Implementation of analytical technologies in a pharmaceutical

Shared by: taw15849

Tags

process analytical technology, pharmaceutical industry, pharmaceutical manufacturing, product quality, manufacturing processes, pharmaceutical companies, analytical technologies, manufacturing process, quality assurance, process analytical, process control, continuous improvement, quality by design, product development, white papers

Stats

views:: 15
posted:: 2/16/2010
language:: English
pages:: 5

Public Domain

Document Sample

J ournal of Automated Methods & Management in Chemistry, Vol. 22, No. 2 (March–April 2000) pp. 41–45

Implementation of analytical technologies
in a pharmaceutical development organ-
izationÐ looking into the next millennium
Nigel North optimize the reaction steps during scale-up using equip-
Pharmaceutical T echnolog ies, SmithK line Beecham Pharmaceuticals, Harlow, ment developed in Europe by Anachem Ltd. This
Essex, UK equipment comprises a multiple-reaction block (STEM
corporation) , Gilson 233X L autosampler and an HPLC
M anaging the implementation of new technology in a pharmaceu-
system. Spectroscopic methods including infrared are also
tical development environment has provided challenges and oppor- being employed to monitor reactions on-line in process
tunities to obtain bene ts from technologies, e.g . laboratory
chemistry providing more rapid feedback on the progress
automation. Successful application of new techniques requires a
of the reactions compared to traditional separation
dedicated resource. Within Pharmaceutical T echnolog ies, this was methods. Chromatograph y for drug substance purity
initially a single person, who has since evolved into a team
determination and puri® cation is being performed using
dedicated to the investig ation and development of robotics and
parallel HPLC columns. The equipment supplied by
non-invasive analytical techniques. Pharmaceutical development is Biotage is a typical example of this concept. Fast-gradient
an important interface between research and commercial manu-
HPLC using short columns (2 mm ID £ 10 mm in length)
facturing. In research, the success of genomics and combinatorial
is also being applied to reaction monitoring enabling run
chemistry will result in a signi cant increase in the number of times to be reduced to ¹1 min.
development compounds, and this, combined with the desire of
commercial manufacturing to move towards parametric release, In the analysis of bio¯ uids, the implementation of LC/
puts an emphasis on the need for rapid analytical methods. Some MS technology has increased productivity markedly with
ideas on the techniques that will be required to meet these goals analysis times typically less than 5 min. The chromato-
will be described together with their impact on automation. graphy in this case is used more as a means of sample
introduction , and the `separation’ is performed by the
mass spectrometer. The requirement for high-speed
sample introduction for LC/MS technology has resulted
in the development of faster autosamplers for serial
Introduction sample introduction. More recently, multi-line auto-
samplers have been developed to increase sample
In research, platform technologies including DNA se- throughput into mass spectrometers.
quencing, combinatorial chemistry and high-throughpu t
screening (HTS) have increased the productivity in The techniques described above are by no means an
compound and lead generation. HTS can now provide exhaustive account of the analytical technologies being
the capability of screening 100 000 compounds per week; developed to increase the e ciency of some areas in pre-
this will soon increase to 100 000 compounds per day. clinical development; however, this does provide an over-
Combinatorial chemistry is rapidly developing to provide view of some of the trends in future technologies. The
the libraries of compounds to keep up with the screening main themes that are evident include: automation; rapid
capacity and to increase the quality of development separations; parallel analytical processing; fast sensitive
candidates. This capability has resulted in SB working and selective detectors; and non-invasive analysis.
on some 200 drug targets in 1997, which is nearly half of
These developments will improve the rate of compound
the 500 targets available to the entire pharmaceutical
supply, which in turn will impact the clinical supplies
industry up until 1995. The supply of compounds into
operation. In drug product analysis, well-established
pre-clinical development will therefore see a signi® cant
automation systems for tablet processing and dissolution
increase in the future. The e ect of this potential increase
of solid oral dosage forms have been implemented. These
in development compounds on the pre-clinical organ-
technologies, however, will not be su cient to keep up
ization in terms of the development of automation and
with the increase in development candidates from re-
other new technologies will be discussed with particular
search or with the emerging alternative technologies that
emphasis on managing its implementation in pharma-
will compete with the current analytical methods.
ceutical development.

