WIP/CIP and Closed Equipment Systems

in the Field of Pharmaceutical Solid

Dosage Forms

Axel Schiffmanna, Dr. Bernhard Luyb, Dr. Markus

Bättigb, Prof. Dr. Hans Leuenbergera

University of Basel, Pharmazentrum (Pharmaceutical

Center)a, Basel, Switzerland
Glatt GmbHb, Binzen, Germany

Pharma Technologie Journal, No. 1080,

Pages 206 – 223 (2000)
Summary:
WIP/CIP and Closed Equipment Systems in the Field of Pharmaceutical Solid Dosage Forms
In the pharmaceutical industry production equipments of the manufacture of solid dosage forms are
cleaned manually, semi- or fully-automated. In order to implement a fully automated cleaning process
which meets pharmaceutical (i.e. GMP-) requirements, many conventionally used components must be
changed considerably with respect to construction and design. An important aspect is the so-called
Total Containment: CIP-ability and Total Containment are interdependent and must be considered
equally with the development. The realisation of the Total Containment is an absolute prerequisite for
the implementation of this new type of equipment. By the example of a fluid bed system this article des-
cribes, how this can be achieved by modifications of conventional equipments and of peripheral devi-
ces.

The abbreviations WIP/CIP used in this context are clearly defined. WIP (washing in place) means
semi- or fully-automated cleaning with either undefined result of cleaning or with the result that the
system is not yet clean according to GMP-requirements. CIP (cleaning in place), on the other hand,
stands for the entire process of a fully-automated cleaning to a GMP-conform level, including all factors,
which have influence on the cleaning result. This includes the proof that the acceptance criterion of
the cleaning validation was achieved.

A comparison study between the manual and fully-automated cleaning shows that by systematic modifi-
cation of the individual components in connection with a fully-automated cleaning program a higher
cleaning grade can be achieved. Furthermore, a statement about the reproducibility of cleaning suc-
cess can be met.

Keywords

Equipment systems, containment

Cleaning in place

Pharmaceutical solid dosage forms

Good Manufacturing Practice

Washing in place

2
1. Introduction

In past years, increasing attention has been directed to the cleaning of processing systems and the
corresponding cleaning validation, both by pharmaceutical companies as well as by monitoring
authorities. There are several important reasons for this.
On the one hand, one seeks to lower the risk of cross-contamination for existing products and
production lines. And on the other, highly effective materials are becoming increasingly more
prevalent in production and manufacturing. For some time now, these materials have placed high
demands on cleaning and, in general, have required - in as far as has been possible – a dust-free,
closed manufacturing process (e.g., cytostatic agents, hormones, cardiac glycosides, etc.). Exposure of
the operating staff, environmental protection and safety aspects provide additional reasons to employ
completely closed production lines whenever possible, as is carried out under Total Containment.
Additional factors that must be considered are economic aspects as well as the specific requirements
for the cleaning validation. Idle times, no matter what their reasons, are unproductive. With the
constantly increasing awareness of costs, manual cleaning procedures that are staff-intensive and are
time-consuming are being critically compared to automated cleaning procedures that, whenever
possible, are carried out at night. Another advantage is that, with automated cleaning in closed
systems, fundamentally more aggressive cleaning conditions (detergent types and concentrations,
temperatures, pressures, etc.) can be used than those that are employed with traditional methods.

With regard to validation issues, the reproducibility of an automated cleaning has certain advantages
when compared to manual hand cleaning. Parameters that are relevant for cleaning tend to be more
readily accessible and easier to record and document, thereby facilitating validation. Considering
these aspects, an automated cleaning system offers clear advantages.

2. WIP / CIP Terminology

Systems that are very different, even from their stated goal, are frequently lumped together under the
same CIP (cleaning in place) designation. The description, CIP Process, already well-ensconced in the
liquids sector where it is used in the more narrow sense of the word, has been adapted to use in the
solid forms sector.

