is he fastest growing

research area for
therapeutic innovation
Cell- and Gene Therapy

▪ CGT therapeutics manufacturing is

an emerging field of science and
technology
▪ Quality and GMP standards as well
as guidelines are limited and not
harmonized
▪ The use of cell and genes derived
from patients or donors cause new
challenges for clinical (commercial)
manufacturing and quality control
▪ The demand and investments will
drive the CGT manufacturing
infrastructure
Cell- and Gene Therapy

Gene Therapy Medicinal Products (GTMPs)

▪ in this context DNA is directly used pharmacological substance, as recombinant genes are
inserted into cells. GTMP is defined as “an active substance which contains or consists of a
recombinant nucleic acid used in or administered to human beings with a view to regulating,
repairing, replacing, adding, or deleting a genetic sequence; its therapeutic, prophylactic or
diagnostic effect relates directly to the recombinant nucleic acid sequence it contains, or to the
product of genetic expression of this sequence” (cited from Regulation EC No 1394/2007 of the
European Parliament).

Stem cell production: Processes, practices and regulations Ed. F.A. Khan (2022): https://doi.org/10.1007/978-981-16-7589-8
Cell- and Gene Therapy

Somatic Cell Therapy Medicinal Products (CTMPs)

▪ It is defined as “biological medicinal product which contains or consists of cells or tissues that
have been subject to substantial manipulation so that biological characteristics, physiological
functions or structural properties relevant for the intended clinical use have been altered, or of
cells or tissues that are not intended to be used for the same essential function(s) in the recipient
and the donor” (cited from Regulation EC No 1394/2007 of the European Parliament). Cell
selection, ex vivo expansion of cells, generation of clones with both anti-neoplastic and anti-
infectious activity can be considered for example substantial manipulations. Cells most often
used in the development of CTMPs are adult stem cells, unspecialized cells that can be selected
from various tissues of the body and have high differentiation capacity. These cells are
autologous when obtained from a donor and reinfused in the same person, or allogeneic, when
donor and recipient are two different people.

Stem cell production: Processes, practices and regulations Ed. F.A. Khan (2022): https://doi.org/10.1007/978-981-16-7589-8
Cell- and Gene Therapy

Tissue Engineering Products (TEPs)

▪ It is defined as “a product that contain or consist of cells or tissue engineered by different
methods, including viral or bacterial plasmid vectors” (cited from Regulation EC No 1394/2007 of
the European Parliament). They can be used to repair, regenerate, or replace tissues. The first
products developed include artificial skin, bone, and cartilage.

Stem cell production: Processes, practices and regulations Ed. F.A. Khan (2022): https://doi.org/10.1007/978-981-16-7589-8
Cell- and Gene Therapy

Combined: Advanced Therapy Medicinal Products (ATMP)

▪ It is defined as “a product that contain one or more medical devices as an integral part of the cell
or tissue based medicinal product” (cited from Regulation EC No 1394/2007 of the European
Parliament).

Stem cell production: Processes, practices and regulations Ed. F.A. Khan (2022): https://doi.org/10.1007/978-981-16-7589-8
Cell- and Gene Therapy
▪ CGT is a new and exciting field
for pharmaceutical sciences in
the coming decade
▪ The advantages of in-house GMP
facility in a hospital is the proximity
of a manufacturing facility to the
hospital operating rooms and the
clinical units.
▪ Enable on demand manufacturing
at lower cost, rapid quality control,
product supply and optimal
organization and transport of
cellular material.

Iancu et al Curr Opin Biotechnol 2020, 65:233–241

Cell- and Gene Therapy

▪ Advanced Therapeutic Medicinal Products (ATMP) are emerging from

academic research focusing on life-threatening diseases, rare and orphan
disease
▪ They represent a new type of drug products, which are based on either on
− a gene therapy medicinal product
− a somatic cell therapy medicinal product
− a tissue engineered product
▪ They are handled / approved by the Paul-Ehrlich-Institute (PEI)
▪ They often include a medical device component
Cell- and Gene Therapy

a. ‘Advanced therapy medicinal product’ means any of the following medicinal

products for human use:
a gene therapy medicinal product as defined in Part IV of Annex I to Directive
2001/83/EC,
— a somatic cell therapy medicinal product as defined in Part IV of Annex I to
Directive 2001/83/EC,
— a tissue engineered product as defined in point (b)
b. ‘Tissue engineered product’ means a product that:
— contains or consists of engineered cells or tissues, and
— is presented as having properties for, or is used in or administered to
human beings with a view to regenerating, repairing or replacing a human
tissue.