M anaging the implementation of new analytical techniques in
Discussion Pharmaceutical T echnologies ( UK )
Pharmaceutical Technologies is responsible for the for-
D evelopment of new technologies in pre-clinical development
mulation and analytical development, clinical trials
The departments in pre-clinical development are intro- supply and technology transfer of new products to manu-
ducing technologies to improve their e ciency. In chemi- facturing facilities. Within Pharmaceutical Technologies
cal development, automation is being employed to help (UK) , the e ort devoted to developing and implement-

J ournal of A utomated M ethods & M anagement in Chemistry ISSN 1463± 9246 print/ISSN 1464± 5068 online # 2000 ISLAR 41
http://www.tandf.co.uk/journals/14639246.html
N. North Analytical technologies in a pharmaceutical development organization

ing analytical techniques including automated methods
has gradually increased over the past 6 years. Initially a
single person was assigned to developing and validating Pharmaceutical and
new automated methods using the ZymateT M dissolution Analytical Sciences Group
robot and the BenchmateT M within Analytical Develop- (PASG)
ment. In 1992, the department was reorganized into a
multidisciplinary team-based structure. This resulted in
the traditional analytical and formulation departments
being reorganized into `product’ teams comprising for-
mulators and analysts together with several specialist
`support’ teams that perform tasks, e.g. microbiology
and automated methods (Analytical Support) . A central
team to develop and validate was considered to be most
e ective for several reasons. (i) The experience gained Technology Sub-Group
from performing this process could be built up in the
Analytical Support team and then shared with the Prod- - Company Reps
uct teams on a timely basis enabling faster implementa- - Instrument Companies
tion of automated methods. (ii) Quali® cation and change - Innovation Group
control of robotic systems is easier to perform using a -----------------------------------
central team. (iii) Consistent approaches to validation of - Academic Links Regulators
automated methods are more likely when the same
scientists are involved in the process.
The reorganization of the department into this team-
based structure also resulted in the Analytical Support
team facing a signi® cant change in the interfaces and
provided communication challenges. The team has Fig ure 1. PASG technolog y development process.
moved from being in a department in which all the
analytical work was concentrated to interfacing with
more than 10 individual teams with projects at various SB. There are several reasons for these links. Firstly, to
stages of development and each progressing at di erent increase the level of resource devoted to new technology
speeds. Although establishing the initial outline for each development without increasing the headcount. Sec-
project was time consuming, keeping the information ondly, to promote the development and regulatory ac-
up to date is relatively easy. The information typically ceptance of non-invasive technologies, e.g. near-infrared
recorded is shown below. and more recently ultrasound. An example of this
scenario is collaborations coordinated by the Pharma-
(1) Compound. ceutical and Analytical Sciences Group (PASG) in the
(2) Priority. UK (® gure 1) . This group represents the major research-
(3) Quali® cation batch manufacturing date. based pharmaceutical companies in the UK with the
(4) Status of automated method for: main aims of: (i) providing a uni® ed voice on analytical
TPWII T M issues via the Association of British Pharmaceutical In-
Powder dustry to international regulatory agencies; (ii) enhan-
Bulk assay cing awareness of analytical science in education; and
Content uniformity (iii) promoting the image and profession of pharmaceu-
Degradation pro® le tical analysis. The PASG has sponsored the setting up of
MultiDoseT M dissolution. the Near Infrared Centre of Excellence at the London
(5) Product team. School of Pharmacy by supporting four PhD students.
The Centre has been working in four key areas in which
In 1997, Pharmaceutical Technologies (UK) was con- the regulators have concerns to aid the acceptance of the
solidated from three di erent sites into a single building technology. Collaboration has also just been initiated on
at Harlow, UK. During the past 18 months, the resource the application of ultrasound technology to pharmaceu-
and remit of this team involved with new analytical tical analysis.
technologies has now been widened to cover the follow-
ing.
(1) Implementation of automated methods (three scien- D evelopment of new technologies in pharmaceutical analysis
tists) .
Automation of the analysis of pharmaceutical drug prod-
(2) Development of non-invasive techniques, e.g. near-
ucts over the past 30 years has resulted in relatively small
infrared and ultrasound (two scientists) .
stepwise improvements in e ciency. In the 1970s, auto-
(3) Investigation of in vitro `dissolution’ systems for
mation of traditional colorimetric assays for drugs, e.g.
improving the prediction of in vivo performance of
penicillins was achieved by the use of autoanalysers. This
poorly soluble drugs (one scientist) .
was followed by the advent of HPLC, and in the early
To drive the development of new technology (e.g. non- 1980s unattended analysis was implemented utilizing
invasive analytical methods) forward faster in the UK, autosamplers. Dissolution robots became available in
collaborations have been set up with groups external to the mid-1980s followed by the tablet processing systems

