You are on page 1of 54





EDITORIAL 4 Excipient Quality and Selection

Editorial Director Rita Peters Irwin B. Silverstein
Senior Editor Agnes Shanley
Managing Editor Susan Haigney EXCIPIENT DATABASES
Science Editor Adeline Siew, PhD
Manufacturing Editor Jennifer Markarian 16 Unifying Excipient Databases
Science Editor Randi Hernandez Agnes Shanley
Community Manager Caroline Hroncich
Contributing Editors Jill Wechsler; Jim Miller; Hallie Forcinio; Susan J. Schniepp 20 Establishing a New Performance; Eric Langer; Standard for HPMC Capsules
and Cynthia A. Challener, PhD
Correspondent Sean Milmo (Europe, Agnes Shanley
485 Route One South, Building F, Second Floor, Iselin, NJ 08830, USA
Tel. 732.596.0276, Fax 732.647.1235, 22 Modeling and Simulation
SALES Move Downstream
Publisher Mike Tracey Agnes Shanley
Mid-West Sales Manager Irene Onesto
East Coast Sales Manager Joel Kern TABLET COMPRESSION
European Sales Manager Linda Hewitt
European Senior Sales Executive Stephen Cleland
26 Optimizing Tablet Compression
Executive Assistant Barbara Sefchick
Frederick J. Murray
485 Route One South 34 Filling the Analysis Gap in
Building F, Second Floor the Move to Continuous Processing
Iselin, NJ 08830, USA Jamie Clayton
Tel. 732.596.0276, Fax 732.647.1235 ELEMENTAL IMPURITIES

40 Meeting USP Guidelines for

Sr. Production Manager Karen Lenzen Elemental Impurity Analysis with
International Licensing Maureen Cannon, X-ray Fluorescence Spectrometry
tel. 440.891.2742 or toll-free 800.225.4569 ext 2742, fax. 440.756.5255
Andrew Fussell
Audience Development Manager Rochelle Ballou

Pharmaceutical Technology does not verify any claims or other information appearing in any of 46 Regulatory Considerations for Controlling
the advertisements contained in the publication and cannot take any responsibility for any losses
or other damages incurred by readers in reliance on such content. Pharmaceutical Technology Intermediates in Type-II Drug Master Files for
welcomes unsolicited articles, manuscripts, photographs, illustrations, and other materials but
cannot be held responsible for their safekeeping or return. UBM Life Sciences provides certain
the Manufacture of Generic Drug Substances
customer contact data (such as customers’ names, addresses, phone numbers and e-mail addresses) Kandasamy Subburaj, Brian T. Connell, Srinivasa Murthy, Humcha
to third parties who wish to promote relevant products, services, and other opportunities which may
be of interest to you. If you do not want UBM Life Sciences to make your contact information available Hariprakasha, Deborah F. Johnson, Huyi Zhang, and David J. Skanchy
to third parties for marketing purposes, simply call toll-free 866.529.2922 between the hours of 7:30
am and 5 pm CT and a customer service representative will assist you in removing your name from
UBM Life Sciences’ lists. Outside the United States, please phone 218.740.6477. To subscribe: Call
54 Ad Index
toll-free 888.527.7008. Outside the US, call 218.740.6477. Single issues, back issues: Call toll-free Issue Editor: Agnes Shanley.
800.598.6008. Outside the US call 218.740.6480. Reprints of all articles in this issue and past issues of On the Static Cover: Science Photo Library/Getty Images; Dan Ward.
this publication are available. Call 877-652-5295 ext. 121 or email Outside On the Animated cover: Nanihta photography/DAJ/Tetra Images/Yuri_Arcurs/
Empato/E+/Image Source/Still Factory/Getty Images; Dan Ward.
US, UK, direct dial: 281-419-5725. Ext. 121. Direct mail lists: Contact Tamara Phillips, Marketing
Services, tel. 440.891.2773, Display, Web, Classified, and Recruitment ©2016 UBM. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any
Advertising: Contact Tod McCloskey, tel. 440.891.2739, Permissions: means, electronic or mechanical including by photocopy, recording, or information storage and retrieval without
Contact Maureen Cannon, tel. 440.891.2742 or toll-free 800.225.4569 ext 2742, fax. 440.756.5255, permission in writing from the publisher. Authorization to photocopy items for internal/educational or personal use, or the internal/educational or personal use of specific clients is granted by UBM for libraries and other users registered with
the Copyright Clearance Center, 222 Rosewood Dr. Danvers, MA 01923, 978-750-8400 fax 978-646-8700 or visit http:// online. For uses beyond those listed above, please direct your written request to Permission Dept.
fax 440-756-5255 or email:

P h a r mTe c h . c o m
A holistic, systems-based approach to drug
development, manufacture and delivery



At PSE we are working with leading pharma Download the PSE, Pfizer
and Eli Lilly SbP phase 1
companies to develop a holistic approach to progress report

drug development, manufacture and delivery.

We call it Systems-based Pharmaceutics (SbP).
The Advanced Process
A systems-based approach considers the entire formulation- Modelling Company

to-delivery cycle as a single integrated system, relating

formulation and manufacturing decisions directly to drug
delivery KPIs. Operations in UK, USA,
Switzerland, Japan, Korea, China,
SbP is helping pharma companies to accelerate development, Taiwan, Malaysia, Thailand and UAE.
improve manufacturing, enhance drug product performance and UK Head Office
increase R&D efficiency. Contact us to find out more. t: +44 20 8563 0888

More online

Excipient Quality
and Selection
Irwin B. Silverstein

Choosing the right ow does one define quality as applied to excipients? If
excipient manufacturer we pose the same question for APIs, the response would
can help ensure the use
be to produce the ingredient under appropriate GMPs,
of quality excipients.
and to the compendial monograph and the API assay.
Because the monograph provides the minimum requirements, API
quality is improved by reducing the presence of all materials other
than the desired chemical. This is logical because, by definition, the
API is “intended to furnish pharmacological activity or other direct
effect in the diagnosis, cure, mitigation, treatment, or prevention
of disease or to affect the structure and function of the body” (1).
Extraneous substances may be harmful to the patient in that they
may lead to side effects, or they are inert, thus reducing API purity
and thereby compromising efficacy.
Excipient quality is described quite differently. While one would
again refer to compliance with the compendial monograph (if there
is one) or the manufacturer’s specification, a higher assay is not al-
ways better. While this may seem counterintuitive, excipients are
often complex mixtures that include constituents arising from raw
materials, catalyst, solvent, initiator residue, or side reactions. The

International Pharmaceutical Excipients Council (IPEC) refers to

these other unavoidable substances in the excipient as concomi-
tant components (2). The performance of many excipients in the
drug formulation may rely on the presence of such substances in
the excipient. Concomitant components in the excipient may aid
in solvating drug components, improving excipient functionality,
etc. Excipient quality, therefore, is characterized as compliance to
Irwin B. Silverstein, PhD, is a the monograph or specification and having a consistent concomitant
consultant to IPEC-Americas.
4 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
Veltek Associates, Inc. offers two garment product lines, which are both pre-folded in
our system. Comfortably styled and fitted with elastic thumb loops to reduce
shifting, as well as tunnelized elastic wrists and ankles.

1600 Garments 1700 Garments Face Masks

Breathable High filtration efficiency Breathable
Comfortable Low particulate and shedding Reduces goggle fogging due
High bacterial efficiency performance to absorption efficiency
Excellent water repelency Soft and comfortable
Specified limits for excipients Povidone is more likely than hydrazine to impact
As required by clause of the ANSI excipi- performance of the excipient in some drug formula-
ent GMP standard, excipient manufacturers are tions. A test method suitable for detecting low levels
expected to identify concomitant components of this monomer was developed, and the compen-
present in the excipient whenever possible and to dium was updated accordingly.
specify limits for those components that have been An additional substance, 2-pyrollidone, was sub-
shown to be either important to excipient perfor- sequently found in Povidone. This substance is
mance or known to have an adverse impact to the formed during the polymerization reaction when
patient (3). Impurities known to be present in the some vinyl pyrrolidone decomposes rather than
excipient are also required to have specified upper polymerizes. While it is unlikely that the presence
limits based upon safety considerations, regulatory of hydrazine or vinyl pyrrolidone beneficially im-
requirements, customer requirements, and, if ap- pacts the performance of Povidone in the drug for-
plicable, the compendium. mulation, the same conclusion cannot be drawn for
Povidone and its monograph illustrate these 2-pyrrolidone. 2-Pyrrolidone is often used as a sol-
points. Povidone is the homo-polymerized mono- vent, and therefore, its presence in the excipient may
mer vinyl pyrrolidone. It is sold in various mo- play a beneficial role in certain drug formulations
lecular weights. In the late 1980s, GAF Chemicals, by helping to solvate the API. While it is possible to
a manufacturer of Povidone, was made aware of remove this substance through further processing, it
the presence of hydrazine, a toxic substance, in is not feasible for the manufacturer to asses the im-
Povidone. The company identified the mechanism pact on performance for all drug formulations that
of hydrazine formation as a by-product of the po- use Povidone. Therefore, it is important to control,
lymerization reaction. Through modification of but not limit, the quantity of 2-pyrrolidone so that
the process, the level of hydrazine was reduced the performance of each lot of Povidone is consis-
to what was deemed acceptable for safe use of Po- tent in the various drugs that use this excipient.
vidone in pharmaceuticals. Because Hydrazine
is not expected to be beneficial in Povidone, it Non-homogeneity
is thus considered an undesirable component. A These examples with one excipient illustrate how
test method was developed, appropriate specified control of all the components in the material are
limits were established, and the compendium was needed in order to assure consistent quality. An-
updated accordingly. other aspect that needs equivalent control is the
In the early 1990s, vinyl pyrrolidone was iden- degree of homogeneity of solid excipients, par-
tified as a suspect carcinogen. As a consequence, ticularly those supplied in powder form. However,
manufacturing methods were again modified many excipients are also manufactured in much
to reduce the level of residual vinyl pyrrolidone larger volumes for other markets where a larger
monomer to a toxicologically safe level. Vinyl pyr- degree of variation is tolerable.
rolidone has solvating properties and is a reactive To illustrate a common cause of non-homoge-
molecule. Therefore, residual vinyl pyrrolidone in neity, consider that excipient manufacture often
6 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
Advancing Drug Delivery
Through Superior Polymer Design
Safe and effective delivery of today’s complex drug molecules is increasingly
a function of collaboration between polymer experts and pharmaceutical
product development teams. Acting on enabling science, producing
superior polymer excipient technology, and creating systems for
optimum bioavailability, Ashland works alongside formulators to
advance drug-specific delivery platforms.

Explore how you can solve intricate drug formulation

and development challenges with polymer
expertise from Ashland.

Visit us at

®Registered trademark, Ashland or its subsidiaries, registered in various countries
™Trademark, Ashland or its subsidiaries, registered in various countries
©2016, Ashland
involves drying the substance. The ability to Common impurities in excipients, which are not
dry material to a consistent residual moisture needed for excipient performance, may include
throughout the lot is inherently difficult due to residual process aids, additives, by-products, and
the many operating variables. Spray drying is a material that sheds from filter media. In addition,
case in point. Operating parameters for the spray contaminants, which are to be avoided, can occur
dryer include the temperature and dew point of from environmental factors such as personnel hy-
inlet air, burner temperature, concentration of giene, equipment failure, contact with packaging,
the excipient in the aqueous solution, spray pat- etc. and include rust, oil, grease, insect fragments,
tern of the excipient solution, rate of drying, and extractable and leachable materials, etc.
outlet air temperature. During a 24-hour cycle, Excipient quality is, therefore, best expressed
as conformance to GMPs
The stability of the excipient can pose a risk as well as to compendia or
if the material is likely to degrade during a specification and consis-
tent composition, lot to lot.
storage or shipment when temperature
Consistent composition
and/or humidity are not controlled within within each lot is also an
acceptable limits. expression of excipient
quality, but oftentimes
the ambient air temperature and humidity may such consistency is difficult to achieve without
differ considerably from day to night. Also the ex- a blending step.
cipient concentration may vary due to prior man- Generally, it is expected that a more consis-
ufacturing steps. Achieving a consistent moisture tent excipient composition will result in a more
level requires frequent sampling of dried material predictable performance in the final drug for-
and adjustment of spray-dryer operating param- mulation. In the selection of an excipient for a
eters. As drying conditions become more severe drug formulation, consideration should be to
in order to maintain constant residual moisture, include an excipient whose composition profile
however, it is possible to cause some degradation has known and tolerable variation with mini-
manifest as charring of the excipient. This is typi- mal number of concomitant components and
cally manifested as burnt particles (4). Consistent impurities.
moisture content in the excipient, therefore, may
be a tradeoff with the quantity of burnt particles Selection of excipient suppliers
in the product. The European Union Directive Guidelines on the
Formalized Risk Assessment for Ascertaining the
Excipient impurities Appropriate Good Manufacturing Practice for Ex-
Excipient impurities are specific entities that cipients of Medicinal Products of Human Use (5)
should not be present and/or need to be con- provides the following characteristics for assess-
trolled for safety, toxicological, or other reasons. ing the manufacture and supply of excipients:
8 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
t Potential presence of transmissible spongiform or molecular weight, therefore, can be considered
encephalopathy (TSE) dedicated. Also equipment used to produce vari-
t Potential for viral contamination ous grades of an excipient that are then sold in dif-
t Potential for microbiological or endotoxin con- ferent markets (e.g., food, cosmetic, or industrial
tamination applications), but produced using the same chem-
t Potential for the presence of impurities istry and raw materials, should also be considered
t Supply chain complexity and security dedicated.
t Excipient stability Using dedicated equipment reduces the risk
t Tamper-evident packaging. that the excipient will be contaminated by the
Each of these characteristics can be related to presence of other substances (e.g., other raw ma-
excipient “quality” and used to assess excipient terial, intermediate, or finished product residue in
suppliers. Note that each of these considerations the production equipment). Using multi-purpose
is in addition to manufacturing the excipient in equipment relies heavily on verifying cleaning ef-
conformance to excipient GMP. fectiveness and the ability to detect potential re-
In addition, the ANSI excipient GMP standard (3) sidual contaminants to assure the minimization
highlights the following criteria to assess for risk to of potential cross-contaminants in the excipient.
protect an excipient from contamination: Where multi-use equipment is used, it is advisable
t Hygienic practices: excipient contamination due to review the excipient manufacturer’s cleaning
to personnel hygiene, illness, attire, unauthor- validation report.
ized access, food, medication, tobacco, etc. When possible, it is preferable to source the ex-
t Infrastructure, building: excipient contamina- cipient from a supplier that does not use animal-
tion, cross-contamination, mix-ups derived raw materials at risk for bovine spongiform
t Infrastructure, equipment: excipient contamina- encephalopathy (BSE)/TSE in the manufacture of
tion due to material of construction, utilities, the excipient. Otherwise, the excipient user will
water, process materials, and work environment have to ensure the excipient presents minimal risk
(air handling, cleaning/sanitation, pest control from TSE contaminants. A risk assessment should
and drainage). include confirmation the animals used in the man-
ufacture of the animal-derived raw material come
Minimizing contamination risk from a country designated as negligible TSE risk.
Using an excipient manufacturer that produces Alternatively, the excipient manufacturer should
the excipient in dedicated equipment is a lower demonstrate that the animal-derived raw material
risk to excipient quality as a result of reduced risk was processed under conditions that have been de-
of cross-contamination. Equipment can be con- fined to inactivate the TSE risk materials if present.
sidered dedicated when it is used to manufacture TSE risks are also present when the excipient is
products utilizing the same chemistry and raw manufactured in multi-purpose equipment where
materials. Equipment used to manufacture an ex- the other products are animal derived. If there is a
cipient in various particle size, density, viscosity, risk of TSE material residue on equipment, the ex-
10 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
Vcaps® Plus Capsules