REGULATION (EC) No 1394/2007 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

Cell- and Gene Therapy

c. Cells or tissues shall be considered ‘engineered’ if they fulfil at least one of the
following conditions:
— the cells or tissues have been subject to substantial manipulation, so that
biological characteristics, physiological functions or structural properties relevant
for the intended regeneration, repair or replacement are achieved. The
manipulations listed in Annex I, in particular, shall not be considered as
substantial manipulations,
— the cells or tissues are not intended to be used for the same essential function
or functions in the recipient as in the donor.

REGULATION (EC) No 1394/2007 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

Cell- and Gene Therapy

d. ‘Combined advanced therapy medicinal product’ means an advanced therapy

medicinal product that fulfils the following conditions:
— it must incorporate, as an integral part of the product, one or more medical
devices within the meaning of Article 1(2)(a) of Directive 93/42/EEC or one or
more active implantable medical devices within the meaning of Article 1(2)(c) of
Directive 90/385/EEC, and
— its cellular or tissue part must contain viable cells or tissues, or
— its cellular or tissue part containing non-viable cells or tissues must be liable to
act upon the human body with action that can be considered as primary to that of
the devices referred to.

REGULATION (EC) No 1394/2007 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

 Cell- and Gene Therapy
▪ Tissue Engineered Products (TEP) are individualized therapies manufactured
manually or by automated systems

The manufacturing process is a complex

sequence of different process steps

1. cell preparation, eg thawing & washing

2. selection,
3. activation,
4. transduction,
5. expansion,
6. harvest and
7. final formulation of the cells
Haeusner et al Product. Front. Med. 8:712917
Cell- and Gene Therapy

▪ Manufacturing of viral
vectors for cell
transfection

https://www.nature.com/articles/d42473-021-00020-x
Cell- and Gene Therapy

▪ Viral vectors used

https://bioprocessintl.com/2016/emerging-platform-bioprocesses-for-viral-vectors-and-gene-therapies/
Cell- and Gene Therapy

Overview of the main types of capsids generated during rAAV production

Gimpel et al Mol Ther Meth Clin Dev 20, 740-54 (2021) https://doi.org/10.1016/j.omtm.2021.02.010.
Cell- and Gene Therapy

The importance of each characteristic during

selection of analytical methods during process
development is based on guidance for validation of
analytical methods and qualification plans during
early-stage process development.
a. Accuracy can be inferred from precision, linearity, and
specificity. It may be difficult to establish due to a lack of
adequate standards.
b. Reproducibility only gains importance for lab-to-lab transfer
and method standardization.
c. The detection and quantification limits are relevant mainly
to assays quantifying low levels of impurities.
d. Range strongly overlaps with linearity, precision, and
accuracy.
e. Robustness is considered in later stages of assay
development.

Gimpel et al Mol Ther Meth Clin Dev 20, 740-54 (2021) https://doi.org/10.1016/j.omtm.2021.02.010.
Cell- and Gene Therapy

Analytical methods
used, their
application and
suitability

The methods must

be selected
specifically for each
program and
qualified & validated
accordingly

https://doi.org/10.1016/j.omtm.2021.02.010.
Cell- and Gene Therapy

▪ CAR-T cell therapy is an

individualized drug therapy
▪ Patient leukocytes are harvested
and „equipped“ with a Chimeric
Antigen Receptor (CAR)
▪ They are reinfused into the same
patient under strict clinical survey
conditions and procedures

https://www.ema.europa.eu/en/documents/presentation/presentation-chimeric-antigen-receptor-car-t-cells-david-lebwohl_en.pdf
 Cell- and Gene Therapy

Cleanroom classification for cell therapy facility based on ISO 14644-1 and EU GMP, PIC/S
Cell- and Gene Therapy