42
N. North Analytical technologies in a pharmaceutical development organization

in the mid-1990s. These technologies have helped im- consuming and laborious. The sample size taken for
prove e ciency in sample throughput but these are the release testing of a drug product is generally very
unlikely to be su cient to keep pace with the develop- small compared with the batch size, e.g. 100 tablets
ments in research and the needs of commercial manu- from a 100 000 tablet batch. Furthermore, 30 tablets
facturing. or less are typically used to provide the assay and
content uniformity data to release the batch. Tech-
The weaknesses of current pharmaceutical analytical niques that enable the analysis to be performed non-
processes will be outlined to enable better understanding destructively without sampling in the manufacturing
of how appropriate newer technologies will be used in our area either at-line or on-line are needed. This would
laboratories in the future. A generic process for drug enable a much larger and more representative
product analysis is outlined in ® gure 2. sample size to be taken.
(1) Request for analysis. In the team-based organizations (3) Analysis. The release of a batch of drug product is
operated by Pharmaceutical Technologies in which generally dependent on limited in-process data and
analysts and formulators work closely together, a on testing of the ® nished product against a speci® ca-
traditional request for analysis form is not required tion. Several improvements to this process are re-
as the team inherently knows the status of its own quired. Firstly, more information on the quality of
samples. However, a laboratory information man- the product during the manufacturing process is
agement system (LIMS) is used to track information required, and as discussed above, rapid non-destruc-
on samples. This information is entered manually tive on-line techniques are required. The ultimate
together with the tests required for the sample. aim of this type of testing is to have su cient in-
Improvements needed include entry of sample in- process controls to enable a product to be released
formation by bar code and speci® cation/test informa- with little or no ® nal product testing, i.e. parametric
tion already available in the system enabling release. Secondly, the current time for automated
automatic generation. tablet preparation (e.g. 15 min) is relatively long
(2) Sampling. This is critical in any analysis process and is compared with the time for required HPLC or UV
generally performed manually, which can be time analysis, therefore more rapid sample preparation
systems are required. Non-destructive methods
would elegantly overcome this problem, but this will
not be possible for all products. For example, for low-
Request for analysis dose products, linking the separation with a sensitive
detector (e.g. LC/MS) would o er a good solution.
Finally, automatic results transfer to LIMS is typi-
cally available for outputs from HPLC systems. Links
Sampling between robotic systems, e.g. the TPWII T M and
MultiDose T M and the LIMS are also needed to
provide easy transfer of sample information during
the analysis. In the future, facile results transfer of the
results from spectroscopic techniques, e.g. NIR and
Analysis mass spectroscopy will also be needed.
· Sample preparation (4) D ocumentation. The documentation of practical work
· Measurement from an analysis is time consuming and has resulted
· Data Analysis in analysts spending more time out of the laboratory.
Typically, 15% of an analysts’ time can be spent
entering data into laboratory notebooks. In many
laboratories, LIMS are used in an attempt to reduce
the documentation burden, but the success and
Documentation savings in time are largely dependent on the environ-
ment in which the system is used. In development,
the rapidly changing project requirements tend to
make the use of LIMS problematic such that in-
Review of data formation is recorded in laboratory notebooks, and
the summary of results and associated references are
transferred to LIMS. The net result from this process
is that a proportion of information is duplicated as
hardcopy and electronic ® les. In a commercial
Approval of data manufacturing QC, laboratories’ work tends to be
more predictable than in R&D, which has resulted in
LIMS being implemented more successfully. Inter-
facing instruments including balances, HPLCs, UV
spectrophotometers and associated robots to LIMS
Reporting of results are needed to reduce the time spent writing-up in
laboratory notebooks.
To reduce the wasted e ort in transcribing in-
Figure 2. Generic process for drug product analysis. formation, electronic notebooks are being developed.

