AN EXAMPLE FAILURE MODES AND EFFECTS ANALYSIS
There are a number of potential risks associated with performing operations in a CNC environment. The magnitude of these risks,
however, depends on the type of drug substance that is being manufactured, the particular unit operation that is being performed,
and the position of a particular step in the overall process. The risk is context-specific, and mitigations should be employed
that are commensurate with the magnitude of the risk. Failure modes and effects analysis (FMEA) is a tool that can be used
to identify risks for a particular process and to direct design teams toward appropriate mitigating measures. In this section,
an example failure mode is examined for illustration.
Consider a process in which the formulation of a BDS is performed by ultrafiltration and diafiltration. The authors are interested
in evaluating the risk of an adventitious virus contaminating the BDS at this point in the process. This unit operation has
no ability to remove or inactivate viruses and there are no viral reduction steps downstream of this step. The analysis assumes
that the manufacturer has many years of experience performing this operation in a Grade C environment and now wishes to perform
this same operation in a CNC environment.
Table II shows an FMEA analysis of this failure mode. The impact of the failure is assigned a value of HIGH because viruses
could be potentially harmful to the patient. In the next column, possible causes of adventitious viral contamination are listed,
and measures designed to mitigate the identified causes are listed in the adjacent column.
Table II: An example failure modes and effects analysis to identify and evaluate risks associated with adventitious viral
contamination of a bulk drug substance formulation operation using an ultrafiltration/diafiltration unit operation. (QI is
quality impact, QR is quality risk, BI is business impact, BR is business risk, UF is ultrafiltration, PCR is polymerase chain
reaction, AVA is adventitious agent assay)
Next, the probability of occurrence is evaluated given the measures that are in place. A value of LOW is assigned to the probability,
and the rationale for this value is that there has been no known incident of viral contamination in the history of the operation,
and operating in a CNC environment does not increase the probability of contamination.
The detection point is given a value of NO. This means that if the event were to occur, it is possible that the event would
not be detected prior to the release of the final container product. Adventitious virus assays may not detect low-titer contaminations
and polymerase chain reaction (PCR) testing will only detect selected viruses.
In this case, the team assigned a risk of MEDIUM to this failure mode. The impact is HIGH, the probability of occurrence is
LOW, and there is possibly no detection point. The team indicated that more data are needed regarding the effectiveness of
caustic sanitization of the ultrafiltration cassettes after installation. They also identify that the raw materials for the
diafiltration buffer could potentially carry a virus and that treatment, or other control options for the buffer should be
explored. It should be noted that this risk of viral contamination exists regardless of the area classification under which
the UF/DF operation is performed.
In this example, the operators wear dedicated shoes and change into a plant uniform. The UF skid is sanitized prior to each
batch and the connections to the skid are steamed after the connections are made. The mitigations are specific to a particular
product, operation, facility, and manufacturer. Different facilities and manufacturers will implement this process step differently.
For example, a different manufacturer may not choose steam sanitization for the skid. The FMEA tool allows manufacturers to
evaluate the details of their particular operation and determine whether or not additional mitigating measures are required
to enable CNC operation.