Not all capsules are

created equal.

Proven Bioequivalence to Gelatin

The new standard in HPMC capsules:
Immediate release in-vivo performance
Molecule = Acetaminophen Human
Mean Plasma Acetaminophen Concentration (ng/mL)
7000 cross-over study in 24 subjects

bioequivalent to gelatin and consistent 6000

dissolution across changing pH and 5000

Vcaps Plus
ionic strength. 3000

Vcaps® Plus capsules can optimize capsule disintegration for pharmaceutical
products with the ability to release contents independent of pH and ionic
strength of the test media. And a new in-vivo study demonstrates that
0 2 4 6 8 10 12
Nominal Hours (hrs) Post Dose
Vcaps Plus capsules are equivalent to gelatin in terms of human
Mean in-vivo acetaminophen plasma
pharmacokinetics profile. concentration profiles over 12 hours.

Finally, a superior specialty polymer capsule with proven gelatin-like

True pH and Ionic Media Independence
performance without cross-linking potential, Vcaps Plus is the new capsule
standard with benefits of narrow weight variability, impressive resistance to 100
heat and humidity, and validated manufacturing experience. By reducing 80

variability, Vcaps Plus capsules are quickly becoming a powerful new tool to
% Caffeine dissolved

60 pH 1.2 USP
reduce timelines in drug product development. 50 pH 6.8 USP
pH 6.8 JP2
Simulated milk fluid
Contact us at 800-845-6973 for our full technical package and 20
pH 1.2 – 2 g KCI/L
pH 1.2 – 9 g KCI/L

samples for your development review. 10

0 3 6 9 12 15 18 21 24 27 30 35 40 45 50 55 60 75
Time in minutes
New in-vivo results comparing Vcaps Plus HPMC In-vitro dissolution of caffeine in
Capsules and gelatin capsules demonstrate Vcaps Plus capsules
bioequivalency of HPMC in multiple dissolution profiles.

Read the full report at

© 2016 Capsugel Belgium NV All rights reserved.

cipient manufacturer should demonstrate cleaning borne particulate to reduce the risk of microbial
procedures that show residual TSE risk material is contamination only requires use of a HEPA filter
either reduced to an acceptable level on the equip- if the excipient is purported to be sterile.
ment surface or is inactivated. Contamination of the excipient with undersir-
The risk of viral, microbiological, or endotoxin able components can arise from such sources as
contamination arises from raw materials, water, nearby manufacturing operations, processing
and the environment. Where the manufacture of equipment (e.g., filters and traps), and utilities.
the excipient uses viral agents or there is a risk of Filters pose a risk from shedding their material of
contamination with viral agents, adequate mea- construction and from traps that are improperly
sures of sanitation or sterilization by the supplier maintained, allowing trapped impurities through.
are expected. Utilities such as nitrogen, compressed air, and
steam may contami-
nate the excipient with
The complexity of the supply chain from
i mpu r it ie s such a s
excipient manufacturer to pharmaceutical compressor oi l a nd
facility is a consideration in selecting a supplier. boiler additives.
It is common for the
Excipient manufacturers should use at least pota- excipient to be produced at a site where many
ble water where water is used in the process after the other products are also manufactured. Some of
starting point for GMP or when water is a potential these other products may be toxic (e.g., herbicides
source of microbial contamination in the finished or pesticides) or they may use toxic ingredients
excipient. Water that is used for temperature control in their manufacture. Where toxic substances are
that does not contact excipient during manufacture volatile enough to become airborne contaminants,
poses minimal risk under normal operating condi- manufacturers should take appropriate measures
tions and therefore need not be potable. For excipi- to minimize the risk of contamination. It is im-
ents that are intended for drug products where the portant for the user to assess the risk of airborne
presence of endotoxin poses a risk to patient safety contamination and the measures taken to protect
and water comes into direct contact with the ex- the excipient during an onsite audit.
cipient during processing, higher purity water such
as United States Pharmacopieal Convention (USP) Supply chain considerations
water for injection may be expected to be used. The complexity of the supply chain from excipi-
There is also the potential for airborne microor- ent manufacturer to pharmaceutical facility is also
ganisms to contaminate the excipient. Generally, a consideration in selecting a supplier. Although
airborne microbes that can contaminate the ex- delivery from an excipient manufacturing site
cipient can be controlled by filtering the air, such directly to the pharmaceutical manufacturing
as when the excipient is exposed to the air during facility provides the least opportunity for the ex-
packaging, to remove particles. Removal of air- cipient to become contaminated or tampered with
12 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
en route; generally direct delivery is uncommon should verify through the paper trail that the ship-
and only applies to full truckloads of the excipi- ment of an excipient lot has come from the excipi-
ent. More often, less-than-truckload quantities are ent manufacturer.
shipped by common carrier. Oftentimes, the ship- The stability of the excipient can pose a risk if
ment goes from the manufacturer to a warehouse the material is likely to degrade during storage or
of the transport carrier. There, the shipment may shipment when temperature and/or humidity are
be cross-docked to a truck heading to the desired not controlled within acceptable limits. Generally,
destination or another intermediate destination. excipients such as inorganic salts, minerals, modi-
While tampering with the excipient at the trans- fied food ingredients, and synthetic substances are
port warehouse is unlikely, there is the possibil- stable materials. Also, many excipients have been
ity for the packaged excipient to be exposed to in commerce for an extended number of years and,
extremes of weather (tem-
perature, humidity, and Where delivery is not direct from the
precipitation), for t he
excipient manufacturer, the pharmaceutical
packaging to be damaged
through mishandling, or
company should periodically establish the
for the tamper-evident seal pedigree of the excipient.
to be accidentally broken.
Excipients are often sold through a distributor. therefore, their stability has been well established
Distributors can sell the excipient in the unopened and characterized. Stability issues occur more
excipient manufacturer’s package or the distributor frequently from exposure to moisture or oxygen
may repackage the excipient into smaller packages. rather than temperature extremes. However, un-
Excipients may also be shipped in bulk to a manu- less studies have shown the excipient to be affected
facturer’s terminal or a distributor where the excip- by extremes of temperature, humidity, or exposure
ient is either stored in bulk tanks or packaged from to oxygen, there is little cause for concern regard-
the tank truck or rail car into discrete containers. ing excipient storage.
Any time the excipient is handled other than in For moisture- and/or oxygen-sensitive excipients,
the original container is an opportunity for the the excipient packaging should be considered and
excipient to become contaminated, adulterated, or assessed when selecting a supplier. The excipient
otherwise compromised. Therefore, the fewer such supplier should provide evidence for the suitability
activities in the supply chain, the lower the risk. of the packaging used to protect the excipient from
Where delivery is not direct from the excipi- moisture and oxygen.
ent manufacturer, the pharmaceutical company Finally, tamper-resistant packaging is an im-
should periodically establish the pedigree of the portant consideration in the selection of a sup-
excipient. As discussed in the IPEC-Americas plier. Though packages can be sealed with tam-
and IPEC-Europe Excipient Pedigree position per-evident closures, the package materials can be
paper of 2008 (6), the pharmaceutical company susceptible to tampering via a puncture. However,
impervious packaging such as steel drums may not been made and the risk is kept to an accept-
be appropriate for some excipients. High-density able maximum.
polyethylene (HDPE) drums may be compatible t No viral risk material is used in excipient man-
with the excipient and are more resistant to prod- ufacture unless an adequate assessment has
uct tampering than fiber drums or bags. Though been made and the risk is kept to an acceptable
bags and supersacks are a challenge to make tam- maximum.
per resistant due to their material of construction, t Adequate measures are taken to control the con-
oftentimes they are the only packaging available. tamination risk from microbes or endotoxin un-
Package openings should be protected with tam- less an adequate assessment has been made and
per-evident seals unique to the excipient manufac- the risk is kept to an acceptable maximum.
turer. Such seals are characterized by their having t The risk of contamination is mitigated through
to be opened to gain access to the excipient, can- the implementation of GMP controls.
not be reapplied if broken or otherwise removed, t Transport of the excipient from the manufacturer
and are unique in that they have the manufac- is directly to the user.
turer’s name, logo, or inherent design characteris- t The excipient is stable under the conditions of
tic. While numbered seals are desirable, they are storage and shipment.
impractical for excipients where the number of t Each excipient package is tamper resistant and, if
containers in a lot often exceeds 100 and can be feasible, closed with a tamper-evident seal.
in excesses of 1000. Using a tamper-evident seal, These considerations will help to ensure the use
however, provides no benefit if incoming inspec- of quality excipients in the manufacture of phar-
tion by the pharmaceutical firm fails to match maceuticals.
the appearance of the seal to an authentic seal (or
photo) provided by the excipient manufacturer. Acknowledgement
The author would like to thank those colleagues
Conclusion at IPEC who took the time to provide comments
Excipient quality is best characterized as confor- to this article.
mance to GMPs and the compendial monograph
or specification with a consistent composition pro- References
1. ICH, Q7, Good Manufacturing Practice Guide for Active Phar-
file lot to lot and within lot.
maceutical Ingredient (ICH, November 2000).
In the selection of an excipient supplier, the fol- 2. IPEC, Excipient Composition Guide (IPEC, 2009).
lowing characteristics minimize the risk to excipi- 3. NSF/IPEC/ANSI Standard for Pharmaceutical Excipients, Good
Manufacturing Practices (GMP) for Pharmaceutical Excipients,
ent quality:
t The excipient is manufactured using dedicated 4. IPEC-Americas, Technically Unavoidable Particle Profile Guide
equipment or mixed-use equipment with suffi- (IPEC, 2013).
5. EUR-Lex, Official Journal of the European Union, 2015/C95/02,
cient cleaning validation.
21.3.2015, C95/10-C95/13. Website:
t No TSE risk material is used in excipient man- 6. IPEC, IPEC-Europe Excipient Pedigree position paper, www.
ufacture unless an adequate assessment has PT

14 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
An obsession with quality definitely works to your advantage. Thatʼs
why Mikart is the Contract Development and Manufacturing Organization
(CDMO) that delivers the attention to detail you need, plus the speed
and responsiveness you want.

A Textbook Case For

(Quality Control) OCD.
Weʼre quick to respond with personalized
solutions that get your projects com-
pleted faster and meet your individual
requirements. All while remaining
fixated on quality.

In addition to Formulation Develop-

ment, Mikart also provides Clinical
Trial Supplies, Regulatory Filing
Support, plus Manufacturing and
Packaging solutions.

Youʼll find Mikart has everything

youʼre looking for in a pharma
partner: innovative technology and
equipment, highly skilled people and
38 years of solid CDMO experience.

To experience just how responsive

we can be, contact us at
1-888-4MIKART or send us an
email to

Ready. Responsive. Right for You.

Mikart, Inc. | 1750 Chattahoochee Avenue | Atlanta GA 30318 | 1-888-4MIKART
Excipient Databases

Unifying Excipient Databases

Properties data critical, says NIPTE’s Hoag
Agnes Shanley

More databases currently nce taken for granted as inactive fillers, excipients play a
provide information on
vital role in any pharmaceutical formulation. The wrong
pharmaceutical excipients,
but are they providing excipient can reduce a drug’s efficacy or interact with
enough information of the APIs or other ingredients, resulting in side effects or worse.
right kind, and can they help Excipients can also be major sources of variability. Improved analy-
users address variability?
tics have shown significant differences in measured physical proper-
ties such as moisture content, particle size, or bulk density, between
the same excipient sourced from different suppliers, and even be-
tween different lots sold by the same vendor. Differences that may
seem small can have an impact on final drug product quality.