But which systems in this latter sector actually carry out a CIP process? Installing one or more
cleaning nozzles in a processing system, e.g., such as in a fluid bed apparatus that has a conventional
construction, may be designated as a CIP system. Nevertheless, it certainly does not compose a
system that is clean enough, according to a truly fully-automated, validated cleaning based on GMP
standards, that would enable the processing of a formulation with another active agent. Based on these
reasons, and especially with conventionally designed systems, the corresponding cleaning systems
have been designated WIP (washing in place) systems. Here, a good - or even a very good -
pre-cleaning is achieved; however, this must always be followed by additional manual cleaning steps.
Even when a solid dosage form production system is set up for optimal cleanability and has the
corresponding peripheral devices, one can only speak of a system with CIP cleaning in the GMP sense
if, in addition to the system design, the respective product, the manufacturing process and the cleaning
procedures are also known. Even though in some cases, system manufacturers can claim
CIP-compatibility, the following must be considered:
One particular system that has proven to be CIP-compatible for a certain product may not be
CIP-compatible for a different product. The system itself can only be laid out for the best possible
cleaning, but the actual degree of cleaning achievable must always be determined in each individual
case.

3
3. Design and Technology Requirements

Even while a multitude of processes with correspondingly different processing technology are
employed in solid dosage form technology, the critical design principles and design characteristics of
the different apparatuses are the same with regard to their cleaning and containment aspects. Flange
joints, sensors, unions, viewing ports, filter media with air or gas carrying apparatuses, etc. must be
employed throughout the system and product transfer executed. Using fluid bed technology as an
example, we aim to demonstrate how the requirements for cleanability and finished product handling
can be met and implemented.

3.1. WIP Cleaning

With fluid bed systems having a traditional design, certain components are not suitable, even
from the start, for a completely automated, GMP-conforming CIP; therefore, only a WIP process
can even be considered.
This will be demonstrated using the example of some of the system parts found in a fluid bed
apparatus.

3.1.1. Seals and Gaskets

These are usually found at flange joints
(Figs. 1 and 2), sight glasses, sensors, nozzle
flanges and blind flanges on nozzle connection
pieces, etc. Even here it must be kept in
mind that, in addition to the solid ingredient,
solutions or suspensions that arise from wet
cleaning can also enter into the sealing system.
In addition to the actual cleaning (i.e., including
the removal of such fluids), the drying of these
types of sites is important in order to prevent
microbial growth, wherever and whenever
possible.

3.1.2. Bottom Screen of the Product Container

A multi-layered design consisting of a supporting
floor is frequently used. On top of this is placed
a (fine) sieve mesh and a stabilizing construction
that eases the strain on the fine sieve. This set-up Fig. 1: A conventionally-designed blind flange. When
the seal is positioned as shown, the product can accu-
must be manually dismantled and each part mulate in the gap.
must be individually cleaned in order to remove
those particles that have found their way in
between the layers (Fig. 3).

Fig. 2: A conventionally-designed flange. With this Fig. 3: A conventionally-constructed bottom screen

shape, either the product or the cleaning compound can (multi-layer mesh construction)
infiltrate into the seal.
4
3.1.3. Sensors
Particularly important to mention in this regard
are those instruments used to measure temperature
and sites where pressure differences are recorded.
Especially when measuring pressure differences,
a direct, open connection exists between the
measuring site (e.g., directly in the product space)
and the pressure transmitter. Cleaning solutions
can infiltrate these sites and, therefore, a hand
cleaning is almost always necessary.

3.1.4. Product Filters

As a rule, these textile filters are usually in the
shape of bag filters that can be shaken or blown
out (Figs. 4 and 5). For WIP cleaning, one can
first carry out a pre-wash in the system; this helps
avoid excessive dust formation during removal
from the system. Subsequently, a cleaning in
the washing machine takes place while the Fig. 4: Conventional bag type shaking filters
rest of the system is further cleaned. These
filters are very frequently dedicated equipment.
An (economically sound) CIP cleanability can
usually be excluded because of the types of
attachments (such as a clamped flange joint
or similar attachment). Specifically, cleaning
liquids containing dissolved or suspended
agents are transported via capillary forces to
sites from which their removal proves most
difficult.

