PROCESS LIFECYCLE VALIDATION - Risk Management Applications in Pharmaceutical and Biopharmaceutical Manufacturing - page 180

8.4.3 Level of Testing and Number of Studies
The study type and testing required for PV will depend on a number of factors,
such as the processing goal, input and output parameters, amount and type of
data needed to demonstrate process control, process variability, routine testing of
in-process and final product, risk control strategy, and process monitoring.
The number of studies required to fully validate a process will also depend
on the complexity of the manufacturing steps. The biologic drug substance man-
ufacturing process is very complex; hence, a large number of studies are usually
required for its PV [16,9]. Application of an appropriate RA tool such as risk
ranking of the activities based on complexity, robustness, previous laboratory and
pilot studies, and knowledge from similar products (e.g., platform technology for
biologics) can be helpful in prioritization of risks. The decision to omit a study
creates its own risk and should be assessed for risk of limited (or lack of) process
knowledge and risks to product approval.
A sampling plan (i.e., number of samples and locations) for a validation study
should be based on statistical consideration, process/equipment design, and/or
potential worst-case location considerations. For example, samples in a filling pro-
cess should include first and last vials in addition to other samples to account for
variation at the start and end of processing. A mixing study needs to be performed
in product pool tanks to ensure homogeneous conditions. For lyophilization pro-
cesses, worst case location(s) should be established from temperature mapping
and that location(s) must be part of the lyophilizer sampling plan. One key con-
sideration during sampling is patient (or product quality) risk versus business
risk. For example, a nonrepresentative sample is a potential patient/product qual-
ity risk in that it is not able to capture worst-case conditions. Bioburden sampling
for CV in a noncontrolled area may pose a business risk for possible false failure
(i.e., false positive).
8.4.4 Validation Approach
Grouping strategies (e.g., family, bracketing, worst-case, modular, and generic
approach) may be justified through a scientific and risk-based approach to deter-
mine an appropriate level of testing. The modular approach used in biologics
(also known as platform strategy) is the use of data from a study performed on
a specific unit operation for one product to support the manufacturing process
for a different product. A science- and risk-based approach related to the process
parameters should be performed to apply the modular approach. To accomplish
this, a set of scientific criteria should be developed to compare the process
parameters for the process under development with the process parameters of
previously validated product(s). A unit operation in the manufacturing process
should meet the following conditions/requirements in order to apply modular
validation.
• The unit operation is robust and comparable raw materials are used.