▪ Major equipment for cell laboratorium

Leemhuis et al Bone Marrow Transplant (2014), 49, 1098 – 1105

Cell- and Gene Therapy
Airlift Bioreactors
▪ Airlift bioreactors are used where the injection of a gas is made in the culture medium which can cause
the broth to circulate between the riser and an interconnected down comer compartment of the
bioreactor.
Fluidized Bed Bioreactors
▪ In a fluidized bed bioreactor, mixture of culture medium is moved in upward direction through a packed
bed of immobilized cells suspends them inducing a fluid-like behavior.
Packed Bed Bioreactors
▪ Packed bed bioreactors are used in cell and tissue engineering applications. These bioreactors
support the growth and expansion of different types of cell lines for long period of time under various
culture conditions.
Photo-Bioreactors
▪ In photo-bioreactor, light source is used to cultivate phototrophic microorganisms. These phototrophic
microorganisms apply photosynthesis process to generate biomass by converting light energy into
biomass using light and CO2.
Cell- and Gene Therapy
Different technologies used to
expand CTMPs. Production
processes development begins
using flasks or multilayer
systems at preclinical level
(standard culture systems), then
the scale-up of the processes
occur mainly using bioreactors,
to obtain adequate cells number
for patients treatment. The blue
table indicates the best system
concerning the expansion
capacity, the timing of
expansion, the scalability, the
safety of the final product, and
the costs.
Stem cell production: Processes, practices and regulations Ed. F.A. Khan (2022): https://doi.org/10.1007/978-981-16-7589-8
Cell- and Gene Therapy

Conceptualized automation of ATMP

manufacturing
▪ The central six axis dual arm robot
can reach the circumference.
▪ Necessary equipment, disposables
and liquids are safely channeled in
through an air lock linked glove box
without the need of personnel
entering the isolator directly from an
unclassified maintenance back side.
▪ Pre-packed, sanitized disposables,
materials or biologicals can be
unpacked easily and set in place for
robot-driven procession. Devices for
cell culture (green) and quality control
(blue) are included in the design.
 Cell- and Gene Therapy

01 Ventilation system 14 Cell counting device

02 Freezer (-20◦C) 15 Storage for Cell culture
03 Fridge (4◦C) tubes
16 Six axis dual arm robot
04 Disposables storage 17 Centrifuge
05 Packaging material 18 Plate handling positions
06 Incubator (37◦C) 19 Decapper (centrifuge
07 Storage for plates flask)
and membranes 20 Decapper (Cell culture
tubes)
08 Gate 21 Tissue grinder
09 Barcode reader 22 Pipettes
10 Washing station 23 Liquid waste
11 Sampling station 24 Sealing machine
12 Shaker 25 Solid waste
26 Microscope
13 Air-lock
Cell- and Gene Therapy

Manufacturing of an allogeneic T-cell product genetically modified to express a chimeric

antigen receptor T cell.

M. Scott et al. Cytotherapy 22 (2020) 669676

Cell- and Gene Therapy
ClinMACS Prodigy platform
(for cell manufactring)

MACSQuant Tyto
(for cell sorting)
Cell- and Gene Therapy

▪ MACSQuant is a flow cytometry

tool for mining and harvestig rare
cells, analyzing the efficiency of
cell manufacturing processes or
investigating signaling pathways

Miltenyi Biotech MACSQuant Tyto Cartridge (cell sorting)

Cell- and Gene Therapy
Ambr 250 high throughput perfusion (Sartorius) - fully-automated bioreactor system for
intensified cell culture process development

High Throughout Perfusion is a

parallel bioreactor system for rapid
development of scalable perfusion
processes using 100 – 250 mL
single-use bioreactors and a fully
automated liquid handling platform.
Cell- and Gene Therapy
Ambr 250 high throughput perfusion (Sartorius) - fully-automated bioreactor system for
intensified cell culture process development

Single-use pressure sensors to

Cell retention filter standards are
continuously monitor pressure at the
0,2 µm or 30kDa
culture fluid inlet and permeate outlet

Single-use pump chambers to exchange

culture fluid between the bioreactor
and the hollow fibre filter.

