Biologics Consulting Group Order No.

FDA199305A

MID ATLANTIC REGION INSPECTION GUIDE

CLEANING VALIDATION
May 7, 1993

DIVISION OF FIELD INVESTIGATIONS

OFFICE OF REGIONAL OPERATIONS
OFFICE OF REGULATORY AFFAIRS
U.S. FOOD AND DRUG ADMINISTRATION
DEPARTMENT OF HEALTH AND HUMAN SERVICES

Biologics Consulting Group, 1407 King Street, Alexandria, VA. 22314/ Tel: 703-739-5695/ Fax: 703-548-7457
 MID ATLANTIC REGION
INSPECTION GUIDE

CLEANING
VALIDATION

p
REVISED
MAY 7, 1993
 INSPECTION GUIDELINES
VALIDATION OF CLEANING PROCESS
FS

Validation of cleaning procedures has generated considerable discussion since agency documents,
including the Inspection Guide for Bulk Drug Substances and the Biotechnology Inspection Guide,
have briefly addressed this issue. These Agency documents clearly establish the expectation that
cleaning procedures (processes) be validated.

In US vs. Barr Labs, the court ruled that a firm cannot wait for contamination and other problems to
reveal inadequate cleaning procedures. In order for cleaning rules to be effective, the specific methods
chosen must be shown to be effective. Judge Wolin ruled that firms must identify the cleaning agents
used in its cleaning process. When these agents are known to cause residue, the company must check for
the residue. With respect to equipment, he ruled that section 211.67 aplies to "equipment" and there is no
regulatory support for validating only "major pieces of equipment".

This inspection guide is designed to establish inspection consistency and uniformity by discussing
practices that have been found acceptable (or unacceptable). At the same time, one must recognize that for
cleaning validation, as with validation of other processes, there may be more than one way to validate a
process. In the end, the test of any validated process is whether scientific data shows that the system
consistently performs as expected and produces a result that consistently meets predetermined
specifications.

Scone of Evaluation

The first step is to focus on the objective of the validation process, and we have observed that some
companies have failed to develop such objectives. It is not unusual to see manufacturers use extensive
sampling and testing programs following the cleaning process without ever really evaluating the
effectiveness of the steps used to clean the equipment or systems. Several questions need to be addressed
when evaluating the cleaning process. For example, at what point does a piece of equipment or system
become clean? Does it have to be scrubbed by hand? What is accomplished by hand scrubbing rather than
just a solvent wash? How variable are manual cleaning processes from batch to batch and product to
product? The answers to these questions are obviously important to the inspection and evaluation of the
cleaning process since one must determine the overall effectiveness of the process. Answers to these
questions may, also identify steps that can be eliminated for more effective measures and result in resource
savings for the company:

Determine the number of cleaning processes for pieces of equipment. Obviously, one can reduce the
resources required to validate the cleaning process and cleaning agent if

Page 1
one cleaning procedure can be used for each piece of equipment or system. Also, such action will
minimize the chance that an inappropriate procedure or agent will be used. However in some cases,
particularly in Research and Development manufacturing areas for clinical products, multiple processes
and various solvents may have to be employed.

Equipment Design

Examine the design of equipment, particularly in those large systems that may employ clean-in-place
(CIP) or automation since they represent significant concerns. For example, sanitary type piping without
ball valves should be used. When such nonsanitary type ball valves are used, as is common in the bulk
drug industry, the cleaning process may be more difficult. Check the plant's documentation that the types
of valves used in the CIP system are suitable for use in these systems. When such systems are identified, it
is important that operators performing cleaning operations be aware of problems and have special training
in cleaning these systems and valves.

Determine whether the cleaning operators have knowledge of CIP systems and the level of training and
experience in cleaning these systems. Also check the written and validated cleaning process to determine
if these systems have been properly identified and validated.

Consider both microbiological and chemical contaminants in these CIP systems. For example, we have
already documented the presence of Pseudomonas sp. contamination in a drug substance because the
piping between reactors and bulk drug substance manufacturing equipment was improperly sanitized.

Cleaning Process Written Procedure and Documentation

Examine the detail and specificity of the procedure for the (cleaning) process being validated, and the
amount of documentation required. We have seen general SOPS, while others utilize a batch record or log
sheet system that requires some type of specific: documentation for performing each step. Depending upon
the complexity of the system and cleaning process and the ability and training of operators, the amount of
documentation necessary for executing various cleaning steps or procedures will vary.