43
N. North Analytical technologies in a pharmaceutical development organization

The Collaborative Electronic Notebook Consortium Analytical techniques for the next millennium
Association (CENSA) is one such group that is
The discussion above indicates that there are emerging
moving the development of this technology forward
trends in analytical technologies that will increase pro-
[1] . A key element in the implementation of electro-
ductivity in pharmaceutical development, and these will
nic information management is greater standardiza- be outlined below.
tion of the working practices in laboratories. The
increased use of robotics will help achieve this goal. (1) Parallel analytical processing. To provide traditional
(5) Review of data. Reviewing of data is the backbone of analysis, e.g. HPLC at-line to monitor a manu-
the GMP process. Automated checking by the facturing process, much faster systems are required.
LIMS of key information in an analytical method This could be achieved using fast HPLC systems;
should be possible. For example, LIMS could alert however, sample preparation instrumentation would
the analyst if the standard weights, sample weights, have to be re-engineered to keep up with the
system suitability and the associated results are not separation speeds. The application of ultrasonic tech-
nology may help solve this problem. Parallel sample
within predetermined limits. This scenario is best
preparation and separation systems that enable the
suited to a standardized process operated by auto-
analysis to be completed in ¹1 min will be required
mated systems.
to compete with near-infrared.
(6) Approval of data. For a product manufactured by a (2) M iniaturization. Computer `chip’ technology is being
well-de® ned process, checking of results against a developed to perform analyses. The so-called `micro-
typical release speci® cation can be undertaken auto- TAS’ (total analysis system) concept [2] combines
matically by LIMS. This has been performed sampling, separation, reaction and detection onto an
successfully for many years in commercial manu- integrated platform. This device is typically fabri-
facturing laboratories. Implementation in a develop- cated on silicon, glass or polymer wafers, on which
ment environment has not been widespread, which is micro minute features that are typically between 1
probably due to the less well-de® ned processes and and 50 microns in size, are photoetched. The key
procedures for manufacturing and analysing the bene® ts of this technology to pharmaceutical analysis
product. are very rapid analysis time (seconds) and the small
(7) Reporting of results. In R&D, results are typically size of the equipment. This technology would lend
required for a number of main purposes. Firstly, to itself to at-line analysis and to early development
release product for clinical trials, which requires the work when little compound is available. Sample
generation of a Certi® cate of Analysis. Secondly, the preparation systems to match these devices are
generation of stability data to support product regis- required before these devices are commercialized.
tration, which requires the data from several storage The recent investment by Hewlett Packard in Cali-
conditions tabulated in a form suitable for direct per Technologies is an indication of the level of
transfer into the regulatory document. The report interest in `chip’ technology for analysis.
generation capabilities of the LIMS system must be (3) N on-invasive analysis. Near-infrared [3] spectroscopy
more ¯ exible to provide data in the required form. and Raman spectroscopy [4] are alternative tech-
Achieving this goal would dramatically decrease niques for the analysis of pharmaceutical products.
the time spent compiling and checking regulatory No sample preparation is required, therefore analysis
dossiers. is rapid and lends the technologies to in-process
testing as well as ® nal product analysis. These tech-
nologies in most instances rely on building libraries of
acceptable samples and comparing the unknown
T echnolog y needs of manufacturing against this library. Linking these techniques to
microscopes is now providing the capability of ima-
Rapid, reliable, easy-to-use and cost-e ective methods ging materials with resolution down to about 10±
are required by quality control (QC) laboratories in 20 microns. This technology is being used for trou-
manufacturing. Methods based on robotics exhibit some bleshooting; however, there may be utility for em-
of these desired attributes. The QC departments typically ploying imaging in the routine testing of products
now require robotic methods for assay, content unifor- and providing early warning of stability problems.
mity and dissolution methods from R&D prior to the Perkin Elmer have developed an automated system
manufacture of the quali® cation batches. In addition, for NIR imaging instrument. For highly potent low-dose
the validation batches, automated methods samples are drug products these techniques are generally not
needed for the analysis of powder samples taken at sensitive enough, and more traditional separation
various time points from the mixers to demonstrate the methods would be required for analysis.
homogeneity of the blending process. This work typically (4) Sensor technology. Low levels of volatile residues in
generates hundreds of powder samples for assay using materials, e.g. packaging components are increas-
TPWII T M . More recently, QC laboratories are request- ingly a ecting the stability of drug products. Form-
ing NIR methods for assay, content uniformity and aldehyde is a compound that has been shown to
powder blend analysis. NIR o ers rapid and non-de- cross-link gelatin capsules and to cause stability
structive analysis compared to automated UV and problems with products. Neural nose technology
HPLC technologies. This technique also has the added has been used to monitor volatile residues in
advantage of generating little waste compared with materials, particularly in the food industry. This
conventional techniques. technology uses electrically conducting organic poly-