Single use perfusion The single-use pinch valve cassette at the

reactor base of the perfusion tower to guide the fluids
through the perfusion tubing for automated
collection.
Cell- and Gene Therapy

Rate of mixing in the bioreactor

Temperature range in different organisms

pH range for different organisms

Stem cell production: Processes, practices and regulations Ed. F.A. Khan (2022)
https://doi.org/10.1007/978-981-16-7589-8
Cell- and Gene Therapy

▪ Scaling up equipment for routine larger scale production (Sartorius)

Qbd for Medical devices and combination
products
 QbD for Medical Devices and combination products
A ‘medical device’ means any instrument, apparatus, appliance, software, implant, reagent, material or
other article intended by the manufacturer to be used, alone or in combination, for human beings for one
or more of the following specific medical purposes:
− diagnosis, prevention, monitoring, prediction, prognosis, treatment or alleviation of disease,
− diagnosis, monitoring, treatment, alleviation of, or compensation for, an injury or disability,
− investigation, replacement or modification of the anatomy or of a physiological or pathological process
or state,
− providing information by means of in vitro examination of specimens derived from the human body,
including organ, blood and tissue donations,
− and which does not achieve its principal intended action by pharmacological, immunological or
metabolic means, in or on the human body, but which may be assisted in its function by such means.
− The following products shall also be deemed to be medical devices:
− devices for the control or support of conception;
− products specifically intended for the cleaning, disinfection or sterilisation of devices as referred to in
Article 1(4) and of those referred to in the first paragraph of this point.
Regulation EU 2017/745
QbD for Medical Devices and combination products
Usability studies in the targeted
patient population and their
environment is required
▪ If the device has not been used in the
proposed patient population before
or if the setting of use is new and
different from the intended use as
confirmed by the certificate of
conformity or NBOp (e.g. a prefilled
syringe used for the first time in an
outpatient setting or used for the first
time in patients with conditions which
could impair use), a usability study –
to evaluate whether the DDC can be
used safely to deliver the medicinal
product to the target population - is
expected
QbD for Medical Devices and combination products
FDA regulation

This guidance recommends that

manufacturers follow human factors or
usability engineering processes during
the development of new medical
devices, focusing specifically on the
user interface, where the user interface
includes all points of interaction
between the product and the user(s)
including elements such as displays,
controls, packaging, product labels,
instructions for use, etc
QbD for Medical Devices and combination products
Hazards traditionally considered in risk
analysis include:
▪ Physical hazards (e.g., sharp corners or
edges),
▪ Mechanical hazards (e.g., kinetic or
potential energy from a moving object),
▪ Thermal hazards (e.g., high-temperature
components),
▪ Electrical hazards (e.g., electrical current,
electromagnetic interference (EMI)),
▪ Chemical hazards (e.g., toxic chemicals),
▪ Radiation hazards (e.g., ionizing and non-
ionizing), and
▪ Biological hazards (e.g., allergens, bio-
incompatible agents and infectious agents).
 QbD for Medical Devices and combination products
The ability of a user to operate a medical device depends on his or her personal characteristics, including:

▪ Physical size, strength, and stamina, ▪ Mental and emotional state,

▪ Physical dexterity, flexibility, and coordination, ▪ Level of education and health literacy
▪ Sensory abilities (i.e., vision, hearing, tactile relative to the medical condition involved,
sensitivity), ▪ General knowledge of similar types of
▪ Cognitive abilities, including memory, devices,

▪ Medical condition for which the device is being ▪ Knowledge of and experience with the
used, particular device,

▪ Comorbidities (i.e., multiple conditions or ▪ Ability to learn and adapt to a new device,
diseases), ▪ Willingness and motivation to learn to use a
▪ Literacy and language skills, new device

▪ General health status,

QbD for Medical Devices and combination products

Similar to the Risk Assessment for the drug development, HFD includes a task
analysis to answer the following questions:

▪ What use errors might users make on each task?

