■ By Michele Levenson and Bikash Chatterjee, Pharmatech Associates
A framework for implementing a stage 3 process validation program
In January 2011 the FDA issued the New Process Validation Guidance. The modern definition of Process Validation has abandoned the concept of a one-off activity where success consists
of obtaining three commercial batches of product. Instead, the new guidance describes
Process Validation as a continuous lifecycle founded upon the principles
of scientific understanding, and
divides it into three stages.
Stage 1 tackles product and
process development by identifying critical process parameters
and critical material attributes, to
establish the Processes Proven
Acceptable range (PAR) and
Normal Operating Range (NOR).
Stage 2 is the closest thing to the
old concept: demonstrating process predictability.
But it is Stage 3 that introduces the most significant
departure from the 1987 concept: that is, Continuous
Process Verification throughout a product’s lifecycle. This
article will examine the components of a successful Stage
3 Continuous Process Verification program and discuss a
framework for integrating this new activity as part of an
organization’s current Quality Management System (QMS).
A STAGE 3 FRAMEWORK
To understand the Stage 3 methodology, it is important
to understand the activities that comprise the total PV lifecycle. Figure 1 summarizes the major tasks that constitute
the three stages.
CONTINUOUS VERIFICATION AND THE APR
All pharmaceutical Quality Management Systems include an
Annual Product Review (APR) process. The GMPs require an
annual evaluation of the quality standards of a drug product to
determine the need for adjustments in drug product specifications, manufacturing and control procedures. Subpart J of 21
CFR 211.180 mandates establishing a written procedure for the
APR process and recommends review of a representative number of approved as well as rejected batches.
The concept behind the APR process is a regular periodic
analysis exercise to evaluate the overall performance of a
product based upon its Critical to Quality Attributes (CQAs).
The concept behind the Stage 3 Continuous Monitoring pro-
Figure 1 – The Three Stages of Process Validation
cess is to extend the same phi-
SHOP FLOOR AND PROCESS
losophy to those factors within
a process that affects product
performance. The challenge
with the Stage 3 portion of the
new process validation guidance
is two-fold; how to acquire the
data from the shop floor and
what rules to apply in terms
of adjustments to the process.
The four activities of establish-
ing a continuous verification sys-
tem will be discussed as follows.
Implementing an effective veri-
fication and control strategy to ensure process and
product performance is the heart of the new Process
Validation Guidance. Prior to Stage 2, the CPPs that drive
product performance variability and the control strategy for
those parameters that affect process predictability should
be clearly identified. Similarly, critical material attributes
(CMAs) should have been identified and measurement tools
implemented in Stage 1 to capture the appropriate attri-
butes. The goal of Stage 3 is to ensure the validated process
remains in a state of control that translates to analyzing the
behavior of any CPPs and CMAs identified against the pro-
cess understanding derived from Stages 1 and 2.
Specifically, for new products and processes, Stage 3 builds
upon the process knowledge gained in Stage 1 and Stage 2.
For legacy products, process understanding is a hybrid of
historical control and performance data and Stage 2 confirmation data. The APR process is a good place to obtain historical
product performance data as well as any discussion regarding
Out of Specification( OOS) and Out of Trend (OOT) results.
When designing the sampling and testing approach for
the control strategy it is important to consider the Stage
3 data acquisition component. Many organizations have
limited or no Statistical Process Control (SPC) programs in
place. Capturing documentation within the batch record is
typical. However, the systems for extracting that data for
use as part of a Continuous Verification Program do not