When more complex cleaning procedures are required, it is important to document the critical cleaning
steps (one example is certain bulk drug synthesis processes). However, for relatively simple cleaning
operations, the mere documentation that the overall cleaning process was performed might be sufficient.

Other factors such as history of cleaning, residue levels found after cleaning, and variability of test
results may also dictate the amount of documentation required. For example, when variable residue
levels are detected following cleaning, particularly for a process that is believed to be acceptable, one
must establish the effectiveness of the

Page 2
process and operator performance. Appropriate evaluations must be made and when operator
performance is deemed a problem, more extensive documentation (guidance) and training may be
required.

In some of the larger systems, such as those employing long transfer lines or piping, check the flow charts
and piping diagrams for the identification of valves and written cleaning procedures. Piping and valves
should be tagged and easily identifiable by the operator performing the cleaning function. In some cases,
inadequately identified valves, both on prints and physically, have led to incorrect cleaning practices.

Always check for the presence of an often critical element in the documentation of the cleaning
processes; identifying and controlling the length of time between the end of processing and each cleaning
step. This is especially important for topicals, suspensions, and bulk drug operations. In such operations,
the drying of residues will directly affect the efficiency of a cleaning process.

alvtical Method

Determine the specificity and sensitivity of the analytical method used to detect residuals or contaminants.
With advances in analytical technology, residues from the manufacturing and cleaning processes can be
detected at very low levels. It should be pointed out that if levels of contamination or residual are not
detected, it does not mean that there is no residual contaminant present after cleaning. It only means that
levels of contaminant greater than the sensitivity or detection limit of the analytical method are not present
in the sample. It may also indicate deficiencies in the sampling methods. For example, we have seen the
use of a W scan of a sample with little or no knowledge of the detection limit for the test, i.e., at what
concentration will the test identify the contaminate.

SAMPLING

There are two general types of sampling that have been found acceptable. The most desirable is the direct
method of sampling the surface of the equipment. A less desirable method is the use of rinse solutions.
The use of a pseudoproduct or placebo i<, not generally acceptable.

a. Direct Surface Sa_mnlin~ - Determine the type of sampling material used

and its impact on the test data since the sampling material may interfere
with the test. For example, the adhesive used in swabs has been found to
interfere with the analysis of samples. Therefore, early in the validation
program, it is important to assure that the sampling media and solvent
(used for extraction from the media) are satisfactory and can be readily
utilized.

Page 3
 Advantages of direct sampling are that areas hardest to clean and which are reasonably
accessible can be evaluated, leading to establishing a level of contamination or residue per
given surface area. Additionally, residues that are "dried out" or are insoluble can be sampled
by physical removal.

b. Rinse Samples - Two advantages of using rinse samples is that a larger

surface area may be sampled, and inaccessible systems or ones that cannot
be routinely disassembled can be sampled and evaluated.

A disadvantage of rinse samples is that the residue or contaminant may not be soluble or may be
physically occluded in the equipment. An analogy that can be used is the "dirty pot". In the
evaluation of cleaning of a dirty pot, particularly with dried out residue, one does not look at the
rinse water to see that it is clean; one looks at the pot.

Check to see that a direct measurement of the residue or contaminant has been made for the
rinse water when it is used to validate the cleaning process. For example, it is not acceptable to
simply test rinse water for water quality (does it meet the oompendial tests) rather than test it
for potential contaminates.

Indirect testing, such as conductivity testing, may be of some value for routine monitoring once a
cleaning process has been validated. This would be particularly true for the bulk drug substance
manufacturer where reactors and centrifuges and piping between such large equipment can be
sampled only using rinse solution samples.

c. Placebo Product - We generally object to the use of a placebo product to

evaluate and validate cleaning processes, although some manufacturers
have processed a placebo batch in the equipment under essentially the
same operating parameters used for processing product. A sample of the
placebo batch is then tested for residual contamination. We have
documented several significant problems in using placebo product to
validate cleaning processes.

One cannot assure that the contaminate will be uniformly distributed throughout the system. For
example, if the discharge valve or chute of a blender are contaminated, the contaminant would
probably not be uniformly disbursed in the placebo; it would most likely be concentrated in the
initial discharged portion of the batch. Additionally, if the contaminant or residue was of a larger
particle size, it may not be uniformly dispersed.

Some have made the assumption that a residual contaminant would be worn off the equipment
surface uniformly; this is also an invalid conclusion.

Page 4
r

Finally, the analytical power may be greatly reduced through dilution of the contaminate.