3.2. CIP Cleaning

In order to achieve the stated goal of having an

optimal fluid bed apparatus that can be cleaned
without any manual action and which, additionally,
adheres to a Total Containment concept, one sees
that it is not only the main components that must
be critically examined, and if necessary, modified,
but also each small detail, no matter how minute,
Fig. 5: Conventional pleated filter cartridge with a lami-
from the charge ports over the sensors and the nated PTFE membrane.
viewing ports up to the cleaning nozzles themselves.
Indeed, for setting up the best possible GMP-conforming
CIP for fluid bed systems, the following modifications
must be carried out in the various system parts:

3.2.1. Flange Connections, Sealing Systems

To make CIP cleaning possible here, a sealing system was developed which, in the beginning, does not
permit the product or the cleaning fluid to enter the flange system at all (Figs. 6 and 7). This is achie-
ved by a correspondingly milled flange in combination with inflatable or mechanically expanding seals
which reliably prevent any escape from the inside of the container. The seal is placed in an exposed
location, i.e., even though it is in direct contact with the product, it nevertheless guarantees a free
inflow and outflow of the cleaning liquids) during CIP cleaning. This is also important with regard to
drying, which is only efficient in a free flow-through system.

5
Fig. 6: Example of a transition junction on a CIP–confor- Fig. 7.- Inflatable seal, CIP-capable, in the sealed condi-
ming system (Glatt, Binzen). tion (view from the inside of the container) (Glatt,
Binzen).

Viewing ports/blind flanges, etc.

Ideally, the viewing ports are solidly built into the system (Figs. 8 and 9). The welded construction thus
ensures a smooth and flat surface.

Fig. 8: Welded-in viewing port, CIP-design, after conta- Fig. 9: Viewing port of a conventional design (Glatt,
mination (view from the product side) (Glatt, Binzen). Binzen).

With exchangeable flange systems, the seal should be exposed in direct contact with the product and
easily positionable.

Both of these conditions have been fulfilled with a special flange design (O-Plus™-System, Glatt,
Binzen; Figs. 10 and 11). Here, the O-ring is hollow and, using a variable collar, can be held in two posi-
tions:
a) In an unloaded condition, a twisting or an incorrect positioning of the seal seat is avoided,
thereby ensuring a facile installation or dismantling of the flange.
b) In the final sealed state; the O-ring is pressed into the proper position.

6
Fig. 10: CIP-compatible blind flange. Fig. 11: Set-up showing the seal installed, CIP-compati-
(O-Plus™ System, Glatt, Binzen). ble design (O-Plus™ System, Glatt, Binzen).

3.2.2. Bottom Screen

Instead of the multi-layered construction having a
perforated bottom upon which a fine sieve is placed,
a so-called wedge wire sieve is installed (Figs. 12
and 13). This is a single layer leading bottom which
is completely accessible from both sides for clea-
ning. With a defined gap distance (which forms the
free surface), stainless steel rods having a triangular
cross-sectional shape are arranged parallel to each
other. The apex of the triangle thus faces downward
so that, when seen from above, the surface is flat
and smooth.

Basically, other systems might also be suitable, for

example, perforated metal plates with a directed air
flow (e.g., screen mesh pierced sheets, such as
Conidur® etc.). All floors have the common potential
problem that the product can drop into the lower
part of the system. Fig. 12: CIP-compatible wedge wire bottom plate (Glatt,
Binzen).

As soon as the leading bottom shows openings that

are larger than the smallest particles contained in
the product, the product can fall through. Through
proper process control, even from the charging
stage onwards, this can be reliably prevented as
long as the leading bottoms ensure a sufficient
pressure difference.

Fig. 13: Cross-section of a wedge wire bottom plate, CIP-

compatible design (Glatt, Binzen).