Variability means risk

“We have variability from the excipients, the API, the manufactur-
ing, the processing conditions, and all of these … result in variabil-
ity of the product,” Stephen Hoag, a professor at the University of
Maryland at Baltimore’s school of pharmacy, commented before an
FDA panel (1). The panel had asked for recommendations for future
generic-drug research funding through the Generic Drug User Fee
Amendments of 2012 (GDUFA). “In terms of risk analysis, that’s
where harm to the patient can come in,” Hoag said.
As recognition increases of the pivotal role that excipients play
in drug safety and quality, and the increasing quality and supply-
chain risks involved, a growing number of databases are being
developed to provide more information on these materials. One

of the first such projects began in 2009 when FDA funded the
creation of a database by the National Institute for Pharmaceutical
Technology and Education (NIPTE), a consortium of universities
and industry, to catalog physical properties and other information
16 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
Excipient Databases
on excipients. Professor Hoag has been leading introduced. IPEC has continued to provide the
these efforts. agency with feedback on the effort’s progress (4, 5).
The goal of the NIPTE–FDA database, the Ex- A separate excipient data-gathering effort is the
cipients Knowledge Base (2), is to offer detailed STEP (Safety and Toxicity of Excipients for Pediat-
information on the physical material properties rics) database (6), designed to help pharmaceutical
of commercial excipients. This is exactly the type formulation scientists screen excipients for use in
of data that engineers routinely use in other in- children’s medications.
dustries, including chemical, automotive, and
aerospace, when developing new products. Knowledge gaps
Despite the increased focus on gathering informa-
“There’s a real lack of knowledge tion on excipients, significant gaps in knowledge
remain. At the FDA GDUFA hearing, Professor
of how to go from material to
Hoag said, “There’s a real lack of knowledge of how
clinical properties.” to go from material to clinical properties.”
—Stephen Hoag, Most of the industry’s knowledge of these rela-
University of Maryland and NIPTE tionships, he said, is gained in a hit-or-miss fashion,
based on empirical observation. More fundamen-
The database, which is housed on Pharmahub, tal knowledge, and development of models, would
a data sharing community developed at Purdue help with change control, he said, and would allow
University, is designed to be a shared resource for the industry to deal with unexpected factors.
the global industry with community features that, Hoag has suggested that separate databases be
for example, allow users to input their experiences. merged to provide increased functionality and
So far, it has between 500 and 1000 regular users. more information on material properties. He dis-
cussed formulation issues and plans for the NIPTE
FDA’s Inactive Ingredients Database database with Pharmaceutical Technology.
At around the same time that NIPTE started to
work on their database, FDA began to catalog the
inactive ingredients used in products (in final dos- Future plans for the database
age form) that it had approved. A working group PharmTech: What progress has been made with the
was established in 2011, involving the International NIPTE database, and what are the plans for its fu-
Pharmaceutical Excipients Council (IPEC), the Ge- ture, both short and long term?
neric Pharmaceuticals Association (GPhA), and a Hoag: FDA funding has come to an end, so we
cross-disciplinary team from FDA’s Office of Ge- are looking for funding that would help us to add
neric Drugs, to develop FDA’s Inactive Ingredients more data to improve it. So far, we have had dis-
Database (IID) (3). cussions on this topic with both GPhA and IPEC.
In 2015, the database’s IT underpinnings were We are seeing that more companies are starting to
improved, and a more user-friendly interface was use the database. Some major upgrades have been
18 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
made to the underlying Pharma Hub software, in- PharmTech: Are companies concerned, as they
cluding new enhancements to improve security, data were in the early days of the process analytical
visualization, and calculations based on the data. technology and quality by design implementations,
IPEC is interested in our work and some mem- that sharing best practices would amount to giving
ber companies have been using it, and adding data, away competitive advantage?
but in a password-protected way. In March 2016, Hoag: Some suppliers worry about how posting
NIPTE will be exploring the idea of a new center data will affect their competitive positions. In the
for pharmaceutial technology and drug-develop- end, some people recognize the need for this data-
ment research with FDA. base, while others worry about it. We have set it up
as a community tool and a place where people can
Need for greater depth and functionality share individual experiences and best practices.
PharmTech: You’ve noted a ‘proliferation’ of excipi- There is still a lot of ‘reinventing the wheel,’ and
ent databases today. Do they provide sufficient redundant testing and evaluation going on within
information to help industry deal with problems individual pharmaceutical companies. I’ve seen
such as variability and complexity? cases within a company where one lab in the US
Hoag: IPEC is primarily interested in FDA’s IID, does a characterization study, then another of the
and we’ve talked about how the two databases company’s labs in Europe runs the same study
might be merged. without realizing that it had already been done.
PharmTech: How does the philosophy behind The NIPTE database could help prevent wasted
NIPTE’s database differ from that behind the IID? efforts like that.
Hoag: IID is not really a database but more of Our goal is to make people aware of this tool,
a spreadsheet with useful information in it. Our and to fund research and work that would improve
database is focused on properties and on allowing it. At this point, the project needs more support.
users to mine data.
PharmTech: You have repeatedly mentioned the References
1. Transcript, Public Hearing, FDA Generic Drug User Fee
need for better modeling in pharmaceutical de- Amendments of 2012 Regulatory Science Initiative, Request for
velopment and formulation, based on an under- Public Input for Fiscal Year 2015, Generic Drug Research, Part
15, Public Hearing, Friday, June 5, 2015,
standing of material properties. Is its reliance on loads/ForIndustry/UserFees/GenericDrugUserFees/
batch processes the reason why the pharmaceutical 2. The NIPTE-FDA Excipients Knowledge Base, www.pharmahub.
industry has not developed more innovative ap- org/excipientsexplore.
3. The FDA Inactive Ingredients Database, www.accessdata.fda.
proaches to modeling? gov/scripts/cder/iig/index.cfm.
4. R. Iser, “Inactive Ingredient Database-FDA Update,” GPhA Fall
Hoag: Batch focus is not the main reason, but Technical Conference, October 30, 2013,
growing interest in continuous manufacturing loads/AboutFDA/CentersOffices/OfficeofMedicalProduct-
will drive home the need for some of this work. 5. “IPEC-Americas suggests improvements to FDA’s Inactive In-
gredients Database,” Pharmaceutical Technology Sourcing Man-
Currently, the biggest holdup appears to be inertia agement, October 30, 2015,
and investment in the status quo and established suggests-improvements-fda-s-inactive-ingredients-database
6. S. Salunke et al., International Journal of Pharmaceutics, 457, (1)
ways of doing things. (November 20, 2013), pp. 310-322. PT


Non-Gelatin Capsules

Establishing a New
Performance Standard
for HPMC Capsules
Agnes Shanley

A new generation of or nearly a hundred years, two-piece hard shell capsules
cellulose-derived materials made from gelatin have been the preferred medium for en-
addresses the variability in capsulating solid oral-dosage forms. The market, however,
disintegration and product began to recognize the need for alternatives decades ago.
dissolution that were seen in Gelatin, derived from pork and beef byproducts, posed a concern
the first generation of for patients with religious or dietary restrictions, but it could also
gelatin alternative HPMC pose formulation problems when encapsulating hygroscopic and
capsules, while offering moisture-sensitive ingredients.
in-vitro and in-vivo The plant-based material, hydroxypropyl methylcellulose (HPMC),
performance comparable to was developed as an alternative to gelatin for two-piece hard shell
that of gelatin capsules. capsules. The first manufacture of the material in capsule form, how-
ever, required the use of secondary gelling agents, which resulted
in variability in both disintegration and product dissolution. Ap-
proximately eight years ago, a second generation of HPMC-polymer
capsules was introduced. Matt Richardson, PhD, manager, Pharma-
ceutical Business Development at Capsugel, and Michael Morgen,
PhD, senior principal scientist, Bend Research, a division of Capsugel
Dosage Form Solutions, discussed these next-generation materials
with Pharmaceutical Technology.

HPMC vs. conventional materials

PharmTech: What makes HPMC capsules different, in terms

of performance properties, from conventional gelatin

Richardson (Capsugel): HPMC capsules, specifically those made by
thermo-gelation, differ from conventional gelatin capsules in several
ways. Gelatin, derived from animal collagen, is composed of amino
acid chains, while HPMC is cellulose-based. The respective poly-
20 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
mers impart different properties to each capsule. second-generation HPMC capsules offer disinte-
Compared to gelatin, HPMC has lower moisture gration/product dissolution profiles that have been
absorption, more flexibility, higher thermal stabil- shown to provide equivalent in-vivo performance
ity, and better resistance to chemical crosslinking. to gelatin capsules for a range of drug molecules.
The manufacture of second-generation HPMC PharmTech: What are their limitations?
capsules by thermo-gelation is quite different than Morgen (Bend Research): HPMC capsules have
that of first-generation HPMC capsules and gelatin higher oxygen permeability than gelatin capsules,
capsules. Both first-generation HPMC and gelatin so they may not be ideal for oxidation-sensitive
capsules are made by dipping stainless-steel pins into formulations, or those that have an odor.
a hot polymer solution. The polymer is deposited on
the stainless-steel pin and then dried to achieve a tar- Recent in-vivo and dissolution test results
geted water content. In the thermo-gelation process PharmTech: Please discuss the significance of recent
for second-generation HPMC capsules, heated stain- in-vivo and dissolution test results.
less-steel pins are dipped into a solution of HPMC to Morgen (Bend Research): The most recent study
form capsules without use of gelling agents. focused on investigating the clinical performance
PharmTech: What were the specific performance of second-generation HPMC capsules relative to
limitations with first-generation HPMC capsules? that of gelatin capsules. It aimed to demonstrate
Morgen (Bend Research): First-generation HPMC that, in-vivo, the slight disintegration time delay for
capsules avoided gelatin’s potential for crosslink- second-generation HPMC capsules does not lead
ing. They were formulated with gelling systems, to differences in pharmacokinetics for relatively
however, and variability in disintegration and soluble, well-absorbed drugs, but that the perfor-
product dissolution rates was often observed, de- mance would be equivalent to that of gelatin cap-
pending on pH and/or ionic strength of the test sules for these applications. This equivalence was
media. Second-generation HPMC capsules were shown, using three different compounds. Formula-
designed to reduce the performance variability of tion scientists now have more options for optimiz-
first-generation HPMC capsules. ing capsule disintegration and product release.
PharmTech: What other specific advantages does PharmTech: What other innovations are being
the second-generation material offer? investigated for second-generation capsules?
Richardson (Capsugel): HPMC capsules have a lower Morgen: For some applications, the capsules
average water content (6% compared with roughly may provide superior bioavailability, particularly
14% for gelatin capsules), making them an excellent for low solubility drugs, by sustaining supersatu-
choice for low moisture or hygroscopic formula- rated concentrations of dissolved drug in the gas-
tions. They also demonstrate high levels of thermal trointestinal (GI) tract through crystallization in-
stability and are quite flexible, compared with gela- hibition. This is a potential advantage,specifically
tin capsules, which can occasionally become brittle, for high-energy salt and amorphous drug forms,
especially with low-moisture APIs and excipients. as well as for weakly basic drugs that are capable
Perhaps their most important attribute is that of supersaturating in the GI tract. PT
Process Simulation

Modeling and Simulation

Move Downstream
Agnes Shanley

Computer-based tools any industries use computer-based simulation and
promise better quality modeling routinely to troubleshoot and improve their
products, improved process
processes. Pharmaceutical manufacturers have not
control, and increased R&D
efficiencies, but will require yet embraced the technology, at least not for down-
different workflows. stream applications. Some companies may use non-computer based
simulation to optimize conditions, for example, compaction simula-
tors for tablet compression. Generally, though, the industry’s most
advanced computer-based systems are reserved for pharmacokinet-
ics and early research.
For chemists, engineers, and technicians who work in develop-
ment and manufacturing, part of the challenge to using models is
that they can be more difficult to develop for batch processes, which
can involve more variability in materials, process conditions, and
other factors. In addition, the physical and material properties data
required for this work, which are freely available to engineers in
petrochemical processing, for instance, via references such as Perry’s
Chemical Engineering Handbook, are not readily available to profes-
sionals in pharmaceutical manufacturing and development. Most of
the data on materials depend on the specific run, batch, equipment,
and facility.
Increased interest in continuous processing, however, is convinc-
ing more professionals in the industry to study and apply in-silico
technologies and models to gain process knowledge and improve
manufacturing. Modeling could also improve drug development,
Yuri_Arcurs/GettY imAGes

enabling development of higher quality, more robust products, and

increasing R&D efficiency.
Indicative of the trend is a new four-year initiative that was
launched in the United Kingdom in January 2016 to help stream-
line drug development and manufacturing by leveraging better
22 Pharmaceutical Technology Solid doSage drug development and manufacturing 2016 P h a r mTe c h . c o m

black ES752106_PTEBOOK0316_022.pgs 03.16.2016 18:38 ADV
& Brand Protection Solutions
• Covert, Inert and Easy to Apply
• Machine-Readable Microtags
• Custom Codes by Product
Click on the video link below to learn how on-dose




Contact us now:

Process Simulation
computer-based process modeling and simula- facturing. Its software is being used by Big Pharma
tion. The $29-million program, called Advanced companies, but as of yet, only by 2–20 people within
Digital Design of Pharmaceutical Therapeutics each company. “Our biggest challenge,” notes Berm-
(ADDoPT), aims to develop the tools required for ingham, “is integrating modeling tools into heavily
an “efficient, knowledge-based, quality by design- experiment-focused workflows.” It is important that
oriented [pharmaceutical] supply chain” (1). Astra- modeling be used before numerous experiments are
Zeneca, Bristol Myers Squibb, GlaxoSmithKline, performed that, individually, yield little information
and Pfizer are participating in the program, as are content, he says.
the universities of Cambridge, Leeds, and Strath- Academic programs, including the Center for
clyde, all of which are doing advanced research in Structured Organic Particulate Systems (C-SOPS),
materials science and continuous manufacturing. involving Rutgers and Purdue, use the software for
Cambridge’s Crystallographic Data Center and the advancing science and transferring knowledge to
Hartree Center, a high-performance computing fa- industry, Bermingham says. For manufacturing,
cility, are also partners. process modeling software allows users to ana-
Participating solutions providers include Process lyze reasons why a process may not be inherently
Systems Enterprise (PSE), a specialist in process robust. It also permits the cost-effective develop-
modeling and optimization; the process-control ment of process-control systems. This approach
company Perceptive Engineering; and Britest, a can help make R&D more efficient.
nonprofit organization that focuses on using quali- “It’s inefficient to have process characterization
tative methods to improve process understanding and scale-up depend overwhelmingly on experi-
and value. ments, and, in some cases, statistical modeling.
One of ADDoPT’s goals will be to diversify the The days when armies of chemists could run these
modeling techniques available to pharmaceutical experiments each day are over,”Bermingham says.
manufacturers, says Sean Bermingham, principal He points to mechanistic model-based scaleup
consultant and head of PSE’s Life Sciences division, of a spray-drying process as an example. Instead of
and the technical lead of ADDoPT. doing 20 different spray-drying experiments, one
PSE is a spinoff of Imperial College UK and was would be able to start with vapor sorption experi-
set up in 1997 to sell process modeling and optimi- ments, estimate isotherm parameters, then work
zation software for the chemicals and oil and gas on drying kinetics using single-droplet studies,
sectors. In 2010, it set up a pharmaceutical and then move on to spray drying using a handful of
life-sciences division. experiments to characterize fluid and particle flow
Process optimization and modeling is widely at the various scales of interest.
used in petrochemicals, and, with companies such Another factor that is driving interest in simula-
as Aspen Technologies and Scale-up Systems, it has tion is continuous processing. Most of the petro-
moved into some facets of API development and leum industry’s processes are run continuously. As
manufacturing. PSE has focused on expanding use more pharmaceutical manufacturers explore the
of modeling technologies into drug-product manu- potential benefits of 24/7 operation, modeling has
24 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
become more important. Among the pharmaceu- Another computer-based simulation and mod-
tical companies that use mechanistic modeling eling platform for downstream pharmaceutical
tools, Bermingham says, 50–60% of their use is to manufacturing application is F-CAD, developed
support continuous manufacturing efforts. by the Center for Innovation in Computer-Aided
In contrast to mechanistic modeling, statisti- Pharmaceutics (CINCAP). Utilizing artificial in-
cal models, such as JMP’s, create black-box type telligence, and the concept of “cellular automata”
models to identify critical process parameters and (2) to model complex systems, the tool looks like
critical quality attributes, all essential for quality- any CAD platform used to model new car and air-
by-design (QbD) work. Mechanistic modeling cuts plane prototypes (3).
through the noise and multivariate data by using A number of companies are evaluating the
first-principles equations that represent the process system. However, as one of its developers notes,
and can be used not only for QbD work but also to using this type of technology will require that
control and improve the process and to adjust to pharmaceutical manufacturers adopt workflows
unanticipated changes. more often seen in the automotive and aviation
However, using this approach requires a differ- industries (3). Specifically, he notes, samples used
ent mindset and new workflows. It also requires in Phase I and II work will have to be prepared at
new types of data, less process data (i.e., less use conditions used in Phases III and IV (i.e., using a
of APIs), but more materials property data, a point mechanical simulator of high-speed rotary presses).
that thought leaders in the industry continue to This would require engineers and pharmaceutical
emphasize (see Feature, p. 16). High-performance scientists to collaborate more closely. It would also
computing will also be needed to sort through require closer connection between the IT and doc-
massive amounts of material property data. One umentation systems used by different functional
characteristic that would help develop solid dos- groups in development and manufacturing.
age-form manufacturing processes more efficiently
is “compressability.” At this point, that variable can References
1. A. Shanley, “Can Better Modeling Reduce Pharmaceutical De-
vary significantly when the blend of API particles
velopment and Manufacturing Costs?”, March
and excipients is changed. “You may run tests on 1, 2016,
a compactor simulator, but when you change the pharmaceutical-development-and-manufacturing-costs
2. M. Puchkov et al., “3-D Cellular Automata in Computer Aided
blend composition or the particle size distribution Design of Pharmaceuticals,” in Formulation Tools for Pharma-
of the API and/or excipients, the blend’s behavior ceutical Development, J. Aguilar Editor, Woodhead Series in
Biomedicine, 2013, Elsevier, p. 155. Chapter available via
will change. One of the research aims in ADDoPT
Google Books,
is to generate correlations for compressability and QBAJ&pg=PA155&lpg=PA155&dq=CINCAP+switzerland+pha
other material properties,” says Bermingham. AD-
DoPT’s pharma partners will provide real-world hXDnYMKHW6uC8QQ6AEIIjAB#v=onepage&q=CIN
measurements from their labs of different com- CAP%20switzerland%20pharma&f=false
3. H. Leuenberger, European Journal of Pharmaceutical Science,
pressability values for different blends, in order to Vol. 16 (February 2016),
develop a compressibility correlation. pubmed/26876764. PT


Tablet Compression

Optimizing Tablet
Frederick J. Murray

A process optimization ll aspects of pharmaceutical manufacturing face in-
template provides a practical creased pressure to improve production efficiency and
approach for maximizing
uptime, while maintaining the highest standards for
output and product quality
of an existing tablet product quality. For tablet compression equipment,
compression process. many companies seek to increase productivity of the existing
equipment platform prior to considering additional equipment. A
structured optimization template provides a practical approach to
maximizing both output and product quality of an existing tablet
compression process.

Tablet weight control

The key process parameter for any tablet compression operation is
tablet weight control. Assuming uniform distribution of the blend,
the ability to hold precise tablet weight is an absolute requirement
to delivering the prescribed dosage of active material. Tablet weight
control is influenced by a number of factors including press speed,
flow properties of the granulation, filling depth, feeder paddle con-
figuration, and working lengths of the upper and lower punches.
Press speed. The press speed and pitch circle diameter of the tablet
press die table will determine the tangential velocity of the press
tools. The lower press tools pass under the feed frame to fill the dies,
and based on the length of the feeder opening and tangential velocity
of the tools, the feeder dwell time may be determined. It is obvious
that a longer feeder dwell-time will permit more time to fill the dies

and that there may be a critical speed limit where it is impossible to

achieve uniform die filling because the feeder dwell time is too short.
Frederick J. Murray is
Flow properties of the granulation. A product with robust flow proper-
President of KORSCH America
Inc., tel: 508.238.9080, fred. ties will fill the dies uniformly at high speeds. Products with mar-
ginal or poor flow properties need long feeder dwell times and often
26 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
Tablet Compression
require a modified feeder paddle configuration to in the tablet weight. Precision tablet weight control
fill the dies uniformly. thus mandates excellent press-tool tolerances and
Filling depth. All tablet presses use a volumetric a corresponding press-tool maintenance program.
fill to obtain the desired tablet weight. The fill Understanding the issues that impact tablet weight
depth (i.e., position of the lower punch in the die control and can cause tablet weight variability is cru-
to achieve the desired tablet weight) is determined cial to a successful process optimization effort.
by the shape of the tablet and the bulk density of
the material. A small tablet diameter with a deep Tablet hardness
filling depth is obviously more difficult to fill than In addition to tablet weight, thickness and hard-
a larger tablet diameter with a small filling depth. ness are key quality parameters for tablet produc-
The ability to fill dies with deep filling depths can tion, and these attributes can be measured in real
certainly be rate limiting. time as the tablet press is producing tablets. The
Feeder paddle configuration. Most modern tablet tablet hardness will determine the dissolution rate,
presses use variable-speed power feeders with ro- which is crucial to ensure effective drug delivery.
tating feeder paddles that assist the transfer of ma- The tablet hardness is a function of the volume of
terial into the die. Feeder paddles generally come material in the die, and the magnitude and rate of
in a rectangular profile; however, many tablet compression force applied to the tablet. In most
presses are available with alternate feeder paddle cases, increasing the press force applied to the
designs, including round profiles and beveled pro- tablet will increase the corresponding tablet hard-
files. There is no real handbook to define the best ness. Some products, however, have a maximum
feeder paddle for any given product, and empirical hardness threshold, and higher press forces actu-
testing, such as described in this article, is required ally cause the tablet hardness to be reduced. Un-
to identify the optimal paddle configuration. derstanding the relationship between press force
Upper- and lower-punch working lengths. The role of and tablet hardness is a crucial component to any
press-tool working lengths on tablet weight control process optimization study.
is often overlooked. Variability in the lower-punch The rate that press force is applied to the tablet
working length will directly impact the volume of is a function of the tangential velocity of the press
material in each die and can, in itself, cause tablet tools, the diameter of the compression rollers, and
weight variation. Variability in the upper-punch the geometry of the press tool, and it is generally
working length will not directly impact the volu- referred to as compression dwell time. In simple
metric fill; however, because most modern tablet terms, the compression process imparts energy
presses are using press-force control systems as the into the tablet by applying a force over a period of
basis for tablet weight control and because upper time. If the time is reduced (when the press speed
punch working length variability will alter the tab- is increased), then the force must be increased
let thickness (and therefore, the press force), the to impart the same amount of energy. The press
force control system is now reacting to tolerance is- speed is thus another dimension to the press force–
sues in the upper punches and not actual variability tablet hardness relationship.
28 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
Process optimization Figure 1: Tablet weight variation shown as relative standard deviation (Srel%) of individual tablet
template weight as a function of press speed.

With this fundamental 6%

understanding of the 5%
key process parameters
and those factors that
Weight Srel %
can impact quality pa-
rameters, one can use
the process optimization
template, which consists
20 30 40 50 60 70 80 90
of the following four sets
Press speed (RPM)
of empirical testing.
Press speed/tablet qual-
ity. This testing consists Figure 2: Tablet thickness variation shown as relative standard deviation (Srel%) of individual
tablet thickness as a function of press speed.
of making the product to
specification (i.e., target
weight, thickness, and
hardness) and then mea- 4%
Thickness Srel %

suring the standard devi- 3%

ation of individual tablet 2%

weight, thickness, and 1%

hardness values across 0%

20 30 40 50 60 70 80 90
a defined speed range.
Graphs of standard de- Press speed (RPM)

viation of individual tab-

let weight, thickness, and Figure 3: Tablet hardness variation shown as relative standard deviation (Srel%) of individual
hardness variation versus tablet hardness as a function of press speed.

press speed will clearly 40%

define the press speed 35%

range in which quality
Hardness Srel %

tablets can be produced. 20%

Figure 1 plots the rela- 15%

tive standard deviation 10%

(Srel %) of individual tab-
let weight as a function of 20 30 40 50 60 70 80 90

press speed. In this ex- Press speed (RPM)

ample, there is a clear in-
Tablet Compression
crease in the tablet weight Figure 4: Tablet hardness vs. compression force at different press speeds (i.e., compression
variability at press speeds dwell times).

higher than 60 RPM, thus 12

60 RPM is the maximum 10

Hardness (KP)
press speed given the flow 8

properties of the material 6

tested. Better flow prop-
erties would likely permit
higher press speeds. 2 4 6 8 10
30 RPM 45 RPM 60 RPM 75 RPM 90 RPM
Figure 2 plots the rela- Main compression force (kN)
tive standard deviation
(Srel %) of individual tab-
let thickness as a function of press speed and shows (i.e., compression dwell time) and will confirm
consistent control across the press speed range. the press force value that corresponds to the de-
Figure 3 plots the standard deviation (Srel %) of sired tablet hardness.
individual tablet hardness as a function of press The representative graph shown in Figure 4 indi-
speed. These data mirror the tablet weight data, cates that the compression dwell time does impact
with consistent control up to the 60 RPM press the tablet hardness, especially at the higher press
speed level. force levels. At 90 RPM, the tablet hardness is, on
The process capability (Cp) may be calculated average, 80% lower than the same tablet produced
for tablet weight, thickness, and hardness at each at the 30 RPM press speed level. Based on required
press speed using Equation 1. tablet hardness range, the expected press force
range can be easily determined.
Cp = (USL – LSL) / 6 * ∑ [Eq. 1] Feeder speed/feeder paddle optimization. This test
consists of running the press at different press
Where Cp is process capability index, USL is speeds and adjusting the feeder speed across the
upper specification limit, LSL is lower specifica- range, while recording the relative standard de-
tion limit, and Σ is standard deviation. viation of individual tablet weights. This measure-
Nominal process capability values of 1.33 or 1.50, ment is performed at each press speed and with
or higher, are generally indicative of a process that different feeder paddles designs (standard, round,
is under control. beveled) to determine the optimal feeder speed
Tablet hardness/compression force. This testing and feeder-paddle configuration.
consists of running the press at different com- If the results show that the standard deviation of
pression forces and measuring tablet hardness at tablet weight is not impacted significantly by the
the different press-speed levels. The force versus feeder speed, then it makes sense to run the feeder
hardness plot will confirm the ability to achieve speed at the lowest value to avoid overmixing or
the desired tablet hardness at each press speed shearing the granulation in the feeder.
30 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
ic S
ro pe
en ci
ca ali
ps sts
Does your current CDMO
leave you with blurred vision?


in the door then seem to go in multiple directions. There is no clear direction;
and the invoices are piling up.
At Coating Place our experienced staff believes the best results come with a
clear understanding of your ideas and goals. Trust the knowledge, quality and



Tablet Compression
lock in the optimal feeder
Figure 5: Weight variation shown as relative standard deviation (Srel%) vs. feeder speed at a series
of press speeds (30–90 RPM indicated as different colors) using a rectangular feeder paddle. speed at each press speed.
Figure 7 shows the im-
pact of the feeder-paddle
configuration on the rela-
tive standard deviation of
Weight Srel %


individual tablet weight at
1% a variety of different press
0% speeds and feeder speeds.
10 20 30 40 50 60
The results show compa-
30 RPM 45 RPM 60 RPM 75 RPM 90 RPM rable performance of the
Feeder speed (RPM), rectangular paddles two paddles at the 30 RPM
press speed (see Figure 7a).
Figure 6: Weight variation shown as relative standard deviation (Srel%) vs. feeder speed at a
At the 60 RPM press speed,
series of press speeds (30–90 RPM indicated as different colors) using a round feeder paddle. the rectangular paddle
clearly performs better
(see Figure 7b). Results are

mixed at the 90 RPM press
speed (see Figure 7c).
Weight Srel %


2% Fill cam optimization.