▪ What circumstances might cause users to make use errors on each task?
▪ What harm might result from each use error?
▪ How might the occurrence of each use error be prevented or made less frequent?
▪ How might the severity of the potential harm associated with each use error be
reduced?
QbD for Medical Devices and combination products

Medical device user interface framework

QbD for Medical Devices and combination products
Medical device regulations

https://eur-lex.europa.eu/legal-
content/EN/TXT/PDF/?uri=CELEX:020
17R0745-20170505&from=EN

Medicines Agency Guidance

https://www.ema.europa.eu/en/hum
an-regulatory/overview/medical-
devices
QbD for Medical Devices and combination products

▪ Medical Device classification is based on potential risk of the device (I = low; III = high)
▪ Conformity assessment is the method by which a manufacturer demonstrates that their
devices comply with the requirements of Directive 93/42/EEC
QbD for Medical Devices and combination products
There are two types of combination:
− integral: the medicinal product and device form a
single integrated product e.g. pre-filled syringes
and pens, patches for transdermal drug delivery
and pre-filled inhalers;

− co-packaged: the medicinal product and the device

are separate items contained in the same pack e.g.
reusable pen for insulin cartridges, tablet delivery
system with controller for pain management.
QbD for Medical Devices and combination products
▪ Combination products in the USA are drug-device products (fixed dose combinations are
drug-drug products)
▪ The regulations emerged early this century with the increasing number of targeted and
personalized threatments
QbD for Medical Devices and combination products

FDA Guidance treats the device part and the

drug part as separate constituents
▪ The cGMP requirements for constituent parts
of cross-labeled combination products that are
entirely manufactured at separate facilities are
the same as those that would apply if these
constituent parts were not part of a
combination product (e.g., for a drug/device
combination product, only parts 210 and 211
(21 CFR parts 210 and 211) would apply to the
manufacture of the drug constituent part(s) of
the cross-labeled combination product, and
only part 820 (21 CFR part 820) would apply to
the device constituent part(s)).
QbD for Medical Devices and combination products

Typical combination products are Inhalers

▪ three different types exist
− pMDI (pressurized metered dose inhaler)
• Propellant driven
• Mechanical pressure driven pMDI
Nebulizer device
− Nebulizer solutions
− Dry Powder Inhaler
• Capsule based metered
• Blister based metered
• Reservoir based metering chambers

Blister based metered

Reservoir based
metering chamber
QbD for Medical Devices and combination products
The DPI product consist of four elements of equal importance for the performance
1. Interactive powder mixture (API, powder blend, porous particles)

2. Powder container/primary packaging (capsule, blister, reservoir)

3. Device

4. Processing

(1) (2) (3)

QbD for Medical Devices and combination products

The critical factors for dry powder manufacturing

Input materials
▪ Material propertise (particle size, size distribution, surface propertise, triboelectrics, physical
state, carrier coarse-fine particle ratio, storage history, moisture content, etc)
Room conditions
▪ Room temperature, relative humidity
Particle size reduction/mixing
▪ Type and geometry of miller/mixer, load, order of entry, time & intensity of milling/mixing, etc)
Filling
▪ Type of filling equipment, nozzle size, diameter, length, machine speed/vibration, powder
transport, powder flow, powder bed height, nozzle diving depth, etc
QbD for Medical Devices and combination products
Harro Höfliger microdosator filling

Courtesy of K-H Seyfang from Harro Höfliger

QbD for Medical Devices and combination products
Harro Höfliger drum filling

Courtesy of K-H Seyfang from Harro Höfliger

QbD for Medical Devices and combination products
Propellant free liquid inhaler systems
▪ Mechanical pressure driven inhalation device, manufactured and assembled in one operation
▪ Respimat, the first propellant free liquid inhaler system spring

nozzle piston=capillary cartridge

Cross-section of Respimat®
H. Wachtel, Patient-centric inhalation product design AAPS Workshop Oct. 2012
QbD for Medical Devices and combination products

▪ The main element of the

Respimat is the Uniblock
whereby two fine jets of
liquid converge at a preset
angle
▪ The collision of the jet
liquids form 65-80 % of fine
mist particles of < 5.8 µm,
delivering a dose of 2.5 µg
tiotropium
QbD for Medical Devices and combination products

▪ Excellent delivery performance across 120

actuations...