7
3.2.3. Sensors
All sensors are installed flush against the wall and do not allow product penetration (Fig. 14).
Temperature sensors are welded in, for example, however, they can still be replaced from the outside
(Fig. 15). When measuring pressure differences, a completely different measuring principle is employ-
ed. An absolute pressure measurement is carried out at the measuring site under consideration. The
specially developed sensor is located behind a ceramic pane which is installed flush against the wall
and sealed with an O-ring. The difference in the absolute pressures is continuously calculated in the
background of the system control and this value is then displayed.

Fig. 14: Pressure sensor, installed in a CIP-compatible Fig. 15: Temperature sensor in the product container,
construction (Glatt, Binzen). installed in a CIP-compatible construction (Glatt,
Binzen).

3.2.4. CIP-Compatible Filter Cartridge

Stainless steel filters cartridges, having a specific con-
struction and design for ensuring cleanability and
long-term stability, are used; this is a most important
aspect, especially with regard to wet cleaning.
Pleated filter cartridges have a significant disadvanta-
ge with respect to product accumulation and a corre-
sponding decrease in the filter surface, accessibility
for the cleaning media as well as long-term stability
with higher blow-out pressures (Figs. 16, 17 and 18).

Fig. 16: Example of a pleated stainless steel filter

without a corresponding support mesh Demonstrates
good cleanability, however, material fatigue arises as a
consequence of the blowing out process.

Fig. 17. Example of a pleated stainless steel filter with a Fig. 18: Example of a very dirty pleated filter (spray gra-
corresponding supporting mesh. Demonstrates good nulation process). Severe decrease in the filter area on
8
mechanical stability, however, product residues are dif- the product side, experiences considerable cleaning
ficult to remove. problems.
Therefore, a round design was developed where the circumference is fixed (unlike the case with plea-
ted cartridges which experience a significant expansion in their circumference when the pleats are
separated out); this round design is extremely stable (Figs. 19 and 20). This design permits one to use a
special blow out mechanism that allows considerably higher pressures within the cartridge than those
that can be borne by a pleated cartridge, especially with wet cleaning. Due to the capillary forces of
the water with wet cleaning, the sieve material forms a considerably higher resistance than it does in
the dry state.

Fig. 19: Filter housing, CIP-compatible design (SC™ fil- Fig. 20: Stainless steel filter cartridge. CIP-compatible
ters, Glatt, Binzen). design (SC™ filters, Glatt, Binzen).

Additionally important is that the filtering layer (as

needed, up to 2 µm mesh size of the stainless steel
sieve mesh) comes to lie in a completely external
location and is not covered by an additional outer,
coarser supporting membrane (for stabilization pur-
poses), as is usually seen with CIP-compatible plea-
ted stainless steel filter cartridges. With a coarser
supporting mesh, the product can actually claw its
way into the filter and even find its way between the
membranes; this poses great difficulty when trying
to remove the product.

With regard to the cleaning validation, it should be

noted that, here, the usual swab tests fail. Thus pro-
duct residues present in the filter or supporting
mesh cannot be reached. For the cleaning valida-
tion, it is recommended that, after cleaning, the car-
tridge be completely immersed in a corresponding
container using the appropriate solvent and that the
analysis be made from this (Fig. 21). Likewise, the Fig. 21: Example of an immersion tank. With the use of
accumulation of residues should be examined. a displacement body, the required solvent volume is
reduced.

9
4. Total Containment

Beside the CIP compatibility of a system, a completely closed product handling is another essential
part of the total concept. Both aspects complement, respectively, necessitate each other. When com-
pared to conventional systems, changes in the following areas are necessary in order to achieve con-
tainment.

4.1. System Safety Concept

As part of containment, with regard to the outside, a completely closed system design can be achieved
without explosion pressure release. This is achieved by performing 3 or 10 bar pressure surge resi-
stant executions, depending on the product characteristics (presence of solvent: yes/no).
What is important is the proven reliability (certification by an independent institution) of the chosen
design as well as a corresponding realistic pressure test on the finished single system.

