SAMPLING METHODS
Another key part of a cleaning-validation program is appropriate sampling methods for the equipment surfaces and for the nature
of the study, including the analytical methods used. The principles for sampling methods for biotech manufacture are not fundamentally
different from those for sampling in nonbiotech cleaning validation. Sampling methods discussed in Technical Report No. 49
include "direct surface" sampling (e.g., using a fiberoptic probe), swabbing, rinse-water sampling, and placebo sampling.
In practice, sampling for biotech manufacture may more likely involve rinse samples because much of the equipment is hard-piped
and not readily accessible for swab sampling. Furthermore, some biotech companies like to use mock runs or blank runs (i.e.,
a type of placebo sampling) to provide an accurate picture of total carryover throughout the entire process of bulk active
manufacture.
Table 5
|
Sampling recovery studies for biotech cleaning validation are not different in principle from sampling recovery studies in
nonbiotech cleaning validation. However, residues used for spiking surfaces for recovery studies in biotech cleaning validation
may include not only the bulk active, but also soils representative of early-stage harvesting steps. Furthermore, these recovery
studies represent worst cases, in that residues actually sampled in cleaning-validation qualification protocols are actually
degraded fragments of the active, which, being smaller in molecular weight and more polar, should be easier to remove in a
sampling recovery study using water as the solvent. Table V provides some of the important considerations for choosing the
sampling method for cleaning.
MAINTENANCE OF VALIDATED STATE
A key part of the validation life cycle for any system is maintaining the validated state on an ongoing basis. Any change
in the validated state of a cleaning process might detract from the quality, safety, and purity of manufactured products.
Tools for validation maintenance covered in Technical Report No. 49 include change control, risk-based periodic monitoring,
and data trending review. Training and retraining for manual cleaning processes are also significant because they are the
primary mechanisms for obtaining consistency in manual cleaning processes.
For biotech and nonbiotech cleaning validation, actual values for residues (e.g., in rinse-water samples) are significantly
below the acceptance criterion limit. The reason is that most manufacturers design their cleaning processes with a reasonable
margin of safety so that any samples taken during qualification protocols or during routine maintenance will pass the acceptance
criteria with a good margin of safety (e.g., a robust cleaning process is designed). Therefore, the fact that actual residue
values are significantly below acceptance limits should not by itself be a reason for making qualification protocol limits
more stringent. This situation is often addressed by establishing action or alert levels for residues for routine monitoring
samples. Routine monitoring results above such action or alert levels provide an indication of a possible change in the cleaning
process, thus requiring an investigation into the cause.
|