1% Most presses offer a vari-
0% ety of fill cams to cover
10 20 30 40 50 60
a range of filling depths.
30 RPM 45 RPM 60 RPM 75 RPM 90 RPM The purpose of the fill
Feeder speed (RPM), round paddles cam is to overfill each
die and then push some
material back into the
The following example shows the impact of feeder feeder to ensure optimal die filling as the lower
speed on tablet weight variation, using a rectangular punch moves through the dosing cam. For exam-
feeder paddle (see Figure 5) and a round feeder paddle ple, a standard EURO or TSM B turret will offer
(see Figure 6), at different press speeds. This infor- a fill depth range of 0–18 mm. This filling depth
mation can be used to establish the optimal feeder is achieved by a range of filling cams, as follows:
speed at each press speed, to be stored in the product
recipe. In this example, it can be concluded that the 6 mm Fill Cam Fill Cam Range 0–6 mm
higher feeder speeds (> 40 RPM) and the rectangular 10 mm Fill Cam Fill Cam Range 0–10 mm
feeder paddle generally produce better results (i.e., 14 mm Fill Cam Fill Cam Range 4–14 mm
less weight variation), and the data can be used to 18 mm Fill Cam Fill Cam Range 8–18 mm.
32 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
Because there is over- Figure 7: Weight variation shown as relative standard deviation (Srel%) of tablet weight vs.
lap in the range of differ- feeder speed and feeder paddle configuration for rectangular paddles (blue lines) and round
paddles (red lines) at (a) 30 RPM, (b) 60 RPM, and (c) 90 RPM press speeds.
ent fill cams, the optimal
(a) 30 RPM press speed
fill cam must be selected
empirically. For a prod- 5%

uct that requires a final

Weight Srel %
filling depth of 8 mm, 2%

the press can operate 1%

with either the 10-mm 10 20 30 40 50 60

fill cam or the 14-mm (b) 60 RPM press speed

f ill cam. Although it
may seem logical that a 4%
Weight Srel %

deeper fill cam will be 3%

better, that is not always
the case. In summary, 0%
10 20 30 40 50 60
products that have fill-
(c) 90 RPM press speed
ing depths in the over-
lap between multiple fill 6%
cams can be optimized
Weight Srel %

only by testing each fill 3%
cam across the desired 1%
speed range.
10 20 30 40 50 60
Figure 8 plots the rela-
Rectangle Paddles Round Paddles
t ive s t a nd a rd de v ia-
Feeder speed (RPM)
tion of tablet weight at
different press speeds
Figure 8: Weight variation shown as relative standard deviation (Srel %) of tablet weight at
using two different fill
different press speeds using two different fill cams (10 mm, blue line and 14 mm, red line) at the
cams (10 mm and 14 same dosing setting and tablet weight.
mm) at the same dos- 6%

i ng s e t t i ng a nd t ab - 5%

let weight. The results
Weight Srel %

show compa rable re-
sults through 60 RPM,
but the deeper fill cam 0%
30 45 60 75 90
(14 mm) clearly extends
10 mm Fill Cam 14 mm Fill Cam
the process control (i.e.,
Press speed (RPM)
Contin. on page 45


Continuous Manufacturing

Filling the Analysis Gap

in the Move to
Continuous Processing
Jamie Clayton

Effective analysis is key for he benefits of continuous manufacturing justify signifi-
the successful continuous cant investment, as evidenced by collaborations such as
manufacturing of solid- the MIT/Novartis Center for Continuous Manufacturing
dosage pharmaceuticals. (1). Batch production dominates within the pharmaceuti-
cal industry but many expect continuous processing to contribute a
substantial share of manufacturing capacity.
The move to continuous manufacturing requires changes in ana-
lytical practices in order to support new manufacturing models. Dy-
namic powder testing can contribute to the development of efficient
continuous processes.

The benefits of continuous manufacturing

Although innovative in many areas, the pharmaceutical industry
has historically focussed less on processing than other research
and development areas. Patent protection previously ensured that
R&D costs could be properly recouped, but as costs rise and time to
market increases, profitability cannot be guaranteed. When patents
expire, profitability relies on efficient production. Furthermore, a
regulatory focus on risk suggests a need for greater understanding of
processes and improved quality control. A shift from batch processes,
which are heavily dependent on manual intervention, to automated,
continuous operation is highly attractive.
In batch production, sequential steps are undertaken, with analysis

performed in between. Batch-to-batch variability and products out

Jamie Clayton is
operations director at of specification (OOS) are common problems. The necessary rework
Freeman Technology.
and waste levels are unacceptably high.
34 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
THINI Download
Optimize your capability
Taking your processing capabilities to the next level is a strategy for
success; one that often demands upgrading equipment to meet new
production goals and future requirements. For immediate access to
the best technologies at the right price while maximizing returns from
surplus idle equipment, Federal Equipment Company is ready to help
you optimize operations while recovering equipment value through
accurate appraisals, strategic liquidations and full logistical support.

When you think equipment,

think Federal Equipment

1.877.536.1538 >

Continuous Manufacturing
Continuous processing is widely used in chemi- of a problem. With continuous manufacture, there
cal and food industries and offers the following is a question of how to define a ‘batch’. A batch be-
important advantages: comes associated with an operating period, which
tReduce costs, labor, and waste begins when start-up completes and ends at a de-
tOptimize asset utilization fined point. Any OOS products are therefore as-
t Simplify scale-up sociated with a time period rather than a discrete
tImprove containment. batch number, which can make problems difficult
These are environmental and economic im- to isolate.
provements, but there are also technical benefits. Optimized processing relies on understanding
A batch step has a beginning and an end; be- and controlling the materials and process variables
tween these points the product continuously that define clinical efficacy. The pharmaceutical
changes (e.g., a blending process starts with the industry has traditionally focused on developing
unmixed constituents and proceeds to a homo- and adhering to repeatable processes. This ap-
geneous state). A well-controlled continuous pro- proach relies on consistent feed and provides little
cess should operate at steady state for the major- flexibility to respond to variation. This is a critical
ity of the time. A steady-state operation requires limitation, as feed variability is a major source of
effective monitoring, as exemplified by widely failure. The transition from batch processing to-
used techniques such as in-line particle size and ward knowledge-based continuous manufacture,
near-infrared (NIR) analysis. Steady-state opera- however, has its challenges.
tion means continuous processing is associated
with consistent output, which equates to consis- A toolkit for more efficient manufacture
tent product quality. Scale-up is also simplified as Operations such as milling, roller compaction,
smaller units can be run for longer, avoiding the and tableting can be considered semi-continuous,
implications of changes in geometry and volume. as they are constantly fed during a batch cam-
Batch production does, however, have benefits. paign. The challenge involves engineering the
One benefit is flexibility; a suite of batch equip- equipment for reliable, prolonged operation and
ment can easily be reconfigured for different prod- successfully integrating the necessary compo-
ucts. Batch production also simplifies containment nents into an optimized continuous process. Au-

Center for Structured Organic Particulate Systems (C-SOPS)

For 10 years, researchers at the Center for Structured Organic Particulate Systems Test Bed 1 focuses on the simultaneous development of formulations, continuous
(C-SOPS) have worked to transform pharmaceutical manufacturing into a science- manufacturing, and analytical control methodologies for solid oral products. Test
driven discipline, in the areas of materials formulation and characterization, Bed 2 is designed to create an integrated continuous manufacturing platform to
design and scale up of material structuring, structural characterization and produce film-based drug products with controlled-release properties. Test Bed 3
modeling, and integrated systems science. C-SOPS also operates three test beds is based on drop-on-demand manufacturing, and uses liquid-phase processing to
to develop new continuous manufacturing processes. avoid challenges normally associated with conventional powder-based processes.
A National Science Foundation Engineering Research Center, C-SOPS is a Information about consortium members, research projects, and test bed
consortium of academic institutions and more than 40 industry partners from programs can be found at
both bio/pharmaceutical companies and industry equipment suppliers. —The Editors of Pharmaceutical Technology

36 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
tomation is important but so are analytical tools Figure 1: Measuring flow energy to quantifying changes in flow
that provide the knowledge required to optimize properties following consolidation.

multi-component systems. The Engineering Re-

search Center for Structured Organic Particulate
Systems (C-SOPS) is a group at the forefront of


research in this area. The group applies modeling,
Flowability Change
in-line analysis, and techniques such as powder
rheology to integrate sequential blending, dry
Density Change
granulation, lubrication, and tableting. A key ≈40% (max)

focus is to develop solutions to avoid three com- INCREASED TAPPING

mon tableting issues: segregation, agglomeration,

and compaction (2). of powders; however, the need for accurate, process-
Whether improving batch processes, or design- relevant data exposes limitations and highlights the
ing, monitoring, and controlling a continuous merits of techniques such as dynamic testing.
process, analytical tools are needed that deliver In dynamic testing, axial and rotational forces
relevant data and expand understanding of how acting on a blade are measured as it rotates through
processes work, reinforcing FDA’s process analyti- a powder sample to determine values of flow en-
cal technology (PAT) initiative. PAT is defined as ergy that quantify how a powder f lows under
“a system for designing, analyzing, and controlling conditions that reflect processing environments.
manufacturing through timely measurement of Powders can be characterized in consolidated, con-
critical quality and performance attributes of raw ditioned, aerated, or even fluidized states to mea-
and in-process materials and processes, with the sure the response to stress and air content. The
goal of ensuring final product quality” (3). Real- impact of moisture, flow additives, compaction,
time analysis is therefore important but so are tech- attrition, and segregation can be evaluated.
niques that, for example, provide robust analysis of Figure 1 contrasts the change in bulk density in-
feeds prior to introduction to the plant. Identifying duced by tapping with the corresponding change
techniques that provide the information required to in flow energy. Flow energy increases by an order
achieve process efficiency is fundamental. of magnitude greater than density suggesting that
flow energy measurements are significantly more
Focus on powder testing sensitive in quantifying the impact of the change.

Tablets are the most common drug-delivery vehicle, Furthermore, this indicates how density changes
and most drugs are handled in solid form at some could be misleading when quantifying how con-
point, demonstrating the need for suitable powder solidation impacts a process.
testing tools. Numerous methods for characteriz- This experiment emphasizes the importance
ing powders exist, including angle of repose, flow of selecting a suitable analytical technique for a
through an orifice, and tapped density. These sim- given application. It is increasingly acknowledged
ple techniques provide some insight into the nature that no single powder test suits every application
Continuous Manufacturing
Figure 2: Dynamic testing can provide differentiation of materials that shear tests classify as identical.

2 vanillin 1800
2 ethylvanillin
5 1600


Total energy, mJ
Shear stress. kPa




1 2 vanillin
200 2 ethylvanillin

Test number 0 2 4 6 8 10
0 1 2 3 4 5 6 7 8 9
Tip speed, mm/s -100 -100 -100 -100 -100 -100 -100 -100 -70 -40 -10
Applied normal stress, kPa

and that tests should represent the conditions to process design and enable effective monitoring
which the material is exposed. and control. It is essential to consider what in-
Shear testing, for example, is a well-estab- formation is required and how to obtain it. Ap-
lished technique developed to support hopper plying this approach to powder characterization
design protocols advanced by Jenike (4, 5). It is highlights limitations with traditional techniques.
still widely applied, with modern instrumenta- Innovative techniques such as dynamic testing
tion delivering improved reproducibility, and is and, in particular, instruments that combine dy-
a valuable tool for characterizing a powder in a namic testing with methods such as shear and
static, consolidated state, but there are limitations. bulk property analysis, present an efficient and
Figure 2 shows shear and flow energy data for versatile choice for those demanding new levels
two excipients: vanillin and ethylvanillin. Shear of efficiency.
testing suggests these materials are identical,
while dynamic testing identifies clear differ- References
1. Novartis-MIT Center for Continuous Manufacturing website,
ences. In this case, the f low energy correlated, accessed March 1, 2016.
with in-process behavior, illustrating how mate- 2. Center for Structured Organic Particulate Systems, Test Bed 1,
Continuous Powder Manufacturing,
rials classified as identical by shear testing may research/test-beds/test-bed-1, accessed March 1, 2016.
3. FDA, Progress Report on Process Analytical Technology,
actually process differently. This highlights the
importance of employing methods that simulate turing/questionsandanswersoncurrentgoodmanufacturing-
practicescgmpfordrugs/ucm072006.htm, accessed March 1,
process conditions. 2016.
4. A.W. Roberts, Basic Principles of Bulk Solids Storage, Flow and
Handling (The Institute for Bulk Materials Handling Re-
Looking ahead search, Callaghan, NSW, Australia, 1993).
5. The Institution of Chemical Engineers/European Federation
The rise of continuous manufacture, and the of Chemical Engineering, Standard Shear Testing Technique
drive for greater efficiency, increases the need for Particulate Solids Using the Jenike Shear Cell, A Report of
the EFCE Working Party on the Mechanics of Particulate Sol-
for analytical techniques that support intelligent ids (1989). PT

38 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
Case Study

“I need a CDMO
that’s a true PARTNER.”
With Rottendorf Pharma as your CDMO, you can be certain you are working with
a financially stable company committed to you in a long-term relationship. We
will know your goals and bring proven ideas to the table on how to achieve them.
Need a true partner for your solid oral doses? Rottendorf owns it.