4.2. Charge Ports / Closed Removal

The system must have corresponding set-ups that permit a completely closed handling also to be car-
ried out here. For example, this can be achieved by the docking of containers and gravity charging;
one alternative would be a possible pneumatic advancement if the housing (-concept) does not permit
charging from the top. Indeed, with the pneumatic feeding, the same CIP design requirements that are
found with the fluid bed are also necessary here. The product is transported away from the system via
floor discharge, a closed side removal or again, pneumatic advancement. With respect to seals, clea-
ning nozzles, etc., all parts must be designed for fully automated cleaning.

4.3. Cleaning Nozzles

At all relevant locations, the system has self-cleaning cleaning nozzles that remain permanently fixed
to the system but which automatically extend during the cleaning process; subsequently, using com-
pressed air, they are dried and then again retract to their original position (flush with the wall) (Figs.
22 and 23). Contrary to the usual WIP process, no manual action before the cleaning is required (ope-
ning of the dirty system, inserting the WIP nozzles), so that here, too, work in an enclosed system is
possible.

Fig. 22: Extendible cleaning nozzles, CIP-compatible Fig. 23: Extendible cleaning nozzles (in the retracted
design (Glatt, Binzen). state); as seen from the inside of the container;
CIP-compatible design (Glatt, Binzen).

10
5. CIP Cleaning Station

For the cleaning itself, there is naturally a corresponding CIP preparation system that is necessary.
This system permits the selection of appropriate cleaning parameters and records the required values.
Because various system functions (flaps, valves, heating and fan for drying, etc.) must be controlled
during cleaning, whenever possible, the CIP program should be managed by the same system control
that directs the actual manufacturing process.

The cleaning process itself is determined by the:

various types of cleaning:
a) Lost cleaning – usually as a pre-rinse or post-rinse step: Here the water that is brought into
the processing system flows directly out of the system after it has had contact with the
surface of the system; it is not reused. This ensures that coarse dirt (pre-rinsing step) is
removed, as are tenside residues and/or lime-containing water (as post-rinsing steps).
b) Recirculating cleaning as the main cleaning step. Due to economic reasons, the cleaning
medium is pumped in a cycle.

Type and concentration of detergents:

Products used include bases, acids, emulsifiers, oxidizing agents, complex-forming agents in
appropriate concentrations.

Temperature, pressure and volume flow of the cleaning fluid.

Duration of the cleaning.

Cleaning sequence (of the cleaning and the nozzles employed), etc.

In this manner, the cleaning process can first be optimized, which is a requirement for each validation.
The effort involved in the validation of a CIP cleaning should not be underestimated; indeed, it can
even be as costly as the process validation itself.

With regard to the costs and the profitability of such a system, lump sum answers cannot be provided;
the actual requirement of a Total Containment Concept is, at the moment, much less a question of
profitability. Even with products that are not highly toxic, the reduction in cleaning time from between
0.5 to 2 days down to less than 4 hours, with considerably less personnel involvement, can present a
most considerable savings potential.

It is also well possible, and even makes sense, to use individual components of the CIP design in a
conventional system. Each improvement in procedure and design will also facilitate and speed up the
manual cleaning process.

11
6. Case Study: CIP Cleaning in
Comparison to Manual Cleaning

6.1. Introduction
The comparison of these two types of cleaning is currently being examined by A. Schiffmann as part of
his dissertation thesis. The results gathered should critically assess the differences in cleaning of a
CIP-compatible fluid bed system compared to a conventional fluid bed system.

On both systems, swab and rinse samples were taken after granulation with acetaminophen (paraceta-
mol). The absolute residue of respectively comparable installation parts was examined while conside-
ring the variability of the cleaning success.

6.2. Experimental Design

The study was carried out on the following fluid bed apparatus:

GPCG 15 SC (CIP-compatible construction)

SC metal filter with 10 mm mesh width

100 mm wedge wire sieve

Exclusively fully automated cleaning

GPCG 15 (conventional construction)

This system is a fluid bed apparatus such as is still commonly used in pharmaceutical production.