Establishment of Limits

Check the manner in which limits are established, since this is one of the most important steps in cleaning
validation and should be one of the first steps in the development of a. cleaning validation protocol. The
objective is to ensure that the basis for any limit is scientifically justifiable.

The manner in which limits have been established by the industry as a whole has been varied and largely
arbitrary. For example, some project levels of contamination based on the concept of uniformity of
contaminants throughout a system or piece of equipment. They also assume that the contaminant will be
uniformly disbursed in the next substance or product processed through the system and/or piece of
equipment. Some even calculate the dose of (diluted) contaminant based on the therapeutic dose of the
following substance or product. Once the theoretical level of contaminant/dose is determined, a safety or
toxicological evaluation is made.

While there are some who believe that this hypothetical exercise must be performed, others are not all
that concerned with toxicological calculations. They are more concerned that their cleaning processes are
evaluated and optimized and are consistent. Once this has been determined, then toxicological (safety)
assessments can be made.

Unlike finished pharmaceuticals where the chemical identity of residuals are know, bulk: processes may
have many partial reactants and unwanted by-products which may never have been chemically identified.
When establishing residual limit, it may not be adequate to focus only on the principal reactant since other
chemical variations may be more difficult to remove. There may be instances where TLC screening, in
addition to chemical analysis, may be needed. Where very potent chemicals are processed, such as some
steriods, the issue of by-products needs very careful attention and evaluation. Such attention is not
necessary when a piece of equipment is dedicated to a single product and process.

Detergent

Detergents are not part of the manufacturing process and are only added to facilitate cleaning during the
cleaning process. Thus, they should be easily removed. If a detergent or soap is used for cleaning,
determine and consider the difficulty that may arise when attempting to test for residues. A common
problem associated with detergent use is its composition. Many detergent suppliers will not provide
specific composition, which makes it difficult for the user to evaluate residues. As with product residues, it
is. important and it is expected that the manufacturer evaluate the efficiency of the cleaning process for the
removal of residues.

Page 5
 - w

IrIi

Test Until Clean

I
.............. Examine and evaluate the level of testing and the retest results since .testing until clean
i, .............. is a concept utilized by some manufacturers, including the bulk drug substance industry.
I .............. They test, resample, and retest equipment or systems until an "acceptable" residue level
is attained. For the system or equipment with a validated cleaning process, this practice;
of resampling should not be utilized and is acceptable only in rare cases. Constant
retesting and resampling can demonstrate that the cleaning process is not validated since
these retests actually document the presence of unacceptable residue and contaminates
and the ineffective cleaning process.

For example, in a recent inspection, it was not uncommon for a manufacturer to perform three or four
retests on "cleaned" equipment until a satisfactory value was obtained. This firm's cleaning procedures
were never evaluated and validated, and the company was basically relying solely on the test result. That
is, they were assuming that the test is infallible and that the sample is truly representative. As previously
discussed, a major problem associated with cleaning validation is the lack of uniformity of the sample and
the inability of the test to identify certain level of residue and contaminates.

Obviously, the amount of residue left on equipment produces a significant impact on the; effectiveness of
the cleaning process. One can expect that a validated cleaning process may not perform consistently when
larger than normal amounts of residue are left on the equipment over unusual periods of time. Therefore,
an important step in the cleaning process is to reduce the residues on equipment to the minimum amount
possible prior to using solvents and other cleaning agents.

In another example, we found a company monitoring its cleaning process with multiple samples collected
throughout the system. The firm's procedures provided for resampling the entire system if some positive
results were obtained; therefore, some resamples were, found to have unacceptable residue levels after
they were initially found to be satisfactory. This outcome again demonstrates that the cleaning process is
not validated„

references

Until recently, there has not been a lot of published literature on the topic of cleaning validation. Papers
that reference this topic include The Validation of Cleaning Procedures by S.W. Harden and Cleaning
Validation by Douglas Mendenhall.

While the word "validation" associated with process development has been a relatively recent buzzword,
the evaluation and justification of cleaning processes has been employed in the pharmaceutical industry
for many years. In a 7/16/66 et.al. seminar on the topic of Control Procedures in Drug Production, Harley
Rhodehamel discussed the development and evaluation of cleaning procedures at a large pharmaceutical
manufacturing facility. The concepts and methods discussed in his paper presented in

Page 6
1966 are still current today.

4/21 /92 Henry L. Avallone

7/22/92 Richard J. Davis
7/27/92 Paul N. D'Eramo
7/27/92 Joseph X. Phillips
S/7/93Richard J. Davis

Page 7