M A N U FA C T U R I N G • PA C K A G I N G • F O R M U L AT I O N D E V E L O P M E N T • A N A LY T I C A L M E T H O D S
Elemental Impurities

Meeting USP Guidelines

for Elemental Impurity
Analysis with X-ray
Fluorescence Spectrometry
Andrew Fussell

This article discusses the here are strict limits and clear regulatory guidelines for the
USP guidelines for monitoring of elemental impurities in drug products. An
monitoring elemental
important source of these impurities is the catalysts, which
impurities in drug products.
are necessary during drug production. These catalysts in-
clude palladium, platinum, rhodium, iridium, and ruthenium—all
elements that, at high concentrations, can be harmful to patients.
Previously, the United States Pharmacopeia (USP) chapter <231> (1)
set the standard for the control of elemental impurities; however,
the qualitative approach lacked selectivity and sensitivity and could
fail to detect certain elements, such as mercury, at toxicologically
relevant levels.
For some time, it has been widely recognized within the phar-
maceutical industry that the methodology set out in USP <231> is
outdated, compared to more modern quantitative techniques. In
response, the US Pharmacopeial Convention (USP) has developed
new regulations, chapters <232> and <233> (2, 3), through a pro-
cess of consultation with the industry, chemists, and toxicologists.
USP <232> and <233> contain changes to the concentration limits
for elemental impurities, as well as introducing flexibility in the

choice of testing methods. In an effort to harmonize the regula-

tions, the International Council for Harmonization (ICH) worked
together with representatives from the European Pharmacopoeia
Dr. Andrew Fussell is the
pharmaceutical segment (Ph. Eur.), the Japanese Pharmacopeia, and USP to create the ICH
manager at PANalytical.
Q3D guideline for elemental impurities (4). ICH Q3D has been ad-
40 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
opted verbatim by the European Medicines Agency Table I: Twenty elements spiked into pure cellulose, lactose, or
(EMA). The ICH Q3D guideline does not stipu- calcium carbonate matrix. International Council for Harmonization
(ICH) oral delivery risk assessment elements shown in bold.
late analytical procedures but instead states that Concentration PAN Concentration
PAN standards
range (μg/g) standards range (μg/g)
pharmacopeial procedures or suitable alternative
As 0-100 Se 0-200
procedures should be used. These regulations (USP
Cd 0-100 TI 0-200
and ICH Q3D) will apply both to products newly Pb 0-100 Cu 0-1000
released onto the market and legacy products, which Hg 0-100 Zn 0-1000
will need to be retested. Co 0-200 Mo 0-200

In a significant move, USP <233> allows manu- V 0-200 Ru 0-200

facturers choice in the analytical method used for Ni 0-1000 Rh 0-200

the evaluation of levels of elemental impurities so Cr 0-500 Pd 0-200

Mn 0-1000 Ir 0-200
long as the technique has been validated in line
Fe 0-1000 Pt 0-200
with the requirements. One option emerging as
a favored alternative is X-ray fluorescence spec- ing. It can be used for the elemental and chemical
trometry (XRF). analysis of solid, powdered, and liquid samples,
making it particularly valuable to pharmaceuti-
X-ray analysis in pharmaceuticals cals. One of the main advantages of XRF is that it
Reliable, accurate analytical processes are the basis requires minimal if any sample preparation prior
for most activities in the pharmaceutical industry to analysis. In addition, the instrumentation is well
and X-ray-based analytical techniques underpin adapted to automation.
many of these procedures. X-ray diffraction (XRD) This is in sharp contrast to inductively coupled
is an established technique in the industry and is plasma atomic emission spectroscopy (ICP–AES)
widely used for qualitative and quantitative analy- and inductively coupled plasma mass spectrometry
sis of solid phases (5). XRD has long been the ac- (ICP–MS), which feature in the two sample meth-
cepted technique for establishing the crystalline ods outlined in USP <233>. ICP–AES uses ICP to
drug “fingerprint” needed for drug approval, pat- excite the atoms in a sample to emit electromag-
ent descriptions, and for the identification of dif- netic radiation—the wavelength indicates the pres-
ferent drug batches (6). The use of XRD has high- ence of an element while the intensity indicates the
lighted the general benefits of X-ray analysis to the concentration. In ICP–MS, ICP is used to ionize
pharmaceutical industry, including the following: a sample, while mass spectrometry separates and
t Speed of analysis quantifies the elements present. Both of these tech-
t Simple (or no) sample preparation niques are already used in parts of the industry;
tNon-destructive measurement. however, the sample preparation is intensive, and
In parallel, XRF is considered a proven tech- the dilutions required can lead to errors in analysis.
nique for material analysis in a broad range of in- XRF analysis can be divided into wavelength
dustries and applications from measuring sulfur in dispersive (WD) and energy dispersive (ED) tech-
oil to analyzing coating thickness in metal finish- niques. The difference between these two tech-
Elemental Impurities

Table II: International Council for Harmonization (ICH) and United States Pharmacopeia (USP) limits and validation sample set up. PDE is
permissible daily exposure.
USP conc. limit Calc. USP max.
ICH threshold Lower spike conc. Upper spike conc.
Element ICH PDE (μg/g) for oral drug daily dose (g/
(30%) (μg/g) (μg/g)
products (μg/g) day)
As 15 4.5 3 5 1 9

Cd 5 1.5 2 2.5 1 9

Pb 5 1.5 2 2.5 1 9

Co 50 15 10 5 5 20

V 100 30 10 10 5 20

Ni 200 60 20 10 5 60

Cu 3000 900 300 10 15 150

Cr 11000 3300 1100 10 5 75

Mo 3000 900 300 10 5 120

Ru 100 30 10 10 5 15

Rh 100 30 10 10 5 15

Pd 100 30 10 10 5 15

Ir 100 30 10 10 5 15

Pt 100 30 10 10 5 15

inner-shell electrons of the elements in the sample.

Table III: Limits of detection—Epsilon 3X; PDE is permissible
daily exposure; LLD is lower limit of detection. The interaction of the high energy X-rays with the
ICH PDE ICH threshold LLD (μg/g) Measurement
(μg/g) (30%) (3m of blank) time (min) electrons results in the emission of X-rays of char-
As 15 4.5 0.1 acteristic energies for the elements in the sample.
Pb 5 1.5 0.1 The WD technique uses an analyzing crystal to
Cd 5 1.5 0.4 30 separate the characteristic X-rays for the different
Co 50 15 0.1
30 elements, which are then detected by an X-ray de-
Cr 11000 3300 0.2
tector. The ED techniques do not require an ana-
Ni 200 60 0.1
lyzing crystal, but instead, the X-ray detector itself
Cu 3000 900 0.2 15

Mo 3000 900 0.4

determines both the energy and the intensity of
V 100 30 0.2 15 the characteristic X-rays (7).
Ru 100 30 0.4 In this study, EDXRF systems were used. EDXRF
Rh 100 30 0.9 30 spectrometers discriminate each specific X-radia-
Pd 100 30 0.9 tion line based on the energy of the produced pho-
Ir 100 30 0.3
5 ton. Compared to other techniques, EDXRF spec-
Pt 100 30 0.3
trometers offer a number of advantages in that they
niques is due to the different detection methods. tend to be smaller, simpler in design, faster, have
Both techniques typically use an X-ray tube to fewer engineered parts, and are typically cheaper.
generate high energy X-rays that interact with the In addition, the advances in XRF technology in
42 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
recent years make this method particularly attrac- Table IV: Limits of detection—Epsilon 5; PDE is permissible daily
tive to the pharmaceutical industry, particularly exposure; LLD is lower limit of detection
ICH PDE ICH threshold LLD (μg/g) Measurement
as it is now a pharmacopeial method with USP Element
(μg/g) (30%) (3m of blank) time (min)

<735> X-ray fluorescence spectrometry and Ph. As 15 4.5 0.8 10

Eur. chapter 2.2.37. Cd 5 1.5 0.3 10

Pb 5 1.5 0.3 10

Co 50 15 1.7
In practice
Ni 200 60 0.4 5
It is recognized that the greatest risks in drug sub-
Cu 3000 900 7.2
stances is from intentionally added metals, namely V 100 30 0.4
metal catalysts. To combat this risk, the industry Cr 11000 3300 0.3
has adopted a risk-based approach to assessing the Mo 3000 900 0.2 5

potential presence of elemental impurities in drug Ru 100 30 0.2

products. During the manufacture of drugs, risk Rh 100 30 0.1
Pd 100 30 0.4 5
assessments have to be conducted that are in ac-
Ir 100 30 0.3
cordance with USP <232> or ICH Q3D, which lists 5
Pt 100 30 0.4
the permissible daily exposure (PDE) for potential
impurity elements in pharmaceutical materials. Table V: Calibration results for the Epsilon 5. RMS is root mean squared.
To demonstrate how XRF fulfills the guideline Element
Conc. range
R value RMS error (μg/g)
criteria, pharmaceutical calibration standards for As 0-100 0.9998 0.6
elemental impurity analysis were developed by Cd 0-100 0.9999 0.3
spiking 20 elements into pure cellulose, lactose, Pb 0-100 0.9989 1.6

or calcium carbonate matrix, as shown in Table I. Co 0-200 0.9986 3.7

The calibration samples were prepared in triplicate V 0-200 0.9999 1.0

Ni 0-1000 0.9998 6.7

using 1 g of loose powder, using a 6 μm polypro-
Cu 0-1000 0.9998 7.0
pylene foil. The results presented in this work are
Cr 0-500 0.9999 2.9
for the cellulose matrix.
Mn 0-1000 0.9996 10.8
The validation samples were also prepared in Mo 0-200 0.9994 2.5
triplicate by mixing pure paracetamol (acetamino- Ru 0-200 1.0000 0.6
phen) with known concentrations of test elements Rh 0-200 1.0000 0.5

from the calibration materials. Each test element Pd 0-200 0.9997 1.7

was validated at two different concentrations as Ir 0-200 0.9999 1.1

Pt 0-200 0.9998 1.5

required by USP <233>. In most cases, the vali-
dation samples were above and below the target
concentration. The powders were mixed for one was prepared and filled with 1 g of sample. Next,
minute in plastic vials with ZrO2 milling balls. The the sample was inserted into the instrument for
total sample preparation took less than five min- measurement. Table II shows the sample set up used.
utes. For analysis, a loose powder sample holder The first system used was the Epsilon 3X, which
Elemental Impurities
Table VI: Validation spike results for Epsilon 5. USP is United States Pharmacopeia.
Expected spike conc. Measurement conc. Within USP<233> Measurement time
Element Percentage recovery
(μg/g) (μg/g) limit (70-150%) (min)
As 1 1.19 ± 0.03 119 YES 5

9 8.87 ± 0.22 99 YES 10

Cd 1 0.77 ± 0.14 77 YES 5

9 8.30 ± 0.12 92 YES 10

Pb 3 2.78 ± 0.21 93 YES 5

9 10.33 ± 0.15 115 YES 10

Co 5 6.90 ± 1.32 138 YES 5

20 16.29 ± 0.01 81 YES 3

V 5 6.52 ± 0.13 130 YES 5

20 23.03 ± 0.31 115 YES 2.5

Ni 5 5.42 ± 1.29 108 YES 5

60 57.23 ± 2.34 95 YES 3

Cu 15 16.30 ± 1.58 109 YES 5

150 144.36 ± 6.87 96 YES 3

Pd 5 4.98 ± 0.11 100 YES 3

15 14.48 ± 0.04 97 YES 3

Figure 1: Calibration lines for As (left), Cd (middle), and Pd (right) show good linearity.