Textile shaking filter·

Bottom screen: PZ mesh between support design and floor plate

Exclusively manual cleaning

6.3. Contamination
In both systems, a fluid bed granulation using a 33 % active ingredient portion (Acetaminophen) was
carried out. After each granulation, a CIP cleaning was performed with the GPCG 15 SC, and with the
conventional GPCG 15, a manual cleaning based on a written SOP was undertaken. The tests were
similarly repeated twice in both fluid bed apparatuses.

6.4. Cleaning
Conventional GPCG 15
After each batch, a manual cleaning was carried out by the system operator. All three cleanings were
carried out by the same system operator. The textile shaking filters were cleaned in a customary was-
hing machine.

CIP-compatible GPCG 15 SC
After each batch, the system was fully automatically cleaned (Tab. 1).

6.5. Sampling
Using a sampling plan, which clearly defined the areas to be sampled, a swab sample and/or rinse
sampling was made.
Sampling occurred in the following manner:
a) Swab sample of one defined area (e.g., 10 x 10 cm), representative for a certain system part.
b) Swab sample of a complete system part.
c) Rinsing/immersion of a certain system part in a defined amount of the corresponding solvent.
Cellulose swabs of a defined size were used as the swabs, and were moistened with 1ml methanol. The
wiping of the area resulted by successively using two moistened swabs. Subsequently, both swabs
were placed together in a test tube.

12
Tab. 1: CIP-Program.

Step Sequence Medium V (m 3) T (°C) t (min) Remarks

1 Rinsing
(lost) tap water 0.28 15 2.5 pre-cleaning step
2 Washing
(recirculated cleaning) tap water + additive 0.28 80 25 Main cleaning step alkalinea)
3 Blowing out Compressed air - - 4 Blowing out the tenside
solution from the pipelines
4 Rinsing (lost) tap water 0.28 15 2.5 Rinsing out the tenside from
the system
5 Rinsing (lost) demineralized water 0.28 20 2.5 "Final rinse”
6 Blowing out compressed air - - 20 Drying the cleaning nozzles
7 Drying air 1300 m3/h - 105 25
Total 1.12 99.5

Water consumption: 0.84 m3: tap water; 0.28 m3 demineralized water.

a) Mixture of 1% Henkel p3 cosa cip 95 (alkaline) + 1.5 % Henkel p3 cosa cip 92 (alkaline + tenside)

6.6. Analysis
The test tubes were filled with 10 ml methanol and mixed for 15 minutes in the ultrasound bath. The
collected samples were then evaluated based on the method described in the USP 23 (Acetaminophen
Capsules) for HPLC (UV detection).

6.7. Results
The results obtained were calculated by averaging the three cleaning runs (Tables 2 and 3).

Tab. 2: Agent residue

Sample description Conventional GPCG 15 GPCG 15 SC (CIP)

Representative µg/100 cm2 Total residue µg/100 cm2 Total residue

system part (mg) (mg)

S1 Product container 2.3 90 <0.3a) 13.3a)

S2 Expansion zone 2.4 922 <0.3a) 186a)
S3 Seal surface in contact
with product 156 538 1.9 10
S4 Lower plenum 354 44250 <0.3a) 25a)
S5 Sample port 164 115 2.8 1.4
R1 Bottom screen 30.1 289 <1.5a) 12.3a)
R2 Filter 10.9 2507 <3.9a) 1170a)

Total residue in the

system (surfaces in
contact with the product) 48711 1418b)

S = Swab samples; R: Rinse samples

a) = For calculating the residue amounts, the detection limit is used.
b) = The amount is calculated from the theoretical amounts. The actual concentration can actually lie well below
the number listed.