44
N. North Analytical technologies in a pharmaceutical development organization

mers as sensors to monitor the aroma from materials sive technologies; sensors; fast sensitive and selective
together with neural networks to classify the aromas. detectors together with associated information manage-
Monitoring of raw materials and products on stabi- ment systems. New technologies are providing challenges
lity may be a rapid means of ® ngerprinting materials and opportunities in pharmaceutical development. Suc-
and predicting problems with volatile residues. cessful application of automated methods and new
(5) Fast sensitive and selective detectors. The rapid develop- analytical techniques requires a central team dedicated
ment of LC/MS in terms of lower cost, ease of use and to the investigation and development of automated
quantitative performance has meant that these methods and non-invasive analytical techniques. Colla-
systems will realize their potential to become routine borations with other pharmaceutical companies and
detectors for HPLC. This may enable fast `universal’ academic centres are providing more resources to intro-
liquid chromatograph y methods to be used in which duce the technologies in a highly regulated pharmaceu-
the chromatograph y system is more of a means of tical environment.
sample introduction and any incomplete separations
could be compensated by the mass spectrometer.
This technology should lead to reduced time for A cknowledgements
method development and validation for quantitative
methods, and provide more sensitive detection for Nigel North wishes to thank Les Brockhurst, Chris
degradation products, thus giving earlier warning of Banton, John Gostick, Kevin Smith and John Stanley
stability problems. for their help in developing new analytical techniques in
(6) Informatics. The technologies described above in (1) ± Pharmaceutical Technologies.
(5) will provide more information on an increased
number of samples. Experience to date with LIMS
systems in a development environment indicates that Trademarks
signi® cant improvements will be required to inte-
grate information collection from both instruments
and scientists (electronic laboratory notebooks) . Zymate T M , Zymark TPWII T M and Zymark Multi-
Once the information resides in the information Dose T M are registered trademarks of the Zymark
management system, better reporting packages and Corporation.
links to regulatory management documents are
necessary.
References

1. Collaborative Electronic Notebooks ConsortiumÐ www.censa.org
Summary 2. A CHE, H. J., 1996, Chemical microanalytical systems: objectives and
latest developments. Fresenius J . Anal. Chem., 355, 467 ± 474.
The increased ¯ ow of development compounds from 3. DRENNEN, J. K., 1995, Near infrared spectroscopy: applications in
research and the needs of commercial manufacturing will the analysis of tablets and pharmaceutical dosage forms. Applied
Spectroscopy Reviews, 30, 139± 174.
require new technologies to be implemented. The trends 4. N IEMCZYK, T. M., 1998, Quantitative determination of Bucindolol
in analytical techniques to meet these needs include: concentration in intact gelatin capsules using Raman spectroscopy.
parallel analytical processing; miniaturization; non-inva- Anal. Chem., 70, 2762 ± 2765.

Related docs