0.05 0.02


0.00 0.00 0.00

0 50 100 0 50 100 0 50 100 150 200
Concentration (μg/g) Concentration (μg/g) Concentration (μg/g)

is a benchtop EDXRF spectrometer with a 10-posi- The samples were also analyzed with the Epsi-
tion sample changer. The analysis was performed in lon 5, a floor standing EDXRF spectrometer with
air atmosphere not requiring any external gasses. A a 133-position sample changer. For this instrument,
total one off calibration time was 15 hours for all 20 the measurements were performed in helium at-
elements. The measurement time used to obtain the mosphere. The total one off calibration time was
detection limits (Table III) was between 5–30 minutes significantly reduced when compared with the Ep-
per element, while the validation spike sample mea- silon 3X at 6 hours for all 20 elements. The limit
surement was between 2.5–30 minutes per element. of detection measurement time and the validation
Measurement time can be shortened depending on spiked sample measurement time were also re-
the number of elements, sample size, and accuracy duced to 5–10 minutes and 2.5–10 minutes per ele-
and precision requirements needed. ment, respectively. Again, this time can be reduced
44 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
depending on the number of elements, sample size, as reduced potential for error through minimized
and the accuracy and precision requirements. The sample preparation, the non-destructive nature
limits of detection for the Epsilon 3X and Epsilon of the analysis, reduced chemical waste, and the
5 are shown in Tables III and IV. lower total cost of ownership and operation. Many
Good linearity was achieved for all the elements companies are picking up the opportunity to use
tested (as noted in Table V), Examples of the cali- alternative techniques and taking advantage of the
bration curves for As, Cd, and Pd are shown in methods now available. It is important for these
Figure 1. Repeatability was ascertained by analyz- techniques to be fully explored, as it will allow
ing the standards measured 20 times consecutively the pharmaceutical industry to find the technique
as unknowns. Further, the validation spike results and methods that best suit their requirements.
shown in Table VI show a reasonable recovery per-
centage for a number of elements mixed as pow- References
1. USP, Chapter <231>, USP 35–NF 30, (USP, 2012).
ders with paracetamol. 2. USP, Chapter <232>, USP 35–NF 30, (USP, 2012).
3. USP, Chapter <233>, USP 35–NF 30, (USP, 2012).
4. ICH, Q3D Guideline for Elemental Impurities (ICH, 2014),
Guidelines/Quality/Q3D/Q3D_Step_4.pdf, accessed Feb. 22
By allowing the use of alternative techniques, such 2016.
as XRF, USP is now supporting pharmaceutical 5. Vyas K., X-ray Diffraction in Pharma Industry (2013), www., accessed Feb. 22, 2016.
manufacturers who want to make their own ana- 6. Randall C., Rocco W., Ricou P. , XRD in Pharmaceutical Analy-
sis: A Versatile Tool for Problem-Solving (2010), www.american-
lytical choices. EDXRF was shown to be a pow-
erful technique capable of reaching the required Pharmaceutical-Analysis-A-Versatile-Tool-for-Problem-Solv-
ing/, accessed Feb. 22, 2016.
USP <232> and ICH Q3D PDE limits for oral 7. Brouwer P., Theory of XRF–Getting acquainted with the prin-
drugs. XRF also provides a number of advantages ciples (booklet) (2013),
over the USP <233> wet chemical techniques, such principles-booklet.htm, accessed Feb. 22, 2016). PT

Tablet Compression — contin. from page 33

maintains low weight variation) through the 75 necessary to expand the process optimization tem-
RPM press speed. plate to incorporate representative granulations for
nominal and more difficult batches.
Conclusion Most modern tablet presses incorporate a prod-
Applying a methodical approach to tablet-com- uct recipe capability that will permit optimized
pression process optimization and evaluating the press settings to be stored and automatically re-
process control under a range of different condi- trieved to eliminate redundant optimization efforts
tions will permit a full understanding of the pro- and expensive losses at start-up. In many cases,
cess and will facilitate the opportunity to achieve there are opportunities to implement process opti-
optimal tablet quality and production output. This mization measures on existing products and obtain
template assumes limited batch-to-batch variabil- benefits from the resulting improvements in tablet
ity. If batch-to-batch variability is high, it may be quality and production efficiencies. PT
Generic Drugs

Regulatory Considerations
for Controlling Intermediates
in Type-II Drug Master Files
for the Manufacture of
Generic Drug Substances
Kandasamy Subburaj, Brian T. Connell, Srinivasa Murthy, Humcha
Hariprakasha, Deborah F. Johnson, Huyi Zhang, and David J. Skanchy

The authors have he Generic Drug User Fee Amendments of 2012 (Public
undertaken an analysis of Law 112–144, Title III), commonly referred to as GDUFA,
the current state of control was signed into law on July 9, 2012 and is designed to
of intermediate identity
and quality, based on speed the delivery of safe and effective generic drugs to
information submitted in a the public. Since the law came into effect, Type II drug master file
pseudorandom sample of 120 (DMF) and abbreviated new drug application (ANDA) submissions
drug master files that have have reached unprecedented levels.
been submitted to FDA.
Quality by design (QbD) is an essential part of the modern ap-
proach to pharmaceutical quality (1). With the introduction and
implementation of the International Council on Harmonization
David J. Skanchy is a senior (ICH) guidelines (2–5) and FDA initiatives (6) in the past few years,
supervisory regulatory review
pharmaceutical manufacturers have gradually and continuously ad-
officer and Kandasamy
Subburaj, Brian T. Connell, opted QbD principles and elements in the field. The ICH Q11 guide-
Srinivasa Murthy, Humcha
line further clarifies the principles and concepts described in ICH
Hariprakasha, Deborah F.
Johnson, and Huyi Zhang, Q8 Guidelines on Pharmaceutical Development, ICH Q9 Quality Risk
are chemists, all at the
Management, and ICH Q10 Pharmaceutical Quality System as they

Office of New Drug Products

under the US Food and Drug pertain to the development and manufacture of drug substances.
Administration’s Office of
ICH Q11 also describes approaches to developing and understand-
Pharmaceutical Quality within
the Center for Drug Evaluation ing the manufacturing process of the drug substance. That guidance
and Research.
recognizes that critical quality attributes (CQAs) of a drug substance
This article represents the views of can be “(1) included on the specification and confirmed through test-
the authors and not of FDA.
ing the final drug substance, or (2) included on the specification and
46 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
confirmed through upstream controls, or (3) not undertaken an analysis of the current state of con-
included on the specification but ensured through trol of intermediate identity and quality, based on
upstream controls.” One of the upstream controls information submitted in a pseudorandom sample
is isolation of intermediate(s), where applicable, of 120 DMFs that have been submitted to FDA.
with appropriate characterization and controls in While the results of this analysis are not intended
place for impurities, which include starting ma- to be extrapolated to all other Type II DMFs cur-
terials, previous intermediates, reagents, solvents, rently available for reference, they do provide some
catalysts, and reaction by-products. ICH Q7 (7) insight into the current state of intermediate con-
defines an intermediate as “a material produced trol. The results of this analysis are presented in
during steps of the manufacturing of an API that the appropriate section throughout the remainder
undergoes further molecular change or purifica- of this article.
tion before it becomes an API. Intermediates may
or may not be isolated.” This article, however, only Identification and
addresses the intermediates that can be isolated, so characterization of intermediates
that intermediate specifications can be established Regulation 21 Code of Federal Regulations (CFR)
and documented to ensure that quality attributes 314.50(d)(1)(i) describes the regulatory require-
are met. ments of drug substance information that should
The isolation of intermediates, with correspond- be provided in an application and one of the key
ing specifications, provides significant opportu- components is the proof of identity (The regulation
nities to purge potential impurities that may be requires “a full description of the drug substance
carried from the starting material(s) or earlier including its physical and chemical characteristics
intermediate(s). This purging is especially impor- and stability; the name and address of its manu-
tant for regulatory starting materials that may be facturer; the method of synthesis (or isolation) and
outsourced from multiple suppliers. Isolation and purification of the drug substance; the process con-
control of intermediates are ways to help mitigate trols used during manufacture and packaging; and
the risk to drug substance quality when new start- the specifications necessary to ensure the identity,
ing material suppliers have been changed/added. strength, quality, and purity of the drug substance
Additionally, intermediate controls can serve as and the bioavailability of the drug products made
the control point(s) for some CQA(s) of the drug from the substance, including, for example, tests,
substance. Because intermediates are considered analytical procedures, and acceptance criteria re-
to be critical materials in API production, the lating to stability, sterility, particle size, and crys-
specifications should be designed to control the talline form”) (8). The ICH common technical doc-
critical material attributes (CMAs) that may have ument (CTD) format dedicates section 3.2.S.3.1 to
an impact on the CQAs of a drug substance. The Elucidation of Structure and other Characteristics
CMAs of an intermediate are its chemical identity of the drug substance (9, 10). Comprehensive struc-
and quality, which equate to its characterization tural elucidation information is typically provided,
and level of impurities present. The authors have including but not limited to infrared ultraviolet/
Generic Drugs
Figure 1: Routine assay and regioisomer control. niques should be used to
distinguish these isomers,
No accompanied by a precise
13% Intermediate
API interpretation of this sup-
Yes 44%
87% Unknown porting information.
For example, to estab-
lish the E/Z configura-
Routine Intermediate tion of an alkene that is
Assay/Purity Test Regioisomer Control Point
introduced by a Wittig
reaction in an intermedi-
visible spectroscopy, mass spectrometry, elemental ate stage, both 1D and 2D NMR spectra can be re-
analysis, one- and two-dimensional 1H and 13C nu- corded and interpreted. Proton coupling constants,
clear magnetic resonance spectroscopy, and X-ray nuclear Overhauser effect (NOE) enhancements,
crystallography in some cases. For many synthetic and 1H-13C long-range multiplet bond correlation
small-molecular drug substances, the synthetic (HMBC) are methods that may provide solid evi-
route, including the choice of starting materials dence of the configuration of a particular alkene
and reagents, and the process design and controls, isomer. Only 87% of the intermediates examined
also serves as supplementary confirmation of the included one assay or purity test (Figure 1). In cases
drug substance structure. Therefore, the character- where regioisomeric impurities were possible, anal-
ization of intermediates is often considered to be a ysis revealed that the control point was not explic-
verification of structure and can be done by an IR itly discussed in 39% of the sampled intermediates,
spectra and a chromatographic comparison to an while 17% had a control for these impurities in the
established reference standard. Although it would intermediate and the regioisomeric impurity, or
be rare that the manufacturer has synthesized a downstream regioisomeric API analogs, was con-
completely different molecule, or even the wrong trolled in the final APIs for the remaining 44% of
enantiomer or diastereomer, there are cases where the intermediates studied.
the identity of the wrong compound was taken for It is important for the applicant to sufficiently
granted, such as the one described recently (11). address the characterization issue of the interme-
In fact, 96% of the intermediates examined in the diate especially when certain structural features

authors’ study included at least one identity test. would not be unambiguously determined by sim-
However, often times an enantiomer, diastereo- ple techniques. It is equally important for the appli-
mers, or regioisomers are possible at a given step, cant to provide the interpretation of the supporting
or a semi-synthetic process isolates/extracts a com- information, describing how the identity and/or
plex structure as an intermediate. In such cases the challenging structural features are assured by
where the manufacturer relies on the identity of the data provided. Data can also be collected on
the intermediate to infer a structural feature of the the undesired isomer(s) to highlight the contrast-
final drug substance, appropriate analytical tech- ing spectral features. If an undesired isomer is not
48 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
Figure 2: Manufacturing process of the drug substance.

Step 1 Step 2 C Step 3 Step 4 Step 5 F (Crude Step 6 G (Final

(RSM) drug substance) Purification drug substance)

available, the applicant should consider describing turer can consider implementing single or multiple
the data that would be expected if any undesired points of control for a specific CQA, particularly
isomers were formed, so that it is clear that the impurities, depending on the risk associated with
proposed specifications are specific to the desired the CQA and the ability of individual controls to
intermediate and not to its closely related isomers. detect potential problems. There are many options
This notion also applies to the full characterization available to the manufacturer for controlling im-
of final drug substance. purities in a drug substance. The decision about
where to control each particular impurity is up
Impurity control in intermediates to the applicant. Some of the more common op-
Impurities are unwanted materials present in the tions include controlling all impurities in the final
API that have no therapeutic value, may be harm- drug substance at their respective ICH limits (the
ful, and therefore, need to be controlled. Potential traditional approach); at an intermediate stage in
impurities at a given intermediate stage include the process at their respective drug substance ICH
residual starting materials and previous interme- limits; or at an intermediate stage in the process at
diates, impurities carried over from the starting higher than their respective drug substance ICH
materials, and previous intermediates, reagents, limits when coupled with demonstration that the
solvents, catalysts, and reaction by-products. The impurities are absent in multiple batches of the
identification of reaction by-products requires drug substance and thus do not require routine
some chemistry knowledge, previous experience, drug substance testing.
and literature knowledge. Investigation of batch Impurities may be divided into three categories:
analyses for observed impurities at ICH Q3A (12) organic impurities, residual solvents, and inor-
levels is the key step to determine what impurities ganic impurities. Each category is discussed sepa-
need to be controlled. It is essential for applicants rately in the following sections.
to demonstrate process understanding by examin- Organic impurities. In the fictional example pro-
ing the origins and fates of impurities to support vided in Figure 2, intermediate C is the proposed
the proposed controls at various stages. regulatory starting material (RSM). The crude
The control of impurities should always be a drug substance F, an intermediate, was manu-
component of a drug substance overall control factured in three steps (steps 3, 4, and 5) from
strategy. Under the QbD paradigm, the manu- the RSM C via intermediates D and E. RSM C is
facturing process of the drug substance and the considered to be a critical material in the overall
impurity profile must be understood step-by-step. manufacturing process and cGMP guidance ap-
When developing a control strategy, a manufac- plies from this step of the manufacturing process.
Generic Drugs
Figure 3: Routine impurity control. cess based on multiple
Greater than 5%
batch analyses. Inclusion
Greater than 10%
8% Between 7% of a study demonstrating
Between 2%-5%
3%-10% Content less 25% Content less how the amount of a par-
27% than 2% than 2%
65% 68% ticular organic impurity
decreases throughout the
manufacturing process
Content of each individual impurity Total impurity content from the proposed limit
to an acceptable level in
CQAs of RSM C include assay, organic impu- the final drug substance would provide stronger
rities, residual solvents, and residual metal from justification than the submission of end testing
catalysts carried over from Steps 1 and 2. These data alone. As long as RSM C is a process-related
attributes should be controlled with appropriate impurity, and not also a degradant that can be
limits in RSM C, thereby reducing the risk of fur- (re)generated in later steps, further control in the
ther carryover. ICH Q11 clearly states that the ana- drug substance may not be warranted. Multi-point
lytical methods for the regulatory starting material control of impurities can also be implemented for
should be capable of detecting impurities in the intermediates up to and including the drug sub-
starting material, and the fate and purge of these stance. A relatively high limit of residual reagent X
impurities and their derivatives in subsequent pro- in intermediate D can be controlled at a lower limit
cessing steps should be understood. in intermediate E, and then shown to be absent in
Often the proposed specification of the RSM C the drug substance. The impurity, therefore, will
is too wide or does not list all possible impurities. be well controlled and the chance of carry over to
When the assay of the starting material is set to a the drug substance will be slim.
wide range (e.g., Not less than 80%), the applicant It is often beneficial for the sponsor to propose
should discuss what makes up the remaining (20%) a limit for impurity Y in the RSM C specification
of the mass balance and how this will affect, via with an acceptance criteria at or below the accept-
the downstream reaction steps (Steps 3 thru 6), the able drug substance limit (based on ICH Q3A and
ultimate purity of the drug substance. MDD of the drug substance), so that monitoring
When the limit for a particular impurity (e.g., re- of Y is optional in the drug substance release speci-
sidual RSM C in intermediate D) is set higher than fication (unless the impurity is also a degradant).
the ICH limit (based ICH Q3A and maximum Alternatively, the limit for impurity Y in RSM C
daily dose of the drug substance), then the appli- can be proposed at a higher limit than the accept-
cant needs to demonstrate that residual RSM C able drug substance limit. In this case, the spon-
(and any by-products formed due to its continued sor must demonstrate by a spike and purge study
reaction) will be purged out to the acceptable ICH that impurity Y is purged out during the successive
level in the drug substance during the successive steps of the manufacturing process when present at
steps (Steps 4, 5, and 6) of the manufacturing pro- the maximum acceptable limit in the RSM C. The
50 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
same procedure can be used when setting a limit Figure 4: Control of potentially genotoxic impurities (when
for an organic process impurity in intermediates discussed).