13
Table 3: Relative standard deviation (rounded) of the individual measurements

Sample description Rel. SD (%) Rel. SD (%)

Conventional GPCG 15 CIP GPCG 15 SC

S1 Product container 52 - a)
S2 Expansion zone 39 - a)
S3 Seal area
contacting product 52 23
S4 Lower plenum 87 - a)
S5 Sample port 137 55
R1 Bottom screen 129 - a)
R2 Filter 70 - a)
a)
Measurements below the detection limit.

6.8. Discussion
The following can be concluded from the results obtained:

The CIP-compatible fluid bed system shows fewer analytically demonstrable contaminated zones
than does the manually cleaned system.

The residue amounts of the individual samples are higher with the conventional construction fluid
bed system that with manual cleaning.

The reproducibility of cleaning success is higher with the CIP-compatible fluid bed system than
with manual cleaning.

Other projects are currently investigating the interactions among different groups of formulation com-
ponents with the various cleaning methods of a technical nature, as well as with regard to the other
parameters that determine the cleaning results. The defined goal is, with knowledge of the formulation,
to rapidly optimize the cleaning in order to achieve a true CIP process that conforms to GMP.

7. References

Harder, S. W., The Validation of Cleaning Procedures, Pharm. Technol. 8, 29-34 (1984) Mendenhall, D. W., Cleaning Validation,
Drug Dev. Ind. Pharm. 15, 2105-2114 (1989)
Baffi, R., A Total Organic Carbon Analysis Method for Validating Cleaning between Products in Biopharmaceutical
Manufacturing, J. Parent. Sci. Technol. 45 (1), (1991)
Agalloco, J., Points to consider in the Validation of Equipment Cleaning Procedures, J. Parent. Sci. Technol. 46 (5), (1992)
FDA, Cleaning Validation Mid Atlantic Region Inspection Guide, July 1992
The Rules Governing Medicinal Products in the European Community, Vol. IV, Good Manufacturing Practice for Medicinal
Products Commission of the European Communities, January 1992
Code of Federal Regulations, Parts 200 to 299, National Archives and Records Administration, April 1992
FDA, Guide to Inspections of Validation of Cleaning Processes, July 1993
Fourman, G. L., Mullen, M. V., Determining Cleaning Validation Acceptance Limits for Pharmaceutical Manufacturing Operations,
Pharm. Technol. 17 (4), 54-56 (1993)
Barr, D. B., Crabbs, W. C., Cooper, D., FDA Regulation of Bulk Pharmaceutical Chemical Production, Pharm. Technol. 17 (9), 54-70
(1993)
Zeller, A. O., Cleaning Validation and Residue Limits: A Contribution to Current Discussions, Pharm. Technol. 17 (11), (1993)
LeBlanc, D. A., Danforth, D. D., Smith, J. M., Cleaning Technology for Pharmaceutical Manufacturing, Pharm. Technol. 17 (10), 118-
124 (1993)
Jenkins, K. M., Vanderwielen, A. J., Cleaning Validation: An overall Perspective, Pharm. Technol. 18 (4), (1994)
Hwang, R.-C., Kowalski, D. L., Truelove, J. E., Process Design and Data Analysis for Cleaning Validation, Pharm. Technol. 21 (2),
21-25 (1997)
Throm, S., Reinigungsvalidierung, Kommentar des VFA zu den PIC-Empfehlungen zur Reinigungsvalidierung, Pharm. Ind. 59 (8),
664-667 (1997)
Haga, R., Murakawa, S., Ostrove, S., Weiss, S., Cleaning Mechanism Study for BioPharmaceutical Plant Design, Pharm. Engin. 17
(5), 8-21 (1997)
Brutsche, A., Reinigungsvalidierung unter besonderer Berücksichtigung der analytischen Methodenentwicklung und -validie-
rung, Pharm. Ind. 59 (11), 998-1001 (1997) Forsyth, R. J, Haynes, D. V., Cleaning Validation in a Pharmaceutical Research Facility,
Pharm. Technol. 22 (9),104-112 (1998)

Correspondence Address: Axel Schiffmann, Pharmaceutical Center, University of Basel,

Klingelbergstraße 50, 4056 Basel (Switzerland)