▪ ...but very expensive and complex and

prone to errors for patients
 QbD for Medical Devices and combination products
Performance testing of DPI products
▪ Three different DPI product concepts
▪ Targeted performance is verified in e.g. Next Generation Impactors

Blister-based system Multidose reservoir Capsule – based

system system
QbD for Medical Devices and combination products

Engineered particles for pulmonary delivery

▪ An emerging field in pulmonary drug delivery is the manufacturing of engineered
particles in the size range < 5 µm using of Spray Drying
▪ Engineered particles are typically used for biologics, drugs with poor crystalline
geometry and high dosed drugs
▪ They contain excipients that provide specific functionality to achieve a desired
aerodynamic particle size/shape
▪ The suitable excipients are limited due to safety and toxicity issues/concerns for
pulmonary applications
▪ Engineered particles are used for e.g. tobramycin (TOBI™ Podhaler™) or have been
used for e.g. insulin (Exubera™ – withdrawn from the market),
QbD for Medical Devices and combination products
Key process parameter

Spray Solution
▪ Stability versus process time
▪ Shear, pH, concentration, interactions
2 Atomizer
Atomization
▪ Define target particle size
▪ ƒ (geometry, pressure) 4
Drying Conditions Cyclone
▪ Product morphology Separator
3 Drying
▪ Water content Chamber
▪ Physical state
Powder
▪ ƒ (TIn Tout Msoln Mgas) Solution 5 Handling
1
Collection Efficiency Tank
▪ High Value Product
▪ ƒ (geometry, product properties)
 QbD for Medical Devices and combination products
Chemical classes of excipients used for particle engineering of DPI
Excipients selection for
spray drying DPI Amino acids Sugars/polyols Polymers Lipids
Alanine lactose Chitosan Cholesterol
formulation arginine Mannitol Dextran/dextran-10 Phospholipon
cysteine Raffinose Hypromellose Phospholipids
▪ should be kept at a Glycine Sucrose Polyethylenglycole Phosphatidylcholine
minimum Leucine/isoleucine Trehalose Polyvenylalcohol
Lysine Polyvenylpyrrolidone
▪ should provide a Methionine
required functionality Phenylalanine
Serine
Threonine
Trileucine
Tryptophan
tyrosine
valine
Possible functionalities provided
Stabilization of API Stability of API Stability of API Aerodynamic Properties
Surface properties Particle Morphology Morphology Dispersibility
Particle and surface Solubility Particle density
Morphology Dispersability Drug release
Flowability Dispersibility
dispersability
QbD for Medical Devices and combination products
▪ Increasing drug load in carrier based powder mixtures
can be targeted release higher doses and emitted FPF
▪ Jet milled (JMSS) and Spray dried salbutamol (SDSS)
mixtures with lactose (Lactohale 100) at 1 % and 10 %
drug load were filled at 5 and 10 mm powder bed height
▪ The FPFED was 29 % (152 µg) for 1 % drug load and 62 %
(3633 µg) for 10 % drug load

Pinto et al J Drug Del Sci Technol 48: 466–477 (2018)

QbD for Medical Devices and combination products
Manufacturing of DPI multidose inhalers (e.g. NEXThaler)
▪ Complex design, with > 22 high precision pieces
▪ Requires an integrated manufacturing line
▪ Single use - waste
QbD for Medical Devices and combination products
Line concept for manufacture of blister–based DPIs
3
Blister manufacturing (1)
Blister strip buffer (2)
Blister coiling and Inhaler assembly (3)

1 2

The manufacturing line are product specific and are not easily scalable
Conclusion
Conclusion
Conclusion
“Using traditional pharmaceutical manufacturing technology, a U.S.-based company
could never offset the labor and other cost advantages that China enjoys simply by
achieving higher productivity. However, FDA believes that advanced manufacturing
technologies could enable U.S.-based pharmaceutical manufacturing to regain its
competitiveness with China and other foreign countries, and potentially ensure a
stable supply of drugs critical to the health of U.S. patients. Advanced manufacturing is
a collective term for new medical product manufacturing technologies that can improve
drug quality, address shortages medicines, and speed time-to-market. Every field has
a different set of production techniques that are considered advanced. Examples of
some cross-cutting advanced manufacturing technologies include continuous
manufacturing and 3D printing. Advanced manufacturing technology, which FDA
supports through its Emerging Technology Program (ETP), has a smaller facility
footprint, lower environmental impact, and more efficient use of human resources than
traditional technology, as will be explained later in this testimony.”
Janet Woodcock Oct 30, 2019
Univ.-Prof. Dr. Sven Stegemann
Graz University of Technology
Inffeldgasse 13
8010 Graz
Austria
e-mail: [email protected]
Phone: +43 316 873 0422
Mobile: +49 172 6054869
Fax: +43 (316) 873 - 1030422