D, E, or F. It is important to note that spike/purge

studies should be performed in a manner represen- Partially
tative of the commercial process in order for the 37%
results to be predictive. 19%
It is often the case that the manufacturing of Fully adequate

RSM C or the intermediates involves the use of

potentially genotoxic material (e.g., substituted
amino- or nitro-aromatics, alkyl halides, sulfonyl
chlorides, or other potentially genotoxic reagents). based the process data provided in the DMFs.
In the case that impurity Z in RSM C possesses Overall, 79% of the proposed individual impu-
a genotoxic alerting functionality, it can be con- rity limits were adequately justified. Additionally,
trolled in RSM C at a permitted daily exposure, total impurity content was controlled at less than
based on ICH M7 Assessment and Control of DNA 2% in 68% of the intermediates examined, at be-
Reactive (Mutagenic) Impurities in Pharmaceuti- tween 2%–5% for 25% of the intermediates, and
cals to Limit Potential Carcinogenic Risk (13) and at greater than 5% for 7% of the intermediates.
the MDD of the drug substance. Such control Overall, 92% of the proposed total impurity lim-
is sufficient to negate the need for any further its were adequately justified based on the process
downstream control of impurity Z. Alternatively, data provided in the DMFs.
a higher limit can be set in RSM C, only if it has However, only 66% of the sampled DMFs that
been established by a spike and purge study that use potentially genotoxic impurities (PGIs) as raw
this impurity Z is purged out during the successive materials or intermediates provided any discussion
steps of the manufacturing process when present of their control. Of those that did, 63% provided at
at the maximum acceptable limit in the RSM. As least some appropriate discussion and justification
another option, the sponsor can provide pharma- of their control strategy. The remaining 37% were
cology/toxicology data to justify a higher proposed judged to be inadequate based on the available in-
limit. The same procedures can be used when set- formation (Figure 4).
ting a limit for a potentially genotoxic process im- In approximately 25% of the sample of DMFs,
purity in intermediates D, E, or F. the undesired enantiomeric impurity is a possible
In the authors’ analysis, 65% of intermediates process impurity. More than 97% of the time, the
controlled individual impurities at not more than resulting enantiomeric impurity is controlled in
2% each, 27% of intermediates controlled indi- the API rather than in an intermediate.
vidual impurities between 3%–10%, and the re- As can be seen in these examples, it is necessary
mainder (8%) had at least one impurity controlled to have a thorough understanding of the manu-
above 10% (Figure 3). However, 21% of the proposed facturing process of the drug substance, including
intermediate impurity limits were not justified the steps leading to the RSM, so that the potential
Generic Drugs
Figure 5: Potential unwanted side reactions.

HOMe + CH3SO2C1 CH3SO2OMe (requires genotoxic impurity control)

HOEt + CH3SO2H CH3SO2OEt (requires genotoxic impurity control)

for carryover of impurities to the drug substance higher than the USP <467> limit if it can be demon-
can be minimized. strated that the process is capable of removing that
Residual solvents. The control of residual solvents solvent to the appropriate level in the final drug sub-
should always be a component of a drug substance stance. A decision not to test for a particular solvent
overall control strategy. Residual solvents may be needs to be justified through adequate supportive
controlled through final testing on a drug substance data. Inclusion of a study demonstrating how the
or during the manufacturing process as in-process amount of a particular residual solvent decreases
controls or components of an intermediate specifi- throughout the manufacturing process from the
cation. The decision about where to control a par- proposed limit to an acceptable level in the final
ticular residual solvent is up to the applicant. drug substance would provide stronger justification
The first option simply requires that all solvents than the submission of end testing data alone.
used in the process be specified and tested for in There are times when controlling a residual sol-
the final drug substance with acceptance criteria vent at an intermediate stage may be especially im-
less than or equal to the United States Pharma- portant. Tightly controlling residual amounts of
copeia (USP) chapter <467> limit for that solvent. lower chain alcohols in intermediates, for example,
Class III solvents can be controlled through in- may be important when sulfonic acids (e.g., meth-
dividual testing or by a loss on drying test with a ane sulfonic, benzene sulfonic, or p-toluene sul-
limit of not more than 0.5%. fonic acids) or their corresponding acid chlorides
It may be more efficient to control solvents that are used in the next step(s) of the manufacturing
are used early in the drug substance manufactur- process (Figure 5). Such tight control, at levels far
ing process through in-process tests or as part of below the USP <467> limits, will mitigate the risk
the specification of a RSM or intermediate, as op- of potentially genotoxic alkyl sulfonic ester impu-
posed to including them in the drug-substance rity formation and carryover to the drug substance.
release specification. This is especially true if a Conversely, controlling residual amounts of
solvent is used only during the early steps of the alkyl sulfonic acid or acid chlorides may be im-
manufacturing process and not at any later point. portant when alcohols are used in the following
Residual solvents that can be shown to be at or steps of the process. There may also be cases in
below their USP <467> limit in an intermediate which residual alcohols or other solvents have the
do not warrant testing in the final drug substance, potential to react with the drug substance to form
if they are not used in the later part of the manu- undesired side products.
facturing process. When these types of materials are part of a man-
A Class II or III residual solvent may be controlled ufacturing process, it is important that a thorough
upstream in the manufacturing process at a level discussion is presented about the possibility of
52 Pharmaceutical Technology SOLID DOSAGE DRUG DEVELOPMENT AND MANUFACTURING 2016 P h a r mTe c h . c o m
sulfonic acid esters being present and the control Figure 6: Potentially genotoxic impurities byproduct control.
strategies put in place to ensure that the risk of
carryover to the final drug substance is minimized.
Not controlled
This should be a preemptive discussion on part of 31% Controlled
the applicant, not the result of a question coming
from a regulatory agency. In this analysis, only
69% of the intermediates, where this type of PGI
was possible, were adequately controlled (Figure 6).
The remainder required additional justification.
Class I solvents are rarely used, but if a process controlled in the drug substance through a residue
does use this type of solvent, it must be tested, ei- on ignition test. Transition metals and some main
ther at an intermediate stage or in the final drug group metals, due to their higher level of toxicity,
substance. Non-testing of a Class I solvent is not require specific controls. The US Pharmacopeial
acceptable. Most Class I solvents, such as benzene, Convention is in the process of introducing Gen-
however, can be introduced into a process as im- eral Chapter <232> Elemental Impurities, which
purities of another solvent. If large quantities of will update the types of analytical methods used
solvents such as toluene, hexane, acetone, or meth- to monitor residual metals and set safe limits for
anol are used near the end of a manufacturing pro- patient exposure. The ICH Q3D guideline will pro-
cess, then residual benzene needs to be controlled. vide “a global policy for limiting metal impurities
Benzene can be controlled as part of an intermedi- qualitatively and quantitatively in drug products
ate’s specifications or in the final drug substance. A and ingredients” (14).
discussion and supporting data should be provided Many transition metals, such as palladium
to justify the proposed benzene limit in the process or platinum, are introduced into a process as
solvent in order to ensure that the benzene level in catalysts absorbed onto carbon. The removal of
the drug substance is less than 2ppm. these metals is usually accomplished by filtra-
Whatever control strategy is chosen, it should tion through some type of filter aid. If the filtra-
be accompanied by a well thought-out discussion tion process is not thorough, residual amounts
of the facts and not justified by simply submitting of these metals may carry through and would
the analytical results of multiple batches of drug most likely not be removed by other purification
substance showing the absence of the solvent. methods such as extraction or recrystallization.
Inorganic impurities. The control of metal impuri- Other metals may be removed from the process
ties in a drug substance is equally as important as by precipitation of an insoluble salt or extrac-
the control of organic and residual solvent impuri- tion using a chelating agent. Whatever process
ties. Residual metals are typically introduced into is used for the removal of the metal, a control
a drug substance as part of a reagent or catalyst. strategy for residual amounts of that metal must
Some metals may be completely innocuous, in- be in place. This control may be done by adding
cluding sodium and potassium, and may simply be a test for the metal to an intermediate specifica-
Pharmaceutical Technology Solid doSage drug development and manufacturing 2016 53
Generic Drugs
tion or including an in-process test. The tradi- References
1. R. A. Lionberger et al., The AAPS Journal 10 (2) (June 2008).
tional route of testing for the metal in the final 2. ICH, Q8 (R2) Pharmaceutical Development (ICH, August 2009).
drug substance is also an option. If a decision is 3. ICH, Q9 Quality Risk Management (ICH, November 2005)
4. ICH, Q10 Pharmaceutical Quality System (ICH June 2008).
made to not routinely test for a residual metal, 5. ICH, Q11 Development and Manufacture of Drug Substances
(ICH, May 2012).
then an applicant needs to submit adequate sup- 6. FDA, Pharmaceutical cGMPs for the 21st Century–A Risk-
porting data to justify that decision. As stated Based Approach; Pharmaceutical Quality Initiative–A Risk-
Based Approach (September 2004),
earlier in the residual solvent section, inclusion opmentApprovalProcess/Manufacturing/
of a study demonstrating how the amount of a escGMPforDrugs/ucm137175.htm
particular residual metal decreases throughout 7. ICH, Q7 Good Manufacturing Practice Guide for Active Phar-
maceutical Ingredients (ICH, November 2000).
the manufacturing process from the proposed 8. 21 CFR 314.50(d)(1)(i) Drug substance.
limit to acceptable level in the final drug sub- 9. ICH, M4Q (R1), The Common Technical Document: Quality.
10. FDA, Guidance for Industry: Submitting Marketing Applications
stance would provide stronger justification than According to the ICH-CTD Format—General Considerations
(FDA, 2001).
the submission of end testing data alone. 11. Chem. Eng. News, Bosutinib Buyer Beware, May 11, 2012.
Approximately 25% of the DMFs in the sample 12. ICH, Q3A (R2) Impurities in New Drug Substances (ICH, June
used a transition metal either during the starting 13. ICH, M7 Assessment and Control of DNA Reactive (Mutagenic)
Impurities in Pharmaceuticals to Limit Potential Carcinogenic
material synthesis or during the manufacturing Risk (ICH, June 2014).
process. In approximately 30% of these DMFs, 14. ICH, Q3D Guideline for Elemental Impurities (ICH, December
2014). PT
multiple batch analyses of the API were provided
to demonstrate process control. In the remaining
70%, specified tests were performed either in the Ad Index
starting material, the API, or both. COMPANY PAGE

Ashland Specialty Ingredients��������������������������������������������������������������7

The quality of the final manufactured drug sub- Capsugel ��������������������������������������������������������������������������������������������11
stance is the highest priority. Proper characteriza- Coating Place Inc ��������������������������������������������������������������������������������31
tion and impurity control of isolated intermediates
Federal Equipment Co ������������������������������������������������������������������������35
contribute to the quality of the drug substance.
Based on adequate control in isolated intermediate(s) Fette Compacting America Inc �����������������������������������������������������������27

of a defined manufacturing process, organic impu- Jost Chemical Co ����������������������������������������������������������������������������������9

rities, including genotoxic impurities, can be shown
Mikart ������������������������������������������������������������������������������������������������15
to be adequately controlled in the final drug sub-
stance. Residual solvents and inorganic impurities Natoli Engineer Co Inc ������������������������������������������������������������������������17

can be controlled in a similar manner. A clear jus- Process Systems Enterprise Limited �����������������������������������������������������3
tification and discussion of control strategies that
Rottendorf Pharmaceuticals ��������������������������������������������������������������39
include routine testing of intermediates should be
Trutag Technologies, Inc� �������������������������������������������������������������������23
provided by the manufacturer who wishes to use
such techniques in their manufacturing process. Veltek Associates����������������������������������������������������������������������������������5
54 Pharmaceutical Technology Solid doSage drug development and manufacturing 2016 P h a r mTe c h . c o m

black ES753623_PTEBOOK0316_054.pgs 03.22.2016 20:30 ADV